United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued March 19, 2007 Decided June 1, 2007
No. 05-5448
OCEANA, INC.,
APPELLANT
v.
CARLOS GUTIERREZ, IN HIS OFFICIAL CAPACITY AS
SECRETARY OF THE U.S. DEPARTMENT OF COMMERCE, ET AL.,
APPELLEES
Appeal from the United States District Court
for the District of Columbia
(No. 04cv01155)
Eric A. Bilsky argued the cause and filed the briefs for
appellant. James F. Simon entered an appearance.
Andrew Mergen, Attorney, U.S. Department of Justice,
argued the cause for appellees. With him on the brief was John
E. Arbab, Attorney.
Before: GINSBURG, Chief Judge, and RANDOLPH and
KAVANAUGH, Circuit Judges.
Opinion for the Court filed by Circuit Judge RANDOLPH.
2
RANDOLPH, Circuit Judge: Leatherback sea turtles are so
named because of their unusual, rubber-like shell. They are
found throughout the oceans of the world, in the Atlantic,
Pacific, and Indian oceans, the Caribbean Sea, and the Gulf of
Mexico. Leatherbacks are the largest living sea turtles; their
front flippers can span nearly 9 feet, and they range in weight
from 450 to 1,500 pounds. Since 1970, the Leatherback has
been listed as an endangered species pursuant to the Endangered
Species Act.
Pelagic – that is, open ocean – longline fishing poses a
threat to leatherback turtles. This type of fishing in the Atlantic
Ocean and the Gulf of Mexico, and specifically the adequacy of
federal measures to reduce the threat to leatherbacks, is the
focus of this appeal. Longline fishermen concentrate on
swordfish and tuna. They determine where to fish by locating
temperature fronts between cooler and warmer water masses.
Fishing vessels deploy a monofilament line five to forty miles
long across these fronts. The mainline is rigged with hooks
baited with squid or mackerel and float configurations
depending on the targeted species. If the intended catch is
swordfish, the line is put out at dusk and retrieved at dawn; if the
target is tuna, the line is put out at dawn and picked up at dusk.
Leatherback turtles typically feed on jellyfish and are not
attracted to the bait on the longlines. But they are prone to
getting entangled in the lines or becoming foul hooked.
Entangled or hooked turtles can drown if they cannot surface to
breathe. Turtles that disentangle themselves may retain gear
such as hooks or line on their flippers or shoulders. This can
cause them to die either from trauma or by impairing their
swimming and foraging abilities. Between 1992 and 1999,
United States longline fishermen in the Atlantic Ocean – who
account for only five to eight percent of the hooks fished
3
there1 – caught an estimated 6,363 leatherbacks. By one
estimate there are only 27,600 nesting female leatherbacks in the
Atlantic basin.
The Endangered Species Act directs the Secretary of
Commerce and the Secretary of the Interior to determine
whether a species is endangered or threatened. 16 U.S.C.
§ 1533(a)(1). Once a species is so designated, each federal
agency must ensure that “any action authorized, funded, or
carried out” by the agency is not “likely to jeopardize the
continued existence” of the species. Id. § 1536(a)(2). The
federal agency first determines whether any action “may affect
listed species,” 50 C.F.R. § 402.14(a), and, if it may, the agency
initiates a formal consultation with either the Fish and Wildlife
Service or the National Marine Fisheries Service, depending on
the species.
During formal consultation, the Service – here the National
Marine Fisheries Service2 – produces a “biological opinion”
describing how the proposed action will affect the species, 16
U.S.C. § 1536(b)(3)(A), and a statement concerning incidental
1
The Service’s regulations are not binding on international
vessels operating on the high seas and therefore do not apply to the
vast majority of pelagic longline fishing in the Atlantic Ocean. See
Ocean Conservancy v. Gutierrez, 394 F. Supp. 2d 147, 154 n.11
(D.D.C. 2005).
2
The action agency is the Highly Migratory Species
Management Division of the National Marine Fisheries Service’s
Office of Sustainable Fishery. The Service’s Southeast Regional
Office is the expert, or consulting, agency. See Ocean
Conservancy, 394 F. Supp. 2d at 154 n.12.
4
“take” of the species, 50 C.F.R. § 402.14(g)(7).3 The biological
opinion contains a determination regarding whether the
proposed action is likely to jeopardize the continued existence
of an endangered or threatened species, id. § 402.14(g)(4), and,
if the Service finds a likelihood of jeopardy to the species,
“reasonable and prudent alternatives” for the agency to
implement in order to avoid the likelihood of jeopardy. 16
U.S.C. § 1536(b)(3)(A); 50 C.F.R. § 402.14(h)(3). A
“reasonable and prudent alternative” – an RPA – is something
“that can be implemented in a manner consistent with the
intended purpose of the action, that can be implemented
consistent with the scope of the Federal agency’s legal authority
and jurisdiction, that is economically and technologically
feasible, and that the Director believes would avoid the
likelihood of jeopardizing the continued existence of listed
species.” 50 C.F.R. § 402.02.
In its June 2001 biological opinion, the Fisheries Service
determined that pelagic longline fishing in the Atlantic fishery
was likely to jeopardize the continued existence of leatherback
sea turtles. It therefore included an RPA to avoid jeopardy to
leatherbacks while allowing longline fishermen to continue their
operations. The RPA required the closure of the entire
Northeast Distant section of the pelagic longline fishery, an area
approximately due east of New Jersey. Longline vessels fishing
in the remaining open areas were also required to carry dipnets
and line-cutters to minimize entanglement and post-release
mortality of sea turtle bycatch.
3
“The term ‘take’ means to harass, harm, pursue, hunt,
shoot, wound, kill, trap, capture, or collect, or to attempt to engage
in any such conduct.” 16 U.S.C. § 1532(19). The “take” therefore
includes not only the turtles that are killed, but also those that are
hooked or entangled and then let go.
5
The 2001 biological opinion also stated that the Fisheries
Service would conduct a cooperative research program to
develop, modify, and test gear technologies and fishing
strategies to “(1) reduce the likelihood of interactions between
fishing gear and sea turtles and (2) dramatically reduce
immediate and delayed mortality rates of turtles captured in the
fisheries.” The research lasted for three seasons, and took place
aboard commercial longline vessels working in the Northeast
Distant under a scientific research permit. The studies evaluated
the effectiveness of various fishing gear and techniques in
reducing both the sea turtle bycatch and the mortality rate of sea
turtles captured in the fishery and released alive.
The Northeast Distant experiment led the Fisheries Service
to draw several conclusions. The Service found that vessels
could significantly reduce loggerhead and leatherback bycatch
by replacing the industry-wide standard J-hook with circle
hooks.4 Circle hooks reduce turtle bycatch by a range of fifty to
ninety percent, depending on the type of hook, bait, and turtle
involved. The gear removal procedures the experiment tested
were also successful. The post-release mortality rate of
leatherbacks was 13.1 percent, down from the 32.8 percent
estimated for leatherbacks outside the experiment.
Upon completion of the experiment, the Fisheries Service
planned to issue new regulations requiring gear modifications
and to reopen longline fishing in the Northeast Distant. The
proposed rule prohibited vessels from using J-hooks. The rule
also included new requirements for gear removal and handling
to reduce post-release mortality. The rule required that vessels
carry line cutters and dipnets, that vessel operators have
additional handling and release equipment, and that operators
comply with additional guidelines regarding removal of gear.
4
See Ocean Conservancy, 394 F. Supp. 2d at 153 n.8.
6
The regulations and reopening of the Northeast Distant
would constitute “agency action,” and the Endangered Species
Act therefore again required consultation between different
offices within the Fisheries Service. In early 2004, the Service
began assembling a new biological opinion to assess the effects
of reopening the Northeast Distant subject to the proposed
regulations. The Service completed the opinion in June 2004
and, having concluded the action threatened the leatherback,
produced a new RPA. The 2004 RPA consists of four elements:
(1) a reduction in the post-release mortality rate of leatherbacks;
(2) improvement of the monitoring of the effects of the fishery;
(3) confirmation of the effectiveness of the hook and bait
combinations; and (4) management action to avoid long-term
elevations in leatherback takes or mortality.
The 2004 RPA is detailed and specific. The 2004 biological
opinion first establishes a maximum incidental take level. Given
this projected number of takes, the RPA then targets a level of
post-release mortality that would result in a sufficiently low
number of sea turtle deaths. The total number of turtles captured
multiplied by the post-release mortality rate results in the “total
estimated mortality,” which is the estimated total number of
turtles killed by the vessels under the regulatory authority of the
Fisheries Service. The anticipated take for the 2004-2006 period
was 1,981, or 805 leatherbacks in 2004 and 588 thereafter. That
was an increase from the goal in the 2001 biological opinion of
438 per year, but a substantial decrease from the actual takes in
2001 (1,208) and 2002 (962). The anticipated three-year take
starting in 2007 was 1,764, or 588 per year. The Service
expected the post-release mortality rates to decline each year as
fishing crews become better trained in gear removal. The
biological opinion predicted mortality rates of 32.8 percent in
2004, 26.2 percent in 2005, and 19.6 percent in 2006.
Beginning in 2007, the Service expects the post-release
mortality rate to be 13.1 percent – the rate achieved in the
7
Northeast Distant experiment. Thus, the Service expects the
2004 RPA to bring down total estimated mortality of sea turtles
from about 264 per year in 2004 to about 77 per year starting in
2007 and continuing indefinitely.
The 2004 RPA details how to implement the other elements
as well. It requires the Fisheries Service to establish a
comprehensive outreach program to ensure fishermen are
“aware of the safe handling and gear removal requirements of
the proposed action, understand how to use the required gear,
and understand the importance of maximizing gear removal to
maximizing post-release survival of sea turtles.” The Service
must develop and distribute training materials, hold training
workshops, and establish a point of contact for the fishery.
Training and certification is mandatory for vessel captains.
The Fisheries Service must increase observer coverage to at
least eight percent. The 2004 RPA imposes deadlines on the
submission of quarterly reports, and requires the Service to
compile data throughout the year as vessels submit their
logbooks. The RPA also requires the Service to review each
annual and quarterly report as soon as it becomes available and
reevaluate whether the fishery is likely to stay within the
authorized take levels. If the reports indicate that this is
unlikely, the Service “must take protective/corrective action to
avoid long-term elevations in sea turtle takes and ensure that
take levels . . . are not exceeded.” This ongoing monitoring also
applies to the post-release mortality rate: “If fleet-wide gear
removal rates are not sufficient to meet the performance targets
in [the RPA], [the Fisheries Service] must take immediate action
to offset the increased mortality rates and bring overall
anticipated mortality back down to the level specified in the first
element of the RPA.” This may involve closing large areas of
the fishery.
8
Oceana repeats a claim the district judge, Leon, J., rejected
in his thorough opinion. See Ocean Conservancy v. Gutierrez,
394 F. Supp. 2d 147, 164 (D.D.C. 2005). The claim is that the
Fisheries Service acted arbitrarily when it predicted that the
measures it was putting in place would result in a 13.1 percent
mortality rate by 2007 for leatherbacks caught in longlines. The
Northeast Distant experiment achieved this rate with well-
trained crews. The RPA requires training to be available for all
longline crews, but crew members are not required to enroll.
Oceana sees other defects in the 2004 RPA: it lacks an
enforcement mechanism and contains no positive incentives for
compliance. The 2000 RPA had requirements similar to those
in the 2004 RPA, but very few vessels complied because the
requirements were never enforced. Oceana also argues the
Service did not adequately account for “observer effect.” The
2004 RPA requires only eight percent observer coverage, while
in the experiment there were observers on every longline vessel.
As evidence of the Fisheries Service’s unreasonable
expectations, Oceana points to the Service’s observation of
vessels outside the Northeast Distant experiment that were
unable to reach the 13.1 percent mortality rate. During the
2002-2003 fishing season, the Service observed twelve such
vessels, ten of which had participated in the experiment. The
resulting overall post-release mortality rate of leatherbacks was
31.9 percent – significantly higher than the 13.1 percent the
Service expects vessels to attain beginning this year. Yet
according to Oceana, the measures the Service is relying on to
get to the 13.1 percent rate were “substantially present” during
observation period.
The Fisheries Service disagrees with this evaluation. The
twelve vessels observed during the 2002-2003, the Service
notes, “were not subject to any of the requirements imposed on
the experimental [Northeast Distant] vessels and included in the
9
2004 RPA.” The vessels “were not required to remove gear to
meet any specific mortality limits” because they were operating
under the 2001 biological opinion, which “contained no express
limits on mortality levels.” The 2004 RPA also imposed, for the
first time, the threat of large-scale fishery closure. The Fisheries
Service thinks these aspects of the RPA are sufficient to
replicate the 13.1 percent mortality rate it attained in the
experiment.
The Service also disputes Oceana’s assertion that there will
be a wide disparity in performance between the eight percent of
vessels the Service observes and the ninety-two percent it does
not. As government counsel explained at oral argument, the
Service’s regulations require all vessels to keep logbooks that
record every interaction with sea turtles. Cf. 50 C.F.R.
§ 648.7(b)(1)(I). Vessel captains give their logbooks to the
Service, which compiles and analyzes the data. Having an
observer on board helps ensure that a vessel will not falsify
reports regarding interactions with sea turtles. But it does not
follow that the lack of an observer will result in false reports. It
is unlawful for any person to “[f]ail to comply in an accurate and
timely fashion with the log report, reporting, record retention,
inspection, and other requirements of [50 C.F.R.] § 648.7, or
submit or maintain false information in records and reports
required to be kept or filed under § 648.7.” 50 C.F.R.
§ 648.14(a)(4). Captains and crewmembers can be fined up to
$100,000 for each violation, see 16 U.S.C. § 1858(a), and may
be criminally prosecuted, see id. § 1857(1)(A), (I); 18 U.S.C.
§ 1001; accord United States v. Tomeny, 144 F.3d 749, 756
(11th Cir. 1998). In light of these considerations, the Service
believes it was justified in concluding that vessels would reach
the 13.1 percent rate as planned in the RPA schedule.
Like the district court, we cannot find the Service’s
judgment is arbitrary or capricious. See 5 U.S.C. § 706(2)(A).
10
Oceana’s objection is directed not at the RPA’s goal of reducing
the post-release mortality to 13.1 percent, but at the Service’s
determination that the goal could be achieved. That
determination is a prediction resting on the agency’s evaluation
of past performance and its expert judgment about how the
measures it implemented will operate in the future. Agencies,
like legislators, make predictive judgments like this all the time.
Sometimes the predictions are realized; sometimes they are not
and adjustments must be made. So long as the agency’s
judgment is within the bounds of reason courts will uphold it, as
the Supreme Court and this court have done in many analogous
cases. See, e.g., FCC v. WNCN Listeners Guild, 450 U.S. 582,
594 (1981); Time Warner Entm’t Co. v. FCC, 240 F.3d 1126,
1133 (D.C. Cir. 2001); In re Core Commc’ns, Inc., 455 F.3d
267, 282 (D.C. Cir. 2006); Village of Bensenville v. FAA, 457
F.3d 52, 72 (D.C. Cir. 2006).
Oceana’s reference to the twelve vessels the Fisheries
Service observed outside the Northeast Distant experiment is
effectively a claim that far from uncertainty over whether the
2004 RPA will reach its goal, there is concrete evidence that it
will not. But this does not fully take into account the measures
the Service expects will affect the vessels’ practices. The
Service believes the incentives and obligations in the RPA will
change the behavior of the fishing crews. While it its true that
the Service cannot guarantee this will occur, it is also true that
Oceana cannot know that the RPA will fail.
In the event the Fisheries Service is wrong about the
expected level of compliance, the RPA contains a backstop to
ensure that its take and mortality levels of leatherbacks are not
exceeded. The RPA requires ongoing monitoring, and
adjustment in the event that anticipated take and mortality levels
are not met. Section 8.1.4.2 of the RPA mandates that “[i]f
fleet-wide gear removal rates are not sufficient to meet the
11
performance targets in [the RPA], [the Service] must take
immediate action to offset the increased mortality rates and
bring overall anticipated mortality back down to the level
specified in the first element of the RPA.” And since the RPA
already includes hook and gear removal requirements, “the only
remaining way to achieve further reductions in leatherback
mortality in the pelagic longline fishery would be through
closures that reduce fishing effort in areas of high leatherback
bycatch.”
For these reasons the judgment of the district court is
affirmed.
So ordered.