Colorado Regulatory Program, 14986-14999 [05-5807]

Agencies

• DEPARTMENT OF THE INTERIOR
• Office of Surface Mining Reclamation and Enforcement
• Surface Mining Reclamation and Enforcement Office

[Federal Register Volume 70, Number 56 (Thursday, March 24, 2005)]
[Rules and Regulations]
[Pages 14986-14999]
From the Federal Register Online via the Government Printing Office [www.gpo.gov]
[FR Doc No: 05-5807]


=======================================================================
-----------------------------------------------------------------------

DEPARTMENT OF THE INTERIOR

Office of Surface Mining Reclamation and Enforcement

30 CFR Part 906

[CO-033-FOR]


Colorado Regulatory Program

AGENCY: Office of Surface Mining Reclamation and Enforcement, Interior.

ACTION: Final rule; approval of amendment.

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

SUMMARY: We are approving an amendment to the Colorado regulatory 
program (the ``Colorado program'') under the Surface Mining Control and 
Reclamation Act of 1977 (SMCRA or the Act). Colorado proposed revisions 
to its rules concerning prime farmland, revegetation, hydrology, 
enforcement, topsoil, historic properties, bond release and permit 
requirements. The State intends to revise its program to be consistent 
with the corresponding Federal regulations, provide additional 
safeguards, clarify ambiguities, and improve operational efficiency.

EFFECTIVE DATE: March 24, 2005.

FOR FURTHER INFORMATION CONTACT: James F. Fulton, Telephone: (303) 844-
1400, extension 1424; Internet address: JFulton@osmre.gov.

SUPPLEMENTARY INFORMATION:

I. Background on the Colorado Program
II. Submission of the Amendment
III. Office of Surface Mining Reclamation and Enforcement's (OSM) 
Findings
IV. Summary and Disposition of Comments
V. OSM's Decision
VI. Procedural Determinations

I. Background on the Colorado Program

    Section 503(a) of the Act permits a State to assume primacy for the 
regulation of surface coal mining and reclamation operations on non-
Federal and non-Indian lands within its borders by demonstrating that 
its State program includes, among other things, ``a State law which 
provides for the regulation of surface coal mining and reclamation 
operations in accordance with the requirements of this Act * * *; and 
rules and regulations consistent with regulations issued by the 
Secretary pursuant to this Act.'' See 30 U.S.C. 1253(a)(1) and (7). On 
the basis of these criteria, the Secretary of the Interior 
conditionally approved the Colorado program on December 15, 1980. You 
can find background information on the Colorado program, including the 
Secretary's findings, the disposition of comments, and conditions of 
approval in the December 15, 1980, Federal Register (45 FR 82173). You 
can also find later actions concerning Colorado's program and program 
amendments at 30 CFR 906.10, 906.15, 906.16, and 906.30.

II. Submission of the Amendment

    By letter dated March 27, 2003, Colorado sent us an amendment to 
its program (Administrative Record No. CO-696-1) under SMCRA (30 U.S.C. 
1201 et seq.). Colorado sent the amendment in response to May 7, 1986, 
June 9, 1987, and March 22, 1990, letters that we sent to it in 
accordance with 30 CFR 732.17(c), as well as to include changes made at 
its own initiative. On April 4, 2003, Colorado sent us an addition to 
its March 27, 2003, amendment. Finally, Colorado submitted to us 
further revisions to its March 27, 2003, amendment on July 23, 2003.

[[Page 14987]]

    We announced receipt of the proposed amendment in the June 3, 2003, 
Federal Register (68 FR 33032). In the same document, we opened the 
public comment period and provided an opportunity for a public hearing 
or meeting on the amendment's adequacy (Administrative Record No. CO-
696-6). We did not hold a public hearing or meeting because no one 
requested one. The public comment period ended on July 3, 2003. We did 
not receive any comments.
    In the November 20, 2003, Federal Register (68 FR 65422), we 
reopened the public comment period to allow for comments on Colorado's 
July 23, 2003, additional submittal which is as follows: Colorado 
recently amended its Noxious Weed Act which necessitated a revision to 
proposed rules 4.15.1(5), Rule 1.04(78), and also amended for 
consistency the earlier version of the draft rules. In addition, the 
earlier proposed revision to Rule 4.15.4 adding (5) was withdrawn. We 
did not receive any comments on the additional submittal.
    Then in the October 1, 2004, Federal Register (69 FR 58873), we 
reopened the public comment period again to allow comments on 
Colorado's July 23, 2003, additional submittal. We received comments 
from the Rocky Mountain Director of ``Public Employees for 
Environmental Responsibility'' (PEER).
    The amendment concerns revegetation, prime farmland, hydrology, 
enforcement, topsoil, historic properties, and bond release 
requirements.

III. OSM's Findings

    Following are the findings we made concerning the amendment under 
SMCRA and the Federal regulations at 30 CFR 732.15 and 732.17.

A. Minor Revisions to Colorado's Rules

    Colorado proposed minor editorial changes to the following 
previously-approved rules.
    1. 2.06.8(4)(a)(i) and (5)(b)(i), Alluvial Valley Floors;
    2. 2.06.8(5)(b)(i), Permit approval or denial;
    3. 2.07.6(1)(a)(ii), Permit review;
    4. 2.07.6(2)(n), Criteria for permit approval or denial;
    5. 2.08.4(6)(c)(iii), Minor revision;
    6. 3.03.2(5)(a), Decision by the Division; and
    7. 4.03.1(4)(e), Culverts and bridges.
    Because these changes are minor, we find that they will not make 
Colorado's rules less effective than the corresponding Federal 
regulations.

B. Revisions to Colorado's Rules That Have the Same Meaning as the 
Corresponding Provisions of the Federal Regulations

    Colorado proposed revisions to the following rules containing 
language that is the same as or similar to the corresponding sections 
of the Federal regulations.
    1. Rule 2.06.6(2)(a) and (g), [30 CFR 785.17(c)((1)], Prime 
farmland soil survey;
    2. Rule 3.03.2(1)(e), [30 CFR 800.40(a)(3)], Release of performance 
bonds;
    3. Rule 4.05.2(2), [30 CFR 816/817.46(b)(5), Sedimentation pond 
removal;
    4. Rule 4.15.7(2), [30 CFR 780.18(b)(5)(vi),780.13(b)(5)(vi)], 
Revegetation monitoring plan;
    5. Rule 4.15.8(3)(a), [30 CFR 816/817.116(a)(2)], Ground cover 
standard;
    6. Rule 4.15.8(4), [30 CFR 816/817.116(a)(2)], Production standard;
    7. Rule 4.15.8(8), [30 CFR 816/817.116(b)(3)], Forestry success 
standards; and
    8. Rule 4.25.2(4), [30 CFR 785.17(e)(5)], Prime Farmland issuance 
of permit.
    Because these proposed rules contain language that is the same as 
or similar to the corresponding Federal regulations, we find that they 
are no less effective than the corresponding Federal regulations.

C. Revisions of Colorado's Rules That Are Not the Same as the 
Corresponding Provisions of the Federal Regulations

1. Rule 4.15.1(5), Revegetation--Weed Control and 1.04(78), Noxious 
Weeds
    The Federal regulations at 30 CFR 816/817.111(b)(5) require that 
the reestablished plant species shall meet the requirements of 
applicable State and Federal seed, poisonous and noxious plant, and 
introduced species laws or regulations.
    The Federal definition of noxious plants at 30 CFR 701.5 means 
species that have been included on official State lists of noxious 
plants for the State in which the surface coal mining and reclamation 
operation occurs.
    Colorado is adding a new rule requiring a weed management plan. The 
plan is designed to deal with noxious weeds and other weed species that 
could threaten development of the desired vegetation.
    While there is no direct Federal counterpart to the proposed rule, 
it implements the Federal requirement at 30 CFR 816/817.111(b)(5) and, 
as proposed, is no less effective than the Federal regulation.
2. Rule 4.15.7(1), Determining Revegetation Success
    The Federal regulations at 30 CFR 816/817.116(a)(1) require that 
standards for success and statistically valid sampling techniques for 
measuring success shall be selected by the regulatory authority and 
included in an approved regulatory program. The proposed revision 
simply adds a reference to ``the techniques identified in these 
rules.''
    By revising 4.15.7(1) as proposed, along with the other changes 
proposed in this amendment, Colorado is including standards for success 
and statistically valid sampling techniques for measuring success in 
its approved regulatory program. This is consistent with and no less 
effective than the Federal regulations. Specific standards and 
techniques are addressed in other Findings in this document.
3. Rule 4.15.7(3)(b), Use of Reference Areas
    The Federal regulations at 30 CFR 816/817.116(a)(1) require that 
standards for success and statistically valid sampling techniques for 
measuring success shall be selected by the regulatory authority and 
included in an approved regulatory program.
    The Federal regulations at 30 CFR 816/817.116(a)(2) require that 
standards for success shall include criteria representative of unmined 
lands in the area being reclaimed to evaluate the appropriate 
vegetation parameters of ground cover, production, or stocking.
    The Federal regulations at 30 CFR 816/817.116(b) require, in part, 
that (1) for areas developed for use as grazing land or pasture land, 
the ground cover and production of living plants on the revegetated 
area shall be at least equal to that of a reference area or such other 
success standards approved by the regulatory authority; and (2) for 
areas developed for use as cropland, crop production on the revegetated 
area shall be at least equal to that of a reference area or such other 
success standards approved by the regulatory authority.
    In support of its proposal, Colorado proposes to reorganize and 
amend Rule 4.15.7(3)(b) to specify exceptions to the requirement that 
reference areas be demonstrated to be statistically comparable to 
equivalent pre-mine vegetation types in terms of vegetation cover and 
herbaceous production.
    Rule 4.15.7(3)(b)(i) is proposed to be recodified to identify 
cropland post-mine land use as one exception to this requirement. The 
content of the existing rule is not changed by the recodification.

[[Page 14988]]

    Rule 4.15.7(3)(b)(ii) is proposed to be added to identify 
situations in which the post-mining land use will be different than 
pre-mining land use as a second exception to the pre-mine equivalency 
requirement. This amendment is in recognition of the fact that when 
there is a change in land use, such as from forestry or wildlife 
habitat to pasture land or cropland, assumptions upon which the 
traditional reference area concepts are based may no longer be valid or 
applicable. Selection of a reference area that reflects the alternative 
post-mining land use and planned vegetation community structure may be 
a more practical approach in such cases, when suitable areas occur in 
the vicinity of the mine.
    Rule 4.15.7(3)(b)(iii) is added to identify situations in which the 
planned post-mining vegetative community structure will differ 
significantly from the pre-mining vegetative community structure as a 
third exception to the pre-mining equivalency requirement. In such 
cases, Colorado does not require selection of separate reference areas 
representative of each plant community present within the area to be 
disturbed. In these situations, selection of a reference area that 
reflects the planned vegetation community structure may be more 
appropriate and practical than the traditional reference area approach 
when suitable areas are identified in the vicinity of the mine.
    We concur with Colorado's proposal. The use of reference areas 
representative of unmined lands in the area as success standards is in 
compliance with the Federal regulations. The selection of reference 
areas that allow direct comparisons between communities with the same 
postmining land uses or similar plant community structures, rather than 
dissimilar communities, is appropriate and biologically and 
statistically valid. The use of multiple reference areas for developing 
weighted success standards based on relative premine ecological site 
acreages ensures restoration of premine capability. The provision 
requiring the permittee to demonstrate that management of the reference 
area will be under its control and will remain under its control 
throughout the period of extended liability, regardless of location, 
ensures the long-term protection of the reference areas. We have 
reviewed the proposed rule change and have determined it is consistent 
with and no less effective than the Federal regulations at 30 CFR 816/
817.116(a)(2) and (b)(1) and (2).
4. Rule 4.15.7(3)(f), Reference Area Management
    There is no Federal counterpart to this requirement.
    The proposed change to this rule would require equivalent 
management of the reclaimed and reference areas in any year vegetation 
sampling will be conducted. In discussing this proposed change, 
Colorado indicated that rule 4.15.7(3)(f) was amended to be consistent 
with the proposed amendment to rule 4.15.7(5), which will allow 
vegetation sampling in two out of any four consecutive years beginning 
in year nine of the liability period.
    This change is appropriate because it assures that similar 
management will be applied to both the reference and reclaimed areas 
during any year bond release evaluation of vegetation occurs. Moreover, 
the change maintains the statistical validity of any direct comparison. 
The proposed change is consistent with the intent of SMCRA and no less 
effective than the Federal regulations.
5. Rule 4.15.7(4), Use of Reference Areas
    The Federal regulations at 30 CFR 816/817.116(a)(1) require that 
standards for success and statistically valid sampling techniques for 
measuring success shall be selected by the regulatory authority and 
included in an approved regulatory program.
    The Federal regulations at 30 CFR 816/817.116(a)(2) require that 
standards for success shall include criteria representative of unmined 
lands in the area being reclaimed to evaluate the appropriate 
vegetation parameters of ground cover, production, or stocking.
    The Federal regulations at 30 CFR 816/817.116(b) require, in part, 
that (1) for areas developed for use as grazing land or pasture land, 
the ground cover and production of living plants on the revegetated 
area shall be at least equal to that of a reference area or such other 
success standards approved by the regulatory authority; and (2) for 
areas developed for use as cropland, crop production on the revegetated 
area shall be at least equal to that of a reference area or such other 
success standards approved by the regulatory authority. Essentially, 
the revisions to the rule simply address how reference areas may be 
used to determine revegetation success.
    In other words, the proposed revisions to rule 4.15.7(4) provide 
additional guidance in the use of reference areas for the evaluation of 
revegetation success. In discussing the proposed revisions, Colorado 
stated that rule 4.15.7(4) is amended to address reference area 
comparison approaches applicable to each of the reference area types 
identified in proposed rule 4.15.7(3).
    The inclusion of approaches for using established reference areas 
helps further define acceptable success standards for evaluating 
revegetation success. As proposed, the approaches represent valid 
methods for using reference areas. There is no direct Federal 
counterpart to the proposed rule. As proposed, the State rule is 
consistent with and no less effective than the Federal regulations. 
Therefore, we approve it.
6. Rule 4.15.7(5), Timeframes for Demonstration of Revegetation 
Success--Sections of the State Regulation Proposed for Amendment: 
4.15.7(5) and 4.15.9 [30 CFR 816/817.116(c)(3)]
    Colorado proposes in Rule 4.15.7(5) that revegetation success 
criteria shall be met for at least two of the last four years of the 
liability period and that sampling for final revegetation success shall 
not be initiated prior to year nine of the liability period. The 
responsibility period for Colorado is a minimum of ten years, the 
proposed rule thus allows for measurements to occur in any four year 
period beginning in year nine.
    The Federal regulations at 30 CFR 816.116(c)(3), which are 
applicable for areas with less than 26 inches of annual precipitation, 
including Colorado, require that revegetation success standards be met 
during the last two consecutive years of the revegetation 
responsibility period. The major difference between the Federal 
regulations and Colorado's proposal is that Colorado's proposal would 
allow measurement in nonconsecutive years.
    Originally the Federal regulations applicable for areas with 
greater than 26 inches of annual precipitation at 30 CFR 816.116(c)(2) 
required success standards to be met for the last two consecutive years 
of the responsibility period. These regulations were amended (53 FR 
34636, September 7, 1988) to allow the standard to be met during any 
two years of the five year responsibility period excluding the first 
year. The change eliminated the requirement to measure revegetation 
success during the last two (consecutive) years of the responsibility 
period. The basis for the change was that measurements in 
nonconsecutive years avoid unduly penalizing the permittee for negative 
effects of climatic variability.
    We previously approved New Mexico regulations stating ground cover 
and productivity shall equal the approved standard for at least two of 
the last four

[[Page 14989]]

years, starting no sooner than year eight of the responsibility period. 
New Mexico, like Colorado, experiences less than 26 inches of annual 
precipitation. We based our approval on the fact that the climatic 
variability of New Mexico was greater than that in areas with greater 
than 26 inches of precipitation. We stated that we believe it is 
appropriate to avoid penalizing permittees in New Mexico for the 
negative effects of climatic variability (the same reasoning used for 
areas receiving greater than 26 inches of precipitation). See New 
Mexico's approval at 65 FR 65770, November 2, 2000.
    Similar to New Mexico, Colorado submitted climatic data. The 
Colorado mines are located in areas that represent variable 
precipitation ranges as shown on the table below. The data in the 
following table is from the monthly climate data, Colorado Climate 
Center at Colorado State University (https://ccc.atmos.colostate.edu), 
the Trapper Mine Annual Reclamation Report and the Federal Register: 
November 2, 2000 (Volume 65, Number 213), pages 65776-65777.

                                            Historical Precipitation
----------------------------------------------------------------------------------------------------------------
                                                   Precipitation
        Geographical area            Years of          range           Mean          Standard     Coefficient of
                                      record         (inches)                        deviation       variation
----------------------------------------------------------------------------------------------------------------
Trapper Mine....................       1980-2000  ..............           16.56            3.54            0.21
Craig...........................       1937-1974      7.42-20.83           13.29            3.26            0.25
Hayden..........................       1932-1999     10.89-26.40           16.38            3.39            0.21
Trinidad........................       1938-1999      5.42-22.24           13.42            3.36            0.25
Grand Junction..................       1963-1999      5.69-15.02            8.89            3.39            0.29
Henderson, KY...................       1978-1998     30.94-63.27           45.64            8.89            0.19
----------------------------------------------------------------------------------------------------------------

    As seen in the table above, the coefficient of variation (a measure 
of the variability of the data) for the Colorado locations is greater 
than the Henderson, Kentucky location, which is representative of 
conditions in the east. Given the variability in precipitation, a dry 
year may present an obstacle to the second year of revegetation success 
sampling. Flexibility in sampling is needed to skip the drought 
year(s), and allow the operator to sample in one of the two following 
non-consecutive years. A demonstration of successful revegetation 
following a drought would clearly indicate the revegetation could 
withstand drought and the variable climatic conditions. Revegetation 
that is capable of meeting the performance standards both before and 
after a period of drought or pestilence would provide a better 
demonstration of resilience, effectiveness, and permanence than 
revegetation that could meet the standards during two consecutive (and 
fortuitous) years of more or less normal precipitation and damage. The 
likelihood of drought in Colorado needs to be recognized. The proposed 
rule changes ensure that performance standards will be met without 
undue costs or extensions of the ten-year liability period.
    Colorado's proposed rules prohibit the inclusion of measurements 
taken during the first eight years of the responsibility period. This 
ensures that the plants will have the opportunity to become well 
established prior to any evaluation of the vegetation. This also 
provides the same level of flexibility in evaluating revegetation 
success provided by the Federal regulations for States receiving more 
than 26 inches of precipitation. Further, Colorado has asserted that if 
revegetation success were not demonstrated the second year of sampling, 
the operator would be required to take the necessary actions to achieve 
revegetation success. The liability period would then be reinitiated. 
The proposed rules do not affect the length of the extended period of 
responsibility, which is 10 years in Colorado. It should also be 
pointed out that because the proposed rules clearly state that the 
demonstration of success must be done for at least two of the last four 
years, the proposed rules provides the opportunity for requiring 
additional demonstrations as needed.
    The current regulation at 30 CFR 816.116(c)(3)(i) pertaining to 
areas of 26 inches or less average precipitation does provide that 
success equal or exceed the approved success standard during the last 
two consecutive years of the responsibility period. However, the 
preamble to that rule published in the Federal Register on March 23, 
1982, (47 FR 12600) does not provide rationale for measurement of 
revegetation success in consecutive years. OSM does state that for 
areas of less than 26 inches average annual precipitation, because of 
the greater variability in climatic conditions in these Western States, 
especially precipitation, it is difficult to base success on a single 
year's data. Thus, there is support for considering climatic 
variability in measuring revegetation success and for requiring two 
years of success, but not necessarily for consecutive years.
    Colorado's proposed rules at 4.15.7(5) and 4.15.9 are as effective 
as the corresponding Federal regulations at 30 CFR 816.116(c)(3) in 
achieving the revegetation requirements of sections 515(b)(19) and 
(b)(20) of SMCRA.
7. Rule 4.15.7(5)(a)-(f), Normal Husbandry Practices [30 CFR 816/
817.116(c)(4)]
    The Federal regulations at 30 CFR 816.116(c)(1) require that the 
period of extended responsibility for successful revegetation shall 
begin after the last year of augmented seeding, fertilizing, 
irrigation, or other work, excluding husbandry practices that are 
approved by the regulatory authority in accordance with 30 CFR 
816.116(c)(4). The Federal regulations at 30 CFR 816.116(c)(4) require 
that a State may approve selective husbandry practices, excluding 
augmented seeding, fertilization, or irrigation, provided it obtains 
prior approval from us that the practices are normal husbandry 
practices. In addition, a State may also approve selective husbandry 
practices, without extending the period of responsibility for 
revegetation success and bond liability, if such practices can be 
expected to continue as part of the post-mining land use or if 
discontinuance of the practices after the liability period expires will 
not reduce the probability of permanent vegetation success. Approved 
practices shall be normal husbandry practices within the region for 
unmined land having land uses similar to the approved postmining land 
use of the disturbed area, including such practices as disease, pest, 
and vermin control, and any pruning, reseeding, and transplanting 
specifically necessitated by such actions.
    Colorado proposed to add rules identifying normal husbandry 
practices that will not be considered augmented practices and will not 
result in

[[Page 14990]]

restarting the responsibility period. In support of the proposed normal 
husbandry practices, Colorado indicated that several management 
practices are also addressed in this proposed rule. In rule 
4.15.7(5)(a), repair of minor erosion (including revegetation) is 
allowed under certain conditions, to reflect the fact that minor 
erosion affecting limited areas is common during the early stages of 
reclamation, even when appropriate reclamation and stabilization 
measures are applied. The provision specifies that the operator's 
liability period for a reclaimed parcel subject to erosion repair 
extend for a minimum of five years after completion of such repair. 
This will allow the Colorado Division of Minerals and Geology 
(hereinafter DMG or Division) to determine that the repair has been 
successful in stabilizing the area prior to final bond release. 
Documentation of the repair work in the annual reclamation report will 
ensure accurate tracking for bond release purposes.
    In Colorado's proposed rule at 4.15.7(5)(b), weed control measures 
are considered normal husbandry practices provided they are conducted 
in compliance with the Colorado Weed Management Act and the Division's 
Guidelines for Management of Noxious Weeds. A copy of the ``Colorado 
Noxious Weed Act'' [Sec.  5-5.5-115, C.R.S. (1996 Supp.)] and rules 
established pursuant thereto, and a copy of DMG's ``Guideline for the 
Management of Noxious Weeds on Coal Mine Permit Areas'' were included 
in the March 27, 2003, submission by Colorado (see Exhibits A and D).
    Rules 4.15.7(5)(c), (d), and (e) identify specific practices 
recognized as normal husbandry practices for annual crop production, 
perennial cropland, and pasture land forage production, respectively. 
These land uses are characterized by more intensive management than is 
typical of rangeland or wildlife habitat. The Federal regulations 
require that all normal husbandry practices be identified in the 
approved State program.
    Rule 4.15.7(5)(f) limits transplanting to a period within the first 
four years of the ten year liability period. The limitation on the 
number of trees or shrubs transplanted is 20 percent of the approved 
standard. These limitations will insure that transplanting to replace 
initial mortality loss during the liability period is of a limited 
nature and that artificially seeded or transplanted woody plants will 
have been in place for a minimum of six years prior to final bond 
release. Such limited transplanting is a normal husbandry practice 
associated with intensive woody plant establishment efforts such as 
wildlife plantings, windbreaks, etc. The U.S. Department of 
Agriculture's Natural Resources Conservation Service (NRCS) (formerly 
known as the Soil Conservation Service), the Colorado Soil Conservation 
Board, and the Colorado Division of Wildlife (DOW) submitted comments 
supporting this approach (Exhibit F to Colorado's March 27, 2003, State 
Program Amendment submission).
    We consider, on a practice-by-practice basis, the administrative 
record supporting each normal husbandry practice proposed by a 
regulatory authority (53 FR 34641, September 7, 1988). We have also 
provided specific guidance concerning the repair of rills and gullies 
by stating that a regulatory authority could allow the repair of rills 
and gullies as a husbandry practice that would not restart the 
liability period if the general standards of 30 CFR 816.116(c)(4) are 
met, and after consideration of the normal conservation practices 
within the region (48 FR 40157, September 2, 1983).
    In support of the proposed rule at 4.15.7(5)(a), allowing for the 
repair of rills and gullies, Colorado has provided a copy of a letter 
from the State Resource Conservationist with the NRCS. The letter 
clearly supports the repair of rills and gullies as a normal husbandry 
practice.
    We reviewed the proposed normal husbandry practices and supporting 
documentation contained in Exhibit G of Colorado's March 27, 2003, 
submission for weed control, crop management and tree and shrub 
replanting. Exhibit G includes correspondence regarding normal 
husbandry practices and comments received from resource agencies.
    Based on our review, we have determined that Colorado has provided 
sufficient supporting documentation to demonstrate that the normal 
husbandry practices described under rules 4.15.7(5)(a), (b), (c), (d), 
(e) and (f) are acceptable for unmined lands having land uses similar 
to the approved postmining land use of the disturbed area. In addition, 
in (a) and (b), Colorado limits the real extent of affordable repair of 
rills and gullies and weed control measures to no more than five 
percent of the acreage revegetated in any one year. If these limits are 
exceeded, the permittee would be required to restart the liability 
period.
    We have determined that the proposed normal husbandry practices 
meet the criteria to be approved under 30 CFR 816/817.116(c)(4) and are 
no less effective than the Federal regulations.
8. Rule 4.15.7(5)(g), Normal Husbandry Practices--Interseeding [30 CFR 
816/817.116(c)(4)]
    Proposed rule 4.15.7(5) requires, in part, that the liability 
period shall re-initiate whenever augmented seeding, planting, 
fertilization, irrigation, or other augmentive work is required or 
conducted. Colorado proposes that management activities that are not 
augmentive, are approved as normal husbandry practices, and may be 
conducted without re-initiating the liability period.
    At rule 4.15.7(5)(a), Colorado proposed that interseeding is 
considered a normal husbandry practice to enhance species or life form 
diversity on rangeland or wildlife habitat. Interseeding is not an 
allowable substitute for complete reseeding when a stand is dominated 
by species that do not support the approved post mine land use, or when 
vegetation cover is deficient and excessive erosion has resulted. 
Interseeding shall be permitted within the first four years of any ten-
year liability period, upon approval by the Division. The nature, 
location and extent of the interseeding must be fully described in the 
annual reclamation report.
    Colorado defines interseeding as a tool to enhance the diversity of 
established vegetation. Forb, shrub, and grass species native to the 
area are considered acceptable. The exact species to be used depends 
upon the post mining land use. Interseeding only applies to lands where 
vegetation is established and no other management tools are necessary. 
In contrast, augmented seeding is reseeding with fertilizer or 
irrigation, or is in response to an unsuccessful germination and 
establishment. If a reclaimed parcel is deficient in vegetative cover 
due to insufficient moisture, poor germination or improper planting 
methodologies, augmented seeding would be necessary and the ten-year 
liability period would be re-initiated.
    The Federal regulations at 30 CFR 816.116(c)(1) require that the 
period of extended responsibility for successful revegetation shall 
begin after the last year of augmented seeding, fertilizing, 
irrigation, or other work, excluding husbandry practices that are 
approved by the regulatory authority in accordance with 30 CFR 
816.116(c)(4). The Federal regulations at 30 CFR 816.116(c)(4) require 
that a State may approve selective husbandry practices, excluding 
augmented seeding, fertilization, or irrigation, provided it obtains 
prior approval from OSM that the practices are normal husbandry

[[Page 14991]]

practices without extending the period of responsibility for 
revegetation success and bond liability, if such practices can be 
expected to continue as part of the post-mining land use or if 
discontinuance of the practices after the liability period expires will 
not reduce the probability of permanent vegetation success. Approved 
practices shall be normal husbandry practices within the region for 
unmined land having land uses similar to the approved postmining land 
use of the disturbed area, including such practices as disease, pest, 
and vermin control, and any pruning, reseeding, and transplanting 
specifically necessitated by such actions.
    In support of the proposed normal husbandry practice, Colorado 
states that interseeding on rangelands and wildlife habitat is a normal 
husbandry practice recommended by biologists and land managers to 
enhance established vegetation. In Rule 4.15.7(5)(g), the Division is 
proposing the use of interseeding. A. Perry Plummer, in ``Restoring Big 
Game Range in Utah'' (1968) states that ``interseeding (seeding 
directly into established vegetation usually with only partial 
reduction in competition) is a widely successful means of improving 
vegetative cover for game and livestock.'' He indicates that 
interseeding can be an effective means to establish shrubs and forbs in 
perennial grass stands and notes that the approach is especially useful 
on steep slopes where it is desirable to establish shrubs in 
predominantly herbaceous cover.
    Many of the Conservation Reserve Program (CRP) lands in 
northwestern Colorado lack spatial, structural and vegetative 
diversity. To improve the diversity of some grass-dominated CRP lands 
for sharp-tailed grouse habitat, the DOW recommended, ``adding legumes 
and bunchgrasses and reducing sod-forming grasses within these fields 
to enhance the suitability for sharp-tailed grouse.'' Some reclaimed 
lands resemble CRP fields and interseeding is one of the tools DOW 
recommends to improve habitat diversity as documented in the DOW letter 
in Exhibit H of Colorado's March 27, 2003, State Program Amendment 
submission. To further implement this recommendation, the DOW assisted 
with the formation of the Habitat Partnership Program.
    The Habitat Partnership Program is designed to protect and enhance 
the condition of public and private rangeland through the use of 
interseeding technology to modify species composition. Working 
cooperatively together in this program are representatives of the Rio 
Blanco Cooperative Extension Service, Douglas Creek Soil Conservation 
District, the White River Soil Conservation District, the DOW, and the 
NRCS.
    Through funding made available by the DOW, an interseeding drill 
was purchased. The drill is available to landowners based on the 
priority list found in the Habitat Partnership Program Proposal. Of 
highest priority are wildlife forage improvement projects to improve 
wildlife habitat. The DMG believes that the use of interseeding on 
reclaimed lands can enhance the established vegetation similar to CRP 
lands and native rangelands to improve wildlife habitat.
    Additional applicable references include Yoakum et. al. (1980), 
Monsen and Shaw (1983), Frischknecht (1983), and Soil Conservation 
Service (now known as NRCS) ``Range Seeding Standards and 
Specifications for Colorado'' (1987). In this latter reference, NRCS 
limits the practice to the eastern plains. Two coal mines on the 
eastern plains have successfully applied this practice to increase the 
warm season grass cover. Specifically, at the Bacon Mine and at the 
CCMC mine, warm season grasses were interseeded after it became 
apparent that the presence of those grasses was not as high as desired. 
Interseeding was a very effective technique for increasing the warm 
season grass component in the reclaimed community. Both of these mines 
have successfully achieved Phase III bond release criteria.
    In rule 4.15.7(5)(g), Colorado defines interseeding as a tool used 
to enhance the diversity of established vegetation. Forb, shrub, and 
grass species native to the area will be considered acceptable. The 
exact species to be used will depend upon the post mining land use. 
Interseeding will only apply to lands where vegetation is established 
and no other management tools are necessary. In contrast, augmented 
seeding is reseeding with fertilizer or irrigation, or in response to 
an unsuccessful reclaimed parcel. If a reclaimed parcel is deficient in 
vegetative cover due to insufficient moisture, poor germination or 
improper planting methodologies, augmented seeding would be necessary.
    Based on these references and practices, it is clear that in 
certain cases interseeding is desirable to increase the structural and 
vegetative diversity of the reclaimed lands for wildlife habitat and 
for rangeland improvement.
    We consider, on a practice-by-practice basis, the administrative 
record supporting each normal husbandry practice proposed by a 
regulatory authority (53 FR 34641, September 7, 1988). In 1983, we 
considered and rejected the idea of allowing interseeding and 
supplemental fertilization during the first 5 years of the 10-year 
responsibility period. While allowing replanting of trees and shrubs 
``to utilize the best technology available'' without extending the 
responsibility period, we determined that augmented seeding, 
fertilizing or irrigation are not allowed during the responsibility 
period. (See 48 FR 40156, September 2, 1983.)
    However, in 1988 (53 FR 34641, September 7, 1988) we stated, in the 
context of the Federal regulation at 30 CFR 816.116(c)(4), that 
seeding, fertilization, or irrigation performed at levels that do not 
exceed those normally applied in maintaining comparable unmined land in 
the surrounding area would not be considered prohibited augmentative 
activities.
    Further, in the response to comments received concerning an Ohio 
program amendment, OSM stated that ``[t]he legislative history of the 
Act [SMCRA] reveals no specific Congressional intent in the use of the 
term augmented seeding.'' Accordingly, our interpretation of augmented 
seeding is given deference so long as it has a rational basis (see 63 
FR 51832, September 29, 1998).
    Included in the proposal to allow interseeding as a normal 
husbandry practice are proposed definitions for ``augmented seeding'' 
and ``interseeding'' to distinguish the differences between them. 
Interseeding is clearly aimed at establishing species that require 
special conditions for germination and the establishment or altering of 
species composition. Colorado's discussion of interseeding as a normal 
husbandry practice in the ``Coal Mine Reclamation Program Vegetation 
Standards'' guidance document further clarifies that interseeding is 
done to enhance revegetation, rather than to augment revegetation. 
Colorado reiterates that interseeding is defined as a secondary seeding 
into established revegetation in order to improve diversity. In 
contrast, augmented seeding is reseeding with fertilization or 
irrigation, or in response to unsuccessful revegetation in terms of 
adequate germination or establishment or permanence. Thus, Colorado's 
goal for interseeding is not only to ensure that the reclaimed area 
will meet the success standards, but to go beyond the minimum standards 
of the regulations and improve the overall diversity of the reclaimed 
area.
    Colorado also proposes to limit interseeding as a normal husbandry 
practice to the first four years of any ten year liability period. Such 
interseeding may consist of only native species and

[[Page 14992]]

approved introduced species contained in the original seed mix.
    To support interseeding as a normal husbandry practice, Colorado 
submitted the documents identified above. Colorado also proposed 
interseeding as a method to improve wildlife habitat and grazing 
values. Further, all referenced publications support the use of 
interseeding as a normal husbandry practice.
    We previously approved Indiana's definition of ``augmented seeding, 
fertilization, or irrigation'' as seeding, fertilizing, or irrigation 
in excess of normal agronomic practices within the region. Our approval 
was based on the concept that the proposed definition made a 
distinction between normal conservation practices that were not 
augmented seeding, fertilizing, irrigation or other work, and augmented 
husbandry practices (60 FR 53512, October 16, 1995).
    We also previously approved the use of interseeding as a normal 
husbandry practice in New Mexico (65 FR 65770, November 2, 2000). The 
Colorado proposal is based on language in the approved New Mexico 
program.
    Based on Colorado's proposed restrictions on ``interseeding,'' and 
the differentiation between ``interseeding'' and ``augmented seeding'' 
and the guidance provided for using interseeding as a normal husbandry 
practice, and other documentation and publications supporting 
interseeding as a normal husbandry practice in Colorado, we find that 
Colorado has demonstrated that the proposed use of interseeding is not 
augmented seeding. Because the use of interseeding proposed by Colorado 
clearly supports a key goal of SMCRA, which is the establishment of a 
permanent, diverse, and effective vegetative cover without compromising 
compliance of the State program with the Act, we also find that 
Colorado's proposed use of interseeding in rule 4.15.7(5)(g) is 
consistent with and no less effective than the Federal regulations at 
30 CFR 816.116(c)(1) and (4).
9. Rules 4.15.11 and 4.15.8(7), Revegetation Sampling Methods and 
Statistical Demonstrations for Revegetation Success [30 CFR 816/
817.116(a)(1)].
    The Federal regulations at 30 CFR 816/817.116(a)(1) require that 
standards for success and statistically valid sampling techniques for 
measuring success shall be selected by the regulatory authority and 
included in an approved regulatory program.
    The Federal regulations at 30 CFR 816/817.116(a)(2) require that 
standards for successes shall include criteria representative of 
unmined lands in the area being reclaimed to evaluate the appropriate 
vegetation parameters of ground cover, production, or stocking. Ground 
cover, production, or stocking shall be considered equal to the 
approved success standard when they are not less than 90 percent of the 
success standard. The sampling techniques for measuring success shall 
use a 90-percent statistical confidence interval (i.e., one-sided test 
with a 0.10 alpha error).
    Colorado indicates that existing rule 4.15.8(7) is reorganized to 
correspond to proposed rule 4.15.11. Reference to a specific confidence 
level is deleted, and detailed statistical requirements including 
confidence levels are addressed in rule 4.15.11. Reference to a 
demonstration that ``woody plant density exceeds 90 percent * * *'' is 
added to allow for use of the ``reverse null'' approach to a success 
demonstration, an option further detailed in rule 4.15.11. The amended 
rules at 4.15.11(1)(b) require DOW consultation and approval for shrub 
plantings, address statistical approaches appropriate to woody plant 
density evaluation, and address the ``80/60'' requirement of 30 CFR 
816/817.116(b)(3)(ii).
    Colorado states that rule 4.15.8(7) also allows for a reverse null 
success demonstration based on the median for woody plant density, with 
a success threshold of ``70% of the approved technical standard.'' 
These changes correspond to the provisions of rule 4.15.11, and a 
detailed justification for use of the median-based reverse null 
approach, supported by data and analyses included in Exhibit I (found 
in the March 27, 2003, State Program Amendment submission), is 
presented within the statement of basis and purpose sections 
corresponding to pertinent provisions of rule 4.15.11. The current rule 
states that the ``establishment of woody plants shall be considered 
acceptable if the density is not less than 90% of the approved 
reference area or standard with 90% statistical confidence.'' This 
language is essentially identical to the Federal requirement at 30 CFR 
816/817.116(a)(2). The ``not less than'' language implies use of the 
standard, or the traditional formulation of the null hypothesis, in 
which the inherent assumption is that reclamation has been successful 
for the parameter in question and the assumption of success must be 
upheld unless demonstrated to be false with statistical certainty. In 
this formulation, the ``burden of proof'' could be thought of as 
falling on the ``opponent'' of bond release. The current rule does not 
specify the use of the mean or median, but traditionally the population 
mean as estimated by the sample mean with associated confidence 
interval has been applied.
    Colorado states that the amended rule allows for the traditional 
approach of the current rule, but would also allow for an alternative 
median-based reverse null approach for a woody plant density success 
demonstration (as specified in proposed rule 4.15.11(3)(a)). The 
reverse null approach is inherently more stringent than the traditional 
null formulation because the assumption is that reclamation has been 
unsuccessful for the parameter in question. The assumption of failure 
must be upheld unless demonstrated to be false with statistical 
certainty. In this formulation, the ``burden of proof'' falls on the 
``proponent'' of bond release to demonstrate with statistical certainty 
that the reclaimed area parameter exceeds the specified success 
threshold. The median has certain advantages compared to the mean as a 
measure of central tendency, as the median is more stable or robust 
than the mean and it is impacted less by extreme data values. As a 
result, it is generally possible to estimate the population median with 
relatively high precision based on a relatively small sample size. 
However, as demonstrated by data included in Exhibit I, the median is a 
more stringent standard of success than the mean for woody plant 
density due to the typically skewed data distributions associated with 
woody plant samples on reclaimed lands. Because of the influence of a 
relatively small percentage of extremely high data values, the woody 
plant density mean almost always exceeds the woody plant density median 
by a substantial margin.
    For woody plant density, the reverse null approach, combined with 
use of the median as a specified measure of central tendency, is more 
stringent than the Federal requirements at 30 CFR 816/817.116(a)(2), 
which allow for the traditional null formulation using the mean as the 
specified measure of central tendency. The increased stringency is due 
to the effects of both the reverse null formulation and use of the 
median. In order to offset this excess stringency, proposed rule 
4.15.8(7) (in combination with proposed 4.15.11(3)(a)) allows for a 
success demonstration to be based on a threshold of 70% of a technical 
standard rather than 90% of the standard. Documentation in Exhibit I 
supports the reduction of the success threshold when the median is the 
specified parameter of comparison. The reduced stress threshold is 
further

[[Page 14993]]

justified by the requirement to employ the more stringent reverse null 
formulation to demonstrate success.
    Colorado states that rule 4.15.11 is being added to be no less 
effective than 30 CFR 816/817.116(a)(1) and to specify the 
statistically valid sampling methods and testing techniques that 
operators must use in demonstrations of revegetation success. 
Acceptable sampling methods and approaches for estimates of vegetation 
cover, herbaceous production, and woody plant density are addressed in 
proposed rule 4.15.11(1).
    We have reviewed rule 4.15.11(1). As proposed, this identifies the 
sampling methods that can be used to evaluate vegetation cover, 
herbaceous production and woody plant density. For vegetation cover, 
point intercept, line intercept or quadrat sampling are listed. For 
herbaceous production, quadrat sampling or total harvest are the 
identified methods. For woody plant density, identified methods include 
belt transects and circular or rectangular quadrats. Sampling can be 
either random or systematic. We have determined that these are all 
standard sampling techniques used throughout the country and have been 
previously approved in multiple State programs. Thus, subsection 
4.15.11(1) is consistent with and no less effective than the 
requirements of 30 CFR 816.116(a) and therefore should be approved.
    The State indicates that statistical testing and sample adequacy 
approaches acceptable for vegetation cover, herbaceous production, and 
woody plant density are addressed in proposed rule 4.15.11(2). The 
amended rule ensures that tests for success will employ a 90 percent 
confidence level (alpha error probability = .10) for ``standard null 
hypothesis-based'' demonstrations of success, and that tests will 
employ an 80 percent confidence level (alpha error probability = .20) 
for ``reverse null hypothesis-based'' demonstrations of success. Data 
and analyses in Exhibit I of the program amendment demonstrate that 
reverse null tests at the 80% level of confidence are no less effective 
(and in fact are more stringent) than standard null tests at the 90% 
level of confidence. Selected revegetation success standards are 
addressed in rules 4.15.7(2)(d), 4.15.7(3), 4.15.7(4), 4.15.8, 4.15.9, 
and 4.15.10. Justification for the 70% success threshold of proposed 
rule 4.15.11(3)(a) for woody plant density is provided in the 
discussion under Rule 4.15.8(7) above, and pursuant to associated 
amendments to Rule 4.15.8(7). Additional justification is included in 
Exhibit I.
    Colorado states that proposed rule 4.15.11(2)(a) incorporates into 
its regulations the standard statistical sample adequacy formula and 
direct success comparison approach previously specified in DMG 
guidelines. A notable modification is that the rule allows for use of a 
precision level of 0.15, rather than 0.10, in the standard sample 
adequacy formula for woody plant density estimation. Larson (1980) used 
a precision level of 0.10 in example data sets, and that level of 
precision has subsequently been widely specified in State regulations 
and guidelines. However, no specific level of statistical precision is 
required by the Federal regulations in 30 CFR 816/817.116. In Colorado, 
we have found the 0.10 precision level to be appropriate and 
practicable in the majority of cases for statistical evaluation of 
cover and production success. However, due to the high variability and 
skewed distributions typical of reclaimed area woody plant density 
data, extremely large sample sizes are typically necessary to 
demonstrate sample adequacy for woody plant density at the 0.10 level 
of statistical precision. The time and expense associated with 
obtaining estimates of woody plant density that are precise to within 
10% of the true mean are not justified for coal reclamation lands in 
Colorado. Colorado enclosed, as Exhibit I, a package containing woody 
plant density data from representative mine reclamation areas in the 
Yampa Basin and North Park, Colorado. The package includes detailed 
analyses of the data, and presents justification for use of a precision 
level of 0.15 in the standard sample adequacy formula for woody plant 
density estimation. Colorado asserts that use of the 0.15 precision 
level rather than 0.10 will significantly reduce required sample sizes 
for reclaimed area woody plant density estimates. In Colorado's 
judgment, the increased precision associated with use of 0.10 for woody 
plant density estimation is not critical, and the relatively small 
increase in precision comes at too high a price in terms of the time 
and effort associated with the additional data collection. Colorado 
also asserts that the use of a 0.15 precision level rather than 0.10 
for demonstrating woody plant density success will negligibly affect 
the extent to which reclaimed shrublands provide desired wildlife cover 
and forage on reclaimed landscapes. In Colorado, woody plant density 
standards are set based on consultation with DOW personnel and reflect 
the consideration of a wide range of variables typically involving 
negotiation among DOW and DMG staff, operators and consultants. It is 
not an exact science and necessary or optimum levels of woody plant 
density to meet applicable habitat requirements are not precisely 
defined. Colorado believes that the application of such a high degree 
of precision to a standard that is based on professional 
recommendations and negotiation is unwarranted.
    Our review affirms that rule 4.15.11(2) identifies the statistical 
analysis and sample adequacy procedures to be used in evaluating 
vegetative cover, herbaceous production and woody plant density. Rule 
4.15.11(2)(a) gives the standard sample adequacy formula for use in 
direct comparisons when the value for the reclaimed area is greater 
than the standard, or when the reclaimed value is less than the 
standard but not significantly different. It sets sampling precision at 
0.10 for vegetative cover and herbaceous production and 0.15 for woody 
plant density. In discussing the setting of precision levels, OSM 
indicates that it has not stated a level of sampling precision in the 
final rules but will instead evaluate on a case-by-case basis the 
adequacy of predetermined sample sizes or methods of sample size 
selection proposed for use in State programs (48 FR 40150, September 2, 
1983). Colorado's proposal to set precision levels at 0.10 for 
vegetative cover and herbaceous production is consistent with 
previously approved precision levels used in States in the West. 
Colorado has also demonstrated that the proposal to use a precision 
level of 0.15 for woody plant density is appropriate given the high 
variability in shrub density across a reclaimed area. The proposed rule 
is consistent with and no less effective than the Federal requirements 
of 30 CFR 816.116(a) and should be approved.
    We note that rule 4.15.11(2)(b) includes the standard method for 
comparing vegetative parameters from the reclaimed area to 90% of the 
success standard. This approach makes use of the classic null 
hypothesis, which is that the vegetation on the reclaimed land is equal 
to or greater than that of the success standard. Under this approach, 
the vegetation on the reclaimed area may be less than the success 
standard; but statistically, it is not significantly different and the 
null hypothesis is not rejected. The minimum sample size is 15 and all 
sampling must meet sample adequacy using the formula in Subsection 
4.15.11(2)(a). This is the standard approach used by State Regulatory 
Authorities throughout the United States and is the approach discussed 
in

[[Page 14994]]

the 1983 preamble (48 FR 40152, September 2, 1983). As proposed, this 
subsection is consistent with and no less effective than the Federal 
regulations and should be approved.
    As discussed in the State's supporting justification, subsection 
4.15.11(2)(c) proposes to allow the use of the ``reverse null'' 
hypothesis when the vegetation parameter from the reclaimed area is 
greater than the success standard, but the number of samples taken is 
not sufficient to meet sample adequacy. The reverse null hypothesis 
states that vegetation on the reclaimed area is less than 90% of the 
success standard. OSM has previously approved use of the reverse null 
hypothesis in the New Mexico program. Under the Colorado proposal, the 
confidence interval is set at 80% (alpha = 0.20) and a minimum of 30 
samples is required. The proposed alpha (error probability) of 0.20 is 
greater than the 0.10 in the Federal regulations. However, in order to 
demonstrate that the revegetation meets the success standard under the 
reverse null hypothesis, the operator must show that the vegetative 
parameter of concern is significantly greater than 90% of the success 
standard. That is, the mean value for a given parameter must be well 
above the success standard because to be significantly greater than the 
success standard, the lower tail of the 80% confidence interval must 
also be greater than 90% of the success standard. Therefore, even 
though the error probability is slightly larger under the State's 
proposal, the requirement to exceed the success standard ensures 
consistency with the Federal regulations. To support this approach, 
data in Exhibit I shows that a comparison of (1) statistical testing 
using the standard null hypothesis and a 90% confidence interval and 
(2) the reverse null hypothesis using an 80% confidence interval either 
gave the same results or the reverse null was more stringent. For this 
reason, the use of an 80% percent confidence interval with an alpha of 
0.20 is consistent with and no less effective than the Federal 
regulations and should be approved.
    In discussing rule 4.15.11(3), the State indicates that it allows 
for additional optional approaches for demonstrations of sample 
adequacy and revegetation success that are solely applicable to woody 
plant density. The approaches include (1) a median based reverse null 
confidence limit comparison, (2) a mean based pre-determined sample 
size direct comparison, and (3) an approach based on stabilization of 
the running sample mean. The range of options presented for woody plant 
density is warranted, due to the extremely large sample sizes that have 
frequently been necessary in order for operators to demonstrate success 
for this parameter using traditional statistical methods. Based on the 
discussion below, the approaches specified in rules 4.15.11(3)(a), (b), 
and (c) are no less effective than the applicable Federal requirements 
of 30 CFR 816.116(a)(1) and (a)(2). However, depending on 
characteristics of the data, the range of options may allow for 
operators to select a success demonstration approach that requires a 
less intensive sampling effort than would be required if limited to 
only one or two approaches.
    Colorado included, in Exhibit I, data and arguments in support of 
these approaches.
    Rules 4.15.8(7) and 4.15.11(3)(a) propose using the reverse null 
hypothesis and nonparametric rank-sum test to demonstrate that the 
median value for the reclaimed area is greater than 70% of the success 
standard using an 80% confidence interval. In discussing this proposal 
in Exhibit I, the State indicates that, based on the literature and its 
observations, woody plant density data from reclaimed lands are seldom 
normally distributed and typically exhibit lognormal or similar 
distributions with a strong skewness to the right. Parametric 
statistics based on means and standard deviations include the 
assumption that the data come from a normal distribution. This limits 
the use of normal statistics in these type of populations. The median 
is a relatively ``robust'' or ``resistant'' measure of central 
tendency. It is not influenced by a few extreme values and so it does 
not get pulled toward the right tail. As a result, in a right-skewed 
distribution, the median is always lower than the mean. Because 
reclaimed parcel woody plant density data sets typically exhibit right-
skewed distributions, the requirement to demonstrate woody plant 
density success based on a comparison of the median to a technical 
standard is more stringent than a demonstration based on a comparison 
of the mean to the same technical standard. Review of the various data 
sets and summary statistics submitted by Colorado in Exhibit I 
indicates that, on average, the medians for data averaged less than 75% 
of the mean for those same data sets. Based on this information, it is 
reasonable to use 70% (e.g., 90% of 75%) of the success standard when 
making comparisons to the median value of the reclaimed area. The fact 
that amended rule 4.15.11(3)(a) also requires a reverse null confidence 
limit comparison on the median adds an additional layer of stringency. 
To be judged successful, the one tailed 80% lower confidence interval 
on the sample median would have to exceed the success threshold.
    Based on a review of the data submitted by the State, OSM has 
determined that proposed rules 4.15.8(7) and 4.15.11(3)(a) are 
consistent with the intent of SMCRA and no less effective than 30 CFR 
816.116(a)(2) in establishing success standards and ensuring that 
statistically valid comparisons are made during the evaluation of 
revegetation success. Accordingly, the rule should be approved.
    In discussing rule 4.15.11(3)(b)(i) in Exhibit I, Colorado 
indicates that an approach that may in certain situations allow for a 
smaller sample size than indicated by the standard sample adequacy 
formula, without a corresponding reduction in stringency, is a non-
statistical predetermined (or maximum) sample size.
    Rule 4.15.11(3)(b)(i) allows for an empirically derived, 
predetermined sample size of 75 that operators could use for a success 
demonstration in cases where sample adequacy has not been demonstrated 
by approved statistical formulas. In this approach, the woody plant 
density sample mean obtained from a sample of at least 75 100-square-
meter quadrats is compared directly against the approved success 
threshold (90% of the approved standard) with no consideration of 
statistical error or confidence level). The specified quadrat size 
restriction is necessary because a high percentage of the data that 
comprise the basis for the proposed sample size of 75 were obtained 
using a 2-meter by 50-meter quadrat.
    Again, the State has included in Exhibit I a review of several data 
sets to demonstrate that a sample size of 75 is generally adequate to 
ensure that the sample mean would be within the 90% confidence interval 
of a statistically adequate sample. The 75 sample size was no less 
effective than using the sample adequacy formula to determine sample 
size more than 90% of the time. It should also be noted that in the 
preamble to the Federal regulations at 30 CFR 816.116(a)(1), OSM stated 
that we will evaluate on a case-by-case basis the adequacy of 
predetermined sample sizes (48 FR 40150, September 2, 1983). Based on 
the information submitted as part of this program amendment, we 
determined that the use of a maximum of 75 samples to evaluate the 
success of woody plant density is consistent with the intent of SMCRA 
and no less effective than the Federal regulations.
    Rule 4.15.11(3)(b)(ii) will allow the use of a sample adequacy 
calculation that is based on the variance of the

[[Page 14995]]

running mean, a minimum sample size of 40 samples, a precision of 0.03, 
and an alpha of 0.10. In Exhibit I of this amendment, Colorado 
evaluated the variance of the running mean sample adequacy approach 
based on a number of the data sets. The running mean approach results 
in drastically reduced sample sizes compared to the standard sample 
adequacy approach (as specified in 4.15.11(2)(a)), when the same level 
of precision is specified in the formulas. This is due to the fact that 
successive running mean values are much less variable than successive 
sample observations. As such, the variance of the sample mean is 
correspondingly smaller than the sample variance.
    As discussed in Exhibit I of the amendment, Colorado compared three 
different levels of precision, 0.10, 0.05, and 0.03, to determine the 
effect on sample size and estimates of the mean and to ensure that 
reduced sample size will not weaken the ability of hypothesis testing 
to detect a true difference between the reclaimed area mean and the 
approved standard (success threshold). The two lower levels of 
precision (i.e., 0.10 or 0.05) do not appear to result in reliable 
estimates of the mean when applied to the Colorado data, even when a 
minimum sample size of 40 is imposed. At the .03 level of statistical 
precision, and with a minimum sample size of 40, the modified sample 
adequacy formula provides for a modest reduction in average sample size 
compared to average sample size resulting from application of the 
standard sample adequacy formula with a 0.15 precision level. Further, 
success demonstration stringency is comparable when the modified 
standard deviation term is substituted in the t-test formula.
    We have reviewed the proposed alternati