Marine Mammals; Incidental Take of Polar Bears and Pacific Walruses in the Beaufort Sea and North Slope of Alaska

The FWS published a proposed ITR in the Federal Register on March 9, 2026, with a 30-day period seeking comments on both the proposed ITR and draft EA (91 FR 11240). The comment period closed on April 8, 2026. The FWS received 85,589 comments. Comments were received from the requester, the State of Alaska, the North Slope Borough, various advocacy organizations, student groups, a corporation, and interested members of the public. Submissions with multiple signatures were counted as multiple comments (one comment per individual signing).

The majority of comments (77,468) were part of mass mailing campaigns, where members of public added information and signatures to letters led by environmental organizations expressing opposition to the promulgation of the regulations. Outside of these mass campaigns, an additional 8,104 comments were submitted individually. Many of these submissions were identical or highly similar variations of nine different comment letters all expressing opposition to the regulations. A small number of comments were submitted to this docket by mistake, either as duplicate entries or as comments intended for a different docket. Nearly all individually submitted comments expressed opposition to the regulations.

We reviewed all comments, which are part of the docket for these ITRs, for substantive issues, new information, and recommendations regarding these ITRs and EA. The FWS used “DocketScope” to aggregate comments submitted by the public. The comments are aggregated by subject matter, summarized and addressed below, and changes have been incorporated into the final ITRs and final EA as appropriate. A summary of the changes to these final ITRs from the proposed ITR is found in the preamble section, entitled Summary of Changes from the Proposed Regulations.

Comment (1): Many commenters responded to FWS's statement in the proposed ITR that “[t]he FWS is now considering whether the best reading of the MMPA's provisions concerning ITRs requires the FWS to allow (1) all types of incidental take that result from the specified activities; (2) only the types of incidental take that FWS anticipated during the rulemaking process; or (3) only the types of incidental take that were requested to be allowed.” Commenters offered different views on how FWS should proceed in this regard.

Our Response: After considering the various interpretations and perspectives provided via public comments, FWS finds that approach (1) as described in this comment and in the proposed ITR reflects the best reading of the MMPA. Where FWS receives an adequate and complete request for an ITR, satisfies applicable procedural requirements and makes findings that the incidental taking from the specified activities will comply with applicable MMPA standards (i.e., small numbers, negligible impact, and no unmitigable adverse impacts on the availability of the species or stock for taking for subsistence uses), FWS will issue an ITR allowing for all types of incidental take that result from the specified activities (to the extent they are conducted in compliance with ITR and LOA requirements).

This interpretation flows from section 101(a)(5)(A) which provides that in the circumstances described above, the Secretary “shall allow . . . the incidental, but not intentional, taking by citizens while engaging in [a specified activity within a specified geographical region. . . .]” This statutory directive to allow “the incidental . . . taking” is non-discretionary (i.e., the Secretary “shall” allow it) and is not limited or caveated by any language concerning the types or numbers of incidental taking to be allowed.

Additional support for this interpretation is provided in comment responses below. Since the proposed regulatory text provided in the proposed ITR reflected the interpretation described above, no related changes to that regulatory text are made in this final ITR.

Comment (2): One comment agreed that this interpretation reflects the best reading of the MMPA, noting that the exemption from the MMPA's general take prohibition afforded by section 101(a)(5)(A) focuses on “citizens . . . who engage in a specified activity . . . within a specified geographical region,” and then exempts “the incidental, but not intentional taking by citizens while engaging in that activity within that region.” The comment also notes case law recognizing that there is nothing in section 101(a)(5)(A) that requires FWS, when promulgating ITRs, to quantify or estimate the number of animals that would be taken.

Our Response: FWS agrees with this interpretation of section 101(a)(5)(A). FWS also recognizes that the absence of a requirement to quantify or estimate the number of animals that would be taken when promulgating ITRs further undercuts the notion that ITRs may only allow that incidental take which was quantified or estimated. FWS relatedly notes that the statutory provision establishing the required contents of ITRs expressly requires FWS to set forth, inter alia, “permissible methods of taking pursuant to such activity” but says nothing of permissible types or numbers of taking pursuant to such activity (16 U.S.C. 1371(a)(5)(A)(i)(II)).

Comment (3): Some comments asserted that because FWS cannot authorize incidental take unless FWS finds the total taking will comply with certain MMPA standards, it necessarily follows that FWS cannot authorize any additional take beyond what it anticipated in making its MMPA findings.

Our Response: The FWS agrees that it cannot authorize incidental take under section 101(a)(5)(A) unless it first makes certain MMPA findings. However, FWS disagrees with the comments' ensuing conclusion, in part because it conflates the separate statutory standards governing (1) when ITRs may (or more precisely, must) be issued and (2) what an ITR (if issued) must allow.

When reviewing requests for ITRs, FWS makes MMPA findings that are inherently predictive in that they concern anticipated or expected takings from all future activities anticipated or expected to be conducted under the ITR (i.e., the total taking). Where FWS makes the requisite MMPA findings, it must issue an ITR. The ITR must then “allow . . . the incidental, but not intentional, taking by citizens engaging in that activity within that region. . . .” (section 101(a)(5)(A)(i)). The statutory language establishing the scope of incidental taking to be allowed in ITRs does not mention predictive take estimates.

FWS recognizes that estimates of impacts from future activities rarely align perfectly with the actual impacts from those activities once conducted. It is always possible that FWS may overestimate or underestimate impacts resulting from the specified activities whenever it reviews ITR requests and makes its MMPA findings. The comments here seem to argue for a mechanism to ensure that the scope of allowable incidental take does not exceed the scope of previously estimated incidental take. However, the

text of section 101(a)(5)(A) does not accommodate such a mechanism and Congress has already described in a separate provision what FWS must do if the specified activities cause unforeseen impacts.

Section 101(a)(5)(B) of the MMPA establishes the circumstances under which the permission to incidentally take marine mammals under subparagraph (A) must be withdrawn or suspended, either on an individual or class basis. One of these circumstances is where “the taking allowed under [the ITR] pursuant . . . is having, or may have, more than a negligible impact on the species or stock concerned” (16 U.S.C. 1371(a)(5)(B)(ii)). This fail-safe provision demonstrates Congress' awareness that the Secretary's predictive MMPA findings under subparagraph (A) might not account for the full extent of actual impacts from the specified activities, due to unforeseen circumstances or other reasons. The fact that a fail-safe provision that provides FWS a recourse in the event of unforeseen impacts is already built into section 101(a)(5)(B) further argues against reading an additional, unspoken precautionary mandate into the text of section 101(a)(5)(A).

The commenters' interpretation of what MMPA incidental take can be authorized pursuant to an ITR also results in misalignment between MMPA and Endangered Species Act (ESA; 16 U.S.C. 1531-1544) provisions concerning incidental take of ESA-listed marine mammals. Under the ESA's section 7 consultation provisions, where a biological opinion concludes that incidental take will result from the proposed action but would not violate the substantive standards of section 7(a)(2), the biological opinion must provide an incidental take statement (ITS) (16 U.S.C. 1536(b)(4); see also 50 CFR 402.14(i)). The ITS must specify, inter alia, reasonable and prudent measures to minimize the impact of the taking and set forth terms and conditions that the action agency or any applicant must comply with to implement those reasonable and prudent measures. Congress made clear that any taking “that is in compliance with the terms and conditions specified in [the ITS] shall not be considered to be a prohibited taking of the species concerned” (16 U.S.C. 1536(o)(2)). Nothing in the ESA or its implementing regulations allows (much less requires) the consulting agency to limit the scope of this exemption to correspond to the types or numbers of incidental takings that it anticipated in its biological opinion. The exemption plainly applies to all incidental taking that is in compliance with the terms and conditions of the ITS. This is important here because a further requirement applies where the consultation involves ESA-listed marine mammals, i.e., incidental take of marine mammals cannot be exempted under the ESA unless the taking is authorized pursuant to section 101(a)(5) of the MMPA (see 16 U.S.C. 1536(b)(4)(C)). Limiting the scope of an MMPA authorization in the manner suggested by commenters serves to preclude full implementation of the ESA's incidental take exemption provision. Under the commenters' interpretation, a citizen could be issued an MMPA section 101(a)(5) authorization, fully comply with the mitigation, monitoring, and reporting requirements of that authorization, fully comply with the terms and conditions of an ITS, and yet remain liable for certain incidental take under both the MMPA and ESA. This result does not square with the plain meaning of section 101(a)(5) of the MMPA or section 7(o)(2) of the ESA.

The commenters' interpretation also results in misalignment between the MMPA and a key ESA provision specific to polar bears. FWS's ESA section 4(d) rule for polar bears serves to generally prohibit take of polar bears but makes a key exception: “None of the prohibitions in Sec. 17.31 of this part apply to any activity that is authorized or exempted under the [MMPA]. . . provided that the person carrying out the activity has complied with all the terms and conditions that apply to that activity under the provisions of the MMPA. . . and [its] implementing regulations” (50 CFR 17.40(q)(2) and 73 FR 28306, May 15, 2008). Like the MMPA's ITR provisions, this 4(d) rule provision applies to categories of “activities” as opposed to the types or numbers of take anticipated by FWS in its predictive analyses of such activities. Here again, limiting the extent of allowable MMPA take in the manner suggested by the commenters serves to preclude full implementation of an ESA provision. In this case, declining to implement an “activity”- based approach when allowing incidental take under section 101(a)(5) undercuts FWS's intent in promulgating the polar bear 4(d) Rule. See FWS's ESA section 4(d) rule for polar bears which states (“This special rule under section 4(d) of the ESA aligns ESA incidental take provisions for polar bears with incidental take provisions of the MMPA and its implementing regulations” (73 FR 28306 at 28310, May 15, 2008).

There are also some important policy and enforcement considerations at play here. FWS's interpretation promotes fairness and regulatory certainty in that LOA recipients who conduct their specified activities in compliance with all ITR and LOA requirements (including the mitigation measures that ensure the least practicable adverse impact on marine mammals) remain insulated from legal liability where FWS's predictive estimates prove inaccurate. Also, in FWS's prior experience, where the extent of allowable incidental take is limited to correspond to FWS's predictive estimates, it can be very difficult for FWS to adequately support enforcement actions concerning any “extra” and thus unauthorized incidental takes. This is largely because incidental take can be difficult to detect. The March 9, 2026, proposed ITR touched on this issue in the context of polar bears when it explained that “ambient conditions common to the ITR area (e.g., darkness, snow, fog, and wind) can reduce the ability to observe polar bears (and impacts, if any) in the vicinity of industrial activities”, and certain types of potential impacts “would likely manifest after [the polar bears have] departed from the observable area . . .” (91 FR 11240 at 11252). Incomplete information concerning the extent of an LOA recipient's prior incidental taking makes it difficult for FWS to sufficiently establish that one or more “extra” and thus unauthorized incidental takings have occurred. The notion (which seems implicit in these comments) that FWS could reliably identify and enforce against all “extra” and thus unauthorized takings is undercut by these practical realities.

Comment (4): Several comments stated that issuing an ITR that allows all incidental take resulting from the specified activities is inconsistent with the conservation-based purposes of the MMPA.

Response: The FWS disagrees. Whether an incidental taking is authorized pursuant to an ITR, or remains prohibited under the MMPA's general take prohibition, is immaterial to the extent of any associated impact to the marine mammal itself. The only change concerns the legal liability of the person or entity that commits the take. Meanwhile, FWS does not believe that allowing all incidental taking resulting from specified activities conducted in accordance with ITR and LOA requirements incentivizes or otherwise results in more frequent or more severe taking. ITR and LOA requirements already ensure the least practicable adverse impact on the species or stock and its habitat, and any incidental taking that results from activities that

are not conducted in full compliance with ITR and LOA requirements remain prohibited and subject to enforcement.

Comment (5): Several comments expressed confusion stemming from the proposed ITR's inclusion of the word “Nonlethal” in the proposed title of a new Subpart J, despite other text indicating that all forms of incidental take could be allowed.

Response: FWS regrets any confusion; this word appears to have been accidentally carried over as an artifact from a prior iteration of the Beaufort Sea ITR. While no lethal take is expected to result from the specified activities, the proposed regulatory provisions did contemplate allowing all forms of incidental take that could result from the specified activities, and FWS received many comments on that aspect of the proposed ITR. The referenced use of “Nonlethal” is omitted from this final rule.

Comment (6): One comment suggested that allowing for all incidental take that results from the specified activities (not just the take that was anticipated when developing the ITR) is incompatible with FWS's implementing regulations which require analysis of the “total taking.”

Our Response: FWS disagrees. The comment quotes from the preamble to the 1989 final rule (54 FR 40338, September 29, 1989) that established FWS's implementing regulations. The referenced language calls for FWS to estimate the “total takings expected” from all the activities expected to be covered by the ITR. FWS has done so in its analysis here.

The comment then suggests that this preamble language, which concerns what is to be estimated during the rulemaking process, dictates that scope of incidental take that an ITR (if issued) can allow. There is no support for that conclusion in the preamble much less in the regulatory provisions themselves. FWS's implementing regulations address the required contents of ITRs at 50 CFR 18.27(e)(1). This provision (which essentially repeats the statutory provisions of section 101(a)(5)(A)(i)(II)) requires ITRs to set forth permissible methods of taking pursuant to the relevant activities but does not require ITRs to set forth permissible types or numbers of taking pursuant to those activities. The ensuing provision at 50 CFR 18.27(e)(1)(2) then provides that “. . . [a]s new information is developed, through monitoring, reporting, or research, the regulations may be modified, in whole or in part. . . .”. Implicit here is the recognition that if monitoring and reporting data indicates that unexpected degrees of take are occurring from the specified activities, FWS can reassess the ITR and, if appropriate, take corrective action.

The referenced preamble is instructive as to what the phrase “total taking” means in the ITR context. Relevant portions of the preamble explain:

Before the Service may allow a taking of marine mammals under the authority of section 101(a)(5) of the MMPA, it must find that the total taking expected from the specified activity will have a negligible impact on that species or stock. . . . In determining impact, the Service must evaluate the “total taking” expected from the specified activity in a specific geographic area. The estimate of total taking involves the accumulation of impacts from all anticipated activities that are expected to be covered by the specific regulations. In other words, the applicant's anticipated taking from its own activities is only one part of the story; the total takings expected from all persons conducting the activities to be covered by the regulations must be determined. (54 FR 40338 at 40340 and 40342, September 29, 1989).

In other words, “total taking” simply clarifies the scope of activities (i.e., all those “anticipated” or “expected” to be covered by the ITR) and potential taking (i.e., the “anticipated” or “expected” taking from those activities) to be considered when rendering predictive MMPA findings (e.g., “negligible impact”). It does not circumscribe the scope of incidental take that ITRs (if issued, after the requisite MMPA findings are made) must allow.

Comment (7): Several commenters stated the FWS must elaborate on the three potential interpretations of the MMPA presented in the proposed ITR, stating what they find to be the best reading of the MMPA. They state FWS appears to have given more weight to option 1, wherein all incidental take associated with the specified activities would be authorized. One commenter stated that should the FWS decide to use one of the other two options, it should publish a revised proposed ITR with an additional opportunity for public comment.

Our Response: The proposed ITR specifically alerted the public to this interpretive question and specifically requested public comment concerning the most appropriate resolution. FWS received and considered numerous public comments offering different perspectives on this issue. FWS then determined that the regulatory text of the proposed ITR already reflected the best reading of the statute. That text remains unchanged in the final rule, and there was no need to publish a revised proposed ITR.

Comment (8): The requester stated the final rule should clarify the LOA process and FWS should issue LOAs for activities within the scope and type of activities described in the Request without detailed data requirements or new model runs.

Our Response: Section 18.123 describes FWS's process for reviewing LOA requests. This process entails determining “whether the level of activity identified in the request exceeds that analyzed by [FWS] in considering the number of animals estimated to be taken and evaluating whether there will be a negligible impact on the species or stock and an unmitigable adverse impact on the availability of the species or stock for subsistence uses.” All specific projects and activities described in the Request and reflected in the requester's supporting materials are inherently within the “level of activity. . . considered by [FWS] . . .” in issuing this rule. FWS will thus grant LOA requests for such projects and activities without performing further, duplicative impact analysis. However, FWS will also consider LOA requests pertaining to other projects and activities so long as they would be conducted by an eligible entity within the specified geographical region and are of same classes and types of activities considered in this Rule. These projects or activities were not considered by FWS in issuing this rule, and additional analysis would be required to determine if the level of activity associated with the LOA request would exceed the level of activity already analyzed by FWS. In making this determination, FWS would consider the scope of activities described in the LOA request under review in combination with the scope of any contemporaneous activities covered by existing LOAs issued pursuant to this rule. The appropriate method for conducting such analyses may vary depending on the specific factual circumstances presented, and FWS will not require detailed data requirements such as activity footprint “shapefiles” and associated occupancy rates and remains committed to making decisions on LOA requests expeditiously. Clarifying language has been added to 50 CFR 18.123.

Comment (9): The requester noted that there is language in the preamble that suggests LOA requests will be evaluated against the total taking allowable under the ITRs and asserts that this is inconsistent with a proposed regulatory provision and FWS's stated approach to authorize “all types of incidental take that result from the specified activities.” The requester

stated the FWS should revise the preamble text and clarify that this Rule does not set a “total taking allowable” number.

Our Response: The referenced preamble text and the FWS implementing regulation it refers to (50 CFR 18.27(f)(2)) concern the standard that FWS applies when deciding whether to issue a requested LOA. FWS's stated approach of authorizing “all types of incidental take that result from the specified activities” pertains to a different topic, i.e., the extent of incidental take that would be allowable under the LOA, if issued. The LOA review process described by Sec. 18.123 of this Rule squares the regulatory phrase “total taking allowable” as it appears in 50 CFR 18.27(f)(2) with the best reading of the statute by ensuring that the scope of activities and estimated impacts associated with approved LOA requests will not exceed what was considered in this Rule.

Comment (10): The requester stated the FWS should clarify the potential length of LOAs issued under the proposed ITR, specifically highlighting the FWS should allow multi-year LOAs.

Our Response: The FWS will process and may approve LOA requests for periods of up to 5 years under this Rule.

Comment (11): The requester suggested a revision in the preamble text that describes specified activities to include “. . . some projects not explicitly described may occur during the ITR period. All activities will be of the same classes and types of activity described in the Request as required by the MMPA.”

Our Response: The FWS will process and may approve LOA requests pertaining to projects not explicitly described in the Request or preamble text, so long as the project activities described therein would be conducted by an eligible entity within the specified geographical region and are of same classes and types of activities considered in this Rule. In lieu of modifying the “Description of . . . Specified Activities” section as between the proposed and final rules, the FWS has retained the original preamble text, which adequately accommodates LOA requests of this nature.

Comment (12): One commenter stated FWS cannot defer identification and analysis of specified activities to the LOA stage, as this would call into question the determination that based on the best available scientific evidence, the total taking from the specified activities included in the request are not reasonably expected to, and are not reasonably likely to, adversely affect the species or stock through effects on annual rates of recruitment or survival. They argued that the lack of opportunity for public review of proposed LOAs demonstrates that no significant new information or analysis was contemplated to occur at the LOA stage.

Our Response: The FWS has not proposed a mechanism wherein MMPA findings are deferred to the LOA stage. This rule fully considers the specified activities described in the Request and makes the requisite MMPA findings. FWS and reviewing courts have long recognized that specified activities may be described broadly at the ITR stage. As we have explained above, FWS's subsequent reviews at the LOA stage will ensure that sufficient project-specific information was considered at the ITR or LOA stage before an LOA is granted, and that levels of taking will be consistent with the findings made in this rule.

Comment (13): One commenter suggested all activities eligible for LOAs must be fully analyzed in the ITR to be consistent with the MMPA and its implementing regulations. The commenter recommends that the FWS rejects deferring identification or analysis of specified activities, take estimates, and impacts assessment to the Letter of Authorization (LOA) stage.

Our Response: This rule fully analyzes the potential impact of the specified activities described in the Request. FWS's analysis considered the descriptions of activities provided in the Request as well as the requester's supporting materials, which included spatial files reflective of certain ongoing and planned activities. The requester has stated some projects not explicitly described may occur during the ITR period, and all additional activities will be of the same classes and types of activity described in the Request. Should an entity request an LOA for projects not previously analyzed, the FWS would evaluate the request in the manner described in section 18.123 of this rule.

In a broader sense, the commenter seems to question FWS's use of a two-step approach (i.e., ITRs and then LOAs) to authorizing incidental take pursuant to section 101(a)(5)(A) of the MMPA. FWS has implemented regulations utilizing this approach for decades and it has proven successful in authorizing incidental take where warranted, imposing mitigation measures that effect the least practicable adverse impact on marine mammals, and ensuring that impacts to marine mammal species and stocks remain negligible.

Comment (14): Commenters stated that the proposed ITR is not clear on what entities will be eligible for LOAs, as it states that AOGA requested eligibility for LOAs be conditioned on active AOGA membership but also states that AOGA requested regulations be applicable to AOGA members and a non-member company that have applied for the regulations.

Our Response: The entities listed referenced in Sec. 18.122(b) of this final rule, as well as their corporate affiliates, respective contractors, subcontractors, partners, owners, co-lessees, designees, or successors-in-interest will be eligible to apply for LOAs.

Comment (15): Some commenters noted that in the requester's petition, AOGA requested that eligibility under the proposed ITR be conditioned upon active AOGA membership. The proposed ITR extends eligibility slightly, but still restricts it to AOGA members, 8 Star Alaska, LLC, and their respective corporate affiliates and contractors, subcontractors, partners, etc. The commenter then urges the FWS to remove all eligibility requirements related to requiring membership in AOGA or other industry groups.

Our Response: The entities listed referenced in Sec. 18.122(b) of this final rule, as well as their corporate affiliates, respective contractors, subcontractors, partners, owners, co-lessees, designees, or successors-in-interest will be eligible to apply for LOAs. If a party seeks an incidental take authorization for an activity not included in the specified activities, they can apply for a separate take authorization.

Comment (16): Several comments focused on who may apply for an LOA. One commenter stated that the FWS providing more flexibility in which companies are eligible for LOAs would support conservation and compliance objectives such as reducing the likelihood that Industry activities would proceed without incidental take coverage due to eligibility barriers, ensure all operators are subject to consistent mitigation, monitoring, and reporting requirements, and more flexible eligibility requirements would not pose any risks to the MMPA goals or the FWS's authority to process LOAs. One commenter stated that the FWS restricting LOA eligibility to entities based on their membership with the requester raises administrative and legal concerns such as unequal treatment of operators conducting similar activities in the same geographic region, potential exclusion of entities that are willing to comply with ITR requirements but elect not to join a private association for independent business or governance reasons or who lack a qualifying relationship to an eligible member entity, and unclear implications about

continuity of authorization, transferability, and enforcement if membership with the requester changes during the ITR period. They further state that the FWS's ability to ensure effective mitigation and monitoring requirements relies on clearly enforceable regulatory conditions and not whether an entity has a membership with the third- party organization. Finally, a commenter stated that the FWS restricting LOA eligibility to entities that have membership or a qualifying relationship with the requester conflicts with the MMPA statute of authorizing incidental take by specified activities within a defined geographic region, subject to findings regarding small numbers, negligible impact, and mitigation.

Our Response: The entities listed referenced in Sec. 18.122(b) of this final rule, as well as their corporate affiliates, respective contractors, subcontractors, partners, owners, co-lessees, designees, or successors-in-interest will be eligible to apply for LOAs. As is reflected in the rule, FWS will not require confirmation of membership in AOGA prior to LOA issuance. If a party seeks an incidental take authorization for an activity not included in the specified activities, they can apply for a separate take authorization.

Comment (17): One commenter stated that, as noted in the proposed ITR's small numbers discussion, the population survey data used to calculate the minimum population estimate (Nmin) for the SBS stock of polar bears is over 10 years old. They recommended the FWS address if and how the age of the survey data impacts their confidence in their calculation of potential biological removal (PBR) as 14 bears.

Our Response: We acknowledge that having a more recent population estimate would be ideal; unfortunately these data are difficult to obtain. The FWS is currently working with the U.S. Geological Survey (USGS) and partners to develop an updated population estimate for the SBS stock. The FWS is confident that its calculation of PBR as 14 bears in the SBS SAR represents the best available scientific evidence at this time. Best available science also indicates the population decline that was noted in the mid-2000s appears to have stabilized (Bromaghin et al. 2021).

Comment (18): One commenter stated FWS should reassess its small numbers and negligible impact determinations using the minimum population estimate from the stock assessment reports (Nmin). They state Nmin is more appropriate for FWS to use the minimum population estimate (Nmin) for SBS polar bears in its small numbers determination rather than a population point estimate because the Nmin value provides reasonable assurance that the stock size is equal to or greater than that estimate.

Our Response: The MMPA does not require FWS to use Nmin when calculating small numbers. While Nmin is a valuable conservation metric, we find the population estimate to be the most appropriate denominator when calculating the percentage of the stock that is estimated to be subjected to incidental take as a result of specified activities.

Comment (19): One commenter recommended the FWS commission updated population surveys before finalizing the rule because the most recent population data is nearly a decade old.

Our Response: The FWS strives to obtain updated abundance estimates for polar bear stocks under its jurisdiction every 10 years. There has been an ongoing effort in the past year to use more recent data to update the stock's abundance estimate. The FWS anticipates these results to be available by the end of 2026. However, the estimate the FWS uses for the current ITR represents the best available scientific evidence available at this time.

Comment (20): One commenter recommended the FWS require meaningful, independently verified population recovery benchmarks before any authorization is issued.

Our Response: The FWS declines to implement this recommendation as the best available scientific information indicates the population decline of the SBS Stock in the mid-2000s has appeared to stabilize (Bromaghin et al. 2021). Separate from this rulemaking, the FWS, with a wide variety of partners from non-governmental conservation organizations to oil and gas industry representations, created and published the polar bear conservation management plan which presents certain benchmarks for polar bears.

Comment (21): Commenters argued that any additional industrial activity or harassment of polar bears in the Southern Beaufort Sea stock would be unacceptably harmful since the population is facing conservation pressure from declines in sea ice, which force polar bears to spend more time on shore. Commenters cited declines in foraging opportunities, body condition, cub survival, and overall numbers of Southern Beaufort Sea polar bears, sometimes highlighting specific population estimates (e.g., estimates from the 1970s, 1990s, a 2021 stock assessment report, and a 2025 draft stock assessment report) or a period of low cub survival (2004-2007). Highly similar comments submitted by many individuals stated that peer-reviewed research has shown that the polar bear population in Alaska is likely lower than it has been at any point since the passage of the MMPA in 1972, and that further declines can be expected in the future if protections are not strengthened. Commenters stated that strengthened protections, including minimizing disturbance and protecting critical habitat, are needed to support recovery and prevent further population declines, given the vulnerability of the population. Some commenters similarly argued that the Pacific walrus population is already vulnerable due to sea ice declines and stronger protection is needed, as any additional industrial disturbance--especially noise, vessel traffic, and habitat disruption--exacerbates existing pressures.

Our Response: The FWS has carefully assessed anticipated impacts of the specified activities on the Southern Beaufort Sea stock of polar bears and the Pacific walrus population in light their current statuses and the baseline of existing impacts and concluded that incidental take associated with specified activities will have no more than a negligible impact on these populations. Our analytical methods account for the increased industrial footprint and for changing space use patterns (e.g., more polar bear denning on shore and summer land use) associated with sea ice declines. The ITR includes mitigation and monitoring requirements that will limit potential disturbances and impacts to habitat. Commenters are correct that the SBS stock has declined since the MMPA was passed. Currently, however, existing evidence suggests the subpopulation's decline has stabilized and survival rates are much higher than during the early 2000s (Bromaghin et al. 2021). Some commenters incorrectly suggested that the lower population estimate included in the 2025 SBS draft stock assessment report indicated a biological decline in the population since the publication of the 2021 stock assessment report, but that is not the case as the updated population estimate was lower due to a change in the boundary that defines the SBS stock (movement of the eastern boundary of the SBS stock in Canada to the west).

Comment (22): Some commenters stated that the relatively late age of sexual maturity, slow reproductive rates, and heavy investment in each reproductive event for polar bears and walruses make populations vulnerable to impacts from oil and gas activities. Regarding polar bears, commenters

stated that any additional cub mortality resulting from den disturbance could harm recruitment and population trends due to polar bears' slow reproductive cycles, low numbers of dens, and low cub survival rates. One commenter stated that in light of the slow reproduction in polar bear and walrus life histories, the FWS should explain with much greater rigor how additional disturbance to denning, feeding, or maternal care can still be considered negligible.

Our Response: The FWS considered the relatively late age of sexual maturity, slow reproductive rates, and heavy investment in each reproductive event, population dynamics, among other relevant considerations, of SBS polar bears and walruses when analyzing potential impacts and developing its MMPA findings. The FWS does not anticipate lethal take of polar bears or Pacific walruses as a result of the specified activities. The SBS polar bear stock exhibits marked natural interannual variation in cub survival rates (Bromaghin et al. 2021) which are much higher than the anticipated effect of potential disturbance. Even with this variation, the most recent and best available science has supported that the population has stabilized (Bromaghin et al. 2021). Impacts to Pacific walrus are estimated to be limited to temporary behavioral disturbances of a very small portion of the population that may be encountered outside their typical home range.

Comment (23): One commenter states that the FWS must ensure that take authorized under individual LOAs already issued does not exceed the extent of take FWS anticipates or authorizes in the proposed ITR.

Our Response: LOAs already issued will not carry over into this ITR period. The FWS will evaluate new requests for LOAs under the provisions of this ITR.

Comment (24): Commenters stated that climate change and increases in human presence and activity are reducing habitat available to polar bears and walruses, leading to increases in encounters that are potentially dangerous.

Our Response: We account for the increased industrial footprint and changing space use patterns (e.g., more denning on shore, more summer land use) in our analytical methods, as described in the SUPPLEMENTAL INFORMATION.

Comment (25): Comments stated that walrus populations are vulnerable to negative impacts from the rule due to their increased use of large coastal haulouts in association with reductions in sea ice. They stated that walruses at haulouts, particularly calves, face risks of disturbance-related mortality from stampedes and trampling triggered by aircraft noise, industrial activity, or human presence, in addition to risks of harmful behavioral disturbance.

Our Response: We agree that large coastal haulouts can be particularly vulnerable, however walrus are extralimital (i.e., outside their range) in the ITR area and are only encountered in small numbers. There are no known walrus haulouts in the ITR area. However, should haulouts occur, LOA holders must maintain a mitigative offset in vessels or in aircraft.

Comment (26): One commenter stated that at a time when scientists warn of further declines and potential local extinctions in some polar bear subpopulations, Federal policy must prioritize habitat protection and disturbance minimization, not industrial convenience.

Our Response: Consistent with MMPA requirements, the final rule sets forth, among other things, means of effecting the least practicable adverse impact on SBS polar bears, walruses, and their habitat.

Comment (27): Several commenters recommended FWS conduct a more robust cumulative impact analysis accounting for prior, ongoing, and proposed authorizations, stating that considering individual authorizations in isolation obscures long term effects and is inconsistent with the MMPA. One commenter stated the FWS must analyze the consequences of multiple disturbances from all identified human activities with the potential to cause incidental take on individual polar bears and the SBS stock overall. One commenter stated the FWS must fully analyze the impacts of past take in considering how future annual, unknown levels of take authorized for the same population of bears pursuant to the proposed ITR will impact the SBS stock.

Our Response: FWS considered the impacts of the specified activities in light of the baseline of existing impacts, which includes impacts from ongoing, authorized activities.

Comment (28): One commenter stated that research shows extended ice-free fasting periods and decrease in sea ice extent are causing energetic stress to polar bears, and additional energetic costs from disturbance could therefore affect survival or reproduction.

Our Response: The FWS thoroughly discusses the potential effects of climate change on polar bears, including changes in sea ice extent, in the supplemental information to the proposed ITR. FWS considered this information when conducting its analysis and reaching its MMPA findings. We acknowledge that changes in the sea ice may lead to longer swims between the ice edge and the shore. However, we do not anticipate human-polar bear interactions to increase in severity during the 5-year ITR period. We anticipate polar bear reactions to continue to be short- term, temporary disruptions of behavior.

Comment (29): Some commenters stated that because the steep decline in the SBS polar bear population between 2000 and 2010 was driven by low cub survival, maximizing cub survival by preventing den disturbance is critical. One commenter specifically stated that the FWS failed to address the best available scientific information showing cub survival rates' influence on population trajectory, citing a study that showed high influence of cub survival on SBS polar bear population trajectory from 2002-2005.

Our Response: Bromaghin et al. (2021) noted that after the decline, survival rates increased to much higher levels. Further, the SBS polar bear stock exhibits marked natural interannual variation in cub survival rates (Bromaghin et al. 2021) which are much higher than the anticipated effect of potential disturbance from the specified activities. Even with this variation, the most recent and best available science has supported that the population has stabilized (Bromaghin et al. 2021). The FWS has analyzed the possibility of cub mortality and the extent to which it would affect the SBS stock. Preventing den disturbance is indeed important and these ITRs require various mitigation measures to effect the least practicable adverse impact on denning polar bears.

Comment (30): A commenter stated that one reason the loss of one additional cub due to the specified activities is very significant to the long-term survival of the stock is the greater importance of land dens to the population given their greater likelihood of yielding cubs that survive and reproduce themselves. They stated the FWS failed to acknowledge the particular importance of land-denning bears.

Our Response: While it is true that Rode et al. (2018) have found support for a higher rate of denning success in land-based dens, the available science does not provide a basis for discounting the value of cubs born in sea ice dens. Rode et al. (2018) indicated the likely reason for increased litter survival of land-based dens was later average emergence dates. We incorporated this effect into our model as empirical data

from land-based dens were used to create denning parameters in our analysis.

Comment (31): One commenter stated that in the proposed ITR, the FWS is “glossing over a substantial probability of injury or death rather than reckoning with the impacts of such an injury or loss on the stock.”

Our Response: As the FWS has stated in our negligible impact determinations, we do not anticipate loss of a cub through den abandonment. Further, the litters that experience Level A take over a 5-year period are anticipated to have an average 12.6 percent decrease in survival probability. This change in survival probability is not anticipated to have more than a negligible impact on annual rates of recruitment for the SBS stock. The stock exhibits marked natural interannual variation in cub survival rates (Bromaghin et al. 2021) which are much higher than the anticipated effect of potential disturbance. Even with this variation, the most recent and best available science has supported that the population has stabilized (Bromaghin et al. 2021).

Comment (32): One commenter stated that FWS cannot assume the SBS population is stable because the most recent draft stock assessment shows a continued decline in population estimates.

Our Response: The commenter is incorrect. The draft stock assessment report for the SBS (citing Bromaghin et al. 2021) found evidence that the decline that occurred in the early 2000s had likely abated and that the population had appeared to stabilize. As noted in the stock assessment report, the lower number (819) reported is not the result of a decline in the population, but rather a boundary shift in Canada that reduced the overall footprint of the SBS and thus the number of bears considered to be part of the subpopulation.

Comment (33): Commenters stated that take considered in the rule does not meet the definition of incidental take, which is defined as takings which are infrequent, unavoidable or accidental. One commenter stated that when disruption is anticipated, modeled, and permitted in advance, it is not incidental in any meaningful sense, and another stated that “incidental” is not a standard the FWS can uphold or monitor.

Our Response: In accordance with the MMPA, this rule may only authorize incidental take. The fact that future incidental take can be predicted does not mean it is not infrequent, unavoidable, or accidental. FWS's definition of “[i]ncidental, but not intentional, taking” makes clear that such taking need not be “unexpected.” The FWS can and does make distinctions between take that is intentional (such as take that occurs when polar bears are intentionally deterred from an area to promote bear and human safety) and incidental take.

Comment (34): One commenter stated that the FWS does not explain what it is using for a threshold to characterize a probability as low, nor does it explain how the 45 percent likelihood of lethal take occurring within the 5-year period, which was estimated by the FWS predictive model for denning bears, compares to this threshold.

Our Response: The qualifier (“low”) does not involve any thresholds and was used generally to characterize FWS's overall assessment of the likelihood of lethal take and was not necessary to support our assessment. In light of the probability from the FWS predictive model of one or more lethal takes occurring either annually (0.12) or in 5-year aggregate (0.43), along with a median value of 0 for either time period, we do not anticipate the specified activities will result in lethal take of polar bears.

Comment (35): The commenter believes that the FWS is not properly considering the MMPA's definitions of take specifically with regard to the phrasing of “potential” threat of take. Therefore, the FWS is overlooking the potential for lethal take of cubs when characterizing impacts as Level A harassment.

Our Response: The FWS's application of the term Level A harassment is appropriate. The MMPA defines “take” to include, among other things, “harass” and “kill.” The MMPA further defines two types of harassment (Level A and Level B) and specifically incorporated the concept of “potential” into those definitions. The MMPA does not define “kill.” The Service therefore applies the ordinary meaning of the word “kill,” which is to deprive of life or cause the death of, not to potentially deprive of life or cause the death of.

The limited numbers of Level A harassment contemplated by this rule are estimated to entail an average 12.8 percent marginal reduction in litter survival probability, which represents a possibility of death, but not a certainty or even likelihood of death. This is not an adequate basis for assuming lethal (i.e., “kill”) take. As previously noted, the existing dataset contains no definitive evidence that an Industry activity caused a maternal den abandonment that in turn resulted in cub mortality. We also note that such an assumption would be unworkable in the enforcement context in terms of meeting the applicable burden of proof.

Comment (36): One commenter states that the FWS improperly deems impacts that may result in cub mortality as Level A harassment instead of lethal take. They specifically state the science shows impacts during late denning could result in cub death.

Our Response: The FWS's application of the term Level A harassment is appropriate. The MMPA defines “take” to include, among other things, “harass” and “kill.” The MMPA further defines two types of harassment (Level A and Level B) and specifically incorporated the concept of “potential” into those definitions. The MMPA does not define “kill.” The Service therefore applies the ordinary meaning of the word “kill,” which is to deprive of life or cause the death of, not to potentially deprive of life or cause the death of.

The limited numbers of Level A harassment contemplated by this rule are estimated to entail an average 12.8 percent marginal reduction in litter survival probability, which represents a possibility of death, but not a certainty or even likelihood of death. This is not an adequate basis for assuming lethal (i.e., “kill”) take. We also note that such an assumption would be unworkable in the enforcement context in terms of meeting the applicable burden of proof.

Comment (37): One commenter stated the Request included estimates of potential take that used FWS methodology, that these estimates are different than those presented by FWS in the proposed ITR, and that FWS must explain and address this discrepancy.

Our Response: The estimates referenced by the commenter are not those presented by AOGA as their likely estimate of take. They were part of AOGA's discussion of FWS predictive modeling methodology and appear to have used input values that are no longer used by FWS. There is no statutory or regulatory provision that requires any applicant to use any FWS predictive model.

Comment (38): One commenter stated that the FWS failed to consider some additional level of lethal and non-lethal take in the ITR that is likely to occur over the ITR period.

Our Response: The take estimates considered by the FWS for this ITR account for the potential of lethal and non-lethal take to occur due to the applicant's proposed activities based on the best available scientific information. Assuming additional take beyond what was analyzed would be unduly speculative.

Comment (39): Commenters stated that previous Beaufort Sea ITRs (e.g.,

rules finalized in 2016 and 2021) did not permit Level A harassment and that permitting Level A harassment now is unwanted regulatory rollback.

Our Response: In consideration of the best available scientific information, the FWS has updated our approaches to estimating and authorizing Level A harassment since 2021. Level A harassment was authorized under the revised Beaufort Sea ITRs issued in 2025. We disagree that authorizing similar take under this ITR amounts to regulatory rollback, and is consistent with the requirements of the MMPA.

Comment (40): One commenter states that the FWS cannot lawfully issue an ITR or LOA that allows unintentional, incidental taking, including undefined amounts of unanticipated lethal take, when the ITR findings rely on the assertion that lethal take is not sufficiently likely to warrant consideration in the findings. The commenter further stated their belief that this approach violates the MMPA, is beyond the Service's authority, and is arbitrary and capricious.

Our Response: FWS specifically considered the potential for lethal take in its analysis and associated findings, including the negligible impact finding. A conclusion, based on detailed analysis that a potential consequence is not anticipated to occur should not be mistaken for a conclusion the issue does not warrant analysis.

Comment (41): The requester has stated that the FWS statement that the 43 Level B harassment events reported by two operators in calendar year 2025 are a standard deviation above the annual Level B take requested is a mischaracterization because the request did not seek take authorization on a certain number of takes/year basis.

Our Response: We have included some clarifying language to more accurately reflect the take authorization request and note that the purpose of the statement at issue was to illustrate inter-annual variability in polar bear use of the ITR area, not to characterize how take was estimated or requested.

Comment (42): The requester stated that FWS's modeling is unreliable and model results should not be used to support the rule. The requester further suggested that FWS disavow the notion that an ITR request could be disapproved based on modeled take estimates.

Our Response: The FWS's model-derived take estimates, like requester's take estimates, support issuance of the requested ITR. It is beyond the scope of this rulemaking to delineate the potential basis for disapproving a request for an ITR.

Comment (43): The requester states that any discussion of the FWS predictive estimates in the final rule should clarify (1) the results and associated determinations are not likely to occur, (2) they are not intended to be used as a regulatory measure, and (3) they are not necessary to support the final rule.

Our Response: The FWS has made a number of revisions that help provide context concerning the nature of FWS's methodologies and appropriate consideration of resulting take estimates.

Comment (44): One commenter states that the weight of evidence approach employed by the requester and summarized in the proposed ITR is a farce because the most heavily weighted line of evidence is from AOGA's own field observations and LOA database. The commenter further claims that the FWS is leaving industry to police themselves.

Our Response: AOGA is required by the MMPA to submit estimates of likely take in their Request. The FWS clearly explained their rationale for considering both sets of estimates in the proposed ITR (91 FR 11240 at 11251, March 9, 2026).

Comment (45): One commenter stated the FWS acknowledges what appears to be skepticism regarding the estimates presented in the request and does not go on to explain its rationale for relying on the estimates in the Request.

Our Response: It is unclear what part of the proposed ITR suggests skepticism to the commenter. It was not our intent to express skepticism of a particular analytical approach to estimating take. We clearly explain our rationale for considering both sets of estimates in the Rule.

Comment (46): One commenter stated that before finalizing the proposed ITR, the FWS should provide clearer quantitative estimates of anticipated take levels.

Our Response: The FWS provided all the quantitative estimates of anticipated take levels as part of the proposed ITR and also published all of the code and model output that these results were derived from.

Comment (47): One commenter urged the FWS to adopt the take estimates produced by its own predictive model rather than AOGA's “weight of evidence” methodology.

Our Response: We considered both estimates along with qualitative factors in our analysis regarding the required MMPA determinations.

Comment (48): One commenter stated FWS should complete a retrospective analysis of polar bear observations and takes reported by Industry operators over several decades to more accurately predict future incidental take through an objective, transparent, scientifically rigorous, and uniformly applicable process.

Our Response: The FWS has considered predictive estimates of take from both the requester and FWS predictive model estimates which both incorporate a retrospective analysis of polar bear observations by Industry operators over several decades. The FWS considers the observational data a key component of the best available scientific information. As such, we continue to analyze observations alongside additional scientific information to inform and refine its predictive model estimates, as is explained in the proposed ITR and associated supplemental information.

Comment (49): One commenter stated that the ITR request relies on past opportunistic observations to predict future take without providing details on how past activities compare to potential future projects, some of which are not yet identified. They stated that to justify using past take levels as a predictor of future take, the requester would need to show that exposure levels are likely comparable between them by documenting intensity, methods, locations, and timing of past and proposed activities, and these data are not provided in the proposed ITR or in the ITR request.

Our Response: We considered both the estimates provided in the Request and our own independent predictive take analysis that resulted in the best available scientific evidence which included, but was not limited to, the past records of polar bear observations discussed in the request. The FWS considered additional information, such as the geospatial files referenced by AOGA in their request.

Comment (50): One commenter states that while the FWS should explore and assess different approaches to estimating potential take, the FWS has an obligation to determine which method and associated take estimate is based on best available scientific information and present that to the public.

Our Response: The MMPA requires the FWS to utilize best available scientific evidence when developing MMPA findings; however, nothing in the MMPA requires the FWS to quantify take estimates or identify a single, best methodology for estimating take or making MMPA findings. For instance, prior iterations of the Beaufort Sea ITR have made determinations based on both qualitative and quantitative factors.

Here, the FWS considered both sets of take estimates along with qualitative factors in making its MMPA findings.

Comment (51): One commenter stated that the proposed ITR and EA are unclear in how the FWS, when making required determinations, applied take estimates generated by its quantitative predictive methods or applied take estimates provided in the ITR request. The commenter stated it is implied the FWS relied on the requester's incidental take estimates to make determinations, which is problematic because there is evidence these values are underestimates.

Our Response: Each set of take estimates supported FWS's preliminary MMPA findings in the proposed ITR, and support FWS's findings in the final rule.

Comment (52): One commenter suggests the proposed ITR is vague, does not give enough guidelines or estimates of what may occur, and focuses on the applicant's own methodology.

Our Response: We disagree. We considered both the estimates provided in the Request and its own independent predictive take estimates alongside a number of qualitative factors, incorporating the best available scientific evidence. The preamble contains a summary of both sets of quantitative estimates with the details of each methodology available in the requester's petition and the FWS Supplemental Information.

Comment (53): One commenter claims the proposed ITR is an example of regulatory capture, suggesting a linkage between fossil fuel lobbying and the current administration. The commenter recommends that all industry communications related to the proposed ITR be disclosed and all lobbyist-written language be excluded from the final rule.

Our Response: During the ITR process, the FWS must review and speak to the Request we have received. We must also invite, consider, and respond to public comments, and in this process we must be open to receiving comments from all members of the public, regardless of status. As such, we may be required to present and respond to comments from lobbyists. When information from the requester or other external parties is included in FWS documents, we clearly attribute the source.

Comment (54): The requester states the FWS should explain why the result of their analysis to understand the frequency of incidental take preceding intentional take is different than those in their own request.

Our Response: The comment references an analysis that FWS conducted prior to and separate from its consideration of the Request at issue in this rulemaking. The decision rules and overall conclusions of that referenced analysis were explained in the Supplemental Materials provided in the docket associated with the proposed ITR. Meanwhile, the Request reports its own conclusions concerning this issue. Rather than focus on differences between the take estimates provided in the Request versus those produced via FWS's predictive modeling, this rulemaking considers both sets of estimates when rendering MMPA findings.

Comment (55): The requester suggests the denning model does not accurately reflect real-world conditions or animal reaction to oil and gas activities.

Our Response: The FWS uses the best available science which is largely collected from polar bears in the SBS stock thus they are representative of bears in the region. As previously explained, the FWS considers qualitative and quantitative considerations and data, the Request's estimates of likely incidental take, and FWS predictive model results as the “best available scientific evidence” used to support the determinations in this ITR. The FWS has been explicit about their analytical assumptions and why any assumptions are needed.

Comment (56): The requester and one commenter provided several criticisms of the FWS predictive model for estimating potential take of denning polar bears. They stated the model was inaccurate, seriously flawed, and resulted in intentional overestimates of take that far exceed historical data. They state the model has not been validated, and its use in MMPA authorizations does not meet the standard set by E.O. 14303, “Restoring Gold Standard Science” because it undermines required agency transparency due to the use of assumptions and disregard of available data due to uncertainty. They also state that use of the model will unnecessarily constrain industry by decreasing oil and gas activities and adding costs and mitigation measures. Finally, they conclude estimates generated using the model should not be used in the FWS's negligible impact determinations.

Our Response: We acknowledge the criticisms of our modeling methodologies and results from various stakeholders and appreciate that all predictive analyses and models have limitations. We further recognize that our predictive models contribute to annual take estimates that have exceeded (and are expected to exceed) the number of events where biological consequences indicative of take are observed to result from Industry activities. One reason is that the predictive models are based on annual activities forecast for 5 years of Industry activity in advance, and the annual activities, and thus the corresponding take, has historically been of a smaller scope than that analyzed. Additionally, the predictive models are one way to account for any unobserved take and account for certain acts which have “the potential to” result in certain biological consequences. Accounting for unobserved take and “the potential for” certain consequences involve a level of scientific uncertainty. When addressing uncertainties, the models incorporate assumptions that attempt to strike a balance between potential overestimates and potential underestimates, but, overall, are more likely to contribute to an overestimate, rather than an underestimate, of the incidental take of polar bears.

The FWS recognizes the continuing uncertainty about the scope of unobserved take and “the potential to” result in certain biological consequences, because these components, definitionally, do not avail themselves of an opportunity to evaluate the estimates against observations. Therefore, in evaluating the Request, we have avoided giving our predictive models, and quantification of take in general, undue weight in evaluating the Request. More specifically, we considered two separate predictive methodologies and resulting sets of take estimates, as well as qualitative factors, when making our small numbers and negligible impact determinations. The FWS thus did not rely solely on any single source of numerical estimates.

The FWS recognizes the importance of transparency required by E.O. 14303, “Restoring Gold Standard Science” and has sought to provide transparency in reporting assumptions of its predictive modeling, as well as providing all of the model code, input, and rationale provides transparency. In any scientific endeavor, control rules must be established to direct the inclusion of data, so incomplete information does not bias the results. We have explained our control rules to clearly communicate why certain data may not be incorporated into our analyses. Using different control rules will lead to differing data included in estimates using the same model.

The FWS does not agree that consideration of its predictive model results unnecessarily constrains oil and

gas activities through increased costs and mitigation measures. Any restrictions or mitigation measures imposed by ITRs stem from implementation of the MMPAs requirement that an ITR sets forth the means of effecting the least practicable adverse impact on the stock. This requirement applies to all ITRs, regardless of the extent of take estimated or how take was estimated (or whether take was quantitatively estimated at all). In implementing this requirement, we do not require mitigation measures that are so restrictive as to make specified activities unduly burdensome or impossible to undertake. Requesters are afforded an opportunity to comment on these issues when FWS-proposed mitigation measures are identified in proposed ITR.

As noted in the preamble, in light of the inherent uncertainties and challenges with quantifying take and the disparate results produced through different methodologies, the FWS considers its predictive model estimates alongside the requester's predictive estimates, which it derived from a distinct methodology, and various qualitative factors.

Comment (57): The requester urged FWS to provide a more “complete and balanced discussion” of the methods presented in the Request in the final rule.

Our Response: The FWS has updated the preamble of this rule to provide a more thorough summary of the requester's methods and take estimates alongside FWS's predictive modeling methods and results.

Comment (58): Commenters stated that greater numbers of SBS polar bears are projected to spend more time on shore in the vicinity of oil and gas activities in the future (and cited a proportion of the SBS stock estimates to spend prolonged period onshore by the year 2040), meaning current mitigation measures may not be sufficient and take numbers may be underestimated in the proposed ITR.

Our Response: The FWS disagrees that the mitigation measures are not sufficient and disagrees that the predicted take numbers are underestimated. The take estimates considered adequately account for the anticipated number of polar bears expected to utilize onshore areas within the 5-year ITR period that will expire in 2031. We continue to use and consider the best available scientific information.

Comment (59): Several commenters stated the FWS appears to have conducted its predictive modeling analysis and subsequent small numbers and negligible impact determinations using speculative spatio-temporal data, data from similar projects, or with a general lack of specificity.

Our Response: FWS requested and AOGA supplied a geodatabase that was used to conduct our predictive modeling estimates. The geodatabase included location, occupancy rate, and start/end dates for future proposed activities. We supplied the public with our analyses, which included these files, incorporated in an R environment datafile, in the docket of the proposed ITR.

Comment (60): One commenter stated that FWS estimates concerning Level A harassment indicate a 95 percent probability that oil and gas activities on the North Slope could be lethal to polar bears over a 5- year period.

Our Response: FWS's predictive take estimate indicates the greatest annual simulated probability of lethal take throughout the ITR period is 0.12, and the aggregated probability of potential lethal take over a 5-year period is 0.43. The median number of lethal takes in both the annual and 5-year aggregated modeling outputs was zero. Our predictive take estimates do estimate a high probability of one or more Level A harassments occurring, and that Level A harassment of cubs in the late denning and post-emergence period entail some (albeit limited) possibility of survival consequences to affected cubs. The extent to which these potential impacts could affect the SBS stock was considered in the proposed ITR and is considered in the final rule. We also considered AOGA's estimates for the aggregated probability of lethal take throughout the 5-year period of 0.000 and of Level A take of 0.103.

Comment (61): One commenter states that the proposed ITR only considers to possibility of Level A harassment or lethal take occurring to cubs and arbitrarily fails to consider the possibility of Level A harassment or lethal take occurring of an adult female polar bear while denning.

Our Response: The FWS has detailed the information it relies on to inform Level A harassment and lethal take. The scientific literature is clear that the impacts to dens are restricted to the cubs because the early emergence/departure relate to cub fitness and ability to keep up with their mother and survive outside the den. The mother does not have these same constraints and would only experience a short-term behavioral reaction as a result of such disturbance. The only instances where the impact could be as severe as to cause lethal take, or even Level A harassment to a mother would be if a den was run over by equipment and the female crushed. However, we have already shown that the probability of this happening is so low (60 days old. Additionally, our approach does not consider the presence of non-maternal dens, so it makes sense that we would rely on an analysis that purposefully attempted to exclude those. Further, it is an incorrect assumption that bears observed without cubs later in the spring that emerged prior to February 27 meant they emerged without cubs. While that is one possibility, another is that the bear emerged with cub(s) but lost them prior to being later observed in spring.

Comment (66): Two commenters stated that the FWS's den disturbance model is flawed or arbitrary and capricious because the way it handles emergence or departure dates creates a second opportunity for there to be no response to disturbance. One commenter specified that even after application of the probabilities in table 8 of the Request (p. 41 (supplemental information)), indicates that a Level A harassment occurs (resulting in assignment of a new “disturbed” emergence and/or departure date for the den), the model arbitrarily creates a second opportunity for there to be no response to the disturbance by allowing the new “disturbed” emergence and/or departure date to be the same as, or very close to, the original “undisturbed” date. They stated the FWS does not provide any explanation for creating this second opportunity in the model for there to be no response to the disturbance, nor does the FWS explain how this squares with its rationale that whether exposure to disturbance results in Level A take from early emergence and/or departure or not is determined based on probabilities calculated in its case studies analysis.

Our Response: We disagree. In our approach, if a Level A harassment occurs, dens are not given another opportunity for no response to occur. Once a den is simulated to have a take by Level A harassment occur, nothing in our modeling approach allows that to be taken away. We believe the commenter is confused by our inclusion of two sets of post-emergence disturbance probabilities. Significant thought, based on polar bear ecology, went into informing this portion of the model. The reason there are two post-emergence probabilities is because if a den is not disturbed until after the bears emerge from the den, based on our case studies, bears tend to show higher levels of disturbance outside the den than if they had been disturbed leading to an early emergence. Part of this is related to mitigation measures being imposed to limit disturbance once a den is discovered. The other post-emergence set of probabilities is for the case where a den was disturbed leading to an early emergence. Again, these probabilities were derived from the case studies and showed lower probabilities of disturbance during the late denning period. Because we do not have very precise information on when anthropogenic disturbance might occur, we simulate the new, disturbed, emergence/time-at-den values, and then calculate the updated litter survival rate from those. It's true that those new “disturbed” emergence/time-at-den values could be very close to when they were “naturally” meant to be. However, this would occur with the same probability as a den being disturbed much earlier than its “undisturbed” values. This is because we utilize a uniform distribution to simulate the updated phenology values.

Comment (67): One commenter explained from their reading of the FWS's analyses, that about 61 percent of dens in the Colville-Canning region are on barrier islands, or 15.25, leaving 9.75 other land-based dens annually in the Colville-Canning region. Those, plus the 12 estimated for the National Petroleum Reserve--Alaska (NPRA), indicate at least 21.75 expected dens occur in the ITR area annually. The reduction of that number to just 6.5 dens expected to be within a mile of industry activities is unexplained. The commenter noted that it is unclear to them how the FWS calculated 6.5 dens within the ITR area and requested that the FWS better explain how the 6.5 dens was calculated.

Our Response: The reduction to 6.5 dens expected to be within a mile of Industry activities was accomplished by examining the spatio- temporal files provided to FWS by AOGA upon our request. These files were used to complete our predictive modeling estimates as described in the proposed ITR and supplemental information at https://www.regulations.gov under Docket No. FWS-R7-ES-2026-0694.

Comment (68): One commenter states that the FWS does not properly explain how it estimated the number of dens to be located within a mile of industry activities, when the FWS does not know when and where various activities would take place. Additionally, the commenter suggests it was an error by the FWS to remove certain dens from the analysis.

Our Response: The FWS provided the code and model inputs for public comment. In the model documentation (R workspace), the spatial layers that the FWS used are provided.

Comment (69): One commenter notes that in its analysis of denning impacts the FWS defines the post-emergence period as “the date of den emergence until permanent departure from the den site.” The commenter further suggests that defining the post-emergence period in this manner does not account for the time spent by the mother and cubs after

they depart the den site and move toward the coast and sea ice.

Our Response: The commenter is correct that the FWS den disturbance model does not account for things that happen to the family group after they permanently depart the den site. That is because the information we use to inform impacts to denning bears is related to events that occur at the den site. However, we do account for these impacts in the surface analysis. The LOA database captures observations of sows and cubs of the year as they head to the sea ice, and therefore the surface analysis predictive estimates sufficiently address their potential harassment.

Comment (70): One commenter notes that the FWS's predictive models account for MMPA's harassment definitions, which encompass certain acts which have “the potential to” result in certain biological consequences, and to account for any unobserved take. However, the commenter disagrees with how the FWS takes these into account and suggests that the FWS's assessment of activities should be based on best available science and must avoid speculation.

Our Response: The FWS's analytical approaches do use the best available scientific evidence when assessing the nature of certain acts and the potential consequences of those acts to marine mammals.

Comment (71): Several commenters expressed preference for take estimation methods based on “observed data” and stated the FWS predictive estimates are not based on observed data.

Our Response: We use observational data, as well as other data sources, in our predictive modeling. The use of predictive modeling does not abandon observed data, rather it incorporates it into a broader estimate that also includes other empirical biological data.

Comment (72): The requester states the FWS should adopt the analyses presented in the Request, which they claim are better suited than those produced by FWS.

Our Response: We have chosen to retain their predictive model estimates in the ITR alongside the estimates presented in the Request.

Comment (73): The requester states the MMPA does not require applicants to prove harassment didn't occur, yet the burden has shifted in the proposed ITR.

Our Response: The required components of requests for ITRs are specified in the implementing regulations at 50 CFR 18.27(d). Here, we deemed AOGA's request adequate and complete.

Comment (74): One commenter states that in the FWS's denning impacts analysis, a disturbance during the late denning period can only cause assignment of a new “disturbed” emergence date, but cannot result in an early departure unless disturbance also occurs again between the time of the new “disturbed” emergence date and the original departure date, in which case only 0.30 of such dens will then have a new departure date assigned in response to the second disturbance. The commenter then states that disturbance during the late denning period can result not only in premature emergence, but in premature emergence and departure, even absent additional disturbance following the emergence. The commenter attributes this assertion to best-available science. However, no information sources or data are provided to indicate where this conclusion is drawn from or why it should be considered best available science. The commenter further states that the FWS does not explain why a den must be disturbed at least twice before a disturbance during the late denning period results in premature departure.

Our Response: We are not aware of any science that indicates that departure from the den will be shortened even in the absence of a disturbance stimuli after emergence (i.e., during the post-emergence period). We have been clear that for disturbance to be categorized in our analysis of case studies, a source of disturbance within 1 mile must be present.

Comment (75): One commenter states the FWS use of 60 days as the threshold at which polar bear cubs would not die if they were exposed to the environment outside the den is arbitrary, because the source study that discusses this threshold references cubs walking at 60-70 days of age. They state using a 60-day cutoff results in arbitrarily underestimating the probability and extent of lethal take.

Our Response: The FWS disagrees with this assessment. While it is true that Kenney and Bickel (2005) note that bears were generally walking by 60 days old, this is not the only information we use to support our position that cubs can be viable outside of the den once they reach 60 days of age. Harington (1968) provides evidence that cubs begin to sleep away from their mothers in the day at about 60 days of age, indicating their thermoregulatory abilities have improved. This is further confirmed by Blix and Lentfer (1979) that showed thermoregulatory ability of cubs to ambient air temperatures is well- established at den emergence. Finally, Rode et al. (2018) found “the earliest estimated date of emergence from a den for a female later observed with cubs was 8 January, a duration of 61 days.” They further show that, “[t]he shortest estimated denning duration for a reproductively successful female was 42 days.” Thus, there are multiple lines of evidence that at 60 days, cubs could be viable outside of the den. The FWS is not claiming that their survival wouldn't be impacted by such an early emergence, but this reduced survival would be captured based on our litter survival rates calculated based on the date of emergence and time spent at the den post emergence.

Comment (76): One commenter stated the denning bear behavioral reaction probabilities are not supported by reliable scientific studies, but instead the LOA database of human observations. They state the use of the LOA database and case studies is fundamentally flawed because (1) the information was not obtained through a systematic or peer-reviewed study and (2) FWS acknowledges polar bears can move cryptically through their environment, avoiding detection by humans even when in close proximity to infrastructure and activities. Finally, the commenter states FWS cannot logically rely on the absence of reported take of surface bears by industry to conclude that little harassment has occurred in the past.

Our Response: The commenter is incorrect that the FWS approach to estimating disturbance to denning bears was not obtained through a peer-review study. We developed the den case studies methods, submitted them to a peer-reviewed scientific journal for review, and the manuscript was later published in a peer-reviewed scientific journal (i.e., Woodruff et al. 2022). We agree that having more precise information on how polar bears respond to disturbance through an experimental study would be best. However, that would lead to an unacceptable number of dens being actively disturbed to obtain those data, including the potential for cubs to be abandoned. We will not propose to conduct such invasive research. Thus, we relied on existing information from the peer-reviewed literature, LOA reports, and other case studies to develop an informed estimate of disturbance from these. We discuss the potential for unobserved bears in the context of our predictive modeling estimates, which are designed to provide estimates inclusive of both observed and unobserved take.

Comment (77): The commenters cite a letter from Dr. Steve Amstrup that

discusses a non-zero probability that family groups en route from dens to the sea ice will encounter human activities that may cause mothers to run and separate from their cubs. They also state that this potential source of Level A harassment or cub mortality should be included in the predictive estimates of impacts to surface bears.

Our Response: The FWS's approach to assessing disturbance to denning bears only considers the period of time bears are at the den site, and not what happens after they depart the den site. However, we also consider the potential for disturbance to bears once they leave the den through their surface level disturbance analysis. We have no records from the LOA database that indicate a sow was permanently separated from her cubs. While we cannot state there is no chance of this occurring, it has not been reported in decades of operations, therefore we do not anticipate it occurring during the ITR period.

Comment (78): One commenter states that the FWS chose to re-examine the denning model and denning case studies, which resulted in a lower likelihood estimate of serious injury or mortality, to address least practicable adverse impact on the species or stock rather than work with the requester to examine how the frequency, methodology, timing or location of activities could be altered to reduce the probability of serious injury or mortality and thereby ensure that the take met the definition of incidental take which requires it to be infrequent, unavoidable, and accidental to address least practicable adverse impact on the species or stock.

Our Response: The FWS has always recognized that its analytical methods are not set in stone but are updated as new information becomes available or when issues with our current ways of interpreting information are identified. That is what happened with our reassessment of the case studies. Upon reviewing our case study assessments prior to the creation of the proposed ITR, we realized that we were not being consistent in how we applied some of our decision criteria, such that the methodology only allowed case studies to be classified as having an impact based on information other than phenology dates for dens that occurred after peak emergence date, but did not allow a similar thing to happen for dens with emergence prior to the peak emergence date. We also recognized that there was not sufficient information in the scientific literature to support the contention that a bear captured during (or just prior to) the den establishment period would have no impact on the denning process. Thus, we further updated our case study inclusion criteria to address this. We also worked with the requester to incorporate a new data-led strategy for AIR surveys that prioritizes high density denning areas, leading to a decrease in survey time while increasing the amount of habitat surveyed.

Comment (79): One commenter noted that while the FWS predictive denning model incorporates the relationship between early emergence, early departure, and a decrease in litter survival rate presented by Rode et al. 2018, the authors of that study also noted that the earlier den emergences were likely of function of less healthy sows, indicating early emergence and/or departure may be a combination of natural factors, and not human disturbance.

Our Response: The relationship between earlier den emergence and decreased litter survival has been documented in multiple studies (Rode et al. 2018, Andersen et al. 2024). These studies did not include investigations of the mechanisms underlying the decision by bears to depart dens early vs. late, but the authors did speculate that maternal condition could be influential. Decreased maternal condition at emergence could be the result of poor condition prior to denning, but it is also possible that stress or changes in phenology during earlier denning periods caused by disturbance led to decreases in body condition. Further, the case studies document multiple instances of sows that appear to be in good body condition emerging from dens concurrent with a disturbance, strongly suggesting that the disturbance caused the bear to emerge earlier than it would have under undisturbed circumstances.

Comment (80): One commenter discussed the use of median thresholds for time at den between emergence and departure when evaluating denning model case studies. They highlighted the difference between the Andersen et al. 2024 dataset (hereby referred to as the reference dataset), which was used to establish a median value threshold, and the case study dens. Specifically, the commenter said the bears included in reference dataset were collared, and their den emergence and den departure timelines were measured from collar data, which is more precise than the human observation data available for case study dens. They also state collar data can potentially include time that was post den departure. The commenter states these differences contribute to the reference dens having a longer time at the den before departure. The commenter further describes the reference dataset as having a bimodal distribution and large 50 percent confidence interval. They state these statistical traits may be an artifact of the small sample size and not representative of true post-emergence behavior. Finally, the commenter states almost all case studies would be classified as staying longer than the lower end of the 50 percent confidence interval, and almost all case studies would be classified as departing early if we relied on the upper end of the 50 percent confidence interval. They state it is unclear whether the modest effect of human activity on post-emergence den duration in the distribution data is due to low sample size in the undisturbed reference group, different measures of den emergence and departure between ITR case studies and the undisturbed reference group, or a true effect.

Our Response: The temperature-based methodology used to estimate emergence in Andersen et al. (2024) can lead to earlier emergence dates than the observation-based estimates in the case studies if bears open dens before they first leave the den and are observed on the surface. This could cause an overestimation of take in the post-emergence period, but for the same reason, this would cause an underestimation of take in the late denning period. We acknowledge this possibility, but because any overestimation in one period would necessarily be equal to an underestimation in the other, we will continue to use this approach unless a more appropriate alternative is revealed in future peer- reviewed studies. It is correct that the median post-emergence duration (6.25 days) in the Andersen et al. (2024) dataset was based on a sample of 20 dens. This is because we limited the data to dens that occurred on land in the SBS subpopulation for which cubs were observed in the spring following denning, which we considered to be most representative of the AOGA ITR simulation process. The Andersen et al. (2024) dataset also demonstrates that durations do not differ significantly (p=27 Feb) from both subpopulations, the sample size increases to 36, the distribution of durations is not bimodal, and the median duration increases to 9.5 days. For future iterations of the simulation model, we will continue to consider the appropriateness of including a larger

sample of dens to address the commenter's concerns.

Comment (81): One commenter stated using a den emergence threshold for case study evaluation that is based on the median den emergence date of the dens included in Andersen et al. (2024) is difficult to justify. The commenter points to the standard deviation of den emergence dates in the referenced dataset, which they say is 19.5 days, and constitutes a spread so large that the median-based threshold would classify a substantial proportion of any sample of dens as early emergers regardless of whether disturbance occurred. They also state the observed 28.9 percent of case study dens classified as early emergers falls 21 percentage points below the 50 percent baseline expected by chance alone, which they say is the opposite relationship one would expect when comparing an undisturbed reference median to exposed dens in the case studies. Finally, the commenter states these statistical results indicate there is no clear evidence of early emergence, and instead the estimate appears to simply be the result of the median-based threshold cutting off the early end of what appears to be a typical distribution of emergences for undisturbed dens.

Our Response: We appreciate these comments and agree that the median-based approach can be improved if additional data or scientific studies become available to support revision or a viable alternative. However, we must account for the peer-reviewed scientific studies that have demonstrated that early den emergences and short durations at the den post-emergence are associated with reduced survival of polar bear litters (Rode et al. 2018, Andersen et al., 2024) and evidence from multiple case studies that document instances of den emergence or departure quickly following a disturbance event. In some cases, the reaction seems clearly causal, primarily because of the timing of the reaction compared to the human activity (e.g., a bear emerged from a den for the first time, looked towards a loader operating 146 m away, then departed for the sea ice with 1 cub within 5 minutes). The case studies, however, often lack the detailed information necessary for temporally linking an emergence or departure to an activity, and attributing the response to the activity becomes increasingly subjective as the temporal threshold increases. For example, a response concurrent with an activity is very likely to have been influenced by that activity, but this link becomes less clear if the response occurs days later. Our approach of using median dates of emergence and departure to inform probabilities of disturbances resulting in early responses was developed to avoid this subjectivity in classifying case study responses. We continually review and refine the process based on received input and new scientific information.

Comment (82): One commenter references a response to comments in a prior rulemaking. In that comment response, the FWS compared den emergence times derived from collar temperature sensor data (such as those in Rode et al. (2018) and Andersen et al. (2024)) to those derived from case study data. Collar temperature sensors detect the moment the den is opened to outside air, whereas case study data, which are based on visual observations of bears on the surface, are likely to be recorded after initial emergence. However, the commenter disagrees that the comparison should not be made at all (i.e., that the comparison is completely uninformative). Rather, the commenter believes it is appropriate to keep in mind the difference in how emergence times are measured when making broad comparisons between the two distributions.

Our Response: From the beginning, the FWS has considered how this difference in methodology could influence the structure and outcomes on the simulation model. For example, if a bear in a case study was not observed until after March 12 (hence not classified as an early emergence), but the den was actually opened earlier than March 12 (or the bear was on the surface earlier, but not observed due to darkness, weather conditions, lack of observers in the area, etc.), then the probabilities of disturbance derived from the case studies could underestimate take during the late denning period. This underestimation, however, would likely be balanced by a concomitant overestimation in the late denning period because the imprecise observer-based emergence estimate would underestimate the duration of the post-emergence period. We update the ITR approach continually based on input from concerned parties and as new empirical data and scientific studies become available.

Comment (83): One commenter states that using the median as a threshold for classifying Level A harassment creates a baseline classification problem in assessing how industry activities alter polar bear behavior.

Our Response: We appreciate this point and agree that the median- based approach can be improved if additional data or scientific studies become available to support revision or a viable alternative. However, we must account for the peer-reviewed scientific studies that have demonstrated that early den emergences and short durations at the den post-emergence are associated with reduced survival of polar bear litters (Rode et al. 2018, Andersen et al., 2024) and evidence from multiple case studies that document instances of den emergence or departure occurring as a response to a disturbance event. In some cases, the reaction seems clearly causal, primarily because of the timing of the reaction compared to the human activity (e.g., a bear emerged from a den for the first time, looked towards a loader operating 146 m away, then departed for the sea ice with 1 cub within 5 minutes). The case studies, however, often lack the detailed information necessary for temporally linking an emergence or departure to an activity, and attributing the response to the activity becomes increasingly subjective as the temporal threshold increases. For example, a response concurrent with an activity is very likely to have been influenced by that activity, but this link becomes less clear if the response occurs days later. Our approach of using median dates of emergence and departure to inform probabilities of disturbances resulting in early responses was developed to avoid this subjectivity in classifying case study responses. We continually review and refine the process based on received input and new scientific information.

Comment (84): One commenter states that poor maternal body condition may affect post-emergence duration and reduced cub survival.

Our Response: We agree that poor maternal condition could cause females to depart the den site early. Although it's possible that poor body condition could be unrelated to disturbance, it is also possible that disturbance during denning could contribute to that condition (e.g., through stress or disruptions to denning phenology or activities earlier in the denning period). Presently, the mechanisms behind this relationship are not well understood, but the scientific literature demonstrates that decreased time at the den site post-emergence is associated with reduced litter survival and disturbance can decrease that time. At every step, we incorporate the best available science into the ITR process and amend that process as new information becomes available.

Comment (85): One commenter states that they think research activities are typically inappropriate proxies for predicting incidental take related to oil and gas activities. The commenter also states that they believe the inclusion of research case studies as a proxy for

industry activities likely inflates the incidental take probabilities in the denning analysis.

Our Response: Sources of disturbance in research case studies include snowmachine, helicopter, fixed-wing aircraft, and people on foot. Each of these modes of travel are used by Industry and produce sound profiles similar to other types of Industry activities. Recognizing polar bear research may focus on activities involving closer proximity to polar bears than Industry activities focused on avoidance or minimization, the distributions of distances at which the activities occurred overlap substantially.

Comment (86): One commenter states that they disagree with the FWS classification of Case Study 32 as a cub mortality.

Our Response: The FWS did not conclude that a cub mortality did occur, just that there was likely den abandonment during the early denning period, which may have resulted in the loss of one or more cubs due to the early emergence date before cubs would have been viable.

Comment (87): One commenter states that they disagree with assertions described in Woodruff et al. (2022) that say the use of median- or mean-based values should minimize bias associated with individual variation. The commenter believes that using median- or mean-based threshold is unlikely to capture true disturbance rates and leads to an overestimation of disturbance.

Our Response: It isn't possible to know when a bear that was exposed to disturbance would have emerged or departed in the absence of that disturbance, but multiple case studies document instances of den emergence or departure occurring very soon after disturbance events, so we disagree that it is possible that true rates are low or close to zero. In some cases, the reaction seems clearly causal, primarily because of the timing of the reaction compared to the human activity (e.g., a bear emerged from a den for the first time, looked towards a loader operating 146 m away, then departed for the sea ice with 1 cub within 5 minutes). The case studies, however, often lack the detailed information necessary for temporally linking an emergence or departure to an activity, and attributing the response to the activity becomes increasingly subjective as the temporal threshold increases. For example, a response concurrent with an activity is very likely to have been influenced by that activity, but this link becomes less clear if the response occurs days later. Our approach of using median dates of emergence and departure to inform probabilities of disturbances resulting in early responses was developed to avoid the potential for bias in subjectively classifying case study responses. We agree, however, that the median-based approach can be improved if additional data or scientific studies become available to support revision or a viable alternative.

Comment (88): One commenter presented an analysis and concluded that using a median-based threshold to estimate disturbance of denning bears likely overestimates actual den disturbance rates.

Our Response: The smallest effect considered in the simulations is 5 days, so it isn't possible to gauge results directly for a 1-day change in the behavior. Nevertheless, these results suggest that our approach may be as likely to underestimate take as it is to overestimate it based on the true level of disturbance. Based on the emergence simulations, if >50 percent of bears emerge a day early due to disturbance, these simulations suggest our method underestimates take, whereas take would be overestimated when 60 days old. Additionally, our approach does not consider the presence of non-maternal dens, so it makes sense that we would rely on an analysis that purposefully attempted to exclude those. Further, it is an incorrect assumption that because bears were observed without cubs later in the spring if they emerged prior to February 27 that means they emerged without cubs. While that is one possibility, another is that she emerged with cub(s) but lost them prior to being later observed in spring.

Comment (93): One commenter states the FWS has failed to consider the potential impacts of early emergence and/or early departure on cub survival into the autumn, citing information from Rode et al. (2018) on the positive relationship between sow body weight at emergence and cub and litter survival.

Our Response: The FWS disagrees that we did not incorporate information from Rode et al. regarding duration in the den and its effects on litter survival. In actuality, we utilized the underlying data in Rode et al. (2018) and Andersen et al. (2024) to estimate the relationship between time in the den (and at the den post emergence) and litter survival in the spring. Instead of relying on a single metric, we rely on the underlying data to estimate this relationship and apply it to our analytical methods.

Comment (94): One commenter suggests that the FWS's rationales for revising its assessments of case studies were not justified and resulted in underestimation of the probability and extent of lethal take and Level A harassment.

Our Response: The FWS took a very detailed look at our previous set of methods for analyzing case studies and realized there were some areas where they could be improved. So, we reassessed them based on these updated criteria. We articulated why we made changes to each of the case studies and justified those decisions based on polar bear biology.

Comment (95): One commenter points out that FWS revised its assessment of Case Studies 8 and 9 and removed these Case Studies from its denning analysis inputs because the polar bears in these studies had been previously captured for research purposes. The commenter, citing the opinions of Dr. John Whitman and Geoff York, implies that it was arbitrary for the FWS to remove these studies because the best available science indicates that research captures have minimal effects, and there is no evidence indicating that the research capture influenced the subsequent response to disturbance. The commenter further suggests that the FWS's removal of Case Study 9 is not consistent with the rationale FWS offered to justify its removal because the research capture occurred during the den establishment period, since the record indicates the bear was collared prior to den establishment.

Our Response: The FWS has always excluded case studies involving research activities from the calculation of disturbance probabilities if those research activities were deemed to not be representative of activities that the oil and gas industry would conduct. We had previously included two case studies (8 and 9) in our analysis where polar bears were captured by researchers during the den establishment period, although both were not included as disturbance during the den establishment period. Both, however, were used to estimate disturbance during subsequent den periods (i.e., early denning and late denning, respectively). We acknowledged that we do not know the effect of intense, capture-related disturbance during the den establishment period on subsequent denning behavior or phenology and therefore cannot be sure that these observations would reflect conditions experienced by bears disturbed by oil and gas activities. The commenter indicates that by excluding Case Study 9, we're inconsistent with our own justifications for dropping it. Yet, as described above, our criteria for dropping this den was that she was collared during the den establishment period. This was in the case study record: “The bear was satellite radio collared and denned on a north coast barrier island on 20 October 1988.” This is consistent with our stated rationale because capture is a stressful event for bears and thus could impact denning activity and potentially phenology. That is why it was dropped.

Comment (96): One commenter noted that upon revaluation the FWS reclassified Case Study 17 from early emergence and departure in response to disturbance to a classification of early departure only. The commenter cites the opinion of Dr. Joh Whiteman and Geoff York to conclude that the reclassification is not rational and should be considered arbitrary and capricious.

Our Response: Upon reviewing our case study assessments prior to the creation of the 2025 revised ITR, we realized that we were not being consistent in how we applied our decision criteria, such that it only allowed case studies to be classified as having an impact based on information other than phenology dates for dens that occurred after peak emergence date, but we did not allow a similar thing to happen for dens with emergence prior to the peak emergence date. Thus, we are actually being more consistent with our stated procedures for assessing case studies and not being arbitrary or capricious.

Comment (97): One commenter suggests that the FWS improperly concluded that any observed emergence after the median emergence date must be a natural emergence rather than an early emergence in response to a disturbance event. In the commenter's opinion this will lead to 50 percent mischaracterization of observations, therefore it is arbitrary and capricious for the FWS to make this conclusion.

Our Response: We agree that our case study methods can lead to some dens being classified as disturbed when they were not, or some that were disturbed as not being disturbed. Because we do not know when polar bears gave birth to cubs, we needed to rely on population-level metrics to make our assessments. Because we do not bias our assessment to be more prone to be disturbed (or less disturbed), our approach allows for the average effects to be identified over the course of 10,000 simulations. These procedures went through peer-review and are published in the scientific literature (Woodruff et al. 2022).

Comment (98): One commenter stated that past reports should not be the only basis of determining the type of harassment denning polar bears experience. They pointed out that FWS states reports of lethal take of cubs via den abandonment are rare, but that den abandonment would be difficult to observe.

Our Response: The FWS agrees that den abandonment would be difficult to observe. This, combined with the other

uncertainties listed in the proposed ITR, is why the FWS has also presented estimates generated using predictive modeling which considers not only past observational reports but also scientific studies illustrating various cause and effect mechanisms. As stated in the proposed ITR, “FWS's predictive model results provide, among other things, a basis for assessing the probability and severity of potential impacts that the scientific literature suggests are possible but have not been observed or are rarely observed in the ITR area.”

Comment (99): The requester stated that the FWS predictive modeling of impacts to denning bears does not reflect best available scientific information because it relies on case studies that do not represent the impact of industry activities and relies on precautionary assumptions.

Our Response: The FWS has provided clear justification for its use of case studies that have a research component. We have always removed case studies where the research efforts did not have a parallel with Industry activities (e.g., capture). However, just because a researcher walked in the general vicinity of a den does not mean that the bear would perceive a different risk than if it were an Industry employee walking in the general vicinity. And we have case studies where dens have been located very close (~50 feet) from active industrial activities all winter before the den was discovered on emergence. So, including certain research-related case studies is appropriate because they represent human activity around polar bear dens and how bears respond to it. Additionally, a number of case studies had researchers periodically near the den (e.g., setting up cameras), but then there is documentation of a bear responding to Industry activity in the distance, but not to the researchers setting up the camera. Thus, we still believe these are representative case studies that should not be ignored.

Comment (100): The requester states the aircraft-related case studies used to calculate disturbance probabilities are not representative of industrial activity, even pipeline surveys. They state aircraft-based surveys of pipelines are unlikely to overlap with maternal polar bear dens because there are few polar bear dens that occur in the region relative to the suitable habitat in the coastal plain, spills are rare and distributed throughout a large network of infrastructure, and only remote stretches of pipelines are surveyed via aircraft.

Our Response: The FWS has provided clear justification for its use of case studies that have a research component. We have always removed case studies where the research efforts did not have a parallel with Industry activities (e.g., capture). Including certain research-related case studies is appropriate because they represent human activity around polar bear dens and how bears respond to it.

Comment (101): The requester states that the case studies associated with research activities (both ground-based and aerial) are not representative of industrial activities and therefore their inclusion in the disturbance probabilities creates bias in the denning model.

Our Response: The FWS has provided clear justification for its use of case studies that have a research component. We have always removed case studies where the research efforts did not have a parallel with Industry activities (e.g., capture). However, just because a researcher walked in the general vicinity of a den does not mean that the bear would perceive a different risk than if it were an Industry employee walking in the general vicinity. We have case studies where dens have been located very close (~50 feet) from active industrial activities all winter before the den was discovered on emergence. So, including some research-related case studies is appropriate because they represent human activity around polar bear dens and how bears respond to it. Additionally, a number of case studies had researchers periodically near the den (e.g., setting up cameras), but then there is documentation of a bear responding to Industry activity in the distance, but not to the researchers setting up the camera. Thus, we still believe these are representative case studies that should not be ignored.

Comment (102): The requester cites three publications: Stankowich (2008), Blumstein et al. (2003), and Fernandez-Jurcic et al. (2005) in support of their statement that wildlife responses are strongly influenced by approach behavior and perceived intent with direct, targeted approaches eliciting significantly greater disturbance than tangential or passing movements.

Our Response: The cited references are not directly applicable to draw the commenter's conclusions or are counter to the argument being made by the commenter. A primary consideration that must be made in evaluating the applicability of these sources is that all three of the cited papers specifically focus on prey species, either small birds or ungulates (e.g., hooved animals such as deer, moose, or caribou). The behaviors of prey vs. predator wildlife species display when confronted with humans are not interchangeable and they cannot be assumed to act similarly. Many prey species view humans as predators and elicit similar responses, whereas this cannot be assumed true for predator species such as polar bears. Therefore, the response of small birds and ungulates to humans should not be used as proxy for the response of apex predators (e.g., polar bears) as the commenter has done.

Furthermore, the conclusion drawn from these specific articles does not directly support the claim that wildlife is significantly more disturbed by direct approaches than tangential movements. In Fernandez- Jurcic et al. (2005) the authors found that four of the five birds in their study were disturbed (measured as “flight initiation distance” (FID)) at greater distances from tangential approaches than in direct approaches. The second referenced paper that focuses on birds, Blumstein et al. (2003), did not compare nor attempt to compare direct and tangential approaches. This study only assessed the impact of direct approach on various bird species. Secondly, a major conclusion drawn by the authors is that the distance that elicits disturbance response in wildlife is species-specific and cannot be easily applied interspecifically (between species). A conclusion that is particularly pertinent when using the results of Blumstein et al. (2003) to make assumptions about polar bear behavior. The third article cited by the commenter concludes that ungulates, which are a prey species, are more likely to be disturbed by humans on foot than other stimuli (e.g., vehicle, human noises). However, as previously mentioned, responses of prey species (which often view humans as predators) are not a valid proxy for the response of apex predators like polar bears to humans. When other researchers have compared the behavioral impacts of humans on multiple predator (including grizzly and black bears) and prey species (all ungulates) they found different behavioral reactions between the two groups of wildlife (Granados et al. 2023).

Comment (103): The requester states that there are no Industry activities, even pipeline surveys, that intentionally or repeatedly hover a helicopter above a known polar bear den as they state was done in Case Study 4. The requester suggests that conflating the research activities described in Case Study 4 with industry activities inflates estimated probabilities of harassment and the projected potential for take, specifically the risk of Level A or lethal take. They further state aircraft-based surveys of pipelines are unlikely to overlap with maternal polar bear dens

because there are few polar bear dens that occur in the region relative to the suitable habitat in the coastal plain, spills are rare and distributed throughout a large network of infrastructure, and only remote stretches of pipelines are surveyed via aircraft.

Our Response: The FWS has provided clear justification for its use of case studies that have a research component. We agree that case studies concerning activities that are not analogous to the specified activities in terms of their potential to impact polar bears should not be incorporated into the inputs of predictive modelling. Over the years, we have carefully scrutinized the case studies applying this standard and have in some instances removed case studies that after further review were deemed non-analogous. However, wholesale removal of all case studies concerning activities entailing a research component is not appropriate, as some activities conducted for the purpose of research closely parallel activities conducted by Industry for other purposes in terms of their capacity to impact polar bears. In other words, the status of the person conducted the case study activities, and the purpose of such activities, matter less than the extent to which the case study activities parallel the specified activities in terms of factors such as equipment used and stimuli produced. With respect to the example cited in this comment, while routine pipeline surveys generally occur at a higher altitude and in a linear fashion, weather conditions or other operational needs may require aircraft to fly lower or to revisit or hover over certain areas. Considering a case study that improves understanding of how such activities may affect polar bears helps us more accurately estimate take from analogous specified activities in the future.

Comment (104): Commenters suggested that the case studies used in the FWS's analyses resulted in an overestimation of effects of oil and gas activities to denning polar bears. The commenters cited differences in distances from and nature of sources of anthropogenic disturbances, seasonality, den phenology, and location as factors in the FWS's interpretation of the data in a manner that the commenter deemed overly conservative. Commenters stated that the den case studies used to estimate den disturbance rates are substantially closer to industry activities compared to the distances of simulated dens to industry activities, thus resulting in the denning model overestimating take.

Our Response: We agree with commenters that there is a relationship between the distance between a den and the source of disturbance. This factor was apparent in the recent study of aircraft overflights that we published on the relationship between altitude and polar bear take. Unfortunately, these data are challenging to collect in a designed study because they would require purposefully disturbing dens, which could lead to significant demographic impacts to the population. Instead, we must rely on existing case study data to help inform this critical model parameter. The arguments made by the commenter highlight this difficulty. While it is true that the distances between simulated dens and Industry are on balance higher than in the case study dataset, this is likely attributable, at least in part, to detection bias toward dens found closer to infrastructure. But just because fewer dens were detected farther from human activity in the case studies does not mean that dens farther out were not disturbed; it is possible such dens existed and were disturbed but were not detected. Numerous case studies show that bears can be disturbed by human activity at distances greater than those suggested by the commenter, and we appropriately account for this possibility.

Comment (105): One commenter stated 805-m buffers are inadequate to prevent disturbance of polar bears, especially since bears cannot distinguish between industrial activity and research monitoring.

Our Response: The FWS has not included 805-m buffers in the mitigation measures for this ITR. The ITR requires 1.6 km (1 mi) operational mitigation zone around putative polar bear dens.

Comment (106): One commenter stated that the fact that female polar bears have rarely been observed abandoning a den during early denning, which is the FWS's stated rationale for restricting the denning impact area to half a mile during the early denning stage, does not necessarily mean that polar bears have not experienced harassment expressed in another form (i.e., physiological injuries) during this stage. Their comment suggested the FWS should consider this possibility and discuss information such as studies of noise propagation and detection probabilities as a function of distance from dens (Owen 2021), highlighting that the FWS supplemental analysis states that “successful denning, birthing, and rearing activities require a relatively undisturbed environment.”

Our Response: It is possible that denning females could react inside the den in response to outside activities. However, the FWS has no data to determine if this occurs, how it would manifest (in the small space of a maternal den there are few ways behavioral patterns may be disturbed), and if it does, how frequently. The commenter is also correct that the referenced study provides information on how noise from different sources of anthropogenic disturbance propagates into the den at different distances. However, this study provides no information on how bears respond to the different activities, thus cannot be used to inform response rates to maternal females while inside the den (a time where their responses to activity cannot be observed).

Comment (107): One commenter states the louder and closer an exposure is to a den, the more likely is to affect a denning bear, and den Case Study 32 should not be included in the case studies to classify den disturbance rates.

Our Response: The FWS agrees that there is a higher chance of disturbance occurring to dens closer to sources of disturbance. However, to claim that the level of disturbance experienced by this bear is an outlier is not supported by other observations in the collection of case studies. For example, Case Studies 58 and 59, which both involve exposure to only industrial activities, both describe dens that had significant sources of disturbance occurring for prolonged periods, immediately adjacent to the dens. Thus, the den from Case Study 32 is relevant to retain.

Comment (108): The requester stated that the FWS-proposed 1 mi mitigation zone around denning polar bears is not practicable on the basis that it is unnecessary. The requester stated the FWS should adopt the 805 m mitigation zone around polar bear dens included in their Request, referencing data showing the reduced likelihood of disturbance of denning bears beyond 805 m. The requester also pointed to FWS's use of a 805 mitigation zone in recently-issued an incidental harassment authorization as evidence that a 805 m mitigation zone is more than sufficient to effect the least practicable adverse impact on denning polar bears.

Our Response: As stated in the proposed ITR, the FWS “ensure[s] the least practicable adverse impact by requiring mitigation measures that are effective in reducing the impact of specified activities, but not so restrictive as to make specified activities unduly burdensome or impossible to undertake and complete.” While we agree that likelihood of disturbance decreases as the distance between human activities and dens increases, the best available scientific evidence shows that denning polar bears can be disturbed at distances

greater than 805 m (0.5 mi). We thus find that expanding the 805 m mitigation zone proposed in the Request to a 1 mi mitigation zone (1,610 m) as contemplated in the proposed ITR will be effective in reducing the impact of the specified activities. We have not been provided any information demonstrating why implementing a 1 mi mitigation zone would be so restrictive as to make the specified activities here unduly burdensome or impossible to complete, and note that a 1 mi mitigation zone is required and has been successfully implemented via the current 2021-2026 Beaufort Sea ITRs. With respect to the one recent instance where we imposed a smaller, 805 m mitigation zone in an incidental take authorization, our determination to do so was based on project-specific safety considerations due to a specific sea ice route that required grounded ice.

Comment (109): The requester cited the conclusion of Larson et al. (2020), that denning polar bears on Alaska's North Slope are overtly unreactive (i.e., largely tolerant) of human activity near den sites (10 years. The commenter is incorrect that the SBS stock is in rapid decline. While it's true that the stock did suffer a significant decline in the early 2000s, Bromaghin et al. (2021) found that that decline had largely leveled off and the subpopulation's abundance had likely stabilized. We are not aware of any more recent research that has shown that the subpopulation's decline has likely begun again.

Comment (169): Commenters stated other forms of anthropogenic take must be factored into the baseline for the species in considering whether take will have a negligible impact. One

commenter stated the proposed ITR does not connect essential information on the annual number of human-caused mortalities for SBS polar bears (e.g., polar bear harvest under the I[ntilde]upiat- Inuvialuit Agreement) to its analysis of impacts on the SBS stock from oil and gas activities, including impacts to the population and to subsistence uses. They stated the FWS's failure to discuss whether take under the proposed ITR would have a negligible impact (i.e., an impact that is consistent with the stock reaching or maintaining its optimal sustainable population) in light of ongoing take and other mortality events renders the proposed ITR arbitrary and capricious.

Our Response: We have considered impacts to SBS polar bears in light of the environmental baseline, which includes ongoing subsistence harvest. The analysis draws heavily from the stock assessment report for SBS polar bears. This report includes descriptions of the stock and its range, a minimum population estimate, net productivity rate, and current population trend, an estimate of the potential biological removal level, an estimate of human-caused mortality and serious injury, and a categorization of the status of the stock. Through these stock assessment reports, we analyze the baseline impacts to SBS polar bears. We have taken this information into account in small numbers and negligible impact determinations. The most recent draft polar bear stock assessment report is included by reference into the proposed ITR.

Comment (170): One commenter stated that FWS's estimate of 20 walruses being taken per year underrepresents walruses' sensitivity to noise disturbance and that FWS should further assess the extent of harm to Pacific walruses in the Beaufort Sea and account for potential cumulative effects from repeated disturbances.

Our Response: Given the scarcity of walruses in the ITR area and the mitigation measures incorporated into the Request, it is highly unlikely that the specified activities could take more than 20 walruses per year.

Comment (171): One commenter stated that there could be a good opportunity to conduct research in association with the proposed ITR on how repeated takings affect overall health of polar bears and Pacific walruses.

Our Response: Whereas we agree with the commenter that these types of data would be valuable to informing future incidental take authorizations, research to obtain them is outside of the scope of our authorization. We also agree with the commenter that these type of data are not yet available for polar bears or walrus and would thus not be possible to include in our analyses because we don't know if this effect exists, and if it does, what the magnitude of the effect would be.

Comment (172): One commenter states that the FWS does not explain why it used the estimate of 907 polar bears in the SBS stock when the FWS's draft SAR changes that estimate to 819 polar bears in the SBS stock.

Our Response: The FWS does explain its presentation of both 907 and 819 estimates. The draft stock assessment report has not been finalized, thus it would be pre-decisional to use draft numbers for our formal assessment of small numbers. However, we provided the small numbers estimates with respect to 819 as well, in the event the SAR are approved during the ITR period, so that the public may have the information.

Comment (173): One commenter states the FWS has failed to consider whether the total take being authorized over a 5-year period (422 takes by Level B harassment of 199 individual polar bears) can reasonably be considered small numbers.

Our Response: The numbers reported represent a reduction of repeated bears within a year, but we do not have any data to estimate the rate of repeats between years, so the 5-year totals reflected in our estimates are simply derived by adding up the total number of individuals for each year without regard to any of them being potential repeats. The probability that none of these are repeats is extremely low. If we assume 49 individuals per year, the probability of choosing 49 new individuals every year for 5 years is 0.00000000001 (1 e-11). This also supposes that the bears can come randomly from the entire SBS stock of 819 bears. If we presume that only 75 percent (614) of the bears visit land, that probability reduces to 4.8e-16, 50 percent (410) even further to 4.8e-26.

Comment (174): One commenter stated they are concerned the structure of FWS's small numbers determinations are vulnerable and warrant consideration due to the Loper Bright v. Raimondo, 144 S. Ct. 2244 (2024) decision that ended Chevron deference.

Our Response: The FWS's approach to evaluating small numbers is consistent with the best reading of the MMPA.

Comment (175): Commenters stated that by not considering the age and sex of polar bears that may be taken as part of its small numbers determination, the proposed ITR does not adequately protect the genetic integrity of the SBS polar bear population.

Our Response: Genetic variation of among polar bear stocks, globally, is very low because of the level of gene flow across the Arctic and the general lack of geographic barriers to movement. Thus, concerns about genetic diversity are not significant.

Comment (176): Commenters stated that it is against best scientific practice for FWS reach a small numbers determination while there is a lack of reliable population data for SBS bears within the past decade.

Our Response: The FWS uses the best available scientific evidence to inform its small numbers estimate. While it is true that the most recent polar bear population estimate for the SBS stock used data through 2015, it is not uncommon to have decade or more time lags between polar bear stock estimates given the challenges of studying them. Further, while not specific to the SBS subpopulation, Regehr et al. (2021) found that 10+ year intervals between abundance estimates can be reasonable and not lead to mismanagement of populations. The most recent information from Bromaghin et al. (2021) found evidence that the previously-documented population decline of the SBS had likely stabilized so we are most likely not in a situation where the population has continued to decline since 2010. There is currently an effort led by non-FWS employees to update the SBS abundance estimate with more recent data and that is anticipated to be available by late 2026.

Comment (177): Commenters stated that FWS should consider the impacts of harassment of 20 walruses relative to the total of 49 Pacific walruses that have been observed in the South Beaufort Sea between 1995 and 2023 (40 percent of animals), rather than relative to the full Pacific walrus population stock (0.0008 percent of animals), since the Beaufort Sea represents a specific portion of the walruses' range and effects are experienced by individuals present in the local area.

Our Response: The FWS's small numbers determinations evaluate if the number of marine mammals estimated to be incidentally taken is small relative to the size of the species or stock. Pacific walruses have not been found to have population substructuring, and are one large, panmictic population. As is stated in the supplemental information for the proposed ITR, the walruses found in the Beaufort Sea are extralimital, not a unique subpopulation.

Comment (178): Commenters stated that under 16 U.S.C. 1371(a)(5)(A),

small numbers must be relative to both population size and biological significance, and must account for underlying population trend and status. For a threatened, declining marine mammal population already at less than half its historic abundance, such as SBS polar bears, the killing of four cubs or even one maternal individual, and harassing up to 5.7 percent of the entire population in any single year, cannot be considered small.

Our Response: “Small numbers” is not defined in the MMPA, as the commenter suggests. As we stated in the proposed ITR, “small numbers” is defined in 50 CFR 18.27. However, we do not rely on that definition as it conflates “small numbers” with “negligible impact.” Instead, for our small numbers determination, we considered estimates of the likely number of annual takes of polar bears and determined that the estimates were small relative to the size of the SBS stock.

Comment (179): One commenter states the FWS should reassess their small numbers analysis of the Request's estimate of Level B harassment to include both the estimate of 24 bears on the surface and 3 denning bears.

Our Response: Thank you for the suggestion. We have revised the text in the ITR.

Comment (180): Commenters stated that FWS's estimates of the percentage of populations impacted are uncertain because estimates of total population size are changing.

Our Response: The justification for the proposed change in abundance estimate was articulated in the Draft SBS Stock Assessment Report. It had nothing to do with the size being uncertain, but rather accounting for a shift in the SBS boundary which reduced the overall area represented by the SBS stock.

Comment (181): One commenter believes it is irrelevant to consider the size of the specified geographical region in relation to the entire range of polar bears when the FWS makes its “small numbers” determination.

Our Response: The FWS qualitative discussion of the size of the specified geographical area in relation to the entire range of SBS polar bears is presented in addition to a quantified estimate of potential take from both AOGAs Request and FWS predictive modeling. We disagree that the discussion is irrelevant, as it helps speak to the proportion of SBS polar bears anticipated to be taken. We also provide additional discussion in the ITR about the number of estimated takes and the size of the SBS stock, supporting the finding that no more than a small number of individuals are estimated to be harassed.

Comment (182): One commenter states that the FWS is conflating its “small numbers” finding with it “negligible impacts” finding.

Our Response: As we stated in the proposed ITR, while the term “small numbers” is defined in 50 CFR 18.27, we do not rely on that definition as it conflates “small numbers” with “negligible impact.” We recognize “small numbers” and “negligible impact” as two separate and distinct requirements for promulgating ITRs under the MMPA (see Ctr. for Biological Diversity v. Salazar, 695 F.3d 893, 904 (9th Cir. 2012)). Instead, for our small numbers determination, we evaluate if the number of marine mammals estimated to be incidentally taken is small relative to the size of the species or stock.

Comment (183): Some commenters stated spatial and temporal restrictions on activities to avoid denning habitat and sensitive denning timing could protect polar bears while also avoiding limitations for projects or briefly limiting project timelines.

Our Response: Spatial restrictions on activities would not be practicable in the existing oil fields because the infrastructure is already established and operating. It would not be possible to shut down those operations annually to allow denning to occur without disturbance. Due to the large amount of potential denning habitat in northern Alaska, spatial restrictions would lead to nearly the entire operations within this application not being possible until after denning, which would not be practicable. Activities outside of the existing oil field can include seismic surveys, exploration drilling, and ice roads. Seismic surveys must necessarily be conducted in certain areas in order to acquire sufficient data to meet project objectives. Exploration drilling must necessarily be conducted within the vicinity of the target reservoir. And ice roads typically cannot be routed to avoid all areas in or near denning habitat, which is prevalent in the ITR area. Meanwhile, environmental and regulatory constraints (such as a limited season for tundra travel) dictate that these activities can only be conducted during seasonal timeframes that largely overlap with the denning season. Categorically imposing spatial and temporal restrictions at the ITR phase would make specified activities unduly burdensome or impossible to undertake and complete. That said, project- specific spatial or temporal restrictions may be imposed, if appropriate, at the LOA phase, when operational plans and specific locations are provided.

Comment (184): One commenter states that the FWS should not rely on mitigation measures to reach a conclusion that the authorized take would have the least practicable adverse impact without accurately assessing the effectiveness of each mitigation measure.

Our Response: We disagree. We have carefully explained the numerous peer-reviewed studies on AIR efficacy. There are several mitigation measures that the FWS knows are beneficial but their efficacy is variable. These include handheld FLIR surveys and personnel training. While the benefits of these measures cannot be quantified, and are therefore not included in quantitative take estimates, they can be included as measures that can go toward meeting the least practicable adverse impact standard. We have also clearly described why other possible mitigation measures may not be practicable. The mitigation measures incorporated into the Request are similar to those implemented under current incidental take authorization and have proven to be successful in minimizing impacts. To ensure the least practicable adverse impact, we evaluated various additional mitigation strategies and imposed one additional mitigation measure in these ITRs.

Comment (185): Some commenters stated mitigation measures are not analyzed or explained enough with supporting data to show that the proposed ITR meets the “least practicable adverse impact” requirement.

Our Response: We disagree. We have carefully explained the numerous peer-reviewed studies on AIR efficacy. There are several mitigation measures that we know are beneficial but their efficacy is variable. These include handheld FLIR surveys and personnel training. While the benefits of these measures cannot be quantified, and are therefore not included in quantitative take estimates, they can be included as measures that can go toward meeting the least practicable adverse impact standard. We have also clearly described why other possible mitigation measures may not be practicable. The mitigation measures incorporated into the Request are similar to those implemented under current incidental take authorization and have proven to be successful in minimizing impacts. To ensure the least practicable adverse impact, we evaluated various additional mitigation strategies and imposes one additional mitigation measure in these ITRs.

Comment (186): A commenter stated the proposed ITR does not ensure the least practicable adverse impact and comply with the MMPA's protective intent due to its insufficient restrictions on seismic surveys and drilling operations during sensitive periods such as denning and migration, its inadequate monitoring requirements that rely heavily on unreliable visual detection, and a lack of enforceable, real-time shutdown protocols when animals are detected near operations.

Our Response: A mitigation measure preventing seismic exploration during denning season would not be practicable, as it would completely prevent seismic exploration. Polar bears do not migrate, however during the critical period when sows are headed to the sea ice with their cubs, the FWS requires LOA holders to cease activity to provide the family group with an unimpeded path to the sea ice. We have incorporated measures requiring notification and coordination in the event of polar bear injury or death.

Comment (187): A commenter stated that it seems difficult to guarantee impacts will be minimal given the long-term effects of habitat disruption and increased industrial activity. Another commenter questioned whether the least practicable adverse impact standard could be met, given the current ecological conditions in the Arctic and the scientific evidence regarding the vulnerability of polar bears and walruses.

Our Response: We have presented a multifaceted discussion on the potential impacts to polar bears and walruses from the specified activities. This included a discussion of polar bear and walrus habitat use. The proposed ITR area is outside the range of Pacific walruses. The footprint of the specified activities does overlap with polar bear habitat, however large untouched areas of habitat remain. Thus, we feel potential habitat disruption will not prevent the impacts from being negligible.

Comment (188): One commenter asserts that the FWS treats data deficient areas as if the areas are biologically less significant than areas with greater amounts of data.

Our Response: Our predictive modeling is designed to account for areas where there has been little to no human activity and are therefore data deficient. We use relationships derived from data rich areas, such as harassment rates, spatial impact zones, and encounter rates, as well as data from outside sources such as collar data and satellite-derived habitat maps, to make an informed estimate of the harassment anticipated in newly occupied areas.

Comment (189): One commenter requests that the FWS withdraw or reject the proposed ITR and adopt a more precautionary, science-based approach that ensures enforceable protections for subsistence communities.

Our Response: The FWS's analysis is based on the best available scientific evidence. We do not anticipate the specified activities will overlap with polar bear harvest, because the vast majority of specified activities are located in an active oilfield where hunting is generally prohibited.

Comment (190): A commenter expressed concern about the level of cooperation between the oil and gas industry and Alaska Native communities, stating that without specific rules on what LOA applicants must disclose to potentially affected communities, they may leave out important information. They requested FWS provide specific information on what it will consider to be adequate safeguards in Plans of Cooperation that ensure the least practicable adverse impact on the availability of walruses and polar bears for subsistence use.

Our Response: The requirements for a plan of cooperation are listed in the proposed ITR. The FWS declines to pre-judge what project- specific safeguards are adequate prior to considering Alaska Native community input.

Comment (191): One commenter recommended that the rule fully consider interconnected impacts between wildlife disturbance and Indigenous community food security, cultural practices, and long- standing relationships with the environment when evaluating disturbance thresholds and mitigation requirements.

Our Response: For reasons explained in its analysis, FWS does not anticipate any impacts to Indigenous community food security, cultural practices, or long-standing relationships with the environment.

Comment (192): Commenters stated the proposed ITR does not demonstrate sufficiently that no unmitigatable adverse impacts to subsistence will occur, particularly as distributions of animals shift and spatial use between subsistence users and industrial activity increases. One commenter stated that effects from industry-caused shifts in animal migration routes should be considered, and they suggested more analysis on impacts to subsistence should occur at the proposed ITR stage as opposed to at a later stage when Plans of Cooperation are developed. Another commenter stated subsistence impacts should be evaluated not only in terms of direct displacement, but also in terms of ecological disruption and unpredictability.

Our Response: There is no scientific evidence to suggest shifting distributions of polar bears away from where subsistence harvest occurs. Polar bears are still primarily drawn to subsistence whaling locations each autumn (i.e., Cross Island, Kaktovik) where much of the subsistence harvest of polar bears occurs. The FWS has not noted any changing movement patterns of polar bears along the coast that would indicate their movements have been hindered at the scale of seasonal movements along the coast. Industrial activities have occurred in the region for over 50 years so far and there has been no changes noted in the ability of local communities to harvest polar bears.

Comment (193): One commenter recommended that FWS solicit additional scientific review, local and Indigenous Knowledge, and expertise to refine and validate its assumptions and methodologies to ensure that estimated take levels and potential impacts to polar bears are accurately described and appropriately quantified. The commenter also stated that the FWS's analysis for take is overly precautious and that an overestimate of potential impacts on polar bears could create an unfounded conflict between industry activities and continued subsistence harvest of marine mammals.

Our Response: The FWS has frequently invited comment and discussion concerning its assumptions and methodologies and looks forward to further opportunities to continue these discussions. We does not view our analysis as overly precautious and finds that Industry activities do not conflict with continued subsistence harvest of marine mammals.

Comment (194): One commenter stated that the FWS should require the requester to notify any potentially impacted subsistence community rather than just the subsistence communities that the requester identified in their Request.

Our Response: There are limited subsistence communities in the proposed ITR area. The nearby communities of Utqiagvik, Nuiqsit, and Kaktovik are regularly notified of LOA holders' activities and would be notified when appropriate under these ITRs.

Comment (195): One commenter recommended the FWS seek to promote consistency and predictability in the issuance and administration of LOAs since clear and standardized expectations across operators enhance compliance, facilitate planning, and

reduce the administrative burden associated with case-by-case variation.

Our Response: Each LOA issued pursuant to these ITRs will include ITR-required mitigation measures and monitoring and reporting requirements. FWS works closely with LOA applicants to ensure they have the information needed to submit successful, complete LOA applications.

Comment (196): The requester states 8 Star Alaska, LLC should be eligible for coverage for LOAs under these ITRs regardless of their AOGA membership status.

Our Response: The preamble text specifically identifies that 8 Star Alaska, LLC is an eligible party to request an LOA under these ITRs.

Comment (197): One commenter requested that the FWS withdraw or reject the proposed ITR and adopt a more precautionary, science-based approach that applies conservative thresholds consistent with statutory intent.

Our Response: The FWS has relied on science-based approaches to reach negligible impact and small numbers determinations, as is outlined in its analysis and supplemental information.

Comment (198): One commenter stated that the proposed regulations appear to violate the 1973 International Agreement on the Conservation of Polar Bears.

Our Response: The FWS is not authorizing anything that qualifies as a “taking” as that term is defined by the 1973 Agreement on the Conservation of Polar Bears. Article VI.2 of the 1973 International Agreement on the Conservation of Polar Bears states that the Agreement shall not prevent a Contracting Party from maintaining or amending existing legislation or other measures or establishing new measures on the taking of polar bears so as to provide more stringent controls than those required under the provisions of this Agreement. Consequently, the proposed regulations do not violate the terms of the 1973 Agreement.

Comment (199): One commenter stated that the proposed ITR should not move forward until lawsuits related to recent National Petroleum Reserve-Alaska lease sales are resolved.

Our Response: The FWS declines this suggestion. ITRs do not authorize oil and gas activities themselves, they only facilitate the authorization of incidental, unintentional take of small numbers of polar bears and Pacific walruses during certain oil and gas industry activities. Resolution of those lawsuits will neither inform, nor impact the validity of, these ITRs.

Comment (200): One commenter states the FWS should evaluate whether polar bears in the SBS stock are demographically independent from polar bears in adjacent areas, because MMPA authorizations are based on stock boundaries and the stock estimates affect small numbers and negligible impact determinations.

Our Response: The FWS has previously assessed the data that exist to inform stock boundary delineations between the CS and SBS stocks. These were provided in the most recently published CS and SBS stock assessment reports. Further, there is ample evidence from multiple, independent, lines of data to support the current boundary separating the CS and the SBS. These included assessments of movement data (e.g., Scharf et al. 2019), capture data (i.e. very limited [i.e., 1-2] recapture of SBS bears in the CS and vice versa), den locations (Rode et al. 2015, Patil et al. 2022), summer land use locations (Rode et al. 2015, Atwood et al. 2016), demography and responses to sea ice loss (Rode et al. 2014, 2018), and stable isotope data (Smith et al. 2022). The commenter is therefore incorrect that the boundaries are largely based on harvest management. Although the boundaries are used for that purpose (one of many purposes), they were not developed for that purpose.

Comment (201): One commenter states there are studies that indicate there is a potential for meaningful dispersal between the SBS and adjacent stocks. They cite studies indicating broad areas of overlap between stocks (Amstrup et al. 2004, Amstrup et al. 2005), SBS bears denning within the bounds of adjacent stocks (Olson et al. 2017), and spatial overlap of GPS-tracked bears (Scharf et al. 2019). They state these studies indicate there are not strong barriers to dispersal between stocks, and there may be relatively high rates of dispersal between the stocks. Finally, they state that if rates of dispersal are high between currently identified stocks, they may not be demographically independent units that should be treated as separate stocks.

Our Response: It is true that there can be overlapping space use between the CS and SBS, but often this overlap is not accurately reflected by static maps (e.g., Amstrup 2004, 2005) because often one group has already moved away from that area before the other arrives in meaningful numbers. This is likely the case for the SBS and CS subpopulations based on sea ice dynamics in the region. Retreating sea ice in the CS causes most bears to move northwest towards Wrangel Island, whereas ice retreat in the SBS tends to move north or northeast, away from the CS bears. There are only rare instances of a collared CS bear moving beyond a “mixing zone” into the heart of the SBS stock boundary. The same is true for the SBS moving into the CS stock boundary. Denning (based on animals collared in their respective stock boundaries) in the CS primarily occurs on Wrangel Island (Rode et al. 2015), whereas in the SBS it occurs on the northern Alaska coastal plain or the sea ice north of Alaska. Similarly, land use in summer for bears collared in the CS is largely in Russia, whereas SBS bears collared inside the SBS boundary largely summer on land in northern Alaska. Lastly, only 1-2 bears captured in the CS have been recaptured in the SBS. The same is true for SBS bears being captured in the CS. If there was significant interaction between these two stocks, we wouldn't expect to see this much space use differentiation and would further expect significantly more captures of bears outside of their respective stock boundaries. Thus, based on the commenters own metrics, we do not believe there are high rates of dispersal between the two subpopulations, further supporting their discreteness.

Comment (202): Some commenters stated that the proposed ITR is not consistent with the MMPA objective of preventing stocks from falling below their optimum sustainable population level because of the risk to denning mothers and cubs and ongoing decline of the population. Commenters also stated that an impact that is not consistent with the stock reaching or maintaining its optimum sustainable population cannot rationally be deemed negligible.

Our Response: The best available science indicates the population decline that was noted in the mid-2000s has appeared to stabilize (Bromaghin et al. 2021). The FWS does not anticipate lethal take occurring and thus there would be no anticipated reduction in the population to move the subpopulation away from OSP. Further, we disagree that any additional take of SBS bears obstructs recovery. Regehr et al. (2017) showed that a polar bear population declining due to sea ice loss can still have a level of removal that would not hinder the persistence of the population.

Comment (203): One commenter stated the Endangered Species Act will be violated should the proposed ITR become final.

Our Response: Nothing that could be authorized pursuant to the ITR would violate the ESA.

Comment (204): One commenter states that all authorized and ongoing

take of polar bears, including incidental and subsistence take from the United States and Canada, as well as unknown mortality events, must be incorporated into the environmental baseline for the FWS's analysis of take. The commenter further states that any Level A takes of polar bears resulting from oil and gas activities can reasonably be expected to adversely affect the SBS stock, have a negative population level impact, and adversely impact the availability of the SBS stock for subsistence use.

Our Response: We analyzed the impacts of the specified activities in light of the environmental baseline, which includes the factors noted by the commenter.

Comment (205): One commenter states that for the FWS to comply with its section 7 consultation requirements it must consult on multiple gas and oil related impacts including seismic exploration, construction of oil and gas production and development facilities, as well as impacts from the construction, use, and maintenance of access routes, aircraft and vehicle traffic, field crews and other human activities, and noise disturbance.

Our Response: The FWS has completed intra-service section 7 consultation for the final ITRs. The consultation considered the factors noted by the commenter.

Comment (206): One commenter suggested that the FWS should reassess its take estimate to ensure that AOGA's petition does not take polar bears in violation of Section 9 of the ESA.

Our Response: Through our section 7 consultation and corresponding incidental take statement, the FWS has ensured polar bear harassment will not violate section 9 of the ESA.

Comment (207): One commenter recommended the FWS strengthen protections for polar bear denning habitat and walrus haulouts and prioritize the long-term conservation of Arctic wildlife and ecosystems.

Our Response: The FWS is not aware of any walrus haulouts in the specified project area. Polar bear denning habitat (i.e., areas with suitable terrain to capture snow sufficient for creating a den) is widely distributed across the north slope of Alaska. Even with the proposed activities, a large area of denning habitat will remain available. Only new permanent infrastructure could modify existing denning habitat, but would likely create new habitat through the creation of elevated pads/roads. Denning habitat would not be affected that occurs along tundra travel routes and ice roads because those would be created after denning was initiated by a female and would not be altered such that it couldn't be used in subsequent years. The ITRs include measures to ensure the least practicable adverse impacts on SBS polar bears and walruses and their habitat. We do not view the ITRs as inconsistent with the long-term conservation of Arctic wildlife and ecosystems, and notes that the mitigation, monitoring and reporting requirements prescribed by ITRs have a conservation benefit.

Comment (208): Commenters stated that the proposed ITR did not sufficiently consider potential impacts of repeated disturbances of polar bears or Pacific walruses across years, including potential displacement and loss of functional habitat. One commenter stated that multiple exposures of the same polar bear within and across authorizations and activities can increase the likelihood of a response rising to the level of an incidental take and/or can increase the severity of the response (e.g., making an injury or serious injury and mortality more likely than from a single exposure). Other commenters stated that impacts from repeated disturbances cannot be reasonably considered as negligible. They argued that even low levels of repeated disturbance can lead to long term population consequences, particularly for populations that reproduce slowly and are already facing other pressures, including limited energetic flexibility.

Our Response: The FWS's analyses take into account climate-related changes in the distribution of bears (i.e., during denning and summer land use) as well as the most recent estimate of the size of the subpopulation to reach its negligible impact determination. We consider these types of disturbance in our analyses and they are classified accordingly into Level B and Level A harassments. We are not aware of any scientific literature or data for polar bears that indicated repeated low-level disturbance (i.e., Level B harassment) can have long-term population consequences.

Comment (209): Commenters stated more analysis should be completed on how the proposed ITR will affect polar bears at longer time intervals, possibly through the application of additional climate change projections, extending the regulatory timeframe by 10 to 20 years, and incorporating cumulative impacts analysis into projections.

Our Response: Per the MMPA, FWS cannot issue ITRs for periods exceeding 5 years (outside of the military readiness context). As we describe in our analysis, impacts to polar bears from Level B and Level A harassment are expected to be temporary.

Comment (210): One commenter stated any incidental take estimate included in the proposed ITR should explicitly account for detection limitations, unobserved interactions, environmental conditions that reduce monitoring effectiveness, and the likelihood and consequences of repeated exposures to the same individuals within and across years from the activities within the scope of the proposed ITR and other sources of human-caused impacts.

Our Response: The FWS agrees that accounting for 1018-0070, Incidental Take of Marine Mammals During Specified Activities (50 CFR 18.27 and 50 CFR 18, Subpart J) (expires 08/31/2028), and

1018-0203, Incidental Take of Marine Mammals During Specified Activities (50 CFR 18.27 and 50 CFR 18, Subpart L) (expires 08/31/2028).