Taking and Importing Marine Mammals; Taking Marine Mammals Incidental to Geophysical Surveys in the Gulf of America

NMFS published a notice of proposed rulemaking in the Federal Register on February 24, 2026 (91 FR 9014), beginning a 30-day comment period. In that notice, we requested public input on the proposed rule, including but not limited to NMFS' proposed or preliminary findings, determinations, or conclusions regarding the MMPA standards, and the information NMFS relies on in support of those findings, determinations, or conclusions; and NMFS' preliminary decision to not make changes to the 2021 final rule, and the information NMFS relies on in support of those preliminary decisions, and requested that interested persons submit relevant information, suggestions, and comments.

During the 30-day comment period, we received nine comment letters. Of this total, we received four letters from various organizations (described below) and five submissions from private citizens. Submissions from private citizens expressed a mix of generic support for or opposition to the proposed issuance of the regulations, as well as generic support for protection of marine mammals. We do not address these comments further. A letter was submitted jointly by EnerGeo, the American Petroleum Institute, the Louisiana Mid-Continent Oil and Gas Association, the National Ocean Industries Association, and the Offshore Operators Committee (hereafter, the “Associations”). A separate letter was submitted jointly by the Natural Resources Defense Council (NRDC), Association of Zoos and Aquariums, Center for Biological Diversity, Earthjustice, Sierra Club, and Turtle Island Restoration Network (hereafter, “NRDC”). Turtle Island Restoration Network (TIRN) also submitted a letter independently, and an additional letter was submitted by Animal Counsel. TIRN's independent letter provided a generic statement in opposition to issuance of the rule, and we do not address it further.

The Associations express support for issuance of the reimplemented ITRs, stating that “the record should reflect that the request is also made on behalf of the regulated industry participants.” The Associations emphasize that they preserve all previously expressed objections to the current ITRs, urge prompt issuance of new ITRs based on the EnerGeo petition, and request that requested LOAs be issued promptly under the reimplemented ITRs. NMFS does not further address the Associations' letter.

NMFS has reviewed all public comments received on the 2026 proposed rule. All relevant comments and our responses are described below. All comments received are available online at: https://www.regulations.gov. A direct link to these comments is provided at: https://www.fisheries.noaa.gov/action/incidental-take-authorization-oil-and-gas-industry-geophysical-survey-activity-gulf-america.

Comment 1: Animal Counsel indicated a belief that NMFS' proposal to issue the rule contradicts Congressional intent behind the MMPA and would be incompatible with the original intent of the MMPA. Animal Counsel additionally stated that the legislative history of the MMPA makes clear that the precautionary principle must be applied and bias must favor marine mammals.

Response 1: NMFS believes that this action is fully consistent with the original public meaning of the relevant provisions of the statute, and NMFS disagrees that this action contradicts any requirement of the MMPA or is contrary to Congressional intent as expressed through the enactment of the relevant provisions of the statute. Neither the MMPA nor NMFS' implementing regulations include references to, or requirements for, the precautionary approach, nor is there a clear, agreed-upon description of what the precautionary approach is or would entail in the context of the MMPA or any specific activity. Nevertheless, the MMPA by nature is inherently protective, including the requirement to mitigate to the lowest level practicable (“least” practicable adverse impacts, or “LPAI,” on species or stocks and their habitat). This requires that NMFS assess measures in light of the LPAI standard. To ensure that we fulfill that requirement, NMFS considers all potential measures (e.g., from recommendations or review of available data) that have the potential to reduce impacts on marine mammal species or stocks, their habitat, or subsistence uses of those stocks, regardless of whether those measures are characterized as “precautionary.”

Comment 2: Animal Counsel requests that NMFS extend the comment period for the proposed rule to a minimum of 60 days and hold public hearings. The request is primarily founded on the contention that an announced March 2026 meeting of the Endangered Species Committee demands delay, with Animal Counsel stating that the meeting “is likely to drastically change the outlook for marine mammals in the Gulf.” Elsewhere in its letter, Animal Counsel highlights the independence of the MMPA from the Endangered Species Act (ESA).

Response 2: We do not find Animal Counsel's rationale persuasive; NMFS finds the 30-day comment period for the proposed rule sufficient. Supporting material for this rule has been available for review during public comment periods for prior rulemakings associated with oil and gas industry geophysical survey activity in the GOA (see 83 FR 29212, June 22, 2018; 88 FR 916, January 5, 2023). This rule covers the same activity and reimplements the measures that were considered in those rulemakings, which were the subject of extensive comment. As to the Endangered Species Committee meeting, we disagree that the Committee's March 31, 2026, action to exempt GOA oil and gas activities from the requirements of ESA section 7 is likely to drastically change the outlook for marine mammals in the GOA. Specifically, the oil and gas activities considered by the Committee included oil and gas exploration, development, and production activities in the Gulf as well as the avoidance or minimization measures that are described in the agency action analyzed in NMFS' 2025 biological opinion and in the U.S. Fish and Wildlife Service's 2018 and 2025 consultation decisions. The Committee's exemption order explicitly states that those avoidance and minimization measures shall continue to be implemented. See Dep't of the Interior, March 31, 2026, Endangered Species Committee Order, https://www.doi.gov/endangered-species-committee. Thus we do not agree that the Committee's order has drastically changed the outlook for marine mammals in the GOA such that we should extend the public comment period or hold public hearings on that basis.

Comment 3: NRDC claims that use of BOEM's programmatic estimates of ongoing geophysical survey activity is arbitrary. NRDC's claim is based on the age of BOEM's original projections, which were initially included in its 2017 PEIS issued under the National Environmental Policy Act (NEPA) before being revised as part of BOEM's 2020 revision of the geographic scope of the program addressed in NMFS' 2021 final rule, and on what NRDC describes as “recent developments that are likely to influence industry development of the Gulf,” including prospective future schedules for offshore oil and gas lease sales.

Response 3: NMFS disagrees with NRDC's claim. The programmatic

estimates of ongoing geophysical survey activity included in this rule remain a reasonable representation of ongoing and future activity over the 5-year duration of this rule. While the 10-year projections in the PEIS indeed extended through 2027, they were based on generic expectations regarding BOEM's program over the 10-year window of analysis provided at that time, not on specific understanding of conditions dictating oil and gas industry levels of effort. When revising its level of effort projections in 2020 in support of its 2016 petition for ITRs, BOEM affirmed that the projections should be considered an adequate representation of generic conditions informing the need for survey activity in the GOA. Subsequently, EnerGeo (an industry trade organization dedicated to the geophysical and exploration industry and best positioned to understand future industry demand for survey activity) affirmed the validity of these projections by adopting them as an ongoing representation of future survey activity in its 2025 petition for new ITRs. We also note that actual survey effort under the current ITRs from 2021 to 2025 has been less than the association projections, suggesting that the projections remain reasonable and accommodating of potential future fluctuations in demand.

NRDC's references to new developments relating to schedules for future lease sales are not relevant to our analysis of expected survey activity during the course of the next 5-year period. Future lease sales may influence survey activity beyond the next 5-year period but are unlikely to increase demand on the timeframe imagined by NRDC. NRDC also suggests that speculative future lease sales in areas outside of the geographic scope of this rule, i.e., in the eastern GOA, would invalidate the level of effort projections considered here. However, should this demand in the eastern GOA arise, it would necessarily be subject to additional MMPA incidental take authorization requests, as the area is not covered under this rule.

Comment 4: NRDC expressed concerns regarding NMFS' use of the probabilistic response function described by Wood et al. (2012), in which 10 percent, 50 percent, and 90 percent of individuals exposed are assumed to produce a behavioral response (of a sufficient degree of severity to constitute Level B harassment) at exposures of 140, 160, and 180 dB root mean square (rms), respectively. (The function is shifted for the more behaviorally sensitive beaked whales such that 50 percent and 90 percent response probabilities are assumed to occur at 120 and 140 dB rms, respectively.) NRDC stated that the function is inconsistent with the best available science, asserting that behavioral disruptions occur at higher percentages at lower noise exposure levels than those suggested by Wood et al. (2012). NRDC's criticism of the function also focused on the use of horizontal displacement studies as the supposed basis of analysis for Wood et al. (2012), as well as on the function's nature as a series of step functions. In addition, NRDC expressed concerns that the use of frequency weighting in the Wood et al. (2012) approach is inappropriate. Overall, NRDC claims that reliance on this function results in underestimation of impacts and is arbitrary and capricious. Similarly, Animal Counsel urges NMFS to adopt different harassment thresholds.

Response 4: NMFS has historically been criticized for the use of the single-step 160-dB rms approach. Those criticisms are based on the idea that an approach reflecting a more complex multi-step probabilistic function would more effectively represent the known variation in responses at different levels due to differences in the receivers, the context of the exposure, and other factors, as well as the science indicating that animals may react in ways constituting Level B harassment when exposed to lower received levels. In developing the acoustic exposure analysis for the 2021 rule, we reviewed relevant past public comments as well as the best available science, determining that a more complex probabilistic function is indeed better reflective of available scientific information and that it was appropriate to take the fundamental step of recognizing the potential for Level B harassment occurring at exposures to received levels below 160 dB rms (as well as the potential for no Level B harassment occurring at exposures above 160 dB rms). This approach also accounts for differential hearing sensitivity by incorporating frequency-weighting functions, as behavioral responses in cetaceans are best explained by the interaction between sound source type and functional hearing group (Gomez et al., 2016). NMFS has determined that the general approach used for this rule--a probabilistic risk function that allows for the likelihood of differential response probability at given received levels on the basis of multiple factors, including behavioral context and distance from the source, and that addresses particularly sensitive species--is appropriate in light of the best available scientific information.

However, because behavioral responses to sound depend on the context in which an animal receives the sound, including the animal's behavioral mode when it hears sounds, prior experience, additional biological factors, and other contextual factors, predicting when sound levels will be likely to disrupt behavioral patterns is extremely difficult. Even experts have not previously been able to suggest specific new criteria due to these difficulties (e.g., Southall et al., 2007; Gomez et al., 2016). While we acknowledge our approach reduces a complex suite of interactions to make reasonable inferences, it is consistent with the best available science.

NRDC referenced “recent” research they claim is not consistent with the recommendations of Wood et al. (2012). We note that, of the nine studies cited by NRDC, five were published prior to the Wood et al. (2012) study, and were therefore available for those authors' consideration (and some were specifically referenced by those authors in discussion of their recommendations). Further, we disagree that the referenced findings are inconsistent with Wood et al. (2012). For example, a detectable change in motion or vocalization in response to noise exposure does not necessarily mean that a take by Level B harassment under the MMPA has occurred, given that the definition focuses on “disruption of behavioral patterns.” NRDC also suggests that some of these studies were not incorporated into Wood et al.'s recommendations, or our consideration of those and other potential approaches in context of the available science, and criticize what they view as an over-reliance on horizontal displacement studies as the supposed basis of analysis. While it is true that the majority of available behavioral data focus on avoidance responses, Wood et al. (2012) does not mention excluding behavioral studies involving vocal changes, and Southall et al. (2007) specifically incorporates numerous studies that do mention changes in vocalization associated with sound exposure. Thus, these datasets were not excluded and as discussed in our notice of proposed rulemaking, we adequately considered all studies addressed by NRDC.

Regarding baleen whales, we acknowledge that changes in vocalization have been observed in association with exposure to airgun surveys within migratory and non-migratory contexts (e.g., Castellote et al., 2012; Blackwell et al., 2013; Cerchio et

al., 2014). The potential for such effects to occur over relatively large spatial scales is not surprising for species with large communication spaces (e.g., Clark et al., 2009), but we reiterate our disagreement with NRDC's apparent contention that every detected change to vocalizations rises to the level of a take. NRDC cites reports of changes in vocalization, typically for baleen whales, as evidence in support of lower thresholds, claiming these reactions result in biological consequences indicating that the reaction was indeed a take. However, NMFS is not aware of research that provides a well-supported link between the reported reactions at lower received levels and the putative consequences. In conflict with NRDC's interpretation of the literature are documented instances of marine mammal exposure to greater received levels that did not elicit any response (e.g., Malme et al., 1983, 1984, 1985, 1988; McCauley et al., 1998, 2000a, 2000b; Barkaszi et al., 2012; Stone, 2015a; Gailey et al., 2016; Barkaszi and Kelly, 2018).

The received level associated with stoppage of calling for bowhead whales (Balaena mysticetus) observed by Blackwell et al. (2013, 2015)-- a response that may arguably rise to the level of harassment--is consistent with the Wood et al. (2012) scheme, in which the potential for take upon exposure to received levels as low as 140 dB is accounted for. Similarly, the findings of Pirotta et al. (2014) for harbor porpoise (Phocoena phocoena) are consistent with the treatment of behaviorally sensitive species by Wood et al. (2012), in which the potential for take at even lower received levels is accounted for (though irrelevant here, as harbor porpoise are not found in the GOA). The response levels reported by McDonald et al. (1995) and Di Iorio and Clark (2009) for blue whales (Balaenoptera musculus) also comport with the Wood et al. (2012) function, if we assume that the observed responses equate to harassment (though it is not clear that they do). With regard to NRDC's citation of Clark and Gagnon (2006), a non-peer reviewed white paper, NRDC incorrectly calculated the area over which the effect was observed by an order of magnitude (the paper discusses an area of 100 x 100 nmi, which equates to 10,000 nmi\2\--not 100,000 nmi\2\).

In regard to Cerchio et al. (2014), it is important to note that received levels provided in this study are those recorded at locations of their underwater recording devices. The authors indicated “we did not have the ability to locate the singers or the seismic survey vessel, estimate the source level of the pulses, the distance between the source and potentially impacted singers, or the received level of the pulses at the singers.” The same situation, i.e., actual received levels at the location of the animals are unknown, is true for Castellote et al. (2012) and Clark and Gagnon (2006), which provide average background sound levels with and without the presence of airgun surveys. Thus, not having the location of the animals at the time of exposure makes it difficult to draw conclusions based strictly on received level. NMFS has evaluated the papers and determined they are not informative about appropriate Level B harassment thresholds.

Regarding sperm whales, NMFS disagrees that assuming a 100 percent probability of take of sperm whales upon exposure to survey noise at 135 dB--as suggested by NRDC--is an accurate reflection of the results of the Miller et al. (2009) study. While we agree that the work of Miller et al. (2009) suggests that sperm whales in the GOA may be susceptible to disruption of foraging behavior upon exposure to relatively moderate sound levels, NRDC incorrectly interprets results of the study in claiming that sperm whale “foraging success” was found to “decline significantly.” Instead, the authors report that buzz rates (a proxy for attempts to capture prey) were approximately 20 percent lower, meaning that the appropriate interpretation would be that foraging activity (versus foraging success) was reduced by 20 percent (Jochens et al., 2008). Of the eight whales tagged in that study, only one was observed to actually cease foraging.

Moreover, while we do believe that these study results support a conclusion that exposure to survey noise can impact foraging activity, others have interpreted them differently, e.g., by focusing on the finding that exposed whales did not change behavioral state during exposure or show horizontal avoidance (a finding replicated in other studies, e.g., Madsen et al., 2002a; Winsor et al., 2017). Importantly, the observed effect was not statistically significant and, as reported by the authors, constituted “subtle effects on their foraging behavior.” Furthermore, the authors of the Wood et al. (2012) study explicitly described their consideration of Miller et al. (2009) in the development of their recommended criteria. Therefore, the Wood et al. (2012) recommendation is indeed consistent with the Miller et al. (2009) study. In referencing Bowles et al. (1994), NRDC fails to state that the observed cessation of vocalization was likely in response to a low-frequency tone (dissimilar to airgun signals), though a distant airgun survey was noted as producing signals that were detectable above existing background noise.

Regarding other odontocetes, NRDC's representation of the available scientific information is also inaccurate. Miller et al. (2005) specifically state that “[s]ighting rates at distances of 10-20 km from the airgun array were significantly lower than those in areas 20- 30 km from the airgun array, where sighting rates were unexpectedly high” (i.e., the study indicates sighting rates of beluga whales (Delphinapterus leucas) were lower, not “100% avoidance” as claimed by NRDC). Miller et al. (2005) reported seven aerial beluga whale sightings from 8 to 18 km from the survey vessel and two vessel-based beluga whale sightings at 1.5 and 2.5 km from the survey vessel. Furthermore, Southall et al. (2007) described the findings of the Miller et al. (2005) study as temporary avoidance behaviors at these lower received levels, while Gomez et al. (2016) (which NRDC agrees reflects the best available science) evaluated Miller et al. (2005) based on a received level of 150 dB. Thus, the Wood et al. (2012) approach does capture responses associated with this study.

Additionally, Wood et al. (2012) has the advantage of accounting for sensitive species such as beaked whales, meaning that a response of a beaked whale at 140 dB (as cited by NRDC) is covered within the Wood et al. (2012) recommended criteria (e.g., Wood et al. assumes 90 percent of an exposed beaked whale population will respond at 140 dB).

It should be noted that the systematic review by Gomez et al. (2016), cited by NRDC and Animal Counsel in support of their positions, found that received level was not appropriate as the sole indicator of behavioral response. For example, this review shows that “low” effects were actually found to reach peak probability at a higher received level than “moderate” effects for baleen whales. The results of the Gomez et al. (2016) review are not inconsistent with Wood et al. (2012). With regard to NRDC's comment that the authors consider their results “non-conservative,” Gomez et al. (2016) only indicates that they may have scored the severity of vocal responses higher if they had more information on the ecological significance of these types of responses. There is no indication elsewhere in Gomez et al. (2016) that their overall results and analysis are “non-conservative.” Commenters also

reference the “more conservative approach” of Nowacek et al. (2015), suggesting that this approach is more appropriate. However, the more substantive reviews presented by Southall et al. (2007) and Gomez et al. (2016) were unable to present any firm recommendations, as noted above. We note that Animal Counsel recommends that NMFS update thresholds to “reflect the findings of Gomez et al. (2016) and Nowacek et al. (2015),” but provide no specific recommendations. We addressed Gomez et al. in further detail in this response, and note that NMFS has addressed the brief, scientifically unsupported recommendation of Nowacek et al. (2015) in detail in prior comment responses (e.g., 86 FR 5322, January 19, 2021).

NRDC also criticizes the use of weighting functions in evaluating potential Level B harassment, and specifically criticizes use of the M- weighting scheme of Southall et al. (2007). Gomez et al. (2016) suggest that incorporation of frequency-weighting is necessary to account for differential hearing sensitivity, as behavioral responses in cetaceans are best explained by the interaction between sound source type and functional hearing group. That is, implementing weighting functions allows for consideration that different marine mammal groups do not hear varying frequencies of sound equally well. Thus, it is appropriate to account for sounds below a group's best hearing range having a lower likelihood of resulting in a behavioral response (let alone that animals are likely unable to effectively detect sounds at frequencies completely outside their hearing range).

The M-weighting functions are described in Southall et al. (2007) as “intentionally precautionary (wide)” (as opposed to the weighting functions used in NMFS' Technical Guidance to account for noise-induced hearing loss) and are used to account for the functional hearing ranges of different marine mammal hearing groups. This frequency weighting scheme (i.e., M-weighting) was intentionally selected because it is more conservative in accounting for hearing sensitivity (as is appropriate in evaluating potential Level B harassment) than are more recently developed filters designed to better assess potential noise- induced hearing loss.

NRDC asserts that because M-weighting assumes that high- and very high-frequency (HF and VHF) cetaceans are relatively insensitive to noise below 1 kHz, it is likely that the incorporation of M-weighting has a significant downwards effect on take estimates. This is incorrect. As demonstrated in NMFS' 2021 final rule, at 250 Hz and above, the M-weighting functions do not result in a significant reduction for HF or VHF cetaceans. Furthermore, the lower bound of the functional hearing range of these groups is 150 Hz for MF cetaceans and 200 Hz for HF cetaceans (i.e., sounds below 100 Hz, where most energy in airgun noise is found and where M-weighting results in the greatest reductions, are outside functional hearing range). At 1 kHz, where these species are most likely to be able to detect and respond to airgun noise, there is very little assumed reduction in sensitivity.

Finally, NRDC advocates for the use of a linear risk function as opposed to the multiple step function of Wood et al. (2012), stating that linear risk functions are scientifically accepted methodology that better acknowledge individuals may vary in responsiveness. Although NRDC does not specifically define what they mean by “linear risk function,” NMFS assumes a linear risk function is a smooth, continuous function, as opposed to a function defined by multiple steps, as is the case of Wood et al. (2012). NRDC states that Wood et al. (2012) “has a significant negative bias on take estimates” where “all exposures from 140 dB to 159.9 dB are considered to produce the same risk.” Wood et al. (2012) does acknowledge that responsiveness varies with received levels, while relying on broad steps, rather than a continuous function. These broad steps allow for easier implementation of a risk function and are more practical for most users, which is an important consideration, especially in the context of users that may not have the ability or access to more sophisticated modeling (in contrast to the Navy). Therefore, if new linear risk functions become available, NMFS may still provide a more simplistic function broken down in broad steps, so that it can be applied by all users.

Overall, we reiterate the lack of scientific consensus regarding what criteria might be most appropriate for evaluating Level B harassment. Defining sound levels that disrupt behavioral patterns is difficult because responses depend on complex, difficult to predict contextual factors much more so than received level. Therefore, levels at which responses occur are not necessarily consistent and can be difficult to predict. However, although better methods of assessing likely behavioral response to acoustic stimuli than the relatively simple multi-step function used here may be forthcoming from the scientific community, NMFS has determined that the recommendations of Wood et al. (2012) adequately reflect the available science, and commenters offer no alternative recommendation.

Comment 5: NRDC raised concerns regarding NMFS' Acoustic Technical Guidance for evaluating the potential for marine mammal hearing loss, claiming that the Technical Guidance is not based on the best available science and leads to underestimation of potential auditory effects. NRDC also asserts that NMFS fails to account for alternative mechanisms of auditory injury, stating that permanent hearing loss (i.e., Level A harassment) may occur through mechanisms other than PTS. NRDC states that NMFS must account for these mechanisms in its assessment of potential injury. Animal Counsel similarly expresses concern regarding the potential for accumulated auditory effects resulting from repeated temporary threshold shift exposures.

Response 5: We first acknowledge our use of the 2018 Revised Technical Guidance (NMFS, 2018) in evaluating the potential for auditory effects in the context of estimating take and refer the reader to the Estimated Take section below for discussion of NMFS' use of the 2018 guidance for take estimation. Moreover, NMFS did consider the 2024 Technical Guidance. However, NRDC's critique relates to aspects of the Technical Guidance that predate the 2024 update and, indeed, NMFS considered similar comments from NRDC for issuance of the 2021 final rule.

Regarding the updated 2024 Technical Guidance, the 2024 updated acoustic thresholds (cumulative sound pressure level and peak sound pressure level) for cetaceans either are identical or are higher than those in the 2018 Revised Technical Guidance (i.e., level of exposure has to be higher to result in PTS or AUD INJ compared to the 2018 thresholds). Additionally, the weighting functions were also updated, but this update only resulted in minor changes for LF cetaceans and no changes for the other two cetacean hearing groups, where the peak sound pressure level metric thresholds, which are unweighted, result in the largest isopleths. Furthermore, despite these changes, the Level A harassment isopleths for LF cetaceans are still expected to be within the shutdown zone and would not result in a change in the amount of Level A harassment take expected for this species or any other cetacean in the Gulf.

Each of NMFS' Technical Guidance documents (NMFS, 2016, 2018, 2024) is a compilation, interpretation, and

synthesis of the scientific literature that provides the best available information regarding the effects of anthropogenic sound on marine mammals' hearing. In 2016, the initial iteration of the Technical Guidance was classified as a Highly Influential Scientific Assessment and, as such, underwent three independent peer reviews, at three different stages in its development, including a follow-up to one of the peer reviews, prior to its dissemination by NMFS. In addition, there were three separate public comment periods, during which time NMFS received and responded to similar comments on the guidance (81 FR 51694), and subsequent public and interagency review under Executive Order 13795. While new information may help to improve the guidance in the future, and NMFS will continue to review the available literature to determine when revisions are appropriate, the science underlying the Technical Guidance (2018 and 2024 update) reflects the best available science, and all information received through peer review and public comment. The concerns raised by NRDC have been addressed by NMFS in responses associated with publication of the guidance (see www.fisheries.noaa.gov/national/marine-mammal-protection/marine-mammal-acoustic-technical-guidance).

In light of these considerations, NRDC's argument that estimates based on the guidance “are based on methods and assumptions that have no scientific basis” is unpersuasive. NMFS considers the guidance to represent the best scientific information currently available and, given the incorporation of multiple peer reviews and public comment opportunities during its development, we did not solicit and are not responding in detail to comments concerning the contents of the Technical Guidance (NMFS, 2016, 2018, 2024), as such comments are outside the scope of this rulemaking.

NRDC also referenced information related to occupational noise standards established by the National Institute of Occupational Safety and Health (NIOSH). Human noise risk assessments (NIOSH, 1998) are not equivalent (or applicable) to thresholds provided in the guidance, because they are used to predict hearing loss based on a daily 8-h occupational noise exposure over 40 years (i.e., current marine mammal TTS data are only available to predict exposure periods of 24-h or less and cannot be used to assess or predict risk associated with a lifetime of exposure) and are based on larger sample sizes of human listeners (e.g., NIOSH 1972 and 1997 risk assessments were based on a sample size of 1,172 people). As pointed out in Wright (2015), NIOSH criteria provide a 95 percent confidence interval for their human noise standards but also allow for an excess risk of material hearing impairment, defined as an average threshold elevation for both ears that exceeds 25 dB, of eight percent (i.e., human noise standards limits do allow for some risk; risk is not zero percent and specifically that eight percent of the population is still capable of developing noise-induced hearing loss exceeding 25 dB when exposed to the 85 dB NIOSH level).

Finally, we note that a group of scientists published an update to their original, seminal publication concerning noise exposure criteria to predict the onset of auditory effects in marine mammals (Southall et al., 2007, 2019a), the topic of this comment. The newer peer-reviewed publication evaluates the recommendations of the original publication in light of subsequent scientific findings, including those findings that form the basis for the recommendations of NMFS (2018, 2024). While Southall et al. (2019a) provide recommendations for future research that could lead to revisions, the fundamental aspects of an evaluation of the onset of auditory effects for the marine mammals considered in these ITRs (i.e., auditory weighting functions and noise exposure criteria) are identical to those presented by NMFS (2018, 2024) and incorporated into the modeling process developed for these ITRs.

Regarding NRDC's suggestion that NMFS fails to account for alternative mechanisms of auditory injury, NMFS is aware of the work by Kujawa and Liberman (2009), which is cited by NRDC. The authors report that in mice, despite completely reversible threshold shifts that leave cochlear sensory cells intact, there were synaptic level changes and delayed cochlear nerve degeneration. However, the large threshold shifts measured (i.e., maximum 40 dB) that led to the synaptic changes shown in this study are within the range of the large shifts used by Southall et al. (2007, 2019a) and in NMFS' Technical Guidance to define PTS onset (i.e., 40 dB). There are no data indicating that smaller levels of TTS would lead to similar changes or what may be the long- term implications of irreversible neural degeneration. The effects of sound exposure on the nervous system are complex, and this will be re- examined as more data become available. It is important to note that NMFS' Technical Guidance incorporated various conservative factors, such as a 6-dB threshold shift to represent TTS onset (i.e., minimum amount of threshold shift that can be differentiated in most experimental conditions); the incorporation of exposures only with measured levels of TTS (i.e., did not incorporate exposures where TTS did not occur); and assumed no potential of recovery between intermittent exposures. NMFS disagrees that consideration of likely PTS is not sufficient to account for reasonably expected incidents of auditory injury.

NRDC also suggested that NMFS should “revise its take estimates” following the update to the definition of injury in the 2024 Technical Guidance to include indirectly induced injury in addition to PTS. However, this update in terminology is simply an acknowledgement that in addition to PTS, there could be the potential for auditory injury (which may or may not result in PTS). This updated terminology acknowledges data available for terrestrial mammals and that, since the mammalian cochlea is highly conserved, there is the potential for similar auditory injury in marine mammals. However, this terminology change does not numerically affect our acoustic criteria and, therefore, nothing would change in terms of estimated take (i.e., the criteria remain based on a 40 dB threshold shift, reflecting the same threshold shifts associated with auditory injury seen in terrestrial mammal studies). Thus, numerically the 2018 and 2024 criteria account for the potential for auditory injury in the same way, but the updated guidance has changed the terminology to more directly reflect this.

Comment 6: NRDC commented that the aversion adjustment applied to estimates of Level A harassment proposed by NMFS for low- and very high-frequency cetaceans is not supported. NRDC cites NMFS' statements attributing the original modeling decision not to include a quantitative adjustment for aversion to lack of information regarding species-specific degree of aversion and level of onset as suggestive of lack of an “evidentiary basis” for making such an adjustment.

Response 6: NMFS disagrees with these comments and clarifies our position. Aversion is a known real-world phenomenon. It is well-known that animals will avoid unpleasant stimuli, such as very high received levels of sound. A large and growing literature has demonstrated behavioral aversion in a number of contexts for many marine mammal species in increasingly controlled and well-documented contexts. While considerable species, individual, and

context-dependencies exist in terms of received noise levels associated with behavioral aversion, clear patterns of behavioral aversion have been demonstrated empirically within odontocetes and mysticetes (e.g., Miller et al., 2012, 2014; DeRuiter et al., 2013; Southall et al., 2019b). This is particularly true for exposure scenarios in which animals occur relatively close to sources and at the high levels that would be required for even TTS (much less PTS) to occur. In some instances, in these and other studies, behavioral avoidance has been measured at received levels many orders of magnitude below those required for predicted PTS onset and even below the nominal, 50 percent behavioral response probability at 160 dB rms that NMFS has applied historically.

However, accounting for aversion quantitatively in an acoustic exposure modeling process is a significantly data-heavy endeavor and, despite the growing body of evidence there still not sufficient data regarding the specific degree of aversion and level of onset on a species-specific basis. That is, in order to account for aversion within the modeling process, one must program individual simulated animals (animats) representing different species to respond at a specific received level by changing their direction of travel by a specific degree and assuming a specific rate of speed. Through a test scenario evaluation (discussed in detail in the 2021 rulemaking and in associated modeling reports that support this rule), we determined that while this is possible to do, the specific values that must be used in programming the animat response could not be adequately derived. Instead, a nominal offset factor was applied to the modeled injurious exposures based on published model result evaluation to account for aversion.

Ellison et al. (2016) modeled scenarios using animal movement models to evaluate predicted PTS in which no aversion was assumed relative to scenarios where reasonable assumptions were made about aversion, in line with historical response probability assumptions and that existing scientific literature suggest are appropriate. Scenarios where no aversion probability was used overestimated the potential for high levels of exposure required for PTS by about five times. Accordingly, for the 2021 final rule, total modeled injurious exposures calculated without accounting for behavioral aversion (for low- and high-frequency species) were multiplied by 0.2. NMFS consulted scientific experts (referred to in the 2021 final rule as the “Expert Working Group,” contracted by NMFS and BOEM to provide scientific expertise in developing a risk assessment framework), including the lead author of Ellison et al. (2016), in selecting the specific offset factor, and discussed that selection again in context of the public comments received on that rule. These experts agreed that the approach and specific offset factor was a reasonable approach to addressing the issue of aversion.

NRDC acknowledges that aversion is a meaningful real-world phenomenon that is significantly influential on actual occurrence of Level A harassment, stating that “it is certainly true that some marine mammals will flee the sound.” Ignoring this phenomenon would result in unrealistically high estimates of auditory injury for marine mammals in the GOA. As described above, there is extensive information supporting the aversion concept in marine mammals, but limited quantitative data with which to develop precise, species-specific offset factors. Accordingly, utilizing the available data and expert input, NMFS applied its professional judgement in order to account for this meaningful phenomenon, and has determined it appropriate to retain this approach for the present rule.

Comment 7: Animal Counsel expressed concern that density estimates available for Rice's whale are based on visual survey data and do not include acoustic detections.

Response 7: We agree that ideally, passive acoustic data could be incorporated to the spatial density models to improve the model predictions. However, incorporation of visual and acoustic data to spatial density models remains cutting edge science, and such models have only rarely been produced. The long-term cetacean density modeling effort presented by Roberts et al. (2016) and subsequent, related publications provides an example of the difficulty of doing so. This U.S. Navy-funded effort has been responsible for continually improved iterations of spatial density models for cetaceans along the U.S. East Coast since 2015. However, to date, acoustic data have been incorporated only into models for beaked whales and sperm whales (two species that are most amenable to acoustic surveys and for which acoustic detections are most important to understanding occurrence), and only in the most recently updated model iterations. This required 7 years and a model version 7 for beaked whales and model version 8 for sperm whales (https://seamap.env.duke.edu/models/Duke/EC/). Acoustic data have been used to qualitatively verify density model predictions for certain mysticetes, but have not been incorporated to date into any East Coast mysticete density model. Efforts to evaluate the feasibility and utility of combining visual and acoustic survey data in the GOA have only recently been conducted as a pilot study (Frasier et al., 2021).

We note that the same areas in which the acoustic detections were made are predicted by the spatial density model as being suitable Rice's whale habitat and, in fact, density predictions within areas expected to provide suitable habitat for Rice's whale increased compared with the predictions provided by Roberts et al. (2016) (e.g., Rice's whale density value in Zone 5, which includes areas of the central GOA where acoustic detections were made, increased by 71 percent; see Appendix A of Weirathmueller et al., 2022).

Comment 8: NRDC and Animal Counsel assert that NMFS has failed to account adequately for the effects of masking or stress on marine mammals.

Response 8: The potential impacts of masking were properly considered. We acknowledge that masking may impact marine mammals, particularly baleen whales such as the Rice's whale, and particularly when considered in the context of the full suite of regulated and unregulated anthropogenic sound contributions overlaying an animal's acoustic habitat. We acknowledge that masking can constitute a take, depending on the particular circumstances, but do not agree that masking effects from the incremental noise contributions of individual activities or sound sources always rise to the level of take. Further, not all takes are readily quantifiable. In this case, while masking is considered in the analysis, we do not believe it will result in take of marine mammals beyond those that have already been quantified as taken by behavioral harassment. Specifically, in the case of these proposed activities, in the event that some masking incidents rise to the level of a take, we would expect them to be accounted for in the quantified exposures above the harassment thresholds. Given the short duration of expected noise exposures, any take by masking in the case of these surveys would be most likely to be incurred by individuals either exposed briefly to notably higher levels or those that are generally in the wider vicinity of the source for comparatively longer times. Both of these situations would be captured in the enumeration of takes by Level B harassment, which accounts for

takes that may occur upon exposure at relatively low levels of received sound (e.g., 140 dB).

In addition, Animal Counsel suggests that NMFS should evaluate “the proposed activity's acoustic contribution against the natural, pre-industrial baseline rather than the existing anthropogenic soundscape;” however, no recommendation is provided as to how it believes this analysis should be accomplished absent data relating to the “natural, pre-industrial baseline.” In addition, this recommendation runs afoul of the MMPA requirement for NMFS to evaluate the effects of the specified activity alone.

In addition to the comment that NMFS has underestimated takes that may result from masking, commenters disregard the consideration given to masking effects in NMFS' risk assessment framework (see Negligible Impact Analysis and Determinations). Broadly, the results of the analysis for any given species are based on the integration of two components: the severity of the impacts (which reflects the extent of the activities overlaid with the presence and distribution of the given species) and the vulnerability of that species based on multiple biological and environmental risk factors, including explicit consideration of masking. Expected consequences of masking account for 20 percent of a species' vulnerability rating, and is a substantial proportion of the vulnerability score.

NMFS recognizes that masking is not necessarily co-extensive with harassment and explicitly recognizes this in our discussion of effects, although we also note that the distances at which behavioral harassment is quantified for this rule are farther than those contemplated in the past, due to the behavioral harassment thresholds used (see the Estimated Take section and comment responses later in this section for further discussion of acoustic thresholds). As discussed in detail in NMFS' 2021 final rule, NMFS designed and supported the implementation of a chronic and cumulative effects analysis, which is incorporated in NMFS' discussion of the potential effects of the specified activity here for the specific purpose of addressing the effects of these activities on the listening space of all species and the communication space of Rice's whales specifically (see Negligible Impact Analysis and Determinations). This modeling effort explicitly considered the effects of masking over realistic spatial scales.

Similarly, we addressed the available literature regarding potential impacts of stress resulting from noise exposure in marine mammals. As described in that discussion, stress responses are complicated and may or may not have meaningful impacts on marine mammals. NRDC implies that NMFS must (1) enumerate takes resulting from stress alone and (2) specifically address stress in its negligible impact analysis. The effects of stress are not straightforward, and there is no information available to inform an understanding of whether it is reasonably likely that an animal may experience a stress response upon noise exposure that would not be accounted for in NMFS' existing enumeration of takes via exposure to noise, which includes an accounting for exposures above received levels as low as 140 dB rms (and as low as 120 dB rms for beaked whales). NRDC provides nothing informative regarding how such an analysis might be carried out. With regard to NMFS' negligible impact analysis, we believe that the potential effects of stress are addressed and subsumed within NMFS' considerations of severity of effect and vulnerability of affected populations. Similarly, NRDC provides no justification as to why stress would appropriately be considered separately in this analysis, and no useful recommendation as to how to do so, if appropriate. We believe we have appropriately acknowledged the potential effects of stress, and that these potential effects are accounted for within our overall assessment of potential effects on marine mammals.

Comment 9: NRDC states that NMFS uses an “arbitrary” method to convert area-specific risk scores into a “basis for making Gulf-wide negligible impact determinations.” NRDC takes issue with NMFS' use of the median of zone-specific risk ratings (for those zones including at least 0.05 percent of GOA-wide abundance for a particular species), suggesting that the application of this method inappropriately minimizes findings of “high” to “very high” risk for certain species in Zone 5, where there is a confluence of relatively high levels of survey activity and high proportions of GOA-wide abundance for some species, resulting in high take numbers. NRDC expressed concern that using the median does not allow for appropriate consideration of the importance of specific areas to a particular species, i.e., that this approach “smooths” away granularity of the risk assessment.

Response 9: We disagree with NRDC's comments on this topic, and note that NRDC provided no alternative recommendation. On the contrary, NMFS' approach explicitly incorporates considerations of the importance of a particular area to a species, or the particular localized threats faced by a species, through the zone-specific vulnerability assessment that contributes to the overall risk rating. In addition, NMFS' approach is specifically designed to retain considerations of zone- specific impacts and vulnerability beyond simply the inclusion of the vulnerability scoring. For example, an alternative approach to generating a GOA-wide risk rating would be to employ a wholly different paradigm in which aggregate GOA-wide vulnerability and severity scores are assessed, versus taking a median value of zone-specific ratings. NMFS retained the median value approach precisely because we believe that evaluating risk for such a large and variable area (i.e., the entire U.S. GOA) with species and activities that are each highly localized would provide only a very general and less informative answer regarding risk. The approach employed by NMFS highlights the fundamental importance of the spatiotemporal intersection of animals and activity as the fundamental driver in evaluating risk, while also allowing us to avoid exactly the effect of concern to NRDC (blurring of localized scoring) by avoiding the influence of areas where a particular species essentially does not occur on the overall risk rating for that species.

NRDC is incorrect that use of the median value is inappropriate or that it has “no biological basis.” We note that mean (or average) values can be more heavily skewed by outliers with small sample size than median values. Thus, we chose the median as a better descriptor of central tendency, which is a more appropriate perspective for the risk analysis. (We also rounded up values of 0.5 (e.g., median score of 3.5 would be rounded to a 4), a mathematically valid approach that builds in a reasonable degree of conservatism.)

One of the fundamental values of the analytical framework is that it is structured in a spatially explicit way that can be applied at multiple scales, based on the scope of the action and the information available, to inform an assessment of the risk associated with the activity (or suite of activities). This allows one to generate overall risk ratings while also evaluating risk on finer scales. In this case, severity ratings were generated on the basis of seven different GOA zones, allowing an understanding not only of the relative scenario- specific risk across the entire GOA, as is demanded for this analysis, but also to better understand the particular zones where risk may be

relatively high (depending on actual future survey effort) and what part of the stock's range may be subject to relatively high risk.

NRDC cites the Expert Working Group (EWG) Report in support of its comment, stating it was “[telling]” that “report did not contrive a Gulf-wide risk assessment” and that “doing so would have belied the very different purpose underlying its design: a relative risk assessment across multiple species and geographies.” Although the initial EWG report (Southall et al., 2017) made available for public review of the framework concept did not derive GOA-wide risk ratings, the EWG did so in a later draft report that NMFS adopted in producing the risk evaluation presented in its 2021 final rule.

Comment 10: NRDC asserts that NMFS has failed to appropriately account for stock structure and the importance of certain habitat in its negligible impact analysis, highlighting bottlenose dolphins and sperm whales.

Response 10: Regarding bottlenose dolphins, NRDC criticizes the use of generic risk ratings that homogenize risk for three shelf and coastal bottlenose dolphin stocks. NRDC's comment dramatically inflates this issue through reference to the total 36 stocks of bottlenose dolphin found in the GOA. In fact, 31 of these stocks are comprised of bay, sound, and estuarine populations that are not expected to be affected by the specified activity considered herein. We described and discounted the potential for impacts to these stocks of bottlenose dolphin in the 2018 notice of proposed rulemaking (83 FR 29212, June 22, 2018). As described there, estuarine stocks of bottlenose dolphin primarily inhabit inshore waters of bays, sounds, and estuaries (BSE), and stocks are defined throughout waters adjacent to the specified geographical region. However, estuarine stock ranges are generally described as including coastal waters (i.e., waters adjacent to shore, barrier islands, or presumed outer bay boundaries and outside of typical inshore ranges) to approximately 1-3 km. For example, bottlenose dolphins that were captured in Texas and outfitted with radio transmitters largely remained within the bays, with three individuals tracked to 1 km offshore (Lynn and W[uuml]rsig, 2002). Radio-tracking of dolphins in the St. Joseph Bay, Florida area showed that most dolphins stayed within the bay and that, although some individuals ranged more than 40 km along the coastline from the study site, they never ventured outside of immediate nearshore waters (Balmer et al., 2008). Dolphins captured in Barataria Bay, Louisiana were fitted with satellite-linked transmitters, showing that most dolphins remained within the bay, while those that entered nearshore coastal waters remained within 1.75 km (Wells et al., 2017). Therefore, these stocks would not generally be expected to be impacted by the described geophysical surveys. If a deep penetration seismic survey were occurring in nearshore Federal waters (i.e., at least 3 miles from shore but 9 miles from shore off Texas and Florida), it is possible that a dolphin belonging to a BSE stock could be affected. However, such surveys are expected to be rare in such shallow waters, and given the fact that BSE dolphins in sheltered inshore waters would largely not be impacted by noise generated offshore, we believe that impacts from the specified activities that could potentially be considered as a “take” (as defined by the MMPA) should be considered discountable. Further, the eastern coastal stock of bottlenose dolphin range is entirely outside the specified geographical region, and the segregation of density models into ocean and shelf/coastal domains means that stock-specific risk ratings may be attributed to the oceanic stock of bottlenose dolphin. Therefore, the issue described by NRDC is limited to three stocks of bottlenose dolphin: the continental shelf and western and northern coastal stocks, for which homogenized risk ratings are provided for Zones 2 and 3.

Overall, species-level take and abundance estimates are used to support risk ratings for bottlenose dolphins in Zones 2 and 3 out of necessity. As acknowledged in our negligible impact analysis, the best available information (Garrison et al., 2023) was used to inform combined stock values for these zones and did not support further quantitative apportionment of estimated take or abundances to stocks. However, NRDC's criticism of the development of risk ratings for these stocks of bottlenose dolphins is unwarranted. The risk rating is comprised of the severity score, reflecting the homogenous take information available for the three stocks in Zones 2 and 3, and the vulnerability score, which is designed to incorporate stock-specific information regarding status and threats. The population portion of the vulnerability score comprises three components: status, i.e., is the stock listed under the ESA and/or designated as depleted under the MMPA; trend, i.e., does information over the available time series of abundance estimates indicate a trend; and size, i.e., is the population defined as small (less than 2,500). None of the designated stocks of bottlenose dolphin in Federal waters of the GOA are listed under the ESA or designated as depleted under the MMPA, and none would be classed as small. Regarding trend, for each of the three stocks, two point estimates of abundance are available based on aerial data from surveys during 2011-2012 and 2017-2018 (Garrison et al., 2021). Each of these surveys had a similar design and was conducted using the same aircraft and observer configuration. For each of the three stocks, the 2017-2018 estimate is larger than the 2011-2012 estimate. While a formal trends analysis is not possible based on two point estimates, the information does suggest an increasing trend for these stocks for the purposes of the trend component of the population score. We recognize that the effects of the Deepwater Horizon (DWH) oil spill included likely population reductions for all GOA marine mammal stocks; however, the best available information indicates that these reductions were likely modest for all bottlenose dolphin stocks other than the northern coastal stock, and no more recent population abundance estimates that might reflect any potential reduction are yet available. While the likely decline in population abundance for northern coastal bottlenose dolphins is subsumed within the population score assigned for bottlenose dolphins at the species level for Zones 2 and 3, vulnerability scoring is necessarily performed at the species level such that it may appropriately be integrated with the take-based severity scoring and used to generate an overall zone-specific risk rating. As mentioned above, the best available scientific information does not allow for stock-specific parsing of take for bottlenose dolphins in Zones 2 and 3. Moreover, the trend component of the population score is a relatively small contribution to the overall vulnerability scoring. The likely decline in population abundance for northern coastal bottlenose dolphins, although not reflected in the existing vulnerability scoring, is insignificant as a contribution to the overall vulnerability score for bottlenose dolphins as a species. As noted above, the effects of the DWH oil spill are separately accounted for in the vulnerability scoring and, in summary, the stock- specific status concerns described by NRDC are incorporated to the risk ratings (though we reiterate that NRDC's references to concern relating to the Barataria Bay stock of bottlenose dolphins is in error, as this stock will not be impacted by the specified activities). Importantly, and also not accounted for in the risk assessment

framework, we include significant mitigation (time-area restriction) intended to alleviate impacts to northern coastal bottlenose dolphins during periods of greatest importance for their reproductive behavior.

NRDC contends that through use of homogenous relative risk ratings addressing the combination of bottlenose dolphin stocks residing in Zones 2 and 3, the actual authorization of take of bottlenose dolphins through LOAs could result in greater than a negligible impact through disproportionate impact to the shelf and two coastal stocks of bottlenose dolphin that may be impacted by survey activity occurring in Zones 2 and 3. However, we reiterate that the issues of concern to NRDC are applicable only to the shelf and coastal bottlenose dolphin stocks and Zones 2 and 3, while the majority of projected bottlenose dolphin takes would be expected to accrue to oceanic stock dolphins, reflecting the greater amount of survey effort projected for Zones 4-7. Moreover, as discussed earlier, the risk ratings are only one component of NMFS' negligible impact analysis for each respective stock. Even for these stocks, the implementation of the rule is bound by the description of the specified activity that informs the analysis. That is, the specified activity is comprised of zone- and year-specific effort projections that inform the take estimates and, therefore, the analysis supporting issuance of the rule. Were actual effort to occur in a manner inconsistent with the specified activity, it would not be within the analysis contemplated by the rule. NRDC's comment addresses a hypothetical that cannot occur and still be within the bounds of the rule.

NRDC's additional suggestion, using sperm whales as a case study, that the “averaging of zones disregards the heightened relevance of certain areas to species ecology,” has no merit. Leaving aside that we do not “average zones”--elsewhere in its letter, NRDC criticizes NMFS' use of median values to derive Gulf-wide risk ratings--as explained above, both the Gulf-wide and zone-specific relative risk ratings factor into NMFS' negligible impact analysis, as does consideration independent of the risk ratings of impacts to habitat.

Comment 11: Animal Counsel states that NMFS must provide species- specific impact analyses for each species within the blackfish, beaked whale, and Kogia guilds rather than relying solely on guild-level estimates.

Response 11: NMFS has performed the requisite species- and stock- level analyses. We acknowledge that stock-specific take estimates cannot be derived based on the available scientific information, which is resolved for the referenced groups only at the guild level. As described in the Estimated Take and Negligible Impact Analysis and Determinations sections, sightings of these species that provide the basis for both abundance estimates provided in NMFS' Stock Assessment Reports (SARs) and for the density models comprising the best scientific information available regarding marine mammal occurrence in the GOA typically cannot be resolved to the species level. These species are first rarely sighted due to their pelagic distribution and typical high availability bias due to deep-diving behavior and cryptic nature when at the surface, and are generally difficult to distinguish between species when visually observed in the field. Accordingly, abundance estimates in NMFS SARs are recorded for Mesoplodon spp. (and, separately, for the goose-beaked whale) and Kogia spp. For the broader beaked whale and blackfish guilds, available sightings data, including often unresolved sightings, must be combined in order to develop habitat-based density models, as were used to inform our acoustic exposure modeling effort. Therefore, density and take estimates in this rule are similarly lumped for the referenced guilds, and there is no additional information by which NMFS could appropriately apportion impacts other than equally/proportionally across the constituent species. Animal Counsel appears to misunderstand our description of these issues, suggesting that these issues are only superficial (“grouping based on superficial similarity in looks”) and therefore unjustified compared with grouping that may be based on ecological similarity. In this case, these issues are indistinguishable: as described above, the reliance on undifferentiated sightings for members of these guilds is in fact due to similarities in ecological function as well as to visual similarity. However, of greater importance is that Animal Counsel appears to misunderstand our treatment of the constituent species of these guilds in the negligible impact analysis. We reiterate and confirm here that, while we are unable to precisely apportion estimated take to the species level (reflecting the scientific information available to inform the take estimates), we import species-specific considerations to qualitative aspects of the negligible impact analysis and, in so doing, reach the requisite species- and stock-specific negligible impact determinations.

Comment 12: NRDC asserts that NMFS has erroneously used the relativistic assessment produced through the risk assessment framework as the basis for the negligible impact determination, incorrectly applying it as though it evaluated absolute risk.

Response 12: NMFS disagrees with the comment. The relative risk ratings produced through the framework did not replace our judgments regarding the absolute risk to stocks from the specified activity represented by the negligible impact determinations. The relative risk ratings, both Gulf-wide and zone-specific, are one factor in NMFS' stock specific negligible impact determinations. Despite thorough discussion of all relevant factors for each stock in the Negligible Impact Analysis and Determinations section, NRDC confuses our reference to the straightforward risk rating terms (e.g., low, moderate, high) as being solely determinative of the negligible impact determinations.

The risk assessment analysis is an important component of the negligible impact analysis, but is not the sole basis for our determination. Similarly, implicit in NRDC's complaints regarding NMFS' method of evaluating Gulf-wide risk is the suggestion that NMFS did not also consider zone-specific risk in its negligible impact determinations. These claims are incorrect. We incorporate the Gulf- wide relative risk ratings to our negligible impact analysis, but also provide appropriate focus to the specific zones for each species or guild where relative risk was evaluated to be highest. While the risk assessment analysis comprehensively considered the spatial and temporal overlay of the activities and the marine mammals in the GOA, as well as the number of takes predicted by the described modeling, there are details about the nature of any “take” anticipated to result from these activities that were not considered directly in the risk assessment analysis and which warrant explicit consideration in the negligible impact analysis. Accordingly, NMFS' analysis considers the results of the risk assessment analysis, the effects of the required mitigation, and the nature and context of the takes that are predicted to occur. NMFS' analysis also explicitly considers the effects of predicted Level A harassment, duration of Level B harassment events, and impacts to marine mammal habitat, which respectively were not integrated into or included in the risk ratings. These components of the full analysis, along with any germane species or stock-specific information, are integrated and

summarized for each species or stock in the Species and Stock-specific Negligible Impact Analysis Summaries section of the negligible impact analysis.

In addition, while the risk assessment framework comprehensively considers the aggregate impacts to marine mammal populations from the activities addressed in this rule in the context of both the severity of the impacts and the vulnerability of the affected species, it does not fully consider the absence of survey activity in the eastern GOA. While this is to some degree reflected in the updated take estimates, and thereby incorporated into the risk ratings, the absence of survey activities within areas of greater biological importance for certain species benefits those species GOA-wide beyond what is simply reflected in the updated take numbers. The negligible impact analysis considers the beneficial effects of the absence of survey activity in the eastern GOA regarding both acute and chronic effects.

Also, we note that while the risk assessment framework produces relativistic risk ratings, its components consist of absolute concepts, some of which are also absolutely quantified (e.g., whether the specified activity area contains greater than 30 percent of total region-wide estimated population, between 30 and 15 percent, between 15 and 5 percent, or less than 5 percent). Further, NMFS provided substantive input into the scoring used in implementing the EWG framework for the GOA, to ensure that the categories associated with different scores, the scores themselves, and the weight of the scores within the overall risk rating all reflected meaningful biological, activity, or environmental distinctions that would appropriately inform the negligible impact analysis. Accordingly, and as intended, we used our understanding of the framework and best professional judgment to interpret the relativistic results of the risk assessment analysis appropriately into the larger negligible impact analysis, with the other factors discussed above, to make the necessary findings specific to the effects of the total taking on the affected species and stocks.

Comment 13: NRDC asserts that the rule allows for take in excess of the estimates used in its negligible impact determinations.

Response 13: NRDC's comment suggests a misunderstanding of the acoustic exposure modeling process. The comment assumes that because the modeling involved placement of one survey simulation area in each of the seven original modeling zones, real-world surveys taking place at other locations would take place under conditions (e.g., acoustic propagation, marine mammal density) different than those modeled. That comment confuses the smaller survey boxes with the simulation areas within which sound exposure was modeled, stating that “none of the nominal surveys, save for the survey modeled for zone 4, appear to take place within the proposed critical habitat of Rice's whale” despite the fact that six out of the seven simulation areas overlap the 100-400 m isobaths that mark the proposed designation of Rice's whale critical habitat.

A fundamental component of the modeling approach was the thoroughness with which Gulf-wide conditions were represented in the modeling. Key modeling aspects with geospatial dependence include acoustic propagation conditions and marine mammal density. For the former, propagation conditions were modeled at each of 10 locations representing a full range of static conditions (depths and bottom type) in both the central and western Gulf, with conditions at each of these 10 locations modeled in multiple seasons to understand the full range of dynamic conditions (sound speed profiles). The resulting sound fields were then used in animal movement modeling within the simulation areas to represent the full range of potential conditions that marine mammals may encounter throughout the Gulf and throughout the year. Similarly, average marine mammal density values for each zone were then used to scale the modeled exposure estimates according to real-world density values. While the use of zone-wide average density values does smooth spatial variability in likelihood of encountering particular species of marine mammal, the overall result, both within and across surveys, is a reasonable representation of the likelihood of exposing particular species to sound exceeding harassment criteria.

NMFS acknowledges that a real-world survey occurring in an area with marine mammal occurrence higher than the zonal average may encounter more marine mammals than expected, just as a survey occurring in an area with marine mammal occurrence lower than the zonal average may encounter fewer marine mammals than expected. Overall, during the course of rule implementation, there is no fundamental aspect of the modeling that would result in directionally more marine mammals being impacted than we assume here, regardless of location or season.

Comment 14: NRDC states that NMFS' analysis of the consequences of exposure in support of the negligible impact analysis “contradicts the factual record,” asserting that NMFS wrongly assumed low severity for certain exposure durations and disregarded repeated exposures. Animal Counsel similarly suggests that NMFS ignores the consequences of exposure, referring to the “cumulative impacts of repeat harassment.”

Response 14: While NMFS evaluated exposure durations, which are critical to understanding how the authorized takes are likely to impact individual marine mammals, the negligible impact analysis is not dependent on this or any other single factor, as described fully in the Negligible Impact Analysis and Determinations section. Exposure duration was not addressed in the risk assessment framework but was incorporated into the negligible impact analysis.

As we indicate in the Negligible Impact Analysis and Determinations discussion of this final rule, to put the predicted amount of take into meaningful context, it is useful to understand the duration of exposure at or above a given level of received sound (as well as the likely number of repeated exposures across days). While even a momentary exposure above the criteria for Level B harassment counts as an instance of take, that accounting does not make any distinction between fleeting exposures and encounters in which an animal may be exposed to that received level of sound for a longer period of time. This information is meaningful to an understanding of the likely severity of the exposure, which is relevant to the negligible impact evaluation. For example, for bottlenose dolphin exposed to noise from 3D WAZ surveys in Zone 6, the modeling report shows that approximately 72 takes (Level B harassment) would be expected to occur in a 24-hr period. However, each animat modeled has a record or time history of received levels of sound over the course of the modeled 24-hr period. The 50th percentile of the cumulative distribution function indicates that the time spent exposed to levels of sound above 160 dB rms SPL (i.e., the 50 percent midpoint for Level B harassment) would be only 1.8 minutes--a minimal amount of exposure carrying little potential for significant disruption of behavioral activity.

The Species and Stock-specific Negligible Impact Analysis Summaries discussion considers the relative impact ratings in conjunction with required mitigation and other relevant contextual information-- including exposure durations at the various thresholds--to produce an assessment of impact to the

stock or species, i.e., the negligible impact determinations. For beaked whales, take is estimated on the basis of a risk function shifted down such that 90 percent of the animals exposed to received levels above 140 dB and 50 percent exposed to received levels above 120 dB are predicted to be harassed. We used this approach based on the documented behavioral sensitivity of beaked whales. However, as NRDC acknowledges, context is important when assessing behavioral responses to sound. The exposures above 120 dB here occur at significant distance from the source (i.e., greater than 50 km). It is generally accepted that an animal's distance from the sound source plays an important role in the animal's behavioral response to a received sound level (e.g., Gomez et al., 2016). NMFS believes that exposures to the relevant harassment thresholds at significant modeled distances from the actual sound source, although included in the take estimates based on the risk function, will not carry significant consequences for the potentially exposed animals. Rather, these exposures are likely to result in significantly less severe responses (if any). Examples provided by NRDC purporting to demonstrate greater severity of response than we have assumed include irrelevant examples--beaked whales are known to respond with greater severity to mid-frequency active military sonar than to other sources, as discussed in greater detail in the Potential Effects of the Specified Activities on Marine Mammals and Their Habitat section of the proposed rule--and examples of “responses” entailing changes to vocalization patterns over longer durations, but these responses do not necessarily rise to the level of a take, much less a take event of significant severity.

Regarding repeated exposures, despite the figures cited by NRDC concerning potential days of activity, it is unlikely that any given individual animal would in fact experience repeated take events of the magnitude suggested. Each of the GOA zones is an extremely large area (average zone size approximately 100,000 km\2\), and the likely harassment “footprint” of any given survey would be relatively small. Modeled isopleth distances to the 160-dB threshold are approximately 12 km for low-frequency cetaceans (i.e., the Rice's whale), 7 km for high- frequency cetaceans (i.e., sperm whales, beaked whales, dolphins), and 6 km for very high-frequency cetaceans (i.e., Kogia spp.). Distances to the 140-dB isopleths are substantially larger, but we again emphasize that only 10 percent of the animals exposed at that level would be expected to incur harassment, while 50 percent of the animals exposed at the 160-dB level would be expected to incur harassment. It is clear that, in reality, there is a relatively low chance of any given individual marine mammal being repeatedly taken within relatively short timeframes, much less that such events would result in fitness consequences for those individuals. Additionally, NRDC suggests that NMFS fails to consider repeated takes at all, when in fact this likelihood is inherently addressed through the severity rating of the risk assessment.

Comment 15: NRDC claims that NMFS' negligible impact analysis is inappropriately reliant upon the prescribed mitigation and, further, that the mitigation will be ineffective. NRDC further expressed concern regarding the efficacy of the prescribed visual and acoustic monitoring methods, stating that species could go undetected. NRDC also repeats an erroneous claim from prior letters that acoustic shutdowns for sperm whales are not required under the ITRs.

Response 15: NMFS did not rely on the mitigation in the negligible impact analysis to the degree NRDC implies. As is stated in the analysis, consideration of the implementation of prescribed mitigation is one factor in the analysis but is not determinative in any case. In certain circumstances, mitigation is more important in reaching the negligible impact determination, e.g., when mitigation helps to alleviate the likely significance of taking by avoiding or reducing impacts in important areas.

NRDC misunderstands the degree to which NMFS relies on shutdowns for sensitive or vulnerable species, including beaked whales, at extended distances. We agree that these measures in and of themselves will have limited benefit for cryptic species such as beaked whales that are unlikely to be observed. However, we believe it makes sense to minimize the duration and intensity of exposure for these species when they are observed, and because shutdowns are practicable we include them in the suite of prescribed measures and discuss them where appropriate. For more readily detected species, such as the sperm whale, which is easily detected when at the surface and vocalizes frequently while underwater, the extended distance shutdowns (for both visual and acoustic detections) should appropriately be considered influential in our assessment of impacts to affected individuals and, therefore, ultimately on the stock. In summary, we consider these measures appropriately as mitigating factors when considering context as part of our negligible impact analysis.

While NMFS disagrees with some specific comments regarding efficacy, we generally agree with the overall point that there are limitations on what may reasonably be expected from either visual or acoustic monitoring. While visual and acoustic monitoring effectively complement each other, and acoustic monitoring is the more effective monitoring method (for certain species) during periods of impaired visibility, there is no expectation that these methods will detect all marine mammals present. In general, NRDC appears to misunderstand what NMFS claims with regard to what such monitoring may reasonably be expected to accomplish and/or the extent to which we rely on assumptions regarding the efficacy of monitoring in reaching the necessary findings. We acknowledge these limitations in prescribing these monitoring requirements, while stating why NMFS believes that visual and acoustic monitoring, and the related protocols we have prescribed, are an appropriate part of the suite of mitigation measures here that satisfy the MMPA's least practicable adverse impact standard. However, the negligible impact finding is not conditioned on the presumption of a specific degree of monitoring efficacy.

Regarding NRDC's claim that “NMFS hasn't included an acoustic shutdown requirement for sperm whales in its proposed regulation,” we clarify that the requirement to shut down “upon detection of a sperm whale,” which was described in detail in the notice of proposed rulemaking and is included in this final rule, includes detection by either visual or acoustic means.

Comment 16: NRDC asserts that NMFS “arbitrarily dismissed high number of injuries” for Kogia spp. Animal Counsel also expressed concern regarding Kogia spp., suggesting similarly that the estimated takes of these species were significantly greater than the total population.

Response 16: As basis for its claim that the assessed Level A harassment events are of such magnitude as to call into question NMFS' negligible impact determinations, NRDC states that the annual number of injuries “would amount to 2.4 times the size” of the Kogia spp. population and “exceed potential biological removal by some 6700 percent.” (We note that although irrelevant, NRDC's calculations for both values are incorrect.) However, these values are based on the severely negatively biased abundance estimate of 336 provided in NMFS' SAR. Animal

Counsel similarly compares the estimated takes of these species to the SAR abundance estimate of 336. As discussed in the notice of proposed rulemaking (and below), but omitted by NRDC, NMFS' SARs state that the abundance estimate provided for Kogia spp. is likely a severe underestimate because it was not corrected for the probability of detection on the trackline, and because Kogia spp. are often difficult to see, present little of themselves at the surface, do not fluke when they dive, and have long dive times. In addition, they exhibit avoidance behavior towards ships and changes in behavior towards approaching survey aircraft. As a result, as stated in the notice of proposed rulemaking, we appropriately refer to the model-generated estimated abundance of 1,385 as the most appropriate estimate of abundance available for Kogia spp.

NRDC's argument is based entirely on its premise that if the estimated instances of Level A harassment are so large in comparison with the affected population, it cannot be true that the population- level effects of the assessed Level A harassment events can be mild, i.e., that a majority of the population would experience repeated hearing loss, leading to greater injury. Therefore, the invalid basis for the premise renders the comment irrelevant.

Comment 17: Animal Counsel suggests that NMFS' analysis related to Rice's whale is deficient, stating that NMFS must “conduct a separate, dedicated Rice's whale impact analysis incorporating post-2024 passive acoustic data, the multi-year NOAA trophic ecology study, and a population viability model.” Animal Counsel separately suggests that NMFS' consideration of prey species impacts is deficient.

Response 17: NMFS fully considered the information Animal Counsel references. In our view, these investigations solidify NMFS' previous understanding of the importance of continental slope waters between approximately 100-400 m water depth as Rice's whale habitat. The previously used spatial density model for Rice's whale (Roberts et al., 2016) identified waters of approximately 100-400 m depth on the continental slope throughout the GOA as potential habitat, and the updated density model (which, as discussed previously, incorporates new data on Rice's whale habitat associations) predictions do not markedly differ (Garrison et al., 2023).

NMFS fully considered new information concerning acoustic detections of Rice's whales in areas along the shelf break in the central and western GOA, which demonstrates year-round Rice's whale occurrence in areas outside of the previously identified core habitat. Soldevilla et al. (2022) detected Rice's whale calls at sites in the central GOA south of Louisiana. A subsequent study placed acoustic recorders in shelf break waters in the same central GOA area and added a location in the western GOA offshore of Texas (Soldevilla et al., 2024). This information provides additional evidence of the regular occurrence of Rice's whales outside the northeastern GOA, with Rice's whale calls recorded on 33 and 25 percent of days at the central and western GOA sites, respectively. As in the prior study, calls were recorded throughout the year. Continued study has demonstrated the persistence of Rice's whale presence in the western GOA (Debich et al., 2025a,b).

In summary, available data related to marine mammal presence and habitat, including impacts to prey species, were considered in the negligible impact analysis. Animal Counsel does not provide adequate detail regarding either the manner in which it believes we did not adequately consider the cited factors, or provide sufficiently detailed recommendations regarding the impact analysis it suggests be conducted, including how it should be conducted differently than or incorporated differently to the impact analysis we present in the Negligible Impact Analysis and Determinations section.

Comment 18: Animal Counsel raises concerns regarding NMFS' negligible impact analysis through comparison of take estimates to NMFS' Potential Biological Removal values.

Response 18: PBR is defined in the MMPA (16 U.S.C. 1362(20)) as “the maximum number of animals, not including natural mortalities, that may be removed from a marine mammal stock while allowing that stock to reach or maintain its optimum sustainable population” and is a measure to be considered when evaluating the effects of mortality or serious injury on a marine mammal species or stock. Harassment is not equivalent to serious injury and does not “remove” an individual from a stock. Therefore, the PBR metric is not relevant to an evaluation of the effects of harassment on a stock in the manner suggested by the commenter.

Comment 19: Animal Counsel suggests that NMFS should discount consideration of practicability in determining the appropriate suite of mitigation requirements.

Response 19: Animal Counsel states that practicability should not hold equal weight with consideration of the effectiveness of a measure when evaluating measures towards achieving the MMPA standard of “least practicable adverse impact.” Because mitigation requirements are inherently burdensome, Animal Counsel suggests that consideration of this factor is in conflict with the MMPA. However, this reading is in contradiction with the statutory requirement to prescribe the “means of effecting the least practicable adverse impact.” Consideration of practicability is ingrained in the very standard that Animal Counsel urges NMFS to meet, and disregarding practicability would contradict the requirements of the MMPA.

Animal Counsel also suggests that consideration of practicability is at odds with the need to based decisions on the best scientific information available. However, relevant data are not limited to consideration of mitigation effectiveness, but also contribute to an understanding of the operational impacts of a measure, i.e., practicability. In either case (consideration of mitigation effectiveness and practicability), NMFS must consider the available data and make judgments regarding these two primary factors.

In the evaluation of specific measures, the details of the specified activity will necessarily inform each of the two primary factors discussed above (expected reduction of impacts and practicability) and are carefully considered to determine the types of mitigation that are appropriate under the least practicable adverse impact standard. Analysis of how a potential mitigation measure may reduce adverse impacts on a marine mammal stock or species and practicability of implementation are not issues that can be meaningfully evaluated through a yes/no lens. The manner in which, and the degree to which, implementation of a measure is expected to reduce impacts, as well as its practicability, can vary widely. For example, a time-area restriction could be of very high value for reducing the potential for, or severity of, population-level impacts (e.g., avoiding disturbance of feeding females in an area of established biological importance) or it could be of lower value (e.g., decreased disturbance in an area of high productivity but of less firmly established biological importance). Regarding practicability, a measure might involve restrictions in an area or time that impede the operator's ability to acquire necessary data (higher impact), or it could mean incremental delays that increase operational costs but still allow the activity to be conducted (lower impact). A

responsible evaluation of “least practicable adverse impact” will consider the factors along these realistic scales. Expected effects of the activity and of the mitigation as well as status of the stock all weigh into these considerations. Accordingly, the greater the likelihood that a measure will contribute to reducing the probability or severity of adverse impacts to the species or stock or their habitat, the greater the weight that measure is given when considered in combination with practicability to determine the appropriateness of the mitigation measure, and vice versa.

We acknowledge that in some cases, certain mitigation may be necessary in order to make a “negligible impact” finding for an affected species or stock, which is a fundamental requirement of issuing an authorization--in these cases, consideration of practicability may be a lower priority for decision-making if impacts to marine mammal species or stocks would not be negligible in the measure's absence. However, this circumstance is not present in the analysis presented herein, and practicability must be given due consideration.

Comment 20: NRDC comments that, overall, NMFS has not prescribed mitigation measures sufficient to meet the MMPA's LPAI standard, and specifically that NMFS must prescribe an area closure for Rice's whale in order to meet this standard. Animal Counsel expresses similar concerns regarding NMFS' assessment of the available science regarding Rice's whale habitat and its LPAI analysis.

Response 20: NRDC provides no actionable recommendation beyond the assertion that NMFS must “provide an area closure” for Rice's whale. As discussed at length in the Mitigation section of this final rule, we find that, while there is now robust evidence that Rice's whales utilize areas of the Gulf within roughly the 100-400 m depth isobaths outside of their northeastern GOA core habitat area, key questions remain about Rice's whale abundance, density, habitat use, demography, and stock structure in the central and western GOA. As was acknowledged by NMFS in its proposed rule to designate critical habitat for Rice's whale, aside from visual and acoustic detections indicating presence, this area hosts features that comprise characteristics of habitat that is essential to the species. However, it does not follow that the area must be closed to activity of a certain sector, in this case geophysical survey activity. And as described in the notice of proposed rulemaking and later in this rule, there is no available scientific information that would enable NMFS to identify any particular portion of this broad swath of suitable habitat throughout the GOA that is of particular importance, or to identify any time of year when this habitat may be of particular importance. Absent this information, NMFS is left to contemplate a complete, year-round closure to geophysical survey effort of the 100-400 m isobaths across the entire central to western GOA, based solely on data establishing that the habitat contains features making it suitable for Rice's whale use, and scant information establishing that Rice's whales are present in unknown numbers in this area. NRDC does not dispute that the available information does not enable identification of any area or time that could be subject to meaningful consideration as a closure, and it does not attempt to identify and recommend any such area. Instead, NRDC implies that the entire area should be closed by asserting that the MMPA “places no ceiling on the amount of habitat that may be subject to mitigation” and that the entire area should be closed as a result of its assertion that the MMPA demands “overprotection” rather than “underprotection,” without acknowledging that the measures included in the rule may be the right amount of protection.

NRDC's argument in favor of some sort of area closure rests on a misunderstanding of the acoustic exposure modeling supporting the rule's take estimates. NRDC wrongly asserts that NMFS' modeling “placed its nominal surveys entirely or almost entirely outside the [Rice's] whale's habitat,” leading to the misguided assertion that “the rule itself is insufficient to ensure that actual take remains within authorized numbers.” As we address in greater detail above, the exposure modeling is representative of all habitat, including Rice's whale habitat, that may be traversed by projected survey effort, and the assertion that much greater impacts than have been modeled would result is unsupported.

As discussed in detail in the notice of proposed rulemaking, the rate of call detections throughout the year is considerably higher in the eastern GOA than at the central GOA sites where calls were most commonly detected, with at least 8.3 calls/hour among four eastern GOA sites over 110 deployment days (Rice et al., 2014) compared to 0.3 calls/hour over the 299-day deployment at the central GOA site where calls were detected most frequently during the Soldevilla et al. (2022) study. During that study, approximately 2,000 total calls were detected at the central site over 10 months, compared to more than 66,000 total detections at the eastern GOA deployment site over 11 months (approximately 30 times more calls detected at the eastern GOA site) (Soldevilla et al., 2022). Similarly, Soldevilla et al. (2024) reported detecting 0.2 calls/hour at the western GOA site off Texas (1,694 detections over 8,547 hours of recording). While continued monitoring off Texas demonstrates ongoing, regular presence of whales (Debich et al., 2025a,b), available information continues to suggest that density and abundance of Rice's whales is likely lower in the central and western GOA than in the species' core habitat area in the eastern GOA.

Regarding the suggestion that NMFS has not adequately considered habitat in its consideration of mitigation, we disagree. Habitat value is generally informed by marine mammal presence and use, and the available data can support the consideration and discussion of impacts to (and mitigation for) both marine mammals and their habitat simultaneously. The discussion above clearly considers physical features that can drive habitat use (e.g., depth), as well as detailed information related to relative presence in the eastern versus the central and western GOA, which is indicative of preferred habitat in the east. Because habitat value is generally informed by marine mammal presence and use, in some cases, there may be overlap in measures for the species or stock and for use of habitat. NRDC has not presented any information that would suggest habitat we did not consider for mitigation.

In summary, the newly available data related to marine mammal presence and habitat were considered under the LPAI standard, and we concluded additional mitigation for Rice's whale was not warranted under that standard. NRDC provides no persuasive argument to the contrary.

Comment 21: NRDC finds fault with NMFS' consideration of practicability concerning possible closure of potential Rice's whale habitat in the central and western GOA to future survey activity, suggesting that NMFS' reference to analysis presented in its Regulatory Impact Analysis (RIA) for the 2021 rule is not relevant. Animal Counsel similarly recommends that NMFS must implement restrictions on survey activity in Rice's whale habitat (waters 100-400 m depth throughout the Gulf). NRDC also suggests that NMFS must consider that the Outer Continental Shelf Lands Act (OCSLA) “requires a balancing between the development of

offshore energy resources and the protection of marine resources.”

Response 21: As was acknowledged in the proposed rule, the RIA did not directly evaluate a potential closure of potentially suitable habitat in the central and western GOA outside of the Rice's whale core distribution area. However, we disagree that the RIA is not relevant to our practicability analysis here. The RIA's assessment of potential restrictions in the northeastern GOA provided a useful framework for considering practicability relating to a broad closure of potential Rice's whale habitat to future survey activity.

To bolster that discussion, we turned to the same sources of data referenced in the RIA in analysis of potential closure areas considered therein (see https://www.data.boem.gov/Main/Default.aspx). While areas of Rice's whale habitat (i.e., water depths of 100-400 m on the continental shelf break) contain less oil and gas industry infrastructure than do shallower, more mature waters, and have been subject to less leasing activity than deeper waters with greater expected potential reserves, they nonetheless host significant industry activity. BOEM provides summary information by water depth bin, including water depths of 201-400 m. Omitting information regarding water depths of 100-200 m, the area overlaps 33 active leases, with 17 active platforms and over 1,200 approved applications to drill. In the past 20 years, over 500 wells have been drilled in water depths of 100- 400 m. These data confirm that there is substantial oil and gas industry activity in this area and, therefore, the inability to collect new seismic data could affect oil and gas development given that oil companies typically use targeted seismic to refine their geologic analysis before drilling a well. In addition, year-round occurrence of Rice's whales in waters 100-400 m deep precludes the use of seasonal closures to minimize exposure of Rice's whales. Therefore, we analyzed the potential for a year-round closure, which exacerbates the potential for effects on oil and gas productivity in the GOA because operators have no ability to plan around the closure. While the area is not as important to regional oil and gas productivity as the prospective deepwater central GOA closure analyzed in the RIA (as we acknowledged in the proposed rule), the more area-specific data provided above continue to support NMFS' previous conclusions, which we affirm here: (1) We are unable to delineate specific areas of Rice's whale habitat in the central and western GOA where restrictions on survey activity would be appropriate because there is currently uncertainty about Rice's whale density, abundance, habitat usage patterns and other factors in the central and western GOA; and (2) there is high likelihood that closures or other restrictions on survey activity in all waters of 100-400 m depth in the central and western GOA would have significant economic impacts. Finally, we note that despite NRDC's concerns, it does not recommend any particular closure that it believes NMFS should evaluate.

Regarding NRDC's suggestions concerning OCSLA--a statute administered by BOEM--NMFS' statutory obligations arise under the MMPA (with associated requirements under the ESA, NEPA, and Administrative Procedure Act (APA), among others). NMFS has no statutory obligation relative to OCSLA. Similarly, NMFS' obligations under the MMPA require that we prescribe the means of effecting the LPAI on the affected species or stock and their habitat, which we have done here.

Comment 22: NRDC states that NMFS “fails to consider mitigation measures” for Rice's whale, suggesting that NMFS consider: (1) allowing some survey activities in Rice's whale habitat (i.e., the 100- 400 m isobaths), such as surveys undertaken by leaseholders to develop their lease blocks, while prohibiting others; (2) extending geographically vessel strike avoidance measures “presently in effect for industry”; and (3) requiring use of “lowest practicable source levels within the whales' communication frequencies for activities taking place in the vicinity of the whales' habitat.” Animal Counsel also states that NMFS must impose a mandatory 10-knot vessel speed restriction for all project vessels within Rice's whale habitat.

Response 22: NRDC does not provide supporting detail regarding its recommended mitigation requirements. As such, NMFS is unable to fully evaluate the suggested measures.

Regarding the suggestion to allow some surveys but prohibit others, section 101(a)(5)(A) of the MMPA requires NMFS to make a determination that the take incidental to a “specified activity” will have a negligible impact on the affected species or stocks of marine mammals, and will not result in an unmitigable adverse impact on the availability of marine mammals for taking for subsistence uses. NMFS' implementing regulations require applicants to include in their request a detailed description of the specified activity or class of activities that can be expected to result in incidental taking of marine mammals. 50 CFR 216.104(a)(1). Thus, the “specified activity” for which incidental take coverage is being sought under section 101(a)(5)(A) is generally defined and described by the applicant. Here, BOEM, having joined the NMFS Policy request as co-petitioner, is an applicant for the ITRs in support of industry operators, and we are responding to the specified activity as requested, i.e., carrying forward the specified activity underlying the 2021 rule (and making the necessary findings on that basis). BOEM's initial petition made no distinction between surveys that may be speculative or otherwise fall into a category of surveys that NRDC suggests should be prohibited, and those that are not.

Moreover, NRDC does not describe any useful metric for determining which surveys should be allowed, aside from reference to “surveys undertaken by leaseholders to develop their lease blocks.” The suggestions are not sufficiently developed to allow for adequate consideration.

Regarding vessel strike avoidance measures, NRDC does not specify what measures it is referring to. However, the ITRs already contains a suite of vessel strike avoidance measures that apply wherever survey activity is occurring and, regarding Animal Counsel's recommendation, these measures include practicable measures expected to avoid the risk of vessel strike, including separation distance requirements from whales and recommendations to slow speeds in the presence of whales. Animal Counsel suggests that the rule is internally contradictory in its “reliance on 5-knot towing speeds;” however, we are not reliant on this fact in determining that strike is unlikely. We acknowledge that support vessels and other vessels not towing gear may travel at speeds in excess of 5 knots. However, it remains that a contributing factor to our assessment of vessel strike risk is the fact that survey vessels are typically slow-moving.

Finally, NRDC does not describe any useful scheme by which “lowest practicable source levels within the whales' communication frequencies” might be defined. An expert panel, convened by BOEM to determine whether it would be feasible to develop standards to determine a lowest practicable source level, determined that it would not be reasonable or practicable to develop such metrics (see Appendix L in BOEM, 2017). The subject matter is outside NMFS' expertise, and we have no basis upon which to doubt the panel's published findings.

Comment 23: Animal Counsel presents certain mitigation recommendations, including that NMFS should (1) require dedicated, independent third-party marine mammal observers for all survey operations, not only those where they are currently mandated; (2) prohibit survey operations from continuing during any passive acoustic monitoring (PAM) malfunction beyond the minimum time necessary for troubleshooting; and, (3) increase exclusion and buffer zones to at least 500 meters for all survey types and eliminate the dolphin exception for shutdown requirements.

Response 23: NMFS has thoroughly evaluated all of Animal Counsel's recommendations in this and prior related rulemakings, and has previously addressed all of these issues. With regard to the suggestion that third-party marine mammal observers should be required for all survey operations, this is a requirement of the rule.

Regarding the allowance that survey operations may continue during brief periods of PAM malfunction, we retain this feature as reasonable consideration of both overall impacts of the suite of mitigation requirements as well as practicability. It would serve little purpose to require cessation of operations during such brief periods, as the overall survey duration would be significantly extended due to the frequent brief disruptions. Each cessation of survey operations requires extended time to reposition the source vessels and engage in necessary pre-clearance and ramp-up periods, and the amount of sound energy in the water overall would be increased. Animal Counsel's recommendation is short-sighted. In addition, the comment references at length other issues associated with broader malfunction than is at issue here, including hypothetical safety issues associated with continued operation during malfunction of acoustic source equipment. This provision of mitigation is solely with respect to malfunction of the PAM listening devices.

Regarding its recommendation to increase exclusion and buffer zones, Animal Counsel offers no supporting rationale other than the unsupported premise that larger is better, claiming that NMFS' required zones are arbitrary “by NMFS' own admission.” Nowhere in the record has NMFS stated or suggested that its required zones are arbitrary. NMFS provided a detailed rationale for both the size of the specified zones and for the associated dolphin exemption.

Comment 24: NRDC states that NMFS “fails to reconsider prescribing quieter alternatives to conventional seismic airguns, despite evidence of the availability of such alternatives,” and claims that NMFS has not adequately analyzed the practicability of such a requirement. Animal Counsel states that NMFS must require similar measures.

Response 24: NMFS acknowledges that there are an increasing number of sources that may reasonably be considered as environmentally preferable to conventional airguns, including sources operating at lower frequencies and without the high peak pressure output associated with airguns. In fact, such sources have been used during certain surveys conducted under NMFS-issued LOAs. However, imposing requirements to use certain technologies, or prescribing the manner in which geophysical survey data must be acquired, would exceed NMFS' authority under the MMPA. Survey funders and operators define survey objectives and methodologies, including which acoustic sources are used, on the basis of data needs that are beyond NMFS' technical expertise to judge. NRDC argues that specific mandates are not required, versus a generic “best available technology” requirement, but offers no recommended metrics. NMFS agrees that increased use of environmentally preferable sources is an appropriate goal, but it would be more appropriate to continue working with industry to incentivize use of such sources and techniques rather than require them.

Comment 25: NRDC states that NMFS must consider “measures to eliminate duplicative surveys.”

Response 25: BOEM has historically been the subject matter expert regarding whether and which surveys are “duplicative.” NRDC fails to explain how this is within NMFS' statutory authority or suggest ways to appropriately apportion the amount of effort that it believes should be allowed. NMFS cannot arbitrarily limit planned effort and has no legitimate means of changing the specified activity absent a conclusion that the activity would have more than a negligible impact. However, NMFS has made the necessary findings under the MMPA for issuance of this rule. Similarly, NRDC states that NMFS should “require BOEM to eliminate unnecessary duplication of survey effort” but does not explain how they believe that this suggestion is within NMFS' statutory authority. As the permitting agency, BOEM has the authority to require permit applicants to submit statements indicating that existing data are not available to meet the data needs identified for the applicant's survey (i.e., non-duplicative survey statement), but such requirements are not within NMFS' purview. NMFS may not demand that BOEM discharge its authority under OCSLA in any particular manner. As stated previously, NMFS considers the specified activity described by an applicant in reviewing a request for an incidental take authorization. Nothing in the statute provides authority to direct consolidation or removal of activities based on some presumption of duplication that NMFS is not qualified to judge.

NMFS also notes that, although surveys may be perceived as “duplicative” simply because other surveys have also occurred in the same location, they are in fact designed specifically to produce proprietary data that satisfies the needs of survey funders. As noted by NRDC, BOEM convened an expert panel to study the issue of duplicative surveys (see Appendix L in BOEM, 2017) and developed standards for consideration of what surveys are duplicative. NRDC provides extensive discussion of their thoughts regarding the insufficiency of BOEM's duplicative survey standard and its implementation. These comments are more appropriately directed to BOEM.

Comment 26: Animal Counsel suggests that NMFS must incorporate enforceable adaptive management triggers into the regulatory text, including automatic review upon finalization of Rice's whale critical habitat, completion of a recovery plan, or updated stock assessments.

Response 26: As acknowledged by Animal Counsel, the ITRs include adaptive management provisions that would allow NMFS to modify the existing mitigation, monitoring, or reporting measures if doing so is practicable and creates a reasonable likelihood of more effectively accomplishing the goals of the mitigation and monitoring set forth in this final rule. Separate from the adaptive management provisions, the ITRs allow that NMFS may modify the requirements without regard for practicability if it determines that an emergency exists that poses a significant risk to the well-being of the affected species or stocks of marine mammals. Animal Counsel does not clarify how it believes its recommendation differs from these provisions, which were included in the proposed rule and in this final rule.

Comment 27: Animal Counsel suggests that NMFS cannot satisfy the MMPA's small numbers requirement, as

the estimated take numbers exceed population estimates.

Response 27: We refer the commenter to the detailed description of NMFS' implementation of the small numbers requirement later in this rule, entailing the appropriate evaluation of small numbers at the LOA stage (see Small Numbers). However, we also note Animal Counsel's reference to inappropriate population abundance estimates in support of its comment. NMFS provides detailed discussion, in this rule and in the notice of proposed rulemaking, regarding its approach to evaluation of the most appropriate abundance estimate for purposes of evaluating “small numbers.” As an example, Animal Counsel cites to NMFS' SARs estimate for the Clymene dolphin of 513 (CV = 1.03). As explained later, the most appropriate abundance estimate for this species is the modeled abundance value of over 6,000. NMFS' series of SAR abundance estimates for this stock have fluctuated between 129 and 17,355 animals, i.e., varying by a maximum factor of more than 100. For most species, such fluctuations across these “snapshot” abundance estimates (i.e., that are based on only the most recent year of survey data) reflect interannual variations in dynamic oceanographic characteristics that influence whether animals will be seen when surveying in predetermined locations, rather than any true increase or decline in population abundance. In fact, NMFS' SARs typically caution that trends should not be inferred from multiple such estimates, that differences in temporal abundance estimates are difficult to interpret without an understanding of range-wide stock abundance, and that temporal shifts in abundance or distribution cannot be effectively detected by surveys that only cover portions of a stock's range (i.e., U.S. waters). The corresponding density model for Clymene dolphins predicts a mean abundance of over 6,000 dolphins. Therefore, in this example, Animal Counsel would have us compare takes predicted by a model in which 6,000 dolphins are assumed to exist against an abundance estimate of 513 dolphins. Our goal in assessing predicted takes is to generate a meaningful comparison, which is accomplished through use of the model-predicted abundance.

Comment 28: NRDC asserts that NMFS' interpretation of the MMPA's small numbers requirement is contrary to law, stating its belief that NMFS must make a small numbers determination in the rule, rather than for issuance of individual LOAs; that NMFS must evaluate the same amount of take in order to separately determine that the total take will both meet the small numbers standard and have a negligible impact; that NMFS must analyze whether the estimated amount of takes over the 5-year ITRs will exceed small numbers for each affected species; and that NMFS' approach impermissibly cuts the public out of the agency's findings.

Response 28: Based on NMFS' analysis of the language and structure of section 101(a)(5)(A) and the implementing regulations for that provision, NMFS disagrees that the small numbers finding must be based on the total of all take over the five-year (or less) period from all potential survey activity. The MMPA does not define small numbers or explain how to apply the term in either section 101(a)(5)(A) or the similar provision for incidental harassment authorizations (IHAs) in section 101(a)(5)(D),\4\ including how to apply the term in a way that allows for consistency across those two provisions that are similar but allow for potentially different time and activity scales. (See Small Numbers below.) Especially when taken together with NMFS' implementing regulations, our approach is consistent with the structure of section 101(a)(5)(A), which provides:

\4\ Section 101(a)(5)(D) states in relevant part:

(i) Upon request therefor by citizens of the United States who engage in a specified activity (other than commercial fishing) within a specific geographic region, the Secretary shall authorize, for periods of not more than 1 year, subject to such conditions as the Secretary may specify, the incidental, but not intentional, taking by harassment of small numbers of marine mammals of a species or population stock by such citizens while engaging in that activity within that region if the Secretary finds that such harassment during each period concerned--

(I) will have a negligible impact on such species or stock, and

(II) will not have an unmitigable adverse impact on the availability of such species or stock for taking for subsistence uses . . . .

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

(I) finds that the total of such taking during each five-year (or less) period concerned will have a negligible impact on such species or stock and will not have an unmitigable adverse impact on the availability of such species or stock for taking for subsistence uses . . . .

(Emphasis added.)

Section 101(a)(5)(A)(i)(I) is explicit that the “negligible impact” determination for a specified activity must take into account the “total of such taking” (i.e., all of the taking that the Secretary may conceivably allow (or authorize) under individual LOAs during the five year (or less) period considered for the rule). In contrast, the “small numbers” language in 101(a)(5)(A) is not subject to the same time period requirement of five years (or less in cases where the period being considered for a rule is less than five years).

In our view, the statutory language for small numbers and the negligible impact finding indicates that the negligible impact finding is made based on consideration of an aggregation of potential authorizations (LOAs) for taking small numbers of marine mammals, and allows for different temporal periods in applying the two different standards. The statute contemplates that the Secretary shall allow taking during the five year (or less) period, which, in our view, also implies that there could be multiple allowances or authorizations (i.e., LOAs), so long as the maximum total taking from all of those authorizations combined is considered in the upfront assessment of whether the negligible impact standard is met for the total taking allowable under the regulations.

As we note in our Small Numbers section, the regulatory vehicle for authorizing (i.e., allowing) the take of marine mammals is the LOA, a creature of NMFS' long-standing implementing regulations that is not in the statute. See 50 CFR 216.106. Those 1989 implementing regulations requiring an LOA to effectuate an authorization were in effect when Congress amended the MMPA in 1994 to add section 101(a)(5)(D) for issuance of one-year IHAs, and over the years when Congress amended section 101(a)(5)(A) for various reasons (including most recently in 2018, to extend the maximum authorization period to seven years for military readiness activities, Public Law 115-232 (John S. McCain National Defense Authorization Act for Fiscal Year 2019) (Aug. 13, 2018)).

Under NMFS' approach, the negligible impact analysis for the rulemaking is conducted for the time period covered by the rule (five years in this case, the maximum under the statute for a non-military readiness activity), but the small numbers analysis attaches to the instrument that actually “allows” or authorizes taking, i.e., the LOA. The statute does not preclude

NMFS from issuing an LOA that comports with the small numbers level set forth in the relevant rule for the specified activity. Consistent with the MMPA requirement, here the Secretary (through NMFS) has prescribed the necessary specified activity regulations after notice and comment. At that point, once the regulations are effective, NMFS thereafter may authorize incidental take through the issuance of LOAs, provided that they satisfy the requirements set forth in the rule and regulations, including the small numbers standard articulated in the rule.

NRDC cites Conservation Council for Hawaii v. NMFS, 97 F. Supp. 3d 1210 (D. Hawaii 2015), in stating that the MMPA “plainly requires that the agency evaluate both whether there will be small numbers of take and whether there will be a negligible impact” before issuing regulations, and that these determinations “must be based on the same amount of take.” We disagree with the second proposition. In NMFS' view, Conservation Council for Hawaii stands for the proposition that NMFS cannot authorize more take than it has analyzed under the negligible impact standard. 97 F. Supp. 3d at 1221. There the court found that there were substantial differences between the anticipated take numbers, which were the basis for the negligible impact finding, and the amount of take that NMFS was prepared to authorize incidental to U.S. Navy military readiness activities. That case did not involve the small numbers provision, which does not apply in the case of military readiness activities. 16 U.S.C. 1371(a)(5)(F)(i). The court in Conservation Council for Hawaii did not consider or make any pronouncements about whether the small numbers provision must be applied to the total annual taking under the rule or whether it could be applied at the LOA stage.

We disagree with NRDC's view that the word “such” in the phrase “total of such taking” for determining negligible impact means that the analysis for small numbers must also consider the total taking allowable under the regulations. The phrase “total of such taking” in clause (I) clearly refers to what must be considered in making a negligible impact determination, and does not qualify the small numbers requirement. In addition, the statute's reference to “citizens” in the plural contemplates that there could be more than one entity making a request for take under a regulation for a specified activity. Under the statute, the Secretary (through NMFS) shall allow the take of small numbers of marine mammals by citizens--which, as in this case, could entail more than one requester--provided that NMFS can make the negligible impact finding for the total of all the taking that may be authorized under the regulations.

NRDC repeatedly states that the negligible impact and small numbers provisions must have separate meaning. NMFS agrees that the two provisions have separate meanings, and this rule satisfies that requirement. Each LOA must meet the small numbers requirement as NMFS has interpreted it in this rule. In other words, it is not sufficient for the survey activity described in an LOA application to fall within the scope of the activity analyzed for the rule and NMFS' negligible impact determination. The small numbers limitation set forth in this rule also must be satisfied. For example, NMFS may receive an application for an LOA where the take estimates exceed the small numbers standard identified in the rule. In that case, the request would be denied, even if the amount of taking was considered in the negligible impact evaluation. Thus the negligible impact and small numbers inquiries are separate and have different meanings.

To summarize, the MMPA is silent on how to apply “small numbers” in either section 101(a)(5)(A) or (D), including in a way that allows for consistency across those two very similar provisions. Moreover, NMFS' implementing regulations for section 101(a)(5)(A) make it clear that LOAs are the instrument for authorizing (or “allowing”) take. Thus, the mere existence of regulations issued under section 101(a)(5)(A) for a specified activity is not sufficient to authorize take under that provision. An LOA is required.

As we have previously stated, the small numbers standard has limited biological relevance (i.e., there is a lack of a biological underpinning for the concept), but NMFS' application of the small numbers standard at the LOA stage does not rely on that view for the approach taken here. NMFS' interpretation and approach are based on analysis of the governing section 101(a)(5)(A) and limited legislative history, and our long-standing approach to implementing section 101(a)(5)(A) through separate LOAs, as well as consideration of section 101(a)(5)(D). In the absence of further congressional guidance on the meaning of the term “small numbers” and how it must be applied, we have determined our interpretation and application of small numbers is the best reading of the statute, consistent with the U.S. Supreme Court's opinion in Loper Bright Enters. v. Raimondo, 603 U.S. 369.

Importantly, the final rule, which was subject to notice and comment, sets the small numbers standard for future LOAs issued under the rule. Moreover, contrary to NRDC's assertions, NMFS has set the total taking allowable for all LOAs issued under the rule for this specified activity--i.e., the taking that was analyzed for the negligible impact determination. If an LOA application for a survey provides take estimates that are within the small numbers threshold set in this rule, then the LOA for that survey will be deemed to satisfy the small numbers requirement.

As NRDC correctly points out, NMFS' implementing regulations require issuance of LOAs to be consistent with the “total taking allowable” under the activity-specific regulations. The regulations for the specified activity also reflect this. The rulemaking for these regulations evaluated the level of activity projected in BOEM's update for its original petition as well as the updated take estimates for the 2024 rule and other available information, and NMFS' negligible impact determination is based on consideration of that level (and the corresponding take estimates). Any LOA must be within the amount analyzed for the scope of the rule, and the total amount of take under all issued LOAs combined cannot exceed the amount analyzed and “allowable” under the rule for this activity.

NRDC also states that “NMFS must analyze whether the estimated amount of takes over the 5-year proposed ITRs will rise above small numbers for each affected marine mammal species,” referencing NMFS' use of species-specific scalar ratios (see Estimated Take) and recommending that NMFS “should use the scaled numbers it derived to engage in a small numbers analysis.” Although NRDC's recommendation is unclear, we disagree with NRDC's position regarding the need to conduct a small numbers analysis for the total allowable taking over the 5-year ITRs, as discussed elsewhere in this response. However, we do use the “scaled numbers” for purposes of the LOA-by-LOA small numbers analyses, as described in Small Numbers. In addition, NRDC recommends that NMFS should invite comment on the “details of the mathematical modelling used to produce these ratios.” However, this aspect of the rulemaking has been described in detail and available for public review and comment in NMFS' 2018 (83 FR 29212, June 22, 2018) and 2026 (91 FR 9014, February 24, 2026) rulemakings.

Finally, NRDC's statement that the public is impermissibly cut out of the

agency's findings is incorrect. Both the proposed and this final rule set forth the maximum total taking and annual taking that would be allowable (via the issuance of LOAs) for the five-year period that the regulations will be effective. The proposed rule included a 30-day public comment period. We also believe that our rulemaking afforded a full and focused opportunity for public review of and comment on the full scope of survey activities and proposed mitigation, rather than through dozens of individual IHAs, each with separate public comment periods. Thus the public had a meaningful opportunity to comment.

Comment 29: NRDC states that the interpretation of “small numbers” presented by NMFS in the notice of proposed rulemaking is contrary to the plain meaning and purpose of the MMPA, in part because NMFS allegedly did not provide a reasoned basis for the take limit proposed (i.e., one-third of the best available species or stock abundance estimate). NRDC makes three specific claims. First, NRDC states that one-third cannot be considered a “small number” within the plain meaning of the word. Second, NRDC states that Congress intended that takes be limited to “infrequent, unavoidable, or accidental” occurrences, but that NMFS has not explained why the taking of up to one-third would be infrequent or unavoidable. Third, NRDC contends that NMFS should define different small numbers thresholds on the basis of the conservation status of individual species.

Response 29: NMFS disagrees with these arguments. Although there is limited legislative history available to guide NMFS and an apparent lack of biological underpinning to the concept, we have worked to develop a reasoned approach to small numbers. In the discussion of Small Numbers in this and our 2021 rule for this specified activity, NMFS explains the concept of “small numbers” in recognition that there could also be quantities of individuals taken that would correspond with “medium” and “large” numbers. As such, NMFS has established that one-third of the most appropriate population abundance number--as compared with the assumed number of individuals taken--is an appropriate limit with regard to “small numbers.” This relative approach is consistent with the statement from the legislative history that “[small numbers] is not capable of being expressed in absolute numerical limits” (H.R. Rep. No. 97-228, at 19 (September 16, 1981)), and relevant case law (Center for Biological Diversity v. Salazar, 695 F.3d 893, 907 (9th Cir. 2012) (holding that the U.S. Fish and Wildlife Service reasonably interpreted “small numbers” by analyzing take in relative or proportional terms)).

NRDC claims that a number may be considered small only if it is “little or close to zero” or “limited in degree.” This selectively picks a definition in support of the commenter's favored position. But the definition of “small” in Webster's New Collegiate Dictionary (1981) included “having little size, esp. as compared with other similar things.” See also www.merriam-webster.com/dictionary/small (defining “small” as “having comparatively little size”). These definitions comport with the small numbers method developed by NMFS, which utilizes a proportionality approach. NRDC's comment also ignores the grammar in the relevant legislative history language when stating that NMFS has not explained why the taking of up to one-third would be “infrequent or unavoidable.” The actual statement from the legislative history is that taking of marine mammals should be “infrequent, unavoidable, or accidental.” H.R. Rep. No. 97-228, at 19 (September 16, 1981) (emphasis added). Like the term “small,” infrequent is a relative term that has multiple meanings. In addition, this disjunctive language in the series suggests that taking that is unavoidable or accidental may qualify as small numbers, even if not infrequent.

Finally, the argument to establish a small numbers threshold on the basis of stock-specific conservation status is unnecessarily duplicative of the required negligible impact finding, in which relevant biological and contextual factors are considered in conjunction with the amount of take.

Comment 30: NRDC states that NMFS' severability clause regarding small numbers is inoperable.

Response 30: Our small numbers interpretation and application contains several aspects. In the event a court were to invalidate some but not all aspects of NMFS' small numbers interpretation/application, NMFS intends that the remaining aspects of the rule and ITRs be severable to the extent possible. The extent to which this is possible may depend on which aspect is invalidated. The small numbers standard is a statutory requirement that could be satisfied on an LOA-by-LOA basis in accordance with the ruling of a court if, for example, the court upholds NMFS' LOA-by-LOA aspect of our interpretation but rules adversely on other aspects of our small numbers interpretation. Importantly, the negligible impact analysis for this five-year rule is the biologically relevant inquiry, and that analysis is based on the total annual estimated taking for all activities the regulations will govern over the 5-year period. Our ability to issue LOAs to allow the incidental take of marine mammals, subject to the mitigation, monitoring, and reporting requirements, is based on our findings in this final rule that the total taking over the 5-year period of the rule will have a negligible impact on the affected species or stocks; that the mitigation and related monitoring required in the ITRs will effect the least practicable adverse impact on those species or stocks; and our determination in this rule that any LOA we issue must not involve taking that exceeds one-third of the abundance for that species or stock.

Comment 31: NRDC asserts that the MMPA “does not authorize NMFS to reissue an incidental take regulation after five consecutive years have passed” and that NMFS is ignoring new information that has become available over the past 5 years. Commenters also suggest that, simply because of the nature of NMFS' action, i.e., reimplementation of the rule, that its necessary determinations (which incorporate all relevant newly available information) are not “fresh.”

Response 31: NRDC mischaracterizes the nature of this action, stating that NMFS is “attempting to extend the duration of its 2021 rule beyond 5 years by freezing its previous findings in amber.” To the contrary, NMFS has followed all requirements under both MMPA section 101(a)(5)(A) and the APA by conducting a new notice and comment rulemaking; we do not seek to extend the duration of the 2021 rule beyond 5 years without performing the required analysis, making the necessary determinations, and engaging the public as required. While it is correct that NMFS is using the same estimated take numbers from our recent 2024 rule and reimplementing the previously prescribed mitigation, monitoring, and reporting requirements, all necessary analyses have been performed in light of new information. Commenters ignore the substantive changes to estimated take numbers considered for NMFS' 2024 final rule, which incorporated new information available since issuance of the 2021 final rule. No new information is available since 2024 that would impact NMFS' take estimates. Similarly, NMFS has not frozen its previous findings “in amber,” but has revised its negligible impact analysis and least practicable adverse impact analysis to consider all relevant new information available since issuance of the 2021 rule. NMFS has appropriately conducted a new action, has performed

new analysis incorporating new information, and has made the necessary findings to issue this final rule and ITRs.

Comment 32: NRDC asserts that NMFS cannot rely upon BOEM's 2017 PEIS to support issuance of a new rule, stating that a new NEPA analysis must be prepared for a new action and that, at minimum, NMFS must supplement the analysis. NRDC also states that NMFS cannot rely on the PEIS because it “does not adequately address NMFS' own actions and responsibilities under the MMPA,” given that BOEM's PEIS is “framed around a fundamentally different purpose and need” relating to its mandates under OCSLA that is “incongruent with NMFS obligations under the MMPA.” Finally, NRDC states that BOEM's PEIS, as it relates to marine mammals, is deficient on its face due to the range of alternatives and mitigation considered, significance criteria, take and impact estimates, and cumulative impacts analysis. Animal Counsel similarly suggests that the 2017 PEIS must be supplemented.

Response 32: NRDC suggests that NMFS' adoption of BOEM's PEIS and issuance of a Record of Decision (ROD) is not sufficient analysis to address site-specific actions, claims that “neither NMFS nor BOEM ever envisioned that the 2017 PEIS would itself be sufficient to constitute a hard look at future actions not yet contemplated by the agencies,” and, without evidence, states that the PEIS is not sufficient to satisfy NMFS' NEPA obligations for actions implementing its ongoing program of issuing requested incidental take authorizations. NMFS' ROD refutes these assertions in part. Of relevance, the ROD states that “unless [future] applications vary substantially from the analysis provided in the 2017 Final PEIS and final rule, this ROD forms the basis for NMFS' decision under NEPA, with no further site-specific analysis necessary.” As described in detail herein, the current application and action is the same as that analyzed in the PEIS. The ROD further announces NMFS' decision “to issue . . . future ITAs on a case-by-case basis, if appropriate, consistent with the analyses in the 2017 Final PEIS and mitigation measures specified in the final ITR.” Therefore, NMFS' expressed intentions regarding future reliance on BOEM's PEIS were clearly stated in its ROD.

Regarding the suggestion that NMFS can no longer rely on the 2017 PEIS absent supplementation, NMFS has carefully evaluated relevant new information and circumstances over the intervening period and determined that supplementation is not required. This inquiry, in part, entails an evaluation as to whether new information not previously considered in the PEIS is now available. Given the new information available since the 2017 PEIS, we evaluate whether that information changes the impact analysis for marine mammals contained in the PEIS, and if the impact analyses are different, whether the new information and impact analyses change the impact conclusions provided in the PEIS. One component of new information available since issuance of the PEIS is NMFS' revised take estimates, produced through incorporation of new information in NMFS' 2024 final rule. These revised take estimates, incorporating the current best available scientific information, are uniformly and dramatically lower than those upon which the PEIS analysis is founded. As a result, the PEIS impact conclusions remain robust.

Commenters imply that the PEIS must be supplemented simply due to its age. However, they provide no detailed discussion as to how they believe that the information that is newly available would influence the relevant impact analysis. By design, the PEIS provided analysis of a substantial program of industry survey activity, including evaluation of estimated marine mammal takes substantially greater than those at issue here. Regarding Rice's whale, while the regulatory status of the species has changed in the intervening years, the information underlying that status change, which dates to NMFS' 2016 status review, adequately informed the PEIS's analysis such that it is robust to the evolution of scientific knowledge.

The proposed action at issue in the PEIS is BOEM's issuance of permits or authorizations for survey activities in the GOA. PEIS Chapter 1.1.1. The PEIS also recognizes that NMFS' proposed action is a decision on whether to approve requests for incidental take regulations and is not bound to a particular timeframe or specific request. NOAA was a cooperating agency on BOEM's PEIS, as NOAA has jurisdiction by law and special expertise over marine resources impacted by the proposed action, including marine mammals and federally listed threatened and endangered species. The PEIS explicitly recognizes that the PEIS would be used in support of NMFS' decision on requests for incidental take regulations. See PEIS Appendix B.

It is accepted NEPA practice for NOAA to adopt a lead agency's NEPA analysis when, after independent review, NOAA determines the document to be sufficient. Specifically here, NOAA is satisfied that BOEM's PEIS adequately addresses the impacts of issuing MMPA incidental take authorizations and that NOAA's comments and concerns have been adequately addressed. There is no requirement that NMFS, as a cooperating agency, issue a separate purpose and need statement in order to ensure adequacy and sufficiency for adoption. Nevertheless, the statement of Purpose and Need in the PEIS explicitly acknowledges NMFS' own separate action of issuing an MMPA incidental take authorization, and the PEIS is replete with discussion of issues relating to the issuance of an MMPA authorization, including discussion of marine mammal impacts, mitigation, and take estimates. NMFS' early participation in the NEPA process and the agency's continuing role in shaping and informing analyses using its special expertise ensured that the analysis in the PEIS is sufficient for purposes of NMFS' own NEPA obligations related to its issuance of an incidental take authorization under the MMPA.

Regarding the alternatives, NMFS' early involvement in the development of the PEIS and role in evaluating the effects of incidental take under the MMPA ensured that the PEIS would include adequate analysis of a reasonable range of alternatives for NMFS. The PEIS includes a no action alternative specifically to address what could happen if NMFS did not issue an MMPA authorization. See PEIS, Chapter 2.9.1, pp. 2-20 to 2-22. Some of the alternatives explicitly reference marine mammals or mitigation designed for marine mammals in their title. More importantly, these alternatives fully analyze a comprehensive variety of mitigation measures for marine mammals. This mitigation analysis supported NMFS' evaluation of our options in potentially issuing an MMPA authorization. This approach to evaluating a reasonable range of alternatives is consistent with NMFS' policy and practice for issuing MMPA incidental take authorizations. NOAA independently reviewed and evaluated the PEIS, including the purpose and need statement and range of alternatives, and determined that the PEIS fully satisfies NMFS' NEPA obligations related to its decision to issue the MMPA final rule requested by BOEM, as well as future incidental take authorizations.

Regarding NRDC's reiterating of comments it submitted during the PEIS development process, as a cooperating agency NMFS reviewed all responses to comments on the draft PEIS that were relevant to its management authorities

and provided input where we deemed it appropriate. See Appendix M of the Final PEIS.