Taking and Importing Marine Mammals; Taking Marine Mammals Incidental to Geophysical Surveys Related to Oil and Gas Activities in the Gulf of Mexico, 5322-5450 [2020-27252]

Agencies

• DEPARTMENT OF COMMERCE
• National Oceanic and Atmospheric Administration

[Federal Register Volume 86, Number 11 (Tuesday, January 19, 2021)]
[Rules and Regulations]
[Pages 5322-5450]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2020-27252]



[[Page 5321]]

Vol. 86

Tuesday,

No. 11

January 19, 2021

Part II





Department of Commerce





-----------------------------------------------------------------------





National Oceanic and Atmospheric Administration





-----------------------------------------------------------------------





50 CFR Part 217





Taking and Importing Marine Mammals; Taking Marine Mammals Incidental 
to Geophysical Surveys Related to Oil and Gas Activities in the Gulf of 
Mexico; Final Rule

Federal Register / Vol. 86 , No. 11 / Tuesday, January 19, 2021 / 
Rules and Regulations

[[Page 5322]]


-----------------------------------------------------------------------

DEPARTMENT OF COMMERCE

National Oceanic and Atmospheric Administration

50 CFR Part 217

[Docket No. 201204-0326]
RIN 0648-BB38


Taking and Importing Marine Mammals; Taking Marine Mammals 
Incidental to Geophysical Surveys Related to Oil and Gas Activities in 
the Gulf of Mexico

AGENCY: National Marine Fisheries Service (NMFS), National Oceanic and 
Atmospheric Administration (NOAA), Commerce.

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: NMFS, upon request from the Bureau of Ocean Energy Management 
(BOEM), hereby issues regulations to govern the unintentional taking of 
marine mammals incidental to geophysical survey activities conducted by 
oil and gas industry operators, and those persons authorized to conduct 
activities on their behalf (collectively ``industry operators''), in 
Federal waters of the U.S. Gulf of Mexico (GOM) over the course of five 
years. These regulations, which allow for the issuance of Letters of 
Authorization (LOA) to industry operators for the incidental take of 
marine mammals during the described activities and specified timeframe, 
prescribe the permissible methods of taking and other means of 
effecting the least practicable adverse impact on marine mammal species 
or stocks and their habitat, as well as requirements pertaining to the 
monitoring and reporting of such taking.

DATES: Effective from April 19, 2021 through April 19, 2026.

ADDRESSES: Electronic copies of the application and supporting 
documents, as well as a list of the references cited in this document, 
may be obtained online at: www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico. In case of problems accessing these documents, please call 
the contact listed below.

FOR FURTHER INFORMATION CONTACT: Ben Laws, Office of Protected 
Resources, NMFS, (301) 427-8401.

SUPPLEMENTARY INFORMATION: 

Purpose and Need for Regulatory Action

    These incidental take regulations (ITR) establish a framework under 
the authority of the MMPA (16 U.S.C. 1361 et seq.) to allow for the 
authorization of take of marine mammals incidental to the conduct of 
geophysical survey activities in the GOM. We received a petition from 
BOEM requesting the regulations. Subsequent LOAs may be requested by 
industry operators. Take is expected to occur by Level A and/or Level B 
harassment incidental to use of active acoustic sound sources. Please 
see the Background section below for definitions of harassment.

Legal Authority for the Action

    Section 101(a)(5)(A) of the MMPA (16 U.S.C. 1371(a)(5)(A)) directs 
the Secretary of Commerce to allow, upon request, the incidental, but 
not intentional, taking of small numbers of marine mammals by U.S. 
citizens who engage in a specified activity (other than commercial 
fishing) within a specified geographical region for up to five years 
if, after notice and public comment, the agency makes certain findings 
and issues regulations that set forth permissible methods of taking 
pursuant to that activity and other means of effecting the ``least 
practicable adverse impact'' on the affected species or stocks and 
their habitat (see the discussion below in the Mitigation section), as 
well as monitoring and reporting requirements. Section 101(a)(5)(A) of 
the MMPA and the implementing regulations at 50 CFR part 216, subpart 
I, provide the legal basis for issuing this rule containing the 
regulations, and for any subsequent LOAs. As directed by this legal 
authority, the regulations contain mitigation, monitoring, and 
reporting requirements.

Summary of Major Provisions Within the Regulations

    Following is a summary of the major provisions of these regulations 
regarding geophysical survey activities. These measures include:
     Standard detection-based mitigation measures, including 
use of visual and acoustic observation to detect marine mammals and 
shut down acoustic sources in certain circumstances;
     A time-area restriction designed to avoid effects to 
bottlenose dolphins in times and places believed to be of particular 
importance;
     Vessel strike avoidance measures; and
     Monitoring and reporting requirements.
    These incidental take regulations govern and allow for the 
subsequent issuance of letters of authorization for the take of marine 
mammals incidental to the specified activity described in this Notice, 
within the upper bounds of take that was evaluated for this rule, and 
prescribe measures for mitigation, monitoring, and reporting. They do 
not preclude a U.S. citizen from applying for an incidental take 
authorization for a specified activity with different parameters or 
required measures through a separate request and process.

Background

    Section 101(a)(5)(A) of the MMPA (16 U.S.C. 1361 et seq.) directs 
the Secretary of Commerce (as delegated to NMFS) to allow, upon 
request, the incidental, but not intentional, taking of small numbers 
of marine mammals by U.S. citizens who engage in a specified activity 
(other than commercial fishing) within a specified geographical region 
if certain findings are made, regulations are issued, and notice is 
provided to the public.
    An authorization for incidental taking shall be granted if NMFS 
finds that the taking will have a negligible impact on the species or 
stock(s) and will not have an unmitigable adverse impact on the 
availability of the species or stock(s) for subsistence uses (where 
relevant), and if the permissible methods of taking and requirements 
pertaining to the mitigation, monitoring, and reporting of such takings 
are set forth.
    NMFS has defined ``negligible impact'' in 50 CFR 216.103 as an 
impact resulting from the specified activity that cannot be reasonably 
expected to, and is not reasonably likely to, adversely affect the 
species or stock through effects on annual rates of recruitment or 
survival.
    The MMPA states that the term ``take'' means to harass, hunt, 
capture, or kill, or attempt to harass, hunt, capture, or kill, any 
marine mammal.
    Except with respect to certain activities not pertinent here, the 
MMPA defines ``harassment'' as: Any act of pursuit, torment, or 
annoyance which (i) has the potential to injure a marine mammal or 
marine mammal stock in the wild (Level A harassment); or (ii) has the 
potential to disturb a marine mammal or marine mammal stock in the wild 
by causing disruption of behavioral patterns, including, but not 
limited to, migration, breathing, nursing, breeding, feeding, or 
sheltering (Level B harassment).

Summary of Request

    On October 17, 2016, BOEM submitted a revised petition \1\ to NMFS 
for rulemaking under section 101(a)(5)(A) of the MMPA to authorize take 
of marine mammals incidental to conducting geophysical surveys during

[[Page 5323]]

oil and gas industry exploration and development activities in the GOM. 
This revised petition was deemed adequate and complete based on NMFS' 
implementing regulations at 50 CFR 216.104.
---------------------------------------------------------------------------

    \1\ In the notice of proposed rulemaking (83 FR 29212; June 22, 
2018), NMFS provided a brief history of prior petitions received 
from BOEM's predecessor agencies.
---------------------------------------------------------------------------

    On December 8, 2016 (81 FR 88664), we published a notice of receipt 
of the petition in the Federal Register, requesting comments and 
information related to the request. This 30-day comment period was 
extended to January 23, 2017 (81 FR 92788), for a total review period 
of 45 days. The comments and information received during this public 
review period informed development of the proposed ITR, and all 
comments received are available online at www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico.
    In August 2017, BOEM produced a final Programmatic Environmental 
Impact Statement (PEIS) to evaluate potential significant environmental 
effects of geological and geophysical (G&G) activities on the Outer 
Continental Shelf (OCS) of the GOM, pursuant to the National 
Environmental Policy Act (NEPA). The PEIS is available online at: 
www.boem.gov/Gulf-of-Mexico-Geological-and-Geophysical-Activities-Programmatic-EIS/. NOAA participated as a cooperating agency in the 
development of the PEIS.
    NMFS published a notice of proposed rulemaking in the Federal 
Register for a 60-day public review on June 22, 2018 (83 FR 29212). The 
comments and information received during this public review period 
informed development of the final ITR, and NMFS has responded to all 
comments received (see Comments and Responses).
    On February 24, 2020, BOEM submitted a notice to NMFS of its 
``updated proposed action and action area for the ongoing [ITR] 
process[.]'' This update consisted of removal of the area currently 
under a Congressional leasing moratorium under the Gulf of Mexico 
Energy Security Act (GOMESA) (Pub. L. 109-432, Sec.  104) from 
consideration in the ITR. BOEM stated in its notice to NMFS that G&G 
activities are not likely to be proposed within the area subject to the 
leasing moratorium during the 5-year period of effectiveness for the 
ITR and, therefore, that the ``number, type, and effects of any such 
proposed G&G activities are simply too speculative and uncertain for 
BOEM to predict or meaningfully analyze.'' These Congressional leasing 
restrictions are in place until June 30, 2022. Based on this updated 
scope, BOEM on March 26, 2020, submitted revised projections of 
expected activity levels and corresponding changes to modeled acoustic 
exposure numbers. BOEM's notice and updated information are available 
online at: www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico. These 
changes are addressed as appropriate throughout this final ITR. On 
September 8, 2020, the President effectively extended this moratorium 
through withdrawal under the Outer Continental Shelf Lands Act (OCSLA) 
of the same area covered by the GOMESA moratorium from disposition by 
leasing for 10 years, beginning on July 1, 2022, and ending on June 30, 
2032.
    Geophysical surveys are conducted in support of hydrocarbon 
exploration and development in the GOM, typically by companies that 
provide such services to the oil and gas industry. Broadly, these 
surveys include (1) deep penetration surveys using large airgun arrays 
as the acoustic source; (2) shallow penetration surveys using a small 
airgun array, single airgun, or similar systems as the acoustic source; 
and (3) high-resolution surveys, which may use a variety of acoustic 
sources. Generally speaking, these surveys may occur within Federal 
territorial waters and waters of the U.S. Exclusive Economic Zone (EEZ) 
(i.e., to 200 nautical miles (nmi)) within the GOM, and corresponding 
with BOEM's GOM OCS planning areas (i.e., Western Planning Area (WPA), 
Central Planning Area (CPA), Eastern Planning Area (EPA)). The use of 
these acoustic sources is expected to produce underwater sound at 
levels that have the potential to result in harassment of marine 
mammals. Cetacean species with the potential to be present in the GOM 
are described below (see Table 4).
    These regulations establish a framework under the authority of the 
MMPA (16 U.S.C. 1361 et seq.) and NMFS' implementing regulations (50 
CFR 216.101 et seq.) to allow for the authorization, through LOAs, of 
take of marine mammals incidental to the conduct of geophysical surveys 
for oil and gas activities in the GOM. The regulations are effective 
for five years.

Description of the Specified Activity

Overview

    The specified activity consists of geophysical surveys conducted by 
industry operators for a variety of reasons related to hydrocarbon 
exploration, development, and production. These operators are typically 
companies that provide geophysical services, such as data acquisition 
and processing, to the oil and gas industry, including exploration and 
production companies. The petition describes a five-year period of 
geophysical survey activity and provides estimates of the amount of 
effort by survey type and location. BOEM's PEIS (BOEM, 2017) describes 
a range of potential survey effort. The levels of effort in the 
petition (which form the basis for the modeling effort described later 
in the Estimated Take section) were the high-end estimates. Following 
BOEM's update of the petition's geographic scope, these estimates were 
revised accordingly. Actual total amounts of effort (including by 
survey type and location) would not be known in advance of receiving 
LOA requests from industry operators, but take in excess of what is 
analyzed for this rulemaking would not be authorized. As noted above, 
BOEM has updated the scope of the specified activity/specified 
geographical region by removing the area currently under leasing 
moratorium through GOMESA from consideration. The removed area largely 
covers the EPA, including areas in which NMFS had proposed time-area 
restrictions as mitigation, but also includes a portion of the CPA. 
Applicants seeking authorization for take of marine mammals incidental 
to survey activities within the GOMESA area during the 5-year period of 
effectiveness for this rule will need to pursue a separate MMPA 
incidental take authorization. See Figures 1 and 2.
    Geophysical surveys are conducted to obtain information on marine 
seabed and subsurface geology for a variety of reasons, including to: 
(1) Obtain data for hydrocarbon and mineral exploration and production; 
(2) aid in siting of oil and gas structures, facilities, and pipelines; 
(3) identify possible seafloor or shallow depth geologic hazards; and 
(4) locate potential archaeological resources and benthic habitats that 
should be avoided. In addition, geophysical survey data inform Federal 
government decisions. For example, BOEM uses such data for resource 
estimation and bid evaluation to ensure that the government receives a 
fair market value for OCS leases, as well as to help to evaluate worst-
case discharge for potential oil-spill analysis and to evaluate sites 
for potential hazards prior to drilling.
    Deep penetration seismic surveys using airgun arrays as an acoustic 
source (sound sources are described in the ``Detailed Description of 
Activities'' section) are a primary method of obtaining geophysical 
data used to characterize subsurface structure. These surveys are 
designed to illuminate

[[Page 5324]]

deeper subsurface structures and formations that may be of economic 
interest as a reservoir for oil and gas exploitation. A deep 
penetration survey uses an acoustic source suited to provide data on 
geological formations that may be thousands of meters (m) beneath the 
seafloor, as compared with a shallow penetration or high resolution 
geophysical (HRG) survey that may be intended to evaluate shallow 
subsurface formations or the seafloor itself (e.g., for hazards).
    Deep penetration surveys may be two-dimensional (2D) or three-
dimensional (3D) (see Figure 1-2 of the petition), and there are a 
variety of survey methodologies designed to provide the specific data 
of interest. 2D surveys are designed to acquire data over large areas 
(thousands of square miles) in order to screen for potential 
hydrocarbon prospectivity, and provide a cross-sectional image of the 
structure. In contrast, 3D surveys may use similar acoustic sources but 
are designed to cover smaller areas with greater resolution (e.g., with 
closer survey line spacing), providing a volumetric image of underlying 
geological structures. Repeated 3D surveys are referred to as four-
dimensional (4D), or time-lapse, surveys that assess the depletion of a 
reservoir.
    Shallow penetration and high-resolution surveys are designed to 
highlight seabed and near-surface potential obstructions, archaeology, 
and geohazards that may have safety implications during rig 
installation or well and development facility siting. Shallow 
penetration surveys may use a small airgun array, single airgun, or 
similar sources, while high-resolution surveys (which are limited to 
imaging the seafloor itself) may use a variety of sources, such as sub-
bottom profilers, single or multibeam echosounders, or side-scan 
sonars.

Dates and Duration

    The specified activities may occur at any time during the five-year 
period of validity of these regulations. Actual dates and duration of 
individual surveys are not known. Survey activities are generally 24-
hour operations. However, BOEM estimates that a typical seismic survey 
involves approximately 20 to 30 percent of non-operational downtime due 
to a variety of factors, including technical or mechanical problems, 
standby for weather or other interferences, and implementation of 
mitigation measures.

Specified Geographical Region

    The OCS planning areas are depicted in Figure 1, and the overlap of 
the GOMESA moratorium area with the planning areas (as well as with the 
modeling zones, see discussion of modeling zones below) is depicted in 
Figure 2, showing the updated specified geographical region.
    Only the northern portion of the GOM contains Federal waters. BOEM 
manages development of U.S. Federal OCS energy and mineral resources 
within OCS regions, which are divided into planning areas. Within 
planning areas are lease blocks, on which specific production 
activities may occur. Geophysical survey activities may occur on scales 
ranging from entire planning areas to multiple or specific lease 
blocks, or could occur at specific potential or existing facilities 
within a lease block. NMFS provided a detailed discussion of the 
specified geographical region in the notice of proposed rulemaking (83 
FR 29212; June 22, 2018).
    The prospective survey activities may occur in the U.S. waters of 
the GOM, within BOEM's Western, Central, and Eastern GOM OCS planning 
areas (approximately within the U.S. EEZ; Figure 1), but excluding the 
GOMESA moratorium area (Figure 2). Although survey activity in the 
GOMESA moratorium area is no longer being considered, the region has 
not changed compared with what was described, nor has substantive new 
information regarding the region become available. Therefore, we do not 
reprint that discussion here and refer the reader to that notice of 
proposed rulemaking for additional detail.
BILLING CODE 3510-22-P

[[Page 5325]]

[GRAPHIC] [TIFF OMITTED] TR19JA21.000


[[Page 5326]]


[GRAPHIC] [TIFF OMITTED] TR19JA21.001

BILLING CODE 3510-22-C

Detailed Description of Activities

    An airgun is a device used to emit acoustic energy pulses into the 
seafloor, and generally consists of a steel cylinder that is charged 
with high-pressure air. There are different types of airguns; 
differences between types of airguns are generally in the mechanical 
parts that release the pressurized air, and the bubble and acoustic 
energy released are effectively the same. Airguns are typically 
operated at a firing pressure of 2,000 pounds per square inch (psi). 
Release of the compressed air into the water column generates a signal 
that reflects (or refracts) off the seafloor and/or subsurface layers 
having acoustic impedance contrast. Individual airguns are available in 
different volumetric sizes and, for deep penetration seismic surveys, 
are towed in arrays (i.e., a certain number of airguns of varying sizes 
in a certain arrangement) designed according to a given company's 
method of data acquisition, seismic target, and data processing 
capabilities.
    Airgun arrays are typically configured in subarrays of 6-12 airguns 
each. Towed hydrophone streamers (described below) may follow the array 
by 100-200 m and can be 5-12 kilometer (km) long. The airgun array and 
streamers are typically towed at a speed of approximately 4.5 to 5 
knots (kn). BOEM notes that arrays used for deep penetration surveys 
typically have between 20-80 individual elements, with a total volume 
of 1,500-8,460 in\3\. The output of an airgun array is directly 
proportional to airgun firing pressure or to the number of airguns, and 
is expressed as the cube root of the total volume of the array.
    Airguns are considered to be low-frequency acoustic sources, 
producing sound with energy in a frequency range from less than 10 Hz 
to 2 kHz (though there may be energy at higher frequencies), with most 
energy radiated at frequencies below 500 Hz. Frequencies of interest to 
industry are below approximately 100 Hz. The amplitude of the acoustic 
wave emitted from the source is equal in all directions (i.e., 
omnidirectional) for a single airgun, but airgun arrays do possess some 
directionality due to phase delays between guns in different 
directions. Airgun arrays are typically tuned to maximize functionality 
for data acquisition purposes, meaning that sound transmitted in 
horizontal directions and at higher frequencies is minimized to the 
extent possible.
    When fired, a brief (~0.1 second) pulse of sound is emitted by all 
airguns in an array nearly simultaneously, in order to increase the 
amplitude of the overall source pressure signal. The combined signal 
amplitude and directivity is dependent on the number and sizes of 
individual airguns and their geometric positions within the array. The 
airguns are silent during the intervening periods, with the array 
typically fired on a fixed distance (or shot point) interval. The 
intervals are optimized for water depth and the distance of important 
geological features below seafloor, but a typical interval in 
relatively deep water might be approximately every 10-20 seconds (or 
25-50 m, depending on vessel speed). The return signal is recorded by a 
listening device, and later analyzed with computer interpretation and 
mapping systems used to depict the subsurface. There must be enough 
time between shots for the sound signals to propagate down to and 
reflect from the feature of interest, and then to propagate upward to 
be received on hydrophones or geophones. Reverberation of sound from 
previous shots must also be given time to dissipate. The receiving 
hydrophones can be towed behind or in front of the airgun array (may be 
towed from the source vessel or from a separate receiver vessel), or 
geophone receivers can be deployed on the seabed. Receivers may be 
displaced several kilometers horizontally away from the source, so 
horizontal propagation time is also

[[Page 5327]]

considered in setting the interval between shots.
    Sound levels for airgun arrays are typically modeled or measured at 
some distance from the source and a nominal source level then back-
calculated. Because these arrays constitute a distributed acoustic 
source rather than a single point source (i.e., the ``source'' is 
actually comprised of multiple sources with some predetermined spatial 
arrangement), the highest sound levels measurable at any location in 
the water will be less than the nominal source level. A common analogy 
is to an array of light bulbs; at sufficient distance--in the far 
field--the array will appear to be a single point source of light but 
individual sources, each with less intensity than that of the whole, 
may be discerned at closer distances (Caldwell and Dragoset (2000) 
define the far field as greater than 250 m). Therefore, back-calculated 
source levels are not typically considered to be accurate indicators of 
the true maximum amplitude of the output in the far field, which is 
what is typically of concern in assessing potential impacts to marine 
mammals. In addition, the effective source level for sound propagating 
in near-horizontal directions (i.e., directions likely to impact most 
marine mammals in the vicinity of an array) is likely to be 
substantially lower (e.g., 15-24 decibels (dB); Caldwell and Dragoset, 
2000) than the nominal source level applicable to downward propagation 
because of the directional nature of the sound from the airgun array. 
The horizontal propagation of sound is reduced by noise cancellation 
effects created when sound from neighboring airguns on the same 
horizontal plane partially cancel each other out.
    Survey protocols generally involve a predetermined set of survey, 
or track, lines. The seismic acquisition vessel(s) (source vessel) will 
travel down a linear track for some distance until a line of data is 
acquired, then turn and acquire data on a different track. In some 
cases, data is acquired as the source vessel(s) turns continuously 
rather than moving on a linear track (i.e., coil surveys). The spacing 
between track lines and the length of track lines can vary greatly, 
depending on the objectives of a survey. In addition to the line over 
which data acquisition is desired, full-power operation may include 
run-in and run-out. Run-in is approximately 1 km of full-power source 
operation before starting a new line to ensure equipment is functioning 
properly, and run-out is additional full-power operation beyond the 
conclusion of a trackline (e.g., half the distance of the acquisition 
streamer behind the source vessel, when used) to ensure that all data 
along the trackline are collected by the streamer. Line turns can 
require two to six hours when towed hydrophones are used, due to the 
long trailing streamers, but may be much faster when streamers are not 
used. Spacing and length of tracks varies by survey. Survey operations 
often involve the source vessel(s), supported by a chase vessel. Chase 
vessels typically support the source vessel(s) by protecting the long 
hydrophone streamer (when used) from damage (e.g., from other vessels) 
and otherwise lending logistical support (e.g., returning to port for 
fuel, supplies, or any necessary personnel transfers). Chase vessels do 
not deploy acoustic sources for data acquisition purposes; the only 
potential effects of the chase vessels are those associated with normal 
vessel operations.
    The general activities described here could occur pre- or post-
leasing and/or on- or off-lease. Pre-lease surveys are more likely to 
involve larger-scale activity designed to explore or evaluate geologic 
formations. Post-lease activities may also include deep penetration 
surveys, but would be expected to be smaller in spatial and temporal 
scale as they are associated with specific leased blocks. Shallow 
penetration and HRG surveys are more likely to be associated with 
specific leased blocks and/or facilities, with HRG surveys used along 
pipeline routes and to search for archaeological resources and/or 
benthic communities. Specific types of surveys, including 2D and 3D 
surveys and various survey geometries typically associated with 3D 
surveys (e.g., narrow- and wide-azimuth (NAZ and WAZ) and coil 
surveys), were described in summary in the notice of proposed 
rulemaking (83 FR 29212; June 22, 2018). We also described surveys 
involving the placement of seismic sensors in a drilled well or 
borehole, including various types of vertical seismic profiling and 
other types of borehole seismic surveys. For full detail, please refer 
to that notice or sections 1.2 and 1.3 of BOEM's petition.
    Surveys may be designed as either multi-source (i.e., multiple 
arrays towed by one or more source vessel(s)) or single source. Surveys 
may also be differentiated by the way in which they record the return 
signals using hydrophones and/or geophones. Hydrophones may be towed in 
streamers behind a vessel (either the source vessel(s) or a separate 
vessel) or in some cases may be placed in boreholes (called vertical 
seismic profiling) or spaced at various depths on vertical cables in 
the water column. Sensors may also be incorporated into ocean-bottom 
cables (OBC) or autonomous ocean-bottom nodes (OBN) and placed on the 
seafloor--these surveys are referred to generally as ocean-bottom 
seismic (OBS). Autonomous nodes can be tethered to coated lines and 
deployed from ships or remotely-operated vehicles, with current 
technology allowing use in water depths to approximately 3,000 m. OBS 
surveys are most useful to acquire data in shallow water and obstructed 
areas, as well as for acquisition of four-component survey data (i.e., 
including pressure and 3D linear acceleration collected via geophone). 
For OBS surveys, one or two vessels usually are needed to lay out and 
pick up cables, one ship is needed to record data, one ship tows an 
airgun array, and two smaller utility boats support survey operations.
    In summary, 3D survey design involves a vessel with one or more 
acoustic sources covering an area of interest with relatively tight 
spatial configuration (compared with 2D surveys). In order to provide 
richer, more useful data, particularly in areas with more difficult 
geology, survey designs become more complicated with additional source 
and/or receiver vessels operating in potentially increasingly 
complicated choreographies.
    As compared with 2D and 3D deep penetration surveys, shallow 
penetration and HRG surveys are conducted to provide data informing 
initial site evaluation, drilling rig emplacement, and platform or 
pipeline design and emplacement. Identification of geohazards (e.g., 
gas hydrates, buried channels) is necessary to avoid drilling and 
facilities emplacement problems, and operators are required to identify 
and avoid archaeological resources and certain benthic communities. 
These surveys may use single airguns or small airgun arrays, but 
generally use various types of electromechanical acoustic sources. 
Please see our notice of proposed rulemaking or BOEM's petition for 
additional detail regarding these survey types and electromechanical 
acoustic sources.

Summary of Representative Sound Sources

    Because the specifics of acoustic sources to be used cannot be 
known in advance of receiving LOA requests from industry operators, it 
was necessary to define representative acoustic source parameters, as 
well as representative survey patterns. BOEM determined realistic 
representative proxy sound sources and survey patterns, which were used 
in acoustic exposure

[[Page 5328]]

modeling and more broadly to support the analysis, after discussions 
with individual geophysical companies. Acoustic exposure modeling is 
described in detail in ``Acoustic Propagation and Marine Mammal 
Exposure Modeling of Geological and Geophysical Sources in the Gulf of 
Mexico'' and ``Addendum to Acoustic Propagation and Marine Mammal 
Exposure Modeling of Geological and Geophysical Sources in the Gulf of 
Mexico'' (Zeddies et al., 2015, 2017a), hereafter referred to 
collectively as ``the modeling report,'' as well as in ``Gulf of Mexico 
Acoustic Exposure Model Variable Analysis'' (Zeddies et al., 2017b), 
which evaluated a smaller, alternative airgun array. The reports are 
available online at: www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico.
    Representative sources for the modeling include a single airgun, an 
airgun array, and multiple electromechanical sources. Two major survey 
types were considered: Large-area seismic (including 2D, 3D NAZ, 3D 
WAZ, and coil surveys) and small-area, high-resolution geotechnical 
(including single airgun surveys and surveys using a CHIRP sub-bottom 
profiler in combination with multibeam echosounder and side-scan sonar; 
the single airgun was used as a reasonable proxy for surveys using a 
boomer). The nominal airgun sources used for analysis of the proposed 
action include a small single airgun (90 in\3\ airgun) and a large 
airgun array (8,000 in\3\). In addition, the supplemental Model 
Variable Analysis (Zeddies et al., 2017b) provides analysis of an 
alternative 4,130 in\3\ array (see Letters of Authorization section). 
We note that while high-resolution geophysical sources were 
conservatively included for consideration in this rule to allow for 
take authorization if necessary, some of these types of sources would 
not necessarily be expected to cause the incidental take of marine 
mammals, depending on the source type and/or the manner in which it is 
operated (e.g., operational settings, mitigation measures), and Letters 
of Authorization would not be necessary in those cases.
    Additional characteristics of the representative acoustic sources 
and representative operational parameters of the different survey types 
that were used in the modeling simulations to predict the exposure of 
marine mammals to different received levels of sound are described in 
the modeling report and in our notice of proposed rulemaking. Please 
see those documents for additional detail.
    We note that while it was necessary to identify representative 
sources for the purposes of modeling the number of takes to be included 
in the analysis under the rule, the analysis is intended to be, and is 
appropriately, applicable to takes resulting from the use of other 
sizes or configurations of airguns (e.g., the alternative, smaller 
airgun array modeled in the ``Gulf of Mexico Acoustic Exposure Model 
Variable Analysis'' report (Zeddies et al., 2017b) referenced in the 
proposed rule and available for public review as supplementary material 
to the proposed rule).
    While these descriptions reflect existing technologies and current 
practice, new technologies and/or uses of existing technologies may 
come into practice during the period of validity of these regulations. 
NMFS will evaluate any such developments on a case-specific basis to 
determine whether expected impacts on marine mammals are consistent 
with those described or referenced in this document and, therefore, 
whether any anticipated take incidental to use of those new 
technologies or practices may appropriately be authorized under the 
existing regulatory framework. We also note here that activities that 
may result in incidental take of marine mammals, and which would 
therefore appropriately require authorization under the MMPA, are not 
limited to those activities requiring permits from BOEM. There may be 
some activities that do not require permits from BOEM, such as certain 
ancillary activities, for which an LOA under this rule may be 
appropriate. Operators should consult NMFS regarding the 
appropriateness of applying for an LOA under this rule prior to 
conducting such activities.

Estimated Levels of Effort

    As noted previously, actual total amounts of effort by survey type 
and location cannot be known in advance of receiving LOA requests from 
industry operators. Therefore, BOEM's PEIS provided projections of 
survey level of effort for the different survey types for a 10-year 
period (and BOEM's updated scope refined those projections to a five-
year period). In order to construct a realistic scenario for future 
geophysical survey effort, BOEM evaluated trends in permit applications 
as well as industry estimates of future survey activity. In addition, 
GOMESA precludes leasing, pre-leasing, or any related activity (though 
not geophysical surveys) in the GOM east of 86[deg]41' W, in BOEM's 
Eastern Planning Area (EPA) and within 125 mi (201 km) of Florida, or 
in BOEM's Central Planning Area (CPA) and within 100 mi of Florida (and 
according to certain other detailed stipulations). These leasing 
restrictions are in place until June 30, 2022. On September 8, 2020, 
the President effectively extended this moratorium through withdrawal 
under OCSLA of the same area covered by the GOMESA moratorium from 
disposition by leasing for 10 years, beginning on July 1, 2022, and 
ending on June 30, 2032. This withdrawal prevents consideration of 
these areas for any leasing for purposes of exploration, development, 
or production during the 10-year period beginning on July 1, 2022, and 
ending on June 30, 2032. Although the withdrawal does not preclude 
geophysical survey activity, similar to the moratorium under GOMESA, 
the lack of leasing opportunities may be expected to curtail interest 
in exploratory surveys to some degree.
    In order to provide some spatial resolution to the projections of 
survey effort and to provide reasonably similar areas within which 
acoustic modeling might be conducted, the geographic region was divided 
into seven zones, largely on the basis of water depth, seabed slope, 
and defined BOEM planning area boundaries. Shelf regions typically 
extend from shore to approximately 100-200 m water depths where 
bathymetric relief is gradual (off Florida's west coast, the shelf 
extends approximately 150 km). The slope starts where the seabed relief 
is steeper and extends into deeper water. In the GOM water deepens from 
100-200 m to 1,500-2,500 m over as little as a 50 km horizontal 
distance. As the slope ends, water depths become more consistent, 
though depths can vary from 2,000-3,300 m. Three primary bathymetric 
areas were defined as shelf (0-200 m water depth), slope (200-2,000 m), 
and deep (>2,000 m).
    Available information regarding cetacean density in the GOM (e.g., 
Roberts et al., 2016) shows that, in addition to water depth, animal 
distribution tends to vary from east to west in the GOM and appears 
correlated with the width of shelf and slope areas from east to west. 
The western region is characterized by a relatively narrow shelf and 
moderate-width slope. The central region has a moderate-width shelf and 
moderate-width slope, and the eastern region has a wide shelf and a 
very narrow slope. Therefore, BOEM's western, central, and eastern 
planning area divisions provide appropriate longitudinal separations 
for the shelf and slope areas. Due to relative consistency in both 
physical properties

[[Page 5329]]

and predicted animal distribution, the deep area was not subdivided. As 
shown in Figure 3, Zones 1-3 represent the shelf area (from east to 
west), Zones 4-6 represent the slope area (from east to west), and Zone 
7 is the deep area.
BILLING CODE 3510-22-P
[GRAPHIC] [TIFF OMITTED] TR19JA21.002

BILLING CODE 3510-22-C
    Table 1 in the notice of proposed rulemaking provided the 10-year 
estimated levels of effort from BOEM's PEIS, estimated as 24-hr survey 
days, including annual totals by survey type and by zone for deep 
penetration and shallow penetration surveys, respectively. As the basis 
for the analysis supporting the proposed rulemaking, NMFS selected one 
high survey effort scenario and two each of moderate and low survey 
effort scenarios from the ten survey effort scenarios provided by BOEM. 
Of the ten ``years'' or effort scenarios, Year 1 (high), Years 4 and 5 
(moderate), and Years 8 and 9 (low) were selected as representative 
effort scenarios and carried forward for further evaluation.
    However, as noted previously, BOEM subsequently revised its 
proposed action by removing the area subject to leasing moratorium 
under GOMESA from consideration in the rule. In support of this 
revision, BOEM provided revised 5-year level of effort predictions and 
associated acoustic exposure estimates. BOEM's process for developing 
this information, described in detail in ``Revised Modeled Exposure 
Estimates,'' available online, was straightforward. Rather than using 
the PEIS's 10-year period, BOEM provided revised levels of effort for a 
5-year period, using Years 1-5 of the original level of effort 
projections. BOEM stated that the first five years were selected to be 
carried forward ``because they were contiguous, they included the three 
years with the most activity, and they were the best understood in 
relation to the historical data upon which they are based.'' NMFS 
concurs with this choice. Levels of effort were revised based on the 
basic assumption that if portions of areas are removed from 
consideration, then the corresponding effort previously presumed to 
occur in those areas also is removed from consideration. Revised 
estimates of future effort and associated acoustic exposures draw upon 
the prior projections and modeling approach, which were subject to 
notice and comment. Table 1 shows the percentage reduction in survey 
area for each modeling zone that results from BOEM's scope revisions, 
and Table 2 provides the subsequent revised level of effort projections 
for the 5-year period.

   Table 1--Percentage Reduction in Survey Area for Each Modeled Zone
------------------------------------------------------------------------
                                                            Percentage
                      Modeling zone                        reduction in
                                                               area
------------------------------------------------------------------------
1.......................................................           100.0
2.......................................................             2.7
3.......................................................             0.0
4.......................................................            98.2
5.......................................................             4.0
6.......................................................             0.0
7.......................................................            33.0
------------------------------------------------------------------------


[[Page 5330]]


                                              Table 2--Projected Levels of Effort in 24-hr Survey Days for Five Years, by Zone and Survey Type \1\
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                                                                    Shallow                              Total
                            Year                               Zone \2\     2D \3\    3D NAZ \3\  3D WAZ \3\   Coil \3\     VSP \3\      Total      hazards   Boomer \4\    HRG \4\    (shallow)
                                                                                                                                      (deep) \3\      \4\                                 \4\
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1...........................................................           1           0           0           0           0           0           0           0           0           1           1
                                                                       2           0           7           0           0           0           7           0           0           1           1
                                                                       3           0           0           0           0           0           0           0           0           0           0
                                                                       4           0           0           0           0           0           0           0           0           0           0
                                                                       5           2          16           8           3           0          29           0           0           1           1
                                                                       6           0           0           0           0           0           0           0           0           0           0
                                                                       7          23         170          82          35           1         311           0           0          11          11
                                                             -----------------------------------------------------------------------------------------------------------------------------------
    Total...................................................  ..........          25         193          90          38           1         347           0           0          14          14
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
2...........................................................           1           0           0           0           0           0           0           0           0           1           1
                                                                       2           0          10           1           0           0          11           0           0           1           1
                                                                       3           0           0           0           0           0           0           0           0           0           0
                                                                       4          27           0           0           0           0          27           0           0           0           0
                                                                       5           0          16           8           3           0          27           0           0           1           1
                                                                       6           0           0           0           0           0           0           0           0           0           0
                                                                       7          10         166          79          34           1         290           0           0          11          11
                                                             -----------------------------------------------------------------------------------------------------------------------------------
    Total...................................................  ..........          37         192          88          37           1         355           0           0          14          14
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
3...........................................................           1           0           0           0           0           0           0           0           0           1           1
                                                                       2           0          10           1           0           0          11           0           0           0           0
                                                                       3           0           0           0           0           0           0           0           0           0           0
                                                                       4          54          50          21           9           0         134           0           0           1           1
                                                                       5           1          10           4           2           0          17           0           0           1           1
                                                                       6           0           0           0           0           0           0           0           0           0           0
                                                                       7          31         125          46          20           1         223           0           0          12          12
                                                             -----------------------------------------------------------------------------------------------------------------------------------
    Total...................................................  ..........          86         195          72          31           1         385           0           0          12          12
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
4...........................................................           1           0           0           0           0           0           0           0           0           0           0
                                                                       2           0          10           1           0           0          11           0           0           0           0
                                                                       3           0           0           0           0           0           0           0           0           0           0
                                                                       4          54          50          21           9           0         134           0           0           1           1
                                                                       5           1          10           4           2           0          17           0           0           1           1
                                                                       6           0           0           0           0           0           0           0           0           0           0
                                                                       7          31         125          46          20           1         223           0           0          12          12
                                                             -----------------------------------------------------------------------------------------------------------------------------------
    Total...................................................  ..........          86         195          72          31           1         385           0           0          14          14
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
5...........................................................           1           0           0           0           0           0           0           0           0           0           0
                                                                       2           0           7           0           0           0           7           0           0           1           1
                                                                       3           0           0           0           0           0           0           0           0           0           0
                                                                       4           0          75           0           0           0          75           0           0           0           0
                                                                       5           0          12           8           3           0          23           0           0           1           1
                                                                       6           0           0           0           0           0           0           0           0           0           0
                                                                       7           0         154          79          34           1         268           1           1          11          13
                                                             -----------------------------------------------------------------------------------------------------------------------------------
    Total...................................................  ..........           0         248          87          37           1         373           1           1          13          15
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Projected levels of effort in 24-hr survey days.
\2\ Zones follow the zones depicted in Figure 3.
\3\ Deep penetration survey types include 2D, which uses one source vessel with one large array (8,000 in\3\); 3D NAZ, which uses two source vessels using one large array each; 3D WAZ and
  coil, each of which uses four source vessels using one large array each (but with differing survey design); and VSP, which uses one source vessel with a large array. ``Deep'' refers to
  survey type, not to water depth.
\4\ Shallow penetration/HRG survey types include shallow hazards surveys, assumed to use a single 90 in\3\ airgun or boomer, and high-resolution surveys using the multibeam echosounder, side-
  scan sonar, and chirp sub-bottom profiler systems concurrently. ``Shallow'' refers to survey type, not to water depth.

    This description of the specified activity is a summary of critical 
information. The interested reader should refer to the notice of 
proposed rulemaking (83 FR 29212; June 22, 2018), as well as BOEM's 
petition (with recent addenda) and PEIS, for additional detail 
regarding these prospective activities and the region. Required 
mitigation, monitoring, and reporting measures are described in detail 
later in this document (please see Mitigation and Monitoring and 
Reporting).

Changes From the Proposed Rule

    This section provides a summary of changes from the proposed rule. 
Each section in which changes were made (e.g., Mitigation) includes a 
more detailed list of changes made and a fuller description of the 
rationale. The following Comments and Responses section also provides 
additional detail relating to changes, in cases where the change 
resulted from a public comment.
    Most notably, as described in greater detail above, BOEM updated 
the scope of the specified activity/specified geographical region that 
is the subject of this rule by removing from consideration the area 
that is subject to the GOMESA leasing moratorium. In accordance with 
this updated spatial scope, BOEM provided revised activity level 
projections and revised estimated acoustic exposure numbers based on 
the same modeling that informed the numbers evaluated in the proposed 
rule. BOEM's revised activity level projections correspond with Years 
1-5 of the original 10-year projections (see Table 1 of the notice of 
proposed rulemaking), which is a conservative choice as these years 
contained higher levels of effort than Years 6-10. In the proposed 
rule, NMFS selected years that were representative of different levels 
of effort as the basis for the total taking over five years, including 
one year of relatively high effort (Year 1), two years of relatively 
moderate effort (Years 4 and 5), and two years of relatively low effort 
(Years 8 and 9). This selection is now in part supplanted (with the two

[[Page 5331]]

representative ``low effort'' years replaced by one relatively high 
effort year and one relatively moderate effort year) by BOEM's 
selection of Years 1-5 and the associated updated levels of effort.
    The revised acoustic exposure numbers form the basis for our 
analyses in this final rule. Of note, the maximum total taking, as well 
as the annual maximum, that would be allowable under the regulations 
has decreased for most species and stocks, with the exception of the 
annual maximums for Atlantic spotted dolphin and bottlenose dolphins, 
and the total taking over five years for the Atlantic spotted dolphin, 
which have increased slightly (please see Estimated Take for additional 
information). These changes (largely decreases) in the take numbers do 
not have a meaningful effect on the analysis (except where impacts are 
significantly reduced, e.g., for Bryde's whales) and do not change any 
of the findings.
    In the proposed rule, NMFS included several time-area restrictions, 
including a seasonal restriction on airgun survey activity in the 
``Bryde's whale core habitat'' area (as well as alternatives to this 
proposal that were offered for public comment, including a year-round 
restriction in the same area). Following BOEM's update to the scope of 
the rule, two of these areas (the Bryde's whale area and the ``Dry 
Tortugas'' area that was, in part, designed to provide protection for 
sperm whales and beaked whales) were removed from consideration, as the 
specified activity/specified geographical region no longer includes 
surveys in the areas where these proposed restrictions are located.
    A third time-area restriction--the ``Coastal Restriction,'' 
designed to protect bottlenose dolphins in coastal waters most heavily 
impacted by the Deepwater Horizon oil spill--has been modified in 
consideration of public comments. The restriction was proposed to be 
GOM-wide within coastal waters inside the 20-m isobath, and to be in 
effect from February through May. The area encompassed by the 
restriction has been reduced to match the assumed range of the northern 
coastal stock of bottlenose dolphins (i.e., between 90-84[deg] W, but 
in effect only to the eastern extent of the coastal waters portion of 
BOEM's updated specified geographic region) while the temporal window 
has been expanded to include January. In addition, a proposed 13-km 
buffer to this area has been removed.
    In the proposed rule, NMFS defined ``deep penetration'' surveys as 
those using arrays greater than 400 in\3\ total volume. That 
delineation has been revised to include surveys using arrays greater 
than 1,500 in\3\ total volume, with arrays of 1,500 in\3\ total volume 
and less considered ``shallow penetration'' surveys.
    In the notice of proposed rulemaking, NMFS proposed an exception to 
the general shutdown requirements for certain species of dolphins in 
relation to airgun surveys, in which the acoustic source would be 
powered down to the smallest single element of the array. Power-down 
conditions would be maintained until the animal(s) is observed exiting 
the exclusion zone or for 15 minutes beyond the last observation of the 
animal, following which full-power operations may be resumed without 
ramp-up. NMFS also provided an alternative proposal for consideration 
by the public, in which no shutdown or power-down would be required 
upon observation of the same species of dolphins. Following review of 
public comments, NMFS removes the power-down measure for small 
delphinids, in favor of the no-shutdown and no power-down alternative. 
No shutdown or power-down is required for these species.
    NMFS proposed a number of extended distance shutdown requirements 
on the basis of detections of certain species deemed particularly 
sensitive (e.g., beaked whales) or of particular circumstances deemed 
to warrant the extended distance shutdown requirement (e.g., whales 
with calves). These extended distance shutdowns were all conditioned 
upon observation or detection of these species or circumstances ``at 
any distance'' from the vessel. However, NMFS also included as an 
alternative proposal for public consideration a distance limit of 1,000 
m for these shutdown requirements. Following review of public comments, 
NMFS determined that a distance limit on extended shutdown zones for 
relevant species or circumstances was appropriate, but determined 1,500 
m was the appropriate distance (rather than 1,000 m).
    The proposed rule included an extended distance shutdown for sperm 
whales that was applicable upon acoustic detection, but was not 
applicable to visual detection. Following review of public comments, 
the shutdown requirement has been expanded to include any detection of 
sperm whales within the extended distance shutdown zone, including 
visual detection.
    For shallow penetration surveys, NMFS reduces the standard 
exclusion zone from 200 m to 100 m, while including an extended 
distance shutdown requirement mirroring the requirements for deep 
penetration surveys, but within a distance of 500 m. NMFS eliminates 
shutdown requirements for HRG surveys (defined here as surveys using 
electromechanical sources such as multi-beam echosounders, side-scan 
sonars, and chirp sub-bottom profilers). The proposed regulations 
required shutdown for marine mammals within the proposed exclusion zone 
for surveys operating in water depths greater than 200 m.
    NMFS eliminates proposed requirements for visual observation during 
nighttime ramp-up and pre-clearance, and for the use of third-party 
PSOs aboard node retrieval vessels.
    In the proposed rule, NMFS discussed the use of an extrapolation 
method recommended by the Marine Mammal Commission for use in 
estimating potential unobserved takes. NMFS agrees with public 
commenters that the appropriateness of the method for application to 
observations conducted from working source vessels (versus research 
vessels) is unknown and, as suggested through public comment, NMFS will 
not require use of this method but will continue to evaluate approaches 
for assessment of effects to marine mammal stocks, including those 
based on extrapolation of marine mammal detections, through the 
adaptive management process and subsequently apply them through LOAs as 
appropriate.
    NMFS has revised requirements relating to reporting of injured or 
dead marine mammals and has added newly crafted requirements relating 
to actions that should be taken in response to notification of live 
stranding events in certain circumstances, in order to reflect current 
best practice.
    The proposed rule indicated that LOA applications with take 
estimates based on modeling other than that specifically included in 
the modeling report used to support the EIS and the proposed rule (the 
modeling report; Zeddies et al., 2015, 2017a) would necessarily be 
published for public comment prior to the issuance of an LOA. Upon 
consideration of public comment and related supplemental materials, the 
final rule more flexibly allows that if applicants do not use the 
modeling provided by the rule, NMFS will publish a notice in the 
Federal Register soliciting public comment, when the model or inputs 
differ substantively from those that have been reviewed by NMFS and the 
public previously. Please see the Letters of Authorization section for 
more detail.

[[Page 5332]]

Comments and Responses

    NMFS published a notice of proposed rulemaking in the Federal 
Register on June 22, 2018 (83 FR 29212), beginning a 60-day comment 
period. In that notice, we requested public input on the proposed rule 
and regulations, including the variations of the proposed rule, two 
economic baselines, and other information provided in the Regulatory 
Impact Analysis and associated appendices, and requested that 
interested persons submit relevant information, suggestions, and 
comments. In response to BOEM's change in scope and in consideration of 
public comments, we modified our action, as discussed in the following 
responses to comments. Please also see the Changes from the Proposed 
Rule section, above. We note that one area of significant concern for 
some members of the public was potential impacts to Bryde's whales and 
related mitigation measures. The reduced geographic scope eliminates 
the need to consider activity in the Bryde's whale ``core habitat 
area'' and eliminates the majority of the incidental take of Bryde's 
whale that was evaluated in the proposed rule.
    During the 60-day comment period, we received 17 comment letters. A 
letter was submitted jointly by the International Association of 
Geophysical Contractors, the American Petroleum Institute, the National 
Ocean Industries Association, and the Offshore Operators Committee 
(hereafter, the ``Associations''). A separate letter was submitted 
jointly by the Natural Resources Defense Council (NRDC), Center for 
Biological Diversity, Earthjustice, Gulf Restoration Network, Humane 
Society Legislative Fund, The Humane Society of the United States, and 
Sierra Club (hereafter, ``NRDC''). Additional letters were submitted by 
the following: BP Exploration & Production Inc. (BP), Consumer Energy 
Alliance, CGG, Chevron USA Inc. (Chevron), the Center for Regulatory 
Effectiveness (CRE), the Florida Department of Environmental 
Protection, the Marine Mammal Commission (MMC), and eight private 
citizens. NMFS has reviewed all public comments received on the 
proposed rulemaking. All relevant comments and our responses are 
described below, with comment responses outlined by major categories. 
All comments received are available online at: www.regulations.gov. A 
direct link to these comments is provided at: www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico.

General Comments

    As an initial matter, we note that under the MMPA, NMFS generally 
does not have discretion regarding issuance of requested incidental 
take authorizations for small numbers of marine mammals, provided that 
(1) the total taking associated with a specified activity will have a 
negligible impact on the affected species or stock(s); (2) the total 
taking associated with a specified activity will not have an 
unmitigable adverse impact on the availability of the species or 
stock(s) for subsistence uses (not relevant here); and (3) mitigation, 
monitoring, and reporting of such takings are set forth, including 
mitigation measures sufficient to meet the standard of least 
practicable adverse impact on the affected species or stocks and their 
habitat.
    In addition, NMFS' proposed action--the issuance of the ITR and any 
subsequent LOAs authorizing incidental take of marine mammals--
addresses only marine mammals (and their habitat). As such, effects of 
the surveys on other aspects of the marine environment are not relevant 
to NMFS' analyses under the MMPA.
    The MMPA does require that we evaluate potential effects to marine 
mammal habitat, which includes prey species (e.g., zooplankton, fish, 
squid). However, consideration of potential effects to taxa other than 
marine mammals and their prey, or consideration of effects to potential 
prey species in a context other than the import of such effects on 
marine mammals, is not relevant to our action under the MMPA. We have 
appropriately considered effects to marine mammal habitat. Separately, 
BOEM evaluated effects to all relevant aspects of the human environment 
(including marine mammals and other taxa) through the analysis 
presented in BOEM's PEIS (available online at: www.boem.gov/Gulf-of-Mexico-Geological-and-Geophysical-Activities-Programmatic-EIS/), and 
effects to all potentially affected species that are listed under the 
Endangered Species Act (ESA) and any critical habitat designated for 
those species were addressed through consultation between BOEM and NMFS 
pursuant to section 7 of the ESA. That Biological Opinion, which 
evaluated NMFS' proposed action (issuance of the ITR and any subsequent 
LOAs) as well as all BOEM and Bureau of Safety and Environmental 
Enforcement (BSEE) approvals of activities associated with the OCS oil 
and gas program in the GOM, is available online at: 
www.fisheries.noaa.gov/national/endangered-species-conservation/biological-opinions-issued-noaa-fisheries-office-protected. We do not 
further address taxa other than marine mammals and marine mammal prey.
    Comment: The Associations comment that the proposed ITR is a well-
structured and thorough document that appropriately concludes that 
geophysical activities in the GOM would have no more than a negligible 
impact on marine mammal populations, and that they appreciate NMFS' 
effort in preparing the proposed ITR and consideration of some of the 
Associations' previous comments.
    Response: NMFS appreciates the comment.
    Comment: The Associations comment that geophysical surveys play a 
critical role in the safe and orderly development of the oil and gas 
resources of the GOM.
    Response: We acknowledge the background operational information 
provided by the Associations.
    Comment: BP comments that the ITR is a much-needed process to 
govern the authorization of incidental takes of marine mammals 
associated with geophysical survey activity in the GOM. Chevron also 
indicates support for promulgation of the ITRs.
    Response: NMFS appreciates the comments.
    Comment: BP comments that projected survey efforts are 
underestimated but did not provide specific justification or 
recommendations.
    Response: Projected levels of survey effort were formulated by BOEM 
and included in their PEIS. BOEM's PEIS stated, ``the scenarios contain 
projections based on the analysis of recent historic activity levels 
and trends made by BOEM's subject-matter experts who also considered 
industry-projected activity levels in their estimates.'' These 
projected levels of survey effort were made available for public review 
on multiple occasions during the development of the PEIS, as well as 
during the notice of receipt comment period, in which the public was 
given the opportunity to review and comment on the petition itself (81 
FR 88664; December 8, 2016). Neither BP nor other industry stakeholders 
submitted comments on the BOEM-developed effort levels, and no evidence 
was provided that projected survey efforts are underestimated. The 
projected levels of effort were subsequently updated by BOEM based on 
the removal of the GOMESA area from consideration.
    Comment: The Florida Department of Environmental Protection (FLDEP) 
expressed its concern regarding the potential impacts of OCS oil and 
gas

[[Page 5333]]

activities on marine and coastal environments and the biological 
resources and critical habitats associated with them. The FLDEP also 
indicated that former Secretary of the Interior Zinke had made a 
commitment to former Governor Scott to remove the State of Florida from 
future consideration for offshore drilling.
    Response: NMFS acknowledges the comments. Assuming that the 
requirements of the MMPA are met, e.g., findings of negligible impact 
and small numbers are made, NMFS does not have discretion as to whether 
it may issue ITRs and LOAs under those ITRs, and NMFS has no authority 
to limit oil and gas activities outside of prescribing appropriate 
mitigation requirements.

Marine Mammal Impacts

    Comment: The Associations (as well as other industry commenters and 
the CRE) stated, in summary, that there is no scientific evidence that 
geophysical survey activities have caused adverse consequences to 
marine mammal stocks or populations, and that there are no known 
instances of injury to individual marine mammals as a result of such 
surveys, stating that similar surveys have been occurring for years 
without significant impacts. The Associations stated that surveys have 
been ongoing in the GOM for years and have not resulted in any negative 
impacts to marine mammals, including reducing fitness in individuals or 
populations. Referring to other regions, the commenters stated that 
bowhead whale numbers have increased in the Arctic despite survey 
activity. The Associations go further in claiming that ``NMFS 
misconstrues its legal obligations'' and ``NMFS violates the MMPA's 
best available science requirement.''
    Response: Disruption of behavioral patterns (i.e., Level B 
harassment) has been documented numerous times for marine mammals in 
the presence of airguns, in the form of avoidance of areas, notable 
changes in vocalization or movement patterns, or other shifts in 
important behaviors. See Potential Effects of the Specified Activity on 
Marine Mammals and Their Habitat in the notice of proposed rulemaking. 
In addition, there is growing scientific evidence demonstrating the 
connections between sub-lethal effects, such as behavioral disturbance, 
and population-level effects on marine mammals (e.g., Lusseau and 
Bedjer, 2007; New et al., 2014; Pirotta et al., 2018). Disruptions of 
important behaviors, in certain contexts and scales, have been shown to 
have energetic effects that can translate to reduced survivorship or 
reproductive rates of individuals (e.g., feeding is interrupted, so 
growth, survivorship, or ability to bring young to term may be 
compromised), which in turn can adversely affect populations depending 
on their health, abundance, and growth trends.
    With specific regard to sound, as a 2017 report from the National 
Academy of Sciences noted, while it is true that ``[n]o scientific 
studies have conclusively demonstrated a link between exposure to sound 
and adverse effects on a marine mammal population,'' this is largely 
because such impacts are very difficult to demonstrate (NRC, 2005; NAS, 
2017), not because they do not exist. Population[hyphen]level effects 
are inherently difficult to assess because of high variability, 
migrations, and multiple factors affecting the populations. Appropriate 
studies are exceedingly difficult to carry out, and no appropriate 
study and reference populations have yet been established. Nonetheless 
there is a growing body of literature and science illustrating the 
connections between prolonged behavioral disturbance and impacts to 
reproductive success and survivorship. Accordingly, it is not 
defensible to conclude that sub-lethal acoustic stressors cannot have 
population level consequences. Based on the available evidence, a 
sufficient analysis of the potential impacts of airgun noise requires 
consideration of impacts on individuals and the potential for 
population level effects. NMFS has carefully considered the available 
evidence in making the necessary determinations (see Negligible Impact 
Analysis and Determinations) and determining the most appropriate suite 
of mitigation measures.
    Because some commenters repeatedly cite (and misunderstand) public 
statements by BOEM in support of a contention that there is ``no harm 
from seismic,'' we clarify the record by citing BOEM's own responses to 
similar comments on their PEIS (BOEM, 2017). BOEM stated: ``It is 
critically important to understand that BOEM's . . . Science Note . . . 
refers to impacts on marine mammal . . . population sustainability 
rather than effects on individual animals. Studies have shown that 
marine mammals may and do react to sound through physical displacement 
from or avoidance of the area of ensonification and/or by altering 
their vocalizations. This [PEIS] acknowledges that significant acute 
physical injury to or death of marine mammals is not likely to be a 
direct result of seismic noise. It does, however, acknowledge that 
sublethal injurious effects are possible and may, over time, result in 
the eventual death of the individual(s) from these physical injuries 
and/or loss of hearing with (as in the case of marine mammals) the 
resultant inability to forage and communicate with conspecifics. 
Another prominent concern is whether anthropogenic sounds such as those 
generated during seismic survey activities may ``mask'' communications 
between some marine mammals. Depressed survival rates related to 
energetic effects or other impacts of noise are difficult to determine. 
BOEM, however, does not assume that lack of demonstrated adverse 
population-level effects from seismic surveys means that those effects 
may not occur.''
    In support of assertions that there are ``no effects'' to marine 
mammals from seismic surveys and that there is a ``lack of any harm'' 
to marine mammals, CRE cites statements made by NMFS, in which we 
conclude that there is no evidence that serious injury, death, or 
stranding is reasonably likely to occur as a result of such surveys, 
and that Level A harassment is not reasonably likely to occur for mid-
frequency cetaceans. CRE's assertion that there are ``no effects'' and 
``no harm'' to marine mammals as a result of seismic surveys is based 
on the fact that marine mammals still exist in the GOM despite survey 
activity. CRE overlooks the evidence put forward for Level B 
harassment, and the potential effects of behavioral disruption, as well 
as the additional effects of noise that do not rise to the level of a 
take, but which nevertheless must be considered when evaluating the 
effects of a specified activity on a species or stock.
    The Associations assert that we premise our decisions on the idea 
that we must act conservatively because effects that have not been 
conclusively proven--which the Associations claim, without evidence, do 
not and cannot occur--could occur in the future. The Associations state 
that we misconstrue our legal obligations via the application of ``an 
additional layer of precautionary bias'' beyond that established in the 
MMPA standards themselves, though they do not demonstrate that the bias 
exists. The Associations acknowledge that the MMPA requires mitigation 
sufficient to meet the standard of least practicable adverse impact. 
Therefore, some portion of the mitigation requirements contained in the 
proposed ITR would be necessary to meet that standard. However, they 
provide no analysis to support the contention that specific mitigation 
requirements exceed that standard. In fact, we have declined to adopt 
the recommendations of other commenters that are based on vague and 
unexplained standards of

[[Page 5334]]

``conservatism'' that are not required in the MMPA. Here, we conducted 
the requisite analyses of mitigation and found that the requirements 
contained in this final ITR, as modified on the basis of new 
information and review of public comments, meet the least practicable 
adverse impact (LPAI) standard.
    We base our conclusions, relating to the potential effects of the 
specified activity on the affected species and stocks, on reasonable 
interpretation of the available science, which we summarize in this 
preamble and described in detail in our notice of proposed rulemaking. 
While we acknowledge the lack of conclusive evidence for population-
level consequences, this is an artifact of the extreme difficulty of 
empirically demonstrating such effects (as concluded by the National 
Academies of Science, stated above). The best available scientific 
information provides considerable evidence that the activities 
evaluated in this ITR have the potential to adversely affect the 
fitness of individual animals. The best available science clearly 
demonstrates that, given adverse impacts to an animal's fitness, 
population-level effects are plausible. The Associations' comments on 
this topic treat the lack of empirical evidence as evidence that such 
effects do not occur. However, NMFS does not agree that absence of 
evidence is evidence of absence of effects. The comments further 
incorrectly frame our decision-making as being premised on the idea 
that such effects could occur in the future, when they are actually 
based on a reasonable interpretation of the best available scientific 
information regarding what the effects of the specified activity are 
likely to be in the absence of prescribed mitigation. Despite the 
paucity of empirical research on population effects, the best available 
information demonstrates impacts at the individual level that, at a 
high enough level of take, have reasonably foreseeable population-level 
impacts.
    Similarly, the Associations imply that our interpretation of the 
existing scientific information reflects speculation about what future 
research might demonstrate. The Associations' statements that NMFS 
dismissed current scientific findings and premised decisions on 
hypothesized future impacts are inaccurate, and their assertion that 
NMFS ``has effectively required conclusive scientific proof that 
seismic surveys do not impact marine mammal populations'' 
misunderstands NMFS' use of the scientific literature. The best 
available information demonstrates that the effects of seismic surveys 
on marine mammals may include adverse impacts on behavior in ways that 
can also have energetic consequences. To draw different conclusions 
regarding the need for the strong suite of mitigation requirements 
included in this final ITR, NMFS would require scientific evidence that 
demonstrates that seismic surveys do not have energetic consequences 
or, alternatively, do not reach a point where there are population-
level consequences. NMFS is not aware of such evidence. NMFS' final 
rule is based on the best available scientific information and the 
requirements of the MMPA.
    Chevron states that we do not account for ``real-world'' protected 
species observer (PSO) observations, calling this ``arbitrary and 
capricious,'' and seems to imply that these ``ignored'' PSO 
observations of marine mammals are evidence that seismic activities 
produce no more than ``negligible effects on species.'' Chevron does 
not provide evidence to support its comment or otherwise develop the 
suggestion to enable a specific response. However, we incorporated the 
best available scientific information for our analysis, as evidenced 
(for example) by our references in the notice of proposed rulemaking to 
BOEM's synthesis study of PSO data from 2002-08 (Barkaszi et al., 2012) 
(as well as other similar syntheses from other locations). In this 
final rulemaking, we have incorporated analysis of a newly available 
study of PSO data from 2009-15 (Barkaszi and Kelly, 2018). These data 
are also key to the evaluation of direct costs found in our RIA. We 
disagree with Chevron's apparent contention that we ``ignore[d]'' 
BOEM's earlier ``admissions that no scientific evidence exists 
contradicting the real-world observations of negligible impact'' 
(citing to BOEM's ``Science Notes''). NMFS addressed BOEM's ``Science 
Notes'' in some detail in our notice of proposed rulemaking (83 FR 
29264-65). Chevron misinterprets a statement from BOEM regarding the 
absence of evidence (``no documented evidence of noise from air guns . 
. . adversely affecting animal populations) as evidence itself of no 
adverse effects. According to Chevron, our ``failure to account for'' 
this is ``arbitrary and capricious.'' These issues have been addressed 
both above and in the notice of proposed rulemaking.
    Comment: NRDC referenced studies showing that noise from airgun 
surveys can travel great distances underwater, suggesting that due to 
the scale of this propagation, marine mammals in the GOM are 
consistently compromised in their ability to perform important life 
functions.
    Response: NMFS acknowledges that relatively loud, low-frequency 
noise (as is produced by airgun arrays) has the potential to propagate 
across large distances. However, propagation and received sound levels 
are highly variable based on many biological and environmental factors. 
For example, while one commonly cited study (Nieukirk et al., 2012) 
described detection of airgun sounds almost 4,000 km from the acoustic 
source, the sensors were located within the deep sound channel (SOFAR), 
where low-frequency signals may travel great distances due to the 
advantageous propagation environment. While sounds within this channel 
are unlikely to be heard by most marine mammals due to the depth of the 
SOFAR channel--which is dependent primarily on temperature and water 
pressure and therefore variable with latitude--it is arguable whether 
sounds that travel such distances may be heard by whales as a result of 
refraction to shallower depths (Nieukirk et al., 2012; McDonald et al., 
1995). Regardless, while the extreme propagation distances cited in 
some comments may not be realistic, we acknowledge that contraction of 
effective communication space for Bryde's whales, which vocalize and 
hear at frequencies overlapping those emitted by airgun arrays, can 
occur at distances on the order of tens to hundreds of kilometers 
(e.g., Hatch et al., 2012). However, attenuation to levels below which 
more acute effects are likely to occur is expected over much shorter 
distances (Zeddies et al., 2015, 2017a) and, therefore, we do not agree 
with the contention that the GOM would be ensonified to a degree that 
marine mammals would find it an unsuitable habitat or would be 
consistently compromised in their ability to perform important life 
functions. Rather, it is likely that displacement would occur within a 
much smaller region in the vicinity of the acoustic source (e.g., 
within 10-20 km of the source, depending on season and location). 
Overall, the specific geographic region and marine mammal use of the 
area is sufficiently large that, although some displacement may occur 
(i.e., Level B harassment as a result of acoustic exposure beyond the 
exclusion zone), the GOM offers enough habitat for marine mammals to 
seek temporary viable habitat elsewhere, if necessary. Many of the 
affected species occupy a wide portion of the GOM, and it is expected 
that individuals of these species can reasonably find temporary

[[Page 5335]]

foraging grounds or other suitable habitat areas consistent with their 
natural use of the region. Further, although the surveys are expected 
to occur over large portions of the GOM, they will only be transitory 
in any given area. Therefore, NMFS does not expect displacement to 
occur frequently or for long durations. Please see Negligible Impact 
Analysis and Determinations for additional analysis.
    Comment: NRDC states that airgun surveys have been linked to 
significant reductions in the probability of calf survival in western 
Pacific gray whales (an endangered baleen whale population), implying 
that these findings indicate that such surveys would similarly have 
significant negative effects on whales in the GOM.
    Response: Commenters cite a preliminary report (Cooke et al., 2015) 
that documented a reduction in calf survival that the authors suggested 
may be related to disruption of foraging from airgun survey activity 
and pile driving in Russia due to presumed avoidance of foraging areas. 
However, a more recent analysis (Cooke et al., 2017) invalidated these 
findings, showing that this was a sampling effect, as those calves that 
were assumed dead in the 2015 study have since been observed alive 
elsewhere. The new study found no significant annual variation in calf 
survival. Johnson et al. (2007) had previously reported that foraging 
gray whales exposed to airgun sounds during surveys in Russia did not 
experience any biologically significant or population-level effects.
    Comment: NRDC asserts that we have not adequately accounted for 
vessel collision risk, stating that the surveys will drive marine 
mammals into shipping lanes, thereby increasing their risk of ship 
strike. Relatedly, NRDC noted that NMFS' conclusion that ship strikes 
will not occur indicates an assumption that required ship-strike 
avoidance procedures will be effective. NRDC disagrees that the ship-
strike avoidance measures will be effective.
    Response: NMFS is not aware of any scientific information 
suggesting that the surveys would drive marine mammals into shipping 
lanes and disagrees that this would be a reasonably anticipated effect 
of the specified activities. While the primary stressor to marine 
mammals from the specified activities is acoustic exposure to the sound 
source, NMFS takes seriously the risk of vessel strike and has 
prescribed measures sufficient to avoid the potential for ship strike 
to the extent practicable (see Mitigation). NMFS has required these 
measures despite a very low likelihood of vessel strike; vessels 
associated with the surveys will add a discountable amount of vessel 
traffic to the specific geographic region and, furthermore, vessels 
towing survey gear travel at very slow speeds (i.e., roughly 4-5 kn).
    Comment: The MMC criticizes one aspect of the methodology for the 
analysis of chronic effects to Bryde's and sperm whales conducted by 
NMFS with the support of JASCO Applied Sciences (JASCO), i.e., removing 
the top ten percent of the greatest pulse exposures. (JASCO is a 
consulting company contracted by NMFS and BOEM to model acoustic 
exposures of marine mammals to noise produced by industry survey 
activity.) The MMC recommends re-estimation of the various lost 
listening and communication space parameters without removing the 
greatest ten percent of pulse exposures.
    Response: The goal of this modeling exercise was to create a tool 
that could help evaluate loss of ability to detect signals of 
biological importance over spatial scales relevant to the sources and 
hearing capabilities of a wide variety of regional animals. In order to 
do so, we attempt to examine the portion of low-frequency acoustic 
energy lost from seismic surveys that has been empirically measured in 
many contexts around the world to generate higher chronic, longer-term 
average noise levels. Masking experienced by individual calling and 
receiving animals due to noise at relatively close proximity to a 
single intermittent source is an important but limited aspect of the 
real-world contexts within which populations of marine mammals are 
exposed to noise from multiple seismic surveys in a region like the 
GOM. This modeling sought to account for the known attributes of airgun 
noise, by which low-frequency energy lost laterally attenuates over 
large spatial scales with loss of impulsive features, leading to 
elevated background noise conditions, particularly when multiple 
surveys are concurrent within an acoustic region. Close range pulse 
energy would entirely drown out such evaluation, and would not account 
for the different acoustic characteristics of the signal and potential 
masking at such scales. Thus, while masking of specific signals 
relative to the near-field of operating airgun arrays is an impact that 
may occur, for the purposes of the analysis conducted for this rule, 
near-field impacts have been addressed through the modeling of acoustic 
exposures. The chronic and cumulative impacts analysis that is the 
subject of this comment addresses far-field chronic impacts. 
Additionally, there are technical concerns with modifying the analysis 
specifically as recommended and, accordingly, we disagree with the 
recommendation for purposes of this analysis of potential chronic 
effects.
    The purpose of this modeling exercise was not to evaluate exposure 
implications for animals close to the modeling locations (i.e., 
``acute'' effects). Evaluation of acute effects, such as injury and 
behavioral disruption, was achieved through the primary acoustic 
exposure modeling effort (Zeddies et al., 2015, 2017a). These evaluated 
effects (evaluated through the primary acoustic modeling effort) are 
separate and separable from loss of hearing opportunities experienced 
by animals farther from source locations, which are evaluated through 
the chronic and cumulative effects modeling discussed here.

Marine Mammal Impacts--Habitat

    Comment: NRDC expressed concern regarding potential impacts to 
marine mammal prey and/or food webs from the planned surveys. NRDC 
provided numerous citations in claiming that the surveys could impact 
marine mammal prey through the following: (1) Cause severe physical 
injury and mortality; (2) damage hearing and sensory abilities of fish 
and marine invertebrates; (3) impede development of early life history 
stages; (4) induce stress that physically damages marine invertebrates 
and compromises fish health; (5) cause startle and alarm responses that 
interrupt vital behaviors; (6) alter predator avoidance behavior that 
may reduce probability of survival; (7) affect catchability of prey 
species; (8) mask important biological sounds essential to survival; 
(9) reduce reproductive success, potentially jeopardizing long-term 
sustainability of fish populations; (10) interrupt feeding behaviors 
and induce other species-specific effects that may increase risk of 
starvation, reduce reproduction, and alter community structure; and 
(11) compromise orientation of fish larvae with potential ecosystem-
level effects. Additionally, NRDC cited a publication by McCauley et 
al. (2017) as evidence that the surveys could potentially impact 
zooplankton and consequently marine mammal food webs.
    Response: NMFS strongly disagrees with the suggestion that we 
ignored effects to prey species. In fact, we considered relevant 
literature (including that cited by NRDC) in finding that the most 
likely impact of survey activity to prey species such as fish and 
invertebrates would be temporary avoidance of an area, with a rapid 
return to pre-survey distribution and behavior,

[[Page 5336]]

and minimal impacts to recruitment or survival anticipated. While there 
is a lack of specific scientific information to allow an assessment of 
the duration, intensity, or distribution of effects to prey in specific 
locations at specific times and in response to specific surveys, NMFS' 
review of the available information does not indicate that such effects 
could be significant enough to impact marine mammal prey to the extent 
that marine mammal fitness would be affected. We agree that seismic 
surveys could affect certain marine mammal prey species, and addressed 
these potential effects, as well as the potential for those effects to 
impact marine mammal populations, in our notice of proposed rulemaking 
(83 FR 29241-29242). As stated in that notice, our review of the 
available information and the specific nature of the activities 
considered herein suggest that the activities evaluated in this ITR are 
not likely to have more than short-term adverse effects on any prey 
habitat or populations of prey species. Further, any impacts to prey 
species are not expected to result in significant or long-term 
consequences for individual marine mammals, or to contribute to adverse 
impacts on their populations. In support of this conclusion, we refer 
the commenter to discussion provided in our notice of proposed 
rulemaking. Additional information is summarized below.
    In summary, fish react to sounds which are especially strong and/or 
intermittent low-frequency sounds, and behavioral responses such as 
flight or avoidance are the most likely effects. However, the reaction 
of fish to airguns depends on the physiological state of the fish, past 
exposures, motivation (e.g., feeding, spawning, migration), and other 
environmental factors. While we agree that some studies have 
demonstrated that airgun sounds might affect the distribution and 
behavior of some fishes, potentially impacting foraging opportunities 
or increasing energetic costs (e.g., Fewtrell and McCauley, 2012; 
Pearson et al., 1992; Skalski et al., 1992; Santulli et al., 1999; 
Paxton et al., 2017), our review shows that the weight of evidence 
indicates either no or only a slight reaction to noise (e.g., Miller 
and Cripps, 2013; Dalen and Knutsen, 1987; Pena et al., 2013; Chapman 
and Hawkins, 1969; Wardle et al., 2001; Sara et al., 2007; Jorgenson 
and Gyselman, 2009; Blaxter et al., 1981; Cott et al., 2012; Boeger et 
al., 2006), and that, most commonly, while there may be impacts to fish 
as a result of noise from nearby airguns, any effects will be 
temporary. For example, investigators reported significant, short-term 
declines in commercial fishing catch rate of gadid fishes during and 
for up to five days after seismic survey operations, but the catch rate 
subsequently returned to normal (Engas et al., 1996; Engas and 
Lokkeborg, 2002). Other studies have reported similar findings (e.g., 
Hassel et al., 2004). Skalski et al. (1992) also found a reduction in 
catch rates--for rockfish (Sebastes spp.) in response to controlled 
airgun exposure--but suggested that the mechanism underlying the 
decline was not dispersal but rather decreased responsiveness to baited 
hooks associated with an alarm behavioral response. A companion study 
showed that alarm and startle responses were not sustained following 
the removal of the sound source (Pearson et al., 1992). Therefore, 
Skalski et al. (1992) suggested that the effects on fish abundance may 
be transitory, primarily occurring during the sound exposure itself. In 
some cases, effects on catch rates are variable within a study, which 
may be more broadly representative of temporary displacement of fish in 
response to airgun noise (i.e., catch rates may increase in some 
locations and decrease in others) than any long-term damage to the fish 
themselves (Streever et al., 2016).
    SPLs of sufficient strength have been known to cause injury to fish 
and fish mortality and, in some studies, fish auditory systems have 
been damaged by airgun noise (McCauley et al., 2003; Popper et al., 
2005; Song et al., 2008). However, in most fish species, hair cells in 
the ear continuously regenerate and loss of auditory function likely is 
restored when damaged cells are replaced with new cells. Halvorsen et 
al. (2012b) showed that a temporary threshold shift (TTS) of 4-6 dB was 
recoverable within 24 hours for one species. Impacts would be most 
severe when the individual fish is close to the source and when the 
duration of exposure is long--both of which are conditions unlikely to 
occur for surveys that are necessarily transient in any given location 
and likely result in brief, infrequent noise exposure to prey species 
in any given area. For these surveys, the sound source is constantly 
moving, and most fish would likely avoid the sound source prior to 
receiving sound of sufficient intensity to cause physiological or 
anatomical damage. In addition, ramp-up may allow certain fish species 
the opportunity to move further away from the sound source.
    NMFS considered the research provided by NRDC and disagrees with 
its interpretation of the literature. A recent comprehensive review 
(Carroll et al., 2017) found that results are mixed as to the effects 
of airgun noise on the prey of marine mammals. While some studies 
suggest a change in prey distribution and/or a reduction in prey 
abundance following the use of seismic airguns, others suggest no 
effects or even positive effects in prey abundance. As one specific 
example--regarding Paxton et al. (2017), which describes findings 
related to the effects of a 2014 seismic survey on a reef off of North 
Carolina--NRDC asserts that the study supports a conclusion that 
seismic surveys ``cause significant shifts in distribution that may 
compromise life history behaviors.'' However, our own review of this 
work shows that a reasonable interpretation leads to a more moderate 
conclusion. While the study did show a 78 percent decrease in observed 
nighttime abundance for certain species--which NRDC interprets as a 
significant shift in distribution that could compromise life history 
behaviors--it is important to note that the evening hours during which 
the decline in fish habitat use was recorded (via video recording) 
occurred on the same day that the seismic survey passed, and no 
subsequent data is presented to support an inference that the response 
was long-lasting. Additionally, given that the finding is based on 
video images, the lack of recorded fish presence does not support a 
conclusion that the fish actually moved away from the site or suffered 
any serious impairment because fish may remain present yet not be 
recorded on video. In summary, this particular study corroborates prior 
studies demonstrating a startle response or short-term displacement.
    Available data suggest that cephalopods are capable of sensing the 
particle motion of sounds and detect low frequencies up to 1-1.5 kHz, 
depending on the species, and so are likely to detect airgun noise 
(Kaifu et al., 2008; Hu et al., 2009; Mooney et al., 2010; Samson et 
al., 2014). Auditory injuries (lesions occurring on the statocyst 
sensory hair cells) have been reported upon controlled exposure to low-
frequency sounds, suggesting that cephalopods are particularly 
sensitive to low-frequency sound (Andre et al., 2011; Sole et al., 
2013). Behavioral responses, such as inking and jetting, have also been 
reported upon exposure to low-frequency sound (McCauley et al., 2000b; 
Samson et al., 2014). Similar to fish, however, the transient nature of 
the surveys leads to an expectation that effects will be largely 
limited to

[[Page 5337]]

behavioral reactions and would occur as a result of brief, infrequent 
exposures.
    We discussed impacts to benthic communities from impulsive sound 
generated by active acoustic sound sources in our notice of proposed 
rulemaking, including one study showing that exposure to airgun signals 
was found to significantly increase mortality in scallops, in addition 
to causing significant changes in behavioral patterns and disruption of 
hemolymph chemistry during exposure (although the authors state that 
the observed levels of mortality were not beyond naturally occurring 
rates) (Day et al., 2017). In addition, Fitzgibbon et al. (2017) found 
significant changes to hemolymph cell counts in spiny lobsters 
subjected to repeated airgun signals, with the effects lasting up to a 
year post-exposure. However, despite the high levels of exposure, 
direct mortality was not observed. Further, in reference to the study, 
Day et al. (2016) stated that ``[s]eismic surveys appear to be unlikely 
to result in immediate large scale mortality [. . .] and, on their own, 
do not appear to result in any degree of mortality'' and that ``[e]arly 
stage lobster embryos showed no effect from air gun exposure, 
indicating that at this point in life history, they are resilient to 
exposure and subsequent recruitment should be unaffected.'' A majority 
of the studies reviewed by NMFS have observed no increased mortality in 
invertebrates exposed to airgun noise (e.g., Wardle et al., 2001; Parry 
et al., 2002; Christian et al., 2003; Andriguetto-Filho et al., 2005; 
Parry and Gason, 2006; Payne et al., 2007; Harrington et al., 2010; 
Przeslawski et al., 2018).
    With regard to potential impacts on zooplankton, McCauley et al. 
(2017) found that exposure to airgun noise resulted in significant 
depletion for more than half the taxa present and that there were two 
to three times more dead zooplankton after airgun exposure compared 
with controls for all taxa, within 1 km of the airguns. However, the 
authors also stated that in order to have significant impacts on r-
selected species (i.e., those with high growth rates and that produce 
many offspring) such as plankton, the spatial or temporal scale of 
impact must be large in comparison with the ecosystem concerned, and it 
is possible that the findings reflect avoidance by zooplankton rather 
than mortality (McCauley et al., 2017). In addition, the results of 
this study are inconsistent with a large body of research that 
generally finds limited spatial and temporal impacts to zooplankton as 
a result of exposure to airgun noise (e.g., Dalen and Knutsen, 1987; 
Payne, 2004; Stanley et al., 2011). Most prior research on this topic, 
which has focused on relatively small spatial scales, has showed 
minimal effects (e.g., Kostyuchenko, 1973; Booman et al., 1996; 
S[aelig]tre and Ona, 1996; Pearson et al., 1994; Bolle et al., 2012).
    A modeling exercise was conducted as a follow-up to the McCauley et 
al. (2017) study (as recommended by McCauley et al.), in order to 
assess the potential for impacts on ocean ecosystem dynamics and 
zooplankton population dynamics (Richardson et al., 2017). Richardson 
et al. (2017) found that a full-scale airgun survey would impact 
copepod abundance within the survey area, but that effects at a 
regional scale were minimal (2 percent decline in abundance within 150 
km of the survey area and effects not discernible over the full 
region). The authors also found that recovery within the survey area 
would be relatively quick (3 days following survey completion), and 
suggest that the quick recovery was due to the fast growth rates of 
zooplankton, and the dispersal and mixing of zooplankton from both 
inside and outside of the impacted region. The authors also suggest 
that surveys in areas with more dynamic ocean circulation in comparison 
with the study region and/or with deeper waters (i.e., typical GOM 
survey locations) would have less net impact on zooplankton.
    Notably, a recently described study produced results inconsistent 
with those of McCauley et al. (2017). Researchers conducted a field and 
laboratory study to assess if exposure to airgun noise affects 
mortality, predator escape response, or gene expression of the copepod 
Calanus finmarchicus (Fields et al., 2019). Immediate mortality of 
copepods was significantly higher, relative to controls, at distances 
of 5 m or less from the airguns. Mortality one week after the airgun 
blast was significantly higher in the copepods placed 10 m from the 
airgun but was not significantly different from the controls at a 
distance of 20 m from the airgun. The increase in mortality, relative 
to controls, did not exceed 30 percent at any distance from the airgun. 
Moreover, the authors caution that even this higher mortality in the 
immediate vicinity of the airguns may be more pronounced than what 
would be observed in free-swimming animals due to increased flow speed 
of fluid inside bags containing the experimental animals. There were no 
sublethal effects on the escape performance or the sensory threshold 
needed to initiate an escape response at any of the distances from the 
airgun that were tested. Whereas McCauley et al. (2017) reported an SEL 
of 156 dB at a range of 509-658 m, with zooplankton mortality observed 
at that range, Fields et al. (2019) reported an SEL of 186 dB at a 
range of 25 m, with no reported mortality at that distance.
    Regardless, if we assume a worst-case likelihood of severe impacts 
to zooplankton within approximately 1 km of the acoustic source, the 
typically wide dispersal of survey vessels and brief time to 
regeneration of the potentially affected zooplankton populations does 
not lead us to expect any meaningful follow-on effects to the prey base 
for odontocete predators (the region considered in this rule is not an 
important feeding area for taxa that feed directly on zooplankton, 
i.e., mysticetes).
    Given the inconsistency of the McCauley et al. (2017) results with 
prior research on impacts to zooplankton as a result of exposure to 
airgun noise and with the research of Fields et al. (2019), further 
validation of those findings would be necessary for NMFS to reach a 
determination that these impacts are likely to occur. Moreover, a 
single study is not sufficient to evaluate the potential impacts, and 
further study in additional locations must be conducted. Therefore, 
BOEM proposed to fund such a study as part of their 2019-21 Studies 
Development Plan (www.boem.gov/FY-2019-2021-SDP/).
    A recent review article concluded that, while laboratory results 
provide scientific evidence for high-intensity and low-frequency sound-
induced physical trauma and other negative effects on some fish and 
invertebrates, the sound exposure scenarios in some cases are not 
realistic to those encountered by marine organisms during routine 
seismic operations (Carroll et al., 2017). The review finds that there 
has been no evidence of reduced catch or abundance following seismic 
activities for invertebrates, and that there is conflicting evidence 
for fish with catch observed to increase, decrease, or remain the same. 
Further, where there is evidence for decreased catch rates in response 
to airgun noise, these findings provide no information about the 
underlying biological cause of catch rate reduction (Carroll et al., 
2017).
    NRDC's assertions regarding the likely effects of airgun survey 
noise on marine mammal prey include, for example, the assertion that 
the specified activity would harm fish and invertebrate species over 
the long-term, cause reductions in recruitment and effects to behavior 
that may reduce reproductive potential and foraging success and

[[Page 5338]]

increase the risk of predation, and induce changes in community 
composition via such population-level impacts. We have addressed these 
claims both in this response and in our review of the available 
literature. We also reviewed available information regarding 
populations of representative prey stocks in the northern GOM, i.e., 
the only U.S. location where marine seismic surveys are a routinely 
occurring activity. While we recognize the need for caution in assuming 
correlation between the ongoing survey activity in the GOM and the 
health of assessed stocks there, we also believe this information has 
some value in informing the likelihood of population-level effects to 
prey species and, therefore, the likelihood that the specified activity 
would negatively impact marine mammal populations via effects to prey. 
We note that the information reported below is in context of managed 
commercial and recreational fishery exploitation, in addition to any 
other impacts (e.g., noise) on the stocks. The species listed below are 
known prey species for marine mammals and represent groups with 
different life histories and patterns of habitat use. Numerous other 
managed stocks are similarly healthy.
     Red snapper (Lutjanus campechanus): Red snapper are 
bottom-dwelling fish generally found at approximately 10-190 m deep 
that typically live near hard structures on the continental shelf that 
have moderate to high relief (for example, coral reefs, artificial 
reefs, rocks, ledges, and caves), sloping soft-bottom areas, and 
limestone deposits. Larval snapper swim freely within the water column. 
Increases in total and spawning stock biomass are predicted beginning 
in about 1990 (Cass-Calay et al., 2015). Regional estimates suggest 
that recruitment in the west has generally increased since the 1980s, 
and has recently been above average, while recruitment in the east 
peaked in the mid-2000s, and has since declined. However, the most 
recent assessment suggests a less significant decline (to moderate 
levels) (Cass-Calay et al., 2015).
     Yellowfin tuna (Thunnus albacares): Yellowfin tuna are 
highly migratory, living in deep pelagic waters, and spawn in the GOM 
from May to August. However, we note that a single stock is currently 
assumed for the entire Atlantic, with additional spawning grounds in 
the Gulf of Guinea, Caribbean Sea, and off Cabo Verde. The most recent 
assessment indicates that spawning stock biomass for yellowfin tuna is 
stable or increasing somewhat and that, overall, the stock is near 
levels that produce the maximum sustainable yield (ICCAT, 2016).
     King mackerel (Scomberomorus cavalla): King mackerel are a 
coastal pelagic species, found in open waters near the coast in waters 
from approximately 35-180 m deep. King mackerel migrate in response to 
changes in water temperature, and spawn in shelf waters from May 
through October. Estimates of recruitment demonstrate normal cyclical 
patterns over the past 50 years, with a period of higher recruitment 
most recently (1990-2007) (SEDAR, 2014). Long-term spawning stock 
biomass patterns indicate that the spawning stock has been either 
rebuilding or remained relatively consistent over the last 20 years, 
with nothing indicating that the stock has declined in these recent 
decades (SEDAR, 2014).
    In summary, the scientific literature demonstrates that impacts of 
seismic surveys on marine mammal prey species will likely be limited to 
behavioral responses, the majority of prey species will be capable of 
moving out of the area during surveys, a rapid return to normal 
recruitment, distribution, and behavior for prey species is 
anticipated, and, overall, impacts to prey species, if any, will be 
minor and temporary. Prey species exposed to sound might move away from 
the sound source, experience TTS, experience masking of biologically 
relevant sounds, or show no obvious direct effects. Mortality from 
decompression injuries is possible in close proximity to a sound, but 
only limited data on mortality in response to airgun noise exposure are 
available (Hawkins et al., 2014). The most likely impacts for most prey 
species in a given survey area would be temporary avoidance of the 
area. Surveys using towed airgun arrays move through an area relatively 
quickly, limiting exposure to multiple impulsive sounds. In all cases, 
sound levels would return to ambient once a survey moves out of the 
area or ends and the noise source is shut down and, when exposure to 
sound ends, behavioral and/or physiological responses are expected to 
end relatively quickly (McCauley et al., 2000b). The duration of fish 
avoidance of a given area after survey effort stops is unknown, but a 
rapid return to normal recruitment, distribution, and behavior is 
anticipated. While the potential for disruption of spawning 
aggregations or schools of important prey species can be meaningful on 
a local scale, the mobile and temporary nature of most surveys and the 
likelihood of temporary avoidance behavior suggest that impacts would 
be minor.
    Finally, and relevant to NMFS' findings under the MMPA, NRDC does 
not demonstrate that even the asserted worst-case effects on prey 
species would have any meaningful impact on marine mammals or their 
respective populations. Referencing a single study on zooplankton 
effects (i.e., McCauley et al., 2017), NRDC implies that airgun surveys 
will definitively reduce ``the abundance and diversity of zooplankton 
over vast areas and induc[e] changes in community composition due to 
the aggregation of individual- and population-level impacts across 
multiple fish and invertebrate species,'' thereby leading to ecosystem-
level effects that would harm marine mammal populations. NMFS disagrees 
with this interpretation of the scientific literature and notes the 
presence of healthy stocks of marine mammal prey species currently 
found in the GOM, despite decades of routine geophysical survey 
operations, but also a devastating oil spill (discussed in detail in 
the notice of proposed rulemaking). NMFS believes that no evidence is 
presented to contradict our conclusions regarding likely impacts to 
marine mammals due to effects on prey species, i.e., that impacts of 
the specified activity are not likely to have more than short-term 
adverse effects on any prey habitat or populations of prey species, and 
that any effects that do occur are not expected to result in 
significant or long-term consequences for individual marine mammals, or 
to contribute to adverse impacts on their populations.
    Comment: The Associations object to NMFS' use of an analysis of 
chronic and cumulative impacts of noise on marine mammals in the GOM 
(i.e., the CCE report), which was described in detail in our notice of 
proposed rulemaking. The Associations state that ``[c]oncepts such as 
``soundscape,'' ``communication space,'' or ``acoustic footprint'' have 
no basis in any existing statutory or regulatory authorities, and are 
therefore inapplicable to this rulemaking.''
    Response: The purpose of the analysis was to evaluate the more 
cumulative nature of low-frequency, long-distance propagation of 
relatively low-intensity energy from multiple seismic surveys operating 
concurrently in a region, and to evaluate potential loss of ability to 
detect signals of biological importance over spatial scales relevant to 
the sources and hearing capabilities of representative species. NMFS is 
required to evaluate the effects of the specified activity on the 
potentially impacted marine mammal stocks and their habitat. Noise can 
disrupt marine mammals' behavioral patterns through the contraction of 
their communication

[[Page 5339]]

space, among other impacts. Moreover, NMFS is required to mitigate 
impacts on marine mammal species or stocks and their habitat. Concepts 
such as listening area and communication space are not novel, having 
been published in peer-reviewed literature and previously applied in 
impact assessment contexts. NMFS is required to consider these effects.
    Comment: NRDC argues that NMFS fails to consider chronic harm, 
including masking effects and impacts on acoustic habitat. For example, 
NRDC asserts that the consideration of masking in NMFS' negligible 
impact analysis was cursory in that it only came through the 
vulnerability ratings and ``seems to misapprehend the spatial and 
temporal scope of the effects'' of masking. Similarly, in addition to 
citing general concerns about chronic effects to Bryde's whales and 
other species, NRDC asserts that acoustic habitat is discussed, but not 
factored into the negligible impact analysis.
    Response: The potential impacts of masking were properly 
considered. NRDC significantly understates the consideration given to 
masking effects in the Expert Working Group (EWG) risk assessment 
framework (see Negligible Impact Analysis and Determinations). Broadly, 
the results of the EWG analysis for any given species are based on the 
integration of two components: The severity of the impacts (which 
reflects the extent of the activities overlaid with the presence and 
distribution of the given species) and the vulnerability of that 
species based on multiple biological and environmental risk factors, 
including explicit consideration of masking. The maximum possible 
vulnerability score any species can attain under the assessment across 
all of these factors is 30. The masking component of the vulnerability 
score considers communication masking, foraging masking, and 
navigation/orientation masking--for a total of seven points. The 
minimum score that any species assessed in the context of these survey 
activities could (and did) attain is 1, while the Bryde's whale was 
given the maximum scores across all types of masking for a score of 7. 
The differential across the highest and lowest possible masking scores 
is 6, out of a maximum possible total of 30 for the overall 
vulnerability score, which means that masking accounts for twenty 
percent of a species' vulnerability rating. Twenty percent is an 
appropriate and not insubstantial proportion of the vulnerability 
score, given that the total score (with its 30-point maximum) also 
accounts for behavioral impacts, whether there are biologically 
important areas or times overlaying the activities, whether there are 
additional chronic anthropogenic (e.g., other anthropogenic noise) or 
chronic biological factors (e.g., disease), and the status and trends 
of the population.
    NMFS recognizes that masking is not necessarily co-extensive with 
harassment and explicitly recognizes this in our discussion of effects, 
although we also note that the distances at which behavioral harassment 
is quantified for this rule are farther than those contemplated in the 
past, due to the behavioral harassment thresholds used (see the 
Estimated Take section and comment responses later in this section for 
further discussion of acoustic thresholds). As discussed elsewhere, 
NMFS designed and supported the implementation of a chronic and 
cumulative effects analysis (the CCE report, discussed later in this 
preamble) for the specific purpose of addressing the effects of these 
activities on the listening space of all species and the communication 
space of Bryde's whales specifically. This modeling effort explicitly 
considered the effects of masking over realistic spatial scales. In 
their 2017 public comments on incidental harassment authorizations NMFS 
had proposed for seismic survey activities in the Atlantic Ocean, NRDC 
specifically recommended that NMFS conduct a modeling exercise like the 
effort conducted here for the GOM rule to better support those findings 
(see 83 FR 63268; December 7, 2018), yet they now suggest that this 
analysis is inadequate, even paired with the quantitative analysis 
included in the EWG analysis as it is here. See Potential Effects of 
the Specified Activities on Marine Mammals and Their Habitat in the 
notice of proposed rulemaking for additional discussion.
    Comment: A private citizen offers commentary and clarifications 
regarding the discussions of acoustic masking and acoustic habitat 
provided in our notice of proposed rulemaking.
    Response: We appreciate the discussion provided by the commenter, 
but note that no specific recommendations are provided towards an 
improved assessment of the effects of chronic aggregate noise from 
survey activity, as the commenter suggests is needed.

Cumulative Impacts and Related Issues

    Comment: NRDC expressed concern regarding cumulative impacts, 
claiming that NMFS' negligible impact determination underestimates 
impacts to marine mammal species and populations because it fails to 
consider the effects of other anticipated activities on the same marine 
mammal populations. NRDC also stated that NMFS must include geophysical 
surveys occurring within state waters within the scope of the ITR.
    Response: Neither the MMPA nor NMFS' codified implementing 
regulations address consideration of other unrelated activities and 
their impacts on populations. However, the preamble for NMFS' 
implementing regulations (54 FR 40338; September 29, 1989) states in 
response to comments that the impacts from other past and ongoing 
anthropogenic activities are to be incorporated into the negligible 
impact analysis via their impacts on the baseline. Consistent with that 
direction, NMFS has factored into its negligible impact analysis the 
impacts of other past and ongoing anthropogenic activities via their 
impacts on the baseline, e.g., as reflected in the density/distribution 
and status of the species, population size and growth rate, the chronic 
and cumulative effects analysis (the ``CCE report'' discussed later in 
this preamble), and other relevant stressors. Some of these are 
addressed explicitly through the environmental risk factor scoring in 
the population vulnerability analysis of the Expert Working Group 
Assessment (including consideration of Deepwater Horizon (DWH) oil 
spill effects and risk from other anthropogenic activities). In 
addition, we consider these factors as relevant contextual elements of 
the analysis. See the Negligible Impact Analysis and Determinations 
section of this notice for full detail.
    Our 1989 final rule for the MMPA implementing regulations also 
addressed public comments regarding cumulative effects from future, 
unrelated activities. There we stated that such effects are not 
considered in making findings under section 101(a)(5) concerning 
negligible impact. We indicated (1) that NMFS would consider cumulative 
effects that are reasonably foreseeable when preparing a NEPA analysis, 
and (2) that reasonably foreseeable cumulative effects would also be 
considered under section 7 of the ESA for ESA-listed species.
    Here, we recognize the potential for cumulative impacts, as 
analyzed through BOEM's PEIS, which addressed the impacts of an 
extended time period of survey activity that may be permitted by BOEM 
(ten years versus the five years that the ITR is limited to), and which 
NMFS adopted as the basis for its Record of Decision. In that analysis, 
the assessment was focused on whether the predicted level of take from 
the forecasted level of survey effort, when considered in context, 
would have a

[[Page 5340]]

meaningful biological consequence at a species or population level. 
NMFS, therefore, assessed and integrated other contextual factors 
(e.g., species' life history and biology, distribution, abundance, and 
status of the stock; mitigation and monitoring; characteristics of the 
surveys and sound sources) in determining the overall impact of 
issuance of the ITR and subsequent LOAs on the human environment. Key 
considerations included the nature of the surveys and the required 
mitigation. In all cases, it is expected that sound levels will return 
to previous background levels once the acoustic source moves a certain 
distance from the area, or the surveys cease. The proposed rule also 
identified several time-area restrictions to minimize risk or severity 
of impacts to the extent practicable, consistent with the MMPA's least 
practicable adverse impact standard. In the final rule, two of those 
areas were removed from consideration based on the reduction in the 
scope of the rule per BOEM's request. The other proposed mitigation 
area remains (as modified; see Mitigation). Although those two areas 
have been removed from consideration as mitigation due to the reduction 
in scope of the rule, the practical effect on GOM stocks is similar, in 
that no survey activity within those areas may be considered for take 
authorization pursuant to the rule. The similar result is a reduction 
in the overall numbers of take but also, importantly, elimination or 
minimization of impacts to marine mammal species or stocks in the areas 
most important to them for feeding, breeding, and other important 
functions. Therefore, the severity of takes that may occur pursuant to 
the rule is expected to be meaningfully lower due to the reduction in 
impacts that could reduce reproductive success or survivorship.
    In summary, NMFS does not expect aggregate impacts from the 
forecast level of survey effort to affect rates of recruitment or 
survival for marine mammals, either alone or in combination with other 
past, present, or ongoing activities. The cumulative impacts of these 
surveys (i.e., the incremental impact of the action when added to other 
past, present, and reasonably foreseeable future actions) were 
addressed as required through the NEPA documents cited above. These 
documents, as well as the relevant Stock Assessment Reports, are part 
of NMFS' Administrative Record for this action, and provided the 
decision-maker with information regarding other activities in the 
action area that affect marine mammals, an analysis of cumulative 
impacts, and other information relevant to the determinations made 
under the MMPA.
    Separately, cumulative effects were analyzed as required through 
NMFS' required intra-agency consultation under section 7 of the ESA, 
which concluded that NMFS' action of issuing the ITR and subsequent 
LOAs was not likely to jeopardize the continued existence of listed 
marine mammals.
    We disagree with NRDC's suggestion that we include geophysical 
surveys in state waters within the scope of this rulemaking. Section 
101(a)(5)(A) of the MMPA requires NMFS to make a determination that the 
take incidental to a ``specified activity'' will have a negligible 
impact on the affected species or stocks of marine mammals, and will 
not result in an unmitigable adverse impact on the availability of 
marine mammals for taking for subsistence uses. NMFS' implementing 
regulations require applicants to include in their request a detailed 
description of the specified activity or class of activities that can 
be expected to result in incidental taking of marine mammals. 50 CFR 
216.104(a)(1). Thus, the ``specified activity'' for which incidental 
take coverage is being sought under section 101(a)(5)(A) is generally 
defined and described by the applicant. Here, BOEM is the applicant for 
the ITR in support of industry operators, and we are responding to the 
specified activity as described in that petition (and making the 
necessary findings on that basis). As BOEM's PEIS makes clear, BOEM 
does not have a regulatory role regarding surveys occurring in state 
waters. (See, e.g., BOEM's PEIS, Chapter 1.1.3)
    NRDC's representation of our action--``The agency's decision to 
evaluate the impacts of state water surveys separately as if they would 
occur in isolation''--also ignores the fact that we have no information 
about the possible extent of potential future geophysical survey 
activity in state waters, including type, amount, duration, timing, 
location, etc., even if such activity were to occur. Although it may be 
reasonable to assume that such activity occurs, we have no specific 
knowledge of any past, present, or reasonably foreseeable future survey 
activity in state waters. No prospective applicant has contacted NMFS 
to request incidental take authorization for any such survey activity 
planned or expected within state waters, on either a programmatic or 
specific basis. NRDC did not provide any information about the expected 
future extent of survey activity in state waters.

Acoustic Thresholds

    Comment: NRDC expressed concerns regarding NMFS' proposed use of 
the probabilistic response function described by Wood et al. (2012), in 
which 10 percent, 50 percent, and 90 percent of individuals exposed are 
assumed to produce a behavioral response (of a sufficient degree of 
severity to constitute Level B harassment) at exposures of 140, 160, 
and 180 dB root mean square (rms), respectively. (The function is 
shifted for the more behaviorally sensitive beaked whales such that 50 
percent and 90 percent response probabilities are assumed to occur at 
120 and 140 dB rms, respectively.) NRDC stated that the function is 
inconsistent with the best available science, asserting that behavioral 
disruptions occur at higher percentages at lower noise exposure levels 
than those suggested by Wood et al. (2012). NRDC's criticism of the 
function also focused on the use of horizontal displacement studies as 
the supposed basis of analysis for Wood et al. (2012), as well as on 
the function's nature as a series of step functions. In addition, NRDC 
expressed concerns that the use of frequency weighting in the Wood et 
al. (2012) approach is inappropriate. NRDC requested that NMFS revise 
the threshold as suggested in Nowacek et al. (2015), which recommended 
a similar function (but centered on 140 dB rms rather than 160 dB rms), 
while simultaneously stating that the use of such step-based risk 
functions is ``biologically irrational.'' Overall, NRDC claims that 
reliance on this function results in underestimation of impacts. A 
private citizen echoed some of NRDC's comments on this topic while CRE 
supports use of the Wood et al. approach.
    Response: NMFS has been criticized in the past for the use of the 
single-step 160-dB rms approach. Those criticisms are based on the idea 
that an approach reflecting a more complex multi-step probabilistic 
function would more effectively represent the known variation in 
responses at different levels due to differences in the receivers, the 
context of the exposure, and other factors, as well as the science 
indicating that animals may react in ways constituting Level B 
harassment when exposed to lower received levels. In developing the 
acoustic exposure analysis for the proposed rulemaking, we reviewed 
relevant past public comments as well as the best available science, 
determining that a more complex probabilistic function is indeed better 
reflective of available scientific information, and that it was 
appropriate

[[Page 5341]]

to take the fundamental step of recognizing the potential for Level B 
harassment occurring at exposures to received levels below 160 dB rms 
(as well as the potential for no Level B harassment occurring at 
exposures above 160 dB rms). This approach necessarily also accounts 
for differential hearing sensitivity by incorporating frequency-
weighting functions, as behavioral responses in cetaceans are best 
explained by the interaction between sound source type and functional 
hearing group (Gomez et al., 2016). NMFS has determined that the 
general approach used for this rule--a probabilistic risk function that 
allows for the likelihood of differential response probability at given 
received levels on the basis of multiple factors, including behavioral 
context and distance from the source, and that addresses particularly 
sensitive species--is appropriate in light of the best available 
scientific information.
    However, because behavioral responses to sound depend on the 
context in which an animal receives the sound, including the animal's 
behavioral mode when it hears sounds, prior experience, additional 
biological factors, and other contextual factors, defining sound levels 
that disrupt behavioral patterns is extremely difficult. Even experts 
have not previously been able to suggest specific new criteria due to 
these difficulties (e.g., Southall et al. 2007; Gomez et al., 2016). 
Agency expertise is appropriate in defining the particular steps at 
which specific response probabilities are assumed to occur, and while 
we acknowledge our approach reduces a complex suite of interactions to 
make reasonable inferences, it is consistent with the best available 
science.
    NRDC expressed concerns regarding our approach by noting the size 
discrepancy between the area ensonified to 140 dB versus that 
ensonified to 160 dB, implying that we ignore potential responses at 
the lower received level. To clarify, the difference between our 
approach and NRDC's recommendation is solely in the proportion of a 
population assumed to be taken upon exposure to the specified received 
level which, as stated above, is determined on the basis of expert 
judgement based on the best available science. We believe that the Wood 
et al. (2012) function is consistent with the best available science, 
and is therefore an appropriate approach. Below, we address NRDC's 
concerns in greater detail.
    NRDC referenced ``recent'' research they claim is not consistent 
with the recommendations of Wood et al. (2012). We note that, of the 
nine studies cited by NRDC, five were published prior to the Wood et 
al. (2012) study, and were therefore available for those authors' 
consideration (and some were specifically referenced by those authors 
in discussion of their recommendations). Further, we disagree that the 
referenced findings are inconsistent with Wood et al. (2012). First, a 
mere reaction to noise exposure does not mean that a take by Level B 
harassment, as defined by the MMPA, has occurred. For a take to occur 
requires that an act have ``the potential to disturb by causing 
disruption of behavioral patterns,'' not simply result in a detectable 
change in motion or vocalization. NRDC also suggests that some of these 
studies were not incorporated into Wood et al.'s recommendations, or 
our consideration of those and other potential approaches in context of 
the available science, and criticize what they view as an over-reliance 
on horizontal displacement studies as the supposed basis of analysis. 
While it is true that the majority of available behavioral data focus 
on avoidance responses, Wood et al. (2012) does not mention excluding 
behavioral studies involving vocal changes, and the precedent Southall 
et al. (2007) specifically incorporates numerous studies that do 
mention changes in vocalization associated with sound exposure. Thus, 
these datasets were not excluded and, as discussed in our notice of 
proposed rulemaking, we adequately considered all studies addressed by 
NRDC.
    Regarding baleen whales, we acknowledge that changes in 
vocalization have been observed in association with exposure to airgun 
surveys within migratory and non-migratory contexts (e.g., Castellote 
et al., 2012; Blackwell et al., 2013; Cerchio et al., 2014). The 
potential for such effects to occur over relatively large spatial 
scales is not surprising for species with large communication spaces 
(e.g., Clark et al., 2009), but we reiterate our disagreement with 
NRDC's apparent contention that every detected change to vocalizations 
rises to the level of a take. NRDC cites reports of changes in 
vocalization, typically for baleen whales, as evidence in support of 
lower thresholds, claiming these reactions result in biological 
consequences indicating that the reaction was indeed a take. However, 
NMFS is not aware of research that provides a well-supported link 
between the reported reactions at lower received levels and the 
putative consequences. In conflict with NRDC's interpretation of the 
literature are documented instances of marine mammal exposure to 
greater received levels that did not elicit any response (e.g., Malme 
et al., 1983, 1984, 1985, 1988; McCauley et al., 1998, 2000a, 2000b; 
Barkaszi et al., 2012; Stone, 2015a; Gailey et al., 2016; Barkaszi and 
Kelly, 2018).
    The received level associated with stoppage of calling for bowhead 
whales (Balaena mysticetus) observed by Blackwell et al. (2013, 2015)--
a response that may arguably rise to the level of harassment--is 
consistent with the Wood et al. (2012) scheme, in which the potential 
for take upon exposure to received levels as low as 140 dB is accounted 
for. Similarly, the findings of Pirotta et al. (2014) for harbor 
porpoise (Phocoena phocoena) are consistent with the treatment of 
behaviorally sensitive species by Wood et al., in which the potential 
for take at even lower received levels is accounted for (though 
irrelevant here, as harbor porpoise are not found in the GOM). The 
response levels reported by McDonald et al. (1995) and Di Iorio and 
Clark (2009) for blue whales (Balaenoptera musculus) also comport with 
the Wood et al. function, if we assume that the observed responses 
equate to harassment (though it is not clear that they do). With regard 
to NRDC's citation of Clark and Gagnon (2006), a non-peer reviewed 
white paper, NRDC incorrectly overestimated the area over which the 
effect was observed by an order of magnitude (the paper discusses an 
area of 100 x 100 nmi, which equates to 10,000 nmi\2\--not 100,000 
nmi\2\).
    In regard to Cerchio et al. (2014), it is important to note that 
received levels provided in this study are those recorded at locations 
of their underwater recording devices. The authors indicated ``we did 
not have the ability to locate the singers or the seismic survey 
vessel, estimate the source level of the pulses, the distance between 
the source and potentially impacted singers, or the received level of 
the pulses at the singers.'' The same situation, i.e., actual received 
levels at the location of the animals are unknown, is true for 
Castellote et al. (2012) and Clark and Gagnon (2006), which provide 
average background sound levels with and without the presence of airgun 
surveys. Thus, not having the location of the animals at the time of 
exposure makes it difficult to draw conclusions based strictly on 
received level. NMFS has evaluated the papers and determined they are 
not informative about appropriate Level B harassment thresholds.
    Regarding sperm whales, NMFS disagrees that assuming a 100 percent

[[Page 5342]]

probability of take of sperm whales upon exposure to survey noise at 
135 dB--as suggested by NRDC--is an accurate reflection of the results 
of the Miller et al. (2009) study. While we agree that the work of 
Miller et al. (2009) suggests that sperm whales in the GOM may be 
susceptible to disruption of foraging behavior upon exposure to 
relatively moderate sound levels, NRDC incorrectly interprets results 
of the study in claiming that sperm whale ``foraging success'' was 
found to ``decline significantly.'' Instead, the authors report that 
buzz rates (a proxy for attempts to capture prey) were approximately 20 
percent lower, meaning that the appropriate interpretation would be 
that foraging activity (versus foraging success) was reduced by 20 
percent (Jochens et al., 2008). Of the eight whales tagged in that 
study, only one was observed to actually cease foraging.
    Moreover, while we do believe that these results support a 
conclusion that exposure to survey noise can impact foraging activity, 
other commenters have interpreted them differently, e.g., by focusing 
on the finding that exposed whales did not change behavioral state 
during exposure or show horizontal avoidance (a finding replicated in 
other studies, e.g., Madsen et al., 2002a; Winsor et al., 2017). 
Importantly, the observed effect was not statistically significant and, 
as reported by the authors, constituted ``subtle effects on their 
foraging behavior.'' Furthermore, the authors of the Wood et al. (2012) 
study explicitly described their consideration of Miller et al. (2009) 
in the development of their recommended criteria. Therefore, the Wood 
et al. (2012) recommendation is indeed consistent with the Miller et 
al. (2009) study.
    In referencing Bowles et al. (1994), NRDC fails to state that the 
observed cessation of vocalization was likely in response to a low-
frequency tone (dissimilar to airgun signals), though a distant airgun 
survey was noted as producing signals that were detectable above 
existing background noise. NRDC recommends that NMFS base a sperm whale 
threshold on the findings of a separate study of exposure of sperm 
whales and other species to sonar signals (Miller et al., 2012). NMFS 
disagrees that behavioral response data for sperm whales exposed to 
mid-frequency active sonar (Miller et al., 2012) is more appropriate 
than using data from the airgun exposures described by Miller et al. 
(2009) and already considered within the Wood et al. function. 
Furthermore, the alternative recommendation of Nowacek et al. (2015), 
which is repeatedly mentioned by NRDC as a more appropriate alternative 
to Wood et al. (2012), does not make a distinction between sperm whales 
and other odontocetes and instead advocates for a criteria that treats 
all marine mammal species the same (we address this in greater detail 
below).
    Regarding other odontocetes, NRDC's representation of the available 
scientific information is also inaccurate. Miller et al. (2005) 
specifically state that ``[s]ighting rates at distances of 10-20 km 
from the airgun array were significantly lower than those in areas 20-
30 km from the airgun array, where sighting rates were unexpectedly 
high'' (i.e., the study indicates sighting rates of beluga whales 
(Delphinapterus leucas) were lower, not ``100% avoidance'' as claimed 
by NRDC). Miller et al. (2005) reported seven aerial beluga whale 
sightings from 8 to 18 km from the survey vessel and two vessel-based 
beluga whale sightings at 1.5 and 2.5 km from the survey vessel. 
Furthermore, Southall et al. (2007) described the findings of the 
Miller et al. (2005) study as temporary avoidance behaviors at these 
lower received levels, while Gomez et al. (2016) (which NRDC agrees 
reflects the best available science) evaluated Miller et al. (2005) 
based on a received level of 150 dB. Thus, the Wood et al. (2012) 
approach does capture responses associated with this study.
    Additionally, Wood et al. (2012) has the advantage of accounting 
for sensitive species such as beaked whales, meaning that a response of 
a beaked whale at 140 dB (as cited by NRDC) is covered within the Wood 
et al. (2012) recommended criteria (e.g., Wood et al. assumes 90 
percent of an exposed beaked whale population will respond at 140 dB). 
If Nowacek et al. (2015) was instead used, as advocated by NRDC, the 
probability of response would only be 50 percent at 140 dB.
    It should be noted that the systematic review by Gomez et al. 
(2016), cited by NRDC in support of their position, found that received 
level was not appropriate as the sole indicator of behavioral response. 
For example, this review shows that ``low'' effects were actually found 
to reach peak probability at a higher received level than ``moderate'' 
effects for baleen whales. As we discussed in our notice of proposed 
rulemaking, the results of the Gomez et al. (2016) review are not 
inconsistent with Wood et al. (2012). With regard to NRDC's comment 
that the authors consider their results ``non-conservative,'' Gomez et 
al. (2016) only indicates that they may have scored the severity of 
vocal responses higher if they had more information on the ecological 
significance of these types of responses. There is no indication 
elsewhere in Gomez et al. (2016) that their overall results and 
analysis are ``non-conservative.''
    NRDC repeatedly cites Nowacek et al. (2015) in public comments. We 
note first that while NRDC repeatedly refers to this paper as a 
``study'' (implying that it presents new scientific data or the results 
of new analyses of existing scientific data), the paper (which is co-
authored by the author of NRDC's comment letter) in fact makes policy 
recommendations rather than presenting any new science. The more 
substantive reviews presented by Southall et al. (2007) and Gomez et 
al. (2016) were unable to present any firm recommendations, as noted 
above. We addressed the Nowacek et al. (2015) approach relative to the 
Wood et al. (2012) approach, in context of the best available 
scientific information, in detail in our notice of proposed rulemaking. 
Then, as now, we found that those recommendations are not justified by 
the available scientific evidence.
    Other than suggesting a 50 percent midpoint for a probabilistic 
function, Nowacek et al. (2015) offer minimal detail on how their 
recommended probabilistic function should be derived/implemented or 
exactly how this midpoint value (i.e., 140 dB rms) was derived (i.e., 
what studies support this point). In contrast with elements of a Level 
B harassment function that NRDC indicates as important, Nowacek et al. 
(2015) does not make distinctions between any species or species groups 
and provides no quantitative recommendations for acknowledging that 
behavioral responses can vary by species group and/or behavioral 
context. In summary, little substantive support is provided by Nowacek 
et al. (2015) for the proposal favored by NRDC. Few studies are offered 
in support of the recommended midpoint and the proposal is offered only 
in a one-page supplementary document. The Nowacek et al. (2015) 
approach is not well-supported scientific consensus, as NRDC's comment 
suggests.
    Additionally, the application of the Nowacek et al. (2015) approach 
disregards the important role that distance from a source plays in the 
likelihood that an animal will respond to a given received level from 
that source type in a particular manner. By assuming, for example, a 50 
percent midpoint at 140 dB rms, the approach implies an unrealistically 
high probability of marine mammal response

[[Page 5343]]

to signals received at very far distances from a source (e.g., greater 
than 50 km). DeRuiter et al. (2013) found that beaked whales exposed to 
similar received levels responded when the sound was coming from a 
closer source and did not respond to the same level received from a 
distant source. Although the Wood et al. (2012) approach does not 
specifically include a distance cut-off, the distances at which marine 
mammals are predicted to respond better comport with the distances at 
which behavioral responses have been detected and reported in the 
literature.
    NRDC also criticizes the use of weighting functions in evaluating 
potential Level B harassment, and specifically criticizes use of the M-
weighting scheme of Southall et al. (2007). Gomez et al. (2016) suggest 
that incorporation of frequency-weighting is necessary to account for 
differential hearing sensitivity, as behavioral responses in cetaceans 
are best explained by the interaction between sound source type and 
functional hearing group. That is, implementing weighting functions 
allows for consideration that different marine mammal groups do not 
hear varying frequencies of sound equally well. Thus, it is appropriate 
to account for sounds below a group's best hearing range having a lower 
likelihood of resulting in a behavioral response (let alone that 
animals are likely unable to effectively detect sounds at frequencies 
completely outside their hearing range).
    The M-weighting functions are described in Southall et al. (2007) 
as ``intentionally precautionary (wide)'' (as opposed to the weighting 
functions used in NMFS' 2018 Revised Technical Guidance \2\ to account 
for noise-induced hearing loss) and are used to account for the 
functional hearing ranges of different marine mammal hearing groups. 
This frequency weighting scheme was intentionally selected because it 
is more conservative in accounting for hearing sensitivity (as is 
appropriate in evaluating potential Level B harassment) than are more 
recently developed filters designed to better assess potential noise-
induced hearing loss.
---------------------------------------------------------------------------

    \2\ NMFS. 2018. 2018 revision to: Technical guidance for 
assessing the effects of anthropogenic sound on marine mammal 
hearing (Version 2.0). NOAA Technical Memorandum NMFS-OPR-59, 
National Marine Fisheries Service: 178.
---------------------------------------------------------------------------

    NRDC asserts that because M-weighting assumes that mid- and high-
frequency (MF and HF) cetaceans are relatively insensitive to noise 
below 1 kHz, it is likely that the incorporation of M-weighting has a 
significant downwards effect on take estimates. This is incorrect. The 
table below illustrates the impact of M-weighting functions on 
frequencies ranging from 100 Hz to 1 kHz.

              Table 3--Impact of M-Weighting Functions on Frequencies Ranging From 100 Hz to 1 kHz
----------------------------------------------------------------------------------------------------------------
                                                 Weighting (-dB)
-----------------------------------------------------------------------------------------------------------------
                  Hearing group                        1 kHz          500 Hz          250 Hz          100 Hz
----------------------------------------------------------------------------------------------------------------
Mid-frequency Cetaceans.........................       -0.186 dB        -0.76 dB        -2.77 dB         -10 dB.
High-frequency Cetaceans........................       -0.034 dB        -1.33 dB        -4.45 dB       -13.6 dB.
----------------------------------------------------------------------------------------------------------------

We see that, at 250 Hz and above, the M-weighting functions do not 
result in a significant reduction (less than 3 dB for MF cetaceans and 
less than 5 dB for HF cetaceans). Furthermore, the lower bound of the 
functional hearing range of these groups is 150 Hz for MF cetaceans and 
275 Hz for HF cetaceans (i.e., sounds below 100 Hz, where most energy 
in airgun noise is found and where M-weighting results in the greatest 
reductions, are outside functional hearing range). At 1 kHz, where 
these species are most likely to be able to detect and respond to 
airgun noise, there is very little assumed reduction in sensitivity.
    Finally, NRDC advocates for the use of a linear risk function as 
opposed to the multiple step function of Wood et al. (2012), stating 
that linear risk functions are scientifically accepted methodology that 
better acknowledge individuals may vary in responsiveness. Although 
NRDC does not specifically define what they mean by ``linear risk 
function,'' NMFS assumes a linear risk function is a smooth, continuous 
function, as opposed to a function defined by multiple steps, as is the 
case of Wood et al. (2012) (and Nowacek et al. (2015), which NRDC 
recommends as an alternative to Wood et al.). NRDC states that Wood et 
al. (2012) ``has a significant negative bias on take estimates'' where 
``all exposures from 140 dB to 159.9 dB are considered to produce the 
same risk.'' While it is true that relying upon Wood et al. (2012) 
results in all exposures within a particular step (e.g., 140 dB and 
159.9 dB) having the same risk, and future risk functions may be 
further refined by incorporating more steps, Wood et al. (2012) better 
represents known variation in behavioral responses at different 
received levels than Nowacek et al. (2015), which provides only a 
suggested midpoint for a risk function without any guidance on what 
should be done above or below this midpoint, much less the linear risk 
function NRDC states should be used. Wood et al. (2012) does 
acknowledge that responsiveness varies with received levels, while 
relying on broad steps, rather than a continuous function. These broad 
steps allow for easier implementation of a risk function and are more 
practical for most users, which is an important consideration, 
especially in the context of users that may not have the ability or 
access to more sophisticated modeling (i.e., non-Navy users). 
Therefore, if new linear risk functions become available, NMFS may 
still provide a more simplistic function broken down in broad steps, so 
that it can be applied by all users.
    In referencing NMFS' proposal to use the recommendations of Wood et 
al., and prior to even attempting to characterize the scientific 
evidence, NRDC states, ``Incredibly [NMFS'] approach produces take 
estimates that are substantially lower than the much-criticized, non-
conservative, 160 dB threshold [. . .].'' NRDC (1) mischaracterizes 
criticism of the historic 160-dB threshold as being about the results 
of its use, rather than being about whether it adequately represents 
the best available science; (2) introduces an MMPA standard that does 
not exist in the statute (implying that NMFS is being unlawfully or 
improperly ``non-conservative''); and (3) suggests that NRDC favors 
whichever method of evaluating potential Level B harassment returns the 
highest estimate. This is repeated later in their comment when they 
assert that use of the Wood et al. recommendations are ``arbitrary and 
capricious'' because use of the recommendations ``appears, in its 
results, even less conservative than the outdated 160 dB threshold.'' 
However,

[[Page 5344]]

selection of an evaluation scheme on the basis of the results it 
returns, rather than on how well the scheme reflects the available 
scientific literature, would be truly arbitrary and capricious and run 
counter to our mandates.
    Overall, we reiterate the lack of scientific consensus regarding 
what criteria might be most appropriate for evaluating Level B 
harassment. Defining sound levels that disrupt behavioral patterns is 
difficult because responses depend on complex, difficult to predict 
contextual factors much more so than received level. Therefore, levels 
at which responses occur are not necessarily consistent and can be 
difficult to predict. However, although better methods of assessing 
likely behavioral response to acoustic stimuli than the relatively 
simple multi-step function used here may be forthcoming from the 
scientific community, NMFS is compelled to move forward with the best 
available information. We believe the recommendations of Wood et al. 
(2012) reflect the best available science.
    Comment: NRDC notes NMFS' reference to a ``preliminary analysis'' 
in the discussion of acoustic thresholds for Level B harassment and 
asserts that NMFS must make the analysis publicly available and allow 
opportunity for public comment before finalizing the rule.
    Response: Our use of the phrase ``preliminary analysis'' in the 
notice of proposed rulemaking merits some clarification. The particular 
analysis we referred to is not in and of itself pre-decisional or 
preliminary. Rather, it is a discrete analytical product with a result 
that will not change--it is one way (non-parametric regression method) 
of looking at one subset (Malme et al., 1984, 1988; Houser et al., 
2013; Antunes et al., 2014; Moretti et al., 2014) of the data related 
to marine mammal behavioral responses to intermittent sound. NMFS 
conducted an analysis of relevant data starting with the premise of 
deriving a generic exposure-response curve using previously published 
exposure-response curves. This exercise was conducted as part of an 
ongoing separate and broader agency effort to evaluate behavioral 
response data. We also clarify that the Level B harassment criteria for 
this rule did not substantively rely upon that analysis.
    Comment: NRDC claims that NMFS misapplies the MMPA's statutory 
definition of harassment by adopting a probability standard other than 
``potential'' in setting thresholds for auditory injury, stating that a 
take estimate based on ``potential'' should either count take from the 
lowest exposure level at which hearing loss can occur or establish a 
probability function that accounts for variability in the acoustic 
sensitivity of individual marine mammals. NRDC states that NMFS instead 
derived auditory injury thresholds from average exposure levels at 
which tested marine mammals experience hearing loss, which discounts 
instances of hearing loss at lower levels of exposure. The comment 
further states that for purposes of take estimation, thresholds based 
on mean or median values will lead to roughly half of an exposed cohort 
experiencing the impacts that the threshold is designed to avoid, at 
levels that are considered ``safe,'' therefore resulting in substantial 
underestimates of auditory injury. NRDC makes similar statements with 
regard to the criteria for Level B harassment.
    Response: The 2018 Revised Technical Guidance's (NMFS, 2018) onset 
thresholds for TTS for non-impulsive sounds encompass more than 90 
percent of available TTS data (i.e., for mid-frequency cetaceans, only 
two data points are below the onset threshold, with maximum point only 
2 dB below), and in some situations 100 percent of TTS data (e.g., 
high-frequency cetaceans; although this group is data-limited). Thus, 
the 2018 Revised Technical Guidance thresholds provide realistic 
predictions, based on currently available data, of noise-induced 
hearing loss in marine mammals. For impulsive sounds, data are limited 
to two studies, and NMFS directly adopted the TTS onset levels from 
these two studies for the applicable hearing groups.
    Our Federal Register notice announcing the availability of the 
original 2016 Technical Guidance (81 FR 51694; August 4, 2016; NMFS, 
2016), indicated that onset of auditory injury (i.e., permanent 
threshold shift (PTS)) equates to Level A harassment under the MMPA. We 
explained in that notice that because the acoustic thresholds for PTS 
conservatively predict the onset of PTS, they are inclusive of the 
``potential'' language contained in the definition of Level A 
harassment. See 81 FR 51697, 51721.
    Regarding Level B harassment, based on the language and structure 
of the definition of Level B harassment, we interpret the concept of 
``potential to disturb'' as embedded in the assessment of the 
behavioral response that results from an act of pursuit, torment, or 
annoyance (collectively referred to hereafter as an ``annoyance''). The 
definition refers to a ``potential to disturb'' by causing disruption 
of behavioral patterns. Thus, an analysis that indicates a disruption 
in behavioral patterns establishes the ``potential to disturb.'' A 
separate analysis of ``potential to disturb'' is not needed. In the 
context of an ITR such as this, our analysis is forward-looking. The 
inquiry is whether we would reasonably expect a disruption of 
behavioral patterns; if so, we would conclude a potential to disturb 
and therefore expect Level B harassment. We addressed NRDC's concerns 
regarding the scientific support for the Level B harassment criteria in 
a previous comment response.
    Comment: NRDC raised concerns regarding use of NMFS' 2018 Revised 
Technical Guidance (NMFS, 2018), claiming that the guidance is not 
based on the best available science and underestimates potential 
auditory injury. We also note that NRDC's comment references an 
attachment that was not provided.
    Response: The 2018 Revised Technical Guidance (NMFS, 2018) is a 
compilation, interpretation, and synthesis of the scientific literature 
that provides the best available information regarding the effects of 
anthropogenic sound on marine mammals' hearing. The 2016 Technical 
Guidance was classified as a Highly Influential Scientific Assessment 
and, as such, underwent three independent peer reviews, at three 
different stages in its development, including a follow-up to one of 
the peer reviews, prior to its dissemination by NMFS. In addition, 
there were three separate public comment periods, during which time 
NMFS received and responded to similar comments on the guidance (81 FR 
51694), and more recent public and interagency review under Executive 
Order 13795. While new information may help to improve the guidance in 
the future, and NMFS will review the available literature to determine 
when revisions are appropriate, the final guidance reflects the best 
available science and all information received through peer review and 
public comment. The concerns raised by NRDC have been addressed by NMFS 
in responses associated with the guidance (see www.fisheries.noaa.gov/national/marine-mammal-protection/marine-mammal-acoustic-technical-guidance). In light of these considerations, NRDC's argument that use 
of the guidance is ``arbitrary and capricious'' is unpersuasive. As was 
stated in our notice of proposed rulemaking, NMFS considers the 2018 
Revised Technical Guidance to represent the best scientific information 
currently available and, given the incorporation of multiple peer 
reviews and public comment opportunities during its development, we did 
not solicit and are not

[[Page 5345]]

responding in detail to comments concerning the contents of the 
Technical Guidance (NMFS, 2016, 2018), as such comments are outside the 
scope of this rulemaking.
    NRDC also referenced information related to occupational noise 
standards established by the National Institute of Occupational Safety 
and Health (NIOSH). Human noise risk assessments (NIOSH, 1998) are not 
equivalent (or applicable) to thresholds provided in the guidance, 
because they are used to predict hearing loss based on a daily 8-h 
exposure over 40 years (i.e., current marine mammal TTS data are only 
available to predict exposure periods of 24-h or less and cannot be 
used to assess or predict risk associated with a lifetime of exposure) 
and are based on larger sample sizes of human listeners (e.g., NIOSH 
1972 and 1997 risk assessments were based on a sample size of 1,172 
people). As pointed out in Wright (2015), NIOSH criteria provide a 95 
percent confidence interval for their human noise standards but also 
allow for an excess risk of material hearing impairment, defined as an 
average threshold elevation for both ears that exceeds 25 dB, of eight 
percent (i.e., human noise standards limits do allow for some risk; 
risk is not zero percent and specifically that eight percent of the 
population is still capable of developing noise-induced hearing loss 
exceeding 25 dB when exposed to the 85 dB NIOSH level).
    Finally, we note that a group of scientists recently published an 
update to their original, seminal publication concerning noise exposure 
criteria to predict the onset of auditory effects in marine mammals 
(Southall et al., 2007, 2019a), the topic of this comment. The newer 
publication evaluates the recommendations of the original publication 
in light of subsequent scientific findings, including those findings 
that form the basis for the recommendations of NMFS (2018). While 
Southall et al. (2019a) provide recommendations for future research 
that could lead to revisions, the fundamental aspects of an evaluation 
of the onset of auditory effects for the marine mammals considered in 
this ITR (i.e., auditory weighting functions and noise exposure 
criteria) are identical to those presented by NMFS (2018) and 
incorporated into the modeling process developed for this ITR.

Sound Field Modeling

    Comment: NRDC asserts that NMFS has not appropriately accounted for 
hard-bottom habitat in our propagation analysis, stating that there are 
areas of hard bottom in the GOM and that we cannot assume that proposed 
surveys will take place entirely in areas with soft or sandy bottoms.
    Response: Sound propagation modeling performed in support of BOEM's 
PEIS and this ITR was developed to adequately represent a wide range of 
conditions for a variety of parameters, including bottom composition. 
NMFS does not assume that hard bottom does not exist in the GOM, but 
rather that it is not sufficiently predominant to warrant specific 
representation in a propagation modeling exercise covering the whole 
GOM. As shown in Figure 50 of the modeling report--depicting a 
compilation of surficial sediment composition available through a 
hydrographic survey database from NOAA's National Ocean Service and its 
predecessor, the U.S. Coast and Geodetic Survey--muds and sands are the 
dominant substrate types throughout the GOM (as stated in our notice of 
proposed rulemaking), with only small, scattered areas of hard bottom. 
The Minerals Management Service (MMS) report cited by NRDC, which 
concerns conversion of seafloor maps existing at the time to MMS-
approved GIS format for use in geohazards evaluations, does not 
contradict this.
    Substrate types for propagation modeling are based on grain size, 
porosity, and shear velocity, etc., and do not include ``hard'' or 
``coral'' bottom. It is also important to note that, while some hard 
bottom habitats would increase propagation due to increased 
reflectivity, NRDC's statement that coral bottom can ``significantly 
increase propagation of airgun noise'' is erroneous. In fact, the 
roughness of the coral habitat would cause severe bottom loss due to 
scattering. As noted above, bottom composition in the region is mostly 
mud and sand and, therefore, selection of parameter values associated 
with these bottom types for propagation modeling is appropriate. We 
also note that, for the shelf region of the eastern GOM, where sand is 
predominant, a larger grain size value was selected to account for 
this. The acoustic modeling provided by Zeddies et al. (2015, 2017a) 
appropriately and reasonably accounts for variability in bottom 
composition throughout the region.
    The modeling process requires the use of simplifying assumptions 
about oceanographic and seabed parameters, and these assumptions carry 
some uncertainty, which may lead to uncertainty in the form of variance 
or error in individual model outputs and in the final estimates of 
marine mammal acoustic exposures. It is for this reason that parametric 
uncertainty analysis was performed to evaluate the effects of this 
uncertainty ``envelope.'' (This analysis was summarized in our notice 
of proposed rulemaking and described in detail in the modeling report. 
NRDC does not reference this assessment.) Uncertainties in the results 
of acoustic propagation modeling were estimated by examining the 
variation in model outputs when model inputs were offset by realistic 
errors. The environmental properties were selected so that the median, 
or expected, value could be compared to a worst-case outcome (e.g., 
assuming an extreme case of a more reflective bottom), which was 
generated by selecting extreme values for several input parameters. 
These comparisons represent the maximum errors in the predicted sound 
fields that result from incorrect specification of the parameters 
tested. As described in the modeling report, the greatest uncertainty 
due to geoacoustic parameters of the sea bottom is 4 dB (in the deep 
zone). The effect of the geoacoustic uncertainty increased when the 
sound speed profile was downwardly refracting. In the case of a surface 
channel (slope zone, winter season), the average difference between the 
median and worst-case was only 0.5 dB, i.e., in this case the 
geoacoustic parameters had virtually no effect on the sound levels at 
the top of the water column (where marine mammals are likely to be 
present).

Marine Mammal Densities

    Comment: NRDC criticized NMFS' use of the Roberts et al. (2016) 
model outputs for purposes of deriving abundance estimates, as used for 
comparison to exposure estimates herein. NRDC states that we should use 
the NMFS Stock Assessment Report (SAR) abundance estimates for this 
purpose, while allowing that model-predicted abundance estimates may be 
used for ``data-deficient'' stocks. NRDC implies that use of model-
predicted abundances would overestimate actual abundances, apparently 
based on the fact that the density models are informed by many years of 
data rather than only the most recent year of data. Where model-
predicted abundance estimates are used, NRDC recommends that we adjust 
the averaged model outputs to the lower bound of the standard deviation 
estimated by the model for each grid cell.
    Response: The approach recommended by NRDC is inappropriate. 
Comparing take estimates generated through use of the outputs of a 
density model to an unrelated abundance estimate provides a meaningless 
comparison. As explained in our notice of proposed rulemaking,

[[Page 5346]]

we compare the take estimates generated through use of the density 
outputs to the abundance predicted through use of the model precisely 
to provide a meaningful comparison of predicted takes to predicted 
population.
    The two potential sources of abundance data--the output of cetacean 
density models (Roberts et al., 2016) and the available SARs data--
provide different results, with the SARs estimates typically much 
lower. Differences between the two separate sets of abundance estimates 
result from key methodological differences. In order to produce 
sufficiently reliable and detailed density surfaces (maps), Roberts et 
al. (2016) combined multiple NMFS cetacean surveys and modeled density 
using a habitat-based approach (Miller et al., 2013), while the SARs 
estimates utilized only the most recent NMFS survey and estimated 
density using traditional distance sampling (Buckland et al., 2001). 
The two approaches, while compatible and based on a common statistical 
framework (distance sampling), can yield different results, depending 
on complex factors such as whether population sizes have changed, or 
species habitat preferences have shifted over time. Neither approach 
will necessarily yield a higher abundance estimate than the other, but 
use of multiple years of data in developing an abundance estimate 
minimizes the influence of interannual variation in over- or 
underestimating actual abundance. By linking sightings with 
environmental conditions, habitat-based density layers represent 
smoothed surfaces that are not biased by anomalous conditions. This 
makes them particularly appropriate for the five-year timeframe of this 
ITR, which will span varying environmental conditions.
    To illustrate why this smoothing of interannual variation helps to 
create a meaningful comparison to take estimates, we provide the 
extreme example of the GOM Clymene dolphin. NMFS' three most recent SAR 
abundance estimates for this stock have fluctuated between 129 and 
17,355 animals, i.e., varying by a maximum factor of more than 100. For 
most species, such fluctuations across these ``snapshot'' abundance 
estimates (i.e., that are based on only the most recent year of survey 
data) reflect interannual variations in dynamic oceanographic 
characteristics that influence whether animals will be seen when 
surveying in predetermined locations, rather than any true increase or 
decline in population abundance. In fact, NMFS' SARs typically caution 
that trends should not be inferred from multiple such estimates, that 
differences in temporal abundance estimates are difficult to interpret 
without an understanding of range-wide stock abundance, and that 
temporal shifts in abundance or distribution cannot be effectively 
detected by surveys that only cover portions of a stock's range (i.e., 
U.S. waters). The corresponding density model for Clymene dolphins 
predicts a mean abundance of 11,000 dolphins. Therefore, in this 
example, NRDC would have us compare takes predicted by a model in which 
11,000 dolphins are assumed to exist against an abundance estimate of 
129 dolphins. Our goal in assessing predicted takes is to generate a 
meaningful comparison, which is accomplished through use of the model-
predicted abundance.
    A second key methodological difference explains the tendency for 
the model-predicted abundance estimates to be higher than the SARs 
estimates. SAR abundance estimates are typically underestimates of 
actual abundance because they do not account for bias on the ability of 
observers to detect animals--in contrast, Roberts et al. (2016) do 
account for availability bias and perception bias on the probability of 
sighting an animal. Availability bias occurs when a model assumes that 
animals are always available to be observed by the survey team when, in 
fact, they are not. Cetaceans are diving animals; while submerged, they 
are unavailable. Assuming diving animals are always available results 
in an underestimation of abundance, because while they are diving they 
are present but not counted by the survey team. Perception bias occurs 
when a model assumes that animals will always be detected when they are 
on the survey trackline, when, in fact, detection is not certain.
    With regard to bias correction, NRDC suggests that such corrections 
are incorporated into NMFS' GOM SARs. However, some correction has been 
performed only for the more-recently surveyed shelf and coastal stocks 
of bottlenose dolphin, i.e., four out of the 25 stocks of GOM marine 
mammals considered herein. NRDC also strangely suggests that ``NMFS 
doesn't show that applying trackline correction factors consistent with 
Barlow (2015), who reported trackline detection probabilities for 
marine mammals in the Pacific, would result in population estimates 
consistent with the ones the agency has derived from Roberts et al. 
(2016).'' We can safely assume that these would be consistent, given 
that the models developed by Roberts et al. (2016) considered, and in 
some cases directly incorporated, the correction factors of Barlow 
(1999) and Barlow and Forney (2007) (the original work upon which 
Barlow (2015) builds).
    These issues, which are typical for NMFS' SAR abundance estimates, 
are particularly exacerbated for GOM stocks. For the majority of 
stocks, the most recent abundance estimates are derived from the 
results of vessel-based surveys in 2009, i.e., even if one believes 
that such ``snapshot'' estimates are most appropriate, the GOM 
estimates are out of date and NMFS' guidelines state that data greater 
than eight years old should not be used for abundance estimates. (We 
note that more recent survey effort has been conducted, but 
corresponding abundance estimates have only recently been made 
available via unpublished draft SARs for most stocks that have yet to 
be available for public comment or finalized at the time the analyses 
were completed for these regulations.) More important for cryptic 
species, i.e., those species that spend little time at the surface and/
or are difficult to detect when at the surface, is the lack of any bias 
correction. For example, the Cuvier's beaked whale--a cosmopolitan 
species and perhaps the most widespread and most commonly observed 
species of beaked whale--is officially estimated by NMFS to number 74 
individuals in the GOM, a clear underestimate. For purposes of 
reference, current abundance estimates for the U.S. Pacific and 
Atlantic stocks--for which some bias corrections have been made--are 
3,274 and 5,744 individuals, respectively. Marine mammal scientists 
working in the GOM have acknowledged that the likely abundance of 
beaked whales (and other cryptic species, such as Kogia spp.) should be 
expected to be closer to the values predicted by Roberts et al. (2016) 
than those given in the SARs. For example, Dias and Garrison (2016) 
state that current abundance estimates for Kogia spp. may be 
considerably underestimated due to the cryptic behavior of these 
species and difficulty of detection in Beaufort sea state greater than 
one, while density estimates for certain species derived from long-term 
passive acoustic monitoring are much higher than are estimates derived 
from visual observations (e.g., Hildebrand et al., 2015). Separately, 
NMFS' announcement of a negative 90-day finding on a petition to list 
the GOM Cuvier's beaked whale as endangered (84 FR 11058) included 
adoption of the abundance estimate of Roberts et al. (2016) as being 
most appropriate. Roberts et al. (2015b) summarize this situation: 
``Because [NMFS' SAR] estimates are very low relative to the

[[Page 5347]]

abundance we estimated, it is likely that if our [density] results are 
used to estimate population-level impacts from potentially harmful 
human activities (i.e. ``takes'', as defined by the Marine Mammal 
Protection Act), the estimated impacts will be very high [. . .].''
    NRDC suggests that the SARs are an appropriate representation of 
``actual'' abundance, whereas the Roberts et al. (2016) predictions are 
not. NRDC also appears to claim, without substantiation, that an 
abundance estimate derived from multiple years of data would typically 
overestimate actual abundance. However, these estimates are not 
directly comparable--not because one represents a ``snapshot,'' while 
one represents multiple years of data--but because one does not correct 
for one or more known biases against the probability of observing 
animals during survey effort, while the other does. Because of this 
important caveat, NMFS' SAR abundance estimates should not be 
considered ``actual'' abundance more than any other accepted estimate. 
Therefore, when multiple estimates of a stock's abundance are 
available, they should be evaluated based on quality, e.g., does the 
estimate account for relevant biases, does it minimize the effect of 
interannual variability, and, importantly, should provide a meaningful 
comparison. In this light, our use of the Roberts et al. (2016) 
abundance estimates are not a ``radical departure from past practice,'' 
as claimed by NRDC. Our practice, as mandated by our implementing 
regulations, is to use the best scientific evidence available. NRDC 
states that ``NMFS cannot simply discard this Congressionally mandated 
estimate in favor of the larger population estimates derived from its 
misapplication of the [. . .] model.'' The statute does not mandate use 
of the SARs for comparison with take estimates.
    Aside from their failure to explain the claim of 
``misapplication,'' and the unwarranted implication that we must make 
use of the model-generated abundance estimates simply because they are 
larger (and not because they are the best available scientific 
information), NRDC errs in asserting that the MMPA requires that we use 
SAR abundance estimates. Section 117 of the MMPA requires the 
development of SARs, and dictates certain information that SARs must 
provide. However, there is no part of the MMPA that requires the 
population abundance estimates given in a SAR to be used in any 
specific application and, importantly, the MMPA does not even require 
that the SAR include a best population estimate. The MMPA requires only 
that SARs provide a minimum population estimate, which is used in the 
formulation of a potential biological removal (PBR) level, which is 
then required by section 118 of the MMPA for certain uses in the 
management of marine mammal take incidental to commercial fisheries. In 
summary, NRDC's comment reflects an inaccurate interpretation of the 
available information, and NMFS disagrees with the approach recommended 
by the comment.

Take Estimates

    Comment: The Associations state that ``NMFS substantially 
overestimates the number of incidental takes predicted to result'' from 
the specified activity. The comment goes on to discuss the modeling 
that is ``intentionally designed to overestimate takes,'' and discusses 
the findings of the Acoustic Exposure Model Variable Analysis (Zeddies 
et al., 2017b) (which was provided for public review in association 
with the proposed rule). Other industry commenters and the CRE echo 
these points.
    Response: The commenters' statements that NMFS has substantially 
overestimated takes are incorrect. We used current scientific 
information and state-of-the-art acoustic propagation and animal 
movement modeling to reasonably estimate potential exposures to noise. 
Chevron stated that the modeling used ``admittedly erroneous models'' 
but provides no supporting information or citation. Chevron further 
describes ``errors in methodology'' and ``admissions'' that the 
modeling methodology and the data used are not ``rigorous science,'' 
while asserting that NMFS ``repeatedly rejects and omits science that 
is available.'' Chevron's comments do not provide any illumination as 
to what specifically these errors may be, what data it believes is 
flawed, or what ``science'' NMFS has rejected or omitted. NMFS has 
considered all relevant available scientific information.
    To summarize in a basic way, it is foreseeable that a large amount 
of noise-producing activity, such as BOEM's application and PEIS 
describe, results in a substantial number of predicted acoustic 
exposures. Despite recommending that ``a better approach would be to 
use the best and most likely values for all of the input variables to 
the model,'' the Associations' comments do not include substantive 
recommendations for improvement. They do not specify which of the many 
data inputs are ``conservative'' or to what degree, nor do they 
recommend alternatives to the choices that were painstakingly 
documented in developing the modeling.
    As was noted in the notice of proposed rulemaking, NMFS disagrees 
with the Associations' characterization of the modeling and with 
certain statements in BOEM's draft PEIS regarding the modeling that are 
frequently cited by the Associations. As we stated in the notice, 
BOEM's draft PEIS included unsupported and erroneous statements that 
characterized the modeling results--which BOEM and NMFS developed 
collaboratively--as ``unrealistically high,'' ``overly conservative,'' 
and representative of a ``worst-case scenario,'' among other things. 
These statements were included in that document without NMFS' prior 
knowledge. Importantly, as a result of NOAA's public comments on that 
draft PEIS in its role as a cooperating agency, the statements 
referenced by the commenters were properly removed from the final PEIS, 
which more accurately characterizes the modeling process and results.
    The Associations take out of context a number of statements from 
the discussion in NMFS' notice of proposed rulemaking of the modeling 
process, data inputs, and user selections. We address these in turn:
     The Associations quote NMFS as stating that our modeling 
likely ``leads to substantial overestimates of the numbers of 
individuals potentially disturbed [and] . . . to an overestimation of 
the population-level consequences of the estimated exposures'' and 
that, even with the application of a correction factor, the modeling 
still represents an ``overestimate.'' (83 FR 29261, 29291). But the 
full statement in our notice is as follows: ``While the modeling 
provides reasonable estimates of the total number of instances of 
exposure exceeding Level B harassment criteria, it is likely that it 
leads to substantial overestimates of the numbers of individuals 
potentially disturbed, given that all animals within the areas modeled 
are unlikely to be completely replaced on a daily basis. Therefore, in 
assuming an increased number of individuals impacted, these results 
would lead to an overestimation of the potential population-level 
consequences of the estimated exposures.'' Our point was that, although 
the modeling provides reasonable estimates of the total amount of 
acoustic exposures, it would be an overestimate to interpret this total 
as representative of the number of individuals impacted. We then 
discussed our development of a correction factor to address this issue

[[Page 5348]]

(see Table 12 of the notice of proposed rulemaking).
     The Associations highlight our ``admission'' that the 
modeling is purposely conservative. We address this below by explaining 
why, in some cases, it is appropriate to make reasonably conservative 
choices.
     The Associations mischaracterize the Wood et al. (2012) 
Level B harassment criteria as ``expressly rejecting the best available 
science.'' In discussing different versions of frequency weighting 
functions, we stated that ``Type III filters'' are better designed to 
predict the onset of auditory injury, while explaining why use of 
``Type I filters'' (or M-weighting) was appropriate for use in 
evaluation of Level B harassment (83 FR 29248). The Type III filters, 
as adopted by NMFS (2018), were appropriately used for evaluation of 
Level A harassment (which includes auditory injury).
     Although characterized as ``conservative'' NMFS has made 
reasonable choices through the application of professional judgment by 
subject matter experts. For example, using single airgun modeling 
results in lieu of boomer results was a choice made for computational 
efficiency precisely because it was not significantly influential on 
the results. (83 FR 29251). And selecting an estimate for standard 
deviation in an investigation of model sensitivity to source level 
variance was in response to a concern of the commenters--overall 
sensitivity of the model to uncertainty in input parameters and the 
resulting uncertainty in model results--and had no bearing on the model 
results (83 FR 29257).
     The Associations also highlight our statement that ``the 
lack of aversion within the animal movement modeling process results in 
overestimates of potential injurious exposure,'' without noting that we 
corrected this issue through a post-hoc correction to reasonably 
account for aversion.
    The modeling required that a number of assumptions and choices be 
made by subject matter experts and, in most cases, the most 
representative data or methods were used. As we acknowledged, in some 
cases, some assumptions or choices are purposely conservative (where 
the conservative choice is reasonable) to minimize the likelihood of 
underestimating the potential impacts on marine mammals represented by 
a specified level of survey effort. These are reasonable, 
scientifically acceptable choices that do not create, as the 
Associations state, ``multiplicatively accumulating bias as the 
conservative assumptions interact with each other to multiply 
uncertainty''). To the extent that the results of the modeling may be 
conservative, they are the most credible, science-based information 
available at this time (assuming the notional 8,000 in\3\ array and 
activity level projections specified by BOEM in the petition).
    These comments provide no reasonable justification as to why the 
modeling results in overestimates of take. The Associations instead 
seem to rely on the incorrect premise that real-time mitigation would 
somehow reduce actual levels of acoustic exposure (versus reducing the 
duration and/or intensity of exposure). NMFS disagrees that ``each of 
the inputs is purposely developed to be conservative''--again, the 
Associations do not provide any support for this assertion, and none is 
to be found in the administrative record for this action. Although it 
may be correct that some conservativeness accumulates throughout the 
analysis, the Associations do not adequately describe the nature of 
conservativeness associated with model inputs or the degree to which 
such conservativeness ``accumulates'' (either quantitatively or 
qualitatively), nor do they offer more appropriate alternatives.
    The modeling effort incorporated representative sound sources and 
projected survey scenarios (both based on the best available 
information obtained by BOEM), physical and geological oceanographic 
parameters at multiple locations within the GOM and during different 
seasons, the best available information regarding marine mammal 
distribution and density, and available information regarding known 
behavioral patterns of the affected species. Current scientific 
information and state-of-the-art acoustic propagation and animal 
movement modeling were used to reasonably estimate potential exposures 
to noise. The notice of proposed rulemaking described all aspects of 
the modeling effort in significant detail, including numerous 
investigations (test scenarios) designed by the agencies to understand 
various model sensitivities and the effects of certain choices on model 
results. The modeling report itself was provided for public review, in 
association with both BOEM's PEIS and NMFS' notice of proposed 
rulemaking.
    We quote the Marine Mammal Commission's public comment on this 
topic: ``Complex sound propagation and animat modeling was used to 
estimate the numbers of potential takes from various types of 
geophysical surveys in the Gulf. NMFS received comments from industry 
operators suggesting that the modeling results were overly conservative 
and that the take estimates were `higher than BOEM expects would 
actually occur in a real world environment.' However, the Commission 
has reviewed the modeling approach and parameters used to estimate 
takes and believes they represent the best available information 
regarding survey scenarios, sound sources, physical and oceanographic 
conditions in the Gulf, and marine mammal densities and behavior. As 
such, the Commission agrees with NMFS and BOEM that the resulting take 
estimates were conservative but reasonable, thereby minimizing the 
likelihood that actual takes would be underestimated.''
    The CRE says, absent citation or reference, that ``everyone 
agrees'' that takes are overestimated. Their assertion that we 
``greatly overestimate both exposures and takes'' is based on their 
view that we relied on ``flawed models and on Risk Assessment 
Frameworks that are unfinished and have not been peer reviewed.'' While 
the Associations focus on supposed conservatism built into the modeling 
process, the CRE appears to believe that there is some unknown process 
by which modeled exposures are ``converted'' to takes. (``These take 
overestimates stem primarily from [NMFS'] use of various models to 
convert exposures to takes [. . .]. They have no credible framework for 
converting exposure to takes.'') We believe the CRE is likely referring 
to the EWG risk assessment framework, which is a systematic analysis 
used as an aid to understanding the significance of the modeled takes 
to the affected stocks. However, this framework plays no role in the 
estimation of takes (takes are an input to the EWG framework) and is 
not itself a ``model.'' CRE also makes the claim, addressed elsewhere 
in this response, that the take estimates ``do not include the impact 
of mitigation measures.''
    Regarding the modeling variable analysis submitted by the 
Associations (Zeddies et al., 2017b), we have fully considered the 
results in developing this final ITR,\3\ but do not find that the

[[Page 5349]]

analysis supports any changes to the modeling. IAGC and API contracted 
with JASCO Applied Sciences, which performed the original modeling 
effort, to conduct additional analysis regarding the effect that 
various acoustic model parameters or inputs have on the outputs used to 
estimate numbers of animals exposed to threshold levels of sound from 
geophysical sources used in the GOM. The analysis investigated five 
factors:
---------------------------------------------------------------------------

    \3\ The Associations misunderstand the timeline relating to the 
availability of the report for NMFS' consideration for developing 
the proposed rule. (``NMFS inexplicably dismisses [the report] as 
being provided too late despite the fact that it was provided to 
NMFS 11 months ago''). We must correct the record on this point. The 
analysis was submitted by IAGC for NMFS' consideration on September 
6, 2017, well after the total 45-day comment period on the petition 
had closed (81 FR 88664 (December 8, 2016, notice of receipt of 
petition providing for 30-day comment period); 81 FR 92788 
(extending comment period an additional 15 days to January 23, 
2017)). The final PEIS was then issued in August 2017. Subsequent 
materials could no longer be considered as NMFS prepared the draft 
proposed rule for interagency review. The rule was submitted to OMB 
on October 3, 2017. Upon submission, no further changes could be 
made to the rule other than those arising pursuant to the 
interagency review. The Office of Information and Regulatory Affairs 
cleared the proposed rule on June 11, 2018, whereupon it was 
submitted to the Federal Register on June 12 and published on June 
22. Therefore, the analysis was not able to be considered by NMFS in 
the notice of proposed rulemaking despite the length of time between 
submission of the report to NMFS and publication of the proposed 
rule.
---------------------------------------------------------------------------

     Airgun array size (including total volume, number of array 
elements, element air pressure, array geometry and spacing) used in 
source and propagation models;
     Acoustic threshold criteria and associated weighting used 
to calculate exposures;
     Animal densities used for adjusting simulated computer 
model exposures to potential real-world animal exposures;
     Natural aversive behaviors of marine mammals; and
     The addition of mitigative measures that lessen the 
potential for animals' exposure to threshold levels of seismic sound.
    The primary finding of the Acoustic Exposure Model Variable 
Analysis is that use of appropriate acoustic injury criteria (i.e., 
NMFS, 2016, 2018) decreased predictions of injurious exposure. At the 
time the Associations submitted this report, they were apparently 
unaware that, as described herein, NMFS had already made the change 
that the Associations' analysis indicates is most significant: The 
appropriate acoustic injury criteria (i.e., NMFS, 2016, 2018), 
representing the best available science, were used in NMFS' analysis in 
the proposed rule. Other significant investigations in the 
Associations' modeling variable analysis included an alternative array 
size and quantitative consideration of animal aversion and mitigation 
effectiveness. We address these below.
    The Associations state that the selected array (8,000 in\3\) is 
unrealistically large, resulting in an overestimation of likely source 
levels and, therefore, size of the sound field with which marine 
mammals would interact. Zeddies et al. (2017b) evaluated the use of a 
substitute 4,130 in\3\ array, finding that reduction in array volume 
reduces the number of predicted exposures. Use of a smaller airgun 
array volume with lower source level unsurprisingly creates a smaller 
ensonified area resulting in fewer numbers of animals expected to 
exceed exposure thresholds. However, selection of the representative 
array to be used in the modeling was directed by the ITR applicant 
(i.e., BOEM). Given that the array used was selected by the applicant 
and included in the petition for the ITR (which was available for 
public comment in our Federal Register notice of receipt of BOEM's 
application), any complaint regarding this or other aspects of the 
specified activity, including activity level projections and 
representative source characteristics or survey geometry, should be 
addressed to BOEM. According to BOEM, the particular array was selected 
as a realistic representative proxy after BOEM's discussions with 
individual geophysical companies. An 8,000-in\3\ array was considered 
reasonable, as it falls within the range of typical airgun arrays 
currently used in the GOM, which are roughly 4,000-8,475 in\3\ (BOEM, 
2017). According to BOEM's permitting records, approximately one-third 
of arrays used in a recent year were 8,000 in\3\ or greater. Also, as 
noted previously, regardless of the representative airgun array size 
used to model the number of takes of marine mammals for the purposes of 
the analysis conducted in this rule, the analysis of the take and the 
associated findings are applicable to take incurred from the use of 
other sizes of airgun arrays, including smaller ones such as those 
modeled in the Acoustic Exposure Model Variable Analysis report.
    The Associations' comments also focus significantly on the need to 
incorporate quantitative adjustments to account for aversion and 
mitigation. As discussed in the notice of proposed rulemaking, the 
effects of mitigation and aversion on exposure estimates were 
investigated via test scenarios, and NMFS acknowledges that both of 
these factors would lead to a reduction in likely injurious exposure to 
some degree. (As noted above, the issue of aversion was addressed via 
post-hoc quantitative adjustment). Ultimately these factors were not 
quantified in the modeling because, in summary, there is too much 
inherent uncertainty regarding the effectiveness of detection-based 
mitigation for these activities to support any reasonable 
quantification of its effect in reducing injurious exposure, and there 
is too little information regarding the likely level of onset and 
degree of aversion to justify its use in the modeling via precise 
quantitative control of animat movements (as compared to post-hoc 
adjustment of the modeling results, as is done here). Zeddies et al. 
(2017b) found that incorporation of aversion into the modeling process 
appears to reduce the number of predicted injurious exposures, though 
the magnitude of the effect was variable. The authors state that this 
variability is likely because there are few samples of injurious 
exposure exceedance, meaning that the statistical variability of re-
running simulations is evident.
    While aversion and mitigation implementation are expected to reduce 
somewhat the modeled levels of injurious exposure, it is important to 
note that they would not be expected to result in any meaningful 
reduction in assumed exposures resulting in Level B harassment, nor in 
total takes by harassment, as any averted injurious (Level A 
harassment) takes would not be alleviated, but rather would be 
appropriately changed to behavioral disturbance (Level B harassment) 
takes. The Associations, acknowledging the analysis we have done to 
produce more realistic estimates of potential Level A harassment, are 
focused on the supposed overestimation of Level B harassment. Yet their 
focal areas of complaint are limited to array size, which is a decision 
made by BOEM, and mitigation effectiveness, a factor that would have no 
effect on the amount of predicted Level B harassment. With regard to 
the large number of other data inputs and/or choices made in the 
modeling, the Associations conclude that ``NMFS has admittedly chosen 
conservative numerical values to assess allegedly uncertain variables 
to overestimate adverse effects,'' without specifically identifying a 
single issue where they feel a meaningful data or process error was 
made.
    Comment: The CRE recommends a different method of estimating 
potential take of marine mammals, stating that NMFS ``should continue 
to use Line Transect to estimate exposures and takes.''
    Response: Although CRE does not actually describe the method they 
recommend, we infer that they are referencing a relatively simplistic 
method historically used in estimating acoustic exposures, typically on 
a survey-specific basis. Essentially, this methodology consists of: (1) 
Determination of estimated isopleth ranges from the source for a 
specified acoustic threshold (nominally this threshold was historically 
the 160 dB

[[Page 5350]]

rms received level for Level B harassment); (2) assumption that a 
cylinder whose radius matched the range to these isopleths and 
encompassed the entire water column was ensonified to that threshold; 
(3) calculating the surface area ensonified by this water column as the 
source moved along its track; and (4) multiplying that resultant 
ensonified surface area by the density of each marine mammal species 
present to estimate potential harassment takes. (Note that this process 
is somewhat more complicated for evaluation of 3D surveys.) In this 
case, following a modeling workshop held in 2014 as a collaborative 
effort between the American Petroleum Institute (API) and the 
International Association of Geophysical Contractors (IAGC), NMFS, and 
BOEM, the agencies determined that it would be most appropriate to 
collaborate on a more sophisticated approach, in which more detailed 
modeling of the source and its properties, the acoustic propagation 
field in three dimensions, and three dimensional animal placement and 
movement is used to better calculate the potential impacts to marine 
mammals. To summarize aspects of the process:
     Operational Scenario Development: According to BOEM, the 
source and operations scenarios presented in the petition and which 
underlie the modeling effort were based on historical permit 
information. BOEM sought industry input and used historical data to 
develop the specification of the nominal airgun array. The array 
specifications and level of survey effort were intended to be 
representative of future activity, not a conservative over-estimate.
     Acoustic Modeling: The propagation model output has been 
compared with measured data and been shown to be reliable. The physical 
inputs to the model are the best available data. The full sound field 
was used to predict exposures, not a `maximum over depth' 
simplification.
     Animal Modeling: The animal movement model used is one of 
the few models available that incorporates full four-dimensional 
movement. Properly applied, such models provide the most accurate 
predictions of acoustic exposure.
     Animal Density: The density and distribution data used 
were the best available and represent the latest synthesis and 
analysis.
     Effects Criteria: The historical Level B harassment 
threshold of 160 dB has been criticized for multiple reasons, and the 
use of the Wood et al. (2012) criteria in this analysis allows for the 
application of current scientific information to address some of the 
issues raised. The best available science relating to potential 
auditory injury, as synthesized in NMFS (2018) and more recently 
described by Southall et al. (2019a), was used in the modeling effort.
    Taking advantage of these more sophisticated tools allows for a 
more accurate and detailed model of the exposures of a population of 
marine animals in the three dimensions and time, and also provides: (1) 
Statistical data on each individually modeled animal and the population 
as a whole; (2) rate of exposure (threshold exceedance per unit time) 
over the duration of a survey; and (3) the data necessary to determine 
effects based on more sophisticated thresholds, such as cumulative 
sound exposure level. A comparison of these methods--animat method 
involving three-dimensional animal movement modeling and static 
distribution, in which a static two-dimensional density is overlaid on 
a simplified representation of the sound field--found that differences 
consistently arise between the two methods. The static distribution 
method was found to consistently underestimate the number of takes by 
Level B harassment compared with the animat method. In addition, 
repeating many simulations with the animat method provides a more 
robust risk assessment and provides a better measure of variability 
(Schecklman et al., 2011).
    We agree with CRE (and our own statements, as cited by CRE) that 
sophisticated modeling is not a requirement of the MMPA process. 
However, all take estimation requires the use of modeling; the 
difference between various approaches to estimating take is the degree 
of sophistication of the modeling approach employed. We note that the 
National Science Foundation (NSF) typically utilizes the method 
espoused by CRE in take authorization requests for specific surveys. In 
order to derive the necessary estimated isopleth distance, NSF 
applications typically use Nucleus (a source model) in conjunction with 
ray trace modeling to approximate propagation of the acoustic 
signatures. The modeling developed by BOEM and NMFS supports both 
BOEM's 2017 PEIS and the analyses conducted for this rulemaking, and 
additionally is available for use in supporting LOA applications to 
maximize efficiency of the LOA process for disparate applicants. 
However, we have made clear that LOA applicants are free to pursue a 
different method of estimating takes than the modeling effort developed 
collaboratively by NMFS and BOEM. Use of a different analytical method 
in support of an LOA application will necessarily require additional 
review.
    CRE compares ``Line Transect'' modeling performed in support of a 
2004 Minerals Management Service Environmental Assessment to that 
developed in support of this effort, stating that the take estimates 
generated in that effort are ``orders of magnitude smaller than the 
take estimates'' evaluated here. CRE's erroneous implication is that 
the only difference between the two efforts is the modeling approach. 
(``The great difference between GOM takes as estimated by Line 
Transect, and as estimated by [NMFS]'s current models, demonstrates 
just how inaccurate and exaggerated the model take estimates are.'') 
However, the inputs to the two efforts are significantly different. 
Most notably, the assumptions relating to projected effort, animal 
occurrence, and sound source output are not comparable. Effort 
projections for the 2004 modeling were roughly 53 percent of those 
given by BOEM for the high effort scenario in the PEIS, and included 
only relatively archaic 3D survey geometries, versus the more complex 
azimuth designs and coil surveys considered herein. Advances in 
cetacean density modeling provide estimates for use here that are, in 
some cases, multiple orders of magnitude greater than the poor 
estimates used in the 2004 effort. The 15-year old modeling held up by 
CRE as a good example assumed a 4,550 in\3\ acoustic source with a 
uniform 3 km isopleth distance to the 160-dB rms threshold. BOEM 
specified use of an 8,000 in\3\ acoustic source for the modeling effort 
here, with a mean distance to the 160 dB isopleths of 12.7 km, but even 
more recent modeling of a more comparable source (4,130 in\3\) shows 
that the isopleth distance may be as large as 8.4 km, depending on the 
season (Zeddies et al., 2017b). Moreover, the 2004 modeling reduced 
even that ensonified area by an arbitrary 50 percent to account for an 
``elliptical zone of ensonification.'' It is clear that the two 
modeling efforts are in no way comparable.
    Comment: NRDC asserts that NMFS fails to account for forms of 
injury that are reasonably anticipated, stating that permanent hearing 
loss (i.e., Level A harassment) may occur through mechanisms other than 
PTS, and that behaviorally-mediated injury may occur as a result of 
exposure to airgun noise. NRDC states that NMFS must account for these 
mechanisms in its assessment of potential injury.

[[Page 5351]]

    Response: NMFS is aware of the work by Kujawa and Liberman (2009), 
which is cited by NRDC. The authors report that in mice, despite 
completely reversible threshold shifts that leave cochlear sensory 
cells intact, there were synaptic level changes and delayed cochlear 
nerve degeneration. However, the large threshold shifts measured (i.e., 
maximum 40 dB) that led to the synaptic changes shown in this study are 
within the range of the large shifts used by Southall et al. (2007, 
2019a) and in NMFS' 2018 Revised Technical Guidance to define PTS onset 
(i.e., 40 dB). It is unknown whether smaller levels of TTS would lead 
to similar changes or what may be the long-term implications of 
irreversible neural degeneration. The effects of sound exposure on the 
nervous system are complex, and this will be re-examined as more data 
become available. It is important to note that NMFS' 2018 Revised 
Technical Guidance incorporated various conservative factors, such as a 
6-dB threshold shift to represent TTS onset (i.e., minimum amount of 
threshold shift that can be differentiated in most experimental 
conditions); the incorporation of exposures only with measured levels 
of TTS (i.e., did not incorporate exposures where TTS did not occur); 
and assumed no potential of recovery between intermittent exposures. 
NMFS disagrees that consideration of likely PTS is not sufficient to 
account for reasonably expected incidents of auditory injury.
    There is no conclusive evidence that exposure to airgun noise 
results in behaviorally-mediated forms of injury. Behaviorally-mediated 
injury (i.e., mass stranding events) has been primarily associated with 
beaked whales exposed to mid-frequency active (MFA) navy sonar. 
Military tactical sonar and the alerting stimulus used in Nowacek et 
al. (2004) are very different from the noise produced by airguns. One 
should therefore not expect the same reaction to airgun noise as to 
these other sources. Yet NRDC infers that because strandings of beaked 
whales have been correlated with navy MFA sonar use, strandings are 
also likely to occur due to seismic surveys. As explained below, navy 
MFA sonar is very different from airguns, and it is not reasonable to 
assume that airguns will cause the same effects as navy MFA sonar 
(including strandings).
    To understand why navy MFA sonar affects beaked whales differently 
than airguns do, it is important to note the distinction between 
behavioral sensitivity and susceptibility to auditory injury. To 
understand the potential for auditory injury in a particular marine 
mammal species in relation to a given acoustic signal, the frequency 
range the species is able to hear is critical, as well as the species' 
auditory sensitivity to frequencies within that range. Current data 
indicate that not all marine mammal species have equal hearing 
capabilities across all frequencies and, therefore, species are grouped 
into hearing groups with generalized hearing ranges assigned on the 
basis of available data (Southall et al., 2007, 2019a). Hearing ranges 
as well as auditory sensitivity/susceptibility to frequencies within 
those ranges vary across the different groups. For example, in terms of 
hearing range, the high-frequency cetaceans (e.g., Kogia spp.) have a 
generalized hearing range of frequencies between 275 Hz and 160 kHz, 
while mid-frequency cetaceans--such as dolphins and beaked whales--have 
a generalized hearing range between 150 Hz to 160 kHz. Regarding 
auditory susceptibility within the hearing range, while mid-frequency 
cetaceans and high-frequency cetaceans have roughly similar hearing 
ranges, the high-frequency group is much more susceptible to noise-
induced hearing loss during sound exposure, i.e., these species have 
lower thresholds for these effects than other hearing groups (NMFS, 
2018). Referring to a species as behaviorally sensitive to noise simply 
means that an animal of that species is more likely to respond to lower 
received levels of sound than an animal of another species that is 
considered less behaviorally sensitive. So, while dolphin species and 
beaked whale species--both in the mid-frequency cetacean hearing 
group--are assumed to (generally) hear the same sounds equally well and 
be equally susceptible to noise-induced hearing loss (auditory injury), 
the best available information indicates that a beaked whale is more 
likely to behaviorally respond to that sound at a lower received level 
compared to an animal from other mid-frequency cetacean species that is 
less behaviorally sensitive. This distinction is important because, 
while beaked whales are more likely to respond behaviorally to sounds 
than are many other species (even at lower levels), they cannot hear 
the predominant, lower frequency sounds from seismic airguns as well as 
sounds that have more energy at frequencies that beaked whales can hear 
better (such as navy MFA sonar).
    Navy MFA sonar affects beaked whales differently than airguns do 
because it produces energy at different frequencies than airguns. Mid-
frequency cetacean hearing is generically thought to be best between 
8.8 to 110 kHz, i.e., these cutoff values define the range above and 
below which a species in the group is assumed to have declining 
auditory sensitivity, until reaching frequencies that cannot be heard 
(NMFS, 2018). However, beaked whale hearing is likely best within a 
higher, narrower range (20-80 kHz, with best sensitivity around 40 
kHz), based on a few measurements of hearing in stranded beaked whales 
(Cook et al., 2006; Finneran et al., 2009; Pacini et al., 2011) and 
several studies of acoustic signals produced by beaked whales (e.g., 
Frantzis et al., 2002; Johnson et al., 2004, 2006; Zimmer et al., 
2005). While precaution requires that the full range of audibility be 
considered when assessing risks associated with noise exposure 
(Southall et al., 2007, 2019a), animals typically produce sound at 
frequencies where they hear best. More recently, Southall et al. 
(2019a) suggested that certain species amongst the historical mid-
frequency hearing group (beaked whales, sperm whales, and killer 
whales) are likely more sensitive to lower frequencies within the 
group's generalized hearing range than are other species within the 
group and state that the data for beaked whales suggest sensitivity to 
approximately 5 kHz. However, this information is consistent with the 
general conclusion that beaked whales (and other mid-frequency 
cetaceans) are relatively insensitive to the frequencies where most 
energy of an airgun signal is found. Navy MFA sonar is typically 
considered to operate in the frequency range of approximately 3-14 kHz 
(D'Amico et al., 2009), i.e., outside the range of likely best hearing 
for beaked whales but within or close to the lower bounds, whereas most 
energy in an airgun signal is radiated at much lower frequencies, below 
500 Hz (Dragoset, 1990).
    It is important to distinguish between energy (loudness, measured 
in dB) and frequency (pitch, measured in Hz). In considering the 
potential impacts of mid-frequency components of airgun noise (1-10 
kHz, where beaked whales can be expected to hear) on marine mammal 
hearing, one needs to account for the energy associated with these 
higher frequencies and determine what energy is truly ``significant.'' 
Although there is mid-frequency energy associated with airgun noise (as 
expected from a broadband source and as we acknowledged in the notice 
of proposed rulemaking), airgun sound is predominantly below 1 kHz 
(Breitzke et al., 2008; Tashmukhambetov et al., 2008; Tolstoy et al., 
2009). As stated by Richardson et al. (1995), ``[. . .] most emitted 
[seismic airgun] energy is at 10-120 Hz, but the pulses contain some

[[Page 5352]]

energy up to 500-1,000 Hz.'' Tolstoy et al. (2009) conducted empirical 
measurements, demonstrating that sound energy levels associated with 
airguns were at least 20 decibels (dB) lower at 1 kHz (considered 
``mid-frequency'') compared to higher energy levels associated with 
lower frequencies (below 300 Hz) (``all but a small fraction of the 
total energy being concentrated in the 10-300 Hz range'' [Tolstoy et 
al., 2009]), and at higher frequencies (e.g., 2.6-4 kHz), power might 
be less than 10 percent of the peak power at 10 Hz (Yoder, 2002). 
Energy levels measured by Tolstoy et al. (2009) were even lower at 
frequencies above 1 kHz. In addition, as sound propagates away from the 
source, it tends to lose higher-frequency components faster than low-
frequency components (i.e., low-frequency sounds typically propagate 
longer distances than high-frequency sounds) (Diebold et al., 2010). 
Although higher-frequency components of airgun signals have been 
recorded, it is typically in surface-ducting conditions (e.g., DeRuiter 
et al., 2006; Madsen et al., 2006) or in shallow water, where there are 
advantageous propagation conditions for the higher frequency (but low-
energy) components of the airgun signal (Hermannsen et al., 2015). This 
should not be of concern because the likely behavioral reactions of 
beaked whales that can result in acute physical injury would result 
from noise exposure at depth (because of the potentially greater 
consequences of severe behavioral reactions) and because, even if near-
surface exposure to such higher-frequency components were of concern, 
oceanographic conditions in the GOM do not consistently support such 
ducting conditions. In summary, the frequency content of airgun signals 
is such that beaked whales will not be able to hear the signals well 
(compared to MFA sonar), especially at depth where we expect the 
consequences of noise exposure could be more severe.
    Aside from frequency content, there are other significant 
differences between MFA sonar signals and the sounds produced by 
airguns that minimize the risk of severe behavioral reactions that 
could lead to strandings or deaths at sea, e.g., significantly longer 
signal duration, horizontal sound direction, typical fast and 
unpredictable source movement. All of these characteristics of MFA 
sonar tend towards greater potential to cause severe behavioral or 
physiological reactions in exposed beaked whales that may contribute to 
stranding. Although both sources are powerful, MFA sonar contains 
significantly greater energy in the mid-frequency range, where beaked 
whales hear better. Short-duration, high energy pulses--such as those 
produced by airguns--have greater potential to cause damage to auditory 
structures (though this is unlikely for mid-frequency cetaceans, as 
explained later in this document), but it is longer duration signals 
that have been implicated in the vast majority of beaked whale 
strandings. Faster, less predictable movements in combination with 
multiple source vessels are more likely to elicit a severe, potentially 
anti-predator response. Of additional interest in assessing the 
divergent characteristics of MFA sonar and airgun signals and their 
relative potential to cause stranding events or deaths at sea is the 
similarity between the MFA sonar signals and stereotyped calls of 
beaked whales' primary predator: The killer whale (Zimmer and Tyack, 
2007). Although generic disturbance stimuli--as airgun noise may be 
considered in this case for beaked whales--may also trigger 
antipredator responses, stronger responses should generally be expected 
when perceived risk is greater, as when the stimulus is confused for a 
known predator (Frid and Dill, 2002). In addition, because the source 
of the perceived predator (i.e., MFA sonar) will likely be closer to 
the whales (because attenuation limits the range of detection of mid-
frequencies) and moving faster (because it will be on faster-moving 
vessels), any antipredator response would be more likely to be severe 
(with greater perceived predation risk, an animal is more likely to 
disregard the cost of the response; Frid and Dill, 2002). Indeed, when 
analyzing movements of a beaked whale exposed to playback of killer 
whale predation calls, Allen et al. (2014) found that the whale engaged 
in a prolonged, directed avoidance response, suggesting a behavioral 
reaction that could pose a risk factor for stranding. Overall, these 
significant differences between sound from MFA sonar and the mid-
frequency sound component from airguns and the likelihood that MFA 
sonar signals will be interpreted in error as a predator are critical 
to understanding the likely risk of behaviorally-mediated injury due to 
seismic surveys.
    The available scientific literature also provides a useful contrast 
between airgun noise and MFA sonar regarding the likely risk of 
behaviorally-mediated injury. There is strong evidence for the 
association of beaked whale stranding events with MFA sonar use, and 
particularly detailed accounting of several events is available (e.g., 
a 2000 Bahamas stranding event for which investigators concluded that 
MFA sonar use was responsible; Evans and England, 2001). D'Amico et al. 
(2009) reviewed 126 beaked whale mass stranding events over the period 
from 1950 (i.e., from the development of modern MFA sonar systems) 
through 2004. Of these, there were two events where detailed 
information was available on both the timing and location of the 
stranding and the concurrent nearby naval activity, including 
verification of active MFA sonar usage, with no evidence for an 
alternative cause of stranding. An additional ten events were at 
minimum spatially and temporally coincident with naval activity likely 
to have included MFA sonar use and, despite incomplete knowledge of 
timing and location of the stranding or the naval activity in some 
cases, there was no evidence for an alternative cause of stranding.\4\ 
Separately, the International Council for the Exploration of the Sea 
reported in 2005 that, worldwide, there have been about 50 known 
strandings, consisting mostly of beaked whales, with a potential causal 
link to MFA sonar (ICES, 2005). In contrast, very few such associations 
have been made to seismic surveys, despite widespread use of airguns as 
a geophysical sound source in numerous locations around the world.
---------------------------------------------------------------------------

    \4\ The U.S. Navy has publicly stated its agreement that five 
such events since 1996 were associated in time and space with MFA 
sonar use, either by the U.S. Navy alone or in joint training 
exercises with the North Atlantic Treaty Organization. The U.S. Navy 
additionally noted that, as of 2017, a 2014 beaked whale stranding 
event in Crete coincident with naval exercises was under review and 
had not yet been determined to be linked to sonar activities (DoN, 
2017).
---------------------------------------------------------------------------

    A more recent review of possible stranding associations with 
seismic surveys (Castellote and Llorens, 2016) states plainly that, 
``[s]peculation concerning possible links between seismic survey noise 
and cetacean strandings is available for a dozen events but without 
convincing causal evidence.'' The authors' ``exhaustive'' search of 
available information found ten events worth further investigation via 
a ranking system representing a rough metric of the relative level of 
confidence offered by the data for inferences about the possible role 
of the seismic survey in a given stranding event. Only three of these 
events involved beaked whales. Whereas D'Amico et al. (2009) used a 1-5 
ranking system, in which ``1'' represented the most robust evidence 
connecting the event to MFA sonar use, Castellote and Llorens (2016) 
used a 1-6 ranking system, in which ``6'' represented the most robust 
evidence

[[Page 5353]]

connecting the event to the seismic survey. As described above, D'Amico 
et al. (2009) found that two events were ranked ``1'' and ten events 
were ranked ``2'' (i.e., 12 beaked whale stranding events were found to 
be associated with MFA sonar use). In contrast, Castellote and Llorens 
(2016) found that none of the three beaked whale stranding events 
achieved their highest ranks of 5 or 6.\5\ However, we acknowledged in 
the notice of proposed rulemaking that one of these stranding events, 
involving two Cuvier's beaked whales, was contemporaneous with and 
reasonably associated spatially with a 2002 seismic survey in the Gulf 
of California, and here acknowledge the same for the 2007 Gulf of Cadiz 
seismic survey discussed by Castellote and Llorens (also involving two 
Cuvier's beaked whales). However, neither event was considered a ``true 
atypical mass stranding'' (according to Frantzis [1998]) as used in the 
analysis of Castellote and Llorens (2016). While we agree with the 
authors that this lack of evidence should not be considered conclusive, 
it is clear that there is very little evidence that seismic surveys 
should be considered as posing a significant risk of acute harm to 
marine mammals.
---------------------------------------------------------------------------

    \5\ Of the ten total events, none achieved the highest rank of 
6. Two events were ranked as 5: One stranding in Peru involving 
dolphins and porpoises and a 2008 stranding in Madagascar. This 
latter ranking can only broadly be associated with the survey 
itself, as opposed to use of seismic airguns. An exhaustive 
investigation of this stranding event, which did not involve beaked 
whales, concluded that use of a high-frequency mapping system (12-
kHz multibeam echosounder) was the most plausible and likely initial 
behavioral trigger of the event, which was likely exacerbated by 
several site- and situation-specific secondary factors. The review 
panel found that seismic airguns were used after the initial 
strandings and animals entering a lagoon system, that airgun use 
clearly had no role as an initial trigger, and that there was no 
evidence that airgun use dissuaded animals from leaving (Southall et 
al., 2013).
---------------------------------------------------------------------------

    Comment: NRDC asserts that NMFS has failed to account adequately 
for the effects of stress on marine mammals.
    Response: As NRDC acknowledges, we addressed the available 
literature regarding potential impacts of stress resulting from noise 
exposure in marine mammals. As described in that discussion, stress 
responses are complicated and may or may not have meaningful impacts on 
marine mammals. NRDC implies that NMFS must (1) enumerate takes 
resulting from stress alone and (2) specifically address stress in its 
negligible impact analysis. The effects of stress are not 
straightforward, and there is no information available to inform an 
understanding of whether it is reasonably likely that an animal may 
experience a stress response upon noise exposure that would not be 
accounted for in NMFS' existing enumeration of takes via exposure to 
noise, which includes an accounting for exposures above received levels 
as low as 140 dB rms (and as low as 120 dB rms for beaked whales). NRDC 
provides nothing informative regarding how such an analysis might be 
carried out. With regard to NMFS' negligible impact analysis, we 
believe that the potential effects of stress are addressed and subsumed 
within NMFS' considerations of severity of effect and vulnerability of 
affected populations. Similarly, NRDC provides no justification as to 
why stress would appropriately be considered separately in this 
analysis, and no useful recommendation as to how to do so, if 
appropriate. We believe we have appropriately acknowledged the 
potential effects of stress, and that these potential effects are 
accounted for within our overall assessment of potential effects on 
marine mammals.
    Comment: NRDC states that masking results in take of marine mammals 
and that NMFS must account for this in its take estimates.
    Response: We addressed our consideration of masking in greater 
detail in a previous response. We acknowledge that masking may impact 
marine mammals, particularly baleen whales such as the Bryde's whale, 
and particularly when considered in the context of the full suite of 
regulated and unregulated anthropogenic sound contributions overlaying 
an animal's acoustic habitat. We acknowledge that masking can 
constitute a take, depending on the particular circumstances, but do 
not agree that masking effects from the incremental noise contributions 
of individual activities or sound sources always rise to the level of 
take. Further, not all takes are readily quantifiable. In this case, 
while masking is considered in the analysis, we do not believe it will 
result in take of marine mammals beyond those that have already been 
quantified as taken by behavioral harassment. Specifically, in the case 
of these proposed activities, in the event that some masking incidents 
rise to the level of a take, we would expect them to be accounted for 
in the quantified exposures above the harassment thresholds. Given the 
short duration of expected noise exposures, any take by masking in the 
case of these surveys would be most likely to be incurred by 
individuals either exposed briefly to notably higher levels or those 
that are generally in the wider vicinity of the source for 
comparatively longer times. Both of these situations would be captured 
in the enumeration of takes by Level B harassment, which accounts for 
takes that may occur upon exposure at relatively low levels of received 
sound (e.g., 140 dB).
    Comment: MMC commented that the aversion adjustment applied to 
estimates of Level A harassment proposed by NMFS for low- and high-
frequency cetaceans is not supported. NRDC provided similar comments.
    Response: NMFS disagrees with these comments, and clarifies our 
position given the misunderstanding evident in the comments. The MMC 
cites NMFS' statements that ``too little is known about the factors 
that lead to avoidance of sounds to quantify aversive behavior for 
survey activities when modeling marine mammal exposure to sound'' and 
that ``aversion is a context-dependent behavioral response affected by 
biological factors, including energetic and reproductive state, 
sociality, and health status of individual animals'' in characterizing 
our subsequent use of a post-hoc correction factor to account for 
aversion as an ``apparent contradiction.'' Similarly, NRDC cites NMFS' 
statement that aversion was not quantified in the modeling process due 
to lack of information regarding species-specific degree of aversion 
and level of onset in criticizing the adjustment that was later made.
    Aversion is a known real-world phenomenon. It is well-known that 
animals will avoid unpleasant stimuli, such as very high received 
levels of sound. A large and growing literature has demonstrated 
behavioral aversion in a number of contexts for many marine mammal 
species in increasingly controlled and well-documented contexts. While 
considerable species, individual, and context-dependencies exist in 
terms of received noise levels associated with behavioral aversion, 
clear patterns of behavioral aversion have been demonstrated 
empirically within odontocetes and mysticetes (e.g., Miller et al., 
2012, 2014; DeRuiter et al., 2013; Southall et al., 2019b). This is 
particularly true for exposure scenarios in which animals occur 
relatively close to sources and at the high levels that would be 
required for even TTS (much less PTS) to occur. In some instances, in 
these and other studies, behavioral avoidance has been measured at 
received levels many orders of magnitude below those required for 
predicted PTS onset and even below the nominal, 50 percent behavioral 
response probability at 160 dB rms that NMFS has applied historically.
    However, accounting for aversion quantitatively in an acoustic 
exposure modeling process is a significantly data-

[[Page 5354]]

heavy endeavor and, as we noted, despite the growing body of evidence 
there is at this time still not sufficient data regarding the specific 
degree of aversion and level of onset on a species-specific basis. That 
is, in order to account for aversion within the modeling process, one 
must program individual animats representing different species to 
respond at a specific received level by changing their direction of 
travel by a specific degree and assuming a specific rate of speed. 
Through a test scenario evaluation (discussed in the notice of proposed 
rulemaking), we determined that while this is possible to do, the 
specific values that must be used in programming the animat response 
could not be adequately derived. Instead, a nominal offset factor was 
applied to the modeled injurious exposures based on published model 
result evaluation to account for aversion.
    Ellison et al. (2016) modeled scenarios using animal movement 
models to evaluate predicted PTS in which no aversion was assumed 
relative to scenarios where reasonable assumptions were made about 
aversion, in line with historical response probability assumptions and 
that existing scientific literature suggest are appropriate. Scenarios 
where no aversion probability was used overestimated the potential for 
high levels of exposure required for PTS by about five times. 
Accordingly, total modeled injurious exposures calculated without 
accounting for behavioral aversion (for low- and high-frequency 
species) were multiplied by 0.2 as part of the EWG risk analysis. NMFS 
consulted the EWG in selecting the specific offset factor, and 
discussed that selection again in context of the public comments 
received. The EWG--which is composed of some of the foremost scientists 
in the field of marine mammal behavioral response study, and includes 
the lead author of the Ellison et al. (2016) study--agreed that the 
approach and specific offset factor was a reasonable and likely 
conservative approach to addressing the issue of aversion.
    The commenters do not dispute that aversion is a meaningful real-
world phenomenon that is significantly influential on actual occurrence 
of Level A harassment. As NRDC acknowledges, ``it is certainly true 
that some marine mammals will flee the sound.'' Yet the commenters 
would have us ignore this phenomenon and assume unrealistically high 
amounts of auditory injury for marine mammals in the GOM. NMFS does not 
agree that this would be appropriate. As described above, there is 
extensive information supporting the aversion concept in marine 
mammals, but limited quantitative data with which to develop precise, 
species-specific offset factors. Accordingly, utilizing the available 
data and expert input, NMFS applied its professional judgement in order 
to account for this meaningful phenomenon.
    Comment: NRDC disagrees with NMFS' conclusion that Level A 
harassment is not likely to occur for mid-frequency (MF) cetaceans and 
states that this ``problem [. . .] must be addressed.''
    Response: As was explained in the notice of proposed rulemaking, 
the number of modeled incidents of Level A harassment for MF cetacean 
species is not realistic. The modeled isopleth distance to the relevant 
Level A harassment threshold, i.e., the predominant MF peak pressure 
threshold, is only 18 m. As we explained in the notice of proposed 
rulemaking, it is understandable that even such a small assumed area 
could lead to the results given when a real-world density value is 
sufficiently high to lead to non-zero scaled 24-hr modeled exposure 
results, which are then multiplied by large numbers of notional survey 
days. We explain in greater detail below why relatively small zones, 
i.e., zones contained within the near-field of an airgun array, should 
not be expected to result in actual injurious exposure. NRDC also 
appears to be under the impression that the conclusion was based on 
what they refer to as ``shorter injurious take distances assumed in the 
Gulf of Mexico modeling than in modeling for seismic in other regions, 
such as the Atlantic,'' an apparent misunderstanding on the part of the 
commenter that they refer to as a ``discrepancy'' that is ``never 
explained'' and ``appears arbitrary.'' Given the lack of detail 
provided, NMFS cannot be sure what NRDC is referring to. However, we do 
know that state-of-science propagation modeling performed for a 
notional array here provided the 18 m result described above. For five 
different, real-world arrays evaluated for use in the Atlantic Ocean 
(83 FR 63268; December 7, 2018), the calculated isopleth distance to 
the 230 dB peak sound pressure level (SPL) MF Level A harassment 
threshold was an average 27 m (range 14-63 m), in keeping with the 
value calculated here.
    For MF cetaceans, the only potential injury zones will be based on 
the peak pressure metric, as such zones will be larger than those 
calculated on the basis of the cumulative sound exposure level (SEL) 
metric (which are essentially non-existent for MF and HF cetaceans). As 
noted, the estimated zone size for the 230 dB peak threshold for MF 
cetaceans is only 18 m. In a theoretical modeling scenario, it is 
possible for animats to engage with such a small assumed zone around a 
notional point source and, subsequently, for these interactions to 
scale to predictions of real-world exposures given a sufficient number 
of predicted 24-hr survey days in confluence with sufficiently high 
predicted real-world animal densities--i.e., the modeling process that 
resulted in the predicted exposure estimates for MF cetaceans in the 
modeling report. However, this is not a realistic outcome. The source 
level of the array is a theoretical definition assuming a point source 
and measurement in the far-field of the source (MacGillivray, 2006). As 
described by Caldwell and Dragoset (2000), an array is not a point 
source, but one that spans a small area. In the far-field, individual 
elements in arrays will effectively work as one source because 
individual pressure peaks will have coalesced into one relatively broad 
pulse. The array can then be considered a ``point source.'' For 
distances within the near-field, i.e., approximately 2-3 times the 
array dimensions, pressure peaks from individual elements do not arrive 
simultaneously because the observation point is not equidistant from 
each element. The effect is destructive interference of the outputs of 
each element, so that peak pressures in the near-field will be 
significantly lower than the output of the largest individual element. 
Here, the 230 dB peak isopleth distances would be expected to be within 
the near-field of the arrays where the definition of source level 
breaks down. Therefore, actual locations within this distance (i.e., 
within 18 m) of the array center where the sound level exceeds 230 dB 
peak SPL would not necessarily exist. In general, Caldwell and Dragoset 
(2000) suggest that the near-field for airgun arrays is considered to 
extend out to approximately 250 m.
    In order to provide quantitative support for this theoretical 
argument, we calculated expected maximum distances at which the near-
field would transition to the far-field for five specific, real-world 
arrays proposed for use in the Atlantic Ocean (83 FR 63268). The 
average distance to the near-field calculated for the five arrays, 
following the process described below, was 203 m (range 80-417 m).
    For a specific array one can estimate the distance at which the 
near-field transitions to the far-field by:

[[Page 5355]]

[GRAPHIC] [TIFF OMITTED] TR19JA21.003

with the condition that D >> [lgr], and where D is the distance, L is 
the longest dimension of the array, and [lgr] is the wavelength of the 
signal (Lurton, 2002). Given that [lgr] can be defined by:
[GRAPHIC] [TIFF OMITTED] TR19JA21.004

where f is the frequency of the sound signal and v is the speed of the 
sound in the medium of interest, one can rewrite the equation for D as:
[GRAPHIC] [TIFF OMITTED] TR19JA21.005

and calculate D directly given a particular frequency and known speed 
of sound (here assumed to be 1,500 meters per second in water, although 
this varies with environmental conditions).
    To determine the closest distance to the array at which the modeled 
source level prediction is valid (i.e., maximum extent of the near-
field), we calculated D based on an assumed frequency of 1 kHz. A 
frequency of 1 kHz is commonly used in near-field/far-field 
calculations for airgun arrays (Zykov and Carr, 2014; MacGillivray, 
2006; NSF and USGS, 2011), and based on representative airgun spectrum 
data and field measurements of an airgun array used on the R/V Marcus 
G. Langseth, nearly all (greater than 95 percent) of the energy from 
airgun arrays is below 1 kHz (Tolstoy et al., 2009). Thus, using 1 kHz 
as the upper cut-off for calculating the maximum extent of the near-
field should reasonably represent the near-field extent in field 
conditions.
    If the largest distance to the peak sound pressure level threshold 
was equal to or less than the longest dimension of the array (i.e., 
under the array), or within the near-field, then received levels that 
meet or exceed the threshold in most cases are not expected to occur. 
This is because within the near-field and within the dimensions of the 
array, the specified source level is overestimated and not applicable. 
In fact, until one reaches a distance of approximately three or four 
times the near-field distance, the average intensity of sound at any 
given distance from the array is still less than that based on 
calculations that assume a directional point source (Lurton, 2002). For 
example, an airgun array used on the R/V Marcus G. Langseth has an 
approximate diagonal of 29 m, resulting in a near-field distance of 140 
m at 1 kHz (NSF and USGS, 2011). Field measurements of this array 
indicate that the source behaves like multiple discrete sources, rather 
than a directional point source, beginning at approximately 400 m (deep 
site) to 1 km (shallow site) from the center of the array (Tolstoy et 
al., 2009), distances that are actually greater than four times the 
calculated 140-m near-field distance. Within these distances, the 
recorded received levels were always lower than would be predicted 
based on calculations that assume a directional point source, and 
increasingly so as one moves closer towards the array (Tolstoy et al., 
2009). Given this, relying on the calculated distances as the distances 
at which we expect to be in the near-field is a conservative approach 
because even beyond this distance the acoustic modeling still 
overestimates the actual received level.
    Within the near-field, in order to explicitly evaluate the 
likelihood of exceeding any particular acoustic threshold, one would 
need to consider the exact position of the animal, its relationship to 
individual array elements, and how the individual acoustic sources 
propagate and their acoustic fields interact. Given that within the 
near-field and dimensions of the array source levels would be below the 
modeled notional source level, we believe exceedance of the peak 
pressure threshold would only be possible under highly unlikely 
circumstances.
    Therefore, we expect the potential for Level A harassment of MF 
cetaceans to be de minimis, even before the likely moderating effects 
of aversion and/or other compensatory behaviors (e.g., Nachtigall et 
al., 2018) are considered. We do not believe that Level A harassment is 
a likely outcome for any MF cetacean.
    Comment: The MMC comments that the estimated numbers of Level B 
harassment must be increased to account for the incidents of acoustic 
exposure that were modeled as injurious but subsequently discounted due 
to aversion. NRDC commented similarly.
    Response: NMFS agrees that animals that avoid Level A harassment 
through aversive behavior should be considered as having been subject 
to Level B harassment and increased the Level B harassment estimates 
accordingly. However, these estimates have been superseded by the 
revised estimates submitted by BOEM in support of their revised scope 
of activity.

Marine Mammal Protection Act--General

    Comment: The MMC recommended that any ``formal interpretation'' by 
NMFS of MMPA standards, such as the least practicable adverse impact 
standard and small numbers standard, be issued in stand-alone, 
generally applicable rulemakings (e.g., in amendments to 50 CFR 216.103 
or 216.105) or in a separate policy directive, rather than in the 
preambles to individual proposed rules.
    Response: We appreciate the recommendation and may consider the 
recommended approaches in the future. However, providing directly 
relevant explanations of programmatic approaches or interpretations 
related to the incidental take provisions of the MMPA in a proposed 
incidental take authorization is an effective and efficient way to 
provide information to and solicit focused input from the public. 
Further, this approach ultimately affords the same

[[Page 5356]]

opportunities for public comment as a stand-alone rulemaking would.
    Regarding the least practicable adverse impact standard, NMFS has 
provided similar explanations in other recent section 101(a)(5)(A) 
rules. See, e.g., 83 FR 66846 (December 27, 2018) (U.S. Navy Training 
and Testing Activities for Hawaii-Southern California Study Area).

Least Practicable Adverse Impact

    Comment: NRDC believes NMFS relies on a ``flawed interpretation'' 
of the least practicable adverse impact standard. They state that NMFS 
(1) wrongly imports a population-level focus into the standard, 
contrary to the ``clear'' holding of the Ninth Circuit in NRDC v. 
Pritzker, 828 F.3d 1125 (9th Cir. 2016); and (2) inappropriately 
``balances'' or weighs effectiveness against practicability without 
sufficient analysis, counter to Pritzker.
    Response: NMFS carefully evaluated the Ninth Circuit's opinion in 
Pritzker and believe we have fully addressed the court's concerns. 
NMFS' discussion of the least practicable adverse impact standard in 
the Mitigation section explains why we believe a population focus is a 
reasonable interpretation of the standard.
    With regard to the second point, NMFS disagrees that the analysis 
is insufficient. NMFS' interpretation of the LPAI standard is a 
reasonable interpretation that gives effect to the language in the 
statute and the underlying legislative intent. Congress intended the 
agencies administering section 101(a)(5)(A) to consider practicability 
when determining appropriate mitigation, and we do not believe the 
analysis must be conducted in a rigid sequential fashion. There is a 
tension inherent in the phrase ``least practicable adverse impact'' in 
that ``least [. . .] adverse impact'' pulls in favor of one direction 
(i.e., expanding mitigation) while ``practicable'' pulls in favor of 
the other direction (i.e., limiting mitigation), and weighing the 
relative costs and benefits is, in NMFS' view, a meaningful way to 
address and resolve this tension. Further, as described in the proposed 
rule and augmented in this final rule in both the Mitigation section 
and the response to comments, NMFS considered all recommended 
mitigation in the context of both the reduction of impacts on marine 
mammal species and stocks and their habitat and the practicability of 
such mitigation in reaching the required set of measures that we 
believe satisfy the least practicable adverse impact standard.
    Comment: The Associations assert that NMFS failed to provide 
sufficient practicability analyses for the proposed mitigation 
requirements.
    Response: No guidance is provided by the MMPA or NMFS' implementing 
regulations as to what constitutes ``practicability'' for the non-
military readiness activities considered here, or how to ascertain 
whether a proposed measure is practicable. Neither the term 
``practicable'' nor the phrase ``least practicable adverse impact'' is 
defined by the MMPA or in NMFS' implementing regulations. (See 
Mitigation, later in this document, for extensive discussion on NMFS' 
interpretation of the meaning of ``least practicable adverse impact.'') 
Therefore, while the MMPA's requirement to prescribe mitigation 
achieving the ``least practicable adverse impact'' demands 
consideration of practicability, the need for additional ``analysis'' 
of unspecified scope, detail, or methodology, as demanded by the 
Associations, cannot be found in the statute, legislative history, 
regulations, or case law.
    However, NMFS does not start from scratch. Our implementing 
regulations at 50 CFR 216.104(a)(11) require applications for 
incidental take authorizations to include information about the 
availability and feasibility (economic and technological) of equipment, 
methods, and manner of conducting such activity or other means of 
effecting the least practicable adverse impact upon the affected 
species or stocks, their habitat, and on their availability for 
subsistence uses, paying particular attention to rookeries, mating 
grounds, and areas of similar significance. This often provides the 
foundation of NMFS' proposed mitigation, after consideration of the 
objectives of those and other possible measures and how they may 
achieve those objectives as well as, when possible, what we know about 
the practicability of the proposed measures.
    As a general matter, where an applicant proposes measures that are 
likely to reduce impacts to marine mammals, the fact that they are 
included in the proposal and application indicates that the measures 
are practicable, and it is not necessary for NMFS to conduct a detailed 
analysis of the measures the applicant proposed (rather, they are 
simply included). However, it is incumbent on NMFS to consider whether 
there are other practicable measures that would contribute to the 
reduction of risk or severity of adverse effects on the species or 
stocks.
    We then seek public comment on the proposal and, if contradictory 
information is presented by members of the public (including 
prospective applicants), the information is considered in making a 
decision regarding whether to retain, modify, or eliminate a proposed 
measure.
    Our notice of proposed rulemaking presented specific discussion of 
practicability considerations, including both the monetized direct 
costs of proposed measures as well as what we understand about 
potential indirect costs, and provided detailed discussion relating to 
certain measures. While much of this analysis was conducted under a 
regulatory impact analysis (RIA) conducted pursuant to Executive Order 
12866, as stated by the Associations, the utility of the analysis is 
not limited to use there. For example, while the Associations claim 
that NMFS fails to ``consider impacts beyond immediate operational 
impacts,'' the RIA provides a detailed analysis of the sort of 
speculative indirect costs of concern to industry, and the RIA's 
analysis is incorporated into NMFS' consideration of practicability. 
Overall, we note that the Associations' comments are peppered with 
reference to cost increases, both vague (``resulting in millions of 
dollars of added cost'') and specific (``increase costs an estimated 5% 
to 20%''), but without sufficient supporting data.
    NMFS interprets ``practicable'' simply as capable of being put into 
practice or of being done or accomplished. Practicability of the 
standard operational protocols was reasonably assumed in consideration 
of the fact that they are included in many incidental take 
authorizations and that we did not receive any specific public comments 
to the contrary. Moreover, many of these measures were proposed by the 
applicant (BOEM) in their petition for regulations, including ramp-up 
and shutdown requirements and a requirement to observe a time-area 
restriction in coastal waters to protect bottlenose dolphins during the 
time of their reproductive activity peak. The Associations claim that 
our proposal applies these standard measures in such a way as to extend 
their ``geographic and temporal scope or to circumstances where they 
are unnecessary or impossible to implement,'' but provide no specific 
information as to what measures they specifically refer to, in what 
circumstances they believe specific measures are unnecessary, or in 
what circumstances specific measures are impossible to implement. The 
Associations assert that NMFS' considerations of practicability ``fail 
to adequately estimate levels of current and future geophysical work or 
consider costs and impacts beyond the immediate survey work,'' but 
their

[[Page 5357]]

comments provide no specific information to enable NMFS to assess its 
consideration of practicability. NMFS' consideration of practicability 
was sufficient and in accordance with law, and the Associations 
provided no specific contradictory information for NMFS' evaluation.
    Comment: The MMC recommended that NMFS rework its evaluation 
criteria for applying the least practicable adverse impact standard to 
separate the factors used to determine whether a potential impact on 
marine mammal species or stocks or their habitat is adverse and whether 
possible mitigation measures would be effective. In this regard, the 
MMC asserted that it seems as though the proposed ``effectiveness'' 
criterion more appropriately fits as an element of practicability and 
should be addressed under that prong of the analysis. In other words, a 
measure not expected to be effective should not be considered a 
practicable means of reducing impacts.
    Response: In the Mitigation section, NMFS has explained in detail 
its interpretation of the least practicable adverse impact standard, 
the rationale for the interpretation, and our approach for implementing 
the interpretation. The ability of a measure to reduce effects on 
marine mammals is entirely related to its ``effectiveness'' as a 
measure, whereas the effectiveness of a measure is not connected to its 
practicability. The MMC did not support its argument with scientific 
information, and NMFS has not implemented the suggestion.
    Comment: The MMC recommended that NMFS address the habitat 
component of the least practicable adverse impact provision in greater 
detail. It asserted that NMFS' discussion of critical habitat, marine 
sanctuaries, and biologically important areas (BIA) in the proposed 
rule is not integrated with the discussion of the least practicable 
adverse impact standard. As stated by the MMC, it would seem that, 
under the least practicable adverse impact provision, adverse impacts 
on important habitat should be avoided whenever practicable. Therefore, 
to the extent that activities would be allowed to proceed in these 
areas, NMFS should explain why it is not practicable to constrain them 
further. The MMC also suggests that NMFS intends to defer consideration 
of measures to protect habitat to individual LOAs, rather than 
addressing such measures in the regulations, as the MMC contends is 
required.
    Response: Marine mammal habitat value is informed by marine mammal 
presence and use and, in some cases, there may be overlap in mitigation 
measures for the species or stock directly and for use of habitat. In 
this rule, NMFS has identified one time-area restriction (carried 
forward from the proposed rule) based on a combination of factors that 
include higher densities and observations of specific important 
behaviors of marine mammals themselves, but also that clearly reflect 
preferred habitat. In addition to being delineated based on physical 
features that drive habitat function (e.g., bathymetric features, among 
others for some BIAs), the high densities and concentration of certain 
important behaviors (e.g., feeding) in these particular areas indicate 
the presence of preferred habitat. The MMC seems to suggest that NMFS 
must always consider separate measures aimed at marine mammal habitat. 
However, the MMPA does not specify that effects to habitat must be 
mitigated in separate measures, and NMFS has identified measures that 
provide significant reduction of impacts to both ``marine mammal 
species and stocks and their habitat,'' as required by the statute. 
Finally, we clarify here that all measures to reduce impacts to both 
marine mammal species and stocks and their habitat are included in the 
regulations and then implemented through activity-specific LOAs.

Negligible Impact

    Comment: The Associations and Chevron concur with NMFS' finding 
that the incidental taking that may be authorized under the ITR will 
have a negligible impact on the affected marine mammal stocks. The 
Associations additionally specify their agreement with NMFS' 
conclusions that Level A harassment will not play a meaningful role in 
the overall degree of impact experienced by marine mammal populations 
as a result of the projected survey activity and that mid-frequency 
cetaceans are unlikely to incur Level A harassment, as well as with 
NMFS' use of the Wood et al. (2012) probabilistic risk function.
    Response: NMFS appreciates the comments.
    Comment: NRDC claims that NMFS did not define the total amount of 
take it evaluated in making the negligible impact determination and 
asserts that the proposed rule is unclear about the data and 
calculations that informed the basis of the negligible impact finding.
    Response: NMFS disagrees with these comments. NMFS explicitly 
defined the basis, as well as the process, for the negligible impact 
analysis. Although the negligible impact analysis was built upon 
relatively sophisticated acoustic exposure modeling, and incorporated 
advances in the science of risk assessment, the informational inputs to 
the analysis and the analytical framework were clearly elucidated and 
the supporting documentation identified and provided as companion 
documents to the public for review in association with our negligible 
impact analysis. The notice of proposed rulemaking identified a point 
of contact available to provide further information or answer questions 
if necessary.
    NMFS stated that the ``specified activity'' for the proposed 
regulations is a broad program of geophysical survey activity that 
could occur at any time of year in U.S. waters of the GOM. This 
conceptual program, as defined by BOEM through projected levels of 
survey effort, was described and shown in Table 1 of the notice of 
proposed rulemaking. These annual survey projections aligned generally 
with ``low,'' ``moderate,'' and ``high'' effort years (83 FR 29224). 
These projected levels of survey effort informed the acoustic modeling 
report (Zeddies et al. 2015, 2017a), which was extensively and clearly 
summarized in the notice of proposed rulemaking, while the report 
itself was made available for public review concurrently with the 
notice of proposed rulemaking. In order to reasonably estimate the 
actual effort that might occur over the five-year timeframe of the 
proposed ITR, NMFS determined for the proposed rule analysis that it 
would be appropriate to assume that one high-effort year, two moderate-
effort years, and two low-effort years (and, therefore, associated 
acoustic exposure estimates) would occur. NMFS then selected and 
identified the specific effort scenarios that formed the basis for the 
analysis in association with Table 9 of the notice of proposed 
rulemaking, titled ``Scenario-Specific Expected Take Numbers and Mean 
Annual Take Level.'' Table 9 of the notice identified the annual and 
total amounts of take that NMFS expected to occur under the ITR. The 
preliminary negligible impact analysis then referred back to Table 9 as 
the basis for the analytical process and discussion provided therein 
(See 83 FR 29290-29291).
    NRDC complains that ``NMFS never defines the total amount of take 
it proposes to authorize.'' However, as is typical for a programmatic 
analysis, the ITR and its associated analysis (including negligible 
impact) do not propose to authorize take per se, but rather to provide 
a description of the upper bound within which take may be authorized 
via LOAs. The upper bounds of the instances of take that may be 
authorized under this rule are indicated

[[Page 5358]]

in Table 9 in this final rule. The actual amount of take authorized 
through LOAs under the ITR will be determined by applicant interest 
(subject to the upper bound).
    NMFS also identified the Expert Working Group (EWG) report 
(Southall et al., 2017) as an essential companion to the notice of 
proposed rulemaking and, similar to the acoustic modeling report, 
provided the document for concurrent public review. The EWG report 
describes the systematic risk assessment framework that, in part, forms 
the basis for the negligible impact analysis. We concisely described 
the analytical framework in the notice and provided the results of that 
analysis. Ultimately, the EWG report provides overall evaluated 
relative risk for each of the three effort scenarios (low, moderate, 
high) for each species in each of seven different zones. As stated in 
the notice, the severity and vulnerability ratings (facets of the 
analytical framework that are also clearly explained both in the EWG 
report and the notice) are integrated to provide relative impact 
ratings of overall risk. These zone-specific relative impact ratings 
for each species were then integrated using basic calculations to 
produce species-specific, GOM-wide overall evaluated relative risk 
ratings for each of the three effort scenarios. Overall vulnerability 
scores for each species were produced by summing the zone-specific 
vulnerability scores, as scaled to the zone-specific population. For 
example, the Zone 1 vulnerability score is multiplied by the ratio of 
the Zone 1 population to the total population. These zone-specific 
products are then summed. Overall severity scoring is calculated as the 
proportion of the sum of scenario-specific takes to the total 
population. These two factors are then integrated as described in the 
EWG report.
    NRDC also states that the ``actual percentages of populations 
affected by takes'' are not provided. NMFS disagrees, as this 
information can be replicated using information that was provided to 
the public via the acoustic modeling report. Additional underlying data 
are necessary to replicate zone-specific findings. Excel workbooks 
containing these data were made publicly available by BOEM during 
review of their PEIS. NMFS did not view these additional data as 
essential to understanding the modeling report or the proposed ITR and 
did not publish these data on its website. Members of the public 
interested in further exploration of the information provided in the 
modeling report, or in need of assistance regarding their independent 
analysis of the modeling report, could have contacted the NMFS point of 
contact identified in the notice of proposed rulemaking.
    In sum, NMFS provided sufficient information in support of its 
negligible impact analysis affording the public meaningful opportunity 
to comment. Further, consistent with a potential alternative scope 
identified in the proposed rule that would remove the Eastern Planning 
Area (EPA), the scope of this final rule has been modified to remove 
the GOMESA area, which includes most of the EPA (and a small portion of 
the Central Planning Area), based on BOEM's update to its action. This 
has resulted in a reduction in the upper bounds of the instances of 
take that may be authorized for all species pursuant to this final rule 
(see Tables 8 and 9).
    Comment: The MMC commented similarly to NRDC, expressing some 
concern regarding the risk assessment framework and asserting 
``apparent inconsistencies,'' while recommending that NMFS (1) provide 
the final risk assessment framework, underlying results, and its 
interpretation of those results to the public and (2) allow for an 
additional 30-day comment period to review the findings sufficiently in 
advance of issuing the final rule.
    Response: NMFS disagrees with the MMC comments. They state that the 
EWG report and analysis ``has some apparent inconsistencies'' as 
compared against the preamble to the proposed rule because the 
scenario-specific high, moderate, and low values presented in Table 3 
of the EWG report do not align with the summary minimum, maximum, and 
mean values given in Table 2 of the notice of proposed rulemaking. We 
note that the MMC provided clarifying questions to NMFS during the 
public comment period in advance of submitting a formal comment letter 
and expressed some confusion regarding Table 2 of that notice at that 
time. As was explained to the MMC then, Table 2 of the preamble was 
provided for illustrative purposes only, as a way of providing a more 
concise look at the information given in Table 1 of the preamble. As 
was explained, the values given in Table 2 were not consequential with 
regard to anything that followed in the preamble. NMFS regrets any 
confusion caused by inclusion of Table 2 in the notice of proposed 
rulemaking but explained clearly to the MMC that the table was not 
related to the analysis. It has been removed from this final rule.
    Separately, the MMC states that ``neither NMFS nor BOEM stipulated 
why only certain years were selected for analysis,'' claiming that NMFS 
indicated that years 1, 4, and 9 were used in the analysis ``upon 
further inquiry.'' This is incorrect. In the notice of proposed 
rulemaking, we stated that ``Year 1 provides an example of what might 
be a high-effort year in the GOM, while Year 9 is representative of a 
low-effort year. A moderate level of effort in the GOM, according to 
these projections, would be similar to the level of effort projected 
for Year 4.'' (83 FR 29224.) NMFS provided explanation of its choices 
in the notice of proposed rulemaking (see, e.g., 83 FR 29261-29262, 
29290).
    This portion of the MMC's recommendation regarding representative 
years is no longer relevant to this final rule. As discussed 
previously, BOEM revised the scope of the activity and provided revised 
effort projections and resulting take estimates accordingly. The 
revised take estimates provided by BOEM reflect years 1-5 of their 
original level of effort projections and, therefore, the question of 
rationale behind the selection of years 1, 4, and 9 is no longer 
relevant.
    Regarding the notice of proposed rulemaking, the MMC also states 
that supposed discrepancies between zone-specific risk ratings and risk 
derived per year across the GOM are ``inconsistencies.'' Zone-specific 
risk ratings for any given effort scenario are driven by the actual 
effort within that zone for that scenario, while the overall level of 
effort GOM-wide underlies the labeling of scenarios as ``high,'' 
``moderate,'' and ``low.'' For example, although year 1 was designated 
as the ``high'' effort scenario and year 4 the ``moderate'' effort 
scenario on the basis of the total projected GOM-wide survey days 
(2,286 and 1,902, respectively), the ``high'' effort scenario actually 
includes significantly less projected effort in zones 2 and 4 than does 
the ``moderate'' effort scenario. Therefore, risk ratings for certain 
species were higher in those specific zones for the ``moderate'' effort 
scenario than they were for the ``high'' effort scenario. This was 
explained in our notice of proposed rulemaking: ``[P]er-zone ranges can 
provide a different outlook than does an assessment of total year 
projected effort across zones. For example, in the ``high'' effort 
annual scenario (Year 1; considering total projected survey days across 
zones), there are 263 projected survey days in Zone 2, while the 
``moderate'' effort annual scenario (Year 4) projects 446 survey days 
in Zone 2.'' This was explained directly to the MMC upon its informal 
inquiry during the public comment period. The MMC also stated to NMFS 
at that time that ``the

[[Page 5359]]

relative risk scores for certain species [. . .] do not make sense, 
presumably because they are based on the incorrect number of estimated 
survey days,'' giving as an example that ``rough-toothed dolphins in 
Zone 5 have an overall Moderate risk in the High and Low scenario 
years, but a Low risk in the Moderate scenario year.'' We reiterated to 
the MMC at that time that what the MMC viewed as illogical and 
erroneous did not in fact reflect errors, but rather the confluence of 
zone-specific activity levels and species presence for a given year. 
The effort scenarios used as the basis for the analysis were clearly 
identified, and there were no inconsistencies in terms of risk ratings 
in consideration of the zone-specific information underlying those 
ratings (which was explained in the notice of proposed rulemaking).
    Separately, the MMC stated its view that ``the basis for 
determining the relative risk thresholds, relative rating thresholds, 
species-specific biological risk factors, and environmental risk 
factors was not provided'' and that ``many of the quantitative aspects 
have not been substantiated.'' While NMFS disagrees with this statement 
and refers the reader to the EWG report (Southall et al., 2017), we 
also point out that, in the absence of precise quantitative information 
on these aspects of the risk assessment framework (on a species- and 
zone-specific basis), the application of the framework necessarily 
requires the application of professional judgment. As NMFS 
acknowledged, ``[e]lements of this approach are subjective and relative 
within the context of this program of projected actions and, overall, 
the analysis necessarily requires the application of professional 
judgment.'' (83 FR 29290.) The MMC comments do not find fault with any 
specific element or attribute of the framework or with any specific 
value chosen to represent a particular risk threshold or a particular 
species' vulnerability. NMFS does not agree that the MMC's 
recommendation to allow for an additional 30-day comment period for the 
public to review the risk assessment framework findings in advance of 
issuing the final rule is warranted and has not implemented the 
suggestion.
    Comment: NRDC asserts that NMFS has erroneously used the 
relativistic assessment presented in the EWG report as the basis for 
the negligible impact determination, incorrectly applying it as though 
it evaluated absolute risk. A private citizen offers similar comments.
    Response: NMFS disagrees with the comment. The EWG analysis is an 
important component of the negligible impact analysis, but is not the 
sole basis for our determination. While the EWG analysis 
comprehensively considered the spatial and temporal overlay of the 
activities and the marine mammals in the GOM, as well as the number of 
takes predicted by the described modeling, there are details about the 
nature of any ``take'' anticipated to result from these activities that 
were not considered directly in the EWG analysis and which warrant 
explicit consideration in the negligible impact analysis. Accordingly, 
NMFS' analysis considers the results of the EWG analysis, the effects 
of the required mitigation, and the nature and context of the takes 
that are predicted to occur. NMFS' analysis also explicitly considers 
the effects of predicted Level A harassment and impacts to marine 
mammal habitat, which were, respectively, not integrated into or 
included in the EWG risk ratings. These components of the full 
analysis, along with any germane species or stock-specific information, 
are integrated and summarized for each species or stock in the Species 
and Stock-specific Negligible Impact Analysis Summaries section of the 
negligible impact analysis.
    In addition, while the EWG framework comprehensively considers the 
aggregate impacts to marine mammal populations from the activities 
addressed in this rule in the context of both the severity of the 
impacts and the vulnerability of the affected species, it does not 
fully consider the absence of survey activity in the eastern GOM 
(within the GOMESA moratorium area), following BOEM's update to the 
scope of activity. While this is to some degree reflected in the 
updated take estimates, and thereby incorporated into the EWG 
framework's risk ratings, the absence of survey activities within areas 
of increased biological importance for certain species benefits those 
species GOM-wide beyond what is simply reflected in the updated take 
numbers. The negligible impact analysis considers the reduction of both 
acute and chronic effects afforded through the revised scope of the 
rule.
    Also, we note that while the EWG framework produces relativistic 
risk ratings, its components consist of absolute concepts, some of 
which are also absolutely quantified (e.g., whether the specified 
activity area contains greater than 30 percent of total region-wide 
estimated population, between 30 and 15 percent, between 15 and 5 
percent, or less than 5 percent). Further, NMFS provided substantive 
input into the scoring used in implementing the EWG framework for the 
GOM, to ensure that the categories associated with different scores, 
the scores themselves, and the weight of the scores within the overall 
risk rating all reflected meaningful biological, activity, or 
environmental distinctions that would appropriately inform the 
negligible impact analysis. Accordingly, and as intended, we used our 
understanding of the framework and best professional judgment to 
interpret the relativistic results of the EWG analysis appropriately 
into the larger negligible impact analysis, with the other factors 
discussed above, to make the necessary findings specific to the effects 
of the total taking on the affected species and stocks.
    Comment: NRDC asserts that the vulnerability ratings used in the 
EWG framework fail to account for several factors appropriately, which 
undermine the framework's ability to contribute accurately to the 
overall evaluation of relative risk. NRDC cites the following as 
problematic factors: Application of vulnerability ratings on a zone-by-
zone basis, which they state negatively biases the habitat use and 
temporal overlap factors; unaccountably low ratings for non-seismic 
stressors (specifically citing the DWH oil spill); relatedly, failure 
to account appropriately for all other stressors; and failure to fully 
account for stock structure and status.
    Response: NMFS first notes that the application of the EWG 
framework, and specifically the development of appropriate 
vulnerability ratings, necessarily involves the use of professional 
judgment, here on the part of a group of experts in the fields of 
marine mammal biology, ocean acoustics, and the effects of noise on 
marine mammals, among other things (and in consultation with NMFS and 
BOEM). Reasonable people may disagree about the specific numerical 
values assigned to any one of the 11 different factors contributing to 
the overall species-specific vulnerability score generated for each of 
the seven zones (with seven factors that are static GOM-wide and four 
that vary spatially, scoring for 18 taxa and seven zones means that 630 
individual numerical value selections underlie the vulnerability 
scores); but this does not imply that any of the specific values 
selected are unreasonable. All relevant factors were considered in 
generating the species- and zone-specific vulnerability scores.
    NRDC misapprehends one of the fundamental values of the analytical 
framework, in that it is structured in a spatially explicit way that 
can be applied at multiple scales, based on the scope of the action and 
the information available to inform an assessment of the risk 
associated with the activity (or suite

[[Page 5360]]

of activities). This allows one to generate overall risk ratings while 
also evaluating risk on finer scales. In this case, severity ratings 
were generated on the basis of seven different GOM zones, allowing an 
understanding not only of the relative scenario-specific risk across 
the entire GOM, as is demanded for this analysis, but also to better 
understand the particular zones where risk may be high (depending on 
actual future survey effort) and what part of the stock's range may be 
subject to relatively high risk. The framework recognizes, 
fundamentally, that the spatial, temporal, and spectral overlaps 
between noise-generating activities and animal distribution are the 
primary factors that drive the type, magnitude, and overall evaluated 
risk of potential noise effects on marine mammals. These considerations 
are inherent and fundamental in both the severity and vulnerability 
ratings and are deliberately integrated into both the vulnerability and 
severity assessments; in fact, key features of the analytical framework 
include explicit recognition of the importance of species distribution 
relative to activity spatial distribution and temporal and contextual 
differences in exposure scenarios. If the spatially explicit nature of 
the framework were removed, as it seems NRDC is suggesting, there would 
be no value in generating a ``habitat use'' factor (i.e., the spatial 
scale would be the GOM, and it would necessarily contain 100 percent of 
the estimated population). Spatial overlap is a central consideration 
for the extent of physical overlap between species and other 
environmental stressors, with consideration of species distribution 
across all zones, as well as the extent of population concentration and 
habitat specialization (as expressed through zone-specific 
vulnerability assessment). Regarding the temporal overlap factor 
referenced by NRDC, overall activity duration is a limited 
consideration within the vulnerability assessment rating but is 
expressed as a central consideration within magnitude-duration 
functions used to evaluate severity.
    Despite the explanations provided in the EWG report, NRDC 
characterizes certain aspects of the vulnerability scoring as 
``unaccountably low.'' However, NRDC does not provide specific 
recommendations for revisions to the assigned numerical values, or 
justification for their contention that scoring is too low. All 
relevant stressors were accounted for in the vulnerability scoring and 
specific scores were reasonably made on the basis of expert 
professional judgment. Contrary to NRDC's assertion, the effects of the 
DWH oil spill were considered in the vulnerability scoring (as well as 
in our development of mitigation in consideration of the MMPA's least 
practicable adverse impact standard). Overall, NRDC seems to provide a 
blanket suggestion, without adequate justification or evidence, that 
for all species, impacts should be considered to be higher than we have 
determined. We believe that we have satisfied the statutory standards 
after careful consideration of the available science.
    Regarding stock structure, NRDC criticizes the treatment of 
bottlenose dolphins in the vulnerability scoring. Overall, species-
level take and abundance estimates are used to support findings for 
bottlenose dolphins out of necessity. The best available information 
(Roberts et al., 2016) was used to inform combined species values and 
did not support further quantitative apportionment of estimated take or 
abundances to stocks. However, NRDC's specific criticism of the 
``population'' vulnerability scoring for bottlenose dolphins is 
unwarranted. The population score comprises three components: Status, 
i.e., is the stock listed under the ESA and/or designated as depleted 
under the MMPA; trend, i.e., does information over the available time 
series of abundance estimates indicate a trend; and size, i.e., is the 
population defined as small (less than 2,500). None of the five 
designated stocks of bottlenose dolphin in Federal waters of the GOM 
are listed under the ESA or designated as depleted under the MMPA, and 
none would be classed as small. Regarding trend, multiple SAR abundance 
estimates are available for three of the five stocks (oceanic stock and 
northern and western coastal stocks); and available information does 
show an increasing trend for these stocks. We recognize that the 
effects of the DWH oil spill included likely population reductions for 
all GOM marine mammal stocks (other than the eastern coastal stock of 
bottlenose dolphins, which was not impacted by the spill); however, the 
best available information indicates that these reductions were likely 
modest for all bottlenose dolphin stocks other than the northern 
coastal stock (Table 5), and no more recent population abundance 
estimates that might reflect any potential reduction are yet available. 
While the likely decline in population abundance for northern coastal 
bottlenose dolphins is subsumed within the population score assigned 
for bottlenose dolphins at the species level, vulnerability scoring is 
necessarily performed at the species level such that it may 
appropriately be integrated with the take-based severity scoring and 
used to generate an overall risk rating. As mentioned above, the best 
available scientific information does not allow for stock-specific 
parsing of take for bottlenose dolphins. Moreover, the trend component 
of the population score is a relatively small contribution to the 
overall vulnerability scoring, accounting for a maximum of two out of 
30 potential points. The likely decline in population abundance for 
northern coastal bottlenose dolphins, although not reflected in the 
existing vulnerability scoring, is insignificant as a contribution to 
the overall vulnerability score for bottlenose dolphins as a species. 
As noted above, the effects of the DWH oil spill are separately 
accounted for in the vulnerability scoring. Importantly, and also not 
accounted for in the EWG framework, we include significant mitigation 
(time-area restriction) intended to alleviate impacts to northern 
coastal bottlenose dolphins during periods of greatest importance for 
their reproductive behavior.
    Comment: NRDC states that NMFS' use of daily exposure durations 
``to justify its negligible impact determination'' is arbitrary and 
capricious. They state that we incorrectly used exposure times above 
the 160-dB threshold (rather than the lower threshold associated with 
the multi-step probabilistic risk function); assumed low severity for 
certain exposure durations; and disregarded repeated exposures. A 
private citizen offers similar comments.
    Response: As an initial matter, while it is true that NMFS 
evaluated exposure durations for the negligible impact analysis, it is 
not the only factor that we considered ``to justify'' the 
determination, as described fully in the Negligible Impact Analysis and 
Determinations section. Moreover, the consideration of exposure 
duration is entirely appropriate in assessing the severity of a likely 
exposure, which is critical to understanding how the authorized takes 
are likely to impact individual marine mammals. This was not addressed 
in the EWG assessment but was incorporated into the negligible impact 
analysis.
    NMFS appreciates NRDC's comments regarding use of exposure times 
above the 160-dB threshold, and we have re-evaluated the exposure 
duration information and better integrated discussion of this 
information into the negligible impact analysis (see Negligible Impact 
Analysis and Determinations and Table 16 for more

[[Page 5361]]

information). However, it is incorrect that ``NMFS' time-exposure 
analysis is predicated on its use of 160 dB as the operative threshold 
of harm'' and that our use of exposure information above the 160-dB 
threshold is a ``back-door return'' of the ``outdated 160 dB 
threshold.'' Inherent in the concept of a multi-step probabilistic risk 
function is the assumption that varying proportions of an exposed 
population will be harassed upon exposure at the different steps of the 
function. We presented the 160-dB exposure durations in the notice of 
proposed rulemaking because exposure above this step represents the 50 
percent midpoint of the function (for all species other than beaked 
whales) and, therefore, was deemed an appropriate representation of 
durations where a significant proportion of exposed animals would be 
expected to experience harassment (versus 10 percent of the population 
exposed to received sound levels between 140 and 160 dB). In Table 16 
of this final rule, we present these durations for both the 160-dB and 
140-dB steps of the function. It is important to keep in mind that, of 
the animals exposed above the 160-dB threshold for the indicated 
species-specific durations, not all are considered harassed. The risk 
function assumes 50 percent of animals exposed between 160-dB and 180-
dB will be harassed. For the longer exposure durations associated with 
the 140-dB threshold, only 10 percent are expected to be harassed.
    As we indicate in the Negligible Impact Analysis and Determinations 
discussion of this final rule, to put the predicted amount of take into 
meaningful context, it is useful to understand the duration of exposure 
at or above a given level of received sound (as well as the likely 
number of repeated exposures across days). While a momentary exposure 
above the criteria for Level B harassment counts as an instance of 
take, that accounting does not make any distinction between fleeting 
exposures and more severe encounters in which an animal may be exposed 
to that received level of sound for a longer period of time. This 
information is meaningful to an understanding of the likely severity of 
the exposure, which is relevant to the negligible impact evaluation. 
For example, for bottlenose dolphin exposed to noise from 3D WAZ 
surveys in Zone 6, the modeling report shows that approximately 72 
takes (Level B harassment) would be expected to occur in a 24-hr 
period. However, each animat modeled has a record or time history of 
received levels of sound over the course of the modeled 24-hr period. 
The 50th percentile of the cumulative distribution function indicates 
that the time spent exposed to levels of sound above 160 dB rms SPL 
(i.e., the 50 percent midpoint for Level B harassment) would be only 
1.8 minutes--a minimal amount of exposure carrying little potential for 
significant disruption of behavioral activity.
    The Species and Stock-specific Negligible Impact Analysis Summaries 
discussion considers the relative impact ratings in conjunction with 
required mitigation and other relevant contextual information--
including exposure durations at the various thresholds--to produce an 
assessment of impact to the stock or species, i.e., the negligible 
impact determinations. For beaked whales, take is estimated on the 
basis of a risk function shifted down such that 90 percent of the 
animals exposed to received levels above 140 dB and 50 percent exposed 
to received levels above 120 dB are expected to be harassed. We used 
this approach based on the documented behavioral sensitivity of beaked 
whales. However, as NRDC acknowledges, context is important when 
assessing behavioral responses to sound. The exposures above 120 dB 
here occur at significant distance from the source (i.e., greater than 
50 km). It is generally accepted that an animal's distance from the 
sound source plays an important role in the animal's behavioral 
response to a received sound level (e.g., Gomez et al., 2016). NMFS 
believes that exposures to the relevant harassment thresholds at 
significant modeled distances from the actual sound source, although 
included in the take estimates based on the risk function, will not 
carry significant consequences for the potentially exposed animals. 
Rather, these exposures are likely to result in significantly less 
severe responses (if any). Examples provided by NRDC purporting to 
demonstrate greater severity of response than we have assumed include 
irrelevant examples--beaked whales are known to respond with greater 
severity to mid-frequency active military sonar than to other sources, 
as discussed in greater detail in a previous comment response--and 
examples of ``responses'' entailing changes to vocalization patterns 
over longer durations, but these responses do not necessarily rise to 
the level of a take, much less a take event of significant severity.
    Regarding repeated exposures, despite the figures cited by NRDC 
concerning potential days of activity, it is unlikely that any given 
individual animal would in fact experience repeated take events of the 
magnitude suggested. Each of the seven GOM zones is an extremely large 
area (average zone size approximately 100,000 km\2\), and the likely 
harassment ``footprint'' of any given survey would be relatively small. 
Modeled isopleth distances to the 160-dB threshold are approximately 12 
km for low-frequency cetaceans (i.e., the Bryde's whale), 7 km for mid-
frequency cetaceans (i.e., sperm whales, beaked whales, dolphins), and 
6 km for high-frequency cetaceans (i.e., Kogia spp.). Distances to the 
140-dB isopleths are substantially larger, but we again emphasize that 
only ten percent of the animals exposed at that level would be expected 
to incur harassment, while 50 percent of the animals exposed at the 
160-dB level would be expected to incur harassment. It is clear that, 
in reality, there is a relatively low chance of any given individual 
marine mammal being repeatedly taken within relatively short 
timeframes, much less that such events would result in fitness 
consequences for those individuals. Additionally, NRDC suggests that 
NMFS fails to consider repeated takes at all, when in fact this 
likelihood is inherently addressed through the severity rating of the 
EWG assessment.
    NRDC concludes their comment by claiming that NMFS failed to 
undertake sufficient analysis in support of the negligible impact 
determinations. We disagree with this assertion, and refer to the 
Negligible Impact Analysis and Determinations section in support of 
this final rule. NRDC focuses in particular on sperm whales, implying 
that they are likely to incur impacts to reproductive fitness and 
stating that NMFS cannot make a negligible impact finding for sperm 
whales without additional mitigation requirements. NMFS agrees that the 
bioenergetics simulations of Farmer et al. (2018a)--cited by NRDC in 
support of their argument--show that frequent disruptions in foraging 
can have potentially severe fitness consequences for individual sperm 
whales. However, a follow-up study (Farmer et al., 2018b), which 
additionally accounted for the population-level effects of the DWH oil 
spill on GOM sperm whales, modeled the potential population level 
consequences of the specific disturbance events underlying this 
analysis (i.e., the acoustic exposure modeling of Zeddies et al., 2015, 
2017a). This follow-up study found that, under realistic modeled 
scenarios, no sperm whales were projected to reach terminal starvation 
and no fetal abortions were predicted as a result of long-term 
disturbance effects (i.e., over ten years of projected survey 
activity). Similarly,

[[Page 5362]]

predicted declines in relative body condition (expressed as the 
percentage of available reserves for a disturbed individual whale 
relative to an undisturbed whale with identical characteristics) as a 
result of long-term disturbance effects were not significant under 
realistic modeled scenarios. When evaluating the additional effects of 
modeled disturbance on the DWH oil spill-impacted trajectory, the 
modeling did not predict any significant additional stock declines 
(Farmer et al., 2018b). We believe the administrative record for this 
final rule amply demonstrates that NMFS used the best available science 
during our administrative process to inform our analyses and satisfy 
the standards under section 101(a)(5)(A). Of note, and as indicated in 
Changes from the Proposed Rule, as a result of BOEM's updated scope of 
the activities and the associated revisions to the levels of effort, 
both the maximum allowable amount of take and the maximum annual take 
under the rule have decreased (significantly in some cases, including 
for Bryde's whales and sperm whales) for all except two species/stocks. 
For the two exceptions these figures increased only slightly, and the 
severity of many of the impacts has been lessened via the removal and/
or reduction of take in areas of greater biological importance 
previously considered as mitigation areas.
    Comment: Chevron comments that NMFS should make the final version 
of the EWG report available to the public for review and suggests 
expanding the description of the inputs of the analysis. Chevron states 
that the ``vulnerability'' assessment, in particular, would benefit 
from additional discussion to explain how professional judgments led to 
specific rankings for each species. Chevron also comments that NMFS 
should provide an additional plain language discussion of the risk 
analysis process, including background on the development of the risk 
analysis framework, including any relevant analogues in other 
ecosystems or regulatory contexts, the ways in which species may be 
considered ``vulnerable,'' and the meaning of the ``risk'' discussed.
    Response: NMFS appreciates the comment. The final report is 
available to the public online at: www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico. The content of the final report was determined by 
NMFS and BOEM in conjunction with the EWG. We believe that we have 
provided sufficient plain language discussion of the EWG framework.
    Comment: NRDC claims that NMFS' negligible impact analysis is 
inappropriately reliant upon the prescribed mitigation and, further, 
that the mitigation will be ineffective.
    Response: First, NMFS did not rely solely on the mitigation in 
order to reach its findings under the negligible impact standard. As is 
stated in the analysis, consideration of the implementation of 
prescribed mitigation is one factor in the analysis but is not 
determinative in any case. In certain circumstances, mitigation is more 
important in reaching the negligible impact determination, e.g., when 
mitigation helps to alleviate the likely significance of taking by 
avoiding or reducing impacts in important areas. Second, while NRDC 
dismisses the importance of the prescribed mitigation by stating 
(mistakenly) that it is ``unsupported by evidence,'' NRDC offers no 
support for their conclusions.
    NRDC misunderstands the degree to which NMFS relies on shutdowns 
for sensitive or vulnerable species, including beaked whales, at 
extended distances. We agree that these measures in and of themselves 
will have limited benefit for cryptic species such as beaked whales 
that are unlikely to be observed. However, we believe that it makes 
sense to minimize the duration and intensity of disturbance for these 
species when they are observed, and because they are practicable we 
include them in the suite of prescribed measures and discuss them where 
appropriate. For more readily detected species, such as the sperm 
whale, which is easily detected when at the surface and which vocalizes 
frequently while underwater, the extended distance shutdowns (for both 
visual and acoustic detections) should appropriately be considered 
influential in our assessment of impacts to affected individuals and, 
therefore, ultimately on the stock. Despite NRDC's dismissal of these 
requirements, we presume they would agree that the duration and 
intensity of disturbance of sensitive species should be minimized where 
practicable.
    In summary, we consider these measures appropriately as mitigating 
factors when considering context as part of our negligible impact 
analysis.
    Comment: The Associations state that the Expert Working Group 
framework was applied without following all of the recommended steps, 
such as conducting expert elicitation to derive risk functions for 
species that do not have parameterized Population Consequences of 
Disturbance (PCOD) models. The Associations recommend that NMFS seek 
input and advice on the framework and its conclusions from independent 
experts.
    Response: There is extensive scientific interest in forecasting how 
short-term behavioral responses by individual animals may aggregate and 
result in population-level consequences. The concept was introduced by 
the National Research Council (2005) as Population Consequences of 
Acoustic Disturbance. However, given the lack of data on acoustic 
responses, research studies have generalized the issue to look at 
environmental and anthropogenic stressors in general and renamed the 
concept Population Consequences of Disturbance. New et al. (2014) 
presented a modified conceptual framework to help forecast long-term 
impacts. Conceptually, a series of transfer functions connect 
increasingly broader impacts from the initial disturbance to effects on 
individual health, individual vital rates, and finally population 
dynamics. The concept has been demonstrated with a few species for 
which there are extensive data from tagged or photo-identified animals 
so that effects on individuals can be quantified. Northern elephant 
seals were the first study species for which the data from time-depth 
recorders were able to be linked to an individual animal's body fat 
condition (Aoki et al., 2011; Adachi et al., 2014), which provided 
insight into foraging success and ultimately individual health and 
vital rates (Robinson et al., 2010). Rolland et al. (2016) used 
photographic data of North Atlantic right whales to evaluate individual 
health and link it to demographic groups and population status. 
Additional studies exploring population consequences are ongoing, but a 
common theme is that extensive data documenting individual health and 
population vital rates are necessary for such analyses. These are 
considered the gold standards for future studies, but, at present, 
studies within the GOM have not occurred in sufficient detail for such 
analyses.
    For purposes of the analysis contained herein, the disturbance 
severity rating facet of the EWG framework involves a relativistic 
framework relating Level B harassment to the zone-specific population 
size and then evaluating this proportion to specified severity criteria 
common across species. In the idealized framework discussed by the EWG 
(Southall et al., 2017), the severity rating involves consideration of 
the magnitude of population affected and the duration of disturbance, 
i.e., by deriving magnitude-duration risk functions that describe the 
potential effects of

[[Page 5363]]

exposure to noise on affected populations. The EWG considered that a 
better approach would apply values obtained using software developed to 
implement the Interim PCOD approach (Harwood et al., 2014; King et al., 
2015).
    While various models have been developed implementing the PCOD 
approach (e.g., New et al., 2013), the approach is problematic for 
general application because it is very data-heavy, and sufficient data 
specific to a taxon and/or disturbance context is not typically 
available. Few marine mammal populations have been as intensively 
studied as the PCOD case study populations, and the lack of appropriate 
datasets that link exposure to disturbance with behavioral change, and 
behavioral change with health, currently limits the general 
applicability of the full PCOD model. This difficulty led to 
development of the Interim PCOD approach, which uses results from an 
expert elicitation process, rather than empirical data, to predict the 
effects that a specific amount of disturbance will have on the vital 
rates of an individual marine mammal. In evaluating potential use of 
the Interim PCOD approach for developing magnitude-duration curves 
suitable for use in assessing risk associated with the projected survey 
activity considered here, the EWG used the results of an expert 
elicitation process that considered potential effects of pile driving 
noise associated with the construction of offshore wind farms on 
bottlenose dolphins, harbor porpoises, and minke whales in the North 
Sea. While this evaluation provided proof-of-concept and highlighted 
areas for future improvement of the process, such evaluations are not 
appropriately extrapolated to a risk assessment involving dissimilar 
species, stressors, and locations. For example, demographic rates and 
population growth rates specific to those species in U.K. waters of the 
North Sea were used and, further, even in that expert elicitation the 
authors warned that the results for the minke whale were likely not 
reliable due to a lack of available data. The EWG recommended that the 
available elicitation results not be used towards the current analysis, 
and NMFS and BOEM concurred. Currently, results of these expert 
elicitation processes are additionally viewed as potentially unreliable 
because experts may misinterpret the questions they are asked (Booth et 
al., 2016).
    Overall, while we agree with the Associations that it would be 
ideal to evaluate the effects of the specified activity on the affected 
populations by incorporating a PCOD or Interim PCOD approach to the EWG 
framework, sufficient data are not available to conduct a PCOD 
approach, and sufficient resources were not available to NMFS to 
develop and implement an expert elicitation process specific to seismic 
and the affected GOM populations on a timeline amenable to this ITR. 
With regard to the Associations' suggestion that outside experts review 
the EWG framework, we note that the EWG comprises experts outside NMFS 
and BOEM who were contracted for the express purpose of developing the 
framework. We do not believe it necessary to engage outside experts to 
review the work of other experts outside NMFS and BOEM, which is itself 
subject to review by experts within both NMFS and BOEM.
    Comment: The Associations object to the terminology used for the 
relative severity ratings in the EWG framework approach, stating their 
disagreement with the implications of rating descriptors such as 
``severe,'' and reiterating their belief that the modeled exposure 
levels are incompatible with the available data. Relatedly, the 
Associations assert that there is ``little scientific support'' for the 
relative risk ratings for sperm and beaked whales.
    Response: Respectfully, NMFS believes this comment involves a 
semantic issue. The Associations do not suggest alternative terminology 
for the relative risk ratings. Regarding the risk ratings for sperm 
whales and beaked whales, these ratings are a product of a relatively 
straightforward analysis of severity (i.e., amount of predicted 
disturbance relative to population size) and vulnerability (i.e., 
consideration of factors inherent to the population that make it more 
or less vulnerable to the disturbance considered via the severity 
rating). The Associations provide no specific critique of any of these 
aspects of the analysis. We have addressed the Associations' criticism 
of the acoustic exposure modeling elsewhere in these comment responses.
    Comment: The Associations object to use of the potential biological 
removal (PBR) metric as the basis for evaluating severity of Level A 
harassment within the EWG framework, stating that its use in evaluating 
non-serious injury is inappropriate because the metric was developed 
for evaluation of the significance of serious injury and mortality.
    Response: We acknowledge that the PBR metric defines a level of 
removals from a population (i.e., mortality) that would allow that 
population to remain at its optimum sustainable population level or, if 
depleted, would not increase the population's time to recovery by more 
than 10 percent, and therefore that it is inappropriate to make 
comparisons between Level B harassment takes and the PBR value for any 
stock. However, as discussed in the EWG report and in the notice of 
proposed rulemaking, while NMFS does not expect PTS (Level A 
harassment) that might be accrued through noise exposure to result in 
mortality of marine mammals, PBR can serve as a good surrogate for 
population vulnerability/health. Accordingly, PBR or a related metric 
can be used appropriately as a value against which to evaluate the 
potential severity to the population of a permanent impact such as PTS 
on a given number of individuals, and it is only in this sense that we 
use the PBR value. The Associations do not provide an alternative 
recommendation.

Small Numbers

    Comment: The Associations and other industry commenters express 
agreement with NMFS' interpretation of the small numbers requirement as 
allowing that the finding may be made at the individual LOA level.
    Response: We thank the Associations for their comment in support of 
the small numbers approach. NMFS' analysis generally comports with many 
of the points they raise, as discussed in this preamble.
    Comment: NRDC states that the interpretation of ``small numbers'' 
presented by NMFS in the notice of proposed rulemaking is contrary to 
the plain meaning and purpose of the MMPA, in part because NMFS 
allegedly did not provide a reasoned basis for the take limit proposed 
(i.e., one-third of the best available species or stock abundance 
estimate). NRDC makes four specific claims. First, NRDC states that 
one-third cannot be considered a ``small number.'' Second, NRDC states 
that Congress intended that takes be limited to ``infrequent, 
unavoidable'' occurrences, and that NMFS has not explained why the 
taking would be infrequent or unavoidable. Third, NRDC contends that 
NMFS should define different small numbers thresholds on the basis of 
the conservation status of individual species. Finally, NRDC believes 
that NMFS must account for ``additive and adverse synergistic effects'' 
that may occur due to multiple concurrent surveys in conducting a small 
numbers analysis. Industry commenters suggest that additional detail is 
necessary regarding the basis for NMFS' small numbers threshold.
    Response: NMFS disagrees with NRDC's arguments on this topic. 
Although there is limited legislative history available to guide NMFS 
and an apparent lack of biological

[[Page 5364]]

underpinning to the concept, we have worked to develop a reasoned 
approach to small numbers. As discussed in the section of the notice of 
proposed rulemaking entitled Small Numbers, NMFS explains the concept 
of ``small numbers'' in recognition that there could also be quantities 
of individuals taken that would correspond with ``medium'' and 
``large'' numbers. As such, NMFS has established that one-third of the 
most appropriate population abundance number--as compared with the 
assumed number of individuals taken--is an appropriate limit with 
regard to ``small numbers.'' This relative approach is consistent with 
the statement from the legislative history that ``[small numbers] is 
not capable of being expressed in absolute numerical limits'' (H.R. 
Rep. No. 97-228, at 19 (September 16, 1981)), and relevant case law 
(Center for Biological Diversity v. Salazar, 695 F.3d 893, 907 (9th 
Cir. 2012) (holding that the U.S. Fish and Wildlife Service reasonably 
interpreted ``small numbers'' by analyzing take in relative or 
proportional terms)).
    NRDC claims that a number may be considered small only if it is 
``little or close to zero'' or ``limited in degree.'' We note that the 
comment selectively picks a definition in support of NRDC's favored 
position. For example, the definition of ``small'' in Webster's New 
Collegiate Dictionary (1981) included ``having little size, esp. as 
compared with other similar things.'' See also www.merriam-webster.com/dictionary/small (defining ``small'' as ``having comparatively little 
size''). These definitions comport with the small numbers 
interpretation developed by NMFS, which utilizes a proportionality 
approach. The comment also selectively quotes the relevant legislative 
history language, stating that Congress ``intended that the agency 
limit takes to `infrequent, unavoidable' occurrences.'' The actual 
statement from the legislative history is that taking of marine mammals 
should be ``infrequent, unavoidable, or accidental.'' H.R. Rep. No. 97-
228, at 19 (September 16, 1981) (emphasis added). This language 
suggests that taking that is unavoidable (or accidental) may qualify as 
small numbers, even if not infrequent.
    The argument to establish a small numbers threshold on the basis of 
stock-specific context is unnecessarily duplicative of the required 
negligible impact finding, in which relevant biological and contextual 
factors are considered in conjunction with the amount of take. 
Similarly, NRDC's assertion that NMFS' proposed approach fails to 
account for ``additive and adverse synergistic effects'' from multiple 
surveys is not required by section 101(a)(5)(A) of the MMPA, and it is 
unclear how NRDC defines this concept or how it may be related to the 
``small numbers'' concept. These suggestions are not founded in any 
relevant requirement of statute or regulation, discussed in relevant 
legislative history, or supported by relevant case law.
    A private citizen echoed certain of NRDC's comments on this topic, 
adding that NMFS' approach is ``embarrassing and scientifically 
indefensible.'' However, the commenter does not provide a more 
scientifically defensible interpretation of small numbers, suggesting 
only that ``[o]ne could approach this in many ways.''
    Regarding the comment that additional explanation is needed for 
NMFS' interpretation of the small numbers standard, we believe the 
proposed and final rule provide sufficient explanation for setting one-
third as the upper limit for small numbers where reasonably reliable 
quantified take estimates are available. See the Small Numbers section 
later in this preamble.
    Comment: Several commenters suggest that the small numbers finding 
need not be based on a quantitative threshold.
    Response: NMFS agrees that a more qualitative small numbers finding 
may be permissible. See, e.g., Center for Biological Diversity v. 
Salazar, 695 F.3d at 906-908. However, in this case, where take 
estimates can be predicted with relative confidence, we have elected to 
set a quantitative threshold. Moreover, the commenters do not provide 
any specific recommendations for an appropriate qualitative approach in 
this case.
    Comment: The MMC recommended that any ``formal interpretation'' of 
the small numbers standard by NMFS be issued in a stand-alone, 
generally applicable rulemaking (e.g., in amendments to 50 CFR 216.103 
or 216.105) or in a separate policy directive, rather than in the 
preambles to individual proposed rules.
    Response: We appreciate the MMC's recommendation and may consider 
the recommended approaches in the future. We note, however, that 
providing relevant explanations in a proposed ITR is an effective and 
efficient way to provide information to the reader and solicit focused 
input from the public, and ultimately affords the same opportunities 
for public comment as a stand-alone rulemaking would.
    Comment: NRDC asserts that NMFS' interpretation of the MMPA's small 
numbers requirement is contrary to law, stating their belief that NMFS 
must make a small numbers determination in the rule, rather than for 
issuance of individual LOAs; that NMFS must evaluate the same amount of 
take in order to separately determine that the total take will both 
meet the small numbers standard and have a negligible impact; and that 
NMFS' approach impermissibly cuts the public out of the agency's 
findings.
    Response: Based on NMFS' analysis of the language and structure of 
section 101(a)(5)(A) and the implementing regulations for that 
provision, NMFS disagrees that the small numbers finding must be based 
on the total of all take over the five-year (or less) period from all 
potential survey activity. The MMPA does not define small numbers or 
explain how to apply the term in either section 101(a)(5)(A) or the 
similar provision for incidental harassment authorizations (IHAs) in 
section 101(a)(5)(D),\6\ including how to apply the term in a way that 
allows for consistency across those two provisions that are similar but 
allow for potentially different time and activity scales. (See Small 
Numbers below.) Especially when taken together with NMFS' implementing 
regulations, our approach is consistent with the structure of section 
101(a)(5)(A), which provides:
---------------------------------------------------------------------------

    \6\ Section 101(a)(5)(D) states in relevant part:
    (i) Upon request therefor by citizens of the United States who 
engage in a specified activity (other than commercial fishing) 
within a specific geographic region, the Secretary shall authorize, 
for periods of not more than 1 year, subject to such conditions as 
the Secretary may specify, the incidental, but not intentional, 
taking by harassment of small numbers of marine mammals of a species 
or population stock by such citizens while engaging in that activity 
within that region if the Secretary finds that such harassment 
during each period concerned--
    (I) will have a negligible impact on such species or stock, and
    (II) will not have an unmitigable adverse impact on the 
availability of such species or stock for taking for subsistence 
uses[.]
---------------------------------------------------------------------------

    (i) Upon request therefor by citizens of the United States who 
engage in a specified activity (other than commercial fishing) within a 
specified geographical region, the Secretary shall allow, during 
periods of not more than five consecutive years each, the incidental, 
but not intentional, taking by citizens while engaging in that activity 
within that region of small numbers of marine mammals of a species or 
population stock if the Secretary, after notice (in the Federal 
Register and in newspapers of general circulation, and through 
appropriate electronic media, in the coastal areas that may be affected 
by such activity) and opportunity for public comment--
    (I) finds that the total of such taking during each five-year (or 
less) period

[[Page 5365]]

concerned will have a negligible impact on such species or stock and 
will not have an unmitigable adverse impact on the availability of such 
species or stock for taking for subsistence uses [. . .]. (emphasis 
added).
    Section 101(a)(5)(A)(i)(I) is explicit that the ``negligible 
impact'' determination for a specified activity must take into account 
the ``total of such taking'' (i.e., all of the taking that the 
Secretary may conceivably allow (or authorize) under individual LOAs 
during the five year (or less) period considered for the rule). In 
contrast, the ``small numbers'' language in 101(a)(5)(A) is not subject 
to the same time period requirement of five years (or less in cases 
where the period being considered for a rule is less than five years).
    In our view, the statutory language for small numbers and the 
negligible impact finding indicates that the negligible impact finding 
is made based on consideration of an aggregation of potential 
authorizations (LOAs) for taking small numbers of marine mammals, and 
allows for different temporal periods in applying the two different 
standards. The statute contemplates that the Secretary shall allow 
taking during the five year (or less) period, which in our view also 
implies that there could be multiple allowances or authorizations 
(i.e., LOAs), so long as the maximum allowable total taking from all of 
those authorizations combined is considered in the upfront assessment 
of whether the negligible impact standard is met.
    As we have noted, the regulatory vehicle for authorizing (i.e., 
allowing) the take of marine mammals is the LOA, a creature of NMFS' 
long-standing implementing regulations that is not in the statute. See 
50 CFR 216.106. Those 1989 implementing regulations requiring an LOA to 
effectuate an authorization were in effect when Congress amended the 
MMPA in 1994 to add section 101(a)(5)(D) for issuance of one-year IHAs, 
and over the years when Congress amended section 101(a)(5)(A) for 
various reasons (including most recently in 2018, to extend the maximum 
authorization period to seven years for military readiness activities, 
Pub. L. 115-232 (John S. McCain National Defense Authorization Act for 
Fiscal Year 2019) (Aug. 13, 2018)). Presumably Congress was aware of 
these implementing regulations and the framework they created for 
authorizing take under section 101(a)(5)(A) and could have invalidated 
those regulations had it so desired.
    Under NMFS' approach, the negligible impact analysis for the 
rulemaking is conducted for the time period covered by the rule (five 
years in this case, the maximum under the statute for a non-military 
readiness activity), but the small numbers analysis attaches to the 
instrument that actually ``allows'' or authorizes taking, i.e., the 
LOA. The statute does not preclude NMFS from issuing an LOA that 
comports with the small numbers level set forth in the relevant rule 
for the specified activity. Consistent with the MMPA requirement, here 
the Secretary (through NMFS) has prescribed the necessary specified 
activity regulations after notice and comment. At that point, once the 
regulations are effective, NMFS thereafter may authorize incidental 
take through the issuance of LOAs, provided that they satisfy the 
requirements set forth in the rule and regulations, including the small 
numbers standard articulated in the rule.
    NRDC cites Conservation Council for Hawaii v. NMFS, 97 F. Supp. 3d 
1210 (D. Hawaii 2015), in stating that the MMPA ``plainly requires that 
the agency evaluate both whether there will be small numbers of take 
and whether there will be a negligible impact'' before issuing 
regulations, and that these determinations ``must be based on the same 
amount of take.'' We disagree. In NMFS' view, Conservation Council for 
Hawaii stands for the proposition that NMFS cannot authorize more take 
than it has analyzed under the negligible impact standard. 97 F. Supp. 
3d at 1221. There the court found that there were substantial 
differences between the anticipated take numbers, which were the basis 
for the negligible impact finding, and the amount of take that NMFS was 
prepared to authorize incidental to U.S. Navy military readiness 
activities. That case did not even involve the small numbers provision, 
which does not apply in the case of military readiness activities. 16 
U.S.C. 1371(a)(5)(F)(i). The court in Conservation Council for Hawaii 
did not consider or make any pronouncements about whether the small 
numbers provision must be applied to the total annual taking under the 
rule or whether it could be applied at the LOA stage.
    NRDC repeatedly states that the negligible impact and small numbers 
provisions must have separate meaning. NMFS agrees that the two 
provisions do have separate meanings, and this rule satisfies that 
requirement. Each LOA must meet the small numbers requirement as NMFS 
has interpreted it in this rule. In other words, it is not sufficient 
for the survey activity described in an LOA application to fall within 
the scope of the activity analyzed for the rule and NMFS' negligible 
impact determination. The small numbers limitation also must be 
satisfied. For example, NMFS may receive an application for an LOA 
where the take estimates exceed the small numbers standard identified 
in the rule. In that case, the request would be denied, even if the 
amount of taking was considered in the negligible impact evaluation. 
Thus the negligible impact and small numbers inquiries are separate and 
have different meanings.
    To summarize, the MMPA is silent on how to apply ``small numbers'' 
in either section 101(a)(5)(A) or (D), including in a way that allows 
for consistency across those two very similar provisions. Moreover, 
NMFS' implementing regulations for section 101(a)(5)(A) make it clear 
that LOAs are the instrument for authorizing take. Thus, the mere 
existence of regulations under 101(a)(5)(A) for a specified activity is 
not sufficient to authorize take under that provision. An LOA is 
required.
    As we have previously stated, the small numbers standard has 
limited biological relevance (i.e., there is a lack of a biological 
underpinning for the concept), but NMFS' application of the small 
numbers standard at the LOA stage does not rely on that view for the 
approach taken here (and moreover, NMFS did not receive any public 
comments offering an alternative definition that is rooted in 
biological concepts or is not conflated with negligible impact 
considerations). As the notice of proposed rulemaking explained, NMFS' 
interpretation and approach are based on analysis of the governing 
section 101(a)(5)(A) and limited legislative history, as well as 
consideration of section 101(a)(5)(D), and our long-standing approach 
to implementing section 101(a)(5)(A) through separate LOAs. NMFS has 
determined that the statute is ambiguous in terms of what small numbers 
means and how ``small numbers'' must be applied, which affords the 
agency reasonable discretion in how to do so. After weighing various 
policy considerations, NMFS exercised its discretion to define small 
numbers and apply small numbers determinations at the LOA level.
    Importantly, the final rule, which was subject to notice and 
comment, sets the small numbers standard for future LOAs issued under 
the rule. Moreover, contrary to NRDC's assertions, NMFS has set the 
total taking allowable for all LOAs issued under the rule for this 
specified activity--i.e., the taking that was analyzed for the 
negligible impact determination. If an LOA application for

[[Page 5366]]

a survey provides take estimates that are within the small numbers 
threshold set in this rule, then the LOA for that survey will be deemed 
to satisfy the small numbers requirement.
    As NRDC correctly points out, NMFS' implementing regulations 
require issuance of LOAs to be consistent with the ``total taking 
allowable'' under the activity-specific regulations. The regulations 
for the specified activity also reflect this. The rulemaking for these 
regulations evaluated the level of activity projected in BOEM's update 
for its petition, and NMFS' negligible impact determination is based on 
consideration of that level (as are the corresponding take estimates). 
Any LOA must be within the amount analyzed for the scope of the rule, 
and the total amount of take under all issued LOAs combined cannot 
exceed the amount analyzed and ``allowable'' under the rule for this 
activity.
    Regarding the differences between the processes under sections 
101(a)(5)(A) and (D), we did not mean to suggest that section 
101(a)(5)(A) is necessarily or always more protective than and 
preferable to 101(a)(5)(D). Rather, section 101(a)(5)(A), which can 
span a longer period of time and cover multiple applicants through 
issuance of LOAs, allows for a more comprehensive/holistic analysis by 
the agency (one negligible impact analysis for all activities over the 
five-year (or less) period and consideration of mitigation appropriate 
for the full suite of activities). Such an approach has the potential 
to be more protective because it allows for a more comprehensive 
understanding of impacts, as well as a mechanism to include holistic 
mitigation that can more effectively address both acute and chronic 
effects resulting from multiple activities covered under a rule. 
Section 101(a)(5)(A) also focuses public attention on one rulemaking 
(rather than--as would be the case for these survey activities--
potentially dozens of IHA actions per year, each with separate notice 
and comment), and allows for other administrative efficiencies. We note 
that BOEM applied for the regulations in support of the oil and gas 
industry, and prepared an EIS in support of its own program related to 
the permitting of the survey activities that are the subject of this 
MMPA application and rulemaking.
    NRDC claims that the approach ``is a novel interpretation of the 
MMPA.'' However, the rule cited in support of their argument (81 FR 
47240; July 20, 2016) is consistent with one aspect of our approach 
here, in that the small numbers determinations in both contexts are 
based on annual take estimates, not total take over the five-year 
period of the regulations.\7\ We acknowledge that we have not 
previously determined that small numbers could be applied at the 
individual LOA stage where more than one LOA applicant may apply under 
the activity-specific regulations. However, that is simply because the 
issue had not previously presented itself. In nearly all cases to 
date,\8\ there has been a single operator who is the sole applicant for 
both the LOA (or LOAs if they cover less than the five-year period) and 
the governing specified activity regulations. As a result, in such a 
scenario, the small numbers determination by default corresponds to the 
maximum annual taking covered by the regulation (and the LOA). But even 
when there is only one applicant for LOAs under a regulation, NMFS does 
not tally take across the five-year period for purposes of assessing 
small numbers. Rather, NMFS assesses annual levels of take. (This also 
promotes consistency between 101(a)(5)(A) and 101(a)(5)(D) to avoid 
incentivizing IHAs at the expense of LOAs issued under more 
comprehensive rules.)
---------------------------------------------------------------------------

    \7\ We further note that population biology often focuses on 
annual cycles. See, e.g., 50 CFR 216.103 (negligible impact defined 
in terms of impacts on annual rates of recruitment or survival); 16 
U.S.C. 1386(a), (c) (requiring stock assessment reports to estimate 
the annual human-caused mortality and serious injury of the stock, 
and annual review of stock assessments when significant new 
information is available that may indicate the stock assessment 
should be revised); 16 U.S.C. 1362(26) (defining ``net productivity 
rate'' as the annual per capita rate of increase in a stock 
resulting from additions due to reproduction, less losses due to 
mortality); 16 U.S.C. 1383a(l)(ii) (requiring MMC's recommended 
guidelines to govern the incidental taking of marine mammals in the 
course of commercial fishing operations, to the maximum extent 
practicable, to include as a factor to be considered and utilized in 
determining permissible levels of taking ``the abundance and annual 
net recruitment of such stocks'').
    \8\ The one exception to date is NMFS' regulation governing the 
incidental take from explosive removal of offshore structures in the 
GOM (EROS), promulgated at the request of the Minerals Management 
Service on behalf of multiple private removal companies that 
individually submitted LOA requests. NMFS' rulemaking for the EROS 
regulations evaluated the estimated annual take based on MMS' 
projections for the specified activity as a whole, i.e., for all 
operators combined. Our rule here is consistent with the EROS 
rulemaking as it relates to the approach for the negligible impact 
evaluation. However, the EROS rule also concluded the total annual 
taking (by species) for all operators combined met the small numbers 
requirement. Thus NMFS did not have occasion before now to consider 
whether it could apply the small numbers provision at an individual 
LOA level where there are multiple concurrent LOA holders. Having 
now considered the question, NMFS believes the MMPA affords the 
discretion to do so.
---------------------------------------------------------------------------

    Finally, NRDC's statement that the public is impermissibly cut out 
of the agency's findings is incorrect. The proposed rule set forth the 
maximum total taking and annual taking that would be allowable (via the 
issuance of LOAs) for the five-year period that the regulations will be 
effective, which was based on information contained in BOEM's publicly 
available application and PEIS. Those figures decreased for all but two 
species. For the two species where the figures increased, we evaluated 
those changes and determined they do not represent a meaningful change 
for our analyses. See Changes From the Proposed Rule.
    The proposed rule included a 60-day public comment period. We also 
believe that our rulemaking afforded a full and focused opportunity for 
public review of and comment on the full scope of survey activities and 
proposed mitigation, rather than through dozens of individual IHAs, 
each with 30-day public comment periods and shorter timeframes for NMFS 
to consider the public comments. Thus the public had a meaningful 
opportunity to comment.
    Comment: Citing their interpretation of the statute and multiple 
judicial decisions, the MMC suggests that NMFS' interpretation and 
implementation of the small numbers standard is contrary to law and 
further recommends that NMFS adopt a policy interpreting the small 
numbers requirement of section 101(a)(5)(A) such that it:
     Requires determinations be made when issuing incidental 
take regulations (as opposed to when LOAs are issued);
     makes such determinations based on the total take 
authorized incidental to the specified activity and for the full 
duration covered by those regulations (as opposed to for each LOA and 
on an annual basis); and
     provides an opportunity for public notice and comment on 
all small numbers determinations.
    Response: As explained in the responses above and discussion under 
the Small Numbers section of this preamble, NMFS disagrees, based on 
our analysis of the statute, the legislative history, the implementing 
regulations, and relevant case law.
    NMFS issues incidental take authorizations under section 
101(a)(5)(A) through LOAs, provided that we satisfy the relevant 
statutory standards. Analysis of that statutory provision and relevant 
legislative history, including when read in conjunction with section 
101(a)(5)(D), leads NMFS to conclude that the small numbers limitation 
may be applied at the LOA stage, provided that we make the negligible 
impact finding for the total taking allowable under the regulations for 
the specified activity and

[[Page 5367]]

set the small numbers standard for future LOAs in the notice and 
comment rulemaking.
    As noted above, the term ``small numbers'' is not defined in the 
statute. Over the years NMFS has grappled with how to define the term, 
particularly given the limited legislative history (i.e., ``accidental, 
infrequent, or unavoidable''; ``not capable of being expressed in 
absolute numerical terms''). Recent court decisions lend support for 
NMFS' proportional approach to the concept. See Center for Biological 
Diversity v. Salazar, 695 F.3d 893 (9th Cir. 2012). In terms of what 
proportion may constitute ``small numbers'' for purposes of what the 
Secretary may authorize, NMFS has determined that small numbers means 
up to one-third of a species or stock. NMFS has further determined that 
this limit can be applied at the LOA level, subject to a finding that 
the total taking allowable (through any and all LOAs issued under the 
activity-specific rule and corresponding regulations) satisfies the 
negligible impact standard.
    The MMC inaccurately states that the ``interpretation of the small 
numbers requirement proposed by NMFS in many ways seeks to maximize the 
numbers of takes of marine mammals that may be authorized under a 
single rulemaking.'' With one exception, the points raised by the MMC 
reflect NMFS' existing practice. The decision to make small numbers 
findings on an LOA-specific basis is the only new development and, as 
explained in the response to the previous comment, came about only when 
the issue arose for the first time in the context of this rulemaking. 
NMFS considered the specific issue, determined that section 
101(a)(5)(A) does not unambiguously speak to it, and reasonably 
exercised its discretion in determining that small numbers findings 
could apply at the LOA stage, provided that the standard is set forth 
in the rule itself, which it is.
    We acknowledge that section 101(a)(5)(A) does not expressly 
contemplate the issuance of LOAs, which are a creature of NMFS and U.S. 
Fish and Wildlife Service (FWS) joint implementing regulations for 
section 101(a)(5)(A). (See 50 CFR 216 subpart I (NMFS regulations); 50 
CFR 18.27 (FWS regulations)). Those implementing regulations, in effect 
since 1989, established LOAs as the regulatory instrument to authorize 
lawful incidental take under section 101(a)(5)(A), after the 
promulgation of activity-specific regulations that undergo notice and 
comment rulemaking.
    Although not the typical scenario, NMFS' implementing regulations 
allow for the issuance of LOAs to more than one ``U.S. citizen'' taking 
marine mammals under a specified activity regulation, see, e.g., 50 CFR 
216.105(a); 216.106(e); (54 FR 40338 (September 29, 1989)), provided 
that the negligible impact finding is made for the total taking for the 
specified activity as a whole, by all entities conducting that 
activity.
    NMFS also administers section 101(a)(5)(D), a very similar 
provision enacted in 1994 that established an expedited process for the 
issuance of one-year incidental take authorizations for the taking of 
small numbers of marine mammals by harassment only when the taking from 
the specified activity is found to have a negligible impact on the 
affected species or stocks of marine mammals (referred to as incidental 
harassment authorizations, or ``IHAs''). See the Small Numbers section 
later in this Notice. The small numbers standard in section 
101(a)(5)(D) applies to each individual one-year IHA, yet the same 
small numbers language also appears in section 101(a)(5)(A). In NMFS' 
view, the statute is silent on how to apply the same small numbers 
limitation in these two provisions across potentially different scales 
and timeframes. In the case such as here, where serious injury or 
mortality is not expected from the activity (and would not be 
authorized in any LOA), each prospective LOA applicant could instead 
opt to apply for an IHA under section 101(a)(5)(D). It would be an 
absurd result to deny an LOA for a single geophysical survey on the 
sole basis that small numbers is not satisfied because the take numbers 
from that survey must be aggregated with the takes from other surveys 
occurring under the same regulations, only to turn around and issue an 
IHA for the same survey, simply because the applicant has decided to 
avail itself of section 101(a)(5)(D) instead. But that would be the 
result under the MMC's approach. Given NMFS' implementing regulations 
for section 101(a)(5)(A), which are authorized under 16 U.S.C. 1382(a), 
and when viewed in light of section 101(a)(5)(D) and applying our 
administrative experience, NMFS has determined our approach is a 
reasonable interpretation of how to carry out section 101(a)(5)(A) and 
the implementing regulations in the context of these two statutory 
provisions. This is a reasoned approach that draws on NMFS' expertise.
    Further, authorization of take incidental to geophysical survey 
activity within the covered regions of the GOM under this ITR allows 
for the more comprehensive evaluation and management of take of marine 
mammals than if NMFS were to authorize take for those same activities 
under IHAs. NMFS worked with BOEM and its predecessor agency over many 
years to ensure a process that holistically analyzed the impacts from 
expected geophysical surveys in the GOM. This is preferable first and 
foremost for its greater likelihood of achieving the best substantive 
impact analysis and comprehensive management (including mitigation and 
monitoring) scheme, but the process is also efficient for stakeholders 
(regulated industry and interested members of the public) and results 
in more efficient use of administrative agency resources.
    The MMC argues that NMFS' implementing regulations support the 
MMC's view of the application of small numbers, because ``whereas the 
regulatory section governing the issuance of incidental take 
regulations (50 CFR 216.105) includes a reference to the small numbers 
requirement, the section governing LOAs (50 CFR 216.106) omits any 
reference to that requirement.'' However, the implementing regulations 
originally defined small numbers as synonymous with negligible impact. 
NMFS no longer interprets small numbers in that way, but as a result of 
that original approach, the MMC's particular citations do not shed 
light on the permissible approach for making a small numbers 
determination as that term is now interpreted.
    NMFS agrees with the MMC that workload alone would not be a 
sufficient basis for our interpretation, and it is not what we rely on. 
Rather, the analysis we presented leads us to conclude that NMFS has 
discretion to apply small numbers at the LOA level and, in this case, 
policy considerations supported that approach.
    Comment: NRDC states that NMFS' interpretation of small numbers 
``leads to absurd results and permits excessive take.''
    Response: NMFS' negligible impact assessment evaluated the risk to 
the affected species and stocks of marine mammals, taking into account 
the amount and severity of anticipated take (and take the agency is 
prepared to authorize) as well as the status of the species and 
mitigation/monitoring. Of note, and as indicated in Changes from the 
Proposed Rule, as a result of BOEM's updated scope of the activities 
and the associated revisions to the levels of effort, both the maximum 
allowable amount of take under the rule, as well as the maximum annual 
take,

[[Page 5368]]

has decreased (significantly in some cases) for all species and stocks 
except two, for which maximum allowable take and/or maximum annual take 
increased slightly, and the severity of many of the impacts has been 
lessened via the elimination and/or reduction of take in areas of 
greater biological importance previously considered as mitigation 
areas.
    The numbers of potential incidents of take or animals taken are 
only part of an assessment and are not, alone, decisively indicative of 
the degree of impact. In order to adequately evaluate the effects of 
noise exposure at the population level, the total number of take 
incidents must be further interpreted in context of relevant biological 
and population parameters and other biological, environmental, and 
anthropogenic factors and in a spatially and temporally explicit 
manner. The effects to individuals of a ``take'' are not necessarily 
equal. Some take events represent exposures that only just exceed a 
Level B harassment threshold, which would be expected to result in 
lower-level impacts, while other exposures (fewer, as the exposure 
modeling effort illustrates) occur at higher received levels and would 
typically be expected to have comparatively greater potential impacts 
on an individual. Further, responses to similar received levels may 
result in significantly different impacts on an individual dependent 
upon the context of the exposure or the status of the individuals 
(e.g., if it occurred in an area and time where concentrated feeding 
was occurring, or to individuals weakened by other effects). Last, 
impacts of a similar degree on a proportion of the individuals in a 
stock may have differing impacts to the stock based on its status, 
i.e., smaller stocks may be less able to absorb deaths or reproductive 
suppression and maintain similar growth rates as larger stocks.
    Comment: The MMC recommended that if such determinations are made 
based on a proportion of a species' or stock's abundance, NMFS adopt a 
policy interpreting the small numbers requirement of section 
101(a)(5)(A) such that it: (1) Include a sliding scale, such that a 
lower proportion is allowed as stock size increases, and (2) include an 
evaluation of the relative risk that the established threshold would be 
exceeded if the best available population estimate or some other 
metric, such as a minimum or intermediate population, is used.
    Response: NMFS disagrees with these recommendations. Under the 
``one-third'' interpretation offered here, and on which we will base 
our small numbers analyses when evaluating LOA applications under this 
rule, take equating to greater than one-third of the predicted 
individuals in the population would generally not be considered small 
numbers. The MMC presents an example from a very large population, 
asserting that an amount of take that would meet NMFS' proportional 
small numbers standard would not appropriately be considered ``small'' 
because it is large in terms of absolute magnitude. The MMC does not 
present a rationale for why its proposed sliding scale approach is more 
appropriate, nor does it provide an explanation of what the drawbacks 
are (biological or otherwise) of authorizing takes of large numbers of 
marine mammals (in the absolute sense) from a significantly large (and 
arguably healthier and more robust) population (even where still less 
than one-third of the population under NMFS' proportional approach). We 
have determined that a proportional approach is the appropriate way to 
interpret small numbers, not an absolute ``on its face'' numeric 
standard. Accordingly, absolute numbers would not be relevant to our 
small numbers determinations. There is no meaningful way to define what 
should be considered as a ``small'' number on the basis of absolute 
magnitude, and the MMC offers no such recommendation.

Mitigation, Monitoring, and Reporting

    Comment: NRDC states that NMFS should include a year-round area 
closure for Bryde's whales. Specifically, NRDC states that this should 
include the following: (1) Excluding airgun surveys year-round from the 
whales' occupied habitat; (2) excluding airgun surveys from areas 
identified, through modeling, as most likely to propagate low-frequency 
sound into the Bryde's whales' habitat; and (3) establishing mitigation 
to reduce noise in the whales' unoccupied habitat, i.e., areas they are 
likely to have inhabited according to the whaling records and have 
habitat characteristics similar to those of the De Soto canyon. The MMC 
also recommends that NMFS include a year-round area closure for Bryde's 
whales, while agreeing that the area defined by NMFS in the proposed 
rule is appropriate. In addition, a private citizen commented that a 
year-round closure is more appropriate than a seasonal closure, because 
Bryde's whales use the area year-round. The Associations and other 
industry commenters argue to the contrary, stating that there should be 
no restriction within the Bryde's whale area and that, if a restriction 
is required, it should be seasonal rather than year-round. The 
Associations also state that if implemented, the restriction area 
should be smaller. With regard to the other alternative offered by NMFS 
for comment--no restriction but a requirement to conduct real-time 
whale detection through use of a moored listening array--the 
Associations state, ``the final ITR should not impose a moored array 
requirement because the limits inherent in such data are outweighed by 
the impracticability of such arrays.'' The CRE also comments, with no 
supporting information, that there should be no restriction on survey 
effort in the Bryde's whale core habitat area.
    Response: As described in the proposed rule, NMFS agrees with NRDC 
and the MMC that the status (e.g., small population size, restricted 
distribution, anthropogenic effects, small population effects) of the 
recently ESA-listed GOM Bryde's whale warranted the consideration of a 
year-round closure to airgun surveys within the area described as core 
habitat for the whale (Area #3). We disagree with the Associations' 
arguments that no requirement is warranted. However, the comments 
specifically relating to the need (or lack thereof) to impose a 
restriction on survey effort in Bryde's whale core habitat, the 
duration of any such restriction, or any additional requirements in the 
core habitat area, are no longer relevant following BOEM's updated 
scope of activity. This update means that no survey effort within 
Bryde's whale core habitat is considered through this rulemaking and 
the vast majority of any anticipated or authorized impacts to this 
species have been eliminated. Please see Table 1 and Figure 2, earlier 
in this notice.
    Regarding NRDC's recommendations for establishing Bryde's whale 
mitigation measures beyond the core habitat area identified in the 
notice of proposed rulemaking, NMFS does not believe these are 
warranted. We initially note that the comment uses the terms ``occupied 
and ``unoccupied'' to describe habitat. These are terms of art in the 
Endangered Species Act and implementing regulations for designation of 
``critical habitat.'' For this MMPA rulemaking, the correct standard is 
measures to effect the ``least practicable adverse impact'' on the 
affected species or stocks and their habitat. NMFS has now determined 
that additional geographic-based mitigation for Bryde's whales is not 
warranted. Following BOEM's update to the scope of their specified 
activity, expected takes of Bryde's whales are significantly reduced in 
the remaining area where the specified activity will occur under this

[[Page 5369]]

rule (i.e., there are now no more than 10 anticipated instances of take 
annually; see Table 9).
    Regarding NRDC's comments that additional protections are needed in 
areas that are ``unoccupied'' by the Bryde's whale, we disagree. NMFS' 
objective in requiring a closure would be to minimize the effects of 
airgun surveys on Bryde's whales while in important habitat. In areas 
where modeling and/or observational data show a species or stock is 
unlikely to occur during the period of the rule, it is generally 
unlikely that a geographic or other mitigative restriction would reduce 
impacts from the specified activities on the species or stock and its 
habitat, and therefore is not justifiable absent some other compelling 
basis. Finally, we are unsure of what NRDC might mean in recommending 
exclusion of surveys from areas identified as most likely to propagate 
low-frequency sound into Bryde's whale habitat, or whether such areas 
are still covered by the rule given BOEM's updated scope, and NRDC 
provides no meaningful justification for the recommendation, nor any 
useful recommendations for how such areas could be identified.
    Comment: In reference to NMFS' statement that the agency does not 
consider towed passive acoustic monitoring (PAM) to be a useful tool 
with regard to detection of Bryde's whales, the Associations state that 
they do believe more typical real-time detection-based mitigation, such 
as use of towed PAM, should provide sufficient protection for Bryde's 
whales, and assert that we did not provide sufficient information to 
meaningfully comment on the conclusion.
    Response: It is generally well-accepted fact that, even in the 
absence of a firing airgun, using a towed passive acoustic sensor to 
detect baleen whales (including Bryde's whales) is not typically 
effective because the noise from the vessel, the flow noise, and the 
cable noise are in the same frequency band and will mask the vast 
majority of baleen whale calls. Further, Bryde's whales have relatively 
short calls, further exacerbating the problem. As background, airguns 
produce loud, broadband, impulsive signals at low frequencies (e.g., 
Hildebrand, 2004). Source characteristics are variable but typically 
peak pressures are in the 5-300 Hz frequency range, with source levels 
as high as 260 dB peak re 1 [mu]Pa at 1 m output pressure (Hildebrand, 
2009). Pulse rates are typically one per 10-20 s (Hildebrand, 2009). 
Seismic survey noise can raise background noise levels by 20 dB or more 
over large areas while present. Because the seismic pulse and the 
whale's call are within the same frequency range, and the seismic pulse 
is much louder than the whale's call (see below), it is extremely 
unlikely that a baleen whale can be detected during the pulse. In 
addition to the actual seismic pulse (approximately every 10-20 s), the 
background noise level is expected to be significantly increased as a 
result of the reverberant field generated from seismic pulses (Guerra 
et al., 2011; Guan et al., 2015), i.e., during the inter-pulse 
interval. The level of elevated inter-pulse noise levels can be as high 
as 30-45 dB within 1 km of an active 3,147 in\3\ airgun array (Guerra 
et al., 2011). Given that towing hydrophones for PAM used for marine 
mammal monitoring would be within 1 km from the airgun source, the 
received noise spectral density during the inter-pulse interval is 
expected to be very high.
    Vessels also produce low-frequency noise, primarily through 
propeller cavitation, with main energy also in the 5-300 Hz frequency 
range. Source levels range from about 140 to 195 dB re 1 [mu]Pa at 1 m 
(NRC, 2003; Hildebrand, 2009), depending on factors such as ship type, 
load, and speed, and ship hull and propeller design. Studies of vessel 
noise show that it appears to increase background noise levels in the 
71-224 Hz range by 10-13 dB (Hatch et al., 2012; McKenna et al., 2012; 
Rolland et al., 2012). PAM systems employ hydrophones towed in streamer 
cables approximately 500 m behind a vessel. Noise from water flow 
around the cables and from strumming of the cables themselves is also 
low-frequency and typically masks signals in the same range.
    GOM Bryde's whale calls have relatively low source levels (155 dB 
re 1[mu]Pa) and frequency ranges (78-110 Hz; [Scaron]irovi[cacute] et 
al., 2014) that overlap the sounds described above. In addition, GOM 
Bryde's whales call only infrequently (i.e., a 3.5 hour research 
encounter with 4 whales resulted in detections of 14 calls). The 
chances of acoustically detecting these whales is low under ideal 
research circumstances, is much lower with elevated background noise 
from the ship and towing cable, and essentially impossible with an 
airgun array shooting. Whales are routinely detected acoustically using 
moored systems and sonobuoys, or using autonomous gliders. However, 
these platforms are all quiet. A leading provider of observer services 
for the seismic industry, including PAM, reports that they have never 
detected a baleen whale (other than rare detections of humpback whales, 
which have significantly higher frequency content in their call) using 
PAM aboard a working seismic vessel (S. Milne, RPS Group, pers. comm.). 
Experienced PAM operators participating in a recent workshop (Thode et 
al., 2017) emphasized that a PAM operation could easily report no 
acoustic encounters, depending on species present, simply because 
background noise levels rendered any acoustic detection impossible. The 
same workshop report stated that a typical eight-element array towed 
500 m behind a seismic vessel could be expected to detect delphinids, 
sperm whales, and beaked whales at the required range, but not baleen 
whales, due to expected background noise levels (including seismic 
noise, vessel noise, and flow noise).
    Comment: The Associations provided comments regarding NMFS' 
proposed power-down exception to the general shutdown requirements for 
certain species of dolphin, as well as the related alternative of no 
shutdown or power-down requirement. The Associations stated that no 
shutdowns for dolphins are warranted, and added that an exception 
should not be limited to small dolphins but rather should be expanded 
to all delphinid species. The MMC recommended that NMFS not require a 
shutdown or power-down when small delphinids enter the exclusion zone, 
and relatedly suggested that NMFS should provide clarification as to 
the basis for exempting only small delphinids from shutdowns. The MMC 
stated their agreement with NMFS that shutting down when small 
delphinids enter the exclusion zone is not warranted and may result in 
additional survey activity. Furthermore, as indicated in the MMC 
comments, power-down may not be effective. The MMC stated that, given 
the variation in array characteristics and configuration, a requirement 
to ``power-down'' does not provide sufficient assurance that the 
resulting received levels would be below the Level B harassment 
threshold. CGG provided a detailed analysis of the potential 
operational costs associated with dolphin shutdowns or power-downs, 
supporting their comment that these costs would be substantial and that 
shutdown or power-down should not be required. NRDC provided multiple 
objections to NMFS' proposals, stating that of the two proposals they 
favor power-down.
    Response: Following review of the available information and public 
comments, NMFS agrees that a general exception to the standard shutdown 
requirement is warranted for small delphinids, and that the alternative 
power-down requirement may not be effective and yet could impose costs 
on

[[Page 5370]]

operators. (Here we refer to ``large delphinids'' and ``small 
delphinids'' as shorthand for generally deep-diving versus surface-
dwelling/bow-riding groups, respectively, as the important distinction 
is their dive behavior rather than their size.) As NMFS discussed in 
the notice of proposed rulemaking, mid- and high-frequency cetaceans 
are relatively insensitive to the frequencies where the most energy in 
an airgun signal is found. In order to demonstrate this quantitatively, 
a ``spectral ratio'' may be calculated for each hearing group. This 
ratio essentially compares the energy in a group-specific weighted 
airgun source spectrum with the energy in an unweighted airgun source 
spectrum, providing a representation of the proportion of total energy 
from the unweighted airgun spectrum that is available for animals to 
hear based on their group-specific general auditory filter shapes, 
which presumably influences the probability of behavioral response. 
Using M-weighting (i.e., Type I filters), spectral ratios for the three 
hearing groups are as follows: LF, 0.71; MF, 0.03; HF, 0.02.
    However, NMFS does not agree that the available evidence supports 
certain commenters' assertions that seismic surveys do not have any 
adverse effects on dolphin species. As discussed in Mitigation, 
auditory injury is not expected for dolphins, but the reason for 
dolphin behavior around vessels (when they are attracted) is not 
understood and cannot be assumed to be harmless. In fact, the analyses 
of Barkaszi et al. (2012), Stone (2015a), Stone et al. (2017), and 
Barkaszi and Kelly (2018) show that dolphins do avoid working vessels. 
That said, the available information does not suggest that such 
reactions are likely to have meaningful energetic effects to 
individuals such that the effectiveness of such measures outweighs the 
practicability concerns raised by commenters, in terms of the 
operational costs as well as the difficulty of implementation.
    As noted above, the proposed rule included an alternative in which 
a power-down requirement would be required. However, following review 
of public comments, NMFS believes that a power-down requirement would 
potentially lead to the need for termination of survey lines and infill 
of the line where data were not acquired if a power-down was performed 
according to accepted practice, in which the power-down condition would 
last until the dolphin(s) are no longer observed within the exclusion 
zone. The need to revisit missed track line to reacquire data is likely 
to result in an overall increase in the total sound energy input to the 
marine environment and an increase in the total duration over which the 
survey is active in a given area.
    NMFS disagrees with comments that no shutdown requirements should 
apply to any delphinid species regardless of behavior. As noted above, 
industry commenters have asserted that no shutdown requirements are 
warranted for any delphinid species, stating that the best available 
science does not support imposing such requirements. The industry 
comments acknowledge that small delphinids are more likely to approach 
survey vessels than large delphinids, but claim without supporting data 
that there is no evidence that large delphinids will benefit from a 
shutdown requirement. In contrast to the typical behaviors of (and 
observed effects on) the small delphinid species group, the typical 
deep diving behavior of the relatively rarely occurring large delphinid 
group of species makes these animals potentially susceptible to 
interrupted/delayed feeding dives, which can cause energetic losses 
that can accrue to affect fitness. As described in greater detail in 
the notice of proposed rulemaking, there are ample data illustrating 
the responses of deeper diving odontocetes (including large delphinids) 
to loud sound sources (including seismic) to include interrupted 
foraging dives, as well as avoidance with increased speed and stroke 
rate, both of which may contribute to energetic costs through lost 
feeding opportunities and/or increased energy demands. Significant 
advances in study of the population consequences of disturbance are 
informing our understanding of how disturbances accrue to effects on 
individual fitness (reproduction and survival) and ultimately to 
populations via the use of energetic models, where data are available 
for a species, and expert elicitation when data are still limited. The 
link between behavioral disturbance, reduced energy budgets, and 
impacts on reproduction and survival is clear, as is the value in 
reducing the probability or severity of these behavioral disturbances 
where possible. Therefore, NMFS finds that there is support for the 
effectiveness of the standard shutdown requirement as applied to the 
large delphinid species group.
    Further, the claim that shutdowns for these deep-diving species 
would be impracticable was not accompanied by supporting data. The data 
available to NMFS demonstrates that this requirement is practicable. 
For example, recent synthesis of observer data in the GOM shows that 
large delphinids were sighted only rarely, and that of these sightings, 
almost half were not within the 500-meter exclusion zone. We note that 
the Associations provided a quantitative analysis of ``historical PSO 
and PAM data from over 32,000 survey activity hours conducted in the 
GOM between 2007 and 2017,'' but provide no citation for these data 
(nor the data itself). Therefore, we cannot verify or meaningfully 
evaluate the industry-supplied analysis. Nevertheless, as detailed 
herein, NMFS agrees in substantial part with the comments received and 
accordingly do not require shutdown or power-down for small delphinids 
detected within the exclusion zone.
    Comment: Several commenters criticized our proposal to require 
shutdowns upon detection of certain species or circumstances (e.g., 
beaked whales, Bryde's whales) at any distance. The Associations 
suggest that such requirements are ``arbitrary and unlawful'' because 
they require shutdowns in ``circumstances in which no disturbance or 
harassment will occur.'' The Associations contend that PSOs are likely 
to make frequent ``precautionary'' shutdown calls for uncertain 
observations ``at any distance,'' and that these measures will have 
negative impacts on the effectiveness of visual PSOs. CGG makes similar 
claims, stating that ``there is no proven or likely efficacy to 
initiate a shutdown for cetaceans that are well outside of incidental 
take range'' and concluding that the standard 500-m exclusion zone 
should be applied in these circumstances. The MMC commented that, in 
reference to the two proposals of ``at any distance'' or ``within 1 
km,'' they support the implementation of shutdowns for detections at 
any distance (rather than within 1 km of the airgun array), based on 
the status of the applicable species, their small population sizes, and 
their sensitivity to seismic sound.
    Response: As discussed below and in Mitigation, an extended 
shutdown distance of 1.5 km is included in the final rule, in lieu of 
the ``at any distance'' shutdown included in the proposed rule. We 
first note that the industry comments against proposed shutdowns for 
certain species, in their view beyond the range at which harassment may 
occur, appears to reflect an assumption that the single-step 160-dB 
threshold is the relevant metric for harassment. Even if this were the 
case, the minimum distance to the 160-dB isopleth, based on 60 
different propagation modeling scenarios, would be beyond the likely 
detection distance for visual observers. The smallest

[[Page 5371]]

threshold radius to the 160-dB isopleth is more than 7 km. However, the 
multi-step probabilistic risk function used here assumes that 10 
percent of the population exposed above 140 dB would experience 
harassment; isopleth distances to 140 dB, based on the same modeling 
exercise, are typically greater than 50 km (minimum of approximately 29 
km). Even the 90 percent harassment isopleth (i.e., 180 dB) has a mean 
distance of 1.6 km. Therefore, the claims that shutdowns upon detection 
``at any distance'' would occur in circumstances where there is no 
harassment are incorrect. The Associations' comments are also 
inconsistent in that they imply both that marine mammals are likely to 
be detected at ranges significantly distant from the vessel, where 
shutdowns would be effected on detection of animals not subject to 
harassment, and that marine mammals cannot be adequately identified 
beyond close distances, resulting in unnecessary ``precautionary'' 
shutdowns. NMFS agrees that visual monitoring under typical 
circumstances is unlikely to be effective at ranges much beyond the 
extended distance shutdown of 1.5 km, while under ideal circumstances 
acoustic detectability will also be limited to within the exclusion 
zone distance. (NMFS presented a detailed analysis in the notice of 
proposed rulemaking demonstrating that acoustic detections of sperm 
whales during active firing of an airgun array are not likely beyond 
approximately 500 m). Moreover, we specify in these regulations that 
shutdowns are required on positive identification of relevant species 
(as determined through professional judgment), meaning that there is no 
real likelihood that there would be numerous shutdowns based on false 
positive detections. Overall, it is unlikely that there will be 
``unnecessary'' shutdowns to any significant degree.
    The MMC provided the following supporting rationale to their 
comment: ``Bryde's whales are LF cetaceans with particular sensitivity 
to the predominantly low-frequency energy output of airguns. Beaked 
whales are well-documented to react behaviorally to sound levels well 
below those thought to cause injury, and larger exclusion zones have 
been recommended for beaked whales and other deep-diving whales (such 
as Kogia spp. and sperm whales) as they are more likely to exhibit a 
stress response when disturbed (Wright et al., 2011).'' NMFS agrees 
with these comments. In these cases, we have identified species or 
circumstances with particular sensitivities for which we determined it 
appropriate to minimize the duration and intensity of the behavioral 
disruption, as well as to minimize the potential for auditory injury 
(for low- and high-frequency cetaceans).
    NMFS disagrees with industry comments regarding the likelihood that 
trained, experienced professional PSOs would misunderstand the intent 
of a requirement to shut down upon detection ``at any distance'' and 
would therefore spend undue time focusing observational effort at 
distances beyond approximately 1,000 m from the acoustic source (i.e., 
the zone within which we assume that monitoring is typically focused, 
though not necessarily exclusively). Nevertheless, in order to ensure 
that this potential is minimized, and to address commenters' concerns 
regarding the potential costs associated with shutdowns at any 
distance, especially in light of the diminished benefits of the measure 
beyond 1.5 km, we limit these shutdowns to within 1.5 km (versus at any 
distance). The rationale for this distance is explained later in this 
document in Mitigation.
    Comment: NRDC states that NMFS should require that ramp-up occur 
over several stages in order to minimize exposure.
    Response: NMFS agrees with NRDC on this point, which appears to 
restate the ramp-up procedures described by NMFS in the notice of 
proposed rulemaking. NMFS believes this approach is consistent with the 
Australian study referenced by NRDC.
    Comment: NRDC states that the standard 500-m exclusion zone is 
``not conservative,'' asserting that NMFS did not explain why the 
proposed zone achieves the least practicable adverse impact and stating 
that NMFS must consider other exclusion zone distances.
    Response: NMFS has acknowledged that some limited occurrence of 
auditory injury is likely, for low- and high-frequency cetaceans. 
However, we disagree that a larger standard exclusion zone is 
warranted. As explained in the notice of proposed rulemaking, NMFS' 
intent in prescribing a standard exclusion zone distance is to (1) 
encompass zones for most species within which auditory injury could 
occur on the basis of instantaneous exposure; (2) provide additional 
protection from the potential for more severe behavioral reactions 
(e.g., panic, antipredator response) for marine mammals at relatively 
close range to the acoustic source; (3) provide consistency and ease of 
implementation for PSOs, who need to monitor and implement the 
exclusion zone; and (4) to define a distance within which detection 
probabilities are reasonably high for most species under typical 
conditions. The use of 500 m as the zone is not based directly on any 
quantitative understanding of the range at which auditory injury would 
be entirely precluded or any range specifically related to disruption 
of behavioral patterns. Rather, NMFS believes it is based on a 
reasonable combination of factors. In summary, a practicable criterion 
such as this has the advantage of familiarity and simplicity while 
still providing in most cases a zone larger than relevant auditory 
injury zones, given realistic movement of source and receiver. 
Increased shutdowns, without a firm idea of the outcome the measure 
seeks to avoid, simply displace survey activity in time and increase 
the total duration of acoustic influence as well as total sound energy 
in the water, which NMFS seeks to avoid.
    NMFS agrees that, when practicable, the exclusion zone should 
encompass distances within which auditory injury is expected to occur 
on the basis of instantaneous exposure. For high-frequency cetaceans, 
this distance was modeled as 457 m (though we acknowledged that the 
actual distance would be dependent on the specific airgun array and 
could be larger). However, we require an extended exclusion zone of 1.5 
km for certain sensitive species, including Kogia spp. Potential 
auditory injury for low-frequency cetaceans is based on the 
accumulation of energy, and is therefore not a straightforward 
consideration. However, the extended exclusion zone is required for the 
only low-frequency cetacean in the GOM (Bryde's whale). In keeping with 
the four broad goals outlined above, and in context of the information 
given here, the standard 500-m exclusion zone is appropriate. NRDC does 
not provide any substantive reasoning for a larger zone.
    Comment: Several industry commenters criticized the requirement for 
use of buffer zones in addition to the standard exclusion zones, 
claiming in part that there is no scientific basis for monitoring a 
zone larger than the exclusion zones.
    Response: NMFS disagrees with the suggestion that there is no 
scientific basis for this requirement. It is important to implement a 
larger zone during pre-clearance, when na[iuml]ve animals may be 
present and potentially subject to severe behavioral reactions if 
airguns begin firing at close range. While the delineation of zones is 
typically associated with shutdown, the period during which use of the 
acoustic source is being initiated is critical, and

[[Page 5372]]

in order to avoid more severe behavioral reactions it is important to 
be cautionary regarding marine mammal presence in the vicinity when the 
source is turned on. This requirement has broad acceptance in other 
required protocols: The Brazilian Institute of the Environment and 
Natural Resources previously required a 1,000-m pre-clearance zone 
before recently extending the exclusion zone to encompass the entire 
1,000-m zone (IBAMA, 2005, 2018), the New Zealand Department of 
Conservation requires that a 1,000-m zone be monitored as both a pre-
clearance and a shutdown zone for most species (DOC, 2013), and the 
Australian Department of the Environment, Water, Heritage and the Arts 
requires an even more protective scheme, in which a 2,000-m ``power 
down'' zone is maintained for higher-power surveys (DEWHA, 2008). 
Broker et al. (2015) describe the use of a precautionary 2-km exclusion 
zone in the absence of sound source verification (SSV), with a minimum 
zone radius of 1 km (regardless of SSV results). We believe that the 
simple doubling of the exclusion zone required here is appropriate for 
use as a pre-clearance zone.
    Comment: CGG comments that shutdowns based on acoustic detections 
should be required only when the acoustic PSO is confident that the 
vocalization is from a non-delphinid species within the exclusion zone, 
as opposed to when the PSO is confident that the animal is outside of 
the exclusion zone.
    Response: We are unclear as to the practical impact of what appears 
to be a fairly nuanced difference, but clarify that shutdown upon 
acoustic detection of non-delphinids within the exclusion zone is 
required when the animal is detected acoustically and localized within 
the exclusion zone. However, we also note that PSO decision-making 
regarding shutdown implementation shall be informed to a reasonable 
extent by professional judgment.
    Comment: NRDC suggests that NMFS is remiss in not limiting the 
amount of activity that can occur overall (to a lesser amount than 
analyzed in the rule). Relatedly, NRDC suggests that NMFS must consider 
``placing a cap on the amount of allowable seismic activity.''
    Response: Such a requirement is not within NMFS' authority under 
the MMPA, assuming that the requisite findings are made. NMFS' 
responsibility is to evaluate the potential effects of the specified 
activity as presented by the applicant (BOEM in this case, acting on 
behalf of future industry applicants) and to determine whether the 
total taking will have a negligible impact on the affected species or 
stocks (among other things). If NMFS is unable to make the necessary 
finding, the applicant may then consider a revision to the specified 
activity that could lead to NMFS being able to make the necessary 
finding of negligible impact (or in some cases additional mitigation 
may enable a negligible impact finding). However, in this case, NMFS 
has made a finding of negligible impact, and it is not within NMFS' 
authority to unilaterally impose a reduction in activity levels to some 
degree (NRDC does not specify the degree or distribution of reduction 
in time or space that they would find acceptable).
    Comment: NRDC expressed concern regarding the efficacy of the 
prescribed visual and acoustic monitoring methods, stating that species 
could go undetected.
    Response: While NMFS disagrees with some specific comments 
regarding efficacy, we generally agree with the overall point that 
there are limitations on what may reasonably be expected of either 
visual or acoustic monitoring. While visual and acoustic monitoring 
effectively complement each other, and acoustic monitoring is the more 
effective monitoring method (for certain species) during periods of 
impaired visibility, there is no expectation that these methods will 
detect all marine mammals present. In general, NRDC appears to 
misunderstand what NMFS claims with regard to what such monitoring may 
reasonably be expected to accomplish and/or the extent to which we rely 
on assumptions regarding the efficacy of monitoring in reaching the 
necessary findings. We acknowledge these limitations in prescribing 
these monitoring requirements, while stating why NMFS believes that 
visual and acoustic monitoring, and the related protocols we have 
prescribed, are an appropriate part of the suite of mitigation measures 
here that satisfy the MMPA's least practicable adverse impact standard. 
However, the negligible impact finding is not conditioned on the 
presumption of a specific degree of monitoring efficacy.
    Comment: The MMC recommends that NMFS expand its shutdown 
requirement for sperm whales to include both visual and acoustic 
detections at extended distance, stating that vital functions of sperm 
whales, including both foraging and resting, should be afforded the 
additional protection of the extended shutdown zone. NRDC asserts that 
acoustic shutdowns for sperm whales, which they believe are not 
required under the ITR, would not be effective. The CRE comments that 
they ``agree with [NMFS] that sperm whale shutdowns are not 
warranted.''
    Response: NMFS agrees with the MMC's recommendation and has made 
the recommended change (albeit within a revised extended distance 
shutdown zone of 1.5 km; see Mitigation). However, we note the MMC's 
statement that ``[t]he requirement for implementing shut-down 
procedures upon acoustic detection of a sperm whale was inadvertently 
omitted from the proposed regulatory text.'' NMFS disagrees with this 
statement. The proper interpretation of the proposed regulatory text 
was that such shutdowns would be required. Nevertheless, the revised, 
final regulatory text makes this requirement clearer, in addition to 
making the change to be inclusive of visual detections at the greater 
distance. Regarding the CRE's comment, NMFS did not determine that 
``sperm whale shutdowns are not warranted.'' Shutdowns for sperm whales 
have been required in the GOM for over a decade, and NMFS does not make 
any findings that this should change.
    With regard to NRDC, we reference this comment only to provide 
necessary clarification. Because NRDC mistakenly claims that ``NMFS 
hasn't included an acoustic shutdown requirement for sperm whales in 
its proposed regulation,'' we refer the reader to the notice of 
proposed rulemaking, in which we state that shutdown of the acoustic 
source is required upon acoustic detection of a sperm whale (29274-
29275). (``We are proposing that shutdown of the acoustic source should 
also be required in the event of certain other observations [. . .]. 
Circumstances [. . .] include [. . .] acoustic detection of a sperm 
whale.'') This requirement is carried forward in this final ITR, as 
modified (see Mitigation).
    With regard to the efficacy of the measure, we are confused as to 
NRDC's comments. NRDC first asserts that sperm whales are the only 
species for which acoustic detection may reasonably be assumed, but 
then seemingly states that implementation of the measure is not 
sufficiently effective as to be considered in context of reducing 
impacts to sperm whales. As discussed in greater detail elsewhere, NMFS 
believes that shutdowns for sperm whales at an extended distance, on 
the basis of both acoustic and visual detections (the latter added in 
this final ITR), will meaningfully reduce impacts to the species.
    Comment: NRDC asserts that NMFS does not fulfill the MMPA's 
requirement to prescribe mitigation achieving the

[[Page 5373]]

``least practicable adverse impact'' to marine mammal habitat, and 
specifically notes that NMFS does not separately consider mitigation 
aimed at reducing impacts to marine mammal habitat, as the MMPA 
requires.
    Response: NMFS disagrees with this comment. Our discussion of least 
practicable adverse impact points out that because habitat value is 
informed by marine mammal presence and use, in some cases there may be 
overlap in measures for the species or stock and for use of habitat. In 
the notice of proposed rulemaking, NMFS identified time-area 
restrictions based on a combination of factors that include higher 
densities and observations of specific important behaviors of the 
animals themselves, but also clearly reflect preferred habitat. In 
addition to being delineated based on physical features that drive 
habitat function (e.g., bathymetric features, among others), the high 
densities and concentration of certain important behaviors (e.g., 
feeding) in these particular areas clearly indicates the presence of 
preferred habitat. Also, NRDC asserts that NMFS must ``separately'' 
consider measures aimed at marine mammal habitat. The MMPA does not 
specify that effects to habitat must be mitigated in separate measures, 
and the notice of proposed rulemaking clearly identified measures that 
provide significant reduction of impacts to both ``marine mammal 
species and stocks and their habitat,'' as required by the statute. 
Last, we note that NRDC acknowledges that the measures identified in 
the notice of proposed rulemaking measures would reduce impacts on 
``acoustic habitat.'' Following BOEM's update to the scope of activity, 
two of the three time-area restrictions identified and proposed by NMFS 
now fall outside the area in which survey activity may be considered 
under this rule.
    Comment: NRDC recommends that NMFS should consider a year-round 
restriction on geophysical survey activity within coastal waters in the 
footprint of the DWH oil spill, and that NMFS must expand its proposed 
GOM-wide coastal restriction temporally to include the month of 
January. Conversely, the Associations state that no coastal restriction 
should be required. The MMC recommends that the proposed coastal 
closure be expanded temporally such that the timeframe is from January 
through August.
    Response: NMFS finds aspects of both NRDC's and the Associations' 
statements with which we agree and disagree and, as discussed in 
Mitigation, have revised the time-area restriction. This restriction on 
airgun survey activity (``Area 1'') was proposed as including all GOM 
waters inside the 20-m isobath, from February through May. The revised 
restriction is limited to those waters inside the 20-m isobath from 90 
[deg] to 84 [deg] W. Temporally, the restriction is expanded to be in 
effect from January through May.
    The Associations provide extensive comments relating to the impacts 
on practicability presented by the proposed restriction. The potential 
economic consequences of the measure are addressed in greater detail in 
the regulatory impact analysis (RIA), which analysis we adopt as a 
portion of our practicability assessment for the revised measure. NMFS 
agrees that there will likely be negative economic and operational 
consequences of the restriction, though these consequences are 
difficult to assess (and cannot reasonably be assessed quantitatively) 
(see the RIA for full analysis). While the Associations express 
concerns regarding the practicability analysis as being too vague, they 
fail to provide additional specific information that would help to 
improve the analysis. For example, the Associations state that data 
from the area contained within the restriction are outdated and that 
the restriction will impede industry's ability to identify prospects in 
coastal areas, but provide no specific information to support these 
claims, such as information about the data that do exist or the areas 
where industry anticipates having interest in identifying prospects. 
Despite the lack of information provided in support of the 
practicability concerns, NMFS takes seriously the Associations' 
concerns, and therefore did consider eliminating the restriction.
    The Associations also assert that the restriction would not result 
in any meaningful benefit to coastal bottlenose dolphin populations. 
NMFS disagrees that this is the case. Although dolphins are less 
sensitive to the frequencies at which the greatest energy in an airgun 
signal is found, we have described the large body of evidence of 
adverse or aversive behavior by various dolphin species during airgun 
firing (e.g., Goold and Fish, 1998; Stone and Tasker, 2006; Barkaszi et 
al., 2012; Stone, 2015a; Barkaszi and Kelly, 2018). Considered in 
context of a generic dolphin population with no notable issues 
affecting the population as part of the environmental baseline, it may 
be reasonable to assume that such effects are not indicative of any 
response of a severity such that the need to avoid it outweighs the 
impact on practicability for the industry and operators. However, as 
was described in the notice of proposed rulemaking, and as discussed in 
NRDC's comment, coastal bottlenose dolphins in the GOM--particularly 
the northern coastal stock of bottlenose dolphins--were severely 
impacted by the DWH oil spill.
    As explained in the notice of proposed rulemaking, while none of 
the dolphin strandings or deaths have been attributed to airgun survey 
activities, stocks in the area are stressed and the northern coastal 
stock in particular is in extremely poor health. The Associations' 
discussion of NMFS' analysis--claiming that our justification for the 
restriction was premised merely on ``the broad understanding that 
`marine mammals react to underwater noise' ''--is factually mistaken. 
As we stated, behavioral disturbance or stress may reduce fitness for 
individual animals and/or may exacerbate existing declines in 
reproductive health and survivorship. For example, stressors such as 
noise and pollutants may be expected to induce responses involving the 
neuroendocrine system, which controls reactions to stress and regulates 
many body processes (NAS, 2017), and there is strong evidence that 
petroleum-associated chemicals can adversely affect the endocrine 
system, providing a potential pathway for interactions with other 
stressors (Mohr et al., 2008, 2010). Romano et al. (2004) found that 
upon exposure to noise from a seismic watergun, bottlenose dolphins had 
significantly elevated levels of a stress-related hormone and, 
correspondingly, a decrease in immune cells. As we stated, the 
restriction is intended specifically to avoid additional stressors to 
these coastal bottlenose dolphin populations during the time period 
believed to be of greatest importance as a reproductive period. The 
Associations do not contradict this information, instead weakly 
relating the concern to the potential for dolphins to experience damage 
to auditory structures (which NMFS agrees is unlikely) or to the idea 
that ``reactions'' to noise are innocuous.\9\
---------------------------------------------------------------------------

    \9\ The Associations also apparently misunderstand some 
discussion of stranding events (which have occurred primarily as a 
result of military use of mid-frequency active sonar) provided in 
the notice of proposed rulemaking, interpreting this discussion as 
NMFS' ``suggestion that seismic surveys are similar to mid-frequency 
sonar (which has been implicated in strandings) simply because 
seismic signatures include a mid-frequency component.'' We suggested 
no such thing and agree with the Associations that airguns and sonar 
are very different sound sources with very different potential to 
cause strandings.
---------------------------------------------------------------------------

    Population-level impacts related to energetic effects or other 
impacts of noise are difficult to determine, but the addition of other 
stressors can add

[[Page 5374]]

considerable complexity due to the potential for interaction between 
the stressors or their effects (NAS, 2017). When a population is at 
risk, NAS (2017) recommends identifying those stressors that may 
feasibly be mitigated. We cannot undo the effects of the DWH oil spill, 
but the potentially synergistic effects of noise due to the activities 
that are the subject of this rule may be mitigated. However, NMFS does 
acknowledge that the two populations of greatest concern--the western 
and northern coastal stocks of bottlenose dolphin--do not have the same 
status. As identified in the notice of proposed rulemaking, while both 
stocks were impacted by the DWH oil spill, the northern coastal stock 
in particular was perhaps the single most heavily impacted stock, with 
82 percent of animals belonging to the stock expected to have been 
exposed to oil, resulting in a possible population reduction of 50 
percent (this latter figure was only five percent for the western 
stock). The northern coastal stock was also subject to a recent Unusual 
Mortality Event (UME), described later in this notice (see Description 
of Marine Mammals in the Area of the Specified Activity). NMFS 
acknowledges the uncertainty associated with predicting the ways in 
which different stressors may interact, or how the effects of a 
stressor might be exacerbated in an unhealthy population. However, as 
an example, Schwacke et al. (2014a) described findings indicating that 
a significant proportion of the population is expected to exhibit 
adrenal insufficiency as a result of oil exposure. Adrenal 
insufficiency can lead to adrenal crisis and death in animals that are 
challenged with other stressors (Venn-Watson et al., 2015b). NMFS 
agrees that the potential practicability concerns warrant consideration 
and, in light of the differential baselines for the potentially 
affected coastal stocks, has determined it appropriate to contract the 
restriction. However, the post-DWH oil spill baseline condition of the 
northern coastal stock, as exacerbated by the recent UME, requires 
caution. This restriction may reasonably be anticipated to provide 
additional protection to these populations during their peak 
reproductive activity. We note that NRDC's proposed focus area for 
heightened restriction aligns generally with this area of concern, but 
that in aligning with the footprint of the spill rather than with the 
stock boundaries, this recommendation would not necessarily encompass 
the animals of greatest concern and which we assume are the population 
targeted by the proposal.
    With regard to the timing of the closure, there is no definitive 
definition of the ``peak reproductive activity'' associated with the 
stock and, additionally, there is some uncertainty as to whether the 
more important focus is on effects to pregnant mothers or on the post-
partum period when energetic or stress effects would lead to greater 
risk for lactating mothers and/or disruption of mother-calf bonding and 
ultimate effects on rates of neonate and/or calf survivorship. We 
acknowledged this uncertainty in discussing the recommendations of 
NMFS' subject matter experts and describing the proposed temporal 
extent of February through May in the notice of proposed rulemaking. 
Upon review of the information presented in the comments of NRDC (e.g., 
reference to the data presented by, e.g., Carmichael et al., 2012; 
Mattson et al., 2006; Urian et al., 1996), which supported NRDC's 
assertion that, in summary, inclusion of January would cover the 
remainder of the dolphins' peak calving and late gestation periods as 
well as the beginning of the period of highest reproductive failure, 
NMFS agrees that this temporal expansion is appropriate (within the 
contracted region of our revised restriction area). In contrast, the 
MMC does not provide compelling information in support of the 
recommendation to expand the restriction by an additional three months 
(through August), stating only that ``calves can be born at any time of 
the year'' and referencing a bimodal peak in neonate strandings from 
the Sarasota Bay area. Given the exacerbation of practicability 
concerns that this expansion would entail and the lack of information 
to support it, NMFS does not believe it appropriate to expand the 
restriction through August.
    We do note that one concern of the Associations, which is that the 
restriction may result in an inability to complete surveys within one 
year, may be alleviated to some degree by the ability under this ITR to 
issue LOAs for any term up to five years. The Associations recommend 
that, if the restriction is included in the ITR, NMFS allow for multi-
year LOAs, which we have done.
    Comment: The Associations state that the proposed time-area 
restriction in the Dry Tortugas region of the eastern GOM should not be 
required. However, the MMC concurs with NMFS' proposal, stating that 
the imposition of this restriction is appropriate.
    Response: NMFS appreciates the comments. The proposed time-area 
restriction referenced here is no longer relevant following BOEM's 
update to the geographic scope of activity, as no survey activity 
within this area can be considered through this rule.
    Comment: NRDC comments that NMFS must consider restrictions and 
limitations on survey activity in the Central Planning Area (CPA) 
restriction area analyzed in the proposed rule. NRDC states that NMFS' 
practicability analysis must focus on (1) how much oil and gas 
development is projected to occur within the proposed areas over the 
next five years; (2) what effect the proposed mitigation area would 
have on that projected development; and (3) whether that projected 
development would be offset by exploration in other parts of the GOM.
    Response: NRDC accurately characterizes the area as being important 
for sperm whales \10\ and beaked whales, as was described by NMFS in 
the notice of proposed rulemaking, and accurately describes that this 
area is projected to be subject to significant survey effort. NMFS 
acknowledges these issues. However, NRDC provides no serious rebuttal 
of NMFS' practicability analysis, which includes incorporation by 
reference of the findings of the RIA for this rule, instead providing 
only a cursory rejection of the analysis as inadequate. We also note 
that the third prong of NRDC's suggested analysis is not reasonable: 
Development foregone due to a lack of survey data in the closure areas 
cannot be ``offset by exploration'' elsewhere.
---------------------------------------------------------------------------

    \10\ However, we note that NRDC mischaracterizes sperm whale 
buzz rates as ``a measure of foraging success,'' as opposed to a 
measure of foraging effort. The study referenced by NRDC did not 
find that sperm whale foraging success ``declined substantially'' on 
exposure to airgun noise. Moreover, the measured decline in foraging 
effort was not a statistically significant result and, therefore, 
cannot appropriately be referred to as a substantial decline. See 
Miller et al. (2009).
---------------------------------------------------------------------------

    As discussed in detail in the RIA, there are significant 
uncertainties associated with assessing the indirect costs of 
restricting survey effort within the described area. Notable areas of 
uncertainty include the demand for and timing of oil and gas production 
in the GOM over the next five years, the suitability of existing data 
to direct oil and gas production in the closure areas, and the most 
likely substitute sites for oil and gas production. These uncertainties 
foreclose the possibility of the analysis demanded by NRDC. However, 
what information is available strongly suggests that the economic 
impacts of the evaluated CPA restrictions would be significant. A 
mitigation requirement that could lead

[[Page 5375]]

to regional- to national-scale economic impacts is not practicable.
    The impacts of year-round area closures are highly dependent on 
volatile oil and gas market conditions over the next five years, which 
dictate the demand for activities in the GOM. The greater the demand 
for oil and gas, the greater the expected impacts of the restrictions. 
The extent to which oil and gas production is delayed because of the 
need for newer, better data is a key source of uncertainty. Some sites 
may be able to employ existing data from recent surveys. However, even 
for relatively recent data, the inability to collect new seismic data 
could affect oil and gas development given that oil companies typically 
use targeted seismic data to refine their geologic analysis before 
drilling a well.
    It is possible that some fraction of reductions in production from 
the closure areas may be made up for with production in other areas in 
the GOM, mitigating potential regional economic impacts. However, 
uncertainty with regard to the location of ``substitute'' production 
has potentially critical impacts on the ultimate economic impacts of 
the closure. If a closure requirement reduces exploration and 
development activity in the GOM, the displaced capital expenditures 
would likely shift to the next-lowest-cost opportunities promising the 
greatest development potential. Given that oil is produced and sold in 
a global market, the next-lowest-cost areas may be elsewhere within the 
GOM, but also may be international locations. To the extent that 
substitute areas are outside of the GOM but within the United States, 
national-level impacts of the closure areas would likely be limited. 
However, to the extent that industry moves displaced activities outside 
of the United States, national-level impacts associated with industry 
income and employment could be substantial. Recent levels of leasing 
and drilling activity in the CPA indicate that the closure areas 
considered are among the most productive in the entire GOM.\11\ Given 
this, it is less likely that other GOM areas will offer equivalent 
alternative opportunities. As a result, the analyzed area closures have 
greater potential to reduce domestic oil and gas production, industry 
income, and related regional employment opportunities.
---------------------------------------------------------------------------

    \11\ Leases within the closure areas considered within the 
Central Planning Area accounted for approximately 50 percent of 
total oil production in the GOM between 2012 and 2016 and 24 percent 
of total gas production. Existing reserves within the closure areas 
represent 57 percent of estimated oil reserves and 37 percent of 
estimated gas reserves in the GOM.
---------------------------------------------------------------------------

    NRDC asks NMFS to conduct analyses that cannot be supported by 
existing data. Further, NRDC asks NMFS to speculate as to the impacts 
of restricting exploration activity outside the development of existing 
leases. However, such a restriction, while less impactful than a 
complete area-wide restriction, would necessarily foreclose the ability 
of both the government and industry to assess fair market value of 
leases already planned for sale. While NMFS believes that the evaluated 
restriction area would be beneficial for sperm whales and beaked 
whales, such restrictions are at this time not practicable. NRDC does 
not provide any information contradicting this conclusion, and provides 
no specific, viable alternatives for NMFS' evaluation.
    Comment: NRDC states that NMFS should consider time-area closures 
for additional species.
    Response: NMFS did consider habitat-based protections for species 
additional to those discussed in the time-area restrictions section of 
Mitigation. For all affected species, NMFS evaluated the environmental 
baseline (i.e., other population-level stressors), the nature and 
degree of effects likely to be the result of the specified activities, 
and the information available to support the development of appropriate 
time-area restrictions. NMFS determined that the available information 
supported development of the measures described in the notice of 
proposed rulemaking for the Bryde's whale, sperm whales, and beaked 
whales. For other species, context does not justify additional 
protections and/or the available information does not support the 
designation of any specific area for protection, when considered in 
combination with practicability concerns.
    NRDC asserts that ``marine mammal populations in the northern Gulf 
of Mexico can no longer be considered by the agency to be too `data 
poor' or broadly distributed to justify specific mitigation measures 
for their protection, including time-area closures.'' This is not a 
representation NMFS made in the notice of proposed rulemaking. NRDC 
then erroneously claims that NMFS ``limits its analysis to two deep-
diving species, sperm whales and beaked whales [. . .]. In doing so, 
however, it omitted other populations whose conservation status or 
modeled impacts pose particular concern.'' First, NMFS did conduct core 
abundance analyses for all GOM stocks. Second, NRDC declines to 
elaborate on which stocks they believe ``pose particular concern,'' 
other than noting that Kogia spp. may be subject to Level A harassment. 
However, despite NRDC's statement that species can no longer be 
considered to be too broadly distributed to justify specific time-area 
mitigation measures, our core abundance analysis for Kogia spp. shows 
exactly that. Based on the Roberts et al. (2016) models, the two 
species are broadly distributed in shelf-break waters essentially 
throughout the GOM, and there is no identified biologically important 
area or specific bathymetric feature that would allow us a more refined 
understanding of an area suitable for protection (if it were 
warranted). NRDC does not suggest any specific area for protection of 
Kogia spp.
    NRDC also suggests that NMFS should prohibit seismic activity in 
the Flower Garden Banks National Marine Sanctuary (FGBNMS) but offers 
no strong justification other than stating that marine mammals occur 
there. In addition, BOEM and/or BSEE will consult with NOAA's Office of 
National Marine Sanctuaries when they receive an application that 
indicates that survey activity may occur within or near the FGBNMS.
    Overall, NRDC offers no useful recommendation as to the designation 
of protections for additional species. NMFS' consideration of habitat-
based protections was conducted appropriately in light of relevant 
information regarding the environmental baseline, expected effects of 
the specified activities, and information regarding species use of the 
GOM.
    Comment: Several commenters recommended establishing wider buffer 
zones around the proposed time-area closures. The Associations state 
that no buffers should be required around any time-area restriction (if 
required; the Associations also disagree that any restrictions should 
be required, as discussed previously).
    Response: NRDC indicates that NMFS' stated objective in 
establishing the proposed buffer zones around time-area restrictions 
was unclear in terms of evaluating the proposed buffer zone relative to 
the objective. The stated objective was to exclude noise that is likely 
to result in harassment, which NMFS interpreted to mean site-specific 
modeled distances to the 160-dB isopleth (i.e., 50 percent midpoint of 
the Level B harassment risk probability function). Following review of 
public comments, NMFS provides further context here regarding the 
multi-step Level B harassment risk function employed for purposes of 
evaluating modeled noise exposures.

[[Page 5376]]

    With regard to the establishment of a buffer zone, NMFS agrees with 
certain commenters that it is generally appropriate to buffer an area 
to be avoided by some degree, as discussed in the notice of proposed 
rulemaking. However, we disagree that a buffer must be developed to 
fully eliminate the potential for Level B harassment, as some 
commenters may have inferred from our use of the distance to the 160-dB 
isopleth (i.e., historically used as a 100 percent single-step function 
for evaluation of Level B harassment; here the 50 percent midpoint of 
the Level B harassment risk function). Rather, the buffer concept, as 
described in the notice of proposed rulemaking, serves to reasonably 
minimize the extent and severity of what limited harassment may occur 
as a result of acoustic exposure to relatively low received levels of 
noise.
    The Associations asserted that NMFS did not consider the use of 
buffer areas in the practicability analyses and provides no biological 
basis for including buffers. We disagree. As noted earlier, the RIA 
analysis (which forms a substantial part of the practicability analysis 
for these measures) includes analysis of the economic impacts of the 
time-area restrictions inclusive of the buffer. As noted above, the 
logical biological rationale is to provide a buffer around an area 
determined to be of particular biological importance such that the 
effects of noise from outside the restriction area intruding within the 
area is minimized.
    However, BOEM's update of the geographic scope for this rule 
eliminates the need for proposed time-area restrictions #3 and 4 (i.e., 
the Bryde's whale core habitat area and the ``Dry Tortugas area'' 
designed for protection of beaked whales and sperm whales). Therefore, 
comments addressing the proposed buffers for those areas are no longer 
relevant. Regarding the coastal bottlenose dolphin restriction (Area 
#1), NMFS has determined that the addition of a buffer to this area is 
not warranted, based on the objectives of the restriction (described in 
detail in a previous response to comment) and on the manner in which 
the area was delineated. Areas #3 and 4 were delineated based on NMFS' 
review of the available scientific information and expert opinion and 
in order to denote areas expected to be of particular biological 
importance for particular species. In contrast, the coastal dolphin 
restriction area was based simply on the stock boundaries for coastal 
bottlenose dolphins (i.e., the seaward extent of the area is set at the 
20-m isobath). As this boundary does not mark an area of specific 
biological importance or high density for the stock, but is rather an 
approximation of stock presence, NMFS has determined following review 
of public comments, in which valid practicability concerns were raised, 
that the inclusion of a buffer to this area is not warranted.
    Comment: Noting that the proposed ITR included requirements to 
conduct visual monitoring following conclusion of active shooting, the 
MMC recommends that NMFS require operators to also continue conducting 
acoustic monitoring following conclusion of active shooting.
    Response: The proposed ITR stated that acoustic monitoring must 
occur for 30 minutes prior to and during all active firing of airguns 
for deep penetration surveys, but was silent on the issue of acoustic 
monitoring following the survey. However, visual monitoring is required 
to continue for 60 minutes following cessation of survey activity 
during good visibility. NMFS agrees with the MMC that ``both visual and 
acoustic monitoring should occur concurrently, as acoustic detections 
can provide additional information not readily available via visual 
detections alone regarding changes in foraging and social behavior 
during survey activities and after activities cease.'' Accordingly, 
acoustic monitoring is also required to continue following cessation of 
survey activity for a period of 60 minutes.
    Comment: BP comments that they welcome use of industry standard 
PAM/operator software such as PAMGuard. Noting the operational 
challenges associated with accommodating increased numbers of PSOs on 
survey vessels, BP also comments that they would welcome the inclusion 
of an option to implement PAM during survey activities using remote 
shore-based operators.
    Response: NMFS agrees that this may be appropriate, depending on 
various factors. While we are not currently aware of the state of 
existing technology towards achieving this end, NMFS would consider the 
use of remote PAM monitoring, assuming reliability and the ability to 
achieve the same performance as shipboard PAM monitoring. NMFS believes 
the adaptive management process will be an appropriate venue for 
further consideration of this approach.
    Comment: BP comments that, while they recognize the value of 
prospective formal standards for PAM operations (e.g., hardware, 
software, training), the standards have not yet been finalized. BP 
requests that the standards be considered for use only after an initial 
draft has been circulated via relevant standards development and 
issuance processes.
    Response: NMFS may adopt elements of the prospective standards, as 
it deems appropriate (as discussed in Monitoring and Reporting). 
However, we agree that wholesale adoption of the standards would not be 
appropriate until appropriate review and other necessary processes are 
complete.
    Comment: Industry commenters state that non-airgun high-resolution 
geophysical (HRG) surveys should not be subject to pre-clearance and 
shutdown requirements. Relatedly, BP and Chevron comment that exclusion 
zones should not be required for HRG surveys, as these surveys 
typically operate using acoustic sources deployed on an automated 
underwater vehicle (AUV) running 40 m above the seafloor. Therefore, 
they state that there is no environmental benefit to a requirement for 
a surface exclusion zone.
    Response: The Associations note that the acoustic footprint of 
sources typically used in non-airgun HRG surveys are too small to 
warrant the proposed exclusion and buffer zone distances and that, more 
importantly, due to the typically highly directional nature of these 
acoustic sources, animals observed at the surface will generally not be 
exposed to the signal. NMFS agrees with these comments, and notes that 
the proposed shutdown and exclusion zone requirements were offered in 
accordance with BOEM's HRG survey protocols (Appendix B of BOEM, 2017). 
Following review of these comments, as well as the available scientific 
information regarding the typical interaction of these signals with the 
environment and likely lack of efficacy of typical standard operational 
protocols developed for omnidirectional sources, NMFS has eliminated 
these requirements. However, we also clarify that certain 
electromechanical sources may be subject to the pre-clearance and 
shutdown requirements associated with shallow penetration surveys. In 
addition, the exclusion and buffer zone distances for shallow 
penetration surveys have been reduced (while adding an extended 
distance shutdown zone for certain circumstances) in recognition of the 
typically smaller harassment zones associated with use of the acoustic 
sources considered here to be used in shallow penetration surveys.
    As noted here, NMFS has eliminated the requirement for 
implementation of an exclusion zone during HRG surveys. We also agree 
with BP's comment that exclusion zones should not be required for 
surveys using an AUV-deployed acoustic source running at short 
distances above the seafloor.
    Comment: NRDC states that NMFS fails to prescribe adequate 
mitigation for

[[Page 5377]]

HRG surveys and, relatedly, that NMFS must not issue LOAs for use of 
lower-frequency multibeam echosounders (MBES).
    Response: As evidenced by the previous comment response, in which 
describing elimination of certain mitigation measures that were 
proposed for HRG surveys, NMFS disagrees with NRDC. NRDC provides no 
reasonable justification for the recommendation to consider additional 
mitigation requirements. They reference the 2008 Madagascar stranding 
of melon-headed whales, implying that a similar occurrence may be a 
reasonably anticipated outcome of HRG survey work in the GOM. Although 
it is correct that an investigation of the event indicated that use of 
a high-frequency mapping system (12-kHz MBES) was the most plausible 
and likely initial behavioral trigger of the event (with the caveat 
that there was no unequivocal and easily identifiable single cause), 
the panel also noted several site- and situation-specific secondary 
factors that may have contributed to the avoidance responses that led 
to the eventual entrapment and mortality of the whales (Southall et 
al., 2013). Specifically, regarding survey patterns prior to the event 
and in relation to bathymetry, the vessel transited in a north-south 
direction on the shelf break parallel to the shore, ensonifying deep-
water habitat prior to operating intermittently in a concentrated area 
offshore from the stranding site. This may have trapped the animals 
between the sound source and the shore, thus driving them towards the 
lagoon system. Shoreward-directed surface currents and elevated 
chlorophyll levels in the area preceding the event may also have played 
a role.
    The relatively lower output frequency, higher output power, and 
complex nature of the system implicated in this event, in context of 
the other factors noted here, likely produced a fairly unusual set of 
circumstances that indicate that such events would likely remain rare 
and are not necessarily relevant to use of more commonly used lower-
power, higher-frequency systems such as those evaluated for this 
analysis. The risk of similar events recurring is expected to be very 
low, given the extensive use of active acoustic systems used for 
scientific and navigational purposes worldwide on a daily basis and the 
lack of direct evidence of such responses previously reported. The only 
report of a stranding that may be associated with this type of sound 
source is the one reported in Madagascar.
    NRDC also references Cholewiak et al. (2017), stating that 
virtually no beaked whale vocalizations were detected acoustically 
during the time that the shipboard echosounder was operational. NRDC 
mischaracterizes the literature, including a speculative description of 
what they imagine the beaked whales were doing while not vocalizing 
(``suggesting that the whales broke off their foraging behavior and 
engaged in [. . .] silent flight''). Cholewiak et al. (2017) do 
describe finding that beaked whales were significantly less likely to 
be detected acoustically while echosounders were active. However, it is 
not clear that this response should be considered as Level B harassment 
when considered in the context of what is likely a brief, transient 
effect, given the mobile nature of the surveys and the fact that some 
beaked whale populations are known to have high site fidelity. In 
support of this conclusion, Quick et al. (2017) describe an 
experimental approach to assess potential changes in short-finned pilot 
whale behavior during exposure to an echosounder. Tags attached to the 
animals recorded both received levels of noise as well as orientation 
of the animal. Results did not show an overt response to the 
echosounder or a change to foraging behavior of tagged whales, but the 
whales did increase heading variance during exposure. The authors 
suggest that this response was not a directed avoidance response but 
was more likely a vigilance response, with animals maintaining 
awareness of the location of the echosounder through increased changes 
in heading variance (Quick et al., 2017). Visual observations of 
behavior did not indicate any dramatic response, unusual behaviors, or 
changes in heading, and cessation of biologically important behavior 
such as feeding was not observed. More recently, Varghese et al. (2020) 
reported the results of an investigation of the effects of a 12-kHz 
MBES system on beaked whale foraging behavior off of California. 
Echolocation clicks from Cuvier's beaked whales were detected and 
classified into foraging events called group vocal periods (GVP), and 
compared across exposure periods before, during, and after MBES 
activity. Of the metrics used to assess beaked whale foraging behavior, 
only the number of GVPs per hour was statistically different during 
MBES activity versus a non-MBES period. GVPs per hour increased during 
MBES activity compared with before MBES activity, demonstrating that 
beaked whales did not stop foraging and were not displaced by the 
activity. These results suggest that there was not a negative impact of 
MBES activity on foraging behavior of this sensitive species (Varghese 
et al., 2020).
    Finally, NRDC references the work of Deng et al. (2014) and Hastie 
et al. (2014) in describing ``leakage'' of ``substantial noise'' at 
frequencies within marine mammal hearing range during use of active 
acoustic systems that are operated at higher frequencies. The 
referenced studies reported some behavioral reaction by marine mammals 
to acoustic systems operating at user-selected frequencies above 200 
kHz. This work was discussed in the notice of proposed rulemaking. In 
general, the referenced literature indicates only that sub-harmonics 
could be detectable by certain species at distances up to several 
hundred meters. As NMFS has noted elsewhere, behavioral response to a 
stimulus does not necessarily indicate that Level B harassment, as 
defined by the MMPA, has occurred. Source levels of the secondary peaks 
considered in these studies--those within the hearing range of some 
marine mammals--mean that these sub-harmonics would either be below the 
threshold for Level B harassment or would attenuate to such a level 
within a few meters. The work cited by the commenters is consistent 
with previously observed occurrences of sub-harmonics. Essentially, the 
first sub-harmonic's source level (e.g., if the primary frequency is 
200 kHz, the first sub-harmonic is 200/2 or 100 kHz, the second is 200/
3 or 66.7 kHz) is at least 20-30 dB less than the primary frequency's 
source level, with each subsequent sub-harmonic's source level 
decreasing rapidly from there. These sub-harmonics are typically so 
reduced in source level that, for most side-scan and multi-beam sonar 
systems, they are not strong enough to produce impacts beyond tens of 
meters from the source (distances at which reactions to the vessel 
itself are likely to supersede reactions to an acoustic signal). 
Additionally, for any potential impacts to occur, an animal must be 
within this range and within the very narrow beams produced by the 
systems (for these sub-harmonic frequencies).
    In addition, recent sound source verification testing of these and 
other similar systems did not observe any sub-harmonics in any of the 
systems tested under controlled conditions (Crocker and Fratantonio, 
2016). While this can occur during actual operations, the phenomenon 
may be the result of issues with the system or its installation on a 
vessel rather than an issue that is inherent to the output of the 
system. As concluded in the notice of proposed rulemaking, there is no 
evidence to suggest that Level B harassment of marine mammals should be 
expected in

[[Page 5378]]

relation to use of active acoustic sources at frequencies exceeding 180 
kHz. NRDC's comments did not address NMFS' prior statements regarding 
this topic.
    NRDC fails to adequately support the claims of harm to marine 
mammals that are reasonably likely to occur as a result of HRG surveys 
and, thus, fails to justify their recommendation for enhanced 
mitigation. The recommended measures include ``extended safety zone and 
monitoring requirements'' and a ``bar on nighttime operations.'' Even 
when animals are receiving echosounder signals, they may not be 
harassed, as described above. However, given the directional nature of 
these sources, animals observed at the surface will almost certainly 
not be within the acoustic beam, thus negating the benefit of 
detection-based measures such as shutdowns. Any exposure to the 
echosounder would likely be only in the ensonified cone below the 
vessel, and responses to the vessel itself at such close ranges would 
influence likelihood of acoustic exposure. The package of active 
acoustic systems modeled as representative of a typical HRG survey 
included a 200-kHz echosounder. Regarding the suggestion that this bars 
use of any system with a lower frequency output, NMFS disagrees. NMFS' 
analysis also includes use of different lower-frequency sources (i.e., 
single airguns and boomers). Moreover, the specific sources selected 
for analysis do not limit the actual sources that may be used, assuming 
the actual sources are reasonably similar to the full suite of analyzed 
sources, as is the case here.
    Comment: The Associations and other industry commenters claim that 
the proposed PSO staffing requirements compromise personnel safety, 
cannot be effectively implemented, and are unnecessary and unsupported.
    Response: In the notice of proposed rulemaking, NMFS described in 
detail the importance of detection-based mitigation as a component of 
standard operational mitigation protocols. Detection-based mitigation 
cannot occur effectively without both visual and acoustic monitoring, 
with the latter being the only effective method of detection during 
periods of poor visibility or at any time for cryptic species (e.g., 
beaked whales) or species with high availability bias (e.g., sperm 
whales). Therefore, visual monitoring is required during daylight hours 
and acoustic monitoring is required throughout the period of survey 
operations. When these monitoring techniques are required, two visual 
PSOs must be on duty in order to effectively monitor 360 degrees around 
the vessel, communicate with the operator as necessary, and record 
data, and an acoustic PSO must be on duty to monitor the PAM system. In 
order to effectively carry out monitoring duties, PSOs must have 
sufficient periods of rest to minimize fatigue that would compromise 
their performance. Based on these considerations, and in consideration 
of the literature relating to mitigation and monitoring requirements 
and standard practice for scientific surveys, NMFS proposed minimum 
duty requirements.
    While NMFS agrees that there is likely to be some increased 
logistical burden associated with these requirements, which are 
expanded to some degree from current practice in the GOM in the absence 
of compliance with the MMPA, the Associations do not demonstrate that 
this burden is so large as to be impracticable. Similarly, they do not 
provide information supporting claims that these requirements would 
compromise personnel safety (and certainly do not support the claim 
that the requirements are ``unnecessary and unsupported''). The 
Associations' comment states that survey vessels are typically at 
maximum capacity. NMFS acknowledges that in some cases, increased PSO 
staffing may result in a need for operators to balance staffing in 
other areas, such as in the seismic crew (25 to 30) or the three to 
seven client representatives that the Associations state are typically 
aboard a survey vessel, in order to accommodate necessary PSO staffing 
while not exceeding a vessel's maximum capacity. However, assuming that 
a vessel's maximum capacity is not exceeded, the claim that increasing 
the number of people aboard necessarily increases ``the risk of 
injuries, illnesses, and evacuation for medical reasons'' is 
unsupported. The comment is inconsistent regarding the number of PSO 
staff that the requirements would add, at various places stating that 
the requirements would result in the addition of six to eight or three 
to five PSO staff. Overall, the Associations state that only three (and 
possibly up to four) PSOs should be allowed, without explaining how 
this may achieve the objective of the proposed detection-based 
mitigation requirements.
    However, in recognition of the likely increase in logistical burden 
and the possibility that individual LOA applicants may be able to 
demonstrate legitimate practicability issues, NMFS allows for the 
potential that an exception may be obtained specifically for the 
requirement that PSOs may be on duty for a maximum period of two hours, 
followed by a minimum period of one hour off. If an exception is 
granted based on practicability, the historical practice of a maximum 
on-duty period of four hours, followed by a minimum period of two hours 
off, would be substituted.
    Comment: The Associations and other industry commenters comment 
that the proposed requirement for visual observation before and during 
nighttime ramp-ups would be ineffective and potentially present safety 
concerns.
    Response: NMFS agrees that reduced efficacy should be expected for 
visual monitoring at night and, in consideration of comments asserting 
that this may present a safety concern, we have eliminated this 
requirement (noting that passive acoustic monitoring is still required 
for all nighttime operations of large airgun arrays). NMFS also agrees 
with the Associations' comment that employment of a PSO for the 
dedicated purpose of documenting entanglements with ocean-bottom node 
(OBN) cables is unnecessary and has eliminated this requirement. 
Elimination of these requirements is expected to help somewhat in 
alleviating the logistical concerns expressed by the Associations.
    Comment: The Associations suggest that entanglement avoidance 
requirements should be removed from the ITR. The MMC comments that they 
support these requirements, and that the requirements are consistent 
with best management practices developed for avoiding entanglements.
    Response: The Associations' comment, offered only in a footnote, is 
unclear as to whether the Associations' suggestion is to remove all 
entanglement avoidance requirements or only the requirement to use 
negatively buoyant coated wire-core tether cable. (Note that NMFS does 
agree with the suggested elimination of a requirement for use of a 
dedicated PSO for purposes of documenting entanglement.) Regardless, 
the Associations' suggestion that this requirement should be removed is 
keyed only to concern regarding practicability. NMFS disagrees that 
this requirement is impracticable, and the Associations offer no 
information to the contrary. Moreover, this measure is designed to 
prevent serious injury or mortality, which cannot be authorized under 
this rule.
    Here, no mortality was requested or proposed for authorization and, 
therefore, potential for death by entanglement must be avoided. There 
is demonstrated potential for entanglement of protected species in 
association with OBN survey operations. As described in the notice of 
proposed rulemaking, a GOM OBN

[[Page 5379]]

operator remarkably entangled three different protected species within 
a year--including an Atlantic spotted dolphin, as well as an ESA-listed 
turtle and a manta ray. BSEE subsequently issued two enforcement 
actions against the operator for incidents of non-compliance, 
indicating that it is appropriate to be stringent regarding 
requirements relating to entanglement avoidance. Specific appropriate 
measures were determined in consultation between NMFS, BOEM, and BSEE, 
including consultation with NMFS' gear engineering experts, and were 
subsequently included in permits issued by BOEM (e.g., OCS Permit L17-
009, issued July 11, 2017). NMFS proposed these specific measures for 
this ITR and no comments offering useful suggestions regarding 
potential modifications to the measures were received. A generic 
suggestion that no entanglement avoidance requirements are necessary is 
not credible.
    Comment: NRDC claims that NMFS fails to consider mitigation to 
reduce ship strike, particularly within Bryde's whale habitat. 
Separately, NRDC states that NMFS should consider extending ship-speed 
requirements to all project vessels. The Associations state that vessel 
strike avoidance measures should not be required, or that there should 
be modifications and/or exemptions to the measures.
    Response: NMFS disagrees with NRDC's contention. NMFS' required 
vessel strike avoidance protocol is expected to further minimize any 
potential interactions between marine mammals and survey vessels, 
relative to the already low likelihood of vessel strike in relation to 
the activities considered herein. Please see ``Vessel Strike 
Avoidance'' for a full description of requirements, which include: 
Vessel operators and crews must maintain a vigilant watch for all 
marine mammals and must take necessary actions to avoid striking a 
marine mammal; vessels must reduce speeds to 10 kn or less when mother/
calf pairs, pods, or large assemblages of cetaceans are observed near a 
vessel; and vessels must maintain minimum separation distances.
    We also note that NRDC's comment that ``vessels supporting the 
seismic operation are not similarly constrained'' is in error. All 
project vessels are required to adhere to vessel strike avoidance 
requirements, including speed requirements in certain circumstances. As 
stated clearly in the proposed and final regulatory text, ``[v]essel 
operators and crews must maintain a vigilant watch for all marine 
mammals and slow down or stop their vessel or alter course, as 
appropriate and regardless of vessel size, to avoid striking any marine 
mammal [. . .].'' Regarding whether ship speed requirements are 
warranted for all project vessels in the Bryde's whale core habitat 
area to minimize risk of strike for Bryde's whales, the consideration 
is no longer relevant to this rule, as activity within the Bryde's 
whale core habitat area referenced by NRDC can no longer be considered 
through this rule following BOEM's update to the scope of action. 
Further, we disagree that similar risk exists for sperm whales and 
beaked whales, as would be necessary to warrant vessel speed 
restrictions in the Mississippi Canyon area. There are very few records 
of vessel strikes for sperm whales, as compared with mysticete whales 
in general, and Bryde's whales' dive behavior in particular makes them 
potentially more susceptible to vessel strike.
    The Associations' comments state that they are not aware of any 
incidence of ship strike associated with a geophysical survey, implying 
that no strike avoidance measures are necessary. The lack of recorded 
incidents of strike does not mean that none have occurred and, more 
importantly, does not mean that none will occur. Therefore, it is NMFS' 
responsibility to prescribe measures that will achieve the least 
practicable adverse impact via avoidance of vessel strike. The comments 
go on to assert that there is ``no evidence'' that strike avoidance 
measures benefit marine mammals, despite the wealth of scientific 
evidence described in the notice of proposed rulemaking that slower 
vessel speeds result in fewer strikes and, if a strike does occur, a 
significantly lower likelihood of mortality. Separation requirements 
are common sense--a moving vessel should maintain some minimum distance 
from a whale to avoid striking it--and are similar to generic strike 
avoidance guidelines found elsewhere. The comments implying that these 
requirements are unwarranted and burdensome are unpersuasive, 
particularly given that BOEM has required essentially identical strike 
avoidance measures in the GOM via notices to lessees for many years 
(currently, via BOEM NTL No. 2016-G01).
    Nevertheless, NMFS recognizes that there are legitimate concerns 
regarding vessels towing gear and human safety issues. We have 
clarified in the strike avoidance measures that vessel strike avoidance 
requirements do not apply in any case where compliance would create an 
imminent and serious threat to a person or vessel or to the extent that 
a vessel is restricted in its ability to maneuver and, because of the 
restriction, cannot comply.
    Comment: The Associations and other industry commenters raise 
concerns regarding the PSO requirements, including that NMFS' 
requirements for PSOs will result in labor shortages, and make an 
accompanying recommendation that these be ``preferred'' training 
requirements that LOA-holders would not have to meet.
    Response: NMFS disagrees with the comments. The Associations' 
statement that requirements for PSOs to have bachelor's degrees or to 
satisfy a positive experience requirement are ``difficult, if not 
impossible, to achieve'' is not persuasive. As explained in the notice 
of proposed rulemaking, NMFS discussed these PSO requirements with BSEE 
and with third-party observer providers. Both parties indicated that 
the requirements should not be expected to result in any labor 
shortage. We pointed out that during a period when a significantly 
greater amount of survey activity was occurring in the GOM than at 
present (i.e., with as many as 30 source vessels), requirements similar 
to those proposed did not result in any labor shortage. Moreover, NMFS 
specifically requested comment on the assumption that the requirements 
would not result in any labor shortage. The Associations' expressed 
concern regarding the potential for a labor shortage, but do not 
provide any specific information to support the claims. We also clarify 
that not all PSOs must have a minimum of 90 days at-sea experience, 
with no more than 18 months elapsed since the conclusion of the most 
recent relevant experience. As described in the notice of proposed 
rulemaking and herein, a minimum of one visual PSO and two acoustic 
PSOs must have such experience (rather than all PSOs).
    Comment: The Associations provide a list of specific proposed 
monitoring and reporting requirements that they assert are unreasonable 
or otherwise problematic. Some of these comments are echoed by other 
industry commenters.
    Response: We address the specific issues raised by the Associations 
in turn.
    1. The Associations state that bigeye binoculars should not be 
required, because they are expensive, require installation on the 
vessel, and are not appropriate for monitoring of the exclusion zone.
    NMFS disagrees with this comment. While it is correct that 
procurement of bigeye binoculars will incur costs, these costs were 
analyzed in NMFS' RIA.

[[Page 5380]]

While bigeye binoculars may not be an individual PSO's tool of choice 
for observing marine mammals at close range to the vessel, they are an 
indispensable tool for observing marine mammals at greater distance 
upon initial detection, are a standard component of marine mammal 
observation (for scientific purposes, but also as a part of standard 
mitigation monitoring conducted aboard surveys for which incidental 
take authorizations are issued), and will be helpful in more accurately 
identifying animals at greater distances, such that the precautionary 
shutdowns of concern to the Associations are avoided.
    2. The Associations state that PSOs should not report on factors 
that may be contributing to impaired observations, as such reporting 
may be speculative, unverifiable, and/or incorrect.
    NMFS disagrees with this comment. Reporting on such conditions is 
not connected to any requirement for action, but it is important to 
understand whether visual observation is able to be conducted in an 
effective fashion, whether it be due to weather conditions or to 
conditions on the vessel.
    3. The Associations suggest that the reporting requirement to 
estimate numbers of animals observed by cohort is overly complicated, 
and that the rule should require only recording of juveniles and 
adults.
    NMFS agrees with this comment and has made corresponding edits to 
the regulatory text.
    4. The Associations express some confusion regarding language 
addressing the information that visual PSOs should be compiling on a 
daily basis and whether these daily ``reports'' include estimates of 
actual animals taken.
    NMFS clarifies that the language cited by the Associations was not 
intended to mean that PSOs should be estimating ``takes'' on a daily 
basis, and confirm that the Associations' statement that such 
information should be included only in annual reporting is correct.
    5. Regarding NMFS' consideration of an approach recommended by the 
MMC to produce estimates of actual take from observations of animals 
during survey effort, the Associations express concern about the 
appropriate application of this process, and suggest that the protocol 
be applied at the end of a period long enough to accumulate sufficient 
data to adequately evaluate the appropriateness and proper application 
of the process as part of the adaptive management process.
    NMFS shares many of the Associations' concerns on this subject and 
regarding the specific methodology proposed by the MMC. NMFS looks 
forward to working with the Associations (as well as BOEM and BSEE) 
towards the development of appropriate methods through the adaptive 
management process.
    6. In reference to the requirement for the lead PSO to submit to 
NMFS a statement concerning mitigation and monitoring implementation 
and effectiveness, CGG adds that, because there is a lead PSO on each 
offshore rotation, the LOA-holder should submit collated statements.
    NMFS agrees that this may be a more practical approach.
    Comment: The MMC recommends that NMFS require LOA-holders to 
implement electronic reporting systems for field-based PSO data entry 
and expedited reporting.
    Response: NMFS agrees that this would be appropriate and would 
better ensure expedited field entry and quality control checking of PSO 
data, as well as facilitate data transfer, quality control, data 
analysis, and automated report generation. Overall, such a requirement 
is helpful to ensure the efficient synthesis of data, as required by 
the comprehensive reporting process.
    Comment: The Associations express support for NMFS' proposed 
approach to comprehensive monitoring and development of a structured 
adaptive management process, and highlight their support for efforts 
that improve the quantity and quality of information related to 
determining the nature and magnitude of the potential effects of 
offshore geophysical activities on marine mammals, including industry-
supported independent third-party research.
    Response: NMFS appreciates the comments and looks forward to 
continued engagement with the regulated community, as well as BOEM and 
BSEE, to improve the collection and use of the best available science 
consistent with the requirements and limits of the MMPA.
    Comment: The MMC comments that they support an annual adaptive 
management process for the issuance of LOAs in the GOM and recommend 
that they be included in the process along with representatives from 
BOEM, BSEE, and industry.
    Response: NMFS appreciates the comments and will ensure that the 
adaptive management process includes participation of the parties 
noted, where appropriate.
    Comment: NRDC asserts that NMFS fails to prescribe requirements 
sufficient to monitor and report takings of marine mammals. The MMC 
recommends that NMFS and BOEM work together to develop a coordinated 
long-term monitoring and research plan, and further recommends that, to 
facilitate the completion of the plan, NMFS and BOEM establish a GOM 
scientific advisory group, composed of agency and industry 
representatives and independent scientists, to assist in the 
identification and prioritization of monitoring needs and hypothesis-
driven research projects to better understand the short- and long-term 
effects of geophysical surveys on marine mammals in the GOM. Commenters 
also noted that there are many research gaps that need to be filled and 
suggested that NMFS should include monitoring requirements that fill 
those gaps.
    Response: Section 101(a)(5)(A) of the MMPA indicates that any 
regulations NMFS issues shall include ``requirements pertaining to the 
monitoring and reporting of such taking.'' This broad requirement 
allows for a high degree of flexibility in what NMFS may accept or 
include as a monitoring requirement, but is not specific in identifying 
a threshold of what should be considered adequate monitoring. Contrary 
to NRDC's comments, except for IHAs in Arctic waters, NMFS' 
implementing regulations do not provide a specific standard regarding 
what required monitoring and reporting measures ``must'' accomplish. 
However, NMFS' implementing regulations require incidental take 
applications to include suggested means of accomplishing the necessary 
monitoring and reporting that will result in increased knowledge of the 
species, the level of taking, or impacts on populations of marine 
mammals that are expected to be present while conducting activities, as 
well as suggested means of minimizing burdens by coordinating such 
reporting requirements with other schemes already applicable to persons 
conducting such activity. 50 CFR 216.104(a)(13). The comment extracts 
pieces of this language to suggest that future LOA applicants are 
required to coordinate with each other's monitoring efforts, ignoring 
the fact that the relevant regulation points to this coordination only 
in support of minimizing the burden on the applicant and that it refers 
to coordination with ``schemes already applicable to persons conducting 
such activity.'' 50 CFR 216.104(a)(13). NRDC attempts to further the 
argument that coordination across projects is required by statute by 
referencing a monitoring plan that they state is in development by 
BOEM. The MMC also references development of a ``long-term monitoring 
plan'' that they

[[Page 5381]]

attribute to BOEM. NMFS is not aware that any such monitoring plan has 
been developed and, therefore, such a plan is not ``already applicable 
to persons conducting such activity.''
    NRDC discusses a litigation settlement agreement related to the 
activities that are the subject of this rule, stating that ``BOEM must 
analyze `the development of a long-term adaptive monitoring plan that 
addresse[s] cumulative and chronic impacts from seismic surveys on 
marine mammal populations in the Gulf of Mexico.' '' NRDC et al. v. 
Bernhardt et al., 2:10-cv-1882, ECF No. 118 (E.D. La. June 18, 2013). 
NRDC also cites BOEM's PEIS in discussing this plan. That requirement 
in the settlement agreement does not pertain to NMFS' statutory 
authority under the MMPA, which does not provide authority for NMFS to 
require the development of a ``long-term monitoring plan'' via the 
promulgation of ITRs or as a condition of an incidental take 
authorization. As noted above, NMFS' statutory authority is to 
prescribe ``requirements pertaining to the monitoring and reporting of 
such taking.'' Although applicants that anticipate the need for 
consecutive periods of five-year regulations to cover ongoing 
activities may develop monitoring and reporting plans that extend past 
the five-year effectiveness period of a rule, section 101(a)(5)(A) 
requires only monitoring and reporting to cover the specified 
activities undertaken during the period of the rule. Were a long-term 
monitoring plan to be developed by BOEM, it would therefore be a 
voluntary undertaking on the part of participants, rather than a 
requirement under the MMPA. While certainly an exemplar of what a 
strong comprehensive monitoring plan can look like, the U.S. Navy's 
Integrated Comprehensive Monitoring Program (ICMP), which NRDC 
references as a relevant analogue to the monitoring plan that they 
assert is required in the GOM to satisfy the requirements of the MMPA, 
should not be hailed as a model that should always be copied or a 
standard that must be achieved for all MMPA ITRs. The Navy's ICMP was 
developed in close coordination with NMFS and reflects several factors 
that are not present for all ITRs (including these regulations) and 
that lay the groundwork for what is an exceptionally comprehensive 
program. Specifically, as the single entity for which take is 
authorized and that has the responsibility for implementing a 
monitoring program, the Navy has an existing organizational and funding 
structure that can support a truly integrated and comprehensive plan 
that would be far more difficult under a rule allowing for 
authorization of take by disparate applicants with varying activity 
levels, resource availability, and familiarity with regulatory 
requirements and marine mammal issues. Also, the Navy has an 
independent environmental stewardship mandate that influences their 
monitoring approach and supports a robust program intended to work in 
concert with the work funded through their Office of Naval Research and 
Living Marine Resources programs to create essentially full coverage of 
the science necessary to support vigorous environmental assessment and 
compliance across all Navy actions. Last, Navy training and testing 
utilize a large variety and number of platforms and sound sources, many 
of which can result in the take of marine mammals but cannot be 
monitored at the source. Accordingly, the Navy employs the robust, 
problem-based, often off-site monitoring program currently in place in 
order to answer targeted questions with controlled studies.
    Although NMFS' authority with regard to the prescription of 
monitoring requirements does not include mandating long term 
monitoring, the MMPA does require an assessment of impacts from the 
total taking by all persons conducting the activity. Thus, meaningful 
monitoring and reporting for a specified activity under section 
101(a)(5)(A) should be designed to help us better understand the total 
taking that is considered for authorization under the regulations for 
all persons conducting the specified activities under the five-year 
regulations. This necessitates coordination across applicants with 
regard to comprehensive analysis and reporting of information collected 
in relation to ``the level of taking or impacts on populations of 
marine mammals that are expected to be present while conducting 
activities.'' 50 CFR 216.104(a)(13). We discuss these comprehensive 
reporting requirements in greater detail in Monitoring and Reporting. 
These requirements are appropriate to the necessary function of 
informing the assessment of the overall impact of the incidental take 
allowable under the regulations and acknowledge the need to conduct 
aggregation and analysis of the data in a manner that directly informs 
the question of whether and the degree to which marine mammal 
populations addressed may be affected by the incidental take authorized 
by LOAs.
    We appreciate the MMC's acknowledgement of the investments made by 
BOEM and industry (via the E&P Sound and Marine Life Joint Industry 
Program) towards better understanding of marine mammal abundance and 
distribution, the characterization of anthropogenic sound sources in 
the GOM, and the effects of sound on marine mammal hearing and 
behavior, among other initiatives. We also note that much of the 
research recommended by NRDC has been conducted via the BOEM-sponsored 
Gulf of Mexico Marine Assessment Program for Protected Species 
initiative. However, while these voluntary efforts are commendable, 
section 101(a)(5)(A) does not require hypothesis-driven, focused 
research pertaining to the impact and mitigation of chronic noise 
exposure on populations of special concern, nor does it require a 
``coordinated long-term monitoring and research plan,'' as expressed by 
the commenters.
    Regarding the MMC's recommendation that NMFS establish a GOM 
``scientific advisory group, composed of agency and industry 
representatives and independent scientists, to assist in the 
identification and prioritization of monitoring needs and hypothesis-
driven research projects,'' NMFS would be willing to explore with the 
MMC the appropriate mechanisms for convening such a group, including 
consideration of the MMC's authorities under the MMPA.
    The monitoring approach described in this preamble includes LOA-
specific monitoring and reporting set forth in the regulations and, 
separately, outlines a framework for potential data collection, 
analysis, research, or collaborative efforts that are not specified in 
these regulations but which work towards satisfying the information 
elements identified in our implementing regulations. NMFS is committed 
to working with industry and BOEM through the adaptive management 
process to ensure that LOA-specific monitoring and reporting will be 
used appropriately to help better understand the impacts of the total 
taking from the specified activity contemplated in this ITR on the 
affected populations, as well as how the more overarching voluntary 
efforts will be identified and carried out.
    Comment: The Associations reiterate their belief that NMFS, as the 
regulating agency, has the responsibility to collect, organize, and 
assess all of the data reported to NMFS under the terms of issued LOAs.
    Response: NMFS disagrees with this comment. The MMPA requires NMFS 
to prescribe regulations setting forth requirements pertaining to the 
monitoring and reporting of ``such

[[Page 5382]]

taking.'' 16 U.S.C. 1371(a)(5)(A)(i)(II)(bb). (In contrast, the other 
required component of activity-specific regulations, relating to 
mitigation requirements, refers to ``taking pursuant to such 
activity.'' 16 U.S.C. 1371(a)(5)(A)(i)(II)(aa). In NMFS' view, these 
monitoring and reporting requirements in our activity-specific 
regulations refer to the total taking from the specified activity as a 
whole, and they are requirements that can be imposed on those entities 
availing themselves of LOAs issued under the activity specific 
regulations. Therefore, it is incumbent upon LOA-holders, collectively, 
to provide this information to NMFS in a reasonably synthesized form 
such that NMFS may adequately assess the effects of the specified 
activity on an ongoing basis. This information may in some cases be 
essential to NMFS' ability to carry out 50 CFR 216.105(e) (``Letters of 
Authorization shall be withdrawn or suspended, either on an individual 
or class basis, as appropriate, if, after notice and opportunity for 
public comment, the Assistant Administrator determines that: (1) The 
regulations prescribed are not being substantially complied with; or 
(2) The taking allowed is having, or may have, more than a negligible 
impact on the species or stock or, where relevant, an unmitigable 
adverse impact on the availability of the species or stock for 
subsistence uses'' (emphasis added).) While NMFS recognizes that the 
Associations are not subject to the ITR (including any reporting 
requirements in the ITR or related LOAs), LOA-holders (many of which 
are likely to be Association members) will collectively be responsible 
for the comprehensive reporting requirements described herein. The 
Associations in their comment commit to participate in the annual 
assessment process, and NMFS welcomes that participation.
    Comment: The MMC recommends that NMFS require industry operators to 
measure and report the horizontal leakage of their various airgun 
arrays and investigate options to minimize horizontal sound leakage 
from those array configurations.
    Response: As stated in the notice of proposed rulemaking, NMFS 
encourages the minimization of unnecessary horizontal propagation. 
However, while the MMC's recommendation would likely lead to a better 
understanding of actual horizontal propagation (or ``leakage'') that 
does occur, it is not clear that the product of such measurements 
(termed ``waste ratios'' by the MMC) would necessarily lead to a viable 
path to reducing such leakage. In addition, the MMC does not specify 
what it recommends as a sufficient amount of data concerning waste 
ratios to allow consideration of a potential threshold. Thus, the 
comment implies that all operators would be required to conduct field 
measurements of the acoustic output of airgun arrays under this 
recommendation, which NMFS believes would not be practicable. NMFS 
appreciates the comment and will further consider the utility of the 
recommendation, and methods of implementation, through the adaptive 
management process.
    We do note that BOEM currently requires operators to confirm 
through the permitting process that the airgun arrays used have been 
calibrated or tuned to maximize subsurface illumination and to 
minimize, to the extent practicable, horizontal propagation of noise.
    Comment: NRDC suggests that NMFS should consider requiring use of 
thermal detection as a supplement to visual monitoring.
    Response: NMFS appreciates the suggestion and agrees that 
relatively new thermal detection platforms have shown promising 
results. Following review of NRDC's letter, we considered these and 
other supplemental platforms as suggested. However, to our knowledge, 
there is no clear guidance available for operators regarding 
characteristics of effective systems, and the detection systems cited 
by NRDC are typically extremely expensive, and are therefore considered 
impracticable for use in most surveys. For example, one system cited by 
NRDC (Zitterbart et al., 2013)--a spinning infrared camera and an 
algorithm that detects whale blows on the basis of their thermal 
signature--was tested through funding provided by the German government 
and, according to the author at a 2015 workshop concerning mitigation 
and monitoring for seismic surveys, the system costs hundreds of 
thousands of dollars. We are not aware of its use in any commercial 
application. Further, these systems have limitations, as performance 
may be limited by conditions such as fog, precipitation, sea state, 
glare, water- and air-temperatures and ambient brightness, and the 
successful results obtained to date reflect a limited range of 
environmental conditions and species. NRDC acknowledges certain of 
these limitations in their comment, including that the systems have 
lesser utility in warmer temperatures. The GOM, however, is a warm 
environment. NRDC does not provide specific suggestions with regard to 
recommended systems or characteristics of systems. NMFS does not 
consider requirements to use systems such as those recommended by NRDC 
to currently be practicable.
    Comment: NRDC states that NMFS should prescribe requirements for 
use of ``noise-quieting'' technology. NRDC elaborates that in addition 
to requiring noise-quieting technology (or setting a standard for 
``noise output''), NMFS should ``prescribe targets to drive research, 
development, and adoption of alternatives to conventional airguns.''
    Response: NMFS agrees with NRDC that development and use of 
quieting technologies, or technologies that otherwise reduce the 
environmental impact of geophysical surveys, is a laudable objective 
and may be warranted in some cases. However, here the recommended 
requirements either are not practicable or are not within NMFS' 
authority to require. To some degree, NRDC misunderstands the 
discussion of this issue as presented in the notice of proposed 
rulemaking. NMFS recognizes, for example, that certain technologies, 
including the Bolt eSource airgun, are commercially available, and that 
certain techniques such as operation of the array in ``popcorn'' mode 
may reduce impacts when viable, depending on survey design and 
objectives. However, a requirement to use different technology from 
that planned or specified by an applicant--for example, a requirement 
to use the Bolt eSource airgun--would require an impracticable 
expenditure to replace the airguns planned for use. NRDC offers no 
explanation for why such a large cost imposition (in the millions of 
dollars) should be considered practicable.
    Separately, NRDC appears to suggest that NMFS must require or 
otherwise incentivize the development of wholly new or currently 
experimental technologies. We note that BOEM's PEIS concluded that 
alternative technologies are in various stages of development, and that 
none of the systems with the potential to replace airguns as a seismic 
source are currently commercially available for use on a scale of 
activity such as that considered herein. Although some alternative 
technologies are available now, or will be in the next several years, 
for select uses, none are, or will likely be in the next five years, at 
a stage where they can replace airgun arrays outright. However, some 
may be used in select environments when commercially available. 
According to BOEM, the suggestion in this comment would not provide the 
oil and gas industry or the government with sufficiently accurate data 
on the location, extent, and properties of hydrocarbon resources or the 
character

[[Page 5383]]

of formation fluids or gases, or information on shallow geologic 
hazards and seafloor geotechnical properties, in order to explore, 
develop, produce, and transport hydrocarbons safely and economically. 
Such technologies may be evaluated in the future as they become 
commercially available and on a scale commensurate to the need. In 
summary, while NMFS agrees that noise quieting technology is 
beneficial, the suggestions put forward by NRDC are either 
impracticable or outside the authority provided to NMFS by the MMPA. 
However, NMFS would consider participating in or learning about related 
efforts by parties interested in investigating these technologies. We 
note that NMFS has described a process by which new and unusual 
technologies may be considered for use under this rule (see Letters of 
Authorization).
    Comment: NRDC recommended that NMFS consider compensatory 
mitigation for the adverse impacts of the specified activity on marine 
mammals and their habitat that cannot be prevented or mitigated.
    Response: NMFS has prescribed a robust comprehensive suite of 
measures that are expected to reduce the amount of Level A and Level B 
harassment take, as well as the severity of any incurred impacts on the 
species or stock and their habitat. Compensatory mitigation is not 
required under the MMPA. Importantly, NRDC did not recommend any 
specific measure(s), rendering it impossible to evaluate their 
recommendation. In addition, many of the methods of compensatory 
mitigation that have proven successful in terrestrial settings (e.g., 
purchasing or preserving land with important habitat, improving habitat 
through plantings) are not applicable in a marine setting with such 
far-ranging species. NMFS concludes that the concept is too speculative 
at this time to warrant specific action.

Letters of Authorization

    Comment: The Associations assert that it is ``arbitrary and 
inappropriate'' for NMFS to provide an opportunity for public notice 
and comment in the event that an LOA applicant wishes to deviate from 
the modeling approach used herein (which was subject to public review 
and comment). The Associations state that such a requirement is 
contrary to the legal requirement to base the authorization of 
incidental take under the MMPA on the best available science, as better 
information may become available during the period of effectiveness for 
the ITR.
    Response: LOAs issued under the authority of section 101(a)(5)(A) 
and NMFS' implementing regulations must be preceded by both substantive 
findings (including a negligible impact finding) and a process that 
includes rulemaking after notice and comment. In the case of LOA 
applications whose take estimates are not based on the modeling used 
for the rulemaking, NMFS has determined that it may be appropriate to 
subject those to notice and comment in certain circumstances. Such a 
process requirement does not impede or contradict the requirement to 
use the best available information.
    Comment: The Associations and other industry commenters assert that 
there is no legal justification for NMFS to use the ITR as a mechanism 
to limit the number of activities that may occur in the GOM, stating 
that authorization of the activities themselves are subject to BOEM's 
jurisdiction.
    Response: NMFS clarifies that we do not intend to use the ITR in 
the manner suggested by the Associations, and that the language cited 
in the Associations' comment (``cap on the number of authorizations 
that could be issued'') was inartful. We also acknowledge BOEM's 
jurisdiction regarding the authorization of the subject activities 
themselves. However, the total taking analyzed in the negligible impact 
analysis necessarily bounds the taking that may be authorized under 
these activity-specific regulations, as described in the Estimated Take 
section.
    Comment: Referencing a cap on the number of authorizations that 
could be issued, the MMC recommends that NMFS (1) provide details to 
the public on how NMFS plans to implement the proposed cap and the 
basis for it; and (2) allow for an additional 30-day comment period to 
review such details sufficiently in advance of issuing the final rule.
    Response: As discussed in the preceding comment response, the 
language referenced by the MMC was meant only to affirm what is 
inherent in the regulations, i.e., that the amount of take analyzed for 
making a finding of negligible impact necessarily bounds the amount of 
take that may be authorized through LOAs issued under this rule 
(provided they also satisfy the small numbers requirement). The MMC 
places undue emphasis on this aspect of rule implementation. In 
claiming a ``lack of transparency,'' the MMC assigns complexity that 
does not exist, and no additional details exist to give. Therefore, we 
do not implement the MMC's recommendation for additional public 
comment.
    Comment: The Associations and other industry commenters express 
concern regarding the implementation of the ITR, including the 
evaluation and processing of LOA requests. The Associations recommend 
that the final ITR clearly address how NMFS plans to process LOA 
applications in a timely and efficient manner, and encourage NMFS to 
retain flexibility in the final ITR for the development of efficient 
and effective LOA processes through workshops or other engagement with 
BOEM and the regulated community.
    Response: NMFS appreciates the concerns expressed by the 
commenters. We believe we have addressed these issues in the updated 
preamble to this ITR (see Letters of Authorization) and are committed 
to ongoing, proactive engagement with BOEM, BSEE, and the regulated 
community towards efficient implementation of the ITR.
    Comment: BP comments that a low-frequency geophysical survey source 
they refer to as ``Wolfspar'' should be considered to be within the 
range of potential impacts modeled in the ITR and, therefore, able to 
be used under an LOA issued pursuant to the ITR.
    Response: NMFS will look forward to evaluating the Wolfspar source 
according to the ``New and Unusual Technology'' review process detailed 
in the Letters of Authorization section of this preamble. Only upon 
review of additional information regarding the source can NMFS make a 
determination in this regard.

Regulatory Impact Analysis

    Comment: The Associations provide a bulleted list of criticisms of 
the RIA. We summarize these here and provide brief responses below. For 
full detail, we refer the reader to the final RIA, available online at: 
www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico. The Associations' 
critiques of the RIA are not accompanied by any specific 
recommendations regarding potential changes to the analysis or 
additional data.
     The Associations assert that the RIA assumes that the only 
indirect costs of closures are delays, and state that such measures 
``may render some survey proposals economically unattractive to the 
point at which prospects will not be explored.'' The Associations also 
state that closures may be assumed to be permanent, thereby having an 
additional dampening impact on exploration activity.
    Response: The RIA accompanying the proposed rule did not assume 
that the costs of closures are simply delays. The RIA stated that the 
``closures have the potential to affect the overall levels of G&G 
[geological and geophysical]

[[Page 5384]]

activities that occur in the GOM over the five-year timeframe of the 
analysis. In the case that the closures delay or reduce the ability of 
industry to collect the necessary data to identify and recover oil and 
gas resources, the overall level of oil and gas production in the GOM 
may in turn be delayed or reduced.'' The RIA for the proposed rule 
discussed the possibility of both delays and reductions in activity due 
to the uncertainty surrounding rule impacts. In addition, NMFS 
reiterates that closures and any other measures are in effect only 
during the five-year period of effectiveness of the ITR. It is unclear 
why closures may be ``assumed to be permanent'' if the regulations 
requiring them are effective only for five years. We note here that two 
of the closure areas included in the proposed rule and evaluated in the 
RIA are no longer in the final rule because they fall outside the area 
considered for this rule, following BOEM's update of the rule's scope.
     The Associations assert that the RIA ``incorrectly assumes 
that the costs of closures are highly uncertain or even low because 
geologic potential of some areas is low,'' stating that geophysical 
surveys are essential to understanding the geologic potential of the 
areas.
    Response: There is significant uncertainty regarding the 
development potential of the areas considered for closure, and the RIA 
did not simply assume that it is ``low.'' The RIA accompanying the 
proposed rule provided the best available information regarding lease 
activity and reserves in the proposed closure areas and characterized 
the associated uncertainty.
     The Associations assert that the RIA ``wrongly assumes 
that the GOMESA moratorium prevents exploration of the Eastern GOM,'' 
when in fact the moratorium is currently set to expire in 2022. The 
Associations go on to state that ``the RIA seriously misleads readers 
about the costs of closure and increased restrictions in the Eastern 
GOM,'' and that the potential cost of the closure should be considered 
equivalent to that of the closure considered for the Central GOM 
because ``high-potential resources may underlie'' the Eastern GOM 
closure.
    Response: The RIA made no such assumption, and in fact acknowledged 
that the moratorium does not restrict exploration per se. However, the 
Eastern GOM closure referenced by the commenter is no longer part of 
the rule, as BOEM's update to the scope of the rule has removed the 
area from consideration.
     The Associations state that the RIA fails to account for 
the environmental benefit associated with avoiding unnecessary drilling 
via use of geophysical surveys. Chevron echoes this comment. Neither 
commenter provides any specific recommendation as to how they believe 
it should be considered.
    Response: The RIA does acknowledge the benefit of geophysical 
technology. However, we note that the magnitude of this benefit depends 
on the extent to which exploration and development companies move 
forward with drilling in cases where they have less seismic data than 
they otherwise would because of the requirements of the ITR.
     The Associations assert that the RIA for the proposed rule 
incorrectly assumes current geophysical data in the Eastern GOM is 
suitable, stating that ``there is high demand for state-of-the-art new 
data for Eastern GOM frontier areas where older data is considered 
unsuitable to support new investment.''
    Response: The RIA did not state that current geophysical data for 
the Eastern GOM is ``suitable.'' Rather, the RIA stated that ``the 
suitability of existing G&G data to direct oil and gas production in 
the closure areas is unknown.'' As noted herein, the area of concern to 
the commenter is no longer considered through this rule.
     The Associations assert that the RIA ``fails to account 
for possible increased industry interest in Eastern GOM geophysical 
surveys'' and, therefore, that the RIA inappropriately relies on old 
statistics on survey interest for estimating costs.
    Response: The RIA acknowledged that industry interest in Eastern 
GOM geophysical surveys is likely to increase leading up to the 
expiration of the moratorium. As noted herein, the area of concern to 
the commenter is no longer considered through this rule.
    Comment: Chevron comments that NMFS should ensure in the final ITRs 
that all costs are evaluated, including the cost of reduced 
environmental benefits from effective geophysical surveys. The 
Associations echo these concerns.
    Response: NMFS has appropriately evaluated the regulatory impacts 
of the ITR according to the requirements of E.O. 12866. See section 5.3 
of the Final RIA, which describes this benefit of geophysical 
technology. The magnitude of this benefit depends on the extent to 
which exploration and development companies move forward with drilling 
in cases where they have less seismic data than they otherwise would 
because of the rule.

National Environmental Policy Act

    Comment: NRDC reiterates (and resubmitted) comments that it 
submitted on BOEM's draft PEIS, stating that as it relates to marine 
mammals, the PEIS is deficient on its face due to the range of 
alternatives and mitigation considered, significance criteria, take and 
impact estimates, and cumulative impacts analysis.
    Response: As a cooperating agency, NMFS reviewed all responses to 
comments on the draft PEIS that were relevant to its management 
authorities and provided input where we deemed it appropriate. See 
Appendix M of the Final PEIS.
    Comment: NRDC also states that NMFS cannot rely on the PEIS because 
it ``does not adequately address NMFS' own actions and responsibilities 
under the MMPA,'' given that BOEM's PEIS is ``framed around a 
fundamentally different purpose and need'' relating to its mandates 
under the Outer Continental Shelf Lands Act (OCSLA) that is 
``incongruent with NMFS obligations under the MMPA.''
    Response: The proposed action at issue is BOEM's issuance of 
permits or authorizations for G&G activities in the GOM. PEIS Chapter 
1.1.1. The PEIS also recognizes that NMFS' proposed action is a 
decision on whether to approve BOEM's petition for incidental take 
regulations. NOAA is a cooperating agency on BOEM's PEIS, as NOAA has 
jurisdiction by law and special expertise over marine resources 
impacted by the proposed action, including marine mammals and federally 
listed threatened and endangered species. The PEIS explicitly 
recognizes that the PEIS would be used in support of NMFS' decision on 
BOEM's petition for incidental take regulations. See PEIS Appendix B.

[[Page 5385]]

    Consistent with the Council on Environmental Quality's (CEQ's) 
regulations, it is accepted NEPA practice for NOAA to adopt a lead 
agency's NEPA analysis when, after independent review, NOAA determines 
the document to be sufficient in accordance with 40 CFR 1506.3. 
Specifically here, NOAA is satisfied that BOEM's PEIS adequately 
addresses the impacts of issuing the MMPA incidental take authorization 
and that NOAA's comments and concerns have been adequately addressed. 
There is no requirement in CEQ regulations that NMFS, as a cooperating 
agency, issue a separate purpose and need statement in order to ensure 
adequacy and sufficiency for adoption. Nevertheless, the statement of 
Purpose and Need in the PEIS explicitly acknowledges NMFS' own separate 
action of issuing an MMPA incidental take authorization, and the PEIS 
is replete with discussion of issues relating to the issuance of an 
MMPA authorization, including discussion of marine mammal impacts, 
mitigation, and take estimates. NMFS' early participation in the NEPA 
process and the agency's continuing role in shaping and informing 
analyses using its special expertise ensured that the analysis in the 
PEIS is sufficient for purposes of NMFS' own NEPA obligations related 
to its issuance of an incidental take authorization under the MMPA.
    Regarding the alternatives, NMFS' early involvement in the 
development of the PEIS and role in evaluating the effects of 
incidental take under the MMPA ensured that the PEIS would include 
adequate analysis of a reasonable range of alternatives. The PEIS 
includes a no action alternative specifically to address what could 
happen if NMFS did not issue an MMPA authorization. Some of the 
alternatives explicitly reference marine mammals or mitigation designed 
for marine mammals in their title. More importantly, these alternatives 
fully analyze a comprehensive variety of mitigation measures for marine 
mammals. This mitigation analysis supported NMFS' evaluation of our 
options in potentially issuing an MMPA authorization. This approach to 
evaluating a reasonable range of alternatives is consistent with NMFS' 
policy and practice for issuing MMPA incidental take authorizations. 
NOAA independently reviewed and evaluated the PEIS, including the 
purpose and need statement and range of alternatives, and determined 
that the PEIS fully satisfies NMFS' NEPA obligations related to its 
decision to issue the MMPA final rule and associated Letters of 
Authorization. Accordingly, NMFS has adopted the PEIS.
    Finally, we disagree with the notion that the district court's 
decision in Conservation Council for Hawaii v. NMFS somehow would 
preclude NMFS from adopting the PEIS here. In Conservation Council, the 
court concluded that the FEIS NMFS adopted was deficient because it did 
not consider a true ``no action'' alternative from NMFS' perspective, 
in that the ``no action'' alternative assumed continuation of Navy's 
baseline activities, and therefore avoided the task facing NMFS, i.e., 
whether to authorize the requested take. 97 F. Supp. 3d at 1236. In 
contrast, the PEIS here for NMFS' rule for GOM geophysical surveys 
includes a ``no action'' alternative from the perspectives of both NMFS 
and BOEM. See PEIS, Chapter 2.9.1, pp. 2-20 to 2-22.

Information Quality Act

    Comment: The CRE states that NMFS' Technical Guidance violates 
Information Quality Act (IQA) requirements, because it (1) does not 
include an IQA Pre-dissemination Review Certification; (2) relies 
heavily on models that have not been peer reviewed to determine whether 
they are validated and comply with the Environmental Protection 
Agency's Council for Regulatory Environmental Modeling (CREM) guidance; 
and (3) relies heavily on models that were not peer reviewed in 
compliance with the Office of Management and Budget's (OMB) Final 
Information Quality Bulletin for Peer Review (70 FR 2664; January 14, 
2005).
    Response: The CRE is incorrect. NMFS performed appropriate pre-
dissemination review and documentation according to relevant agency 
guidance (NMFS Policy Directive PD 04-108, Policy on the Data Quality 
Act; NMFS Instruction 04-108-03, Section 515 Pre-Dissemination Review 
and Documentation Guidelines). All aspects of development of the 2016 
Technical Guidance were peer reviewed (www.cio.noaa.gov/services_programs/prplans/ID43.html). Also of note, the same 
information and methodology that supported development of NMFS' 
Technical Guidance (NMFS, 2016, 2018) were more recently published in a 
peer-reviewed journal (Southall et al., 2019a).
    Comment: CRE states that NMFS' use of models in the acoustic 
exposure modeling process for this rule violates the IQA because ``they 
are incomplete, unfinished, inaccurate, unreliable, have never been 
validated, and have never been peer reviewed.'' CRE also asserts that 
NMFS has not conducted pre-dissemination review and documentation as 
required by the IQA and implies that, because NMFS did not address the 
IQA in the notice of proposed rulemaking, we must be in violation of 
it.
    Response: CRE is incorrect; NMFS is in compliance with the 
requirements of the IQA. NMFS conducted the required pre-dissemination 
review at both the proposed and final stages of this rulemaking and 
appropriate documentation is included in the administrative record for 
this action.
    CRE asserts that the models used in NMFS' rulemaking process are 
not properly evaluated or validated. CRE asserts that as a result, NMFS 
``grossly overestimate[s] exposures and takes.'' According to the CRE, 
the supposed failings of the modeling necessarily lead to the 
overestimation of takes, as opposed to error in potentially different 
directions and of different magnitude in association with the various 
components of the modeling process. CRE comments at length that NMFS 
should use only the relatively simple approach of ``Line Transect,'' 
which they believe will result in lower numbers of estimated takes (see 
more detailed response to these suggestions earlier in Comments and 
Responses).
    In asserting that the models used in support of this rule have not 
been adequately validated or peer reviewed, CRE refers to a similarly 
sophisticated, proprietary modeling package (Marine Acoustics, Inc.'s 
Acoustic Integration Model (``AIM'')) that underwent a dedicated 
external peer review, stating that AIM is ``therefore properly 
validated and acceptable for regulatory use.'' However, the AIM package 
functions virtually the same as the models used for this analysis, and 
was used for an essentially identical modeling process developed in 
support of BOEM's 2014 PEIS for geological and geophysical survey 
activities on the Mid- and South Atlantic Outer Continental Shelf.
    The IQA concerns expressed in the comment are unfounded. As stated 
in the NOAA Information Quality Guidelines, information quality is 
composed of three elements: Utility, integrity, and objectivity.
    Utility means that disseminated information is useful to its 
intended users. The disseminated information at issue here--modeled 
exposures of marine mammals to underwater noise--is useful to NMFS in 
that it forms the basis for subsequent analysis allowing NMFS to make 
determinations

[[Page 5386]]

necessary under the MMPA. It is useful to the public in that it enables 
appropriate review of NMFS' action and supporting determinations. It is 
useful to the regulated entities in that it will allow for an efficient 
regulatory regime, in which potential LOA applicants may make use of 
the existing modeling effort (while being afforded the opportunity to 
engage in different modeling if desired) in service of a streamlined 
LOA application process.
    Integrity refers to security, i.e., the protection of information 
from unauthorized access or revision, to ensure that the information is 
not compromised through corruption or falsification. The integrity of 
the information disseminated herein was not questioned, but it meets 
all relevant standards for integrity (as demonstrated in the 
administrative record for this action).
    Finally, objectivity ensures that information is accurate, 
reliable, and unbiased, and that information products are presented in 
an accurate, clear, complete, and unbiased manner. Objectivity consists 
of two distinct elements: Presentation and substance. The presentation 
element includes whether disseminated information is presented in an 
accurate, clear, complete, and unbiased manner and in a proper context. 
NMFS has appropriately presented the disseminated information, and CRE 
does not assert otherwise. The substance element involves a focus on 
ensuring accurate, reliable, and unbiased information. Disseminated 
information reflects the inherent uncertainty of the scientific 
process, which is inseparable from the concept of statistical 
variation. In assessing information for accuracy, the information is 
considered accurate if it is within an acceptable degree of imprecision 
or error appropriate to the particular kind of information at issue and 
otherwise meets commonly accepted scientific and statistical standards, 
as applicable. This concept is inherent in the definition of 
``reproducibility,'' as used in the OMB IQA Guidelines and adopted by 
NOAA. Therefore, original and supporting data that are within an 
acceptable degree of imprecision, or an analytic result that is within 
an acceptable degree of imprecision or error, are by definition within 
the agency standard and are therefore considered correct. CRE does not 
assert that the modeling results disseminated by NMFS are outside the 
bounds of an acceptable degree of imprecision or error.
    The modeling report goes into great detail regarding potential 
error associated with different facets of the modeling process, and 
provides specific analysis of uncertainty in both the acoustic and 
animal phases of the modeling process (discussed in detail in the 
notice of proposed rulemaking and in the modeling report). Uncertainty 
associated with all aspects of the modeling was clearly identified and 
evaluated as to the effect on the overall modeling results. In order to 
best represent the overall uncertainty associated with the modeling, 
the report presents the exposure estimates as a distribution. The 
exposure estimate distribution provides the public with an 
understanding of the probability of certain events occurring, including 
the probability that an operation would not result in any animals being 
exposed above a defined threshold.
    Regarding reproducibility and transparency, the NOAA Information 
Quality guidelines state that ``reproducibility means that the 
information is capable of being substantially reproduced, subject to an 
acceptable degree of imprecision [. . .] With respect to analytic 
results, `capable of being substantially reproduced' means that 
independent analysis of the original or supporting data using identical 
methods would generate similar analytic results, subject to an 
acceptable degree of imprecision or error.'' We have no reason to 
believe that similar modeling (for example, using the AIM modeling 
package) using the same data inputs would not return similar analytic 
results, and CRE provides none. Transparency is not defined in the OMB 
Guidelines, but is at the heart of the reproducibility standard. At its 
most basic, transparency--and ultimately reproducibility--is a matter 
of showing how you got the results you got. NMFS has produced a 
painstakingly detailed accounting of the modeling process and decisions 
made, such that an independent party using a different set of models 
would be able to perform a similar modeling effort in order to evaluate 
the similarity of the results. The modeling report includes a full 
description of all assumptions and reference material used for both 
sound sources and species of interest. CRE provides no meaningful 
argument to the contrary.
    The NOAA Information Quality guidelines expressly address and allow 
for the use of proprietary models and other supporting information 
which cannot be disclosed. In such cases, the guidelines call for 
``especially rigorous robustness checks.'' As summarized below and 
described in detail in the notice of proposed rulemaking and modeling 
report, NMFS has conducted rigorous robustness checks of the 
proprietary models used in support of this rule.
    The models used in estimating the acoustic exposures described 
herein have been appropriately validated and reviewed. As described in 
detail in the notice of proposed rulemaking and in the modeling report, 
the acoustic exposure modeling effort requires the use of a package of 
models. Acoustic exposure modeling in general is not novel, 
controversial, or precedent-setting, and similar modeling has been 
performed for various applications for over 15 years. This type of 
modeling requires modeling of the acoustic output of a source, in this 
case a specified airgun array (as well as a single airgun and certain 
electromechanical sources that were modeled separately). The output of 
the source model is an input to a model or models used to model 
underwater sound propagation as a function of range from the source. 
The output of this process is a 3D sound field. Subsequently, an animal 
movement model is used to simulate the behavior of virtual animats in 
relation to the modeled sound field. Each animat acts as a virtual 
dosimeter, producing individual records of exposure history. There were 
many animats in the simulations, and together their received levels 
represent the probability, or risk, of exposure for each survey.
    In this case, the source model used was JASCO Applied Sciences' 
proprietary Airgun Array Source Model (AASM). The AASM accepts airgun 
volume, pressure, and depth and has internal parameters that must be 
fit to real signature data. The model was originally fit to a large 
library of empirical airgun data spanning a range of airgun volumes and 
operating depths. Subsequently, the model was improved to better 
predict airgun radiation at frequencies above 1 kHz. Development and 
validation of this improved version were made possible by high quality 
airgun source signature data from field studies conducted under the 
industry-sponsored Joint Industry Program on Sound and Marine Life. 
Desktop evaluation and validation of AASM have been conducted against 
commercial geophysical source models such as Gundalf and Nucleus.
    JASCO's proprietary Marine Operations Noise Model (MONM) was used 
to generate the 3D sound fields necessary for sound exposure estimates. 
MONM is based on standard and proven acoustic propagation models. In 
this case, propagation at frequencies less than 2 kHz was computed 
using a version of the U.S. Naval Research Laboratory's Range-dependent 
Acoustic

[[Page 5387]]

Model (RAM), which is based on a parabolic equation (PE) solution to 
the wave equation and extensively verified and validated under the Navy 
Ocean and Atmospheric Master Library process. The PE method has been 
extensively benchmarked and is widely employed in the underwater 
acoustics community (Collins et al., 1996), and RAM's predictions (as 
generated within the MONM infrastructure) have been validated against 
experimental data in several underwater acoustic measurement programs 
conducted by JASCO (e.g., Aerts et al., 2008; Funk et al., 2008; 
Ireland et al., 2009; Blees et al., 2010; Warner et al., 2010). At 
frequencies greater than 2 kHz, increased sound attenuation due to 
volume absorption at higher frequencies is accounted for with the 
widely-used BELLHOP Gaussian beam ray-trace propagation model (Porter 
and Liu, 1994). Both of these complementary, non-proprietary 
propagation models (RAM and BELLHOP) have been extensively tested over 
many years and are accepted by the acoustics community. Implementation 
of these codes within the MONM infrastructure has been evaluated and 
validated against other PE codes including RAMS, RAM-Geo and original 
RAM, and against normal mode or wavenumber integration (fast field) 
methods in standard codes. Finally, JASCO has conducted end-to-end 
validation of source and propagation modeling against field data 
collected in sound source verification experiments, demonstrating that 
the results of the acoustic field modeling are in agreement with field 
data. The comparison of model results and measurements show that MONM 
can produce reliable results in challenging acoustic propagation 
conditions (Hannay and Racca, 2005).
    The non-proprietary, peer-reviewed Marine Mammal Movement and 
Behavior (3MB) model (Houser, 2006) was used to generate realistic 
paths of simulated animals (animats) in the modeled area. JASCO's 
Exposure Modeling System (JEMS) was used to combine animal movement 
data (i.e., the output from 3MB), with pre-computed acoustic fields 
(i.e., the output from MONM described above). The JEMS is a relatively 
simple piece of software that acts as an indexer that finds the sound 
level from the computed fields for the location of each animat through 
time. The numerous, rigorous robustness checks described for the 
multiple modeling components are sufficient to comply with the IQA 
requirements, and no additional peer review is required.
    While certain components of the modeling process (AASM, MONM, and 
JEMS) are proprietary in the sense that JASCO does not make the code 
publicly available, they are all based on standard physics or 
mathematical models generally accepted in the field and based on peer-
reviewed models (e.g., 3MB). In addition, ample opportunity has been 
provided for public input and review of the underlying scientific 
information and modeling efforts contained herein (including by 
scientists, peer experts at other agencies, and non-governmental 
organizations). Relevant data is provided such that an entity using 
similar models could reproduce or challenge the results. While the 
modeling results disseminated here may reasonably be considered to be 
influential for purposes of the OMB Peer Review Bulletin--meaning that 
the information may reasonably be considered to have a clear and 
substantial impact on important public policies, such as this ITR--the 
modeling is not a ``highly influential scientific assessment,'' (HISA) 
which is defined as a scientific assessment that: (i) Could have a 
potential impact of more than $500 million in any year, or (ii) is 
novel, controversial, or precedent-setting or has significant 
interagency interest. As described above, similar approaches to 
acoustic exposure modeling have been performed by numerous disparate 
entities for multiple applications. In 2014, during the aforementioned 
modeling workshop co-sponsored by the American Petroleum Institute and 
International Association of Geophysical Contractors, at least a half-
dozen expert presenters (representing private and governmental entities 
from both the United States and Europe) discussed various available 
packages that function much the same way as what is described here. 
There is nothing novel, controversial, or precedent-setting about the 
modeling described here, and the additional peer review requirements 
associated with HISAs are not applicable.

Miscellaneous

    Comment: NRDC contends that NMFS must consider a standard requiring 
analysis and selection of minimum source levels. In furtherance of this 
overall quieting goal, NRDC also states that NMFS should consider 
requiring that all vessels employed in the survey activities undergo 
regular maintenance to minimize propeller cavitation and be required to 
employ the best ship-quieting designs and technologies available for 
their class of ship, and that NMFS should require these vessels to 
undergo measurement for their underwater noise output.
    Response: An expert panel, convened by BOEM to determine whether it 
would be feasible to develop standards to determine a lowest 
practicable source level, determined that it would not be reasonable or 
practicable to develop such metrics (see Appendix L in BOEM, 2017). 
NMFS does not believe it appropriate to address disagreements with 
these conclusions to us. NRDC further claims that NMFS' deference to 
the findings of an expert panel convened specifically to consider this 
issue is ``arbitrary under the MMPA.'' The bulk of NRDC's comment 
appears to be addressed to BOEM, and NMFS encourages NRDC to engage 
with BOEM regarding these alleged shortcomings of the panel's findings. 
The subject matter is outside NMFS' expertise, and we have no basis 
upon which to doubt the panel's published findings.
    With regard to the recommended requirements to measure or control 
vessel noise, or to make some minimum requirements regarding the design 
of vessels used in the surveys, NMFS disagrees that these requirements 
would be practicable. While NMFS agrees that vessel noise is of concern 
in a cumulative and chronic sense, it is not of substantial concern in 
relation to the MMPA's least practicable adverse impact standard for 
this specified activity, given the few vessels used in any given survey 
and relative to commercial shipping. NMFS looks forward to continued 
collaboration with NRDC and others towards ship quieting.
    Comment: NRDC states that NMFS must consider mitigation that limits 
and reduces the amount of survey activity, including ``prohibit[ing]'' 
duplicative surveys, and should consider ``consolidating'' surveys. 
Similarly, the MMC recommends that NMFS ``work with BOEM'' to require 
industry operators to increase collaboration on seismic surveys 
whenever possible.
    Response: NRDC states that NMFS should ``require and enforce a 
cap'' on surveys, without explaining how they believe this is within 
NMFS' statutory authority or suggesting ways to appropriately apportion 
the amount of effort that might be allowed. NMFS cannot arbitrarily 
limit planned effort and has no legitimate means of changing the 
specified activity absent a conclusion that the activity would have 
more than a negligible impact. However, NMFS has made the necessary 
findings under the MMPA for issuance of this rule. NRDC goes on to 
state that NMFS should ``require BOEM to eliminate unnecessary 
duplication of survey effort'' but does not explain how they

[[Page 5388]]

believe that this suggestion is within NMFS' statutory authority. As 
the permitting agency, BOEM has the authority to require permit 
applicants to submit statements indicating that existing data are not 
available to meet the data needs identified for the applicant's survey 
(i.e., non-duplicative survey statement), but such requirements are not 
within NMFS' purview. NMFS may not demand that BOEM discharge its 
authority under OCSLA in any particular manner. As stated previously, 
NMFS considers the specified activity described by an applicant in 
reviewing a request for an incidental take authorization. Nothing in 
the statute provides authority to direct consolidation or removal of 
activities based on some presumption of duplication that NMFS is not 
qualified to judge. NRDC claims erroneously that NMFS ``has authority 
under the mitigation provision of the MMPA to consider directing the 
companies to consolidate their surveys,'' placing such a requirement 
under the auspices of practicability. Leaving aside that directing any 
given applicant to abandon their survey plans would not in fact be 
practicable, it is inappropriate to consider this suggested requirement 
through that lens.
    The MMC specifically cites a number of collaborative surveys 
conducted in foreign waters and recommends that NMFS ``work with BOEM'' 
to require such collaboration. However, the MMC provides no useful 
recommendations as to how such collaboration might be achieved. Given 
the absence of appropriate statutory authority, NMFS is willing to 
explore with the MMC possible mechanisms for fostering such 
collaboration between geophysical data acquisition companies and 
relevant Federal agencies, within the context of our respective 
authorities.
    NMFS also notes that, although surveys may be perceived as 
``duplicative'' simply because other surveys have also occurred in the 
same location, they are in fact designed specifically to produce 
proprietary data that satisfies the needs of survey funders. As noted 
by NRDC, BOEM convened an expert panel to study the issue of 
duplicative surveys (see Appendix L in BOEM, 2017) and developed 
standards for consideration of what surveys are duplicative. NRDC 
provides extensive discussion of their thoughts regarding the 
insufficiency of BOEM's duplicative survey standard and its 
implementation. We respectfully suggest that these comments are more 
appropriately directed at BOEM.
    Comment: Chevron states that NMFS ``must be mindful of the mandates 
under OCSLA to assess and then balance the costs and benefits of 
alternative restrictions on geophysical activities against a 
requirement for `expeditious and orderly development' of GOM 
resources.''
    Response: NMFS' statutory obligations arise under the Marine Mammal 
Protection Act (with associated requirements under the Endangered 
Species Act, National Environmental Policy Act, and Administrative 
Procedure Act, among others). NMFS has no statutory obligation relative 
to OCSLA.
    Comment: CRE provides several comments relating to E.O. 12866. CRE 
reiterates their view that there is ``no harm from seismic,'' and 
therefore, that it is not surprising that NMFS has not produced a 
quantitative statement of benefits. They also conclude that ``[s]ince 
the benefits of the proposed rule are minimal at best, the resultant 
benefit-cost ratio is less than one, making the proposed rule non-
compliant'' with E.O. 12866.
    Response: NMFS disagrees with the commenter's premise that there is 
no potential for harm, and accordingly evaluated the impacts of the 
specified activity and prescribed appropriate mitigation in the ITR, as 
required under the MMPA. With respect to E.O. 12866, the RIA provides a 
qualitative description of potential ecological benefits and their 
economic implications due to uncertainty preventing quantification. 
Similar to the qualitative evaluation of costs associated with the 
proposed area closures, the qualitative treatment of benefits does not 
indicate a lesser magnitude, but rather more data limitations or 
uncertainty.
    Comment: Regarding E.O. 13211, Chevron comments that NMFS has 
provided inconsistent statements that should be resolved.
    Response: NMFS has clarified its discussion regarding E.O. 13211. 
Overall, within the five-year timeframe of the analysis, the ITR is not 
expected to constitute a significant adverse effect on energy supply, 
distribution or use, according to the thresholds described by E.O. 
13211, given that the direct compliance costs represent a small 
fraction (on the order of less than one percent) of the total costs of 
exploration and development in the GOM.
    Comment: Chevron notes that E.O. 13795 required evaluation of NMFS' 
2016 Technical Guidance (review of which was ongoing at the time of 
publication of the notice of proposed rulemaking). Chevron also asserts 
that assumptions of the 2016 Technical Guidance ``are multiplied with 
those in other elements of the modeling to reach `unrealistic' 
conclusions.'' Because the 2016 Technical Guidance was used in the 
modeling, Chevron asserts that the modeling is inconsistent with the 
requirements of E.O. 13795. The CRE also claims that use of the 2016 
Technical Guidance is in violation of E.O. 13795 and that the guidance 
should be rescinded or substantially revised. CRE also states that NMFS 
must emphasize that use of the Technical Guidance is not required.
    Response: Review of the Technical Guidance under E.O. 13795 was 
completed in 2018. In response to the feedback received during the 
public comment period and the Interagency Consultation meeting, the 
Secretary of Commerce approved NMFS to issue a 2018 Revised Technical 
Guidance for Assessing the Effects of Anthropogenic Sound on Marine 
Mammal Hearing: Acoustic Thresholds for Onset of Permanent and 
Temporary Threshold Shifts (2018 Revised Technical Guidance) (NOAA 
Technical Memorandum NMFS-OPR-59) (June 21, 2018). NMFS' use of the 
guidance is, therefore, in compliance with E.O. 13795.
    The 2018 Revised Technical Guidance retains the thresholds and 
weighting functions presented in the original 2016 Technical Guidance. 
Chevron's comment that the Technical Guidance somehow contributes to 
what they characterize as ``unrealistic'' conclusions is, in context of 
industry's overall comments on the modeling effort, unpersuasive. The 
industry-funded supplementary modeling variable analysis (Zeddies et 
al., 2017b) found that use of the Technical Guidance was the single 
most influential factor in reducing the modeled exposures (for Level A 
harassment).
    We acknowledge that the Technical Guidance is indeed guidance, and 
its use is voluntary (as stated in the Executive Summary of the 
Technical Guidance). The Technical Guidance provides more detail on if/
when an alternative approach may be used.

Description of Marine Mammals in the Area of the Specified Activities

    Sections 3 and 4 of the petition summarize available information 
regarding status and trends, distribution and habitat preferences, and 
behavior and life history of the potentially affected species. NMFS 
refers the reader to those descriptions, descriptions of the affected 
environment in Appendix E of BOEM's PEIS, as well as NMFS' Stock 
Assessment Reports (SAR; www.fisheries.noaa.gov/national/

[[Page 5389]]

marine-mammal-protection/marine-mammal-stock-assessments), incorporated 
here by reference, instead of reprinting the information. Additional 
general information about these species (e.g., physical and behavioral 
descriptions) may be found on NMFS' website (www.fisheries.noaa.gov/find-species).
    Table 4 lists all species with expected potential for occurrence in 
the GOM and summarizes information related to the population or stock, 
including potential biological removal (PBR). For taxonomy, we follow 
Committee on Taxonomy (2020). PBR, defined by the MMPA as the maximum 
number of animals, not including natural mortalities, that may be 
removed from a marine mammal stock while allowing that stock to reach 
or maintain its optimum sustainable population, is considered in 
concert with known sources of ongoing anthropogenic mortality (as 
described in NMFS' SARs). For status of species, we provide information 
regarding U.S. regulatory status under the MMPA and ESA.
    Marine mammal abundance estimates presented in this document 
represent the total number of individuals that make up a given stock or 
the total number estimated within a particular study area. NMFS' stock 
abundance estimates for most species represent the total estimate of 
individuals within the geographic area, if known, that comprises that 
stock. For some species, this geographic area may extend beyond U.S. 
waters. Survey abundance (as compared to stock or species abundance) is 
the total number of individuals estimated within the survey area, which 
may or may not align completely with a stock's geographic range as 
defined in the SARs. These surveys may also extend beyond U.S. waters. 
For many GOM stocks, information regarding distribution and range-wide 
abundance is limited, as available data are generally limited to U.S. 
waters of the northern GOM. Abundance and distribution for GOM stocks 
occurring in the Mexican EEZ or the high seas are poorly understood. As 
discussed in additional detail below, U.S. waters only comprise about 
40 percent of the entire GOM, and 65 percent of GOM oceanic waters are 
south of the U.S. EEZ. Studies based on abundance and distribution 
surveys restricted to U.S. waters are unable to detect temporal shifts 
in distribution beyond U.S. waters that might account for any changes 
in abundance within U.S. waters.
    In some cases, species are treated as guilds. In general ecological 
terms, a guild is a group of species that have similar requirements and 
play a similar role within a community. However, for purposes of stock 
assessment or abundance prediction, certain species may be treated 
together as a guild because they are difficult to distinguish visually 
and many observations are ambiguous. For example, NMFS' GOM SARs assess 
stocks of Mesoplodon spp. and Kogia spp. as guilds. Here, we consider 
beaked whales and Kogia spp. as guilds. In the following discussion, 
reference to ``beaked whales'' includes the Cuvier's, Blainville's, and 
Gervais beaked whales, and reference to ``Kogia spp.'' includes both 
the dwarf and pygmy sperm whale.
    Twenty-one species (with 24 managed stocks) have the potential to 
co-occur with the prospective survey activities. Extralimital species 
or stocks unlikely to co-occur with survey activity include 31 
estuarine bottlenose dolphin stocks, the blue whale (Balaenoptera 
musculus), fin whale (B. physalus), sei whale (B. borealis), minke 
whale (B. acutorostrata), humpback whale (Megaptera novaeangliae), 
North Atlantic right whale (Eubalaena glacialis), and the Sowerby's 
beaked whale (Mesoplodon bidens). All mysticete species listed here (as 
well as Sowerby's beaked whale) are considered only of accidental 
occurrence in GOM and are generally historically known only from a very 
small number of strandings and/or sightings (W[uuml]rsig et al., 2000; 
W[uuml]rsig, 2017). In addition, following BOEM's update to the scope 
of activity considered through this rule, the eastern coastal stock of 
bottlenose dolphin, which was considered in the notice of proposed 
rulemaking, would no longer be potentially impacted by activities that 
may be authorized under this rule. For detailed discussion of these 
species, please see the notice of proposed rulemaking (83 FR 29212; 
June 22, 2018). In addition, the West Indian manatee (Trichechus 
manatus latirostris) may be found in coastal waters of the GOM. 
However, manatees are managed by the U.S. Fish and Wildlife Service and 
are not considered further in this document. All managed stocks in this 
region are assessed in NMFS' U.S. Atlantic SARs.
    All values presented in Table 4, which are available in the most 
recent final SARs (Hayes et al., 2020) and have not changed since the 
proposed rule was published, are the most recent available at the time 
the analyses for this final rule were completed. We also reviewed new 
information for many GOM stocks in unpublished draft 2020 SARs. The 
unpublished draft SARs include updates to most GOM stocks, including to 
abundance estimates, PBR values, and annual mortality and serious 
injury (M/SI) estimates. The most notable change is that, through the 
introduction of M/SI estimates related to the Deepwater Horizon (DWH) 
oil spill, M/SI values are generally larger than in past SARs and in 
some cases are larger than the PBR values. NMFS has considered this 
information and determined that it is previously accounted for as part 
of the baseline, through our existing analysis of the effects of the 
DWH oil spill. We have fully considered the underlying information in 
our analysis and have determined that the unpublished draft SAR updates 
do not impact our conclusions.

                                    Table 4--Marine Mammals Potentially Present in the Specified Geographical Region
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                NMFS stock
                                                                               ESA/MMPA       abundance (CV,      Predicted mean               Annual M/
           Common name              Scientific name           Stock             status;      Nmin, most recent     (CV)/maximum        PBR      SI (CV)
                                                                             Strategic (Y/       abundance         abundance \3\                  \4\
                                                                                N) \1\         survey)\2,7\
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                          Order Cetartiodactyla--Cetacea--Superfamily Mysticeti (baleen whales)
--------------------------------------------------------------------------------------------------------------------------------------------------------
Family Balaenopteridae
 (rorquals):
    Bryde's whale...............  Balaenoptera edeni.  Gulf of Mexico.....  E/D; Y          33 (1.07; 16;       44 (0.27)/n/a.....       0.03        0.8
                                                                                             2009).
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                            Superfamily Odontoceti (toothed whales, dolphins, and porpoises)
--------------------------------------------------------------------------------------------------------------------------------------------------------
Family Physeteridae:
    Sperm whale.................  Physeter             GOM................  E/D; Y          763 (0.38; 560;     2,128 (0.08)/2,234        1.1          0
                                   macrocephalus.                                            2009).

[[Page 5390]]

 
Family Kogiidae:
    Pygmy sperm whale...........  Kogia breviceps....  GOM................  -; N            186 (1.04; 90;      2,234 (0.19)/6,117        0.9  0.3 (1.0)
                                                                                             2009) \5\.          \5\.
    Dwarf sperm whale...........  K. sima............  GOM................  -; N
Family Ziphiidae (beaked
 whales):
    Cuvier's beaked whale:......  Ziphius cavirostris  GOM................  -; N            74 (1.04; 36;       2,910 (0.16)/3,958        0.4          0
                                                                                             2009).              \5\.
    Gervais beaked whale........  Mesoplodon           GOM................  -; N            149 (0.91; 77;      ..................        0.8          0
                                   europaeus.                                                2009) \5\.
    Blainville's beaked whale...  M. densirostris....  GOM................  -; N
Family Delphinidae:
    Rough-toothed dolphin.......  Steno bredanensis..  GOM................  -; N            624 (0.99; 311;     4,853 (0.19)/n/a..        2.5  0.8 (1.0)
                                                                                             2009).
    Common bottlenose dolphin...  Tursiops truncatus   GOM Oceanic........  -; N            5,806 (0.39;        138,602 (0.06)/            42        6.5
                                   truncatus.          GOM Continental      -; N             4,230; 2009).       192,176 \5\.             469     (0.65)
                                                        Shelf.                              51,192 (0.10;       ..................                   0.8
                                                                                             46,926; 2011-12).
                                                       GOM Coastal,         -; N            7,185 (0.21;                                   60        0.4
                                                        Northern.                            6,044; 2011-12).
                                                       GOM Coastal,         -; N            20,161 (0.17;                                 175        0.6
                                                        Western.                             17,491; 2011-12).
Clymene dolphin.................  Stenella clymene...  GOM................  -; N            129 (1.00; 64;      11,000 (0.16)/            0.6          0
                                                                                             2009).              12,115.
Atlantic spotted dolphin........  S. frontalis.......  GOM................  -; N            37,611 (0.28;       47,488 (0.13)/         Undet.  42 (0.45)
                                                                                             29,844; 2000-       85,108.
                                                                                             01)\6\.
Pantropical spotted dolphin.....  S. attenuata         GOM................  -; N            50,880 (0.27;       84,014 (0.06)/            407        4.4
                                   attenuata.                                                40,699; 2009).      108,764.
Spinner dolphin.................  S. longirostris      GOM................  -; N            11,441 (0.83;       13,485 (0.24)/             62          0
                                   longirostris.                                             6,221; 2009).       31,341.
Striped dolphin.................  S. coeruleoalba....  GOM................  -; N            1,849 (0.77;        4,914 (0.17)/5,323         10          0
                                                                                             1,041; 2009).
Fraser's dolphin................  Lagenodelphis hosei  GOM................  -; N            726 (0.7; 427;      1,665 (0.73)/n/a..     Undet.          0
                                                                                             1996-2001)\6\.
Risso's dolphin.................  Grampus griseus....  GOM................  -; N            2,442 (0.57;        3,137 (0.10)/4,153         16        7.9
                                                                                             1,563; 2009).                                        (0.85)
Melon-headed whale..............  Peponocephala        GOM................  -; N            2,235 (0.75;        6,733 (0.30)/7,105         13          0
                                   electra.                                                  1,274; 2009).
Pygmy killer whale..............  Feresa attenuata...  GOM................  -; N            152 (1.02; 75;      2,126 (0.30)/n/a..        0.8          0
                                                                                             2009).
False killer whale..............  Pseudorca            GOM................  -; N            777 (0.56; 501;     3,204 (0.36)/n/a..     Undet.          0
                                   crassidens.                                               2003-04)\6\.
Killer whale....................  Orcinus orca.......  GOM................  -; N            28 (1.02; 14;       185 (0.41)/n/a....        0.1          0
                                                                                             2009).
Short-finned pilot whale........  Globicephala         GOM................  -; N            2,415 (0.66;        1,981 (0.18)/n/a..         15  0.5 (1.0)
                                   macrorhynchus.                                            1,456; 2009).
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ ESA status: Endangered (E)/MMPA status: Depleted (D). A dash (-) indicates that the species is not listed under the ESA or designated as depleted
  under the MMPA. Under the MMPA, a strategic stock is one for which the level of direct human-caused mortality exceeds PBR or which is determined to be
  declining and likely to be listed under the ESA within the foreseeable future. Any species or stock listed under the ESA is automatically designated
  under the MMPA as depleted and as a strategic stock.
\2\ NMFS marine mammal stock assessment reports online at: www.fisheries.noaa.gov/national/marine-mammal-protection/marine-mammal-stock-assessments. CV
  is coefficient of variation; Nmin is the minimum estimate of stock abundance.
\3\This information represents species- or guild-specific abundance predicted by habitat-based cetacean density models (Roberts et al., 2016). These
  models provide the best available scientific information regarding predicted density patterns of cetaceans in the U.S. Gulf of Mexico, and we provide
  the corresponding abundance predictions as a point of reference. Total abundance estimates were produced by computing the mean density of all pixels
  in the modeled area and multiplying by its area.
\4\ These values, found in NMFS' SARs, represent annual levels of human-caused mortality plus serious injury from all sources combined (e.g., commercial
  fisheries, ship strike). A CV associated with estimated mortality due to commercial fisheries is presented in some cases.
\5\ Abundance estimates are in some cases reported for a guild or group of species when those species are difficult to differentiate at sea. Similarly,
  the habitat-based cetacean density models produced by Roberts et al. (2016) are based in part on available observational data which, in some cases, is
  limited to genus or guild in terms of taxonomic definition. NMFS' SARs present pooled abundance estimates for Kogia spp. and Mesoplodon spp., while
  Roberts et al. (2016) produced density models to genus level for Kogia spp. and as a guild for beaked whales (Ziphius cavirostris and Mesoplodon
  spp.). Finally, Roberts et al. (2016) produced a density model for bottlenose dolphins that does not differentiate between oceanic, shelf, and coastal
  stocks. The modeled abundance estimate provided here for all bottlenose dolphins includes abundance that may be attributed to the eastern coastal
  stock.
\6\ NMFS' abundance estimates for these species are not considered current. PBR is therefore considered undetermined, as there is no current minimum
  abundance estimate for use in calculation. We nevertheless present the most recent abundance estimate.
\7\ We note that Dias and Garrison (2016) present abundance estimates for oceanic stocks that were calculated for use in DWH oil spill injury
  quantification. For most stocks, these estimates are based on pooled observations from shipboard surveys conducted in 2003, 2004, and 2009 and
  corrected for detection bias. Estimates for beaked whales and Kogia spp. were based on density estimates derived from passive acoustic data collection
  (Hildebrand et al., 2012). The abundance estimate for Bryde's whales incorporated the results of additional shipboard surveys conducted in 2007, 2010,
  and 2012. Here we retain NMFS' official SAR information for comparison with model-predicted abundance (Roberts et al., 2016).

    For the majority of species potentially present in the specified 
geographical region, NMFS has designated only a single generic stock 
(i.e., ``Gulf of Mexico'') for management purposes, although there is 
currently no information to differentiate the stock from the Atlantic 
Ocean stock of the same species, nor information on whether more than 
one stock may exist in the GOM (Hayes et al., 2020).
    For the bottlenose dolphin, NMFS defines an oceanic stock, a 
continental shelf stock, and three coastal stocks. As in the 
northwestern Atlantic Ocean, there are two general bottlenose dolphin 
ecotypes: ``coastal'' and ``offshore.''

[[Page 5391]]

These ecotypes are genetically and morphologically distinct (Hoelzel et 
al., 1998; Waring et al., 2016), though ecotype distribution is not 
clearly defined and the stocks are delineated primarily on the basis of 
management rather than ecological boundaries. The offshore ecotype is 
assumed to correspond to the oceanic stock, with the stock boundary 
(and thus the de facto delineation of offshore and coastal ecotypes) 
defined as the 200-m isobath. The continental shelf stock is defined as 
between two typical survey strata: The 20- and 200-m isobaths. While 
the shelf stock is assumed to consist primarily of coastal ecotype 
dolphins, offshore ecotype dolphins may also be present. There is 
expected to be some overlap with the three coastal stocks as well, 
though the degree is unknown and it is not thought that significant 
mixing or interbreeding occurs between them (Waring et al., 2016). The 
coastal stocks are defined as being in waters between the shore, 
barrier islands, or presumed outer bay boundaries out to the 20-m 
isobath and, as a working hypothesis, NMFS has assumed that dolphins 
occupying habitats with dissimilar climatic, coastal, and oceanographic 
characteristics might be restricted in their movements between 
habitats, thus constituting separate stocks (Waring et al., 2016). 
Shoreward of the 20-m isobath, the eastern coastal stock extends from 
Key West, FL to 84[deg] W longitude; the northern coastal stock from 
84[deg] W longitude to the Mississippi River delta; and the western 
coastal stock from the Mississippi River delta to the Mexican border. 
The latter is assumed to be a trans-boundary stock, though no 
information is available regarding abundance in Mexican waters. As 
noted above, the eastern coastal stock will not be affected by 
activities considered through this rule.
    At the time of publication of the notice of proposed rulemaking, 
the GOM Bryde's whale was proposed for listing as an endangered species 
under the ESA (81 FR 88639; December 8, 2016). Since that time, NMFS 
has listed the GOM Bryde's whale as endangered under the ESA, effective 
on May 15, 2019 (84 FR 15446; April 15, 2019). The proposed listing was 
based largely on NMFS' status review of Bryde's whales in the GOM 
(Rosel et al., 2016), and no significant new information has become 
available since that time. No critical habitat has yet been designated 
for the species, and no recovery plan has yet been developed. NMFS' 
analysis related to the GOM Bryde's whale in the notice of proposed 
rulemaking was conducted in context of the same information that 
informed the proposal to list the GOM Bryde's whale and, therefore, the 
final listing decision itself does not introduce new information for 
consideration in the analysis for this final rulemaking.
    In Table 4 above, NMFS reports two sets of abundance estimates: 
Those from NMFS' SARs and those predicted by Roberts et al. (2016)--for 
the latter, we provide both the annual mean and the monthly maximum 
(where applicable). Please see footnotes 2-3 of Table 4 for more 
detail. NMFS' SAR estimates are typically generated from the most 
recent shipboard and/or aerial surveys conducted. GOM oceanography is 
dynamic, and the spatial scale of the GOM is small relative to the 
ability of most cetacean species to travel. As an example, no groups of 
Fraser's dolphins were observed during dedicated cetacean abundance 
surveys during 2003-2004 or 2009, yet the SAR states that it is 
probable that Fraser's dolphins were present in the northern GOM but 
simply not encountered, and therefore declines to present an abundance 
estimate of zero (Waring et al., 2013). U.S. waters only comprise about 
40 percent of the entire GOM, and 65 percent of GOM oceanic waters are 
south of the U.S. EEZ. Studies based on abundance and distribution 
surveys restricted to U.S. waters are unable to detect temporal shifts 
in distribution beyond U.S. waters that might account for any changes 
in abundance within U.S. waters. NMFS' SAR estimates also typically do 
not incorporate correction for detection bias. Therefore, they should 
generally be considered underestimates, especially for cryptic or long-
diving species (e.g., beaked whales, Kogia spp., sperm whales). Dias 
and Garrison (2016) state, for example, that current abundance 
estimates for Kogia spp. may be considerably underestimated due to the 
cryptic behavior of these species and difficulty of detection in 
Beaufort sea state greater than one, and density estimates for certain 
species derived from long-term passive acoustic monitoring are much 
higher than are estimates derived from visual observations (Mullin and 
Fulling, 2004; Mullin, 2007; Hildebrand et al., 2012).
    The Roberts et al. (2016) abundance estimates represent the output 
of predictive models derived from multi-year observations and 
associated environmental parameters and which incorporate corrections 
for detection bias. Incorporating more data over multiple years of 
observation can yield different results in either direction, as the 
result is not as readily influenced by fine-scale shifts in species 
habitat preferences or by the absence of a species in the study area 
during a given year. NMFS' abundance estimates show substantial year-
to-year variability in some cases. For example, NMFS-reported estimates 
for the Clymene dolphin vary by a maximum factor of more than 100 (2009 
estimate of 129 versus 1996-2001 estimate of 17,355), indicating that 
it may be more appropriate to use the model prediction versus a point 
estimate, as the model incorporates all available data (from 1992-
2009). The latter factor--incorporation of correction for detection 
bias--should systematically result in greater abundance predictions. 
For these reasons, the Roberts et al. (2016) estimates are generally 
more realistic and, for these purposes, represent the best available 
information. For purposes of assessing estimated exposures relative to 
abundance--used in this case to understand the scale of the predicted 
takes compared to the population--NMFS generally believes that the 
Roberts et al. (2016) abundance predictions are most appropriate 
because they were used to generate the exposure estimates and therefore 
provide the most relevant comparison. Roberts et al. (2016) represents 
the best available scientific information regarding marine mammal 
occurrence and distribution in the Gulf of Mexico.
    As a further illustration of the distinction between the SARs and 
model-predicted abundance estimates, the current NMFS stock abundance 
estimates for most GOM species are based on direct observations from 
shipboard surveys conducted in 2009 (from the 200-m isobath to the edge 
of the U.S. EEZ) and not corrected for detection bias, whereas the 
exposure estimates presented herein for those species are based on the 
abundance predicted by a density surface model informed by observations 
from surveys conducted over approximately 20 years and covariates 
associated at the observation level. To directly compare the estimated 
exposures predicted by the outputs of the Roberts et al. (2016) model 
to NMFS' SAR abundance would therefore not be meaningful.
    Biologically Important Areas (BIA)--As part of our description of 
the environmental baseline, we discuss any known areas of importance as 
marine mammal habitat. These areas may include designated critical 
habitat for ESA-listed species (as defined by section 3 of the ESA) or 
other known areas not formally designated pursuant to any statute or 
other law. Important areas may include areas of known importance for 
reproduction, feeding, or migration, or areas where small and

[[Page 5392]]

resident populations are known to occur.
    Although there is no designated critical habitat for marine mammal 
species in the specified geographical region, BIAs for marine mammals 
are recognized. For example, the GOM Bryde's whale is a very small 
population that is genetically distinct from other Bryde's whales and 
not genetically diverse within the GOM (Rosel and Wilcox, 2014). 
Further, the species is typically observed only within a narrowly 
circumscribed area within the eastern GOM. Therefore, this area is 
described as a year-round BIA by LaBrecque et al. (2015). Although 
survey effort has covered all oceanic waters of the U.S. GOM, whales 
were observed only between approximately the 100- and 300-m isobaths in 
the eastern GOM from the head of the De Soto Canyon (south of 
Pensacola, Florida) to northwest of Tampa Bay, Florida (Maze-Foley and 
Mullin, 2006; Waring et al., 2016; Rosel and Wilcox, 2014; Rosel et 
al., 2016). NOAA subsequently conducted a status review of the GOM 
Bryde's whale (Rosel et al., 2016). The review expanded this 
description by stating that, due to the depth of some sightings, the 
area is more appropriately defined to the 400-m isobath and westward to 
Mobile Bay, Alabama, in order to provide some buffer around the deeper 
sightings and to include all sightings in the northeastern GOM. 
However, the recorded Bryde's whale shipboard and aerial survey 
sightings between 1989 and 2015 have mainly fallen within the BIA 
described by LaBreque et al. (2015). The entirety of this area is now 
excluded from the scope of this rule following BOEM's update to that 
scope.
    LaBrecque et al. (2015) also described eleven year-round BIAs for 
small and resident BSE bottlenose dolphin populations in the GOM. 
Additional study would likely allow for identification of additional 
BIAs associated with other GOM BSE dolphin stocks.
    Deepwater Horizon Oil Spill--In 2010 the Macondo well blowout and 
explosion aboard the Deepwater Horizon drilling rig (also known as the 
Deepwater Horizon explosion, oil spill, and response; hereafter 
referred to as the DWH oil spill) caused oil, natural gas, and other 
substances to flow into the GOM for 87 days before the well was sealed. 
Total oil discharge was estimated at 3.19 million barrels (134 million 
gallons), resulting in the largest marine oil spill in history (DWH 
NRDA Trustees, 2016). In addition, the response effort involved 
extensive application of dispersants at the seafloor and at the 
surface, and controlled burning of oil at the surface was also used 
extensively as a response technique. The oil, dispersant, and burn 
residue compounds present ecological challenges in the region. NMFS 
discussed the impacts of the DWH oil spill on marine mammals in detail 
in the notice of proposed rulemaking (83 FR 29212; June 22, 2018) and 
we refer the reader to that document for additional detail.
    At its maximum extent, oil covered over 40,000 km\2\ of ocean. 
Cumulatively, over the course of the spill, oil was detected on over 
112,000 km\2\ of ocean. Currents, winds, and tides carried these 
surface oil slicks to shore, fouling more than 2,100 km of shoreline, 
including beaches, bays, estuaries, and marshes from eastern Texas to 
the Florida Panhandle. In addition, some lighter oil compounds 
evaporated from the slicks, exposing air-breathing organisms like 
marine mammals to noxious fumes at the sea surface.
    The Oil Pollution Act requires that a natural resource damage 
assessment (NRDA) be conducted following oil pollution incidents. An 
injury assessment undertaken as part of the NRDA first requires a 
determination of whether an incident injured natural resources. 
Trustees assessing natural resource injuries must establish that a 
pathway existed from the oil discharge to the resource, confirm that 
resources were exposed to the discharge, and evaluate the adverse 
effects that occurred as a result of the exposure (or response 
activities). Subsequently, the assessment requires injury 
quantification (including degree and spatiotemporal extent), 
essentially by comparing the post-event conditions with the pre-event 
baseline. For a fuller overview of the injury assessment process in 
this case, please see Takeshita et al. (2017). Critical pathways of 
exposure for marine mammals included the contaminated water column, 
where they swim and capture prey; the surface slick at the air to water 
interface, where they breathe, rest, and swim; and contaminated 
sediment, where they forage and capture prey.
    DWH oil was found to cause problems with the regulation of stress 
hormone secretion from adrenal cells and kidney cells, which will 
affect an animal's ability to regulate body functions and respond 
appropriately to stressful situations, thus leading to reduced fitness. 
Bottlenose dolphins living in habitats contaminated with DWH oil showed 
signs of adrenal dysfunction, and dead, stranded dolphins from areas 
contaminated with DWH oil had smaller adrenal glands (Schwacke et al., 
2014a; Venn-Watson et al., 2015b). Other factors were ruled out as a 
primary cause for the high prevalence of adverse health effects, 
reproductive failures, and disease in stranded animals. When all of the 
data were considered together, the DWH oil spill was determined to be 
the only reasonable cause for the full suite of observed adverse health 
effects.
    Due to the difficulty of investigating marine mammals in pelagic 
environments and across the entire region impacted by the event, the 
injury assessment focused on health assessments conducted on bottlenose 
dolphins in nearshore habitats and used these populations as case 
studies for extrapolating to coastal and oceanic populations that 
received similar or worse exposure to DWH oil, with appropriate 
adjustments made for differences in behavior, anatomy, physiology, life 
histories, and population dynamics among species. Investigators then 
used a population modeling approach to capture the overlapping and 
synergistic relationships among the metrics for injury, and to quantify 
the entire scope of DWH marine mammal injury to populations into the 
future, expressed as ``lost cetacean years'' due to the DWH oil spill 
(which represents years lost due to premature mortality as well as the 
resultant loss of reproductive output). This approach allowed for 
consideration of long-term impacts resulting from immediate losses and 
reproductive failures in the few years following the spill, as well as 
expected persistent impacts on survival and reproduction for exposed 
animals well into the future (Takeshita et al., 2017). For a more 
detailed overview of the injury quantification for these stocks and 
their post-DWH population trajectory, please see Schwacke et al. 
(2017), and for full details of the overall injury quantification, see 
DWH MMIQT (2015).
    The results of the quantification exercise for each affected shelf 
and oceanic stock, and for northern and western coastal stocks of 
bottlenose dolphin, are presented in Table 5. This is likely a 
conservative estimate of impacts, because: (1) Shelf and oceanic 
species experienced long exposures (up to 90 days) to very high 
concentrations of fresh oil and a diverse suite of response activities, 
while estuarine dolphins were not exposed until later in the spill 
period and to weathered oil products at lower water concentrations; (2) 
oceanic cetaceans dive longer and to deeper depths, and it is possible 
that the types of lung injuries observed in estuarine dolphins may be 
more severe for oceanic cetaceans; and (3) cetaceans

[[Page 5393]]

in deeper waters were exposed to very high concentrations of volatile 
gas compounds at the water's surface near the wellhead. No analysis was 
performed for Fraser's dolphins or killer whales; although they are 
present in the GOM, sightings are rare and there were no historical 
sightings in the oil spill footprint during the surveys used in the 
quantification process. These stocks were likely injured, but no 
information is available on which to base a quantification effort.

                                                  Table 5--Summary of Modeled Effects of DWH Oil Spill
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                          % Females with   % Population      % Maximum
                                                           % Population    % Population    reproductive    with adverse     population       Years to
                       Common name                        exposed to oil    killed (95%    failure (95%   health effects  reduction (95%   recovery (95%
                                                             (95% CI)           CI)             CI)          (95% CI)           CI)            CI) b
--------------------------------------------------------------------------------------------------------------------------------------------------------
Bryde's whale...........................................     48 (23-100)       17 (7-24)      22 (10-31)       18 (7-28)             -22              69
Sperm whale.............................................      16 (11-23)         6 (2-8)        7 (3-10)         6 (2-9)              -7              21
Kogia spp...............................................       15 (8-29)         5 (2-7)        7 (3-10)         6 (2-9)              -6              11
Beaked whales...........................................       12 (7-22)         4 (2-6)         5 (3-8)         4 (2-7)              -6              10
Rough-toothed dolphin...................................     41 (16-100)       14 (6-20)       19 (9-26)       15 (6-23)             -17              54
Bottlenose dolphin, oceanic.............................       10 (5-10)         3 (1-5)         5 (2-6)         4 (1-6)              -4             n/a
Bottlenose dolphin, northern coastal....................     82 (55-100)      38 (26-58)      37 (17-53)      30 (11-47)     -50 (32-73)      39 (23-76)
Bottlenose dolphin, western coastal.....................      23 (16-32)         1 (1-2)       10 (5-15)        8 (3-13)        -5 (3-9)             n/a
Shelf dolphins a........................................       13 (9-19)         4 (2-6)         6 (3-8)         5 (2-7)              -3             n/a
Clymene dolphin.........................................        7 (3-15)         2 (1-4)         3 (2-5)         3 (1-4)              -3             n/a
Pantropical spotted dolphin.............................      20 (15-26)        7 (3-10)        9 (4-13)        7 (3-11)              -9              39
Spinner dolphin.........................................      47 (24-91)       16 (7-23)      21 (10-30)       17 (6-27)             -23             105
Striped dolphin.........................................       13 (8-22)         5 (2-7)         6 (3-9)         5 (2-8)              -6              14
Risso's dolphin.........................................        8 (5-13)         3 (1-4)         3 (2-5)         3 (1-4)              -3             n/a
Melon-headed whale......................................       15 (6-36)         5 (2-7)        7 (3-10)         6 (2-9)              -7              29
Pygmy killer whale......................................       15 (7-33)         5 (2-8)        7 (3-10)         6 (2-9)              -7              29
False killer whale......................................       18 (7-48)         6 (3-9)        8 (4-12)        7 (3-11)              -9              42
Short-finned pilot whale................................         6 (4-9)         2 (1-3)         3 (1-4)         2 (1-3)              -3             n/a
--------------------------------------------------------------------------------------------------------------------------------------------------------
Modified from DWH NRDA Trustees (2016).
CI = confidence interval. No CI was calculated for population reduction or years to recovery for shelf or oceanic stocks.
a ``Shelf dolphins'' includes Atlantic spotted dolphins and the shelf stock of bottlenose dolphins (20-200 m water depth). These two species were
  combined because the abundance estimate used in population modeling was derived from aerial surveys and the species could not generally be
  distinguished from the air.
b It is not possible to calculate YTR for stocks with maximum population reductions of less than or equal to 5 percent.

    Coastal and oceanic marine mammals were injured by exposure to oil 
from the DWH spill. Nearly all of the stocks that overlap with the oil 
spill footprint have demonstrable, quantifiable injuries, and the 
remaining stocks (for which there is no quantifiable injury) were also 
likely injured, though there is not currently enough information to 
make a determination. Injuries included elevated mortality rates, 
reduced reproduction, and disease. Due to these effects, affected 
populations may require decades to recover absent successful efforts at 
restoration (e.g., DWH NRDA Trustees, 2017). The ability of the stocks 
to recover and the length of time required for that recovery are tied 
to the carrying capacity of the habitat, and to the degree of other 
population pressures. NMFS treats the effects of the DWH oil spill as 
part of the baseline in considering the likely resilience of these 
populations to the effects of the activities considered in this 
regulatory framework.
    Unusual Mortality Events (UME)--A UME is defined under Section 
410(6) of the MMPA as ``a stranding that is unexpected; involves a 
significant die-off of any marine mammal population; and demands 
immediate response.'' From 1991 to the present, there have been 
fourteen formally recognized UMEs affecting marine mammals in the 
region and involving species under NMFS' jurisdiction. These have 
primarily impacted coastal bottlenose dolphins, with multiple UMEs 
determined to have resulted from biotoxins and one from infectious 
disease. One relevant UME was declared since publication of the notice 
of proposed rulemaking and is discussed below.
    Most significantly, a UME affecting multiple cetacean species in 
the northern GOM occurred from 2010-2014. NMFS discussed this UME in 
the notice of proposed rulemaking (83 FR 29212; June 22, 2018). Please 
see that document for additional information regarding the 2010-2014 
UME. Additional information on the UME is also available online at: 
www.fisheries.noaa.gov/national/marine-life-distress/2010-2014-cetacean-unusual-mortality-event-northern-gulf-mexico. In summary, the 
event included all cetaceans stranded during this time in Alabama, 
Mississippi, and Louisiana and all cetaceans other than bottlenose 
dolphins stranded in the Florida Panhandle (Franklin County through 
Escambia County), with a total of 1,141 cetaceans stranded or reported 
dead offshore. For reference, the same area experienced a normal 
average of 75 strandings per year from 2002-09 (Litz et al., 2014). The 
majority of stranded animals were bottlenose dolphins, though at least 
ten additional species were reported as well. Since not all cetaceans 
that die wash ashore where they may be found, the number reported 
stranded is likely a fraction of the total number of cetaceans that 
died during the UME. There was also an increase in strandings of 
stillborn and newborn dolphins (Colegrove et al., 2016). The UME 
investigation and the Deepwater Horizon Natural Resource Damage 
Assessment determined that the DWH oil spill (discussed above) is the 
most likely explanation of the persistent, elevated stranding numbers 
in the northern GOM after the 2010 spill. The evidence to date supports 
that exposure to hydrocarbons released during the DWH oil spill was the 
most likely explanation of adrenal and lung disease in dolphins, which 
has contributed to increased deaths of dolphins living within the oil 
spill footprint and increased fetal loss. The longest and

[[Page 5394]]

most prolonged stranding cluster was in Barataria Bay, Louisiana in 
2010-11, followed by Mississippi and Alabama in 2011, consistent with 
timing and spatial distribution of oil, while the number of deaths was 
not elevated for areas that were not as heavily oiled. Subsequent 
health assessments of live dolphins from Barataria Bay and comparison 
to a reference population found significantly increased adrenal 
disease, lung disease, and poor health, while histological evaluations 
of samples from dead stranded animals from within and outside the UME 
area found that UME animals were more likely to have lung and adrenal 
lesions and to have primary bacterial pneumonia, which caused or 
contributed significantly to death (Schwacke et al., 2014a, 2014b; 
Venn-Watson et al., 2015b). The chronic adrenal gland and lung diseases 
identified in stranded UME dolphins are consistent with exposure to 
petroleum compounds (Venn-Watson et al., 2015b). Colegrove et al. 
(2016) found that the increase in perinatal strandings resulted from 
late-term pregnancy failures and development of in utero infections 
likely caused by chronic illnesses in mothers who were exposed to oil.
    While the number of dolphin mortalities in the area decreased after 
the peak from March 2010-July 2014, it does not indicate that the 
effects of the oil spill on these populations have ended. Researchers 
still saw evidence of chronic lung disease and adrenal impairment four 
years after the spill (in July 2014) and saw evidence of failed 
pregnancies in 2015 (Smith et al., 2017). These follow-up studies found 
a yearly mortality rate for Barataria Bay dolphins of roughly 13 
percent (as compared to annual mortality rates of 5 percent or less 
that have been previously reported for other dolphin populations) and 
found that only 20 percent of pregnant dolphins produced viable calves 
(compared with 83 percent in a reference population) (Lane et al., 
2015; McDonald et al., 2017). In addition, compromised health may make 
dolphins more susceptible to additional environmental stressors.
    Since the publication of the proposed rule, another UME involving 
bottlenose dolphins in the northern GOM was declared. Elevated 
bottlenose dolphin strandings occurred in Louisiana, Mississippi, 
Alabama, and the panhandle of Florida (Alabama border through Taylor 
County) from February 1, 2019, through November 30, 2019. A total of 
337 confirmed strandings were documented, with a majority occurring 
from February through May. Excluding prior UMEs, the annual average for 
February through May in the affected area is 57 dolphins; at least 260 
standings were documented during this period in 2019. The cause of the 
UME was determined to be environmentally driven by exposure to low 
salinity waters resulting from extreme freshwater discharge from 
watersheds that drain into the GOM, including rivers in Florida, 
Alabama, Mississippi and Louisiana. This unprecedented amount of 
freshwater discharge during the winter, spring, and summer months of 
2019 resulted in a drop in salinity levels across the coastally 
associated waters in the region. Prolonged exposure to low salinity 
water has been documented to have harmful health impacts on bottlenose 
dolphins, ranging from skin lesions and serum electrolyte abnormalities 
to acute mortality. The location of the UME and the dolphin stocks 
affected, including the western and northern coastal stocks of 
bottlenose dolphin, are the same as those impacted by the 2010-2014 
UME. For additional information, please visit www.fisheries.noaa.gov/national/marine-life-distress/2019-bottlenose-dolphin-unusual-mortality-event-along-northern-gulf.

Marine Mammal Hearing

    Hearing is the most important sensory modality for marine mammals 
underwater, and exposure to anthropogenic sound can have deleterious 
effects. To appropriately assess the potential effects of exposure to 
sound, it is necessary to understand the frequency ranges that marine 
mammals are able to hear. Current data indicate that not all marine 
mammal species have equal hearing capabilities (e.g., Richardson et 
al., 1995; Wartzok and Ketten, 1999; Au and Hastings, 2008). To reflect 
this, Southall et al. (2007) recommended that marine mammals be divided 
into functional hearing groups based on directly measured or estimated 
hearing ranges on the basis of available behavioral response data, 
audiograms derived using auditory evoked potential techniques, 
anatomical modeling, and other data. Note that no direct measurements 
of hearing ability have been successfully completed for mysticetes 
(i.e., low-frequency cetaceans). NMFS (2018) describes generalized 
hearing ranges for these marine mammal hearing groups. Generalized 
hearing ranges were chosen based on the approximately 65 dB threshold 
from the normalized composite audiograms, with the exception for lower 
limits for low-frequency cetaceans where the lower bound was deemed to 
be biologically implausible and the lower bound from Southall et al. 
(2007) retained. The functional groups and the associated frequencies 
are indicated below (note that these frequency ranges correspond to the 
range for the composite group, with the entire range not necessarily 
reflecting the capabilities of every species within that group):
     Low-frequency cetaceans (mysticetes): Generalized hearing 
is estimated to occur between approximately 7 Hz and 35 kHz;
     Mid-frequency cetaceans (larger toothed whales, beaked 
whales, and most delphinids): Generalized hearing is estimated to occur 
between approximately 150 Hz and 160 kHz;
     High-frequency cetaceans (porpoises, river dolphins, and 
members of the genera Kogia and Cephalorhynchus; including two members 
of the genus Lagenorhynchus, on the basis of recent echolocation data 
and genetic data): Generalized hearing is estimated to occur between 
approximately 275 Hz and 160 kHz.
    For more detail concerning these groups and associated frequency 
ranges, please see NMFS (2018) for a review of available information. 
Twenty-one species of cetacean have the reasonable potential to co-
occur with the proposed survey activities. Please refer to Table 4. Of 
the cetacean species that may be present, one is classified as a low-
frequency cetacean (i.e., the Bryde's whale), 18 are classified as mid-
frequency cetaceans (i.e., all delphinid and ziphiid species and the 
sperm whale), and two are classified as high-frequency cetaceans (i.e., 
Kogia spp.).

Potential Effects of the Specified Activities on Marine Mammals and 
Their Habitat

    In NMFS' notice of proposed rulemaking (83 FR 29212; June 22, 
2018), this section included a comprehensive summary and discussion of 
the ways that components of the specified activity may impact marine 
mammals and their habitat, including general background information on 
sound and specific discussion of potential effects to marine mammals 
from noise produced through use of airgun arrays. We incorporate by 
reference that information and do not repeat that discussion here, 
instead referring the reader to the notice of proposed rulemaking.
    The Estimated Take section later in this document includes a 
quantitative analysis of the number of individuals that are expected to 
be taken by the specified activity. The Negligible Impact Analysis and 
Determinations section includes an analysis of how these

[[Page 5395]]

activities will impact marine mammals and considers the content of this 
section, the Estimated Take section, and the Mitigation section, to 
draw conclusions regarding the likely impacts of these activities on 
the reproductive success or survivorship of individuals and from that 
on the affected marine mammal populations.

Description of Active Acoustic Sound Sources

    In the notice of proposed rulemaking, this section contained a 
brief technical background on sound, the characteristics of certain 
sound types, and on metrics used in the proposal inasmuch as the 
information is relevant to the specified activity and to a discussion 
of the potential effects of the specified activity on marine mammals 
found later in this document. Here, we summarize key information 
relating to terminology used in this notice.
    Amplitude (or ``loudness'') of sound is typically described using 
the relative unit of the decibel (dB). A sound pressure level (SPL) in 
dB is described as the ratio between a measured pressure and a 
reference pressure (for underwater sound, this is 1 microPascal 
([mu]Pa)). The source level (SL) represents the SPL referenced at a 
distance of 1 m from the source (referenced to 1 [mu]Pa), while the 
received level is the SPL at the listener's position (referenced to 1 
[mu]Pa).
    Root mean square (rms) is the quadratic mean sound pressure over 
the duration of an impulse. This measurement is often used in the 
context of discussing behavioral effects, in part because behavioral 
effects, which often result from auditory cues, may be better expressed 
through averaged units than by peak pressures. Sound exposure level 
(SEL; represented as dB re 1 [mu]Pa\2\-s) represents the total energy 
contained within a pulse, and considers both intensity and duration of 
exposure. Peak sound pressure (also referred to as zero-to-peak sound 
pressure or 0-p) is the maximum instantaneous sound pressure measurable 
in the water at a specified distance from the source, and is 
represented in the same units as the rms sound pressure. Another common 
metric is peak-to-peak sound pressure (pk-pk), which is the algebraic 
difference between the peak positive and peak negative sound pressures. 
Peak-to-peak pressure is typically approximately 6 dB higher than peak 
pressure (Southall et al., 2007).
    As described in more detail in the notice of proposed rulemaking, 
airgun arrays are in a general sense considered to be omnidirectional 
sources of pulsed noise. Pulsed sound sources (as compared with non-
pulsed sources) produce signals that are brief (typically considered to 
be less than one second), broadband, atonal transients (ANSI, 1986, 
2005; Harris, 1998; NIOSH, 1998; ISO, 2003) and occur either as 
isolated events or repeated in some succession. Pulsed sounds are all 
characterized by a relatively rapid rise from ambient pressure to a 
maximal pressure value followed by a rapid decay period that may 
include a period of diminishing, oscillating maximal and minimal 
pressures, and generally have an increased capacity to induce physical 
injury as compared with sounds that lack these features. Airguns 
produce sound with energy in a frequency range from about 10-2,000 Hz, 
with most energy radiated at frequencies below 200 Hz. Although the 
amplitude of the acoustic wave emitted from the source is equal in all 
directions (i.e., omnidirectional), airgun arrays do possess some 
directionality due to different phase delays between guns in different 
directions. Airgun arrays are typically tuned to maximize functionality 
for data acquisition purposes, meaning that sound transmitted in 
horizontal directions and at higher frequencies is minimized to the 
extent possible.
    Acoustic sources used for high-resolution geophysical (HRG) surveys 
generally produce higher frequency signals with highly directional beam 
patterns. These sources are generally considered to be intermittent, 
with typically brief signal durations. Boomers, considered to be 
impulsive sources, generate a high-amplitude broadband (100 Hz-10 kHz) 
acoustic pulse with high downward directivity, though may be considered 
omnidirectional at frequencies below 1 kHz. Other typical HRG sources 
are considered non-impulsive. Sub-bottom profiler systems generally 
project a chirp pulse spanning an operator-selectable frequency band, 
usually between 1 to 20 kHz, with a single beam directed vertically 
down. Multibeam echosounders use an array of transducers that project a 
high-frequency, fan-shaped beam under the hull of a survey ship and 
perpendicular to the direction of motion. Side-scan sonars use two 
transducers to project high-frequency beams that are usually wide in 
the vertical plane (50[deg]-70[deg]) and very narrow in the horizontal 
plane (less than a few degrees). Other, similar impulsive or non-
impulsive sources may be used in conducting shallow penetration or HRG 
surveys.

Acoustic Habitat

    NMFS also included a detailed discussion and analysis of potential 
impacts to acoustic habitat. Acoustic habitat is the soundscape--which 
encompasses all of the sound present in a particular location and time, 
as a whole--when considered from the perspective of the animals 
experiencing it. Animals listen for sounds produced by conspecifics 
(communication during feeding, mating, and other social activities), 
other animals (finding prey or avoiding predators), and the physical 
environment (finding suitable habitats, navigating). Together, sounds 
made by animals and the geophysical environment (e.g., produced by 
earthquakes, lightning, wind, rain, waves) make up the natural 
contributions to the total acoustics of a place. These acoustic 
conditions, termed acoustic habitat, are one attribute of an animal's 
total habitat.
    That discussion summarized a report titled ``Cumulative and Chronic 
Effects in the Gulf of Mexico: Estimating Reduction of Listening Area 
and Communication Space due to Seismic Activities,'' (``Cumulative and 
Chronic Effects report'') as well as a subsequent addendum to the 
report presenting additional analysis relating to sperm whales. The 
initial report (originally presented as Appendix K in BOEM (2017)) as 
well as the addendum ((hereafter, ``the CCE report''), are available 
online at www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico. The CCE 
report presented a first-order cumulative and chronic effects 
assessment for noise produced by oil and gas exploration activities in 
the U.S. GOM.
    The term ``listening area'' refers to the region of ocean over 
which sources of sound can be detected by an animal at the center of 
the space. Loss of communication space concerns the area over which a 
specific animal signal, used to communicate with conspecifics in 
biologically-important contexts (e.g., foraging, mating), can be heard, 
in noisier relative to quieter conditions (Clark et al., 2009). Lost 
listening area concerns the more generalized contraction of the range 
over which animals would be able to detect a variety of signals of 
biological importance, including eavesdropping on predators and prey 
(Barber et al., 2009). Implications for acoustic masking (i.e., when a 
sound interferes with or masks the ability of an animal to detect a 
signal of interest that is above the absolute hearing threshold; see 
notice of

[[Page 5396]]

proposed rulemaking at 29239 for explanation of masking) and reduced 
communication space resulting from noise produced by airgun surveys in 
the GOM are expected to be particularly heightened for animals that 
actively produce low-frequency sounds or whose hearing is attuned to 
lower frequencies (i.e., Bryde's whales).
    Acoustic modeling was conducted for ten locations (``receiver 
sites'') within the study area to examine aggregate noise produced over 
a full, generic year. The locations of the receiver sites were chosen 
to reflect areas of biological importance to cetaceans, areas of high 
densities of cetaceans, and areas of key biological diversity. The CCE 
report analyzed multiple scenarios, including a baseline scenario in 
which no geophysical surveys are conducted and noise consists of 
natural sounds and a minimum estimate of commercial vessel noise; a 
survey activity scenario in which projected activities were uniformly 
distributed throughout the study area, with the exception of coastal 
waters from February to May; and a closure scenario in which no 
activities are conducted in certain restriction areas, 25 percent of 
the activity that would have occurred in the restriction areas is 
redistributed into non-restriction areas of the same activity zone, and 
75 percent of the activities that would have occurred in the 
restriction areas are not conducted at all. For additional 
methodological details, see discussion in the notice of proposed 
rulemaking or the CCE report.
    Regarding sperm whales, the analysis shows that the survey 
activities do not significantly contribute to the soundscape in the 
frequency band relevant for their lower-frequency slow-clicks, and that 
there will be no significant change in communication space for sperm 
whales. Because other sperm whale calls are higher-frequency, they 
would not be expected to be affected. Please see the CCE report for 
further discussion of the findings for sperm whales. The remaining 
discussion that follows is in reference to the findings for Bryde's 
whales and to general findings for other hearing groups.
    The methods used in the CCE report were meant to average the 
conditions generated by low-frequency dominant noise sources throughout 
a full year, during which animals of key management interest rely on 
habitats within the study area. Considered as a complement to 
assessments of the acute effects of the same types of noise sources in 
the same region (discussed below in the Estimated Take section), the 
CCE assessment estimates noise produced by the same sources over much 
larger spatial scales, and considers how the summation of noise from 
these sources relates to levels without the proposed activity 
(ambient). The lost listening area method calculates a fractional 
reduction in listening area due to the addition of anthropogenic noise 
to ambient noise. Results are presented as a percentage of the original 
listening area remaining due to the increase in noise levels relative 
to no activity and between activity scenarios. The communication space 
assessment provides relative losses of communication space (in both 
areas and percentages) between the activity scenarios.
    At most sites, lost listening area was greater for deeper waters 
than for shallower waters, which is attributed to the downward-
refracting sound speed profile near the surface, caused by the 
thermocline, which steers sound to deeper depths. Shallow water noise 
levels were reduced due to surface interactions that increase 
transmission loss, particularly for low frequencies. Listening area 
reductions were also generally most severe when weighted for low-
frequency hearing cetaceans. Both low- and mid-frequency weighted 
losses were high in the Mississippi Canyon, while only low-frequency 
weighted values were high for the De Soto Canyon. Both of these sites 
are considered important to sperm whales as well as other deep-diving 
odontocetes. These modeling results suggest that accumulations of noise 
from survey activities below 5 kHz and often heightened at depth could 
be degrading the ability of animals that forage at great depths in the 
GOM to use acoustic cues to find prey as well as to maintain 
conspecific contact.
    Comparison between results provided for the two metrics applied in 
the CCE report highlights important interpretive differences for 
evaluating the biological implications of background noise. The 
strength of the communication space approach is that it evaluates 
potential contractions in the availability of a signal of documented 
importance to a population of animals of key management interest in the 
region. In this case, losses of communication space for Bryde's whales 
were estimated to be higher in eastern and central GOM canyons and 
shelf break areas. In contrast, relative maintenance of listening area 
and communication space was seen within the Bryde's whale core habitat 
area in the eastern GOM (an area that has since been removed from 
consideration through this rule). In areas where larger amounts of 
survey activity were projected, significant loss of low-frequency 
listening area and communication space for Bryde's whale calls was 
estimated, though we emphasize that these are not areas where Bryde's 
whales are expected to occur.
    The CCE report is described here in order to summarize information 
presented in the proposed rule regarding potential longer-term and 
wider-range noise effects from sources such as airguns. Please see the 
notice of proposed rulemaking, as well as the CCE report and addendum, 
for additional information.

Estimated Take

    This section provides an estimate of the number and type of 
incidental takes that may be expected to occur under the specified 
activity (as it has been revised in scope), which informed NMFS' 
negligible impact determination. Realized incidental takes would be 
determined by the actual levels of activity at specific times and 
places that occur under any issued LOAs.
    Except with respect to certain activities not pertinent here, 
section 3(18) of the MMPA defines ``harassment'' as: Any act of 
pursuit, torment, or annoyance which (i) has the potential to injure a 
marine mammal or marine mammal stock in the wild (Level A harassment); 
or (ii) has the potential to disturb a marine mammal or marine mammal 
stock in the wild by causing disruption of behavioral patterns, 
including, but not limited to, migration, breathing, nursing, breeding, 
feeding, or sheltering (Level B harassment). Harassment is the only 
type of take expected to result from these activities.
    Anticipated takes would primarily be by Level B harassment, as use 
of the described acoustic sources, particularly airgun arrays, is 
likely to disrupt behavioral patterns of marine mammals. There is also 
some potential for auditory injury (Level A harassment) to result for 
low- and high-frequency species due to the size of the predicted 
auditory injury zones for those species. NMFS does not expect auditory 
injury to occur for mid-frequency species, as discussed in greater 
detail on the notice of proposed rulemaking (83 FR 29212; June 22, 
2018) and in responses to public comments. The required mitigation and 
monitoring measures are expected to minimize the severity of such 
taking to the extent practicable. It is unlikely that lethal takes 
would occur even in the absence of the mitigation and monitoring 
measures, and no such takes are anticipated or will be authorized. 
Below we summarize how the take that may be authorized was estimated 
using acoustic thresholds, sound field modeling, and

[[Page 5397]]

marine mammal density data. Detailed discussion of all facets of the 
take estimation process was provided in the notice of proposed 
rulemaking (83 FR 29212; June 22, 2018), and nothing has changed since 
that time. Therefore, that full discussion is not repeated. Please see 
that notice, and associated companion documents available online, for 
additional detail.

Acoustic Thresholds

    NMFS uses acoustic thresholds that identify the received level of 
underwater sound above which exposed marine mammals generally would be 
reasonably expected to exhibit disruption of behavioral patterns (Level 
B harassment) or to incur permanent threshold shift (PTS) of some 
degree (Level A harassment).
    Level B Harassment--Although available data are consistent with the 
basic concept that louder sounds evoke more significant behavioral 
responses than softer sounds, defining precise sound levels that will 
potentially disrupt behavioral patterns is difficult because responses 
depend on the context in which the animal receives the sound, including 
an animal's behavioral mode when it hears sounds (e.g., feeding, 
resting, or migrating), prior experience, and biological factors (e.g., 
age and sex). Some species, such as beaked whales, are known to be more 
highly sensitive to certain anthropogenic sounds than other species. 
Other contextual factors, such as signal characteristics, distance from 
the source, duration of exposure, and signal to noise ratio, may also 
help determine response to a given received level of sound. Therefore, 
levels at which responses occur are not necessarily consistent and can 
be difficult to predict (Southall et al., 2007; Ellison et al., 2012; 
Bain and Williams, 2006). Typically, and especially in cases where PTS 
is predicted, NMFS anticipates that some number of individuals may 
incur temporary threshold shift (TTS) (considered Level B harassment). 
However, it is not necessary to separately quantify those takes, as it 
is unlikely that an individual marine mammal would be exposed at the 
levels and duration necessary to incur TTS without also being exposed 
to the levels associated with behavioral harassment and, therefore, 
NMFS expects any potential TTS takes to be captured by the estimated 
takes by behavioral harassment.
    Based on the practical need to use a relatively simple threshold 
based on available information that is both predictable and measurable 
for most activities, NMFS has historically used a generalized acoustic 
threshold based on received level to estimate the onset of Level B 
harassment. These thresholds are 160 dB rms (intermittent sources, 
which include impulsive sources) and 120 dB rms (continuous sources). 
Airguns are impulsive sound sources and electromechanical sources used 
for HRG surveys are intermittent sources. Therefore, the 160 dB rms 
threshold has typically been used in evaluating effects from the 
sources planned for use in the specified activities. However, in the 
notice of proposed rulemaking, NMFS identified a more complex 
probabilistic risk function for use in evaluating the potential effects 
of the specified activity considered herein. That function, described 
in Wood et al. (2012), is better reflective of available scientific 
information (as discussed in detail in the notice of proposed 
rulemaking, as well as in comment responses provided earlier in this 
preamble). Such an approach takes the fundamental step of acknowledging 
the potential for Level B harassment at exposures to received levels 
below 160 dB rms (as well as the potential that animals exposed to 
received levels above 160 dB rms will not respond in ways constituting 
Level B harassment). The approach described by Wood et al. (2012) also 
accounts for differential hearing sensitivity by incorporating 
frequency-weighting functions. The analysis of Gomez et al. (2016) 
indicates that behavioral responses in cetaceans are best explained by 
the interaction between sound source type and functional hearing group. 
Southall et al. (2007) proposed auditory weighting functions for 
species groups based on known and assumed hearing ranges (Type I). 
Although newer filters are better designed to predict the onset of 
auditory injury (as discussed below and used for evaluation of 
potential Level A harassment), the broader Type I filters were retained 
for use in evaluating potential behavioral disturbance in conjunction 
with the Wood et al. (2012) probabilistic response function.
    NMFS received public comments on this topic, including some 
criticizing the proposed use of the Wood et al. (2012) risk function. 
We responded to all comments received on this topic and, in addition to 
the more detailed discussion provided in the Estimated Take section of 
the notice of proposed rulemaking, we provide detailed discussion of 
these concerns in the responses to comments, provided earlier in this 
preamble. NMFS retains use of the Wood et al. (2012) approach as the 
basis for estimating take and considering the effects of the specified 
activity on marine mammal behavior. The Level B harassment criteria 
upon which the analysis presented herein is based are presented in 
Table 6.

                                      Table 6--Behavioral Exposure Criteria
----------------------------------------------------------------------------------------------------------------
                                                       Probability of response to frequency-weighted rms SPL
                      Group                      ---------------------------------------------------------------
                                                      120 (%)         140 (%)         160 (%)         180 (%)
----------------------------------------------------------------------------------------------------------------
Beaked whales...................................              50              90             n/a             n/a
All other species...............................             n/a              10              50              90
----------------------------------------------------------------------------------------------------------------

    Level A Harassment--NMFS' Technical Guidance for Assessing the 
Effects of Anthropogenic Sound on Marine Mammal Hearing (NMFS, 2018) 
(2018 Revised Technical Guidance) identifies dual criteria to assess 
the potential for auditory injury (Level A harassment) to occur for 
different marine mammal groups (based on hearing sensitivity) as a 
result of exposure to noise. The 2018 Revised Technical Guidance 
identifies the received levels, or thresholds, above which individual 
marine mammals are predicted to experience changes in their hearing 
sensitivity for all underwater anthropogenic sound sources and reflects 
the best available science on the potential for noise to affect 
auditory sensitivity by:
     Dividing sound sources into two groups (i.e., impulsive 
and non-impulsive) based on their potential to affect hearing 
sensitivity;
     Choosing metrics that best address the impacts of noise on 
hearing sensitivity, i.e., peak sound pressure level (peak SPL) 
(reflects the physical properties of impulsive sound sources to affect 
hearing sensitivity) and

[[Page 5398]]

cumulative sound exposure level (cSEL) (accounts for not only level of 
exposure but also duration of exposure); and
     Dividing marine mammals into hearing groups and developing 
auditory weighting functions based on the science that indicates that 
not all marine mammals hear and use sound in the same manner.
    The premise of the dual criteria approach is that, while there is 
no definitive answer to the question of which acoustic metric is most 
appropriate for assessing the potential for injury, both the received 
level and duration of received signals are important to an 
understanding of the potential for auditory injury. Therefore, peak SPL 
is used to define a pressure criterion above which auditory injury is 
predicted to occur, regardless of exposure duration (i.e., any single 
exposure at or above this level is considered to cause auditory 
injury), and cSEL is used to account for the total energy received over 
the duration of sound exposure (i.e., both received level and duration 
of exposure) (Southall et al., 2007, 2019a; NMFS, 2018). As a general 
principle, whichever criterion is exceeded first (i.e., results in the 
largest isopleth) would be used as the effective injury criterion 
(i.e., the more precautionary of the criteria). Note that cSEL acoustic 
threshold levels incorporate marine mammal auditory weighting 
functions, while peak pressure thresholds do not (i.e., flat or 
unweighted). Weighting functions for each hearing group (e.g., low-, 
mid-, and high-frequency cetaceans) are described in NMFS (2018).
    The 2018 Revised Technical Guidance recommends 24 hours as a 
maximum accumulation period relative to cSEL thresholds. These 
thresholds were developed by compiling and synthesizing the best 
available science, and are provided in Table 7 below. The references, 
analysis, and methodology used in the development of the thresholds are 
described in NMFS (2018), and more information is available online at: 
www.fisheries.noaa.gov/national/marine-mammal-protection/marine-mammal-acoustic-technical-guidance.

                                 Table 7--Exposure Criteria for Auditory Injury
----------------------------------------------------------------------------------------------------------------
                                                                                     Cumulative sound exposure
                                                                                             level \2\
                          Hearing group                           Peak  pressure -------------------------------
                                                                     \1\ (dB)                      Non-impulsive
                                                                                  Impulsive (dB)       (dB)
----------------------------------------------------------------------------------------------------------------
Low-frequency cetaceans.........................................             219             183             199
Mid-frequency cetaceans.........................................             230             185             198
High-frequency cetaceans........................................             202             155             173
----------------------------------------------------------------------------------------------------------------
\1\ Referenced to 1 [mu]Pa; unweighted within generalized hearing range.
\2\ Referenced to 1 [mu]Pa\2\-s; weighted according to appropriate auditory weighting function. Airguns and the
  boomer are treated as impulsive sources; other HRG sources are treated as non-impulsive.

    NMFS considers these updated thresholds and associated weighting 
functions to be the best available information for assessing whether 
exposure to specific activities is likely to result in changes in 
marine mammal hearing sensitivity.

Modeling Overview

    Zeddies et al. (2015, 2017a) (i.e., ``the modeling report'') 
provides estimates of the annual marine mammal acoustic exposure caused 
by sounds from geophysical survey activity in the GOM for ten years of 
notional activity levels. Here we provide a brief overview of key 
modeling elements, with more detail provided in the notice of proposed 
rulemaking (83 FR 29212; June 22, 2018). For full details of the 
modeling effort, the interested reader should see the report (available 
online at: www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico).
    Initial phases of the modeling effort involved preliminary modeling 
of a typical 3D WAZ survey, which was simulated at two locations in 
order to establish the basic methodological approach and to provide 
results used to evaluate test scenarios that could influence exposure 
estimates. We discussed each of the six evaluated test scenarios in the 
notice of proposed rulemaking. Please see that discussion and the 
modeling report for full details.
    The modeling effort produced exposure estimates computed from 
modeled sound levels as received by simulated animals (animats) in a 
specific modeling area. The GOM was divided into seven modeling zones 
with six survey types simulated within each zone to estimate the 
potential effects of each survey. The zones were designed as described 
previously (Description of the Specified Activity; Figure 3)--shelf and 
slope waters were divided into eastern, central, and western zones, 
plus a single deep-water zone--to account for both the geospatial 
dependence of acoustic fields and the geographic variations of animal 
distributions. The selected boundaries considered sound propagation 
conditions and species distribution to create regions of optimized 
uniformity in both acoustic environment and animal density. Survey 
types included deep penetration surveys using a large airgun array (2D, 
3D NAZ, 3D WAZ, and coil survey types), shallow penetration surveys 
using a single airgun (which were assumed to be a reasonable proxy for 
surveys conducted using a boomer), and high resolution surveys 
concurrently using a CHIRP sub-bottom profiler, side-scan sonar, and 
multibeam echosounder. The results from each zone were summed to 
provide GOM-wide estimates of take for each marine mammal species for 
each survey type for each notional year. To get these annual aggregate 
exposure estimates, 24-hr average exposure estimates from each survey 
type were multiplied by the number of expected survey days from BOEM's 
effort projections. Because these projections are not season-specific, 
surveys were assumed to be equally likely to occur at any time of the 
year and at any location within a given zone.
    Acoustic source emission levels and directivity of a single airgun 
and an airgun array were modeled using JASCO Applied Sciences' Airgun 
Array Source Model (AASM). AASM is capable of predicting airgun source 
levels at frequencies up to 25 kHz, and produces a set of notional 
signatures for each array element based on array layout; volume, tow 
depth, and firing pressure for each element; and interactions between 
different elements in the array. The signatures are summed to obtain 
the far-field source signature of the entire array in the horizontal 
plane, which is then filtered into one third-octave frequency bands to 
compute the source

[[Page 5399]]

levels of the array as a function of frequency band and azimuthal angle 
in the horizontal plane (at the source depth), after which it is 
considered to be an azimuth-dependent directional point source in the 
far field. Source levels for high-resolution sources were obtained from 
manufacturer's specifications for representative sources. 
Electromechanical sources were modeled on the basis of transducer beam 
theory, which is often used to estimate beam pattern of the source in 
the absence of field measurements, and which is described in detail in 
the modeling report.
    Underwater sound propagation (i.e., transmission loss) as a 
function of range from each source was modeled using JASCO's Marine 
Operations Noise Model (MONM) for multiple propagation radials centered 
at the source to yield 3D transmission loss fields in the surrounding 
area. The MONM computes received per-pulse SEL for directional sources 
at specified depths. MONM uses two separate models to estimate 
transmission loss. At frequencies less than 2 kHz, MONM computes 
acoustic propagation via a wide-angle parabolic equation (PE) solution 
to the acoustic wave equation (Collins, 1993), based on a version of 
the U.S. Naval Research Laboratory's Range-dependent Acoustic Model 
(RAM) modified to account for an elastic seabed (Zhang and Tindle, 
1995). MONM-RAM incorporates bathymetry, underwater sound speed as a 
function of depth, and a geoacoustic profile based on seafloor 
composition, and accounts for source horizontal directivity. At 
frequencies greater than 2 kHz, MONM accounts for increased sound 
attenuation due to volume absorption at higher frequencies (Fisher and 
Simmons, 1977) with the widely-used BELLHOP Gaussian beam ray-trace 
propagation model (Porter and Liu, 1994). This component incorporates 
bathymetry and underwater sound speed as a function of depth with a 
simplified representation of the sea bottom, as sub-bottom layers have 
a negligible influence on the propagation of acoustic waves with 
frequencies above 1 kHz. MONM-BELLHOP accounts for horizontal 
directivity of the source and vertical variation of the source beam 
pattern. Both propagation models account for full exposure from a 
direct acoustic wave, as well as exposure from acoustic wave 
reflections and refractions (i.e., multi-path arrivals at the 
receiver).
    In order to accurately estimate exposure, a simulation must 
adequately cover the various location- and season-specific 
environments. The surveys may be conducted at any location within the 
planning area and occur at any time of the year, so simulations must 
adequately cover each area and time period. The seven zones within 
which potential exposures were modeled, corresponding with shelf and 
slope environments subdivided into western, central, and eastern areas, 
as well as a single deep zone, were previously introduced (Figure 3). 
The subdivision depth definitions are: Shelf, 0-200 m; slope, 200-2,000 
m; and deep, greater than 2,000 m. Within each of the seven zones, a 
set of representative survey-simulation rectangles for each of the 
survey types was defined, with larger areas for the ``large-area'' 
surveys (i.e., deep penetration airgun) and smaller areas for the 
``small-area'' surveys (i.e., shallow penetration airgun and HRG). In 
Figure 3, the smaller numbered boxes represent the survey area extents 
for the different survey types. The stars represent acoustic modeling 
sites along western, central, and eastern transects (Figure 3).
    A set of 30 sites was selected to calculate acoustic propagation 
loss grids as functions of source, range from the source, azimuth from 
the source, and receiver depth. These were then used as inputs to the 
acoustic exposure model. The environmental parameters and acoustic 
propagation conditions represented by these 30 modeling sites were 
chosen to be representative of the prevalent acoustic propagation 
conditions within the survey extents. To account for seasonal variation 
in propagation, winter (most conservative) and summer (least 
conservative) were both used to calculate exposure estimates. 
Propagation during spring and fall was found to be almost identical to 
the results for summer, so those seasons were represented with the 
summer results. The primary seasonal influence on transmission loss is 
the presence of a sound channel, or duct, near the surface in winter.

Marine Mammal Density Information

    The best available scientific information was considered in 
conducting marine mammal exposure estimates (the basis for estimating 
take). Roberts et al. (2016) provided several key improvements over 
information previously available for the GOM, by incorporating NMFS 
aerial and shipboard survey data collected over the period 1992-2009; 
controlling for the influence of sea state, group size, availability 
bias, and perception bias on the probability of making a sighting; and 
modeling density from an expanded set of eight physiographic and 16 
dynamic oceanographic and biological covariates. There are multiple 
reasons why marine mammals may be undetected by observers. Animals are 
missed because they are underwater (availability bias) or because they 
are available to be seen, but are missed by observers (perception and 
detection biases) (e.g., Marsh and Sinclair, 1989). Negative bias on 
perception or detection of an available animal may result from 
environmental conditions, limitations inherent to the observation 
platform, or observer ability. Therefore, failure to correct for these 
biases may lead to underestimates of cetacean abundance (as is the case 
for NMFS' SAR abundance estimates for the GOM). Additional data was 
used to improve detection functions for taxa that were rarely sighted 
in specific survey platform configurations. The degree of 
underestimation would likely be particularly high for species that 
exhibit long dive times or are cryptic, such as sperm whales, beaked 
whales, or Kogia spp. In summary, consideration of additional survey 
data and an improved modeling strategy allowed for an increased number 
of taxa modeled and better spatiotemporal resolutions of the resulting 
predictions. More information concerning the Roberts et al. (2016) 
models, including the model results and supplementary information for 
each model, is available online at seamap.env.duke.edu/models/Duke-EC-GOM-2015/.

Description of Exposure Estimates

    The sound received by an animal when near a sound source is a 
function of the animal's position relative to the source, and both 
source and animals may be moving. To a reasonable approximation, we 
know, predict, or specify the location of the sound source, a 3D sound 
field around the source, and the expected occurrence of animals within 
100 km\2\ grid cells (Roberts et al., 2016). However, because the 
specific location of animals within the modeled sound field is unknown, 
agent-based animal movement modeling is necessary to complete the 
assessment of potential acoustic exposure. Realistic animal movement 
within the sound field can be simulated, and repeated random sampling 
(Monte Carlo)--achieved by simulating many animals within the 
operations area--used to estimate the sound exposure history of animals 
during the operation. Animats are randomly placed, or seeded, within 
the simulation boundary at a specified density, and the probability of 
an event's occurrence is determined by the frequency with which it 
occurs in the simulation. Higher densities provide a finer resolution 
for an estimate of the probability distribution function (PDF),

[[Page 5400]]

but require greater computational resources. To ensure good 
representation of the PDF, the animat density is set as high as is 
practical, with the resulting PDF then scaled using the real-world 
animal density (Roberts et al., 2016) to obtain the real-world number 
of modeled acoustic exposures.
    Several models for marine mammal movement have been developed 
(e.g., Frankel et al., 2002, Gisiner et al., 2006; Donovan et al., 
2013). Animats transition from one state to another, with user-
specified parameters representing simple states, such as the speed or 
heading of the animal, or complex states, such as likelihood of an 
animal foraging, playing, resting, or traveling. This analysis uses the 
Marine Mammal Movement and Behavior (3MB) model (Houser, 2006). 
Parameter values to control animat movement are typically determined 
using available species-specific behavioral studies, but the amount and 
quality of available data varies by species. While available data often 
provides a detailed description of the proximate behavior expected for 
real individual animals, species with more available information must 
be used as surrogates for those without sufficient available 
information. In this study, pantropical spotted dolphins are used as a 
surrogate for Clymene, spinner, and striped dolphins; short-finned 
pilot whales are surrogates for Fraser's dolphins, Kogia spp., and 
melon-headed whales; and rough-toothed dolphins are surrogates for 
false killer whales and pygmy killer whales. Observational data for all 
remaining species in the study were sufficient to determine animat 
movement.
    Species-specific animats were created with programmed behavioral 
parameters describing dive depth, surfacing and dive durations, 
swimming speed, course change, and behavioral aversions (e.g., water 
too shallow). The programmed animats were then randomly distributed 
over a given bounded simulation area. Because the exact positions of 
sound sources and animals are not known in advance for proposed 
activities, multiple runs of realistic predictions are used to provide 
statistical validity to the simulated scenarios. Each species-specific 
simulation was seeded with approximately 0.1 animats/km\2\ which, in 
most cases, represents a higher density of animats in the simulation 
than occurs in the real environment. A separate simulation was created 
and run for each combination of location, survey movement pattern, and 
marine mammal species. Animats were only allowed to be `taken' once 
during a 24-hour evaluation period. That is, an animat whose received 
level exceeds the peak SPL threshold more than once during an 
evaluation period was only counted once. Energy accumulation for SEL 
occurred throughout the 24-hour integration period and was reset at the 
beginning of each period. Similarly, the maximum received rms SPL was 
determined for the entirety of the evaluation period and reset at the 
beginning of each period.
    The JASCO Exposure Modeling System (JEMS) combined animal movement 
data (i.e., the output from 3MB), with pre-computed acoustic fields. 
The JEMS output was the time-history of received levels and slant 
ranges (the three-dimensional distance between the animat and the 
source) for all animats of the 3MB simulation. Animat received levels 
and slant ranges are used to determine the risk of acoustic exposure. 
There were many animats in the simulations and together their received 
levels represent the probability, or risk, of exposure for each survey.
    All survey simulations were for 7 days and a sliding 4-hr window 
approach was used to get the average 24-hr exposure. In this sliding-
window approach, 42 exposure estimate samples are obtained for each 
seven-day simulation, with the mean value then used as the 24-hr 
exposure estimate for that survey. The 24-hr exposure levels were then 
scaled by the projected level of effort for each survey type (i.e., 
multiplied by the number of days) to calculate associated annual 
exposure levels. The number of individual animals expected to exceed 
threshold during the 24-hr window is the number of animats exposed to 
levels exceeding threshold multiplied by the ratio of real-world animal 
density to model animat density.
    Injury--To evaluate the likelihood an animal might experience 
auditory injury as a result of accumulated sound energy, the cSEL for 
each animat in the simulation was calculated. To obtain that animat's 
cSEL, the SEL an animat received from each source over the 24-hr 
integration window was summed, and the number of animats whose cSEL 
exceeded the specified thresholds (Table 7) during the integration 
window was counted. To evaluate the likelihood an animal might be 
injured via exposure to peak SPL, the range at which the specific peak 
SPL threshold (Table 7) occurs for each source based on the broadband 
peak SPL source level was estimated. For each 24-hr integration window, 
the number of animats that came within this range of the source was 
counted.
    Behavior--To evaluate the likelihood an animal might experience 
disruption of behavioral patterns (i.e., a ``take''), the number of 
animats that received a maximum rms SPL exposure within the specified 
step ranges (Table 6) was calculated. The number of animats with a 
maximum rms SPL received level categorized into each bin of the step 
function was multiplied by the probability of the behavioral response 
specific to that range (Table 6). Specifically, 10 percent of animals 
exposed to received levels from 140-159 dB rms would be assumed as 
``takes,'' while 50 percent exposed to levels between 160-179 dB rms 
and 90 percent exposed to levels of 180 dB rms and above would be. The 
totals within each bin were then summed as the total estimated number 
of exposures above Level B harassment thresholds. This process was 
repeated for each 24-hr integration window. For beaked whales, for 
which lower behavioral harassment thresholds are designated, 50 percent 
of animals exposed to received levels from 120-149 dB rms would be 
assumed as ``takes,'' while 90 percent exposed to levels of 140 dB rms 
and above would be.

Take Estimates

    In summary, BOEM provided estimated levels of effort for 
geophysical survey activity in the GOM for a notional ten-year period. 
Exposure estimates were then computed from modeled sound levels 
received by animats for several representative types of geophysical 
surveying. Because animals and acoustic sources move relative to the 
environment and each other, and the sound fields generated by the 
sources are shaped by various physical parameters, the sound levels 
received by an animal are a complex function of location and time. The 
basic modeling approach was to use acoustic models to compute the 3D 
sound fields and their variations in time. Animats were modeled moving 
through these fields to sample the sound levels in a manner similar to 
how real animals would experience these sounds. From the time histories 
of the received sound levels of all animats, the numbers of animals 
exposed to levels exceeding effects threshold criteria were determined 
and then adjusted by the number of animals expected in the area, based 
on density information, to estimate the potential number of real-world 
marine mammal exposures to levels above the defined criteria. The 
acoustic exposure history of many simulated animals (animats) allows 
for the estimation of potential exposures due to operations. These 
modeled exposures are summed and represent the aggregate exposures that 
may result

[[Page 5401]]

from future surveys given the specified levels of effort for each 
survey type in each year and may vary according to the statistical 
distribution associated with these mean annual exposures.
    Exposure estimates above Level A and Level B harassment criteria, 
developed by Zeddies et al. (2015, 2017a) in association with the 
activity projections for the various annual effort scenarios, were 
generated based on the specific modeling scenarios (including source 
and survey geometry), i.e., 2D survey (1 x 8,000 in\3\ array), 3D NAZ 
survey (2 x 8,000 in\3\ array), 3D WAZ survey (4 x 8,000 in\3\ array), 
coil survey (4 x 8,000 in\3\ array), shallow penetration survey (either 
single 90 in\3\ airgun or boomer), and HRG surveys (side-scan sonar, 
multibeam echosounder, and sub-bottom profiler). Annual effort 
scenario-based pooled exposure estimates are therefore available by 
species.
    NMFS presented BOEM's original 10-year activity projections in 
Table 1 of the notice of proposed rulemaking under ``Detailed 
Description of Activities.'' For purposes of analysis in the notice of 
proposed rulemaking, NMFS identified representative ``high,'' 
``moderate,'' and ``low'' effort years. Because the duration of these 
regulations are limited to five years, NMFS needed to determine a 
reasonable basis for evaluating acoustic exposures that might occur 
during that timeframe (rather than ten years). Therefore, for the 
proposed rule, in recognition of relatively low recent levels of 
geophysical survey activity, from the ten notional years of projected 
survey effort provided by BOEM, NMFS selected five representative years 
representing three different potential levels of survey effort as the 
basis for the assessment. These included one ``high-activity'' year, 
two separate ``moderate-activity'' years, and two separate ``low-
activity'' years. Because the first 5 years of BOEM's original effort 
projections were relatively high-effort years, NMFS' level-of-effort 
selections for the proposed rule corresponded with the detailed per-
survey type effort projections given for Years 1, 4, 5, 8, and 9, 
respectively. Exposure estimates resulting from the process summarized 
here and corresponding with those activity scenarios were shown in 
Table 8 of the notice of proposed rulemaking. These exposure estimates 
were then further evaluated to provide an estimate of takes of marine 
mammals that could occur as a result of a reasonably expected level of 
geophysical survey activity in the GOM over the course of five years. 
Take estimates associated with those scenarios, which informed the 
analysis in the proposed rule, are shown in Table 8 of this document 
for reference. These values have been updated from those shown in Table 
8 of the notice of proposed rulemaking by correctly incorporating 
discounted estimates of Level A harassment into the estimates of Level 
B harassment (as pointed out by public commenters).

Level A Harassment

    As we explain here, the modeled exposure estimates for onset of 
permanent threshold shift (i.e., Level A harassment), are not expected 
to represent realistic results for any species. Overall, there is a low 
likelihood of take by Level A harassment for any species, though the 
degree of this low likelihood is primarily influenced by the specific 
hearing group. For mid- and high-frequency cetaceans, potential 
auditory injury would be expected to occur on the basis of 
instantaneous exposure to peak pressure output from an airgun array 
while, for low-frequency cetaceans, potential auditory injury would 
occur on the basis of the accumulation of energy output over time by an 
airgun array. Importantly, the modeled exposure estimates do not 
account for either aversion or the beneficial impacts of the required 
mitigation measures.
    Of even greater import for mid-frequency cetaceans is that the 
small calculated Level A harassment zone size in conjunction with the 
properties of sound fields produced by arrays in the near field versus 
far field leads to a logical conclusion that Level A harassment is so 
unlikely for species in this hearing group as to be discountable. As 
stated in the notice of proposed rulemaking, for all mid-frequency 
cetaceans, following evaluation of the available scientific literature 
regarding the auditory sensitivity of mid-frequency cetaceans and the 
properties of airgun array sound fields, NMFS does not expect any 
reasonable potential for Level A harassment to occur. We discussed this 
issue in detail earlier in the response to public comments. NMFS 
expects the potential for Level A harassment of mid-frequency cetaceans 
to be discountable, even before the likely moderating effects of 
aversion are considered. When considering potential for aversion, NMFS 
does not believe that Level A harassment is a likely outcome for any 
mid-frequency cetacean (as reflected in Table 9).
    As discussed in greater detail in the notice of proposed 
rulemaking, NMFS and BOEM considered the possibility of incorporating 
quantitative adjustments within the modeling process to account for the 
effects of mitigation and/or aversion, as both of these factors would 
lead to a reduction in likely injurious exposure. However, these 
factors were ultimately not quantified in the modeling because, in 
summary, there is too much inherent uncertainty regarding the 
effectiveness of detection-based mitigation to support any reasonable 
quantification of its effect in reducing injurious exposure and there 
is too little information regarding the likely level of onset and 
degree of aversion to quantify this behavior in the modeling process. 
This does not mean that mitigation is not effective (to some degree) in 
avoiding incidents of Level A harassment, nor does it mean that 
aversion is not a meaningful real-world effect of noise exposure that 
should be expected to reduce the number of incidents of Level A 
harassment. However, certain public commenters misconstrued statements 
in the notice of proposed rulemaking regarding the strictly modeling-
related investigations of aversion (i.e., that there is not sufficient 
quantitative data to inform decisions regarding the programming of 
animats as far as received levels of noise that provoke aversive 
response, and the degree of response, for relevant species) as meaning 
that there is not sufficient information to support that aversion 
happens at all. To the contrary, there is ample evidence in the 
literature that aversion is one of the most common responses to noise 
exposure across varied species, though the onset and degree may be 
expected to vary across individuals and in different contexts. 
Therefore, NMFS proposed to incorporate a reasonable adjustment to 
modeled Level A harassment exposure estimates to account for aversion 
for low- and high-frequency species. That adjustment is retained here, 
as discussed in greater detail in the responses to public comments. 
Specifically, NMFS assumes here that an eighty percent reduction in 
modeled exposure estimates for Level A harassment for low- and high-
frequency cetaceans is reasonable and likely conservative in terms of 
the overall numbers of actual incidents of Level A harassment for these 
species, as the adjustment does not explicitly account for the effects 
of mitigation.
    As discussed previously, BOEM provided an update to the scope of 
their proposed action through removal of the area subject to leasing 
moratorium under GOMESA from consideration in the rule. In support of 
this revision, BOEM provided revised 5-year level of effort predictions 
and associated acoustic exposure estimates. BOEM's process for 
developing this information,

[[Page 5402]]

described in detail in ``Revised Modeled Exposure Estimates,'' 
available online, was straightforward. Rather than using the PEIS's 10-
year period, BOEM provided revised levels of effort for a 5-year 
period, using Years 1-5 of the original level of effort projections. 
BOEM stated that the first five years were selected to be carried 
forward ``because they were contiguous, they included the three years 
with the most activity, and they were the best understood in relation 
to the historical data upon which they are based.'' NMFS concurs with 
this choice. Levels of effort were revised based on the basic 
assumption that if portions of areas are removed from consideration, 
then the corresponding effort previously presumed to occur in those 
areas also is removed from consideration. The revised levels of effort 
are shown in Table 2. Associated revised take estimates, which were 
generated utilizing the methods described above and in the proposed 
rule and inform the analysis in this final rule, are shown in Table 9. 
These estimates have been modified from the values provided by BOEM 
(available online; ``Revised Modeled Exposure Estimates'') in that we 
have correctly accounted for the type of taking expected, i.e., for 
mid-frequency cetaceans, Level A harassment is not expected to occur 
and the calculated takes have been shifted into the totals for Level B 
harassment. No incidents of Level A harassment for Bryde's whales were 
predicted under the revised effort scenarios, which exclude the area 
where most Bryde's whales would be expected to be found. For Kogia 
spp., estimates of Level A harassment were adjusted as discussed 
previously to account for likely aversion, and the portion of estimated 
Level A harassment events not expected to occur were shifted into the 
totals for Level B harassment for these species.
    Estimated instances of take, i.e., scenario-specific acoustic 
exposure estimates incorporating the adjustments to Level A harassment 
exposure estimates discussed here, are shown in Table 9. This 
information regarding total number of takes (with Level A harassment 
takes based on assumptions relating to mid-frequency cetaceans in 
general as well as aversion), on an annual basis for five years, 
provides the bounds within which incidental take authorizations may be 
issued in association with this regulatory framework.
    Typically, and especially in cases where PTS is predicted, NMFS 
anticipates that some number of individuals may incur TTS. However, it 
is not necessary to separately quantify those takes, as it is unlikely 
that an individual marine mammal would be exposed at the levels and 
duration necessary to incur TTS without also being exposed to the 
levels associated with behavioral disruption and, therefore, NMFS 
expects any potential TTS takes to be captured by the estimated takes 
by behavioral disruption (discussed below).

[[Page 5403]]



                         Table 8--Scenario-Specific Instances of Take (by Level A and Level B Harassment) and Mean Annual Take Levels Evaluated in the Proposed Rule \1\
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                         Survey effort scenario \2\
                                           -----------------------------------------------------------------------------------------------------------------------------------------------------
                  Species                             High                 Moderate #1              Moderate #2                 Low #1                   Low #2              Mean annual take
                                           -----------------------------------------------------------------------------------------------------------------------------------------------------
                                               A            B           A            B           A            B           A            B           A            B           A            B
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Bryde's whale.............................        3             572        2             422        2             510        2             395        2             411        2             462
Sperm whale...............................        0          43,549        0          27,300        0          33,378        0          26,681        0          27,689        0          31,719
Kogia spp.................................      728          19,101      475          13,328      636          16,188      472          12,629      562          13,414      575          14,932
Beaked whale..............................        0         235,667        0         162,172        0         190,824        0         151,745        0         156,622        0         179,406
Rough-toothed dolphin.....................        0          37,816        0          30,306        0          31,231        0          28,775        0          26,420        0          30,910
Bottlenose dolphin........................        0         655,345        0         979,905        0         598,607        0         941,001        0         581,121        0         751,196
Clymene dolphin...........................        0         111,211        0          73,225        0          87,995        0          69,913        0          73,051        0          83,079
Atlantic spotted dolphin..................        0         133,758        0         175,128        0         116,988        0         165,221        0         110,126        0         140,244
Pantropical spotted dolphin...............        0         609,653        0         421,786        0         513,572        0         401,568        0         421,856        0         473,687
Spinner dolphin...........................        0          83,041        0          59,818        0          73,259        0          56,735        0          59,448        0          66,460
Striped dolphin...........................        0          44,232        0          30,069        0          36,431        0          28,652        0          30,023        0          33,881
Fraser's dolphin..........................        0          13,910        0           9,690        0          11,438        0           9,162        0           9,426        0          10,725
Risso's dolphin...........................        0          27,165        0          18,197        0          22,005        0          17,380        0          18,166        0          20,583
Melon-headed whale........................        0          69,152        0          47,719        0          57,004        0          45,011        0          46,801        0          53,137
Pygmy killer whale........................        0          18,112        0          12,335        0          14,859        0          11,733        0          12,198        0          13,847
False killer whale........................        0          25,622        0          17,708        0          20,922        0          16,849        0          17,239        0          19,668
Killer whale..............................        0           1,498        0           1,034        0           1,262        0             987        0           1,039        0           1,164
Short-finned pilot whale..................        0          19,326        0          12,198        0          14,214        0          11,565        0          11,942        0          13,849
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ A and B refer to expected scenario-based instances of take by Level A and Level B harassment, respectively. For the Bryde's whale and Kogia spp., expected takes by Level A harassment
  represent modeled exposures adjusted to account for aversion. For Kogia spp., exposures above Level A harassment criteria were predicted by the peak SPL metric. For the Bryde's whale,
  exposures above Level A harassment criteria were predicted by the cSEL metric.
\2\ High survey effort scenario corresponds with level of effort projections given previously for Year 1 (see Table 1 of the notice of proposed rulemaking). Moderate #1 and #2 and Low #1 and
  #2 correspond with Years 4, 5, 8, and 9, respectively.


                                       Table 9--Revised Instances of Take (by Level A and Level B Harassment) (Years 1-5) and Mean Annual Take Levels \1\
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                                     Year 1                   Year 2                   Year 3                   Year 4                   Year 5              Mean annual take
                  Species                  -----------------------------------------------------------------------------------------------------------------------------------------------------
                                               A            B           A            B           A            B           A            B           A            B           A            B
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Bryde's whale.............................        0              10        0               8        0               8        0               6        0               7        0               8
Sperm whale...............................        0          16,405        0          14,205        0          13,603        0           9,496        0          12,388        0          13,219
Kogia spp.................................      371          10,383      337           9,313      310           8,542      209           6,238      314           8,318      308           8,559
Beaked whale..............................        0         191,566        0         162,301        0         158,328        0         111,415        0         142,929        0         153,308
Rough-toothed dolphin.....................        0          30,640        0          27,024        0          25,880        0          19,620        0          23,219        0          25,277
Bottlenose dolphin........................        0         603,649        0         973,371        0         567,962        0       1,001,256        0         567,446        0         742,737
Clymene dolphin...........................        0          85,828        0          67,915        0          73,522        0          47,332        0          60,379        0          66,995
Atlantic spotted dolphin..................        0         128,299        0         183,717        0         112,120        0         191,495        0         111,305        0         145,387
Pantropical spotted dolphin...............        0         478,490        0         436,047        0         391,363        0         311,316        0         395,987        0         402,641
Spinner dolphin...........................        0          75,953        0          71,873        0          61,098        0          48,775        0          64,357        0          64,411
Striped dolphin...........................        0          33,573        0          29,275        0          27,837        0          20,136        0          26,056        0          27,375
Fraser's dolphin..........................        0           4,522        0           3,843        0           3,792        0           2,726        0           3,455        0           3,668
Risso's dolphin...........................        0          21,859        0          18,767        0          18,218        0          12,738        0          16,634        0          17,643
Melon-headed whale........................        0          55,813        0          47,784        0          46,584        0          32,581        0          42,224        0          44,997
Pygmy killer whale........................        0           8,079        0           6,964        0           6,764        0           4,970        0           6,277        0           6,611
False killer whale........................        0          16,165        0          13,710        0          13,604        0           9,664        0          12,269        0          13,082
Killer whale..............................        0              60        0              56        0              50        0              42        0              52        0              52
Short-finned pilot whale..................        0          15,045        0           9,824        0          13,645        0           7,459        0           8,959        0          10,986
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ A and B refer to expected instances of take by Level A and Level B harassment, respectively, for Years 1-5. For Kogia spp., expected takes by Level A harassment represent modeled exposures
  adjusted to account for aversion. For the Bryde's whale, no takes by Level A harassment are predicted to occur. Therefore, no adjustment to modeled exposures to account for aversion was
  necessary. For Kogia spp., exposures above Level A harassment criteria were predicted by the peak SPL metric. For the Bryde's whale, the cSEL metric is used to evaluate the potential for
  Level A harassment.


[[Page 5404]]

Level B Harassment

    NMFS has determined the estimated values shown in Table 9 are a 
reasonable representation of the potential instances of take that may 
occur (more specifically, each of these ``takes'' representing a day in 
which one individual is exposed above the Level B harassment criteria, 
even if only for seconds). However, these take numbers do not represent 
the number of individuals expected to be taken, given they are higher 
than the estimated abundance for all species. Accordingly, as described 
in the notice of proposed rulemaking, NMFS references Test Scenario 1 
in the modeling report (``Long-Duration Surveys and Scaling Methods'') 
to inform two important parts of the analyses. Comparing the results of 
modeling simulations that more closely match longer survey durations 
(30 days) to the results of 24-hour take estimates scaled up to 30 days 
(as the instances of take in Table 9 were calculated) provides the 
comparative ratios of number of individuals taken/calculated (within a 
30-day survey) and instances of take, in order to better understand the 
comparative distribution of exposures across individuals of different 
species. First, in NMFS' analyses in this rule, the ratio and its 
inverse are used to inform a better understanding of the nature in 
which individuals are taken across the multiple days of a longer 
duration survey given the different behaviors that are represented in 
the animat modeling, i.e., looking at the ratio of (number of 
individuals taken in 30-day modeling scenario)/(number of instances of 
take when 1-day average multiplied by 30 days), if all else is equal 
within one survey, for the species with a smaller ratio (larger 
inverse), fewer individuals will be taken but each will be exposed 
above the threshold on a higher number of days (see Table 16). Second, 
this ratio may be appropriately be used in combination with the 
calculated instances of take to predict the number of individuals taken 
for surveys of similar duration (noting that for surveys of notably 
longer than 30-day duration, it will still likely result in some degree 
of overestimate of individuals), in order to support evaluation of take 
estimates in requests for Letters of Authorization, given the need to 
meet the ``small numbers of marine mammals'' standard, which is based 
on the number of individuals taken. A summary of this, which was 
included in the notice of proposed rulemaking along with a description 
of the other Test Scenarios and how they inform this analysis, is 
included below.
    Although some survey operations may continue for months, survey 
simulations were conducted for seven days in order to derive mean 24-hr 
exposure averages, with these averages then used to scale according to 
the total number of survey days projected by BOEM. This approach was 
necessary due to the more computationally-intensive modeling required 
to model more realistic durations (i.e., 30 days). As summarized above 
and discussed in detail in the notice of proposed rulemaking, a test 
scenario was used to evaluate methods for scaling results from shorter-
duration simulations to longer duration operations. Results from test 
modeling conducted for a suite of six representative species over 30-
day simulations of a hypothetical 3D WAZ survey were compared to the 
results of a shorter 5-day simulation, i.e., the number of animats 
exposed to levels exceeding threshold for 24-hr time periods multiplied 
by the number of days in the simulations was compared to the number of 
animats exposed to levels exceeding threshold for the entire duration 
of the simulations. The results of the test scenario indicated that 
undesired systematic biases in the modeling procedure, if present, were 
small relative to the survey design and would not affect scaling up the 
results in time (i.e., the shorter 7-day simulations ultimately used in 
the modeling would provide unbiased results). However, the results also 
indicated that scaling up the 24-hr average SPL exposure estimates to 
30 days greatly overestimates the number of notional marine mammals 
(i.e., animats) exposed to levels exceeding threshold when determined 
over the entire simulation (although the estimated instances of 
exposure are reasonably accurate). This occurs because animats were 
commonly exposed to levels exceeding these thresholds, and the 
relatively short reset period of 24 hours means that individual animats 
were, in effect, counted several times during the scale-up (i.e., on 
multiple days) whereas they would only have been counted once when 
evaluating over the entire simulation. When a real-world survey extends 
over longer durations within the same region, it is most likely that 
the same individuals are repeatedly exposed to survey noise. However, 
the modeling assumption that populations of animals were reset for each 
24-hr period is equivalent to an assumption that each survey day is a 
completely independent event, i.e., that new individuals are impacted 
on each subsequent day.
    In order to determine more realistic exposure probabilities for 
individuals across multiple days, modeled results were compared for a 
30-day period versus the aggregation of 24-hr population reset 
intervals (the investigation described above) to determine a species-
typical offset of modeled daily exposures. When conducting 
computationally-intensive modeling over the full assumed 30-day survey 
period (versus aggregating the smaller 24-hr periods for 30 days), 
results showed about 10-45 percent of the total number of takes 
calculated using a 24-hr reset of the population, with differences 
relating to species-typical movement and residency patterns. Given that 
many of the evaluated survey activities occur for 30-day or longer 
periods, particularly some of the larger surveys for which the majority 
of the modeled exposures occur, using such a scaling process is 
appropriate in order to evaluate the likely severity of the predicted 
exposures. This approach is also discussed in more detail in the EWG 
report (Southall et al., 2017), available online at: 
www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico.
    The test scenario modeled six representative GOM species/guilds: 
Bryde's whale, sperm whale, beaked whales, bottlenose dolphin, Kogia 
spp., and short-finned pilot whale. For purposes of this analysis, 
bottlenose dolphin was used as a proxy for other small dolphin species, 
and short-finned pilot whale was used as a proxy for other large 
delphinids. Tables 22-23 in the modeling report provide information 
regarding the number of modeled animals receiving exposure above 
criteria for average 24-hr sliding windows scaled to the full 30-day 
duration and percent change in comparison to the same number evaluated 
when modeling the full 30-day duration. This information was used to 
derive 30-day scalar ratios which, when applied to the total instances 
of take given in Table 9, captures repeated takes of individuals at a 
30-day sampling level. Scalar ratios are as follows: Bryde's whale, 
0.189; sperm whale, 0.423; beaked whales, 0.101; bottlenose dolphin, 
0.287; Kogia spp., 0.321; and short-finned pilot whale, 0.295. 
Application of the re-scaling method reduced the overall magnitude of 
modeled takes for all species by slightly more than double to up to 
ten-fold (Table 10).
    These adjusted take numbers (shown in Table 10) provide a more 
realistic basis upon which to evaluate severity of

[[Page 5405]]

the expected taking. Please see the Negligible Impact Analysis and 
Determinations section, later in this document, for additional detail. 
It is important to recognize that while these scaled numbers better 
reflect the number of individuals likely to be taken within a single 
30-day survey than the number of instances in Table 9, they will still 
overestimate the number of individuals taken across the aggregated GOM 
activities, because they do not correct for (i.e., further reduce take 
to account for) individuals exposed to multiple surveys or fully 
correct for individuals exposed to surveys significantly longer than 30 
days.
    As noted in the beginning of this section and in the Small Numbers 
section, using modeled instances of take (Table 9) and the method 
described here to scale those numbers (based on Test Scenario 1) allows 
one to more accurately predict the number of individuals that will be 
taken as a result of exposure to one survey and, therefore, these 
scaled predictions should be considered in requests for LOAs to assess 
whether a resulting LOA would meet the small numbers standard. However, 
for the purposes of ensuring that the take authorized pursuant to all 
issued LOAs is within the scope of the analysis conducted to support 
the negligible impact finding in this rule, authorized instances of 
take (which are the building blocks of the analysis) also must be 
assessed. Specifically, reflecting Table 9 and what has been analyzed, 
the total take authorized for any given species or stock over the 
course of the five years covered under these regulations should not 
exceed the sum of the five years of take indicated for the five 
scenarios in that table, and in any given year, the take of any species 
should not exceed the highest annual take listed for any of the five 
scenarios.

                                Table 10--Expected Total Take Numbers, Scaled \1\
----------------------------------------------------------------------------------------------------------------
             Species                  Year 1          Year 2          Year 3          Year 4          Year 5
----------------------------------------------------------------------------------------------------------------
Bryde's whale...................               2               2               2               1               1
Sperm whale.....................           6,939           6,009           5,754           4,017           5,240
Kogia spp.......................           3,452           3,098           2,841           2,069           2,771
Beaked whale....................          19,348          16,392          15,991          11,253          14,436
Rough-toothed dolphin...........           8,794           7,756           7,428           5,631           6,664
Bottlenose dolphin..............         173,247         279,357         163,005         287,360         162,857
Clymene dolphin.................          24,633          19,492          21,101          13,584          17,329
Atlantic spotted dolphin........          36,822          52,727          32,178          54,959          31,945
Pantropical spotted dolphin.....         137,327         125,145         112,321          89,348         113,648
Spinner dolphin.................          21,799          20,628          17,535          13,998          18,470
Striped dolphin.................           9,635           8,402           7,989           5,779           7,478
Fraser's dolphin................           1,298           1,103           1,088             782             992
Risso's dolphin.................           6,448           5,536           5,374           3,758           4,907
Melon-headed whale..............          16,465          14,096          13,742           9,611          12,456
Pygmy killer whale..............           2,383           2,054           1,995           1,466           1,852
False killer whale..............           4,769           4,044           4,013           2,851           3,619
Killer whale....................              18              17              15              12              15
Short-finned pilot whale........           4,438           2,898           4,025           2,200           2,643
----------------------------------------------------------------------------------------------------------------
\1\ Scalar ratios were applied to values in Table 9 as described in preceding text to derive scaled take numbers
  shown here.

Mitigation

``Least Practicable Adverse Impact'' Standard

    Under section 101(a)(5)(A) of the MMPA, NMFS must set forth the 
permissible methods of taking pursuant to such activity, and other 
means of effecting the least practicable adverse impact on such species 
or stock and its habitat, paying particular attention to rookeries, 
mating grounds, and areas of similar significance, and on the 
availability of such species or stock for subsistence uses (hereinafter 
referred to as ``LPAI'' or ``least practicable adverse impact''). NMFS 
does not have a regulatory definition for least practicable adverse 
impact. However, NMFS' implementing regulations require applicants for 
incidental take authorizations to include information about the 
availability and feasibility (economic and technological) of equipment, 
methods, and manner of conducting such activity or other means of 
effecting the least practicable adverse impact upon the affected 
species or stocks and their habitat (50 CFR 216.104(a)(11)). We note 
that in some cases, certain mitigation may be necessary in order to 
make a ``negligible impact'' finding for an affected species or stock, 
which is a fundamental requirement of issuing an authorization--in 
these cases, consideration of practicability may be a lower priority 
for decision-making if impacts to marine mammal species or stocks would 
not be negligible in the measure's absence.
    In Conservation Council for Hawaii v. NMFS, 97 F. Supp. 3d 1210, 
1229 (D. Haw. 2015), the district court stated that NMFS ``appear[s] to 
think [it satisfies] the statutory `least practicable adverse impact' 
requirement with a `negligible impact' finding.'' Later, expressing 
similar concerns in a challenge to an incidental take rule for U.S. 
Navy Operation of Surveillance Towed Array Sensor System Low Frequency 
Active (SURTASS LFA) Sonar (77 FR 50290, August 20, 2012), the Ninth 
Circuit in Natural Resources Defense Council (NRDC) v. Pritzker, 828 
F.3d 1125, 1134 (9th Cir. 2016), stated, ``[c]ompliance with the 
`negligible impact' requirement does not mean there [is] compliance 
with the `least practicable adverse impact' standard.'' NMFS is in full 
agreement that the ``negligible impact'' and ``least practicable 
adverse impact'' requirements are distinct, even though both statutory 
standards refer to species and stocks. With that in mind, we provide 
further explanation of NMFS' interpretation of least practicable 
adverse impact and explain what distinguishes it from the negligible 
impact standard. This discussion is consistent with, and expands upon, 
previous rules issued by NMFS, such as the Navy Gulf of Alaska rule (82 
FR 19530; April 27, 2017); the Navy Atlantic Fleet Testing and Training 
rule (83 FR 57076; November 14, 2018); the Navy Hawaii-Southern 
California Training and Testing rule (83 FR 66846; December 27, 2018); 
and the SURTASS

[[Page 5406]]

LFA sonar rule (84 FR 40132; August 13, 2019).
    Before NMFS can issue incidental take regulations under section 
101(a)(5)(A) of the MMPA, it must make a finding that the total taking 
will have a ``negligible impact'' on the affected ``species or stocks'' 
of marine mammals. NMFS' and the U.S. Fish and Wildlife Service's 
implementing regulations for section 101(a)(5) both define ``negligible 
impact'' as an impact resulting from the specified activity that cannot 
be reasonably expected to, and is not reasonably likely to, adversely 
affect the species or stock through effects on annual rates of 
recruitment or survival (50 CFR 216.103 and 50 CFR 18.27(c)). 
Recruitment (i.e., reproduction) and survival rates are used to 
determine population growth rates \12\ and, therefore, are considered 
in evaluating population level impacts.
---------------------------------------------------------------------------

    \12\ A growth rate can be positive, negative, or flat.
---------------------------------------------------------------------------

    As NMFS stated in the preamble to the final rule for the incidental 
take implementing regulations, not every population-level impact 
violates the negligible impact requirement. The negligible impact 
standard does not require a finding that the anticipated take will have 
``no effect'' on population numbers or growth rates: ``The statutory 
standard does not require that the same recovery rate be maintained, 
rather that no significant effect on annual rates of recruitment or 
survival occurs. [T]he key factor is the significance of the level of 
impact on rates of recruitment or survival.'' (54 FR 40338, 40341-42; 
September 29, 1989).
    While some level of impact on population numbers or growth rates of 
a species or stock may occur and may still satisfy the negligible 
impact requirement--even without consideration of mitigation--the least 
practicable adverse impact provision separately requires NMFS to 
prescribe means of effecting the least practicable adverse impact on 
such species or stock and its habitat, paying particular attention to 
rookeries, mating grounds, and areas of similar significance, 50 CFR 
216.102(b), which are typically identified as mitigation measures.\13\
---------------------------------------------------------------------------

    \13\ For purposes of this discussion, we omit reference to the 
language in the standard for least practicable adverse impact that 
says that NMFS also must mitigate for subsistence impacts, because 
subsistence impacts are not at issue in this action.
---------------------------------------------------------------------------

    The negligible impact and least practicable adverse impact 
standards in the MMPA both call for evaluation at the level of the 
``species or stock.'' The MMPA does not define the term ``species.'' 
However, Webster's New Collegiate Dictionary (1981) defines ``species'' 
to include ``a group of intimately related and physically similar 
organisms that actually or potentially interbreed [. . .], that 
ordinarily comprise differentiated populations limited geographically 
[. . .] or ecologically [. . .]'' (emphasis added). See also Merriam-
Webster Dictionary, which defines ``species'' to include ``related 
organisms or populations potentially capable of interbreeding.'' 
www.merriam-webster.com/dictionary/species (emphasis added). The MMPA 
defines ``stock'' as a group of marine mammals of the same species or 
smaller taxa in a common spatial arrangement that interbreed when 
mature (16 U.S.C. 1362(11)). The definition of ``population'' includes 
``a group of interbreeding biotypes that represents the level of 
organization at which speciation begins.'' Webster's New Collegiate 
Dictionary (1981). See also www.merriam-webster.com/dictionary/population, which defines population as ``a group of interbreeding 
organisms that represents the level of organization at which speciation 
begins.'' The definition of ``population'' is strikingly similar to the 
MMPA's definition of ``stock,'' with both involving groups of 
individuals that belong to the same species and are located in a manner 
that allows for interbreeding. In fact, the term ``stock'' in the MMPA 
is interchangeable with the statutory term ``population stock.'' (16 
U.S.C. 1362(11)). Both the negligible impact standard and the least 
practicable adverse impact standard call for evaluation at the level of 
the species or stock, and the terms ``species'' and ``stock'' both 
relate to populations. Therefore, it is appropriate to view both the 
negligible impact standard and the least practicable adverse impact 
standard as having a population-level focus.
    This interpretation is consistent with Congress's statutory 
findings for enacting the MMPA, nearly all of which are most applicable 
at the species or stock (i.e., population) level. See 16 U.S.C. 1361 
(finding that it is species and population stocks that are or may be in 
danger of extinction or depletion; that it is species and population 
stocks that should not diminish beyond being significant functioning 
elements of their ecosystems; and that it is species and population 
stocks that should not be permitted to diminish below their optimum 
sustainable population level). Annual rates of recruitment (i.e., 
reproduction) and survival are the key biological metrics used in the 
evaluation of population-level impacts, and accordingly these same 
metrics are also used in the evaluation of population-level impacts for 
the least practicable adverse impact standard.
    Recognizing this common focus of the least practicable adverse 
impact and negligible impact provisions on the ``species or stock'' 
does not mean that NMFS conflates the two standards; despite some 
common statutory language, we recognize the two provisions are 
different and have different functions.
    First, a negligible impact finding is required before NMFS can 
issue an incidental take authorization. Although it is acceptable to 
use mitigation measures to reach a negligible impact finding (see 50 
CFR 216.104(c)), no amount of mitigation can enable NMFS to issue an 
incidental take authorization for an activity that would not meet the 
negligible impact standard.
    Second, even where NMFS can reach a negligible impact finding--
which we emphasize does allow for the possibility of some 
``negligible'' population-level impact--the agency must still prescribe 
measures that will effect the least practicable amount of adverse 
impact upon the affected species or stock.
    Section 101(a)(5)(A)(i)(II) requires NMFS to issue, in conjunction 
with its authorization, binding--and enforceable--restrictions (in the 
form of regulations) setting forth how the activity must be conducted, 
thus ensuring the activity has the ``least practicable adverse impact'' 
on the affected species or stocks and their habitat. In situations 
where mitigation is specifically needed to reach a negligible impact 
determination, section 101(a)(5)(A)(i)(II) also provides a mechanism 
for ensuring compliance with the ``negligible impact'' requirement.
    Finally, as noted above, the least practicable adverse impact 
standard requires consideration of measures for marine mammal habitat, 
with particular attention to rookeries; mating grounds; and other areas 
of similar significance, and for subsistence impacts. By contrast, the 
negligible impact standard is concerned solely with conclusions about 
the impact of an activity on annual rates of recruitment and 
survival.\14\
---------------------------------------------------------------------------

    \14\ Mitigation may also be appropriate to ensure separate 
compliance with the ``small numbers'' language and negligible impact 
standard in MMPA sections 101(a)(5)(A) and (D).
---------------------------------------------------------------------------

    In NRDC v. Pritzker, the Ninth Circuit stated, ``[t]he statute is 
properly read to mean that even if population levels are not threatened 
significantly, still the agency must adopt mitigation measures aimed at 
protecting marine mammals to

[[Page 5407]]

the greatest extent practicable in light of military readiness needs.'' 
Pritzker, 828 F.3d at 1134 (emphases added). This statement is 
consistent with our understanding stated above that even when the 
effects of an action satisfy the negligible impact standard (i.e., in 
the court's words, ``population levels are not threatened 
significantly''), still the agency must prescribe mitigation under the 
least practicable adverse impact standard. However, as the statute 
indicates, the focus of both standards is ultimately the impact on the 
affected ``species or stock''; the standards are not solely focused on 
or directed at the impact on individual marine mammals.
    NMFS has carefully considered the Ninth Circuit's opinion in NRDC 
v. Pritzker in its entirety. While the court's reference to ``marine 
mammals'' rather than ``marine mammal species or stocks'' in the 
italicized language above might be construed as a holding that the 
least practicable adverse impact standard applies at the individual 
``marine mammal'' level, i.e., that NMFS must require mitigation to 
minimize impacts to each individual marine mammal unless impracticable, 
we believe that such an interpretation reflects an incomplete 
appreciation of the court's decision. In NMFS' view, the decision as a 
whole turned on the court's determination that the agency had not given 
separate and independent meaning to the least practicable adverse 
impact standard apart from the negligible impact standard. NMFS further 
believes that the court's use of the term ``marine mammals'' was not 
addressing the question of whether the standard applies to individual 
animals as opposed to the species or stock as a whole. We recognize 
that while consideration of mitigation can play a role in a negligible 
impact determination, consideration of mitigation measures extends 
beyond that analysis. In evaluating what mitigation measures are 
appropriate, NMFS considers the potential impacts of the specified 
activity, the availability of measures to minimize those potential 
impacts, and the practicability of implementing those measures, as 
described below.

Implementation of Least Practicable Adverse Impact Standard

    In light of the NRDC v. Pritzker decision, we discuss here how NMFS 
determines whether a measure or set of measures meets the ``least 
practicable adverse impact'' standard. Our separate analysis of whether 
the take anticipated to result from the specified activities meets the 
``negligible impact'' standard appears in the Negligible Impact 
Analysis and Determinations section below.
    NMFS' evaluation of potential mitigation measures includes 
consideration of two primary factors:
    (1) The manner in which, and the degree to which, implementation of 
the potential measure(s) is expected to reduce adverse impacts to 
marine mammal species or stocks, their habitat, and their availability 
for subsistence uses (where relevant). This analysis considers such 
things as the nature of the potential adverse impact (such as 
likelihood, scope, and range), the likelihood that the measure will be 
effective if implemented, and the likelihood of successful 
implementation.
    (2) The practicability of the measures for applicant 
implementation. Practicability of implementation may consider such 
things as cost, impact on activities, personnel safety, and 
practicality of implementation.
    While the language of the least practicable adverse impact standard 
calls for minimizing impacts to affected species or stocks and their 
habitat, NMFS recognizes that the reduction of impacts to those species 
or stocks accrues through the application of mitigation measures that 
limit impacts to individual animals. Accordingly, NMFS' analysis 
focuses on measures that are designed to avoid or minimize impacts on 
individual marine mammals that are likely to increase the probability 
or severity of population-level effects.
    While direct evidence of impacts to species or stocks from a 
specified activity is rarely available, and additional study is still 
needed to understand how specific disturbance events affect the fitness 
of individuals of certain species, there have been improvements in 
understanding the process by which disturbance effects are translated 
to the population. With recent scientific advancements (both marine 
mammal energetic research and the development of energetic frameworks), 
the relative likelihood or degree of impacts on species or stocks may 
often be inferred given a detailed understanding of the activity, the 
environment, and the affected species or stocks. This same information 
is used in the development of mitigation measures and helps us 
understand how mitigation measures contribute to lessening effects (or 
the risk thereof) to species or stocks. NMFS also acknowledges that 
there is always the potential that new information, or a new 
recommendation that had not previously been considered, becomes 
available and necessitates re-evaluation of mitigation measures (which 
may be addressed through adaptive management) to see if further 
reductions of population impacts are possible and practicable.
    In the evaluation of specific measures, the details of the 
specified activity will necessarily inform each of the two primary 
factors discussed above (expected reduction of impacts and 
practicability) and are carefully considered to determine the types of 
mitigation that are appropriate under the least practicable adverse 
impact standard. Analysis of how a potential mitigation measure may 
reduce adverse impacts on a marine mammal stock or species and 
practicability of implementation are not issues that can be 
meaningfully evaluated through a yes/no lens. The manner in which, and 
the degree to which, implementation of a measure is expected to reduce 
impacts, as well as its practicability, can vary widely. For example, a 
time-area restriction could be of very high value for reducing the 
potential for, or severity of, population-level impacts (e.g., avoiding 
disturbance of feeding females in an area of established biological 
importance) or it could be of lower value (e.g., decreased disturbance 
in an area of high productivity but of less firmly established 
biological importance). Regarding practicability, a measure might 
involve restrictions in an area or time that impede the operator's 
ability to acquire necessary data (higher impact), or it could mean 
incremental delays that increase operational costs but still allow the 
activity to be conducted (lower impact). A responsible evaluation of 
``least practicable adverse impact'' will consider the factors along 
these realistic scales. Expected effects of the activity and of the 
mitigation as well as status of the stock all weigh into these 
considerations. Accordingly, the greater the likelihood that a measure 
will contribute to reducing the probability or severity of adverse 
impacts to the species or stock or their habitat, the greater the 
weight that measure is given when considered in combination with 
practicability to determine the appropriateness of the mitigation 
measure, and vice versa. Consideration of these factors is discussed in 
greater detail below.
    1. Reduction of adverse impacts to marine mammal species or stocks 
and their habitat.\15\
---------------------------------------------------------------------------

    \15\ NMFS recognizes the least practicable adverse impact 
standard requires consideration of measures that will address 
minimizing impacts on the availability of the species or stocks for 
subsistence uses where relevant. Because subsistence uses are not 
implicated for this action, we do not discuss them. However, a 
similar framework would apply for evaluating those measures, taking 
into account both the MMPA's directive that we make a finding of no 
unmitigable adverse impact on the availability of the species or 
stocks for taking for subsistence, and the relevant implementing 
regulations.

---------------------------------------------------------------------------

[[Page 5408]]

    The emphasis given to a measure's ability to reduce the impacts on 
a species or stock considers the degree, likelihood, and context of the 
anticipated reduction of impacts to individuals (and how many 
individuals) as well as the status of the species or stock.
    The ultimate impact on any individual from a disturbance event 
(which informs the likelihood of adverse species- or stock-level 
effects) is dependent on the circumstances and associated contextual 
factors, such as duration of exposure to stressors. Though any proposed 
mitigation needs to be evaluated in the context of the specific 
activity and the species or stocks affected, measures with the 
following types of effects have greater value in reducing the 
likelihood or severity of adverse species- or stock-level impacts: 
Avoiding or minimizing injury or mortality; limiting interruption of 
known feeding, breeding, mother/young, or resting behaviors; minimizing 
the abandonment of important habitat (temporally and spatially); 
minimizing the number of individuals subjected to these types of 
disruptions; and limiting degradation of habitat. Mitigating these 
types of effects is intended to reduce the likelihood that the activity 
will result in energetic or other types of impacts that are more likely 
to result in reduced reproductive success or survivorship. It is also 
important to consider the degree of impacts that are expected in the 
absence of mitigation in order to assess the added value of any 
potential measures. Finally, because the least practicable adverse 
impact standard gives NMFS discretion to weigh a variety of factors 
when determining appropriate mitigation measures and because the focus 
of the standard is on reducing impacts at the species or stock level, 
the least practicable adverse impact standard does not compel 
mitigation for every kind of take, or every individual taken, if that 
mitigation is unlikely to meaningfully contribute to the reduction of 
adverse impacts on the species or stock and its habitat, even when 
practicable for implementation by the applicant.
    The status of the species or stock is also relevant in evaluating 
the appropriateness of potential mitigation measures in the context of 
least practicable adverse impact. The following are examples of factors 
that may (either alone, or in combination) result in greater emphasis 
on the importance of a mitigation measure in reducing impacts on a 
species or stock: the stock is known to be decreasing or status is 
unknown, but believed to be declining; the known annual mortality (from 
any source) is approaching or exceeding the PBR level; the affected 
species or stock is a small, resident population; or the stock is 
involved in a UME or has other known vulnerabilities, such as 
recovering from an oil spill.
    Habitat mitigation, particularly as it relates to rookeries, mating 
grounds, and areas of similar significance, is also relevant to 
achieving the standard and can include measures such as reducing 
impacts of the activity on known prey utilized in the activity area or 
reducing impacts on physical habitat. As with species- or stock-related 
mitigation, the emphasis given to a measure's ability to reduce impacts 
on a species or stock's habitat considers the degree, likelihood, and 
context of the anticipated reduction of impacts to habitat. Because 
habitat value is informed by marine mammal presence and use, in some 
cases there may be overlap in measures for the species or stock and for 
use of habitat.
    NMFS considers available information indicating the likelihood of 
any measure to accomplish its objective. If evidence shows that a 
measure has not typically been effective nor successful, then either 
that measure should be modified or the potential value of the measure 
to reduce effects should be lowered.
    2. Practicability.
    Factors considered may include those costs, impact on activities, 
personnel safety, and practicality of implementation.
    In carrying out the MMPA's mandate for this action, NMFS applies 
the previously described context-specific balance between the manner in 
which and the degree to which measures are expected to reduce impacts 
to the affected species or stocks and their habitat and practicability 
for operators. The effects of concern (i.e., those with the potential 
to adversely impact species or stocks and their habitat), addressed 
previously in the Potential Effects of the Specified Activity on Marine 
Mammals and Their Habitat section of the notice of proposed rulemaking, 
include auditory injury, severe behavioral reactions, disruptions of 
critical behaviors, and to a lesser degree, masking and impacts on 
acoustic habitat (see discussion of this concept in the ``Anticipated 
Effects on Marine Mammal Habitat'' section in the notice of proposed 
rulemaking). Here, we focus on measures with proven or reasonably 
presumed ability to avoid or reduce the intensity of acute exposures 
that have potential to result in these anticipated effects with an 
understanding of the drawbacks or costs of these requirements, as well 
as time-area restrictions that would avoid or reduce both acute and 
chronic impacts. To the extent of the information available to NMFS, we 
considered practicability concerns, as well as potential undesired 
consequences of the measures, e.g., extended periods using the acoustic 
source due to the need to reshoot lines. NMFS also recognizes that 
instantaneous protocols, such as shutdown requirements, are not capable 
of avoiding all acute effects, and are not suitable for avoiding many 
cumulative or chronic effects and do not provide targeted protection in 
areas of greatest importance for marine mammals. Therefore, in addition 
to a basic suite of seismic mitigation protocols, we also consider 
measures that may or may not be appropriate for other activities (e.g., 
time-area restrictions specific to the surveys discussed herein), but 
that are warranted here given the spatial scope of these specified 
activities, potential for population-level effects and/or high 
magnitude of take for certain species in the absence of such mitigation 
(see Negligible Impact Analysis and Determinations), and the 
information we have regarding habitat for certain species.
    In order to satisfy the MMPA's least practicable adverse impact 
standard, NMFS evaluated a suite of basic mitigation protocols that are 
required regardless of the status of a stock. Additional or enhanced 
protections are required for species whose stocks are in particularly 
poor health and/or are subject to some significant additional stressor 
that lessens that stock's ability to weather the effects of the 
specified activities without worsening its status. NMFS reviewed the 
mitigation measures proposed in the petition, the requirements 
specified in BOEM's PEIS, seismic mitigation protocols required or 
recommended elsewhere (e.g., HESS, 1999; DOC, 2013; IBAMA, 2018; Kyhn 
et al., 2011; JNCC, 2017; DEWHA, 2008; BOEM, 2016; DFO, 2008; GHFS, 
2015; MMOA, 2016; Nowacek et al., 2013; Nowacek and Southall, 2016), 
recommendations received during the public comment period, and the 
available scientific literature. NMFS also considered recommendations 
given in a number of review articles (e.g., Weir and Dolman, 2007; 
Compton et al., 2008; Parsons et al., 2009; Wright and Cosentino, 2015; 
Stone, 2015b). Certain changes from the mitigation measures described 
in the notice of proposed

[[Page 5409]]

rulemaking were made on the basis of additional information and 
following review of public comments. The required suite of mitigation 
measures differs in some cases from the measures proposed in the 
petition and/or those specified by BOEM in the preferred alternative 
identified in their PEIS in order to reflect what NMFS believes to be 
the most appropriate suite of measures to satisfy the requirements of 
the MMPA. Additionally, two geographic mitigation measures discussed in 
the proposed rule are no longer applicable because of the change in the 
scope of the rule.
    For purposes of defining mitigation requirements, we differentiate 
here between requirements for two classes of airgun survey activity: 
Deep penetration and shallow penetration, with surveys using arrays 
greater than 1,500 in\3\ total airgun volume considered deep 
penetration. This delineation is discussed further below, under 
``Changes from the Proposed Regulations.'' Shallow penetration surveys 
also include those using single airguns, boomers, or equivalent 
sources. A third general class of surveys is also considered, referred 
to here as high-resolution geophysical (HRG) surveys and including 
those surveys using the other electromechanical sources described 
previously. HRG surveys are treated differentially on the basis of 
water depth, with 200 m as the divider between shallow and deep HRG. 
Water depth is used as an indicator for surveys (shallow) that should 
be expected to have less potential for impacts to marine mammals, 
because HRG sources used in shallow waters are typically higher-
frequency, lower power, and/or having some significant directionality 
to the beam pattern. Finally, HRG surveys using only sources operating 
at frequencies greater than or equal to 180 kHz are exempt from the 
mitigation requirements described herein, with the exception of 
adherence to vessel strike avoidance protocols. (Note that this has 
been changed from 200 kHz to reflect the best available scientific 
information regarding generalized hearing ranges for affected marine 
mammal hearing groups (NMFS, 2018).) No distinction in standard 
required mitigations is made on the basis of BOEM's planning areas 
(i.e., Western Planning Area (WPA), Central Planning Area (CPA), 
Eastern Planning Area (EPA)).
    First, we summarize notable changes made to the mitigation 
requirements as a result of review of public comments and/or new 
information and then describe mitigation prescribed in the regulations. 
For additional detail regarding mitigation considerations, including 
expected efficacy and/or practicability, or descriptions of mitigation 
considered but not required, please see the notice of proposed 
rulemaking. Where the practicability analysis was described in the 
notice of proposed rulemaking and nothing has changed, we do not repeat 
the description.

Changes to Mitigation From the Proposed Regulations

    Here we summarize substantive changes to mitigation requirements 
from the proposed regulations. All changes were made on the basis of 
review of public comments received and/or review of new information.
Delineation of Airgun Activity Tiers
    In the notice of proposed rulemaking, for purposes of prescribing 
mitigation, NMFS proposed to define ``deep penetration'' surveys as 
those using arrays greater than 400 in\3\ total volume. As stated in 
that notice, NMFS had little information upon which to base such a 
delineation for purposes of defining appropriate mitigation, but 
considered 400 in\3\ as a reasonable cutoff based on descriptions of 
airgun surveys provided in BOEM's petition. We also noted that the 
Associations stated in their comments on the petition that deep 
penetration array volumes used in the GOM range from approximately 
2,000 to 8,400 in\3\. BOEM has subsequently provided information to 
NMFS supporting a cutoff at 1,500 in\3\. In support of section 3(c) of 
E.O. 13795, BOEM analyzed available data for single airguns and airgun 
arrays, including arrays with known characteristics used by the 
National Science Foundation and U.S. Geological Survey and arrays 
evaluated through BOEM NEPA analyses. See e.g., Richardson et al. 
(1995); NSF and USGS (2011). These data suggest that the output of an 
array, in terms of peak source level, increases at a greater rate at 
volumes above approximately 1,500 in\3\. No public comments addressing 
this issue were received. Therefore, NMFS has elected to redefine the 
transition from ``shallow penetration'' to ``deep penetration'' from 
400 to 1,500 in\3\ total volume of the array.
Time-Area Restrictions
    Bryde's Whale Core Habitat Area: The proposed regulatory text 
included a seasonal restriction within an area we termed Bryde's whale 
core habitat, and the preamble for the proposed rule presented several 
alternatives to the seasonal restriction for consideration by the 
public (83 FR 29281; 29302) including a year-round closure for this 
area, which was considered in the analysis for the preliminary 
determination of negligible impact. See 83 FR 29280-29281; 83 FR 29297.
    However, the entirety of this area is now excluded from 
consideration through this rule following BOEM's update to the scope of 
activity (i.e., removal of the GOMESA moratorium area from the 
geographic scope of the rulemaking). Therefore, consideration of a 
time-area restriction for the Bryde's whale core habitat area 
(including the alternatives described above) is moot, and no 
restriction is included in this final rule.
    Dry Tortugas Area: As with the Bryde's whale core habitat area, the 
entirety of the Dry Tortugas area is now excluded from consideration 
through this rule following BOEM's update to the scope of activity 
(i.e., removal of the GOMESA moratorium area from the geographic scope 
of the rulemaking). Therefore, consideration of a time-area restriction 
for the biologically important area for sperm whales and beaked whales 
in the EPA is moot, and no restriction is included in this final rule.
    Coastal Restriction: NMFS proposed a GOM-wide restriction within 
coastal waters inside the 20-m isobath, to be in effect from February 
through May. For this final rule, NMFS contracted the proposed coastal 
time-area restriction spatially and expanded it temporally. The 
restriction has been reduced to cover the same coastal waters (20-m 
isobath) but between 90[deg] W and the eastern extent of the coastal 
waters portion of BOEM's updated specified geographic region, while 
expanding temporally to include the month of January. NMFS received 
informative public comment on both sides of this issue. Some commenters 
provided information indicating practicability concerns regarding the 
proposed restriction, while other commenters supported the importance 
of the restriction and provided information supporting the temporal 
expansion of the restriction to include January. As described in the 
notice of proposed rulemaking, the stock most heavily impacted by the 
DWH oil spill (of those that may be affected by the specified 
activities) was the northern coastal stock of bottlenose dolphin. Since 
publication of the proposed regulations, an additional UME occurred in 
the area largely overlapping the range of this stock. Therefore, while 
NMFS appreciates the practicability concerns raised by commenters, we 
contracted the restriction spatially but did not eliminate the 
restriction, while expanding it temporally to encompass January through 
May. The change is

[[Page 5410]]

described in more detail under Comments and Responses as well as later 
in this section where the details of the specific closure area is 
discussed.
    Restriction Area Buffer Zones: The proposed regulations included 
buffer zones specific to each time-area restriction that corresponded 
with modeled distances to the 160-dB isopleth (i.e., the midpoint of 
the Level B harassment risk function). These distances were 6 km around 
the EPA Bryde's whale core habitat area (Area #2), 13 km around the 
coastal waters restriction (Area #1), and 9 km around the southern EPA 
area (Area #3). Following BOEM's update to the geographic scope of 
activity considered through this rule, Areas 2 and 3 are excluded from 
consideration. Therefore, consideration of buffer zone size around 
these areas is not relevant. Upon review of public comment, in which 
commenters raised concerns about practicability among others, and re-
evaluation of the nature and extent of mitigation Area #1 as it relates 
to the necessity of an additional buffer area, NMFS determined it 
appropriate to not include a buffer for this area. The rationale for 
the change is described in more detail under Comments and Responses.
Shutdown Requirements
    Delphinid Exception: NMFS does not require shutdown or power-down 
for certain delphinid species. In the notice of proposed rulemaking, we 
proposed an exception to the general shutdown requirements for certain 
species of dolphins in relation to airgun surveys, in which the 
acoustic source would be powered down to the smallest single element of 
the array. Power-down conditions would be maintained until the 
animal(s) is observed exiting the exclusion zone or for 15 minutes 
beyond the last observation of the animal, following which full-power 
operations may be resumed without ramp-up. NMFS also provided an 
alternative proposal for consideration by the public, in which no 
shutdown or power-down would be required upon observation of the same 
species of dolphins. While we are careful to note that the reasons for 
and potential effects of dolphin interaction with vessels, including 
working survey vessels, are not fully understood, we also understand 
that dolphins are unlikely to incur any degree of threshold shift due 
to their relative lack of sensitivity to the frequency content in an 
airgun signal (as well as because of potential coping mechanisms). NMFS 
also recognizes that, although dolphins do in fact react to airgun 
noise in ways that may be considered take (Barkaszi et al., 2012; 
Barkaszi and Kelly, 2018), there is a lack of notable adverse dolphin 
reactions to airgun noise despite a large body of observational data. 
Therefore, the removal of the power-down measure for small delphinids, 
in favor of the no-shutdown or power-down alternative, is warranted in 
consideration of the available information regarding the effectiveness 
of such measures in mitigating impacts to small delphinids and the 
practicability of such measures. No shutdown or power-down is required 
for these species.
    Distance of Extended Shutdowns: NMFS limits extended distance 
shutdowns to within 1,500 m. We proposed a number of shutdown 
requirements on the basis of detections of certain species deemed 
particularly sensitive (e.g., beaked whales) or of particular 
circumstances deemed to warrant particular caution (e.g., whales with 
calves). These were all conditioned upon observation or detection of 
these species or circumstances at any distance from the vessel. 
However, NMFS also included as an alternative proposal for public 
consideration a distance limit of 1,000 m for these extended distance 
shutdown requirements. We received several comments challenging the 
value of extended distance shutdown requirements at all and, while NMFS 
disagrees with these comments, we agree that some reasonable distance 
limit should be placed on these requirements in order to better focus 
the observational effort of protected species observers (PSO) and to 
avoid the potential for numerous shutdowns based on uncertain 
detections at great distance. Therefore, as described in greater detail 
later in this section, NMFS determined that a limit on such extended 
distance shutdown zones for relevant species or circumstances was 
appropriate. However, upon consideration of additional information 
(discussed later in this section), NMFS determined it appropriate to 
limit extended distance shutdown zones to 1,500 m, rather than 1,000 m.
    Sperm Whale Shutdowns: The proposed regulatory text included an 
extended distance shutdown upon acoustic detection of sperm whales, and 
this final ITR explicitly expands that requirement to include any 
detection of sperm whales (i.e., including visual detection) at 
extended distance (i.e., within 1,500 m). As discussed in Comments and 
Responses, NMFS received some comments showing that there was a lack of 
clarity regarding the extended distance shutdowns for acoustic 
detections of sperm whales. NMFS also received comments indicating that 
the proposed division (i.e., extended distance shutdown upon acoustic 
detection of sperm whales but not visual detection) did not make sense 
given the available information regarding both the status of the GOM 
sperm whale population and the potential impacts of airgun noise on 
sperm whale foraging activity. While this measure does not avoid such 
impacts--the observed impacts on foraging behavior were at even greater 
distances (Miller et al., 2009)--it may be expected to practicably 
reduce the occurrence and severity of impacts on foraging behavior.
    Shallow Penetration Surveys: NMFS has reduced the standard 
exclusion zone from 200 m to 100 m, and included an extended distance 
shutdown requirement that mirrors the requirements for deep penetration 
surveys but out to a distance of 500 m. The 200-m shutdown distance was 
proposed on the basis of BOEM's HRG survey protocol (Appendix B of 
BOEM, 2017). However, practicability concerns were raised by public 
commenters and 100-m shutdown zones have been effectively applied in 
the past to afford protection from potential Level A harassment and 
more severe behavioral responses from these types of activities. 
Therefore, rather than defer to BOEM's HRG survey protocol, NMFS re-
evaluated the same information informing development of the proposed 
rule, as well as public comment, and determined that the 200-m shutdown 
distance is not warranted and we reduce the distance accordingly. 
Regarding the extended distance shutdown in special circumstances, NMFS 
proposed this mitigation concept in context of deep penetration surveys 
in the notice of proposed rulemaking. Airgun (and equivalent) surveys 
are considered to have similar effects on exposed marine mammals, and 
the sensitive species for which the extended distance shutdown measure 
was proposed are similarly susceptible to disturbance from shallow 
penetration surveys, if exposed. Therefore, NMFS expands the extended 
distance shutdown measure to shallow penetration surveys in addition to 
deep penetration surveys.
    HRG Surveys: NMFS eliminates shutdown requirements for HRG surveys 
(defined here as surveys using electromechanical sources such as multi-
beam echosounders, side-scan sonars, and chirp sub-bottom profilers). 
The proposed regulations required shutdown for marine mammals within 
the proposed exclusion zone for surveys operating in water depths 
greater than 200 m. As discussed above for shallow penetration surveys, 
this proposal was

[[Page 5411]]

modeled after BOEM's HRG survey protocol. However, NMFS re-evaluated 
the available information, as well as public comment, and has 
determined the requirement to not be warranted. These sources are 
typically higher-frequency and lower-power, and have highly directional 
beam patterns. Effects to marine mammals due to use of these sources, 
if any, are expected to be of very low severity and, therefore, the 
benefits of the proposed shutdown requirement would be minimal 
(especially given that animals observed at the surface are necessarily 
not ensonified by the downward-directed beams from the source at the 
time they are observed).
Monitoring
    Nighttime Ramp-Up: NMFS eliminates the requirement for visual 
observation during nighttime ramp-up and pre-clearance. Public 
commenters indicated that this measure is not likely to be effective, 
and that there are safety concerns associated with PSOs working on deck 
at night. NMFS concurs with this assessment, as described in detail in 
Comments and Responses.
    PSOs for Node Retrieval: The proposed requirement for third-party 
PSOs aboard node retrieval vessels is eliminated due to practicability 
concerns expressed through public comment. NMFS concurs with this 
assessment, as described in detail in Comments and Responses.
    Below, mitigation requirements are described in detail.

Mitigation-Related Monitoring

    Monitoring by dedicated, trained marine mammal observers is 
required in all water depths and, for certain surveys, observers must 
be independent. Additionally, for some surveys, NMFS requires that some 
PSOs \16\ have prior experience in the role. Independent observers are 
employed by a third-party observer provider; vessel crew may not serve 
as PSOs when independent observers are required. Dedicated observers 
are those who have no tasks other than to conduct observational effort, 
record observational data, and communicate with and instruct the survey 
operator (i.e., vessel captain and crew) with regard to the presence of 
marine mammals and mitigation requirements. Communication with the 
operator may include brief alerts regarding maritime hazards. Trained 
PSOs have successfully completed an approved PSO training course (see 
Monitoring and Reporting), and experienced PSOs have additionally 
gained a minimum of 90 days at-sea experience working as a PSO during a 
deep penetration seismic survey, with no more than 18 months having 
elapsed since the conclusion of the relevant at-sea experience. 
Training and experience is specific to either visual or acoustic PSO 
duties (where required). An experienced visual PSO must have completed 
approved, relevant training and must have gained the requisite 
experience working as a visual PSO. An experienced acoustic PSO must 
have completed a passive acoustic monitoring (PAM) operator training 
course and must have gained the requisite experience working as an 
acoustic PSO. Hereafter, we also refer to acoustic PSOs as PAM 
operators, whereas when we use ``PSO'' without a qualifier, the term 
refers to either visual PSOs or PAM operators (acoustic PSOs).
---------------------------------------------------------------------------

    \16\ Note that, although we discuss requirements related only to 
observation of marine mammals, we use the generic term ``protected 
species observer.''
---------------------------------------------------------------------------

    NMFS does not formally administer any PSO training program or 
endorse specific providers but will approve PSOs that have successfully 
completed courses that meet the curriculum and trainer requirements 
specified herein (see Monitoring and Reporting). NMFS will provide PSO 
approvals in the context of the need to ensure that PSOs have the 
necessary training to carry out their duties competently while also 
approving applicant staffing plans quickly. In order for PSOs to be 
approved, NMFS must review and approve PSO resumes indicating 
successful completion of an acceptable training course. Although PSOs 
must be approved by NMFS, third-party observer providers and/or 
companies seeking PSO staffing should expect that observers having 
satisfactorily completed acceptable training and with the requisite 
experience (if required) will be quickly approved and, if NMFS does not 
respond within one week of having received the required information, 
such PSOs shall be considered to have received de facto approval. A PSO 
may be trained and/or experienced as both a visual PSO and PAM operator 
and may perform either duty, pursuant to scheduling requirements. Where 
multiple PSOs are required and/or PAM operators are required, PSO watch 
schedules shall be devised in consideration of the following 
restrictions: (1) A maximum of two consecutive hours on watch followed 
by a break of at least one hour between watches for visual PSOs 
(periods typical of observation for research purposes and as used for 
airgun surveys in certain circumstances (Broker et al., 2015)); (2) a 
maximum of four consecutive hours on watch followed by a break of at 
least two consecutive hours between watches for PAM operators; and (3) 
a maximum of 12 hours observation per 24-hour period. NMFS may grant an 
exception for the requirement that visual PSOs be limited to a maximum 
of two consecutive hours on watch followed by a break of at least one 
hour between watches if requested on the basis of practicability 
concerns by LOA applicants. If an exception is granted, visual PSOs 
would instead be limited to a maximum of four consecutive hours on 
watch followed by a break of at least two hours between watches. 
Further information regarding PSO requirements may be found in the 
Monitoring and Reporting section, later in this document.
    Deep Penetration Surveys--During deep penetration survey \17\ 
operations (e.g., any day on which use of the acoustic source is 
planned to occur; whenever the acoustic source is in the water, whether 
activated or not), a minimum of two independent PSOs must be on duty 
and conducting visual observations at all times during daylight hours 
(i.e., from 30 minutes prior to sunrise through 30 minutes following 
sunset). PSOs should use NOAA's solar calculator (www.esrl.noaa.gov/gmd/grad/solcalc/) to determine sunrise and sunset times at their 
specific location. NMFS recognizes that certain daytime conditions 
(e.g., fog, heavy rain) may reduce or eliminate effectiveness of visual 
observations. However, on-duty PSOs shall remain alert for marine 
mammal observational cues and/or a change in conditions.
---------------------------------------------------------------------------

    \17\ Deep penetration surveys are defined as those surveys using 
airgun arrays with total volume greater than 1,500 in\3\.
---------------------------------------------------------------------------

    All source vessels must carry a minimum of one experienced visual 
PSO, who shall be designated as the lead PSO, coordinate duty schedules 
and roles,\18\ and serve as the primary point of contact for the 
operator. However, while it is desirable for all PSOs to be qualified 
through experience, NMFS is also mindful of the need to expand the 
workforce by allowing opportunity for newly trained PSOs to gain 
experience. Therefore, the lead PSO shall devise the duty schedule such 
that experienced PSOs are on duty with trained PSOs (i.e., those PSOs 
with appropriate training but who have not yet gained relevant 
experience) to the maximum extent practicable in order to provide 
necessary mentorship.
---------------------------------------------------------------------------

    \18\ The coordination of PSO duty schedules and roles may 
alternatively be performed by a third-party, shore-based Monitoring 
Coordinator.

---------------------------------------------------------------------------

[[Page 5412]]

    With regard to specific observational protocols, NMFS largely 
follows those described in Appendix B of BOEM's PEIS (BOEM, 2017). The 
lead PSO shall determine the most appropriate observation posts that 
will not interfere with navigation or operation of the vessel while 
affording an optimal, elevated view of the sea surface. These should be 
the highest elevation available on each vessel, with the maximum 
viewable range from the bow to 90 degrees to port or starboard of the 
vessel. PSOs shall coordinate to ensure 360[deg] visual coverage around 
the vessel, and shall conduct visual observations using binoculars and 
the naked eye while free from distractions and in a consistent, 
systematic, and diligent manner. All source vessels must be equipped 
with pedestal-mounted ``bigeye'' binoculars that will be available for 
PSO use. Within these broad outlines, the lead PSO and PSO team will 
have discretion to determine the most appropriate vessel- and survey-
specific system for implementing effective marine mammal observational 
effort. Any observations of marine mammals by crew members aboard any 
vessel associated with the survey, including receiver or chase vessels, 
should be relayed to the source vessel(s) and to the PSO team.
    All source vessels must use a towed PAM system for potential 
detection of marine mammals at all times when operating the sound 
source in waters deeper than 100 m. The term ``towed PAM system'' 
refers to any combination of hardware and software that uses a towed 
array for operations. The array can be physically separate from other 
in-water hardware, or embedded into other equipment, such as seismic 
streamers. The system must be monitored at all times during use of the 
acoustic source, and acoustic monitoring must begin at least 30 minutes 
prior to ramp-up. PAM operators must be independent, and all source 
vessels shall carry a minimum of two experienced PAM operators. PAM 
operators shall communicate all detections to visual PSOs, when visual 
PSOs are on duty, including any determination by the PSO regarding 
species identification, distance and bearing, and the degree of 
confidence in the determination. Further detail regarding PAM system 
requirements may be found in the Monitoring and Reporting section, 
later in this document. The effectiveness of PAM depends to a certain 
extent on the equipment and methods used and competency of the PAM 
operator, but no formal standards are currently in place regarding PAM 
system hardware/software requirements, or regarding PAM operator 
training.
    Visual monitoring must begin at least 30 minutes prior to ramp-up 
(described below) and must continue until one hour after use of the 
acoustic source ceases or until 30 minutes past sunset. If any marine 
mammal is observed at any distance from the vessel, a PSO would record 
the observation and monitor the animal's position (including latitude/
longitude of the vessel and relative bearing and estimated distance to 
the animal) until the animal dives or moves out of visual range of the 
observer. A PSO would continue to observe the area to watch for the 
animal to resurface or for additional animals that may surface in the 
area. Visual PSOs shall communicate all observations to PAM operators, 
including any determination by the PSO regarding species 
identification, distance, and bearing and the degree of confidence in 
the determination.
    As noted previously, all source vessels must carry a minimum of one 
experienced visual PSO and two experienced PAM operators. The observer 
designated as lead PSO (including the full team of visual PSOs and PAM 
operators) must have experience as a visual PSO. The applicant may 
determine how many additional PSOs are required to adequately fulfill 
the requirements specified here. To summarize, these requirements are: 
(1) 24-hour acoustic monitoring during use of the acoustic source in 
waters deeper than 100 m; (2) visual monitoring during use of the 
acoustic source by two PSOs during all daylight hours; (3) maximum of 
two consecutive hours on watch followed by a minimum of one hour off 
watch for visual PSOs and a maximum of four consecutive hours on watch 
followed by a minimum of two consecutive hours off watch for PAM 
operators; and (4) maximum of 12 hours of observational effort per 24-
hour period for any PSO, regardless of duties.
    Shallow Penetration Surveys--During shallow penetration 
surveys,\19\ operators must follow the same requirements described 
above for deep penetration surveys, with one notable exception. The use 
of PAM is not required.
---------------------------------------------------------------------------

    \19\ Shallow penetration surveys are defined as those using 
airgun arrays with total volume less than or equal to 1,500 in\3\, 
single airguns, boomers, or equivalent sources.
---------------------------------------------------------------------------

    HRG Surveys--HRG survey protocols differ from the previously 
described protocols for deep and shallow penetration surveys, and we 
differentiate between deep-water (greater than 100 m) and shallow-water 
HRG surveys. Water depth in the GOM provides a reliable indicator of 
the marine mammal fauna that may be encountered and, therefore, the 
complexity of likely observations and concern related to potential 
effects on deep-diving and/or sensitive species.
    Deep-water HRG surveys are required to employ a minimum of one 
independent visual PSO during all daylight operations, in the same 
manner as was described for deep and shallow penetration surveys. 
Shallow-water HRG surveys are required to employ a minimum of one 
visual PSO, which may be a crew member. PSOs employed during shallow-
water HRG surveys are only required during a pre-clearance period. PAM 
is not required for any HRG survey.
    PAM Malfunction--Emulating sensible protocols described by the New 
Zealand Department of Conservation for airgun surveys conducted in New 
Zealand waters (DOC, 2013), survey activity may continue for brief 
periods of time when the PAM system malfunctions or is damaged. 
Activity may continue for 30 minutes without PAM while the PAM operator 
diagnoses the issue. If the diagnosis indicates that the PAM system 
must be repaired to solve the problem, operations may continue for an 
additional two hours without acoustic monitoring under the following 
conditions:
     Daylight hours and sea state is less than or equal to 
Beaufort sea state (BSS) 4;
     No marine mammals (excluding delphinids; see below) 
detected solely by PAM in the exclusion zone (see below) in the 
previous two hours;
     NMFS is notified via email as soon as practicable with the 
time and location in which operations began without an active PAM 
system; and
     Operations with an active acoustic source, but without an 
operating PAM system, do not exceed a cumulative total of four hours in 
any 24-hour period.

Exclusion Zone and Buffer Zone

    An exclusion zone is a defined area within which occurrence of a 
marine mammal triggers mitigation action intended to reduce the 
potential for certain outcomes such as auditory injury or more severe 
disruption of behavioral patterns. For deep penetration surveys, the 
PSOs shall establish and monitor a 500-m exclusion zone and additional 
500-m buffer zone (total 1,000 m) during the pre-clearance period (see 
below) and a 500-m exclusion zone during the ramp-up and operational 
periods (see below for description of extended 1,500-m zone in

[[Page 5413]]

special circumstances). PSOs should generally focus their observational 
effort within a 1.5-km zone, to the extent possible, with animals 
observed at greater distances recorded and mitigation action taken as 
necessary (see below). For shallow penetration surveys, the PSOs shall 
establish and monitor a 100-m exclusion zone with additional 100-m 
buffer (total 200-m zone) during the pre-clearance period and a 100-m 
exclusion zone during the ramp-up (for small arrays only, versus single 
airguns) and operational periods (see below for description of extended 
500-m zone in special circumstances). PSOs should generally focus their 
observational effort within a 500-m zone, to the extent possible, with 
animals observed at greater distances recorded and mitigation action 
taken as necessary (see below). These zones shall be based upon radial 
distance from any element of the airgun array (rather than being based 
on the center of the array or around the vessel itself). During use of 
the acoustic source, occurrence of marine mammals within the buffer 
zone (but outside the exclusion zone) should be communicated to the 
operator to prepare for the potential shutdown of the acoustic source. 
Use of the buffer zone in relation to ramp-up is discussed below under 
``Ramp-up.'' Further detail regarding the exclusion zone and shutdown 
requirements is given under ``Exclusion Zone and Shutdown 
Requirements.''

Ramp-Up

    Ramp-up of an acoustic source is intended to provide a gradual 
increase in sound levels, enabling animals to move away from the source 
if the signal is sufficiently aversive prior to its reaching full 
intensity. We infer on the basis of behavioral avoidance studies and 
observations that this measure results in some reduced potential for 
auditory injury and/or more severe behavioral reactions. Although this 
measure is not proven and some arguments have been made that use of 
ramp-up may not have the desired effect of aversion (which is itself a 
potentially negative impact but assumed to be better than the 
alternative), ramp-up remains a relatively low-cost, common-sense 
component of standard mitigation for surveys using airgun arrays. Ramp-
up is most likely to be effective for more sensitive species (e.g., 
beaked whales) with known behavioral responses at greater distances 
from an acoustic source (e.g., Tyack et al., 2011; DeRuiter et al., 
2013; Miller et al., 2015). Ramp-up is required for all surveys using 
airgun arrays. While non-airgun acoustic sources are not typically 
amenable to ``ramping up'' the acoustic output in the way that multi-
element airgun arrays are, power to these sources should be increased 
as feasible in order to effect a ramp-up.
    The ramp-up procedure involves a step-wise increase in the number 
of airguns firing and total array volume until all operational airguns 
are activated and the full volume is achieved. Ramp-up is required at 
all times as part of the activation of the acoustic source (including 
source tests; see ``Miscellaneous Protocols'' for more detail) and may 
occur at times of poor visibility, assuming appropriate acoustic 
monitoring with no detections in the 30 minutes prior to beginning 
ramp-up. Acoustic source activation may only occur at night where 
operational planning cannot reasonably avoid such circumstances. For 
example, a nighttime initial ramp-up following port departure is 
reasonably avoidable and may not occur. Ramp-up must occur at night 
following acoustic source deactivation due to line turn or mechanical 
difficulty. The operator must notify a designated PSO of the planned 
start of ramp-up as agreed-upon with the lead PSO; the notification 
time should be at least 60 minutes prior to the planned ramp-up. A 
designated PSO must be notified again immediately prior to initiating 
ramp-up procedures and the operator must receive confirmation from the 
PSO to proceed.
    Ramp-up procedures follow the recommendations of IAGC (2015). Ramp-
up begins by activating a single airgun (i.e., array element) of the 
smallest volume in the array. Ramp-up continues in stages by doubling 
the number of active elements at the commencement of each stage, with 
each stage of approximately the same duration. Total duration should 
not be less than approximately 20 minutes but maximum duration is not 
prescribed and will vary depending on the total number of stages. Von 
Benda-Beckmann et al. (2013), in a study of the effectiveness of ramp-
up for sonar, found that extending the duration of ramp-up did not have 
a corresponding effect on mitigation benefit. There will generally be 
one stage in which doubling the number of elements is not possible 
because the total number is not even. This should be the last stage of 
the ramp-up sequence. The operator must provide information to the PSO 
documenting that appropriate procedures were followed. Ramp-ups should 
be scheduled so as to minimize the time spent with the source activated 
prior to reaching the designated run-in. This approach is intended to 
ensure a perceptible increase in sound output per increment while 
employing increments that produce similar degrees of increase at each 
step.
    For deep penetration surveys, PSOs must monitor a 1,000-m zone (or 
to the distance visible if less than 1,000 m) for a minimum of 30 
minutes prior to ramp-up (i.e., pre-clearance). For shallow penetration 
surveys, PSOs must monitor a 200-m zone (or to the distance visible if 
less than 200 m) for a minimum of 30 minutes prior to ramp-up or start-
up (for single airgun or non-airgun surveys). (Note that extended 
distance shutdowns, discussed below, may be required if certain species 
or circumstances are detected within greater distances: 1.5 km for deep 
penetration surveys and 500 m for shallow penetration surveys). The 
pre-clearance period may occur during any vessel activity (i.e., 
transit, line turn). Ramp-up must be planned to occur during periods of 
good visibility when possible; operators may not target the period just 
after visual PSOs have gone off duty. Following deactivation of the 
source for reasons other than mitigation, the operator must communicate 
the near-term operational plan to the lead PSO with justification for 
any planned nighttime ramp-up. Any suspected patterns of abuse by the 
operator must be reported by the lead PSO to be investigated by NMFS. 
Ramp-up may not be initiated if any marine mammal is within the 
designated zone. If a marine mammal is observed within the zone during 
the pre-clearance period, ramp-up may not begin until the animal(s) has 
been observed exiting the zone or until an additional time period has 
elapsed with no further sightings (i.e., 15 minutes for small 
delphinids and 30 minutes for all other species). PSOs will monitor the 
exclusion zone during ramp-up, and ramp-up must cease and the source 
shut down upon observation of marine mammals within or approaching the 
zone.

Exclusion Zone and Shutdown Requirements

    Deep Penetration Surveys--The PSOs must establish a minimum 
exclusion zone with a 500-m radius as a perimeter around the outer 
extent of the airgun array (rather than being delineated around the 
center of the array or the vessel itself). If a marine mammal (other 
than the small delphinid species discussed below) appears within or 
enters this zone, the acoustic source must be shut down (i.e., power to 
the acoustic source must be immediately turned off). If a marine mammal 
is detected acoustically, the acoustic source must be shut down, unless 
the PAM operator is confident that the

[[Page 5414]]

animal detected is outside the exclusion zone or that the detected 
species is not subject to the shutdown requirement (see below).
    The 500-m radial distance of the standard exclusion zone is 
expected to contain sound levels exceeding peak pressure injury 
criteria for all hearing groups other than, potentially, high-frequency 
cetaceans, while also providing a consistent, reasonably observable 
zone within which PSOs would typically be able to conduct effective 
observational effort. Although significantly greater distances may be 
observed from an elevated platform under good conditions, NMFS believes 
that 500 m is likely regularly attainable for PSOs using the naked eye 
during typical conditions. In addition, an exclusion zone is expected 
to be helpful in avoiding more severe behavioral responses. Behavioral 
response to an acoustic stimulus is determined not only by received 
level but by context (e.g., activity state) including, importantly, 
proximity to the source (e.g., Southall et al., 2007; Ellison et al., 
2012; DeRuiter et al., 2013). In prescribing an exclusion zone, NMFS 
seeks not only to avoid most potential auditory injury but also to 
reduce the likely severity of the behavioral response at a given 
received level of sound.
    As discussed in the notice of proposed rulemaking, use of 
monitoring and shutdown measures within defined exclusion zone 
distances is inherently an essentially instantaneous proposition--a 
rule or set of rules that requires mitigation action upon detection of 
an animal. This indicates that defining an exclusion zone on the basis 
of thresholds related to the accumulation of energy (i.e., cumulative 
SEL), which require that an animal accumulate some level of sound 
energy exposure over some period of time (e.g., 24 hours), has 
questionable relevance as a standard protocol for mobile sources, given 
the relative motion of the source and the animals. A PSO aboard a 
mobile source will typically have no ability to monitor an animal's 
position relative to the acoustic source over relevant time periods for 
purposes of understanding whether auditory injury is likely to occur on 
the basis of cumulative sound exposure and, therefore, whether action 
should be taken to avoid such potential.
    Cumulative SEL (cSEL) thresholds are more relevant for purposes of 
modeling the potential for auditory injury than they are for dictating 
real-time mitigation, though they can be informative (especially in a 
relative sense). NMFS recognizes the importance of the accumulation of 
sound energy to an understanding of the potential for auditory injury 
and that it is likely that, at least for low-frequency cetaceans, some 
potential auditory injury may be impossible to fully avoid, depending 
on survey location in relation to the areas where these species occur, 
and should be considered for authorization.
    Considering both the dual-metric thresholds described previously 
(and shown in Table 7) and hearing group-specific marine mammal 
auditory weighting functions in the context of airgun sources, auditory 
injury zones indicated by the peak pressure metric are expected to be 
predominant for both mid- and high-frequency cetaceans, while zones 
indicated by cSEL criteria are expected to be predominant for low-
frequency cetaceans. Assuming a source level of 255.2 dB 0-pk SPL for 
the notional 8,000 in\3\ array and spherical spreading propagation, 
distances for exceedance of group-specific peak injury thresholds are 
as follows: 65 m (LF), 18 m (MF), and 457 m (HF) (for high-frequency 
cetaceans, although the notional source parameters indicate a zone less 
than 500 m, we recognize that actual isopleth distances will vary based 
on specific array characteristics and site-specific propagation 
characteristics, and that it is therefore possible that a real-world 
distance to the injury threshold could exceed 500 m). Assuming a source 
level of 227.7 dB 0-pk SPL for the notional 90 in\3\ single airgun and 
spherical spreading propagation, these distances would be 3 m (LF) and 
19 m (HF) (the source level is lower than the threshold criterion value 
for mid-frequency cetaceans). These specific modeled source level 
values were discussed in the notice of proposed rulemaking, and 
additional information may be found in the modeling report.
    Consideration of auditory injury zones based on cSEL criteria is 
dependent on the animal's generalized hearing range and how it overlaps 
with the frequencies produced by the sound source of interest in 
relation to marine mammal auditory weighting functions (NMFS, 2018). As 
noted above, zones based on the cSEL threshold are expected to be 
predominant for low-frequency cetaceans because their most susceptible 
hearing range overlaps the low frequencies produced by airguns, while 
the modeling indicates that zones based on peak pressure criteria 
dominate for mid- and high-frequency cetaceans. As described in detail 
in the notice of proposed rulemaking, NMFS obtained unweighted spectrum 
data (modeled in 1 Hz bands) for a reasonably equivalent acoustic 
source (i.e., a 36-airgun array with total volume of 6,600 in\3\) in 
order to evaluate notional zone sizes and to incorporate NMFS' 
Technical Guidance weighting functions over an airgun array's full 
acoustic band. Using NMFS' associated User Spreadsheet with hearing 
group-specific weighted source levels, and inputs assuming a 231.8 dB 
SEL source level for the notional 8,000 in\3\ array, spherical 
spreading propagation, a source velocity of 4.5 kn, pulse duration of 
100 ms, and a 25-m shot interval (shot intervals may vary, with longer 
shot intervals resulting in smaller calculated zones), distances for 
group-specific threshold criteria are as follows: 574 m (LF), 0 m (MF), 
and 1 m (HF). NMFS also assessed the potential for injury based on the 
accumulation of energy resulting from use of the single airgun and, 
assuming a source level of 207.8 dB SEL, there would be no realistic 
zone within which injury would occur.
    Therefore, the 500-m exclusion zone contains the entirety of any 
potential injury zone for mid-frequency cetaceans (realistically, there 
is no such zone, as discussed above in Estimated Take and in Comments 
and Responses), while the zones within which injury could occur may be 
larger for high-frequency cetaceans (on the basis of peak pressure and 
depending on the specific array) and for low-frequency cetaceans (on 
the basis of cumulative sound exposure).
    In summary, NMFS' goal in prescribing a standard exclusion zone 
distance is to (1) encompass zones for most species within which 
auditory injury could occur on the basis of instantaneous exposure; (2) 
provide protection from the potential for more severe behavioral 
reactions (e.g., panic, antipredator response) for marine mammals at 
relatively close range to the acoustic source; (3) enable more 
effective implementation of required mitigation by providing 
consistency and ease of implementation for PSOs, who need to monitor 
and implement the exclusion zone; and (4) define a distance within 
which detection probabilities are reasonably high for most species 
under typical conditions. NMFS' use of 500 m as the zone is not based 
directly on any quantitative understanding of the range at which 
auditory injury would be entirely precluded or any range specifically 
related to disruption of behavioral patterns. Rather, we believe it is 
a reasonable combination of factors. This zone has been proven as a 
feasible measure through past implementation by operators in the GOM. 
In summary, a practicable criterion such as this has the advantage of 
familiarity and simplicity while still providing in most

[[Page 5415]]

cases a zone larger than relevant auditory injury zones, given 
realistic movement of source and receiver. Increased shutdowns, without 
a firm idea of the outcome the measure seeks to avoid, simply displace 
survey activity in time and increase the total duration of acoustic 
influence as well as total sound energy in the water (due to additional 
ramp-up and overlap where data acquisition was interrupted). The 
shutdown requirement described here would be required for most marine 
mammals, with certain differences. Small delphinids are excepted from 
the shutdown requirement, as described in the following section. 
Certain species are subject to an extended distance shutdown zone, as 
described in the subsequent section entitled ``Other Shutdown 
Requirements.''
    Dolphin Exception--The shutdown requirement described above is in 
place for all marine mammals, with the exception of small delphinids. 
As defined here, the small delphinid group is intended to encompass 
those members of the Family Delphinidae most likely to voluntarily 
approach the source vessel for purposes of interacting with the vessel 
and/or airgun array (e.g., bow-riding). (Here we refer to ``large 
delphinids'' and ``small delphinids'' as shorthand for generally deep-
diving versus surface-dwelling/bow-riding groups, respectively, as the 
important distinction is their dive behavior rather than their size.) 
This exception to the shutdown requirement applies solely to specific 
genera of dolphins--Steno, Tursiops, Stenella, and Lagenodelphis (see 
Table 4)--and applies under all circumstances, regardless of what the 
perception of the animal(s) behavior or intent may be. The proposed 
regulations included a requirement to conduct a power-down upon 
detection of these species within the exclusion zone. However, in the 
preamble to the proposed regulations, NMFS also included an alternative 
proposal for public review and comment in which no shutdown or power-
down would be required. We requested comment on both proposals and 
other variations of those proposals, including NMFS' interpretation of 
the data and any other data that support the necessary findings 
regarding dolphins for no shutdown and no power-down or no shutdown but 
a power-down. Upon review of the public comments received, as well as 
the scientific information summarized below, NMFS has determined that 
the alternative proposal of no shutdown or power-down is appropriate, 
and satisfies the least practicable adverse impact requirement.
    Variations of this measure that include exceptions based on animal 
behavior--e.g., ``bow-riding'' dolphins, or only ``traveling'' 
dolphins, meaning that the intersection of the animal and exclusion 
zone may be due to the animal rather than the vessel--have been 
proposed by both NMFS and BOEM and have been criticized, in part due to 
the subjective on-the-spot decision-making this scheme would require of 
PSOs. If the mitigation requirements are not sufficiently clear and 
objective, the outcome may be differential implementation across 
surveys as informed by individual PSOs' experience, background, and/or 
training. The exception described here is based on several factors: The 
lack of evidence of or presumed potential for the types of effects to 
these species of small delphinid that our shutdown requirement for 
other species seeks to avoid, the uncertainty and subjectivity 
introduced by such a decision framework, and the practicability concern 
presented by the operational impacts. Despite a large volume of 
observational effort during airgun surveys, including in locations 
where dolphin shutdowns have not previously been required (i.e., the 
U.S. GOM and United Kingdom (UK) waters), we are not aware of accounts 
of notable adverse dolphin reactions to airgun noise (Stone, 2015a; 
Barkaszi et al., 2012; Barkaszi and Kelly, 2018) other than one 
isolated incident (Gray and Van Waerebeek, 2011). Dolphins have a 
relatively high threshold for the onset of auditory injury (i.e., PTS) 
and more severe adverse behavioral responses seem less likely given the 
evidence of purposeful approach and/or maintenance of proximity to 
vessels with operating airguns.
    The best available scientific evidence indicates that auditory 
injury as a result of airgun sources is extremely unlikely for mid-
frequency cetaceans, primarily due to a relative lack of sensitivity 
and susceptibility to noise-induced hearing loss at the frequency range 
output by airguns (i.e., most sound below 500 Hz) as shown by the mid-
frequency cetacean auditory weighting function (NMFS, 2018). Criteria 
for TTS in mid-frequency cetaceans for impulsive sounds were derived by 
experimental measurement of TTS in beluga whales exposed to pulses from 
a seismic watergun. Dolphins exposed to the same stimuli in this study 
did not display TTS (Finneran et al., 2002). Moreover, when the 
experimental watergun signal was weighted appropriately for mid-
frequency cetaceans, less energy was filtered than would be the case 
for an airgun signal. More recently, Finneran et al. (2015) exposed 
bottlenose dolphins to repeated pulses from an airgun and measured no 
TTS.
    NMFS cautions that, while dolphins are observed voluntarily 
approaching source vessels (e.g., bow-riding or interacting with towed 
gear), the reasons for the behavior are unknown. In context of an 
active airgun array, the behavior cannot be assumed to be harmless. 
Although bow-riding comprises approximately 30 percent of behavioral 
observations in the GOM, there is a much lower incidence of the 
behavior when the acoustic source is active (Barkaszi et al., 2012), 
and this finding was replicated by Stone (2015a) for surveys occurring 
in UK waters. Some studies have found evidence of aversive behavior by 
dolphins during firing of airguns. Barkaszi et al. (2012) found that 
the median closest distance of approach to the acoustic source was at 
significantly greater distances during times of full-power source 
operation when compared to silence, while Stone (2015a) and Stone and 
Tasker (2006) reported that behavioral responses, including avoidance 
and changes in swimming or surfacing behavior, were evident for 
dolphins during firing of large arrays. Goold and Fish (1998) described 
a ``general pattern of localized disturbance'' for dolphins in the 
vicinity of an airgun survey. However, while these general findings--
typically, dolphins will display increased distance from the acoustic 
source, decreased prevalence of ``bow-riding'' activities, and 
increases in surface-active behaviors--are indicative of adverse or 
aversive responses that may rise to the level of ``take'' (as defined 
by the MMPA), they are not indicative of any response of a severity 
such that the need to avoid it outweighs the impact on practicability 
for the industry and operators.
    Additionally, increased shutdowns resulting from such a measure 
would require source vessels to revisit the missed track line to 
reacquire data, resulting in an overall increase in the total sound 
energy input to the marine environment and an increase in the total 
duration over which the survey is active in a given area. Therefore, 
the removal of such measures for small delphinids is warranted in 
consideration of the available information regarding the effectiveness 
of such measures in mitigating impacts to small delphinids and the 
practicability of such measures.
    Although other mid-frequency hearing specialists (e.g., large 
delphinids) are considered no more likely to incur auditory injury than 
are small delphinids, they are more

[[Page 5416]]

typically deep divers, meaning that there is some increased potential 
for more severe effects from a behavioral reaction, as discussed in 
greater detail in Comments and Responses. Therefore, NMFS anticipates 
benefit from a shutdown requirement for large delphinids, in that it is 
likely to preclude more severe behavioral reactions for any such 
animals in close proximity to the source vessel as well as any 
potential for physiological effects.
    At the same time, large delphinids are much less likely to approach 
vessels. Therefore, a shutdown requirement for large delphinids would 
not have similar impacts as a small delphinid shutdown in terms of 
either practicability for the applicant or corollary increase in sound 
energy output and time on the water.
    Other Surveys--Shutdown protocols for shallow penetration surveys 
are similar to those described for deep penetration surveys, except 
that the exclusion zone is defined as a 100-m radial distance around 
the perimeter of the acoustic source. The dolphin exception described 
above for deep penetration surveys would apply. As described 
previously, no shutdowns would be required for HRG surveys.
    Extended Shutdown Requirements for Special Circumstances--Shutdown 
of the acoustic source is also required in the event of certain other 
detections beyond the standard exclusion zones. In the proposed 
regulatory text, NMFS conditioned these shutdowns upon detection of the 
relevant species or circumstances at any distance. However, in the 
preamble to the proposed regulations, we also included an alternative 
proposal for public review and comment in which shutdown of the 
acoustic source would occur in the circumstances listed below, but only 
within 1 km of the source (for deep penetration surveys). We requested 
comment on both proposals and other variations of those proposals, 
including NMFS' interpretation of the data and any other data that 
support the necessary findings regarding initiating shutdown for 
certain circumstances at any distance or within 1 km. Following review 
of public comments and the relevant scientific information, NMFS 
determined that it is appropriate to limit such shutdown requirements. 
However, as discussed in the next paragraph, we also determined that 
the relevant scientific information better supports 1.5 km as a 
reasonable detection radius (versus 1 km). Placement of a distance 
limit on these requirements maintains the intent of the measures as 
originally proposed, i.e., to provide for additional real-time 
protection by limiting the intensity and duration of acoustic exposures 
for certain species or in certain circumstances, while reducing the 
area over which PSOs must maintain observational effort. As for normal 
shutdowns within the standard exclusion zone, shutdowns at extended 
distance should be made on the basis of confirmed detections (visual or 
acoustic) within the zone.
    For deep penetration surveys, NMFS determined an appropriate 
distance on the basis of available information regarding detection 
functions for relevant species, but notes that, while based on 
quantitative data, the distance is an approximate limit that is merely 
intended to encompass the region within which we would expect a 
relatively high degree of success in sighting certain species while 
also improving PSO efficacy by removing the potential that a PSO might 
interpret these requirements as demanding a focus on areas further from 
the vessel. The appropriate distance limit may vary for different 
regions, depending on the species to which it may apply. For each 
modeled taxon, Roberts et al. (2016) fitted detection functions that 
modeled the detectability of the taxon according to distance from the 
trackline and other covariates (i.e., the probability of detecting an 
animal given its distance from the transect). These functions were 
based on nearly 1.1 million linear km of line-transect survey effort 
conducted from 1992-2014, with surveys arranged in aerial and shipboard 
hierarchies and further grouped according to similarity of observation 
protocol and platform. Where a taxon was sighted infrequently, a 
detection function was fit to pooled sightings of suitable proxy 
species. For example, for the Bryde's whale and shipboard binocular 
surveys (i.e., the relevant combination of platform and protocol), a 
detection function was fit using pooled sightings of Bryde's whales and 
other mysticete species (Roberts et al., 2015c). The resulting 
detection function shows a slightly more than 20 percent probability of 
detecting whales at 2 km, with a mean effective strip half-width (ESHW) 
(which provides a measure of how far animals are seen from the transect 
line; Buckland et al., 2001) of 1,309 m (Roberts et al., 2015c). 
Similarly, Barlow et al. (2011) reported mean ESHWs for various 
mysticete species ranging from approximately 1.5-2 km. The detection 
function used in modeling density for beaked whales provided mean ESHWs 
of 1,462 m and 2,258 m for two NOAA vessels on which visual surveys 
have historically been conducted (Roberts et al., 2015b). Therefore, 
NMFS set the shutdown radius for special circumstances (described 
below) at 1.5 km for deep penetration surveys. The shutdown radius for 
special circumstances is set at 500 m for shallow penetration surveys.
    Comments disagreeing with the proposal to require shutdowns upon 
certain detections at any distance also suggested that the measures did 
not have commensurate benefit for the relevant species. However, it 
must be noted that any such observations would still be within range of 
where behavioral disturbance of some form and degree would be likely to 
occur. While visual PSOs should focus observational effort within the 
vicinity of the acoustic source and vessel, this does not preclude them 
from periodic scanning of the remainder of the visible area or from 
noting observations at greater distances, and there is no reason to 
believe that such periodic scans by professional PSOs would hamper the 
ability to maintain observation of areas closer to the source and 
vessel. Circumstances justifying shutdown at extended distance (e.g., 
within 1.5 km) include:
     Upon detection of a Bryde's whale. On the basis of the 
findings of NMFS' status review (Rosel et al., 2016), NMFS has listed 
the GOM Bryde's whale as an endangered species pursuant to the ESA 
(April 15, 2019; 84 FR 15446). These whales form a small and resident 
population in the northeastern GOM, with a highly restricted geographic 
range and a very small population abundance (fewer than 100)--recently 
determined by a status review team to be ``at or below the near-
extinction population level'' (Rosel et al., 2016). The review team 
stated that, aside from the restricted distribution and small 
population, the whales face a significant suite of anthropogenic 
threats, one of which is noise produced by geophysical surveys. NMFS 
believes it appropriate to eliminate potential effects to individual 
Bryde's whales to the extent practicable. There may be rare sightings 
of vagrant baleen whales of other species in the GOM, and the PSO may 
order a shutdown when observed in the applicable exclusion zone.
     Upon detection of a sperm whale. NMFS provided an expanded 
discussion of the available evidence that supports this measure in the 
notice of proposed rulemaking. In summary, the sperm whale's primary 
means of locating prey is echolocation (Miller et al., 2004), and 
multiple studies have shown that noise can disrupt feeding behavior 
and/or significantly reduce foraging success for sperm whales at 
relatively low levels of exposure (e.g., Miller et al., 2009, 2012; 
Isojunno et al.,

[[Page 5417]]

2016; Sivle et al., 2012; Cure et al., 2016). Effects on energy intake 
with no immediate compensation, as is suggested by disruption of 
foraging behavior without corollary movements to new locations, would 
be expected to result in bioenergetics consequences to individual 
whales. Farmer et al. (2018a) developed a stochastic life-stage 
structured bioenergetic model to evaluate the consequences of reduced 
foraging efficiency in sperm whales, finding that individual resilience 
to foraging disruptions is primarily a function of size (i.e., reserve 
capacity) and daily energetic demands, and that the ultimate effects on 
reproductive success and individual fitness are largely dependent on 
the duration and frequency of disturbance. The bioenergetic simulations 
of Farmer et al. (2018a) show that frequent disruptions in foraging, as 
might be expected when large amounts of survey activity overlap with 
areas of importance for sperm whales, can have potentially severe 
fitness consequences. In addition, the GOM sperm whale population was 
heavily impacted by the DWH oil spill. Therefore, in consideration of 
the potential energetic impacts of survey activity on individual sperm 
whales and the environmental baseline for the GOM sperm whale 
population, NMFS determined that meaningful measures must be taken to 
minimize disruption of foraging behavior. As described earlier in this 
section, the proposed regulations limited this extended distance 
shutdown requirement to acoustic detections of sperm whales. However, 
while stating that NMFS preliminarily did not believe the addition of 
shutdowns for sperm whales based on visual detections at any distance 
were warranted, we also requested any information from the public that 
would be relevant to that determination. NMFS' review of the comments 
and information provided by the public indicates that expansion of this 
requirement to include all sperm whale detections, rather than only 
acoustic detections (as was proposed), is warranted. Please see 
Comments and Responses for further discussion.
     Upon detection of a beaked whale or Kogia spp. These 
species are behaviorally sensitive deep divers and it is possible that 
disturbance could provoke a severe behavioral response leading to 
fitness consequences (e.g., Wursig et al., 1998; Cox et al., 2006). 
NMFS recognizes that there are generally low detection probabilities 
for beaked whales and Kogia spp., meaning that many animals of these 
species may go undetected. Barlow (1999) estimates such probabilities 
at 0.23 to 0.45 for Cuvier's and Mesoplodont beaked whales, 
respectively. However, Barlow and Gisiner (2006) predict a roughly 24-
48 percent reduction in the probability of detecting beaked whales 
during seismic mitigation monitoring efforts as compared with typical 
research survey efforts, and Moore and Barlow (2013) noted a decrease 
in g(0) for Cuvier's beaked whales from 0.23 at BSS 0 (calm) to 0.024 
at BSS 5. Similar detection probabilities have been noted for Kogia 
spp., though they typically travel in smaller groups and are less 
vocal, thus making detection more difficult (Barlow and Forney, 2007). 
As discussed previously in this document (see the Estimated Take 
section), there are high levels of predicted exposures for beaked 
whales in particular. Because it is likely that only a small proportion 
of beaked whales and Kogia spp. potentially affected by the proposed 
surveys would actually be detected, it is important to avoid potential 
impacts when practicable. Additionally for Kogia spp.--the one species 
of high-frequency cetacean likely to be encountered--auditory injury 
zones relative to peak pressure thresholds are significantly greater 
than for other cetaceans--approximately 500 m from the acoustic source, 
depending on the specific real world array characteristics (NMFS, 
2018).
    Shutdown Implementation Protocols--Any PSO on duty has the 
authority to delay the start of survey operations or to call for 
shutdown of the acoustic source. When shutdown is called for by a PSO, 
the acoustic source must be immediately deactivated and any dispute 
resolved only following deactivation. The operator must establish and 
maintain clear lines of communication directly between PSOs on duty and 
crew controlling the acoustic source to ensure that shutdown commands 
are conveyed swiftly while allowing PSOs to maintain watch; hand-held 
UHF radios are recommended. When both visual PSOs and PAM operators are 
on duty, all detections must be immediately communicated to the 
remainder of the on-duty team for potential verification of visual 
observations by the PAM operator or of acoustic detections by visual 
PSOs and initiation of dialogue as necessary. When there is certainty 
regarding the need for mitigation action on the basis of either visual 
or acoustic detection alone, the relevant PSO(s) must call for such 
action immediately.
    Upon implementation of shutdown, the source may be reactivated 
after the animal(s) has been observed exiting the exclusion zone or 
following a 30-minute clearance period with no further detection of the 
animal(s).
    If the acoustic source is shut down for reasons other than 
mitigation (e.g., mechanical difficulty) for brief periods (i.e., less 
than 30 minutes), it may be activated again without ramp-up if PSOs 
have maintained constant observation (including acoustic observation, 
where required) and no visual detections of any marine mammal have 
occurred within the exclusion zone and no acoustic detections have 
occurred (when required). NMFS defines ``brief periods'' in keeping 
with other clearance watch periods and to avoid unnecessary complexity 
in protocols for PSOs. For any longer shutdown (e.g., during line 
turns), pre-clearance watch and ramp-up are required. For any shutdown 
at night or in periods of poor visibility (e.g., BSS 4 or greater), 
ramp-up is required but if the shutdown period was brief and constant 
observation maintained, pre-clearance watch is not required.

Miscellaneous Protocols

    The acoustic source must be deactivated when not acquiring data or 
preparing to acquire data, except as necessary for testing. Unnecessary 
use of the acoustic source should be avoided. Firing of the acoustic 
source at any volume above the stated production volume would not be 
authorized. The operator must provide information to the lead PSO at 
regular intervals confirming the firing volume. Notified operational 
capacity (not including redundant backup airguns) must not be exceeded 
during the survey, except where unavoidable for source testing and 
calibration purposes. All occasions where activated source volume 
exceeds notified operational capacity must be noticed to the PSO(s) on 
duty and fully documented for reporting. The lead PSO must be granted 
access to relevant instrumentation documenting acoustic source power 
and/or operational volume.
    Testing of the acoustic source involving all elements requires 
normal mitigation protocols (e.g., ramp-up). Testing limited to 
individual source elements or strings does not require ramp-up but does 
require pre-clearance.

Restriction Areas

    Discussion of various time-area restrictions was provided in the 
notice of proposed rulemaking. NMFS proposed two time-area restrictions 
located within the area covered by the current GOMESA moratorium. As 
discussed previously, BOEM subsequently updated the scope of the 
specified activity that was the subject of

[[Page 5418]]

the petition for the ITR, removing the area subject to the current 
GOMESA moratorium from consideration through this rule. Therefore, 
consideration of those two proposed restrictions (Areas 2-3 in Figure 4 
below), and any alternatives, is no longer relevant. Figure 4 depicts 
the time-area restrictions, absent consideration of BOEM's removal of 
the GOMESA moratorium area. Areas 2 and 3 are entirely within that 
area, and the eastern extent of Area 1 is functionally reduced through 
the removal of the GOMESA moratorium area.
BILLING CODE 3510-22-P
[GRAPHIC] [TIFF OMITTED] TR19JA21.006

BILLING CODE 3510-22-C
    Coastal Restriction--No airgun surveys may occur from 90-84[deg] W 
(as truncated through removal of the GOMESA moratorium area) and 
shoreward of a line indicated by the 20-m isobath, during the months of 
January through May (Area 1; Figure 4). Waters shoreward of the 20-m 
isobath, where coastal dolphin stocks occur, represent the areas of 
greatest abundance for bottlenose dolphins (Roberts et al., 2016). As 
discussed above, and in greater detail in Comments and Responses, this 
requirement was modified from the proposed regulations by contracting 
the area spatially while expanding the restriction temporally by one 
month, in order to more practicably minimize potential impacts on the 
potentially affected stock most heavily impacted by the DWH oil spill 
(i.e., the northern coastal stock of bottlenose dolphins).
    The restriction is intended specifically to avoid additional 
stressors to the northern coastal stock of bottlenose dolphins during 
the time period believed to be of greatest importance as a reproductive 
period. As described previously, NOAA estimates that potentially 82 
percent of northern coastal dolphins were exposed to DWH oil, resulting 
in an array of long-term health impacts (including reproductive 
failure) and possible population reductions of 50 percent for the stock 
(DWH MMIQT, 2015). The same analysis estimated that these population-
level impacts could require 39 years to recovery, in the absence of 
other additional stressors. More recently, the stock has been subject 
to another declared UME; further discussion of this UME is provided 
under Description of Marine Mammals in the Area of the Specified 
Activity.
    The January-May timeframe is intended to best encompass the most 
important reproductive period for bottlenose dolphins in these coastal 
waters, when additional stress is most likely to have serious impacts 
on pregnancy and/or survival of neonates. Expert interpretation of the 
long-term data for neonate strandings is that February-April are the 
primary months that animals are born in the northern GOM, and that 
fewer but similar numbers are born in January and May. This refers to 
long-term averages and in any particular year the peak reproductive 
period can shift earlier or later.
    Bryde's Whale--The ``Bryde's whale core habitat area'' considered 
in the notice of proposed rulemaking was designated as between the 100- 
and 400-m isobaths, from 87.5[deg] W to 27.5[deg] N (Area 2; Figure 4). 
As summarized at the beginning of this section, and discussed in 
greater detail in Comments and Responses, the proposed regulatory text 
included a seasonal restriction within the same area. The preamble to 
the proposed regulations also included alternative proposals for public 
review and comment. This area is entirely

[[Page 5419]]

located in the GOMESA moratorium area, which is now removed from 
consideration through this rule.
    As described previously, NOAA's status review team determined the 
status of the GOM Bryde's whale to be precarious (Rosel et al., 2016). 
These findings formed, in part, the basis for the analysis presented in 
the preamble to the proposed regulations and subsequently supported 
NMFS' listing of the GOM Bryde's whale as an endangered species 
pursuant to the ESA (84 FR 15446; April 15, 2019). These whales form a 
small and resident population in the northeastern GOM, with a highly 
restricted geographic range and a very small population abundance--
determined by the status review team to be ``at or below the near-
extinction population level'' (Rosel et al., 2016). Aside from the 
restricted distribution and small population, the whales face a 
significant suite of anthropogenic threats, one of which is noise 
produced by airgun surveys.
    While various population abundance estimates are available (e.g., 
Waring et al., 2016; Roberts et al., 2016; Dias and Garrison, 2016), 
the population abundance was almost certainly less than 100 prior to 
the DWH oil spill. NOAA estimated that, as a result of that event, 48 
percent of the population may have been exposed to DWH oil, with 17 
percent killed and 22 percent of females experiencing reproductive 
failure. The best estimate for maximum population reduction was 22 
percent, with an estimated 69 years to recovery (to the precarious 
status prior to the DWH oil spill) (DWH MMIQT, 2015). It is considered 
likely that Bryde's whale habitat previously extended to shelf and 
slope areas of the western and central GOM similar to where they are 
found now in the eastern GOM, and that anthropogenic activity--largely 
energy exploration and production--concentrated in those areas could 
have resulted in habitat abandonment (Reeves et al., 2011; Rosel and 
Wilcox, 2014). Further, the population exhibits very low levels of 
genetic diversity and significant genetic mitochondrial DNA divergence 
from other Bryde's whales worldwide (Rosel and Wilcox, 2014).
    The small population size, restricted range, and low genetic 
diversity alone place these whales at significant risk of extinction 
(IWC, 2017), which has been exacerbated by the effects of the DWH oil 
spill. Additionally, Bryde's whale dive and foraging behavior places 
them at heightened risk of being struck by vessels and/or entangled in 
fishing gear (Soldevilla et al., 2017). NMFS considered a restriction 
in this core habitat area to protect Bryde's whales because of their 
hearing sensitivity in the lower frequency range (which makes them 
generally more susceptible to incurring effects from airgun noise than 
other taxa in the GOM); the potential impacts to important behavioral 
functions such as feeding, breeding, and raising young; their 
dangerously low population size; and other issues discussed previously. 
The absence of survey activity in the area would be expected to protect 
Bryde's whales and their habitat through the alleviation or 
minimization of a range of airgun effects, both acute and chronic, that 
could otherwise accrue to impact the reproduction or survival of 
individuals in the core habitat area. The absence of survey activity in 
the area would not only largely avoid Level B harassment of Bryde's 
whales, but also very importantly minimize other acoustic effects such 
as masking and loss of communication space. Based on Roberts et al., 
2016, this core habitat area is expected to encompass approximately 92 
percent of Bryde's whales in the Gulf of Mexico. The update of the 
scope of the rule eliminates this core area and the corresponding 
impacts of concern from consideration in the analysis.
    Although this area is no longer relevant under the updated 
geographic scope of the specified activity and this rule, the 
discussion above is still important to provide a picture of the 
species' distribution in the GOM and NMFS' work to identify appropriate 
mitigation in this rulemaking. Because NMFS acknowledges that some 
whales may be present at locations other than within this core habitat 
area, we considered additional information in order to evaluate whether 
a different closure area may be warranted. For example, a NOAA survey 
reported observation of a Bryde's whale in the western GOM in 2017 
(NMFS, 2018). There had not previously been a verified sighting of a 
Bryde's whale in the western GOM and, given the importance of this 
observation, additional survey effort was conducted in an attempt to 
increase effort in the area. However, no additional sightings were 
recorded. Overall, Bryde's whales observations have been consistently 
located within the eastern GOM core habitat area, with few whales 
sighted elsewhere despite a large amount of dedicated cetacean survey 
effort that covered both continental shelf and oceanic waters. Whales 
have been sighted in the core habitat area in all seasons, and all 
indications are that the whales inhabit this area year-round as a 
resident population. A tagged whale remained within the area for 38 
days, the entire time the tag was active. Therefore, while it is 
possible that Bryde's whales occur outside the core habitat area, or 
that whales from the eastern GOM occasionally travel outside the area, 
the few existing observations outside the eastern GOM do not affect 
NMFS' determination that the area considered in the proposed rule 
represents core habitat, or identify any additional important habitat 
that may appropriately be subject to a restriction on survey activity.

Entanglement Avoidance

    The use of ocean-bottom nodes (OBN) or similar equipment requiring 
the use of tethers or connecting lines poses an entanglement risk. In 
order to avoid incidents of entanglement, NMFS requires the same 
measures included for the same purpose in permits issued by BOEM. These 
measures apply to operators conducting OBN surveys (or surveys using 
similar equipment), and include: (1) Use negatively buoyant coated 
wire-core tether cable (e.g., \3/4\'' polyurethane-coated cable with 
\1/2\'' wire core); (2) retrieve all lines immediately following 
completion of the survey; and (3) attach acoustic pingers directly to 
the coated tether cable. Acoustic releases should not be used. No 
unnecessary release lines or lanyards may be used and nylon rope may 
not be used for any component of the system. Pingers must be attached 
directly to the nodal tether cable via shackle, with cables retrieved 
via grapnel. If a lanyard is required it must be as short as possible 
and made as stiff as possible, e.g., by placing inside a hose sleeve. 
The notice of proposed rulemaking also included a proposed requirement 
to require operators to employ a third-party PSO aboard the node 
retrieval vessel in order to document any unexpected marine mammal 
entanglement. In consideration of the information provided by public 
commenters, NMFS has determined that this measure is unnecessary and 
eliminates it from the final ITR. Use of a third-party PSO in this 
capacity would not help to avoid entanglement events, and operators 
would be required to report any such events to BSEE. Therefore, the 
requirement provides little benefit while imposing costs on operators.

Vessel Strike Avoidance

    These measures apply to all vessels associated with any survey 
activity (e.g., source vessels, streamer vessels, chase vessels, supply 
vessels). However, NMFS notes that these requirements do not apply in 
any case where compliance would create an imminent and serious

[[Page 5420]]

threat to a person or vessel or to the extent that a vessel is 
restricted in its ability to maneuver and, because of the restriction, 
cannot comply. These measures include the following:
    1. Vessel operators and crews must maintain a vigilant watch for 
all marine mammals and must slow down, stop their vessel, or alter 
course, as appropriate and regardless of vessel size, to avoid striking 
any marine mammal. A visual observer aboard the vessel must monitor a 
vessel strike avoidance zone around the vessel (distances stated 
below). Visual observers monitoring the vessel strike avoidance zone 
may be third-party observers (i.e., PSOs) or crew members, but crew 
members responsible for these duties must be provided sufficient 
training to (1) distinguish protected species from other phenomena and 
(2) broadly to identify a marine mammal as a baleen whale, sperm whale, 
or other marine mammal;
    2. Vessel speeds must be reduced to 10 kn or less when mother/calf 
pairs, pods, or large assemblages of any marine mammal are observed 
near a vessel;
    3. All vessels must maintain a minimum separation distance of 500 m 
from baleen whales;
    4. All vessels must maintain a minimum separation distance of 100 m 
from sperm whales;
    5. All vessels must, to the maximum extent practicable, attempt to 
maintain a minimum separation distance of 50 m from all other marine 
mammals, with an understanding that at times this may not be possible 
(e.g., for animals that approach the vessel); and
    6. When marine mammals are sighted while a vessel is underway, the 
vessel shall take action as necessary to avoid violating the relevant 
separation distance (e.g., attempt to remain parallel to the animal's 
course, avoid excessive speed or abrupt changes in direction until the 
animal has left the area). If marine mammals are sighted within the 
relevant separation distance, the vessel must reduce speed and shift 
the engine to neutral, not engaging the engines until animals are clear 
of the area. This does not apply to any vessel towing gear or any 
vessel that is navigationally constrained.
    NMFS has carefully evaluated the suite of mitigation measures 
described here and considered a range of other measures in the context 
of ensuring that we prescribe the means of effecting the least 
practicable adverse impact on the affected marine mammal species and 
stocks and their habitat. Based on our evaluation of these measures, we 
have determined that the required mitigation measures provide the means 
of effecting the least practicable adverse impact on marine mammal 
species or stocks and their habitat, paying particular attention to 
rookeries, mating grounds, and areas of similar significance.

Monitoring and Reporting

    In order to issue an LOA for an activity, Section 101(a)(5)(A) of 
the MMPA states that NMFS must set forth requirements pertaining to the 
monitoring and reporting of the authorized taking. NMFS' MMPA 
implementing regulations further describe the information that an 
applicant should provide when requesting an authorization (50 CFR 
216.104(a)(13)), including the means of accomplishing the necessary 
monitoring and reporting that will result in increased knowledge of the 
species and the level of taking or impacts on populations of marine 
mammals. Effective reporting is critical both to compliance as well as 
ensuring that the most value is obtained from the required monitoring.
    Section 101(a)(5)(A) allows that incidental taking may be 
authorized only if the total of such taking contemplated over the 
course of five years will have a negligible impact on affected species 
or stocks (a finding based on impacts to annual rates of recruitment 
and survival) and, further, section 101(a)(5)(B) requires that 
authorizations issued pursuant to 101(a)(5)(A) be withdrawn or 
suspended if the total taking is having, or may have, more than a 
negligible impact (or such information may inform decisions on requests 
for LOAs under the specific regulations). Therefore, the necessary 
requirements pertaining to monitoring and reporting must address the 
total annual impacts to marine mammal species or stocks. Effective 
reporting is critical both to compliance as well as ensuring that the 
most value is obtained from the required monitoring.
    These requirements are described below under ``Data Collection'' 
and ``LOA Reporting.'' Additional comprehensive reporting, across LOA-
holders on an annual basis, is also required and is described below 
under ``Comprehensive Reporting.''
    More specifically, monitoring and reporting requirements prescribed 
by NMFS should contribute to improved understanding of one or more of 
the following:
     Occurrence of marine mammal species in action area (e.g., 
presence, abundance, distribution, density);
     Nature, scope, or context of likely marine mammal exposure 
to potential stressors/impacts (individual or cumulative, acute or 
chronic), through better understanding of: (1) Action or environment 
(e.g., source characterization, propagation, ambient noise); (2) 
affected species (e.g., life history, dive patterns); (3) co-occurrence 
of marine mammal species with the action; or (4) biological or 
behavioral context of exposure (e.g., age, calving or feeding areas);
     Individual marine mammal responses (behavioral or 
physiological) to acoustic stressors (acute, chronic, or cumulative), 
other stressors, or cumulative impacts from multiple stressors;
     How anticipated responses to stressors impact either: (1) 
Long-term fitness and survival of individual marine mammals; or (2) 
populations, species, or stocks;
     Effects on marine mammal habitat (e.g., marine mammal prey 
species, acoustic habitat, or important physical components of marine 
mammal habitat); and
     Mitigation and monitoring effectiveness.

Changes To Monitoring and Reporting From the Notice of Proposed 
Rulemaking

    Here we summarize substantive changes to monitoring and reporting 
requirements from the notice of proposed rulemaking. All changes were 
made on the basis of review of public comments received and/or review 
of new information.
     Although NMFS recognizes the importance of producing the 
most accurate estimates of actual take possible, we agree that the 
specific approach described in the proposed rule for correcting 
observations to produce estimates of actual takes is novel in that it 
has not been previously required of applicants conducting similar 
activities and, therefore, its appropriateness for application to 
observations conducted from working source vessels (versus research 
vessels) is unknown. As suggested through public comment, NMFS will 
continue to evaluate the best method for producing accurate estimates 
of actual take, based on marine mammal detections, through the adaptive 
management process, including consideration of the Marine Mammal 
Commission-recommended method included in the proposed regulations.
     NMFS has revised requirements relating to reporting of 
injured or dead marine mammals and has added newly crafted requirements 
relating to actions that should be taken in response to notification of 
live stranding events in

[[Page 5421]]

certain circumstances, in order to reflect current best practice.

PSO Eligibility and Qualifications

    All PSO resumes must be submitted to NMFS and PSOs must be approved 
by NMFS after a review of their qualifications. These qualifications 
include whether the individual has successfully completed the necessary 
training (see ``Training,'' below) and, if relevant, whether the 
individual has the requisite experience (and is in good standing). PSOs 
should provide a current resume and information indicating successful 
completion of an acceptable PSO training course; submitted resumes 
should not include superfluous information. In order for a PSO training 
course to be deemed acceptable by NMFS (in consultation with BOEM/
BSEE), the agencies must, at minimum, review a course information 
packet that includes the name and qualifications (e.g., experience, 
training, or education) of the instructor(s), the course outline or 
syllabus, and course reference material. Absent a waiver (discussed 
below), PSOs must be trained biologists, with the following minimum 
qualifications:
     A bachelor's degree from an accredited college or 
university with a major in one of the natural sciences and a minimum of 
30 semester hours or equivalent in the biological sciences and at least 
one undergraduate course in math or statistics; and
     Successful completion of relevant training (described 
below), including completion of all required coursework and passing (80 
percent or greater) a written and/or oral examination developed for the 
training program.
    In addition, it is recommended that PSOs meet the following 
requirements:
     Experience and ability to conduct field observations and 
collect data according to assigned protocols (may include academic 
experience) and experience with data entry on computers;
     Visual acuity in both eyes (vision correction is 
permissible) sufficient for discernment of moving targets at the 
water's surface with ability to estimate target size and distance; use 
of binoculars may be necessary to correctly identify the target 
(required for visual PSOs only);
     Experience or training in the field identification of 
marine mammals, including the identification of behaviors (required for 
visual PSOs only);
     Sufficient training, orientation, or experience with the 
survey operation to ensure personal safety during observations;
     Writing skills sufficient to prepare a report of 
observations (e.g., description, summary, interpretation, analysis) 
including but not limited to the number and species of marine mammals 
observed; marine mammal behavior; and descriptions of activity 
conducted and implementation of mitigation; and
     Ability to communicate orally, by radio or in person, with 
survey personnel to provide real-time information on marine mammals 
detected in the area as necessary.
    The educational requirements may be waived if the PSO has acquired 
the relevant skills through alternate experience. Requests for such a 
waiver must include written justification, and prospective PSOs granted 
waivers must satisfy training requirements described below. Alternate 
experience that may be considered includes, but is not limited to, the 
following:
     Secondary education and/or experience comparable to PSO 
duties;
     Previous work experience conducting academic, commercial, 
or government-sponsored marine mammal surveys; and
     Previous work experience as a PSO; the PSO should 
demonstrate good standing and consistently good performance of PSO 
duties.
    Training--NMFS does not formally administer any PSO training 
program or endorse specific providers but will approve PSOs that have 
successfully completed courses that meet the curriculum and trainer 
requirements specified herein and, therefore, are deemed acceptable. To 
be deemed acceptable, training should adhere generally to the 
recommendations provided by ``National Standards for a Protected 
Species Observer and Data Management Program: A Model Using Geological 
and Geophysical Surveys'' (Baker et al., 2013). Those recommendations 
include the following topics for training programs:
     Life at sea, duties, and authorities;
     Ethics, conflicts of interest, standards of conduct, and 
data confidentiality;
     Offshore survival and safety training;
     Overview of oil and gas activities (including geophysical 
data acquisition operations, theory, and principles) and types of 
relevant sound source technology and equipment;
     Overview of the MMPA and ESA as they relate to protection 
of marine mammals;
     Mitigation, monitoring, and reporting requirements as they 
pertain to geophysical surveys;
     Marine mammal identification, biology and behavior;
     Background on underwater sound;
     Visual surveying protocols, distance calculations and 
determination, cues, and search methods for locating and tracking 
different marine mammal species (visual PSOs only);
     Optimized deployment and configuration of PAM equipment to 
ensure effective detections of cetaceans for mitigation purposes (PAM 
operators only);
     Detection and identification of vocalizing species or 
cetacean groups (PAM operators only);
     Measuring distance and bearing of vocalizing cetaceans 
while accounting for vessel movement (PAM operators only);
     Data recording and protocols, including standard forms and 
reports, determining range, distance, direction, and bearing of marine 
mammals and vessels; recording GPS location coordinates, weather 
conditions, Beaufort wind force and sea state, etc.;
     Proficiency with relevant software tools;
     Field communication/support with appropriate personnel, 
and using communication devices (e.g., two-way radios, satellite 
phones, internet, email, facsimile);
     Reporting of violations, noncompliance, and coercion; and
     Conflict resolution.
    PAM operators should regularly refresh their detection skills 
through practice with simulation-modeling software and should keep up 
to date with training on the latest software/hardware advances.

Visual Monitoring

    The lead PSO is responsible for establishing and maintaining clear 
lines of communication with vessel crew. The vessel operator shall work 
with the lead PSO to accomplish this and shall ensure any necessary 
briefings are provided for vessel crew to understand mitigation 
requirements and protocols. While on duty, PSOs will continually scan 
the water surface in all directions around the acoustic source and 
vessel for presence of marine mammals, using a combination of the naked 
eye and high-quality binoculars, from optimum vantage points for 
unimpaired visual observations with minimum distractions. PSOs will 
collect observational data for all marine mammals observed, regardless 
of distance from the vessel, including species, group size, presence of 
calves, distance from vessel and direction of travel, and any observed 
behavior (including an assessment of behavioral responses to survey 
activity). Upon observation of marine mammal(s), a

[[Page 5422]]

PSO will record the observation and monitor the animal's position 
(including latitude/longitude of the vessel and relative bearing and 
estimated distance to the animal) until the animal dives or moves out 
of visual range of the observer, and a PSO will continue to observe the 
area to watch for the animal to resurface or for additional animals 
that may surface in the area. PSOs will also record environmental 
conditions at the beginning and end of the observation period and at 
the time of any observations, as well as whenever conditions change 
significantly in the judgment of the PSO on duty.
    For all deep penetration surveys, the vessel operator must provide 
bigeye binoculars of appropriate quality (e.g., 25 x 150; 2.7 view 
angle; individual ocular focus; height control) solely for PSO use. 
These should be pedestal-mounted on the deck at the most appropriate 
vantage point that provides for optimal sea surface observation, PSO 
safety, and safe operation of the vessel. Other required equipment, 
which should be made available to PSOs by the third-party observer 
provider, includes reticle binoculars of appropriate quality (e.g., 7 x 
50), GPS, digital camera with a telephoto lens (the camera or lens 
should also have an image stabilization system) that is at least 300 mm 
or equivalent on a full-frame single-lens reflex, compass, and any 
other tools necessary to adequately perform the tasks described above, 
including accurate determination of distance and bearing to observed 
marine mammals.
    Individuals implementing the monitoring protocol will assess its 
effectiveness using an adaptive approach. Monitoring biologists will 
use their best professional judgment throughout implementation and seek 
improvements to these methods when deemed appropriate. Specifically, 
implementation of shutdown requirements will be made on the basis of 
the PSO's best professional judgment. While PSOs should not insert 
undue precaution into decision-making, it is expected that PSOs may 
call for mitigation action on the basis of reasonable certainty 
regarding the need for such action, as informed by professional 
judgment. Any modifications to protocol will be coordinated between 
NMFS and the applicant.

Acoustic Monitoring

    Use of towed PAM is required for deep penetration surveys. 
Monitoring of a towed PAM system is required at all times for these 
surveys, from 30 minutes prior to ramp-up, throughout all use of the 
acoustic source, and for 60 minutes following cessation of survey 
activity. Towed PAM systems should consist of hardware (e.g., 
hydrophone array, recorder, cables) and software (e.g., data processing 
program and algorithm). Some type of automated detection software must 
be used. Acoustic signals are processed for output to the PAM operator 
with software designed to detect marine mammal vocalizations. Current 
PAM technology has some limitations (e.g., limited directional 
capabilities and detection range, detection of signals due to vessel 
and flow noise, low accuracy in localization) and there are no formal 
guidelines currently in place regarding specifications for hardware, 
software, or operator training requirements.
    NMFS' requirement to use PAM refers to the use of calibrated 
hydrophone arrays with full system redundancy to detect, identify, and 
estimate distance and bearing to vocalizing cetaceans, to the extent 
possible. With regard to calibration, the PAM system should have at 
least one calibrated hydrophone, sufficient for determining whether 
background noise levels on the towed PAM system are sufficiently low to 
meet performance expectations. Additionally, if multiple hydrophone 
types occur in a system (i.e., monitor different bandwidths), then one 
hydrophone from each such type shall be calibrated, and whenever sets 
of hydrophones (of the same type) are sufficiently spatially separated 
such that they would be expected to experience ambient noise 
environments that differ by 6 dB or more across any integrated species 
cluster bandwidth, then at least one hydrophone from each set should be 
calibrated. In terms of calibrating the rest of the system, the signal 
route to the data recorder and monitoring software shall be calibrated 
so that the binary amplitude data written to hard disk can be converted 
into units of acoustic pressure. The configuration of hardware should 
be coupled with appropriate software to aid monitoring and listening by 
a PAM operator skilled in bioacoustics analysis and computer system 
specifications capable of running appropriate software. GPS data 
acquisition is recommended for all PAM operations. If the PAM plan (see 
below) claims an ability to localize, every localization estimate 
obtained from a PAM system must be accompanied by some estimate of 
uncertainty and ambiguity.
    In the absence of formal standards addressing any of these three 
facets of PAM technology, all applicants must provide a PAM plan 
including description of the hardware and software proposed for use 
prior to proceeding with any survey where PAM is required. Following 
the survey, a validation document must be submitted as part of required 
reporting (see below). The purpose of the PAM plan is to demonstrate 
that the PAM system being proposed for use is adequate for addressing 
the mitigation goals. The plan shall include methodology and 
documentation requirements for all stages of the project. As 
recommended by Thode et al. (2017), PAM plans should, at minimum, 
adequately address and describe (1) the hardware and software planned 
for use, including a hardware performance diagram demonstrating that 
the sensitivity and dynamic range of the hardware is appropriate for 
the operation; (2) deployment methodology, including target depth/tow 
distance; (3) definitions of expected operational conditions, used to 
summarize background noise statistics; (4) proposed detection-
classification-localization methodology, including anticipated species 
clusters (using a cluster definition table), target minimum detection 
range for each cluster, and the proposed localization method for each 
cluster; (5) operation plans, including the background noise sampling 
schedule; (6) array design considerations for noise abatement; and (7) 
cluster-specific details regarding which real-time displays and 
automated detectors the operator would monitor. Where relevant, the 
plan should address the potential for PAM deployment on a receiver 
vessel or other associated vessel separate from the acoustic source.
    Species clusters--The plan shall list the species of concern during 
the upcoming operation. While some species may be listed individually 
for special attention, in many circumstances it is expected that for 
the purposes of a PAM operation multiple species can be grouped 
together in a ``cluster'' that shares similar acoustic and behavioral 
characteristics (e.g., sperm whale, beaked whales). The plan must 
specify a target minimum detection (and possibly localization) range 
for each species cluster used in the document. Different ranges can be 
defined for different operational conditions. The PAM system may exceed 
this detection range, but shall always be capable of achieving this 
minimum detection range.
    Hardware and software specifications--The plan shall have a section 
dedicated to demonstrating that the PAM hardware is sensitive enough to 
detect signals from the species clusters of concern at the target 
minimum detection ranges specified. The plan should include a hardware

[[Page 5423]]

specification table and hardware performance diagram. The diagram will 
show the sensitivity and bandwidth of the combined array hardware and 
recording system, as well as the received levels required for a given 
species cluster to be detectable at the target minimum detection range. 
The overall goal of the diagram is to visually demonstrate that the 
planned PAM array/recording system would have the capability of 
detecting various species clusters at required target ranges, provided 
that background noise levels are not an issue.
    Operational conditions--The validation document should demonstrate 
whether the PAM system has been compromised by excessive background 
noise, whether that noise is electronic interference, flow, platform, 
or environmental noise. Therefore, the plan shall define a set of 
``operational conditions'' under which detection statistics (background 
noise profiles) will be categorized during the project. Operational 
conditions consist of three categories: Platform activity and status, 
mitigation (activity) status, and environmental status.
    Operating procedures--The plan shall describe the level of effort 
that is reasonably expected to occur for the monitoring requirements. 
For every species cluster, the plan should detail which part of the PAM 
display would be used for detecting that cluster. For example, if a 
scrolling spectrogram display is being used for a species cluster, then 
the spectrogram's fast Fourier transform sample size, frequency 
bandwidth, and their refresh rate shall be specified. Similar details 
would be provided for other software tools, such as click detectors and 
other automated detectors and classifiers. The plan shall also provide 
a screenshot of the expected monitor display.
    In coordination with vessel crew, the lead PAM operator will be 
responsible for deployment, retrieval, and testing and optimization of 
the hydrophone array. While on duty, the PAM operator must diligently 
listen to received signals and/or monitoring display screens in order 
to detect vocalizing cetaceans, except as required to attend to PAM 
equipment. The PAM operator must use appropriate sample analysis and 
filtering techniques and must report all cetacean detections. While not 
required prior to development of formal standards for PAM use, NMFS 
recommends that vessel self-noise assessments be undertaken during 
mobilization in order to optimize PAM array configuration according to 
the specific noise characteristics of the vessel and equipment 
involved, and to refine expectations for distance/bearing estimations 
for cetacean species during the survey. Copies of any vessel self-noise 
assessment reports must be included with the summary trip report.

Data Collection

    PSOs must use standardized electronic data forms. PSOs will record 
detailed information about any implementation of mitigation 
requirements, including the distance of animals to the acoustic source 
and description of specific actions that ensued, the behavior of the 
animal(s), any observed changes in behavior before and after 
implementation of mitigation, and if shutdown was implemented, the 
length of time before any subsequent ramp-up of the acoustic source to 
resume survey. If required mitigation was not implemented, PSOs should 
submit a description of the circumstances. NMFS requires that, at a 
minimum, the following information be reported:
     Vessel names (source vessel and other vessels associated 
with survey), vessel size and type, maximum speed capability of vessel, 
port of origin, and call signs;
     PSO names and affiliations;
     Dates of departures and returns to port with port name;
     Dates and participants of PSO briefings;
     Dates and times (Greenwich Mean Time) of survey effort and 
times corresponding with PSO effort;
     Vessel location (latitude/longitude) when survey effort 
begins and ends and vessel location at beginning and end of visual PSO 
duty shifts;
     Vessel location at 30 second intervals (if software 
capability allows) or 5 minute intervals (if location must be manually 
recorded);
     Vessel heading and speed at beginning and end of visual 
PSO duty shifts and upon any line change;
     Environmental conditions while on visual survey (at 
beginning and end of PSO shift and whenever conditions change 
significantly), including Beaufort scale and any other relevant weather 
conditions including cloud cover, fog, sun glare, night, and overall 
visibility to the horizon;
     Vessel location when environmental conditions change 
significantly;
     Factors that may have contributed to impaired observations 
during each PSO shift change or as needed as environmental conditions 
change (e.g., vessel traffic, equipment malfunctions);
     Survey activity information, such as acoustic source power 
output while in operation, number and volume of airguns operating in an 
array, tow depth of an acoustic source, and any other notes of 
significance (i.e., pre-clearance, ramp-up, shutdown, testing, 
shooting, ramp-up completion, end of operations, streamers, etc.);
     If a marine mammal is sighted, the following information 
should be recorded:
    [cir] Watch status (sighting made by PSO on/off effort, 
opportunistic, crew, alternate vessel/platform);
    [cir] PSO who sighted the animal and PSO location (including height 
above water) at time of sighting;
    [cir] Time of sighting;
    [cir] Vessel location at time of sighting;
    [cir] Water depth;
    [cir] Direction of vessel's travel (compass direction);
    [cir] Direction of animal's travel relative to the vessel;
    [cir] Pace of the animal;
    [cir] Estimated distance to the animal (and method of estimating 
distance) and its heading relative to vessel at initial sighting;
    [cir] Identification of the animal (e.g., genus/species, lowest 
possible taxonomic level, or unidentified) and PSO confidence in 
identification; also note the composition of the group if there is a 
mix of species;
    [cir] Estimated number of animals (high/low/best);
    [cir] Estimated number of animals by cohort (adults, yearlings, 
juveniles, calves, group composition, etc.);
    [cir] Description (as many distinguishing features as possible of 
each individual seen, including length, shape, color, pattern, scars or 
markings, shape and size of dorsal fin, shape of head, and blow 
characteristics);
    [cir] Detailed behavior observations (e.g., number of blows, number 
of surfaces, breaching, spyhopping, diving, feeding, traveling; as 
explicit and detailed as possible; note any observed changes in 
behavior);
    [cir] Animal's closest point of approach (CPA) and/or closest 
distance from the acoustic source;
    [cir] Platform activity at time of sighting (e.g., deploying, 
recovering, testing, shooting, data acquisition, other); and
    [cir] Description of any actions implemented in response to the 
sighting (e.g., delays, shutdown, ramp-up); time and location of the 
action should also be recorded;
     If a marine mammal is detected while using the PAM system, 
the following information should be recorded:
    [cir] An acoustic encounter identification number, and whether the 
detection was linked with a visual sighting;

[[Page 5424]]

    [cir] Time when first and last heard;
    [cir] Types and nature of sounds heard (e.g., clicks, whistles, 
creaks, burst pulses, continuous, sporadic, strength of signal, etc.); 
and
    [cir] Any additional information recorded such as water depth of 
the hydrophone array, bearing of the animal to the vessel (if 
determinable), species or taxonomic group (if determinable), 
spectrogram screenshot, and any other notable information.

LOA Reporting

    PSO effort, survey details, and sightings data should be recorded 
continuously during surveys. Reports must include all information 
described above under ``Data Collection,'' including amount and 
location of line-kms surveyed and all marine mammal observations with 
closest approach distance. Reports must be submitted to NMFS within 90 
days of survey completion or following expiration of an issued LOA. In 
the event that an LOA is issued for a period exceeding one year, annual 
reports must be submitted during the period of validity. The draft 
report must be accompanied by a certification from lead PSOs as to the 
accuracy of the report. A final report must be submitted within 30 days 
following resolution of any comments on the draft report.
    The report must describe the operations conducted and sightings of 
marine mammals near the operations; provide full documentation of 
methods, results, and interpretation pertaining to all monitoring; 
summarize the dates and locations of survey operations, and all marine 
mammal sightings (dates, times, locations, activities, associated 
survey activities); and provide information regarding locations where 
the acoustic source was used. The LOA-holder shall provide geo-
referenced time-stamped vessel tracklines for all time periods in which 
airguns (full array or single) were operating. Tracklines should 
include points recording any change in airgun status (e.g., when the 
airguns began operating, when they were turned off). GIS files shall be 
provided in ESRI shapefile format and include the UTC date and time, 
latitude in decimal degrees, and longitude in decimal degrees. All 
coordinates should be referenced to the WGS84 geographic coordinate 
system. In addition to the report, all raw observational data shall be 
made available to NMFS.
    This report must also include a validation document concerning the 
use of PAM (if PAM was required), which should include necessary noise 
validation diagrams (NVD) and demonstrate whether background noise 
levels on the PAM deployment limited achievement of the planned 
detection goals. A separate diagram shall be produced for every 
background noise percentile chosen for analysis. Background noise 
percentiles, rather than a simple average of the data, are required 
because the highly non-stationary characteristics of many background 
noise profiles cannot be described by a simple mean. For example, data 
collected during a seismic survey will have short periods of time 
containing high-intensity pulses and longer periods of time dominated 
by lower levels of reverberation. Taking a simple mean of these noise 
data would imply background noise levels substantially higher than what 
may actually have been present between seismic pulses. A validation 
report would typically contain between three to five diagrams, 
depending on the number of percentiles analyzed. At a minimum, the 
validation report should contain three diagrams that include the 50th 
percentile (median), 5th percentile, and 95th percentile. The 25th 
percentile and 75th percentile may also be included. In each percentile 
diagram, a separate background noise curve shall be drawn for each 
defined operational condition. In general, the NVD should be generated 
from the data stream that is used for detecting the presence of marine 
mammal signals. For example, if beamforming or some other form of array 
gain has been applied before invoking signal detection, then the NVD 
should be generated using the beamformed data, and not omnidirectional 
data. The complete set of NVDs, one for each percentile of interest, 
combined with a table that lists the fraction of time the activity was 
in each operational state, provides a means of reviewing the background 
noise-limitations encountered by the PAM system during various 
operational conditions. Actual marine mammal detections should be 
plotted on this diagram for a reasonableness check on the expected 
received levels. Overall, the validation document should reiterate all 
the goals and parameters stated in the planning document and verify 
that goals were/were not met, why, changes, etc. Also, the validation 
document should state whether the planning was suited to the needs of 
the survey and met the required mitigation standards.
    There are multiple reasons why marine mammals may be present and 
yet be undetected by observers. Animals are missed because they are 
underwater (availability bias) or because they are available to be 
seen, but are missed by observers (perception and detection biases) 
(e.g., Marsh and Sinclair, 1989). Negative bias on perception or 
detection of an available animal may result from environmental 
conditions, limitations inherent to the observation platform, or 
observer ability. In this case, we do not have prior knowledge of any 
potential negative bias on detection probability due to observation 
platform or observer ability. Therefore, it may be appropriate to make 
observational data corrections with respect to assumed species-specific 
detection probability as evaluated through consideration of 
environmental factors (e.g., f(0)). Appropriate methods will be 
considered through the adaptive management process.
    The report must include a post-survey estimate of the instances of 
take of each species utilizing the line miles of survey actually 
conducted and the same methods used to initially predict the estimated 
take in the LOA application. Depending on the length and dates of the 
survey, LOA-holders may be required to segment take estimates into 
specific years to support the administration of the rule.

Comprehensive Reporting

    Individual LOA-holders will be responsible for collecting and 
submitting monitoring data to NMFS, as described above. In addition, on 
an annual basis, LOA-holders will also collectively be responsible for 
compilation and analysis of those data for inclusion in subsequent 
annual synthesis reports. Individual LOA-holders may collaborate to 
produce this report or may elect to have their trade associations 
support the production of such a report. These reports would summarize 
the data presented in the individual LOA-holder reports, provide 
analysis of these synthesized results, discuss the implementation of 
required mitigation, and present any recommendations. This 
comprehensive annual report would be the basis of an annual adaptive 
management process (described below in Adaptive Management). The 
following topics will be described in comprehensive reporting:
     Summary of geophysical survey activity by survey type, 
geographic zone (i.e., the seven zones described in the modeling 
report), month, and acoustic source status (e.g., inactive, ramp-up, 
full-power, power-down);
     Summary of monitoring effort (on-effort hours and/or 
distance) by acoustic source status, location, and visibility 
conditions (for both visual and acoustic monitoring);
     Summary of mitigation measures implemented (e.g., delayed 
ramp-ups,

[[Page 5425]]

shutdowns, course alterations for vessel strike avoidance) by survey 
type and location;
     Sighting rates of marine mammals during periods with and 
without acoustic source activities and other variables that could 
affect detectability of marine mammals, such as:
    [cir] Initial sighting distances of marine mammals relative to 
source status;
    [cir] Closest point of approach of marine mammals relative to 
source status;
    [cir] Observed behaviors and types of movements of marine mammals 
relative to source status;
    [cir] Distribution/presence of marine mammals around the survey 
vessel relative to source status; and
    [cir] Analysis of the effects of various factors influencing the 
detectability of marine mammals (e.g., wind speed, sea state, swell 
height, presence of glare or fog).
     Estimates of total take across all activities for which 
take is authorized based on actual survey effort and original 
estimation method;
     Summary and conclusions from monitoring in previous year; 
and
     Recommendations for adaptive management.
    Each annual comprehensive report should cover one full year of 
monitoring effort and must be submitted for review each year. Each 
report should analyze survey and monitoring effort described in reports 
submitted by individual LOA-holders during a given one-year period, 
beginning from the date of effectiveness of these regulations. Each 
annual comprehensive report must be submitted for review 90 days 
following conclusion of the annual reporting period.

Reporting Injured or Dead Marine Mammals

    Discovery of Injured or Dead Marine Mammal--In the event that 
personnel involved in the survey activities covered by the 
authorization discover an injured or dead marine mammal, the LOA-holder 
shall report the incident to the Office of Protected Resources (OPR), 
NMFS and to the regional stranding network as soon as feasible. The 
report must include the following information:
     Time, date, and location (latitude/longitude) of the first 
discovery (and updated location information if known and applicable);
     Species identification (if known) or description of the 
animal(s) involved;
     Condition of the animal(s) (including carcass condition if 
the animal is dead);
     Observed behaviors of the animal(s), if alive;
     If available, photographs or video footage of the 
animal(s); and
     General circumstances under which the animal was 
discovered.
    Vessel Strike--In the event of a ship strike of a marine mammal by 
any vessel involved in the activities covered by the authorization, the 
LOA-holder shall report the incident to OPR, NMFS and to the regional 
stranding network as soon as feasible. The report must include the 
following information:
     Time, date, and location (latitude/longitude) of the 
incident;
     Species identification (if known) or description of the 
animal(s) involved;
     Vessel's speed during and leading up to the incident;
     Vessel's course/heading and what operations were being 
conducted (if applicable);
     Status of all sound sources in use;
     Description of avoidance measures/requirements that were 
in place at the time of the strike and what additional measures were 
taken, if any, to avoid strike;
     Environmental conditions (e.g., wind speed and direction, 
Beaufort sea state, cloud cover, visibility) immediately preceding the 
strike;
     Estimated size and length of animal that was struck;
     Description of the behavior of the marine mammal 
immediately preceding and following the strike;
     If available, description of the presence and behavior of 
any other marine mammals immediately preceding the strike;
     Estimated fate of the animal (e.g., dead, injured but 
alive, injured and moving, blood or tissue observed in the water, 
status unknown, disappeared); and
     To the extent practicable, photographs or video footage of 
the animal(s).

Actions To Minimize Additional Harm to Live-Stranded (or Milling) 
Marine Mammals

    For deep penetration surveys, in the event of a live stranding (or 
near-shore atypical milling) event within 50 km of the survey 
operations, where the NMFS stranding network is engaged in herding or 
other interventions to return animals to the water, the Director of 
OPR, NMFS (or designee) will advise the LOA-holder of the need to 
implement shutdown procedures for all active acoustic sources operating 
within 50 km of the stranding. Shutdown procedures for live stranding 
or milling marine mammals include the following:
     If at any time, the marine mammals die or are euthanized, 
or if herding/intervention efforts are stopped, the Director of OPR, 
NMFS (or designee) will advise the LOA-holder that the shutdown around 
the animals' location is no longer needed.
     Otherwise, shutdown procedures will remain in effect until 
the Director of OPR, NMFS (or designee) determines and advises the LOA-
holder that all live animals involved have left the area (either of 
their own volition or following an intervention).
     If further observations of the marine mammals indicate the 
potential for re-stranding, additional coordination with the LOA-holder 
will be required to determine what measures are necessary to minimize 
that likelihood (e.g., extending the shutdown or moving operations 
farther away) and to implement those measures as appropriate.
    Shutdown procedures are not related to the investigation of the 
cause of the stranding and their implementation is not intended to 
imply that the specified activity is the cause of the stranding. 
Rather, shutdown procedures are intended to protect marine mammals 
exhibiting indicators of distress by minimizing their exposure to 
possible additional stressors, regardless of the factors that 
contributed to the stranding.
    Additional Information Requests--If NMFS determines that the 
circumstances of any marine mammal stranding found in the vicinity of 
the activity suggest investigation of the association with survey 
activities is warranted (example circumstances noted below), and an 
investigation into the stranding is being pursued, NMFS will submit a 
written request to the LOA-holder indicating that the following initial 
available information must be provided as soon as possible, but no 
later than 7 business days after the request for information.
     Status of all sound source use in the 48 hours preceding 
the estimated time of stranding and within 50 km of the discovery/
notification of the stranding by NMFS; and
     If available, description of the behavior of any marine 
mammal(s) observed preceding (i.e., within 48 hours and 50 km) and 
immediately after the discovery of the stranding.
    Examples of circumstances that could trigger the additional 
information request include, but are not limited to, the following:
     Atypical nearshore milling events of live cetaceans;
     Mass strandings of cetaceans (two or more individuals, not 
including cow/calf pairs);
     Beaked whale strandings; or,
     Necropsies with findings of pathologies that are unusual 
for the species or area.

[[Page 5426]]

    In the event that the investigation is still inconclusive, the 
investigation of the association of the survey activities is still 
warranted, and the investigation is still being pursued, NMFS may 
provide additional information requests, in writing, regarding the 
nature and location of survey operations prior to the time period 
above.

Negligible Impact Analysis and Determinations

    NMFS has defined negligible impact as an impact resulting from the 
specified activity that cannot be reasonably expected to, and is not 
reasonably likely to, adversely affect the species or stock through 
effects on annual rates of recruitment or survival (50 CFR 216.103). A 
negligible impact finding is based on the lack of likely adverse 
effects on annual rates of recruitment or survival (i.e., population-
level effects). An estimate of the number of takes alone is not enough 
information on which to base a negligible impact determination. In 
addition to considering estimates of the number of marine mammals that 
might be ``taken'' by mortality, serious injury, and Level A or Level B 
harassment, we consider other factors, such as the type of take, the 
likely nature of any behavioral responses (e.g., intensity, duration), 
the context of any such responses (e.g., critical reproductive time or 
location, migration), as well as effects on habitat, and the likely 
effectiveness of mitigation. We also assess the number, intensity, and 
context of estimated takes by evaluating this information relative to 
population status. Consistent with the 1989 preamble for NMFS' 
implementing regulations (54 FR 40338; September 29, 1989), the impacts 
from other past and ongoing anthropogenic activities are incorporated 
into these analyses via their impacts on the baseline (e.g., as 
reflected in the regulatory status of the species, population size and 
growth rate where known, ongoing sources of human-caused mortality).
    For each potential activity-related stressor, NMFS considers the 
potential effects to marine mammals and the likely significance of 
those effects to the species or stock as a whole. Potential risk due to 
vessel collision and related mitigation measures, as well as potential 
risk due to entanglement and contaminant spills, was addressed under 
Mitigation and in the Potential Effects of the Specified Activity on 
Marine Mammals section of this notice and the notice of proposed 
rulemaking and are not discussed further, as there are minimal risks 
expected from these potential stressors.
    The ``specified activity'' for these regulations is a broad program 
of geophysical survey activity that could occur at any time of year in 
U.S. waters of the GOM, within the specified geographical region as 
updated by BOEM (i.e., excluding the GOMESA leasing moratorium area). 
In recognition of the broad scale of this activity in terms of 
geographic and temporal scales, we use a new analytical methodology--
first described by Ellison et al. (2015) and proposed for use and 
discussed in detail in the notice of proposed rulemaking--through which 
an explicit, systematic risk assessment framework is applied to 
evaluate potential effects of aggregated discrete acoustic exposure 
events (i.e., proposed geophysical survey activities) on marine 
mammals. This risk assessment framework is one component of the overall 
negligible impact analysis. Development of the approach was supported 
collaboratively by BOEM and NMFS, which together provided guidance to 
an expert working group (EWG) in terms of application to relevant 
regulatory processes. The risk assessment framework (or EWG framework) 
is described by Southall et al. (2017), which is available online at: 
www.fisheries.noaa.gov/national/marine-mammal-protection/incidental-take-authorizations-oil-and-gas. That document is a companion to this 
analysis, and is referred to hereafter as the ``EWG report.'' The risk 
assessment framework is also described below. It was developed and 
implemented by the EWG in relation to the specified activity described 
in the proposed rule, provided for public review in association with 
the notice of proposed rulemaking (Southall et al., 2017), and 
subsequently refined in response to public comment and in consideration 
of the updated scope of the activity. We incorporate the framework and 
its results into this analysis.
    The EWG framework described below comprehensively considers the 
aggregate impacts to marine mammal populations from the activities 
addressed in this rule in the context of both (1) the severity of the 
impacts and (2) the vulnerability of the affected species. However, it 
does not consider the effects of the mitigation required through these 
regulations in identifying risk ratings for the affected species. In 
addition, while the EWG framework comprehensively considers the spatial 
and temporal overlay of the activities and the marine mammals in the 
GOM, as well as the number of takes predicted by the described modeling 
(both in the proposed rule, and as updated in this final rule), there 
are details about the nature of any ``take'' anticipated to result from 
these activities that were not considered directly in the EWG framework 
analysis that warrant explicit consideration in the negligible impact 
determination. Last, the EWG framework analysis addresses impacts to 
guilds in some cases where there is not specific information to further 
support species-specific findings. Accordingly, following the 
description of the EWG framework below, NMFS highlights a few factors 
regarding the nature of the predicted ``takes'' and then brings 
together the results of implementation of the EWG framework, these 
additional factors, and the anticipated effects of the mitigation to 
summarize the negligible impact analysis for each of the affected 
species or stocks.

EWG Risk Assessment

    The acoustic exposure modeling (Zeddies et al., 2015, 2017a) 
provided marine mammal noise exposure estimates based on BOEM-provided 
projections of future survey effort and best available modeling of 
sound propagation, animal distribution, and animal movement. This 
provided a conservative but reasonable best estimate of potential acute 
noise exposure events that may result from the described suite of 
activities, and formed the basis for the analysis in the proposed rule. 
BOEM subsequently updated the scope of its activity, which reduced the 
amount of activity overall through removal of projected activity in the 
eastern GOM (see Table 1 and Figure 2). Acoustic exposure estimates 
were updated by BOEM accordingly (based on the same modeling presented 
in the proposed rule) and these revised estimates form the basis for 
this updated analysis.
    The primary goal of this new analytical effort was to develop a 
systematic risk assessment framework that would use the modeling 
results to put into biologically-relevant context the level of 
potential risk of injury and/or disturbance to marine mammals. The risk 
assessment framework considers both the aggregation of acute effects 
and the broad temporal and spatial scales over which chronic effects 
may occur. Previously, Wood et al. (2012) conducted an analysis of a 
proposed airgun survey, in which the authors derived a qualitative risk 
assessment method of considering the biological significance of 
exposures predicted to be consistent with the onset of physical injury 
and behavioral disturbance. Subsequently, Ellison et al. (2015) 
described development of a more systematic and (in some cases)

[[Page 5427]]

quantitative basis for a risk assessment approach to assess the 
biological significance and potential population consequences of 
predicted noise exposures. The approach for this final rule, which 
incorporates the revised acoustic exposure modeling results as an 
input, includes certain modifications to and departures from the 
conceptual approach described by Ellison et al. (2015). These are 
described in greater detail in the EWG report.
    Generally, this approach is a relativistic risk assessment that 
provides an interpretation of the exposure estimates within the context 
of key biological and population parameters (e.g., population size, 
life history factors, compensatory ability of the species, animal 
behavioral state, aversion), as well as other biological, 
environmental, and anthropogenic factors. This analysis as updated 
since BOEM revised the scope of its action was performed on a species-
specific basis within each modeling zone (Figure 3) for a high-effort 
scenario (represented by Year 1 of BOEM's revised effort projections) 
and a moderate-effort scenario (represented by Year 4 of BOEM's revised 
effort projections). (For most species, the maximum annual take occurs 
under the Year 1 scenario. The two exceptions are the bottlenose 
dolphin and Atlantic spotted dolphin, for which the maximum annual take 
occurs under the Year 4 scenario.) The end result provides an 
indication of the biological significance of these exposure numbers for 
each affected marine mammal stock (i.e., yielding the severity of 
impact and vulnerability of stock/population information), and 
forecasts the likelihood of any such impact. This result is expressed 
as relative impact ratings of overall risk that couple potential 
severity of effect on a stock and likely vulnerability of the 
population to the consequences of those effects, given biologically 
relevant information (e.g., compensatory ability).
    Spectral, temporal, and spatial overlaps between survey activities 
and animal distribution are the primary factors that drive the type, 
magnitude, and severity of potential effects on marine mammals, and 
these considerations are integrated into both the severity and 
vulnerability assessments. In discussion with BOEM and NMFS, the EWG 
developed a strategic approach to balance the weight of these 
considerations between the two assessments, specifying and clarifying 
where and how the interactions between potential disturbance and 
species within these dimensions are evaluated. Overall ratings are then 
considered in conjunction with the required mitigation (and any 
additional relevant contextual information) to ultimately inform our 
determinations. Elements of this approach are subjective and relative 
within the context of this program of projected actions and, overall, 
the analysis necessarily requires the application of professional 
judgment.

Severity of Effect

    Level A Harassment--In order to evaluate the potential severity of 
the expected potential takes by Level A harassment (accounting for 
aversion) (Table 9) on the species or stock, the EWG framework uses a 
potential biological removal (PBR)-equivalent metric. As described 
previously, PBR is defined by the MMPA as the maximum number of 
animals, not including natural mortalities, that may be removed from a 
marine mammal stock while allowing that stock to reach or maintain its 
optimum sustainable population. To be clear, NMFS does not expect any 
of the potential occurrences of injury (i.e., permanent threshold shift 
(PTS)) that may be authorized under this rule to result in mortality of 
marine mammals, nor should Level A harassment be considered a 
``removal'' in the context of PBR when used to inform a negligible 
impact determination. PTS is not appropriately considered equivalent to 
serious injury. However, PBR can serve as a gross indicator of the 
status of the species and a good surrogate for population 
vulnerability/health and, accordingly, PBR or a related metric can be 
used appropriately to inform a separate analysis to evaluate the 
potential relative severity to the population of a permanent impact 
such as PTS on a given number of individuals. This analysis is used to 
assess relative risks to populations as a result of PTS; NMFS does not 
expect that Level A harassment could directly result in mortality and 
our use of the PBR metric in this context should not be interpreted as 
such.
    However, exposure estimates generated using habitat-based density 
models (Roberts et al., 2016) cannot appropriately be directly related 
to the PBR values found in NMFS' SARs. Therefore, a modified PBR value 
was derived on the basis of the typical pattern for NMFS' PBR values, 
where the value varies between approximately 0.6-0.9 percent of the 
minimum population abundance depending upon population confidence 
limits (higher with increasing confidence). For endangered species, PBR 
values are typically \1/5\ of the values for non-endangered species due 
to assumption of a lower recovery factor--endangered species are 
typically assigned recovery factors of 0.1, while species of unknown 
status relative to the optimum sustainable population level (i.e., most 
species) are typically assigned factors of 0.5. This basic relationship 
of population size relative to PBR was used to define the following 
relative risk levels due to Level A harassment.
     Very high--Level A harassment greater than 1.5 or 0.3 
percent (the latter figure is used for endangered species) of zone-
specific estimated population abundance.
     High--0.75-1.5 or 0.15-0.3 percent of zone-specific 
population.
     Moderate--0.375-0.75 or 0.075-0.15 percent of zone-
specific population.
     Low--0.075-0.375 or 0.015-0.075 percent of zone-specific 
population.
     Very low--less than 0.075 or less than 0.015 percent of 
zone-specific population.
    Relative severity scores by zone (Figure 3) and species were 
determined for high and moderate annual effort scenarios. As described 
previously, we do not believe that Level A harassment is likely to 
actually occur for mid-frequency cetaceans and therefore do not predict 
(nor will we authorize) any take by Level A harassment for these 
species (i.e., most species in the GOM).
    Bryde's whales (a low-frequency cetacean species) are expected to 
be present primarily in Zones 1 and 4 (though may be present to a 
lesser extent in Zones 2 and 5). BOEM's update to the geographic scope 
of its action removed the entirety of Zone 1 and the majority of Zone 4 
from consideration in this rule. Altogether, no incidents of Level A 
harassment are predicted for Bryde's whales.
    Kogia spp. (high-frequency cetacean species) are primarily present 
in Zones 4-7. We assess the relative severity resulting from injury for 
Kogia spp. to be ``very high'' in Zones 5-7 under both evaluated 
activity scenarios. In Zone 4, relative severity is ``high'' under the 
moderate effort scenario, and no activity is projected in Zone 4 under 
the high effort scenario.
    In summary, we assess that there is no risk of Level A harassment 
for any mid-frequency cetacean species. Overall severity associated 
with take by Level A harassment is expected to be very high for Kogia 
spp. and very low for Bryde's whales, as no incidents of Level A 
harassment are predicted for the stock.
    We note that regardless of the relative risk assessed in this 
framework, because of the anticipated received levels and duration of 
sound exposure expected for any marine mammals exposed above Level A 
harassment criteria, no individuals of any species or stock are

[[Page 5428]]

expected to receive more than a relatively minor degree of PTS, which 
would not be expected to meaningfully increase the likelihood or 
severity of any potential population-level effects. See ``Loss of 
Hearing Sensitivity,'' below, for additional discussion.
    Level B Harassment--As described above in Estimated Take, a 
significant model assumption was that populations of animals were reset 
for each 24-hr period. Exposure estimates for the 24-hr period were 
then aggregated across all assumed survey days as completely 
independent events, assuming populations turn over completely within 
each large zone on a daily basis. In order to evaluate modeled daily 
exposures and determine more realistic exposure probabilities for 
individuals across multiple days, we used information on species-
typical movement behavior to determine a species-typical offset of 
modeled daily exposures, using the exploratory analysis discussed under 
Estimated Take (i.e., Test Scenario 1). In this test scenario, modeled 
results were compared for a 30-day period versus the aggregation of 24-
hr population reset intervals. When conducting computationally-
intensive modeling over the full assumed 30-day survey period (versus 
aggregating the smaller 24-hr periods for 30 days), results showed 
about 10-45 percent of the total number of takes calculated using a 24-
hr reset of the population, with differences relating to species-
typical movement and residency patterns. Given that many of the 
evaluated survey activities occur for 30-day or longer periods, 
particularly some of the larger surveys for which the majority of the 
modeled exposures occur, using such a scaling process is appropriate in 
order to evaluate the likely severity of the predicted exposures and to 
estimate take for the purposes of LOA applications and predicting the 
number of individual marine mammals taken during the course of a single 
survey (although, as noted previously, for surveys significantly longer 
than 30 days, the take numbers with this scaling applied would still be 
expected to overestimate the number of individuals, given the greater 
degree of repeat exposures that would be expected the longer the survey 
goes on). This output was used in a severity assessment. This approach 
is also discussed in more detail in the EWG report.
    Similar to the evaluation of severity for Level A harassment, the 
scaled Level B harassment takes were rated through a population-
dependent binning system. For each species, scaled takes were divided 
by the zone-specific predicted abundance, and these proportions were 
used to evaluate the relative severity of modeled exposures based on 
the distribution of values across species to evaluate risk associated 
with behavioral disruption across species--a simple, logical means of 
evaluating relative risk across species and areas. Relative risk 
ratings using percent of area population size were defined as follows:
     Very high--Adjusted Level B harassment takes greater than 
800 percent of zone-specific population;
     High--Adjusted Level B harassment takes 401-800 percent of 
zone-specific population;
     Moderate--Adjusted Level B harassment takes 201-400 
percent of zone-specific population;
     Low--Adjusted Level B harassment takes 100-200 percent of 
zone-specific population; and
     Very low--Adjusted Level B harassment takes less than 100 
percent of zone-specific population.

Vulnerability of Affected Population

    Vulnerability rating seeks to evaluate the relative risk of a 
predicted effect given species-typical and population-specific 
parameters (e.g., species-specific life history, population factors) 
and other relevant interacting factors (e.g., human or other 
environmental stressors). The assessment includes consideration of four 
categories within two overarching risk factors (species-specific 
biological and environmental risk factors). These values were selected 
to capture key aspects of the importance of spatial (geographic), 
spectral (frequency content of noise in relation to species-typical 
hearing and sound communications), and temporal relationships between 
sound and receivers. Explicit numerical criteria for identifying scores 
were specified where possible, but in some cases qualitative judgments 
based on a reasonable interpretation of given aspects of the proposed 
activity and how it relates to the species in question and the 
environment within the specified area were required. Factors considered 
in the vulnerability assessment were detailed in Southall et al. (2017) 
and are reproduced here (Table 11). Note that the effects of the 
Deepwater Horizon oil spill are accounted for through the non-noise 
chronic anthropogenic risk factor identified below, while the effects 
to acoustic habitat and on individual animal behavior via masking 
(summarized in Potential Effects of the Specified Activity on Marine 
Mammals and Their Habitat and described in detail in that section of 
the notice of proposed rulemaking) are accounted for through the 
masking and chronic anthropogenic noise risk factors. Species-specific 
vulnerability scoring according to this scheme is shown in Table 12. 
Zone-specific vulnerability ratings corresponding with the scores given 
in Table 12 below are provided in Tables 8-10 of the EWG report.

               Table 11--Vulnerability Assessment Factors
------------------------------------------------------------------------
                                                       Score
------------------------------------------------------------------------
Masking: Degree of spectral overlap
 between biologically important acoustic
 signals
and predominant noise source of proposed
 activity (max: 7 out of 30):
    Communication masking: Predominant    +3/+1
     noise energy directly/partially
     overlaps \1\ species-specific
     signals utilized for communication.
    Foraging masking: Predominant noise   +2/+1
     energy directly/partially
     overlaps\1\ species-specific
     signals utilized in foraging
     (including echolocation and other
     foraging coordination signals).
    Navigation/Orientation signal         +2/+1
     masking: Predominant noise energy
     directly/partially overlaps\1\
     signals likely utilized in spatial
     orientation to which species is
     well capable of hearing.
Species population: Stock status, trend,
 and size (max: 7 out of 30):
    Population status: Endangered (ESA)   +3/0
     and/or depleted (MMPA) (Y/N).
    Trend rating: Decreasing/unknown or   +2/+1/0/-1
     data deficient/stable (i.e., within
     5 percent)/increasing (last three
     SARs for which new population
     estimates were updated).
    Population size: Small (less than     +2
     2,500).
Species habitat use and compensatory
 abilities: Degree to which activity
 within a specified area \2\ overlaps
 with species habitat and distribution
 (max: 7 out of 30):
    Habitat use: Survey area contains     +4/+2/+1/0
     greater than 30/15-30/5-15/less
     than 5 percent of total region-wide
     estimated population (during
     defined survey period).

[[Page 5429]]

 
    Temporal sensitivity: Survey          Up to +3
     overlaps temporally with well-
     defined species-specific
     biologically-important period
     (e.g., calving).
Other (chronic) noise and non-noise
 stressors: Magnitude of other potential
 sources of disturbance or other
 stressors that may influence a species
 response to additional noise and
 disturbance of the proposed activity
 (max: 9 out of 30):
    Chronic anthropogenic noise: Species  +2/+1
     subject to high/moderate degree of
     current or known future
     (overlapping activity) chronic
     anthropogenic noise.
    Chronic anthropogenic risk factors    Up to +4/+2
     (non-noise): Species subject to
     high/moderate degree of current or
     known future risk from other
     chronic, non-noise anthropogenic
     activities (e.g., fisheries
     interactions, ship strike).
    Chronic biological risk factors (non- Up to +3
     noise): Known presence of disease,
     parasites, prey limitation, or high
     predation pressure.
------------------------------------------------------------------------
\1\ Direct or partial overlap means that the predominant spectral
  content of received noise exposure from activity specific sources is
  expected to occur at identical frequencies as signals of interest, or
  that secondary (lower-level) spectral content of received noise
  exposure from activity specific sources is expected to occur at
  identical frequencies as signals of interest.
\2\ This is the area over which an activity is evaluated and a local
  population is determined, in this case the seven modeling zones.


                                                                         Table 12--Vulnerability Assessment Scoring \1\
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                                                                                                        Total
               Species                  Communi-     Foraging    Navigation     Status       Trend         Size       Habitat        Time       Chronic      Chronic     Biological     score
                                         cation                                                                                                  noise        other         risk        range
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Bryde's whale.......................            3            2            2            3            2            2          0-4          0-1          1-2          0-3            0        16-23
Sperm whale.........................            1            1            2            3            2            2          0-4          0-1          1-2          0-3            0        14-18
Kogia spp...........................            0            0            1            0            2            2          0-4          0-1          1-2          0-3            0         8-13
Beaked whale........................            0            0            1            0            1            0          0-4          0-1          1-2          0-3            1         6-13
Rough-toothed dolphin...............            0            0            1            0            2            0          1-4          0-1          1-2          0-3            0         6-10
Bottlenose dolphin..................            1            0            1            0           -1            0          0-4          0-1          1-2          0-3            0         2-10
Clymene dolphin.....................            0            0            1            0            2            0          0-4          0-1          1-2          0-3            0         6-10
Atlantic spotted dolphin............            1            0            1            0            1            0          0-4          0-1          1-2          0-3            2         6-14
Pantropical spotted dolphin.........            0            0            1            0            2            0          0-4          0-1          1-2          0-3            0         6-10
Spinner dolphin.....................            0            0            1            0            0            0          0-4          0-1          1-2          0-3            0          3-9
Striped dolphin.....................            0            0            1            0            2            0          0-4          0-1          1-2          0-3            0         6-10
Fraser's dolphin....................            0            0            1            0            1            2          0-4          0-1          1-2          0-3            0         7-11
Risso's dolphin.....................            0            0            1            0           -1            0          0-4          0-1          1-2          0-3            1          4-9
Melon-headed whale..................            0            0            1            0            2            0          0-4          0-1          1-2          0-3            0         6-10
Pygmy killer whale..................            0            0            1            0            2            2          0-4          0-1          1-2          0-3            0         8-12
False killer whale..................            0            0            1            0           -1            0          0-4          0-1          1-2          0-3            0          3-7
Killer whale........................            1            0            1            0            2            2          0-4          0-1          1-2          0-3            0         9-12
Short-finned pilot whale............            1            0            1            0            0            2          0-4          0-1          1-2          0-3            1         7-13
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Factors with a single value presented are those that remain constant across zones; other factors vary based on zone and a range of values is presented.


                                      Table 13--Vulnerability Rating Scheme
----------------------------------------------------------------------------------------------------------------
                                                 Risk
                Total score                   probability                   Vulnerability rating
                                             (% of total)
----------------------------------------------------------------------------------------------------------------
24-30.....................................          80-100  Very high.
18-23.....................................           60-79  High.
12-17.....................................           40-59  Moderate.
6-11......................................           20-39  Low.
0-5.......................................            0-19  Very low.
----------------------------------------------------------------------------------------------------------------

Risk

    In the final step of the framework, severity and vulnerability 
ratings are integrated to provide relative impact ratings of overall 
risk. Severity and vulnerability assessments each produce a numerical 
rating (1-5) corresponding with the qualitative rating (i.e., very low, 
low, moderate, high, very high). A matrix is then used to integrate 
these two scores to provide an overall risk assessment. The matrix is 
shown in Table 2 of Southall et al. (2017).
    The likely severity of effect was assessed as the percentage of 
total population affected based on scaled modeled Level B harassment 
takes relative to zone population size. There is no risk due to the 
effects of survey activity when there is no survey activity in a given 
zone for a given effort scenario. However, a stock's inherent zone-
specific vulnerability score drives the risk rating in those zones 
(Zone 1 under any activity scenario and Zone 4 under the high effort 
scenario), and risk ratings for all zones are considered

[[Page 5430]]

together in generating scenario-specific GOM-wide risk ratings for each 
species. Also, zones predicted to contain abundance of less than 0.05 
percent of the GOM-wide population for a species were considered to 
have de minimis risk and are not included in derivation of the stock-
specific GOM-wide rating.
    Table 14 provides relative impact ratings by zone, and Table 15 
provides GOM-wide relative impact ratings, for overall risk associated 
with predicted takes by Level B harassment, for representative high and 
moderate effort scenarios.

                         Table 14--Overall Evaluated Risk by Zone and Activity Scenario
----------------------------------------------------------------------------------------------------------------
                                Zone 1 \1\    Zone 2      Zone 3    Zone 4 \1\    Zone 5      Zone 6     Zone 7
            Species            ---------------------------------------------------------------------------------
                                  H     M     H     M     H     M     H     M     H     M     H     M    H    M
----------------------------------------------------------------------------------------------------------------
Bryde's whale.................  L     L     L     L     n/a   n/a   L     L     L     L     n/a   n/a   n/a  n/a
Sperm whale...................  n/a   n/a   n/a   n/a   n/a   n/a   L     L     VH    VH    M     L     L    L
Kogia spp.....................  VL    VL    n/a   n/a   n/a   n/a   L     L     H     M     L     VL    VL   VL
Beaked whale..................  n/a   n/a   n/a   n/a   n/a   n/a   VL    VL    VH    VH    H     M     M    L
Rough-toothed dolphin.........  VL    VL    L     M     VL    VL    VL    VL    H     M     M     L     VL   VL
Bottlenose dolphin............  VL    VL    H     H     VL    VL    VL    VL    VL    VL    n/a   n/a   n/a  n/a
Clymene dolphin...............  n/a   n/a   n/a   n/a   n/a   n/a   VL    VL    H     M     M     L     VL   VL
Atlantic spotted dolphin......  L     L     H     VH    VL    VL    VL    VL    M     L     n/a   n/a   n/a  n/a
Pantropical spotted dolphin...  VL    VL    n/a   n/a   n/a   n/a   VL    VL    H     M     L     VL    L    L
Spinner dolphin...............  VL    VL    n/a   n/a   n/a   n/a   VL    VL    H     M     n/a   n/a   VL   VL
Striped dolphin...............  n/a   n/a   n/a   n/a   n/a   n/a   VL    VL    H     M     L     VL    L    VL
Fraser's dolphin..............  VL    VL    VL    VL    VL    VL    VL    VL    M     L     L     VL    VL   VL
Risso's dolphin...............  VL    VL    n/a   n/a   n/a   n/a   VL    VL    H     M     M     VL    VL   VL
Melon-headed whale............  VL    VL    n/a   n/a   n/a   n/a   VL    VL    H     M     M     VL    VL   VL
Pygmy killer whale............  VL    VL    n/a   n/a   n/a   n/a   VL    VL    M     L     L     VL    VL   VL
False killer whale............  VL    VL    VL    VL    VL    VL    VL    VL    H     M     L     VL    VL   VL
Killer whale..................  VL    VL    VL    VL    VL    VL    VL    VL    L     L     VL    VL    L    L
Short-finned pilot whale......  n/a   n/a   n/a   n/a   n/a   n/a   VL    VL    M     M     M     VL    VL   VL
----------------------------------------------------------------------------------------------------------------
H = Year 1 (representative high effort scenario); M = Year 4 (representative moderate effort scenario).
n/a = less than 0.05% of GOM-wide population predicted in zone.
VL = very low; L = low; M = moderate; H = high; VH = very high.
\1\ No activity would occur in Zone 1, and no activity is projected in Zone 4 under the high effort scenario.


     Table 15--Overall Evaluated Risk by Activity Scenario, GOM-Wide
------------------------------------------------------------------------
                                                        Moderate effort
             Species                   High effort      scenario  (Year
                                   scenario  (Year 1)          4)
------------------------------------------------------------------------
Bryde's whale....................  Low...............  Low.
Sperm whale......................  Moderate \1\......  Low.
Kogia spp........................  Low...............  Very low.
Beaked whales....................  High \1\..........  Moderate.\1\
Rough-toothed dolphin............  Very low..........  Very low.
Bottlenose dolphin (shelf/         Very low..........  Very low.
 coastal).
Bottlenose dolphin (oceanic).....  Very low..........  Very low.
Clymene dolphin..................  Low...............  Low.\1\
Atlantic spotted dolphin.........  Low...............  Low.
Pantropical spotted dolphin......  Low...............  Very low.
Spinner dolphin..................  Very low..........  Very low.
Striped dolphin..................  Low...............  Very low.
Fraser's dolphin.................  Very low..........  Very low.
Risso's dolphin..................  Very low..........  Very low.
Melon-headed whale...............  Very low..........  Very low.
Pygmy killer whale...............  Very low..........  Very low.
False killer whale...............  Very low..........  Very low.
Killer whale.....................  Very low..........  Very low.
Short-finned pilot whale.........  Low...............  Low.\1\
------------------------------------------------------------------------
\1\ For these ratings, the median value across zones for the scenario
  fell between two ratings, and the higher rating is presented.

    In order to characterize the relative risk for each species across 
their entire range in the GOM, the EWG analysis used the median of the 
seven zone-specific risk ratings for each activity scenario (high and 
moderate effort), not counting those in which less than 0.05 percent of 
the GOM-wide abundance occurred, to describe a GOM-wide risk rating for 
each of the representative activity scenarios (Table 15).
    Overall, the results of the risk assessment show that (as 
expected), risk is highly correlated with effort and density. Areas 
where little or no survey activity is predicted to occur or areas 
within which few or no animals of a particular species are believed to 
occur have very low or no potential risk of negatively affecting marine 
mammals, as seen across activity scenarios in Zones 1, 3, and 4. Areas 
with consistently high levels of effort (Zones 2, 5, 6, and 7) are 
generally predicted to have higher overall evaluated risk across all 
species. However, fewer species of animals are expected to be present 
in Zone 2, where we primarily expect shelf species such as bottlenose 
and Atlantic spotted dolphins. In Zone 7, animals are expected to be 
subject to less other chronic noise and non-noise stressors, which is 
reflected in the vulnerability scoring for that zone. Therefore, 
despite consistently high levels of projected effort, overall rankings 
for that zone are lower than for Zones 5 and 6.
    Zones 2 and 5 were the only zones with ``very high'' levels of risk 
due to behavioral disturbance, identified for two species of particular 
concern in Zone 5 (beaked and sperm whales) and for Atlantic spotted 
dolphins in Zone 2 (moderate effort scenario only). As particularly 
sensitive species, beaked whales and sperm whales consistently receive 
relatively high severity scores. For sperm whales, this sensitivity is 
manifest through typically higher vulnerability scoring, whereas the 
assumed sensitivity of beaked whales to

[[Page 5431]]

noise exposure is expressed through the application of behavioral 
harassment criteria (Table 6) and, therefore, relatively high assumed 
take numbers. Bottlenose dolphins and Atlantic spotted dolphin are 
generally the only species expected to commonly occur in relatively 
shallow waters of the continental shelf (Zones 1-3) and relatively high 
risk is assessed for these species in Zone 2, across activity 
scenarios. Relatively moderate levels of risk were also identified for 
other species in some contexts, and these are generally explained by 
the interaction of specific factors related to survey effort 
concentration and areas of heightened geographic distribution or 
specific factors related to population trends or zone-related 
differences in vulnerability. Overall, following BOEM's update to the 
geographic scope of activity (with the entirety of Zone 1, most of Zone 
4, and one-third of Zone 7 removed from consideration here; see Table 
1) the greatest relative risk across species is generally seen in Zones 
5 and 6.
    When considered across both representative activity scenarios 
(Table 15), only beaked whales are considered to have relatively high 
risk (under the high effort scenario only). Relatively moderate risk is 
assessed for beaked whales under the moderate effort scenario. 
Relatively moderate risk is also assessed for sperm whales under the 
high effort scenario. The rest of the species have no more than low to 
very low risk under either scenario. Shelf/coastal and oceanic 
bottlenose dolphin stocks, rough-toothed dolphins, spinner dolphins, 
Fraser's dolphins, Risso's dolphins, melon-headed whales, pygmy killer 
whales, false killer whales, and killer whales are assessed as having 
no greater than very low relative risk under any scenario.
    Although the scores generated by the EWG framework and further 
aggregated across zones (as described above) are species-specific, 
additional stock-specific information can be gleaned through the zone-
specific nature of the analysis. For example, with some bottlenose 
dolphin stocks, the zones align with stock range edges. The oceanic 
stock of bottlenose dolphins occurs within Zones 4-7, while coastal and 
shelf stocks occur within Zones 1-3 (sufficient information is not 
available to attribute takes on a stock-specific basis in Zones 1-3). 
These species-specific risk ratings are broadly applied in NMFS' 
negligible impact analysis to all of the multiple stocks that are 
analyzed in this rule (Table 4). However, NMFS is also considering 
additional stock-specific information in our analysis, where 
appropriate, as indicated in our Description of Marine Mammals in the 
Area of the Specified Activity, Potential Effects of the Specified 
Activity on Marine Mammals and Their Habitat, and Mitigation sections 
(e.g., coastal bottlenose dolphins were heavily impacted by the DWH oil 
spill, and we have therefore required a time/area restriction to reduce 
impacts).

Duration of Level B Harassment Exposures

    In order to more fully place the predicted amount of take into 
meaningful context, it is useful to understand the duration of exposure 
at or above a given level of received sound, as well as the likely 
number of repeated exposures across days. While a momentary exposure 
above the criteria for Level B harassment counts as an instance of 
take, that accounting does not make any distinction between fleeting 
exposures and more severe encounters in which an animal may be exposed 
to that received level of sound for a longer period of time. Yet this 
information is meaningful to an understanding of the likely severity of 
the exposure, which is relevant to the negligible impact evaluation and 
not directly incorporated into the risk assessment framework described 
above. For example, for bottlenose dolphins exposed to noise from 3D 
WAZ surveys in Zone 6, the modeling report shows that approximately 72 
takes (Level B harassment) would be expected to occur in a 24-hr 
period. However, each animat modeled has a record or time history of 
received levels of sound over the course of the modeled 24-hr period. 
The 50th percentile of the cumulative distribution function indicates 
that the time spent exposed to levels of sound above 160 dB rms SPL 
(i.e., the 50 percent midpoint for Level B harassment) would be only 
1.8 minutes--a minimal amount of exposure carrying little potential for 
significant disruption of behavioral activity. We provide summary 
information regarding the total average time in a 24-hr period that an 
animal would spend with received levels above 160 dB and between 140 
and 160 dB in Table 16.
    Additionally, as we discussed in the Estimated Take section of the 
notice of proposed rulemaking for Test Scenario 1 (and summarized 
above), by comparing exposure estimates generated by multiplying 24-hr 
exposure estimates by the total number of survey days versus modeling 
for a full 30-day survey duration for six representative species, we 
were able to refine the exposure estimates to better reflect the number 
of individuals exposed above threshold within a single survey. Using 
this same comparison and scalar ratios described above, we are able to 
predict an average number of days each of the representative species 
modeled in the test scenario were exposed above the Level B harassment 
thresholds within a single survey. As with the duration of exposures 
discussed above, the number of repeated exposures is important to an 
understanding of the severity of effects. For example, the ratio for 
beaked whales indicates that the 30-day modeling showed that 
approximately 10 percent as many individual beaked whales (compared to 
the results produced by multiplying average 24-hr exposure results by 
the 30-day survey duration) could be expected to be exposed above 
harassment thresholds. However, the approach of scaling up the 24-hour 
exposure estimates appropriately reflects the instances of exposure 
above threshold (which cannot be more than 1 in 24 hours), so the 
inverse of the scalar ratio suggests the average number of days in the 
30-day modeling period that beaked whales are exposed above threshold 
is approximately ten. It is important to remember that this is an 
average and that it is likely some individuals would be exposed on 
fewer days and some on more. Table 16 reflects the average days exposed 
above threshold for the indicated species having applied the scalar 
ratios described previously.

     Table 16--Time in Minutes (Per Day) Spent Above Thresholds (50th Percentile) and Average Number of Days Individuals Taken During 30-Day Survey
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                           Survey type and time (min/day) above    Survey type and time (min/day) above       Average
                                                                   160 dB rms (50% take)                   140 dB rms (10% take)             number of
                         Species                         -------------------------------------------------------------------------------- days ``taken''
                                                                                                                                           during 30-day
                                                             2D      3D NAZ    3D WAZ     Coil       2D      3D NAZ    3D WAZ     Coil         survey
--------------------------------------------------------------------------------------------------------------------------------------------------------
Bryde's whale...........................................       7.6      18.2       6.8      21.4      61.7     163.5      55.4     401.1             5.3
Sperm whale.............................................       5.2      10.3       4.0      20.7      12.0      31.8      10.7      25.2             2.4

[[Page 5432]]

 
Kogia spp...............................................       3.2       7.9       2.8      15.3       7.6      19.0       6.7      13.9             3.1
Beaked whale \1\........................................       6.0      12.4       4.4      24.0      16.2      39.7      14.1      31.1             9.9
Rough-toothed dolphin...................................       3.0       6.3       2.5      11.4      11.2      27.6      10.2      20.9             3.5
Bottlenose dolphin......................................       4.5      11.7       4.0      16.8      22.0      54.6      19.7      53.2             3.5
Clymene dolphin.........................................       1.8       3.9       1.6       8.7       8.0      21.1       7.2      20.4             3.5
Atlantic spotted dolphin................................       7.0      16.0       6.5      25.7      23.4      58.1      20.9      49.3             3.5
Pantropical spotted dolphin.............................       1.8       4.1       1.6       8.7       8.1      21.0       7.1      22.2             3.5
Spinner dolphin.........................................       3.2       8.5       2.7      16.4      12.4      31.0      10.8      22.8             3.5
Striped dolphin.........................................       1.8       4.0       1.6       8.5       8.0      21.0       7.2      21.3             3.5
Fraser's dolphin........................................       2.8       6.4       2.4      13.8       9.4      24.2       8.4      24.0             3.5
Risso's dolphin.........................................       3.4       8.4       2.9      15.3      13.8      37.7      12.2      31.5             3.5
Melon-headed whale......................................       2.6       5.9       2.2      13.1       9.3      24.2       8.3      24.0             3.4
Pygmy killer whale......................................       1.8       3.6       1.4       7.1       7.3      18.5       6.6      17.3             3.4
False killer whale......................................       2.4       4.9       1.9       9.3       8.8      22.0       8.0      17.8             3.4
Killer whale............................................       2.7       6.1       3.3      12.0      16.8      46.1      14.9      73.6             3.4
Short-finned pilot whale................................       3.3       8.1       2.9      17.5      10.9      27.4       9.8      20.8             3.4
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Beaked whales are evaluated according to a different scale where 90% of the population exposed above 140 dB rms is considered taken and 50% of the
  population exposed above 120 dB rms is considered taken.

Loss of Hearing Sensitivity

    In general, NMFS expects that noise-induced hearing loss, whether 
temporary (temporary threshold shift, equivalent to Level B harassment) 
or permanent (PTS, the only form of Level A harassment that may result 
from this action), is only possible as a result of airgun survey 
activity for low-frequency and high-frequency cetaceans. The best 
available scientific information indicates that low-frequency cetacean 
species (i.e., mysticete whales, including the Bryde's whale) have 
heightened sensitivity to frequencies in the range output by airguns, 
as shown by their auditory weighting function, whereas high-frequency 
cetacean species (including Kogia spp.) have heightened sensitivity to 
noise in general (as shown by their lower threshold for the onset of 
PTS) (NMFS, 2018). However, no instances of Level A harassment are 
predicted to occur for Bryde's whales, and Level A harassment of 
Bryde's whales will not be authorized under this rule.
    Level A harassment is predicted to occur for Kogia spp. (as 
indicated in Table 9 and evaluated in the ``Level A harassment'' 
subsection above). However, the degree of injury (hearing impairment) 
is expected to be mild. If permanent hearing impairment occurs, it is 
most likely that the affected animal would lose a few dB in its hearing 
sensitivity, which in most cases would not be expected to affect its 
ability to survive and reproduce. Hearing impairment that occurs for 
these individual animals would be limited to at or slightly above the 
dominant frequency of the noise sources. In particular, the predicted 
PTS resulting from airgun exposure is not likely to affect their 
echolocation performance or communication, as Kogia spp. likely produce 
acoustic signals at frequencies above 100 kHz (Merkens et al., 2018), 
well above the frequency range of airgun noise. Further, modeled 
exceedance of Level A harassment criteria typically resulted from being 
near an individual source once, rather than accumulating energy from 
multiple sources. Overall, the modeling indicated that exceeding the 
SEL threshold is a rare event, and having four vessels close to each 
other (350 m between tracks) did not cause appreciable accumulation of 
energy at the ranges relevant for injury exposures. Accumulation of 
energy from independent surveys is expected to be negligible. This is 
relevant for Kogia spp. because based on their expected sensitivity, we 
expect that aversion may play a stronger role in avoiding exposures 
above the peak pressure PTS threshold than we have accounted for.
    For both Bryde's whales and Kogia spp., some subset of the 
individual marine mammals predicted to be taken by Level B harassment 
may incur some TTS in addition to being behaviorally harassed. For 
Bryde's whales, TTS is more likely to occur at frequencies important 
for communication. However, any TTS incurred would be expected to be of 
a relatively small degree and short duration. This is due to the low 
likelihood of sound source approaches of the proximity or duration 
necessary to cause more severe TTS, given the fact that both sound 
source and marine mammals are continuously moving, the anticipated 
effectiveness of shutdowns, and general avoidance by marine mammals of 
louder sources.
    For these reasons, and in conjunction with the required mitigation, 
NMFS does not believe that Level A harassment (here, PTS) or Level B 
harassment in the form of TTS will play a meaningful role in the 
overall degree of impact experienced by marine mammal populations as a 
result of the projected survey activity. Further, the impacts of any 
TTS incurred are addressed along with behavioral disruption through the 
broader analysis of Level B harassment.

Impacts to Habitat

    Potential impacts to marine mammal habitat, including to marine 
mammal prey, were discussed in detail in the notice of proposed 
rulemaking and summarized herein (see Potential Effects of the 
Specified Activities on Marine Mammals and Their Habitat as well as 
responses to comments concerning these issues).
    Regarding impacts to prey species such as fish and invertebrates, 
NMFS' review of the available information leads to a conclusion that 
the most likely impact of survey activity would be temporary avoidance 
of an area, with a rapid return to pre-survey distribution and 
behavior, and minimal impacts to recruitment or survival anticipated. 
Therefore, the specified activities are not likely to have more than 
short-term adverse effects on any prey habitat or populations of prey 
species. Further, any impacts to prey species are not expected to 
result in significant or long-term consequences for individual marine 
mammals, or to contribute to adverse impacts on their populations.
    Regarding potential impacts to acoustic habitat, NMFS previously 
summarized a detailed analysis of

[[Page 5433]]

potential cumulative and chronic effects to marine mammals (found in 
the CCE report available online at www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico). That analysis focused on potential effects to 
sperm whales (which also provides a conservative proxy regarding 
potential effects to other mid- and high-frequency cetacean species) 
and to Bryde's whales. Regarding sperm whales, the analysis shows that 
the survey activities do not significantly contribute to the soundscape 
in the frequency band relevant for their lower-frequency slow-clicks, 
and that there will be no significant change in communication space for 
sperm whales. Similar conclusions may be assumed for other mid- and 
high-frequency cetacean species.
    Implications for acoustic masking and reduced communication space 
resulting from noise produced by airgun surveys in the GOM are expected 
to be particularly heightened for animals that actively produce low-
frequency sounds or whose hearing is attuned to lower frequencies 
(i.e., Bryde's whales). The strength of the communication space 
approach used here is that it evaluates potential contractions in the 
availability of a signal of documented importance to a population of 
animals of key management interest in the region. In this case, losses 
of communication space for Bryde's whales were estimated to be higher 
in eastern and central GOM canyons and shelf break areas. In contrast, 
relative maintenance of listening area and communication space was seen 
within the Bryde's whale core habitat area in the eastern GOM. The 
result was heavily influenced by the projected lack of survey activity 
in that region, which underscores the importance of maintaining this 
important habitat for the Bryde's whale. Following BOEM's update to the 
scope of activity considered herein, no survey activity will occur 
under this rule within Bryde's whale core habitat, or within the 
broader eastern GOM. In areas where larger amounts of survey activity 
were projected, significant loss of low-frequency listening area and 
communication space for Bryde's whale calls was estimated. However, 
these are areas where Bryde's whales are unlikely to occur (i.e., 
deeper waters of the central and western GOM).

Species and Stock-Specific Negligible Impact Analysis Summaries

    In this section, we consider the relative impact ratings described 
above in conjunction with the required mitigation and other relevant 
contextual information in order to produce a final assessment of impact 
to the stock or species, i.e., the negligible impact determinations. 
The effects of the DWH oil spill are accounted for through the 
vulnerability scoring (Table 12). NMFS developed mitigation 
requirements for consideration in the proposed rule, including time-
area restrictions, designed specifically to provide benefit to certain 
populations for which a relatively high amount of risk is predicted in 
relation to exposure to survey noise. The required time-area 
restrictions, described in detail in Proposed Mitigation in the notice 
of proposed rulemaking and depicted in Figure 4, were designed 
specifically to provide benefit to the bottlenose dolphin, Bryde's 
whale, and beaked and sperm whales, with additional benefits to Kogia 
spp., which are often found in higher densities in the same locations 
of greater abundance for beaked and sperm whales. Two of the three 
time-area restrictions in the proposed rule--the Bryde's whale core 
habitat area and the Dry Tortugas area (Areas #2 and 3; Figure 4)--are 
eliminated from consideration as a result of BOEM's update to the 
geographic scope of action, as these two areas are entirely within the 
portion of the GOM removed from consideration. The bottlenose dolphin 
area, as revised herein (see Mitigation), is included in this final 
rule.
    Although the Bryde's whale core habitat and Dry Tortugas areas are 
not the subject of restrictions on survey activity, as the updated 
scope of activity considered here does not include those two areas, the 
beneficial effect for animals in those areas, and the stocks of which 
they are a part, remains the same. No survey activity in those areas 
can be considered for LOAs issued under this rule. In addition, we 
expect the lack of survey activity in those areas to provide some 
subsidiary benefit to additional species that may be present, as 
indicated in the sections below and reflected in the updated take 
estimates.
    The absence of survey activity in those two areas benefits both the 
primary species for which they were designed and species that may 
benefit secondarily by likely reducing the portion of a stock likely 
exposed to survey noise and avoiding impacts to certain species in 
areas of importance for them. These areas are discussed more 
specifically in the context of the species and stocks they were 
designed to protect in the Proposed Mitigation section of the notice of 
proposed rulemaking, and are summarized in the sections below.

Bryde's Whale

    First, we note that the estimated (and allowable) take of Bryde's 
whales has been reduced as compared to the proposed rule as a result of 
the change in scope. Specifically, both the maximum annual take and the 
average annual take decreased by approximately 98 percent. The EWG 
analysis, which evaluated the relative significance of the aggregated 
impacts of the survey activities across seven GOM zones in the context 
of the vulnerability of each species, concluded that the GOM-wide risk 
ratings for Bryde's whales are low, regardless of activity scenario. We 
note that, although the evaluated severity of take for Bryde's whales 
is very low in all zones where take could occur, vulnerability for the 
species is assessed as high in all zones where the species occurs. When 
integrated through the risk framework as described above, overall risk 
for the species is therefore assessed as low for both the high and 
moderate effort scenarios. Evaluated risk is lower than what was 
considered in the proposed rule, where analysis of the prior take 
estimates resulted in a risk rating of moderate for both scenarios.
    We further consider the likely severity of any predicted behavioral 
disruption of Bryde's whales in the context of the likely duration of 
exposure above Level B harassment thresholds. Specifically, the average 
modeled time per day spent at received levels above 160 dB rms (where 
50 percent of the exposed population is considered taken) ranges from 
6.8-21.4 minutes for deep penetration survey types. The average time 
spent exposed to received levels between 140 and 160 dB rms (where 10 
percent of the exposed population is considered taken) ranges from 55-
164 minutes for 2D, 3D NAZ, and 3D WAZ surveys, and 401 minutes for 
coil surveys (which comprise approximately 10 percent of the total 
activity days).
    Importantly, no survey activity will occur within the Bryde's whale 
core habitat area pursuant to this rule. The absence of survey activity 
in the area is expected to benefit Bryde's whales and their habitat by 
minimizing a range of potential effects of airgun noise, both acute and 
chronic, that could otherwise accrue to impact the reproduction or 
survival of individuals in this area. Absence of survey activity in 
this area will minimize disturbance of the species in the place most 
important to them for critical behaviors such as foraging and 
socialization. Based on Roberts et al. (2016), the area encompasses 
approximately 92 percent of the predicted abundance of Bryde's whales 
in the GOM. Intensive survey effort in

[[Page 5434]]

the region has not resulted in any confirmed Bryde's whale sightings 
outside this core habitat area (aside from a single anomalous sighting 
in the western GOM). Although some sound from airguns may still 
propagate into the Bryde's whale core habitat area from surveys that 
may occur outside of the area (in certain locations where separation 
distance between the core habitat area and the area considered for 
survey activity through this rule is less; see Figure 2), exposure of 
Bryde's whales to sound levels that may be expected to result in Level 
B harassment will be eliminated or reduced for animals within the 
Bryde's whale core area. The absence of survey activity in this area 
and significant reduction in associated exposure of Bryde's whales to 
seismic airgun noise is expected to eliminate the likelihood of 
auditory injury of Bryde's whales. Finally, the absence of survey 
activity in the eastern GOM will reduce chronic exposure of Bryde's 
whales to higher levels of anthropogenic sound and the associated 
effects including masking, disruption of acoustic habitat, long-term 
changes in behavior such as vocalization, and stress.
    As described in the preceding ``Loss of Hearing Sensitivity'' 
section, we have analyzed the likely impacts of potential temporary 
hearing impairment and do not expect that they would result in impacts 
on reproduction or survival of any individuals. The extended shutdown 
zone for Bryde's whales (1,500 m)--to be implemented in the unlikely 
event that a Bryde's whale is encountered outside of the core habitat 
area--is expected to further minimize the severity of any hearing 
impairment incurred as well as reducing the likelihood of more severe 
behavioral responses. Similarly, application of this extended distance 
shutdown requirement when calves are present will minimize the 
potential for and degree of disturbance during this sensitive life 
stage.
    No mortality of Bryde's whales is anticipated or authorized. It is 
possible that Bryde's whale individuals in this stock, if encountered 
in areas not typically considered to be Bryde's whale habitat, will be 
impacted briefly on one or more days during a year of activity by one 
type of survey or another and some subset of those exposures above 
thresholds may be of comparatively long duration within a day. However, 
the significant and critical protection afforded through the absence of 
survey activity in the core habitat area and the associated reduction 
in estimated take ensures that the impacts of the expected takes from 
these activities are not likely to adversely affect the GOM stock of 
Bryde's whales through impacts on annual rates of recruitment or 
survival.

Sperm Whale

    First, we note that the estimated (and allowable) take of sperm 
whales has been reduced as compared to the proposed rule as a result of 
the change in scope. Specifically, the maximum annual take decreased by 
approximately 62 percent and the average annual take decreased by 
approximately 58 percent. The EWG analysis, which evaluated the 
relative significance of the aggregated impacts of the survey 
activities across seven GOM zones in the context of the vulnerability 
of each species, concluded that the GOM-wide risk ratings for sperm 
whales were between moderate and low (equivalent to a 2.5 on a 5-point 
scale, with a 3 equating to ``moderate'') (for the high effort 
scenario) or low (for the moderate effort scenario). Evaluated risk is 
reduced from the proposed rule, where the high effort scenario resulted 
in a very high risk rating and the moderate effort scenario resulted in 
a high risk rating. We further consider the likely severity of any 
predicted behavioral disruption of sperm whales in the context of the 
likely duration of exposure above Level B harassment thresholds. 
Specifically, the average modeled time per day spent at received levels 
above 160 dB rms (where 50 percent of the exposed population is 
considered taken) ranges from 4-10.3 minutes for 2D, 3D NAZ, and 3D WAZ 
surveys and up to 20.7 minutes for coil surveys (which comprise less 
than 10 percent of the total projected activity days) and the average 
time spent between 140 and 160 dB rms (where 10 percent of the exposed 
population is considered taken) is 12-31.8 minutes.
    Odontocetes echolocate to find prey, and while there are many 
different strategies for hunting, one common pattern, especially for 
deeper-diving species, is to conduct multiple repeated deep dives 
within a feeding bout, and multiple bouts within a day, to find and 
catch prey. While exposures of the short durations noted above could 
potentially interrupt a dive or cause an individual to relocate to 
feed, such a short-duration interruption would typically be unlikely to 
have significant impacts on an individual's energy budget. However, the 
moderate risk rating for the high effort scenario reflects the higher 
number of total days across which these singularly more minor impacts 
may occur, as well as other factors, and points to the need for the 
consideration of additional reduction of impacts where possible. In 
years when less effort occurs, as represented by the moderate effort 
scenario, risk will be less.
    Importantly, no survey activity is expected within the Dry Tortugas 
Mitigation Area, which was analyzed and proposed for implementation in 
the proposed rule. The area provides preferred habitat for 
comparatively high densities of sperm whales and is thought to be used 
as a calving area. The absence of survey activity in the area is 
expected to alleviate some of the previous impacts of concern to sperm 
whales (as well as beaked whales and Kogia spp.) and their habitat by 
minimizing a range of potential effects of airgun noise, both acute and 
chronic, that could otherwise accrue to impact the reproduction or 
survival of individuals in this area. Absence of survey activity in 
this area will minimize disturbance of the species in a place of 
importance for critical behaviors such as foraging and socialization 
and, overall, helps to reduce evaluated risk to the stock as a whole.
    Additionally, we note that the extended distance shutdown zone for 
sperm whales (1,500 m) is expected to further reduce the likelihood and 
minimize the severity of more severe behavioral responses. Similarly, 
application of this extended distance shutdown requirement when calves 
are present will minimize the potential for and degree of disturbance 
during this sensitive life stage.
    No mortality or Level A harassment of sperm whales is anticipated 
or authorized. While it is likely that the majority of the individual 
sperm whales will be impacted briefly on one or more days during a year 
of activity by one type of survey or another, based on the nature of 
the individual exposures (shorter duration) and takes, as well as the 
aggregated scale of the impacts across the GOM in consideration of the 
mitigation discussed here, the impacts of the expected takes from these 
activities are not likely to adversely affect the GOM stock of sperm 
whales through adverse impacts on annual rates of recruitment or 
survival.

Beaked Whales

    In consideration of the similarities in the nature and scale of 
impacts, we consider the GOM stocks of Cuvier's, Gervais', and 
Blainville's beaked whales together in this section. First, we note 
that the estimated (and allowable) take of beaked whales has been 
reduced as compared to the proposed rule as a result of the change in 
scope. Specifically, the maximum annual take decreased by approximately 
19 percent and the average annual take decreased

[[Page 5435]]

by approximately 15 percent. The EWG analysis, which evaluated the 
relative significance of the aggregated impacts of the survey 
activities across seven GOM zones in the context of the vulnerability 
of each species, concluded that the GOM-wide risk ratings for beaked 
whales were between high and moderate (equivalent to a 3.5 on a 5-point 
scale, with a 4 equating to ``high'') for the high effort scenario and 
between moderate and low (equivalent to a 2.5 on a 5-point scale, with 
a 3 equating to ``moderate'') for the moderate effort scenario. 
Evaluated risk is reduced from the proposed rule, where the high effort 
scenario resulted in a very high risk rating and the moderate effort 
scenario resulted in a high risk rating. We further consider the likely 
severity of any predicted behavioral disruption of beaked whales in the 
context of the likely duration of exposure above Level B harassment 
thresholds. Beaked whales are considered more behaviorally sensitive to 
sound than most other species, and therefore we utilize different 
thresholds to predict behavioral disturbance. However, this means that 
beaked whales are evaluated as ``taken'' upon exposure to received 
sound levels as low as 120 dB (where 50 percent of the exposed beaked 
whale population is considered taken). These received levels are 
typically reached at extreme distance from the acoustic source (i.e., 
greater than 50 km from the source). Behavioral responses to noise are 
significantly correlated with distance from the source (e.g., Gomez et 
al., 2016); and potential responses to these relatively low received 
levels at such great distances, while conservatively evaluated here as 
take under the MMPA, are unlikely to result in any response of such a 
severity as to carry any cost to the animal. (Additionally, in certain 
circumstances, noise from the surveys at these distances may be 
indistinguishable from other low-frequency background noise). 
Therefore, as for other species, we consider only the average modeled 
time per day spent at received levels above 140 dB rms (where 90 
percent of the exposed beaked whale populations are considered taken) 
and 160 dB rms (where, potentially, all exposed beaked whales are 
taken). The average time spent in a state of exposure above 160 dB rms 
is only 6-12.4 minutes for 2D, 3D NAZ, and 3D WAZ surveys and 24 
minutes for coil surveys. The average time spent in a state of exposure 
above 140 dB rms is 14.1 minutes for 3D WAZ surveys, 16.2 minutes for 
2D surveys, 31.1 minutes for coil surveys, and 39.7 minutes for 3D NAZ 
surveys.
    Odontocetes echolocate to find prey, and while there are many 
different strategies for hunting, one common pattern, especially for 
deeper-diving species, is to conduct multiple repeated deep dives 
within a feeding bout, and multiple bouts within a day, to find and 
catch prey. As we noted, while some of the exposures of the durations 
noted above could interrupt a dive or cause an individual to relocate 
to feed because of the lower thresholds combined with the way exposures 
are distributed across received levels, a higher proportion of the 
total takes (as compared to other taxa) are at the lower end of the 
received levels at which take would be expected to occur and at great 
distance from the acoustic source, where responses (if any) should be 
assumed to be minor. All else being equal, exposures to lower received 
levels and, separately, at greater distances might be expected to 
result in less severe responses, even given longer durations (e.g., 
DeRuiter et al., 2013). Considered individually or infrequently, these 
sorts of feeding interruptions would be unlikely to have significant 
impacts on an individual's energy budget, especially given the likely 
availability of adequate alternate feeding areas relatively nearby. 
However, the high risk rating for the high effort scenario reflects the 
higher number of total days across which these singularly more minor 
impacts may occur, as well as other factors, and points to the need for 
the consideration of additional reduction of impacts where possible. In 
years when less effort occurs, as represented by the moderate effort 
scenario, risk will be less.
    Importantly, no survey activity is expected within the Dry Tortugas 
Mitigation Area, which was analyzed and proposed for implementation in 
the proposed rule. The area provides preferred habitat for 
comparatively high densities of beaked whales. The absence of survey 
activity in this important area is expected to alleviate some of the 
previous impacts of concern to beaked whales (as well as sperm whales 
and Kogia spp.) and their habitat by minimizing a range of potential 
effects of airgun noise, both acute and chronic, that could otherwise 
accrue to impact the reproduction or survival of individuals in this 
area. Absence of survey activity in this area will minimize disturbance 
of the species in a place of importance for critical behaviors such as 
foraging and socialization and, overall, helps to reduce evaluated risk 
to the stocks as a whole.
    Additionally, we note that the extended distance shutdown zone for 
beaked whales (1,500 m) is expected to further reduce the likelihood 
of, and minimize the severity of, more severe behavioral responses.
    Despite the nature and duration of the exposures anticipated, which 
at a smaller scale might not be expected to meaningfully impact 
individual fitness, given the high to moderate EWG risk rating and the 
relatively high number of predicted beaked whale takes (increasing the 
likelihood of some subset of individuals accruing a fair number of 
repeated takes over sequential days--albeit assuming takes at low 
received levels and at distances from the source where responses, if 
any, should be expected to be minor), it is more likely that a small 
number of individuals could be interrupted during foraging in a manner 
and amount such that impacts to the energy budgets of females (from 
either losing feeding opportunities or expending energy to find 
alternative feeding options) could cause them to forego reproduction 
for a year. Energetic impacts to males are generally meaningless to 
population rates unless they cause death, and extreme energy deficits 
(beyond what could be considered reasonably likely to result from these 
activities) are required to cause the death of an adult marine mammal. 
As noted previously, however, foregone reproduction (especially for one 
year, which is the maximum predicted because the relatively small 
number anticipated in any one year makes the probability that any 
individual would be impacted in this way twice in five years very low) 
has far less of an impact on population rates than mortality. And a 
small number of instances of foregone reproduction would not be 
expected to adversely affect these stocks through effects on annual 
rates of recruitment or survival.
    It is worth noting that in similar situations, i.e., where 
individual beaked whales may be exposed to noise above harassment 
thresholds regularly, populations appear to be stable based on multiple 
studies and lines of evidence (e.g., Falcone and Schorr, 2014; DiMarzio 
et al., 2018). In research done at the Navy's fixed tracking range in 
the Bahamas, animals were observed to leave the immediate area of an 
anti-submarine warfare training exercise but return within a few days 
after the event ended (Claridge and Durban, 2009; McCarthy et al., 
2011; Moretti et al., 2009, 2010; Tyack et al., 2010, 2011). It is 
important to note that in these contexts, beaked whales were exposed to 
noise stimuli to which they are significantly more acoustically 
sensitive (i.e., mid-frequency active sonar versus low-frequency airgun 
noise).

[[Page 5436]]

    Of note, due to their pelagic distribution, typical high 
availability bias due to deep-diving behavior and cryptic nature when 
at the surface, beaked whales are rarely sighted during at-sea surveys 
and difficult to distinguish between species when visually observed in 
the field. Accordingly, abundance estimates in NMFS SARs are recorded 
for Mesoplodon spp. Available sightings data, including often 
unresolved sightings of beaked whales, must be combined in order to 
develop habitat-based density models for beaked whales, as were used to 
inform our acoustic exposure modeling effort. Therefore, density and 
take estimates in this rule are similarly lumped for the three species 
of beaked whales, and there is no additional information by which NMFS 
could appropriately apportion impacts other than equally/proportionally 
across the three species.
    No mortality or Level A harassment of any of these three species of 
beaked whales is anticipated or authorized. It is likely that the 
majority of the individual beaked whales will be impacted on one or 
more days during a year of activity by one type of survey or another. 
It is possible that some small number of female beaked whales may 
experience a year of foregone reproduction. However, based on the 
nature of the majority of the individual exposures and the overall 
scale of the aggregate impacts and risk rating in consideration of the 
mitigation discussed here, and noting the continued presence of beaked 
whales in the GOM given the many years of high activity levels and the 
evidence that beaked whales maintain stable or increasing populations 
in other areas with high levels of acoustic activity, the impacts of 
the expected takes from these activities are not likely to adversely 
affect the GOM stocks of Cuvier's, Gervais', or Blainville's beaked 
whales through adverse impacts on annual rates of recruitment or 
survival.

Kogia spp.

    First, we note that the estimated (and allowable) take of Kogia 
spp. has been reduced as compared to the proposed rule as a result of 
the change in scope. Specifically, the maximum annual take by Level B 
harassment decreased by approximately 46 percent and the average annual 
take decreased by approximately 43 percent. (These reductions are 49 
and 46 percent, respectively, for Level A harassment.) The EWG 
analysis, which evaluated the relative significance of the aggregated 
impacts of the survey activities across seven GOM zones in the context 
of the vulnerability of each species, concluded that the GOM-wide risk 
ratings for Kogia spp. were low (for the high effort scenario) and very 
low (for the moderate effort scenario). Evaluated risk is reduced from 
the proposed rule, where the high effort scenario resulted in a 
moderate risk rating and the moderate effort scenario resulted in a low 
risk rating. We further consider the likely severity of any predicted 
behavioral disruption of Kogia spp. in the context of the likely 
duration of exposure above Level B harassment thresholds. Specifically, 
the average modeled time per day spent at received levels above 160 dB 
rms (where 50 percent of the exposed population is considered taken) 
ranges from 2.8-7.9 minutes for 2D, 3D NAZ, and 3D WAZ surveys and up 
to 15.3 minutes for coil surveys (which comprise less than 10 percent 
of the total projected activity days), and the average time spent 
between 140 and 160 dB rms (where 10 percent of the exposed population 
is considered taken) is 6.7-19 minutes.
    Odontocetes echolocate to find prey, and while there are many 
different strategies for hunting, one common pattern, especially for 
deeper diving species, is to conduct multiple repeated deep dives 
within a feeding bout, and multiple bouts within a day, to find and 
catch prey. While exposures of the short durations noted above could 
potentially interrupt a dive or cause an individual to relocate to 
feed, such a short-duration interruption would be unlikely to have 
significant impacts on an individual's energy budget and, further, for 
these species and this open-ocean area, there are no specific known 
reasons (i.e., these species range GOM-wide beyond the continental 
slope and there are no known biologically important areas) to expect 
that there would not be adequate alternate feeding areas relatively 
nearby, especially considering the anticipated absence of survey 
activity in the eastern GOM.
    As described above, no survey activity is expected within the Dry 
Tortugas Mitigation Area, which was analyzed and proposed for 
implementation in the proposed rule. The absence of survey activity in 
the area is expected to afford additional reduction of impacts to Kogia 
spp., in addition to sperm and beaked whales, given their relatively 
high density in that area. Importantly, the absence of survey activity 
in the area will reduce disturbance of these species in places of 
importance to them for critical behaviors such as foraging and 
socialization and, overall, help to reduce evaluated risk to the stocks 
as a whole.
    NMFS has analyzed the likely impacts of potential hearing 
impairment, including the estimated upper bounds of permanent threshold 
shift (Level A harassment) that could be authorized under the rule, and 
do not expect that they would result in impacts on reproduction or 
survival of any individuals. As described in the previous section, the 
degree of injury for individuals would be expected to be mild, and the 
predicted PTS resulting from airgun exposure is not likely to affect 
echolocation performance or communication for Kogia spp. Additionally, 
the extended distance shutdown zone for Kogia spp. (1,500 m) is 
expected to further minimize the severity of any hearing impairment 
incurred and also to further reduce the likelihood of, and minimize the 
severity of, more severe behavioral responses.
    Of note, due to their pelagic distribution, small size, and cryptic 
behavior, pygmy sperm whales and dwarf sperm whales are rarely sighted 
during at-sea surveys and difficult to distinguish between when 
visually observed in the field. Accordingly, abundance estimates in 
NMFS SARs are recorded for Kogia spp. only, density and take estimates 
in this rule are similarly lumped for the two species, and there is no 
additional information by which NMFS could appropriately apportion 
impacts other than equally/proportionally across the two species.
    No mortality of Kogia spp. is anticipated or authorized. While it 
is likely that the majority of the individuals of these two species 
will be impacted briefly on one or more days during a year of activity 
by one type of survey or another, based on the nature of the individual 
exposures and takes, as well as the aggregated scale of the impacts 
across the GOM, and in consideration of the mitigation discussed here, 
the impacts of the expected takes from these activities are not likely 
to adversely impact the GOM stocks of dwarf or pygmy sperm whales 
through adverse impacts on annual rates of recruitment or survival.

Bottlenose Dolphins

    The change in scope did not result in any appreciable change to 
estimated (and allowable) take of bottlenose dolphins compared to the 
proposed rule. Specifically, the maximum annual take increased slightly 
(by approximately 2 percent), while the average annual take decreased 
slightly (by approximately 1 percent). The EWG analysis, which 
evaluated the relative significance of the aggregated impacts of the 
survey activities across seven GOM zones in the context of the 
vulnerability of each species, concluded that the GOM-wide risk ratings 
for both oceanic

[[Page 5437]]

bottlenose dolphins and coastal/shelf bottlenose dolphins are very low 
for both scenarios. In the proposed rule, risk was evaluated for 
bottlenose dolphins GOM-wide (here we have refined the risk evaluation 
to differentiate between oceanic and coastal/shelf stocks). Evaluated 
risk is reduced from the proposed rule, where the high effort scenario 
resulted in a low risk rating and the moderate effort scenario resulted 
in a moderate risk rating. We further considered the likely severity of 
any predicted behavioral disruption of bottlenose dolphins in the 
context of the likely duration of exposure above Level B harassment 
thresholds. Specifically, the average modeled time per day spent at 
received levels above 160 dB rms (where 50 percent of the exposed 
population is considered taken) ranges from 4-11.7 minutes for 2D, 3D 
NAZ, and 3D WAZ surveys and up to 16.8 minutes for coil surveys (which 
comprise less than 10 percent of the total projected activity days) and 
the average time spent between 140 and 160 dB rms is 19.7-54.6 minutes. 
While exposures of the short durations noted above could potentially 
interrupt a dive or cause an individual to relocate to feed, among 
other impacts, such a short-duration interruption would be unlikely to 
have significant impacts on an individual's energy budget or otherwise 
impact reproduction or survival.
    As described earlier in this preamble, the northern coastal stock 
of bottlenose dolphin was particularly severely impacted by the DWH oil 
spill, and was additionally affected by a recent UME. Importantly, as 
described in Mitigation, NMFS is requiring a seasonal time-area 
restriction on airgun survey activity within the coastal waters where 
this stock is likely to be found. The closure area is expected to 
protect coastal bottlenose dolphins and their habitat through the 
alleviation or minimization of a range of potential effects of airgun 
noise, both acute and chronic, that could otherwise accrue to impact 
the reproduction or survival of individuals in this area. The timing of 
the restriction provides protection during the times of year thought to 
be most important for bottlenose dolphin calving and nursing of young. 
Although some sound from airguns may still propagate into the area from 
surveys that may occur outside of the area, exposure of bottlenose 
dolphins to sound levels that would result in Level B harassment will 
be alleviated or reduced for animals within the closure area. Any 
exposure to noise that may increase stress levels and exacerbate health 
problems in bottlenose dolphins still recovering from the effects of 
the DWH spill will be minimized during this important reproductive 
period. This important mitigation results in a reduction in the scale 
of aggregate effects (which, among other things, suggests the 
comparative number of days across which individual bottlenose dolphins 
might be taken within a year) and associated risk assessment.
    Of note, bottlenose dolphins cannot be identified to stock when 
visually observed in the field. Abundance estimates in NMFS SARs are 
based strictly on the location where animals are observed, and 
available sightings data must be combined in order to develop habitat-
based density models for bottlenose dolphins, as were used to inform 
our acoustic exposure modeling effort. Therefore, density and take 
estimates in this rule are provided for bottlenose dolphins as a GOM-
wide species. However, based on NMFS' stock delineations, we can 
reasonably assume that dolphins occurring within Zones 4-7 would be 
from the oceanic stock, while dolphins occurring within Zones 1-3 would 
be from the shelf stock and/or coastal stocks. Therefore, for the 
oceanic stock, we are able to draw stock-specific conclusions in this 
analysis. For coastal/shelf stocks, there is no additional information 
by which NMFS could appropriately apportion impacts other than equally/
proportionally across the stocks, with the exception of predicting 
reduced impacts to the northern coastal stock as described above. We 
note that, as a result of BOEM's update to the scope of activity, the 
eastern coastal stock will not experience any impacts and is 
accordingly no longer considered in this rule.
    No mortality or Level A harassment of bottlenose dolphins is 
anticipated or authorized. While it is likely that the majority of 
individual dolphins may be impacted briefly on one or more days during 
a year of activity by one type of survey or another, based on the 
nature of the individual exposures (shorter duration) and takes, as 
well as the aggregated scale of the impacts across the GOM in 
consideration of the mitigation discussed here, the impacts of the 
expected takes from these activities are not likely to adversely affect 
any affected GOM stock of bottlenose dolphins through adverse impacts 
on annual rates of recruitment or survival.

All Other Stocks

    In consideration of the similarities in the nature and scale of 
impacts, we consider the GOM stocks of the following species together 
in this section: Rough-toothed dolphin, Clymene dolphin, Atlantic 
spotted dolphin, pantropical spotted dolphin, striped dolphin, spinner 
dolphin, Fraser's dolphin, Risso's dolphin, melon-headed whale, pygmy 
killer whale, false killer whale, killer whale, and short-finned pilot 
whale. Estimated (and allowable) take of these stocks (including both 
the maximum annual take and the average annual take) has been reduced 
as compared to the proposed rule as a result of the change in scope 
(with the exception of the Atlantic spotted dolphin). For the Atlantic 
spotted dolphin, the change in scope resulted in increases compared to 
the proposed rule. Specifically, the maximum annual take increased by 
approximately 9 percent, while the average annual take increased by 
approximately 4 percent. These slight increases do not impact our 
analysis for the stock.
    The EWG analysis, which evaluated the relative significance of the 
aggregated impacts of the survey activities across seven GOM zones in 
the context of the vulnerability of each species, concluded that the 
GOM-wide risk ratings for high and moderate effort scenarios ranged 
from very low to low for these species. For all stocks, there was a 
trend of decreased or static risk ratings compared to the proposed 
rule, where the GOM-wide risk ratings for high and moderate effort 
scenarios ranged from low to moderate.
    We further considered the likely severity of any predicted 
behavioral disruption of the individuals of these species in the 
context of the likely duration of exposure above Level B harassment 
thresholds. Specifically, the average modeled time per day spent at 
received levels above 160 dB rms (where 50 percent of the exposed 
population is considered taken) ranges from 1.4-11.7 minutes for 2D, 3D 
NAZ, and 3D WAZ surveys and up to 25.7 minutes for coil surveys (which 
comprise less than 10 percent of the total projected activity days). 
The average time per day spent between 140 and 160 dB rms for 
individuals that are taken is from 8-58.1 minutes, with the one 
exception of killer whales exposed to noise from coil surveys, which 
average 73.6 minutes (though we note that the overall risk rating for 
the species is very low).
    Odontocetes echolocate to find prey, and there are many different 
strategies for hunting. One common pattern for deeper-diving species is 
to conduct multiple repeated deep dives within a feeding bout, and 
multiple bouts within a day, to find and catch prey. While

[[Page 5438]]

exposures of the shorter durations noted above could potentially 
interrupt a dive or cause an individual to relocate to feed, such a 
short-duration interruption would be unlikely to have significant 
impacts on an individual's energy budget and, further, for these 
species and this open-ocean area, there are no specific known reasons 
(i.e., these species range GOM-wide beyond the continental slope and 
there are no known biologically important areas) to expect that there 
would not be adequate alternate feeding areas relatively nearby, 
especially considering the anticipated absence of survey activity in 
the eastern GOM. For those species that are more shallow feeding 
species, it is unlikely that the noise exposure considered herein would 
result in minimal significant disruption of foraging behavior and, 
therefore, the concomitant energetic effects would similarly be 
minimal.
    Of note, the Atlantic spotted dolphin would benefit (via lessening 
of both number and severity of takes) from the coastal waters time-area 
restriction developed to benefit bottlenose dolphins and several 
additional species experience notably reduced effects from the absence 
of survey activity in important eastern GOM habitat. Specifically, 
multiple shelf-break associated and pelagic species (such as Risso's 
dolphin, melon-headed whales, and rough-toothed dolphins) experience a 
reduction estimated take from the absence of survey activity in both 
the Bryde's whale core habitat and Dry Tortugas Areas. Maximum annual 
and average annual take decreased for these species compared with the 
proposed rule by 20 and 14 percent, 19 and 15 percent, and 19 and 18 
percent, respectively. Numerous other species would be expected to be 
present in varying numbers at various times.
    No mortality or Level A harassment of these species is anticipated 
or authorized. It is likely that the majority of the individuals of 
these 13 species will be impacted briefly on one or more days during a 
year of activity by one type of survey or another. Based on the nature 
of the individual exposures and takes, as well as the very low to low 
aggregated scale of the impacts across the GOM and considering the 
mitigation discussed here, the impacts of the expected takes from these 
activities are not likely to adversely impact the GOM stocks of any of 
these 13 GOM stocks of these species through adverse impacts on annual 
rates of recruitment or survival.

Determination

    Based on the analysis contained herein of the likely effects of the 
specified activities on marine mammals and their habitat, and taking 
into consideration the implementation of the monitoring and mitigation 
measures, NMFS finds that the total marine mammal take from the 
specified activities will have a negligible impact on all affected 
marine mammal species and stocks.

Small Numbers

    The sections below provide an explanation of how NMFS interprets 
and applies the small numbers standard and remain substantively 
unchanged from the discussion provided in the notice of proposed 
rulemaking. Additional discussion appears in Comments and Responses to 
address specific comments, questions, or recommendations received from 
the public.

What are small numbers?

    The term ``small numbers'' appears in section 101(a)(5)(A) of the 
MMPA as follows:
    (5)(A)(i) Upon request therefor by citizens of the United States 
who engage in a specified activity (other than commercial fishing) 
within a specified geographical region, the Secretary shall allow, 
during periods of not more than five consecutive years each, the 
incidental, but not intentional, taking by citizens while engaging in 
that activity within that region of small numbers of marine mammals of 
a species or population stock if the Secretary, after notice (in the 
Federal Register and in newspapers of general circulation, and through 
appropriate electronic media, in the coastal areas that may be affected 
by such activity) and opportunity for public comment--
    (I) finds that the total of such taking during each five-year (or 
less) period concerned will have a negligible impact on such species or 
stock and will not have an unmitigable adverse impact on the 
availability of such species or stock for taking for subsistence uses [ 
. . . ] and
    (II) prescribes regulations setting forth--
    (aa) permissible methods of taking pursuant to such activity, and 
other means of effecting the least practicable adverse impact on such 
species or stock and its habitat, paying particular attention to 
rookeries, mating grounds, and areas of similar significance, and on 
the availability of such species or stock for subsistence uses; and
    (bb) requirements pertaining to the monitoring and reporting of 
such taking.
    (Emphasis added.)
    In addition to section 101(a)(5)(A), the MMPA as amended in 1994 
includes a similar provision in section 101(a)(5)(D), which provides 
for the issuance of incidental take authorizations for small numbers of 
marine mammals without the need for regulations, effective for up to 
one year, where the taking is limited to harassment:
    (5)(D)(i) Upon request therefor by citizens of the United States 
who engage in a specified activity (other than commercial fishing) 
within a specific geographic region, the Secretary shall authorize, for 
periods of not more than 1 year, subject to such conditions as the 
Secretary may specify, the incidental, but not intentional, taking by 
harassment of small numbers of marine mammals of a species or 
population stock by such citizens while engaging in that activity 
within that region if the Secretary finds that such harassment during 
each period concerned--
    (I) will have a negligible impact on such species or stock, and
    (II) will not have an unmitigable adverse impact on the 
availability of such species or stock for taking for subsistence 
uses[.]
    (Emphasis added.)
    The MMPA does not define ``small numbers.'' NMFS' and the U.S. Fish 
and Wildlife Service's 1989 implementing regulations defined small 
numbers as a portion of a marine mammal species or stock whose taking 
would have a negligible impact on that species or stock. This 
definition was invalidated in Natural Resources Defense Council v. 
Evans, 279 F.Supp.2d 1129 (2003) (N.D. Cal. 2003), based on the court's 
determination that the regulatory definition of small numbers was 
improperly conflated with the regulatory definition of ``negligible 
impact,'' which rendered the small numbers standard superfluous. As the 
court observed, ``the plain language indicates that small numbers is a 
separate requirement from negligible impact.'' Since that time, NMFS 
has not applied the definition found in its regulations. Rather, 
consistent with Congress' pronouncement that small numbers is not a 
concept that can be expressed in absolute terms (House Committee on 
Merchant Marine and Fisheries Report No. 97-228 (September 16, 1981)), 
NMFS makes its small numbers findings based on an analysis of whether 
the number of individuals authorized to be taken annually from a 
specified activity is small relative to the stock or population size. 
The Ninth Circuit has upheld a similar approach. See Center for 
Biological Diversity v. Salazar, 695 F.3d 893 (9th Cir. 2012).

[[Page 5439]]

However, NMFS has not historically indicated what the agency believes 
to be the upper limit of small numbers.
    To maintain an interpretation of small numbers as a proportion of a 
species or stock that does not conflate with negligible impact, NMFS 
uses a simple approach that establishes equal bins corresponding to 
small, medium, and large proportions of the population abundance. NMFS 
then compares the number of individuals estimated and authorized to be 
taken against the best available abundance estimate for that species or 
stock.
    It can be challenging to predict the numbers of individual marine 
mammals that will be taken by an activity. Many models calculate 
instances of take but are unable to account for repeated exposures of 
individual marine mammals, though the instances of take necessarily 
represent the upper bound of the number of individuals. In some of 
those cases, such as for this rule (see Estimated Take), we are able to 
generate a more refined estimate of the numbers of individuals 
predicted to be taken utilizing a combination of quantitative tools and 
qualitative information. When an acceptable estimate of the individual 
marine mammals taken is available,\20\ the small numbers determination 
is based directly upon whether these estimates exceed one-third of the 
stock abundance. In other words, consistent with past practice, when 
the estimated number of individual animals taken (which may or may not 
be assumed as equal to the total number of takes, depending on the 
available information) is up to, but not greater than, one-third of the 
most appropriate species or stock abundance, NMFS will determine that 
the numbers of marine mammals taken of a species or stock are small.
---------------------------------------------------------------------------

    \20\ We note that although NMFS' implementing regulations 
require applications for incidental take to include an estimate of 
the marine mammals to be taken, there is nothing in section 
101(a)(5)(A) (or (D)) that requires NMFS to quantify or estimate 
numbers of marine mammals to be taken for purposes of evaluating 
whether the number is small. (See CBD v. Salazar.)
---------------------------------------------------------------------------

    Another circumstance in which NMFS considers it appropriate to make 
a small numbers finding is in the case of a species or stock that may 
potentially be taken but is either rarely encountered or only expected 
to be taken on rare occasions. In that circumstance, one or two assumed 
encounters with a group of animals (meaning a group that is traveling 
together or aggregated, and thus exposed to a stressor at the same 
approximate time) should reasonably be considered small numbers, 
regardless of consideration of the proportion of the stock, as 
infrequent or rare encounters resulting in take of one or two groups 
should be considered small relative to the range and distribution of 
any stock.
    In summary, when quantitative take estimates of individual marine 
mammals are available or inferable through consideration of additional 
factors, and the number of animals taken is one-third or less of the 
best available abundance estimate for the species or stock, NMFS 
considers it to be of small numbers. NMFS may also appropriately find 
that one or two predicted group encounters will result in small numbers 
of take relative to the range and distribution of a species, regardless 
of the estimated proportion of the abundance.

Is the small numbers standard evaluated based on total take under 
incidental take regulations or within the context of an individual 
letter of authorization?

    Neither the MMPA nor NMFS' implementing regulations address whether 
the small numbers determination should be based upon the total annual 
taking for (1) all activities occurring under specific incidental take 
regulation or (2) to individual LOAs issued thereunder. The MMPA does 
not define small numbers or explain how to apply the term in either 
paragraph (A) or (D) of section 101(a)(5), including how to apply the 
term in a way that allows for consistency between those two very 
similar provisions in the statute. Whether to apply the small numbers 
finding to each individual LOA under regulations that cover multiple 
concurrent LOA holders is a matter of first impression for NMFS.
    Specifically, section 101(a)(5)(A)(i)(I) explicitly states that the 
negligible impact determination for a specified activity must take into 
account the total taking over the five-year period, but the small 
numbers language is not tied explicitly to the same language. Rather, 
the small numbers provision appears in section 101(a)(5)(A)(i) as a 
limitation on what the Secretary may allow. The regulatory vehicle for 
authorizing (i.e., ``allowing'') the take of marine mammals is the LOA.
    Given NMFS' discretion in light of the ambiguities in the statute 
regarding how to apply the small numbers standard, we have determined 
that the small numbers finding should be applied to the annual take 
authorized per individual LOA, rather than to the total annual taking 
for all activities potentially occurring under the incidental take 
regulations. This per-LOA approach harmonizes section 101(a)(5)(A) with 
the per-IHA application in section 101(a)(5)(D) of the MMPA.\21\ This 
per-LOA approach is not only permissible but also preferable to the 
total annual taking approach because NMFS' per-LOA approach to small 
numbers in section 101(a)(5)(A) affords greater regulatory flexibility 
to utilize section 101(a)(5)(A) when there are benefits to doing so for 
the resource (marine mammals), the public, prospective applicants, and 
administrative efficiency:
---------------------------------------------------------------------------

    \21\ As the court observed in Native Village of Chickaloon v. 
NMFS, 947 F. Supp. 2d 1031, 1049 n.123 (D. Alaska 2013) ``the same 
statutory standards apply'' to incidental take authorization under 
both provisions.
---------------------------------------------------------------------------

     From a resource protection standpoint, it is better to 
conduct a comprehensive negligible impact analysis that considers all 
of the activities covered under the rule (versus considering them 
independently pursuant to individual IHAs) and ensures that the total 
combined taking from those activities will have a negligible impact on 
the affected marine mammal species or stocks and no unmitigable adverse 
impact on subsistence uses. Furthermore, mitigation and monitoring are 
more effective when considered across all activity and years covered 
under regulations.
     From an agency resource standpoint, it ultimately will 
save significant time and effort to cover multi-year activities under a 
rule instead of multiple incidental harassment authorizations (IHAs). 
While regulations require more analysis up front, additional public 
comment and internal review, and additional time to promulgate compared 
to a single IHA, they are effective for up to five years (for non-
military readiness activities) and can cover multiple actors within a 
year. The process of issuing individual LOAs under incidental take 
regulations utilizes the analysis, public comment, and review that was 
conducted for the regulations, and takes significantly less time than 
it takes to issue independent IHAs.
     From an applicant standpoint, incidental take regulations 
offer more regulatory certainty than IHAs (five years versus one year) 
and significant cost savings, both in time and environmental compliance 
analysis and documentation. This is especially true for situations like 
here, where multiple applicants will be applying for individual LOAs 
under regulations. In the case of this rule, the certainty afforded by 
the promulgation of a regulatory framework (e.g., by using previously 
established take estimates, mitigation and monitoring

[[Page 5440]]

requirements, and procedures for requesting and obtaining an LOA) is a 
significant benefit for prospective applicants.
    NMFS' evaluation of past IHAs suggests that bundling together the 
activities covered by two or three IHAs that might be ideal subjects 
for a combined incidental take regulation (e.g., for ongoing 
maintenance construction activities, or seismic surveys in the Arctic 
by different entities) may exceed the taking of small numbers of a 
species if NMFS were to apply the small numbers standard across all 
taking contemplated by the regulation in a year. In other words, if the 
small numbers standard is applied to the total annual taking under a 
rule, NMFS may not be able to make the necessary small numbers finding, 
which would preclude the use of section 101(a)(5)(A) for multiple 
activities, thereby eliminating the opportunity to derive the resource 
and streamlining benefits outlined above. Also, application of the 
small numbers standard across the total annual taking covered by an 
incidental take regulation, inasmuch as prospective applicants can see 
that the total annual take may exceed one-third of species or stock 
abundance, would create an incentive for applicants to pursue 
individual IHAs (again, precluding the ability to gain the benefits 
outlined above).
    Our conclusion is that NMFS can appropriately elect to make a 
``small numbers'' finding based on the estimated annual take in 
individual LOAs issued under the rule. This approach does not affect 
the negligible impact analysis for a rule, which is the biologically 
relevant inquiry and based on the total annual estimated taking for all 
activities the regulations will govern. Making the small numbers 
finding based on the estimated annual take in individual LOAs allows 
NMFS to take advantage of the associated administrative and 
environmental benefits of utilizing section 101(a)(5)(A) that would be 
precluded in many cases if small numbers were required to be applied to 
the total annual taking under the regulations. NMFS finds this method 
of making a small numbers determination to be a permissible 
interpretation of the relevant MMPA provisions.
    Although this application of small numbers may be argued as being 
less protective of marine mammals, NMFS disagrees. As noted previously, 
the small numbers standard has less biological significance as compared 
to the substantive and contextually-specific analysis necessary to 
support the negligible impact determination. The negligible impact 
determination is still controlling, and the maximum total annual taking 
that may be authorized across all LOAs under an incidental take 
regulation still could not exceed the overall amount analyzed for the 
negligible impact determination. Thus, under this option, the 
negligible impact analysis for the rulemaking still would have to be 
conducted for the time period explicitly specified in the statute 
(i.e., up to five years), but the small numbers analysis would attach 
to the instrument itself that authorizes the taking, i.e., the LOA.

How will small numbers be evaluated under this GOM rule?

    In this rule, up-to-date species information is available, and 
sophisticated models have been used to estimate take in a manner that 
will allow for quantitative comparison of the take of individuals 
versus the best available abundance estimates for the species or 
stocks. Specifically, while the modeling effort utilized in the rule 
enumerates the estimated instances of takes that will occur across days 
as the result of the operation of certain survey types in certain 
areas, the modeling report also includes the evaluation of a test 
scenario that allows for a reasonable modification of those generalized 
take estimates to better estimate the number of individuals that will 
be taken within one survey. LOA applicants using modeling results from 
the rule to inform their applications will be able to reasonably 
estimate the number of marine mammal individuals taken by their 
activities. LOA applications that do not use the modeling provided in 
the rule to estimate take for their activities will need to be 
reviewed, and applicants will be required to ensure that their 
estimates adequately inform the small numbers finding. If applicants 
use the modeling provided by this rule to estimate take, additional 
review will not be deemed necessary (unless other conditions 
necessitating review exist, as described in the Letters of 
Authorization section). If applicants do not use the modeling provided 
by the rule, however, NMFS may publish a notice in the Federal Register 
soliciting public comment, if the model or inputs differ substantively 
from those that have been reviewed by NMFS and the public previously, 
if the model or inputs differ substantively from those that have been 
reviewed by NMFS and the public previously. The estimated take of 
marine mammals for each species will then be compared against the best 
available scientific information on species or stock abundance estimate 
as determined by NMFS, and estimates that do not exceed one-third of 
that estimate will be considered small numbers.

Adaptive Management

    The regulations governing the take of marine mammals incidental to 
geophysical survey activities contain an adaptive management component. 
The comprehensive reporting requirements associated with this rule (see 
the Monitoring and Reporting section) are designed to provide NMFS with 
monitoring data from the previous year to allow consideration of 
whether any changes are appropriate. The use of adaptive management 
allows NMFS to consider new information from different sources to 
determine (with input from the LOA-holders regarding practicability) on 
a regular (e.g., annual or biennial) basis if mitigation or monitoring 
measures should be modified (including additions or deletions). 
Mitigation measures could be modified if new data suggests that such 
modifications would have a reasonable likelihood of reducing adverse 
effects to marine mammal species or stocks or their habitat and if the 
measures are practicable. The adaptive management process and 
associated reporting requirements would serve as the basis for 
evaluating performance and compliance.
    The following are some of the possible sources of applicable data 
to be considered through the adaptive management process: (1) Results 
from monitoring reports, as required by MMPA authorizations; (2) 
results from general marine mammal and sound research; and (3) any 
information which reveals that marine mammals may have been taken in a 
manner, extent, or number not authorized through these regulations and 
subsequent LOAs or that the specified activity may be having more than 
a negligible impact on affected stocks.
    Under this rule, NMFS plans to implement an annual adaptive 
management process including BOEM, BSEE, industry operators (including 
geophysical companies as well as exploration and production companies), 
and others as appropriate. Industry operators may elect to be 
represented in this process by their respective trade associations. 
NMFS, BOEM, and BSEE (i.e., the regulatory agencies) and industry 
operators who have conducted or contracted for survey operations in the 
GOM in the prior year (or their representatives) will provide an 
agreed-upon description of roles and responsibilities, as well as 
points of contact, in advance of each year's

[[Page 5441]]

adaptive management process. The foundation of the adaptive management 
process will be the annual comprehensive reports produced by LOA-
holders (or their representatives), as well as the results of any 
relevant research activities, including research supported voluntarily 
by the oil and gas industry and research supported by the Federal 
government. Please see the Monitoring Contribution Through Other 
Research section in the notice of proposed rulemaking for a description 
of representative past research efforts. The outcome of the annual 
adaptive management process would be an assessment of effects to marine 
mammal populations in the GOM relative to NMFS' determinations under 
the MMPA and ESA, recommendations related to mitigation, monitoring, 
and reporting, and recommendations for future research (whether 
supported by industry or the regulatory agencies).
    Data collection and reporting by individual LOA-holders will occur 
on an ongoing basis, per the terms of issued LOAs. In a given annual 
cycle, the comprehensive annual report will summarize and synthesize 
all LOA-specific reports received, with report development (supported 
through collaboration of individual LOA-holders or by their 
representatives) occurring for 90 days following the end of a given 
one-year period. Review and revision of the report, followed by a joint 
meeting of the parties, will occur within 90 days following receipt of 
the annual report. Any agreed-upon modifications will occur through the 
process for modifications and/or adaptive management described in the 
regulatory text following this preamble.

Monitoring Contribution Through Other Research

    NMFS' MMPA implementing regulations require that applicants for 
incidental take authorizations describe the suggested means of 
coordinating research opportunities, plans, and activities relating to 
reducing incidental taking and evaluating its effects (50 CFR 
216.104(a)(14)). Such coordination can serve as an effective supplement 
to the monitoring and reporting required pursuant to issued LOAs and/or 
incidental take regulations. NMFS expects that relevant research 
efforts will inform the annual adaptive management process described 
above, and that levels and types of research efforts will change from 
year to year in response to identified needs and evolutions in 
knowledge, emerging trends in the economy and available funding, and 
available scientific and technological resources. In the notice of 
proposed rulemaking, NMFS described examples of relevant research 
efforts (83 FR 29300-29301). We do not repeat that information here, 
but refer the reader to that notice for more information. The described 
efforts may not be predictive of any future levels and types of 
research efforts. Research occurring in locations other than the GOM 
may be relevant to understanding the effects of geophysical surveys on 
marine mammals or marine mammal populations or the effectiveness of 
mitigation. NMFS also refers the reader to the industry Joint Industry 
Program (JIP) website (www.soundandmarinelife.org), which hosts a 
database of available products funded partially or fully through the 
JIP, and to BOEM's Environmental Studies Program (ESP), which develops, 
funds, and manages scientific research to inform policy decisions 
regarding outer continental shelf resource development (www.boem.gov/studies).

Impact on Availability of Affected Species for Taking for Subsistence 
Uses

    There are no relevant subsistence uses of marine mammals implicated 
by these actions. Therefore, NMFS has determined that the total taking 
of affected species or stocks will not have an unmitigable adverse 
impact on the availability of such species or stocks for taking for 
subsistence purposes.

Endangered Species Act (ESA)

    Section 7 of the ESA requires Federal agencies to insure that their 
actions are not likely to jeopardize the continued existence of 
endangered or threatened species or adversely modify or destroy their 
designated critical habitat. Federal agencies must consult with NMFS 
for actions that may affect such species under NMFS' jurisdiction or 
critical habitat designated for such species.
    At the conclusion of consultation, the consulting agency provides 
an opinion stating whether the Federal agency's action is likely to 
jeopardize the continued existence of ESA-listed species or destroy or 
adversely modify designated critical habitat.
    NMFS's issuance of this final rule, and any subsequent LOAs, is 
subject to the requirements of Section 7 of the ESA. Therefore, NMFS' 
Office of Protected Resources (OPR), Permits and Conservation Division 
requested initiation of a formal consultation with the NMFS OPR, ESA 
Interagency Cooperation Division on the proposed issuance of the rule 
and subsequent LOAs on July 19, 2018. The formal consultation concluded 
and a final Biological Opinion (BiOp) was issued on March 13, 2020. The 
BiOp concluded that the Permits and Conservation Division's proposed 
action is not likely to jeopardize the continued existence of sperm 
whales or the GOM Bryde's whale.

National Environmental Policy Act

    In 2017, BOEM produced a final Programmatic Environmental Impact 
Statement (PEIS) to evaluate the direct, indirect, and cumulative 
impacts of geological and geophysical survey activities on the GOM OCS, 
pursuant to requirements of NEPA. These activities include geophysical 
surveys in support of hydrocarbon exploration, as are described in the 
MMPA petition before NMFS. The PEIS is available online at: 
www.boem.gov/Gulf-of-Mexico-Geological-and-Geophysical-Activities-Programmatic-EIS/. NOAA, through NMFS, participated in preparation of 
the PEIS as a cooperating agency due to its legal jurisdiction and 
special expertise in conservation and management of marine mammals, 
including its responsibility to authorize incidental take of marine 
mammals under the MMPA.
    NEPA, Council on Environmental Quality (CEQ) regulations, and 
NOAA's NEPA implementing procedures (NOAA Administrative Order (NAO) 
216-6A) encourage the use of programmatic NEPA documents to streamline 
decision-making. NMFS reviewed the Final PEIS and determined that it 
meets the requirements of the CEQ regulations (40 CFR part 1500-1508) 
and NAO 216-6A. NMFS further determined, after independent review, that 
the Final PEIS satisfied NMFS' comments and suggestions in the NEPA 
process. In the notice of proposed rulemaking, NMFS stated its 
intention to adopt BOEM's analysis in order to assess the impacts to 
the human environment of issuance of the subject ITR, and that we would 
review all comments submitted in response to the notice as we completed 
the NEPA process, including a final decision of whether to adopt BOEM's 
PEIS and sign a Record of Decision related to issuance of the ITR and 
subsequent LOAs. Following review of public comments received, NMFS 
confirmed that it would be appropriate to adopt BOEM's analysis in 
order to support assessment of the impacts to the human environment of 
issuance of the subject ITR and subsequent LOAs. Therefore NMFS 
prepared a Record of Decision for the following purposes: (1) To adopt 
the Final PEIS to support NMFS' analysis associated with issuance of 
incidental take authorizations pursuant to section 101(a)(5)(A) or (D) 
of the MMPA and the regulations governing the taking and

[[Page 5442]]

importing of marine mammals (50 CFR part 216); and (2) in accordance 
with 40 CFR 1505.2, to announce and explain the basis for NMFS' 
decision to review and potentially issue incidental take authorizations 
under the MMPA on a case-by-case basis, if appropriate.

Letters of Authorization

    Under these incidental take regulations, industry operators may 
apply for and obtain LOAs, as described in NMFS' MMPA implementing 
regulations (50 CFR 216.106). LOAs may be issued for any time period 
that does not exceed the effective duration of the final rule, provided 
the description of the activity in the request includes a sufficient 
degree of specificity with which to evaluate whether the activity falls 
within the scope of the rule. Because the specified activity described 
herein does not provide actual specifics of the timing, location, and 
survey design for activities that would be the subject of issued LOAs, 
such requests must include, at minimum, the information described at 50 
CFR 216.104(a)(1) and (2), and should include an affirmation of intent 
to adhere to the mitigation, monitoring, and reporting requirements 
described in the regulations. The level of effort proposed by an 
operator would be used to develop an LOA-specific take estimate based 
on the results of Zeddies et al. (2015, 2017a).
    The proposed rule indicated that LOA applications with take 
estimates based on modeling other than that specifically included in 
the modeling report used to support the EIS and the proposed rule 
(Zeddies et al., 2015, 2017a) would be published for public comment 
prior to the issuance of an LOA. However, upon further consideration of 
the ``Gulf of Mexico Acoustic Exposure Model Variable Analysis'' 
(Zeddies et al., 2017b; ``Acoustic Exposure Model Variable Analysis'') 
provided by IAGC and API to NMFS prior to the publication of the 
proposed rule and made available to the public with the proposed rule 
and the Associations' public comments, which extensively referenced the 
Acoustic Exposure Model Variable Analysis, the final rule more flexibly 
provides that if applicants do not use the modeling provided by the 
rule, NMFS may publish a notice in the Federal Register soliciting 
public comment, if the model or inputs differ substantively from those 
that have been reviewed by NMFS and the public previously. 
Specifically, the Acoustic Exposure Model Variable Analysis includes 
the results (i.e., take estimates) of a supplemental analysis of the 
same modeling effort used in Zeddies et al. (2015, 2017a) to support 
the proposed rule, but evaluating the effects on the modeling results 
of different variables. One analyzed variable of particular utility was 
the use of a smaller airgun array that could serve as a reasonable 
representative for some of the smaller arrays that are commonly used in 
the GOM. This specific applicable example, in which the model and 
inputs of this Acoustic Exposure Model Variable Analysis have been 
reviewed by NMFS and the public previously (both in that they mirror 
Zeddies et al. (2015, 2017a) and in that NMFS also explicitly made the 
Acoustic Exposure Model Variable Analysis available to the public 
during the comment period), illustrates the need to provide flexibility 
and make efficient use of previous public and agency review. NMFS has, 
therefore, determined it appropriate to allow that additional public 
review is not needed unless the model or inputs differ substantively 
from those that have been reviewed by NMFS and the public previously. 
Further, we explicitly note the utility of the modeling and results 
presented in the Acoustic Exposure Model Variable Analysis report for 
representing smaller airgun arrays that are commonly used in the GOM 
and affirm that further public comment on that report should not be 
necessary prior to the use of its results to support the issuance of 
LOAs.
    Technologies continue to evolve to meet the technical, 
environmental, and economic challenges of oil and gas development. The 
use of ``new and unusual technologies'' (NUT), i.e., technologies other 
than those described herein, will be evaluated on a case-by-case basis 
and may require public review. Some seemingly new technologies proposed 
for use by operators are often extended applications of existing 
technologies and interface with the environment in essentially the same 
way as well-known or conventional technologies. For such evaluations, 
NMFS will follow the existing process used by BOEM, by using the 
following considerations:
     Has the technology or hardware been used previously or 
extensively in the U.S. GOM under operating conditions similar to those 
anticipated for the activities proposed by the operator? If so, the 
technology would not be considered a NUT;
     Does the technology function in a manner that potentially 
causes different impacts to the environment than similar equipment or 
procedures did in the past? If so, the technology would be considered a 
NUT;
     Does the technology have a significantly different 
interface with the environment than similar equipment or procedures did 
in the past? If so, the technology would be considered a NUT; and
     Does the technology include operating characteristics that 
are outside established performance parameters? If so, the technology 
would be considered a NUT.
    NMFS will consult with BOEM as well as with NMFS' ESA Interagency 
Cooperation Division regarding the level of review necessary for 
issuance of an LOA in which a NUT is proposed for use.

Classification

    Pursuant to the procedures established to implement Executive Order 
12866, the Office of Management and Budget (OMB) has determined that 
this rule is economically significant. Accordingly, a regulatory impact 
analysis (RIA) was prepared and made available for review by the 
public. Following review of public comments, a final RIA has been 
prepared and is available online at: www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-mexico. Appendix B of the RIA provides a final regulatory 
flexibility analysis (FRFA, discussed below), while Appendix C 
addresses other compliance requirements.
    The RIA evaluates the potential costs and benefits of these 
incidental take regulations against two baselines, a baseline 
corresponding with regulatory conditions in place since 2013 pursuant 
to a settlement agreement, as amended through stipulated agreement, 
involving a stay of litigation (NRDC et al. v. Bernhardt et al., Civil 
Action No. 2:10 cv-01882 (E.D. La.)), and a baseline corresponding to 
conditions prior to the 2013 settlement agreement. Under the settlement 
agreement that is in effect, industry trade groups representing 
operators agreed to include certain mitigation requirements for 
geophysical surveys in the GOM.
    OMB Circular A-4 provides that agencies may present multiple 
baselines where this would provide additional useful information to the 
public on the projected effects of the regulation. NMFS presented both 
baselines for public information and comment, consistent with the 
Circular A-4 provision allowing agencies to present multiple baselines. 
No information or comments regarding the economic baselines were 
received.
    These regulations require new mitigation measures relative to the 
settlement baseline and, thus, new costs for survey operators. However, 
the rule

[[Page 5443]]

also alleviates the burden of implementing minimum separation distance 
requirements for deep penetration airgun surveys, as required under the 
settlement agreement. The rule also results in certain indirect (but 
non-monetized) costs. However, the RIA analysis demonstrates that these 
costs are not likely to be significant. Moreover, as described in the 
RIA, total costs related to compliance for survey activities are small 
compared with expenditures on other aspects of oil and gas industry 
operations, and direct compliance costs of the regulatory requirements 
are unlikely to result in materially reduced oil and gas activities in 
the GOM.
    The rule also results in certain non-monetized benefits. The 
protection of marine mammals afforded by this rule (pursuant to the 
requirements of the MMPA) benefits the regional economic value of 
marine mammals via tourism and recreation to some extent, as mitigation 
measures applied to geophysical survey activities in the GOM region are 
expected to benefit the marine mammal populations that support this 
economic activity in the GOM. In addition, some degree of benefits can 
be expected to accrue solely via ecological benefits to marine mammals 
and other wildlife as a result of the regulatory requirements. The 
published literature (described in the RIA) is clear that healthy 
populations of marine mammals and other co-existing species benefit 
regional economies and provide social welfare benefits to people. 
However, the literature does not provide a basis for quantitatively 
valuing the cost of anticipated incremental changes in environmental 
disturbance and marine mammal harassment associated with the rule.
    Notably, the rule also affords significant benefit to the regulated 
industry by providing regulatory certainty through an efficient 
framework within which to achieve compliance with the MMPA. In 
particular, cost savings may be generated by the reduced administrative 
effort required to obtain an LOA under the framework established by a 
rule compared to what would be required to obtain an incidental 
harassment authorization (IHA) under section 101(a)(5)(D). Absent the 
rule, to attain equivalent compliance with the MMPA, survey operators 
in the GOM would need to apply for an IHA. Although not monetized in 
the RIA, NMFS' analysis indicates that the upfront work associated with 
the rule (e.g., analyses, modeling, process for obtaining LOA) likely 
saves significant time and money for operators. A conservative cost 
savings calculation, based on estimates of the costs for IHA 
applications relative to LOA application costs and an assumption of the 
number of likely authorizations based on total annual survey days and 
survey estimates included in the RIA, ranges from $500,000 to $1.5 
million annually. In terms of timing, NMFS recommends that IHA 
applicants contact the agency six to nine months in advance of the 
planned activity, whereas NMFS anticipates a timeframe of three months 
or less (depending upon the content of the request and the activities 
covered) for LOA applications under this rule.
    Details regarding cost estimation are available in the RIA. A 
qualitative evaluation of indirect costs related to the regulations is 
also provided in the RIA. Note that these costs would be diffused 
across all operators receiving LOAs.
    NMFS prepared a FRFA, as required by Section 603 of the Regulatory 
Flexibility Act (RFA), for this rule. The FRFA describes the economic 
effects this rule will have on small entities. A description of this 
action, why it is being considered, the objectives of the action, and 
the legal basis for the action are contained in the preamble of this 
rule. A copy of the full analysis is available as an appendix to the 
RIA. The MMPA provides the statutory basis for this rule. No 
duplicative, overlapping, or conflicting Federal rules have been 
identified. A detailed summary of the initial regulatory flexibility 
analysis was provided at the proposed rule stage. No comments or 
information regarding this analysis were received.
    This final rule is expected to directly regulate businesses that 
conduct geophysical surveys in the GOM with the potential to 
incidentally take marine mammals. Some of these businesses may be 
defined as small entities. The FRFA is summarized below.
    The FRFA focuses on identifying small businesses that would bear 
the incremental survey costs associated with the rule. These may 
include entities undertaking, commissioning, or purchasing surveys. In 
order to estimate the number of small entities to which the rule will 
apply, permit applications between 2006 and 2015 were analyzed to 
understand what industries were involved in permit applications for 
geophysical surveys in the Gulf of Mexico and to identify U.S.-based 
permit applicants that would be classified as small according to Small 
Business Administration definitions and the most recent revenue or 
employment data available. In total, 34 U.S.-based small businesses 
applied for geophysical survey permits in the Gulf of Mexico between 
2006 and 2015. By assuming that the same proportion of international, 
large, and small companies will undertake the surveys over the next 
five years as occurred during 2006 to 2015, the likely number of future 
surveys that will include small entity applicants may be estimated. 
Accordingly, NMFS estimates that small entities would apply for 
approximately 32 to 53 surveys over the next five years, or 
approximately six to 11 surveys annually. Historically, there was a 
ratio of approximately 2.2 surveys applied for per small entity. Using 
this ratio, NMFS estimates that approximately 15 to 24 small companies 
will likely apply for permits over the next five years, or 
approximately 3 to 5 small companies each year. The future distribution 
of small survey companies by industry is not known, but the historical 
pattern of surveys suggests that companies involved in oil and gas 
extraction (NAICS 2111) and support activities for oil and gas (213112) 
will account for the majority of the survey applications by small 
companies.
    A review of the reported annual revenues for the 34 small entities 
that applied for survey permits between 2006 and 2015 reveals a wide 
range, with the lowest revenues reported to be $0.04 million and the 
highest revenues reported to be $1.9 billion. Average revenues for the 
small entities who applied for permits were $232 million, with median 
revenues of $12.26 million. We note, however, that the revenues and 
numbers of employees reported for many of these small companies 
appeared to be erroneous, in multiple instances reporting annual 
revenues significantly less than the costs of conducting even the 
lowest cost surveys. As a result, these revenue estimates are likely to 
be inaccurate or, alternatively, permit applicants must pass survey 
costs on to the companies that purchase or commission the seismic data. 
Given that the oil and gas extraction companies are generally the 
entities purchasing the survey data, we expect that it is most likely 
that survey costs are ultimately borne by NAICS 2111 (oil and gas 
extraction), either as the permittees for the survey permit or because 
the other, smaller businesses pass these costs along in the data 
purchase price.
    In summary, the FRFA finds the following: First, in the majority of 
cases (88 percent), survey permit applicants are large businesses. 
Second, when the permit applicants are small businesses, the majority 
of the time (63 percent) they are oil and gas extractors (NAICS 2111). 
Third, together, these permits (for large businesses and small 
businesses with high annual revenues for which

[[Page 5444]]

rule costs are a small fraction) account for 96 percent of the survey 
permits. Fourth, while small entities in other industries occasionally 
apply for permits (four percent historically), these businesses are 
quite small, with average annual revenues in the millions or even less. 
Given their size, it is unlikely that these permit applicants bear 
survey costs; otherwise it would be reflected in their annual revenues 
(i.e., their revenues on average would reflect that they recover their 
costs). Accordingly, NMFS expects it is most likely that survey costs 
are passed on to oil and gas extraction companies who commission the 
surveys or purchase the data. And fifth, overall, up to five small 
businesses (NAICS 2111) per year may experience increased costs of 
between 0.1 and 0.7 percent of average annual revenues.
    The draft version of the RIA and the Initial Regulatory Flexibility 
Analysis considered effects of a more stringent alternative than the 
proposed rule. The more stringent alternative included additional 
shutdown requirements and area closures for surveys, generating costs 
up to 20 percent greater than the proposed rule. NMFS did not elect to 
proceed with these elements of the more stringent alternative in the 
final rule, which reduces the potential for impacts to small 
businesses. NMFS determined that the final rule achieves the statutory 
objectives with a lower regulatory burden. As described above, a 
relatively small portion of total survey activities are undertaken by 
small entities and the FRFA determines that it is unlikely that small 
entities will bear the compliance costs described in the RIA.
    This final rule revises the information collection request (ICR) 
requirement associated with OMB Control Number 0648-0151 to allow for 
the expected increase in applicants/respondents due to this final 
action. This revision is subject to review and approval by OMB under 
the Paperwork Reduction Act (PRA) and has been submitted to OMB. NMFS 
published a 30-day Federal Register notice (85 FR 60765; September 28, 
2020) that provided for an additional comment period. Details on the 
new information collection requirements can be found in the RIA 
Appendix C.2. NMFS anticipates that 95 to 151 geophysical surveys will 
take place annually on average over the five years of the regulations 
in the GOM that would be subject to potential information collection 
requirements. Due to this final rule, NMFS estimates at least 95 new 
LOA applications annually. Because the existing OMB Control Number 
0648-0151 expires less than a year (June 30, 2021) after this final 
rule publishes, there will be less than a year for respondents to carry 
out work under these regulations before this OMB Control Number 
expires. Thus, NMFS estimates no more than one-quarter of respondents 
(24) will complete work to the point of developing an annual report 
prior to when 0648-0151 must be renewed.
    We invite the general public and other Federal agencies to comment 
on proposed and continuing information collections, which helps us 
assess the impact of our information collection requirements and 
minimize the public's reporting burden. Written comments and 
recommendations for this information collection should be submitted at 
the following website: www.reginfo.gov/public/do/PRAMain. Find this 
particular information collection by using the search function and 
entering either the title of the collection or the OMB Control Number 
0648-0151.
    Notwithstanding any other provision of the law, no person is 
required to respond to, nor shall any person be subject to a penalty 
for failure to comply with, a collection of information subject to the 
requirements of the PRA, unless that collection of information displays 
a currently valid OMB Control Number.

List of Subjects in 50 CFR Part 217

    Exports, Fish, Imports, Indians, Labeling, Marine mammals, 
Penalties, Reporting and recordkeeping requirements, Seafood, 
Transportation.

    Dated: December 7, 2020.
Samuel D. Rauch III,
Deputy Assistant Administrator for Regulatory Programs, National Marine 
Fisheries Service.

    For reasons set forth in the preamble, 50 CFR part 217 is amended 
as follows:

PART 217--REGULATIONS GOVERNING THE TAKING AND IMPORTING OF MARINE 
MAMMALS

0
1. The authority citation for part 217 continues to read as follows:

    Authority:  16 U.S.C. 1361 et seq.


0
2. Add Subpart S, consisting of Sec. Sec.  217.180 through 217.189, to 
read as follows:
Subpart S--Taking Marine Mammals Incidental to Geophysical Survey 
Activities in the Gulf of Mexico
Sec.
217.180 Specified activity and specified geographical region.
217.181 Effective dates.
217.182 Permissible methods of taking.
217.183 Prohibitions.
217.184 Mitigation requirements.
217.185 Requirements for monitoring and reporting.
217.186 Letters of Authorization.
217.187 Renewals and modifications of Letters of Authorization.
217.188 [Reserved]
217.189 [Reserved]

Subpart S--Taking Marine Mammals Incidental to Geophysical Survey 
Activities in the Gulf of Mexico


Sec.  217.180  Specified activity and specified geographical region.

    (a) Regulations in this subpart apply only to oil and gas industry 
operators (LOA-holders), and those persons authorized to conduct 
activities on their behalf, for the taking of marine mammals that 
occurs in the area outlined in paragraph (b) of this section and that 
occurs incidental to geophysical survey activities.
    (b) The taking of marine mammals by oil and gas industry operators 
may be authorized in a Letter of Authorization (LOA) only if it occurs 
within U.S. waters in the Gulf of Mexico, outside the area subject to a 
Congressional leasing moratorium under the Gulf of Mexico Energy 
Security Act (GOMESA) (Pub L. 109-432, Sec.  104) as of the effective 
date of these regulations.


Sec.  217.181  Effective dates.

    Regulations in this subpart are effective from April 19, 2021 
through April 19, 2026.


Sec.  217.182  Permissible methods of taking.

    Under LOAs issued pursuant to Sec. Sec.  216.106 of this chapter 
and 217.186, LOA-holders may incidentally, but not intentionally, take 
marine mammals within the area described in Sec.  217.180(b) by Level A 
and Level B harassment associated with geophysical survey activities, 
provided the activity is in compliance with all terms, conditions, and 
requirements of the regulations in this subpart and the appropriate 
LOA.


Sec.  217.183  Prohibitions.

    Notwithstanding takings contemplated in Sec. Sec.  217.180 and 
217.182, and authorized by a LOA issued under Sec. Sec.  216.106 of 
this chapter and 217.186, no person in connection with the activities 
described in Sec.  217.180 may:
    (a) Violate, or fail to comply with, the terms, conditions, and 
requirements of this subpart or a LOA issued under Sec. Sec.  216.106 
of this chapter and 217.186;
    (b) Take any marine mammal not specified in such LOAs;
    (c) Take any marine mammal specified in such LOAs in any manner 
other than as specified; or
    (d) Take a marine mammal specified in such LOAs if NMFS determines 
such

[[Page 5445]]

taking results in more than a negligible impact on the species or 
stocks of such marine mammal.


Sec.  217.184  Mitigation requirements.

    When conducting the activities identified in Sec.  217.180, the 
mitigation measures contained in any LOA issued under Sec. Sec.  
216.106 of this chapter and 217.186 must be implemented. These 
mitigation measures shall include but are not limited to:
    (a) General conditions. (1) A copy of any issued LOA must be in the 
possession of the LOA-holder, vessel operator, other relevant 
personnel, the lead protected species observer (PSO), and any other 
relevant designees operating under the authority of the LOA.
    (2) The LOA-holder must instruct relevant vessel personnel with 
regard to the authority of the protected species monitoring team (PSO 
team), and must ensure that relevant vessel personnel and PSO team 
participate in a joint onboard briefing, led by the vessel operator and 
lead PSO, prior to beginning work to ensure that responsibilities, 
communication procedures, protected species monitoring protocols, 
operational procedures, and LOA requirements are clearly understood. 
This briefing must be repeated when relevant new personnel join the 
survey operations before work involving those personnel commences.
    (3) The acoustic source must be deactivated when not acquiring data 
or preparing to acquire data, except as necessary for testing. 
Unnecessary use of the acoustic source must be avoided. For surveys 
using airgun arrays as the acoustic source, notified operational 
capacity (i.e., total array volume) (not including redundant backup 
airguns) must not be exceeded during the survey, except where 
unavoidable for source testing and calibration purposes. All occasions 
where activated source volume exceeds notified operational capacity 
must be communicated to the PSO(s) on duty and fully documented. The 
lead PSO must be granted access to relevant instrumentation documenting 
acoustic source power and/or operational volume.
    (4) PSOs must be used as specified in this paragraph (a)(4).
    (i) LOA-holders must use independent, dedicated, qualified PSOs, 
meaning that the PSOs must be employed by a third-party observer 
provider, must have no tasks other than to conduct observational 
effort, collect data, and communicate with and instruct relevant vessel 
crew with regard to the presence of protected species and mitigation 
requirements (including brief alerts regarding maritime hazards), and 
must be qualified pursuant to Sec.  217.185(a) (except as specified at 
Sec.  217.184(d)(2)(iii-iv)). Acoustic PSOs are required to complete 
specialized training for operating passive acoustic monitoring (PAM) 
systems and are encouraged to have familiarity with the vessel on which 
they will be working. PSOs may act as both acoustic and visual 
observers (but not simultaneously), so long as they demonstrate that 
their training and experience are sufficient to perform each task.
    (ii) The LOA-holder must submit PSO resumes for NMFS review and 
approval prior to commencement of the survey (except as specified at 
Sec.  217.184(d)(2)(iii)). Resumes should include dates of training and 
any prior NMFS approval, as well as dates and description of last 
experience, and must be accompanied by information documenting 
successful completion of an acceptable training course. NMFS is allowed 
one week to approve PSOs from the time that the necessary information 
is received by NMFS, after which PSOs meeting the minimum requirements 
will automatically be considered approved.
    (iii) At least one visual PSO and two acoustic PSOs (when required) 
aboard each acoustic source vessel must have a minimum of 90 days at-
sea experience working in those roles, respectively, with no more than 
eighteen months elapsed since the conclusion of the at-sea experience 
(except as specified at Sec.  217.184(d)(2)(iii)). One visual PSO with 
such experience must be designated as the lead for the entire PSO team. 
The lead must coordinate duty schedules and roles for the PSO team and 
serve as the primary point of contact for the vessel operator. (Note 
that the responsibility of coordinating duty schedules and roles may 
instead be assigned to a shore-based, third-party monitoring 
coordinator.) To the maximum extent practicable, the lead PSO must 
devise the duty schedule such that experienced PSOs are on duty with 
those PSOs with appropriate training but who have not yet gained 
relevant experience.
    (b) Deep penetration surveys. (1) Deep penetration surveys are 
defined as surveys using airgun arrays with total volume greater than 
1,500 in\3\.
    (2) Visual monitoring must be conducted as specified in this 
paragraph (b)(2).
    (i) During survey operations (i.e., any day on which use of the 
acoustic source is planned to occur, and whenever the acoustic source 
is in the water, whether activated or not), a minimum of two PSOs must 
be on duty and conducting visual observations at all times during 
daylight hours (i.e., from 30 minutes prior to sunrise through 30 
minutes following sunset).
    (ii) Visual monitoring must begin not less than 30 minutes prior to 
ramp-up and must continue until one hour after use of the acoustic 
source ceases or until 30 minutes past sunset.
    (iii) Visual PSOs must coordinate to ensure 360[deg] visual 
coverage around the vessel from the most appropriate observation posts, 
and must conduct visual observations using binoculars and the naked eye 
while free from distractions and in a consistent, systematic, and 
diligent manner.
    (iv) Visual PSOs must immediately communicate all observations of 
marine mammals to the on-duty acoustic PSO, including any determination 
by the PSO regarding species identification, distance, and bearing and 
the degree of confidence in the determination.
    (v) Any observations of marine mammals by crew members aboard any 
vessel associated with the survey must be relayed to the PSO team.
    (vi) During good conditions (e.g., daylight hours; Beaufort sea 
state (BSS) 3 or less), visual PSOs must conduct observations when the 
acoustic source is not operating for comparison of sighting rates and 
behavior with and without use of the acoustic source and between 
acquisition periods, to the maximum extent practicable.
    (vii) Visual PSOs may be on watch for a maximum of two consecutive 
hours followed by a break of at least one hour between watches and may 
conduct a maximum of 12 hours of observation per 24-hour period. NMFS 
may grant an exception for LOA applications that demonstrate such a 
``two hours on/one hour off'' duty cycle is not practicable, in which 
case visual PSOs will be subject to a maximum of four consecutive hours 
on watch followed by a break of at least two hours between watches. 
Combined observational duties (visual and acoustic but not at the same 
time) must not exceed 12 hours per 24-hour period for any individual 
PSO.
    (3) Acoustic monitoring must be conducted as specified in this 
paragraph (b)(3).
    (i) All source vessels must use a towed PAM system at all times 
when operating in waters deeper than 100 m, which must be monitored by 
a minimum of one acoustic PSO beginning at least 30 minutes prior to 
ramp-up, at all times during use of the acoustic source, and until one 
hour after use of the acoustic source ceases. ``PAM system'' refers to 
calibrated hydrophone arrays with full system redundancy to

[[Page 5446]]

detect, identify, and estimate distance and bearing to vocalizing 
cetaceans, coupled with appropriate software to aid monitoring and 
listening by a PAM operator skilled in bioacoustics analysis and 
computer system specifications capable of running appropriate software. 
The PAM system must have at least one calibrated hydrophone (per each 
deployed hydrophone type and/or set) sufficient for determining whether 
background noise levels on the towed PAM system are sufficiently low to 
meet performance expectations. Applicants must provide a PAM plan 
including description of the hardware and software proposed for use 
prior to proceeding with any survey where PAM is required.
    (ii) Acoustic PSOs must immediately communicate all detections of 
marine mammals to visual PSOs (when visual PSOs are on duty), including 
any determination by the PSO regarding species identification, 
distance, and bearing, and the degree of confidence in the 
determination.
    (iii) Acoustic PSOs may be on watch for a maximum of four 
consecutive hours followed by a break of at least two hours between 
watches, and may conduct a maximum of 12 hours of observation per 24-
hour period. Combined observational duties (visual and acoustic but not 
at the same time) must not exceed 12 hours per 24-hour period for any 
individual PSO.
    (iv) Survey activity may continue for 30 minutes when the PAM 
system malfunctions or is damaged, while the PAM operator diagnoses the 
issue. If the diagnosis indicates that the PAM system must be repaired 
to solve the problem, operations may continue for an additional two 
hours without acoustic monitoring during daylight hours only under the 
following conditions:
    (A) Sea state is less than or equal to BSS 4;
    (B) No marine mammals (excluding delphinids) detected solely by PAM 
in the applicable exclusion zone in the previous two hours;
    (C) NMFS is notified via email as soon as practicable with the time 
and location in which operations began occurring without an active PAM 
system; and
    (D) Operations with an active acoustic source, but without an 
operating PAM system, do not exceed a cumulative total of four hours in 
any 24-hour period.
    (4) PSOs must establish and monitor applicable exclusion and buffer 
zones. These zones must be based upon the radial distance from the 
edges of the airgun array (rather than being based on the center of the 
array or around the vessel itself). During use of the acoustic source 
(i.e., anytime the acoustic source is active, including ramp-up), 
occurrence of marine mammals within the relevant buffer zone (but 
outside the exclusion zone) should be communicated to the operator to 
prepare for the potential shutdown of the acoustic source.
    (i) Two exclusion zones are defined, depending on the species and 
context. A standard exclusion zone encompassing the area at and below 
the sea surface out to a radius of 500 meters from the edges of the 
airgun array (0-500 m) is defined. For special circumstances (defined 
at Sec.  217.184(b)(9)(v)), the exclusion zone encompasses an extended 
distance of 1,500 meters (0-1,500 m).
    (ii) During pre-start clearance monitoring (i.e., before ramp-up 
begins), the buffer zone acts as an extension of the exclusion zone in 
that observations of marine mammals within the buffer zone would also 
preclude airgun operations from beginning (i.e., ramp-up). For all 
marine mammals (except where superseded by the extended 1,500-m 
exclusion zone), the buffer zone encompasses the area at and below the 
sea surface from the edge of the 0-500 meter exclusion zone out to a 
radius of 1,000 meters from the edges of the airgun array (500-1,000 
m). The buffer zone is not applicable when the exclusion zone is 
greater than 500 meters, i.e., the observational focal zone is not 
increased beyond 1,500 meters.
    (5) A ramp-up procedure, involving a step-wise increase in the 
number of airguns firing and total active array volume until all 
operational airguns are activated and the full volume is achieved, is 
required at all times as part of the activation of the acoustic source. 
A 30-minute pre-start clearance observation period must occur prior to 
the start of ramp-up. The LOA-holder must adhere to the following pre-
start clearance and ramp-up requirements:
    (i) The operator must notify a designated PSO of the planned start 
of ramp-up as agreed upon with the lead PSO; the notification time 
should not be less than 60 minutes prior to the planned ramp-up.
    (ii) Ramp-ups must be scheduled so as to minimize the time spent 
with source activated prior to reaching the designated run-in.
    (iii) A designated PSO must be notified again immediately prior to 
initiating ramp-up procedures and the operator must receive 
confirmation from the PSO to proceed.
    (iv) Ramp-up must not be initiated if any marine mammal is within 
the applicable exclusion or buffer zone. If a marine mammal is observed 
within the exclusion zone or the buffer zone during the 30-minute pre-
start clearance period, ramp-up must not begin until the animal(s) has 
been observed exiting the zones or until an additional time period has 
elapsed with no further sightings (15 minutes for small delphinids and 
30 minutes for all other species).
    (v) Ramp-up must begin by activating a single airgun of the 
smallest volume in the array and shall continue in stages by doubling 
the number of active elements at the commencement of each stage, with 
each stage of approximately the same duration. Total duration must not 
be less than 20 minutes. The operator must provide information to the 
PSO documenting that appropriate procedures were followed.
    (vi) Ramp-up must cease and the source shut down upon observation 
of marine mammals within the applicable exclusion zone. Once ramp-up 
has begun, observations of marine mammals within the buffer zone do not 
require shutdown.
    (vii) Ramp-up may occur at times of poor visibility, including 
nighttime, if appropriate acoustic monitoring has occurred with no 
detections of a marine mammal other than delphinids in the 30 minutes 
prior to beginning ramp-up. Acoustic source activation may only occur 
at night where operational planning cannot reasonably avoid such 
circumstances.
    (viii) If the acoustic source is shut down for brief periods (i.e., 
less than 30 minutes) for reasons other than implementation of 
prescribed mitigation (e.g., mechanical difficulty), it may be 
activated again without ramp-up if PSOs have maintained constant visual 
and/or acoustic observation and no visual or acoustic detections of any 
marine mammal have occurred within the applicable exclusion zone. For 
any longer shutdown, pre-start clearance observation and ramp-up are 
required. For any shutdown at night or in periods of poor visibility 
(e.g., BSS 4 or greater), ramp-up is required, but if the shutdown 
period was brief and constant observation maintained, pre-start 
clearance watch is not required.
    (ix) Testing of the acoustic source involving all elements requires 
ramp-up. Testing limited to individual source elements or strings does 
not require ramp-up but does require the pre-start clearance 
observation period.
    (6) Shutdowns must be implemented as specified in this paragraph 
(b)(6).
    (i) Any PSO on duty has the authority to delay the start of survey 
operations or to call for shutdown of the acoustic source pursuant to 
the requirements of this subpart.

[[Page 5447]]

    (ii) The operator must establish and maintain clear lines of 
communication directly between PSOs on duty and crew controlling the 
acoustic source to ensure that shutdown commands are conveyed swiftly 
while allowing PSOs to maintain watch.
    (iii) When both visual and acoustic PSOs are on duty, all 
detections must be immediately communicated to the remainder of the on-
duty PSO team for potential verification of visual observations by the 
acoustic PSO or of acoustic detections by visual PSOs.
    (iv) When the airgun array is active (i.e., anytime one or more 
airguns is active, including during ramp-up) and (1) a marine mammal 
appears within or enters the applicable exclusion zone and/or (2) a 
marine mammal (excluding delphinids) is detected acoustically and 
localized within the applicable exclusion zone, the acoustic source 
must be shut down. When shutdown is called for by a PSO, the acoustic 
source must be immediately deactivated and any dispute resolved only 
following deactivation.
    (v) The extended 1,500-m exclusion zone must be applied upon 
detection (visual or acoustic) of a baleen whale, sperm whale, beaked 
whale, or Kogia spp. within the zone.
    (vi) Shutdown requirements are waived for dolphins of the following 
genera: Tursiops, Stenella, Steno, and Lagenodelphis. If a delphinid is 
visually detected within the exclusion zone, no shutdown is required 
unless the PSO confirms the individual to be of a genus other than 
those listed above, in which case a shutdown is required. Acoustic 
detection of delphinids does not require shutdown.
    (vii) If there is uncertainty regarding identification or 
localization, PSOs may use best professional judgment in making the 
decision to call for a shutdown.
    (viii) Upon implementation of shutdown, the source may be 
reactivated after the marine mammal(s) has been observed exiting the 
applicable exclusion zone or following a 30-minute clearance period 
with no further detection of the marine mammal(s).
    (c) Shallow penetration surveys. (1) Shallow penetration surveys 
are defined as surveys using airgun arrays with total volume equal to 
or less than 1,500 in\3\, single airguns, boomers, or equivalent 
sources.
    (2) LOA-holders conducting shallow penetration surveys must follow 
the requirements defined for deep penetration surveys at Sec.  
217.184(b), with the following exceptions:
    (i) Acoustic monitoring is not required for shallow penetration 
surveys.
    (ii) Ramp-up for small airgun arrays must follow the procedure 
described above for large airgun arrays, but may occur over an 
abbreviated period of time. Ramp-up is not required for surveys using 
only a single airgun. For non-airgun sources, power should be increased 
as feasible to effect a ramp-up.
    (iii) Two exclusion zones are defined, depending on the species and 
context. A standard exclusion zone encompassing the area at and below 
the sea surface out to a radius of 100 meters from the edges of the 
airgun array (if used) or from the acoustic source (0-100 m) is 
defined. For special circumstances (Sec.  217.184(b)(6)(v)), the 
exclusion zone encompasses an extended distance of 500 meters (0-500 
m).
    (iv) The buffer zone encompasses the area at and below the sea 
surface from the edge of the 0-100 meter exclusion zone out to a radius 
of 200 meters from the edges of the airgun array (if used) or from the 
acoustic source (100-200 meters). The buffer zone is not applicable 
when the exclusion zone is greater than 100 meters.
    (d) High-resolution geophysical (HRG) surveys. (1) HRG surveys are 
defined as surveys using an electromechanical source that operates at 
frequencies less than 180 kHz, other than those defined at Sec.  
217.184(c)(1) (e.g., side-scan sonar, multibeam echosounder, or chirp 
sub-bottom profiler).
    (2) LOA-holders conducting HRG surveys must follow the requirements 
defined for shallow penetration surveys at Sec.  217.184(c), with the 
following exceptions:
    (i) No shutdowns are required for HRG surveys. Pre-start clearance 
watch is required as defined at Sec.  217.184(c), i.e., for a period of 
30 minutes and over a 200-m radius from the acoustic source.
    (ii) During survey operations (e.g., any day on which use of the 
acoustic source is planned to occur, and whenever the acoustic source 
is in the water, whether activated or not), a minimum of one trained 
and experienced independent PSO must be on duty and conducting visual 
observations at all times during daylight hours (i.e., from 30 minutes 
prior to sunrise through 30 minutes following sunset) when operating in 
waters deeper than 100 m.
    (iii) When operating in waters shallower than 100 m, LOA-holders 
must employ one trained visual PSO, who may be a crew member, only for 
purposes of conducting pre-start clearance monitoring. If PSOs are crew 
members, i.e., are not independent PSOs, the PSOs are not subject to 
NMFS' approval. In these circumstances, LOA requests must describe the 
training that will be provided to crew members filling the role of PSO.
    (iv) PSOs are not required during survey operations in which the 
active acoustic source(s) are deployed on an autonomous underwater 
vehicle.
    (e) Time-area closure. From January 1 through May 31, no use of 
airguns may occur shoreward of the 20-m isobath and between 90-84[deg] 
W.
    (f) Entanglement avoidance. To avoid the risk of entanglement, LOA-
holders conducting surveys using ocean-bottom nodes or similar gear 
must:
    (1) Use negatively buoyant coated wire-core tether cable;
    (2) Retrieve all lines immediately following completion of the 
survey; and
    (3) Attach acoustic pingers directly to the coated tether cable; 
acoustic releases should not be used.
    (g) Vessel strike avoidance. LOA-holders must adhere to the 
following requirements:
    (1) Vessel operators and crews must maintain a vigilant watch for 
all marine mammals and must slow down, stop their vessel, or alter 
course, as appropriate and regardless of vessel size, to avoid striking 
any marine mammal. A visual observer aboard the vessel must monitor a 
vessel strike avoidance zone around the vessel, which shall be defined 
according to the parameters stated in this subsection. Visual observers 
monitoring the vessel strike avoidance zone may be third-party 
observers (i.e., PSOs) or crew members, but crew members responsible 
for these duties must be provided sufficient training to distinguish 
marine mammals from other phenomena and broadly to identify a marine 
mammal as a baleen whale, sperm whale, or other marine mammal;
    (2) Vessel speeds must be reduced to 10 kn or less when mother/calf 
pairs, pods, or large assemblages of marine mammals are observed near a 
vessel;
    (3) All vessels must maintain a minimum separation distance of 500 
m from baleen whales;
    (4) All vessels must maintain a minimum separation distance of 100 
m from sperm whales;
    (5) All vessels must, to the maximum extent practicable, attempt to 
maintain a minimum separation distance of 50 m from all other marine 
mammals, with an exception made for those animals that approach the 
vessel; and
    (6) When marine mammals are sighted while a vessel is underway, the 
vessel must take action as necessary to avoid violating the relevant 
separation distance, e.g., attempt to remain parallel

[[Page 5448]]

to the animal's course, avoid excessive speed or abrupt changes in 
direction until the animal has left the area. If marine mammals are 
sighted within the relevant separation distance, the vessel must reduce 
speed and shift the engine to neutral, not engaging the engines until 
animals are clear of the area. This does not apply to any vessel towing 
gear or any vessel that is navigationally constrained.
    (7) These requirements do not apply in any case where compliance 
would create an imminent and serious threat to a person or vessel or to 
the extent that a vessel is restricted in its ability to maneuver and, 
because of the restriction, cannot comply.


Sec.  217.185  Requirements for monitoring and reporting.

    (a) PSO qualifications. (1) PSOs must successfully complete 
relevant, acceptable training, including completion of all required 
coursework and passing (80 percent or greater) a written and/or oral 
examination developed for the training program.
    (2) PSOs must have successfully attained a bachelor's degree from 
an accredited college or university with a major in one of the natural 
sciences, a minimum of 30 semester hours or equivalent in the 
biological sciences, and at least one undergraduate course in math or 
statistics. The educational requirements may be waived if the PSO has 
acquired the relevant skills through alternate experience. Requests for 
such a waiver must be submitted to NMFS and shall include written 
justification. Requests will be granted or denied (with justification) 
by NMFS within one week of receipt of submitted information. Alternate 
experience that may be considered includes, but is not limited to:
    (i) Secondary education and/or experience comparable to PSO duties;
    (ii) Previous work experience conducting academic, commercial, or 
government-sponsored marine mammal surveys; or
    (iii) Previous work experience as a PSO; the PSO should demonstrate 
good standing and consistently good performance of PSO duties.
    (b) Equipment. LOA-holders are required to:
    (i) Provide PSOs with bigeye binoculars (e.g., 25 x 150; 2.7 view 
angle; individual ocular focus; height control) of appropriate quality 
solely for PSO use. These must be pedestal-mounted on the deck at the 
most appropriate vantage point that provides for optimal sea surface 
observation, PSO safety, and safe operation of the vessel.
    (ii) For each vessel required to use a PAM system, provide a PAM 
system that has been verified and tested by an experienced acoustic PSO 
who will be using it during the trip for which monitoring is required;
    (iii) Work with the selected third-party observer provider to 
ensure PSOs have all equipment (including backup equipment) needed to 
adequately perform necessary tasks, including accurate determination of 
distance and bearing to observed marine mammals. (Equipment specified 
in A. through G. below may be provided by an individual PSO, the third-
party observer provider, or the LOA-holder, but the LOA-holder is 
responsible for ensuring PSOs have the proper equipment required to 
perform the duties specified herein.) Such equipment, at a minimum, 
must include:
    (A) Reticle binoculars (e.g., 7 x 50) of appropriate quality (at 
least one per PSO, plus backups);
    (B) Global Positioning Unit (GPS) (plus backup);
    (C) Digital camera with a telephoto lens (the camera or lens should 
also have an image stabilization system) that is at least 300 mm or 
equivalent on a full-frame single lens reflex (SLR) (plus backup);
    (D) Compass (plus backup);
    (E) Radios for communication among vessel crew and PSOs (at least 
one per PSO, plus backups); and
    (F) Any other tools necessary to adequately perform necessary PSO 
tasks.
    (c) Data collection. PSOs must use standardized electronic data 
forms. PSOs must record detailed information about any implementation 
of mitigation requirements, including the distance of marine mammals to 
the acoustic source and description of specific actions that ensued, 
the behavior of the animal(s), any observed changes in behavior before 
and after implementation of mitigation, and if shutdown was 
implemented, the length of time before any subsequent ramp-up or 
activation of the acoustic source. If required mitigation was not 
implemented, PSOs must record a description of the circumstances. At a 
minimum, the following information should be recorded:
    (1) Vessel names (source vessel and other vessels associated with 
survey), vessel size and type, maximum speed capability of vessel, port 
of origin, and call signs;
    (2) PSO names and affiliations;
    (3) Dates of departures and returns to port with port name;
    (4) Dates of and participants in PSO briefings;
    (5) Dates and times (Greenwich Mean Time) of survey effort and 
times corresponding with PSO effort;
    (6) Vessel location (latitude/longitude) when survey effort began 
and ended and vessel location at beginning and end of visual PSO duty 
shifts;
    (7) Vessel location at 30-second intervals (if software capability 
allows) or 5-minute intervals (if location must be manually recorded);
    (8) Vessel heading and speed at beginning and end of visual PSO 
duty shifts and upon any line change;
    (9) Environmental conditions while on visual survey (at beginning 
and end of PSO shift and whenever conditions changed significantly), 
including Beaufort sea state and any other relevant weather conditions 
including cloud cover, fog, sun glare, and overall visibility to the 
horizon;
    (10) Vessel location when environmental conditions change 
significantly;
    (11) Factors that may have contributed to impaired observations 
during each PSO shift change or as needed as environmental conditions 
change (e.g., vessel traffic, equipment malfunctions);
    (12) Survey activity information, such as acoustic source power 
output while in operation, number and volume of airguns operating in an 
array, tow depth of an acoustic source, and any other notes of 
significance (i.e., pre-start clearance, ramp-up, shutdown, testing, 
shooting, ramp-up completion, end of operations, streamers, etc.); and
    (13) Upon visual observation of a marine mammal, the following 
information:
    (i) Watch status (sighting made by PSO on/off effort, 
opportunistic, crew, alternate vessel/platform);
    (ii) PSO who sighted the animal and PSO location (including height 
above water) at time of sighting;
    (iii) Time of sighting;
    (iv) Vessel coordinates at time of sighting;
    (v) Water depth;
    (vi) Direction of vessel's travel (compass direction);
    (vii) Speed of the vessel(s) from which the observation was made;
    (viii) Direction of animal's travel relative to the vessel;
    (ix) Pace of the animal;
    (x) Estimated distance to the animal (and method of estimating 
distance) and its heading relative to vessel at initial sighting;
    (xi) Identification of the animal (e.g., genus/species, lowest 
possible taxonomic level, or unidentified), PSO confidence in 
identification, and the composition of the group if there is a mix of 
species;

[[Page 5449]]

    (xii) Estimated number of animals (high/low/best);
    (xiii) Estimated number of animals by cohort (adults, juveniles, 
group composition, etc.);
    (xiv) Description (as many distinguishing features as possible of 
each individual seen, including length, shape, color, pattern, scars or 
markings, shape and size of dorsal fin, shape of head, and blow 
characteristics);
    (xv) Detailed behavior observations (e.g., number of blows/breaths, 
number of surfaces, breaching, spyhopping, diving, feeding, traveling; 
as explicit and detailed as possible; note any observed changes in 
behavior), including an assessment of behavioral responses to survey 
activity;
    (xvi) Animal's closest point of approach (CPA) and/or closest 
distance from any element of the acoustic source;
    (xvii) Platform activity at time of sighting (e.g., deploying, 
recovering, testing, shooting, data acquisition, other); and
    (xviii) Description of any actions implemented in response to the 
sighting (e.g., delays, shutdown, ramp-up) and time and location of the 
action.
    (12) Upon acoustic detection of a marine mammal using a PAM system, 
the following information:
    (i) An acoustic encounter identification number, and whether the 
detection was linked with a visual sighting;
    (ii) Date and time when first and last heard;
    (iii) Types and nature of sounds heard (e.g., clicks, whistles, 
creaks, burst pulses, continuous, sporadic, strength of signal); and
    (iv) Any additional information recorded such as water depth of the 
hydrophone array, bearing of the animal to the vessel (if 
determinable), species or taxonomic group (if determinable), 
spectrogram screenshot, and any other notable information.
    (d) Reporting. (1) Annual reporting must be submitted as specified 
in this paragraph.
    (i) LOA-holders must submit a summary report to NMFS on all 
activities and monitoring results within 90 days of the completion of 
the survey or expiration of the LOA, whichever comes sooner, and must 
include all information described above under Sec.  217.185(c). If an 
issued LOA is valid for greater than one year, the summary report must 
be submitted on an annual basis.
    (ii) The report must describe activities conducted and sightings of 
marine mammals, must provide full documentation of methods, results, 
and interpretation pertaining to all monitoring, and must summarize the 
dates and locations of survey operations and all marine mammal 
sightings (dates, times, locations, activities, associated survey 
activities, and information regarding locations where the acoustic 
source was used). In addition to the report, all raw observational data 
must be made available to NMFS.
    (iii) For operations requiring the use of PAM, the report must 
include a validation document concerning the use of PAM, which should 
include necessary noise validation diagrams and demonstrate whether 
background noise levels on the PAM deployment limited achievement of 
the planned detection goals. Copies of any vessel self-noise assessment 
reports must be included with the report.
    (iv) The LOA-holder must provide geo-referenced time-stamped vessel 
tracklines for all time periods in which airguns (full array or single) 
were operating. Tracklines must include points recording any change in 
airgun status (e.g., when the airguns began operating, when they were 
turned off). GIS files must be provided in ESRI shapefile format and 
include the UTC date and time, latitude in decimal degrees, and 
longitude in decimal degrees. All coordinates must be referenced to the 
WGS84 geographic coordinate system.
    (v) The draft report must be accompanied by a certification from 
the lead PSO as to the accuracy of the report, and the lead PSO may 
submit directly to NMFS a statement concerning implementation and 
effectiveness of the required mitigation and monitoring.
    (vi) A final report must be submitted within 30 days following 
resolution of any comments on the draft report.
    (2) Comprehensive reporting must be submitted as specified in this 
paragraph. LOA-holders must contribute to the compilation and analysis 
of data for inclusion in an annual synthesis report addressing all data 
collected and reported through annual reporting in each calendar year. 
The synthesis period shall include all annual reports deemed to be 
final by NMFS in a given one-year reporting period. The report must be 
submitted to NMFS within 90 days following the end of a given one-year 
reporting period.
    (e) Reporting of injured or dead marine mammals. (1) In the event 
that personnel involved in the survey activities discover an injured or 
dead marine mammal, the LOA-holder must report the incident to the 
Office of Protected Resources (OPR), NMFS and to the Southeast Regional 
Stranding Network as soon as feasible. The report must include the 
following information:
    (i) Time, date, and location (latitude/longitude) of the first 
discovery (and updated location information if known and applicable);
    (ii) Species identification (if known) or description of the 
animal(s) involved;
    (iii) Condition of the animal(s) (including carcass condition if 
the animal is dead);
    (iv) Observed behaviors of the animal(s), if alive;
    (v) If available, photographs or video footage of the animal(s); 
and
    (vi) General circumstances under which the animal was discovered.
    (2) In the event of a ship strike of a marine mammal by any vessel 
involved in the survey activities, the LOA-holder must report the 
incident to OPR, NMFS and to the Southeast Regional Stranding Network 
as soon as feasible. The report must include the following information:
    (i) Time, date, and location (latitude/longitude) of the incident;
    (ii) Species identification (if known) or description of the 
animal(s) involved;
    (iii) Vessel's speed during and leading up to the incident;
    (iv) Vessel's course/heading and what operations were being 
conducted (if applicable);
    (v) Status of all sound sources in use;
    (vi) Description of avoidance measures/requirements that were in 
place at the time of the strike and what additional measures were 
taken, if any, to avoid strike;
    (vii) Environmental conditions (e.g., wind speed and direction, 
Beaufort sea state, cloud cover, visibility) immediately preceding the 
strike;
    (viii) Estimated size and length of animal that was struck;
    (ix) Description of the behavior of the marine mammal immediately 
preceding and following the strike;
    (x) If available, description of the presence and behavior of any 
other marine mammals immediately preceding the strike;
    (xi) Estimated fate of the animal (e.g., dead, injured but alive, 
injured and moving, blood or tissue observed in the water, status 
unknown, disappeared); and
    (xii) To the extent practicable, photographs or video footage of 
the animal(s).
    (3) For deep penetration surveys, in the event of a live stranding 
(or near-shore atypical milling) event within 50 km of the survey 
operations, where the NMFS stranding network is engaged in herding or 
other interventions to return animals to the water, the Director of 
OPR, NMFS (or designee) will advise the LOA-holder of the need to

[[Page 5450]]

implement shutdown procedures for all active acoustic sources operating 
within 50 km of the stranding. Shutdown procedures for live stranding 
or milling marine mammals include the following:
    (i) If at any time, the marine mammal(s) die or are euthanized, or 
if herding/intervention efforts are stopped, the Director of OPR, NMFS 
(or designee) will advise the LOA-holder that the shutdown around the 
animals' location is no longer needed.
    (ii) Otherwise, shutdown procedures will remain in effect until the 
Director of OPR, NMFS (or designee) determines and advises the LOA-
holder that all live animals involved have left the area (either of 
their own volition or following an intervention).
    (iii) If further observations of the marine mammals indicate the 
potential for re-stranding, additional coordination with the LOA-holder 
will be required to determine what measures are necessary to minimize 
that likelihood (e.g., extending the shutdown or moving operations 
farther away) and to implement those measures as appropriate.
    (4) If NMFS determines that the circumstances of any marine mammal 
stranding found in the vicinity of the activity suggest investigation 
of the association with survey activities is warranted, and an 
investigation into the stranding is being pursued, NMFS will submit a 
written request to the LOA-holder indicating that the following initial 
available information must be provided as soon as possible, but no 
later than 7 business days after the request for information. In the 
event that the investigation is still inconclusive, the investigation 
of the association of the survey activities is still warranted, and the 
investigation is still being pursued, NMFS may provide additional 
information requests, in writing, regarding the nature and location of 
survey operations prior to the time period above.
    (i) Status of all sound source use in the 48 hours preceding the 
estimated time of stranding and within 50 km of the discovery/
notification of the stranding by NMFS; and
    (ii) If available, description of the behavior of any marine 
mammal(s) observed preceding (i.e., within 48 hours and 50 km) and 
immediately after the discovery of the stranding.


Sec.  217.186  Letters of Authorization.

    (a) To incidentally take marine mammals pursuant to these 
regulations, prospective LOA-holders must apply for and obtain an LOA.
    (b) An LOA, unless suspended or revoked, may be effective for a 
period not to exceed the expiration date of these regulations.
    (c) In the event of projected changes to the activity or to 
mitigation and monitoring measures required by an LOA, the LOA-holder 
must apply for and obtain a modification of the LOA as described in 
Sec.  217.187.
    (d) The LOA shall set forth:
    (1) Permissible methods of incidental taking;
    (2) Means of effecting the least practicable adverse impact (i.e., 
mitigation) on the species or stock and its habitat; and
    (3) Requirements for monitoring and reporting.
    (e) Issuance of the LOA shall be based on a determination that the 
level of taking will be consistent with the findings made for the total 
taking allowable under these regulations and a determination that the 
amount of take authorized under the LOA is of no more than small 
numbers.
    (f) For LOA issuance, where either (1) the conclusions put forth in 
an application (e.g., take estimates) are based on analytical methods 
that differ substantively from those used in the development of the 
rule, or (2) the proposed activity or anticipated impacts vary 
substantively in scope or nature from those analyzed for the rule, NMFS 
may publish a notice of proposed LOA in the Federal Register, including 
the associated analysis of the differences, and solicit public comment 
before making a decision regarding issuance of the LOA.
    (g) Notice of issuance or denial of an LOA shall be published in 
the Federal Register within thirty days of a determination.


Sec.  217.187  Renewals and modifications of Letters of Authorization 
(LOA).

    (a) An LOA issued under Sec.  216.106 of this chapter and Sec.  
217.186 for the activity identified in Sec.  217.180 shall be modified 
upon request by the applicant, provided that:
    (1) The proposed specified activity and mitigation, monitoring, and 
reporting measures, as well as the anticipated impacts, are the same as 
those described and analyzed for these regulations (excluding changes 
made pursuant to the adaptive management provision in paragraph (c)(1) 
of this section); and
    (2) NMFS determines that the mitigation, monitoring, and reporting 
measures required by the previous LOA under these regulations were 
implemented.
    (b) For LOA modification requests by the applicant that include 
changes to the activity or the mitigation, monitoring, or reporting 
(excluding changes made pursuant to the adaptive management provision 
in paragraph (c)(1) of this section) that result in more than a minor 
change in the total estimated number of takes (or distribution by 
species or years), NMFS may publish a notice of proposed LOA in the 
Federal Register, including the associated analysis of the change, and 
solicit public comment before issuing the LOA.
    (c) An LOA issued under Sec.  216.106 of this chapter and Sec.  
217.186 for the activity identified in Sec.  217.180 may be modified by 
NMFS under the following circumstances:
    (1) NMFS may modify (including adding or removing measures) the 
existing mitigation, monitoring, or reporting measures (after 
consulting with the LOA-holder regarding the practicability of the 
modifications) if doing so is practicable and creates a reasonable 
likelihood of more effectively accomplishing the goals of the 
mitigation and monitoring set forth in the preamble for these 
regulations;
    (i) Possible sources of data that could contribute to the decision 
to modify the mitigation, monitoring, or reporting measures in an LOA:
    (A) Results from monitoring from previous years;
    (B) Results from other marine mammal and/or sound research or 
studies; and
    (C) Any information that reveals marine mammals may have been taken 
in a manner, extent or number not authorized by these regulations or 
subsequent LOAs.
    (ii) If, through adaptive management, the modifications to the 
mitigation, monitoring, or reporting measures are substantial, NMFS 
will publish a notice of proposed LOA in the Federal Register and 
solicit public comment.
    (2) If NMFS determines that an emergency exists that poses a 
significant risk to the well-being of the species or stocks of marine 
mammals specified in an LOA issued pursuant to Sec.  216.106 of this 
chapter and Sec.  217.186, an LOA may be modified without prior notice 
or opportunity for public comment. Notice would be published in the 
Federal Register within thirty days of the action.


Sec. Sec.  217.188-217.189  [Reserved]

[FR Doc. 2020-27252 Filed 1-15-21; 8:45 am]
BILLING CODE 3510-22-P