Federal Register Vol. 81, No.3,

Background Acronyms or Abbreviations

In the discussion of this proposed rule, we use the following acronyms or other abbreviations to stand in for certain words or phrases:

Under Section 7(h)(9) of the Act, 7 U.S.C. 2016(h)(9), States have the option to require that a SNAP EBT card contain a photo of one or more household members. The statute also stipulates that if a State agency chooses to place photographs on EBT cards, the State must establish procedures to ensure that any appropriate member of the household or any authorized representative of the household may utilize the card.

Pursuant to this statutory provision, existing regulations in 7 CFR 274.8(b)(5)(iv) provide that should a State agency require a photo on EBT cards, it must also establish procedures to ensure this same participant access is maintained. However, recent State implementation of the photo EBT card option revealed significant legal and operational complexities and challenges associated with having a photo on the card that, in the Department's view, calls for more regulatory guidance in this area. As a result, the Department is proposing to amend regulations in several areas to more explicitly define participant protections that must be maintained as well as implementation requirements if the State agency elects to implement a photo EBT card. In particular, this proposed regulation would clarify that the State option to place a photograph on an EBT card is a function of issuance. Pursuant to this, State agencies would be prohibited from having photo EBT card requirements affect the household eligibility or the certification process. Moreover, this rule would clarify the right of all household members and their authorized representatives to use the EBT card, regardless of whether their photo is on the card, and further define the responsibility of State agencies to ensure that retailers understand photo EBT requirements when processing transactions involving SNAP.

While the photo EBT card option was provided to States through the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, few States have implemented the option to date. However, recent efforts have shown that implementation of the photo EBT card option involves complex legal, operational, and civil rights issues that, if not well planned and implemented, can adversely affect access for program participants. Recent photo EBT card rollouts have had major implementation issues, raising concerns about program access, and leading to confusion in the retailer community.

These issues in turn prompted the Department to issue guidance on the subject on December 29, 2014. This proposed rule would expand on and codify this guidance to help ensure clients' access to benefits is maintained in practice and that sufficient measures are taken to ensure access before the photo EBT option is rolled out.

There has been some question as to whether adding the photo to an EBT card adds value to the integrity of SNAP. Statute requires that any household member or any authorized representative must be able to use the EBT card, regardless of whether their photo is on the card. For this reason, and because trafficking often involves a willing retailer, placing a photo on the EBT card will have limited impact in addressing fraud. In fact, some State agencies have investigated the possible benefits of adding the photo and, upon further analysis, decided it was not worth the cost to do so. Nevertheless, the Act provides States with the option to implement the photo on the EBT card and these proposed regulations are intended to fill the existing policy gaps in this area for the reasons cited above.

This proposed rule provides that State agencies would be required to meet certain requirements in order to implement a photo EBT card policy. The implementation requirements firmly establish SNAP policy that the photo EBT card option is a function of issuance and not a condition of eligibility; certification policy may not be impacted by the implementation of the option; all appropriate household members and authorized representatives, as defined in 7 CFR 273.2(n)(3), shall continue to be able to use the EBT card; program access is not inhibited at retail stores; and program access and program integrity are ensured through all stages of the process. A State considering a photo EBT card policy will need to lay out how it will operationalize the policy and develop and implement an Implementation Plan and photo EBT option that upholds all SNAP requirements.

To establish the requirements for the photo EBT card provisions, a new section would be codified in 7 CFR 274.8(f), addressing all the requirements associated with implementing a photo EBT card policy. Changes are also proposed to paragraphs within 7 CFR part 273 to further clarify that photo EBT card processes do not impact the certification of eligible households.

Implementation of the photo EBT card option takes substantial resources and requires substantial changes to State systems and procedures. Due to these challenges, States that have recently implemented photo EBT card implementations have had significant issues with providing timely, accurate, and fair service to SNAP applicants and participants and with meeting other statutory and regulatory requirements. For this reason, the Department would require that States demonstrate to FNS successful administration of SNAP based on SNAP performance standards to be eligible to implement the photo EBT card option. Successful program administration would be based on an evaluation of metrics related to program access, the State's payment error rate, the State's Case and Procedural Error Rate, application processing timeliness, including 7-day expedited service and the 30-day processing standards, timeliness of recertification actions, and other metrics, as determined by the Secretary, that may be relevant to the implementation of a State's photo EBT card option. States would need to document in the Implementation Plan that they are meeting FNS performance expectations. The Department is interested in comments from the public about other metrics that FNS should consider in the context of determining successful program administration including metrics related to access to benefits. These performance standards will allow FNS to evaluate whether clients are receiving timely, accurate, and fair service before the State may be eligible to implement a photo EBT card option. This provision would be codified in 7 CFR 274.8(f)(1).

The proposed rule clarifies that the photo EBT card option is a function of issuance and not a condition of certification. Any implementation of the photo EBT card option must not impact certification of households. State agencies shall not deny or terminate a household based solely on whether one or more household members comply with the requirement to have a photo on the EBT card.

The State agency's photo EBT card policy must not affect the certification process for purposes of determining eligibility regardless of whether an individual has his/her photo placed on the EBT card. For example, an application would be considered complete and be subject to 7 and 30-day processing timelines regardless of the status of the photo. Application processing timeliness requirements would not be different for photo EBT cards. This provision would be codified in 7 CFR 274.8(f)(2).

The proposed rule would allow for State agencies to implement the SNAP photo EBT card on a mandatory or voluntary basis. Regardless of whether the photo is mandatory or voluntary, clients must be informed that their household's certification will not be impacted by whether they agree to the photo. If the policy is mandatory, State agencies must establish which member(s) of the household would be required to be photographed and must establish procedures that allow eligible nonexempt household members who do not agree to the photo to come into compliance at a later time. The photo may be of the head of household and/or other participating nonexempt household member(s). State agencies must issue benefits to compliant or exempt household members and, as noted earlier, non-compliance with a photo requirement by household members who choose not to be photographed must not negatively impact the household's eligibility determination.

If the policy is voluntary, clients would be able to elect to have their photo on the household EBT card but would not be required to do so and would not have to be in an exempted category to opt out of the photo option. State agencies implementing a voluntary photo EBT card policy would be required to make clients aware of the voluntary nature of the photo and the fact that benefit issuance would not be impacted by their decision to have or not have a photo on their card. In voluntary implementations, households would opt in to have a photo on their cards rather than opt out of the option. Therefore, EBT cards with photos would not be the default.

FNS has concerns that States implementing voluntary photo EBT card policies to date have been unsuccessful in communicating the nature of the program to clients and remains concerned about how such lack of information could affect SNAP households, especially those with non-applicant heads of household. The Department is seeking comment about suggestions for how to strengthen the requirements on States to provide clear and effective information that ensures clients understand the State's photo EBT policies. Additionally, to ensure States' implementation of the photo EBT option does not create disparate impacts on members of any protected class, as proposed, States would not be allowed to photograph non-applicants or put their photo on an EBT card, regardless of whether the State's program is voluntary or mandatory. As proposed, States could not offer non-applicant heads of households the option to opt in to have their photos taken to ensure that clients would not be pressured by States, intentionally or inadvertently, to have a photo taken. Nevertheless, FNS would like to better understand if there is any potential benefit of allowing non-applicants to have their photograph taken under a voluntary implementation and whether such benefits outweigh potential problems.

These provisions would be codified in 7 CFR 274.8(f)(3) and 7 CFR 27.8(f)(4).

Because recent implementation by States showed inconsistency and confusion in the application of State-defined exemption criteria, the Department is making clear in this proposed rule who must be exempted from photo EBT card requirements. State implementation showed there were circumstances where even exempt clients as defined by those States, such as the elderly, disabled, or domestic violence victims, were not actually exempted from the photo EBT requirement. In addition, many clients were not clearly informed that they were exempt under the State's exemption criteria and were under the impression that they had to comply with the State's photo EBT card policy in order to receive program benefits.

To ensure that a State's photo EBT card requirement will not place undue burden on vulnerable clients, the proposed rule requires States implementing a mandatory implementation to exempt, at a minimum, the elderly, the disabled, children under 18, homeless households, and victims of domestic violence. FNS proposes that victims of domestic violence would be able to self-attest and States would not be permitted to require clients to submit documentation to verify that they have been victims of domestic violence. The ability to self-attest must be applied equally regardless of whether the victim is a female or male. A State agency may establish additional exempted categories, including, but not limited to, categories described in hardship criteria as specified in 7 CFR 273.2(e)(2).

As noted in the previous section, as proposed, non-applicants shall not be required nor permitted to have their photographs taken or put on EBT cards. This provision would be codified in 7 CFR 274.8(f)(4).

The proposed rule requires that State agencies have sufficient capacity and a process to issue photo EBT cards, taking into account households that face hardship situations as determined by the State agency and that would receive non photo EBT cards. These hardship conditions include, but are not limited to: Illness, transportation difficulties, care of a household member, hardships due to residency in a rural area, prolonged severe weather, or work or training hours which prevent the household from having photos taken in-office.

There are a variety of SNAP policy and operational questions that States must consider in developing their issuance process for photo EBT cards, including the technical aspects relating to software, hardware, and the taking, storage, and security of photos.

The proposed rule would require that States demonstrate sufficient capacity to issue photo EBT cards before they may receive an authorization from FNS to implement the option. As noted in the previous section on hardship, any State implementing a photo EBT card option would have to establish a process to issue cards to households that may not be able to reach a local office due to a hardship condition. Furthermore, the process for issuing and activating photo EBT cards must not inhibit or delay access to benefits nor cause a gap in access to benefits for any eligible households. Any card issued as part of the implementation of the photo EBT card option may not count against the household as part of the maximum threshold of replacement cards as specified in 7 CFR 274.6(b)(5) and 7 CFR 274.6(b)(6). Most importantly, as noted earlier, non-compliance with a photo requirement cannot impact the eligibility determination of the household as the photo EBT card option is an issuance function, not a condition of eligibility.

States that have recently implemented a photo EBT policy have struggled with operational challenges during the transition from regular EBT cards to photo EBT cards. In one State, some clients lost access to their benefits during the period between the deactivation of their previous EBT card and activation of the new EBT photo card. This proposed rule would require that States implementing the photo EBT card option establish a process to ensure that the replacement of cards does not disrupt households' access to benefits, consistent with the requirements of 7 U.S.C. 2016(h)(9). Additionally, State card issuance procedures developed for new SNAP households would need to ensure adherence to the application processing standards of 7 days in the case of expedited households and 30 days for all other households, as required by 7 CFR 273.2(g) and 7 CFR 273.2(i).

As proposed, if a household meets expedited criteria in 7 CFR 273.2(i), the State must issue benefits and issue the EBT card to the entire household without delay. Regardless of whether the State's photo EBT policy is voluntary or mandatory, the State could not delay, hold in abeyance, or prorate benefits for any household that meets expedited criteria in order to obtain a photo on the EBT card. Under a mandatory implementation, a non-exempt household member could be required to comply at the next recertification after expedited benefits have been issued to the household. If the non-exempt household member is not in compliance by the time the household is recertified, then the State could determine whether that member's share of benefits must be held in abeyance prospectively. Under the proposed rule, State agencies implementing a photo EBT card option must also meet the card replacement issuance card requirements stipulated in 7 CFR 274.6, which, among other things require States to issue replacement EBT cards within 2 business days following notice by the household that the card has been lost, stolen, or damaged.

This provision would be codified in 7 CFR 274.8(f)(6).

State agencies would not be able to deny benefits to an entire household because a nonexempt household member(s), required by the State to be photographed, refuses to be photographed. Unless the household meets expedited criteria, this proposed rule would require the State to issue a prorated share of benefits to the remaining household members, so they can use their share of the benefits that they are entitled to receive.

As proposed, for multi-person households, that would be a straight pro-ration of benefits. The State would divide the household's benefit allotment by the household size and multiply that number by the number of household members to be issued benefits. To illustrate, if a four-person household's monthly benefit allotment is $200 and one nonexempt household member does not comply with the requirement to have a photo placed on the EBT card, the $200 would be divided by 4 to equal $50, and then multiplied by 3 to equal $150. The $150 amount would be posted and available for use by the household. Any decision that impacts benefits under this provision would be subject to fair hearings in accordance with 7 CFR 273.15. For a single person household, the State agency would hold the entire benefit in abeyance until the household complies. This proposed provision is addressed in 7 CFR 274.8(f)(7) of the regulation.

FNS proposes that the pro-rated benefit amounts held for noncompliance with a State's photo EBT card requirement would be tracked and retained for future issuance by the State agency if and when any noncompliant household member(s) that previously chose not to be photographed comes into compliance. The pro-rated benefits withheld for that individual or individuals must promptly be issued within two business days of the time the individual(s) comes into compliance. Benefits withheld for non-compliance would not remain authorized for perpetuity and States must treat such benefits in accordance with the same timeframe used for handling expungements under 7 CFR 274.2(h)(2). This would allow States to better manage benefits that have not been issued.

This proposed provision is addressed in 7 CFR 274.8(f)(8) of the regulation.

The Food and Nutrition Act requires States implementing photo EBT to “establish procedures to ensure that any other appropriate member of the household or any authorized representative of the household may utilize the card.” All household members, authorized representatives as defined in 7 CFR 273.2(n)(3), and non-applicants applying on behalf of others have a right to access SNAP benefits by using the household EBT card with a valid PIN even if their picture is not on the card or there is no picture on the card. The State agency must take steps to ensure that individuals who are not pictured on the card can continue to access SNAP benefits in accordance with the Section 7(h)(9)(B) of the Act.

The ability for authorized representatives to use the card is particularly critical to ensure food access for the elderly, disabled, or other homebound recipients who may have difficulty getting to grocery stores and require assistance in obtaining food.

This provision would be codified in 7 CFR 274.8(f)(9).

This proposed rule would require that States ensure all staff and clients are trained on photo EBT card requirements. At a minimum, this training would include information about whether the photo EBT card is voluntary or mandatory, that all appropriate household members and authorized representatives are able to use the card, and with regards to mandatory implementation, which household members (if any) must comply with the photo requirement, which household members and/or household applicant categories are exempt. This proposed rule would also require that all retailer and client notices pertaining to the photo EBT card must also clearly describe the following statutory and regulatory requirement: All household members and any authorized representative of the household regardless of whether they are pictured on the card, may utilize the card without having to submit additional verification of identity as long as the transaction is secured by the use of the PIN. This proposed rule would also stipulate that State agencies may not specifically reference groups exempt from the photo requirement in any materials designed for retailers, as providing the categories of exempt groups may encourage speculation as to the age or circumstances of cardholders. External stakeholder materials should simply note that EBT cards without a photo are also valid. This provision would be codified in 7 CFR 274.8(f)(10).

Although retailer participation in SNAP is authorized and managed by USDA, this rule, as proposed, recognizes State agencies opting to implement a photo EBT card would impact SNAP transactions at the point of sale. Per the statutory requirement, State agencies are required to ensure access to benefits by household members other than the member(s) whose photo is on the card, as well as authorized representatives. Therefore, in paragraph 7 CFR 274.8(f)(10), this proposed rule would require State agencies to provide information to all retailers about the State agencies' implementation and operational plans so retailers are prepared for the changes, as well as to convey the Federal rules stipulating that all household members and authorized representatives must be allowed to use the EBT card regardless of the picture on the card. Furthermore, Federal rules prohibit retailers from treating SNAP participants differently from other customers at the point of sale. According to the equal treatment regulation in 7 CFR 278.2(b), SNAP customers cannot be singled out for special treatment in any way.

A State agency would need to provide documentary evidence that all retailers in the State and contiguous areas, including smaller independent retailers, have received notices from the State that explain the statutory and regulatory requirements related to photos on EBT cards described above and have a full understanding of those requirements. State agencies would be required to describe in the Implementation Plan when they will provide FNS with this documentary evidence in advance of implementation. This provision would be codified in 7 CFR 274.8(f)(11).

Section 7(j) of the Act requires that EBT cards be interoperable, which means that they can be used in any State regardless of where the benefits were issued. Without sufficient education for clients and retailers in both the implementing State and neighboring States, the implementation of the photo EBT card option could inhibit this required access to benefits. For example, a SNAP recipient who lives in a State with a photo EBT card may find it more convenient or cost effective to shop in another bordering State that does not have photo EBT cards. Likewise, a SNAP recipient may live in a State without a photo EBT card requirement, but shop in a State with such a card requirement. To ensure interoperability, clients, and retailers must be fully informed that the photo EBT cards remain interoperable and that authorized retailers must accept EBT cards from all States as long as the household member or authorized representative presents the valid PIN. Before introducing the new photo EBT cards, this proposed rule would require State agencies to conduct sufficient outreach to clients and retailers, including those in contiguous areas, to ensure access to benefits is not inhibited and all parties understand their rights and responsibilities.

This provision would be codified in 7 CFR 274.8(f)(12).

As the Act allows for photo EBT cards, appropriate implementation and administration of this option is an allowable State administrative cost which FNS would reimburse at approximately 50 percent. To ensure that FNS does not exceed the SNAP budget authority for State administration, States should be aware that any EBT contract modifications that increase costs must be approved by FNS before they may be signed. Increased EBT costs, whether contractual or resulting from other sources, also require an Implementation Advance Planning Document Update. This provision would be codified in 7 CFR 274.8(f)(13).

In 7 CFR 274.8(f)(14), this proposed rule would require State agencies to submit an Implementation Plan prior to photo EBT implementation that delineates how the State will operationalize this option. Upon receipt of the State's Implementation Plan, FNS would review the plan and either issue an approval, request modifications before an approval could be granted, or issue an approval subject to conditions. In cases where the Department finds that the steps outlined in the Implementation Plan are not sufficient for a successful implementation, the Department might issue an approval subject to conditions, such as requiring the State agency to demonstrate a rollout in a pilot in a selected region of the State before the State may be approved to implement statewide or, FNS might approve the Implementation Plan for a statewide implementation upon the completion of an appropriate successful pilot project that establishes the State agency's ability to implement a full statewide rollout. Should a State be required to implement a pilot before statewide implementation, that requirement would be documented as a condition of the State's Implementation Plan approval, along with any information that the State must report to FNS before the State may be granted approval to implement statewide.

FNS expects that the process for FNS review and approval of photo EBT Implementation Plans will take at least 120 days. Obtaining FNS approval of the Implementation Plan is the first step States must take. However, a State may not actually issue EBT cards with photos until FNS has given the State an authorization to do so as described below. The multi-step approval process ensures that the State carries out the steps detailed in the Implementation Plan and has the opportunity to make any adjustments needed prior to issuing EBT cards with photos. Similarly if FNS has approved an Implementation Plan subject to conditions, such as requiring the State agency to conduct a pilot prior to statewide implementation, the State may not issue EBT cards with photos in the context of the pilot until FNS has given the State an authorization to do so. Approval from FNS would also be necessary for a State to proceed from pilot to statewide implementation. Any movement to implement without prior approval would be viewed as a violation of program requirements and could result in additional penalties including a loss of Federal financial participation.

FNS would not consider a State eligible or authorize a State to proceed with a photo EBT card option unless that State meets performance requirements noted earlier and established in 7 CFR 274.8(f)(1). States would need to demonstrate in the Implementation Plan that they are meeting FNS performance expectations. The Implementation Plan would also be required to include a description of the State's card issuance procedures, a detailed description of how client protections such as processing timelines and benefit access will be preserved, specific information about exempted recipients, a description of how the State will obtain photographs for the EBT card, training materials and training plans for State staff, examples of letters and other materials communicating the policy to recipients and to retailers, a proposed timeline for implementation, and any other information as required by the Secretary. If a State agency plans to disclose SNAP applicant or client data in accordance with 7 CFR 272.1(c) for purposes of implementing photo EBT cards, such as to obtain photos from another source like the State's Department of Motor Vehicles the proposed rule requires the State to also include any necessary memoranda of understanding as part of its Implementation Plan. Any information collected for the purpose of SNAP must be securely stored and can only be shared in accordance with 7 CFR 272.1(c).

State Implementation Plans would also be required to describe: The specific action steps that the State agency and its EBT contractor must take in order to implement the photo EBT card option as planned, together with the anticipated timetable for each step; the State's capacity to issue photo EBT cards; and the submission of the documentation that all retailers, including small and independent retailers, would receive notice from the State about the photo EBT card policy. The plan would also need to describe how the State will ensure that the photo EBT cards are provided to clients and activated at the same time or before the active non-photo cards are deactivated. With regard to the State's capacity to issue photo EBT cards, the plan would include the description of the capacity at the facility where photo EBT cards will be produced, both for transitional and ongoing production, and assurance that the State and its EBT contractor will continue to meet regulatory timeliness requirements for all EBT card issuances. The Implementation Plan should also describe measures against which the photo EBT card implementation will be evaluated for the post-implementation evaluation required by 7 CFR 274.8(f)(16), and how the requisite data will be collected.

The State would also be required to include in its plan for FNS review all applicable written policy changes necessary to implement the photo EBT card option, as well as copies of all materials that will be used to inform clients, retailers and other stakeholders regarding photo EBT card implementation. Along with these materials, the States would need to provide a detailed description of how the notifications, communication, policies, and procedures regarding the implementation of the photo EBT card option will comply with all applicable civil rights laws.

Finally, the State would need to provide a description of the mechanisms in place to handle complaint calls and questions from clients, retailers, and external stakeholders and address any other issues related to the photo EBT card option, as well as detail how substantive information about those complaints will be tracked and reported. A State would not be authorized to issue EBT cards with photos until FNS grants the State an authorization to implement as specified by 7 CFR 274.8(f)(15).

Upon approval of the Implementation Plan, the State would be allowed to proceed with tasks described in the Implementation Plan, as modified by the approval, but not proceed to issuing actual cards until FNS provides authorization to implement.

An approved Implementation Plan would be considered public and would be posted on the FNS Web site. The Department is interested in receiving comments on any benefits and concerns of posting the approved Implementation Plan.

The authorization to implement would allow the State agency to begin issuing EBT cards with photos. After the Implementation Plan is approved, FNS will review State actions at an appropriate time interval to ensure that the process and steps outlined by the State agency in the Implementation Plan have in fact been carried out in a satisfactory manner. For example, prior to obtaining authorization to implement, a State would need to confirm and/or demonstrate that robust client and retailer outreach, as detailed in its Implementation Plan, has been completed.

If FNS finds that the State agency has not acted in accordance with the steps outlined in the State's photo EBT Implementation Plan, FNS could deny authorization for the State to issue EBT cards with photos until the State has done so in a satisfactory manner. FNS could also require the State to implement in a phased manner, which may include criteria as determined by the Secretary. This provision would be codified in 7 CFR 274.8(f)(15).

As already noted, 7 CFR 274.8(f)(16) would require States to submit to FNS a post-implementation evaluation conducted by from an independent evaluator, which describes the State's implementation to date, including any issues that arose and how they were addressed, the degree to which State staff, clients and retailers properly understood and implemented the relevant policies and procedures, and in the case of a mandatory implementation, the number of clients that complied with adding the photo or did not comply, and the number that had their share of the benefits withheld from issuance and for how long. The evaluation must include, at a minimum, a survey of retailers and clients to measure their understanding of the State's photo EBT policy, and a report which includes the number of households and percent of households with photo EBT cards in the State and the number and scope of complaints related to photo EBT implementation, including a detailed summary of the types of complaints, the SNAP performance metrics as established in section 7 CFR 274.8(f)(1) and other information as determined by the Secretary.

For States implementing a mandatory implementation, the report must also detail the amounts and percent of benefits withheld for non-compliance, the number of households affected by the withholding of benefits due to noncompliance, the number and percent of persons exempt from the photo EBT card requirement, and the number and percent of exempted households and persons who opted to have the photo on the EBT card.

State agencies would be required to deliver this report to FNS within 120 days of implementation. This report would cover the first 90 days of implementation. The Department reserves the right to conduct its own review of the State's implementation.

Based on observed implementation to date, there is cause for concern about possible impacts of photo EBT implementations, both as they are first implemented and over time. There is a need for additional assurance on an ongoing basis that state implementation of photo EBT cards is carried out in a manner consistent with all relevant laws and regulations, including Federal civil rights laws, that protect households' ability to access or utilize SNAP benefits for which they are eligible, and in a manner that does not adversely impact program participation.

As set forth in the proposed rule, in addition to the post-implementation report, a State agency that has implemented a photo EBT policy would be required to provide to FNS, on an ongoing basis, data on established metrics to monitor the impact of the photo EBT policy. The reporting requirements might require State agencies to conduct additional surveys, evaluations, or reviews of their operations, as determined by the Secretary. These ongoing reporting requirements would include information on the amounts and percent of benefits withheld for non-compliance, the number of households affected by the withholding of benefits due to non-compliance, the number and percent of household exempt from the photo EBT card requirement, benefit redemption rates, participation rates, the number and percent of households exempt from photo EBT cards who opted out of the photo requirement, the number and percent of exempted households who opted to have the photo on the EBT card, and any other information as requested by the Secretary. We are interested in receiving comments on other data that should be required from States on an ongoing basis, how frequently States should be required to report, or any other feedback relevant to the ongoing monitoring of this policy. As with other Program information and plans, this information would be available to the public upon request, subject to the Freedom of Information Act provisions.

While staff, client, and retailer education is a critical component, it is not always a perfect indicator of whether actual barriers to access exist in practice. In the context of housing discrimination, “testers” have been utilized to proactively determine whether fair housing laws are being upheld consistently. One question is whether a similar mechanism should be used to ensure that, in practice, SNAP participants and their authorized representatives are able to use their benefits to purchase food at authorized retailers, regardless of whether they are pictured on the EBT card. We invite comment on this question as well as on the topic of how to verify appropriate implementation on an ongoing basis, particularly on ongoing mechanisms for identifying access issues resulting from photo EBT cards.

As part of FNS's management and oversight responsibilities, FNS regularly conducts management evaluation reviews of State agencies' administration and operation of SNAP to determine compliance with program requirements. FNS will conduct management evaluation reviews, as appropriate, to monitor State implementation of photo EBT cards.

If FNS identifies deficiencies in a State's implementation or operations, FNS may require a corrective action plan consistent with 7 CFR 275.16 to reduce or eliminate deficiencies. If a State does not take appropriate actions to address the deficiencies, FNS would consider possible actions such as requiring an updated photo EBT Implementation Plan, suspension of implementation and/or withholding funds in accordance with 7 CFR 276.4. Along these lines the Department is seeking comments on whether a State should be required to stop or suspend issuing photos on EBT cards if the State agency fails to establish procedures to ensure that all members of the household or any authorized representatives of the household are able to utilize the card, and what requirements, if any, should apply to that process.

Beyond Part 7 CFR 274.8(f), changes are proposed to 7 CFR parts 271, 272, 273, and other paragraphs within Parts 274 and 278. While some of these changes are related to photo EBT card requirements, others involve updating SNAP regulations or enhancing integrity provisions.

The Department proposes to amend the definition of Identification (ID) card in 7 CFR 271.2. ID card in this definition refers to a card that was issued when program benefits were issued in the form of food stamp coupons. This ID card, which was used to establish the recipient as eligible to receive food stamp coupons, such as when picking up coupons at the State office or other central location is no longer widely used in the program. Today, program benefits are automatically deposited into the household's EBT account each month and are redeemed through EBT cards. The PIN on the EBT card establishes whether a household recipient or authorized representative can redeem program benefits. However, this ID is still used in Alaska to identify households that are dependent upon hunting and fishing for subsistence. The definition for Identification (ID) would be amended to reflect only cases in which ID cards are currently used in the Program.

The Department proposes changes to 7 CFR part 272 to ensure that regulatory language is in line with current program operations. In alignment with the change to 7 CFR 271.2, FNS proposes removing all references to “ID card” associated with the obsolete paper coupons. It would result in the removal of 7 CFR 272.1(g)(30) and 7 CFR 272.1(g)(47).

FNS proposes several changes to 7 CFR part 273. The Department proposes adding language in 7 CFR 273.2(a)(1) to clarify that the implementation of the photo EBT card option cannot be treated as a condition of eligibility as it is a function of issuance. Further, this paragraph would be amended to ensure that, for the purpose of certification, States shall not treat households subject to a photo EBT card policy differently from households not subject to a photo EBT card policy. To ensure that expedited and standard application processing requirements are still met in photo EBT card situations, 7 CFR 273.2(a)(2) would be revised by adding that State agencies shall ensure that processing times are not delayed by implementation of the photo EBT card option. Third, the Department proposes to clarify the rules governing interviews in 7 CFR 273.2(e). State agencies may not require an in person interview solely for the purpose of taking a photo. Since this option is a function of issuance and not a condition of eligibility, households must be treated equally with regards to certification activities regardless of whether they are subject to or choose to comply with a photo requirement. However, households may be called in for a photo to be taken as a matter of issuance, not eligibility. In 7 CFR 273.2(n)(2), the reference to the ID card would be removed as it is obsolete. In 7 CFR 273.2(n)(3), the proposed language would change the word ID to EBT card. The proposal would change “its ID card and benefits” to “the EBT card.”

The Department proposes changes to 7 CFR 274.8(b)(5)(ii) to modify EBT cards in States implementing the photo EBT card option, in accordance with 7 CFR 274.8(f). States would be required to add text to all EBT cards to ensure retailers are aware that all household members and authorized representatives must be allowed to use the EBT card even if their photo is not on the card or no photo is on the card. Experience has shown that, when a photo is included on the EBT card, some retailers believe the card may only be used by the person pictured. In concert with other required measures to ensure that retailers understand the State's photo EBT implementation, adding a statement on photo EBT cards would help alleviate confusion at retailer checkout and ensure compliance with the Federal statute that requires all household members and authorized members be able to access program benefits. This rule would propose that the States print the text: “Any user with valid PIN can use SNAP benefits on card and need not be pictured.” or alternative text approved by FNS. The Department is willing to consider alternative language suggested by States as long as it achieves the same goal of clearly informing retailers and clients as to the correct policy in this area.

The Department is proposing changes in Part 278 to remove language that is no longer in line with program operations and update language to enhance program integrity. The Department has recently become aware of instances in which SNAP authorized retailers, unauthorized retailers, and other individuals have purchased multiple EBT cards illegally. Generally, these individuals are not SNAP recipients. Frequently they use three or more EBT cards at a time and use the cards to purchase a large amount of eligible foods that are then used to replenish store inventory or sold as inventory to other retailers or restaurants. To address this area of potential fraud in which individuals use multiple cards they have procured illegally, the Department is proposing new language to require retailers to ask for identification of anyone who presents three or more EBT cards at checkout. Specifically, this proposed rule would require SNAP authorized retailers to ask these individuals for photo identification, such as a driver's license, and an explanation as to why multiple cards are being used. Furthermore, should the store believe there is a potential for fraud, retailers would be allowed to record information from the individual's identification, EBT card number, and reason for using three or more EBT cards. The retailers would be required to report this information to the USDA OIG Fraud Hotline. If the retailer suspects fraud is being committed and the individual refuses to show identification, the retailer has the option to deny a sale when three or more EBT cards are being used during a transaction. The Department understands that occasionally an individual or an individual shopping for an elderly client working for an authorized group home or other authorized facility may use multiple cards in order to purchase food legally for clients. Given these concerns regarding program access and program integrity, the Department is interested in comments from the public on whether there are other possible approaches to preventing individuals from using multiple EBT cards that they have obtained illegally, such as establishing a dollar threshold for individuals using three or more cards. These changes are proposed in 7 CFR 278.2(h).

The Department also proposes the removal of two paragraphs, (i) and (k) in 7 CFR 278.2. The paragraphs refer to an outdated method of establishing identity and operations based on paper coupons. These paragraphs currently represent a redundancy and could cause confusion as they refer to an ID card that is only used in Alaska. This process has been replaced by the EBT system. Furthermore, the proof of eligibility is established through EBT and other systems implemented by State agencies.

7 CFR 274.7(i) already addresses this procedure by establishing that State agencies shall implement a method to ensure that access to prepared meals and hunting fishing equipment is limited to eligible households. Eligible households are defined in 7 CFR 274.7(g) and (h).

Executive Orders 12866 and 13563 direct agencies to assess all costs and benefits of available regulatory alternatives and, if regulation is necessary, to select regulatory approaches that maximize net benefits (including potential economic, environmental, public health and safety effects, distributive impacts, and equity). Executive Order 13563 emphasizes the importance of quantifying both costs and benefits, of reducing costs, of harmonizing rules, and of promoting flexibility.

This proposed rule has been determined to be significant and was reviewed by the Office of Management and Budget (OMB) in conformance with Executive Order 12866.

As required for all rules that have been designated as Significant by the Office of Management and Budget, a Regulatory Impact Analysis (RIA) was developed for this proposed rule. The full RIA is included in the supporting documents of the rule docket at www.regulations.gov. The following summarizes the conclusions of the regulatory impact analysis.

Need for Action: This proposed rule would incorporate into regulation and expand on guidance that was issued December 29, 2014 to certain State agencies. Based on observed implementation to date, there is cause for concern about possible impacts of photo EBT programs, both as they are first implemented and over time. This guidance requires States that intend to implement the photo EBT card option to submit a comprehensive Implementation Plan for FNS approval that addresses key operational issues to ensure State implementation complies with all Federal requirements and that program access is protected for participating households.

In this proposed rule, the Department would clarify that the State option to place a photo on an EBT card is a function of issuance. Pursuant to this, State agencies would be prohibited from having photo EBT requirements affect the eligibility process. This includes ensuring that the photo EBT option is implemented in a manner that does not impose additional conditions of eligibility or adversely impact the ability of eligible Americans to access the nutrition assistance they need.

Benefits: The Department anticipates that this proposed rule will provide qualitative benefits to State Agencies, SNAP participants, and authorized retailers. The Act and existing program regulations provide that States that implement a photo on the EBT card must establish procedures to ensure that any other appropriate member of the household or any authorized representative of the household may use the card. This proposed rule will provide clear parameters for States wishing to implement photo EBT to ensure that State implementation is consistent with all Federal requirements and that program access is protected for participating households, which will safeguard the rights of clients, provide training to staff, clients, and retailers, and improve program administration.

Costs: States choosing the photo EBT option may incur additional administrative costs, which may vary based on the size and scope of the State's operations and whether implementation of the photo EBT card option is mandatory or voluntary. Regardless of whether the option is mandatory or voluntary, all States that implement photo EBT cards will incur certain implementation costs to include: Preparing an implementation plan, communications and training for program staff, clients, and retailers, ongoing training costs to maintain an understanding of Photo EBT policies, programming costs for mandatory policies, and costs for the post-implementation assessment, evaluation and on-going monitoring. States with mandatory photo EBT will also incur costs associated with prorating and storing benefits for noncompliant household members that choose not to be photographed. The Department estimates the total cost to be approximately $9.8 million over five years, assuming six States choose to implement a mandatory Photo EBT policy. Costs would be lower if some or all of these States choose to implement voluntary, rather than mandatory, Photo EBT policies. The estimate of six States is based on information from State legislatures that are either currently considering or discussing the possibility of considering such a policy. Given the projected timelines for these legislative actions, the Department assumes that the costs of implementing a Photo EBT system will be phased in over a five year period, as all six States are unlikely to approve and implement the policy in the same year. The two States that have already implemented photo EBT as a State option will not be required to retroactively submit Implementation Plans, but may continue to incur minimal costs associated with ongoing training and monitoring required for program staff, clients, and retailers.

The Regulatory Flexibility Act (5 U.S.C. 601-612) requires agencies to analyze the impact of rulemaking on small entities and consider alternatives that would minimize any significant impacts on a substantial number of small entities. Pursuant to that review, it has been certified that this proposed rule would not have a significant impact on a substantial number of small entities. This proposed rule primarily impacts State agencies. As part of the requirements, State agencies would have to educate retailers about the photo EBT card. There will not be a substantial impact on small entities such as small retailers since the treatment of clients with EBT cards and photo EBT cards do not vary. Minimal changes will be required of retailers. Retailers will need to be aware that some clients may present photo EBT cards but clients shall not be treated any differently. In addition, retailers will be required to request identification of individuals using three or more EBT cards. This is not expected to create a burden on retailers.

Title II of the Unfunded Mandates Reform Act of 1995 (UMRA), Public Law 104-4, establishes requirements for Federal agencies to assess the effects of their regulatory actions on State, local and Tribal governments and the private sector. Under section 202 of the UMRA, the Department generally must prepare a written statement, including a cost benefit analysis, for proposed and final rules with “Federal mandates” that may result in expenditures by State, local or Tribal governments, in the aggregate, or the private sector, of $100 million or more in any one year. When such a statement is needed for a rule, Section 205 of the UMRA generally requires the Department to identify and consider a reasonable number of regulatory alternatives and adopt the most cost effective or least burdensome alternative that achieves the objectives of the rule.

This proposed rule does not contain Federal mandates (under the regulatory provisions of Title II of the UMRA) for State, local and Tribal governments or the private sector of $100 million or more in any one year. Thus, the proposed rule is not subject to the requirements of sections 202 and 205 of the UMRA.

The Supplemental Nutrition Assistance Program is listed in the Catalog of Federal Domestic Assistance Programs under 10.551. For the reasons set forth in the final rule in 7 CFR part 3015, subpart V, and related Notice (48 FR 29115, June 24, 1983), this program is included in the scope of Executive Order 12372 which requires intergovernmental consultation with State and local officials.

Executive Order 13132 requires Federal agencies to consider the impact of their regulatory actions on State and local governments. Where such actions have federalism implications, agencies are directed to provide a statement for inclusion in the preamble to the regulations describing the agency's considerations in terms of the three categories called for under Section (6)(b)(2)(B) of Executive Order 13121.

The Department has determined that this proposed rule does not have Federalism implications. This rule does not impose substantial or direct compliance costs on State and local governments. Therefore, under Section 6(b) of the Executive Order, a Federalism summary impact statement is not required.

This proposed rule has been reviewed under Executive Order 12988, Civil Justice Reform. This proposed rule is intended to have preemptive effect with respect to any State or local laws, regulations or policies which conflict with its provisions or which would otherwise impede its full and timely implementation. This proposed rule is intended to have retroactive effect unless so specified in the Effective Dates section of the final rule. State agencies that have already implemented a photo EBT card must meet all requirements of regulations except the requirement to submit an Implementation Plan prior to State's planned implementation date. Prior to any judicial challenge to the provisions of the final rule, all applicable administrative procedures must be exhausted.

Executive Order 13175 requires Federal agencies to consult and coordinate with Tribes on a government-to-government basis on policies that have Tribal implications, including regulations, legislative comments or proposed legislation, and other policy statements or actions that have substantial direct effects on one or more Indian Tribes, on the relationship between the Federal Government and Indian Tribes, or on the distribution of power and responsibilities between the Federal Government and Indian Tribes. On February 18, 2015, the agency held a consultation. During the consultation, no comments were received on the proposal. We are unaware of any current Tribal laws that could be in conflict with the proposed rule.

FNS has reviewed this proposed rule in accordance with USDA Regulation 4300-4, “Civil Rights Impact Analysis,” to identify any major civil rights impacts the rule might have on program participants on the basis of religion, age, race, color, national origin, sex, political beliefs, or disability. After a careful review of the rule's intent and provisions and understanding the intent of this rule is to in part to protect the civil rights of recipients, FNS has determined that this rule is not expected to adversely affect the participation of protected individuals in the Supplemental Nutrition Assistance Program.

The Paperwork Reduction Act of 1995 (44 U.S.C. Chap. 35; 5 CFR 1320) requires the Office of Management and Budget (OMB) to approve all collections of information by a Federal agency before they can be implemented. Respondents are not required to respond to any collection of information unless it displays a current valid OMB control number.

In accordance with the Paperwork Reduction Act of 1995, this proposed rule does not contain information collections that are subject to review and approval by the Office of Management and Budget.

This rule proposes reporting requirements for States to submit to FNS an Implementation Plan, a post-implementation evaluation of the photo EBT implementation, and related on-going measures. As the PRA requirements are applicable to collection of information from ten or more respondents, there are no information collection requirements that are subject to OMB review at this time. Should the number of estimated respondents reach ten or more, FNS will publish a notice for comment and submit the applicable requirements to OMB for review and approval.

The Department is committed to complying with the E-Government Act, to promote the use of the Internet and other information technologies to provide increased opportunities for citizen access to Government information and services, and for other purposes.

Accordingly, 7 CFR parts 271, 273, 274, 278 are proposed to be amended as follows:

The additions read as follows:

The additions read as follows:

The revision reads as follows:

Please refer to Docket ID NRC-2015-0213 when contacting the NRC about the availability of information for this action. You may obtain publicly-available information related to this action by any of the following methods:

• Federal Rulemaking Web site: Go to http://www.regulations.gov and search for Docket ID NRC-2015-0057.

• NRC's Agencywide Documents Access and Management System (ADAMS): You may obtain publicly-available documents online in the ADAMS Public Documents collection at http://www.nrc.gov/reading-rm/adams.html. To begin the search, select “ADAMS Public Documents” and then select “Begin Web-based ADAMS Search.” For problems with ADAMS, please contact the NRC's Public Document Room (PDR) reference staff at 1-800-397-4209, 301-415-4737, or by email to [email protected]. The ADAMS accession number for each document referenced (if it is available in ADAMS) is provided the first time that it is mentioned in the SUPPLEMENTARY INFORMATION section.

• NRC's PDR: You may examine and purchase copies of public documents at the NRC's PDR, Room O1-F21, One White Flint North, 11555 Rockville Pike, Rockville, Maryland 20852.

Please include Docket ID NRC-2015-0213 in your comment submission.

The NRC cautions you not to include identifying or contact information that you do not want to be publicly disclosed in your comment submission. The NRC will post all comment submissions at http://www.regulations.gov as well as enter the comment submissions into ADAMS. The NRC does not routinely edit comment submissions to remove identifying or contact information.

If you are requesting or aggregating comments from other persons for submission to the NRC, then you should inform those persons not to include identifying or contact information that they do not want to be publicly disclosed in their comment submission. Your request should state that the NRC does not routinely edit comment submissions to remove such information before making the comment submissions available to the public or entering the comment into ADAMS.

On July 20, 2015, Mr. Kurt T. Schaefer filed a PRM with the Commission, PRM-50-112 (ADAMS Accession No. ML15278A208), which was subsequently supplemented on August 31, 2015 (ADAMS Accession No. ML15278A211). The petitioner states that he is a nuclear engineer with over 40 years of nuclear experience, and 30 years of nuclear power plant licensing experience. The petitioner claims to have taught numerous classes related to § 50.59 of title 10 of the Code of Federal Regulations (CFR), “Changes, test, and experiments.” The petitioner notes that he is a nuclear licensing contractor and consultant, and that he is “supporting utility and vendor implementation of the United Arab Emirates Federal Authority of Nuclear Regulation (FANR) version of 10 CFR 50.59.”

The petitioner requests that the NRC amend 10 CFR 50.2, “Definitions,” to include a definition of “Important to safety” that provides specific criteria for determining what SSCs and functions are “important to safety.”

The petitioner requests that the NRC amend its regulations in 10 CFR 50.2 to include a definition with specific criteria for determining what SSCs and functions are “important to safety.” The petitioner states that “[t]he nuclear industry is on its third generation of engineers and regulators with no clear definition of what is `important to safety' ” and that “there is no excuse for not having a concise set of functional criteria defining such a used term.”

The petitioner notes that the “NRC staff's current position is that SSCs `important to safety' consists of two subcategories, `safety-related' and `nonsafety-related'.” The petitioner asserts that while safety-related SSCs are defined in 10 CFR 50.2, “the regulations do not provide an equivalent set of criteria for determining which nonsafety-related SSCs are `important to safety.' ” The petitioner notes that there is very little agreement about what “nonsafety-related structures, systems and components (SSCs) should be categorized as `important to safety'.” Furthermore, the petitioner states that “there is only a general description of what is `important to safety' in 10 CFR 50 Appendix A, and the regulations do not provide a specific set of criteria for determining which SSCs are `important to safety'.” The petitioner states that NRC Generic Letter 84-01, “NRC use of the terms, `Important to Safety' and `Safety Related',” and its attachments (January 5, 1984; ADAMS Accession No. ML031150515), sought to clarify the NRC staff's use of these terms, but did not “provide a specific set of criteria for determining which nonsafety-related SSCs are to be categorized as `important to safety'.” The petitioner asserts that this lack of clarity is problematic because “important to safety” is used “in numerous regulations and NRC guidance documents.” The petitioner notes that consequently, “there are regulations, regulatory guidance and routinely generated regulatory evaluations, based on SSCs with no specific criteria that determines what are the applicable SSCs.”

The petitioner requests that the NRC define “important to safety” as SSCs and functions that are:

(a) Safety-related SSCs (including supporting auxiliaries) as defined in 10 CFR 50.2 and their associated safety-related functions;

(b) Equipment and function(s) assumed or used to mitigate the anticipated operational occurrences and non-accident events evaluated in the Final Safety Analysis Report (as updated) or Design Control Document Tier 2 safety analyses;

(c) Equipment and functions assumed or used to prevent or mitigate internal events that involve common cause failures and/or failures beyond the 10 CFR part 50, appendix A, single failure criterion, which have been postulated to demonstrate some specific mitigation capability in accordance with regulatory requirements, as described in the Final Safety Analysis Report (as updated) or Design Control Document Tier 2;

(d) Equipment and functions whose failure or malfunction could impair the ability of other equipment to perform a safety-related function;

(e) Equipment and functions requiring (for ensuring nuclear safety) elevated quality assurance or design requirements (i.e., special treatment), but not to full safety-related standards;

(f) Nonsafety-related readiness functions of installed plant equipment and their associated plant condition(s) assumed, prior to the initiation of an accident, in any accident safety analysis described in the Final Safety Analysis Report (as updated) or Design Control Document Tier 2;

(g) Nonsafety-related structures, systems, components and functions specifically included in the plant design to control the release of radioactive materials within 10 CFR part 20 limits, as described in the Final Safety Analysis Report (as updated) or Design Control Document Tier 2;

(h) Specific (10 CFR 50.150) aircraft impact assessment design features and functional capabilities, as described in the Final Safety Analysis Report (as updated) or Design Control Document Tier 2;

(i) Fukushima Dai-ichi accident mitigation related new or modified manual actions and equipment (including associated functional capabilities), as described in the current plant licensing basis; and

(j) Severe accident mitigation related new or modified manual actions and equipment (including associated functional capabilities), as described in the current plant licensing basis.

The NRC is seeking advice and recommendations from the public on the PRM. We are particularly interested in comments and supporting rationale from the public on the following:

1. On January 5, 1984, the NRC issued Generic Letter 84-01, “NRC Use of the Terms, `Important to Safety' and Safety Related',” to address concerns on the NRC use of the terms “important to safety” and “safety related” and provided the NRC staff's position on safety classification. In SECY-85-119, “Issuance of Proposed Rule on the Important-To-Safety Issue,” dated April 5, 1985 (ADAMS Accession No. ML15322A002), the NRC staff requested Commission approval to clarify the terms “important to safety” and “safety related” through rulemaking. The proposed rule would have defined these terms generally and clarified specifically the nature and extent of certain affected quality assurance requirements. The NRC staff also looked at determining what equipment should be classified as important to safety and what requirements are imposed on this class of equipment. In the Staff Requirements Memorandum (SRM) to SECY-85-119, SRM-SECY-85-119, “Issuance of Proposed Rule on the Important-To-Safety Issue,” dated December 31, 1985 (ADAMS Accession No. ML15322A003), the Commission disapproved the NRC staff's proposed rulemaking actions. In the SRM, the Commission informed the NRC staff that the proposed rule did not adequately differentiate nor clarify the terms “Important-to-Safety” and “Safety Related.” The Commission reiterated in the SRM that it continues to believe that it is necessary to resolve the apparent confusion surrounding usage of the term “Important-to-Safety.” In SECY-86-164, “Proposed Rule on the Important-To-Safety,” dated May 29, 1986 (ADAMS Accession No. ML15322A005), the NRC staff recommended changes to the proposed rule in SECY-85-119 that would address the Commission comments in the SRM to SECY-85-119. In a memo from the Secretary of the Commission dated June 24, 1991 (ADAMS Accession No. ML15322A006), the request for rulemaking in SECY-86-164 was withdrawn. Please provide any new information and analysis that could provide the basis for changes to the NRC's regulations.

2. The NRC requests specific examples where the lack of a formal NRC definition (i.e., codified in 10 CFR chapter I) of the terms, “safety related,” and “important to safety” directly resulted in adverse consequences to external stakeholders. The NRC's evaluation of the cost and benefits of adopting a formal definition would be enhanced if commenters provided a quantitative estimate of the costs and/or unachieved benefits due to the lack of formal definitions of these two terms.

3. What regulations would have to be revised to reflect the new definition, and what would be the nature (objective) of the revision for each provision of the regulation which must be revised?

4. What, if any, guidance would be needed to implement the new definition, and what should be the scope, level of detail, and content of the guidance?

The NRC has determined that the petition meets the threshold sufficiency requirements for docketing a petition for rulemaking under 10 CFR 2.802, “Petition for rulemaking,” and the petition has been docketed as PRM-50-112. The NRC will examine the issues raised in PRM-50-112 to determine whether they should be considered in rulemaking.

I. Obtaining Information and Submitting Comments A. Obtaining Information

Please refer to Docket ID NRC-2015-0156 when contacting the NRC about the availability of information for this action. You may obtain publicly-available information related to this action by any of the following methods:

• Federal Rulemaking Web site: Go to http://www.regulations.gov and search for Docket ID NRC-2015-0156.

• NRC's Agencywide Documents Access and Management System (ADAMS): You may obtain publicly-available documents online in the ADAMS Public Documents collection at http://www.nrc.gov/reading-rm/adams.html. To begin the search, select “ADAMS Public Documents” and then select “Begin Web-based ADAMS Search.” For problems with ADAMS, please contact the NRC's Public Document Room (PDR) reference staff at 1-800-397-4209, 301-415-4737, or by email to [email protected]. For the convenience of the reader, instructions about obtaining materials referenced in this document are provided in the “Availability of Documents” section.

• NRC's PDR: You may examine and purchase copies of public documents at the NRC's PDR, Room O1-F21, One White Flint North, 11555 Rockville Pike, Rockville, Maryland 20852.

Please include Docket ID NRC-2015-0156 in the subject line of your comment submission.

The NRC cautions you not to include identifying or contact information that you do not want to be publicly disclosed in your comment submission. The NRC will post all comment submissions at http://www.regulations.gov as well as enter the comment submissions into ADAMS. The NRC does not routinely edit comment submissions to remove identifying or contact information.

If you are requesting or aggregating comments from other persons for submission to the NRC, then you should inform those persons not to include identifying or contact information that they do not want to be publicly disclosed in their comment submission. Your request should state that the NRC does not routinely edit comment submissions to remove such information before making the comment submissions available to the public or entering the comment into ADAMS.

This proposed rule is limited to the changes contained in Amendment No. 9, Revision 1, to CoC No. 1014 and does not include other aspects of the Holtec HI-STORM 100 Cask System design. Because the NRC considers this action noncontroversial and routine, the NRC is publishing this proposed rule concurrently with a direct final rule in the Rules and Regulations section of this issue of the Federal Register. Adequate protection of public health and safety continues to be ensured. The direct final rule will become effective on March 21, 2016. If the NRC receives significant adverse comments on this proposed rule by February 5, 2016, then the NRC will publish a Federal Register notice withdrawing the direct final rule. If the direct final rule is withdrawn, the NRC will address the comments received in response to these proposed revisions in a subsequent final rule. Absent significant modifications to the proposed revisions requiring republication, the NRC will not initiate a second comment period on this action in the event the direct final rule is withdrawn.

A significant adverse comment is a comment where the commenter explains why the rule would be inappropriate, including challenges to the rule's underlying premise or approach, or would be ineffective or unacceptable without a change. A comment is adverse and significant if:

(1) The comment opposes the rule and provides a reason sufficient to require a substantive response in a notice-and-comment process. For example, a substantive response is required when:

(a) The comment causes the NRC staff to reevaluate (or reconsider) its position or conduct additional analysis;

(b) The comment raises an issue serious enough to warrant a substantive response to clarify or complete the record; or

(c) The comment raises a relevant issue that was not previously addressed or considered by the NRC staff.

(2) The comment proposes a change or an addition to the rule, and it is apparent that the rule would be ineffective or unacceptable without incorporation of the change or addition.

(3) The comment causes the NRC staff to make a change (other than editorial) to the rule, CoC, or Technical Specifications.

For additional procedural information and the regulatory analysis, see the direct final rule published in the Rules and Regulations section of this issue of the Federal Register.

Section 218(a) of the Nuclear Waste Policy Act (NWPA) of 1982, as amended, requires that “the Secretary [of the U.S. Department of Energy] shall establish a demonstration program, in cooperation with the private sector, for the dry storage of spent nuclear fuel at civilian nuclear power reactor sites, with the objective of establishing one or more technologies that the [U.S. Nuclear Regulatory] Commission may, by rule, approve for use at the sites of civilian nuclear power reactors without, to the maximum extent practicable, the need for additional site-specific approvals by the Commission.” Section 133 of the NWPA states, in part, that “[the Commission] shall, by rule, establish procedures for the licensing of any technology approved by the Commission under Section 219(a) [sic: 218(a)] for use at the site of any civilian nuclear power reactor.”

To implement this mandate, the Commission approved dry storage of spent nuclear fuel in NRC-approved casks under a general license by publishing a final rule which added a new subpart K in part 72 of title 10 of the Code of Federal Regulations (10 CFR) entitled, “General License for Storage of Spent Fuel at Power Reactor Sites” (55 FR 29181; July 18, 1990). This rule also established a new subpart L within 10 CFR part 72 entitled, “Approval of Spent Fuel Storage Casks,” which contains procedures and criteria for obtaining NRC approval of spent fuel storage cask designs. The NRC subsequently issued a final rule (65 FR 25241; May 1, 2000) that approved the HI-STORM 100 Cask System design and added it to the list of NRC-approved cask designs in 10 CFR 72.214, “List of approved spent fuel storage casks,” as CoC No. 1014. Most recently, the NRC issued a final rule effective on March 11, 2014 (78 FR 73379), that approved the HI-STORM 100 Cask System design amendment subject to this rulemaking and added it to the list of NRC-approved cask designs in 10 CFR 72.214 as CoC No. 1014, Amendment No. 9.

The Plain Writing Act of 2010 (Pub. L. 111-274) requires Federal agencies to write documents in a clear, concise, well-organized manner that also follows other best practices appropriate to the subject or field and the intended audience. The NRC has written this document to be consistent with the Plain Writing Act as well as the Presidential Memorandum, “Plain Language in Government Writing,” published June 10, 1998 (63 FR 31883). The NRC requests comment on the proposed rule with respect to clarity and effectiveness of the language used.

The documents identified in the following table are available to interested persons through one or more of the following methods, as indicated.

The NRC may post materials related to this document, including public comments, on the Federal rulemaking Web site at http://www.regulations.gov under Docket ID NRC-2015-0156. The Federal Rulemaking Web site allows you to receive alerts when changes or additions occur in a docket folder. To subscribe: (1) Navigate to the docket folder (NRC-2015-0156); (2) click the “Sign up for Email Alerts” link; and (3) enter your email address and select how frequently you would like to receive emails (daily, weekly, or monthly).

For the reasons set out in the preamble and under the authority of the Atomic Energy Act of 1954, as amended; the Energy Reorganization Act of 1974, as amended; the Nuclear Waste Policy Act of 1982, as amended; and 5 U.S.C. 552 and 553; the NRC is proposing to adopt the following amendments to 10 CFR part 72:

In the final rules section of this Federal Register, EPA is approving the state's SIP revision as a direct final rule without prior proposal because the Agency views this as a noncontroversial revision amendment and anticipates no relevant adverse comments to this action. A detailed rationale for the approval is set forth in the direct final rule. If no relevant adverse comments are received in response to this action, no further activity is contemplated in relation to this action. If EPA receives relevant adverse comments, the direct final rule will be withdrawn and all public comments received will be addressed in a subsequent final rule based on this proposed action. EPA will not institute a second comment period on this action. Any parties interested in commenting on this action should do so at this time. Please note that if EPA receives adverse comment on part of this rule and if that part can be severed from the remainder of the rule, EPA may adopt as final those parts of the rule that are not the subject of an adverse comment. For additional information, see the direct final rule which is located in the rules section of this Federal Register.

Table of Contents I. General Information A. Does this action apply to me? B. What action is the Agency taking? C. What is the Agency's authority for taking this action? II. Background A. Statutory and Regulatory Overview B. MS4 Permitting Requirements III. Judicial Review of the Phase II Rule and Partial Remand A. Decision in Environmental Defense Center et al. v. EPA B. EPA Action Following the Partial Remand of the Phase II Rule IV. Scope of This Rulemaking V. EPA's Evaluation and Selection of Rulemaking Options A. Current Permitting Authority Practice B. Description of Process Used To Evaluate Options C. Considerations in Evaluating Options 1. Permitting Authority Review 2. Public Participation Requirements 3. Other Factors Considered VI. Analysis of Options for Proposal A. Option 1—The Traditional General Permit Approach 1. Current Examples of Clear, Specific, and Measurable Permit Requirements 2. Types of Permit Language Lacking Sufficient Detail To Qualify as Clear, Specific, and Measurable 3. Summary/Description of Proposed Rule Changes B. Option 2—Procedural Approach C. Option 3—State Choice Approach VII. Incremental Costs of Proposed Rule Options VIII. Statutory and Executive Orders Reviews A. Executive Order 12866: Regulatory Planning and Review and Executive Order 13563: Improving Regulation and Regulatory Review B. Paperwork Reduction Act (PRA) C. Regulatory Flexibility Act (RFA) D. Unfunded Mandates Reform Act (UMRA) E. Executive Order 13132: Federalism F. Executive Order 13175: Consultation and Coordination With Indian Tribal Governments G. Executive Order 13045: Protection of Children From Environmental Health Risks and Safety Risks H. Executive Order 13211: Actions Concerning Regulations That Significantly Affect Energy Supply, Distribution or Use I. National Technology Transfer and Advancement Act J. Executive Order 12898: Federal Actions To Address Environmental Justice in Minority Populations and Low-Income Populations I. General Information A. Does this action apply to me?

Entities potentially regulated by this proposed action include:

This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely to be regulated by this action. This table lists the types of entities that EPA is now aware could potentially be regulated or otherwise affected by this action. Other types of entities not listed in the table could also be regulated. To determine whether your entity is regulated by this action, you should carefully examine the applicability criteria found in § 122.32 title 40 of the Code of Federal Regulations, and the discussion in the preamble. If you have questions regarding the applicability of this action to a particular entity, consult the person listed in the FOR FURTHER INFORMATION CONTACT section.

EPA is proposing a change to its regulations governing the way in which small MS4s obtain coverage under National Pollutant Discharge Elimination System (NPDES) general permits. The proposal results from a decision by the Ninth Circuit U.S. Court of Appeals in Environmental Defense Center, et al. v. EPA, in 344 F.3d 832 (9th Cir. 2003) (“EDC decision”), which found that EPA regulations for obtaining coverage under a small MS4 general permit did not provide for adequate public notice, the opportunity to request a hearing, or permit authority review to determine whether the BMPs selected by each MS4 in its stormwater management program (SWMP) meets the Clean Water Act (CWA) requirements including the requirement to “reduce pollutants to the maximum extent practicable.” The preamble discusses two options for addressing the remand, and a third option that is a hybrid of the two alternatives. One option (called the “Traditional General Permit Approach”) would align the process for issuing small MS4 general permits with the way NPDES general permits are issued for other categories of discharges. This would entail requiring the permitting authority to establish within the permit all requirements that MS4s must meet within the term of the general permit to meet the standard applicable to MS4s (to reduce pollutants to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the CWA), which would be subject to public notice and comment and an opportunity to request a hearing. A second option (called the “Procedural Approach”) would add procedural requirements to the existing rule structure that would require the MS4 to inform the permitting authority in its Notice of Intent (NOI) to be covered by the permit of the BMPs it would undertake through its SWMP. Under the Procedural Approach, the public would be given an opportunity to comment on the proposed BMPs and request a hearing, and the permitting authority would have the opportunity to require changes to the proposed BMPs before the permitting authority authorizes a discharge under the general permit. A third option (called the “State Choice Approach”) would enable the permitting authority to choose between the Traditional General Permit and Procedural Approaches, or to implement a combination of these approaches in issuing and authorizing coverage under a general permit.

The authority for this rule is the Federal Water Pollution Control Act, 33 U.S.C. 1251 et seq., including sections 402 and 501.

Stormwater discharges are a significant cause of water quality impairment because they contain a variety of pollutants such as sediment, nutrients, chlorides, pathogens, metals, and trash. Furthermore, the increased volume and velocity of stormwater discharges that result from the creation of impervious cover can alter streams and rivers by causing scouring and erosion. These surface water impacts threaten public health and safety due to flooding and pollutants; lead to economic losses to property and fishing industries; increase drinking water treatment costs; and decrease opportunities for recreation, swimming, and wildlife uses.

Stormwater discharges are subject to regulation under section 402(p) of the CWA. Under this provision, Congress required only the following stormwater discharges to be subject to NPDES permitting requirements: Stormwater discharges for which NPDES permits were issued prior to February 4, 1987; discharges “associated with industrial activity”; discharges from MS4s serving populations of 100,000 or more; and any stormwater discharge determined by EPA or a state to “contribute . . . to a violation of a water quality standard or to be a significant contributor of pollutants to waters of the United States.” With respect to MS4s, section 402(p)(3)(B) provides that NPDES permits may be issued on a system-wide or jurisdiction-wide basis, and requires that MS4 NPDES permits “include a requirement to effectively prohibit non-stormwater discharges into the storm sewers” and require “controls to reduce the discharge of pollutants to the maximum extent practicable . . . and such other provisions as the Administrator or the State determines appropriate for the control of such pollutants.”

EPA developed the stormwater regulations under section 402(p) in two phases, as directed by the statute. In the first phase, under section 402(p)(4), EPA promulgated regulations establishing application and other requirements for NPDES permits for stormwater discharges from medium (serving populations of 100,000 to 250,000) and large (serving populations of 250,000 or more) MS4s, and stormwater discharges associated with industrial activity. EPA published the final Phase I rule on November 16, 1990 (55 FR 47990).

The Phase I rule, among other things, defined “municipal separate storm sewer” as publicly-owned conveyances or systems of conveyances that discharge to waters of the U.S. and are designed or used for collecting or conveying stormwater, are not combined sewers, and are not part of a publicly-owned treatment works at 40 CFR 122.26(b)(8). EPA included construction sites disturbing five acres or more in the definition of “stormwater discharges associated with industrial activity” at 40 CFR 122.26(b)(14)(x).

In the second phase, under section 402(p)(5) and (6), EPA was required to conduct a study to identify other stormwater discharges that needed further controls “to protect water quality,” report to Congress on the results of the study, and to designate for regulation additional categories of stormwater discharges not regulated in Phase I on the basis of the study and in consultation with state and local officials. EPA promulgated the Phase II rule on December 8, 1999, designating discharges from certain small MS4s and from small construction sites (disturbing equal to or greater than one acre and less than five acres) and requiring NPDES permits for these discharges (64 FR 68722, December 8, 1999). A regulated small MS4 is generally defined as any MS4 that is not already covered by the Phase I program and that is located within the urbanized area boundary as determined by the latest U.S. Decennial Census. Separate storm sewer systems such as those serving military bases, universities, large hospital or prison complexes, and highways are also included in the definition of “small MS4.” 40 CFR 122.26(b)(16). In addition, the Phase II rule includes authority for EPA (or states authorized to administer the NPDES program) to require NPDES permits for currently unregulated stormwater discharges by a designation process. 40 CFR 122.26(a)(9)(i)(C) and (D). Other small MS4s located outside of an urbanized area may be designated as a regulated small MS4 if the NPDES permitting authority determines that its discharges cause, or have the potential to cause, an adverse impact on water quality. See 40 CFR 122.32(a)(2) and 123.35(b)(3).

The Phase I regulations are primarily application requirements that identify components that must be addressed in applications for individual permits from large and medium MS4s. The regulations at 40 CFR 122.26(d)(2)(iv) require these MS4s to develop a SWMP, which is considered by EPA or the authorized state permitting authority when establishing permit conditions to reduce pollutants to the MEP.

Like the Phase I rule, the Phase II rule requires regulated small MS4s to develop and implement SWMPs. 40 CFR 122.34(a) requires that SWMPs be designed to reduce pollutants discharged from the MS4 “to the maximum extent practicable (MEP), to protect water quality, and to satisfy the appropriate water quality requirements of the Clean Water Act,” and requires that the SWMPs include six “minimum control measures.” The minimum control measures are: Public education and outreach, public participation and involvement, illicit discharge detection and elimination, construction site runoff control, post construction runoff control, pollution prevention and good housekeeping. 40 CFR 122.34(b). Under the Phase II rule, a regulated small MS4 may seek coverage under an available general permit or may apply for an individual permit. To be authorized to discharge under a general permit, the rule requires submission of an NOI to be covered by the general permit containing a description of the BMPs to be implemented and the measurable goals for each of the BMPs, including timing and frequency, as appropriate. 40 CFR 122.33(a)(1), 122.34(d)(1).

EPA anticipated that under the first two or three permit cycles, whether individual permits or general permits, BMP-based SWMPs implementing the six minimum control measures would, if properly implemented, “be sufficiently stringent to protect water quality, including water quality standards, so that additional, more stringent and/or more prescriptive water quality based effluent limitations will be unnecessary.” (64 FR 68753, December 8, 1999). In the final Phase II rule preamble, EPA also stated that it “has intentionally not provided a precise definition of MEP to allow maximum flexibility in MS4 permitting. MS4s need the flexibility to optimize reductions in storm water pollutants on a location-by-location basis. . . . Therefore, each permittee will determine appropriate BMPs to satisfy each of the six minimum control measures through an evaluative process.” (64 FR 68754, December 8, 1999).

The Agency described this process in the preamble to the Phase II rule as an “iterative process” of developing, implementing, and improving stormwater control measures contained in SWMPs. As EPA further stated in the preamble to the Phase II rule, “MEP should continually adapt to current conditions and BMP effectiveness and should strive to attain water quality standards. Successive iterations of the mix of BMPs and measurable goals will be driven by the objective of assuring maintenance of water quality standards. . . . If, after implementing the six minimum control measures there is still water quality impairment associated with discharges from the MS4, after successive permit terms the permittee will need to expand or better tailor its BMPs within the scope of the six minimum control measures for each subsequent permit.” (64 FR 68754, December 8, 1999).

The Phase II rule was challenged in petitions for review filed by environmental groups, municipal organizations, and industry groups, resulting in a partial remand of the rule. Environmental Defense Center v. U.S. Environmental Protection Agency, 344 F.3d. 832 (9th Cir. 2003). The court remanded the Phase II rule's provisions for small MS4 NPDES general permits because they lacked procedures for permitting authority review and public notice and the opportunity to request a hearing on NOIs submitted under general MS4 permits.

In reviewing how the Phase II rule provided for general permit coverage for small MS4s, the court found that NOIs under the rule were not like NOIs for other NPDES general permits. Other general permits contain the specific effluent limitations and conditions applicable to the class of dischargers for which the permit is available, and authorization to discharge under a general permit is obtained by filing an NOI in which the discharger agrees to comply with the terms of the general permit. In contrast, the court held that under the Phase II rule, because the NOI submitted by the MS4 contains the information as to what the MS4 decides it will do to reduce pollutants to the MEP, it is the “functional equivalent” of a permit application. Environmental Defense Center v. U.S. Environmental Protection Agency, 344 F.3d. at 857. Because the CWA requires public notice and the opportunity to request a public hearing for all permit applications, the court held that failure to require public notice and the opportunity for a public hearing for NOIs under the Phase II rule is contrary to the Act. 344 F.3d. at 858.

Similarly, the court found the Phase II rule allows the MS4 to identify the BMPs that it will undertake in its SWMP without any permitting authority review. The court held that the lack of review “to ensure that the measures that any given operator of a small MS4 has decided to undertake will in fact reduce discharges of pollutants to the maximum extent practicable” also does not comport with CWA requirements. The court stated, “That the Rule allows a permitting authority to review an NOI is not enough; every permit must comply with the standards articulated by the Clean Water Act, and unless every NOI issued under general permit is reviewed, there is no way to ensure that such compliance has been achieved.” 344 F.3d. at 855 n.32.

The court therefore vacated and remanded “those portions of the Phase II Rule that address these procedural issues . . . so that EPA may take appropriate action to comply with Clean Water Act.” 344 F.3d. at 858.

EPA issued interim guidance to address the need for permitting authority review of NOIs and to provide for public notice and opportunity for public hearing in April 2004. This guidance memorandum, Implementing the Partial Remand of the Stormwater Phase II Regulations Regarding Notices of Intent and NPDES General Permitting for Phase II MS4s, outlined recommendations as to how permitting authorities should retroactively provide for public notice and the opportunity to request a hearing, provided options for holding a public hearing if granting a request, and highlighted ways to conduct appropriate review of NOIs already submitted.1 The memorandum also provided guidance on ways to ensure the requisite public notice and review opportunities and permitting authority review of NOIs under new general permits. As a result of the EDC decision, EPA Regions that issue NPDES permits have taken various approaches to provide opportunity for public review. For example, EPA Region 1, the permitting authority for Massachusetts and New Hampshire, uses its Web site to post NOIs and notices of availability for public comment, as well as the annual reports submitted by each permitted MS4.2 EPA Region 6, the permitting authority in New Mexico and in Indian Country in Oklahoma and New Mexico, has established a Web site with information on how to submit comments and opportunity to request a public hearing, and posts the NOI and each MS4's SWMP on its Web site.3 EPA Region 10, the permitting authority in Idaho, has only issued individual permits to small MS4s in that state.

In addition, the EPA Regions and some authorized state permitting authorities have included more specific and definitive requirements in small MS4 general permits, rather than leaving the identification of stormwater controls needed to reduce pollutants to the MEP, protect water quality and meet the water quality requirements of the CWA up to the permittees. In the time since promulgation of the Phase II rule and the partial remand of the rule, permits for small MS4 discharges have evolved, both to reflect the advancement and improvement in stormwater management approaches and techniques and to reflect the need for the specific requirements for compliance with the CWA to be incorporated into MS4 permits. Please see Section V.A of this preamble for a detailed discussion of current EPA and state permitting practices for small MS4 NPDES permits.

The proposed revisions to the Phase II MS4 NPDES permitting requirements are solely for the purpose of responding to the partial remand of the Phase II rule in Environmental Defense Center v. U.S. Environmental Protection Agency, 344 F.3d. 832 (9th Cir. 2003) with respect to small MS4 general permits. To conform to the court's decision, the rule needs to ensure that permitting authorities determine what requirements are needed to reduce pollutants from each permitted small MS4 “to the maximum extent practicable (MEP), to protect water quality, and to satisfy the appropriate water quality requirements of the Clean Water Act,” as currently required for small MS4 permits under 40 CFR 122.34(a). The proposed rule must also require NPDES permitting authorities to provide the public with the opportunity to review, submit comments, and request a public hearing on these requirements.

EPA is not reopening any of the substantive requirements that were promulgated in the Phase II rule (nor is EPA reopening or seeking comment on any aspect of the Phase I rule, which is described in this preamble for informational purposes only). In addition, EPA will address the other aspect of the Ninth Circuit's remand regarding possible regulation of stormwater discharges from forest roads in a separate action.

The EPA collected information on how NPDES permitting authorities have been administering their small MS4 general permits in the years since the EDC decision and the issuance of the EPA's guidance on implementing the remand and compiled this information in a state-by-state spreadsheet (titled Current NPDES Authority Practices in Administering Small MS4 General Permits, EPA, 2015), which is available in the docket for the proposed rule at http://www.regulations.gov under Docket ID No. EPA-HQ-OW-2015-0671. This information provides a basis for understanding how and to what degree different rule options would affect the current MS4 general permit programs in different states.

This research indicates that permitting authorities are using an array of approaches to provide permit coverage to their small MS4s, many of which are unique to the specific state. EPA's guidance following the EDC decision suggested ways to implement a general permit program that would be consistent with the court's ruling. As mentioned, some states chose to develop more definitive general permits that do not rely on MS4 identification of BMPs to establish requirements that meet the applicable CWA standards. Other states require that each NOI undergo individualized permitting authority review and a dedicated public comment period prior to authorizing the discharge. Still other states require the MS4 to provide for public notice and the opportunity to submit comments on the NOI and the SWMP document being submitted. Notwithstanding the disparity in approaches between NPDES authorities, this information has equipped EPA with a sense of how the different options under consideration would be implemented if promulgated, and what types of adjustments may be necessary in some programs depending on the rule approach that is adopted. EPA used the approaches being implemented in certain states to inform the proposed rule options.

Not surprisingly, general permits are used as the permitting vehicle to authorize small MS4 discharges in the vast majority of states (i.e., 43 of 50 states, which represents 94 percent of the 6789 permitted small MS4s). In the remaining states, individual permits are issued to their small MS4 permittees. In the 43 states where general permits are used, 26 of these permitting authorities make their NOIs publicly available through a Web site or some other means, and 27 indicate that they provide a “waiting period” of some length between the time the NOI is submitted and discharge authorization. Currently, most states are not providing a second public comment period for individual NOIs (in addition to the public comment period for the draft general permit). However, 12 states have established such a comment period. EPA notes that four states require the prospective small MS4 permittee to provide for its own public comment period for the NOI and, in some cases, the SWMP. In 23 states, the permitting authority requires the SWMP document to be submitted for review along with the NOI; in 14 of these states, the permitting authority reviews and approves the SMWP document. See Current NPDES Authority Practices in Administering Small MS4 General Permits, EPA, 2015.

EPA also found some states that have moved to develop general permits with more clear and specific requirements as a way of cutting down on the need for additional review procedures for individual NOIs. For instance, rather than requiring NOIs with information on BMPs and measurable goals, California and Washington include in their general permits the specific tasks, milestones, and schedules that are to be met by each permittee. Therefore, once coverage under the general permit in these states is authorized, the enforceable components of the permit are locked in place for each permittee, and the permitting authority is no longer required to review the information submitted by individual MS4s prior to authorizing the discharge. What matters is whether the permittee is complying with the specific requirements of the permit.

EPA met separately with various categories of stakeholders during the development of the proposed rulemaking. The purpose of these meetings was to obtain individual feedback from stakeholders on the type of regulatory changes that would best address the court remand, and which would work best considering how Phase II general permits have been administered to date. The following is a summary of what EPA learned from these meetings.

EPA participated in several meetings with the Association of Clean Water Administrators and their member state stormwater coordinators, and met with the Environmental Council of the States. Many state permitting authority staff appeared receptive to the idea of clarifying in the regulations that the general permit should define all of the applicable requirements necessary to reduce the discharge of pollutants from the MS4 to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the CWA. At the same time, some state staff questioned how they would incorporate requirements into their general permits in a way that would work for all MS4s within their state, given the large number and diversity of the municipal entities regulated. Other state staff indicated a concern for retaining the correct balance between establishing detailed, prescriptive requirements and providing flexibility where appropriate. There are also a few state permitting authorities that are implementing an approach similar to what is being described as the “Procedural Approach” (see Section VI.B), and some expressed the interest in finding a way in the proposed rule to accommodate this approach. Most state permitting staff appeared concerned with the prospect of spending additional time and resources to implement a procedural approach requiring individualized review and public notice of all NOIs, as discussed in the court's decision. Other state permitting staff suggested exploring the concept of allowing permitting authorities to choose which option to follow, without restricting the rule to one approach. Alternatively, a few state permitting staff suggested that permitting authorities be allowed to apply a hybrid of the two approaches, whereby a state could implement one permit using the Traditional General Permit Approach (e.g., for traditional MS4s) and another permit using the Procedural Approach (e.g., for non-traditional MS4s), or use a blend of the options for issuing a general permit and authorizing coverage under the permit.

EPA met with organizations representing state and local elected officials, as well as with small MS4 permittees and organizations that include small MS4s as members. MS4s, in particular, are interested in retaining the flexibility of the existing Phase II regulations, where they are able to make decisions on which BMPs are implemented locally based on factors that are unique to their municipality and environmental concerns. At the same time, many of these same MS4s understand the need for permit requirements that are clear to all parties and the public.

EPA also met with representatives from a number of environmental, non-profit organizations. Many of the representatives expressed an interest in seeing the quality of small MS4 permits improve, and appeared to be supportive of the concept of adopting the Traditional General Approach as a way of addressing the remand. Asked at what point in the current permitting process their organizations tend to provide input, most indicated that they focus their attention on providing comments at the proposed permit stage, as compared to submitting comments on individual NOIs. That being said, a few representatives indicated that they have submitted comments on individual NOIs pertaining to the proposed water quality implementation plans of several small MS4s.

Any option for responding to the remand must meet the CWA requirements for public participation and transparency in section 402(b)(3), consistent with the Ninth Circuit's decision. When individual permits are issued to small MS4s, the standard process for issuing an NPDES permit applies. This process provides for public participation and permitting authority determination as to what set of permit terms and conditions satisfy the requirement to reduce the discharge of pollutants from the MS4 to the MEP, to protect water quality, and to meet the applicable water quality requirements of the CWA. While the court's opinion focused on the Phase II rule's requirement for the NOI to be covered by a general permit, and the procedural steps that need to be taken with respect to the NOI in order for the rule to comply with the CWA, the court's fundamental concern was that the permitting authority must determine which MS4 permit requirements are sufficient to reduce the discharge of pollutants to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the CWA, and that the public have the opportunity to review and comment on those permit requirements and to request a hearing. For example, the court stated that “every permit must comply with the standards articulated by the Clean Water Act, and unless every NOI issued under a general permit is reviewed, there is no way to ensure that such compliance has been achieved.” EDC v. EPA. 344 F.3d at 855, n. 32. Accordingly, EPA has determined that certain factors must be met by any option to revise the rule, as discussed in subsections 1 (Permitting Authority Review), 2 (Public Participation Requirements), and 3 (Other Factors Considered).

The court viewed the NOI as the document that identifies the requirements necessary to meet the MEP standard: “Because a Phase II NOI establishes what the discharger will do to reduce discharges to the `maximum extent practicable,' the Phase II NOI crosses the threshold from being an item of procedural correspondence to being a substantive component of a regulatory scheme.” 344 F.3d at 853. As a result, the role of the permitting authority to determine which requirements are necessary to meet the applicable statutory standard is not, according to the court, accomplished under this scheme. In addition, the court observed that because 40 CFR 122.34(a) in the 1999 Phase II rule states that compliance with the SWMP written by the MS4 constitutes compliance with the MEP standard (without providing for further action by the permitting authority), the regulation put the MS4 in charge of establishing its own requirements. “Therefore, under the Phase II Rule nothing prevents the operator of a small MS4 from misunderstanding or misrepresenting its own stormwater situation and proposing a set of minimum measures for itself that would reduce discharges by far less than the maximum extent practicable.” 344 F.3d at 855.

While EPA has always expected the permitting authority to establish the necessary requirements for reducing discharges to the MEP, protecting water quality, and satisfying the appropriate water quality requirements of the CWA, the existing regulations do not fully address the permitting authorities' responsibilities in this regard. To be consistent with the court's decision, one criterion that any option must meet is that it must ensure the permitting authority provides a final determination on whether the requirements to which the MS4 is subject, whether articulated fully in the permit itself or defined in whole or part by the MS4 operator in the NOI, meet the NPDES requirements to reduce discharges to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the Act.

The court's other concern was that MS4s would choose what requirements apply to them, without being subject to the public participation procedures applicable to all NPDES permit applications and permits, which is contrary to CWA section 402(b)(3). As discussed, the court found the NOI to be the “functional equivalent” of a permit application. The importance of the NOI as identified by the court was that the NOI contained the requirements that would be considered to meet the applicable standards and therefore this was the document that needed to be subject to public notice. See 344 F.3d at 857. To be consistent with the court's decision, any option chosen must provide for public notice and the opportunity to request a public hearing on what is considered necessary for a permitted MS4 to meet the requirement to reduce discharges to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the CWA, regardless of where those requirements are defined.

General permits are premised on the idea that the terms and conditions of the permit are the same for all entities covered by the general permit and that handling permitting for multiple entities in one proceeding is more efficient. In the context of MS4 permits, the Phase II rule sought to establish a general permit scheme that allows each MS4 to address the specific conditions that prevail in its jurisdiction. As stated in the Phase II preamble, “The pollutant reductions that represent MEP may be different for each small MS4, given the unique local hydrologic and geologic concerns that may exist and the differing possible pollutant control strategies. Therefore, each permittee will determine appropriate BMPs to satisfy each of the six minimum control measures through an evaluative process.” (64 FR 68754, December 8, 1999). While the court clearly rejected EPA regulations to the extent that the court found they established a system of MS4 self-regulation, it also recognized the value in having MS4 input on what it could do to meet the MEP standard. “Involving regulated parties in the development of individualized stormwater pollution control programs is a laudable step . . . But EPA is still required to ensure that the individual programs adopted are consistent with the law.” 344 F.3d at 856. There is a need for strong MS4 input into the implementation of the program, and for that reason EPA made flexibility an underlying principle of the Phase II regulations. Individual permits provide the greatest ability to define MS4-specific requirements and small MS4s always have the option of seeking an individual permit if this would best accommodate their specific circumstances. However, with over 94 percent of regulated small MS4s currently covered by general permits, an important consideration for this rulemaking is how to provide flexibility to MS4s while retaining the general permit option in a manner that comports with the remand. The challenge is to balance the flexibility provided to the MS4 to determine how best it can meet the applicable regulatory requirements with the permitting authorities' responsibility to ensure that the terms and conditions to which MS4s will be held accountable are adequate to reduce the discharge to the MEP, protect water quality, and satisfy the appropriate water quality requirements of the CWA. In selecting any regulatory option to comport with the court remand, EPA will consider the need for maintaining this balance in light of the nearly 15-year history of implementing the Phase II program, and the considerable knowledge and expertise about implementing stormwater controls that have emerged during that time.

Another factor requiring consideration is the impact on existing authorized NPDES state permitting programs. Currently 46 states and one territory are authorized under section 402(b) to administer the NPDES permit program in their jurisdictions. EPA recognizes that states have limited resources and face different challenges in meeting the permitting demands within their various NPDES programs. Immediately after the EDC decision, EPA sought to provide state permitting authorities with potential interim strategies that would balance the need to move forward with implementing the Phase II program, while acknowledging the need for state flexibility in how permitting decisions need to be made. See Implementing the Partial Remand of the Stormwater Phase II Regulations Regarding Notices of Intent & NPDES General Permitting for Phase II MS4s (EPA, 2004).4 As discussed more fully elsewhere in this preamble, authorized states [and EPA regional permitting authorities] have taken a variety of approaches in response to the court's decision (and in some cases, decisions by state courts) and EPA guidance. A significant consideration in this rulemaking is the extent to which states would need to make changes to comply with the rule and consideration of the need to minimize disruption to existing state programs, particularly for those states that have chosen approaches that already comport with the EDC decision. EPA clarifies that if, upon promulgation of the final rule, a state is already implementing an approach that is consistent with the final rule EPA would not expect that the permitting authority would need to make any changes to its current approach. Similarly, it is EPA's intention that permitting authorities that only issue individual permits to small MS4s (e.g., EPA Region 10 in Idaho, Delaware, Michigan, and Oregon) would not need to make any changes because the process for issuing individual permits already encompasses the necessary permitting attributes found missing in the Phase II regulations by the Ninth Circuit (i.e., permitting authority determination, public notice, and opportunity to request a hearing). However, state permitting authorities that are using general permits and are currently not implementing strategies that address the core problems found by the court will need to make some degree of change to their general permit process for small MS4s to comply with the modified regulations.

EPA is proposing three rule options for public comment, each of which would address the Ninth Circuit remand. Each of these options shares in common the fact that, as a result of the permitting process, the permitting authority must determine which requirements a small MS4 must meet in order to satisfy the Phase II regulatory requirement “to reduce the discharge of pollutants from [the] MS4 to the maximum extent practicable, to protect water quality, and to satisfy the appropriate water quality requirement of the Clean Water Act.” The key difference between the options, especially between the “Traditional General Permit Approach” (Option 1) and the “Procedural Approach” (Option 2), is that they make this determination at different points in time during the permitting process. For Option 1 (the “Traditional General Permit Approach”), the determination as to what requirements are needed to reduce the discharge of pollutants to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the CWA is made as part of the initial issuance of the general permit. By contrast, under Option 2 (the “Procedural Approach”), the permitting authority would make this determination after reviewing each individual NOI and after public comment and the opportunity for a hearing on the NOI. Each of these options is described more fully in this section, as is a third option (the “State Choice Approach”), which would give the permitting authority the discretion to determine whether it will administer Option1 or Option 2, or a hybrid of options chosen for the final rule.

The “Traditional General Permit Approach” provides a mechanism for addressing the procedural deficiencies identified by the court by requiring all substantive permit requirements to be in the general permit. The rationale behind the Traditional General Permit Approach is that by requiring permitting authorities to include any and all requirements that establish what is necessary to “. . . reduce the discharge of pollutants from the MS4 to the maximum extent practicable (MEP), to protect water quality, and to satisfy the appropriating water quality requirements of the Clean Water Act,” the minimum required procedural steps to issue a final general permit, including providing public notice and the minimum 30-day comment period on the draft permit, and the opportunity to request a public hearing, will fulfill the permitting authority review and public participation requirements of the CWA that the court found missing from the Phase II regulations.

Under the proposed Traditional General Permit Approach, the NPDES authority must establish in any small MS4 general permit the full set of requirements that are deemed adequate “to reduce the discharge of pollutants from the MS4 to the maximum extent practicable (MEP), to protect water quality, and to satisfy the appropriate water quality requirements of the Clean Water Act,” and the administrative record would explain the rationale for its determination. The permittee would have the opportunity, as it always has had, to provide feedback on what requirements are established in the general permit during the development of the draft permit and to submit comments during the public comment period. Furthermore, the permittee could continue to have flexibility in determining how it will implement the permit requirements based on considerations such as pollutant removal and cost effectiveness. However, once the permit is issued, and the terms and conditions in the permit are fixed for the term of the permit, neither the development of a SWMP document nor the submittal of an NOI for coverage would represent new permit requirements. In turn, because the permit contains all of the requirements that will be used to assess permittee compliance, the permitting authority would no longer need to rely on the MS4's NOI as the mechanism for ascertaining what will occur during the permit term. Under this approach, the function of the NOI would be more similar to that of any other general permit NOI, and more specifically other stormwater general permits, where the NOI is used to establish certain minimum facts about the discharger, including the operator's contact details, the discharge location(s), and confirmation that the operator is eligible for permit coverage and has agreed to comply with the terms of the permit. By removing the possibility that effluent limits could be proposed in the NOI (and for that matter in the SWMP) and made part of the permit once permit coverage is provided, the NOI would no longer look and function like an individual permit application, as the court found with respect to MS4 NOIs under the Phase II regulations currently in effect. Therefore, it would not be necessary to carry out the type of additional permitting authority review and public participation steps contemplated by the court.

Under the proposed Traditional General Permit Approach, 40 CFR 122.34(a) would be revised to expressly require the permitting authority to articulate in sufficient detail in the permit what is required to meet the minimum statutory and regulatory requirements, and to ensure that the applicable requirements are enforceable and understandable to the permittee and the public. A general permit would need to make it clear to all what level of effort is expected of the permittee during the permit term for each permit provision. These proposed revisions to 40 CFR 122.34(a) respond to the court's finding that under the Phase II rule, “the operator of a small MS4 has complied with the requirement of reducing discharges to the `maximum extent practicable' when it implements its stormwater management program, i.e., when it implements its Minimum Measures. 40 CFR 122.34(a).” 344 F.3d at 856. The court continued, “Nothing in the Phase II regulations requires that NPDES permitting authorities review these Minimum Measures to ensure that the measures that any given operator of a small MS4 had decided to undertake will in fact reduce discharges to the maximum extent practicable.” 344 F.3d at 855. By clearly shifting the decision as to what is needed to meet the MEP standard and water quality requirements from the permittee to the permitting authority, the Traditional General Permit Approach would address the court's concern.

EPA continues to view MEP as iterative, in that each successive permit needs to define what is required to meet the MEP standard for that permit term. The Traditional General Permit Approach would clarify that the requirements for meeting MEP (and to protect water quality and satisfy CWA water quality requirements) would be required to be established in each successive permit by the permitting authority, while the SWMP implemented by the MS4 would be a planning and programmatic document that the MS4 would be able to update and revise during the permit term as necessary to comply with the terms of the permit. In other words, this option would make it clear that the SWMP document would not contain enforceable requirements. Likewise, it would be unnecessary for the NOI to identify the BMPs selected in the SWMP for each minimum control measure nor for it to undergo public or permitting authority review prior to discharge authorization under the general permit.

Moreover, it was never EPA's intent that the SWMP required by 40 CFR 122.34(a) itself be considered enforceable under the permit. Rather, the SWMP was intended to be the means for the MS4 to engage in an adaptive management process during the term of the permit. “EPA envisions application of the MEP standards as an iterative process. MEP should continually adapt to current conditions and BMP effectiveness and should strive to attain water quality standards.” (64 FR 68754, December 8, 1999).

The Traditional General Permit Approach would include regulatory text to reflect EPA's guidance to permitting authorities regarding the types of permit requirements for MS4s that are considered most effective. For instance, EPA advises permitting authorities to use permit conditions that are “clear, specific, and measurable.” See MS4 Permit Improvement Guide 5 (p. 5-6), and Revisions to the November 22, 2002 Memorandum Establishing Total Maximum Daily Load (TMDL) Wasteload Allocations (WLAs) for Storm Water Sources and NPDES Permit Requirements Based on Those WLAs 6 (p. 5). The MS4 Permit Improvement Guide explains EPA's recommendation as follows:

The proposed rule for the Traditional General Permit Approach would obligate the permitting authority to establish requirements that are “clear, specific, and measurable.” See proposed 40 CFR 122.34(a). The proposed rule further explains that effluent limitations may be expressed as BMPs that include, but are not limited to, “specific tasks, BMP design requirements, performance requirements or benchmarks, schedules for implementation and maintenance, and frequency of actions.” Id. Where permits incorporate clear, specific, and measurable requirements, EPA expects there to be greater certainty and understanding as to what must be accomplished during each permit term.

A foundational principle of MS4 permits is that from permit term to permit term iterative progress will be made towards meeting water quality objectives, and that adjustments in the form of modified permit requirements will be made where necessary to reflect current water quality conditions, BMP effectiveness, and other current relevant information. This principle is incorporated into the proposed Traditional General Permit Approach in the requirement for NPDES authorities to revisit permit requirements during the permit issuance process, and to make any necessary changes in order to ensure that the subsequent permit continues to meet the NPDES requirements “to reduce the discharge of pollutants from the MS4 to the maximum extent practicable (MEP), protect water quality, and to satisfy the water quality requirements of the Clean Water Act.” Thus, in advance of issuing any successive small MS4 general permit, the permitting authority would need to review, among other things, information on the relative progress made by permittees to meet applicable milestones, compliance problems that may have arisen, the effectiveness of the required activities and selected BMPs under the existing permit, and any improvements or degradation in water quality. Sources of this information include, but are not limited to:

• Past annual reports;

• Current SWMP documents;

• NPDES MS4 audit reports, construction/industrial/commercial site inspection reports;

• Monitoring and other information on quality of receiving waters;

• Existing MS4 permit requirements; and

• Approved TMDLs that include wasteload allocations applicable to small MS4s.

As discussed in the previous section, a key component of the proposed Traditional General Permit Approach is that permits be written with sufficient clarity and specificity to enable permittees, the public, and regulatory authorities alike to understand what is required to measure progress. EPA acknowledges that meeting the requirement to include more detailed terms and conditions in small MS4 permits and to ensure, among other things, that the permit terms satisfy the regulatory requirement to reduce pollutant discharges from the MS4 to the MEP (and meet the requirement to protect water quality and meet the appropriate water quality requirements of the CWA) will not be easy for some states. States that have not already written permits in this way would need to evaluate the quality of the existing SWMPs, the track record of each MS4 in implementing their respective SWMPs, the types of BMPs that have proven effective, and information that may suggest what is necessary to address existing water quality conditions, including whether additional requirements are needed to address an applicable TMDL. Among other factors that the state would need to consider when issuing a new, or the next, general permit are how long the MS4 has been permitted, the degree of progress made by the small MS4 permittees as a whole and for individual MS4s as well, the reasons for any lack of progress, and the capability of these MS4s to achieve more focused requirements. EPA finds promise in some of the strategies that EPA and state permitting authorities are already implementing, which will serve as useful models to those permitting authorities needing advice on how to write their permits under the proposed Traditional General Permit Approach. For example, permitting authorities may find that subcategorizing MS4s by experience, size, or other factors, and creating different requirements for each subcategory, may be desirable. Permitting authorities may also consider whether watershed-wide general permits may be an option, especially where the receiving waters are impaired.

In addition to the model permit language in the MS4 Permit Improvement Guide, EPA recently compiled a number of examples where small MS4 general permits have already included requirements that are clear, specific, and measurable in a document entitled MS4 General Permits and the Six Minimum Control Measures: A National Compendium of Clear, Specific, and Measurable Requirements, which can be accessed in the docket for this proposed rule. Additional examples of clear, specific, and measurable permit requirements in MS4 general permits, focusing on post-construction requirements and water quality-based effluent limits, are included in EPA's Municipal Separate Storm Sewer System Permits: Post-Construction Performance Standards & Water Quality-Based Requirements: A Compendium of Permitting Approaches.7 The fact that many permitting authorities have already included provisions that would qualify as clear, specific, and measurable under the proposed rule indicates that making this a requirement for all permits is reasonable and achievable. EPA requests comment on what additional examples should be highlighted as being clear, specific, and measurable in current small MS4 general permits.

Just as there are a number of examples to be highlighted where states are already writing their permits consistent with the proposed Traditional General Permit Approach, EPA also found permits that lack adequate detail and would not qualify as clear, specific, and measurable under the proposed rule modifications. Permit requirements that do not appear to have the type of detail that would be needed under the proposed rule approach may have some of the following characteristics:

• Permit provisions that simply copy the language of the Phase II regulations verbatim without providing further detail on the level of effort required or that do not include the minimum actions that must be carried out during the permit term. For instance, where a permit includes the language in 40 CFR 122.34(b)(4)(ii)(B) (i.e., requiring “. . . construction site operators to implement appropriate erosion and sediment control best management practices”) and does not provide further details on the minimum set of accepted practices, the requirement would not provide clear, specific, and measurable requirements within the intended meaning of the proposed Traditional General Permit Approach. The same would also be true if the permit just copies the language from the other minimum control measure provisions in 40 CFR 122.34(b) without further detailing the particular actions and schedules that must be achieved during the permit term.

• Permit requirements that include “caveat” language, such as “if feasible,” “if practicable,” “to the maximum extent practicable,” and “as necessary” or “as appropriate” unless defined. Without defining parameters for such terms (for example, “infeasible” means “not technologically possible or not economically practicable and achievable in light of best industry practices”), this type of language creates uncertainty as to what specific actions the permittee is expected to take, and is therefore difficult to comply with and assess compliance.

• Permit provisions that preface the requirement with non-mandatory words, such as “should” or “the permittee is encouraged to . . . .” This type of permit language makes it difficult to assess compliance since it is ultimately left to the judgment of the permittee as to whether it will comply. EPA notes that the Phase II regulations include “guidance” in places (e.g., 40 CFR 122.34(b)(1)(ii), (b)(2)(ii), and (b)(3)(iv)), which suggest practices for adoption by MS4s and within permits, but does not mandate that they be adopted. This guidance language is intended for permitting authorities to consider in establishing their permit requirements. While permitting authorities may find it helpful to their permittees to include guidance language within their permits in order to provide suggestions to their permittees, such language would not qualify as a permit requirement under the proposed Traditional General Permit Approach.

• Permit requirements that lack a measurable component. For instance, several permits include language implementing the construction minimum control measure that requires inspections “at a frequency determined by the permittee” based on a number of factors. This type of provision includes no minimum frequency that can be used to measure adequacy and, therefore, would not constitute a measurable requirement for the purposes of the proposed rule.

• Permit requires the development of a plan to implement one of the minimum control measures, but does not include details on the minimum contents or requirements for the plan, or the required outcomes, deadlines, and corresponding milestones. For example, some permits require the MS4 to develop a plan to implement the public education minimum control measure, which informs the public about steps they can take to reduce stormwater pollution. The requirement leaves all of the decisions on what specific actions will be taken during the permit term to comply with this provision to the MS4 permittee, thus enabling almost any type of activity, no matter how minor or insubstantial, to be considered compliance with the permit. In EPA's view, this type of permit provision would not qualify as a clear, specific, and measurable requirement under the proposed Traditional General Permit Approach.

The following is a section-by-section summary of the proposed regulatory changes.

The following changes to 40 CFR 122.33 are proposed to complement the changes made to implement the Traditional General Permit Approach option:

• Throughout the section references to “you” or “your” would be replaced with references to “the operator.” This change is proposed for consistency with revisions to 40 CFR 122.34 and 40 CFR 122.35.

• The requirements for obtaining coverage under a general permit would now be the same as those for any other general permit in 40 CFR 122.28(b)(2). The NOI would no longer be required to include information on the MS4's BMPs and measurable goals.

• The requirements for applying for an individual permit would be consolidated in 40 CFR 122.33(b)(2), whereas these requirements now appear in both 40 CFR 122.31 and in 40 CFR 122.34(d).

• The deadline of March 10, 2003 for MS4s wishing to implement a program that differed from 40 CFR 122.34 to submit an individual permit application would be removed since the date has passed and is no longer relevant. Similarly, the deadline of March 10, 2003 for MS4s designated for regulation by 40 CFR 122.32(a)(1) would be deleted since the date has passed and is no longer relevant.

Most of the proposed changes to 40 CFR 122.34 are made to clarify that it is the permitting authority's responsibility, and not that of the small MS4 permittee, to establish permit terms that meet the small MS4 regulatory standard (i.e., “. . . to reduce the discharge of pollutants from the MS4 to the maximum extent practicable (MEP), to protect water quality, and to satisfy the appropriate water quality requirements of the Clean Water Act.”), and to delineate the requirements for implementing the six minimum control measures, other more stringent effluent limitations as necessary, as well as other requirements. The proposed modifications do not alter the existing, substantive requirements of the six minimum control measures in 40 CFR 122.34(b), but instead emphasize the way in which the permitting authority makes the determination as to what requirements are included in small MS4 permits, including general permits. For instance, a typical change in the proposed Traditional General Permit Approach is made in 40 CFR 122.34(b)(3)(ii), which transfers the obligation to address certain categories of non-stormwater discharges from the small MS4 operator (referred to as “you”) to the permitting authority by requiring that “the permit must require the permittee to address the following categories of non-storm water discharges.” Otherwise, unless specified, there is no change to the language of the existing rule.

The following changes to 40 CFR 122.34(a) are proposed:

• The proposed regulatory text clarifies that the permitting authority is required to include in any small MS4 permit conditions that ensure pollutant discharges from the MS4 are reduced to the MEP, are protective of water quality, and satisfy the water quality requirements of the CWA. In order to ensure that these permit conditions are of adequate detail and their meaning is clear to all parties, the proposed rule emphasizes that permit requirements must be written in a “clear, specific, and measurable” form. This language is consistent with the recommendation in EPA's MS4 Permit Improvement Guide (2010), which advised permitting authorities to write MS4 permits with permit provisions that are “clear, specific, measurable, and enforceable.” In addition, the proposed regulatory text for the Traditional General Permit Approach emphasizes that the permit requirements must be adequate to collectively meet the regulatory standard, that is: “to reduce the discharge of pollutants from the MS4 to the maximum extent practicable (MEP), to protect water quality, and to satisfy the appropriate water quality requirements of the Clean Water Act (CWA).” EPA notes that no changes are proposed to the wording of this regulatory standard.

• The proposed regulatory text reiterates that effluent limitations may be in the form of BMPs, and provides examples of how these BMP requirements may appear in the permit, such as in the form of specific tasks, BMP design requirements, performance requirements or benchmarks, schedules for implementation and maintenance, and the frequency of actions. This list of examples is not intended to be exclusive, and EPA anticipates that permitting authorities will, over time, develop other ways to establish requirements that are consistent with this language. It is EPA's view that this proposed language serves the same underlying purpose as the provision it modifies in the current regulation (i.e., “. . . narrative effluent limitations requiring implementation of best management practices (BMPs) are generally the most appropriate form of effluent limitations when designed satisfy technology requirements . . . and to protect water quality.”)

• The following provision from the existing regulations is proposed to be removed: “Implementation of best management practices consistent with the provisions of the storm water management program required pursuant to this section and the provisions of the permit required pursuant to § 122.33 constitutes compliance with the standard of reducing pollutants to the `maximum extent practicable.' ” The court in EDC found this sentence to be particularly problematic in light of the lack of permitting authority review of NOIs. Based in part on this language, the court observed that “the operator of a small MS4 needs to do nothing more than decide for itself what reduction in discharges would be the maximum practical reduction.” EDC at 855. Furthermore, the court found that “under the Phase II Rule, nothing prevents the operator of a small MS4 from misunderstanding or misrepresenting its own stormwater situation and proposing a set of minimum measures for itself that would reduce discharges by far less than the maximum extent practicable.” Id. EPA addresses these concerns by removing this language, and instead clarifying, as it does through the other proposed changes to 40 CFR 122.34(a), that it is the permitting authority who is responsible for establishing requirements that constitute compliance with requirement to reduce the discharge of pollutants from the MS4 to the MEP, to protect water quality, and to satisfy the water quality requirements of the CWA.

• The language in the existing regulations providing permittees with up to five years from the date of permit issuance to implement their SWMPs is modified to apply to new permittees, recognizing that this 5-year period has passed for existing permittees. Another clarification is included to explain that when a permit is expiring and a new permit is being developed, the permitting authority must ensure that the new permit meets the requirements of 40 CFR 122.34(a) based on current water quality conditions, the record of BMP effectiveness, and other current relevant information. This revision would not change the status quo; it merely recognizes that first-time small MS4 permittees have up to five years to develop and implement their SWMPs, while small MS4s that have already been permitted will have developed and implemented their SWMP when they reapply for permit coverage or submit an NOI under the next small MS4 general permit.

The following changes are proposed to be made to 40 CFR 122.34(b):

• In the proposed regulatory text, the small MS4 operator is still required to develop a SWMP; however, the stated purpose of the SWMP is clarified to emphasize the fact that it is a tool for describing how the permittee will comply with the permit requirements implementing the six minimum control measures, and does not contain effluent limitations or permit conditions. The effluent limitations and other enforceable conditions would be stated in the permit itself. The proposed regulatory text for the Traditional General Permit Approach would clarify that for general permits, documentation of the measurable goals in the SWMP should include schedules that are consistent with any deadlines already established in the general permit. The purpose of this proposed requirement is to preserve the SWMP as a tool for permittees to describe [in more detail] how the MS4 will implement the BMPs required by the permit and to document updates to the SWMP as needed during the permit term if changes are called for to comply with the permit. This language is intended to support the underlying clarification in the proposal that it is in the permit where the enforceable requirements are established, while the role of the SWMP document or other document(s) is to describe in writing how the permittee will comply with these requirements. Under this formulation, a permittee's failure to develop a SWMP document would constitute a violation of the permit, but a permittee's failure to install a specific control measure that is described in the SWMP document would not be a violation of the permit, unless the permit required that this specific control measure be installed as a required BMP. EPA notes that the proposed regulatory text also includes language to clarify that whether or not the SWMP can be found in one document or a series of documents, there should be a written description in some form that explains how the permittee will comply with the permit's minimum control measure requirements. In other words, the “SWMP document” refers to the documentation, whether located in one place or comprised of multiple documents (e.g., ordinances, manuals, documented procedures, and other documentation), that is the written form of the permittee's SWMP. Reference to a “document” in the proposed rule is not intended to create a new documentation requirement.

• Changes in various provisions in 40 CFR 122.34(b)(1) through (6) are proposed to emphasize the permitting authority's role in including requirements that address the minimum control measures as compared to the current regulations, which give this responsibility to the MS4. In most instances, the proposed modifications are merely changing a few words to switch from the first person (i.e., “you”) to the third person (i.e., “the MS4”). The proposed modifications do not alter the existing, substantive requirements of the six minimum control measures in 40 CFR 122.34(b).

The following changes are proposed to be made to 40 CFR 122.34(d).

• The proposed regulatory text for the Traditional General Permit Approach would remove existing paragraph (d) from 40 CFR 122.34. The information required to be included in permit applications for individual permits in paragraph (d)(1) would be moved to 40 CFR 122.33(b)(2)(i). This information would no longer be required to be submitted with NOIs. Because EPA and many states have issued menus of BMPs, paragraph (d)(2) is no longer relevant, and under the Traditional General Permit Approach, paragraph (d)(3) would also no longer be needed.

• For general permits, the information required to be included in the NOI would track with the requirements for general permits in 40 CFR 122.28(b)(2)(ii). See discussion on 40 CFR 122.33. There would be no change to the requirement that an MS4 seeking an individual permit must submit an application with its proposed BMPs to implement the six minimum control measures and measurable goals for BMP implementation.

The following changes are proposed to be made to 40 CFR 122.34(e) and (f):

• The proposal would consolidate the current requirements in 40 CFR 122.34(e)(1) and (f) under one section, 40 CFR 122.34(c), entitled “Other applicable requirements.”

• EPA proposes to remove the guidance in the current regulations at § 122.34(e)(2). The guidance reflects EPA's recommendation for the initial round of permit issuance, which has already occurred for all permitting authorities. The phrasing of the guidance language no longer represents EPA policy with respect to including additional, more stringent requirements. EPA has found that a number of permitting authorities are already including specific requirements in their small MS4 permits that address not only wasteload allocations in TMDLs, but also other more stringent requirements that are in addition to the six minimum measures irrespective of the status of EPA's 40 CFR 122.37 evaluation. See EPA's Municipal Separate Storm Sewer System Permits—Post-Construction Performance Standards & Water Quality-Based Requirements: A Compendium of Permitting Approaches (2014). Based on the advancements made by specific permitting programs, and information that points to stormwater discharges continuing to cause waterbody impairments around the country, EPA has advised in guidance that permitting authorities write MS4 permits with provisions that are “clear, specific, measurable, and enforceable,” incorporating such requirements as clear performance standards, and including measurable goals or quantifiable targets for implementation. See EPA's MS4 Permit Improvement Guide (2010). This guidance is a more accurate reflection of the Agency's current views on how the Phase II regulations should be implemented than the guidance currently in 40 CFR 122.34(e)(2).

The following changes are proposed to be made to 40 CFR 122.34(c) and (g):

• The existing “qualifying local program” provision currently in 40 CFR 122.34(c) would be renumbered as 40 CFR 122.34(e).

• The “evaluation and assessment” provision currently in 40 CFR 122.34(g) would be renumbered as 40 CFR 122.34(d). Conforming changes would be made to 40 CFR 122.35 to update the cross-references in that section.

Another option, called the “Procedural Approach,” for which EPA requests comment would address the remand by incorporating additional permitting authority and public review steps into the existing regulatory framework for providing coverage to small MS4s under general permits. EPA is not proposing specific regulatory text for this option, but has included a detailed description of how the Procedural Approach would work. In addition to comments on the merits of the option, EPA solicits comments recommending specific regulatory text for this option.

Under the existing regulation, 40 CFR 122.34(d)(1), MS4s seeking authorization to discharge under a general permit must submit an NOI that identifies the BMPs that the MS4 will implement for each of the six minimum control measures. The NOI must also state the measurable goals for each of the BMPs, including the timing and frequency of their implementation. Under the Procedural Approach, once an MS4 operator submits its NOI requesting coverage under the general permit, an additional step would take place in which the permitting authority would review, and the public would be given an opportunity to comment and request a hearing on, the merits of the MS4's proposed BMPs and measurable goals for complying with the requirement to reduce discharges to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the CWA.

Under the “Procedural Approach” option, the existing regulatory requirement for the small MS4 to submit an NOI with the BMPS and measurable goals as provided in 40 CFR 122.34(d) and the requirement in 40 CFR 122.34(a) to develop, implement, and enforce a SWMP to meet the six minimum measures and to reduce pollutant discharges to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the CWA would be retained. In this option, the NOI would continue to be used in the same way as the court considered the NOI in the EDC case. The NOI would continue to serve as the document that describes the BMPs and measurable goals that would be considered to be the enforceable requirements applicable to the permittee, in addition to the terms and conditions of the general permit. While a SWMP would still need to be developed, it would not establish enforceable requirements beyond those identified in the NOI that would have undergone public notice and comment and permitting authority review.

The process would occur in the following sequence: Following the receipt of an NOI for coverage under the general permit, the permitting authority would review the NOI to assess whether the proposed BMPs and measurable goals meet the requirements to reduce pollutants to the MEP, protect water quality, and satisfy the water quality requirements of the CWA. If not, the permitting authority would request supplemental information or revisions as necessary to ensure that the submission satisfies the regulatory requirements. Once satisfied with the submission, the Procedural Approach would require the permitting authority to provide public notice of the NOI and an opportunity to request a hearing on the NOI, in accordance with 40 CFR 124.10 through 124.13. After consideration of comments received and a hearing, if held, the permitting authority would provide notice of its decision to authorize coverage under the general permit and with the specific requirements each MS4 must meet, in accordance with 40 CFR 124.15, or as provided by state law for providing notice of a final permit decision in authorized states. Upon completion of this process, the MS4-specific requirements in the NOI, together with the terms and conditions set forth in the general permit, would be incorporated as requirements of the permit for the particular MS4.

Where the state is the permitting authority, it would also provide EPA an opportunity to review the individual NOIs and submit comments or objections to the state regarding the adequacy of the NOI before it is made available for public review, consistent with requirements under 40 CFR part 124 for NPDES permit applications and under 40 CFR 123.44 for draft permits. This two-step Procedural Approach is similar to the procedure used to establish “terms of the nutrient management plan” permit requirements proposed by concentrated animal feeding operations (CAFOs) seeking coverage under a general permit under 40 CFR 122.23(h). While Option 2 still relies on the use of a general permit, it follows several of the same process steps as those used for an individual permit.

Some states, including Minnesota and Texas, have used a similar procedural approach as a way to address the problems identified in the EDC decision. In Minnesota, for example, the state has developed a detailed form that must be completed by any small MS4 seeking coverage under the Minnesota general permit, which when completed will become in effect its SWMP document (referred to as a “Stormwater Pollution Prevention Plan Document” of “SWPPP Document”). The state then reviews the MS4's submission and determines whether revisions are needed to meet the requirements of the permit. After any necessary revisions, the state provides public notice of the NOI and SWPPP Document, and makes them available for public review and comment, and for any requests to hold a public hearing. After considering public comments, the state then makes a final determination on whether to authorize coverage under the general permit, and, if authorized, the contents of the SWPPP Document (as revised when necessary following public comment) become enforceable under the general permit. The Minnesota approach gives MS4s flexibility by providing a range of options from which an MS4 can choose for its particular circumstances. It also provides the public with the opportunity to review the MS4's proposed choices and the permitting authority's determination of adequacy, and to provide comment and request a hearing. The MS4's proposed program for implementing the six minimum measures goes into effect only after the state has made an affirmative determination that the MS4's program has met the burden of showing that pollutant discharges will be reduced to the MEP, will be protective of water quality, and will satisfy the appropriate water quality goals of the CWA, thus providing the necessary permitting authority review.

Texas also reviews individual MS4 program documents to determine whether they meet the minimum permit and regulatory requirements. In contrast to the more detailed NOI checklist used by Minnesota, Texas uses a relatively short NOI form but requires the MS4 to submit its entire SWMP document for review after the general permit is issued. It does so with the intent to have the SWMP document identify the MS4-specific enforceable requirements, rather than to have this information contained in the NOI. Texas requires the MS4 to provide the public notice of the state's preliminary determination to authorize coverage under the general permit in accordance with the SWMP document and an opportunity to comment on the SWMP document and request a hearing. Comments on the adequacy of the SWMP document and requests for public hearings are submitted directly to the state and the state also determines whether there is sufficient interest to hold a public hearing on the SWMP document.

Under the Procedural Approach, EPA would preserve one of the core attributes of the existing regulations, that is the flexibility afforded the MS4 to identify the BMPs that it determines are needed to meet the minimum regulatory requirements to reduce pollutant discharges to the MEP, to protect water quality, and to satisfy the water quality requirements of the CWA in its SWMP. This approach may appeal to states that accept the notion that the MS4 should have the initial opportunity to propose the BMPs that it believes will meet the regulatory requirements, and that each program may differ substantially from MS4 to MS4.

However, the need to undergo a second round of public notice and comment at the state level, in addition to the one provided for the general permit, for approximately 6800 small MS4s, may be seen as a drawback due to the additional workload placed on permitting authorities that do not already follow this approach. The value added by the second comment period is also a consideration. Staff in Minnesota's program reported that while they received over 1500 comments in response to proposing the state-level general permits, only a handful of comments were submitted on the individual MS4 NOI and SWPPP Document submissions during the second public comment period. Staff in Texas' program reported that the state received no comments when it provided public notice on the individual MS4 SWMPs.

Another factor to consider is that under the Procedural Approach some changes to the BMPs and measurable goals identified in the NOI during the term of the permit could constitute a modification to the permit, and would be subject to permit modification procedures applicable to all NPDES permits. See 40 CFR 122.62 and 122.63. For example, if the MS4 decides to discontinue implementing a particular BMP that it included in its NOI (and which became an enforceable permit requirement) and to substitute a different BMP, a permit modification would be needed. It is not clear whether states are currently using permit modification procedures to process changes to a MS4's SWMP. One possibility for addressing the need for change would be for the permitting authority to establish in the general permit itself a process for making changes to the SWMP without triggering the permit modification procedures, as long as it identifies what changes could be made and under what circumstances. EPA seeks comment on whether to provide in the regulations the option for modifying the general permit under the minor modification procedures in 40 CFR 122.63 for “nonsubstantial revisions” to BMPs, as provided for changes to terms of a CAFO's nutrient management plan that are “not substantial” under 40 CFR 122.42(e)(6). EPA also seeks comment on what criteria should apply for distinguishing between when a change to BMPs is “substantial” requiring a full public participation process or “not substantial” that would be subject to public notice but not public comment under a permit modification process similar to the process in 40 CFR 122.42(e)(6).

Like several other states, Texas requires the MS4s to provide local public notice and the opportunity to provide comments on individual MS4 NOIs (or the SWMP, as in Texas). What stands out in the Texas approach is that, even though the MS4 must provide the necessary notice, public comments are submitted to the state agency, and the state clearly maintains the decision making over the adequacy of the MS4's SWMP to meet permit and regulatory requirements. The state does so by reviewing the SWMP document before it is public noticed and evaluating for itself any public comments on the SWMP document and whether there is sufficient interest to require a public hearing. EPA seeks comment on whether a rule establishing a procedural approach should enable permitting authorities that rely on the MS4 to public notice its NOI to be able to use this approach to satisfy the public notice requirement for the individual NOIs. If allowed, should it be limited to when the State clearly makes the ultimate decisions about what requirements are sufficient to meet the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of CWA?

The Texas approach appears to differ from the current procedures that apply to NPDES permits outlined in 40 CFR part 124 in the level of detail about the various procedural requirements such as who must be notified of the proposed action. In this respect, the Texas program resembles EPA's approach to establishing or changing terms of nutrient management plans under CAFO general permits by modifying selected elements of the public participation requirements that apply to individual permits, for example, by shortening the length of public comment period or the period for requesting a public hearing (see 40 CFR 122.23(h)(1) and 122.42(e)(6)), or by allowing web-based public notice alternatives in addition to those identified in 40 CFR 124.10 (c). If EPA chooses to adopt this option, it would largely rely on the existing requirements in 40 CFR part 124 to govern what procedures are necessary to approve the BMPs in the NOI as enforceable provisions of the general permit. However, as discussed, EPA is considering some variations in these 40 CFR part 124 procedural requirements similar to those applicable to incorporating terms of the nutrient management plan into CAFO permits.

Based on the experiences of states that use a similar procedural approach, EPA estimates that conducting individualized reviews of NOIs and requiring an additional notice and comment period for the initial authorization and subsequent permit modifications in states that do not already provide it would require a significant dedication of staff time, in an amount estimated at 24 hours per MS4. Based on Minnesota's experience, EPA expects the workload to be greatest in the first permit cycle but to decrease by some amount in subsequent cycles as the permitting authority takes advantage of efficiencies gained from having gone through the process before and as the quality of the MS4 submissions improve over time. For states that already use a two-step process, some modest amount of workload increase may be necessary to ensure that all of the process steps are carried out, including additional time needed to process and approve SWMP modifications that change the BMPs in the NOI that have been approved and have become enforceable terms of the permit.

The following regulatory modifications are envisioned if the Procedural Approach is selected for the final rule.

• Include additional language indicating that to the extent that the permitting authority chooses to rely on the MS4 operator to describe in its NOI the BMPs, measurable goals, schedules, and other activities in its SWMP that it plans to implement to reduce pollutant discharges to the MEP, to protect water quality, and to satisfy the appropriate water quality requirements of the CWA, the permitting authority will need to incorporate these as enforceable elements of the permit in accordance with the procedures for public notice, the opportunity to request a hearing, and permitting authority final determination in 40 CFR part 124.

• With respect to determining the appropriate 40 CFR part 124 procedures to follow, one model that EPA could utilize in crafting applicable rule language is the regulatory procedures in 40 CFR 122.23(h) for CAFO general permits. While the CAFO and MS4 programs differ fundamentally from one another in many ways, there are some aspects of the CAFO general permit procedures that could be modified in a manner that would make them suitable to small MS4 general permits. Thus, based on some of the key elements of the CAFO general permit procedures in 40 CFR 122.23(h), EPA is considering including the following provisions in revised 40 CFR 122.33(b)(1) as subparagraphs (i)-(iii):

• To accompany these regulatory changes, EPA is also considering specifying what specific information the MS4 will need to provide as part of the NOI in order to obtain coverage under a general permit that will use a procedural approach, such as the approach described previously. The MS4 would need to provide the same information as is required for an application for an individual permit under proposed 40 CFR 122.33(b)(2)(ii). This includes general background information as specified in § 122.21(f) as well as the information currently required by 40 CFR 122.34(d), and any other information requested by the permitting authority.

• If the final rule includes the Procedural Approach or allows for a hybrid approach under Option 3 (the “State Choice Approach”), authorized states would need to revise their approved programs to include the option(s) chosen by the permitting authority and to establish or reference the public notice and comment, hearing request, and other procedures necessary to implement the chosen option(s).

For both the Procedural Approach and State Choice Approach (see Section VI.C), the Agency chose to describe the regulatory changes that would accompany these options if promulgated as opposed to providing line-by-line rule text changes as it has for the Traditional General Permit Approach. In EPA's view, presenting the rule language in this way will aid in the public's review of the three different options as compared to presenting three different sets of line-by-line changes.

EPA requests comment on whether the Agency should adopt as its final rule option the procedural approach for permitting small MS4s. EPA has concerns with adopting this approach as the sole rule option since it would require all but a handful of permitting authorities to change their permitting procedures to conform to this new approach. Due to these concerns, EPA also separately requests comment (see next section) on whether the final rule should give permitting authorities a choice of which approach, either the Traditional General Permit Approach or the Procedural Approach, to adopt for their permitting program, or whether there is support for allowing permitting authorities to use a combination of these two approaches.

Among the concerns EPA has with choosing Option 2 for the final rule is the increase in workload for permitting authorities that would be associated with reviewing and approving, and providing for notice and comment, and providing public hearing opportunities, on each individual NOI. For many permitting authorities, the advantage of providing flexibility to MS4s to propose what they believe will meet the applicable regulatory standards will be outweighed by the resource-intensive procedures that this approach requires. In EPA's discussions with state permitting authorities, the Agency heard a number of concerns about their ability to implement new procedures such as these from a staff and resource perspective. Permitting authorities are also concerned about making individual decisions on what set of MS4 actions are sufficient to meet the regulatory requirements without the benefit of established standards to assist them in making these determinations. Concerns were also raised by many MS4 permittees, who emphasized the effects of these procedures on the timeliness of their discharge authorization, and the fear that states will turn to MS4s to conduct more notice and comment procedures on their behalf. EPA notes that there are also those states that are supportive of making the procedural approach a part of the final rule in some way or form.

Beyond the workload concerns raised about this option, EPA observes that the need for flexibility among MS4s to develop and implement individually tailored SWMPs is different than the type of flexibility required for CAFO operators in developing and implementing nutrient management plans. AFO permit operators must consider where several key and interdependent variables must be considered to account for site-specific factors such as type of crop grown, soil type, terrain, choice of method for calculating application rates, in particular with respect to land application requirements. Each MS4 faces unique circumstances, but for the most part, the BMPs used to meet minimum control measures are not interdependent in the same way as choices needed to develop land application rates under CAFO regulations. EPA and states have developed menus of different BMPs for the various minimum control measures. As discussed previously, some states have developed detailed manuals for the selection, design, installation, and maintenance of allowable BMPs, which further standardizes the practices to be used for pollutant control at MS4s. Also, the need for small MS4 flexibility may have been greater when the small MS4 program was first established. However, this flexibility may be less critical now that most small MS4s have established programs, and they and the corresponding permitting authorities have gained experience in implementing various BMPs and evaluating the results. Permitting authorities already have the flexibility to issue different general permits or include different general permit terms and conditions for different categories of MS4, such as when there is a new group of MS4s that have not been previously regulated (for example, because a new Census is published creating additional urbanized areas) and a group of existing MS4s that may be on their third or fourth permit. By including specific requirements that only apply to some of the MS4s, they undergo permitting authority review and public comment as part of the process and can be part of the general permit itself. (This would be analogous to EPA's Multi Sector General Permit for Stormwater from Industrial Activity, in which different requirements apply to different sectors in the Appendices to the permit).8 For truly unique situations or in instances where the MS4 wishes to implement a different program, individual permits are always an alternative. These factors point to the benefit of using the Traditional General Permit Approach as the preferred way to modify the general permitting regulations for small MS4s. Though there would certainly be increases in workload associated with the Traditional General Permit Approach, EPA's permits and a growing number of state general permits are being written in this manner and therefore would not require significant alteration. Additionally, as the list of examples of clear, specific, and measurable provisions in general permits grows, presumably other states should be able to take advantage of these ideas for their own permits, and thereby save on permit development time. Requiring the procedural approach on a national level would impose pressures on state programs that arguably can be handled in the general permit itself, and therefore avoided.

EPA requests comments on a third option, which would allow permitting authorities to choose either the Traditional General Permit Approach or the Procedural Approach, or some combination of the two as best suits their needs and circumstances. For example, a state could choose to use Option 1 for small MS4s that have fully established programs and uniform core requirements, and Option 2 for MS4s that it finds would benefit from the additional flexibility to address unique circumstances, such as some non-traditional MS4s. Alternatively, a state could apply a hybrid of the two approaches within one permit by defining some elements within the general permit, which are deemed to reduce the discharge of pollutants to the MEP, to protect water quality, and to satisfy the water quality requirements of the CWA, and enabling other elements to be established through a separate process that allows for more MS4-specific actions, using the Procedural Approach. An example of such a hybrid approach might be where a state incorporates into its general permit a requirement to implement certain minimum construction BMP requirements, such as implementation of provisions set forth in a separate statewide manual, which constitute compliance with the regulatory requirements, but leaves it to the MS4 to propose the BMPs that it will implement to meet the public education and outreach requirements of the permit. The former permit requirements would implement the Traditional General Permit Approach and would require no further permitting authority review and public participation procedures during the process of authorizing individual MS4 discharges; however, for the management practices that the MS4 proposes for its public education and outreach, the permitting authority would need to follow the Procedural Approach for incorporating these standards into the permit as requirements of the permit. The benefit of the State Choice Approach is that the fundamental CWA requirements of permitting authority review and public participation would be met irrespective of whether this occurs as a result of the permit issuance itself or whether these procedures take place in a second step that occurs after permit issuance but before the MS4 is authorized to discharge under the permit. This approach would provide for more options for permit development other than traditional individual or general permits. EPA will continue to encourage greater specificity in establishing clear, specific, and measurable permit terms and conditions in the general permit itself, and expects to provide guidance to assist permitting authorities in accomplishing this objective. Nevertheless, the Agency recognizes that permitting authorities may prefer some flexibility in determining the balance between the efficiencies of a general permit and the desirability of providing maximum flexibility to small MS4s in how they will meet the MEP standard.

The particular balance between specificity and flexibility a state chooses could evolve over time as the program continues to mature. The benefit of this option may be that it is the least disruptive to how state programs operate now and would impose the least burden on state permitting authorities, unless a state determines that for its situation (e.g., number and variability among small MS4s, available resources, requirements under state law, etc.) more choices in structuring permits would be desirable. If EPA adopts this option as part of the final rule, the following rule changes would be necessary:

• Adopt the rule changes proposed in this document associated with the Traditional General Permit Approach, as modified pursuant to public comment; and

• Adopt the rule changes described in the discussion under Option 2.

EPA requests comment on whether the final rule should adopt Option 3, as opposed to selecting either Option 1 or Option 2 in the final rule. EPA is also interested in comments from permitting authorities as to which approach they are likely to choose (i.e., Option 1 or Option 2, or a hybrid) if Option 3 is finalized.

EPA also requests comment on whether under Option 3, EPA should consider establishing which permit requirements must be developed using the Traditional General Permit Approach (Option 1), and which may be developed using the Procedural Approach (Option 2). For instance, EPA is interested in finding out whether there is support for requiring permitting authorities to use Option 1 to develop permit conditions implementing the minimum control measures in 40 CFR 122.34(b), while providing the permitting authority with the choice of whether to use an Option 2 approach to establish any more stringent effluent limitations, such as those based on an approved TMDL. Using this approach, the general permit would define the specific actions, performance requirements, and implementation schedules considered necessary to reduce pollutant discharges to the MEP, to protect water quality, and to satisfy the water quality requirements of the CWA. However, this approach would provide the permitting authority the additional flexibility to allow the MS4 to propose in its NOI the specific components of a TMDL implementation plan in order to comply with permit requirements based on applicable wasteload allocation(s). To ensure that the specific actions and timelines of the TMDL plan are properly incorporated as elements of the permit, the permitting authority would then be required to review and approve the small MS4's proposed plan using the process required by the Procedural Approach (Option 2). Additionally, with respect to this concept of specifying which aspects of the small MS4 regulations must be incorporated into permits using the Option 1 approach, while allowing some permit conditions to be developed using the Option 2 approach, EPA requests comment on which permit requirements should be required to be established using Option 1 and which should be given the flexibility to be established using Option 2.

The economic analysis estimates the incremental costs of modifying the Phase II MS4 regulations to address the court's remand. EPA assumed that all other costs accrued as a result of the existing small MS4 program, which were accounted for in the economic analysis accompanying the 1999 final Phase II MS4 regulations, remain the same and are not germane to the economic analysis, unless the proposed rule change would affect the baseline program costs. In this respect, EPA focused only on new costs that may be imposed as a result of implementing any of the three options being proposed for comment. It is, therefore, unnecessary to reevaluate the total program costs of the Phase II rule, since those costs were part of the original economic analysis conducted for the 1999 Phase II rule (see 64 FR 68722, December 8, 1999). For further information, refer to the Economic Analysis that is included in the proposed rule docket.

The following table summarizes the estimated costs for each of the proposed rule options under consideration.

These estimates are all below the threshold level established by statute and various executive orders for determining that a rule has a significant or substantial impact on affected entities. See further discussion in Section VIII of this document.

The Economic Analysis assumes that all costs will be borne by NPDES permitting authorities in the form of increased administrative costs to write more detailed permits for Option 1, or to review and approve and process comments on NOIs submitted for general permit coverage for Option 2. Likewise, Option 3 costs reflect the estimated increase in NPDES permitting authority workload (for both EPA and state permitting authorities), which is a function of an assumed amount of NPDES permitting authorities who will choose to implement Option 1 versus Option 2. EPA does not attribute new costs to regulated small MS4s beyond what they are already subject to under the Phase II regulations. This is because the focus of the proposed rule is on the administrative manner in which general permits are issued and/or coverage under those permits is granted. EPA is changing through this rulemaking any of the underlying requirements in the Phase II regulations to which small MS4s are subject.

EPA chose conservative assumptions about impacts on state workloads, meaning that the estimated economic costs of the policy change are most likely lower than what is actually presented. For instance, EPA did not reduce the number of hours necessary for permitting authorities to draft specific permits pursuant to the Option 1 requirements in the second and third permit term despite the fact that the Agency expects that most permitting authorities, after drafting a specific permit to address Option 1 for the first time would spend less time in subsequent rounds reissuing the same permit. Similarly, in its modeling of Option 2, EPA did not reduce the average number of hours to review each NOI in the second and third permit term, even though EPA expects that most NOIs would address any deficiencies after the first review, therefore resulting in less review time needed in subsequent rounds.

EPA considers the cost assumptions in Option 1 to be conservative because as more permitting authorities write general permits to establish requirements consistent with the proposed Option 1, other permitting authorities could use and build on those examples, reducing the amount of time it takes to draft the permit requirements. EPA has issued guidance to permitting authorities on how to write better MS4 permits (EPA 2010 and EPA 2014), and has included additional examples of permit language from existing permits in the docket for this rule. See General Permits and the Six Minimum Control Measures: A National Compendium of Clear, Specific, and Measurable Requirements. EPA also anticipates providing further guidance once the rule is promulgated to assist states in implementing the new rule requirements, which should make permit writing more efficient.

Additional information about these statutes and Executive Orders can be found at http://www2.epa.gov/laws-regulations/laws-and-executive-orders.

This action is a significant regulatory action that was submitted to the Office of Management and Budget (OMB) for review. Any changes made in response to OMB recommendations have been documented in the docket for this action. In addition, EPA prepared an analysis of the potential costs associated with this action. This analysis, “Economic Analysis for the Proposed Municipal Separate Storm Sewer System (MS4) General Permit Remand Rule,” is summarized in Section V.II and is available in the docket.

This action does not impose any new information collection burden under the PRA. OMB has previously approved the information collection activities contained in the existing regulations and has assigned OMB control number 2040-0004.

I certify that this action will not have a significant economic impact on a substantial number of small entities under the RFA. Although small MS4s are regulated under the Phase II regulations, this rule does not propose changes to the underlying requirements to which these entities are subject. Instead, the focus of this rule is on ensuring that the process by which NPDES permitting authorities authorize discharges from small MS4s using general permits. This action will have an impact on state government agencies that administer the Phase II MS4 permitting program. The impact to states that are NPDES permitting authorities may range from $6,792,106 to $11,356,092 annually. Details of this analysis are presented in “Economic Analysis for the Proposed Municipal Separate Storm Sewer System (MS4) General Permit Remand Rule.”

This action does not contain an unfunded mandate of $100 million or more as described in UMRA, 2 U.S.C. 1531-1538. This action does not significantly or uniquely affect small governments because this rulemaking only affects the way in which state permitting authorities administer general permit coverage to small MS4s. Nonetheless, EPA consulted with small governments concerning the regulatory requirements that might indirectly affect them, as described in section V.B.

This rule will not have substantial direct effects on the states, the relationship between the national government and the states, or on the distribution of power and responsibilities among the various levels of government. The rule proposes changes to the way in which NPDES permitting authorities, including authorized state government agencies, provide general permit coverage to small MS4s. The impact to states which are NPDES permitting authorities may range from $6,792,106 to $11,356,092 annually, depending upon the rule option that is finalized. Details of this analysis are presented in “Economic Analysis for the Proposed Municipal Separate Storm Sewer System (MS4) General Permit Remand Rule,” which is available in the docket for the proposed rule at http://www.regulations.gov under Docket ID No. EPA-HQ-OW-2015-0671.

Keeping with the spirit of E.O. 13132 and consistent with EPA's policy to promote communications between EPA and state and local governments, EPA met with state and local officials throughout the process of developing the proposed rule and received feedback on how proposed options would affect them. EPA engaged in extensive outreach via conference calls to authorized states and regulated MS4s to gather input on how EPA's current regulations are affecting them, and to enable officials of affected state and local governments to have meaningful and timely input into the development of the options presented in this proposed rule.

This action does not have tribal implications as specified in Executive Order 13175 since it does not have a direct substantial impact on one or more federally recognized tribes. The proposed rule affects the way in which small MS4s are covered under a general permit for stormwater discharges and primarily affects the NPDES permitting authorities. No tribal governments are authorized NPDES permitting authorities. The rule could have an indirect impact on an Indian tribe that is a regulated MS4 in that the NOI required for coverage under a general permit may be changed as a result of the rule (if finalized) or may be subject to closer scrutiny by the permitting authority and more of the requirements could be established as enforceable permit conditions. However, the substance of what an MS4 must do in its SWMP will not change significantly as a result of this rule. Thus, Executive Order 13175 does not apply to this action.

Consistent with the EPA Policy on Consultation and Coordination with Indian Tribes, EPA conducted outreach to tribal officials during the development of this action. EPA spoke with tribal members during a conference call with the National Tribal Water Council to gather input on how tribal governments are currently affected by MS4 regulations and may be affected by the options in this proposed rule. Based on this outreach and additional, internal analysis, EPA confirmed that this proposed action would have little tribal impact and would be of little interest to tribes.

EPA interprets Executive Order 13045 as applying only to those regulatory actions that concern environmental health or safety risks that EPA has reason to believe may disproportionately affect children, per the definition of “covered regulatory action” in section 2-202 of the Executive Order. This action is not subject to Executive Order 13045 because it does not concern an environmental health risk or safety risk.

This action is not subject to Executive Order 13211, because it does not significantly affect energy supply, distribution or use.

This rulemaking does not involve technical standards.

EPA determined that the human health or environmental risk addressed by this action will not have potential disproportionately high and adverse human health or environmental effects on minority, low-income or indigenous populations. This action affects the procedures by which NPDES permitting authorities provide general permit coverage for small MS4s, to help ensure that small MS4s “reduce the discharge of pollutants to the maximum extent practicable (MEP), to protect water quality and to satisfy the water quality requirements of the Clean Water Act.” It does not change any current human health or environmental risk standards.

For the reasons set forth in the preamble, EPA proposes to amend 40 CFR part 122 as follows:

Section 4(b)(3)(B) of the Act (16 U.S.C. 1531 et seq.), requires that, for any petition to revise the Federal Lists of Endangered and Threatened Wildlife and Plants that contains substantial scientific or commercial information that listing the species may be warranted, we make a finding within 12 months of the date of receipt of the petition. In this finding, we will determine that the petitioned action is: (1) Not warranted, (2) warranted, or (3) warranted, but the immediate proposal of a regulation implementing the petitioned action is precluded by other pending proposals to determine whether species are endangered or threatened, and expeditious progress is being made to add or remove qualified species from the Federal Lists of Endangered and Threatened Wildlife and Plants. Section 4(b)(3)(C) of the Act requires that we treat a petition for which the requested action is found to be warranted but precluded as though resubmitted on the date of such finding, that is, requiring a subsequent finding to be made within 12 months. We must publish these 12-month findings in the Federal Register.

This finding is based upon the “Status Assessment for the Alexander Archipelago Wolf (Canis lupus ligoni)” (Service 2015, entire) (hereafter, Status Assessment) and the scientific analyses of available information prepared by Service biologists from the Anchorage Fish and Wildlife Field Office, the Alaska Regional Office, and the Headquarters Office. The Status Assessment contains the best scientific and commercial data available concerning the status of the Alexander Archipelago wolf, including the past, present, and future stressors. As such, the Status Assessment provides the scientific basis that informs our regulatory decision in this document, which involves the further application of standards within the Act and its implementing regulations and policies.

On December 17, 1993, the Service received a petition, from the Biodiversity Legal Foundation, Eric Holle, and Martin Berghoffen, to list the Alexander Archipelago wolf as an endangered or threatened species under the Act. On May 20, 1994, we announced a 90-day finding that the petition presented substantial information indicating that the requested action may be warranted, and we initiated a status review of the Alexander Archipelago wolf and opened a public comment period until July 19, 1994 (59 FR 26476). On August 26, 1994, we reopened the comment period on the status review to accept comments until October 1, 1994 (59 FR 44122). The Service issued its 12-month finding that listing the Alexander Archipelago wolf was not warranted on February 23, 1995 (60 FR 10056).

On February 7, 1996, the Southwest Center for Biological Diversity, Biodiversity Legal Foundation, Save the West, Save America's Forests, Native Forest Network, Native Forest Council, Eric Holle, Martin Berghoffen, and Don Muller filed suit in the U.S. Court for the District of Columbia challenging the Service's not-warranted finding. On October 9, 1996, the U.S. District Court remanded the 12-month finding to the Secretary of the Interior, instructing him to reconsider the determination “on the basis of the current forest plan, and status of the wolf and its habitat, as they stand today” (96 CV 00227 DDC). The Court later agreed to the Service's proposal to issue a new finding on June 1, 1997. On December 5, 1996, we published a document announcing the continuation of the status review for the Alexander Archipelago wolf and opening a public comment period until January 21, 1997 (61 FR 64496). The comment period was then extended or reopened through three subsequent publications (61 FR 69065, December 31, 1996; 62 FR 6930, February 14, 1997; 62 FR 14662, March 27, 1997), until it closed on April 4, 1997.

Prior to the publication of a 12-month finding, however, the U.S. Forest Service (USFS) issued the 1997 Tongass Land and Resource Management Plan Revision, which superseded the 1979 version of the plan. In keeping with the U.S. District Court's order that a finding be based upon the “current forest plan,” the District Court granted us an extension until August 31, 1997, to issue our 12-month finding so that the petitioners, the public, and the Service could reconsider the status of the Alexander Archipelago wolf under the revised Tongass Land and Resource Management Plan. Therefore, the Service reopened the public comment period on the status review of the Alexander Archipelago wolf from June 12, 1997, to July 28, 1997 (62 FR 32070, June 12, 1997), and we then reevaluated all of the best available information on the Alexander Archipelago wolf, as well as long-term habitat projections for the Tongass National Forest included in the 1997 Tongass Land and Resource Management Plan Revision. On September 4, 1997, we published a 12-month finding that listing the Alexander Archipelago wolf was not warranted (62 FR 46709).

On August 10, 2011, we received a petition dated August 10, 2011, from the Center for Biological Diversity and Greenpeace, requesting that the Alexander Archipelago wolf be listed as an endangered or threatened species under the Act and critical habitat be designated. Included in the petition was supporting information regarding the subspecies' taxonomy and ecology, distribution, abundance and population trends, causes of mortality, and conservation status. The petitioners also requested that we consider: (1) Prince of Wales Island (POW) as a significant portion of the range of the Alexander Archipelago wolf; and (2) wolves on POW and nearby islands as a distinct population segment. We note here that a significant portion of the range is not a listable entity in and of itself, but instead provides an independent basis for listing and is part of our analysis to determine whether or not listing as an endangered or threatened species is warranted. We published the 90-day finding for the Alexander Archipelago wolf on March 31, 2014, stating that the petition presented substantial information indicating that listing may be warranted (79 FR 17993).

On June 20, 2014, the Center for Biological Diversity, Greenpeace, Inc., and The Boat Company (collectively, plaintiffs) filed a complaint against the Service for failure to complete a 12-month finding for the Alexander Archipelago wolf within the statutory timeframe. On September 22, 2014, the Service and the aforementioned plaintiffs entered into a stipulated settlement agreement stating that the Service shall review the status of the Alexander Archipelago wolf and submit to the Federal Register a 12-month finding as to whether listing as endangered or threatened is warranted, not warranted, or warranted but precluded by other pending proposals, on or before December 31, 2015. In Fiscal Year 2015, the Service initiated work on a 12-month finding for the Alexander Archipelago wolf.

On September 14, 2015, the Service received a petition to list on an emergency basis the Alexander Archipelago wolf as an endangered or threatened species under the Act. The petition for emergency listing was submitted by Alaska Wildlife Alliance, Cascadia Wildlands, Center for Biological Diversity, Greater Southeast Alaska Conservation Community, Greenpeace, and The Boat Company. The petitioners stated that harvest of the Alexander Archipelago wolf in Game Management Unit (GMU) 2, in light of an observed recent population decline, would put the population in danger of extinction. On September 28, 2015, the Service acknowledged receipt of the petition for emergency listing to each of the petitioners. In those letters, we indicated that we would continue to evaluate the status of the Alexander Archipelago wolf as part of the settlement agreement and that if at any point we determined that emergency listing was warranted, an emergency rule may be promptly developed.

This document constitutes the 12-month finding on the August 10, 2011, petition to list the Alexander Archipelago wolf as an endangered or threatened species. For additional information and a detailed discussion of the taxonomy, physical description, distribution, demography, and habitat of the Alexander Archipelago wolf, please see the Status Assessment for Alexander Archipelago Wolf (Canis lupus ligoni) (Service 2015, entire) available under Docket No. FWS-R7-ES-2015-0167 at http://www.regulations.gov, or from the Anchorage Fish and Wildlife Field Office (see ADDRESSES).

Goldman (1937, pp. 39-40) was the first to propose the Alexander Archipelago wolf as a subspecies of the gray wolf. He described C. l. ligoni as a dark colored subspecies of medium size and short pelage (fur) that occupied the Alexander Archipelago and adjacent mainland of southeastern Alaska. Additional morphometric analyses supported the hypothesis that wolves in southeastern Alaska were phenotypically distinct from other gray wolves in Alaska (Pedersen 1982, pp. 345, 360), although results also indicated similarities with wolves that historically occupied coastal British Columbia, Vancouver Island, and perhaps the contiguous western United States (Nowak 1983, pp. 14-15; Friis 1985, p. 82). Collectively, these findings demonstrated that wolves in southeastern Alaska had a closer affinity to wolves to the south compared to wolves to the north, suggesting that either C. l. ligoni was not confined to southeastern Alaska and its southern boundary should be extended southward (Friis 1985, p. 78) or that C. l. ligoni should be combined with C. l. nubilus, the subspecies that historically occupied the central and western United States (Nowak 1995, p. 396). We discuss these morphological studies and others in detail in the Status Assessment (Service 2015, “Morphological analyses”).

More recently, several molecular ecology studies have been conducted on wolves in southeastern Alaska and coastal British Columbia, advancing our knowledge of wolf taxonomy beyond morphometric analyses. Generally, results of these genetic studies were similar, suggesting that coastal wolves in southeastern Alaska and coastal British Columbia are part of the same genetic lineage (Breed 2007, pp. 5, 27, 30; Weckworth et al. 2011, pp. 2, 5) and that they appear to be genetically differentiated from interior continental wolves (Weckworth et al. 2005, p. 924; Munoz-Fuentes et al. 2009, p. 9; Weckworth et al. 2010, p. 368; Cronin et al. 2015, pp. 1, 4-6). However, interpretation of the results differed with regard to subspecific designations; some authors concluded that the level of genetic differentiation between coastal and interior continental wolves constitutes a distinct coastal subspecies, C. l. ligoni (Weckworth et al. 2005, pp. 924, 927; Munoz-Fuentes et al. 2009, p. 12; Weckworth et al. 2010, p. 372; Weckworth et al. 2011, p. 6), while other authors asserted that it does not necessitate subspecies status (Cronin et al. 2015, p. 9). Therefore, the subspecific identity, if any, of wolves in southeastern Alaska and coastal British Columbia remained unresolved. As a cautionary note, the inference of these genetic studies depends on the type of genetic marker used and the spatial and temporal extent of the samples analyzed; we review these studies and their key findings as they relate to wolf taxonomy in detail in the Status Assessment (Service 2015, “Genetic analyses”).

In the most recent meta-analysis of wolf taxonomy in North America, Chambers et al. (2012, pp. 40-42) found evidence for differentiating between coastal and inland wolves, although ultimately the authors grouped wolves in southeastern Alaska and coastal British Columbia with wolf populations that historically occupied the central and western United States (C. l. nubilus). One of their primary reasons for doing so was because coastal wolves harbored genetic material that also was found only in historical samples of C. l. nubilus (Chambers et al. 2012, p. 41), suggesting that prior to extirpation of wolves by humans in the western United States, C. l. nubilus extended northward into coastal British Columbia and southeastern Alaska. However, this study was conducted at a broad spatial scale with a focus on evaluating taxonomy of wolves in the eastern and northeastern United States and therefore was not aimed specifically at addressing the taxonomic status of coastal wolves in western North America. Further, Chambers et al. (2012, p. 41) recognized that understanding the phylogenetic relationship of coastal wolves to other wolf populations assigned as C. l. nubilus is greatly impeded by the extirpation of wolves (and the lack of historical specimens) in the western United States. Lastly, Chambers et al. (2012, p. 2) explicitly noted that their views on subspecific designations were not intended as recommendations for management units or objects of management actions, nor should they be preferred to alternative legal classifications for protection, such as those made under the Act. Instead, the authors stated that the suitability of a subspecies as a unit for legal purposes requires further, separate analysis weighing legal and policy considerations.

We acknowledge that the taxonomic status of wolves in southeastern Alaska and coastal British Columbia is unresolved and that our knowledge of wolf taxonomy in general is evolving as more sophisticated and powerful tools become available (Service 2015, “Uncertainty in taxonomic status”). Nonetheless, based on our review of the best available information, we found persuasive evidence suggesting that wolves in southeastern Alaska and coastal British Columbia currently form an ecological and genetic unit worthy of analysis under the Act. Although zones of intergradation exist, contemporary gene flow between coastal and interior continental wolves appears to be low (e.g., Weckworth et al. 2005, p. 923; Cronin et al. 2015, p. 8), likely due to physical barriers, but perhaps also related to ecological differences (Munoz-Fuentes et al. 2009, p. 6); moreover, coastal wolves currently represent a distinct portion of genetic diversity for all wolves in North America (Weckworth et al. 2010, p. 363; Weckworth et al. 2011, pp. 5-6). Thus, we conclude that at most, wolves in southeastern Alaska and coastal British Columbia are a distinct subspecies, C. l. ligoni, of gray wolf, and at least, are a remnant population of C. l. nubilus. For the purpose of this 12-month finding, we assume that the Alexander Archipelago wolf (C. l. ligoni) is a valid subspecies of gray wolf that occupies southeastern Alaska and coastal British Columbia and, therefore, is a listable entity under the Act.

The Alexander Archipelago wolf has been described as being darker and smaller, with coarser and shorter hair, compared to interior continental gray wolves (Goldman 1937, pp. 39-40; Wood 1990, p. 1), although a comprehensive study or examination has not been completed. Like most gray wolves, fur coloration of Alexander Archipelago wolves varies considerably from pure white to uniform black, with most wolves having a brindled mix of gray or tan with brown, black, or white. Based on harvest records and wolf sightings, the black color phase appears to be more common on the mainland of southeastern Alaska and coastal British Columbia (20-30 percent) (Alaska Department of Fish and Game [ADFG] 2012, pp. 5, 18, 24; Darimont and Paquet 2000, p. 17) compared to the southern islands of the Alexander Archipelago (2 percent) (ADFG 2012, p. 34), and some of the gray-colored wolves have a brownish-red tinge (Darimont and Paquet 2000, p. 17). The variation in color phase of Alexander Archipelago wolves is consistent with the level of variation observed in other gray wolf populations (e.g., Central Brooks Range, Alaska) (Adams et al. 2008, p. 170).

Alexander Archipelago wolves older than 6 months weigh between 49 and 115 pounds (22 and 52 kilograms), with males averaging 83 pounds (38 kilograms) and females averaging 69 pounds (31 kilograms) (British Columbia Ministry of Forests, Lands and Natural Resource Operations [BCMO] 2014, p. 3; Valkenburg 2015, p. 1). On some islands in the archipelago (e.g., POW) wolves are smaller on average compared to those on the mainland, although these differences are not statistically significant (Valkenburg 2015, p. 1) (also see Service 2015, “Physical description”). The range and mean weights of Alexander Archipelago wolves are comparable to those of other populations of gray wolves that feed primarily on deer (Odocoileus spp.; e.g., northwestern Minnesota) (Mech and Paul 2008, p. 935), but are lower than those of adjacent gray wolf populations that regularly feed on larger ungulates such as moose (Alces americanus) (e.g., Adams et al. 2008, p. 8).

The Alexander Archipelago wolf currently occurs along the mainland of southeastern Alaska and coastal British Columbia and on several island complexes, which comprise more than 22,000 islands of varying size, west of the Coast Mountain Range. Wolves are found on all of the larger islands except Admiralty, Baranof, and Chichagof islands and all of the Haida Gwaii, or Queen Charlotte Islands (see Figure 1, below) (Person et al. 1996, p. 1; BCMO 2014, p. 14). The range of the Alexander Archipelago wolf is approximately 84,595 square miles (mi2) (219,100 square kilometers [km2]), stretching roughly 932 mi (1,500 km) in length and 155 mi (250 km) in width, although the northern, eastern, and southern boundaries are porous and are not defined sharply.

The majority (67 percent) of the range of the Alexander Archipelago wolf falls within coastal British Columbia, where wolves occupy all or portions of four management “regions.” These include Region 1 (entire), Region 2 (83 percent of entire region), Region 5 (22 percent of entire region), and Region 6 (17 percent of entire region) (see Figure 1, below). Thirty-three percent of the range of the Alexander Archipelago wolf lies within southeastern Alaska where it occurs in all of GMUs 1, 2, 3, and 5, but not GMU 4. See the Status Assessment (Service 2015, “Geographic scope”) for a more detailed explanation on delineation of the range.

The historical range of the Alexander Archipelago wolf, since the late Pleistocene period when the last glacial ice sheets retreated, was similar to the current range with one minor exception. Between 1950 and 1970, wolves on Vancouver Island likely were extirpated by humans (Munoz-Fuentes et al. 2010, pp. 547-548; Chambers et al. 2012, p. 41); recolonization of the island by wolves from mainland British Columbia occurred naturally and wolves currently occupy Vancouver Island.

In southeastern Alaska and coastal British Columbia, the landscape is dominated by coniferous temperate rainforests, interspersed with other habitat types such as sphagnum bogs, sedge-dominated fens, alpine areas, and numerous lakes, rivers, and estuaries. The topography is rugged with numerous deep, glacially-carved fjords and several major river systems, some of which penetrate the Coast Mountain Range, connecting southeastern Alaska and coastal British Columbia with interior British Columbia and Yukon Territory. These corridors serve as intergradation zones of variable width with interior continental wolves; outside of them, glaciers and ice fields dominate the higher elevations, separating the coastal forests from the adjacent inland forest in continental Canada.

Within the range of the Alexander Archipelago wolf, land stewardship largely lies with State, provincial, and Federal governments. In southeastern Alaska, the majority (76 percent) of the land is located within the Tongass National Forest and is managed by the USFS. The National Park Service manages 12 percent of the land, most of which is within Glacier Bay National Park. The remainder of the land in southeastern Alaska is managed or owned by the State of Alaska (4 percent), Native Corporations (3 percent), and other types of ownership (e.g., private, municipal, tribal reservation; 5 percent). In British Columbia (entire), most (94 percent) of the land and forest are owned by the Province of British Columbia (i.e., Crown lands), 4 percent is privately owned, 1 percent is owned by the federal government, and the remaining 1 percent is owned by First Nations and others (British Columbia Ministry of Forests, Mines, and Lands 2010, p. 121).

In this section, we briefly describe vital rates and population dynamics, including population connectivity, of the Alexander Archipelago wolf. For this 12-month finding, we considered a population to be a collection of individuals of a species in a defined area; the individuals in a population may or may not breed with other groups of that species in other places (Mills 2013, p. 3). We delineated wolves into populations based on GMUs in southeastern Alaska and Regions in British Columbia (coastal portions only) because these are defined areas and wolf populations are managed at these spatial scales (see Figure 1). For example, GMU 2 comprises one population of wolves on POW and adjacent islands.

Using the most recent and best available information, we estimate a current, rangewide population of 850-2,700 Alexander Archipelago wolves. The majority (roughly 62 percent) occurs in coastal British Columbia with approximately 200-650 wolves in the southern portion (Regions 1 and 2; about 24 percent of rangewide population) and 300-1,050 wolves in the northern portion (Regions 5 and 6; about 38 percent of rangewide population) (see Figure 1). In southeastern Alaska, we estimate that currently the mainland (GMUs 1 and 5A) contains 150-450 wolves (about 18 percent of rangewide population), the islands in the middle portion of the area (GMU 3) contain 150-350 wolves (about 14 percent of rangewide population), and the southwestern set of islands (GMU 2) has 50-159 wolves (95 percent confidence intervals [CI], mean = 89 wolves; about 6 percent of rangewide population) (Person et al. 1996, p. 13; ADFG 2015a, p. 2). Our estimates are based on a variety of direct and indirect methods with the only empirical estimate available for GMU 2, which comprises POW and surrounding islands. See the Status Assessment (Service 2015, “Abundance and density”) for details on derivation, assumptions, and caveats.

Similar to abundance, direct estimates of population trend of the Alexander Archipelago wolf are available only for GMU 2 in southeastern Alaska. In this GMU, fall population size has been estimated on four occasions (1994, 2003, 2013, and 2014). Between 1994 and 2014, the population was reduced from 356 wolves (95 percent CI = 148-564) (Person et al. 1996, pp. 11-12; ADFG 2014, pp. 2-4) to 89 wolves (95 percent CI = 50-159) (ADFG 2015a, pp. 1-2), equating to an apparent decline of 75 percent (standard error [SE] = 15), or 6.7 percent (SE = 2.8) annually. Although the numerical change in population size over the 20-year period is notable, the confidence intervals of the individual point estimates overlap. The most severe reduction occurred over a single year (2013-2014), when the population dropped by 60 percent and the proportion of females in the sample was reduced from 0.57 (SE = 0.13) to 0.25 (SE = 0.11) (ADFG 2015a, p. 2). In the remainder of southeastern Alaska, the trend of wolf populations is not known.

In British Columbia, regional estimates of wolf population abundance are generated regularly using indices of ungulate biomass, and, based on these data, the provincial wolf population as a whole has been stable or slightly increasing since 2000 (Kuzyk and Hatter 2014, p. 881). In Regions 1, 2, 5, and 6, where the Alexander Archipelago wolf occurs in all or a portion of each of these regions (see Distribution and Range, above), the same trend has been observed (BCMO 2015a, p. 1). Because estimates of population trend are not specific to the coastal portions of these regions only, we make the necessary scientific assumption that the trend reported for the entire region is reflective of the trend in the coastal portion of the region. This assumption applies only to Regions 5 and 6, where small portions (22 and 17 percent, respectively) of the region fall within the range of the Alexander Archipelago wolf; all of Region 1 and nearly all (83 percent) of Region 2 are within the range of the coastal wolf (see Figure 1). Thus, based on the best available information, we found that the wolf populations in coastal British Columbia have been stable or slightly increasing over the last 15 years. See the Status Assessment (Service 2015, “Abundance and density”) for a more thorough description of data assumptions and caveats.

Similar to the gray wolf, sizes of litters of the Alexander Archipelago wolf can vary substantially (1-8 pups, mean = 4.1) with inexperienced breeding females producing fewer pups than older, more experienced mothers (Person and Russell 2009, p. 216). Although uncommon, some packs fail to exhibit denning behavior or produce litters in a given year, and no pack has been observed with multiple litters (Person and Russell 2009, p. 216). Age of first breeding of the Alexander Archipelago wolf is about 22 to 34 months (Person et al. 1996, p. 8).

We found only one study that estimated survival rates of Alexander Archipelago wolves. Based on radio-collared wolves in GMU 2 between 1994 and 2004, Person and Russell (2008, p. 1545) reported mean annual survival rate of wolves greater than 4 months old as 0.54 (SE = 0.17); survival did not differ between age classes or sexes, but was higher for resident wolves (0.65, SE = 0.17) compared to nonresidents (i.e., wolves not associated with a pack; 0.34, SE = 0.17). Average annual rates of mortality attributed to legal harvest, unreported harvest, and natural mortality were 0.23 (SE = 0.12), 0.19 (SE = 0.11), and 0.04 (SE = 0.05), respectively, and these rates were correlated positively with roads and other landscape features that created openings in the forest (Person and Russell 2008, pp. 1545-1546).

In 2012, another study was initiated (and is ongoing) in GMU 2 that involves collaring wolves, but too few animals have been collared so far to estimate annual survival reliably (n = 12 wolves between 2012 and May 2015). Nonetheless, of those 12 animals, 5 died from legal harvest, 3 from unreported harvest, and 1 from natural causes; additionally, the fate of 2 wolves is unknown and 1 wolf is alive still (ADFG 2015b, p. 4). Thus, overall, harvest of Alexander Archipelago wolves by humans has accounted for most of the mortality of collared wolves in GMU 2. Our review of the best available information did not reveal any estimates of annual survival or mortality of wolves on other islands or the mainland of southeastern Alaska and coastal British Columbia.

Similar to gray wolves, Alexander Archipelago wolves either remain in their natal pack or disperse (Person et al. 1996, p. 10), here defined as permanent movement of an individual away from its pack of origin. Dispersers typically search for a new pack to join or associate with other wolves and ultimately form a new pack in vacant territories or in vacant areas adjacent to established territories. Dispersal can occur within or across populations; when it occurs across populations, then population connectivity is achieved. Both dispersal and connectivity contribute significantly to the health of individual populations as well as the taxon as a whole.

Dispersal rates of the Alexander Archipelago wolf are available only for GMU 2, where the annual rate of dispersal of radio-collared wolves was 39 percent (95 percent CI = 23 percent, n = 18) with adults greater than 2 years of age composing 79 percent of all dispersers (Person and Ingle 1995, p. 20). Minimum dispersal distances from the point of capture and radio-collaring ranged between 8 and 113 mi (13 and 182 km); all dispersing wolves remained in GMU 2 (Person and Ingle 1995, p. 23). Successful dispersal of individuals tends to be short in duration and distance in part because survival of dispersing wolves is low (annual survival rate = 0.16) (e.g., Peterson et al. 1984, p. 29; Person and Russell 2008, p. 1547).

Owing to the rugged terrain and island geography across most of southeastern Alaska and coastal British Columbia, population connectivity probably is more limited for the Alexander Archipelago wolf compared to the gray wolf that inhabits interior continental North America. Of the 67 Alexander Archipelago wolves radio-collared in GMU 2, none emigrated to a different GMU (Person and Ingle 1995, p. 23; ADFG 2015c, p. 2); similarly, none of the four wolves collared in northern southeastern Alaska (GMU 1C and 1D) attempted long-distance dispersal, although the home ranges of these wolves were comparatively large (ADFG 2015c, p. 2). Yet, of the three wolves opportunistically radio-collared on Kupreanof Island (GMU 3), one dispersed to Revillagigedo Island (GMU 1A) (USFS 2015, p. 1), an event that required at least four water crossings with the shortest being about 1.2 mi (2.0 km) in length (see Figure 1). Thus, based on movements of radio-collared wolves, demographic connectivity appears to be more restricted for some populations than others; however, few data exist outside of GMU 2, where the lack of emigration is well documented but little is known about the rate of immigration.

Likewise, we found evidence suggesting that varying degrees of genetic connectivity exist across populations of the Alexander Archipelago wolf, indicating that some populations are more insular than others. Generally, of the populations sampled, gene flow was most restricted to and from the GMU 2 wolf population (Weckworth et al. 2005, p. 923; Breed 2007, p. 19; Cronin et al. 2015, Supplemental Table 3), although this population does not appear to be completely isolated. Breed (2007, pp. 22-23) classified most wolves in northern coastal British Columbia (Regions 5 and 6) as residents and more than half of the wolves in the southern portion of southeastern Alaska (GMUs 1A and 2) as migrants of mixed ancestry. Further, the frequency of private alleles (based on nuclear DNA) in the GMU 2 wolf population is low relative to other Alexander Archipelago wolves (Weckworth et al. 2005, p. 921; Breed 2007, p. 18), and the population does not harbor unique haplotypes (based on mitochondrial DNA), both of which suggest that complete isolation has not occurred. Thus, although some genetic discontinuities of Alexander Archipelago wolves is evident, likely due to geographical disruptions to dispersal and gene flow, genetic connectivity among populations seems to be intact, albeit at low levels for some populations (e.g., GMU 2). The scope of inference of these genetic studies depends on the type of genetic marker used and the spatial and temporal extent of the samples analyzed; we review key aspects of these studies in more detail in the Status Assessment (Service 2015, “Genetic analyses,” “Genetic connectivity”).

Collectively, the best available information suggests that demographic and genetic connectivity among Alexander Archipelago wolf populations exists, but at low levels for some populations such as that of GMU 2, likely due to geographical disruptions to dispersal and gene flow. Based on the range of samples used by Breed (2007, pp. 21-23), gene flow to GMU 2 appears to be uni-directional, which is consistent with the movement data from wolves radio-collared in GMU 2 that demonstrated no emigration from that population (ADFG 2015c, p. 2). These findings, coupled with the trend of the GMU 2 wolf population (see “Abundance and Trend,” above), suggest that this population may serve as a sink population of the Alexander Archipelago wolf; conversely, the northern coastal British Columbian population may be a source population to southern southeastern Alaska, as suggested by Breed (2007, p. 34). This hypothesis is supported further with genetic information indicating a low frequency of private alleles and no unique haplotypes in the wolves occupying GMU 2. Nonetheless, we recognize that persistence of this population may be dependent on the health of adjacent populations (e.g., GMU 3), but conclude that its demographic and genetic contribution to the rangewide population likely is lower than other populations such as those in coastal British Columbia.

In this section, we briefly describe the ecology, including food habits, social organization, and space and habitat use, of the Alexander Archipelago wolf. Again, we review each of these topics in more detail in the Status Assessment (Service 2015, entire).

Similar to gray wolves, Alexander Archipelago wolves are opportunistic predators that eat a variety of prey species, although ungulates compose most of their overall diet. Based on scat and stable isotope analyses, black-tailed deer (Odocoileus hemionus), moose, mountain goat (Oreamnos americanus), and elk (Cervus spp.), either individually or in combination, constitute at least half of the wolf diet across southeastern Alaska and coastal British Columbia (Fox and Streveler 1986, pp. 192-193; Smith et al. 1987, pp. 9-11, 16; Milne et al. 1989, pp. 83-85; Kohira and Rexstad 1997, pp. 429-430; Szepanski et al. 1999, p. 331; Darimont et al. 2004, p. 1871; Darimont et al. 2009, p. 130; Lafferty et al. 2014, p. 145). Other prey species regularly consumed, depending on availability, include American beaver (Castor canadensis), hoary marmot (Marmota caligata), mustelid species (Mustelidae spp.), salmon (Oncorhynchus spp.), and marine mammals (summarized more fully in the Status Assessment, Service 2015, “Food habits”).

Prey composition in the diet of the Alexander Archipelago wolf varies across space and time, usually reflecting availability on the landscape, especially for ungulate species that are not uniformly distributed across the islands and mainland. For instance, mountain goats are restricted to the mainland and Revillagigedo Island (introduced). Similarly, moose occur along the mainland and nearby islands as well as most of the islands in GMU 3 (e.g., Kuiu, Kupreanof, Mitkof, and Zarembo islands); moose distribution is expanding in southeastern Alaska and coastal British Columbia (Darimont et al. 2005, p. 235; Hundertmark et al. 2006, p. 331). Elk also occur only on some islands in southeastern Alaska (e.g., Etolin Island) and on Vancouver Island. Deer are the only ungulate distributed throughout the range of the Alexander Archipelago wolf, although abundance varies greatly with snow conditions. Generally, deer are abundant in southern coastal British Columbia, where the climate is mild, with their numbers decreasing northward along the mainland due to increasing snow depths, although they typically occur in high densities on islands such as POW, where persistent and deep snow accumulation is less common.

Owing to the disparate patterns of ungulate distribution and abundance, some Alexander Archipelago wolf populations have a more restricted diet than others. For example, in GMU 2, deer is the only ungulate species available to wolves, but elsewhere moose, mountain goat, elk, or a combination of these ungulates are available. Szepanski et al. (1999, pp. 330-331) demonstrated that deer and salmon contributed equally to the diet of wolves on POW (GMU 2), Kupreanof Island (GMU 3), and the mainland (GMUs 1A and 1B) (deer = 45-49 percent and salmon = 15-20 percent), and that “other herbivores” composed the remainder of the diet (34-36 percent). On POW, “other herbivores” included only beaver and voles (Microtus spp.), but on Kupreanof Island, moose also was included, and on the mainland, mountain goat was added to the other two herbivore prey species. Therefore, we hypothesize that wolves in GMU 2, and to a lesser extent in parts of GMU 3, are more vulnerable to changes in deer abundance compared to other wolf populations that have a more diverse ungulate prey base available to them.

Given the differences in prey availability throughout the range of the Alexander Archipelago wolf, some general patterns in their food habits exist. On the northern mainland of southeastern Alaska, where deer occur in low densities, wolves primarily eat moose and mountain goat (Fox and Streveler 1986, pp. 192-193; Lafferty et al. 2014, p. 145). As one moves farther south and deer become more abundant, they are increasingly represented in the diet, along with correspondingly smaller proportions of moose and mountain goat where available (Szepanski et al. 1999, p. 331; Darimont et al. 2004, p. 1869). On the outer islands of coastal British Columbia, marine mammals compose a larger portion of the diet compared to other parts of the range of the Alexander Archipelago wolf (Darimont et al. 2009, p. 130); salmon appear to be eaten regularly by coastal wolves in low proportions (less than 20 percent), although some variation among populations exists. Generally, the diet of wolves in coastal British Columbia appears to be more diverse than in southeastern Alaska (e.g., Kohira and Rexstad 1997, pp. 429-430; Darimont et al. 2004, pp. 1869, 1871), consistent with a more diverse prey base in the southern portion of the range of the Alexander Archipelago wolf. We review these diet studies and others in the Status Assessment (Service 2015, “Food habits”).

One of the apparently unusual aspects of the Alexander Archipelago wolf diet is consumption of marine-derived foods. However, we found evidence suggesting that this behavior is not uncommon for gray wolves in coastal areas or those that have inland access to marine prey (e.g., spawning salmon). For example, wolves on the Alaska Peninsula in western Alaska have been observed catching and eating sea otters (Enhydra lutris), using offshore winter sea ice as a hunting platform and feeding on marine mammal carcasses such as Pacific walrus (Odobenus rosmarus divergens) and beluga whale (Delphinapterus leucas) (Watts et al. 2010, pp. 146-147). In addition, Adams et al. (2010, p. 251) found that inland wolves in Denali National Park, Alaska, ate salmon in slightly lower but similar quantities (3-17 percent of lifetime diet) compared to Alexander Archipelago wolves (15-20 percent of lifetime diet; Szepanski et al. 1999, p. 327). These findings and others suggest that marine-derived resources are not a distinct component of the diet of the Alexander Archipelago wolf. Nonetheless, marine prey provide alternate food resources to coastal wolves during periods of the year with high food and energy demands (e.g., provisioning of pups when salmon are spawning; Darimont et al. 2008, pp. 5, 7-8) and when and where abundance of terrestrial prey is low.

Wolves are social animals that live in packs usually composed of one breeding pair (i.e., alpha male and female) plus offspring of 1 to 2 years old. The pack is a year-round unit, although all members of a wolf pack rarely are observed together except during winter (Person et al. 1996, p. 7). Loss of alpha members of a pack can result in social disruption and unstable pack dynamics, which are complex and shift frequently as individuals age and gain dominance, disperse from, establish or join existing packs, breed, and die (Mech 1999, pp. 1197-1202). Although loss of breeding individuals impacts social stability within the pack, at the population level wolves appear to be resilient enough to compensate for any negative impacts to population growth (Borg et al. 2015, p. 183).

Pack sizes of the Alexander Archipelago wolf are difficult to estimate owing to the heavy vegetative cover throughout most of its range. In southeastern Alaska, packs range from one to 16 wolves, but usually average 7 to 9 wolves with larger packs observed in fall than in spring (Smith et al. 1987, pp. 4-7; Person et al. 1996, p. 7; ADFG 2015c, p. 2). Our review of the best available information did not reveal information on pack sizes from coastal British Columbia.

Similar to gray wolves in North America, the Alexander Archipelago wolf uses a variety of habitat types and is considered a habitat generalist (Person and Ingle 1995, p. 30; Mech and Boitani 2003, p. xv). Person (2001, pp. 62-63) reported that radiocollared Alexander Archipelago wolves spent most of their time at low elevation during all seasons (95 percent of locations were below 1,312 feet [ft] [400 m] in elevation), but did not select for or against any habitat types except during the pup-rearing season. During the pup-rearing season, radiocollared wolves selected for open- and closed-canopy old-growth forests close to lakes and streams and avoided clearcuts and roads (Person 2001, p. 62), a selection pattern that is consistent with den site characteristics.

Alexander Archipelago wolves den in root wads of large living or dead trees in low-elevation, old-growth forests near freshwater and away from logged stands and roads, when possible (Darimont and Paquet 2000, pp. 17-18; Person and Russell 2009, pp. 211, 217, 220). Of 25 wolf dens monitored in GMU 2, the majority (67 percent) were located adjacent to ponds or streams with active beaver colonies (Person and Russell 2009, p. 216). Although active dens have been located near clearcuts and roads, researchers postulate that those dens probably were used because suitable alternatives were not available (Person and Russell 2009, p. 220).

Home range sizes of Alexander Archipelago wolves are variable depending on season and geographic location. Generally, home ranges are about 50 percent smaller during denning and pup-rearing periods compared to other times of year (Person 2001, p. 55), and are roughly four times larger on the mainland compared to the islands in southeastern Alaska (ADFG 2015c, p. 2). Person (2001, pp. 66, 84) found correlations between home range size, pack size, and the proportion of “critical winter deer habitat”; he thought that the relation between these three factors was indicative of a longer-term influence of habitat on deer density. We review space and habitat use of Alexander Archipelago wolf and Sitka black-tailed deer, the primary prey item consumed by wolves throughout most of their range, in detail in the Status Assessment (Service 2015, “Space and habitat use”).

In summary, we find that the Alexander Archipelago wolf currently is distributed throughout most of southeastern Alaska and coastal British Columbia with a rangewide population estimate of 850-2,700 wolves. The majority of the range (67 percent) and the rangewide population (approximately 62 percent) occur in coastal British Columbia, where the population is stable or increasing. In southeastern Alaska, we found trend information only for the GMU 2 population (approximately 6 percent of the rangewide population) that indicates a decline of about 75 (SE = 15) percent since 1994, although variation around the point estimates (n = 4) was substantial. This apparent decline is consistent with low estimates of annual survival of wolves in GMU 2, with the primary source of mortality being harvest by humans. For the remainder of southeastern Alaska (about 32 percent of the rangewide population), trends of wolf populations are not known.

Similar to the continental gray wolf, the Alexander Archipelago wolf has several life-history and ecological traits that contribute to its resiliency, or its ability to withstand stochastic disturbance events. These traits include high reproductive potential, ability to disperse long distances (over 100 km), use of a variety of habitats, and a diverse diet including terrestrial and marine prey. However, some of these traits are affected by the island geography and rugged terrain of most of southeastern Alaska and coastal British Columbia. Most notably, we found that demographic and genetic connectivity of some populations, specifically the GMU 2 population, is low, probably due to geographical disruptions to dispersal and gene flow. In addition, not all prey species occur throughout the range of the Alexander Archipelago wolf, and, therefore, some populations have a more limited diet than others despite the opportunistic food habits of wolves. Specifically, the GMU 2 wolf population is vulnerable to fluctuations in abundance of deer, the only ungulate species that occupies the area. We postulate that the insularity of this population, coupled with its reliance on one ungulate prey species, likely has contributed to its apparent recent decline, suggesting that, under current conditions, the traits associated with resiliency may not be sufficient for population stability in GMU 2.

Section 4 of the Act (16 U.S.C. 1533) and implementing regulations (50 CFR 424) set forth procedures for adding species to, removing species from, or reclassifying species on the Federal Lists of Endangered and Threatened Wildlife and Plants. Under section 4(a)(1) of the Act, a species may be determined to be endangered or threatened based on any of the following five factors:

(A) The present or threatened destruction, modification, or curtailment of its habitat or range;

(B) Overutilization for commercial, recreational, scientific, or educational purposes;

(C) Disease or predation;

(D) The inadequacy of existing regulatory mechanisms; or

(E) Other natural or manmade factors affecting its continued existence.

In making this finding, information pertaining to the Alexander Archipelago wolf in relation to the five factors provided in section 4(a)(1) of the Act is discussed below. In considering what factors might constitute threats, we must look beyond the mere exposure of the species to the factor to determine whether the species responds to the factor in a way that causes actual impacts to the species. If there is exposure to a factor, but no response, or only a positive response, that factor is not a threat. If there is exposure and the species responds negatively, the factor may be a threat; we then attempt to determine if that factor rises to the level of a threat, meaning that it may drive or contribute to the risk of extinction of the species such that the species warrants listing as an endangered or threatened species as those terms are defined by the Act. This does not necessarily require empirical proof of a threat. The combination of exposure and some corroborating evidence of how the species is likely impacted could suffice. The mere identification of factors that could impact a species negatively is not sufficient to compel a finding that listing is appropriate, however; we require evidence that these factors are operative threats that act on the species to the point that the species meets the definition of an endangered or threatened species under the Act.

In making our 12-month finding on the petition we considered and evaluated the best available scientific and commercial information.

The Alexander Archipelago wolf uses a variety of habitats and, like other gray wolves, is considered to be a habitat generalist. Further, it is an opportunistic predator that eats ungulates, rodents, mustelids, fish, and marine mammals, typically killing live prey, but also feeding on carrion if fresh meat is not available or circumstances are desirable (e.g., large whale carcass). For these reasons and others (e.g., dispersal capability), we found that wolf populations often are resilient to changes in their habitat and prey. Nonetheless, we also recognize that the Alexander Archipelago wolf inhabits a distinct ecosystem, partially composed of island complexes, that may restrict wolf movement and prey availability of some populations, thereby increasing their vulnerability to changes in habitat.

In this section, we review stressors to terrestrial and intertidal habitats used by the Alexander Archipelago wolf and its primary prey, specifically deer. We identified timber harvest as the principal stressor modifying wolf and deer habitat in southeastern Alaska and coastal British Columbia, and, therefore, we focus our assessment on this stressor by evaluating possible direct and indirect impacts to the wolf at the population and rangewide levels. We also consider possible effects of road development, oil development, and climate-related events on wolf habitat. We describe the information presented here in more detail in the Status Assessment (Service 2015, “Cause and effect analysis”).

Throughout most of the range of the Alexander Archipelago wolf, timber harvest has altered forested habitats, especially those at low elevations, that are used by wolves and their prey. Rangewide, we estimate that 19 percent of the productive old-growth forest has been logged, although it has not occurred uniformly across the landscape or over time. A higher percentage of productive old-growth forest has been logged in coastal British Columbia (24 percent) compared to southeastern Alaska (13 percent), although in both areas, most of the harvest has occurred since 1975 (85 percent and 66 percent, respectively). Within coastal British Columbia, the majority of harvest (66 percent of total harvest) has happened in Region 1, where 34 percent of the forest has been logged; in the coastal portions of Regions 2, 5, and 6, timber harvest has been comparatively lower, ranging from 12 to 17 percent of the productive forest in these regions. Similarly, in southeastern Alaska, logging has occurred disproportionately in GMU 2, where 23 percent of the forest has been logged (47 percent of all timber harvest in southeastern Alaska); in other GMUs, only 6 to 14 percent of the forest has been harvested. We discuss spatial and temporal patterns of timber harvest in more detail in the Status Assessment (Service 2015, “Timber harvest”).

Owing to past timber harvest in southeastern Alaska and coastal British Columbia, portions of the landscape currently are undergoing succession and will continue to do so. Depending on site-specific conditions, it can take up to several hundred years for harvested stands to regain old-growth forest characteristics fully (Alaback 1982, p. 1939). During the intervening period, these young-growth stands undergo several successional stages that are relevant to herbivores such as deer. Briefly, for 10 to 15 years following clearcut logging, shrub and herb biomass production increases (Alaback 1982, p. 1941), providing short-term benefits to herbivores such as deer, which select for these stands under certain conditions (e.g., Gilbert 2015, p. 129). After 25 to 35 years, early seral stage plants give way to young-growth coniferous trees, and their canopies begin to close, intercepting sunlight and eliminating most understory vegetation. These young-growth stands offer little nutritional browse for deer and therefore tend to be selected against by deer (e.g., Gilbert 2015, pp. 129-130); this stage typically lasts for at least 50 to 60 years, at which point the understory layer begins to develop again (Alaback 1982, pp. 1938-1939). An understory of deciduous shrubs and herbs, similar to pre-harvest conditions, is re-established 140 to 160 years after harvest. Alternative young-growth treatments (e.g., thinning, pruning) are used to stimulate understory growth, but they often are applied at small spatial scales, and their efficacy in terms of deer use is unknown; regardless, to date, over 232 mi2 (600 km2) of young-growth has been treated in southeastern Alaska (summarized in Service 2015, “Timber harvest”).

We expect timber harvesting to continue to occur throughout the range of the Alexander Archipelago wolf, although given current and predicted market conditions, the rate of future harvest is difficult to project. In southeastern Alaska, primarily in GMUs 2 and 3, some timber has been sold by the USFS already, but has not yet been cut. In addition, new timber sales currently are being planned for sale between 2015 and 2019, and most of this timber is expected to be sourced from GMUs 2 and 3; however, based on recent sales, it is unlikely that the planned harvest will be implemented fully due to lack of bidders. Also, we anticipate at least partial harvest of approximately 277 km2 of land in GMU 2 that was transferred recently from the Tongass National Forest to Sealaska Native Corporation. In coastal British Columbia, we estimate that an additional 17 percent of forest will be harvested by 2100 on Vancouver Island (Region 1) and an additional 39 percent on the mainland of coastal British Columbia; however, some of this timber volume would be harvested from old young-growth stands. See the Status Assessment for more details (Service 2015, “Future timber harvest”).

Since 2013, the USFS has been developing a plan to transition timber harvest away from primarily logging old-growth and toward logging young-growth stands, although small amounts of old-growth likely will continue to be logged. An amendment to the current Tongass Land and Resource Management Plan is underway and is expected to be completed by the end of 2016. Although this transition is expected to reduce further modification of habitat used by wolves and deer, the amendment that outlines the transition is still in the planning phase.

After reviewing the best available information, we determined that the only potential direct effect from timber harvest to Alexander Archipelago wolves is the modification of and disturbance at den sites. Although coastal wolves avoided using den sites located in or near logged stands, other landscape features such as gentle slope, low elevation, and proximity to freshwater had greater influence on den site use (Person and Russell 2009, pp. 217-219). Further, our review of the best available information did not indicate that denning near logged stands had fitness consequences to individual wolves or that wolf packs inhabiting territories with intensive timber harvest were less likely to breed due to reduced availability of denning habitat. Therefore, we conclude that modification of and disturbance at den sites as a result of timber harvest does not constitute a threat to the Alexander Archipelago wolf at the population or rangewide level.

We then examined reduction in prey availability, specifically deer, as a potential indirect effect of timber harvest to the Alexander Archipelago wolf. Because deer selectively use habitats that minimize accumulation of deep snow in winter, including productive old-growth forest (e.g., Schoen and Kirchhoff, 1990, p. 374; Doerr et al. 2005, p. 322; Gilbert 2015, p. 129), populations of deer in areas of intensive timber harvest are expected to decline in the future as a result of long-term reduction in the carrying capacity of their winter habitat (e.g., Person 2001, p. 79; Gilbert et al. 2015, pp. 18-19). However, we found that most populations of Alexander Archipelago wolf likely will be resilient to predicted declines in deer abundance largely owing to their ability to feed on alternate ungulate prey species and non-ungulate species, including those that occur in intertidal and marine habitats (greater than 15 percent of the diet; see “Food Habits,” above) (Szepanski et al. 1999, p. 331; Darimont et al. 2004, p. 1871, Darimont et al. 2009, p. 130). Moreover, in our review of the best available information, we found nothing to suggest that these intertidal and marine species, non-ungulate prey, and other ungulate species within the range of the Alexander Archipelago wolf (i.e., moose, goat, elk) are affected significantly by timber harvest (Service 2015, “Response of wolves to timber harvest”). Therefore, we focus the remainder of this section on predicted response of wolves to reduction in deer numbers as a result of timber harvest and availability of alternate ungulate prey.

In coastal British Columbia, where a greater proportion of productive old-growth forest has been harvested compared to southeastern Alaska, deer populations are stable (Regions 1, 2, and 5) or decreasing (Region 6) (BCMO 2015b, p. 1). Yet, corresponding wolf populations at the regional scale are stable or slightly increasing (Kuzyk and Hatter 2014, p. 881; BCMO 2015a, p. 1). We attribute the stability in wolf numbers, in part, to the availability of other ungulate species, specifically moose, mountain goat, and elk (Region 1 only), which primarily have stable populations and do not use habitats affected by timber harvest. Therefore, we presume that these wolf populations have adequate prey available and are not being affected significantly by changes in deer abundance as a result of timber harvest.

Similarly, throughout most of southeastern Alaska, wolves have access to multiple ungulate prey species in addition to deer. Along the mainland (GMUs 1 and 5A), where deer densities are low naturally, moose and mountain goats are available, and, in GMU 3, moose occur on all of the larger islands and elk inhabit Etolin and Zarembo islands. Also, although we expect deer abundance in these GMUs to be lower in the future, deer will continue to be available to wolves; between 1954 and 2002, deer habitat capability was reduced by only 15 percent in parts of GMU 1 and by 13 to 23 percent in GMU 3 (Albert and Schoen 2007, p. 16). Thus, although we lack estimates of trend in these wolf populations, we postulate that they have sufficient prey to maintain stable populations and are not being impacted by timber harvest.

Only one Alexander Archipelago wolf population, the GMU 2 population, relies solely on deer as an ungulate prey species and therefore it is more vulnerable to declines in deer numbers compared to all other populations. Additionally, timber harvest has occurred disproportionately in this area, more so than anywhere else in the range of the wolf except Vancouver Island (where the wolf population is stable). As a result, in GMU 2, deer are projected to decline by approximately 21 to 33 percent over the next 30 years, and, correspondingly, the wolf population is predicted to decline by an average of 8 to 14 percent (Gilbert et al. 2015, pp. 19, 43). Further, the GMU 2 wolf population already has been reduced by about 75 percent since 1994, although most of the apparent decline occurred over a 1-year period between 2013 and 2014 (see “Abundance and Trend,” above), suggesting that the cause of the decline was not specifically long-term reduction in deer carrying capacity, although it probably was a contributor. These findings indicate that for this wolf population, availability of non-ungulate prey does not appear to be able to compensate for declining deer populations, especially given other present stressors such as wolf harvest (see discussion under Factor B). Therefore, we conclude that timber harvest is affecting the GMU 2 wolf population by reducing its ungulate prey and likely will continue to do so in the future.

In reviewing the best available information, we conclude that indirect effects from timber harvest likely are not having and will not have a significant effect on the Alexander Archipelago wolf at the rangewide level. Although timber harvest has reduced deer carrying capacity, which in turn is expected to cause declines in deer populations, wolves are opportunistic predators, feeding on a variety of prey species, including intertidal and marine species that are not impacted by timber harvest. In addition, the majority (about 94 percent) of the rangewide wolf population has access to ungulate prey species other than deer. Further, currently the wolf populations in coastal British Columbia, which constitute 62 percent of the rangewide population, are stable or slightly increasing despite intensive and extensive timber harvest.

However, we also conclude that the GMU 2 wolf population likely is being affected and will continue to be affected by timber harvest, but that any effects will be restricted to the population level. This wolf population represents only 6 percent of the rangewide population, is largely insular and geographically peripheral to other populations, and appears to function as a sink population (see “Abundance and Trend” and “Dispersal and Connectivity,” above). For these reasons, we find that the demographic and genetic contributions of the GMU 2 wolf population to the rangewide population are low. Thus, although we expect deer and wolf populations to decline in GMU 2, in part as a result of timber harvest, we find that these declines will not result in a rangewide impact to the Alexander Archipelago wolf population.

Road development has modified the landscape throughout the range of the Alexander Archipelago wolf. Most roads were constructed to support the timber industry, although some roads were built as a result of urbanization, especially in southern coastal British Columbia. Below, we briefly describe the existing road systems in southeastern Alaska and coastal British Columbia using all types of roads (e.g., sealed, unsealed) that are accessible with any motorized vehicle (e.g., passenger vehicle, all-terrain vehicle). See the Status Assessment for a more detailed description (Service 2015, “Road construction and management”).

Across the range of the Alexander Archipelago wolf, the majority (86 percent) of roads are located in coastal British Columbia (approximately 41,943 mi [67,500 km] of roads), where mean road density is 0.76 mi per mi2 (0.47 km per km2), although road densities are notably lower in the northern part of the province (Regions 5 and 6, mean = 0.21-0.48 mi per km2 [0.13-0.30 km per km2]) compared to the southern part (Regions 1 and 2, mean = 0.85-0.89 mi per mi2 [0.53-0.55 km per km2]), largely owing to the urban areas of Vancouver and Victoria. In southeastern Alaska, nearly 6,835 mi [11,000 km] of roads exist within the range of the Alexander Archipelago wolf, resulting in a mean density of 0.37 mi per mi2 (0.23 km per km2). Most of these roads are located in GMU 2, where the mean road density is 1.00 mi per mi2 (0.62 km per km2), more than double that in all other GMUs, where the mean density ranges from 0.06 mi per mi2 (0.04 km per km2) (GMU 5A) to 0.42 mi per mi2 (0.26 km per km2) (GMU 3). Thus, most of the roads within the range of the Alexander Archipelago wolf are located in coastal British Columbia, especially in Regions 1 and 2, but the highest mean road density occurs in GMU 2 in southeastern Alaska, which is consistent with the high percentage of timber harvest in this area (see “Timber Harvest,” above). In addition, we anticipate that most future road development also will occur in GMU 2 (46 mi [74 km] of new road), with smaller additions to GMUs 1 and 3 (Service 2015, “Road construction and management”).

Given that the Alexander Archipelago wolf is a habitat generalist, we find that destruction and modification of habitat due to road development likely is not affecting wolves at the population or rangewide level. In fact, wolves occasionally use roads as travel corridors between habitat patches (Person et al. 1996, p. 22). As reviewed above in “Timber Harvest,” we recognize that wolves used den sites located farther from roads compared to unused sites; however, other landscape features were more influential in den site selection, and proximity to roads did not appear to affect reproductive success or pup survival, which is thought to be high (Person et al. 1996, p. 9; Person and Russell 2009, pp. 217-219). Therefore, we conclude that roads are not a threat to the habitats used by the Alexander Archipelago wolf, although we address the access that they afford to hunters and trappers as a potential threat to some wolf populations under Factor B.

We reviewed potential loss of habitat due to oil and gas development as a stressor to the Alexander Archipelago wolf. We found no existing oil and gas projects within the range of the coastal wolf, although two small-scale exploration projects occurred in Regions 1 and 2 of coastal British Columbia, but neither project resulted in development. In addition, we considered a proposed oil pipeline project (i.e., Northern Gateway Project) intended to transport oil from Alberta to the central coast of British Columbia, covering about 746 mi (1,200 km) in distance. If the proposed project was approved and implemented, risk of oil spills on land and on the coast within the range of the Alexander Archipelago wolf would exist. However, given its diverse diet, terrestrial habitat use, and dispersal capability, we conclude that wolf populations would not be affected by the pipeline project even if an oil spill occurred because exposure would be low. Further, oil development occurs in portions of the range of the gray wolf (e.g., Trans Alaska Pipeline System) and is not thought to be impacting wolf populations negatively. We conclude that oil development is not a threat to the Alexander Archipelago wolf now and is not likely to become one in the future.

We considered the role of climate and projected changes in climate as a potential stressor to the Alexander Archipelago wolf. We identified three possible mechanisms through which climate may be affecting habitats used by coastal wolves or their prey: (1) Frequency of severe winters and impacts to deer populations; (2) decreasing winter snow pack and impacts to yellow cedar; and (3) predicted hydrologic change and impacts to salmon productivity. We review each of these briefly here and in more detail in the Status Assessment (Service 2015, “Climate-related events”).

Severe winters with deep snow accumulation can negatively affect deer populations by reducing availability of forage and by increasing energy expenditure associated with movement. Therefore, deer selectively use habitats in winter that accumulate less snow, such as those that are at low elevation, that are south-facing, or that can intercept snowfall (i.e., dense forest canopy). Timber harvest has reduced some of these preferred winter habitats. However, while acknowledging that severe winters can result in declines of local deer populations, we postulate that those declines are unlikely to affect wolves substantially at the population or rangewide level for several reasons.

First, in southern coastal British Columbia where 24 percent of the rangewide wolf population occurs, persistent snowfall is rare except at high elevations. Second, in GMU 2, where wolves are limited to deer as ungulate prey and therefore are most vulnerable to declines in deer abundance, the climate is comparatively mild and severe winters are infrequent (Shanley et al. 2015, p. 6); Person (2001, p. 54) estimated that six winters per century may result in general declines in deer numbers in GMU 2. Lastly, climate projections indicate that precipitation as snow will decrease by up to 58 percent over the next 80 years (Shanley et al. 2015, pp. 5-6), reducing the likelihood of severe winters. Therefore, we conclude that winter severity, and associated interactions with timber harvest, is not a threat to the persistence of the Alexander Archipelago wolf at the population or rangewide level now or in the future.

In contrast to deer response to harsh winter conditions, recent and ongoing decline in yellow cedar in southeastern Alaska is attributed to warmer winters and reduced snow cover (Hennon et al. 2012, p. 156). Although not all stands are affected or affected equally, the decline has impacted about 965 mi2 (2,500 km2) of forest (Hennon et al. 2012, p. 148), or less than 3 percent of the forested habitat within the range of the Alexander Archipelago wolf. In addition, yellow cedar is a minor component of the temperate rainforest, which is dominated by Sitka spruce and western hemlock and neither of these tree species appears to be impacted negatively by reduced snow cover (e.g., Schaberg et al. 2005, p. 2065). Therefore, we conclude that any effects (positive or negative) to the wolf as a result of loss of yellow cedar would be negligible given that it constitutes a small portion of the forest and that the wolf is a habitat generalist.

Predicted hydrologic changes as a result of changes in climate are expected to reduce salmon productivity within the range of the Alexander Archipelago wolf (e.g., Edwards et al. 2013, p. 43; Shanley and Albert 2014, p. 2). Warmer winter temperatures and extreme flow events are predicted to reduce egg-to-fry survival of salmon, resulting in lower overall productivity. Although salmon compose 15 to 20 percent of the lifetime diet of Alexander Archipelago wolves in southeastern Alaska (Szepanski et al. 1999, pp. 330-331) and 0 to 16 percent of the wolf diet in coastal British Columbia (Darimont et al. 2004, p. 1871; Darimont et al. 2009, p. 13) (see “Food Habits,” above), we do not anticipate negative effects to them in response to projected declines in salmon productivity at the population or rangewide level owing to the opportunistic predatory behavior of wolves.

We are not aware of any nonregulatory conservation efforts, such as habitat conservation plans, or other voluntary actions that may help to ameliorate potential threats to the habitats used by the Alexander Archipelago wolf.

Although several stressors such as timber harvest, road development, oil development, and climate-related events may be impacting some areas within the range of the Alexander Archipelago wolf, available information does not indicate that these impacts are affecting or are likely to affect the rangewide population. First and foremost, wolf populations in coastal British Columbia, where most (62 percent) of the rangewide population occurs, are stable or slightly increasing even though the landscape has been modified extensively. In fact, a higher proportion of the forested habitat has been logged (24 percent) and the mean road density (0.76 mi per mi2 [0.47 km per km2]) is higher in coastal British Columbia compared to southeastern Alaska (13 percent and 0.37 mi per mi2 [0.23 km per km2], respectively). Second, we found no direct effects of habitat-related stressors that resulted in lower fitness of Alexander Archipelago wolves, in large part because the wolf is a habitat generalist. Third, although deer populations likely will decline in the future as a result of timber harvest, we found that most wolf populations will be resilient to reduced deer abundance because they have access to alternate ungulate and non-ungulate prey that are not impacted significantly by timber harvest, road development, or other stressors that have altered or may alter habitat within the range of the wolf. Only the GMU 2 wolf population likely is being impacted and will continue to be impacted by reduced numbers of deer, the only ungulate prey available; however, we determined that this population does not contribute substantially to the other Alexander Archipelago wolf populations or the rangewide population. Therefore, we posit that most (94 percent) of the rangewide population of Alexander Archipelago wolf likely is not being affected and will not be affected in the future by loss or modification of habitat.

We conclude, based on the best scientific and commercial information available, that the present or threatened destruction, modification, or curtailment of its habitat or range does not currently pose a threat to the Alexander Archipelago wolf at the rangewide level, nor is it likely to become a threat in the future.

The Alexander Archipelago wolf is harvested by humans for commercial and subsistence purposes. Mortality of wolves due to harvest can be compensated for at the population or rangewide level through increased survival, reproduction, or immigration (i.e., compensatory mortality), or harvest mortality may be additive, causing overall survival rates and population growth to decline. The degree to which harvest is considered compensatory, partially compensatory, or at least partially additive is dependent on population characteristics such as age and sex structure, productivity, immigration, and density (e.g., Murray et al. 2010, pp. 2519-2520). Therefore, each wolf population (or group of populations) is different, and a universal rate of sustainable harvest does not exist. In our review, we found rates of human-caused mortality of gray wolf populations varying from 17 to 48 percent, with most being between 20 and 30 percent (Fuller et al. 2003, pp. 184-185; Adams et al. 2008, p. 22; Creel and Rotella 2010, p. 5; Sparkman et al. 2011, p. 5; Gude et al. 2012, pp. 113-116). For the Alexander Archipelago wolf in GMU 2, Person and Russell (2008, p. 1547) reported that total annual mortality greater than 38 percent was unsustainable and that natural mortality averaged about 4 percent (SE = 5) annually, suggesting that human-caused mortality should not exceed 34 percent annually. In our review, we did not find any other estimates of sustainable harvest rates specific to the coastal wolf.

Across the range of the Alexander Archipelago wolf, hunting and trapping regulations, including reporting requirements, vary substantially. In southeastern Alaska, wolf harvest regulations are set by the Alaska Board of Game for all resident and nonresident hunters and trappers, and by the Federal Subsistence Board for federally-qualified subsistence users on Federal lands. In all GMUs, each hunter can harvest a maximum of five wolves, and trappers can harvest an unlimited number of wolves; all harvested wolves must be reported and sealed within a specified time following harvest. In GMU 2 only, an annual harvest guideline is applied; between 1997 and 2014, the harvest guideline was set as 25 to 30 percent of the most recent fall population estimate, and in 2015, this guideline was reduced to 20 percent in response to an apparent decline in the population (see “Abundance and Trend,” above). If the annual harvest guideline is exceeded, then an emergency order closing the hunting and trapping seasons is issued. In coastal British Columbia, the provincial government manages wolf harvest, following an established management plan. The hunting bag limit is three wolves per hunter annually, and, similar to southeastern Alaska, no trapping limit is set. In Regions 1 and 2, all wolf harvest is required to be reported, but no compulsory reporting program exists for Regions 5 and 6.

In this section, we consider wolf harvest as a stressor to the Alexander Archipelago wolf at the population and rangewide levels. Given that harvest regulations and the biological circumstances (e.g., degree of insularity; see “Dispersal and Connectivity,” above) of each wolf population vary considerably, we examined possible effects of wolf harvest to each population by first considering the current condition of the population. If the population is stable or increasing, we presumed that wolves in that population are not being overharvested; if the population is declining or unknown, we assessed mean annual harvest rates based on reported wolf harvest. Because some wolves are harvested and not reported, even in areas where reporting is required, we then applied proportions of unreported harvest to reported harvest for a given year to estimate total harvest, where it was appropriate to do so. We used the population-level information collectively to evaluate impacts of total harvest to the rangewide population of the Alexander Archipelago wolf. We present our analyses and other information related to wolf harvest in southeastern Alaska and coastal British Columbia in more detail in the Status Assessment (Service 2015, “Wolf harvest”).

In coastal British Columbia, populations of the Alexander Archipelago wolf are considered to be stable or slightly increasing (see “Abundance and Trend,” above), and, therefore, we presume that current harvest levels are not impacting those populations. Moreover, in Regions 1 and 2, where reporting is required, few wolves are being harvested on average relative to the estimated population size; in Region 1, approximately 8 percent of the population was harvested annually on average between 1997 and 2012, and in Region 2, the rate is even lower (4 percent). It is more difficult to assess harvest in Regions 5 and 6 because reporting is not required; nonetheless, based on the minimum number of wolves harvested annually from these regions, we estimated that 2 to 7 percent of the populations are harvested on average with considerable variation among years, which could be attributed to either reporting or harvest rates. Overall, we found no evidence indicating that harvest of wolves in coastal British Columbia is having a negative effect on the Alexander Archipelago wolf at the population level and is not likely to have one in the future.

In southeastern Alaska, the GMU 2 wolf population apparently has declined considerably, especially in recent years, although the precision of individual point estimates was low and the confidence intervals overlapped (see “Abundance and Trend,” above). In our review, we found compelling evidence to suggest that wolf harvest likely contributed to this apparent decline. Although annual reported harvest of wolves in GMU 2 equated to only about 17 percent of the population on average between 1997 and 2014 (range = 6-33 percent), documented rates of unreported harvest (i.e., illegal harvest) over a similar time period were high (approximately 38 to 45 percent of total harvest) (Person and Russell 2008, p. 1545; ADFG 2015b, p. 4). Applying these unreported harvest rates, we estimate that mean total annual harvest was 29 percent with a range of 11 to 53 percent, suggesting that in some years, wolves in GMU 2 were being harvested at unsustainable rates; in fact, in 7 of 18 years, total wolf harvest exceeded 34 percent of the estimated population (following Person and Russell [2008, p. 1547], and accounting for natural mortality), suggesting that harvest likely contributed to or caused the apparent population decline. In addition, it is unlikely that increased reproduction and immigration alone could reverse the decline, at least in the short term, owing to this population's insularity (see “Dispersal and Connectivity,” above) and current low proportion of females (see “Abundance and Trend,” above). Thus, we conclude that wolf harvest has impacted the GMU 2 wolf population and, based on the best available information, likely will continue to do so in the near future, consistent with a projected overall population decline on average of 8 to 14 percent (Gilbert et al. 2015, pp. 43, 50), unless total harvest is curtailed.

Trends in wolf populations in the remainder of southeastern Alaska are not known, and, therefore, to evaluate potential impact of wolf harvest to these populations, we reviewed reported wolf harvest in relation to population size and considered whether or not the high rates of unreported harvest in GMU 2 were applicable to populations in GMUs 1, 3, and 5A. Along the mainland (GMUs 1 and 5A) between 1997 and 2014, mean percent of the population harvested annually and reported was 19 percent (range = 11-27), with most of the harvest occurring in the southern portion of the mainland. In GMU 3, the same statistic was 21 percent, ranging from 8 to 37 percent, but with only 3 of 18 years exceeding 25 percent. Thus, if reported harvested rates from these areas are accurate, wolf harvest likely is not impacting wolf populations in GMUs 1, 3, and 5A because annual harvest rates typically are within sustainable limits identified for populations of gray wolf (roughly 20 to 30 percent), including the Alexander Archipelago wolf (approximately 34 percent) (Fuller et al. 2003, pp. 184-185; Adams et al. 2008, p. 22; Person and Russell 2008, p. 1547; Creel and Rotella 2010, p. 5; Sparkman et al. 2011, p. 5; Gude et al. 2012, pp. 113-116). In our review, we found evidence indicating that unreported harvest occasionally occurs in GMUs 1 and 3 (Service 2015, “Unreported harvest”), but we found nothing indicating that it is occurring at the high rates documented in GMU 2.

Harvest rates of wolves in southeastern Alaska are associated with access afforded primarily by boat and motorized vehicle (85 percent of successful hunters and trappers) (ADFG 2012, ADFG 2015d). Therefore, we considered road density, ratio of shoreline to land area, and the total number of communities as proxies to access by wolf hunters and trappers and determined that GMU 2 is not representative of the mainland (GMUs 1 and 5A) or GMU 3 and that applying unreported harvest rates from GMU 2 to other wolf populations is not appropriate. Mean road density in GMU 2 (1.00 mi per mi2 [0.62 km per km2]) is more than twice that of all other GMUs (GMU 1 = 0.13 [0.08], GMU 3 = 0.42 [0.26], and GMU 5A = 0.06 [0.04]). Similarly, nearly all (13 of 15, 87 percent) of the Wildlife Analysis Areas (smaller spatial units that comprise each GMU) that exceed the recommended road density threshold for wolves (1.45 mi per mi2 [0.9 km per km2]) (Person and Russell 2008, p. 1548) are located in GMU 2; one each occurs in GMUs 1 and 3. In addition, the ratio of shoreline to land area, which serves as an indicator of boat acess, in GMU 2 (1.30 mi per mi2 [0.81 km per km2]) is greater than all other GMUs (GMU 1 = 0.29 [0.18], GMU 3 = 1.00 [0.62], and GMU 5A = 0.19 [0.12]). Lastly, although the human population size of GMU 2 is comparatively smaller than in the other GMUs, 14 communities are distributed throughout the unit, more than any other GMU (GMU 1 = 11, GMU 3 = 4, and GMU 5A = 1).

Collectively, these data indicate that hunting and trapping access is greater in GMU 2 than in the rest of southeastern Alaska and that applying unreported harvest rates from GMU 2 to elsewhere is not supported. Therefore, although we recognize that some level of unreported harvest likely is occurring along the mainland of southeastern Alaska and in GMU 3, we do not know the rate at which it may be occurring, but we hypothesize that it likely is less than in GMU 2 because of reduced access. We expect wolf harvest rates in the future to be similar to those in the past because we have no basis from which to expect a change in hunter and trapper effort or success. Consequently, we think that reported wolf harvest rates for GMUs 1, 3, and 5A are reasonably accurate and that wolf harvest is not impacting these populations nor is it likely to do so in the future.

In summary, we find that wolf harvest is not affecting most populations of the Alexander Archipelago wolf. In coastal British Columbia, wolf populations are stable or slightly increasing, suggesting that wolf harvest is not impacting those populations; in addition, mean annual harvest rates of those populations appear to be low (2 to 8 percent of the population based on the best available information). In southeastern Alaska, we determined that the GMU 2 wolf population is being affected by intermediate rates of reported harvest (annual mean = 17 percent) and high rates of unreported harvest (38 to 45 percent of total harvest), which have contributed to an apparent population decline that is projected to continue. We also find that wolf populations in GMUs 1, 3, and 5A experience intermediate rates of reported harvest, 19 to 21 percent of the populations annually, but that these populations likely do not experience high rates of unreported harvest like those estimated for GMU 2 because of comparatively low access to hunters and trappers. In addition, these GMUs are less geographically isolated than GMU 2 and likely have higher immigration rates of wolves. Therefore, based on the best available information, we conclude that wolf harvest of these populations (GMUs 1, 3, and 5A) is occurring at rates similar to or below sustainable harvest rates proposed for gray wolf (roughly 20 to 30 percent) and the Alexander Archipelago wolf (approximately 34 percent) (Fuller et al. 2003, pp. 184-185; Adams et al. 2008, p. 22; Person and Russell 2008, p. 1547; Creel and Rotella 2010, p. 5; Sparkman et al. 2011, p. 5; Gude et al. 2012, pp. 113-116).

Although wolf harvest is affecting the GMU 2 wolf population and likely will continue to do so, we conclude that wolf harvest is not impacting the rangewide population of Alexander Archipelago wolf. The GMU 2 wolf population constitutes a small percentage of the rangewide population (6 percent), is largely insular and geographically peripheral to other populations, and appears to function as a sink population (see “Abundance and Trend” and “Dispersal and Connectivity,” above). Therefore, although we found that this population is experiencing unsustainable harvest rates in some years, owing largely to unreported harvest, we think that the condition of the GMU 2 population has a minor effect on the condition of the rangewide population. The best available information does not suggest that wolf harvest is having an impact on the rangewide population of Alexander Archipelago wolf, nor is it likely to have an impact in the future.

Our review of the best available information does not suggest that overexploitation of the Alexander Archipelago wolf due to scientific or educational purposes is occurring or is likely to occur in the future.

The ADFG has increased educational efforts with the public, especially hunters and trappers, in GMU 2 with the goal of improving communication and coordination regarding management of the wolf population. In recent years, the agency held public meetings, launched a newsletter, held a workshop for teachers, and engaged locals in wolf research. We do not know if these efforts ultimately will be effective at lowering rates of unreported harvest.

We are not aware of any additional conservation efforts or other voluntary actions that may help to reduce overutilization for commercial, recreational, scientific, or educational purposes of the Alexander Archipelago wolf.

We find that wolf harvest is not affecting most Alexander Archipelago wolf populations. In coastal British Columbia, wolf harvest rates are low and are not impacting wolves at the population level, as evidenced by stable or slightly increasing populations. In southeastern Alaska, we found that the GMU 2 wolf population is experiencing high rates of unreported harvest, which has contributed to an apparent population decline, and, therefore, we conclude that this population is being affected by wolf harvest and likely will continue to be affected. We determined that wolf harvest in the remainder of southeastern Alaska is occurring at rates that are unlikely to result in population-level declines. Overall, we found that wolf harvest is not having an effect on the Alexander Archipelago wolf at the rangewide level, although we recognize that the GMU 2 population likely is being harvested at unsustainable rates, especially given other stressors facing the population (e.g., reduced prey availability due to timber harvest). Thus, based on the best available information, we conclude that overexploitation for commercial, recreational, scientific, or educational purposes does not currently pose a threat to the Alexander Archipelago wolf throughout its range, nor is it likely to become a threat in the future.

In this section, we briefly review disease and predation as stressors to the Alexander Archipelago wolf. We describe information presented here in more detail in the Status Assessment (Service 2015, “Disease”).

Several diseases have potential to affect Alexander Archipelago wolf populations, especially given their social behavior and pack structure (see “Social Organization,” above). Wolves are susceptible to a number of diseases that can cause mortality in the wild, including rabies, canine distemper, canine parvovirus, blastomycosis, tuberculosis, sarcoptic mange, and dog louse (Brand et al. 1995, pp. 419-422). However, we found few incidences of diseases reported in Alexander Archipelago wolves; these include dog louse in coastal British Columbia (Hatler et al. 2008, pp. 88-91) and potentially sarcoptic mange (reported in British Columbia, but it is unclear whether or not it occurred along the coast or inland; Miller et al. 2003, p. 183). Both dog louse and mange results in mortality only in extreme cases and usually in pups, and, therefore, it is unlikely that either disease is having or is expected to have a population- or rangewide-level effect on the Alexander Archipelago wolf.

Although we found few reports of diseases in Alexander Archipelago wolves, we located records of rabies, canine distemper, and canine parvovirus in other species in southeastern Alaska and coastal British Columbia, suggesting that transmission is possible but unlikely given the low number of reported incidences. Only four individual bats have tested positive for rabies in southeastern Alaska since the 1970s; bats also are reported to carry rabies in British Columbia, but we do not know whether or not those bats occur on the coast or inland. Canine distemper and parvovirus have been found in domestic dogs on rare occasions; we found only one case of canine distemper, and information suggested that parvovirus has been documented but is rare due to the high percentage of dogs that are vaccinated for it. Nonetheless, we found no documented cases of rabies, canine distemper, or canine parvovirus in wolves from southeastern Alaska or coastal British Columbia.

We acknowledge that diseases such as canine distemper and parvovirus have affected gray wolf populations in other parts of North America (Brand et al. 1995, p. 420 and references therein), but the best available information does not suggest that disease, or even the likelihood of disease in the future, is a threat to the Alexander Archipelago wolf. We conclude that, while some individual wolves may be affected by disease on rare occasions, disease is not having a population- or rangewide-level effect on the Alexander Archipelago wolf now or in the future.

Our review of the best available information did not indicate that predation is affecting or will affect the Alexander Archipelago wolf at the population or rangewide level. As top predators in the ecosystem, predation most likely would occur by another wolf as a result of inter- or intra-pack strife or other territorial behavior. The annual rate of natural mortality, which includes starvation, disease, and predation, was 0.04 (SE = 0.05) for radio-collared wolves in GMU 2 (Person and Russell 2008, p. 1545), indicating that predation is rare and is unlikely to be having a population or rangewide effect. Therefore, we conclude that predation is not a threat to the Alexander Archipelago wolf, nor is it likely to become one in the future.

We are not aware of any conservation efforts or other voluntary actions that may help to reduce disease or predation of the Alexander Archipelago wolf.

We identified several diseases with the potential to affect wolves and possible vectors for transmission, but we found only a few records of disease in individual Alexander Archipelago wolves, and, to the best of our knowledge, none resulted in mortality. Further, we found no evidence that disease is affecting the Alexander Archipelago wolf at the population or rangewide level. Therefore, we conclude that disease is not a threat to the Alexander Archipelago wolf and likely will not become a threat in the future.

We also determined that the most likely predator of individual Alexander Archipelago wolves is other wolves and that this type of predation is a component of their social behavior and organization. Further, predation is rare and is unlikely to be having an effect at population or rangewide levels. Thus, we conclude that predation is not a threat to the Alexander Archipelago wolf, nor is it likely to become one in the future.

In this section, we review laws aimed to help reduce stressors to the Alexander Archipelago wolf and its habitats. However, because we did not find any stressors examined under Factors A, B, and C (described above) and Factor E (described below) to rise to the level of a threat to the Alexander Archipelago wolf rangewide, we also did not find the existing regulatory mechanisms authorized by these laws to be inadequate for the Alexander Archipelago wolf. In other words, we cannot find an existing regulatory mechanism to be inadequate if the stressor intended to be reduced by that regulatory mechanism is not considered a threat to the Alexander Archipelago wolf. Nonetheless, we briefly discuss relevant laws and regulations below.

The National Forest Management Act (NFMA; 16 U.S.C. 1600 et seq.) is the primary statute governing the administration of National Forests in the United States, including the Tongass National Forest. The stated objective of NFMA is to maintain viable, well-distributed wildlife populations on National Forest System lands. As such, the NFMA requires each National Forest to develop, implement, and periodically revise a land and resource management plan to guide activities on the forest. Therefore, in southeastern Alaska, regulation of timber harvest and associated activities is administered by the USFS under the current Tongass Land and Resource Management Plan that was signed and adopted in 2008.

The 2008 Tongass Land and Resource Management Plan describes a conservation strategy that was developed originally as part of the 1997 Plan with the primary goal of achieving objectives under the NFMA. Specifically, the conservation strategy focused primarily on maintaining viable, well-distributed populations of old-growth dependent species on the Tongass National Forest, because these species were considered to be most vulnerable to timber harvest activities on the forest. The Alexander Archipelago wolf, as well as the Sitka black-tailed deer, was used to help design the conservation strategy. Primary components of the strategy include a forest-wide network of old-growth habitat reserves linked by connecting corridors of forested habitat, and a series of standards and guidelines that direct management of lands available for timber harvest and other activities outside of the reserves. We discuss these components in more detail in the Status Assessment (Service 2015, “Existing conservation mechanisms”).

As part of the conservation strategy, we identified two elements specific to the Alexander Archipelago wolf (USFS 2008a, p. 4-95). The first addresses disturbance at and modification of active wolf dens, requiring buffers of 366 m (1,200 ft) around active dens (when known) to reduce risk of abandonment, although if a den is inactive for at least 2 years, this requirement is relaxed. The second pertains to elevated wolf mortality; in areas where wolf mortality concerns have been identified, a Wolf Habitat Management Program will be developed and implemented, in conjunction with ADFG; such a program might include road access management and changes to wolf harvest limit guidelines. However, this element, as outlined in the Plan, does not offer guidance on identifying how, when, or where wolf mortality concerns may exist, but instead it is left to the discretion of the agencies. The only other specific elements relevant to the Alexander Archipelago wolf in the strategy are those that relate to providing sufficient deer habitat capability, which is intended first to maintain sustainable wolf populations, then to consider meeting estimated human deer harvest demands. The strategy offers guidelines for determining whether deer habitat capability within a specific area is sufficient or not.

We find the 2008 Tongass Land and Resource Management Plan, including the conservation strategy, not to be inadequate as a regulatory mechanism aimed to reduce stressors to the Alexander Archipelago wolf and its habitats. Although some parts of the Tongass National Forest have sustained high rates of logging in the past, the majority of it occurred prior to the enactment of the Plan and the conservation strategy. We think that the provisions included in the current Plan are sufficient to maintain habitat for wolves and their prey, especially given that none of the stressors evaluated under Factors A, B, C, and E constitutes a threat to the Alexander Archipelago wolf.

However, we recognize that some elements of the Plan have not been implemented fully yet, as is required under the NFMA. For example, despite evidence of elevated mortality of wolves in GMU 2 (see discussion under Factor B, above), the USFS and ADFG have not developed and implemented a Wolf Habitat Management Program for GMU 2 to date. The reason for not doing so is because the agencies collectively have not determined that current rates of wolf mortality in GMU 2 necessitate concern for maintaining a sustainable wolf population. Although we think that a Wolf Habitat Management Program would benefit the GMU 2 wolf population, we do not view the lack of it as enough to deem the entire Plan, or the existing regulatory mechanisms driving it, to be inadequate for the Alexander Archipelago wolf rangewide. Thus, we conclude that the 2008 Tongass Land and Resource Management Plan is not inadequate to maintain high-quality habitat for the Alexander Archipelago wolf and its prey.

On January 12, 2001, the USFS published a final rule prohibiting road construction and timber harvesting in “inventoried roadless areas” on all National Forest System lands nationwide (hereafter Roadless Rule) (66 FR 3244). On the Tongass National Forest, 109 roadless areas have been inventoried, covering approximately 14,672 mi2 (38,000 km2), although only 463 mi2 (1,200 km2) of these areas have been described as “suitable forest land” for timber harvest (USFS 2008a, p. 7-42; USFS 2008b, pp. 3-444, 3-449). All of these roadless areas are located within the range of the Alexander Archipelago wolf. However, the Roadless Rule was challenged in court and currently a ruling has not been finalized and additional legal challenges are pending; in the meantime, the Tongass is subject to the provisions in the Roadless Rule, although the outcome of these legal challenges is uncertain. Thus, currently, the Roadless Rule protects 14,672 mi2 (38,000 km2) of land, including 463 mi2 (1,200 km2) of productive forest, from timber harvest, road construction, and other development, all of which is within the range of the Alexander Archipelago wolf.

The Alaska Board of Game sets wolf harvest regulations for all resident and nonresident hunters and trappers, and the ADFG implements those regulations. (However, for federally-qualified subsistence users, the Federal Subsistence Board sets regulations, and those regulations are applicable only on Federal lands.) Across most of southeastern Alaska, State regulations of wolf harvest appear not to be resulting in overutilization of the Alexander Archipelago wolf (see discussion under Factor B, above). However, in GMU 2, wolf harvest is having an effect on the population, which apparently has declined over the last 20 years (see “Abundance and Trend,” above). Although the population decline likely was caused by multiple stressors acting synergistically (see Cumulative Effects from Factors A through E, below), overharvest of wolves in some years was a primary contributor, suggesting that the wolf harvest regulations for GMU 2 have been allowing for greater numbers to be harvested than would be necessary to maintain a viable wolf population.

In March 2014, ADFG and the USFS, Tongass National Forest, as the in-season manager for the Federal Subsistence Program, took emergency actions to close the wolf hunting and trapping seasons in GMU 2, yet the population still declined between fall 2013 and fall 2014, likely due to high levels of unreported harvest (38 to 45 percent of total harvest, summarized under Factor B, above). In early 2015, the agencies issued another emergency order and, in cooperation with the Alaska Board of Game, adopted a more conservative wolf harvest guideline for GMU 2, but an updated population estimate is not available yet, and, therefore, we do not know if the recent change in regulation has been effective at avoiding further population decline. Therefore, based on the best available information, we think that wolf harvest regulations in GMU 2 are inadequate to avoid exceeding sustainable harvest levels of Alexander Archipelago wolves, at least in some years. In order to avoid future unsustainable harvest of wolves in GMU 2, regulations should consider total harvest of wolves, including loss of wounded animals, not just reported harvest. Although we found that regulations governing wolf harvest in GMU 2 have been inadequate, we do not expect their inadequacy to impact the rangewide population of Alexander Archipelago wolf for reasons outlined under Factor B, above.

The Alexander Archipelago wolf receives no special protection as an endangered species or species of concern by the State of Alaska (AS 16.20.180). However, in the draft State Wildlife Action Plan, which is not yet finalized, the Alexander Archipelago wolf is identified as a “species of greatest conservation need” because it is a species for which the State has high stewardship responsibility and it is culturally and ecologically important (ADFG 2015e, p. 154).

In coastal British Columbia, populations of the Alexander Archipelago wolf have been stable or slightly increasing for the last 15 years (see “Abundance and Trend,” above). Nonetheless, we identified several laws that ensure its continued protection such as the Forest and Range Practices Act (enacted in 2004), Wildlife Act of British Columbia (amended in 2008), Species at Risk Act, Federal Fisheries Act, Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), and other regional land use and management plans. We review these laws in more detail in the Status Assessment (Service 2015, “Existing conservation measures”).

In 1999, the gray wolf was designated as “not at risk” by the Committee on the Status of Endangered Wildlife in Canada, because it has a widespread, large population with no evidence of a decline over the last 10 years (BCMO 2014, p. 2). In British Columbia, the gray wolf is ranked as “apparently secure” by the Conservation Data Centre and is on the provincial Yellow list, which indicates “secure.” We note here that Canada does not recognize the Alexander Archipelago wolf as a subspecies of gray wolf that occupies coastal British Columbia, and, therefore, these designations are applicable to the province or country scale.

The laws described above regulate timber harvest and associated activities, protect habitat, minimize disturbance at den sites, and aim to ensure sustainable harvest of Alexander Archipelago wolves in southeastern Alaska and coastal British Columbia. As discussed under Factors A, B, C, and E, although we recognize that some stressors such as timber harvest and wolf harvest are having an impact on the GMU 2 wolf population, we have not identified any threat that would affect the taxon as a whole at the rangewide level. Therefore, we find that the existing regulatory mechanisms authorized by the laws described above are not inadequate for the rangewide population of the Alexander Archipelago wolf now and into the future.

In this section, we consider other natural or manmade factors that may be affecting the continued persistence of the Alexander Archipelago wolf and were not addressed in Factors A through D above. Specifically, we examined effects of small and isolated populations, hybridization with dogs, and overexploitation of salmon runs.

In the petition, island endemism was proposed as a possible stressor to the Alexander Archipelago wolf. An endemic is a distinct, unique organism found within a restricted area or range; a restricted range may be an island, or group of islands, or a restricted region (Dawson et al. 2007, p. 1). Although small, isolated populations are more vulnerable to extinction than larger ones due to demographic stochasticity, environmental variability, genetic problems, and catastrophic events (Lande 1993, p. 921), endemism or “rarity” alone is not a stressor. Therefore, we instead considered possible effects associated with small and isolated populations of the Alexander Archipelago wolf.

Several aspects of the life history of the Alexander Archipelago wolf result in it being resilient to effects associated with small and isolated populations. First, the coastal wolf is distributed across a broad range and is not concentrated in any one area, contributing to its ability to withstand catastrophic events, which typically occur at small scales (e.g., wind-caused disturbance) in southeastern Alaska and coastal British Columbia. Second, the Alexander Archipelago wolf is a habitat and diet generalist with high reproductive potential and high dispersal capability in most situations, making it robust to environmental and demographic variability. However, owing to the island geography and steep, rugged terrain within the range of the Alexander Archipelago wolf, some populations are small (fewer than 150 to 250 individuals, following Carroll et al. 2014, p. 76) and at least partially isolated, although most are not. Nonetheless, we focus the remainder of this section on possible genetic consequences to small, partially isolated populations of the Alexander Archipelago wolf.

The primary genetic concern of small, isolated wolf populations is inbreeding, which, at extreme levels, can reduce litter size and increase incidence of skeletal effects (e.g., Liberg et al. 2005, p. 17; Raikkonen et al. 2009, p. 1025). We found only one study that examined inbreeding in the Alexander Archipelago wolf. Breed (2007, p. 18) tested for inbreeding using samples from Regions 5 and 6 in northern British Columbia and GMUs 1 and 2 in southern southeastern Alaska, and found that inbreeding coefficients were highest for wolves in GMU 1, followed by GMU 2, then by Regions 5 and 6. This finding was unexpected given that GMU 2 is the smaller, more isolated population, indicating that inbreeding likely is not affecting the GMU 2 population despite its comparatively small size and insularity. Further, we found no evidence of historic or recent genetic bottlenecking in the Alexander Archipelago wolf (Weckworth et al. 2005, p. 924; Breed 2007, p. 18), although Weckworth et al. (2011, p. 5) speculated that a severe bottleneck may have taken place long ago (over 100 generations).

Therefore, while we recognize that some populations of the Alexander Archipelago wolf are small and insular (e.g., GMU 2 population), our review of the best available information does not suggest that these characteristics currently are having a measurable effect at the population or rangewide level. However, given that the GMU 2 population is expected to decline by an average of 8 to 14 percent over the next 30 years, inbreeding depression and genetic bottlenecking may be a concern for this population in the future, but we think that possible future genetic consequences experienced by the GMU 2 population will not have an effect on the taxon as a whole. Thus, we conclude that small and isolated population effects do not constitute a threat to the Alexander Archipelago wolf, nor are they likely to become a threat in the future.

We reviewed hybridization with domestic dogs as a potential stressor to the Alexander Archipelago wolf. Based on microsatellite analyses, Munoz-Fuentes et al. (2010, p. 547) found that at least one hybridization event occurred in the mid-1980s on Vancouver Island, where wolves were probably extinct at one point in time, but then recolonized the island from the mainland. Although hybridization has been documented and is more likely to occur when wolf abundance is unusually low, most of the range of the Alexander Archipelago wolf is remote and unpopulated by humans, reducing the risk of interactions between wolves and domestic dogs. Therefore, we conclude that hybridization with dogs does not rise to the level of a threat at the population or rangewide level and is not likely to do so in the future.

As suggested in the petition, we considered overexploitation of salmon runs and disease transmission from farmed Atlantic salmon (Salmo salar) in coastal British Columbia as a potential stressor to the Alexander Archipelago wolf (Atlantic salmon are not farmed in southeastern Alaska). The best available information does not indicate that the status of salmon runs in coastal British Columbia is having an effect on coastal wolves. First, Alexander Archipelago wolf populations in coastal British Columbia are stable or slightly increasing, suggesting that neither overexploitation of salmon runs nor disease transmission from introduced salmon are impacting the wolf populations. Second, in coastal British Columbia, only 0 to 16 percent of the diet of the Alexander Archipelago wolf is salmon (Darimont et al. 2004, p. 1871; Darimont et al. 2009, p. 130). Given the opportunistic food habits of the coastal wolf, we postulate that reduction or even near loss of salmon as a food resource may impact individual wolves in some years, but likely would not result in a population- or rangewide-level effect. Further, our review of the best available information does not suggest that this is happening or will happen, or that coastal wolves are acquiring diseases associated with farmed salmon. Therefore, we conclude that overexploitation of salmon runs and disease transmission from farmed salmon do not constitute a threat to the Alexander Archipelago wolf at the population or rangewide level and are not likely to do so in the future.

We are not aware of any conservation efforts or other voluntary actions that may help to reduce effects associated with small and isolated populations, hybridation with dogs, overexploitation of salmon runs, disease transmission from farmed salmon, or any other natural or manmade that may be affecting the Alexander Archipelago wolf.

We find that other natural or manmade factors are present within the range of the Alexander Archipelago wolf, but that none of these factors is having a population or rangewide effect on the Alexander Archipelago wolf. We acknowledge that some populations of the coastal wolf are small and partially isolated, and therefore are susceptible to genetic problems, but we found no evidence that inbreeding or bottlenecking has resulted in a population or rangewide impact to the Alexander Archipelago wolf. In addition, even though some populations are small in size, many populations of the Alexander Archipelago wolf exist and are well distributed on the landscape, greatly reducing impacts from any future catastrophic events to the rangewide population. We also found that the likelihood of hybridation with dogs is low and that any negative impacts associated with the status of salmon in coastal British Columbia are unfounded at this time; neither of these potential stressors is likely to affect the continued persistence of the Alexander Archipelago wolf at the population or rangewide level. Therefore, based on the best available information, we conclude that other natural or manmade factors do not pose a threat to the Alexander Archipelago wolf, nor are they likely to become threats in the future.

The Alexander Archipelago wolf is faced with numerous stressors throughout its range, but none of these individually constitutes a threat to the taxon as a whole now or in the future. However, more than one stressor may act synergistically or compound with one another to impact the Alexander Archipelago wolf at the population or rangewide level. Some of the identified stressors described above have potential to impact wolves directly (e.g., wolf harvest), while others can affect wolves indirectly (e.g., reduction in ungulate prey availability as a result of timber harvest); further, not all stressors are present or equally present across the range of the Alexander Archipelago wolf.

In this section, we consider cumulative effects of the stressors described in Factors A through E. If multiple factors are working together to impact the Alexander Archipelago wolf negatively, the cumulative effects should be manifested in measurable and consistent demographic change at the population or species level. Therefore, for most populations such as those in coastal British Columbia and in GMU 2, we relied on trend information to inform our assessment of cumulative effects. For populations lacking trend information (e.g., GMUs 1, 3, and 5A), we examined the severity, frequency, and certainty of stressors to those populations and relative to the populations for which we have trend information to evaluate cumulative effects. We then assess the populations collectively to draw conclusions about cumulative effects that may be impacting the rangewide population.

In coastal British Columbia, Alexander Archipelago wolf populations are stable or slightly increasing (see “Abundance and Trend,” above), despite multiple stressors facing these populations at levels similar to or greater than most populations in southeastern Alaska. The stability of the wolf populations in coastal British Columbia over the last 15 years suggests that cumulative effects of stressors such as timber harvest, road development, and wolf harvest are not negatively impacting these populations.

The GMU 2 population of the Alexander Archipelago wolf apparently experienced a gradual decline between 1994 and 2013, and then declined substantially between 2013 and 2014, although the overall decline is not statistically significant owing to the large variance surrounding the point estimates (see “Abundance and Trend,” above). Nonetheless, we found evidence that timber harvest (Factor A) and wolf harvest (Factor B) are impacting this population, and these two stressors probably have collectively caused the apparent decline. Given reductions in deer habitat capability as a result of extensive and intensive timber harvest, we expect the GMU 2 wolf population to be somewhat depressed and unable to sustain high rates of wolf harvest. However, in our review of the best available information, we found that high rates of unreported harvest are resulting in unsustainable total harvest of Alexander Archipelago wolves in GMU 2 and that roads constructed largely to support the timber industry are facilitating unsustainable rates of total wolf harvest. Based on a population model specific to GMU 2, Gilbert et al. (2015, p. 43) projected that the wolf population will decline by another 8 to 14 percent, on average, over the next 30 years, largely owing to compounding and residual effects of logging, but also wolf harvest, which results in direct mortality and has a more immediate impact on the population. These stressors and others such as climate related events (i.e., snowfall) are interacting with one another to impact the GMU 2 wolf population and are expected to continue to do so in the future provided that circumstances remain the same (e.g., high unreported harvest rates).

In the remainder of southeastern Alaska where the Alexander Archipelago wolf occurs (i.e., GMUs 1, 3, and 5A), we lack trend and projected population estimates to inform our assessment of cumulative effects, and, therefore, we considered the intensity, frequency, and certainty of stressors present. We found that generally the stressors facing wolf populations in GMUs 1, 3, and 5A occur in slightly higher intensity compared to populations in coastal British Columbia (Regions 5 and 6), but significantly lower intensity than the GMU 2 population. In fact, the percent of logged forest and road densities are among the lowest in the range of the Alexander Archipelago wolf. Although wolf harvest rates were moderately high in GMUs 1, 3, and 5A, given the circumstances of these populations, we found no evidence to suggest that they were having a population-level effect. Importantly, our review of the best available information did not suggest that unreported harvest was occurring at high rates like in GMU 2, and hunter and trapper access was comparatively lower (i.e., road density, ratio of shoreline to land area). In addition, the populations in GMUs 1, 3, and 5A are most similar biologically to the coastal British Columbian populations; all of these wolf populations have access to a variety of ungulate prey and are not restricted to deer, and none is as isolated geographically as the GMU 2 population. We acknowledge that elements of GMU 3 are similar to those in GMU 2 (e.g., island geography), but ultimately we found that GMU 3 had more similarities to GMUs 1 and 5A and coastal British Columbia.

Therefore, in considering all of the evidence collectively, we presume that Alexander Archipelago wolf populations in GMUs 1, 3, and 5A likely are stable and are not being impacted by cumulative effects of stressors because these populations face similar stressors as the populations in coastal British Columbia, which are stable or slightly increasing. The weight of the available information led us to make this presumption regarding the Alexander Archipelago wolf in GMUs 1, 3, and 5A, and we found no information to suggest otherwise. We think our reasoning is fair and supported by the best available information, although we recognize the uncertainties associated with it.

In summary, we acknowledge that some of the stressors facing Alexander Archipelago wolves interact with one another, particularly timber harvest and wolf harvest, but we determined that all but one of the wolf populations do not exhibit impacts from cumulative effects of stressors. We found that about 62 percent of the rangewide population of the Alexander Archipelago wolf is stable (all of coastal British Columbia), and another 32 percent is presumed to be stable (GMUs 1, 3, and 5A), suggesting that approximately 94 percent of the rangewide population is not experiencing negative and cumulative effects from stressors, despite their presence. Therefore, we conclude that cumulative impacts of identified stressors do not rise to the level of a threat to the Alexander Archipelago wolf and are unlikely to do so in the future.

As required by the Act, we considered the five factors in assessing whether the Alexander Archipelago wolf is an endangered or threatened species throughout all of its range. We examined the best scientific and commercial information available regarding the past, present, and future threats faced by the Alexander Archipelago wolf. We reviewed the petition, information available in our files, and other available published and unpublished information, and we consulted with recognized wolf experts and other Federal, State, and tribal agencies. We prepared a Status Assessment that summarizes all of the best available science related to the Alexander Archipelago wolf and had it peer reviewed by three experts external to the Service and selected by a third-party contractor. We also contracted the University of Alaska Fairbanks to revise an existing population model for the GMU 2 wolf population, convened a 2-day workshop with experts to review the model inputs and structure, and had the final report reviewed by experts (Gilbert et al. 2015, entire). As part of our review, we brought together researchers with experience and expertise in gray wolves and the temperate coastal rainforest from across the Service to review and evaluate the best available scientific and commercial information.

We examined a variety of potential threats facing the Alexander Archipelago wolf and its habitats, including timber harvest, road development, oil development, climate change, overexploitation, disease, and effects associated with small and isolated populations. To determine if these risk factors individually or collectively put the taxon in danger of extinction throughout its range, or are likely to do so in the foreseeable future, we first considered if the identified risk factors were causing a population decline or other demographic changes, or were likely to do so in the foreseeable future.

Throughout most of its range, the Alexander Archipelago wolf is stable or slightly increasing or is presumed to be stable based on its demonstrated high resiliency to the magnitude of stressors present. In coastal British Columbia, which constitutes 67 percent of the range and 62 percent of the rangewide population, the Alexander Archipelago wolf has been stable or slightly increasing over the last 15 years. In mainland southeastern Alaska (GMUs 1 and 5A) and in GMU 3, approximately 29 percent of the range and 32 percent of the rangewide population, we determined that the circumstances of these wolf populations were most similar to those in coastal British Columbia, and, therefore, based on the best available information, we reasoned that the Alexander Archipelago wolf likely is stable in GMUs 1, 3, and 5A. In GMU 2, which includes only 4 percent of the range and 6 percent of the rangewide population, the Alexander Archipelago wolf has been declining since 1994, and is expected to continue declining by another 8 to 14 percent, on average, over the next 30 years. Nonetheless, we conclude that the Alexander Archipelago wolf is stable or slightly increasing in nearly all of its range (96 percent), representing 94 percent of the rangewide population of the taxon.

We then identified and evaluated existing and potential stressors to the Alexander Archipelago wolf. We aimed to determine if these stressors are affecting the taxon as a whole currently or are likely to do so in the foreseeable future, are likely to increase or decrease, and may rise to the level of a threat to the taxon, rangewide or at the population level. Because the Alexander Archipelago wolf is broadly distributed across its range and is a habitat and diet generalist, we evaluated whether each identified stressor was expected to impact wolves directly or indirectly and whether wolves would be resilient to any impact.

We examined several stressors that are not affecting the Alexander Archipelago wolf currently and are unlikely to occur at a magnitude and frequency in the future that would result in a population- or rangewide-level effect. We found that oil and gas development, disease, predation, effects associated with small and isolated populations, hybridization with domestic dogs, overexploitation of salmon runs, and disease transmission from farmed salmon are not threats to the Alexander Archipelago wolf (see discussions under Factors A, C, and E, above). Most of these stressors are undocumented and speculative, rarely occur, are spatially limited, or are not known to impact gray wolves in areas of overlap. Although disease is known to affect populations of gray wolves, we found few reports of disease in the Alexander Archipelago wolf, and none resulted in mortality. Therefore, based on the best available information, we conclude that none of these stressors is having a population- or rangewide-level effect on the Alexander Archipelago wolf, or is likely to do so in the foreseeable future.

Within the range of the Alexander Archipelago wolf, changes in climate are occurring and are predicted to continue, likely resulting in improved conditions for wolves. Climate models for southeastern Alaska and coastal British Columbia project that precipitation as snow will decrease substantially in the future, which will improve winter conditions for deer, the primary prey species of wolves. Although severe winters likely will continue to occur and will affect deer populations, we expect them to occur less frequently. Therefore, based on the best available information, we conclude that the effects of climate change are not a threat to the Alexander Archipelago wolf, nor are they likely to become a threat in the foreseeable future.

We reviewed timber harvest and associated road development as stressors to the Alexander Archipelago wolf and found that they are not affecting wolves directly, in large part because the wolf is a habitat generalist. Although wolves used den sites farther from logged stands and roads than unused sites, den site selection was more strongly influenced by natural features on the landscape such as slope, elevation, and proximity to freshwater. Further, we did not find evidence indicating that denning near logged stands and roads resulted in lower fitness of wolves. Thus, we conclude that timber harvest and associated road development are not affecting wolves at the population or rangewide levels by decreasing suitable denning habitat. We did not identify any other potential direct impacts to wolves as a result of timber harvest or road development, so next we examined potential indirect effects, specifically reduction of deer habitat capability.

Although the Alexander Archipelago wolf is an opportunistic predator that feeds on a variety of marine, intertidal, and terrestrial species, ungulates compose at least half of the wolf's diet throughout its range, and deer is the most widespread and abundant ungulate available to wolves. Timber harvest has reduced deer habitat capability, which in turn is predicted to reduce deer populations, especially in areas that have been logged intensively. However, based largely on the stability of wolf populations in coastal British Columbia despite intensive timber harvest, we conclude that wolves are resilient to changes in deer populations provided that they have other ungulate prey species available to them. We found that nearly all of the Alexander Archipelago wolves (94 percent of the rangewide population) have access to alternate ungulate prey such as mountain goat, moose, and elk, and, based on wolf diet, Alexander Archipelago wolves are consuming these prey species in areas where they are available. We identified only one Alexander Archipelago wolf population as an exception.

In GMU 2, deer is the only ungulate species available to wolves, and, therefore, wolves in this population have a more restricted ungulate diet and likely are being affected by cascading effects of timber harvest. Both deer and wolves are projected to decline in GMU 2 in the future, largely due to long-term reduction in deer habitat capability. However, we find that the GMU 2 population contributes little to the rangewide population because it constitutes only 4 percent of the range and 6 percent of the rangewide population, is largely insular and geographically peripheral, and appears to function as a sink population. Therefore, while we recognize that timber harvest and associated road development has modified a considerable portion of the range of the Alexander Archipelago wolf, and will continue to do so, we find that the taxon as a whole is not being affected negatively, in large part because the wolf is a habitat and diet generalist. Based on the best available information, we conclude that timber harvest and associated road development do not rise to the level of a threat to the Alexander Archipelago wolf, and are not likely to do so in the future.

Throughout its range, the Alexander Archipelago wolf is harvested for commercial and subsistence purposes, and, therefore, we examined overutilization as a stressor at the population and rangewide levels. In coastal British Columbia, we presume that wolf harvest is not having an effect at the population level given that populations there are stable or slightly increasing. This presumption is supported by the comparatively low rates of reported wolf harvest in coastal British Columbia, although reporting of harvest is required only in Regions 1 and 2, and, therefore, we considered these rates as minimum values. Nonetheless, we found no information suggesting that wolf harvest in coastal British Columbia is affecting wolves at the population level, as evidenced by the stability of the populations.

Within southeastern Alaska, where reporting is required, rates of reported harvest on average are similar across all populations (17 to 21 mean percent of population annually). However, in GMU 2, unreported harvest can be a substantial component of total harvest (38 to 45 percent), resulting in high rates of total harvest in some years, which likely has contributed to the apparent population decline in GMU 2. Although unreported harvest probably occurs in other parts of southeastern Alaska, our review of the best available information does not indicate that it is occurring at the same high rate as documented in GMU 2. Further, access by hunters and trappers is significantly greater in GMU 2 compared to elsewhere (see discussion under Factor B, above), and, therefore, we find that applying rates of unreported harvest from GMU 2 to other wolf populations in southeastern Alaska is not appropriate. Thus, based on the best available information, we think that wolf harvest in most of southeastern Alaska (i.e., GMUs 1, 3, and 5A) is not affecting wolves at the population level, but that total wolf harvest in GMU 2 likely has occurred, at least recently, at unsustainable rates, largely due to high rates of unreported harvest, and has contributed to or caused an apparent decline in the population. However, for the same reasons described above, we determined that negative population impacts in GMU 2 do not affect the rangewide population significantly, and, therefore, we conclude that wolf harvest is not having a rangewide-level effect. In conclusion, we find that overutilization is not a threat to the Alexander Archipelago wolf, nor is it likely to become a threat in the foreseeable future.

In summary, we found that the Alexander Archipelago wolf experiences stressors throughout its range, but based on our consideration of the best available scientific and commercial information, we determined that the identified stressors, individually or collectively, do not pose a threat to the taxon at the rangewide level now or in the foreseeable future. We determined that many of the life-history traits and behaviors of the Alexander Archipelago wolf, such as its variable diet, lack of preferential use of habitats, and high reproductive potential, increase its ability to persist in highly modified habitats with numerous stressors. Only one population of the Alexander Archipelago wolf has declined and likely will continue to decline, but this population contributes little to the taxon as a whole, and, therefore, while we acknowledge the vulnerability of this population to stressors such as timber harvest and wolf harvest, we find that its status does not affect the rangewide status significantly. Further, we found that approximately 94 percent of the rangewide population of the Alexander Archipelago wolf is stable or increasing, or presumed with reasonable confidence to be stable. Therefore, based on our review of the best available scientific and commercial information pertaining to the five factors, we find that the threats are not of sufficient imminence, intensity, or magnitude to indicate that the Alexander Archipelago wolf is in danger of extinction (endangered), or likely to become endangered within the foreseeable future (threatened), throughout all of its range.

Under the Act and our implementing regulations, a species may warrant listing if it is in danger of extinction or likely to become so throughout all or a significant portion of its range. The Act defines “endangered species” as any species which is “in danger of extinction throughout all or a significant portion of its range,” and “threatened species” as any species which is “likely to become an endangered species within the foreseeable future throughout all or a significant portion of its range.” The term “species” includes “any subspecies of fish or wildlife or plants, and any distinct population segment [DPS] of any species of vertebrate fish or wildlife which interbreeds when mature.” We published a final policy interpreting the phrase “significant portion of its range” (SPR) (79 FR 37578, July 1, 2014). The final policy states that (1) if a species is found to be endangered or threatened throughout a significant portion of its range, the entire species is listed as an endangered or a threatened species, respectively, and the Act's protections apply to all individuals of the species wherever found; (2) a portion of the range of a species is “significant” if the species is not currently endangered or threatened throughout all of its range, but the portion's contribution to the viability of the species is so important that, without the members in that portion, the species would be in danger of extinction, or likely to become so in the foreseeable future, throughout all of its range; (3) the range of a species is considered to be the general geographical area within which that species can be found at the time the Service or the National Marine Fisheries Service makes any particular status determination; and (4) if a vertebrate species is endangered or threatened throughout an SPR, and the population in that significant portion is a valid DPS, we will list the DPS rather than the entire taxonomic species or subspecies.

The SPR policy is applied to all status determinations, including analyses for the purposes of making listing, delisting, and reclassification determinations. The procedure for analyzing whether any portion is an SPR is similar, regardless of the type of status determination we are making. The first step in our analysis of the status of a species is to determine its status throughout all of its range. If we determine that the species is in danger of extinction, or likely to become so in the foreseeable future, throughout all of its range, we list the species as an endangered (or threatened) species and no SPR analysis will be required. If the species is neither in danger of extinction nor likely to become so throughout all of its range, we determine whether the species is in danger of extinction or likely to become so throughout a significant portion of its range. If it is, we list the species as an endangered or a threatened species, respectively; if it is not, we conclude that listing the species is not warranted.

When we conduct an SPR analysis, we first identify any portions of the species' range that warrant further consideration. The range of a species can theoretically be divided into portions in an infinite number of ways. However, there is no purpose to analyzing portions of the range that are not reasonably likely to be significant and endangered or threatened. To identify only those portions that warrant further consideration, we determine whether there is substantial information indicating that (1) the portions may be significant and (2) the species may be in danger of extinction in those portions or likely to become so within the foreseeable future. We emphasize that answering these questions in the affirmative is not a determination that the species is endangered or threatened throughout a significant portion of its range; rather, it is a step in determining whether a more detailed analysis of the issue is required. In practice, a key part of this analysis is whether the threats are geographically concentrated in some way. If the threats to the species are affecting it uniformly throughout its range, no portion is likely to warrant further consideration. Moreover, if any concentration of threats apply only to portions of the range that clearly do not meet the biologically based definition of “significant” (i.e., the loss of that portion clearly would not be expected to increase the vulnerability to extinction of the entire species), those portions will not warrant further consideration.

If we identify any portions that may be both (1) significant and (2) endangered or threatened, we engage in a more detailed analysis to determine whether these standards are indeed met. The identification of an SPR does not create a presumption, prejudgment, or other determination as to whether the species in that identified SPR is endangered or threatened. We must go through a separate analysis to determine whether the species is endangered or threatened in the SPR. To determine whether a species is endangered or threatened throughout an SPR, we will use the same standards and methodology that we use to determine if a species is endangered or threatened throughout its range.

Depending on the biology of the species, its range, and the threats it faces, it may be more efficient to address the “significant” question first, or the status question first. Thus, if we determine that a portion of the range is not “significant,” we do not need to determine whether the species is endangered or threatened there; if we determine that the species is not endangered or threatened in a portion of its range, we do not need to determine if that portion is “significant.”

We evaluated the current range of the Alexander Archipelago wolf to determine if there is any apparent geographic concentration of potential threats to the taxon. We examined potential threats from timber harvest, oil and gas development, road development, climate change, effects of small and isolated populations, hybridization with dogs, overexploitation of salmon runs, disease transmission from farmed salmon, overutilization, disease, and predation. We found that potential threats are concentrated in GMU 2, where they are substantially greater than in other portions of its range. We considered adjacent parts of the range that are contained in GMUs 1 and 3, but, based on the best available information, we did not find any concentrations of stressors in those parts that were similar in magnitude and frequency to the potential threats in GMU 2. Therefore, we then considered whether GMU 2 is “significant” based on the Service's SPR policy, which states that a portion of its range is “significant” if the taxon is not currently endangered or threatened throughout all of its range, but the portion's contribution to the viability of the taxon is so important that, without the members in that portion, the taxon would be in danger of extinction, or likely to become so in the foreseeable future, throughout all of its range.

We reviewed population and rangewide metrics in relation to GMU 2 to estimate the numerical contribution of GMU 2 to the viability of the Alexander Archipelago wolf. We determined that GMU 2 constitutes only 4 percent of the total range and 9 percent of the range below 1,312 ft (400 m) in elevation where these wolves spend most of their time (see “Space and Habitat Use,” above). In addition, based on the most current population estimate for GMU 2, which was assessed in 2014, we estimated that only 6 percent of the rangewide population occupies GMU 2. Recognizing the apparent recent decline in the GMU 2 population (see “Abundance and Trend,” above), we then estimated that in 2013, the GMU 2 population composed about 13 percent of the rangewide population. We expect wolf abundance to fluctuate annually at the population and rangewide scales, but generally in recent years, we find that the GMU 2 population composes a somewhat small percentage of the rangewide population. Therefore, we conclude that, numerically, the GMU 2 population contributes little to the viability of the taxon as a whole given that it composes a small percentage of the current rangewide population and it occupies a small percentage of the range of the Alexander Archipelago wolf.

We then considered the biological contribution of the GMU 2 population to the viability of the Alexander Archipelago wolf. We found that given its insularity and peripheral geographic position compared to the rest of the range, the GMU 2 population contributes even less demographically and genetically than it does numerically. In fact, it appears to function as a sink population with gene flow and dispersal primarily occurring uni-directionally from other areas to GMU 2 (see “Dispersal and Connectivity,” above). Therefore, overall, we found that GMU 2 represents a small percentage of the range and rangewide population of the Alexander Archipelago wolf, it is insular and geographically peripheral, and it appears to be functioning as a sink population to the Alexander Archipelago wolf. We conclude that, although potential threats are concentrated in GMU 2, this portion's contribution to the viability of the taxon as a whole is not so important that, without the members of GMU 2, the Alexander Archipelago wolf would be in danger of extinction, or likely to become so in the foreseeable future, throughout all of its range.

Our review of the best available scientific and commercial information indicates that the Alexander Archipelago wolf is not in danger of extinction (endangered) nor likely to become endangered within the foreseeable future (threatened), throughout all or a significant portion of its range. Therefore, we find that listing the Alexander Archipelago wolf as an endangered or threatened species under the Act is not warranted at this time.

After determining that the Alexander Archipelago wolf is not endangered or threatened throughout all or a significant portion of its range and is not likely to become so in the foreseeable future, we then evaluate whether or not the GMU 2 wolf population meets the definition of a distinct population segment (DPS) under the Act, as requested in the petition.

To interpret and implement the DPS provisions of the Act and Congressional guidance, we, in conjunction with the National Marine Fisheries Service, published the Policy Regarding the Recognition of Distinct Vertebrate Population Segments (DPS policy) in the Federal Register on February 7, 1996 (61 FR 4722). Under the DPS policy, two basic elements are considered in the decision regarding the establishment of a population of a vertebrate species as a possible DPS. We must first determine whether the population qualifies as a DPS; this requires a finding that the population is both: (1) Discrete in relation to the remainder of the taxon to which it belongs; and (2) biologically and ecologically significant to the taxon to which it belongs. If the population meets the first two criteria under the DPS policy, we then proceed to the third element in the process, which is to evaluate the population segment's conservation status in relation to the Act's standards for listing as an endangered or threatened species. These three elements are applied similarly for additions to or removals from the Federal Lists of Endangered and Threatened Wildlife and Plants.

In accordance with our DPS policy, we detail our analysis of whether a vertebrate population segment under consideration for listing may qualify as a DPS. As described above, we first evaluate the population segment's discreteness from the remainder of the taxon to which it belongs. Under the DPS policy, a population segment of a vertebrate taxon may be considered discrete if it satisfies either one of the following conditions:

(1) It is markedly separated from other populations of the same taxon as a consequence of physical, physiological, ecological, or behavioral factors. Quantitative measures of genetic or morphological discontinuity may provide evidence of this separation.

(2) It is delimited by international governmental boundaries within which differences in control of exploitation, management of habitat, conservation status, or regulatory mechanisms exist that are significant in light of section 4(a)(1)(D) of the Act.

We found that the GMU 2 population is markedly separated as a consequence of physical, physiological, ecological, or behavioral factors from other populations of the Alexander Archipelago wolf. It occupies a portion of the Alexander Archipelago within the range of wolf that is physically separated from adjacent populations due to comparatively long and swift water crossings and the fact that few crossings are available to dispersing wolves. Although low levels of movement between the GMU 2 population segment and other populations likely occur (see “Dispersal and Connectivity,” above), the GMU 2 wolf population is largely insular and geographically peripheral to the rest of the range of the Alexander Archipelago wolf; further, the Service's DPS policy does not require absolute separation to be considered discrete.

In addition, several studies have demonstrated that, based on genetic assignment tests, the GMU 2 wolf population forms a distinct genetic cluster when compared to other Alexander Archipelago wolves (Weckworth et al. 2005, pp. 923, 926; Breed 2007, p. 21). Further, estimates of the fixation index (F ST, the relative proportion of genetic variation explained by differences among populations) are markedly higher between the GMU 2 population and all other Alexander Archipelago wolf populations than comparisons between other populations (e.g., Weckworth et al. 2005, p. 923; Cronin et al. 2015, p. 7). Collectively, these findings indicate genetic discontinuity between wolves in GMU 2 and those in the rest of the range of the Alexander Archipelago wolf. We review these studies and others in more detail in the Status Assessment (Service 2015, “Genetic analyses”).

We found that the GMU 2 population of the Alexander Archipelago wolf is markedly separated as a consequence of physical (geographic) features and due to genetic divergence from other populations of the taxon. Therefore, we conclude that it is discrete under the Service's DPS policy.

If a population is considered discrete under one or more of the conditions described in the Service's DPS policy, its biological and ecological significance will be considered in light of Congressional guidance that the authority to list DPSs be used “sparingly” while encouraging the conservation of genetic diversity. In making this determination, we consider available scientific evidence of the discrete population segment's importance to the taxon to which it belongs. As precise circumstances are likely to vary considerably from case to case, the DPS policy does not describe all the classes of information that might be used in determining the biological and ecological importance of a discrete population. However, the DPS policy describes four possible classes of information that provide evidence of a population segment's biological and ecological importance to the taxon to which it belongs. As specified in the DPS policy (61 FR 4722), this consideration of the population segment's significance may include, but is not limited to, the following:

(1) Persistence of the discrete population segment in an ecological setting unusual or unique to the taxon;

(2) Evidence that loss of the discrete population segment would result in a significant gap in the range of a taxon;

(3) Evidence that the discrete population segment represents the only surviving natural occurrence of a taxon that may be more abundant elsewhere as an introduced population outside its historical range; or

(4) Evidence that the discrete population segment differs markedly from other populations of the taxon in its genetic characteristics.

Given our determination that the GMU 2 wolf population is discrete under the Service's DPS policy, we now evaluate the biological and ecological significance of the population relative to the taxon as a whole. A discrete population segment is considered significant under the DPS policy if it meets one of the four elements identified in the policy under significance (described above), or otherwise can be reasonably justified as being significant. Here, we evaluate the four potential factors suggested by our DPS policy in evaluating significance of the GMU 2 wolf population.

We find that the GMU 2 population does not persist in an ecological setting that is unusual or unique to the Alexander Archipelago wolf. To evaluate this element, we considered whether or not the habitats used by Alexander Archipelago wolves in GMU 2 include unusual or unique features that are not used by or available to the taxon elsewhere in its range. We found that the Alexander Archipelago wolf is a habitat generalist, using a variety of habitats on the landscape and selecting only for those that occur below 1,312 ft (400 m) in elevation (see “Space and Habitat Use,” above). Throughout its range, habitats used by and available to the Alexander Archipelago wolf are similar with some variation from north to south and on the mainland and islands, but we found no unique or unusual features specific to GMU 2 that were not represented elsewhere in the range. Although karst is more prevalent in GMU 2, we found no evidence indicating that wolves selectively use karst; in addition, karst is present at low and high elevations in GMUs 1 and 3 (Carstensen 2007, p. 24).

The GMU 2 wolf population has a more restricted ungulate diet, comprised only of deer, than other populations of the Alexander Archipelago wolf (see “Food Habits,” above). However, given that the coastal wolf is an opportunistic predator, feeding on intertidal, marine, freshwater, and terrestrial species, we find that differences in ungulate prey base are not ecologically unique or unusual. In addition, Alexander Archipelago wolves feed on deer throughout their range in equal or even higher proportions than wolves in GMU 2 (e.g., Szepanski et al. 1999, p. 331; Darimont et al. 2009, p. 130), demonstrating that a diet based largely on deer is not unusual or unique. Thus, compared to elsewhere in the range, we found nothing unique or unusual about the diet or ecological setting of wolves in GMU 2. Further, we did not identify any morphological, physiological, or behavioral characteristics of the GMU 2 wolf population that differ from those of other Alexander Archipelago wolf populations, which may have suggested a biological response to an unusual or unique ecological setting. Therefore, we conclude that the GMU 2 wolf population does not meet the definition of significance under this element, as outlined in the Service's DPS policy.

We find that loss of the GMU 2 population of the Alexander Archipelago wolf, when considered in relation to the taxon as a whole, would not result in a significant gap in the range of the taxon. It constitutes only 6 percent of the current rangewide population, only 4 percent of the range, and only 9 percent of the range below 1,312 (400 m) in elevation where the Alexander Archipelago wolf selectively occurs. In addition, the GMU 2 population is largely insular and geographically peripheral to other populations, and appears to function as a sink population (see “Abundance and Trend” and “Dispersal and Connectivity,” above). For these reasons, we found that the demographic and genetic contributions of the GMU 2 wolf population to the rangewide population are low and that loss of this population would have a minor effect on the rangewide population of the Alexander Archipelago wolf. Also, although rates of immigration to GMU 2 likely are low (see “Dispersal and Connectivity,” above), recolonization of GMU 2 certainly is possible, especially given the condition of the remainder of the rangewide population. Therefore, we conclude that the GMU 2 wolf population does not meet the definition of significance under this element, as outlined in the Service's DPS policy.

The GMU 2 population does not represent the only surviving natural occurrence of the Alexander Archipelago wolf throughout the range of the taxon. Therefore, we conclude that the discrete population of the Alexander Archipelago wolf in GMU 2 does not meet the significance criterion of the DPS policy under this factor.

We find that the GMU 2 population does not differ markedly from other Alexander Archipelago wolves in its genetic characteristics. As noted above in Discreteness, the GMU 2 population exhibits genetic discontinuities from other Alexander Archipelago wolves due to differences in allele and haplotype frequencies. However, those discontinuities are not indicative of rare or unique genetic characterisics within the GMU 2 population that are significant to the taxon. Rather, several studies indicate that the genetic diversity within the GMU 2 population is a subset of the genetic diversity found in other Alexander Archipelago wolves. For example, the GMU 2 population does not harbor unique haplotypes; only one haplotype was found in the GMU 2 population, and it was found in other Alexander Archipelago wolves including those from coastal British Columbia (Weckworth et al. 2010, p. 367; Weckworth et al. 2011, p. 2). In addition, the number and frequency of private alleles in the GMU 2 population is low compared to other Alexander Archipelago wolves (e.g., Breed 2007, p. 18). The lack of unique haplotypes and the low numbers of private alleles both indicate that the GMU 2 population has not been completely isolated historically from other Alexander Archipelago wolves. Finally, these genetic studies demonstrate that wolves in GMU 2 exhibit low genetic diversity (as measured through allelic richness, heterozygosity, and haplotype diversity) compared to other Alexander Archipelago wolves (Weckworth et al. 2005, p. 919; Breed 2007, p. 17; Weckworth et al. 2010, p. 366; Weckworth et al. 2011, p. 2).

Collectively, results of these studies suggest that the genetic discontinuities observed in the GMU 2 population likely are the outcome of restricted gene flow and a loss of genetic diversity through genetic drift or founder effects. Therefore, although the GMU 2 population is considered discrete under the Service's DPS policy based on the available genetic data, it does not harbor genetic characteristics that are rare or unique to the Alexander Archipelago wolf and its genetic contribution to the taxon as a whole likely is minor. Moreover, while we found no genetic studies that have assessed adaptive genetic variation of the Alexander Archipelago wolf, the best available genetic data do not indicate that the GMU 2 population harbors significant adaptive variation, which is supported further by the fact that the GMU 2 population is not persisting in an unusual or unique ecological setting. Therefore, we conclude that the GMU 2 population does not meet the definition of significance under this element, as outlined in the Service's DPS policy.

We determine, based on a review of the best available information, that the GMU 2 population is not significant in relation to the remainder of the taxon. Therefore, this population does not qualify as a DPS under our 1996 DPS policy and is not a listable entity under the Act. Because we found that the population did not meet the significance element and, therefore, does not qualify as a DPS under the Service's DPS policy, we will not proceed with an evaluation of the status of the population under the Act.

Based on the best scientific and commercial information available, as described above, we find that, under the Service's DPS policy, the GMU 2 population is discrete, but is not significant to the taxon to which it belongs. Because the GMU 2 population is not both discrete and significant, it does not qualify as a DPS under the Act.

Our review of the best available scientific and commercial information indicates that the Alexander Archipelago wolf is not in danger of extinction (endangered) nor likely to become endangered within the foreseeable future (threatened), throughout all or a significant portion of its range. Therefore, we find that listing the Alexander Archipelago wolf as an endangered or threatened species under the Act is not warranted at this time.

We request that you submit any new information concerning the status of, or threats to, the Alexander Archipelago wolf to our Anchorage Fish and Wildlife Field Office (see ADDRESSES) whenever it becomes available. New information will help us monitor the Alexander Archipelago wolf and encourage its conservation. If an emergency situation develops for the Alexander Archipelago wolf, we will act to provide immediate protection.

A complete list of references cited is available on the Internet at http://www.regulations.gov and upon request from the Anchorage Fish and Wildlife Field Office (see ADDRESSES).

The primary authors of this document are the staff members of the Anchorage Fish and Wildlife Field Office.

The authority for this section is section 4 of the Endangered Species Act of 1973, as amended (16 U.S.C. 1531 et seq.).