UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA

_________________________________________ ) OCEANA, INC., ) ) Plaintiff, ) ) v. ) ) WILBUR ROSS, ) United States Secretary of Commerce, et al., ) Civil Action No. 08-1881 (PLF) ) Defendants, ) ) and ) ) FISHERIES SURVIVAL FUND, ) ) Defendant-Intervenor. ) _________________________________________ )

OPINION

This matter comes before the Court on defendants’ notice that the National

Marine Fisheries Service had revised its Incidental Take Statement, thus completing its remand

in response to the Court’s August 17, 2018 Opinion and Order. Plaintiff Oceana, Inc. filed a

response to the notice, challenging the adequacy of the agency’s revisions on remand, and the

parties proceeded to brief the matter. Upon consideration of the second revised Incidental Take

Statement, the parties’ briefs and representations at oral argument, the relevant legal authorities,

and the entire record in this case, the Court will remand to the agency for a limited purpose. 1

1 In reaching its decision, the Court has reviewed the following filings, including the exhibits attached thereto: First Amended Complaint for Declaratory and Injunctive Relief (“Am. Compl.”) [Dkt. No. 80]; Defendants’ Second Notice of Completion of Remand (“Second Notice of Completion of Remand”) [Dkt. No. 134]; Oceana’s Response to Second Notice of Completion of Remand (“Oceana’s Response”) [Dkt. No. 137]; Defendants’ Response in I. STATUTORY AND REGULATORY FRAMEWORK

In its 2014 Opinion, the Court described the relevant statutory and regulatory

framework and recounted the factual and procedural history of this case. See Oceana, Inc. v.

Pritzker,

(D.D.C. 2014). The Court thus recites here only those

matters relevant to resolving the parties’ instant dispute.

The Endangered Species Act (“ESA”) of 1973, as amended,

et seq., created a comprehensive legislative and regulatory scheme that seeks to preserve

and protect species of animals facing man-made threats to their continued existence. See Lujan

v. Defs. of Wildlife,

(1992); Tenn. Valley Auth. v. Hill,

(1978). As part of this scheme, Section 7 of the ESA sets forth “the steps that federal agencies

must take to ensure that their actions do not jeopardize endangered wildlife and flora.” See Nat’l

Ass’n of Home Builders v. Defs. of Wildlife,

(2007). In particular, Section

7(a)(2) requires that each federal agency, “in consultation with and with the assistance of [the

National Marine Fisheries Service (“NMFS”) or the U.S. Fish and Wildlife Service (“FWS”)],

insure that any action authorized, funded, or carried out by such agency . . . is not likely to

jeopardize the continued existence of any endangered species or threatened species or result in

Support of Second Notice of Completion of Remand (“Defendants’ Response”) [Dkt. No. 138]; Defendant-Intervenor Fisheries Survival Fund’s Response in Support of Second Notice of Completion of Remand (“Intervenor’s Response”) [Dkt. No. 139]; Oceana’s Reply in Opposition to Second Notice in Completion of Remand (“Oceana Reply”) [Dkt. No. 140]; Transcript of January 29, 2020 Motions Hearing (“Oral Arg. Tr.”) [Dkt. No. 149]; Defendants’ Notice of Reinitiation of Consultation (“Notice of Re. of Consult.”) [Dkt. No. 150]; Oceana’s Response and Opposition to Notice of Reinitiation of Consultation (“Oceana Opp. to Notice”) [Dkt. No. 151]; Defendant-Intervenor Fisheries Survival Fund’s Response to Oceana’s Response and Opposition to Notice of Reinitiation of Consultation (“Intervenor Resp. to Notice”) [Dkt. No. 152]; and Defendants’ Reply in Support of Notice of Reinitiation of Consultation (“Defendants’ Resp. to Notice”) [Dkt. No. 153].

2 the destruction or adverse modification of habitat of such species . . . .” See 16 U.S.C. 2 § 1536(a)(2).

The Section 7 consultation process culminates in the issuance of a Biological

Opinion, or BiOp, in which the consulting agency sets forth its “opinion, and a summary of the

information on which the opinion is based, detailing how the agency action affects the species or

its critical habitat.” See

(b)(3)(A); see also

(h). Where the

consulting agency concludes that the agency action is not likely to jeopardize the continued

existence of the species but is nonetheless likely to result in some “incidental take,” the BiOp

must include an Incidental Take Statement (“ITS”) specifying the permissible extent of this

impact on the species. See

(b)(4);

(i). The ITS must set

forth conditions that include “reasonable and prudent measures” considered “necessary or

appropriate to minimize” the impact of any incidental takings. See

(i)(1)(ii). 3 And if the amount or extent of incidental taking ever exceeds that specified

in the ITS, the action agency must reinitiate Section 7 consultation “immediately.” See 50

2 FWS and NMFS jointly administer the ESA. See

(b). FWS administers the statute with respect to species under the jurisdiction of the Secretary of the Interior, while NMFS covers those species under the jurisdiction of the Secretary of Commerce. See Nat’l Ass’n of Home Builders v. Defs. of Wildlife,

. The agency whose action is at issue is known as the “action agency,” while either FWS or NMFS serves as the “consulting agency.” See Nat’l Wildlife Fed’n v. Nat’l Marine Fisheries Serv.,

(9th Cir. 2008). In this case, NMFS serves as both the “action agency” and the “consulting agency”: NMFS’ Sustainable Fisheries Division of its Northeast Regional Office administers the fisheries management program that governs the Atlantic Sea Scallop Fishery, making it the action agency here, while the Protected Resources Division of the same Regional Office has served as the consulting agency. See Am. Compl. ¶¶ 18-19. 3 As defined by the ESA, to “take” means “to harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect, or to attempt to engage in any such conduct.” See

(19). Although Section 9 of the ESA generally prohibits any taking of a listed species, see

(a)(1), incidental takes are permissible if they occur in accordance with the conditions set forth in an ITS, see

(i)(5).

(i)(4); see also

(a). As a result, incidental take monitoring is a

key component of any ITS – without the ability to monitor incidental takes, these regulatory

requirements become meaningless.

II. FACTUAL AND PROCEDURAL BACKGROUND

This case is over a decade old. In 2008, Oceana, Inc. (“Oceana”), an international

advocacy group focused on ocean conservation, filed a complaint challenging the 2008

Biological Opinion issued by the National Marine Fisheries Service (“NMFS”) that assesses the

impact of the Atlantic Sea Scallop Fishery on the Northwest Atlantic distinct population segment

(“NWA DPS”) of loggerhead sea turtles. See Dkt. No. 1. After NMFS reinitiated consultation,

the proceedings were stayed until July 12, 2012, when NMFS issued a superseding Biological

Opinion (“the 2012 BiOp”). See Dkt. Nos. 70, 73. Oceana filed an Amended Complaint

challenging the updated 2012 BiOp and the parties filed cross-motions for summary judgment.

See Dkt. Nos. 80, 87, 88 and 88-1 (memorandum in support), and 91.

On December 17, 2014, this Court issued an opinion granting in part and denying

in part the parties’ cross-motions for summary judgment and, in turn, remanding the case to the

agency for the limited purpose of addressing two deficiencies in the 2012 BiOp. See Oceana,

Inc. v. Pritzker (“Oceana I”),

(D.D.C. 2014). First, the Court remanded with

regard to the agency’s decision to evaluate loggerhead takes resulting from trawl gear fishing

only once every five years. See

. Because NMFS had given only “terse treatment”

to trawl take monitoring in the 2012 BiOp, the Court remanded in order to allow the agency to

“either provide a more thorough explanation of its choice, or, if unable to do so, reach a different

conclusion.”

.

4 Second, and most relevant to the instant opinion, the Court remanded to allow the

agency to better explain its reliance on a monitoring surrogate to measure loggerhead turtle takes

caused by dredge fishing. Oceana I,

. Because technologies meant to

benefit loggerheads have also made direct observations of takes more difficult, NMFS had

chosen to measure takes by a surrogate – namely, by “dredge hour,” or the number of hours

spent dredge fishing. See

. The agency had considered a variety of monitoring

mechanisms, but ultimately decided to monitor takes by using the number of hours spent dredge

fishing in Mid-Atlantic waters from May through November as a surrogate for actual takes. See

. The Court, however, found that “the 2012 BiOp fail[ed] to sufficiently explain how

the specific number of dredge hours that NMFS ha[d] chosen as a monitoring surrogate

adequately serve[d] as a monitoring proxy for the numerical take limit of 161 loggerheads.”

. And because NMFS had not adequately explained whether and how 252,323 hours spent

dredge fishing equated to 161 takes, it had not shown that the dredge hour surrogate would serve

as an adequate “trigger” for requiring reinitiation of consultation when actual takes surpassed the

take limit. See

. The Court thus remanded the ITS in order for NMFS to “more

clearly explain the connection,” or, if unable to do so, to choose a monitoring mechanism that

“does align with the numerical take limit.” See

. In doing so, the Court directed the

agency to “address Oceana’s valid concern regarding the effectiveness of linking an hour-based

surrogate to a numerical take limit, in the context of a Fishery where conditions change on a

continuous basis.”

NMFS thereafter revised the ITS for the 2012 BiOp and Oceana subsequently

filed a response arguing that the revised ITS remained defective and failed to comply with both

of the Court’s requirements on remand. See Dkt. Nos. 113, 116 and 117. On August 17, 2018,

5 this Court issued an opinion that remanded the case to the agency for a second time. See

Oceana, Inc. v. Ross (“Oceana II”),

(D.D.C. 2018). The Court first held that

NMFS had sufficiently addressed one of the two deficiencies the Court had identified in the 2012

BiOp’s ITS.

. The revised ITS provided rationales for why NMFS had concluded that

the five-year monitoring system for trawl takes was the best available method.

.

The Court remanded to the agency for a second time, however, with respect to the

other deficiency. It told the agency yet again that it needed to better explain its reliance on a

monitoring surrogate to measure loggerhead turtle takes caused by dredge fishing. See

Oceana II,

. The revised ITS included a scatterplot graph displaying a

fitted regression line with an R-squared value of 0.9164 and explained that it “illustrated the

strong positive linear relationship that exists between estimated sea turtle takes and biennial

dredge hours.” SAR 13392-93. In response to Oceana’s criticisms of this updated ITS, however,

NMFS in its briefing “explicitly disclaim[ed] any reliance on a linear regression model.”

Oceana II,

. Instead of a linear relationship, NMFS’s briefing said that the

scatterplot was meant to illustrate only the “positive relationship between commercial dredge

hours and estimated sea turtle takes,” and that the scatterplot was “not intended to be a predictive

model whereby sea turtle takes can be predicted based on a certain level of dredge effort.”

(emphasis added). Because “the revised ITS advance[d] a rationale now expressly

disclaimed by the agency,” the Court found that it “fail[ed] to explain how 359,757 dredge hours

equals 161 takes,” and therefore “suffer[ed] from the same underlying defect that warranted

remand in the first instance.”

. The Court thus directed the agency “either to more

clearly explain its reliance on the dredge hour surrogate and its selection of 359,757 dredge hours

6 as a surrogate for 161 takes or, if unable to do so, to select a more appropriate surrogate or other

mechanism for monitoring loggerhead takes resulting from dredge fishing.”

NMFS revised the ITS for the 2012 BiOp a second time and now contends that it

has completed its required task on remand by more thoroughly explaining its chosen monitoring

method. See Second Notice of Completion of Remand. Oceana filed a response arguing that the

ITS remains defective and, as a result, NMFS still has failed to demonstrate that its monitoring

methods are not arbitrary and capricious. See Oceana’s Response.

III. LEGAL STANDARD

A BiOp constitutes final agency action subject to judicial review under the

Administrative Procedure Act. See Bennett v. Spear,

(1997). “[W]hen a

party seeks review of agency action under the APA . . . the district judge sits as an appellate

tribunal.” Rempfer v. Sharfstein,

(D.C. Cir. 2009) (quoting Am. Bioscience,

Inc. v. Thompson,

(D.C. Cir. 2001)). The general standard for summary

judgment set forth in Rule 56 of the Federal Rules of Civil Procedure does not apply to a review

of agency action. Summary judgment nonetheless “serves as the mechanism for deciding, as a

matter of law, whether the agency action is supported by the administrative record and otherwise

consistent with the APA standard of review.” Sierra Club v. Mainella,

(D.D.C. 2006) (citing Richards v. INS,

, 1177 & n.28 (D.C. Cir. 1977)); accord

UPMC Braddock v. Harris,

(D.D.C. 2013); Cottage Health Sys. v.

Sebelius,

(D.D.C. 2009). In other words, “[t]he entire case on review

is a question of law.” Marshall Cnty. Health Care Auth. v. Shalala,

(D.C.

Cir. 1993).

7 Under the APA, a reviewing court shall “hold unlawful and set aside agency

action, findings, and conclusions found to be . . . arbitrary, capricious, an abuse of discretion, or

otherwise not in accordance with law.”

(2)(A). “The arbitrary and capricious

standard is deferential; it requires that agency action simply be ‘reasonable and reasonably

explained.’” Cmtys. for a Better Env’t v. EPA,

(D.C. Cir. 2014) (quoting

Nat’l Tel. Coop. Ass’n v. FCC,

(D.C. Cir. 2009)); see also Kennecott Greens

Creek Min. Co. v. Mine Safety and Health Admin.,

(D.C. Cir. 2007) (“[The]

standard of review under the arbitrary and capricious test is only reasonableness, not

perfection.”). “[A] court is not to substitute its judgment for that of the agency” if the agency

“examine[d] the relevant data and articulate[d] a satisfactory explanation for its action including

a rational connection between the facts found and the choice made.” Airmotive Eng’g Corp. v.

Fed. Aviation Admin.,

(D.C. Cir. 2018) (quoting Motor Vehicle Mfrs.

Ass’n v. State Farm Mut. Auto. Ins. Co.,

(1983)) (internal quotation marks

omitted).

Furthermore, a court will “give an extreme degree of deference to the agency

when it is evaluating scientific data within its technical expertise.” Cmtys. for a Better Env’t v.

EPA,

(quoting City of Waukesha v. EPA,

(D.C. Cir. 2003))

(internal quotation marks omitted). A court must remain mindful that it reviews an agency’s

scientific judgments “not as the chemist, biologist, or statistician that [the court is] qualified

neither by training nor experience to be,” and thus it may exercise only the “narrowly defined

duty of holding agencies to certain minimal standards of rationality.” Troy Corp. v. Browner,

(D.C. Cir. 1997) (quoting Ethyl Corp. v. EPA,

, 36 (D.C. Cir. 1976)

(en banc)).

8 “The Court’s review, however, must be ‘searching and careful.’” Colorado River

Cutthroat Trout v. Salazar,

(D.D.C. 2012) (quoting Nat’l Envtl. Dev.

Assn’s Clean Air Project v. EPA,

(D.C. Cir. 2012)). “An agency decision is

arbitrary and capricious if it ‘relied on factors which Congress has not intended it to consider,

entirely failed to consider an important aspect of the problem, offered an explanation for its

decision that runs counter to the evidence before the agency, or is so implausible that it could not

be ascribed to a difference in view or the product of agency expertise.’” Cablevision Sys. Corp.

v. FCC,

(D.C. Cir. 2011) (quoting Motor Vehicle Mfrs. Ass’n v. State Farm

Mut. Auto. Ins. Co.,

); accord Agape Church, Inc. v. FCC,

(D.C. Cir. 2013). Just as the Court may not “substitute [its] judgment for that of the agency” to

set aside an agency action, Rural Cellular Ass’n v. FCC,

(D.C. Cir. 2009), it

also may not “affirm an agency decision on a ground other than that relied upon by the agency.”

Manin v. Nat’l Transp. Safety Bd.,

(D.C. Cir. 2011).

In addition, where an administrative agency has been ordered to reconsider or

explain an earlier decision on remand, as is the case here, the agency has an “affirmative duty to

respond to the specific issues remanded” by the Court. See Defs. of Wildlife v. Kempthorne,

No. 04-1230,

, at *12 (D.D.C. Sept. 29, 2006) (first citing Tex Tin Corp. v.

Envtl. Prot. Agency,

(D.C. Cir. 1993); then citing Ass’n of Civilian

Technicians v. Fed. Labor Relations Auth.,

(D.C. Cir. 2004)). The agency

“retains some discretion to determine how it ‘may best proceed to develop the needed evidence

and how its prior decision should be modified in light of such evidence as develops.’” See

at

*11 (quoting Fed. Power Comm’n v. Transcon. Gas Pipe Line Corp.,

(1976)). And the remanding court “may not dictate to the agency the ‘methods, procedures, or

9 time dimension,’ for its reconsideration.” See

(quoting SEC v. Chenery Corp.,

(1947)). Nor may the court demand that an agency reach any particular result. See

(citations omitted). But the Court retains jurisdiction to enforce the terms of its remand order

where the agency does fail to adequately respond. See

at *12 (citing Int’l Union, United

Auto., Aerospace & Agric. Implement Workers of Am., UAW v. Occupational Safety & Health

Admin.,

(D.C. Cir. 1992)).

IV. ANALYSIS

A. Dredge Hour Surrogate and 359,757 Dredge Hours

In its second remand, this Court directed the agency to both (1) “more clearly

explain its reliance on the dredge hour surrogate” and (2) more clearly explain its “selection of

359,757 dredge hours as a surrogate for 161 takes.” Oceana II,

. In its

second notice of completion of remand, NMFS says that it has done just that. Oceana disagrees.

Before delving into the specifics of the parties’ arguments, it is instructive to

highlight two differences between the original ITS with the second revised ITS, as they relate to

incidental takes of loggerheads. First, the loggerhead take limit is expressed slightly differently

between the two. The ITS in the 2012 BiOp said that the anticipated number of incidental takes

from scallop dredge gear was an annual average of 161 loggerhead takes. SAR 13295. In the

second revised ITS, NMFS characterizes the take limit differently, but it is essentially the same

as the limit in the 2012 BiOp. The anticipated number of incidental takes of loggerheads from

scallop dredge gear is expressed as “take[s] of no more than 322 individuals over any

consecutive two-year period in dredge gear.” Incidental Take Statement, amended Nov. 27,

2018 (“2018 revised ITS”) [Dkt. No. 134-2] at 2. NMFS explains that “[t]he multi-year

loggerhead take levels . . . equate to annual average takes of up to 161 individuals in dredge

10 gear” and that the take level was expressed differently in order to correspond with the monitoring

mechanism. See 2018 revised ITS at 2 n. 2.

The second key difference is the number of monitoring mechanisms employed.

While the 2012 BiOp said that it would employ a “combination” of two monitoring approaches –

(1) a “dredge hour surrogate as a measure of actual takes,” and (2) calculating bycatch estimates

using a method in a scientific paper authored by Kimberly T. Murray (the “Murray model”) –

NMFS had not yet finalized how to use the Murray model. See SAR 13302. The ITS in the

2012 BiOp highlighted the limitations of the Murray model and said that “we will continue to

investigate how the methods from Murray (2011) can be used to be [sic] monitor the ITS for sea

turtles in future years.”

So, even though the ITS in the 2012 BiOp said it was using a

“combination” of two monitoring approaches, in reality, it only relied on one: the dredge hour

surrogate.

By contrast, the monitoring mechanisms described in the second revised ITS do in

fact function as a combination of two monitoring methods. In the second revised ITS, NMFS

first says that it will use the Murray (2011) and Warden and Murray (2011) models to “quantify

and monitor sea turtle interactions with dredge gear over a multi-year time scale.” 2018 revised

ITS at 9. Like the original ITS, NMFS again recognized the limitations of these methods.

But this time, the agency found that “these methods are scientifically valid” and had a plan to

rely on these methods for monitoring, saying that the models could “produce a new bycatch

estimate for sea turtles in the dredge fishery every five years.”

NMFS therefore concluded

that the take-limit of 322 loggerheads over a two-year period could serve as a trigger: “[i]f the

next dredge bycatch estimate produced for loggerhead sea turtles indicates that average annual

takes are greater than 322 over a two-year period . . . then consultation would be reinitiated.”

11 In other words, one monitoring surrogate in the second revised ITS is the bycatch estimates that

NMFS will calculate every five years using the Murray models. The second monitoring

surrogate for loggerhead takes is “fishing effort” – specifically dredge hours – as a second

surrogate measure for monitoring actual takes of loggerheads.

The description of the dredge

hour surrogate in the second revised ITS is the same as that in the 2012 BiOp: “[w]e will assume

that the incidental take level for loggerheads exempted by this opinion . . . has been exceeded if,

over any future two-year period, the fishery exceeds the average number of estimated dredge

hours from 2007-2008.”

This dredge hour surrogate can be reviewed at shorter, more routine

intervals than the five-year time frame required for the Murray model monitoring surrogate. Id.

at 10.

1. Dredge Hours as a Surrogate to Monitor the Numerical Take Limit

Oceana argues that the second revised ITS has done nothing to address the

deficiencies that this Court has previously identified. Twice, this Court has directed NMFS to

better explain how its dredge hour surrogate serves as a monitoring proxy for the anticipated

takes of loggerheads from dredge fishing. See Oceana I,

; Oceana II,

. In the 2012 BiOp’s ITS, NMFS had explained that “the theory behind the

dredge hour monitoring surrogate is simple – the more hours spent dredge fishing, the more

turtles will interact with dredge gear.” Oceana I,

(citing SAR 13296,

13299). After remand, the 2015 revised ITS said that “the relationship between dredge hours

and takes [i]s a strong positive linear relationship.” Oceana II,

(internal

citations omitted). But then, in the subsequent briefing before this Court, “NMFS did an about-

face and expressly disclaimed any reliance on a predictive, linear relationship” and in doing so,

“NMFS . . . left things precisely where they were prior to remand – without a sufficient basis to

12 believe that 359,757 dredge hours will serve as an adequate proxy for 161 takes and thus provide

an appropriate mechanism for determining whether and when the take limit has been exceeded.”

Now, in the second revised ITS, NMFS explains that it is using dredge hours as a

surrogate for estimated sea turtle takes because there is a “positive relationship” between the

two. 2018 revised ITS at 10. In other words, “the more time the scallop dredge gear is in the

water during the Mid-Atlantic sea turtle season, the greater the number of sea turtle takes that

will occur in that time and area.” Id. at 11. NMFS has effectively reverted back to the

explanation that it originally provided for the dredge hour surrogate in the 2012 BiOp. See

Oceana I,

(citing SAR 13296, 13299) (“[T]the theory behind the dredge

hour monitoring surrogate is simple – the more hours spent dredge fishing, the more turtles will

interact with dredge gear.”). But this Court has already rejected this positive correlation as

insufficient, crediting Oceana’s point that “[i]n actuality, any number of takes might occur

during a given time period spent dredge fishing.”

The second revised ITS seems to suggest that NMFS thought that the only

problem it needed to address on remand was the inconsistency between the 2015 revised ITS

saying there was a “strong positive linear relationship” between dredge effort and sea turtle take,

SAR 13392, and NMFS having subsequently walked back any defense of a linear relationship in

its briefing before this Court. The second revised ITS includes the same scatterplot graph as the

first revised ITS, just without the fitted regression line. Compare 2018 revised ITS at 11 with

SAR 13393. Disclaiming much of what the first revised ITS said, the second revised ITS

“clarif[ies]” that the “link” between the number of sea turtles taken and the dredge hour

surrogate is a “positive relationship,” but one that is “not intended to be . . . predictive.” 2018

13 revised ITS at 10-11. This purported clarification fails to address what this Court said in

Oceana II: once the agency abandoned its defense of the dredge hour surrogate’s linear

relationship to sea turtle take, NMFS had “left things precisely where they were prior to remand

– without a sufficient basis to believe that 359,757 dredge hours will serve as an adequate proxy

for 161 takes and thus provide an appropriate mechanism for determining whether and when the

take limit is exceeded.” Oceana II,

. In doubling down on the explanation

that the Court already found insufficient, NMFS has failed to address the specific issues

remanded by the Court, as it is required to do. See Defs. of Wildlife v. Kempthorne, No. 04-

1230,

, at *12 (D.D.C. Sept. 29, 2006) (first citing Text Tin Corp. v. Envtl.

Prot. Agency,

(D.C. Cir. 1993); then citing Ass’n of Civilian Technicians v.

Fed. Labor Relations Auth.,

(D.C. Cir. 2004)) (“[A]n agency faced with a

remand order has an affirmative duty to respond to the specific issues remanded.”).

2. 359,757 Dredge Hours as the Benchmark for the Numerical Take Limit

Not only has NMFS failed to explain how its dredge hour surrogate is an adequate

proxy for sea turtle takes generally, NMFS has also failed to improve on its explanation for the

specific dredge hour limit that it set as a benchmark. In the second revised ITS, NMFS starts off

with precisely the same rationale it provided in the 2012 ITS, saying that it chose 359,757 dredge

hours – the two-year average number of dredge hours in the Mid-Atlantic from May through

November of 2007 and 2008 – “because [that time period] is the first two full years after chain

mats were required and the last two years included in the Murray (2011) analysis upon which the

loggerhead sea turtle dredge take estimates in this ITS are based.” 2018 revised ITS at 9; see

also Oceana I,

(citing SAR 13302, 13245). This Court has already said

that “this fails to sufficiently explain how the specific number of dredge hours that NMFS has

14 chosen as a monitoring surrogate adequately serves as a proxy for the numerical take limit of 161

loggerheads.” Oceana I,

.

Both the 2015 and the 2018 revised incidental take statements have elaborated on

the chosen benchmark, by noting that “[f]or loggerheads, the average annual take estimate of 161

individuals in scallop dredge gear considered in this Opinion was . . . specific to the post-chain

mat time frame from September 26, 2006 to 2008 (Murray 2011).” SAR 13391; see also 2018

revised ITS at 9 (same). But simply noting that there were 359,757 dredge hours over a

somewhat similar time period as the data that led to the calculation of an incidental take limit of

161 loggerheads fails to respond to the logical gaps that this Court previously identified and

directed the agency to address: “the agency must address [the] concern regarding the

effectiveness of linking an hour-based surrogate to a numerical take limit, in the context of a

Fishery where conditions change on a continuous basis.” Oceana I,

(noting that “the fact that an hour of dredge fishing has taken place is not automatically

equivalent to observing the occurrence or non-occurrence of a specific number of loggerhead

takes. In actuality, any number of takes might occur during a given time period spent dredge

fishing”).

The agency’s only new defense in the second revised ITS for choosing 359,757

dredge hours as the surrogate for monitoring 322 loggerhead takes biennially is that this effort

level of 359,757 dredge hours “correspond[s] to an estimated take level that we determined

would not jeopardize the continued existence of any sea turtle species.” 2018 revised ITS at 12

(emphasis added); see also Defendants’ Response at 5-6. While this elaboration sounds

promising on its face, it ultimately fails to provide a rational explanation for the chosen effort

15 level. Missing from this defense of the chosen benchmark is the specific “estimated take level”

that NMFS has supposedly determined would not result in jeopardy for sea turtles.

The agency’s briefing in this Court suggests that this undefined “estimated take

level” referenced in the 2018 revised ITS is 161 loggerheads annually or 322 loggerheads

biennially. See Defendants’ Response at 5-6 (referring back to the biennial take limit of 322

loggerheads). If this was what was meant by the agency in the revised ITS, then this statement is

circular and provides no further explanation from what NMFS has said previously. A statement

that the effort level of 359,757 dredge hours “correspond[s]” with the biennial take limit of 322

loggerheads fails to explain how the agency reached the conclusion that these two numbers

“correspond.” Indeed, this is the exact connection that this Court has repeatedly directed the

agency to better explain. See Oceana I,

; Oceana II,

.

At oral argument, counsel representing NMFS said that it is “not necessary that

the agency be able to predict into the future what level of take [might occur] for the purposes of

monitoring and the reinitiation trigger” and that the purpose of monitoring is for the agency to

“assess whether . . . take has occurred, not looking forward and trying to predict whether it will

occur in the future.” Oral Arg. Tr. at 10:14-17; 11:02-04. This argument underscores the fact

that NMFS misunderstood what the Court meant when it asked, “[w]hat good is a surrogate

based on a model that lacks any predictive value?” Oceana II,

. The

agency needs to explain how its chosen monitoring surrogate enables it to assess whether the

incidental take limit has been exceeded in real time. In other words, the monitoring surrogate

must be predictive of the take limit in the sense that the agency must establish how monitoring

the surrogate is an appropriate proxy for the numerical take limit. See Oceana I,

(“[A] surrogate is only as useful as its fit with the actual object of study for which the

surrogate is substituting.”); Wild Fish Conservancy v. Salazar,

(9th Cir. 2010)

(““If a surrogate is used, the agency must articulate a rational connection between the surrogate

and the taking of the species.”). This is the very connection that the Court has said needs to be

better explained and that the agency has repeatedly failed to provide.

NMFS has consistently characterized its dredge hour surrogate as “a surrogate

measure for monitoring actual takes of loggerheads.” 2018 revised ITS at 9; see also SAR

13302. The ITS has set an anticipated take limit of 322 loggerheads biennially, and the agency

proposes to monitor those “actual takes” by means of a surrogate. But “[a] surrogate is only as

useful as its fit with the actual object of study for which the surrogate is substituting.” Oceana I,

75 F. Supp. 3d at 497; see also Wild Fish Conservancy v. Salazar,

(“[A]

numerical cap [for take] is useful only insofar as the action agency is capable of quantifying take

to determine when the trigger has been met.”). NMFS is correct that the relationship between the

numerical take limit and the chosen surrogate does not necessarily need to be linear. See Nov.

27, 2018 Decision Memorandum (“2018 Decision Mem.”) [Dkt. No. 134-1] at 6 (arguing that the

agency does not need to establish a linear relationship between dredge hours and sea turtle

takes); contra Oceana Response at 8 (arguing that the agency’s chosen benchmark – here

359,757 dredge hours – is only relevant if a certain number of dredge hours is “equivalent” to a

certain number of loggerhead takes, specifically 322 biennially). But the agency does need to

establish how the monitoring surrogate it seeks to rely on is an adequate proxy for the numerical

take limit. Just as NMFS has not provided any better explanation for how dredge hours relate to

loggerhead takes generally, see supra section IV(A)(1), NMFS has also not provided a reasoned

basis for why its chosen benchmark of 359,757 dredge hours serves as an adequate proxy for the

17 numerical take limit of 322 loggerheads biennially. See Oceana Reply at 3 (noting that NMFS

has not shown any link between hours fished and either a number or even a range of takes).

****

Two remands and eight years later, the unfortunate reality is that this case is in the

same position as it was when summary judgment briefs were originally filed. NMFS has failed

to explain how its proposed monitoring surrogate of dredge hours is an adequate proxy for the

numerical take limit established in the ITS, generally. See supra section IV(A)(1). And NMFS

has failed to explain specifically how its chosen benchmark for that monitoring surrogate of

359,757 dredge hours is an adequate proxy for the numerical take limit of 322 loggerheads

biennially. See supra section IV(A)(2). Indeed, NMFS has acknowledged its inability to tie

dredge hours to turtle takes other than there being a “positive” relationship between the two, and

it has provided no response to any of the issues this Court has raised over the past two remands

about such a tenuous monitoring surrogate. See 2018 revised ITS at 11.

Accordingly, this Court must remand for a third time. The Court directs NMFS

either to more clearly explain whether there is a correlation between the dredge hour surrogate

and the numerical take limit, and its selection of 359,757 dredge hours or, if unable to do so, to

select a more appropriate surrogate or other mechanism for monitoring loggerhead takes

resulting from dredge fishing. To be clear, the agency cannot present the same rationale in

support of its dredge hour surrogate that it has presented the past three times. This Court has

determined that this explanation is not sufficient. And if the agency cannot adequately address

the concerns that the Court has raised in this opinion and in its two previous opinions on this

issue, then the agency must use a different mechanism to monitor the numerical take limits

established in the ITS. If no credible monitoring mechanism can be identified or devised, then

18 the Biological Opinion loses its legitimacy. See Oceana II,

(“[W]ithout

the ability to monitor incidental takes, [the Section 7] regulatory requirements become

meaningless.”).

B. Another Surrogate or Monitoring Mechanism

When the Court remanded this matter on this single, discrete issue for the second

time, it told the agency that if it could not more clearly explain its dredge hour surrogate, then it

should “select a more appropriate surrogate or other mechanism for monitoring loggerhead takes

resulting from dredge fishing.” Oceana II,

. On remand, the agency’s

efforts to see what other surrogates or monitoring mechanisms may be possible involved NMFS

asking the experts at the Northeast Fisheries Science Center (“NEFSC”) “whether they have

developed any alternative methods for monitoring loggerhead takes in the dredge fishery.” 2018

Decision Mem. at 4. The record indicates, however, that NEFSC’s response was not entirely

responsive to the breadth of the question posed by NMFS and was instead limited to the dredge

hour surrogate: “NEFSC confirmed that there is presently no model available that would allow

us to predict the number of turtle takes based on the number of dredge hours in a future year.”

Id. at 4 (emphasis added). While this response only underscores the problem necessitating a

remand, see supra section IV(A)(1)-(2), at the same time, it suggests that there is still potential

for NMFS and NEFSC to determine that an alternative monitoring method is available.

C. Five-Year Bycatch Estimate as a Monitoring Surrogate

As previously noted, the second revised ITS employs two monitoring

mechanisms. In addition to the dredge hours surrogate just discussed, see supra section IV(A),

19 NMFS also uses the Murray models to calculate bycatch estimates of sea turtle takes from

scallop dredge gear every five years. See 2018 revised ITS at 8-9.

In its briefing before this Court, however, the agency now represents that “NMFS

is unable to use the Murray analysis to predict a specific level of future take in the dredge

fishery” because there are “too few observable takes to allow NMFS to use existing models to

estimate turtle takes in the fishery with a reasonable degree of certainty.” Defendants’ Response

at 5. This is confusing, because it directly contradicts the second revised ITS, which says that

NMFS intends to continue using the five-year bycatch estimates from the Murray (2011) and

Warren and Murray (2011) models, noting that “these two reports are the best available science

for determining sea turtle bycatch rates. Thus, we will continue to utilize the models . . . to

quantify and monitor sea turtle interactions with dredge gear over a multi-year time scale.” 2018

revised ITS at 9. After recognizing the limitations of these models, the agency concluded that

they are still “scientifically valid and should allow us to produce a new bycatch estimate for sea

turtles in the dredge fishery every five years.” Id. at 9. On remand therefore, the agency must

also clarify whether it still finds these models scientifically valid despite their known limitations,

and thus whether the agency will continue to rely on the second part of its monitoring

mechanism for dredge fishing by calculating five-year bycatch estimates from observer data.

D. Notice of Reinitiation of Consultation

On February 20, 2020 – less than one month after the Court heard oral argument

in this case – NMFS filed a notice saying that it had reinitiated consultation for the Atlantic sea

scallop fishery management plan, the subject of the BiOp at issue in this case. See Notice of Re.

of Consult. at 1. In the notice, NMFS explained that the reinitiation of consultation was

prompted by the agency’s realization that, upon discovering a prior data error, the “incidental

20 take surrogate had been exceeded” in two prior years. See id. at 3. NMFS became aware of this

in the spring of 2019 and confirmed it with certainty no later than November 2019, if not earlier.

See February 13, 2020 Memorandum [Dkt. No. 150-1] at 1-2. The governing regulations

unquestionably mandate that NMFS should have reinitiated consultation immediately.

(i)(4). It did not. Instead, the agency internally “considered options regarding

reinitiation of consultation, including whether there was a biological need to reinitiate.” Notice

of Re. of Consult. at 3. Then, “[a]fter the Court issued an Order in November 2019 setting oral

argument,” the agency “decided to pause an official reinitiation determination.”

Even then,

it failed to advise plaintiffs or the Court of its decision.

The most generous interpretation of the record indicates that the agency’s reason

for ignoring its regulatory obligations is that NMFS was unaware of the regulations governing its

own Biological Opinions. The agency says that it wrongly believed that “the reinitiation

regulations provided us with some degree of flexibility to not reinitiate immediately.” See

February 13, 2020 Memorandum at 2-3. The February 13, 2020 memorandum from the

Assistant Regional Administrator for Protected Resources (the consulting agency) to the

Assistant Regional Administrator for Sustainable Fisheries (the action agency) describes staffing

limitations as a primary reason for the initial delay, noting that the agency’s focus was

elsewhere. See

. Then, upon receiving this Court’s Order scheduling oral argument, the

agency “determined that it was prudent and efficient to pause an official reinitiation

determination while final judgment was pending.”

Oceana expresses incredulity at the agency’s purported ignorance of its statutory

obligations. See Oceana Opp. to Notice at 9. But even if the agency’s disregard for its

regulatory obligations was a product of its own ignorance, and not more nefarious reasons, there

21 is no explanation for the agency’s deliberate decision to keep this significant new information

from Oceana and the Court. The agency’s notice makes clear that well before the January 29,

2020 oral argument – indeed, no later than November 2019 – the agency knew that the dredge

hour surrogate, the very surrogate set to be discussed during oral argument, had been exceeded.

See February 13, 2020 Memorandum at 2. Yet, the agency failed to inform either plaintiff or the

Court of this new information before oral argument.

Instead, at oral argument, counsel from the Department of Justice made multiple

representations to this Court that their client agency knew to be wrong. Counsel affirmatively

stated that “the agency . . . has been monitoring the dredge hours [surrogate] since

the . . . implementation of the . . . biological opinion. And since that time it has not exceeded the

trigger” of 359,757 hours. See Oral Arg. Tr. at 8:20-24. Counsel also specifically cited to the

biennial dredge hours calculated each year since the 2012 BiOp that were presented in the

agency’s 2018 Decision Memorandum – the exact numbers that NMFS knew to have been

miscalculated. See Oral Arg. Tr. at 16:14-18; 2018 Decision Mem. at 7. And in response to a

question posed by the Court, counsel confirmed that the dredge hours surrogate had “never

[been] exceeded” and that, if it had been, “that would have triggered the consultation obligation.”

See Oral Arg. Tr. at 9:05-09. Three agency attorneys were present at counsel table and failed to

correct these misstatements. See

.

Counsel for Oceana spent substantial time drafting its briefs in this matter and

preparing for oral argument without the benefit of knowing – as the agency did – that the

surrogate for the numerical take limit had been exceeded. See Oceana Opp. to Notice at 9. The

Court too was denied the opportunity to question counsel for the agency about the effectiveness

22 of its dredge hours surrogate with the benefit of knowing – as the agency did – that the surrogate

for the numerical take limit had been exceeded without the agency realizing it for multiple years.

The Justice Department lawyers before this Court state that they were not aware

until after oral argument of the take limit being exceeded or the failure of the agency to reinitiate

consultation immediately. See Notice of Re. of Consult. at 2 n.2. But attorneys appearing in

Court have a professional responsibility to make reasonable inquiries to ensure that both their

legal and factual representations are supported. Indeed, the agency states: “[D]uring preparation

for the hearing and based on updates and feedback on the hearing itself, it has become clear to us

that upon the numeric exceedance of the ITS trigger, reinitiation of consultation must occur.”

February 13, 2020 Memorandum at 3 (emphasis added). Presumably, the agency’s “preparation

for the hearing” was done in consultation with their litigating counsel, the lawyers from the

Department of Justice.

Furthermore, Oceana is right when it argues that the agency’s hesitation and

deliberation as to whether to reinitiate consultation upon learning that the take limit had been

exceeded is a violation of its statutory and regulatory obligations. See Oceana Opp. to Notice

at 8-9. Once the agency “concluded that the incidental take surrogate had been exceeded” – at a

time that is not specified in the Notice of Reinitiation of Consultation – the agency was obligated

to immediately reinitiate consultation. See

(i)(2)(4); Notice of Re. of Consult.

at 3. At that point, the agency had no discretion to consider “options,” decide “whether there

was a biological need” to reinitiate consultation, or “pause an official reinitiation determination”

in the manner that it did. See Notice of Re. of Consult. at 3.

Ultimately, the notice of reinitiation of consultation does not affect the underlying

merits presently before the Court. It does, however, illuminate the conduct and credibility of the

23 agency, necessitating the need for greater judicial oversight as this case progresses. Accordingly,

the Court will require regular status reports until reinitiation of consultation is complete. Oceana

also asks the Court that it be awarded attorney’s fees for defending against the agency’s

misrepresentations. See Oceana Opp. to Notice at 9 n.6. If Oceana wishes to file a motion for

attorney’s fees or other sanctions it should do so in a separate motion.

V. CONCLUSION

For the foregoing reasons, the Court will remand to NMFS for the third time for

the limited purpose of revising the ITS as it pertains to dredge hour monitoring. The Court

directs NMFS either to more clearly explain whether there is a correlation between the dredge

hour surrogate and the numerical take limit, and its selection of 359,757 dredge hours or, if

unable to do so, to select a more appropriate surrogate or other mechanism for monitoring

loggerhead takes resulting from dredge fishing. The Court also directs NMFS to explain the

discrepancy between its briefing in this Court and the second revised ITS regarding the reliability

of the bycatch estimates for sea turtles from the dredge fishery. An order consistent with this

opinion shall issue this same day.

SO ORDERED.

/s/ ________________________ PAUL L. FRIEDMAN United States District Judge

DATE: October 1, 2020

24