Sierra Club v. National Marine Fisheries Service ( 2024 )

IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF MARYLAND SIERRA CLUB, et al., * Plaintiffs, * v. * Civ. No. DLB-20-3060 NATIONAL MARINE FISHERIES * SERVICE, et al., * Defendants, * and * AMERICAN PETROLEUM INSTITUTE, et al., * Intervenors. * MEMORANDUM OPINION Three years ago, the nonprofit environmental organizations Sierra Club, Center for Biological Diversity, Friends of the Earth, and Turtle Island Restoration Network brought this case against the National Marine Fisheries Service (“NMFS”) and the Assistant Administrator for the National Oceanic and Atmospheric Administration (“NOAA”). ECF 1. According to the plaintiffs, the defendants violated the Administrative Procedure Act (“APA”), 5 U.S.C. § 701 et seq., and the Endangered Species Act (“ESA”), 16 U.S.C. § 1531 et seq., by issuing a flawed biological opinion concerning oil and gas activity in the Gulf of Mexico that underestimated the risks of harm to protected species and took inadequate measures to mitigate those risks. Id. Several entities representing the oil and gas industry—the American Petroleum Institute (“API”), EnerGeo Alliance, the National Ocean Industries Association, and Chevron U.S.A. Inc. (collectively, “the intervenor-defendants”)—intervened as defendants. ECF 19 & 33. Last year, the Court stayed the case at the parties’ request. ECF 154. Now pending before the Court are six motions: the plaintiffs’ motion to lift the stay, ECF 157; the defendants’ motion for remand without vacatur, ECF 108; the intervenor-defendants’ supporting motions for remand without vacatur, ECF 109 & 112; the plaintiffs’ motion for the admission of extra-record evidence, ECF 94; and the plaintiffs’ motion for summary judgment, ECF 93. The motion to lift the stay, motions for remand without vacatur, and motion for the admission of extra-record evidence are fully briefed. ECF 157-1, 161, 162, 163; ECF 108-1, 112- 1, 121, 123, 124, 125; ECF 94-1, 117, 118, 120. No hearing is necessary. See Loc. R. 105.6. For the reasons below, the Court grants the plaintiffs’ motion to lift the stay, denies the defendants’ and intervenor-defendants’ motions for remand without vacatur, and grants the plaintiffs’ motion for the admission of extra-record evidence. I. Background The Gulf of Mexico is home to a range of endangered or threatened marine species protected under the ESA. ECF 65-2, ¶ 43. These include the Rice’s whale (also known as the Gulf of Mexico Bryde’s whale) and the Kemp’s ridley sea turtle, both of whom are at grave risk of extinction. Id. ¶¶ 44–45. The Gulf of Mexico is home to much of the nation’s oil and gas industry as well. Id. ¶ 3. The oil and gas industry conducts extensive extraction activities in land under federal waters known as the Outer Continental Shelf (“OCS”), including a region known as the Gulf OCS that begins about three miles offshore from several Gulf states and extends 200 nautical miles from shore to the outer boundary of the United States’ Exclusive Economic Zone. ECF 65-2, ¶ 34; 43 U.S.C. §§ 1301(a)(2), 1331 et seq.; 48 Fed. Reg. 10,605 (Mar. 14, 1983). The Gulf OCS includes “tens of thousands of active wells, thousands of production platforms, tens of thousands of miles of underwater pipelines,” and a commensurate high volume of vessel trips. ECF 65-2, ¶ 3. A variety of statutes and regulations attempt to harmonize the co-existence of marine life and gas and oil extraction activities in the Gulf. The Outer Continental Shelf Lands Act ( “OCSLA”), 43 U.S.C. § 1331 et seq., regulates the development of the OCS’s oil and gas resources, including the leases the United States extends to private corporations to explore, develop, and produce oil and gas. ECF 65-2, ¶ 34. The Department of the Interior, through its agencies, is responsible for enforcing safety and environmental standards for offshore oil and gas activities. 30 C.F.R. § 550.101; ECF 65-2, ¶ 35. In addition, Section 7 of the ESA applies to federal gas and oil leases extended pursuant to OCSLA. ECF 65-2, ¶ 50. Section 7 mandates that federal agencies ensure that any agency action “is not likely to jeopardize the continued existence of any endangered [] or threatened species or result in the destruction or adverse modification of habitat of such species[.]” 16 U.S.C. § 1536(a)(2); ECF 65-2, ¶ 24. Accordingly, any agency whose action “may affect” ESA protected species must first initiate a consultation with the appropriate wildlife service before taking that action. 50 C.F.R. § 402.14(a); ECF 65-2, ¶ 25. NMFS is the federal agency within NOAA tasked with ensuring that agency action complies with the ESA as applied to marine species. ECF 65-2, ¶ 20. The Assistant Administrator for NOAA leads these efforts. Id. ¶ 21. In a formal consultation, NMFS must ascertain whether the proposed action, in conjunction with the environmental baseline and any cumulative effects, is likely to jeopardize the continued existence of the species or damage their critical habitats. 50 C.F.R. § 402.14(g). At the close of consultation, NMFS sets forth its findings and conclusions in a biological opinion. 16 U.S.C. § 1536(b)(3)(A); 50 C.F.R. § 402.14(h); ECF 65-2, ¶ 27. If NMFS concludes that the action is likely to jeopardize a species or result in harm to its habitat, it must propose “reasonable and prudent alternatives” to avoid those effects. 16 U.S.C. § 1536(b)(3)(A); ECF 65- 2, ¶ 30. Separately, NMFS must determine whether the proposed agency action is likely to incidentally take members of a protected species—where “take” means “to harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect, or attempt to engage in any such conduct,” 16 U.S.C. § 1532(19)—even if the action would not jeopardize the species or its habitat on the whole, and specify the amount or extent of permissible takes. 50 C.F.R. § 402.14(i); ECF 65-2, ¶¶ 31– 33. NMFS has conducted several consultations in recent decades pertaining to federal oil and gas leases in the Gulf OCS. ECF 65-2, ¶¶ 6, 49–53. After one such consultation in 2007, NMFS issued a biological opinion concluding that the proposed oil and gas activities in the Gulf OCS would not jeopardize ESA protected species or habitats. Id. ¶ 52. Notably, NMFS predicted that because the risk of a very large oil spill was low, any related harms to the animals and their habitats would be minimal. Id. ¶ 53. NMFS was wrong: In 2010, the Deepwater Horizon oil rig exploded, releasing nearly five million barrels of oil into the Gulf of Mexico—hundreds of times more oil than the worst-case scenario NMFS predicted in the 2007 biological opinion. Id. ¶ 54. The explosion contaminated over 43,000 square miles of surface waters and over 1,300 miles of shoreline and killed or seriously harmed over 100,000 individuals of species listed as threatened or endangered. Id. ¶ 56. To this day, the affected species and habitats have yet to recover. Id. ¶¶ 54–57. In response, NMFS reinitiated formal consultation to review anew federally authorized OCS oil and gas leases. Id. ¶ 58. After nearly a decade of research and several rounds of litigation, in March 2020 NMFS published a new biological opinion on the impact of oil and gas activities in the region. See ECF 16-4 (the “2020 BiOp” or “BiOp”). The 2020 BiOp concluded that the proposed oil and gas activities in the Gulf would jeopardize the Rice’s whale but that its proposed “reasonable and prudent alternatives” would mitigate the harms. Id. at 624–27. The opinion concluded that none of the other dozens of protected species would be jeopardized by the proposed activity and that their habitats would not be adversely modified. Id. at 624. On October 21, 2020, the plaintiffs filed this action, claiming that the 2020 BiOp— including its environmental analyses, proposed reasonable and prudent alternatives, and take analyses—violates the APA and ESA. ECF 1; ECF 65-2. The plaintiffs assert that the BiOp suffers from several critical shortcomings: failure to adequately correct the flawed analysis that caused NMFS to underestimate the risk of a catastrophic spill like Deepwater Horizon; failure to account for the significant alteration to the endangered species’ habitats caused by the Deepwater Horizon disaster; failure to consider climate-related population shifts that would compound the current threats to the endangered species posed by the proposed leasing activity; and failure to consider certain sublethal harms to, and threats to the recovery of, ESA protected species. ECF 65-2, ¶¶ 63–124. The plaintiffs further allege that the 2020 BiOp’s reasonable and prudent alternatives are inadequate and that its “incidental take” parameters are improper. Id. ¶¶ 125–141; 156–170. As a result, they say, the 2020 BiOp is arbitrary and capricious, constitutes an abuse of discretion, and is not in accordance with Section 7 of the ESA. The plaintiffs seek an order vacating the BiOp, directing NMFS to prepare a “sufficiently protective biological opinion within six months,” and requiring NMFS to comply with “the ESA, APA and every other order of this Court.” Id. at 39. On January 19, 2021—the final full day of the departing presidential administration—the defendants moved to transfer the case to the U.S. District Court for the Southern District of Texas or the U.S. District Court for the Eastern District of Louisiana. ECF 16. In the days that followed, API, EnerGeo Alliance, the National Ocean Industries Association, and Chevron U.S.A. Inc. entered the case as defendants. ECF 19 & 33. The intervenor-defendants joined the motion to transfer. ECF 57 & 59. On May 24, the Court denied the motion. ECF 63. In the months that followed, the defendants and the intervenor-defendants answered the complaint. ECF 73, 76, 77. On May 27, 2022, the plaintiffs moved for summary judgment, ECF 93, and for the admission of extra-record evidence, ECF 94. After extending the briefing schedule multiple times, the Court ordered the defendants to file their first briefs on the two motions by October 26, 2022. ECF 101. The day those briefs came due, the defendants moved to stay the briefing schedule because they would soon be moving for the Court to remand the 2020 BiOp to NMFS without vacating it. ECF 102. On October 25, they had received a letter from the Bureau of Ocean Energy Management (“BOEM”) and the Bureau of Safety and Environmental Enforcement (“BSEE”)— agencies of the Interior Department—requesting to reinitiate consultation on the 2020 BiOp. ECF 102-1, at 7–8; ECF 102-2, 2–4. After briefing and an on-the-record call, the Court granted the motion to stay the briefing schedule as to the plaintiffs’ motion for summary judgment and denied it as to the plaintiffs’ motion to admit extra-record evidence. ECF 106. As they had announced, the defendants and the intervenor-defendants moved for remand without vacatur on November 9, 2022. ECF 108, 109, 112. After briefing and a hearing, on January 17, 2023, the Court referred the case to Magistrate Judge Quereshi for settlement. ECF 133. On July 21, the parties reached an agreement and filed a stipulation to stay proceedings. ECF 147. The Stipulated Agreement to Stay Proceedings had two parts: a series of “WHEREAS” clauses and a series of numbered, operative clauses. Id. Each part was preceded by a recitation that the plaintiffs and the defendants “hereby agree and stipulate as follows,” id. at 1, or “stipulate and agree as follows,” id. at 4, respectively. The “WHEREAS” clauses declared (in relevant part) that BOEM “will” implement three measures to protect the Rice’s whale for the duration of the consultation between BOEM and NMFS. Id. at 3–4. First, BOEM would issue a Notice to Lessees offering non-binding guidance and recommendations on how to protect the Rice’s whale. Id. at 3; ECF 147-1. Second, BOEM would add a stipulation to Gulf of Mexico oil and gas leases, including the then-imminent Lease Sale 261, directing lessees and operators to keep careful watch for Rice’s whales, observe speed restrictions, keep a specified distance from the whales, and document compliance with these and other requirements. ECF 147, at 3; ECF 147-2. Third, “BOEM will exclude” a specified area of the Gulf “from Gulf of Mexico oil and gas lease sales (beginning with Lease Sale 261).” ECF 147, at 4. The agreement then declared that on that basis— “therefore”—the parties agreed to stay the litigation for the duration of the consultation. Id. at 4. Then the operative provisions outlined a timeline for the consultation. The Bureaus “anticipate[d]” providing their consultation package to NMFS by September 1, 2023, and NMFS “agree[d] to conclude the consultation within one year of receiving a complete consultation package from the Bureaus.” Id. NMFS also agreed to “consider and address each of [the plaintiffs’] issues as part of the reinitiated consultation, as appropriate,” but emphasized that it “does not and cannot make any commitment as to the merits of how any particular issue will be addressed or resolved.” Id. at 4 & n.4. In addition, the agreement noted that if the plaintiffs or the defendants “believe[] there is good cause to lift the stay, they may so move the Court” after meeting and conferring with the other parties to the agreement. Id. at 5. The Court invited the intervenor-defendants to file any objections to the agreement. ECF 148. The intervenor-defendants did not object to the agreement’s operative terms. ECF 149, at 1–2. Nevertheless, they “reserve[d] all rights to challenge BOEM’s actions in an appropriate venue and at the appropriate time.” Id. at 3. On August 24, the Court approved the stipulation and stayed the case. ECF 154. The “appropriate time” arrived immediately. The “appropriate venue” turned out to be the U.S. District Court for the Western District of Louisiana. The same day the Court approved the agreement, the intervenor-defendants (accompanied by the State of Louisiana) sued BOEM, challenging the inclusion of the two binding protective measures in Lease Sale 261. Complaint, Louisiana v. Haaland, No. 2:23-cv-01157-JDC-KK (W.D. La. Aug. 24, 2023). Days later, the intervenor-defendants moved for a preliminary injunction. Mot. Prelim. Inj., Louisiana, 2:23-cv- 01157-JDC-KK (W.D. La. Aug. 28, 2023). The plaintiffs intervened in defense of the measures. Order Granting Intervention, Louisiana, 2:23-cv-01157-JDC-KK (W.D. La. Sept. 14, 2023). On September 21, the U.S. District Court for the Western District of Louisiana enjoined BOEM from including the protective measures and ordered the sale to proceed on September 30. Order 30, Louisiana, 2:23-cv-01157-JDC-KK (W.D. La. Sept. 21, 2023). The plaintiffs and BOEM appealed, Notices of Appeal, Louisiana, 2:23-cv-01157-JDC- KK (W.D. La. Sept. 22, 2023), and sought an emergency stay of the injunction, Emergency Motion to Stay, Louisiana v. Haaland, No. 23-30666 (5th Cir. Sept. 22, 2023); Emergency Mot. Partial Stay, Louisiana, No. 23-30666 (5th Cir. Sept. 22, 2023). But while the plaintiffs sought to stay the district court’s injunction of the protective measures, Emergency Motion to Stay, Louisiana v. Haaland, No. 23-30666 (5th Cir. Sept. 22, 2023), BOEM did not. Instead, BOEM challenged only the district court’s order that the sale take place by September 30. Emergency Mot. Partial Stay 16–29, Louisiana, No. 23-30666 (5th Cir. Sept. 22, 2023). The U.S. Court of Appeals for the Fifth Circuit denied the plaintiffs’ motion to stay the injunction but granted BOEM’s, staying the lease sale until November 8. Order, Louisiana, No. 23-30666 (5th Cir. Sept. 25, 2023). Two days later, the plaintiffs moved to expedite the appeal. Emergency Mot. Expedite Appeal, Louisiana, No. 23-30666 (5th Cir. Sept. 27, 2023). The Fifth Circuit granted the motion. Order Granting Mot. Expedite, Louisiana, No. 23-30666 (5th Cir. Sept. 28, 2023). Once again, the plaintiffs argued that the measures to protect the Rice’s whale should remain in the lease sale. And once again, BOEM did not—challenging only the sale timeline. Opening Br. Fed. Defs.- Appellants, Louisiana, 23-30666 (5th Cir. Oct. 6, 2023); Reply Br. Fed. Defs.-Appellants, Louisiana, 23-30666 (5th Cir. Oct. 18, 2023). The Fifth Circuit briefly stayed the injunction to permit briefing and oral argument. One day after oral argument, the Fifth Circuit dismissed the appeal. Op., Louisiana, 23- 30666 (5th Cir. Nov. 14, 2023). The plaintiffs, the court held, did not have standing. Id. And BOEM, the court concluded, had not challenged the injunction anyway. Id. To address BOEM’s concerns about the timeline, the Fifth Circuit modified the injunction, giving BOEM until December 21, 2023 to conduct the sale. Id. With the protective measures thus enjoined, the parties to this case met and conferred four times about lifting this Court’s stay. ECF 157-1, at 9. Now, the plaintiffs have moved to lift it. ECF 157. The defendants oppose the motion. ECF 161. The intervenor-defendants do too. ECF 162. The plaintiffs have replied. ECF 163. II. Motion to Lift the Stay A “District Court has broad discretion to stay proceedings.” Clinton v. Jones, 520 U.S. 681, 706 (1997). This “power to stay proceedings is incidental to the power inherent in every court to control the disposition of the causes on its docket.” Maryland v. Universal Elections, Inc., 729 F.3d 370, 379 (4th Cir. 2013) (quoting Landis v. N. Am. Co., 299 U.S. 248, 254 (1936)). The “burden of showing the necessity for a stay” falls on the party seeking one. Int’l Refugee Assistance Project v. Trump, 323 F. Supp. 3d 726, 731 (D. Md. 2018) (hereinafter “IRAP”). That burden is substantial. “The party seeking a stay must justify it by clear and convincing circumstances outweighing potential harm to the party against whom it is operative.” Id. (quoting Williford v. Armstrong World Indus., Inc., 715 F.2d 124, 127 (4th Cir. 1983)). In August, the Court exercised its discretion to stay the case because the parties agreed to stay the case. ECF 154. The plaintiffs agreed to a stay in reliance on the defendants’ representations, memorialized in their agreement, that three measures to protect the Rice’s whale would be in place for the duration of the stay. ECF 147, at 3–4; ECF 152, at 3. As the defendants represented, the government put those measures in place. But then the intervenor-defendants asked another court to enjoin two of them. And that court did. So the plaintiffs no longer support a stay. ECF 157-1, at 10–12. Because the stay no longer has the support of all of the parties, the stay no longer has the support of the Court. The defendants respond that they did what they could to uphold their end of the agreement. ECF 161, at 17. The Court does not find otherwise. All the Court finds is that some of the measures the plaintiffs were counting on to secure their interests during the stay are no longer in place, and that in the absence of these predicates for their consent to the agreement, there is good cause to lift the stay. The defendants further contend that they “should not now be penalized” for the actions of the intervenor-defendants. Id. Lifting a stay is not a penalty. It is nothing more than a decision to permit the litigation to run its ordinary course. The intervenor-defendants challenge the legal premises of the plaintiffs’ motion to lift the stay. But they have no authority for their positions. The intervenor-defendants seem to think that the plaintiffs bear the burden of justifying proceeding with the litigation. See, e.g., ECF 162, at 12. In fact, the defendants bear the burden of justifying staying the litigation. See IRAP, 323 F. Supp. 3d at 731. The intervenor-defendants seem to think that there is good cause to lift the stay only if a party breached the operative terms of the agreement. In fact, there is no such restriction on the Court’s exercise of its “broad discretion” to manage its docket. See Clinton, 520 U.S. at 706. Even the agreement itself does not limit “good cause” to breach alone. The intervenor- defendants seem to think that the Court cannot consider why the parties consented to the agreement, even though they memorialized the basis for their consent in the document itself. In fact, there is no reason for the Court to feign ignorance about what happened here. The defendants also argue that the factors courts typically consider to determine whether to grant a motion for a stay—judicial economy and the burdens on the parties—counsel in favor of denying the plaintiffs’ motion to lift the stay. See IRAP, 323 F. Supp. 3d at 731. The reinitiated consultation between BOEM and NMFS, they say, “will likely moot, or, at the very least, significantly narrow or clarify, the issues raised by Plaintiffs”—making further litigation over the current BiOp a waste. ECF 161, at 15. Lifting the stay, they also suggest, would prolong the consultative process by diverting the agencies’ resources from reconsidering the BiOp to participating in this case. Id. at 15–16. At this stage of the proceedings, those arguments are better directed to the motion for remand without vacatur than to the motion to lift the stay. If they were sound, then the Court might as well lift the stay and grant the motion for remand. The Court will turn to them below. The defendants and the intervenor-defendants have not put forward “clear and convincing” considerations that establish “the necessity for [the] stay.” See IRAP, 323 F. Supp. 3d at 731. The Court finds that good cause exists to lift the stay. Accordingly, the plaintiffs’ motion to lift the stay is granted. III. Motion for Remand Without Vacatur With the stay lifted, the Court turns to the defendants’ motion for remand without vacatur and the intervenor-defendants’ motions in support.1 Administrative agencies have the authority to reconsider their decisions. See The Last Best Beef, LLC v. Dudas, 506 F.3d 333, 340 (4th Cir. 2007); Makhteshim Agan of N. Am., Inc. v. NMFS, No. PGW-18-961, 2019 WL 5964526, at *2 (D. Md. Oct. 18, 2019) (quoting Friends of Park v. Nat’l Park Serv., No. 13-cv-3453-DCN, 2014 WL 6969680, at *2 (D.S.C. Dec. 9, 2014)). To exercise that authority, an agency facing a legal challenge to one of its decisions may move for the court to remand the matter to the agency without vacating the decision. See Ohio Valley Env’t Coal. v. Aracoma Coal Co., 556 F.3d 177, 215 (4th Cir. 2009) (citing SKF USA, Inc. v. United States, 254 F.3d 1022, 1028 (Fed. Cir. 2001)).2 A motion for voluntary remand without vacatur 1 The plaintiffs and the defendants requested that if the Court were to lift the stay, the Court proceed to rule on the motion for remand. ECF 157, at 4; ECF 161, at 8, 18. By contrast, the intervenor- defendants suggested they would prefer additional briefing. ECF 162, at 16–17. The Court agrees with the plaintiffs and the defendants that additional briefing is unnecessary. No further briefing is required for the Court to recognize that, as the intervenor-defendants suggest, the plaintiffs’ year-old argument that remand would be premature is no longer applicable. The recent briefing on the motion to lift the stay has updated the Court on the other issues the intervenor-defendants identify. 2 The U.S. Court of Appeals for the Fourth Circuit has discussed voluntary remands only a handful of times. Just one case involved the scenario at issue here: a voluntary remand without vacatur where the agency did not confess error. Compare Ohio Valley Env’t Coal. v. Aracoma Coal Co., 556 F.3d 177, 214–15 (4th Cir. 2009), with Am. Paper Inst. v. EPA, 660 F.2d 954, 964–65 (4th Cir. 1981) (voluntary remand with vacatur after confession of error); The Last Best Beef, LLC v. Dudas, 506 F.3d 333, 340 (4th Cir. 2007) (vacatur with confession of error); Sierra Club v. U.S. Army Corps of Eng’rs, 909 F.3d 635, 641 (4th Cir. 2018) (“Sierra Club I”) (voluntary remand with vacatur without confession of error); and Sierra Club v. W. Va. Dep’t of Env’t Prot., 64 F.4th 487, 495 (4th Cir. 2023) (“Sierra Club II”) (discussing the voluntary remand with vacatur in Sierra Club I). The discussion in Ohio Valley is brief. 556 F.3d at 215. In consequence, the Court follows its peers in turning to courts in other circuits. See, e.g., Friends of Park v. Nat’l Park Serv., “should be treated as with any other motion affecting the substantial rights of the plaintiff, by subjecting the government’s position to careful analysis to ensure that the motion is properly supported and justified.” Rahman v. United States, 149 Fed. Cl. 685, 690 (2020). Whether to grant a motion for remand without vacatur is a matter in the Court’s equitable discretion. See Ford Motor Co. v. NLRB, 305 U.S. 364, 373 (1939); Am. Forest Res. Council v. Ashe, 946 F. Supp. 2d 1, 41–46 (D.D.C. 2013). In some circumstances, it is clear how a court should exercise that discretion. When an agency moves for remand without committing to reconsidering the decision the plaintiffs challenge, the court should deny the agency’s motion. See Limnia, Inc. v. U.S. Dep’t of Transp., 857 F.3d 379, 386–87 (D.C. Cir. 2017). By contrast, when an agency moves for remand while confessing to the error the plaintiffs allege, the court generally should grant the motion. See Citizens Against Pellissippi Parkway Extension, Inc. v. Mineta, 375 F.3d 412, 416 (6th Cir. 2004). Otherwise, the question is highly context-sensitive. When the agency intends to reconsider the decision, but does not confess any error, “the reviewing court has discretion over whether to remand.” SKF USA, 254 F.3d at 1029 (citing Sw. Bell Tel. Co. v. FCC, 10 F.3d 892, 896 (D.C. Cir. 1983)). Typically, courts begin by considering whether the agency’s request for reconsidering the decision is “substantial and legitimate,” made in “bad faith or frivolous,” or somewhere in between. See Keltner v. United States, 148 Fed. Cl. 552, 563 (2020) (citations omitted). Then, [w]here the agency’s request is “substantial and legitimate,” a court ordinarily should grant the motion; alternatively, where the agency’s request is in bad faith or frivolous, a court should deny the motion. In between those relative extremes, the case law demonstrates that the trial court has substantial discretion depending on the timing of the government’s motion, its representations regarding the reasons for No. 13-cv-3453-DCN, 2014 WL 6969680, at *2 (D.S.C. Dec. 9, 2014) (citing the Courts of Appeals for the Sixth, Tenth, D.C., and Federal Circuits). a remand, the plaintiff’s factual allegations viewed through the prism of the particular legal issues involved, and the overall fitness and completeness of the administrative record available for the Court’s review. Id. Other courts have emphasized the importance of evaluating whether remand without vacatur would serve the interest of judicial economy or unduly prejudice the plaintiff. See, e.g., Util. Solid Waste Activities Grp. v. EPA, 901 F.3d 414, 436 (D.C. Cir. 2018). No matter how courts label the elements of the inquiry, the fundamental questions are why the agency is moving for remand, whether remand would resolve the dispute more efficiently than litigation, and how remand would impact the plaintiffs and their interests. NMFS has committed to reconsidering the decision the plaintiffs challenge: the 2020 BiOp. ECF 161-1, ¶¶ 3–14. But by its own admission, “NMFS has not confessed error as to the validity of the 2020 BiOp.” ECF 125, at 3. In consequence, the Court has discretion over whether to grant the agency’s motion to remand. See SKF USA, 254 F.3d at 1029. A. The Agency’s Reason for Requesting Remand The agency’s proffered basis for seeking a remand without vacatur is that the Bureaus requested to reinitiate consultation on the BiOp. This reason falls somewhere between the “relative extremes” of “bad faith or frivolous” and “substantial and legitimate.” See Keltner, 148 Fed. Cl. at 563. On the one hand, the defendants’ request is neither frivolous nor in bad faith. The reinitiation of consultation—which NMFS says will result in a new or revised BiOp—is not a frivolous basis for moving for remand. The plaintiffs argue that the timing of the defendants’ motion for remand is suspect. ECF 121, at 27–28. In the abstract, it might be. After all, the defendants waited until the day their first summary judgment brief was due to tell the Court that they intended to move for remand. In context, however, a more charitable explanation is more compelling. The defendants moved for remand the day after the Bureaus requested to reinitiate consultation—that is, as soon as they knew they had reason to move to remand. On the other hand, the defendants’ request nonetheless does not rest on a “substantial and legitimate” concern. See Keltner, 148 Fed. Cl. at 564. “An agency’s concerns are substantial and legitimate where (1) the agency provides a compelling justification for its remand request; (2) the need for finality does not outweigh the justification for voluntary remand presented by the agency; and (3) the scope of the agency’s remand request is appropriate.” Owens & Minor Distrib., Inc. v. United States, 154 Fed. Cl. 349, 353 (2021) (citing Keltner, 148 Fed. Cl. at 564). First, “the agency has not provided a compelling justification for its remand request.” See Owens & Minor, 154 Fed. Cl. at 353 (emphasis in original). True, NMFS has reported that BOEM and BSEE have reinitiated consultation because the Bureaus seek a reevaluation of the oil spill risk analysis and an amendment incorporating the development of conditions of approval related to impact pile driving and transit through the Rice’s Whale Area, ECF 108-2, ¶ 4—the region NMFS believes the species inhabits year-round, ECF 65-2, ¶ 96. And true, the Bureaus have acknowledged that “additional concerns have been raised with respect to the 2020 BiOp,” which consultation would “provide the opportunity for the agencies to reexamine and address . . . as appropriate.” ECF 108-2, ¶ 4. But the “concern” the defendants advance is not “the agency’s concern.” SKF USA, 254 F.3d at 1029 (emphasis added). It is the Bureaus’ concern. And as the defendants and the intervenor-defendants have emphasized repeatedly, the Bureaus are not parties to this case. ECF 149, at 1 (“the Bureau of Ocean Energy Management and Bureau of Safety and Environmental Enforcement—neither of which is a party here”); ECF 161, at 8; ECF 162, at 5. Twice, the defendants have filed declarations in which NMFS could have embraced the Bureaus’ concerns as its own. ECF 108-2; ECF 161-1. Twice, those declarations have reported the Bureaus’ concerns without disclosing whether NMFS shares them, ECF 108-2, ¶ 4; ECF 161- 1, ¶ 3—even after the plaintiffs pressed the agency to do so, ECF 121, at 20. The closest the defendants have come to claiming these concerns as their own has been in briefing, but even then, they did not adopt the concerns. See ECF 125, at 5 (“NMFS (as consulting agency) has not only acknowledged the issues that prompted the Bureaus’ request for reinitiated consultation, but it has also provided a detailed timeline for its review of these issues and its completion of a new or revised BiOp.”). And the Court cannot look to briefing to remedy these omissions from the agency’s affidavits anyway. Owens & Minor, 154 Fed. Cl. at 353–54. True, NMFS has proposed critical habitats for the Rice’s whale and the green sea turtle of its own volition. ECF 161-1, ¶ 5 n.2. But that happened after NMFS moved for remand—it was not part of, let alone the reason for, the reinitiation of consultation on the BiOp. In fact, NMFS does not even commit to addressing those habitat proposals in the BiOp; it says only that its consultation with the Bureaus “is likely” to encompass them. Id. Whether that prediction proves accurate or not, the relevant point is that NMFS did not move for remand on the basis of its own compelling concerns about the BiOp, but rather on the basis of concerns of other, non-party agencies. Second, the Court cannot say that the defendants’ justification for a voluntary remand outweighs the need for finality. “An agency’s professed intent to revisit the challenged decision is a necessary condition to obtain remand, but it is not always a sufficient condition.” Am. Waterways Operators v. Wheeler, 427 F. Supp. 3d 95, 98–99 (D.D.C. 2019). To be sure, the Bureaus’ reinitiation of consultation to update the oil spill risk analysis distinguishes this case from those in which the “only ‘new’ thing before [the agency] is [the] Plaintiff’s Complaint.” Id. at 98. Still, as the U.S. Court of Federal Claims has recognized, there is a significant interest in finality once agency action is ripe for review: The Supreme Court has explained that, with respect to agency action, an “agency must examine the relevant data and articulate a satisfactory explanation for its action including a rational connection between the facts found and the choice made.” When an agency fails to comply with this minimum standard for agency action in the first instance—but the record is otherwise complete and permits judicial review—a party may seek judicial review and a court may order appropriate relief. If the [agency] did not satisfy this minimum standard in the first instance, the Court sees no reason to afford the [agency] a “second bite at the apple” without first reaching the merits of [the plaintiff’s] claim. Keltner, 148 Fed. Cl. at 565 (quoting Motor Vehicle Mfrs. Ass’n, Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983), and then Am. Waterways, 427 F. Supp. 3d at 98). At most, this factor is a draw. Third, even if the Bureaus’ concerns can be imputed to the defendants, the scope of reconsideration is considerably narrower than the scope of the plaintiffs’ challenge. In their supplemental complaint, the plaintiffs challenge eight elements of the 2020 BiOp: (A) the oil spill risk analysis, ECF 65-2, ¶¶ 64–82, (B) the failure to incorporate population and habitat changes since the Deepwater Horizon oil spill, id. ¶¶ 83–89, (C) the failure to account for climate-change related shifts in the population and distribution of threatened and endangered species, id. ¶¶ 90– 94, (D) the analysis of the survival and recovery of several species, id. ¶¶ 95–109, (E) the analysis of impacts on critical habitats, id. ¶¶ 110–20, (F) the conclusion that a variety of coral species will not be harmed, id. ¶¶ 121–24, (G) the reasonable and prudent alternatives, id. ¶¶ 125–33, and (H) the incidental take statement, id. ¶¶ 134–42. See also ECF 121, at 22. The Bureaus reinitiated consultation to address only one of these issues (the oil spill analysis) and another concern of their own. The defendants say that because the plaintiffs have been “singularly focused on the risk of oil spills,” the consultation will address the plaintiffs’ “core” concern. ECF 125, at 6. The complaint says otherwise: “Chief among NMFS’s errors are its failures to correct the admittedly flawed oil spill analyses from the 2007 biological opinion and to incorporate post-Deepwater Horizon population and habitat changes into the jeopardy and adverse modification analyses.” ECF 65-2, ¶ 63 (emphasis added). In any event, there is no formal hierarchy among the plaintiffs’ claims. To be sure, the Bureaus also “state” that “additional concerns have been raised with respect to the 2020 BiOp.” ECF 108-2, ¶ 4. But even if those “additional concerns” include the plaintiffs’, the fact that the Bureaus “acknowledge that this reinitiation of consultation will provide the opportunity for the agencies to reexamine and address those issues, as appropriate,” id., hardly indicates that they intend to reconsider those issues. Even the phrase “address those issues” is ambiguous between something like “consider a new stance” and something like “provide a new justification for our previous stance.” Courts consistently have condemned the latter as an inadequate basis for remand. See, e.g., Keltner, 148 Fed. Cl. at 566. And the caveat “as appropriate” effectively disclaims any commitment to addressing the plaintiffs’ concerns. Of course, “[t]he fact that Federal Defendants have not committed to specifically addressing each of Plaintiffs’ claims on remand does not bar remand.” N. Alaska Env’t Ctr. v. Haaland, Nos. 3:20- cv-187-SLG & 3:20-cv-253-SLG, 2022 WL 1556028, at *4 (D. Alaska May 17, 2022). But here, at least, it makes the agency’s request less persuasive. See Ctr. for Biological Diversity v. Ross, 419 F. Supp. 3d 16, 23 (D.D.C. 2019) (denying motion to stay proceedings because NMFS could not commit to addressing the plaintiffs’ concerns with challenged BiOp); Conservation Council for Haw. v. NMFS, 97 F. Supp. 3d 1210, 1231–32 (D. Haw. 2015) (declining to dismiss or stay ESA claims despite reinitiated consultation where NMFS did not acknowledge or commit to address any specific issues in the challenged BiOp). Making matters worse, the defendants freely admit that they “may disagree with several aspects of Plaintiffs’ claims,” ECF 108-1, at 14— suggesting that they may have already prejudged some of the issues the plaintiffs raise and may not resolve them. That narrows the actual scope of reconsideration even further, making the case for remand even less compelling. See Keltner, 148 Fed. Cl. at 566. “The Keltner test is conjunctive. If the [government] had failed on any one element, the Court would be justified in denying remand.” Owens & Minor, 154 Fed. Cl. at 355. Here, the Court has serious doubts about all three. Because the agency’s concerns fall between “substantial and legitimate” and “bad faith or frivolous,” “the case law demonstrates that the trial court has substantial discretion” over the motion for remand—and reason to exercise it by denying the motion. See Keltner, 148 Fed. Cl. at 563. B. Judicial Economy Voluntary remands can enable “agencies to cure their own mistakes rather than wasting the courts’ and the parties’ resources reviewing a record that both sides acknowledge to be incorrect or incomplete.” Ethyl Corp. v. Browner, 989 F.2d 522, 524 (D.C. Cir. 1993). The defendants argue that is the case here. A voluntary remand, they say, would enable them to revise the BiOp as quickly as possible, addressing the plaintiffs’ challenge faster and at a lower cost than litigation. But when, as here, the agencies do not admit that they have made any mistakes, nor acknowledge that the record is incorrect or incomplete, the Court cannot assume that remanding the matter to the agency actually advances judicial economy. Instead, the Court must make its own assessment. The Court is persuaded by the reasoning of Ross that judicial economy does not warrant remanding this matter to NMFS. 419 F. Supp. 3d 16 (D.D.C. 2019). There, as here, environmental organizations challenged an NMFS BiOp. Id. at 18–19. After the plaintiffs moved for summary judgment, NMFS moved to stay the case on the ground that the agency intended to issue a replacement BiOp in about one year that would moot the plaintiffs’ claims. Id. at 19, 23. In response, the plaintiffs argued that there was no guarantee that NMFS would issue the new BiOp so swiftly, and that even if it did, there was no guarantee that the new BiOp would actually address their objections. Id. at 23. The court concluded that pausing the litigation while the agency revised the BiOp would not serve the interest in judicial economy. “Even if NMFS is able to comply with its [schedule] for the new BiOp, it would have to fully remedy each of Plaintiffs’ claims outlined above in order to render them moot,” the court observed. Id. It further noted that NMFS could not, “of course . . . commit to the conclusion it [would] reach in that BiOp. In addition, it [was] far from clear that NMFS [would] be able to stick to this schedule.” Id. So too here. NMFS’s latest representation is that it will issue a new or revised BiOp in about one year, ECF 161-1, ¶ 5— two years from when the plaintiffs moved for summary judgment. It is not clear that NMFS will meet that timeline. NMFS concedes that it took seven years to complete the 2020 BiOp, and its consultation process is already months behind the schedule it previously outlined. See ECF 108- 2, ¶ 5. And, NMFS has not committed to reconsidering, much less modifying, every element of the BiOp that the plaintiffs challenge. So here, as in Ross, “[t]he Court is not convinced that it should exercise this discretion” to make the plaintiffs wait for NMFS to reconsider the BiOp. 419 F. Supp. 3d at 23. Two considerations cast further doubt on the defendants’ contention that a voluntary remand would resolve this case more efficiently than a decision on the merits. First, the Court already stayed this case to enable the parties to resolve this dispute without the time and expense of litigation. The result was more litigation, not less. Not only has this case resumed, but the plaintiffs, defendants, and intervenor-defendants incurred the costs of an additional, related case halfway across the country and an ensuing appeal. Second, as that history suggests, there is a substantial risk of further litigation on these same questions no matter how the defendants revise the BiOp. If the defendants do not address all of the plaintiffs’ concerns, then the plaintiffs will likely challenge the revised BiOp on the same grounds they challenge the 2020 BiOp. ECF 121, at 25 (“Conservation Groups will almost certainly need to return to this Court to litigate the vast majority, if not all, of the same legal and analytical errors in a new biological opinion that are already teed up here, with more premerits briefing and proceedings.”). If the defendants do address the plaintiffs’ concerns, then the intervenors may challenge the revised BiOp from the other direction—just as they challenged the protective measures BOEM put in place in accordance with the terms of the stay agreement. With all the parties here, the administrative record submitted, and the issues properly presented, the most efficient course is to proceed to the merits and determine whether the 2020 BiOp complies with the law. The defendants argue that remanding the matter is inherently more efficient than proceeding to the merits because even if the plaintiffs prevail, the remedy will include a remand to NMFS to issue a new or revised BiOp anyway. ECF 108-1, at 14–15; ECF 161, at 22. By granting the motion for voluntary remand without vacatur, the defendants say, “the same relief can be provided now.” ECF 161, at 22. But as the preceding discussion reveals, that argument rests on a series of dubious premises: that NMFS will make the same changes whether the Court remands the matter before a judgment on the merits or after; that if the motion for a voluntary remand is granted, NMFS will resolve the plaintiffs’ concerns; that if NMFS resolves the plaintiffs’ concerns, the intervenor-defendants will not bring a new case challenging the new BiOp; and that NMFS will work as quickly on its own as it would on a court-ordered timeline. The Court cannot make those assumptions. Judicial economy weighs in favor of denying the motion to remand without vacatur. C. Undue Prejudice to the Non-Moving Party “In deciding a motion to remand, [the Court] consider[s] whether remand would unduly prejudice the non-moving party.” Util. Solid Waste, 901 F.3d at 436 (citing FBME Bank Ltd. v. Lew, 142 F. Supp. 3d 70, 73 (D.D.C. 2015)). The plaintiffs argue that a voluntary remand without vacatur would prejudice them in two ways: by exposing imperiled species, particularly the Rice’s whale, to serious harm, ECF 121, at 29–34, and by unreasonably delaying the resolution of their claims on the merits, id. at 34. To the first point, the defendants respond that the BiOp is sufficiently protective and that if it were not, the plaintiffs could have sought a preliminary injunction—but did not. ECF 125, at 10–12. To the second point, the defendants respond that the plaintiffs “presume an entitlement to a decision on the merits where none exists.” Id. at 9. First, the Court declines to find that the harm to protected species would unduly prejudice the plaintiffs. Resolving the dispute between the parties on this point all but requires reaching the merits of the case. If the plaintiffs are right that the BiOp unlawfully exposes these species to harm, then presumably permitting that harm to continue unduly prejudices the plaintiffs. But if the defendants are right that the BiOp is lawful and sound, then presumably the plaintiffs’ interest in the welfare of these species is not unduly prejudiced by the BiOp’s persistence. The Court cannot take the plaintiffs’ claims as true at this stage of the litigation. But the Court cannot rely on the government’s claims either: “The problem with relying upon [the BiOp’s] protective measures to justify voluntary remand without vacatur is that Plaintiffs allege that the [measures] relied upon in the . . . BiOp’s no jeopardy finding are inadequate and unlawful.” Nat. Res. Def. Council v. Norton, No. 1:05-CV-01207 OWW LJO, 2007 WL 14283, at *10 (E.D. Cal. Jan. 3, 2007). In addition, even if the plaintiffs are right about the dangers of the status quo, the fact that they have not sought a preliminary injunction undermines their assertion that they would be unduly prejudiced by the persistence of the status quo on remand. See Makhteshim Agan, 2019 WL 5964526, at *4. The Court declines to determine whether the plaintiffs would be unduly prejudiced on this basis.3 Second, the plaintiffs would be unduly prejudiced by further delay. Courts are divided on whether delay alone constitutes undue prejudice to the plaintiffs. Some have held that “delay of potentially unnecessary litigation is not the sort of undue prejudice that defeats a request for voluntary remand.” Friends of Animals v. Williams, 628 F. Supp. 3d 71, 78 (D.D.C. 2022). But others have held that remand would prejudice the plaintiffs simply because the question the agency sought to remand “necessarily implicate[d]” the plaintiffs’ claim and the plaintiffs opposed the remand. Util. Solid Waste, 901 F.3d at 436. Context is key. On the distinctive facts of this case, the Court finds that further delay would unduly prejudice the plaintiffs. The plaintiffs challenged the 2020 BiOp more than three years ago. They moved for summary judgment over 18 months ago. In a good-faith effort to resolve this dispute, they agreed to a stay that postponed judicial resolution of their claims. When another court enjoined several of the protective measures the plaintiffs relied on to protect their interests in the interim, the government did not appeal that ruling on the merits. At this point, the “Plaintiffs are entitled to have their complaint decided on the merits, particularly given the fact that Defendants continue to rely on the challenged BiOp[].” See Nat. Res. Def. Council, 2007 WL 14283, at *16. 3 Several of the intervenor-defendants argue that the plaintiffs lack standing to seek “any relief based on alleged harms to the Rice’s whale . . . and, therefore, Plaintiffs suffer no prejudice from remand without vacatur.” ECF 110, at 22. Yet they cite no case conducting, much less requiring, an independent standing analysis prior to a finding that voluntary remand without vacatur would unduly prejudice the plaintiffs. Because the Court does not determine whether the plaintiffs would be unduly prejudiced by harm to the Rice’s whale during the consultation period, this argument is irrelevant. The Court finds that a voluntary remand without vacatur would unduly prejudice the plaintiffs by gratuitously delaying the resolution of their claims. * * * The Court has significant discretion over whether to grant the motion for remand without vacatur. After due consideration, the Court exercises that discretion to deny the motion. The basis for the defendants’ motion is not frivolous or suspect, but it is not substantial and legitimate either. A remand would not serve the interest in judicial economy. And remand would unduly prejudice the plaintiffs. For these reasons, the motions for remand without vacatur are denied. IV. Motion to Admit Extra-Record Evidence Finally, the Court considers the plaintiffs’ motion for the admission of the declaration of Susan Lubetkin, PhD, as extra-record evidence, ECF 94. Generally, courts must review agency action for compliance with the APA based on the administrative record alone. See Camp v. Pitts, 411 U.S. 138, 141–42 (1973). However, courts may consider evidence outside the record in limited circumstances, including when that evidence “(1) explains technical information or agency action not adequately explained in record; (2) shows an agency failed to consider relevant evidence; or (3) shows an agency, in bad faith, failed to include information it considered in the record.” Hodges v. Abraham, 253 F. Supp. 2d 846, 855 (D.S.C. 2002), aff’d, 300 F.3d 432 (4th Cir. 2002). “A party challenging an agency bears a special burden of demonstrating that the court should reach beyond the record.” Sanitary Bd. of City of Charleston, W. Va. v. Wheeler, 918 F.3d 324, 334 (4th Cir. 2019). Whether to admit extra-record evidence pursuant to an exception is a matter within the Court’s discretion. Piedmont Env’t Council v. U.S. Dep’t of Transp., 159 F. Supp. 2d 260, 270 (W.D. Va. 2001) (citing Valley Citizens for a Safe Env’t v. Aldridge, 886 F.2d 458, 460 (1st Cir. 1989)). The plaintiffs argue that Dr. Lubetkin’s declaration falls into the first and second exceptions: explaining technical information and illuminating how the agency failed to consider relevant evidence. ECF 94-2, at 6–13. They explain that the defendants’ oil spill risk analysis— an element of the BiOp they challenge—rested at least in part on the defendants’ interpretation of a study, Statistics of Extremes in Oil Spill Risk Analysis, 48 Env’t Sci. & Tech 10,505 (2014), by Zhen-Gang Ji, et al. ECF 94-2, at 6; see also AR13997; AR 0067337–0067368. In her declaration, Dr. Lubetkin explains the terminology, methods, and results of that study, ECF 93-4, ¶¶ 4–30, and then, assuming that the study’s data, methods, and results are correct, she identifies their mathematical implications for the likelihood of particular spills in the 50-year window at issue in the BiOp, id. ¶¶ 31–52. Her main conclusion is that, according to the Ji study, there is a 26.2 percent chance that an oil spill of greater than one million barrels will occur in the next 50 years. Id. ¶ 52. To the plaintiffs, the fact that the Ji study has this implication undermines the rationality of the BiOp’s conclusion—based on the same study—that the risk of such a large oil spill was too low to matter. ECF 94-2, at 2. The first substantive portion of Dr. Lubetkin’s declaration, ECF 93-4, ¶¶ 4–30, falls well within the exception for the explanation of technical information. A couple of cases illustrate why. In Tafas v. Dudas, 511 F. Supp. 2d 652 (E.D. Va. 2007), a patent case, the plaintiff submitted the declaration of an expert in patent law. Id. at 661–63. The defendant agency moved to strike the declaration in part on the ground that it exceeded the scope of the administrative record. Id. at 662. Even though the declaration “contain[ed] analysis that [went] beyond mere explanation of the administrative record,” the court granted the motion to admit it. Id. at 662. As the court explained, “given that the subject matter is difficult and there are a number of terms used uniquely in the patent context, the Court finds that some explanatory material is appropriate.” Id. Similarly, in National Wildlife Federation v. National Marine Fisheries Service (“NWF”), No. 3:01-CV- 00640-SI, 2015 WL 423090 (D. Or. Feb. 2, 2015), another APA and ESA challenge to an NMFS BiOp, the plaintiff environmental organization moved to admit hundreds of pages of extra-record declarations and exhibits in support of its motion for summary judgment. Id. at *1. Even though the court had already appointed an expert to assist in its decision-making, the court granted the motion for the admission of the extra-record evidence to “explain complex or technical aspects of the 2014 BiOp.” Id. at *5. In light of the case’s voluminous and complex record, “the Court [found] it reasonable to permit a small number of additional experts to assist in explaining technical or complex information.” Id. at *4. The same is true here. The Ji study is a peer-reviewed statistical publication replete with technical vocabulary and sophisticated modeling. The first portion of the Lubetkin declaration defines the Ji study’s key terms, explains its methodology, and illustrates how it works with accessible examples. In light of the complexity of the subject matter, the opacity of the relevant vocabulary and methods, and the absence of a court-appointed expert, the Court has no trouble finding that the first portion of the Lubetkin declaration is admissible under the exception for evidence that clarifies technical information. The defendants barely challenge that conclusion. The only argument the defendants direct to this exception is that it is “up to the agency,” not the plaintiffs, to provide any supplemental technical explanations the Court might need. ECF 118, at 6, 8–9. However, the cases the defendants cite for that claim say nothing like it. Instead, these cases hold only that despite the general principle disfavoring post-hoc rationalizations of agency action, an agency may supplement the administrative record with material that helps explain the original record. See, e.g., Yale-New Haven Hosp. v. Leavitt, 470 F.3d 71, 82 (2nd Cir. 2006) (“We therefore hold that to the extent that an agency may supplement the record on judicial review of the validity of a rule that is interpretive, it may do so only if the proffered evidence illuminates the original record and does not advance new rationalizations for the agency’s action.”); Sierra Club v. Marsh, 976 F.2d 763, 772–73 (1st Cir. 1992) (“The administrative record may be supplemented, if necessary, by affidavits, depositions, or other proof of an explanatory nature. The new material, however, should be explanatory of the decisionmakers’ action at the time it occurred. No new rationalizations for the agency’s decision should be included . . .”) (internal quotation marks and citations omitted). These cases provide no reason to think that only the agency may introduce extra-record evidence to clarify technical matters. For their part, the intervenor-defendants argue that the Lubetkin declaration is unnecessary to illuminate the Ji study because the math is easy and the study explains its methods anyway. ECF 117, at 5–6. The Court disagrees. The Ji study’s methods are so complex that the authors needed to include a five-page section titled “Mathematical Method” just to explain their work to the sophisticated audience of the journal in which they published. AR0067344–AR0067348. While that explanation may be suitable for that expert audience, Dr. Lubetkin’s explanation makes the study more accessible to the Court. The second portion of the Lubetkin declaration, ECF 93-4, ¶¶ 31–52, is admissible under the second exception—that is, as potentially “illustrat[ing] factors [NMFS] should have considered but did not.” Piedmont Env’t Council, 159 F. Supp. 2d at 270. Courts have regularly admitted declarations like this one for that purpose. In Dyncorp International, LLC v. United States, 125 Fed. Cl. 1 (Fed. Cl. 2016), for example, the court admitted a similar extra-record declaration that “did not introduce extra-record facts, but instead included calculations and explanations based on data in the” administrative record “to aid the Court in better understanding the record.” Id. at 3.4 NWF is much the same. In addition to admitting the declarations at issue there to explain complex scientific questions, the NWF Court admitted the extra-record evidence because “some of the information in the declarations may [have been] necessary to identify factors [NMFS] did or did not consider.” 2015 WL 423090, at *4. The logic of these cases calls for admitting the declaration of Dr. Lubetkin. The plaintiffs argue that the data and methods of the Ji study imply that the defendants failed to consider two important factors: the actual probability of a one-million-barrel oil spill and the risk of other large spills over the 50-year window relevant to the BiOp. ECF 94- 2, at 8–9. Dr. Lubetkin’s declaration would aid the Court in evaluating that argument. Once again, the defendants do not specifically rebut the plaintiffs’ contention that the declaration qualifies for this exception. The defendants assert that the plaintiffs have failed to overcome the “presumption of regularity” that attends agency designation of the administrative record. ECF 118, at 4. But the presumption they allude to is a presumption that the administrative record is the complete basis for the agency’s decision. See Kravitz v. U.S. Dep’t of Com., 336 F. Supp. 3d 545, 570 (D. Md. 2018). It is a presumption against permitting discovery to identify omitted or withheld bases for the agency’s actions, see id., not a presumption against admitting extra-record evidence to evaluate whether the agency failed to consider some factor it should have considered. The defendants also emphasize that it is up to NMFS, not the Court, to weigh the scientific evidence. ECF 118, at 7. Of course, it is. But even the defendants acknowledge that it 4 The intervenor-defendants contend that Dyncorp is irrelevant because the case concerned “the U.S. Air Force’s procurement decision and the alleged competitive harm that a party incurred due to the U.S. Air Force’s actions.” ECF 117, at 10. The case belies that claim. As the Dyncorp court explained, “the standards set forth in the Administrative Procedure Act govern this Court’s review of a protest involving an agency’s procurement action.” Dyncorp Int’l, LLC v. United States, 125 Fed. Cl. 1, 2 (Fed. Cl. 2016). is the role of the Court “to determine whether the agency acted rationally.” Id. (citing Manufactured Housing Inst. v. EPA, 467 F.3d 391, 399 (4th Cir. 2006)). Evaluating whether an agency acted rationally requires determining whether it “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.” Appalachian Voices v. U.S. Dep’t of Interior, 25 F.4th 259, 269 (4th Cir. 2022) (quoting Motor Vehicle Mfrs. Ass’n, Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983)). That is the purpose for which the plaintiffs move for the admission of the declaration. The defendants also insinuate that Dr. Lubetkin’s declaration is nothing more than an expression of her disagreement with the “correctness or wisdom” of the BiOp. ECF 118, at 7 (quoting Price Road Neighborhood Ass’n v. U.S. Dep’t of Transp., 113 F.3d 1505, 1511 (9th Cir. 1997)). That is simply not the case. Nowhere in the declaration does Dr. Lubetkin criticize the BiOp. In fact, nowhere in the declaration does Dr. Lubetkin even criticize the Ji study. As her conclusion states, all her declaration does is identify the Ji study’s own mathematical implications for the risk of certain oil spills in a 50-year period of time. ECF 93-4, ¶ 52. The intervenor-defendants offer a few additional arguments, but they are no more persuasive. To start, they argue that admitting the declaration of Dr. Lubetkin would convert what should be a deferential assessment of the rationality of the BiOp into a “battle of the experts” decided de novo. ECF 117, at 8–9 (quoting Native Ecosystems Council v. Erickson, 330 F. Supp. 3d 1218, 1231 (D. Mont. 2018), aff’d, 804 F. App’x 651 (9th Cir. 2020)). The case they rely on suggests the opposite. There, the court concluded that considering the declarations at issue would result in a “battle of the experts” because the declarants “attack[ed] the science underlying the Forest Service’s decisions”—forcing the court to weigh competing scientific claims and decide who was right. Native Ecosystems Council, 330 F. Supp. 3d at 1231. Here, by contrast, the plaintiffs move to admit a declaration that expressly assumes that the science the government relied on—the Ji study—is correct but identifies an additional upshot of that science. Admitting the declaration would not compel the Court to decide whether the Ji study or Dr. Lubetkin’s declaration is superior. Rather, it would help the Court assess whether the government failed to consider something it should have. The intervenor-defendants also argue that American Wildlands v. Kempthorne, 530 F.3d 991 (D.C. Cir. 2008), stands for the proposition that “a difference in opinion as to how the same set of data should be interpreted is insufficient to meet the high burden of showing that relevant evidence was not considered.” ECF 117, at 9. This argument misses the mark in two ways. For one, the plaintiffs do not move for the admission of Dr. Lubetkin’s declaration because it offers a rival interpretation of the Ji study; they move because it purports to identify an implication of the Ji study that NMFS did not consider but should have. For another, the intervenor-defendants misread Kempthorne. In Kempthorne, the plaintiffs challenged the decision of the Fish and Wildlife Service (“FWS”) not to list a particular fish as a threatened species under the ESA. 530 F.3d at 994. The plaintiffs moved to supplement the record with two letters from scientists whose work FWS relied on in reaching that decision. Id. at 1002. The U.S. Court of Appeals for the District of Columbia Circuit affirmed the district court’s denial of the motion, concluding that the letters merely expressed the scientists’ disagreement with FWS’s findings—and so did not qualify for any of the exceptions to the rule that review should focus on the administrative record. Id. However, contrary to what the intervenor-defendants assert, the reason the court found that the letters “merely disagree[d] with the [FWS’s] conclusions” was not that the letters offered a novel interpretation of data in the record. Id. The reason was that the letters said things like, “[W]e disagree with the [Service’s] interpretation of our data.” Id. (quoting Letter from N.P. Hitt, Professor, Dept. of Fisheries & Wildlife Scis., Va. Polytechnic Inst. & State Univ., to L.R. Keading, Chief, Branch of Native Fishes Mgmt. (July 10, 2004)) (alterations in Kempthorne). Dr. Lubetkin says nothing like that. So Kempthorne does not counsel against the admission of her declaration. Finally, the intervenor-defendants assert that “there can be no gap in the record or an omission of technical scientific information when the Ji report—the very subject of Dr. Lubetkin’s declaration—is already in the administrative record for the Court’s review.” ECF 117, at 10. That is not quite the question. The question is whether NMFS “failed to consider an important aspect of the problem,” Appalachian Voices, 25 F.4th at 269 (quoting State Farm, 463 U.S. at 43), or in slightly different terms, whether the declaration “illustrate[s] factors [NMFS] should have considered but did not,” Piedmont Env’t Council, 159 F. Supp. 2d at 270. As Dyncorp and other cases show, an extra-record declaration can reveal that sort of flaw even if it does “not introduce extra-record facts, but instead include[s] calcuations and explanations based on data in the” record already, Dyncorp, 125 Fed. Cl. at 3—by revealing that the agency missed something important that was right in front of it. The declaration of Dr. Lubetkin qualifies for the exceptions for extra-record evidence that clarifies technical information and that illustrates a factor the agency failed to consider but should have. In consequence, the Court grants the plaintiffs’ motion to admit it as extra-record evidence. V. Conclusion The plaintiffs’ motion to lift the stay, ECF 157, is granted. The defendants’ and intervenor- defendants’ motions for remand without vacatur, ECF 108, 109, and 112, are denied. The plaintiffs’ motion to admit extra-record evidence, ECF 94, is granted. Accordingly, summary judgment briefing resumes. The defendants shall file concurrently their opposition / opening summary judgment brief, and any motion for the admission of extra-record evidence, no later than February 6, 2024. The intervenors shall file concurrently their opposition / opening summary judgment briefs and any briefing on the admission of extra-record evidence no later than February 16, 2024. The plaintiffs shall file concurrently their reply / opposition summary judgment brief and any briefing on the admission of extra-record evidence no later than March 15, 2024. The defendants shall file concurrently their reply summary judgment brief and any briefing on the admission of extra-record evidence no later than April 12, 2024. The intervenors shall file concurrently their reply summary judgment briefs and any briefing on the admission of extra- record evidence no later than April 22, 2024. The Court will not extend these deadlines save for good cause shown. Any extension request must be accompanied by a supporting affidavit that states the reason for requesting the extension and why the length of the extension is necessary. A separate order follows. Date: January 9, 2024 (G2 — Deborah L. Boardman United States District Judge 32