January 20, 1999

Comment Clerk for the TMDL Program Rule
Water Docket (W-98-31)
Environmental Protection Agency
401 M St., S.W.
Washington, D.C. 20460

Re: Proposed Revisions to the Water Quality Planning and Management Regulation, 40 C.F.R. Part 130, 64 Fed. Reg. 46012 (August 23, 1999).

Dear Comment Clerk:

Northwest Environmental Advocates (NWEA) submits the following comments on the proposed revisions to the water quality planning and management regulation concerning the Total Maximum Daily Load program. NWEA represents members throughout Oregon and Washington and has sought implementation of the water quality-based programs of the Clean Water Act, including but not limited to the TMDL program, for many years, through litigation in the two states, advocacy with state agencies, lobbying state legislatures, working with industry, and participation on EPA's TMDL Federal Advisory Committee. While our comments include some harsh criticisms of the proposed rule, we believe that overall EPA's approach is sound and we commend the agency for taking some unpopular positions. In fact, we think these regulations are crucial to ensuring the TMDL program meets environmental goals and attains some modicum of equity between all sources in reaching those goals.

As a general comment, we think EPA should seek efficiencies in the development of lists and TMDLs and in EPA's approval/disapproval process by requiring states to adhere to simple formats and templates. The more consistent work products are, the better able citizens and other agencies including EPA will be to review them. There are a number of places in the comments below where we specifically identify our suggestions in this area. Even if EPA disagrees, the agency can and should still create templates and urge states to use them. If the majority of states do so voluntarily, or at least adapt their own approaches to these templates, the program will achieve efficiencies it will otherwise lose.

Our comments below are arranged in the order presented in the preamble. Where the preamble presents issues underlying proposed rule language, options that were considered but rejected, or EPA's consideration of various issues, our comments are on the preamble. Following this discussion, under sections labeled "Comments Sought" are our comments on the proposed rule language.

Subpart A: SUMMARY, PURPOSE AND DEFINITIONS

Question No. 3: What Definitions are Being Added or Revised by this Proposal?

Existing Requirements

No comment.

Proposed Rule

NWEA agrees with EPA that definitions need improvement in order that the program have consistency both within and between states and EPA regions. EPA's encouragement of states to develop TMDLs addressing all pollutants impairing a listed waterbody simultaneously is flawed, however. 64 Fed. Reg. 46015. Instead, EPA should: 1) require states to develop TMDLs for all pollutants that have additive or synergistic effects on beneficial uses, and 2) discourage the development of TMDLs for low-priority pollutants on higher priority waters unless the data gathering and modeling are significantly similar. With regard to the first issue, by merely encouraging, but not requiring pollutant combinations to be developed together, EPA undermines the requirements of states' water quality standards to support beneficial uses (including existing uses) and apply narrative criteria. There are many instances where the application of numeric criteria alone will not result in the protection of uses and most numeric criteria, although not all, have not been designed to take into consideration either excessive amounts of other pollutants or simply any amount of other pollutants at all. This is true, for example, with toxic pollutants, the numeric criteria for which are established on the basis of the risk to (only some) uses from exposure to the individual pollutant. Yet we know that: 1) toxic contaminants are rarely found in unsafe or questionable levels individually, and 2) multiple toxic contaminants have additive, and sometimes synergistic, effects on uses. Thus, failure to develop TMDLs that take into account real world situations in the context of the legal definition of water quality standards do not meet the requirements of the Clean Water Act. Conventional pollutants are similar. For example, the combined effects of excess temperature and low dissolved oxygen have a greater effect on fish than each pollutant individually. Even if the numeric criteria are attained, it may be necessary to increase the margin of safety to ensure that the narrative criteria and beneficial use support requirements are met by a TMDL, unless the additive effects of the pollutants are addressed in the standards themselves. In other words, states' developing TMDLs for individual pollutants will often entail a failure to ensure that the TMDLs will lead to attainment of standards. Therefore, EPA should require that states evaluate similar or related pollutants together as they develop TMDLs.

On the other hand, EPA's unrestricted promotion of TMDLs for all pollutants in a waterbody at the same time will encourage states to avoid developing high priority TMDLs. States could well end up developing TMDLs to protect the low priority use of livestock watering or aesthetics while failing to develop TMDLs to address endangered species. This is already a problem we have seen in Washington State. For example, although much of Washington contains aquatic endangered species, Washington's 1999 list for TMDL development is most heavily weighted to addressing fecal coliform bacteria issues, not threatened and endangered species. In this state's situation, it is a continuing problem: Washington initially focused on developing §303(d)(3) preventative TMDLs for high quality waters, resulting in our need to file a lawsuit against EPA. Clearly, the state remains insensitive to the needs of the beneficial uses most in need of TMDLs.

EPA notes that some TMDLs will have to be established for different instream flows in order to comply with the statutory requirement of seasonal variation. Id. at 46016. We agree. However, EPA has failed to underscore the importance of this and to ensure that TMDLs are clear about the seasonal variations and their underlying assumptions. This is primarily because, although flow assumptions are at the heart of the calculation of loading capacity of a waterbody, EPA has not proposed that those assumptions be a part of the definition of a TMDL. It is difficult to imagine how a TMDL could be calculated without the information, so it is assumed by everybody that the flow numbers will be in a TMDL. The same position could be taken with regard to any component of the definition but interestingly, EPA does not. Instead, EPA lays out most of the details of the definition, omitting this key one. Setting out clearly what the instream flows are, with seasonal differences if necessary, should include the assumptions upon which the flows are based. This will help establish: 1) the longevity and relevance of the TMDL, 2) implementation planning, and 3) that the TMDL will lead to attainment. For example, if the TMDL assumes that the flows will remain at current levels yet significant water rights for out-of-stream uses are granted in the future, it should be clear in the document that the TMDL must be revised. The more clearly this is stated in the TMDL, the more clear the Implementation Plan can address this circumstance under its provisions for revision. Likewise, by stating this clearly, stakeholders can understand the limitations of the TMDL or alternatively can understand how their interests can be protected by assuring the TMDL remains relevant despite attempts to decrease instream flows. The same is true of the use of stored water. A TMDL could be based upon certain assumptions about how stored water is managed. Those assumptions could be invalidated if management of stored water changes. Likewise, if flow assumptions are based upon needed changes the TMDL identifies are needed - such as increases in instream flows obtained from conservation measures or changes in the management of stored water - then the Implementation Plan needs to include these actions. Finally, if a TMDL fails to address flow issues or fails to establish that its flow assumptions are valid, it will not demonstrate on its face that it will lead to attainment of standards.

The Preamble also notes that TMDLs may be established on a "watershed basis." Id. at 46016. We do not disagree that larger geographic scale analysis is appropriate for pollutants that are generated on such a scale. However, the preamble does not include a statement of EPA's expectations about how TMDLs done on such a large scale will also address smaller-scale needs. For example, widespread overgrazing takes place on a watershed scale and can be usefully addressed on that basis. However, grazing and its effects on water quality and on beneficial uses is not homogeneous. Instead, a useful analysis of grazing must address hazard areas (e.g., steep slopes, landslides, undercut banks) as well as specific sensitive areas within the watershed that are critical to support of the uses (e.g., wetlands, temperature refugia, spawning gravels). Without this smaller-scale approach, a TMDL is not likely to result in the attainment of standards; instead it will be a warmed over version of whatever plan is currently being used to manage those lands. The TMDL, in this case, will add nothing to state-wide or agency-wide approaches and will be a waste of resources. This draft preamble language has already been used to justify the proposed creation of TMDLs that would make a single load allocation to all nonpoint sources on a watershed basis. The staff of the Washington House of Representatives Committee on Agriculture and Ecology has proposed a so-called "WRIA TMDL" that would do just that as a way of avoiding TMDLs' affecting nonpoint source controls. EPA's rule must address specifically why and how it believes that a watershed approach will "be more likely to result in effective control measures than segment-by-segment TMDLs." Id. A mere pronouncement, without details, is more likely to have the opposite effect. Instead of using TMDLs to create more appropriate controls, states will merely point to existing state-wide efforts as sufficient for watersheds. While we agree that watershed processes must be examined at the watershed level, we urge EPA to focus on ways to ensure this happens in an analytical and thorough way.

We agree with EPA that all microbial contaminants fall under the definition of "biological materials." Id. at 46017. However, we are disturbed by the references in the guidance to exotic species not being similar organisms. Guidance at 2-5. As EPA knows from our petition seeking NPDES coverage for ballast water discharges, it is our position that exotic species are pollutants, requiring TMDLs and NPDES coverage. Nowhere in the proposed rule does EPA establish why it believes that discharges of these biological organisms that devastate native habitats and species are not pollutants. It does not explain how it can draw a magic line between microbial and larger species. EPA does not cite to the Act to explain how it arrives at its conclusion. In Dubois v. U.S. Dept. of Agriculture, 102 F.3d 1273 (1st Cir. 1996), the First Circuit held that discharging water from one waterbody to another (to replenish a lake after the water had been used to make snow) constituted an addition of a pollutant because the river contained different life forms than the lake into which its water was discharged. While it would be convenient to conclude that exotic species are not pollutants, to avoid developing programs to address them, there is no basis in law or fact for that conclusion, and EPA must apply the TMDL program to such species.

Comments Sought

§ 130.2(h) We support EPA's proposed definition of TMDLs with one significant exception. The definition would be more clear, and consistent with the preamble, the proposed guidance, and EPA's stated views, if it clearly included the equation that comprised the old definition, namely that the sum of the wasteload allocations plus the sum of the load allocations plus a margin of safety equals the loading capacity. Including this in the definition would also underscore that the TMDL, as defined, remains a quantification process. The use of the words "written plans and analyses" suggests that TMDLs need not be quantified, which would not be consistent with EPA's publically stated views, the recommendations of the FACA Committee, and the Act. Discussion of the subparts will be found below, under section § 133.33.

§ 130.2(m) We support EPA's definition of an impaired waterbody as consistent with the statute.

§ 130.2(n) We support EPA's definition of threatened waters but note that there is an omission. Where a state has information that a significant source of pollution (e.g., a proposed water withdrawal) or pollutants (e.g., a major clear-cut) is proposed that it believes would cause a violation of state water quality standards - even in the absence of adverse declining trends - the state should be required to list this water as threatened. To define threatened in such as way as to excuse a state from acting on such an obvious basis would undermine the purpose of including threatened waters on the §303(d)(1) list, namely to avoid violations and the resources needed to develop TMDLs. It would also be counter to the intent of antidegradation policies included in state standards. Including this in the definition would still require states to have made a finding that the water was relatively close to violating standards.

§ 130.2(p) We support EPA's definition of reasonable assurances for both point and nonpoint sources. EPA's emphasis on certainty, on expeditious actions, and use of common sense (e.g., adequate funding) all should remain the basis for making these findings.

Subpart B: WATER QUALITY MONITORING AND REPORTING

Comments Sought

§ 130.10 This section of the proposed rule does not include any requirements for states to increase the scope of their monitoring programs. As currently many states only monitor a fraction of their waters, EPA needs to increase requirements for monitoring so that states adequately assess all state waters for inclusion in the TMDL program. The most obvious approach would be to create default listings for all waters and/or default listings for waters where data and information are not definitive but indicate impairment. States should have a limited amount of time (e.g., two years) in which to resolve these questions. If the state does not make a determination that the waters are meeting standards, it should be required to list them for TMDL development.

Subpart C: IDENTIFYING IMPAIRED AND THREATENED WATERBODIES AND ESTABLISHING TMDLs

Question No. 4: What are the Proposed Rule's Requirements for Identifying and Listing Impaired or Threatened Waterbodies?

a. Assembling the Data and Documenting the Approach for Considering and Evaluating Existing and Readily Available Data and Information.

Existing Requirements

Although EPA is well aware of the problem, the preamble fails to discuss the reality of states' inconsistent approaches to §303(d)(1) listings. For example, at the EPA/WEF meetings held to discuss the proposed rules, EPA speakers presented a map showing the percentage of impaired waters listed by the states. Percent Impaired Waters in 1998, a 1999 map. While the differences between states no doubt can and do reflect differences between water quality standards and available data, where there are dramatic differences - e.g., between the dark orange color of Oregon and the light yellow of Washington - it is also likely to be differences in listing criteria (methodology) and in the effort expended by the state to seek out "all readily available" data and information. In this particular example, there are differences in standards, such as for temperature, that account or should account for a majority of listings in the region due to excess grazing, logging, water withdrawals, and poor farming practices. We have also been told that federal agencies in Washington have not collected as much temperature data as in Oregon. We do not know if this is true. What we do know, however, is that when Oregon prepared its 1994/96 list, under the terms of a consent decree with us, the state actually visited the field offices of agencies such as the U.S. Forest Service, thus obtaining data that would otherwise not been made available. Washington, on the other hand, merely sends out a notice saying it will accept data. By taking this minimalistic approach, which we would guess is similar to the majority of states, Washington avoids using data that are readily available, thereby not listing streams that are impaired, not listing for all impairments that are known, and avoiding building relationships with other agencies that could support the TMDL program by better understanding their role with it. It is also instructive to compare New York's very pale yellow color with remarks made by that state's representative in a presentation to the FACA Committee. He explained how New York has identified many waters that are unsafe for fish consumption due to toxic contamination but that these waters were not placed on the state's 303(d)(1) list. Rather than struggling to interpret what "readily available" means each time a state submits its list, EPA could clearly define the process in this rule.

Proposed Rule

EPA mentions the need for states to expand the scope of their monitoring programs but fails to discuss how states should accomplish this, in light of continuing reductions in monitoring budgets. Id. at 46018. One obvious way is to ensure that states both solicit and consider, using reasonable criteria, monitoring data from citizens and tribes. This source of information becomes particularly important as certain legislatures pass laws requiring increased monitoring for the specific purpose of delisting streams. These kinds of messages from legislative bodies to state water quality agencies have a chilling effect on any increase in scope of monitoring with existing resources. Problems with monitoring by state water quality agencies can also be practically overcome through use of data and information collected by other state, federal, and local agencies. However, as discussed above, where states fail to obtain these data, this low cost approach is forfeited. The preamble also fails to tie the recommended scope of monitoring improvements to the §303(d)(1) list. In other words, where states have data but also have adopted restrictive listing criteria - such as Nevada's limitation on using data over two years old - the benefits of monitoring are overridden. EPA's proposed rule should address this by creating default positions and national listing criteria.

The preamble to the proposed rule discusses monitored data versus evaluated data and information. Id. at 46018. We are concerned with EPA's comment that monitored data includes only "some" fish tissue analyses and the evaluated data and/or information contains "some" surveys of fish and wildlife. We are unable to think of any fish tissue analyses that would not constitute monitored data unless a state simply felt like ignoring the information. Likewise, we cannot think of surveys of fish or wildlife that would not be relevant to determining whether standards were violated. While on one hand it does not appear that the distinction EPA makes concerning monitored and evaluated data makes a difference, since the proposed rule instructs states to use both, if overlaid with the directive to use "the best" data and information, it may constitute an invitation to states to ignore perfectly relevant data and information. For example, since EPA accepts the hierarchy of monitored "data" over "information" that can include wildlife surveys, a state could apply its numeric criteria to data as a way to demonstrate compliance with standards even in the face of wildlife surveys that demonstrate noncompliance but are discarded as mere "information." The current situation - where states have no numeric criteria to protect wildlife, where they fail to obtain data on wildlife impairments, where they do not have required policies on how to apply their narrative criteria - is bad enough. Adding EPA's vague instructions to that will encourage states to avoid listings that otherwise are required by law.

The preamble states that EPA will comment on the states' proposed methodology for listing. Id. at 46019. However, EPA will not approve or disapprove the methodology, just the submitted lists. We understand the rationale for this scheme but we believe that EPA is wrong in believing that this approach will make much of a dent in the highly inconsistent listing processes between states. The requirement to submit listing methodologies will increase EPA's ability to haggle with states and will make the listing process more open to citizens but it will not necessarily increase the likelihood of appropriate listings unless EPA is willing to disapprove lists. Knowing of the problem, it defies imagination as to why the agency isn't willing to set out in its rules certain listing criteria that apply to all states. States maintain a huge amount of flexibility in setting standards - particularly when they fail to conduct triennial reviews, EPA fails to promulgate where states fail to improve standards, and EPA approves standards that fail legal requirements. Why then, does EPA have to create flexibility in this crucial area of the TMDL program? EPA needs only to look at its own national map to witness the problem.

NWEA agrees with the proposal to eliminate requiring states to provide EPA with the rationale for not using data and information. However, we disagree with EPA's suggestion that since the ANPRM for water quality standards is supposed to address the "need for clear procedures for interpreting and implementing narrative criteria" that this rule should ignore the issue. The ANPRM is not moving quickly enough and the TMDL program is supposed to be on a fast track.

Other Options Considered

EPA presents the two options for default listings as if the agency considered them and rejected both. Id. at 46020. In fact, the latter default - an assumption that all waters are clean unless proven dirty - is currently in place and would not change in the proposed rules. This default position has allowed states to cut monitoring budgets, avoid obtaining readily available data and information, create absurdly restrictive listing criteria, and encouraged legislatures to skew TMDL programs towards delisting. By choosing the other default - that waters are considered impaired unless proven to meet standards - EPA's rule would cause states to implement meaningful monitoring programs and ensure that the TMDL program would address all needed impairments.

Comments Sought

§130.22 We support EPA's inclusion of the current rule's requirement that states assemble and consider all existing and readily available data and information. But see discussion above to address problems with the existing rule.

§130.22(b) The approach states are taking to defining what is readily available and what constitutes information is extremely random and varied. For this reason, EPA's regulations should clarify these two areas. First, the word "information" in the regulation is not spelled out. EPA should clarify that data and information mean scientific measurements or analyses of water quality (physical, chemical, and biological), including instream flows and the effects of pollution and pollutants on beneficial uses. States should not be able to avoid §303(d)(1) listing of streams because they choose to ignore temperature data, analyses of reproductive failure of species due to toxic pollutants, closures of shellfish beds due to high bacteria counts, or insufficient instream flows. This is the current situation because states do not seek out much of the information that has been collected with tax payers money to put it to use in the one program Congress created to data to drive clean- up. EPA's proposed rules need to offer a remedy.

§130.22(b)(6) Second, it is critical that EPA phrase the regulation in the active, not passive ("problems reported by") voice. The regulation should require states to take the necessary actions to obtain this data and information. A pertinent example of how these data can be ignored is the approach Washington uses, which is to solicit and accept data but not to actively seek it. Oregon, on the other hand, when developing its 1994/96 list, was far more aggressive in actively seeking data. The differences show up in the relatively low number of listing in Washington for temperature. While this requires an initial expenditure of more time, it pays off by expanding the scope of the state's water quality information and monitoring network (a goal of EPA's and endorsed by the FACA Committee), ensures the TMDL program applies to waterbodies that require it, and encourages the understanding and participation in the TMDL program by important players such as state and federal fish and wildlife and land management agencies, and universities. At the least, the rule should require states to prepare a document listing each state, federal, and local agency, university, and Tribe operating within the state, explaining how each was contacted and the result. With this information, EPA can better assess compliance with the rules.

§130.23(a) We support the proposed rules' requirement that states develop a methodology for considering and evaluating data and information for listing and determining priority rankings. We reiterate our concern that states should at the very least be required to develop a methodology for obtaining the data and information. We also agree that states should provide a public comment period of not less than 60 days. It has been our experience that public review of the list is extremely difficult in the absence of clear listing criteria and this is a necessary part of the rule.

§130.23(b) We agree with EPA that the methodology must include the factors included. We believe, however, that EPA should spell out more detail about what it means by "physical/chemical," "biological," and "aquatic and riparian habitat." For example, "biological" should include analyses of likely human health effects from pollutants, including environmental justice concerns. If EPA does not state this clearly, states will continue to ignore data and information that demonstrates that existing levels of contamination pose an adverse, and sometimes unequal, risk to certain human populations. Likewise, biological information includes data and information on adverse effects to fish and wildlife that are caused by or contributed to by pollutants and pollution. States generally do not seek out this information. The more clearly specific the requirements, the more consistency EPA will create in this program.

§130.23(c) We disagree with EPA's decision to merely ask states to divulge their methodology rather than present them with some requirements. The great resource and political pressures caused by 303(d)(1) listings result in states' seeking ways to avoid listings. EPA can best guard against this known problem by addressing it in a straightforward manner. Instead, under the proposed rules EPA will have to make case-by-case judgments that are likely to perpetuate inconsistency and be difficult to justify. Just how inconsistent these are will be made more clear because EPA and the public will have better information as a result of this rule even though the rule will stop short of fixing the problem. At the very least, EPA should create default positions on certain listing criteria. If states want to use a different basis for not listing waters, they will have to justify it against the default criteria.

§130.23(c)(1) For example, rather than leaving open the issue of data age, EPA should specify that all data collected within the last 15 years must be considered. Without this approach, states have an incentive to not collect data, allowing older data to age without replacing it, and then to say it is invalid. Some types of monitoring are very expensive, such as toxic contamination, and simply are not done very frequently. The mere age of the data do not suggest that they are no longer valid.

§130.23(c)(3) Likewise, the number and degree of exceedances of criteria is another area ripe for abuse. These sorts of listing criteria on how standards are applied to data are not part of the submissions of new standards to EPA for review and approval/ disapproval in the triennial reviews, and EPA is not planning to approve or disapprove them in the listing process, yet they are critical to how a state implements the standards and TMDL program. In the absence of EPA's willingness to dictate to states what the listing criteria should be, EPA should make them a clear part of either standards or listing. They should be subject to public review and §7 consultation under the Endangered Species Act. The alternative is that there will be increasing litigation over the sufficiency of states' lists as a way of addressing overly narrow state listing criteria.

EPA also errs in not making clear in this section that states must apply their narrative criteria and beneficial use support aspects of standards independently. This is necessary, for example, to avoid what Oregon has done which is to give lip service to these components by creating listing criteria that require violations of numeric criteria in order to find that narratives have been violated. This is contrary to the fundamental concept of these two components as gap-fillers but it has been an effective way of ensuring that the state can keep its list as short as possible. EPA's regulations should guard against this absurd reading of the law and should make clear to those states that act as if narratives and use support are academic concepts that they are actually legal requirements. The language in this section also needs to be fixed; how can there be a "number and degree of exceedances" of narrative criteria and beneficial use support?

§130.23(d) We find this section confusing in general, although we do not disagree with its apparent intent.

§130.23(d)(1) We cannot comment on this because we don't understand what it means. How are "selection factors" different from the methodology?

§130.23(d)(2) We do not agree that states must resolve disagreements with other jurisdictions. States generally will not yield in their interpretations of their own water quality standards. Since it is clear that the most protective of the standards apply, why try to force states to enter into fruitless negotiations?

§130.23(d)(3) We support this requirement.

§130.23(e) While we support the intent of this section, we strongly oppose EPA's not providing states with more instruction on how to remove previously listed waters. There is too great a likelihood that data from samples taken at different times of year, different types of weather years, different parts of waterbodies, different times of day/night, etc., can be used to justify removal of waters that remain out of compliance with standards. EPA must address this issue with clarity.

§130.24 Subject to comments made above, we support this proposed section.

a. Scope of the List

Existing Requirements

It cannot be repeated too often that EPA's proposed rules must do more to address the extreme degree of inconsistency between states listing programs.

Proposed Rule

Listing Requirement: Point/Nonpoint Sources

We agree with EPA's reading of §303(d) to include nonpoint sources in the TMDL program, including listing of waters fully or partially impaired by nonpoint sources.

Pollutant/Pollution

However, we disagree with EPA's reading that the statute limits the development of TMDLs for waters impaired by "pollution." Pollutants and pollution are so inextricably linked that EPA's attempt to make a distinction both cannot be and is not clear. The proposed rules appear to say that when TMDLs are done for pollutants, pollution must also be addressed and that TMDLs must address the physical and biological integrity of streams. If true, it is nonsensical to arbitrarily address some pollution problems and ignore others, based on whether a pollutant has been detected. It also deserves reiterating that most water quality-based requirements cannot be instituted without TMDLs. Without TMDLs, point source effluent limits cannot be set properly. Without TMDLs, nonpoint sources are not regulated to the extent required to meet standards. And, without TMDLs, narrative criteria, beneficial use support, and the antidegradation policy requirement to protect existing uses are ignored. The TMDL program makes sense out of spending millions of dollars a year on collecting information on water pollution because it makes use of that information to provide the public with clean water and to protect beneficial uses.

Threatened Waters

We agree with EPA that proper reading of the antidegradation policy that declining trends in water quality constitute a violation of water quality standards. While we do not believe that threatened waters should be a high priority for TMDL development, considering scare resources, by identifying them on the §303(d)(1) list states should be able to slow the rate of degradation to at least some degree. This identification will also alert point sources that a TMDL is needed to protect their interests, regardless of whether criteria are ultimately violated, to which they may respond with data collection to support such an effort.

Atmospheric Deposition

We strongly support the inclusion of waters partially or fully impaired by atmospheric deposition on the 303(d)(1) list. First, the statute does not provide for an alternative. Second, thoughtful people are well aware that there are inherent gaps between the media-specific federal environmental statutes. The TMDL program is an obvious gap-filler in this regard and should be used, not ignored. Third, the TMDL is an appropriate way to analyze the contributions made by atmospheric sources and to trigger necessary controls. Fourth, doing so brings scientific analysis to solving the problems of atmospheric deposition but does not fundamentally alter the regulatory scheme. We take issue, however, with EPA's suggestion that waterbodies impaired by atmospheric deposition should be assigned a low priority. First, there is a difference between complexity and priority. It should be the case that complex TMDLs are started earlier so that they are completed within the 15 year time frame because they will take longer. If they are postponed, because they have been deemed "low priority," that action will virtually assure they will not be completed by the end of the schedule. Second, waters impaired by atmospheric deposition include high levels of lead, mercury, and other airborne contaminants including acid deposition that is killing lakes across this country, even in Washington. These are not a low priority because they are causing health problems for people, fish, and wildlife.

Relationship to Antidegradation Requirements in Water Quality Standards

We agree with EPA's analysis of the role of the antidegradation policy in the TMDL program. However, we disagree with the statement in the preamble that "Tier 1 waterbodies may be listed if existing uses have been identified pursuant to 40 CFR 131.3." 64 Fed. Reg. 46023. The preamble supports this statement by noting that states "must" incorporate existing uses into their designated uses, pursuant to EPA's rules but in the event that they do not, members of the public can make demonstrations that they should. EPA cannot rely on either of these two approaches to protect existing uses. First, regardless of EPA's regulations, states simply do not protect all existing uses. An excellent example of this is the existing use of bull trout in Oregon and Washington. Notwithstanding either the states' knowledge of this species that requires particularly cold water for survival, or EPA's knowledge (having promulgated bull trout criteria for Idaho), neither state protects it. Oregon has promulgated bull trout criteria but has not identified where the criteria apply, rendering it nearly useless. Washington has not and will not in the foreseeable future done either. Second, the requirement to apply the antidegradation policy as written should fall squarely on the shoulders of the state, not on private citizens. It is ludicrous that something as important as not eliminating existing uses should be a voluntary affair. Instead, EPA's proposed rule should explicitly address this issue by requiring states to identify existing uses that are not protected by numeric criteria and listing them if they are impaired or water quality data demonstrate water quality is not sufficiently high quality to protect them.

Comments Sought

§130.25 We support EPA's reading of the statutory requirements which is also in the best interests of the public need to clean up polluted waters and protect beneficial uses of all kinds. We strongly support EPA's proposal to include waters where the offending pollutants are unknown at the time of listing, waters are impaired by atmospheric deposition and other nonpoint sources.

§130.26 We agree with EPA's interpretation of how antidegradation policies apply to 303(d)(1) listing requirements.

§130.26(a)(3) EPA should clearly explain in its rule that states are obligated to identify existing uses that are not protected, rather than adopt the approach discussed in the preamble.

a. Required Components of the List

Existing Requirements

No comment.

Proposed Rule

In discussing Part 1 of the list, EPA's preamble states that "[i]f the cause of impairment is determined to be pollution, no TMDL is required" and the water should be moved to Part 2. 64 Fed. Reg. 46024. This statement lacks clarity because in many, if not most, instances pollution and pollutants are found together, at least when someone attempts to determine if pollutants are present in unsafe amounts. We incorporate by reference the comments on this subject submitted to EPA by Robert W. Adler in this rulemaking process.

EPA's preamble states that the FACA Committee recommended maintaining waters on the list following TMDL development as a way to ensure a comprehensive accounting. 64 Fed. Reg. 46024. Rather, the recommendation demonstrates the Committee's emphasis on implementation of TMDLs to actually produce results. Comprehensive accounting is just one part of the primary purpose.

We are pleased with EPA's decision to include so-called "expected to meet" waters on the 303(d)(1) list. 64 Fed. Reg. 46024. However, we are very concerned about how Part 4 will function. First, it is worth noting that EPA's current regulation extending the explicit point source exceptions of the statute to a wide range of other programs is contrary to law. The Act is very clear on those exceptions, giving EPA no discretion to create more. This same analysis applies to EPA's new proposal to exclude from TMDL scheduling those waters that are expected to meet standards by the next listing cycle because of "controls enforceable by State or Federal law or regulation." Id. Even more troublesome is EPA's explanation of some controls that fall into this category. Unlike the language about expected to meet, the preamble gives examples of controls "which may achieve water quality standards," including "state regulations or local ordinances requiring erosion control, state laws requiring manure management practices, NPDES controls for point sources based on best available technology, and Habitat Conservation Plans [HCP] adopted under the Endangered Species Act." 64 Fed. Reg. 46025 (emphasis added). It is the "may" language, and the commentary on the large numbers of TMDLs in need of development, that gives away EPA's purpose in establishing Part 4 of the list.

There are very few nonpoint source programs - regardless of their enforceability - that are remotely likely to lead to water quality standards attainment in two years, or even four or five. Putting aside the fact that very few nonpoint source programs include enforcement, will EPA assess the likelihood that a state will actually use enforcement proceedings when it reviews a state's list for approval or disapproval? Will this not put EPA in an awkward situation; having approved §319 programs and grants, and others, will not EPA feel compelled to take an over optimistic view towards the efficacy of these programs? If so, EPA will simply open itself up to more litigation because as states load up Part 4 of their lists, they will be seeking to extend the very 15 year deadline that EPA establishes in these regulations. Although the regulations are not clear on the subject, one can presume that a state need not include Part 4 waters in its priorities and schedule which will lead either to a de facto extension for lack of sufficient time to do all the back-loaded TMDLs or EPA's view that Part 4 waters are excluded from the 15 year time frame in the first place. Both situations are unacceptable.

Moreover, how will a state be able to demonstrate that controls on NPDES permitted point sources will lead to water quality standards attainment in the absence of a TMDL? Only a TMDL can instruct the permit writer on how to set effluent limitations that are appropriate in light of natural contributions, nonpoint sources, other point sources, a margin of safety, and seasonal variations. EPA should not suggest that a wasteload allocation to a point source in the absence of a TMDL to evaluate the cumulative effects of multiple sources will be a sufficient basis upon which to issue an NPDES permit or to make assertions about the likelihood of attainment of standards.
Equally bad is EPA's assertion that any HCP could possible result in attainment of standards in the period of time under five years. First, HCPs generally address nonpoint sources that cause harm to aquatic species, not humans. This means, for example, that they address logging and similar activities that devastate habitat and lay waste to streams. Even in the best case, where no new human activity in an area affected by logging was allowed except restoration actions - which, by its very conditions would negate the benefits of an HCP - a state could not find that it would lead to attainment of standards within a five year time period, let alone two or four. Will states attempt to use this provision to avoid scheduling TMDLs if EPA includes it in the rule? Absolutely. This will, once again, lead either to EPA's having to disapprove lists or risk additional litigation, which is virtually guaranteed. Second, the "no surprises" policy that attaches to HCPs makes them incompatible with TMDLs and the Clean Water Act in general. EPA cannot sanction plans that lock in 50 or even 75 years of certainty in light of its understanding that if nonpoint source controls do not succeed they must be followed up by increasingly stringent controls. This concept is imbedded in the Implementation Plan proposal. Third, the vast majority of HCPs do not fully address Clean Water Act issues, including whether and when they will lead to attainment of water quality standards. For EPA to endorse postponement of scheduling TMDLs when it is clear HCPs are not a basis to believe that standards will be attained, would be inconsistent with the Act. Even the FACA Committee recognized, in the discussion of "TMDL substitutes," a concept it rejected, that HCPs would not constitute the equivalent of a TMDL (but if they did they could be submitted as such to EPA). EPA's approach of seeking to postpone a TMDL under such circumstances is contrary to the FACA's position on TMDL substitutes and its proposal to set priorities to address threatened and endangered species.

The preamble also mentions EPA's belief that "it is appropriate to provide time to allow controls such as these to attain water quality standards, especially in light of the large numbers of TMDLs that need to be established nationally." Id. at 46025. We agree that there are a large number of TMDLs that need to be established and that there are insufficient resources to do this. We disagree that EPA is justified in postponing the scheduling of TMDLs based on the same state, federal, and local laws and NPDES permits that not only have failed to lead to attainment over the years but allowed water quality to erode to the point of violating standards in the first place. Having ignored the legal requirement to develop TMDLs for most of the 25+ years it has been in effect, and having failed to require states to create and apply antidegradation policies and Implementation Plans to slow the rate of degradation of waters, EPA has only itself to blame for the huge number of TMDLs that are now necessary. The agency had best assess how to create efficiencies in the process rather than simply trying to create a method of postponing TMDLs for waters that will need them eventually, and which in the absence of Part 4 are likely to include many high priority waters. States will surely point to a range of failed regulatory programs and HCPs in an attempt to load up Part 4, and at least in the West many of those programs will address the very nonpoint sources that have resulted in the threatened or endangered status of species. How can EPA oversee a state's decision that control programs are sufficient to lead to attainment of standards without the analysis of a TMDL?

The proposed change in the listing cycle also has a significant effect on the use of Part 4. The longer the time frame that can be used to justify "expected to meet" findings, the larger the number of waters that will end up on this list, unscheduled for TMDLs. The longer the Part 4 list, the greater the de facto extension of time to the 15 year schedule. In fact, if EPA concludes that a five year cycle is appropriate, it must admit that it has just created a 20 year schedule for TMDLs, at the minimum. Then, in each five year period, legislatures will tweak nonpoint source programs ever so slightly and with a wink and a nod, the water quality agencies will place another chunk of listed waters on Part 4. The schedule will stretch out far beyond 20 years just to develop TMDLs, let alone implement them. Moreover, it is unclear from the regulations whether states can use Part 4 just once for a given waterbody or keep making new excuses about why a standards attainment is just around the corner. If EPA insists on using this provision, despite its illegality, it must include a caveat that this is a one-time deal, no exceptions.

The preamble also discusses the possibility that EPA could identify and describe situations where the proper technical conditions are not available to establish TMDLs. Id. We strongly oppose this approach, noting that the FACA Committee spent a significant effort in evaluating the full range of difficult conditions that exist. The Committee agreed that in some cases allocations had to be made that accepted current conditions and that in others more time might be needed to address difficult attainment issues, but it based this on a decision that TMDLs were appropriate for all of these situations. FACA Report at 46, 47. In fact, the Committee assessed numerous circumstances and identified which treatment was appropriate. Id. It appears that EPA has ignored this work. The only situation where the Committee could not agree was that of atmospheric deposition. There too, the Committee was able to identify ways of creating efficiencies, recommending that where numerous waterbodies were affected in the same fashion by the same sources, one TMDL could be done for all of them. Is atmospheric deposition a technically difficult source to address? Yes. Is it more difficult than a whole range of sources and pollutants and pollution? No. There are many, many extremely difficult water quality problems that will pose almost insurmountable challenges to the state and federal agencies that develop TMDLs to address them. The choice, however, is to use the sensible program Congress created and stretch the bounds of technical understanding to fit today's and tomorrow's needs, or to abandon these waters from the goals of the Act. If EPA decides in even one instance that a TMDL is not appropriate because of technical difficulty, it has opened the door for the whole program to slide through. There is simply no reason to believe that any water quality problem will be easy to evaluate and solve; it is only degrees of difficulty.

Other Options Considered

EPA asks for comments on the option of "whether to include a category for waterbodies for which there is some evidence of threat or impairment, but which would not be immediately scheduled for establishment of TMDLs." It concludes that this is not needed because the problem of less than conclusive evidence of impairment could be properly addressed during development of the TMDL. While it is correct that TMDLs will require data collection that will address data gaps, it is not correct to assume this solves the problem EPA has identified. The issue is whether these waters will be listed in the first place. In most instances, states will create listing criteria that avoid listing waters where there is only some evidence, but not conclusive evidence, of impairment. The solution to this problem is two-fold. First, EPA should establish listing criteria that prevent states from not listing waters where there is sufficient evidence of impairment. Second, EPA should create a new segment of the list for waters where there is some, but not sufficient, evidence. These waters should default to Part 1 of the list if data have not been collected by the next cycle that demonstrate the waters are in attainment with standards. In this fashion, EPA can avoid forcing states to list waters that may not be impaired, while not abandoning them altogether without the necessary follow-up monitoring.

EPA also considers the option of whether to list waters that violate standards "but for which other pollution control requirements or actions are planned or are being implemented that are expected to provide for standards attainment. Id. at 46025. The provisions in EPA's current regulation that allow for not listing waters violates the Clean Water Act. We strongly suggest that EPA should not continue this option. Moreover, with few exceptions the option will merely result in postponed TMDLs, not avoided TMDLs. In most cases, as has been discussed on related issues above, state, local and federal programs that have allowed standards violations to occur and persist will not remedy existing violations. Thus, such a provision will only serve to extend the 15 year schedule.

Comments Sought

§130.27(a) Overall, we support EPA's proposal, including listing waters where impairment or threat is of unknown origin, where a TMDL has been developed but standards are not yet attained, and those waters that are "expected to meet." Each of these has the benefit of focusing the TMDL program on the ultimate goal, attainment of standards. We have significant concerns with some of the details however.

§130.27(a)(1) Part 1 of the list is an appropriate response to impaired or threatened waters where the state has not yet made a determination of what the cause of the problem is. If, in the alternative, EPA does not include those waters, states will have an incentive to not identify the causes of problems, which will then not be rectified.

§130.27(a)(2) We agree with EPA's view that the statute requires listing of waters impaired by pollution in the absence of pollutants and that doing so creates a comprehensive accounting of the status of the nation's waters. We also agree that the statute does not require TMDLs for these waters. However, we believe that the TMDL's quantitative analysis would offer significant benefits to efforts to restore waters affected by pollution only. If EPA's goal is to meet water quality standards and the goals of the Act and to protect beneficial uses, requiring TMDLs, or even encouraging TMDLs, for waters with pollution only be far better than relegating them to a "dead" list forever. Our greatest concern is that many waters will be listed by states on Part 2 that are affected by pollutants that are related to their impairment by pollution. In particular, waters with insufficient instream flow will generally exhibit excess pollutants, if an attempt is made to monitor them. States more than likely will list these waters where there are no data on water quality on Part 2 and avoid revisiting the question of whether pollutants are an issue as well. EPA's regulations should guard against this by requiring states to collect data on pollutants for Part 2 waters to avoid having them moved to Part 1. That should be the default position.

§130.27(a)(3) We strongly support Part 3 as a way to emphasize the importance of reaching the Act's goals rather than merely focusing on developing the TMDLs.
§130.27(a)(4) We support EPA's rectifying the illegality of its current rule by requiring listing of expected to meet waters. However, we are extremely disturbed by the problems with Part 4, as discussed above. Specifically, the regulation should specify 1) placement on Part 4 is a one-time finding, no exceptions, 2) there must be an analysis that demonstrates why waters are expected to meet standards by the next list that addresses what the necessary controls are and makes concrete assurances that they will be in place. We do not understand how EPA can make such findings in the absence of TMDLs. In addition, we cannot fathom why EPA thinks it is good policy or consistent with the Clean Water Act to allow waters to remain without TMDLs because a state "document[s] that the failure to attain water quality standards is due to failure to comply with applicable technology-based requirements." That basis is not only ridiculous but inconsistent with the Act.

§130.27(b) We support improving the format of the states' lists by requiring them to at a minimum identify the pollutants or the class of pollutants causing the impairments. However, we do not think EPA's proposal goes far enough to specify the format and details of lists. We urge EPA to adopt the format used by Oregon and Washington. This uses a "decision matrix" that includes: the segment number and name of the waterbody, the parameters considered, the source of the information analyzed, a summary of the analysis of the data and information, the rationale of the decision to list or not list, and the ultimate status. By standardizing the format of states' lists, EPA will increase efficiencies in the program, at the expense of very minor loss of flexibility by states. One of the benefits will be that EPA, Congress, and the public can evaluate states' lists against each other, thereby giving clear indications of what the differences are between states' programs, success in attaining and maintaining water quality standards, the standards themselves, and monitoring programs. In this highly computerized age, there is no reason not to have standardized methods allowing such analysis.

a. Assigning Priorites to Listed Waterbodies

Existing Requirements

We agree that the statute does not explain how the statutory factors for priority-setting should be taken into account. However, reading this provision together with others from the Act, demonstrates the following. While it may be argued that the statute's reference to the "uses to be made of the waters" is ambiguous, establishing what those priorities should be is clear in the context of the Act and its implementing regulations. These include: 1) the stated goals of the Act (e.g., CWA § 101(a)(2) "interim goal of water quality which provides for the protection and propagation of fish, shellfish, and wildlife and provides for recreation in and on the water...," 101(a)(3) "discharge of toxic pollutants in toxic amounts be prohibited"); 2) the goals of the TMDL program (e.g., CWA § 303(d)(1)(B) "assure protection and propagation of a balanced indigenous population of fish, shellfish, and wildlife"); 3) the purposes and requirements of establishing water quality criteria (CWA § 303(c)(2)(A) "standards shall be such as to protect the public health or welfare, enhance the quality of water and serve the purposes of this chapter," 40 CFR 131.11(a) "the criteria shall support the most sensitive use"); and 4) the logic of using the TMDL program to support other laws that are designed to protect uses (e.g., the Endangered Species Act).

EPA notes its "belief that the statutory factors are not exclusive." 64 Fed. Reg. 46025. We disagree. The statute does not say, "here are the factors, now proceed to ignore them.". In EPA's list of other factors states could use it includes "vulnerability of particular waterbodies as aquatic habitats." This is consistent with, not different from, the language of the statute. On the other hand, the "aesthetic importance" of a waterbody is not consistent with the analysis presented above. States should not be allowed to set priorities based on public popularity or protection of uses that are not as sensitive and vulnerable as others.

Proposed Rule

We agree that EPA's proposed rule should eliminate the targeted list for two years. However, in doing so, we note that the proposed "comprehensive schedule" requirement is less than clear. This could be rectified if EPA made clear what level of detail and specificity is required to constitute a "schedule" for purposes of this section.

Other Options Considered

EPA considered a number of other options listed in the preamble. 46026. The first of these was a "more prescriptive approach" that would specific factors states would have to consider. We support this approach, especially in light of the long schedules for TMDL development. The second approach was to defer entirely to state discretion on the matter of ranking, not even requiring a high/medium/low ranking. This approach is flawed because states have already amply demonstrated their inability to develop TMDLs, and to develop them on a priority basis. EPA's third option was not to identify either threatened or endangered species presence or public water supplies with violations of MCLs as "high" priority. Even the FACA Committee, which had mixed opinions about the issue of priority setting, was able to agree on certain high priorities. EPA should be able to do the same. Fourth, EPA considered a vague option of making human health and species concerns one, but not a determinative, factor in determining "high" priorities. This approach would be pointless. As a sixth option, EPA considered whether the category of "extremely difficult to solve" problems should be treated differently in priority-setting. EPA appears to have adopted this option by suggesting that difficult TMDLs should be lower priority. We discuss elsewhere the flaws in this argument, namely that because they will take longer, they should be started earlier. This approach is also plainly contrary to the statutory factors.

Evidently, EPA did not consider the detailed approach to ensuring that high priority TMDLs get developed that was recommended by the FACA Committee. FACA Report at 21. That is a shame. We understand the reason is because states, such as Washington, made the point that their list of impaired waters consists of many more high priority waters than other states because of extensive listings of threatened and endangered species. However, Washington has more segments listed for fecal coliform violations (565) than temperature (455). Washington 1998 § 303 (d)(1) list. Western states' protests are more excuses than anything. They are the same states in which TMDL litigation has been ongoing for over ten years and the same states that allegedly have been developing TMDLs. It is also true that the types of impairments that affect threatened and endangered species are the very ones that EPA suggests are best approached using watershed-scale analyses instead of segment-by-segment. So, states in these situations will accomplish TMDL development with many economies of scale.

To implement the FACA Committee's approach would put pressure on these states. But why should there not be pressure? These species are, after all, on the verge of extinction, elimination from the planet. For all the talk about the relevance and importance of integrating the Clean Water and Endangered Species Acts, little has been done to make this a reality. What more important purpose could the TMDL program be put to than to prevent permanent eradication of species from the face of the earth, except perhaps the build-up of persistent, bioaccumulative toxic contaminants? Why should Washington maintain the flexibility is has seized to continue to develop TMDLs for excess bacteria when so many of its waters are impaired by pollutants that are critical to the survival and long-term health of species including people? If EPA evaluates these states, and finds that the FACA proposal really would be too onerous, it could modify the proposal instead of abandoning it altogether. There is no reason to "throw the baby out with the bath water." EPA should return to the FACA recommendation and devise an approach that will ensure states will develop highest priority TMDLs first, and within some reasonably short time frame.

Comments Sought

§130.28(a) We support EPA's attempt to increase the expectations of how states will establish ranking by priority, according to the requirements of the statute. However, we do not believe that grouping the list into three categories constitutes "ranking," which means to put in order. Aside from the dictionary definition, the benefit of a more carefully thought out ranking plan would be to serve as a check on the schedule of TMDL development. It is our experience that states ignore some obvious high priorities in favor of addressing less controversial issues.

We also urge EPA to require states to develop their priorities on a statewide basis rather than leaving open to states to establish priorities basin-by-basin. This latter approach is inconsistent with the law and has the effect of making many waters higher priority based on nothing more than which watershed or basin a state has chosen to address first. This incorrect reading of the law allows states too much latitude within a 15 year time frame for developing TMDLs.

We urge EPA to adopt the following proposal that would divide Part 1 of the list into tiers, differentiated by priority, with waterbodies ranked within each tier. Each tier would be established based on the priority of beneficial uses being protected. Within each tier, the ranking would be done on the basis of the severity of the pollution. The result would be an assessment of the risks posed to waterbodies that is obtained by combining priority uses and severity of pollution, the statutory factors. In addition, a further step would allow for limited consideration of other factors. Because the exception focuses only on harms that may occur due to waterbodies being ranked lower in priority setting, these exceptions ensure that the priority setting system has the flexibility to obtain needed environmental results without entirely turning it on its head.

The combination of the statutory factors must be driven primarily by the uses. Since protecting the uses is the ultimate goal, the highest priority uses should get the first attention, ranked in order of pollution severity. Seen another way, the fact that a waterbody has really terrible pollution problem should not make it a higher priority than a waterbody with a less serious pollution problem that is affecting a highly sensitive and important use.

1. STEP ONE: Determine tiers by priorities based on the uses.

Using this approach, uses that could be deemed high ranking would include: 1) native aquatic species (fish, shellfish, wildlife) including highly sensitive and threatened or endangered species, and 2) human health from fish consumption (normal, and subsistence level), drinking water, and contact recreation. The highest tiers should be where water quality is causing critical impacts to highly sensitive beneficial uses that cannot be treated or prevented. The lowest uses would be where poor water quality is affecting the least sensitive (i.e., non-biological) uses, such as aesthetics (scenic or non-health problems), livestock watering, and industrial water supply. Two or three tiers in between would separate out uses by their sensitivity to pollution.

2. STEP TWO: Determine the ranking within the tiers by severity of the pollution.

Severity of the pollution can be interpreted in several ways including, but not limited to: 1) conditions (levels of a pollutant and whether multiple pollutants are present), 2) types of pollutants such as: persistent bioaccumulative versus short term, severe versus less severe effect on uses, and pollutants that affect sensitive uses (e.g., salmonid) versus pollutants that do not (e.g., aesthetics).

RESULT OF STEPS ONE AND TWO: An assessment of the risks is obtained by combining priority uses and severity of pollution. Combining the highest priority uses and the highest priority severity concerns (pollutants/conditions) results in an assessment of highest risks. For example, if the uses in waterbody X are more sensitive than those in waterbody Y, the same amount of pollution will cause a higher risk to the uses in waterbody X. If the uses are the same in waterbodies X and Y, but the pollution is greater in Y, the risk to the uses will be greater in waterbody Y. This combination of uses and severity should drive the priority-setting process.

3. STEP THREE: Consider harms of all types that may occur due to priority-setting.

There are other considerations in priority-setting that could be used to alter the priorities or assist in making scheduling choices between similarly ranked waters. The approach taken here is to note the downside of not allowing exceptions to a strict ranking by uses and severity rather than diluting the priority-setting process with lots of considerations of equal weight and rendering the process meaningless. These considerations includes, 1) the possibility that priority-setting might result in foregoing opportunities to take advantage of other programs such as FERC relicensing for dams, NPDES permit renewals with potential for large reductions in waste loads, 2) large dredge and fill projects, and 3) the ease with which TMDLs could be done for lower priority parameters at the same time as higher priority parameters for the same waterbody.

EPA could allow states to rank only its highest tiers, forgoing the more specific ranking until later in the schedule.

§130.28(b)(1) We support EPA's designation of certain high priorities. Serious, unavoidable, threats to public health should be among the highest of priorities. So, while we support the use of MCL violations where there are public drinking supplies, there is no basis for treating as anything but high priority those waters that are used for fish consumption where the fish are contaminated with toxic pollutants with long-term deleterious effects. Without suggesting that human pathogens pose insignificant health risks, EPA should focus on unsafe levels of the contaminants in the family of "endocrine disrupters," those pollutants that are persistent in the environment and alter reproductive health and the development of children. In doing so, EPA should require states to place as high priorities, those waters where consumption of contaminated fish is an environmental justice issue, because of the increased risk of irreversible adverse health effects to subpopulations of people. For the same reasons, EPA should instruct states to find that unsafe levels of toxic contamination for wildlife is a high priority. While many species are not threatened or endangered, levels of toxic contamination is having demonstrable, irreversible, long-term health effects on animal populations. For example, in the Lower Columbia River, it appears that mink and otter have stopped reproducing due to high levels of toxic pollutants. Is this anything but a high priority?

§130.28(a)(2) We agree that threatened and endangered species (T&E) are a high priority unless the identified impairment does not affect the species. We take issue with EPA's notation that the T&E species must be "present" in the waterbody to warrant this designation. Many T&E species are not present because they are on the verge of extinction because pollution has made their habitat unusable. EPA must remove the requirement that T&E species be present and replace it with language that indicates the waterbody is essential to support recovery of species.

§130.28(a)(3) Given the substantial threats to the existence of species, to human health, and to the well-being of aquatic and aquatic-reliant species, we strongly disagree that EPA should allow states to include as high priorities "historical, cultural, economic, and aesthetic uses." EPA has and continues to include so many factors in priority setting that it makes the process worthless. On the other hand, EPA has avoided making any reference to obvious high priority issues, namely where if a TMDL is developed too late, it may very well be useless. Examples of this are FERC relicensing projects, opportunities that arise only once every 50 years. Likewise, major dredging operations, such as deepening shipping channels, are an irretrievable commitment of resources with potentially devastating effects on water quality and aquatic life. If a TMDL is done after such major actions, it will not inform the various administrative and legal processes that consider its environmental effects. These should be high priority waters.

§130.28(c) We support this provision to require states to explain their priorities.

§130.28(d) This section is more of the same, EPA watering down the priority-setting process to render it meaningless. While we support obtaining efficiencies, in developing TMDLs, and believe the vulnerability of waters is a key issue, as discussed immediately above, the other considerations in the proposal are not focused on meeting the goals of the Act.

§130.28(e) Because we think this area is ripe for abuse, we support EPA's requirement that states identify and explain each factor used in assigning priority rankings.

§130.29 We support the requirement to maintain impaired and threatened waters on the list until they attain standards as an important policy to drive the implementation of TMDLs, the ultimate goal of the program. However, as discussed above, EPA must ensure that attempts to delist are based on credible data.

§130.30(a) As discussed in the comments on the preamble, we do not support changing the listing cycle to four or five years. If EPA feels this is necessary, we strongly urge the agency to require at least two more two-year cycles prior to instituting a four year cycle. There are such significant problems with states listing methodologies and failure to obtain data and information, that to create such a long period between now and solving those problems is to undermine the progress made to date in implementing the Act. In addition, we oppose the change to the listing cycle because it will alter the definition of threatened waters and expected to meet waters. We are particularly concerned that allowing expected to meet waters to project attainment in four or five years will simply create more baseless analysis than if the period were two years and will extend the overall schedule for TMDL development by that amount of time.

§130.30(d) As with all public activities discussed below, we urge that this proposed provision provide that EPA will send its order to any mailing lists of interested parties maintained by EPA and the state. States currently use mailing lists as the best method of reaching interested citizens and EPA should meet this standard. All stakeholders agree that access to information is key to success of the program.

§130.30(e) We support EPA's ability to establish lists of impaired and threatened waterbodies where states request the action or when EPA believes a state is not likely to do so. This program has been stymied due to failures of states to act and EPA to act in the absence of state action. The program will be far more successful if EPA does not allow this pattern to continue.

a. Establishing a Schedule for TMDL Development

Existing Requirements

EPA notes that the FACA Committee recommended that EPA's regulations require states to develop expeditious schedules of not more than 8-15 years for establishing TMDLs for listed waters. EPA should consider this recommendation carefully because there are a few areas where it applies that have not been considered. One area is how the recommendation applies to Washington, D.C. and other smaller entities, authorized Tribes and Territories. In these cases, smaller entities that have extremely few TMDLs to develop in relation to full-size states should be held to a different time frame. EPA should consider a more appropriate approach to these entities so that they do not take advantage of the extended schedules required by states that have an enormous work load in comparison. Second, EPA must consider how its various proposals such as the Part 4 scheduling deferral, discussed elsewhere in these comments, could significantly alter this outcome of this recommendation, both individually and in combination. As is pointed out above, combining the proposed Part 4 with the proposed change of listing cycle to five years would have a staggering effect on the ultimate scheduling of TMDLs for currently or imminently listed waterbodies.

Proposed Rule

The preamble discusses two ways in which states may choose to schedule TMDLs based, or not based, on their priorities. Id. 46028. The first of these, "focusing on waterbodies concurrently that are impaired by a particular pollutant or category or subcategory of sources of that share common ecosystem characteristics," has merit so long as states are not allowed to interpret this as a go-ahead for so-called TMDLs that purport to address huge swaths of land. For example, a statewide TMDL for logging would not meet the purpose of the Act. There have already been discussions about such approaches in Washington state that the EPA regulations should guard strongly against. As is discussed elsewhere in these comments, TMDLs for large geographic areas will miss the boat because they will not include the specificity required to address real world sources and real world effects in a non-homogenous landscape. Likewise, if a state were to choose a less-than-high priority pollutant or source category, this approach would have a significant effect on whether TMDLs were established in a timely fashion for high priority pollutants and waters.

The second approach EPA envisions is use of the "watershed approach" which it supports so long as it does not cause states to "depart too far" from their priority rankings. 64 Fed. Reg. 46028. The problem with this is how will EPA make sure that states do not depart too far from priorities? EPA has said that it will only comment upon but not change or substitute schedules. In fact, EPA will only approve or disapprove a state's priorities. So, if a state uses any rationale, including but not limited to using the watershed approach not in schedules to avoid doing TMDLs according to its priorities, it will do so with impunity. Assuming that EPA does not want to change its proposal with regard to its review of schedules, the only other option is to require states to develop TMDL schedules that mesh with their priorities and to ensure through its regulations that the priorities are set on basis that meets the needs of the most sensitive beneficial uses first. Otherwise, the entire approach is a fiction, and there is no reason to believe it will make an iota of difference. This too is bound to lead EPA to the courthouse more often than not because the longer a schedule, the more important the priorities and the scheduling of TMDL development become. Looking at the combination of the proposed Part 4 with the option of a four or five year listing cycle and the possibility of states' creating schedules that avoid high priority waters, it becomes clear that environmental organizations should advocate for the shortest possible schedules just to ensure that TMDLs get done, and that TMDLs for high priority uses and waters are done sometime in our lifetimes. Again, EPA's proposal suffers from a failure to take the sum of its suggestions together, whereupon it becomes clear that the entire program may be built on a very unstable foundation.

Last, EPA mentions that "it may not make sense for States * * * to individually schedule every TMDL" recommending that states "schedule groups of TMDLs, on a watershed or some other appropriate basis, for TMDLs to be established in later years of the schedule." 64 Fed. Reg. 46028. We believe that EPA should be extremely clear about what the scheduling expectations are. The combination of the proposed regulation and this preamble language leaves it too unclear and will lead to great inconsistencies between states. We recommend that EPA require schedules backed up by work-load assessments, in conjunction with clear definitions of medium and low priority waters. The reason for scheduling is multi-fold: to have a clear accounting of whether states are maintaining progress, for EPA to ensure that the pace of TMDL development is proportional, to ensure that the public and the legislature understand what the resource needs are, to allow stakeholders to assess what they want to be able to do in advance of TMDL development, and to make sure that the agencies properly assess the work loads associated with TMDLs (because some TMDLs require more work than others and more years of advance data collection). On the last point, it is a mistake for EPA or the states to assume that each and every TMDL will take a certain number of years, such as five. Some simply will be more complex and will require more advance thought and/or data collection. This means that for them to be completed by their scheduled time, they must be started earlier. By encouraging states to postpone thoughtful analysis of the resource needs of future TMDLs, EPA is increasing the likelihood that the schedules will not be met.

Other Options Considered

We agree with EPA's conclusion and its reasoning that states should develop comprehensive schedules rather than two year targets. 64 Fed. Reg. 46028. We support any EPA effort that will assist states in developing comprehensive work plans that will also make reaching these schedules, as discussed above.

Comments Sought

§130.31(a) We agree that EPA should include in its regulations provisions for states to submit schedules for all TMDLs that are required. We are concerned that the requirement of a schedule may be interpreted in different ways, as the current policy guidance (Perciasepe memo on pace and implementation) is. EPA should decide what a workable format is and require all states to use that format for submitting schedules to EPA. This will guarantee that what states provide EPA can adequately evaluate and that the public will also have something it can understand.

§130.31(a)(1) We also agree that a period of not more than 15 years is warranted in order that states and/or EPA develop thoughtful TMDLs and Implementation Plans. We are very concerned, however, about supporting the 15 year period given the other proposals EPA has included in these rules, namely the combination of Part 4 lists with the possible change in listing cycle to four or five years. While it has been and continues to be our position that an adequate amount of time is needed to develop quality products and that that goal must be balanced with the need to implement TMDLs to restore impaired rivers, extending the schedule out to 20 years is excessive. We have no doubt that will be the result if EPA promulgates the above combination of proposals. Moreover, the longer the schedule the more important priority setting is. Based on the progress of TMDL development to date, we are not at all confident that highest priority TMDLs are being set first. Based on the proposed rule language we lack confidence that this rule change will accomplish this goal. The encouragement to make complex TMDLs lower priority also will encourage states to fail to meet deadlines because complex TMDL require more, not less, time and should be started earlier. Finally, we are not confident that the proportional pace of developing TMDLs within schedules is or will take place. This makes difficult our continued support for a 15 year schedule.

§130.31(a)(2) We strongly support the goal of proportional pacing of the TMDL workload. However, we see no indication that EPA will have the ability to scrutinize states' submissions because there is no requirement for states to submit workload estimates for each TMDL. Equally important, there is every indication that if a state fails to submit an appropriate schedule, EPA will do nothing. In a climate of continued state recalcitrance, both intentional and inadvertent, that is hardly the appropriate approach for EPA to take.

§130.31(a)(3) We believe that EPA should require, not suggest, that states schedule TMDL development in accord with priorities, with the sole exception of efficiencies to be obtained by addressing multiple pollutants in one effort. Again, this is based upon observing states that appear to establish TMDLs in a reverse-priority order.

§130.31(b) We support EPA's requirement that states submit schedules with their lists. The proposed rule does not specify, however, whether states should identify changes in priority ranking or scheduling made since the last submittal. This would greatly assist in public and EPA review of new proposed schedules. We support a change of deadline from April to October because we think that it will allow states adequate time in which to incorporate the results of fall water quality monitoring for toxic contaminants into new §303(d)(1) lists.

§130.31(c) EPA's not approving or disapproving states' schedules virtually guarantees that states will not conform to schedules or meet schedules.
a. Submission of Lists, Priority Rankings, Listing Methodologies, and Schedules to EPA

Existing Requirements

No Comment.

Proposed Rule

EPA invites comments on its proposal to assert its discretionary authority to establish lists of impaired waters. Id. at 46029. We support this proposed action because, even in the face of extensive litigation, states continue to seek ways to avoid the TMDL program including by taking no action.

EPA notes that it will not develop a schedule where a state has proposed one that is clearly deficient or where it fails to submit one. Id. at 46029. EPA notes that it will provide comment where it believes schedules are inadequate, and "would expect" the state to address those comments. Id. This is an unrealistic approach to a critical area, as discussed above. Due to the vulnerability of EPA to states' failure to develop TMDLs in a timely way, EPA should create the same discretionary authority to order schedules for TMDL development as it is for creating lists for recalcitrant states. EPA supports its position that approval or disapproval of schedules is not necessary because EPA reviews the priorities in making its determination on the submitted lists. Id. at 46029. As we have already discussed, not only is there room for tremendous abuse in creating of priorities, there is also an almost complete disconnect between the priorities and the schedules. Therefore, this argument is not supported.

Other Options Considered

EPA notes that the FACA endorsed the two year listing cycle but says that it remains open to the possibility of switching to a four or five-year cycle. Id. at 46029. The reason the Committee endorsed the two year schedule is that every participant could agree, no matter how their interests varied, that current lists are not accurate depictions of TMDL needs in states. The process of listing is in its infancy, no matter that most have been blessed by EPA's approval. The majority of states are still wrestling with their listing criteria, have not sought out readily available data and information that must and should be used, and have not submitted their lists to an ideal level of public scrutiny due to lack of sufficient public processes. For example, FACA members could agree that it was often extremely difficult to obtain draft lists from state agencies despite multiple attempts. This is similar to our experience in Oregon. Here, the state initially refused to adopt our recommendation to adopt the public process for listing used by Washington state. This includes key items such as providing the public with an opportunity to comment twice (once to comment on the proposal and to submit new data and the second to comment on any new data) and creating a "decision matrix" that clearly lays out the basic information on a potentially listed waterbody as well as the data and information sources and the agency's analysis of violations. It was only when industry in Oregon objected to the state's process because it did not provide sufficient information that the state adopted our recommendation. The point is that all interests consider themselves benefitted when the public process works. Right now, when relatively few people think the lists are sufficient, is not the correct time to create a significant time delay in making needed corrections, wether that means listing or delisting. If EPA is determined to make this change, it should establish two more cycles of two years each, followed by a change to a four year cycle.

The preamble states that one reason a two year cycle may be inefficient is that generally there are not significant changes in water quality over such a short period of time. Id. at 46029. That really is not the issue, and EPA knows it. The issues are stated above. Again, once lists are more accurate depictions of reality than not, the frequency of the listing cycle is less important. Unfortunately, we are far from that result.

Comments Sought

Comments on the submission to EPA of lists, methodologies, priorities, and schedules are discussed in the appropriate sections of the proposed rule above.

a. Proposal to Change List Submission Deadline to October 1, 2000 in the Existing TMDL Regulations

Proposed Rule

As we understand from EPA that this proposal will be supplanted with a new proposal to postpone the 2000 list, we refrain from commenting on this proposal.

Comments Sought

No comment.

Subpart D: WATER QUALITY PLANNING AND IMPLEMENTATION

Question No. 5: What are the Proposed Rule's Requirements for TMDL Establishment and EPA Review of TMDLS Submitted by States, Territories and Authorized Tribes?

a. Minimum Elements of a TMDL Submitted to EPA

Existing Requirements

No comment.

Proposed Rule

We strongly support EPA every time it reiterates that TMDLs must demonstrate that they will lead to attainment of water quality standards. The preamble, like the guidance, makes a references to TMDLs needing to identify the "pollutant load or [sic] load reduction necessary to assure that the waterbody will attain and maintain water quality standards, including aquatic or riparian habitat, biological, channel, geomorphological, or other appropriate conditions that represent attainment or maintenance of the water quality standard. Id. at 46031. This statement, which we strongly support, seems to indicate EPA's understanding that what might be termed "pollution" issues - the physical and biological characteristics that affect a river - are crucial in many cases to a finding that a TMDL will lead to attainment of standards. For example, the loss of meander in a stream affected by overgrazing or similar activities will prevent attainment of standards including protection of beneficial uses until the meander is restored. Notwithstanding EPA's observations, however, the rule elsewhere appears to undermine this notion that the TMDL must address the physical and biological considerations as well as those that are chemical in nature, specifically with regard to Part 2. We urge EPA to promulgate rules that ensure TMDLs are used to restore all the nation's impaired waters, and that address all the reasons for waters being in violation of water quality standards. To do otherwise is to misunderstand the fundamental dynamics of how water systems work: pollution and pollutants are integrally tied together.

We agree with EPA that TMDLs should be expressed in terms that are appropriate to the characteristics of the waterbody and pollutant combination. We urge EPA to include more discussion of this issue in the guidance, and to create a reservoir of information about this subject in an on-going way, because this issue is both critical to the relevance of TMDLs and is not well understood. EPA must guard against encouraging TMDLs that are essentially paper exercises. Specifically, just as important as the characteristics of the waterbody and the pollutant are the sources and how they are regulated. In other words, the terms used to express a TMDL is used to express must also be tied to implementation details.

We are concerned with EPA's discussion of temperature in the preamble. While we agree that a daily load of temperature is not useful in and of itself, we strongly take issue with EPA's comment that the "resultant temperature in the waterbody is not as important as maintaining the range required by the aquatic life through different seasons and climatological events." 64 Fed. Reg. 46031. The resultant temperature is of critical importance. When daily temperatures exceed those that aquatic life require to survive, especially in light of removal of thermal refugia upon which those species may have relied in naturally warmer water, the daily and weekly temperatures are critical. Temperatures are also key to important life cycle stages such as the beginning of migration and spawning. Temperature ranges are not the issue here, rather specific temperatures that do not inhibit or postpone critical functions are. It is wrong, therefore, for EPA to conclude that "an allocation of pollutants causing changes in temperature is often better expressed as seasonal or monthly averages."

While we helped formulate the FACA recommendation on extremely difficult to solve problems and concur with it, we believe that EPA must pick up where the Committee left off, namely on putting some context for the idea that some problems will take longer to get to attainment. A basic grammatical rule is that if something is "more than" it must be more than a named thing. In this case, longer time frame must be measured against an ordinary time frame. Yet EPA declines to suggest either a default time frame for attainment, a target time frame for most TMDLs which could be adjusted up or down, or even a process by which that time frame could be determined by the state and reviewed by EPA as adequately expeditious. It does not make sense.

Waterbody Name and Geographic Location
No comment.

Identify the Pollutant Load

EPA's discussion in the preamble is troubling because it implies that where numeric criteria are available, there is no other next step. There is a next step. The law does not define standards as including beneficial use support, narrative criteria, and the antidegradation policy for no reason. In fact, all of these components apply to TMDLs, particularly to the setting of the acceptable pollutant load. EPA may make the TMDL process appear more simple to omit a discussion of these required items but it does not enhance the likelihood of success of TMDLs. What purpose will be served if numeric criteria are met - which in many instances EPA knows are not protective of uses - and uses remain impaired? States and EPA will be accused by the regulated community of changing the goal. Expenditures of funds will have been met but efforts still be found wanting. The public will still not have standards met. There is no point in hiding this particular legal ball.

Identify the Deviation from the Pollutant Load

No comment.

Source Categories, Source subcategories or Individual Sources

We strongly support EPA's requirement to identify sources. We are troubled, however, by the omission of one large source of impairment: loss of stream flow. Where removal of instream flows in any way alters the ability of a waterbody to support beneficial uses or to dilute pollutant loads, the TMDL should identify existing and reasonably foreseeable withdrawals. We are also troubled that, despite EPA's correct assessment of the need to address physical and biological characteristics and to tailor total load findings to the characteristics of waterbodies and pollutants, the agency does not include a discussion in the preamble of what that means for source identification. Specifically, TMDLs done to address land use practices in the Northwest have grown in sophistication, and therefore usefulness, by identifying specific hazard areas within watersheds and streams. Without this, the TMDL is only paperwork rather than the helpful and directive action plan EPA wants it to be.

Wasteload Allocation and Load Allocation

EPA's preamble notes that for nonpoint sources "allocation of pollutant loads to categories or subcategories of sources may be appropriate, especially if measures to reduce these loads are implemented for a whole category at once." This is a dangerous area into which for EPA to venture. If nonpoint source allocations are not done with enough specificity, the entire TMDL program will be pointless, no pun intended. As difficult technically as this area is to address, there must be clear goals and milestones against which land practices are measured in order to reap the environmental benefits of the TMDL program and to achieve some measure of equity between point and nonpoint sources in taking responsibility for obtaining clean water. If the regulations are not specific on this topic, states such as Washington, which is poised to pass a law in the 2000 legislative session, will undermine the TMDL program by grouping all nonpoint sources together. Lack of specificity will undermine accountability, from nonpoint sources and the agencies that administer regulatory and nonregulatory programs to control their polluted runoff.

Margin of Safety (MOS)

While we disagree with EPA that states should be allowed to use conservative assumptions as the basis for the margin of safety, preferring the clarity of an allocation to it instead, we urge the agency to require that TMDLs quantify and document the conservatisms. Otherwise, the public and EPA are left with no basis upon which to make the assumptions that the writer of the TMDL has made. We support the language in the preamble with the exception that EPA does not specifically note that quantification is required. Without at least a professional judgment on the quantification, there is no basis upon which to judge whether the margin is adequate with regard to the technical uncertainties that exist elsewhere in the TMDL.

Seasonal Variations

We support EPA's explanation of how seasonal variations can change critical conditions that affect loading.

Allowance for Future Loading

No comment.

Implementation Plan

We strongly support EPA's proposal to include TMDL Implementation Plans as part of TMDLs under 303(d). We will not reiterate the reasons given in the preamble but endorse them all. In addition, the 303(e) program is a moribund program and will require too much time and too many resources to resurrect in any timely manner. Second, requiring Implementation Plans to be submitted as part of TMDLs will ensure that the plans are submitted concurrently with the TMDLs. Without concurrent submission, it is virtually impossible to evaluate whether the TMDL's analysis and load allocations to nonpoint sources is correct and will be meaningful in the real world. Third, plans will be held to a higher standard under (d) than (e) because they will be scrutinized by EPA and the public as part of the TMDL package. Fourth, doing so will ensure that the Plans will get done, as opposed to not being done, because they are part of the definition of an approvable TMDL.

There has been some discussion of the benefits of "decoupling" TMDLs and Implementation Plans. Some argue that doing so will allow TMDLs to be developed in a timely manner without being slowed by the process of determining how the TMDLs will be implemented. We strongly disagree. We are lucky to have experienced reviewing a draft TMDL that EPA prepared that, while not meeting the definitions of the proposed rules, included some level of implementation planning. This was the South Steens TMDL for Oregon. South Steens Water Quality Management Plan, dated June 22, 1998, and Total Maximum Daily Load, Public Notice Dated: July 10, 1998. In our extensive review, we learned much about the benefits of evaluating a TMDL with an Implementation Plan in hand, as well as evaluating an Implementation Plan with a TMDL in hand. We note that just as the public benefits from this, so does EPA, and it is EPA that ultimately must make the judgments about approving or disapproving TMDLs, regardless of whether they are defined to include Implementation Plans. Presumably EPA wants to ensure that this expensive exercise is a productive one.

One set of benefits from evaluating TMDLs and Implementation Plans concurrently is the issue of addressing the types of issues that arise with nonpoint sources including, on the one hand, hazard areas (e.g., landslides, undercut banks, etc.) and sensitive areas that require additional or different protections (e.g., wetlands), on the other. While both the TMDL and the Plan could discuss both of these situations independently, the analysis of the TMDL and control actions of the Plan should be as connected as possible. This connection is far more likely to take place when the documents are prepared together for simultaneous submittal. If the TMDL is prepared for EPA first, EPA will be forced to take action on it without the benefit of knowing what the proposed solutions are. With nonpoint sources, since technical analysis is difficult and in its infancy, it is difficult to draw a bright line between analysis of problems and proposals for solutions. That is why the surrogate measures approach is so appealing; it allows us to venture into a realm where standards and TMDLs are expressed in terms that are closer to the types of control actions that are needed (e.g., temperature loads = amount of shade = tree height, tree species, width of stream side vegetation = fencing, reforestation, changed logging practices, etc.). With the TMDLs and the Plans submitted together, this proximity of analysis with control actions will only improve. In the alternative, EPA will see TMDLs that could be as simple as indecipherable numbers that make allocations to nonpoint sources (but must be approved or disapproved) and Implementation Plans that fail to address the technical analysis of associated TMDLs.

It is all a matter of connecting the dots. The two dots here are: 1) the technical analysis of loading and other (biological and physical) goals that represent standards, alone with a determination of the amount of loading and other improvements that is required to get there, and 2) the controls needed to curtail the loadings and other effects. In making those connections, the agencies must address the efficacy of previous and current attempts to remedy identified problems, and address the uncertainties that abound (what is the baseline goal? what changes in land practices are needed to attain the goal? how do the physical, chemical, and biological factors interrelate? what information will need to be collected as implementation goes forward to address the range of uncertainties?). Again, addressing these issues should be as seamless as possible in order to link identification of problems with the control actions necessary to resolve them. If the two are not side-by-side, it will be virtually impossible for any party, including EPA, to determine that either one has been done appropriately. It will certainly be burdensome for EPA to attempt to review an Implementation Plan submitted a year or so after a TMDL.

The point of tying TMDLs and Implementation Plans together is to ensure that the analysis of the TMDL is translated into the changes that are necessary to control sources. Analysis by itself does not lead to appropriate control actions. That is what we get if we have TMDLs and no Implementation Plans. Control actions proposed without analysis is what we have already; politically-wrangled determinations of how much some land owner/user is willing to do regardless of whether it is sufficient. That is what we will get if we have Implementation Plans that are not intimately connected with TMDLs. Neither one of these options is desirable if the program is to meet the goals of the Clean Water Act and be worth the significant taxpayer and private resources that will need to be invested. EPA's rules correctly focus on two things: attainment (TMDL) and usefulness (Implementation Plans). One without the other will not be worth the pain.

It is also true that Implementation Plans, as defined by EPA, include important aspects of TMDLs. For example, we would prefer to see the timeframe for attainment be a part of a TMDL analysis, rather than in the Implementation Plan. The same is true with regard to TMDL revision and monitoring. These shortcomings in the definition of a TMDL are relieved where the TMDL is defined to include the Implementation Plan.

We are not opposed to Implementation Plans that are submitted with more than one TMDL, however we are concerned that this is a potential area for unraveling what EPA has created for the reasons discussed immediately above. At issue is whether the connections will be close enough and the geographic scope of the Implementation Plans specific enough to make them of value.

Implementation Plan: Implementation Actions

EPA's description in the preamble of what must be included in implementation actions, the first component of an Implementation Plan is more specific and useful than that provided in the proposed regulation. 64 Fed. Reg. 46033. We urge EPA to adopt the specifics it discusses into the regulation language, particularly its expectations:

EPA expects that the implementation plan would also describe what actions
will be implemented by source category, subcategory, or individual sources.
The description of the actions should include an analysis of the anticipated or
past effectiveness of the best management practices and/or controls that are
expected to meet the wasteload and load allocations. The implementation plan
should describe where the best management practices and/or controls will be
implemented. This description should tie the implementation activity to the
pollutant and geographic scale of the TMDL.

Id. (emphasis added).

The preamble also states that the "implementation plan may deal with sources on a watershed basis as long as the scale of the implementation plan is consistent with the geographic scale for which the TMDL allocations are being established." Id. The problem with this statement is that the preamble and the proposed rules are not clear on the limitations of states' using geographic scales for TMDLs and allocations that are too large. Specifically, if a state develops a TMDL for very large areas, and fails to provide adequate specificity, the Implementation Plan will suffer from the same problems, yet EPA indicates that would be acceptable. Instead, to ensure that TMDLs do not become yet one more paperwork exercise in a long line of reports and plans that have accomplished relatively little over the years, EPA must address the issue of geographic scope.

The language of the preamble also undermines the program by stating that "EPA expects that the implementation plan would also describe what actions will be implemented by source category, subcategory or individual sources. The description of the actions should include* * * ." 64 Fed. Reg. 46033 (emphasis added). These expectations must be requirements.

Implementation Plan: Timeline

Omitted is a requirement that states include schedules for whatever administrative processes are required to revise nonpoint source