2020-20304. Pacific Island Fisheries; Sea Turtle Limits in the Hawaii Shallow-Set Longline Fishery

AGENCY:

National Marine Fisheries Service (NMFS), National Oceanic and Atmospheric Administration (NOAA), Department of Commerce.

ACTION:

Final rule.

SUMMARY:

This final rule revises measures that govern interactions between the Hawaii shallow-set pelagic longline fishery and sea turtles. This rule lowers the annual fleet interaction limit (“hard cap”) for leatherback sea turtles from 26 to 16, and removes the annual fleet hard cap for North Pacific loggerhead turtles. This rule also creates individual trip interaction limits of two leatherback and five North Pacific loggerhead turtle interactions, with accountability measures for reaching a limit. This rule provides managers and fishermen with the necessary tools to respond to and mitigate changes in North Pacific loggerhead and leatherback turtle interactions to ensure a continued supply of fresh domestic swordfish to U.S. markets, consistent with the conservation needs of these sea turtles. This action also ensures that the Hawaii shallow-set longline fishery operates in compliance with the conditions of a recent biological opinion (BiOp).

DATES:

This rule is effective September 17, 2020.

ADDRESSES:

Copies of Amendment 10 to the Fishery Ecosystem Plan for Pelagic Fisheries of the Western Pacific (FEP) and supporting documents are available at www.regulations.gov,, or from the Western Pacific Fishery Management Council, 1164 Bishop St., Suite 1400, Honolulu, HI 96813, tel [phone], fax [phone], www.wpcouncil.org.

FOR FURTHER INFORMATION CONTACT:

Joshua Lee, NMFS PIR Sustainable Fisheries, 808-725-5177.

SUPPLEMENTARY INFORMATION:

The Hawaii shallow-set pelagic longline fishery primarily targets swordfish (Xiphias gladius) on the high seas in the North Pacific Ocean. The Council and NMFS manage the fishery under the FEP and implementing regulations, as authorized by the Magnuson-Stevens Fishery Conservation and Management Act. The fishery occasionally hooks or entangles protected species, including sea turtles. To address these interactions, NMFS has implemented conservation and management measures, including limits on the number of interactions allowed between the fishery and leatherback and North Pacific loggerhead sea turtles.

On June 26, 2019, NMFS issued a BiOp on the effects of the shallow-set fishery on marine species listed under the Endangered Species Act (ESA). The BiOp includes measures required to minimize the effects of incidental take. This rule implements some of those measures. This rule revises the annual fleet hard cap for leatherback sea turtles from 26 to 16. If the fleet reaches this limit, NMFS would close the fishery for the remainder of the calendar year. This rule also removes the annual fleet hard cap on North Pacific loggerhead turtle interactions because it is not necessary at this time for the conservation of this species. If the fishery exceeds the Incidental Take Statement (ITS) for any species in the current valid BiOp, NMFS would reinitiate ESA Section 7 consultation for that species. Finally, this rule establishes limits of two leatherback and five loggerhead turtles per vessel per individual fishing trip. If a vessel reaches either sea turtle limit during a fishing trip, it must immediately stop fishing and return to port, and may not resume shallow-setting until it meets certain requirements. Additional restrictions apply to vessels that might reach a trip limit twice in a calendar year.

All other requirements in this fishery continue, and NMFS will continue to monitor the Hawaii shallow-set longline fishery. You may find additional background information on this action in the preamble to the proposed rule (85 FR 6131, February 4, 2020), and it is not repeated here.

Comments and Responses

On January 23, 2020, NMFS published a notice of availability (NOA) for Amendment 10, including an environmental assessment (EA), and request for public comments (85 FR 3889); the comment period ended March 23, 2020. On February 4, 2020, NMFS published a proposed rule that would implement the management measures described in Amendment 10 (85 FR 6131). That comment period ended on March 20, 2020. NMFS received comments from individuals, the fishing industry and non-governmental organizations, and a petition with signatures, and responds below. Additionally, NMFS received and considered all comments requesting additional minor corrections and clarifications when finalizing Amendment 10 and the EA associated with this final action.

Comment 1: NMFS unlawfully failed to apply the best scientific information available when it “failed” to consider a population viability analysis (PVA) model of leatherback and loggerhead trends with and without fishery mortalities. NMFS “refused” to model sea turtle trends with mortalities because it could not explain why the fisheries' impacts would not accelerate the species' decline. As a result, the biological opinion merely describes the proportion of the adult population and total population that the fishery is expected to kill at benchmark intervals, which is the approach invalidated in TIRN v. NMFS, 878 F3d 725 (9th Cir. 2017). Moreover, the Ninth Circuit has held that where baseline conditions already jeopardize a species, an agency may not take action that deepens the jeopardy by causing additional harm. NWF v. NMFS, 524 F3d 918 (9th Cir. 2008). Without any valid scientific analysis, there is no basis for NMFS to conclude that fishery mortalities would not jeopardize loggerhead or leatherback sea turtles. The PVA take model finalized after the biological opinion was completed confirms that the action accelerates species decline and is therefore jeopardizing.

Response: In conducting the consultation required by Section 7 of the ESA, NMFS is required to use the best scientific and commercial data available. NMFS met this mandate. As described in more detail below, the type of analysis envisioned by the requester is neither a singular nor a simple analysis. Rather, it involves the creation of three separate models. By the time the biological opinion was issued in June of 2019, NMFS had two of the three models (including a PVA model) and took them into account in the development of the biological opinion. The final model was not available until March 2020, several months after the biological opinion was issued.

Importantly, the model the commenter alludes to is actually composed of three separate modeling elements, which must occur sequentially and cannot be performed simultaneously. First, a Bayesian model or prediction of the number of future interactions that each species would be likely to have with shallow-set vessels must be developed; then, a PVA must be developed for the entire population; step three is the development of the final model, the so-called “take model.” This is a mortality model that requires backing out information on the fishery that is already incorporated into the PVA, to avoid the “double-counting” of the fishery impact, and recomputing the trend, with and without the fishery. This take model was not available until March 2020.

While the first two elements of this overall modeling were available and considered as part of the biological opinion, NMFS recognized that there were important limitations to the modeling that needed to be taken into account. Initially, NMFS was concerned that drawing inferences from models developed with incomplete trend data representing less than one generation and virtually no demographic data, would give the appearance of precision when, in fact, data on loggerhead and leatherback sea turtles are insufficient to develop reliable models of the effect of “take” pre- and post-fishery.

This issue has long been a source of concern to the scientific community, and is discussed at length in the National Research Council 2010 publication, “Assessment of Sea-Turtle Status and Trends: Integrating Demography and Abundance.” More than 10 years ago, the National Academies of Sciences gathered together a team of international scientists to discuss sea turtle assessments and models, and underlying the entire review is one singular problem—that sea turtle modeling and analysis that has been done has had to “compensate for a debilitating lack of data (NRC 2010).” Although progress has been made, this data problem persists as there continues to be a substantial lack of demographic data available on sea turtles.

Importantly, for most sea turtle populations, there are no or very limited population-specific demographic data, such as life-stage durations or survival rates. This is true of loggerhead and leatherback sea turtles, as considered in the BiOp. Appropriate data on vital rates are critical for sea turtle population estimation, because nest count data and adult nesters represent only a very small fraction of the total population. “These are clear reasons not to put too much confidence in the assessment of trends in nesting numbers, even if it uses the “best available data” in a careful and rational way” (Crowder 2018).

Recognizing the inherent limitations in modeling with limited demographic data, and because NMFS was cautious about the falsely implied precision of converting all individual turtles that interact with the fishery to an estimated number of adult nester equivalents so as to establish a common currency by which to evaluate the effect of the fishery against the PVA, NMFS determined that the information available in June 2019 (i.e., the first two models) was sufficient to conduct a jeopardy analysis without delaying the consultation further until the third model (the take model) was available. NMFS was also concerned that a third model could compound the error inherent in the PVA, discounting the importance of the injury and death of individual turtles at ages younger than adults and give the false appearance of precision around the model estimates.

Contrary to the commenter's suggestion, NMFS did not “fail” to develop the third model. The third model was ultimately developed and produced nine months later. It was peer reviewed and it supported the “no jeopardy” conclusions in the biological opinion. Further, the model was deemed the “best available science” by the Council's Scientific and Statistical Committee (SSC) although their role was to look at its usefulness under the Magnuson Act as opposed to the Endangered Species Act.

The PVA model in question relies solely on trends in annual nest counts from a subset of beaches considered representative for each species (leatherbacks and loggerheads). Nest counts are then converted to individual nesters and these numbers are used to predict trends in the populations. The NRC notes that methods based on reproductive value (or adult equivalents), such as used in the PVA model, are best used for relative comparisons within species to set priorities for research or conservation effort, rather than attempts at quantitative assessment of threats or setting take limits, as this could `discount' takes of some turtles.

Development of the first two models took about nine months to complete, and consultation was initiated after the completion of the first model. Consultation timelines were running while the second (PVA) model was in development. The consultation was extended more than six months to allow completion of the second model. Based on the data and models available at the time, NMFS was able to conclude its consultation without waiting a further nine months on the third model.

The commenter's claim regarding TIRN v. NMFS is also in error. Contrary to the comment, NMFS did not merely employ the same analytical method as addressed in TIRN v. NMFS. The analytical method the commenter refers to describes the proportion of the adult population and total population that the fishery is expected to kill at benchmark intervals. Instead, when developing the BiOp on the shallow-set longline fishery, NMFS analyzed the effect of the action on several demographically important subsets of the total population: The adult population, the portion of the adult population represented by females only, the proportion of the population represented by unique life history types (summer nesters, summer nester adults and summer nester females), and the potential to disproportionately affect a subpopulation or breeding aggregation (e.g., Ryuku loggerhead sea turtles).

Importantly, NMFS evaluated these effects under four scenarios: The current population size, and three different future population numbers (50, 25, and 12.5 percent of the current population size). This was done to ensure that all impacts considered in the Status of the Species, Baseline and Cumulative Effects sections, including other federally authorized fisheries and foreign fisheries, were appropriately factored into the evaluation. In other words, consistent with the ESA implementing regulations and the approach to the assessment as described in the BiOp, NMFS examined the effect of the action on numbers (e.g., total abundance, numbers of adults, numbers of females), reproduction (e.g., numbers of females and reproductive adults), and distribution (e.g., subpopulations and unique life histories) over a 40-year time horizon (under the assumption of continued degradation of the baseline conditions) and each of these analyses led us to conclude that the small number of animals that would be taken by the shallow-set longline fishery would not, directly or indirectly, reduce appreciably the likelihood of both the survival and recovery of any listed species in the wild by reducing the reproduction, numbers or distribution of that species. This analysis did not discount or remove some of the animals from its assessment because they were suspected of being juveniles or sub-adults that would be unlikely to survive to reproduction (adult nester equivalents). Because there is no reliable known size threshold for an adult, and we do not know that age and stage survival rates would apply to a subset of the population that is affected by the fishery, and we do not know age and stage survival rates for loggerhead and leatherback sea turtles, the BiOp assumed that each individual turtle that the fishery interacts with has the same chance of reaching its full reproductive potential as the next. In other words, juvenile sea turtles were not considered less important than an adult and the interaction with animals suspected of being in the juvenile age-class were not discounted in the BiOp.

The commenter also points to the Ninth Circuit's dicta regarding “baseline jeopardy.” NMFS believes that the Court's use of this term misconstrues the analytical standard that must be applied for a valid Section 7 analysis. To determine whether an action will jeopardize the continued existence of a species, NMFS must assess the effects of a Federal agency action by adding those effects to the environmental baseline. Jeopardy occurs when the effects of the action together with the environmental baseline show that the action appreciably reduces the species' likelihood of survival or recovery. The ESA does not recognize a species' status as being in a pre-determined condition of jeopardy. As NMFS explained in the proposed (83 FR 35178, July 25, 2018) and final (84 FR 44976, August 27, 2019) Section 7 rules, the ESA does not recognize a baseline state of jeopardy. Rather, the ESA is concerned with the action's effects, and whether those effects appreciably reduce the likelihood of the species' survival or recovery in the wild.

While our PVA illustrates that long-term persistence of the leatherback sea turtle is precarious, the proper inquiry is whether the action causes new harm that is consequential to the species' viability. Minor impacts to the species' pre-action condition are not jeopardizing if they do not result in consequential reductions in numbers, reproduction, or distribution at the species level. NMFS too is concerned with the long-term status of the leatherback sea turtle. However, to complete its evaluation of the action under ESA Section 7, NMFS appropriately relied upon its understanding of ecological theory and experience with population growth or decline, which is captured by the fundamental equation: N_t = N₀ + (Births + Immigration)−(Deaths + Emigration).

Every population model derives from this equation (the “BIDE” equation). The BIDE equation reveals the error in asserting that the added loss of a few individuals from a population that exhibits a declining trend necessarily “jeopardizes” the continued existence of a population or species. A declining trend means that the ratio between N_t and N₀ is less than 1.0 (or substantially less than 1.0, if we consider year-to-year variation). However, a population experiencing such a decline still has births and, in some cases, immigration. To illustrate, a small number of deaths would not alter the trajectory of even a declining population if the number of births exceeds the number of deaths in the same time interval (or if recruitment into a life history stage exceeds the number of deaths in that stage). The implication of the BIDE equation is that even if “tipping points” are nominally identified and quasi-extinction thresholds (QETs) estimated, factors that influence productivity outside of our knowledge and control can shift abundance upward, making both constructs invalid.

NMFS analyses were complete given the available data, and NMFS correctly analyzed the effects of the action on the species' viability. Because of its concerns about the paucity of data, NMFS examined several reasonable step-down scenarios relative to the numbers, distribution, and reproduction of the species. NMFS remains confident in its conclusion that the small number of mortalities, even for the leatherback sea turtle and even though there is a measurable reduction in numbers associated with the proposed action, would not appreciably reduce the species' likelihood of survival or recovery.

This conclusion is borne out in the third model (the take portion of the PVA model), which the commenter references. Although the take model was not available when the BiOp issued, subsequent analysis using the model confirms the BiOp's conclusions that the action is not expected to directly or indirectly reduce appreciably the likelihood of either the survival or recovery of leatherback or loggerhead sea turtles in the wild. In other words, the likelihood of survival and recovery remains relatively constant with or without the action.

Although the take model suggests that there is a difference between the “no take (PVA)” model and the “take” model for leatherbacks, the modeled differences are not detectable for roughly 40 years (to 2060). The difference predicted by the third model is not discernable at the point when the leatherback population reaches half its current abundance, though there is a minor observed difference as the population gets smaller (0.01 percent difference when the leatherback sea turtles population reaches 25 percent or 12.5 percent of its current size) and time considered is lengthened. We stress the point that the farther out the projection, the more uncertainty we have around the estimates, and that this model and the analysis in our BiOp applies as a protective assumption, a consistent annual amount of take even though, as the population declines over time, the likelihood of take of individuals also declines. In other words, limitations in our predictive capabilities and changes in future management regimes would render predictions over a longer period increasingly speculative. This is true not only for the PVA with take and without take, but is also true of the analysis we did for the BiOp. Shorter term estimates (e.g., 10 years) are expected to provide more accurate predictions of the effect of the action, but estimates at a longer time interval are more uncertain. In addition, an underlying caveat or assumption of the model and the analysis in the BiOp is that as the population continues to decline (50 percent, 25 percent, and 12.5 percent of current size) the actual number of animals taken in the fishery would not change. This assumption is considered protective of the species, but highly unlikely to be true over an extended time. For example, at the prediction point approximately 40 years in the future (2060), when the potential impacts of the shallow-set longline fishery appear to be detected for leatherbacks, the mean number of nesting females in the absence of the shallow-set longline fishery is predicted to be 24, and the continued fishery take of up to two adult female per year therefore becomes detectable. However, as the population declines and a species becomes rarer, we would generally expect that the rate of interaction (take) would also tend to decline. Since we do not know how “rareness” would affect future interaction rates, we opted to assume that interactions would remain constant over time for the purposes of our jeopardy analysis. This assumption alone would tend to cause longer term evaluations to be less reliable, and would warrant careful consideration of perceived mathematical differences in predicted impacts resulting from the action. To highlight this point, the “take” PVA model predicts that the leatherback population will become extinct 5 years earlier than the “non-take” model. However, in the year when the mean “take” model predicts extinction, the number of nesting females remaining in the “no-take” model is one nesting female. Logically, maintaining the unrealistic same level of take at this point makes the population appear to reach extinction levels 5 years sooner under the “take” model, when this is really just a result of our assumption of constant fishery interaction numbers. There was no discernible difference at all for loggerheads between the “no take (PVA)” model and the “take” model.

Both approaches, the analytical approach taken in the BiOp, and the take/no take model completed nine months after the BiOp have the same basic structural limitations. The primary limitation stems from the ability to reliably predict population growth (or decline) and changes in demographics, which are critical to understand species' extinction risk. Both assessment methods are reliant upon female nester abundance predictions from nest counts. Because these data represent a very small fraction of the total population, and little is known about males, juveniles, or population specific demographics, conclusions drawn about the species from these data are likely to be inaccurate. Thus, NMFS took steps in the consultation and the BiOp to develop a thoughtful and appropriately precautionary analytical approach that would not disadvantage the species. NMFS considers the approach in the BiOp to have certain advantages as an assessment tool because it recognized the importance of unique life histories and the role of small subpopulations (independent demographic units). Nevertheless, both the third NMFS model (take model) and the analysis contained in the BiOp support the same conclusion that the proposed action would not directly or indirectly reduce appreciably the likelihood of both the survival and recovery of any listed species in the wild by reducing the reproduction, numbers or distribution of that species.

Comment 2: The de-lifing approach was improperly applied prospectively across multiple generations, and erroneously assumed a 6 percent generational decline for leatherbacks rather than a 6 percent annual decline.

Response: As defined by Coulson et al. (2006), de-lifing is a retrospective analysis that address questions in evolutionary ecology by identifying an individual's observed contributions to the mean fitness of a population in a given year (as opposed to an entire generation). Upon careful reconsideration, we agree that we erred in our application of the de-lifing approach, and therefore cannot rely upon this analytical method as described in the BiOp. Specifically, the approach was improperly applied prospectively across multiple generations, and contained a mathematical error. However, the de-lifing analysis was not an essential component in reaching the no-jeopardy conclusion for leatherbacks. Our BiOp examined the effect of the action on several reasonable and demographically important units, as described above, including females, summer nesters, small subpopulations, and at reduced population sizes. Based on the multiple analytical evaluations, and the recently published model, the action did not materially change the species' pre-action condition—not its reproduction, numbers, or distribution—and did not hasten the species' decline.

Comment 3: By failing to calculate the species' tipping point or QET, the agency failed to adequately examine the action's impacts on recovery.

Response: The commenter asserts that the failure to calculate a tipping point is relevant to the action's impact on recovery. First, a tipping point is not a scientific construct; it is a term that embodies a general concept that beyond a certain threshold, large uncontrolled shifts in ecology will occur. Second, the tipping point concept does not have bona fide relevance to conservation or recovery within the ESA, as is specifically noted in the recent regulations for Interagency Cooperation under the ESA (84 FR 44976, August 27, 2019). As explained in the BiOp, tipping points (and QETs) are theoretical constructs that the commenter suggests serve to identify a defined level beyond which imperiled populations cannot be expected to recover. It is technically impossible to know, in advance, where the “tipping point” that forecloses recovery might lie for free-ranging plants and animals (and even animals in captivity). Similarly, QETs are arbitrary thresholds used in population ecology to identify some non-zero point below which population abundance might fall, and the probability of falling below that non-zero threshold. Importantly, QETs, like tipping points, are only theoretical methods to evaluate extinction, they are not determinative, and while potentially helpful in assessing jeopardy risk relative to survival under the ESA, they are not relevant to the separate assessment of recovery. In a logical analysis, the effect of a proposed action on the potential for recovery is appropriate when the first analysis for jeopardy concludes with “does not reduce the likelihood of survival;” As the recovery standard is a level of abundance and reproduction that allows a species to be self-sustaining in the wild without the protections of the ESA, QETs and tipping points are not pertinent to that portion of the analysis.

In the BiOp, we estimated the probability that that species would become extinct over time, but we do not have predefined thresholds or decision rules as to what point within that probability a “jeopardy threshold” is reached for each species. NMFS has explored the use of quantitative thresholds in listing, in particular, and several such extinction thresholds have been suggested for more than 20 years. The same premise could apply to “jeopardy” evaluations relative to “survival” and “recovery,” yet the agency has declined to predefine policy thresholds for its ESA decisions because such predefined decision rules in data deficient situations would have to be established as general guidelines or rules, and would be arbitrary for most species. No set of decision rules can compensate for information gaps, particularly when trends are poorly known and demographic data are absent. Moreover, in many cases establishing population level thresholds would overshadow understanding and evaluating the threats on the underlying independent demographic units that comprise the listed species.

Our assessment approach in the BiOp recognizes that a species' risk of extinction is affected by the strength or weakness of the populations or independent demographic units that comprise that species. Producing an assessment approach that relies solely on quantifiable metrics at the species level would fail to account for the important role that the underlying independent demographic units play in the species' risk of extinction, particularly where there is insufficient information to adequately develop a credible quantifiable metric.

Early work on PVA and population ecology did include efforts to define minimum viable populations, defined as the smallest number of individuals required for a population to persist at some predefined probability of time. This led to the development of the 50/500 rule in conservation management, which simply states to avoid inbreeding depression (loss of fitness due to genetic problems), an effective population size of at least 50 individuals is necessary. To ensure that the population can maintain its evolutionary potential to cope with environmental change at least 500 individuals are necessary. Following this line of thinking, 50 individuals might be a survival threshold and 500 individuals might be best considered the minimum number necessary to ensure recovery. However, almost 40 years have passed since these concepts were introduced into the field of conservation biology. We now know that these arbitrary thresholds are not broadly useful, because species differ in their needs, reproductive strategies, age at fecundity, et cetera. As discussed at length in the BiOp, some species can dip well below 500 and be recoverable, and many survive after dropping to numbers below 50.

Common tipping point metrics, or QETs, that are often used in PVAs and many scientific analyses include several of the same metrics we used in the development of our PVA for loggerhead and leatherback turtles, and in our “jeopardy” evaluation (e.g., mean and median times until each species declines to 50 percent, 25 percent, and 12.5 percent of current abundance estimates, probability of each species reaching those thresholds in 5, 10, 25, 50, and 100-year time intervals with associated 95 percent confidence intervals). We used these metrics to characterize the current viability of loggerhead and leatherback sea turtles but these predictions, at the species level, did not help characterize the status of the independent demographic units that comprise each species over time. Demographically-independent units (populations, subpopulations, demes, etc.) that comprise each listed species are important to understanding the species' chances for both survival and recovery. The structure and performance of the two species as they have been listed, the sub-populations that comprise these species, the populations that comprise the various sub-populations, and the demes that comprise those sub-populations are addressed in our consultation using both quantitative and qualitative means, and it is in this combined approach we evaluated the impact of the action on the species' chance of both survival and recovery.

As noted in the NRC 2010 report, reference points are used in fisheries management to demark levels of overfishing and the level of stock abundance that results in sustainable populations, however, such analyses require long time series of data and detailed information on a population's demographic rates. Without such demography there is no way to predict the effects of fishery bycatch, especially for animals as long-lived as sea turtles. The NRC also notes that methods based on reproductive value (or adult equivalents), such as used in the PVA “take” model, are best used only for relative comparisons within species to set priorities for research or conservation effort, rather than attempts at quantitative assessment of threats or setting take limits.

While research has been done on identifying “tipping points” in species abundance trends, these have primarily been either theoretical in nature, using laboratory studies of fruit flies in which 20 or more generations of data are available for analysis, or are retroactive studies in which patterns are only realized after they have happened. The generation time for leatherback sea turtles is approximately 22 years assuming age at maturity is 16 years and annual adult survival rate is 0.89. The longest time series available for the PVA was 17 years; hence, identifying tipping points from a time series of abundance of less than one generation is not feasible, would not be a reliable metric, and would not be a relevant metric for the recovery component of the jeopardy analysis.

Comment 4: The proposed individual vessel limits are too high to effectively reduce endangered sea turtle interactions and mortalities as required by Reasonable and Prudent Measure 1 of the ITS in the BiOp. Further, this measure undermines the entire regulatory scheme by allowing a few bad actors to single-handedly exacerbate the likelihood of sea turtle extinction.

Response: This final rule establishes individual trip limits of five loggerhead and two leatherback turtles, as required by terms and conditions of the BiOp, which apply to every vessel in the shallow-set longline fishery. If a vessel reaches either limit, NMFS will require that vessel stop fishing and return to port, and that vessel will be prohibited from shallow-set fishing for 5 days. This provides a 7-10 day cooling-off period given the distance between fishing grounds and ports in Hawaii and California. The cooling-off period may allow the environmental conditions contributing to the high interactions to dissipate and reduce the likelihood of additional interactions in that area in subsequent trips. If a vessel reaches a trip limit twice in a calendar year, NMFS will prohibit that vessel from shallow-set fishing for the remainder of the calendar year. In the following calendar year, that vessel will have a vessel limit of five loggerhead or two leatherback turtles).

The Council's recommendation to specify a loggerhead trip limit of five was based on the finding that it would provide the most meaningful reduction in interactions in years with high interaction rates, such as those observed in 2017-2018. Observed sea turtle interaction data since 2004 indicate that most shallow-set longline trips with loggerhead turtle interactions have one-two interactions per trip, with a small proportion of trips having four or more interactions coinciding with years with the highest total fleet-wide interactions. The NMFS Pacific Islands Fisheries Science Center (PIFSC) simulated different levels of trip limits, ranging from two-five, to past observed interactions. Based on these simulations, a limit of five loggerhead turtles per trip would have reduced loggerhead turtle interactions in 2018 by 30 percent, even without accounting for avoidance behavior by the vessels. The Council, therefore, determined that the loggerhead trip limit of five would provide a mechanism for response to higher interaction rates, and minimize further interactions when such higher interaction rates are detected while helping to ensure year-round supply of swordfish to meet domestic demand. Note the leatherback trip limit is a complement to, and not a replacement of the fishery's hard cap of 16 leatherback turtles, and also serves as preventative measure if higher interaction rates are observed in the future, and may reduce the likelihood of reaching the hard cap if vessels are able to avoid a second interaction after encountering the first leatherback on a given trip.

Individual trip limits are expected to provide early detection to higher interaction rates that may indicate a potential for higher impacts to sea turtle populations in a given year, and are expected to reduce loggerhead and leatherback turtle interactions in such years. Individual trip limits are intended to mitigate a large proportion of loggerhead and leatherback turtle interactions from occurring in a single trip. Observed sea turtle interaction data since 2004 indicate that trips with loggerhead turtle interactions typically have one-two interactions per trip in years with low fleet-wide loggerhead turtle interactions. Conversely, trips with three or more loggerhead turtle interactions have been observed in years with high fleet-wide interactions. In 2018, when the highest number of loggerhead turtle interactions was observed, 16 percent of the trips contributed to 58 percent of the total fleet-wide interactions. Monitoring the number of loggerhead turtle interactions per trip would provide an early detection mechanism for higher fleet-wide interactions, and the individual trip limit is expected to provide a “dampening” response by minimizing further interactions on those trips.

Individual trip limits also provide an individual vessel incentive to avoid sea turtle interactions because shallow-set vessels may fish 500-1,000 nm from port and require considerable up-front costs for each trip, and thus a shortened trip duration may result in net loss for that trip. Given the economic disincentive of reaching the trip limit, vessel operators are more likely to employ additional avoidance strategies if they encounter multiple interactions in a trip, such as moving away from the area and avoiding areas with higher potential for interactions using information from the NMFS TurtleWatch program. If a vessel reaches a trip limit once, that vessel is more likely to avoid fishing in the same area as the previous trip and employ additional avoidance strategies to prevent further economic loss. Thus, conservation benefits are expected even before the individual trip limit is triggered. Because reaching a trip limit twice in a calendar year would result in that vessel being prohibited from fishing for the remainder of the year, there is a direct disincentive to continue fishing practices that might result in additional interactions.

Additionally, the return to port requirement serves as an additional deterrent to reaching a vessel limit due to the distance between fishing grounds and ports in Honolulu and California where shallow-set vessels land their catch. The travel distance from port to the areas where the shallow-set vessels typically operate is at least 2-3 days and may take as long as 5-6 days one-way. If a vessel reaches a trip limit, the travel time back to port, time in port, and travel time to return to fishing grounds would result in a minimum of 7-10 day days of no fishing. This time lag between the last set on the trip in which a vessel reaches a trip limit and the first set on the subsequent trip also provides a cooling-off period that allows for the conditions contributing to the high interactions to dissipate and reduces the likelihood of additional interactions in that area in subsequent trips. The trip limit also places the accountability of interactions on individual vessels and ensures that the consequence burden remains with the vessel that reaches the individual trip limit.

The Council considered the individual vessel limit, as a standalone measure, to be punitive by discouraging participation in the fishery, and thus inconsistent with the purpose and need of the action to help ensure year-round fishing operations and a continued supply of fresh swordfish to U.S. markets.

Comment 5: One hundred percent observer coverage is necessary to enforce interaction limits.

Response: NMFS currently places at-sea observers on 100 percent of shallow-set longline trips, and this action does not change this. Current NMFS observer data-collection protocols instruct observers to report sea turtle interactions using a satellite phone after each observation, which are used to monitor interaction limits. However, NMFS routinely uses statistical modeling as a proven and reliable method for estimating observer coverage necessary to meet management and monitoring objectives, including coverage to monitor for protected species interactions. NMFS will also continue to explore other tools, such as electronic monitoring, to meet monitoring program objectives.

Comment 6: Continued operation of the Hawaii-based shallow-set longline fishery will adversely affect leatherbacks by jeopardizing the species in violation of the ESA and, therefore, NMFS does not have a valid basis to issue a finding of no significant impact, and an environmental impact statement must be prepared to evaluate the significant effects of the fishery on protected species.

Response: NMFS finds that the continued operation of the shallow-set fishery will not adversely affect the leatherback turtle by causing jeopardy to the species, and NMFS is not in violation of the ESA. Under the ESA, NMFS may authorize the fishery to interact with protected species that would otherwise be prohibited, if conducted pursuant to a lawful activity, and if conducted in accordance with the terms and conditions of a no-jeopardy BiOp and ITS. The BiOp concluded the continued operation of the shallow-set fishery is not likely to jeopardize the continued existence of the leatherback turtle, and analyzed up to 21 interactions (3 mortalities) annually when making this determination. Reasonable and Prudent Measure 1 Term and Condition 1a further limits the fishery to 16 interactions annually which represents an approximate 25 percent reduction in the number of turtles from the predicted interaction numbers in this BiOp. If the fishery reaches this limit, the terms and conditions require that NMFS shall close the fishery for the remainder of the calendar year. The hard cap limit, trip limits, and additional accountability measures specified in this rule are consistent with the Reasonable and Prudent Measures and Terms and Conditions contained in the BiOp.

As described in the response to Comment 1, our analysis is further supported by the PIFSC PVA take model to assess the population level impacts of post-interaction mortality of loggerhead and leatherback turtle interactions in the shallow-set fishery (Martin et al. 2020). The model builds upon the PVA considered in the BiOp. Data for the North Pacific loggerhead came from three index beaches in Yakushima, Japan (Inakahama, Maehama, Yotsusehama), which represents 52 percent of the overall population; and data for the western Pacific leatherback population came from two index beaches in Indonesia (Jamursa, Medi, and Wermon), which represent approximately 75 percent of the overall population. These nest counts represent the best scientific and commercial data available for these species. Furthermore, the model is considered to be conservative because the full anticipated take is only applied to the index beaches (approximately 52 percent of the North Pacific loggerhead population and 75 percent of the Western Pacific leatherback population).

For each species, the modeling framework shows the probability of the population being above or below abundance thresholds (50 percent, 25 percent, 12.5 percent of current annual nesters) within a 100-year simulation time frame, and the number of years (mean, median, and 95 percent credible interval) to reach each threshold for both “take” and “no take” scenarios (i.e., the population trends with and without the take associated with the fishery). The take level evaluated in the model was derived from predictions generated by PIFSC using a Bayesian inferential approach (McCracken 2018) and analyzed in the BiOp. Results for both species suggest that the fishery's anticipated take to be negligible on the long-term population trends, with no discernable changes to the probabilities of the populations falling below abundance thresholds between the “no take” and “take” scenarios for the future (Martin et al. 2020). For the leatherback turtle, the difference in the population trend only becomes apparent after the year 2060 and suggests the population would go extinct roughly 5 years sooner than in the “no take” scenario (around Year 2110 vs. 2115). However, this 5-year difference is inconsequential, and the actual population difference of the 5 year divergence represents less than one adult nester. Importantly, the difference seen between the “no take” and “take” scenarios in the 100-year projection is not seen in the 10-year projection (see Martin et al. (2020) Figs. 22 and 23).

As described in the EA and Martin et al. (2020), projections out to 10 years into the future are more relevant biologically for management purposes than to 100 years given the estimated uncertainty in the population parameters. Specifically, the effects of the environmental or anthropogenic drivers on the population would be lagged; therefore, we think the first 10 years is largely based on the previously observed trend but after that we do not have sufficient information to account for uncertainty of the drivers that affect the populations. Additionally, we analyzed the trend with historical impacts from the fishery removed (i.e., by adding back the adult nesters to the population); however, there was no difference between the trends for the “take” and “no take” scenarios for either species for the past.

In summary, while NMFS conservatively estimates the removal of up to three leatherbacks annually by the fishery, this level of take is not expected to have any consequential impacts in terms of reductions in numbers, reproduction, or distribution at the species level. Rigorous terms and conditions that include annual hard caps for leatherbacks and individual trip limits for sea turtle species help ensure that the fishery's already minor impacts are further mitigated. Moreover, NMFS previously completed a comprehensive Environmental Impact Statement on the shallow-set longline fishery in 2008. This action modifies the prior action by implementing new terms and conditions to mitigate impacts to leatherbacks and loggerheads. Accordingly, NMFS properly concluded that an environmental impact statement was not required.

Comment 7: The draft EA is deficient because it does not examine a reasonable range of alternatives. The National Environmental Policy Act (NEPA) requires Federal agencies to “study, develop, and describe appropriate alternatives to recommended courses of action in any proposal which involves unresolved conflicts concerning alternative uses of available resources. Most noticeably, none of the alternatives examined would allow a single “maximum take” trip per year, and another feasible but unexplored alternative is prohibiting fishing in the thermal band between 17.0 and 18.5 degrees Celsius that is preferred habitat for both loggerhead and leatherback sea turtles.

Response: NMFS and the Council complied with all procedures and requirements under NEPA when developing Amendment 10 and this final rule. As described in Section 1.1.2, section 2.1, section 2.3, and Appendix A of the EA, the Council considered a reasonable range of options for managing the loggerhead and leatherback turtle interactions in the shallow-set fishery, including single year hard caps, multi-year hard caps, and removal of hard caps altogether, individual vessel limits as a stand-alone measure, in-season measures (e.g., trip limits and in-season temporary closures), spatial and temporal measures to manage interaction hotspots and non-regulatory measures (e.g., improvements to fleet communication, industry-led initiatives, and furthering research to minimize trailing gear).

In developing these alternatives, the Council considered the following information: Fisheries observer data for loggerhead and leatherback sea turtle interactions since 2004, effort and economic performance trends of the fishery since 2004, population assessments for the North Pacific loggerhead and western Pacific leatherback turtle populations, the BiOp for the shallow-set fishery, the recent characteristics of loggerhead turtle interaction patterns since 2017, the effectiveness of existing mitigation measures such as circle hooks and mackerel-type bait, potential development of industry initiative for a sea turtle avoidance program, impacts of the hard cap closures on fishery performance, and the 9th Circuit Court decision and settlement agreement (Turtle Island Restoration Network et al. v. NMFS; Civil No. 1:12-cv-594-SOM-RLP).

Upon consideration of the broad range of potential management options and available information, and consistent with the action's Purpose and Need, the Council identified individual trip limits as the most practicable and appropriate measure in developing a more responsive management approach that would further minimize impacts to sea turtles while helping to ensure the year-round fishery operations and supply of fresh swordfish to meet market demands. As described in Section 2.3 of the EA, the Council rejected other measures that did not meet the purpose and need, were not practicable, were not necessary or appropriate, or lacked sufficient data to evaluate effectiveness. The measures rejected by the Council include individual vessel limits as a stand-alone measure, real-time spatial management measures, and time-area closures, which are substantially similar to the alternatives identified by the commenter.

Specifically, the Council rejected individual vessel limits as a stand-alone measure because prohibiting vessels from fishing shallow-set for the remainder of the calendar year if vessels reached the established per-vessel limit would not result in meaningful conservation gains compared to the individual trip limits, as the best available information indicate that the likelihood of vessels having multiple trips with high number of turtle interactions in a given year is very low, and individual trip limits are expected to be just as effective in responding to the rapid accumulation of sea turtle interactions as individual vessel limits. The Council also found that individual vessel limits would discourage vessels from participating in the shallow-set sector of the Hawaii longline fishery as the consequence of reaching an individual vessel limit (prohibition from fishing shallow-set gear for the remainder of the year) is expected to act as a disincentive for entering the fishery, and thus would be inconsistent with the purpose and need of the action.

The Council also explored but rejected real-time spatial management measures and time-area closures that included consideration of the TurtleWatch thermal band for loggerhead and leatherback turtles. The Council found that there are insufficient data to conclude that actions to disperse fishing effort from a particular location will positively impact sea turtle conservation. For example, the original TurtleWatch temperature band between 17.5 and 18.5 degree Celsius is intended to encompass approximately 50 percent of the loggerhead turtle interactions, indicating that avoiding effort in that band would shift effort into areas where the remaining interactions have been historically observed. The thermal band identified by TurtleWatch also overlap with productive swordfish fishing grounds during the peak fishing season, and thus prohibiting fishing in such thermal band would likely discourage vessels from shallow-set fishing. Additionally, prohibiting fishing in a non-static thermal band that shifts daily is impractical from both a management and enforcement standpoint, and presents significant challenges in terms of providing fishermen with timely notice.

Following the issuance of the 2019 BiOp, the Council further considered modifying its recommended management action for consistency with the Reasonable and Prudent Measures therein. The alternatives analyzed in the EA represent the final range of alternatives that the Council considered at its 179th Meeting and is a reasonable range based on the purpose and need of the action, history of the development of alternatives, and the need to incorporate the Reasonable and Prudent Measures as part of the Council action.

Comment 8: The Hawaii Longline Association (HLA) supports NMFS and the Council's proposal to eliminate the existing hard cap for loggerhead sea turtles, and although HLA does not actively oppose NMFS and the Council's proposed implementation of a hard cap for leatherback sea turtles, HLA believes it to be unnecessary.

Response: Regarding the loggerhead turtle, NMFS agrees. The annual hard cap was first implemented as a measure to control sea turtle interactions on the model shallow-set longline fishery while NMFS gathered information on the effectiveness of using circle hooks and mackerel-type bait in reducing sea turtle interactions in the fishery. At the time, the best scientific information available indicated that the North Pacific loggerhead turtle population was projected to decline (NMFS 2004). The current best available scientific information indicates that the North Pacific loggerhead population is increasing at an average rate of 2.3 percent, and the total population estimated in the 2019 BiOp is approximately 340,000 turtles. We note that nothing in the ESA requires that fishery hard caps be used as a management tool, and current information strongly suggests that other mitigation measures, including individual trip limits, will be effective in reducing impacts to loggerheads, while allowing for year-round fishing opportunities.

In the absence of a hard cap for loggerhead turtles, the fishery would still be constrained by the individual trip limit of five loggerhead interactions as well as additional restrictions if the trip limit were reached twice in a calendar year. Consistent with the requirements of the ESA, NMFS would reinitiate consultation pursuant to ESA Section 7 if the ITS for loggerhead turtles is exceeded.

Unlike the loggerhead turtle, the current best scientific information available indicates that the western Pacific leatherback population is decreasing at an average rate of −6.1 percent, and the total population estimated in the BiOp is approximately 175,000 turtles. Although NMFS has determined the operation on the fishery is a not likely to jeopardize the leatherback turtle, we have nevertheless taken additional precautions to reduce the hard cap limit to 16, which represents an approximate 25 percent reduction from the ITS, to minimize the impacts, i.e., amount or extent, of incidental take. Furthermore, this term and condition for Reasonable and Prudent Measure 1 set forth in the 2019 BiOp must be undertaken by NMFS for the exemption in ESA section 7(o)(2) to apply to the shallow-set longline fishery.

Comment 9: HLA supports the trip limits of five loggerhead and two leatherback interactions per trip, but objects to the proposed vessel limits that would apply in the subsequent year if a vessel reaches a trip limit twice in a calendar year.

Response: A purpose of this action is to modify sea turtle mitigation measures for effectively managing impacts to leatherback and loggerhead sea turtles from the shallow-set fishery, consistent with the requirements of the reasonable and prudent measures and terms and conditions of the 2019 BiOp. Term and condition 1b states, “. . . NMFS shall require any vessel that reaches a trip limit for either species twice in one calendar year to have an annual vessel limit of 2 leatherbacks or 5 loggerheads for the following year.” As described in response to Comment 6, these measures must be undertaken by NMFS for the exemption in ESA section 7(o)(2) to apply.

Comment 10: The NMFS take estimates and, therefore, its proposed mitigation measures, are based upon overly precautionary incidental take estimates.

Response: For the purpose of ensuring that our analysis is appropriately precautionary, we chose the 95 percent credible intervals when estimating the take level. The 95 percent credible interval fully represents the possible range of takes, and thereby ensures we are not underestimating potential impacts to species over the full period of the action. In terms of take, this means that there is a 95 percent probability in any given year that the true number of animals captured or killed is within the credible interval. While we agree that the fishery is unlikely to capture animals at the 95 percent credible interval year after year, the BiOp accounts for this and examines take at both the 95 percent interval and mean in its analysis.

Comment 11: The PIFSC modeling analysis and report supports and confirms the BiOp “no-jeopardy” conclusion and a determination that the proposed action has no significant impact on the environment.

Response: NMFS agrees the PIFSC modeling analysis and report supports and confirms the BiOp “no-jeopardy” conclusion and a determination that the proposed action has no significant impact on the loggerhead and leatherback sea turtles. See also Response to Comment 1.

Comment 12: Closures and reduced effort in the fishery result in increased domestic reliance on foreign supply and increased adverse impacts on sea turtles.

Response: Our environmental analysis acknowledges fishery closures often result in shallow-set vessels converting to deep-setting gear to target bigeye tuna and continue to fish under the Hawaii longline limited entry permit. Additionally, in the absence of the swordfish supply from the Hawaii shallow-set fishery, it is possible that fish vendors could increase imports of foreign-caught swordfish to fill the market gap in meeting the demand for swordfish in the U.S. (see Chan and Pan 2016; Rausser et al. 2009). NMFS analyzed whether the transferred effect should be treated as an indirect effect of the fishery in the BiOp, and concluded the evidence available does not indicate that the continued operation of the shallow-set fishery is reasonably certain to cause a change in the number of sea turtles captured and killed in foreign fisheries. As a result, we do not treat the number of sea turtles captured and killed in foreign longline fleets as an “indirect effect” of the proposed action. Instead, the BiOp evaluates the effects of other fisheries, including foreign fisheries, in the action area, on threatened and endangered species in the environmental baseline section of the BiOp. Specifically, foreign fisheries that occur in the action area are treated as “other human activities in the action area” that may affect the status of listed species in that action area. At a larger scale, the BiOp evaluated the positive and negative past, present, and future effects of those fisheries in the status of listed resources section to the extent information was available.

Comment 13: Several commenters oppose the Council's recommendation to remove the loggerhead hard cap.

Response: The ESA does not require NMFS to establish hard caps to manage commercial fishery impacts to protected species. The hard caps were first implemented in 2004 as a measure to control sea turtle interactions on the model shallow-set longline fishery while information was being gathered on the effectiveness of using circle hooks and mackerel-type bait in the Hawaii fishery. At that time, the best available scientific information indicated that the North Pacific loggerhead turtle population was projected to decline (WPFMC 2004). The current best available scientific information indicates that the North Pacific loggerhead population is increasing at an average rate of 2.3 percent, and the total population is estimated at approximately 340,000 turtles (Martin et al. 2020).

The Council and NMFS examined the potential long term effects of removing the hard cap as detailed in the EA. In the absence of a hard cap, the shallow-set fishery is expected to have a long-term average of 15.6 loggerhead turtle interactions per year and a low probability (less than 5 percent) of exceeding the ITS of 36 interactions in any given year, based on the predicted distribution of the anticipated level of loggerhead turtle interactions in the shallow-set fishery (McCracken 2018). The probability of exceeding the ITS of 36 is based on the upper range of the predicted distribution that estimated the fishery to have equal to or less than 36 interactions in any given year at the 95th percentile value. The predictions assumed that the fishery operated throughout the year for every year included in the analysis and did not truncate the predicted takes, thus providing a reasonable prediction of future level of interactions in the absence of a hard cap limit.

Under this final rule, if the fishery exceeds the loggerhead ITS of 36 in the BiOp, NMFS would reinitiate consultation pursuant to ESA Section 7. While the ESA requires reinitiation of Section 7 consultation when an ITS is exceeded, it does not necessarily require hard caps or other mechanisms to close the fishery. In this regard, hard caps are only required if NMFS determines such measures are necessary or appropriate to mitigate the amount or extent of take. In the BiOp, NMFS determined that a leatherback hard cap was necessary and appropriate to minimize impacts of incidental take and required that a fleet-wide limit of 16 to be implemented under terms and conditions in the BiOp, but did not find that a hard cap limit or other mechanisms for closing the fishery for loggerhead turtle interactions was either necessary or appropriate. However, the loggerhead hard cap would continue to be available as a management tool under the Pelagic FEP through future Council or NMFS action if necessary to conserve the species.

Also under this final rule, vessels would still be constrained by the individual trip limit of five loggerheads as well as additional restrictions if the trip limit were reached twice in a calendar year. The individual trip limit of five loggerhead turtle interactions per trip would be expected to provide additional reductions and prevent the fishery from approaching or reaching the ITS of 36, especially in years with higher number of interactions are expected, although the extent of reduction expected from the trip limits is uncertain due to the lack of operational data.

Changes From the Proposed Rule

This final rule contains no changes from the proposed rule.

Classification

The Administrator, Pacific Islands Region, NMFS, determined that Amendment 10 is necessary for the conservation and management of the Hawaii shallow-set longline fishery and that it is consistent with the Magnuson-Stevens Fishery Conservation and Management Act and other applicable laws.

The Chief Counsel for Regulation of the Department of Commerce certified to the Chief Counsel for Advocacy of the Small Business Administration during the proposed rule stage that this action would not have a significant economic impact on a substantial number of small entities. The factual basis for the certification was published in the proposed rule and is not repeated here. NMFS did not receive any comments regarding this certification. As a result, a regulatory flexibility analysis was not required, and none was prepared.

There is good cause under 5 U.S.C. 553(d)(3) to waive the 30-day delay in effectiveness, otherwise required by the Administrative Procedure Act, because this rule would remove the current loggerhead annual hard cap (17) that no longer conforms to the best available scientific information in the current BiOp for the fishery. As discussed above, the 2019 BiOp determined that given the current status of the loggerhead and the implementation of the vessel trip limits, an annual hard cap for the species was no longer necessary or appropriate. As of September 8, 2020, the fishery has interacted with 13 loggerheads in 2020, and therefore is at imminent risk of exceeding the current loggerhead hard cap. Failure to implement this rule immediately would likely result in the current loggerhead hard cap of 17 being exceeded prior to peak swordfish season in October, triggering an unnecessary and disruptive fishery closure that is not supported by the BSIA. Accordingly, waiving the 30-day cooling off period is necessary to bring the current regulations into compliance with the biological opinion.

This final rule implements the reasonable and prudent measures, and terms and conditions of the BiOp NMFS completed for the fishery. The Council took final action to implement these terms and conditions in August of 2019, following the release of the final BiOp in June of 2019. Subsequently, on January 23, 2020, NMFS published an NOA for this action, including an EA, and request for public comments which ended March 23, 2020. On February 4, 2020, NMFS published a proposed rule, and that comment period ended on March 20, 2020.

Reasonable and prudent measures are actions that are necessary or appropriate to minimize the impacts, i.e., amount or extent, of incidental take of loggerhead and leatherback sea turtles in the Hawaii shallow-set longline fishery. The associated terms and conditions set out the specific methods by which the reasonable and prudent measures are to be accomplished. Together, these measures must be implemented by NMFS for the take exemption in ESA section 7(o)(2) to apply to the Hawaii shallow-set longline fishery.

Since 2005, NMFS has required an annual hard cap for the fishery as a measure to control sea turtle interactions on the model shallow-set longline fishery while NMFS gathered information on the effectiveness of using circle hooks and mackerel-type bait in reducing sea turtle interactions in the fishery. The current loggerhead limit is 17. However, in light of the current abundance and increasing trend of the population, the individual vessel trip limit, and the accountability measure for vessels that might reach a trip limit twice in a calendar year, NMFS has determined that a hard cap is not necessary at this time for the conservation of the North Pacific loggerhead turtle and removing the limit would help ensure a continued supply of fresh domestic swordfish to U.S. markets. While this rule would not require an annual loggerhead hard cap, this measure would continue to be available to NMFS and the Council as a management tool under the FEP if necessary, to conserve the species.

Furthermore, this rule also reduces the leatherback hard cap limit from 26 to 16, which represents an approximate 25 percent reduction from the ITS, to minimize the impacts, i.e., amount or extent, of incidental take. This term and condition for Reasonable and Prudent Measure 1 in the 2019 BiOp must be immediately undertaken by NMFS for the take exemption in ESA section 7(o)(2) to apply.

This final rule has been determined to be not significant for purposes of Executive Order 12866.

This final rule is not an Executive Order 13771 regulatory action because this rule is not significant under Executive Order 12866.

NMFS initiated formal ESA section 7 consultation for the continued authorization of the fishery on April 20, 2018. In a BiOp dated June 26, 2019, the Regional Administrator determined that fishing activities conducted under FEP and its implementing regulations are not likely to jeopardize the continued existence of any endangered or threatened species.

List of Subjects in 50 CFR Part 665

• Hawaii
• Leatherback sea turtle
• Pelagic longline fishing
• North Pacific loggerhead sea turtle

Dated: September 9, 2020.

Samuel D. Rauch III,

Deputy Assistant Administrator for Regulatory Programs, National Marine Fisheries Service.

For the reasons set out in the preamble, NMFS amends 50 CFR part 665 as follows:

PART 665—FISHERIES IN THE WESTERN PACIFIC

1. The authority citation for 50 CFR part 665 continues to read as follows:

Authority: 16 U.S.C. 1801 et seq.

2. In § 665.802 revise paragraphs (ss) and (tt) to read as follows:

§ 665.802

Prohibitions.

* * * * *

(ss) Engage in shallow-setting from a vessel registered for use under a Hawaii longline limited access permit after the shallow-set longline fishery has been closed, or upon notice that that the vessel is restricted from fishing, in violation of § 665.813(b) and (i).

(tt) Fail to immediately retrieve longline fishing gear upon notice that the shallow-set longline fishery has been closed, or upon notice that that the vessel is restricted from fishing, in violation of § 665.813(b).

* * * * *

3. In § 665.813 revise paragraphs (b) and (i) to read as follows:

§ 665.813

Western Pacific longline fishing restrictions.

* * * * *

(b) Limits on sea turtle interactions in the shallow-set longline fishery— (1) Fleet Limits. There are limits on the maximum number of allowable physical interactions that occur each year between leatherback sea turtles and vessels registered for use under Hawaii longline limited access permits while engaged in shallow-set fishing.

(i) The annual fleet limit for leatherback sea turtles (Dermochelys coriacea) is 16.

(ii) Upon determination by the Regional Administrator that the shallow-set fleet has reached the limit during a given calendar year, the Regional Administrator will, as soon as practicable, file for publication at the Office of the Federal Register a notification that the fleet reached the limit, and that shallow-set fishing north of the Equator will be prohibited beginning at a specified date until the end of the calendar year in which the limit was reached.

(2) Trip limits. There are limits on the maximum number of allowable physical interactions that occur during a single fishing trip between leatherback and North Pacific loggerhead sea turtles and individual vessels registered for use under Hawaii longline limited access permits while engaged in shallow-set fishing. For purposes of this section, a shallow-set fishing trip commences when a vessel departs port, and ends when the vessel returns to port, regardless of whether fish are landed. For purposes of this section, a calendar year is the year in which a vessel reaches a trip limit.

(i) The trip limit for leatherback sea turtles is 2, and the trip limit for North Pacific loggerhead sea turtles (Caretta caretta) is 5.

(ii) Upon determination by the Regional Administrator that a vessel has reached either sea turtle limit during a single fishing trip, the Regional Administrator will notify the permit holder and the vessel operator that the vessel has reached a trip limit, and that the vessel is required to immediately retrieve all fishing gear and stop fishing.

(iii) Upon notification, the vessel operator shall immediately retrieve all fishing gear, stop fishing, and return to port.

(iv) A vessel that reaches a trip limit for either turtle species during a calendar year shall be prohibited from engaging in shallow-set fishing during the 5 days immediately following the vessel's return to port.

(v) A vessel that reaches a trip limit a second time during a calendar year, for the same turtle species as the first instance, shall be prohibited from engaging in shallow-set fishing for the remainder of that calendar year. Additionally, in the subsequent calendar year, that vessel shall be limited to an annual interaction limit for that species, either 2 leatherback or 5 North Pacific loggerhead sea turtles. If that subsequent annual interaction limit is reached, that vessel shall be prohibited from engaging in shallow-set fishing for the remainder of that calendar year.

(vi) Upon determination by the Regional Administrator that a vessel has reached an annual interaction limit, the Regional Administrator will notify the permit holder and the vessel operator that the vessel has reached the limit, and that the vessel is required to immediately stop fishing and return to port.

(vii) Upon notification, the vessel operator shall immediately retrieve all fishing gear, stop fishing, and return to port.

* * * * *

(i) A vessel registered for use under a Hawaii longline limited access permit may not be used to engage in shallow-setting north of the Equator any time during which shallow-set fishing is prohibited pursuant to paragraphs (b)(1) or (2) of this section.

* * * * *