Oceana, Inc. v. Ross

363 F. Supp. 3d 67

• Court: Court of Appeals for the D.C. Circuit
• Decided: March 11, 2019
• Docket: Case No. 17-cv-829 (CRC)
• Status: Published

Opinion

CHRISTOPHER R. COOPER, United States District Judge

"Like the grizzly bear on land, the dusky shark is seated at the top of the food chain and helps to maintain balance in the ecosystem by eliminating weak and sick individuals, providing scavenging species with food, and regulating the diversity, distribution, and behavior of prey species." So begins plaintiff Oceana, Inc.'s description of the formidable marine species at the center of this case. But the predator dusky shark, Oceana says, has become prey, thanks to a deadly combination of rampant overfishing and regulatory neglect. That one of the world's most fearsome species can be rendered among its most vulnerable in the space of a few decades provides rich context for the administrative law dispute this case presents.

As for that dispute, Oceana demands that federal regulators do more to stem the dusky shark's decline. Specifically, it claims that the National Marine Fisheries Service's most recent effort to protect the dusky shark violated the Magnuson-Stevens Act, the National Environmental Policy Act, and the Administrative Procedure Act by: (1) failing to establish management measures to constrain the number of dusky sharks accidentally caught as "bycatch"; (2) ignoring available evidence about the prevalence of bycatch, leading to an underestimation of the overfishing problem and inadequate corrective measures; and (3) failing to take a hard look at a reasonable range of alternatives for achieving the agency's chosen goal for reducing dusky shark mortality. Both sides have moved for summary judgment. After reviewing the parties' submissions and the administrative record on which they are based, the Court finds in favor of Oceana on the first two issues and will therefore order the agency to reconsider its proposed course of action. Because a remand is proper for the first two reasons, the Court need not reach the third.

I. Background

A. Legal Framework

A primer on the two environmental statutes on which Oceana's claims are based provides necessary context for understanding the facts at issue.

*711. The Magnuson-Stevens Act

The Magnuson-Stevens Act ("MSA"), 16 U.S.C. §§ 1801 et seq. is designed in large part to prevent overfishing in U.S. coastal waters and mitigate and reverse its effects where it has already begun. To that end, the MSA empowers federal agencies to "provide for the preparation and implementation, in accordance with national standards, of fishery management plans which will achieve and maintain, on a continuing basis, the optimum yield from each fishery." Id. § 1801(b)(4). A "fishery" is "one or more stocks of fish which can be treated as a unit for purposes of conservation and management and which are identified on the basis of geographical, scientific, technical, recreational, and economic characteristics" and "any fishing for such stocks." Id. § 1802(13). "Optimum yield," generally defined, "means the amount of fish which will provide the greatest overall benefit to the Nation, particularly with respect to food production and recreational opportunities, and taking into account the protection of marine ecosystems." Id. § 1802(33)(A).

The National Marine Fisheries Service ("Fisheries Service" or "agency"), through authority delegated by the Secretary of Commerce, is responsible for enforcing fisheries' compliance with the fishery management plans ("FMP") established under the MSA. See generally C & W Fish Co. v. Fox, 931 F.2d 1556 (D.C. Cir. 1991). Though the MSA establishes regional fishery management councils to develop FMPs for their respective regions, the Fisheries Service itself handles FMPs for highly migratory species ("HMS")-species of tuna, marlin, oceanic sharks, sailfish, and swordfish-that traverse multiple regions. 16 U.S.C. §§ 1852(a)(3), 1854(c). The regional councils and the Fisheries Service are required to create an FMP, or amend an existing one, when the Secretary of Commerce determines that a fishery is "overfished." Id. § 1854(e)(2). A 2006 amendment to the MSA further requires all FMPs to "establish a mechanism for specifying annual catch limits ... at a level such that overfishing does not occur in the fishery, including measures to ensure accountability." Id. § 1853(a)(15).¹

FMPs, and their implementing regulations, are subject to ten "National Standards," id. § 1851(a)(1)-(10), and other MSA requirements, see id. §§ 1853(a), 1854(e). Among those relevant here, National Standard 1 requires FMPs to "prevent overfishing while achieving, on a continuing basis, the optimum yield from each fishery for the United States fishing industry." Id. § 1851(a)(1). National Standard 2, moreover, requires that FMPs "be based upon the best scientific information available." Id. § 1851(a)(2).

The Fisheries Service, pursuant to another MSA command, id. § 1851(b), provides its own gloss on the statute's mandatory National Standards through a set of guidelines, codified at 50 C.F.R. §§ 600.305 - 600.355. The guidelines do "not have the force and effect of law," but the *72various regional councils and Fisheries Service personnel are instructed to use them "to assist in the development of fishery management plans." 16 U.S.C. § 1851(b). Most relevant to this suit, the guidelines clarify how to develop and implement annual catch limits ("ACLs") and accountability measures ("AMs"). See, e.g., id. § 600.310(g)(3).

2. The National Environmental Policy Act

The National Environmental Policy Act ("NEPA"), 42 U.S.C. §§ 4321 - 4370h, ensures that agency decisionmakers and the public at large are apprised of the environmental impact of proposed federal action. Though NEPA does not impose substantive environmental obligations on federal agencies, Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 349, 109 S.Ct. 1835, 104 L.Ed.2d 351 (1989), it does force them to "take a hard look at environmental consequences" and "provide for broad dissemination of relevant environmental information," id. at 350, 109 S.Ct. 1835 (internal quotation marks omitted). That is, "NEPA itself does not mandate particular results, but simply prescribes the necessary process." Id.

One component of that process is the requirement that an agency prepare an environmental impact statement ("EIS") any time it proposes a "major Federal action[ ] significantly affecting the quality of the human environment." 42 U.S.C. § 4332(C). The "heart" of the EIS is its presentation of the "environmental impacts of the proposal and the alternatives in comparative form," which "sharply defin[es] the issues and provid[es] a clear basis for choice among options by the decisionmaker and the public." 40 C.F.R. § 1502.14.

B. Factual Background

1. Dusky Sharks, Overfishing, and Government Intervention

The dusky shark is a coastal-pelagic fish that inhabits temperate and tropical waters. Administrative Record ("AR") at 009233-34.² The species is extremely mobile; in the Atlantic Ocean, for instance, the dusky shark travels from Nova Scotia to Cuba and from Nicaragua to southern Brazil. AR009236. Accordingly, the dusky shark is classed as a "highly migratory species" under the MSA. See 16 U.S.C. §§ 1852(a)(3), 1854(c).

For rather obvious reasons-it grows to an average length of twelve feet and a weight of 400 pounds-the dusky shark functions as an apex predator. AR009087; AR009163; AR009055. Yet for more subtle ones-it is slow-growing, long-lived, slow to sexually mature, and produces few offspring-the dusky shark is particularly vulnerable to predation of a different sort: human overfishing. AR008960; AR007155. Given its slow maturation and reproductive capacity, even relatively light fishing can cause outsized population reduction, AR009209, and once depleted, the population is slow to recover, AR009217; AR000315.

Predictably, the spread of commercial and recreational shark fisheries in the final quarter of the 20th century proved disastrous for dusky sharks. See AR4376 (noting development of a directed shark fishing fleet and recreational shark fishery in the 1970s); AR004399 (describing dusky *73shark mortality as "low from 1960 through the 1980s" before it "ramped up to unsustainably high levels in the 1990s"). The Fisheries Service designated dusky sharks an Endangered Species Act "species of concern" in 1997, AR009094; AR007156, and then a "prohibited species" in 2000, making it illegal for fishermen to possess, sell, take, or retain them, AR007053; AR007655-56; see 50 C.F.R. § 635.24(a)(5). The 1999 Highly Migratory Species Fishery Management Plan ("HMS FMP") hoped that, by "prohibiting possession of dusky sharks, ... fishermen [would] adjust their fishing activities accordingly" and the dusky shark population would rebound. AR007655.

The rebound would not occur overnight. When the agency in 2006 conducted its first individual stock assessment³ of the northwest Atlantic dusky shark population, it concluded that the population was "heavily exploited" and noted several "declining trend[s]." AR001444; AR008050-51; AR008059. The assessment determined that the population was experiencing overfishing and was overfished. A major culprit was bycatch: fishermen didn't want dusky sharks, but they were catching them anyway while targeting other fish. See AR004878; AR004403; see also 16 U.S.C. § 1802(2) (defining bycatch as "fish which are harvested in a fishery, but which are not sold or kept for personal use, andincludes *74economic discard and regulatory discards"). In the Gulf of Mexico Reef Fish Fishery, for example, fishermen target snapper and grouper, but their longline gear-basically, a horizontal line up to 40 miles long with baited hooks at regular intervals-snared an estimated 800 dusky sharks in 2005, according to one data source. AR008426. Though by law dusky shark bycatch had to be quickly released, the damage in many cases was already done. AR007494 ("[T]here is evidence that dusky sharks experience high at-vessel and post-release mortality rates[.]"). The bottom-line figures were startling: the Fisheries Service estimated that the biomass of the dusky shark population had dropped by as much as 80 percent since 1960, AR008050; AR008094-96, and that it could take 100 to 400 years for the population to recover, AR007156.

The Fisheries Service in 2008 responded to this bleak report, with Amendment 2 to the 2006 HMS FMP. Amendment 2 proposed a rebuilding plan for the dusky shark stock, with a particular emphasis on the bycatch problem. See AR007887 ("Many of the final actions in this rule ... should reduce dusky shark bycatch."); AR004811 ("This rebuilding plan ... focus[es] primarily on bycatch of the species[.]"). It cut the number of non-prohibited large coastal sharks that fishing vessels could retain, so as to reduce fishing effort targeting sharks that might incidentally ensnare dusky sharks; it implemented time and area closures, which temporarily halt all fishing, or at least fishing with certain types of gear, in specified ocean sectors;⁴ and it largely banned fishing for sandbar sharks, a target species that led to high levels of dusky shark bycatch. See AR007935-41; AR007087; AR004811. Later, Amendment 3 to the 2006 HMS FMP had the effect of setting the dusky shark annual catch limit at zero. AR007570.

Even with these reforms, the 2011 Fisheries Service stock assessment returned persistently grim results, see AR000070 (Table 4), including its finding that the biomass of dusky sharks capable of reproducing was less than one-fifth of 1960 levels, id. (2009 row, far-right column); AR000066. The assessment concluded that overfishing continued and that dusky sharks remained overfished. AR000427; AR007565.

In 2012, the Fisheries Service released Draft Amendment 5 to the 2006 HMS FMP, which proposed the following changes: lowering commercial quotas; re-defining species groups, which is how species are categorized for management purposes; creating new or changing existing time and area closures; increasing the minimum size of sharks that recreational fishermen could keep, to reduce mortality of sexually immature sharks that had not yet had a chance to reproduce; and establishing recreational reporting requirements for certain species of sharks. See AR001278-1309. After hearing extensive feedback from the HMS Advisory Panel-consisting of interested parties in the fishing industry, environmental community, academia, and non-governmental organizations-and from the public at large, the Fisheries Service split Draft Amendment 5 into separate rulemakings: one specific to dusky sharks (Amendment 5b), the other for an amalgam of species (Amendment 5a). AR003355-56.

By October 2015, the Fisheries Service had failed to take any further action on Amendment 5b, and Oceana-a non-profit *75organization dedicated to protecting and restoring the world's oceans-filed suit to force the agency's hand. Oceana v. Pritzker, No. 1:15-cv-01824-CRC, 2015 WL 6530525 (D.D.C. filed Oct. 27, 2015). Pursuant to a settlement of that suit, the Fisheries Service in October 2016 published the proposed draft of Amendment 5b and the related EIS. AR004828; AR004825. Immediately before issuing the draft amendment, however, the Fisheries Service completed another stock assessment. This one showed some improvement but still bore warning signs. See AR004402-03. On the one hand, the assessment indicated that dusky shark overfishing had declined, AR007079; AR007101, estimated that the species' mortality need only be reduced 12 percent from 2015 levels to end overfishing, AR007101, and projected that a 35 percent reduction from 2015-mortality levels would give the population a 50 percent chance at returning to sustainable levels by 2107 (not 2017), AR007104. On the other hand, the assessment yet again concluded that dusky sharks remained overfished with overfishing occurring. AR007101; AR004804.

The Final Rule implementing Amendment 5b ("Amendment 5b"), promulgated in April 2017 after a public comment period, addressed these latest findings. It sought to end to dusky shark overfishing by implementing a modified rebuilding plan for the population, including the establishment of new accountability measures. AR007111. Amendment 5b requires recreational and commercial fishermen to undergo education on dusky shark identification to reduce bycatch retention rates by fishermen who misidentify and fail to release them; it establishes a release protocol in the pelagic longline⁵ fishery; it mandates training on how to safely handle dusky sharks to reduce mortality due to bycatch; it orders fishermen in the recreational Atlantic shark fisheries and the directed shark bottom longline fishery to use circle hooks, thought to be less harmful and thus less likely to cause death after discard; and it establishes a fleet communication protocol to help fishermen avoid areas where dusky sharks are likely to be ensnared. AR007565. As for hard targets, the Fisheries Service says that Amendment 5b, based on the 2016 stock assessment, "aims to achieve a 35 percent mortality reduction relative to 2015 levels, and rebuild dusky shark stock by the year 2107." Id.

C. Procedural Background

Oceana filed suit one month later, in May 2017, alleging that Amendment 5b violates the MSA and APA and that the EIS accompanying Amendment 5b violates NEPA and the APA. It named as defendants Secretary of Commerce Wilbur Ross, the National Oceanic and Atmospheric Administration ("NOAA"), and the Fisheries Service. Oceana brought five counts in its complaint, but distills from these "three fundamental errors": first, the Fisheries Service failed to "place a definite, enforceable limit on the number of dusky sharks caught and killed as bycatch" despite bycatch being the primary cause of dusky shark overfishing; second, the Fisheries Service erroneously calculated bycatch data, leading it to underestimate the severity of the overfishing problem; and third, the Fisheries Service offered "no scientific analysis or evidence" that Amendment 5b's accountability measures would achieve the mortality-reduction and population-rebuilding goals the amendment *76claimed to be pursuing. Pl's Memorandum of Points and Authorities in Support of Summary Judgment ("Pl's MSJ"), ECF No. 36-1, at 16-17. Both parties have since moved for summary judgment. The Court held a hearing on December 11, 2018, and the matter is now ripe for the Court's resolution.

II. Legal Standard

"Although styled Motions for Summary Judgment, the pleadings in this case more accurately seek the Court's review of an administrative decision." Oceana, Inc. v. Locke, 831 F.Supp.2d 95, 106 (D.D.C. 2011). The APA governs judicial review of a final agency action under the MSA and NEPA. 5 U.S.C. §§ 701 - 06 ; 16 U.S.C. § 1855(f)(1)(B) ; Theodore Roosevelt Conservation P'ship v. Salazar, 661 F.3d 66, 72 (D.C. Cir. 2011). Under the APA, courts "hold unlawful and set aside" agency actions that are "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A). "This arbitrary and capricious standard of review is a highly deferential one, which presumes the agency's action to be valid." Envtl. Def. Fund, Inc. v. Costle, 657 F.2d 275, 283 (D.C. Cir. 1981) (quotation and citation omitted). In applying this "narrow" review, a court may not "substitute its judgment for that of the agency," but instead should ask only "whether the [agency's] decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment." Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971).

III. Analysis

The Court begins its analysis with Oceana's argument that the Fisheries Service must consider other bycatch evidence-in particular, bycatch data recorded in logbooks maintained by fishing vessels-in crafting Amendment 5b. The Court concludes that Oceana is right, and that conclusion informs its analysis on the remaining issues. For if the agency ignored contrary evidence in violation of the APA and failed to base its management decisions on the best scientific evidence in violation of the MSA, it has to go back and incorporate that evidence-or better explain why it chose not to-before the Court can meaningfully assess whether the accountability measures in Amendment 5b are sufficient and whether the agency unreasonably failed to consider alternative accountability measures.

A. Evidentiary Basis for Agency Action

An agency action is arbitrary and capricious under the APA "if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, [or] offered an explanation for its decision that runs counter to the evidence before the agency[.]" Motor Vehicle Mfrs. Ass'n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983). National Standard 2 of the MSA imposes its own process requirement on the Fisheries Service: its conservation and management measures must be based on the "best scientific information available." 16 U.S.C. § 1851(a)(2). While this means the Fisheries Service may not "disregard superior data in reaching its conclusion," a National Standard 2 challenge "will normally fail, unless there is some indication that superior or contrary data was available and that the agency ignored such information." Guindon v. Pritzker, 31 F.Supp.3d 169, 195 (D.D.C. 2014) (internal quotation marks omitted).

Oceana contends that the Fisheries Service ran afoul of these process requirements in two separate, but related, ways. Pl's MSJ at 22; Pl's Combined Opposition and Reply ("Pl's Opp."), ECF No. 40, at 9.

*77First, the Fisheries Service relied on incomplete and inaccurate bycatch data, leading it to underestimate the severity of the bycatch, and thus the overfishing, problem. Pl's Opp. at 8-9 ("In adopting Amendment 5b, the Service ignored key evidence necessary to evaluate the dusky shark bycatch numbers both within the HMS Fishery and outside the HMS fishery."). Second, as a result of the first mistake, the agency failed to explain how Amendment 5b would effectively redress the bycatch problem and stop overfishing. See Pl's Opp. at 9 ("Without a full picture of the magnitude of this entire bycatch problem, the Service could not possible determine that its measures will reduce bycatch mortality by the amount necessary" to rebuild the population.). The Court agrees with Oceana on the first point, which obviates the need for it to reach the second.

Oceana's basic complaint is that the Fisheries Service relied on an unreasonably narrow set of data in crafting Amendment 5b. It claims the agency made two key mistakes: first, it ignored evidence from fishermen's self-reported logbooks that suggested significant bycatch was occurring in non-HMS fisheries, leading to the erroneous conclusion that Amendment 5b's new accountability measures need only address activity in the HMS fishery, Pl's MSJ at 23; and second, it irrationally relied on bycatch data calculated by independent observers on selected vessels to the exclusion of fishermen logbook data in assessing the bycatch problem within the HMS Fishery, id. at 30. These ostensibly discrete blunders are in fact bound up with one another: if it was arbitrary and capricious for the agency to exclude logbook data from its assessment of the bycatch problem, that problem pervades the agency's analysis, whether it pertains to the HMS fishery or otherwise. The Court therefore starts and stops its analysis with Oceana's logbook data argument, ultimately concluding that the agency's decision to ignore that data was an unreasonable one, or at least one the agency did not rationally explain.⁶

Before going any further, the Court should explain the differences between observer and logbook data. Observer data-what the Fisheries Service primarily used-is collected by independent observers who are employed by the Fisheries Service and stationed on a select group of fishing vessels. See, e.g., AR008294. Observers report the number and location of bycatch and whether the catch was released dead or alive. See Defs' Memorandum of Points and Authorities in Support of Summary Judgment ("Defs' MSJ"), ECF No. 38, at 19 (citing AR007087-89) (Observer data is "simply the actual amount of dusky shark bycatch and mortalities occurring on vessels selected for observer coverage."). Observers are present on a small number of vessels, with coverage hovering around five to fifteen percent depending on the fishery. See, e.g., AR008397-98 (estimating five percent coverage in Gulf of Mexico Reef Fish Fishery); AR007185 (reporting "target coverage level of 5-10 percent" in HMS Fishery, not including the five to ten vessels selected for the shark research fishery); AR009546 (indicating HMS Fishery observer coverage of four to six percent, outside the small number of vessels selected for the research fishery). Logbook data-what Oceana says the Fisheries Service *78should have used in conjunction with observer data-consists of the same information but is reported by fishermen themselves rather than independent observers. AR007165. Logbook data is far more comprehensive than observer data: all fishermen reporting through the HMS Logbook Program were required to track bycatch in their reports, while 20 percent of those reporting through the Coastal Fisheries Logbook Program had to do the same. AR007260-61.

There is reason to think that the difference in source makes a difference in substance. For example, the spotty observer reports from the pelagic longline vessels in the HMS Fishery showed that an average of 32 dusky sharks were caught as bycatch each year from 2008 to 2015. AR007093 (Table 1.5, Sixth Row "Atlantic Pelagic Observer Program"). The logbook information painted a starkly different picture, showing that those same vessels caught an average of some 550 dusky sharks per year from 2008 to 2014. AR007170 (Table 3.10). Outside the HMS Fishery, a Fisheries Service report entitled Update 1 to the First Edition of the National Bycatch Report ("First Update")-ostensibly based on self-reported logbook data, though the agency disputes this-estimated that an average of almost 3,800 dusky sharks had been caught as bycatch in non-HMS fisheries in the South Atlantic and Gulf of Mexico from 2006 to 2010. AR007095, Table 1.6; AR008779.⁷ The observer data, by contrast, yielded far smaller figures: non-HMS dusky shark catches numbered in the single digits. AR007093-94.⁸ In the end, the Fisheries Service chose to rely almost exclusively on observer data for its assessment of dusky shark bycatch in both HMS and non-HMS fisheries.

Before delving into the agency's explanation for that decision, the Court will pause to say a bit more about the First Update, which functions as a sort of proxy battleground for the parties' observer versus logbook data debate. Oceana says the First Update data undermines the agency's conclusion that bycatch in non-HMS fisheries is rare and that corrective efforts should thus focus on the HMS Fishery. The agency, however, dismissed the numbers reflected in the First Update as "inflated," AR007095, and too "technically flawed" to provide a valid basis for management measures, AR007177. The agency rejected the data as unreliable for two principal reasons: they (1) wrongly extrapolated the dusky shark catch rate in the shark bottom longline fishery observer program and applied it to vessels in other fisheries, where (the agency says) there is reason to think dusky shark interactions were less common; and (2) risked double counting dusky sharks, since some HMS shark vessels also target reef fish, meaning a dusky shark caught by such a vessel might be recorded as bycatch in data for two fisheries. AR007095-96.

*79The Fisheries Service's first reason highlights a factual dispute preliminary to the observer versus logbook debate. The agency claims the First Update's data on dusky shark bycatch in the non-HMS fisheries was actually based on extrapolations from observer data, not logbook data. Id.; Defs' MSJ at 21. Oceana calls this sheer invention, citing to an appendix to the First Update, which identifies logbook data as the source for bycatch data for these fisheries. AR008777 (link to "Appendix 3"). If the agency is right, that could suggest that Oceana's issue with respect to the First Update data is with the agency's refusal to extrapolate observer data, not with its rejection of logbook data.

Yet the parties spend far too much time and energy on this skirmish. The agency in particular seems to believe that the Court's resolution of these issues-whether the First Update used logbook or observer data, and whether the agency erroneously dismissed the First Update-will dictate the outcome on the underlying question of whether the agency failed to consider important information, especially logbook data, in adopting Amendment 5b. See Defs' MSJ (spending almost all of data argument on the First Update). For what it's worth, the Court is inclined to agree with Oceana regarding the First Update; the record offers more support for its contention that the First Update drew on logbook, not observer, reports.⁹

But even if the Court reached a different conclusion-and took the agency at its word that the First Update bycatch estimates were based on observer data-that would still tee up the question of why the agency dismissed logbook data as a general matter.¹⁰ Indeed, in a footnote elaborating on its contention that the First Update used observer data, the Fisheries Service explained that "[o]nly very limited information about dusky sharks could be reported in [non-HMS] reef fish or snapper-grouper logbooks ." Defs' MSJ at 21 n.16 (emphasis added). In other words, the agency's support for its claim that the First Update bycatch estimates were extrapolated from observer data is an explanation for why alternative logbook data were inadequate. That means that all roads lead to the question of whether it was reasonable for the Fisheries Service to discard logbook data in making its bycatch assessment. For if the agency, as Oceana contends, rejected the First Update's bycatch estimates because they relied on faulty logbook data, the Court would ask whether the logbook data was so defective to render the First Update's use of them disqualifying. Alternatively, suppose that the Fisheries Service is correct, and that the First Update bycatch estimates for the non-HMS fisheries were premised exclusively on observer data. If that were the case, the Court then must ask why the agency chose to ignore alternative *80logbook data, even after it realized the perils in extrapolating from observed catches. Subtle framing differences aside, the fundamental question is the same: has the agency adequately explained its refusal to consider logbook data, both within and outside the HMS Fishery?

The Court concludes it has not. The Fisheries Service offers three sets of reasons for not using logbook data¹¹ -the first explaining its aversion to making any estimates from catch data (logbook, observer, or otherwise), the second attacking the validity of the Coastal Fisheries logbook data specifically, and the third highlighting the methodological problems that it says inhere in logbook reporting generally. The Court takes these in turn.

The first argument warrants the least airtime. The agency emphasizes at the outset that the "high uncertainty" plaguing catch data compelled its decision to use a " 'catch-free' model to determine the status of the stock." Defs' MSJ at 21. According to the agency, this meant it "would not estimate or extrapolate the total amount of dusky shark bycatch occurring using reported catches and would not consider any estimates or extrapolations it received as valid for management purposes." Id.; see Def's Opp. at 13 (stating, as first reason for not using logbook information for the HMS Fishery, that the agency chose not to "estimate dusky shark bycatch"). Setting to one side the question whether the agency justifiably refused to make any estimates or extrapolations from catch data, this point fails to explain the agency's decision to use one subset of catch data (that recorded by independent observers) to the exclusion of the other (that recorded by the fishermen). Its reference to using a "catch-free" model for stock assessments aside-which, again, strikes the Court as a bit of a misnomer, see supra at 73 n.3-the agency admits that it "primarily used observer data" to assess the bycatch problem. Defs' MSJ at 19. And because the decision to use one type of catch data is separate from what to do with that data-including drawing estimates or extrapolations from it-the agency's first-order burden remains to explain why it used only one subset of catch data and refused to consider the other.

That leads to the agency's next two arguments, one addressing its concern about non-HMS Fishery logbook data, the other its distrust of logbook data as a general matter. For its non-HMS Fishery argument, the agency primarily attempts to poke holes in the Coastal Fisheries logbook data, which covers non-HMS fisheries like the South Atlantic snapper-grouper fishery and the Gulf of Mexico reef fishery. AR008057. Many of the Fisheries Service's reasons for disregarding this data have to do with the fact that it is collected differently than the HMS logbook data. AR007260 (describing "different types of data" reported in the two logbooks). For instance, the Fisheries Service complains that the Coastal Fisheries Logbook Program tracks bycatch for 20 percent of vessels, while the HMS Logbook Program records bycatch for 100 percent of vessels, making it difficult to draw meaningful comparisons between the two data sets. Defs' MSJ at 21 n.16 (citing AR007260-61). While the Court grants that these differences could make apples-to-apples *81comparisons with HMS logbook data more difficult, that the data is different from the HMS data does not explain why the Fisheries Service rejected it out of hand.

Most significantly, the agency's answer does not at all explain why logbook data from 20 percent of vessels would not augment the even more limited non-HMS observer data on which the agency eagerly relied. The agency apparently maintains no observer program at all in the Southeastern Atlantic Snapper-Grouper Fishery, even though the National Bycatch Report shows that dusky sharks are regularly caught as bycatch there. Defs' Answer ¶ 87. As for the Gulf of Mexico Reef Fishery, another non-HMS fishery where dusky shark interactions are common, the agency drew observer data from two programs-the Gulf of Mexico Reef Fish Observer Program, AR008397-98, and the HMS Shark Bottom Longline Observer Program, AR009546-47-but both data sets are sharply limited. Observer coverage in the Reef Fish Observer Program checked in at less than five percent. AR008398 (Table 4.2.2, third row). As for the Bottom Longline Observer Program, that data was drawn only from the small number of vessels (about 55 out of 800) with dual permits allowing them to target both HMS and non-HMS species, like snapper and grouper. Defs' MSJ at 23 n.18 (citing AR007087); AR009546-47. Further compromising the data, observers were present on only eight percent of the trips taken by those few vessels from 2009-2011, see AR009547, and observers did not even record data from reef fish trips-as opposed to those targeting sharks-from 2012 to 2015, see AR009631-50. The agency cannot rationally reject one data source largely because of its small sample size and yet embrace another source with an even smaller sample size.¹²

Even if the Court concluded that the incomplete logbook reporting in the non-HMS fisheries rendered the data useless-which it does not-the Fisheries Service would need wholly different reasons for rejecting logbook data from within the HMS Fishery, since it admits that "all fishermen reporting in the HMS Logbook must provide" bycatch information. AR007260. Indeed, the Fisheries Service's argument against the Coastal Fisheries logbook data would appear to undermine any objection to the HMS Fishery logbook data. If the agency's strongest critique of the Coastal Fisheries logbook data is that it is selective, it would seem the agency would embrace the HMS logbook data because it is so complete.

But instead, the Fisheries Service highlights reasons to distrust logbook data no matter how comprehensive it is. In particular, the agency fixates on the problem of misidentification. Because fishermen might simply misidentify catch as dusky shark in vessel logbooks, the agency has "low confidence in their accuracy." AR007177. Two things must be said in response, however. First, according to the agency's own words, misidentification plagues observer data as well, and yet that data was deemed reliable enough to assess bycatch. See, e.g., AR007177 ("There are issues with species misidentification and reporting in many of the available fishery-dependent data sources (e.g., observer, logbooks ....)"). And second, even if misidentification occurs, the record offers no reason tothink *82that necessarily skews the data in one direction or the other; there is also reason to believe such errors might balance one another out, as the agency itself admits. AR007177 ("[I]t is unknown whether the reported values could over- or under-estimate true catch."); AR009463 ("However, under-reporting is possible, which can lead to a negative bias in bycatch estimates.").

What is more, as Oceana points out and the Fisheries Service essentially admits, the observer data on which the agency did rely suffer from comparable infirmities-only they have the certain effect of inaccurately minimizing the bycatch problem rather than exaggerating it. While the logbook data may be compromised by the risk of misidentification, the observer data is undermined by its limited inputs, as the agency's own description of observer data makes clear. Defs' MSJ at 19 (stating that observer data was drawn only from those "vessels selected for observer coverage"); AR007259 ("Observer data also have constraints; they do not cover the entire fleet[.]"); AR009399 (stating "target coverage level of 5-10%" in HMS Fishery); AR008397-98 (pegging observer coverage in Gulf of Mexico reef fishery at about five percent) Defs' Answer ¶ 87 (admitting "there is no standardized, regular observer program" for South Atlantic snapper-grouper fishery).

Perhaps the Fisheries Service could have mitigated this weakness in observer data by extrapolating it, but the agency refused to do so. Defs' MSJ at 23 (asserting that "even [observer] data is insufficient to accurately estimate" dusky shark bycatch).¹³ Exactly why it believes the observer data cannot be extrapolated is unclear. The closest the agency comes to explaining its aversion to estimates and extrapolations is in its reply, when it asserts that any estimate "would lead to highly uncertain and scientifically invalid results." Defs' Reply in Support of Summary Judgment ("Defs' Reply"), ECF No. 43, at 13. But uncertainty hardly seems a reason to eschew estimates or extrapolations altogether. If it were, then there would also be no use in assessing the bycatch problem by reference to unextrapolated observer data; that, too, given the extremely sparse observer coverage, will inevitably yield a highly uncertain assessment of the amount of bycatch actually occurring. The agency's refusal to extrapolate observer data is all the more confounding given that its own scientists have encouraged extrapolations, including those done with the aid of logbook data. A draft Environmental Impact Statement for Amendment 5 (before the dusky shark-specific Amendment 5b emerged) lamented the "constraints of observer data" and explained that "fishery dependent data," i.e. , logbook data, "provides the most straight-forward approach for determining" dusky shark interactions.¹⁴ AR002181. The Southeast Fisheries Science Center pushed back on this claim and urged the agency to make use of observer data-so long as it extrapolated it using logbook information. Id. ("[U]nder this approach, observer data *83are used to estimate a rate of capture and that rate is extrapolated to the entire fishery ... using effort data from the logbooks. This approach is considered appropriate for sea turtles, which means it should also be appropriate for dusky shark.").

This, and other evidence in the administrative record, demonstrates that the rational response would have been to use one data source to complement, corroborate, and correct the findings of the other. The Fisheries Service, unsurprisingly, often does just that; indeed, it did so in Amendment 5b itself. When the agency evaluated the possibility of closing heavily fished "hotspots" to curb dusky shark bycatch, it primarily "used the HMS logbook data rather than observer data to calculate dusky shark interactions because logbook data are collected across all HMS-permitted participants in the pelagic longline fishery, thus alleviating the need to extrapolate interactions for the entire fishery based on observed trips."¹⁵ AR007259. The agency still made some use of the observer data, but only "to validate the use of the logbook data." AR007261. And since the "observer data tended to reflect the logbook data in terms of the locations with the greatest number of dusky shark interactions," the Fisheries Service could be more confident in its conclusions. Id. This part of the Environmental Impact Statement validates both the independent utility of logbook data and its role as a complement to observer data. It demonstrates that each methodology's weakness-incompleteness for observer data, inaccuracy for self-reported logbooks-is buttressed by the other's strength.

The Fisheries Service hardly bothers to explain its inconsistent treatment of logbook data in Amendment 5b. It admits it used logbook data "to the extent it was appropriate" in evaluating the effectiveness of the hotspot closures that Amendment 5b rejected, Defs' MSJ at 19, yet does not explain what rendered the use of logbook data "appropriate" in that context but inappropriate for the purpose of evaluating the dusky shark bycatch problem more generally. Nor does the agency adequately explain its sharp break from past practice, which has long been to use logbook data to monitor and evaluate bycatch. See, e.g., AR009463 (discussing historical use of "self-reported logbook data" to "produce bycatch estimates"); AR009464 (stating that the agency occasionally uses "self-reporting" logbooks as the "primary method of reporting bycatch").

Instead-to repeat the point yet again-the agency has at most shown that it would be justified in taking the logbook data with a grain of salt(water), not dismissing it outright. Even if it would have been unreasonable for the agency to use logbook data as the primary or sole source of bycatch information, the Court is left scratching its head why such data should not be used in conjunction with observer data-as the agency did in evaluating the viability of hotspot closures in Amendment 5b, and as it has often done in the past. AR009464 ("[L]ogbooks are used to provide effort information against which bycatch rates obtained from observers is multiplied to estimate bycatch."); AR008249 ("Logbook data may be used ... as supplemental data for extrapolating to the unobserved portion of the fishery.").Logbook *84data has always been helpful to the Fisheries Service; either the agency must consider such data in Amendment 5b, see State Farm, 463 U.S. at 43, 103 S.Ct. 2856 (agency action is arbitrary and capricious under the APA where it "fail[s] to consider an important aspect of the problem"); Pritzker, 31 F.Supp.3d at 195 (agencies must consider "superior or contrary data" under the MSA), or offer a much more cogent explanation for why its position has changed.¹⁶

In closing, the Court reiterates that it might have reached a different conclusion if the Fisheries Service had used an alternative method for estimating dusky shark bycatch or taken some other step to account for the undeniable gaps in the observer data. Instead, the agency assessed the extent of bycatch by reference to observer data that it admits is incomplete, declined to give any weight to the admittedly more comprehensive logbook data, and then failed to supply any other measurement device that might more accurately depict the extent of the bycatch problem. Had the agency chosen an alternative means of assessing the bycatch problem-say, by extrapolating its limited observer data-the Court might well defer to that decision so long as it was rationally explained. See, e.g., Flaherty v. Bryson, 850 F.Supp.2d 38, 62 (D.D.C. 2012) (deferring to Regional Council's choice to use three-year average to estimate catch). But, as Oceana says, there is nothing rational about dismissing the logbook data for its imperfections and accepting without reservation the observer data despite its manifest flaws. See Pl's MSJ at 28 ("The agency cannot simply assume that bycatch in these non-HMS fisheries is 'rare' by dismissing extrapolated estimates as overestimates, citing small raw numbers, failing to disclose very small observer coverage, and not looking at other data to fill in the gap.")

Because the Court concludes that the Fisheries Service's decision to exclude logbook data renders its assessment of the dusky shark bycatch problem arbitrary and capricious, it need not consider whether Amendment 5b's measures are appropriately responsive to that problem. The Court will test the validity of the agency's prescription once it is confident in the soundness of its diagnosis.

B. Annual Catch Limits and Accountability Measures

The Court turns next to Oceana's first argument-that the Fisheries Service arbitrarily and capriciously failed to establish accountability measures that would strictly enforce the zero-catch limit for dusky sharks. See Pl's MSJ at 17-18. While the Court will reject much of Oceana's argument, it nevertheless concludes that remand is required on this ground, too, as a result of the concerns detailed above regarding the data the agency relied upon in selecting the accountability measures it did.

Three MSA provisions are particularly relevant to this claim. First, the MSA requires the Fisheries Service to enact measures that prevent overfishing "while achieving, on a continuing basis, the optimum yield from each fishery." 16 U.S.C. § 1851(a)(1). Second, the MSA commands the Fisheries Service to establish an annual catch limit ("ACL") and accountability measures ("AMs") sufficient to ensure that *85"overfishing does not occur." Id. § 1853(a)(15). Third, another MSA provision directs the Fisheries Service to "prepare and implement a fishery management plan, plan amendment, or proposed regulations" designed "to end overfishing immediately in the fishery and to rebuild affected stocks of fish." Id. § 1854(e)(3)(A).

Oceana's argument is straightforward: because the Fisheries Service set an ACL of zero for dusky sharks, it had to establish accompanying accountability measures that would strictly enforce that limit-at least so long as overfishing continued to occur. Instead, Oceana says, the Fisheries Service knew its existing accountability measures had failed to result in compliance with the zero-ACL, and yet the Service chose not to meaningfully fortify those measures through Amendment 5b. Pl's MSJ at 21. By its inaction, Oceana claims, the Fisheries Service defied the MSA's statutory mandate and thus acted arbitrarily and capriciously.

Oceana's reasoning errs on both the law and the facts. Its legal theory-that the MSA requires the Fisheries Service to strictly limit bycatch in accord with its self-prescribed ACL of zero-has some intuitive appeal but is ultimately based on a misreading of the MSA and the accompanying guidelines. And its factual contentions-that the Fisheries Service's past accountability measures failed to curb bycatch and that Amendment 5b doesn't meaningfully improve on them-are mostly inconsistent with the administrative record. Just the same, largely because of its conclusion on the data question, the Court still cannot sign off on the adequacy of the accountability measures in Amendment 5b.

The Court's analysis will proceed as follows. First, it will clarify the statutory scheme relating to annual catch limits and accountability measures. Second, it will explain why the agency's regulatory efforts to date and the new measures proposed in Amendment 5b are not as deficient as Oceana suggests. And third, it will explain why-despite the Court's disagreement with Oceana's reading of the statute and its characterization of the administrative record-the Court must nevertheless order a remand on this ground as well.

1. The MSA, annual catch limits, and accountability measures

It is true that the MSA encourages the Fisheries Service to establish accountability measures that ensure compliance with a given annual catch limit. And it is also true that once the Fisheries Service knows an ACL has been exceeded, that encouragement becomes a mandate: it must establish accountability measures to correct for the overage. Oceana, Inc. v. Locke, 831 F.Supp.2d 95, 119-20 (D.D.C. 2011) (" Locke") ("NMFS's Guidelines indicate that ... AMs for when an ACL is exceeded are mandatory."); see 50 C.F.R. § 600.310(g)(3) ("If an ACL was exceeded, AMs must be implemented[.]" (emphasis added) ). Even so, there is space between the (undisputed) requirement that the Fisheries Service establish accountability measures to address ACL overages, and Oceana's favored requirement in this case that those measures must sharply curtail bycatch that contributes to the ACL overages, regardless of other considerations. The Court finds this latter position untenable.

For starters, the same guidelines provision that says accountability measures are required for ACL overages acknowledges that ACLs will be exceeded. Accountability measures in such situations are required, yes, and they should strive to prevent any overage-but they need only "correct or mitigate overages of the ACL if they occur." 50 C.F.R. § 600.310(g)(1). Oceana's hard-line, zero-means-zero approach to the ACL and AMs is thus difficult to square *86with guidelines that clearly build in some wiggle room. Instead, if the agency establishes AMs that are sufficient to "mitigate" the effect of the ACL-overages on the fish population, then that should be enough. Cf. Guindon, 31 F.Supp.3d at 200 ("NMFS need not implement so many accountability measures that overharvesting and overfishing become utterly beyond possibility. That reads too much into the MSA.").

Still more telling is another guidelines provision that addresses the precise situation at issue in this case: what the Fisheries Service must do when a zero-ACL is exceeded. "If an ACL is set equal to zero and the AM for the fishery is a closure that prohibits fishing for a stock, additional AMs are not required if only small amounts of catch (including bycatch) occur, and the catch is unlikely to result in overfishing." 50 C.F.R. § 600.310(g)(3). Here, the agency determined that the catch for the prohibited dusky shark was small and that-through a combination of existing regulations and Amendment 5b's new dusky shark-specific ones-overfishing was unlikely to continue. Defs' MSJ at 14-15 (citing AR007571; AR000417; AR004830-31); see AR007565 (listing past and proposed measures). The agency therefore contends it has acted consistently with the MSA, at least as interpreted in § 600.310(g)(3). Oceana's argument that that the agency must implement accountability measures to guarantee compliance with the ACL appears to run headlong into this guidelines provision. See Pl's MSJ at 18; Pl's Opp. at 6.

Perhaps recognizing that this language spells trouble for its argument, Oceana urges the Court to ignore it. It should do so, Oceana says, because the guidance conflicts with the MSA itself. Pl's Opp. At 6 ("To the extent the Fisheries Service is positing that this provision authorizes it to establish an annual catch limit of zero, yet consistently allow that limit to be exceeded, the provision itself conflicts with the [MSA]."). The Court disagrees.

Oceana is of course correct that the guidelines do not carry the force of law and that, if some portion of the guidelines contradicted the MSA, the statute would trump the guidelines. But the Court finds § 600.310(g)(3), as amended in 2016, fully consistent with the MSA. It makes sense that the requirement for additional accountability measures for ACL-overages would turn, at least in part, on whether those overages were significant and whether they resulted in overfishing. The primary evil the MSA guards against is overfishing; the law's various proscriptions and prescriptions exist to protect fish populations, not to require arbitrary ACLs for their own sake. See 16 U.S.C. § 1801(b)(1)-(7) (listing MSA's purposes). That much is made plain by the 2006 amendment to the MSA, which requires all fishery management plans to "establish a mechanism for specifying annual catch limits" and "measures to ensure accountability," but only to the extent necessary to ensure that "overfishing does not occur in the fishery[.]" Id. § 1853(a)(15). So while ACLs and AMs should-and in some cases, must-be used by the Fisheries Service, they are not ends in themselves, but rather means to end overfishing and rebuild populations. Cf. Locke, 831 F.Supp.2d at 117 (stating that additional AMs are unnecessary if current plan "establishes an overall suite of accountability measures sufficient to prevent overfishing"). Oceana's narrow focus on ACLs and accompanying AMs without reference to overfishing misses the sea for the kelp. Section 600.310(g)(3)'s focus on accountability measures sufficient to constrain overfishing is thus in keeping with the MSA, not contrary to it.

To be sure, that the Court finds § 600.310(g)(3)consistent with the MSA does not necessarily mean that the guidelines'

*87gloss on the MSA should be accorded weight. Because they do not carry the force of law, the guidelines are not automatically entitled to Chevron-style deference. See Locke, 831 F.Supp.2d at 116. Nevertheless, the Court concludes that these particular guidelines "deserve considerable deference." Id. at 117. As Locke noted, they were passed with extensive process and formality, are "detailed [and] reflect NMFS's considerable expertise in the complex field of fishery management, and are routinely cited by courts as persuasive authority on the meaning of the MSA." Id. (discussing factors for determining level of deference set forth in United States v. Mead Corp., 533 U.S. 218, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001) and Skidmore v. Swift & Co., 323 U.S. 134, 65 S.Ct. 161, 89 L.Ed. 124 (1944) ); see also Guindon, 31 F.Supp.3d at 198 ("[T]he guidelines are entitled to considerable deference in light of their thoroughness, the agency's expertise, and the administrative formalities involved in their promulgation." (internal quotation marks omitted) ). Though Locke was decided before the 2016 revisions, these revisions went through the same process-including public comment and response-as the earlier iterations of the guidelines considered in that case. See Magnuson-Stevens Act Provisions; National Standard Guidelines, 81 Fed. Reg. 71858-01, 71877 (responding to comments expressing concern over the revision). Moreover, both sides in this litigation routinely rely on the guidelines, and both suggest they are all but authoritative, save for Oceana's objection to the single sentence in § 600.310(g)(3) regarding zero-ACLs and additional AMs. See Locke, 831 F.Supp.2d at 117 (finding it significant that "Oceana does not challenge the Guidelines and in fact cites them in support of its argument"). The Court therefore has no qualms about leaning heavily on the guidelines to parse this difficult statutory dispute.

Oceana has backup arguments at the ready. It asserts that, even if the Court finds (as it does) § 600.310(g)(3) consistent with the MSA and deserving of this Court's deference, it does not protect the Fisheries Service in this instance. To refresh, the disputed portion of the guidelines provides: "If an ACL is set equal to zero and the AM for the fishery is a closure that prohibits fishing for a stock, additional AMs are not required if only small amounts of catch (including bycatch) occur, and the catch is unlikely to result in overfishing." 50 C.F.R. § 600.310(g)(3). Oceana first says that Amendment 5b does not create a "closure that prohibits fishing for a stock," id., since the Fisheries Service "has not prohibited all fishing that affects dusky sharks," Pl's Opp. at 6 (emphasis added). A clever move, it nevertheless amounts to re-writing the guidelines. To prohibit fishing for a species means that fishermen may not target it, and if they do catch it, they must release it. See 50 C.F.R. § 635.24 ; AR007570 ("[T]he existing AM for all of the stocks in the prohibited shark fishery is a closure that prohibits fishing for the stocks."). That has been true of dusky sharks since 2000, and that is precisely the sort of prohibition § 600.310(g)(3) contemplates.

Oceana's contrary reading of § 600.310(g)(3) is non-sensical. It cannot be that an agency has enacted a "closure that prohibits fishing for a stock" only when it has prohibited any and all fishing that might affect a particular stock, including fishing that results in bycatch. If the agency had already enacted such a drastic management measure, the Court cannot conceive of how bycatch would continue to occur at all. Thus, if the closure condition means what Oceana says it does, it cannot be squared with the rest of the provision discussing the possibility of bycatch continuing to occur. And even assuming that bycatch could somehow still occur with thetype *88of closure Oceana says § 600.310(g)(3) requires, what additional AMs could the agency possibly implement to stop it? The agency would have no need to seek shelter under § 600.310(g)(3) ; it has already done all it can to stop the species from being caught as bycatch. Oceana's interpretation would thus render this portion of the guidelines meaningless.

Oceana's better argument is that § 600.310(g)(3)'s caveat is inapplicable because "there can be no question that any bycatch occurring is 'likely to result in overfishing' " and the agency has not shown that "bycatch of dusky sharks is 'small.' " Pl's Opp. at 7 (quoting 50 C.F.R. § 600.310(g)(3) ). Because the 2016 stock assessment pegged dusky sharks as overfished and currently subject to overfishing, Oceana says that the Fisheries Service cannot claim future bycatch won't lead to overfishing. But the language "likely to result in overfishing" is prospective, not backward-looking, so the question whether the agency can claim this exception turns not on the 2016 stock assessment alone but rather on the likelihood that overfishing will soon end. As for whether the Fisheries Service can justifiably claim that dusky shark bycatch is currently "small," that is a more nebulous matter: the guidelines don't explain what qualifies as a small amount of bycatch. On this view of the law, then, the question becomes whether the administrative record supports the Fisheries Service's conclusion that Amendment 5b's proposed accountability measures, together with existing ones, would constrain bycatch (1) sufficiently to end overfishing and (2) to a small amount.

2. The Adequacy of the Agency's Accountability Measures

Oceana gives two reasons why the agency's efforts come up short. First, it casts the Fisheries Service's past efforts to enforce the annual catch limit and protect the dusky shark population as an abject failure. See Pl's MSJ at 20 (claiming that Fisheries Service has "allow[ed] unlimited bycatch of dusky sharks to occur" and has not "establish[ed] any management measures that actually constrain fishing effort in a way that prevents or limits bycatch"). Second, it claims that new management measures included in Amendment 5b are woefully deficient. See id. at 21 ("[T]he agency's decision to dress up this failed approach with a few additional provisions that may or may not reduce bycatch does not cure its infirmity."). Both of these contentions are overstated. As the Fisheries Service notes, the available evidence points in the other direction: overfishing of the dusky shark has declined significantly from 2000 through 2016, suggesting the agency's prior efforts to limit bycatch had been successful and will continue to yield benefits. Defs' MSJ at 11-12; AR007101 (documenting decline in overfishing and noting 2016 stock assessment's conclusion that twelve percent mortality reduction could end overfishing). And even if some of the new measures in Amendment 5b do not directly prevent bycatch from occurring , they do seek to limit the damage such bycatch causes to the dusky shark population. Id.

A brief refresher on the Fisheries Service's pre-Amendment 5b regulatory efforts is in order. In 2000,¹⁷ when the 1999 HMS FMP listing dusky sharks as a prohibited species went into effect, the directed fishery for dusky sharks was brought to an end. AR007076. The 2006 HMS FMP sought to reduce dusky shark bycatch (and resulting mortality) by requiring shark dealers to attend identification workshops. Id. Amendments to the 2006 HMS FMP brought further measures. Amendment 2, *89in 2008, made particularly important changes to the directed shark fishery, where dusky sharks were often caught as bycatch. See AR007086; supra at 74. Among other things, it cut the commercial fishing quota for large coastal sharks; limited the times and areas where certain fishing could occur; and largely prohibited fishing for sandbar sharks, the targeting of which often led to incidental dusky shark bycatch. See AR007077; AR007087.

The data show that these measures were effective. While Oceana fixates on the fact that the three stock assessments (2006, 2011, and 2016) all concluded that the dusky shark population is in an overfished state with overfishing occurring, see AR004804, those same assessments revealed a dramatic decline in overfishing, see AR007101. The 2011 stock assessment estimated a median fishing mortality rate of 1.59 F/F_MSY(with any value over 1 indicating overfishing); the 2016 assessment, meanwhile, estimated a further reduced rate of 1.18 F/F F_MSY.¹⁸ From its peak in 1999 through 2015, the Fisheries Service estimated that dusky shark fishing mortality rate declined almost 90 percent. Id.; see AR007102, Fig. 1.3. This data belies Oceana's contention that the Fisheries Services' accountability measures "ha[ven't] worked for the past two decades" and that it was irrational for the agency to expect continued improvement. Pl's Opp. at 5. Quite the opposite, the record reveals that the Fisheries Service's "significant regulatory action ... has reduced fishing effort and mortality on dusky sharks" and will continue to do so. See Defs' MSJ at 17; AR007101 (stating that improving stock assessments "demonstrate[e] that management measures implemented since 2000 have had a significant impact on reducing fishing mortality on dusky sharks").

Oceana persists that, even if past management efforts have yielded some benefits, that does not mean they satisfy the MSA's requirement of accountability measures that ensure that "overfishing does not occur." 16 U.S.C. § 1853(a)(15). In its view "merely reducing [overfishing] is not sufficient" since "reducing the speed at which the agency is driving does not mitigate the fact that it is going the wrong way." Pl's Opp at 3. What Oceana's metaphor has in color, it lacks in coherence. For the data does not show the Fisheries Service is headed in the wrong direction. That might be so if the Fisheries Service's efforts since 2000 had shown no impact on the rate of overfishing, or had in fact exacerbated it, but the reverse appears true: the agency's efforts have reduced the rate of overfishing.¹⁹ Assuming that this rate of decrease in overfishing remains constant, or perhaps even quickens-a reasonable assumption given Amendment 5b's additional measures-the agency could reasonably conclude that overfishing of the dusky shark will come to an end in short order. AR007101 (noting 2016 stock assessment's conclusion that twelve percent mortality reduction could end overfishing).

*90To be sure, Oceana is right that evidence of progress cannot alone save the Fisheries Service's actions from being arbitrary or capricious. Progress that comes too slowly can run afoul of the MSA. See 16 U.S.C. § 1854(e)(3)(A) (directing Fisheries Service to "prepare and implement a fishery management plan, plan amendment, or proposed regulations" designed "to end overfishing immediately in the fishery and to rebuild affected stocks of fish" (emphasis added) ). All the same, provisions like § 1854(e)(3)(A) must be read in light of the broader statutory scheme, not in isolation. And other elements of the MSA show that the Fisheries Service cannot pursue an end to overfishing irrespective of collateral consequences. National Standard 8, for instance, requires that "[c]onservation and management measures shall, consistent with the conservation requirements of this chapter (including the prevention of overfishing and rebuilding of overfished stocks), take into account the importance of fishery resources to fishing communities ... in order to ... minimize adverse economic impacts on such communities." Id. § 1851(a)(8); see also 50 C.F.R. § 600.345 (discussing National Standard 8). Elsewhere, the MSA mandates that an FMP for an overfished species strive to rebuild the stock as quickly as possible while also "taking into account the status and biology of any overfished stocks of fish [and] the needs of fishing communities[.]" Id. § 1854(e)(4)(A)(i).

The upshot is that even if more aggressive accountability measures might better prevent bycatch, substantial countervailing costs-ones the MSA and its attendant guidelines require the Fisheries Service to consider-might counsel against their adoption. See Defs' Reply at 6 (admitting that some alternatives might more quickly curtail bycatch but would be "impracticable and unnecessary"); AR007566 (explaining that hotspot closures were disfavored because they have "negative socioeconomic impacts"). It is the agency's job, not this Court's, to strike the appropriate balance, so long as the agency does so in a manner reasonably consistent with its statutory mandate. See Guindon, 31 F.Supp.3d at 200 (stating that "which accountability measures NMFS should have required, or should require in the future" is a "decision ... best left to the expertise and discretion of the agency tasked with carrying out the statute"); Citizens to Pres. Overton Park, 401 U.S. at 416, 91 S.Ct. 814 (stating that reviewing court's inquiry is limited to "whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment").

At any rate, the Fisheries Service is taking additional steps to hasten the pace of progress. AR007101 (acknowledging that, despite the reduction in overfishing, "it has not yet been completely ended, hence the need for additional management measures"). Amendment 5b includes six additional accountability measures that address the persistent bycatch problem. It requires recreational and commercial fishermen to undergo education on dusky shark identification to reduce retention rates by fishermen who misidentify and fail to release them; it mandates training on how to safely handle dusky sharks to reduce mortality due to bycatch; it orders recreational and bottom longline fisheries to use circle hooks, thought to be less harmful and thus less likely to cause death after discard; and it establishes a fleet communication protocol to help fishermen avoid areas where dusky sharks are likely to be ensnared. AR007565. And while it is true that some of Amendment 5b's new measures will not stop bycatch from occurring in the first instance, those measures are nevertheless calculated to reduce the risk that dusky sharks will die as a result of having been incidentally caught. See, e.g., AR007566 (explaining that gear modification *91and release training measures are designed to "reduce at-vessel and post-release mortality rates"). Thus, contrary to Oceana's dim view of the Fisheries Service's management efforts, the agency's past measures appear to have proven effective, and Amendment 5b supplements those measures with several new dusky shark-specific ones to accelerate an end to overfishing and fast-track the rebuilding process.

All the same, Oceana says that two cases decided in this district dictate a result in its favor: Oceana v. Locke and Guindon v. Pritzker. Neither case does so. Start with Locke. There, the court rejected as inadequate an amendment to a New England Fishery Management Council FMP because it lacked responsive (or "post-season") accountability measures that would be triggered when a zero-ACL for five species was exceeded. 831 F.Supp.2d at 119-21. But, for a few reasons, Locke's application to this case is limited. First, Locke was decided before the guidelines to the MSA were amended in 2016, making clear that additional accountability measures were not required for a zero-ACL exceedance so long as "small amounts of catch (including bycatch) occur, and the catch is unlikely to result in overfishing." 50 C.F.R. § 600.310(g)(3). Locke relied heavily on the prior version of § 600.310(g)(3), particularly its statement that "AMs must be triggered and implemented as soon as possible" if an ACL were exceeded. 831 F.Supp.2d at 120. But what seemed like an unequivocal command under the guidelines in effect in 2011 became something very different in 2016, when the guidelines were amended to explain that responsive AMs were not automatically required when zero-ACLs were exceeded. As the Court has already explained, it finds the 2016 guideline amendment consistent with the MSA's predominant focus on management measures that are calibrated to end overfishing. Thus, Locke was decided on a more stringent view of the agency's MSA's obligations than the Court takes today.

Second, unlike here, the Fisheries Service in Locke could point neither to the success of its past bycatch-mitigation measures nor to any new promising ones. Instead, the agency tried to rely on a proposed accountability measure that, for obvious reasons, could not be expected to curtail overfishing of the five species at issue in the case. That measure provided that, if an ACL for one of the five stocks was exceeded, "common-pool vessels" would face reduced days-at-sea to compensate for the overage, but "sector vessels" would not. Id. at 118. The rub, though, was that common-pool vessels accounted for only two percent of the fishing done on those five stocks, while the sector vessels accounted for the remaining 98 percent. "In other words, the best evidence the Agency has pointed to concerning total AMs for the five stocks caught as bycatch by vessels fishing 98% of the Fishery are reductions in fishing effort imposed upon vessels fishing 2% of the Fishery." Id. Oceana has not identified a similar anomaly here: prior accountability measures-including sharply limiting bottom longline fishing for sandbar sharks, slashing the quota for non-prohibited large coastal sharks, and enforcing time and area closures-have proven responsive to the overfishing problem; and Amendment 5b promises still more reforms to reduce bycatch or at least limit the mortality that results from it. See supra at 88-91.

Guindon is likewise distinguishable. In Guindon, which like Locke was decided before the 2016 guidelines amendment, the Fisheries Service chose not to implement an accountability measure called a "buffer"-which sets the amount of permitted catch below the annual catch limit to guard against uncertainty in catch data-to limit *92excessive catch of red snapper. 31 F.Supp.3d at 199-200. A court in this district held that that decision violated the MSA, specifically § 1853(a)(15)'s requirement that the agency "establish a mechanism for specifying annual catch limits ... at a level such that overfishing does not occur in the fishery, including measures to ensure accountability." Id. at 197. But that was because, one, the "single accountability measure" the agency did choose to limit red-snapper overfishing-use of an in-season closure-had proven ineffective in the past, and, two, the agency added nothing new to its management arsenal. Id. at 200 ("Such a dogged belief that somehow 2013 would be different than previous years defies logic."). Not so here. Rather, as detailed above, the record shows that the agency's past efforts have paid dividends, AR007101 ("[M]anagement measures implemented since 2000 have had a significant impact on reducing fishing mortality on dusky sharks."), and Amendment 5b proposes to buttress those efforts, not merely replicate them, AR007071 (acknowledging "need for additional management measures" until overfishing is ended). See supra at 88-91.

What actually appears to frustrate Oceana, then, is not that Amendment 5b offers no new accountability measures-it plainly does-but that it doesn't contain the precise one Oceana thinks the MSA requires. Specifically, Oceana contends that the agency must set an annual catch limit above zero, the exceedance of which must trigger an accountability measure that lowers the level of permissible catch in the following year, or some other similar reactive measure. See Pl's MSJ at 19-22. Oceana says the agency's current practice of setting an annual catch limit of zero without an accountability measure that is triggered when that limit is exceeded allows "effectively unlimited bycatch" that "will perpetuate the overfishing that dusky sharks already experience." Id. at 20.

But the agency gave two cogent reasons for rejecting Oceana's favored alternative: first, that fixing a tolerable catch level above zero would be highly speculative, AR007089 (explaining why "highly variable" catch data "makes it difficult to determine an appropriate number to use"); AR007133-34 (stating that any ACL greater than zero "would have extremely high uncertainty"); and second, that a non-zero ACL might "send a message to fishermen that interactions are permissible at some level and could disincentivize avoidance of interactions," AR007491. The Court, consistent with others in this district, finds the second reason especially compelling. See Locke, 831 F.Supp.2d at 119 (holding that it is rational for agency to believe zero-ACLs "will discourage sectors from targeting" stocks, "which, in turn, will likely reduce their catch levels"). In any event, that the agency did not choose Oceana's preferred alternative does not mean the agency sat on its hands; it implemented AMs in the past and promises to add still more with Amendment 5b. When an agency chooses between alternatives-so long as each alternative can reasonably be expected to advance the agency's objective-it is prudent to defer to its choice, not dictate it. Nat'l Fisheries Inst., Inc. v. Mosbacher, 732 F.Supp. 210, 223 (D.D.C. 1990) (noting that, when choosing among alternatives, it is "especially appropriate for the Court to defer to the expertise and experience of those individuals and entities-the Secretary, the Councils, and their advisors-whom the [MSA] charges with making difficult policy judgments and choosing appropriate conservation and management measures based on their evaluations of the relevant quantitative and qualitative factors").

3. Whether Remand Is Ultimately Required

So where does all this leave us? The foregoing demonstrates two things:

*93first, that Oceana's interpretation of the MSA and its attendant guidelines does not automatically require the agency to impose accountability measures for exceedances of a zero-annual catch limit; and second, that Amendment 5b does not, as Oceana says, "simply apply the same approach that has failed for over fifteen years to effectively limit dusky shark bycatch and end overfishing." Pl's MSJ at 21.

That notwithstanding, the Court still cannot conclude that the Fisheries Service has satisfied its mandate under the MSA. At the end of the day, the agency must show that its collection of accountability measures-those established before and by Amendment 5b-will keep bycatch to a "small" amount that is unlikely to lead to overfishing. 50 C.F.R. § 600.310(g)(3). While it is intuitive-given the improving stock assessments-that whatever level of bycatch is occurring may be less and less likely to cause overfishing, the agency's support for its contention that bycatch is (and will be) "small" remains thin. Indeed, its evidence for the conclusion that bycatch is small is the observer data that the Court has just concluded is an inadequate proxy for assessing bycatch writ large. See Defs' MSJ at 5 n.14 (citing only limited observer data as support for its conclusion that dusky shark bycatch was "small"); AR007085 (citing low "observed bycatch amounts" for support that bycatch was "small"); AR007492-93 (same). So while the Court grants that the agency has reason to think that its management efforts will lead to an end to overfishing and that dusky shark bycatch is and will remain small, there remains the distinct possibility that more comprehensive data might reveal otherwise, thereby undermining the agency's conclusion that different or additional AMs are unnecessary. In this way, the agency's argument that it has complied with the MSA's obligation to implement sufficient accountability measures rises and falls with its argument that it has considered an appropriate universe of data. And so, at least for today, they fall together.

But the Court must emphasize what it is not ordering the Fisheries Service to do. The Court is not concluding at this point that the accountability measures the agency has chosen for Amendment 5b are inadequate. That may well turn out to be the case. For instance, fresh consideration of the available logbook data might show that the agency, as Oceana contends, was wrong to implement management measures solely within the HMS Fishery. But without knowing what the agency's data reveals regarding bycatch in the Southeastern Atlantic Snapper-Grouper Fishery or the Gulf of Mexico Reef Fish Fishery, it would be premature for the Court to pass on that question. See supra at 77 n.6. A more comprehensive data set could also reveal that the agency's chosen accountability measures for the HMS Fishery are not severe enough to constrain bycatch to a "small" level that is "unlikely to result in overfishing." 50 C.F.R. § 600.310(g)(3). If that proves true, Locke may well have something to say about this case: the omission of responsive accountability measures that are triggered when the zero- ACL is exceeded might prove unreasonable. See 831 F.Supp.2d at 119-20 (holding that prospective AMs were insufficient).

But the Court need not reach these issues just yet. Instead, all the Court holds today is that the agency has taken action based on a data set that the agency has not rationally justified. As a consequence, the Court cannot validate the agency's conclusion that bycatch is "small" in the non-HMS fisheries, and that it therefore need not implement additional accountability measures in those fisheries to protect the dusky shark. Nor can the Court sign off on the adequacy of Amendment 5b's new accountability measures in the HMS fishery.

*94Before the Court will resolve either of those issues, the agency must take a fuller view of the bycatch data or better explain to the Court why it need not do so.

C. Consideration of Alternatives

Oceana's final argument is that the Fisheries Service violated NEPA's requirement that it consider and evaluate alternatives to achieving the agency's stated goal. Pl's MSJ at 40 (citing 40 C.F.R. § 1502.14 and 42 U.S.C. § 4332(C) ). In fact, Oceana says the agency failed even to take a hard look at whether its chosen path would lead it to its desired destination. Id. The Court, however, concludes that it need not reach Oceana's NEPA argument given that it does not differ materially from its MSA argument, on which basis the Court has already remanded Amendment 5b. See Guindon, 31 F.Supp.3d at 201 (finding no need to reach NEPA claim when it does not materially differ from relief sought under MSA); Nat. Res. Def. Council, Inc. v. Daley, 209 F.3d 747, 749 (D.C. Cir. 2000) ("Because of our disposition on [MSA] grounds, we have no need to reach appellants' NEPA claims."). Moreover, just as the Court has reserved judgment on the adequacy of Amendment 5b's accountability measures pending the agency's incorporation of logbook data or a more persuasive explanation for why it need not do so, the Court for similar reasons finds it would be premature to evaluate whether the agency considered a reasonable range of alternatives and offered rational reasons for rejecting them.

IV. Conclusion

For the foregoing reasons, the Court concludes that the final rule implementing Amendment 5b is arbitrary and capricious and violates the MSA. The APA requires agencies to examine all relevant data and consider all important aspects of a problem, while the MSA mandates that the Fisheries Service implement management measures to enforce quotas and prevent overfishing. Because the Court finds that the Fisheries Service in Amendment 5b did not fulfill these statutory mandates, Oceana is entitled to summary judgment and the amendment must be remanded to the agency.

The Court will, however, defer final decision on the scope and duration of that remand. In the Order that accompanies this Memorandum Opinion, the Court will order the parties to confer and submit a proposal clarifying the appropriate remedy. In the meantime, nothing in the Court's opinion should be read to vacate any part of Amendment 5b; the agency may proceed with the management measures prescribed therein.

1 "Overfishing" and "overfished" bear technical meanings under the MSA. "Overfishing" means "a rate or level of fishing mortality that jeopardizes the capacity of a fishery to produce the maximum sustainable yield on a continuing basis."16 U.S.C. § 1802(34) ; see also 50 C.F.R § 600.310(e)(2)(i)(B). Maximum sustainable yield is the "largest long-term average catch or yield that can be taken from a stock or stock complex under prevailing ecological, environmental conditions and fishery technological characteristics." 50 C.F.R. § 600.310(e)(1)(i)(A). As Oceana puts it, "overfishing occurs when a fishery removes too many fish, too quickly relative to the fish's population size and ability to reproduce." Pl's MSJ at 4 n.2. When prolonged overfishing occurs, a population reaches an "overfished" state. 50 C.F.R. § 600.310(e)(2)(i)(E).

2 Shark habitats fall into four broad categories: coastal, pelagic, coastal-pelagic, and deep-dwelling. AR007159. Coastal species "inhabit estuaries, the nearshore and waters of the continental shelves," while pelagic species "range widely in the upper zones of the oceans, often traveling over entire ocean basins." Id. Coastal-pelagic species, like the dusky shark, are "intermediate in that they occur both inshore and beyond the continental shelves, but have not demonstrated mid-ocean or transoceanic movements." Id.

3 Stock assessments are used to gauge the abundance of a given species population. The science behind the stocks assessments is highly technical and hardly discussed in the parties' briefs. The Fisheries Service has conducted three dusky shark stock assessments-in 2006, 2011, and 2016-using a so-called "catch-free" model to determine the status of the dusky shark population. Counsel for the Fisheries Service demurred at the hearing when the Court asked her to "explain in plain English what a catch-free model is," Hr'g Tr. 29:7-8, responding that as "a lawyer and not a scientist," she would instead "get NMFS to provide an explanation for the Court" in a later submission, id. at 30:5-6. The agency has since provided that explanation, with helpful citations to the administrative record. Declaration of Karyl Brewster-Geisz ("Brewster-Geisz Decl."), ECF No. 46-1. Since the agency's bycatch data (rather than the stock assessments themselves) are the primary object of this litigation, the Court need not dive too deep into the latter's mechanics. Still, a brief summary of how the assessments work seems in order.

The "catch-free" stock assessments use "biological and abundance data ... to determine the proportional difference in population level from current population levels to 'virgin' stock levels (the population that prevailed at a time at which scientists and fishermen agree was before fishing on the species began)." Brewster-Geisz Decl. ¶ 8. Biological data "includes information about the life history of the species," including life span and reproductive rates. Id. ¶ 5. Abundance data include factors like "catch rates, or catch per unit of effort," id., which the Court understands to mean data regarding the number of fish caught given a certain level of fishing activity. While this method "cannot provide absolute abundance levels" and thus "cannot provide the information needed to determine the exact catch levels that will allow the stock to rebuild," it can be used to "provide estimates of the relative proportion of abundance compared from one time period to another." Id. ¶ 9. It was those comparisons that informed the agency's determination that dusky sharks were overfished and that overfishing was occurring. Id.

Given the agency's description, the Court wonders about the "catch-free" label. While perhaps it is standard in marine conservation circles, it is more than a little misleading. The assessments may not incorporate absolute catch numbers in its modeling, but they certainly seem to rely heavily on catch rates. Id. ¶ 5; AR004391 ("When fitting to indices of abundance and catch rates, the model predicts ...."). And the Court, while no expert, cannot conceive how catch rates can be calculated without first determining the number of fish caught given a certain amount of effort. See AR004387 (2016 stock assessment explaining that "[i]f catch and effort information are available from sampled trips or observer programs, then standardized catch rates can be developed and incorporated into the model").

4 The degree of specificity of the closures is something to behold. See, e.g. 50 C.F.R. § 635.21(d)(iii)(D) ("Charleston Deep Artificial Reef. Bounded by rhumb lines connecting, in order, the following points: 32°9.65' N lat., 79°9.2' W long.; 32°7.155' N lat., 79°5.595' W long.; 32°2.36' N lat., 79°9.975' W long.; 32°5.04' N lat., 79°13.575' W long.")

5 Whereas bottom longline fishing, as its name implies, involves a fishing line being stretched horizontally near the bottom of the ocean, pelagic longline fishing uses a main line that is placed near the surface with baited hooks attached at intervals at the end of branching lines. AR007161-62.

6 Whether fresh consideration of all available logbook data on dusky shark bycatch should compel the agency to expand Amendment 5b's focus beyond the HMS Fishery is a question on which the Fisheries Service should have the first pass. All the Court holds today is that the agency has not rationally explained its decision to exclude logbook data in making that determination.

7 Though the parties dispute whether the 3,800 figure reported in the First Edition is a four-year total or an annual average, the report itself states that the numbers reflect annual averages. See NBR First Edition Update, Table 4.1 "Southeast Region Fish Bycatch by Fishery," available at https://www.st.nmfs.noaa.gov/Assets/Observer-Program/bycatch-report/Table_4.1.pdf (last accessed January 10, 2019).

8 What other logbook evidence shows is largely a mystery. Although the agency collects logbook data, it did not rely extensively on that data in making its management decisions in Amendment 5b, so they are not explored much in the EIS or elsewhere in the administrative record. That goes a long way toward explaining why Oceana is now embroiled in a Freedom of Information Act dispute that appears to be coming to a close. See Oceana, Inc. v. NOAA, et. al., Case No. 18-cv-648, Joint Status Report, ECF No. 24 (indicating intent to voluntarily dismiss suit).

9 The agency suggests that a single paragraph in a letter responding to Oceana proves the agency used observer, not logbook, data to assess dusky shark bycatch in the First Update. See Defs' MSJ at 21 (citing AR005430). That passage, however, provides only oblique support for the agency's claim. True, it discusses the use of observer data in the First Update, but it addresses neither the fact that the First Update itself says it relied on logbook data nor affirmatively states that logbook data was not used. The Court therefore concludes that the Fisheries Service's decision to dismiss the First Update had at least something to do with its reliance on logbook data.

10 As Oceana insists in its opposition, its data argument does not turn on the validity of the First Update data. While Oceana does argue that the data reflected in the National Bycatch Reports should have "sounded an alarm to the agency," its real beef is with "[t]he agency's wholesale, unexplained failure to consider the available and reliable logbook evidence," regardless whether that evidence can be found in the National Bycatch Reports. Pl's Opp. at 10.

11 Again, the agency spends a great deal of space explaining why it did not rely on the First Update to the First Edition of the National Bycatch Report, but that argument by itself does not explain the agency's general refusal to consider available logbook evidence. Whether the bycatch report was based on logbook or observer data, Oceana's point is that the agency should not have outright dismissed the report "without considering (or even consulting) other sources of information that the agency itself collected and which would have provided some understanding of dusky shark bycatch." Pl's MSJ at 28.

12 The agency also complains that the Coastal Fisheries Logbook reports catch data by weight as opposed to number, which is what the HMS Logbook and observers report. AR007260. But why the agency could not, for example, divide the total weight by the average weight of a dusky shark to harmonize the data with the HMS Logbook and observer data is left unexplained.

13 After refusing to draw any extrapolated estimates (either from observer or logbook data) from the commercial fisheries data, the agency took the still-more perplexing step of comparing limited, unextrapolated data for commercial fishing against large, extrapolated estimates for recreational fishing to draw the conclusion that the latter is more harmful to dusky sharks than the former. AR007092-94.

14 This bit of the administrative record reveals the novelty of the agency's position on logbook data. For it clearly used to think that, at least in some respects, logbook data was superior to observer data. Of course, that the agency has essentially done a one-eighty on the issue does not compel the conclusion that it is violating the APA and the MSA, but it does raise some red flags.

15 This statement alone, the Court notes, might provide an ample basis to set aside Amendment 5b, since it shows that the agency's outright dismissal of the logbook data in favor of observer data is patently unreasonable. The agency cannot on the one hand claim logbook data is superior to observer data for some analytical purposes related to bycatch, and on the other hand say it tells us nothing about the extent of the bycatch problem.

16 Of course, this does not mean that the Fisheries Service must factor all contrary data into Amendment 5b; data that is obviously "corrupted," "skewed," or "useless" may legitimately be disregarded, especially when a clearly superior data alternative exists. Pritzker, 31 F.Supp.3d at 197 n.20. But the agency has not persuaded the Court that this is true here, particularly in light of the agency's historical and consistent use of similar data and its admission that observer data suffers from similar flaws.

17 The Court recognizes that regulatory efforts for Atlantic shark species pre-date 2000 but starts its catalog here since the parties focus their discussion on efforts after 2000.

18 F = Fishing mortality rate and MSY = maximum sustainable yield.

19 True, as Oceana points out elsewhere, the catch-free stock assessments do not actually track the number of dusky sharks caught as bycatch. Plaintiff's Response to Defendant's Submission Re Catch-Free Modeling, ECF No. 48, at 2. But that does not mean they do not provide at least some support for the agency's conclusion that bycatch has been declining. While the stock assessments cannot peg bycatch at 100 or 1,000 or 100,000, they can track whether a species is growing or declining over time. And if it is growing, as the three dusky shark stock assessments show that it is, it is reasonable to deduce that bycatch, at least that leading to dusky shark deaths, is declining.