San Luis & Delta-Mendota Water v. Pacific Coast Federation Etc. ( 2014 )

FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
SAN LUIS & DELTA-MENDOTA No. 12-15144
WATER AUTHORITY; WESTLANDS
WATER DISTRICT; STOCKTON EAST D.C. No.
WATER DISTRICT; METROPOLITAN 1:09-cv-01053-
WATER DISTRICT OF SOUTHERN LJO-DLB
CALIFORNIA; OAKDALE IRRIGATION
DISTRICT; SOUTH SAN JOAQUIN
IRRIGATION DISTRICT; KERN
COUNTY WATER AGENCY;
COALITION FOR A SUSTAINABLE
DELTA; STATE WATER
CONTRACTORS,
Plaintiffs-Appellees,
CALIFORNIA DEPARTMENT OF
WATER RESOURCES,
Intervenor-Plaintiff–Appellee,
v.
GARY LOCKE; UNITED STATES
DEPARTMENT OF COMMERCE;
NATIONAL OCEANIC AND
ATMOSPHERIC ADMINISTRATION;
NATIONAL MARINE FISHERIES
SERVICE; JAMES W. BALSIGER;
RODNEY R. MCINNIS; U.S.
DEPARTMENT OF THE INTERIOR;
UNITED STATES BUREAU OF
2 SAN LUIS V. LOCKE
RECLAMATION; MICHAEL L.
CONNOR; DONALD R. GLASER; JANE
LUBCHENCO; SALLY JEWELL,*
Defendants,
and
THE BAY INSTITUTE; CALIFORNIA
TROUT; FRIENDS OF THE RIVER;
NATURAL RESOURCES DEFENSE
COUNCIL; NORTHERN CALIFORNIA
COUNCIL OF THE FEDERATION OF
FLY FISHERS; SAN FRANCISCO
BAYKEEPER; SACRAMENTO RIVER
PRESERVATION TRUST; WINNEMEM
WINTU TRIBE; PACIFIC COAST
FEDERATION OF FISHERMEN’S
ASSOCIATIONS, INC., Institute for
Fisheries Research,
Intervenor-Defendants–Appellants.
SAN LUIS & DELTA-MENDOTA No. 12-15289
WATER AUTHORITY; WESTLANDS
WATER DISTRICT, D.C. No.
Plaintiffs-Appellants, 1:09-cv-01053-
LJO-DLB
and
*
Sally Jewell is substituted for her predecessor, Kenneth Lee Salazar,
as Secretary of the Interior. Fed. R. App. P. 43(c)(2).
SAN LUIS V. LOCKE 3
STOCKTON EAST WATER DISTRICT;
METROPOLITAN WATER DISTRICT OF
SOUTHERN CALIFORNIA; OAKDALE
IRRIGATION DISTRICT; SOUTH SAN
JOAQUIN IRRIGATION DISTRICT;
KERN COUNTY WATER AGENCY;
COALITION FOR A SUSTAINABLE
DELTA; STATE WATER
CONTRACTORS,
Plaintiffs,
CALIFORNIA DEPARTMENT OF
WATER RESOURCES,
Intervenor-Plaintiff,
v.
GARY LOCKE; UNITED STATES
DEPARTMENT OF COMMERCE;
NATIONAL OCEANIC AND
ATMOSPHERIC ADMINISTRATION;
NATIONAL MARINE FISHERIES
SERVICE; JAMES W. BALSIGER;
RODNEY R. MCINNIS; U.S.
DEPARTMENT OF THE INTERIOR;
UNITED STATES BUREAU OF
RECLAMATION; MICHAEL L.
CONNOR; DONALD R. GLASER; JANE
LUBCHENCO; SALLY JEWELL,
Defendants-Appellees,
THE BAY INSTITUTE; CALIFORNIA
TROUT; FRIENDS OF THE RIVER;
4 SAN LUIS V. LOCKE
NATURAL RESOURCES DEFENSE
COUNCIL; NORTHERN CALIFORNIA
COUNCIL OF THE FEDERATION OF
FLY FISHERS; SAN FRANCISCO
BAYKEEPER; SACRAMENTO RIVER
PRESERVATION TRUST; WINNEMEM
WINTU TRIBE; PACIFIC COAST
FEDERATION OF FISHERMEN’S
ASSOCIATIONS, INC., Institute for
Fisheries Research,
Intervenor-Defendants–Appellees.
SAN LUIS & DELTA-MENDOTA No. 12-15290
WATER AUTHORITY; WESTLANDS
WATER DISTRICT; STOCKTON EAST D.C. No.
WATER DISTRICT; METROPOLITAN 1:09-cv-01053-
WATER DISTRICT OF SOUTHERN LJO-DLB
CALIFORNIA; OAKDALE IRRIGATION
DISTRICT; SOUTH SAN JOAQUIN
IRRIGATION DISTRICT; KERN
COUNTY WATER AGENCY;
COALITION FOR A SUSTAINABLE
DELTA,
Plaintiffs,
CALIFORNIA DEPARTMENT OF
WATER RESOURCES,
Intervenor-Plaintiff,
and
SAN LUIS V. LOCKE 5
STATE WATER CONTRACTORS,
Plaintiff-Appellant,
v.
GARY LOCKE; UNITED STATES
DEPARTMENT OF COMMERCE;
NATIONAL OCEANIC AND
ATMOSPHERIC ADMINISTRATION;
NATIONAL MARINE FISHERIES
SERVICE; JAMES W. BALSIGER;
RODNEY R. MCINNIS; U.S.
DEPARTMENT OF THE INTERIOR;
UNITED STATES BUREAU OF
RECLAMATION; MICHAEL L.
CONNOR; DONALD R. GLASER; JANE
LUBCHENCO; SALLY JEWELL,
Defendants-Appellees,
THE BAY INSTITUTE; CALIFORNIA
TROUT; FRIENDS OF THE RIVER;
NATURAL RESOURCES DEFENSE
COUNCIL; NORTHERN CALIFORNIA
COUNCIL OF THE FEDERATION OF
FLY FISHERS; SAN FRANCISCO
BAYKEEPER; SACRAMENTO RIVER
PRESERVATION TRUST; WINNEMEM
WINTU TRIBE; PACIFIC COAST
FEDERATION OF FISHERMEN’S
ASSOCIATIONS, INC., Institute for
Fisheries Research,
Intervenor-Defendants–Appellees.
6 SAN LUIS V. LOCKE
SAN LUIS & DELTA-MENDOTA No. 12-15291
WATER AUTHORITY; WESTLANDS
WATER DISTRICT; STOCKTON EAST D.C. No.
WATER DISTRICT; METROPOLITAN 1:09-cv-01053-
WATER DISTRICT OF SOUTHERN LJO-DLB
CALIFORNIA; OAKDALE IRRIGATION
DISTRICT; SOUTH SAN JOAQUIN
IRRIGATION DISTRICT; STATE
WATER CONTRACTORS,
Plaintiffs,
CALIFORNIA DEPARTMENT OF
WATER RESOURCES,
Intervenor-Plaintiff,
and
KERN COUNTY WATER AGENCY;
COALITION FOR A SUSTAINABLE
DELTA,
Plaintiffs-Appellants,
v.
GARY LOCKE; UNITED STATES
DEPARTMENT OF COMMERCE;
NATIONAL OCEANIC AND
ATMOSPHERIC ADMINISTRATION;
NATIONAL MARINE FISHERIES
SERVICE; JAMES W. BALSIGER;
RODNEY R. MCINNIS; U.S.
DEPARTMENT OF THE INTERIOR;
UNITED STATES BUREAU OF
SAN LUIS V. LOCKE 7
RECLAMATION; MICHAEL L.
CONNOR; DONALD R. GLASER; JANE
LUBCHENCO; SALLY JEWELL,
Defendants-Appellees,
THE BAY INSTITUTE; CALIFORNIA
TROUT; FRIENDS OF THE RIVER;
NATURAL RESOURCES DEFENSE
COUNCIL; NORTHERN CALIFORNIA
COUNCIL OF THE FEDERATION OF
FLY FISHERS; SAN FRANCISCO
BAYKEEPER; SACRAMENTO RIVER
PRESERVATION TRUST; WINNEMEM
WINTU TRIBE; PACIFIC COAST
FEDERATION OF FISHERMEN’S
ASSOCIATIONS, INC., Institute for
Fisheries Research,
Intervenor-Defendants–Appellees.
SAN LUIS & DELTA-MENDOTA No. 12-15293
WATER AUTHORITY; WESTLANDS
WATER DISTRICT; STOCKTON EAST D.C. No.
WATER DISTRICT; OAKDALE 1:09-cv-01053-
IRRIGATION DISTRICT; SOUTH SAN LJO-DLB
JOAQUIN IRRIGATION DISTRICT;
STATE WATER CONTRACTORS; KERN
COUNTY WATER AGENCY;
COALITION FOR A SUSTAINABLE
DELTA,
Plaintiffs,
8 SAN LUIS V. LOCKE
CALIFORNIA DEPARTMENT OF
WATER RESOURCES,
Intervenor-Plaintiff,
and
METROPOLITAN WATER DISTRICT OF
SOUTHERN CALIFORNIA,
Plaintiff-Appellant,
v.
GARY LOCKE; UNITED STATES
DEPARTMENT OF COMMERCE;
NATIONAL OCEANIC AND
ATMOSPHERIC ADMINISTRATION;
NATIONAL MARINE FISHERIES
SERVICE; JAMES W. BALSIGER;
RODNEY R. MCINNIS; U.S.
DEPARTMENT OF THE INTERIOR;
UNITED STATES BUREAU OF
RECLAMATION; MICHAEL L.
CONNOR; DONALD R. GLASER; JANE
LUBCHENCO; SALLY JEWELL,
Defendants-Appellees,
THE BAY INSTITUTE; CALIFORNIA
TROUT; FRIENDS OF THE RIVER;
NATURAL RESOURCES DEFENSE
COUNCIL; NORTHERN CALIFORNIA
COUNCIL OF THE FEDERATION OF
FLY FISHERS; SAN FRANCISCO
BAYKEEPER; SACRAMENTO RIVER
SAN LUIS V. LOCKE 9
PRESERVATION TRUST; WINNEMEM
WINTU TRIBE; PACIFIC COAST
FEDERATION OF FISHERMEN’S
ASSOCIATIONS, INC., Institute for
Fisheries Research,
Intervenor-Defendants–Appellees.
SAN LUIS & DELTA-MENDOTA No. 12-15296
WATER AUTHORITY; WESTLANDS
WATER DISTRICT; STOCKTON EAST D.C. No.
WATER DISTRICT; METROPOLITAN 1:09-cv-01053-
WATER DISTRICT OF SOUTHERN LJO-DLB
CALIFORNIA; OAKDALE IRRIGATION
DISTRICT; SOUTH SAN JOAQUIN
IRRIGATION DISTRICT; KERN OPINION
COUNTY WATER AGENCY;
COALITION FOR A SUSTAINABLE
DELTA; STATE WATER
CONTRACTORS,
Plaintiffs-Appellees,
CALIFORNIA DEPARTMENT OF
WATER RESOURCES,
Intervenor-Plaintiff–Appellee,
v.
GARY LOCKE; UNITED STATES
DEPARTMENT OF COMMERCE;
NATIONAL OCEANIC AND
ATMOSPHERIC ADMINISTRATION;
NATIONAL MARINE FISHERIES
10 SAN LUIS V. LOCKE
SERVICE; JAMES W. BALSIGER;
RODNEY R. MCINNIS; U.S.
DEPARTMENT OF THE INTERIOR;
UNITED STATES BUREAU OF
RECLAMATION; MICHAEL L.
CONNOR; DONALD R. GLASER; JANE
LUBCHENCO; SALLY JEWELL,
Defendants-Appellants,
and
THE BAY INSTITUTE; CALIFORNIA
TROUT; FRIENDS OF THE RIVER;
NATURAL RESOURCES DEFENSE
COUNCIL; NORTHERN CALIFORNIA
COUNCIL OF THE FEDERATION OF
FLY FISHERS; SAN FRANCISCO
BAYKEEPER; SACRAMENTO RIVER
PRESERVATION TRUST; WINNEMEM
WINTU TRIBE; PACIFIC COAST
FEDERATION OF FISHERMEN’S
ASSOCIATIONS, INC., Institute for
Fisheries Research,
Intervenor-Defendants.
Appeal from the United States District Court
for the Eastern District of California
Oliver W. Wanger, District Judge, Presiding
Argued and Submitted
September 15, 2014—San Francisco, California
SAN LUIS V. LOCKE 11
Filed December 22, 2014
Before: Richard C. Tallman and Johnnie B. Rawlinson,
Circuit Judges, and Thomas O. Rice, District Judge.**
Opinion by Judge Tallman
SUMMARY***
Endangered Species Act
The panel affirmed in part and reversed in part the district
court’s summary judgment and remanded for entry of
summary judgment in favor of defendants, federal agencies
and intervenor-environmental groups, in an action pertaining
to a formal Biological Opinion developed by the Commerce
Department’s National Marine Fisheries Service pursuant to
the Endangered Species Act regarding the impact of
continuing water extraction in the California Central Valley
on certain threatened and endangered Salmonid species.
The Marine Fisheries Service in its 2009 Biological
Opinion determined that the Department of Interior Bureau of
Reclamation’s proposed water project in the Central Valley
would jeopardize some of the Delta’s endangered Salmonids.
To remedy this problem, the Marine Fisheries Service
**
The Honorable Thomas O. Rice, United States District Judge for the
Eastern District of Washington, sitting by designation.
***
This summary constitutes no part of the opinion of the court. It has
been prepared by court staff for the convenience of the reader.
12 SAN LUIS V. LOCKE
required the Bureau to change the way it pumps water out of
the Central Valley’s rivers. A number of groups that depend
on the Central Valley’s water sued to halt this change. On
summary judgment, the district court found, in part, that the
Marine Fisheries Service violated the Administrative
Procedure Act’s arbitrary or capricious standard when
developing much of the Biological Opinion.
On an initial evidentiary question, the panel held that the
district court went beyond the exceptions, set forth in Lands
Council v. Powell, 395 F.3d 1019 (9th Cir. 2004), when it
admitted extra-record declarations and substituted the
analysis in those declarations for that provided by the Marine
Fisheries Service.
The panel held that the district court did not give the
Service the substantial deference it was due under the
Administrative Procedure Act. The panel found that the
components of the Biological Opinion invalidated by the
district court were reasonable and supported by the record
and therefore the panel upheld the Biological Opinion in its
entirety.
Specifically, the panel found that: (1) the Service acted
within its substantial discretion when it used raw salvage data
instead of data scaled to fish population to set flows in the
Old and Middle Rivers; (2) the Service’s jeopardy opinion
components were not arbitrary and capricious as they
pertained to the winter-run Chinook, the Southern Resident
orca, the steelhead critical habitat, and the impact of indirect
mortality factors on the listed species; and (3) the Biological
Opinion’s challenged reasonable and prudent alternative
actions were not arbitrary and capricious.
SAN LUIS V. LOCKE 13
Affirming, on cross-appeal, several components of the
district court’s opinion, the panel held that the Marine
Fisheries Service did not need to distinguish between
discretionary and non-discretionary actions; that the
Biological Opinion’s indirect mortality factors were direct
effects under the Endangered Species Act; and that Bureau of
Reclamation was not independently liable under the
Endangered Species Act.
COUNSEL
Rebecca Rose Akroyd, Daniel J. O’Hanlon, Hanspeter
Walter, Kronick, Moskovitz, Tiedemann, & Girard, PC,
Sacramento, California, for Plaintiffs-Appellees/Cross-
Appellants San Luis & Delta-Mendota Water Authority and
Westlands Water District.
David A. Diepenbrock, Eileen Diepenbrock, Jonathan Marz,
and Jon D. Rubin, Diepenbrock Elkin LLP, Sacramento,
California, for Plaintiffs-Appellees San Luis & Delta-
Mendota Water Authority.
Robert D. Thornton and Paul S. Weiland, Nossaman LLP,
Irvine, California, for Plaintiff-Appellee/Cross-Appellant
Kern County Water Agency and Plaintiff-Appellee Coalition
for a Sustainable Delta.
Amelia T. Minaberrigarai, General Counsel, Bakersfield,
California, for Plaintiff-Appellee/Cross-Appellant Kern
County Water Agency.
Martha F. Bauer, Mark J. Mathews, Brownstein Hyatt Farber
Schreck, LLP, Denver, Colorado; Steve Sims, Brownstein
14 SAN LUIS V. LOCKE
Hyatt Farber Schreck, LLP, Albuquerque, New Mexico;
David Longly Bernhardt (argued), Brownstein Hyatt Farber
Schreck, LLP, Washington, D.C.; Harold Craig Manson,
General Counsel, Fresno, California, for Plaintiff-
Appellee/Cross-Appellant Westlands Water District.
Tim O’Laughlin and William C. Paris, III, O’Laughlin &
Paris LLP, for Plaintiffs-Appellees Oakdale Irrigation District
and South San Joaquin Irrigation District.
Steven M. Anderson, Melissa R. Cushman, Steven G. Martin,
and Gregory K. Wilkinson, Best Best & Krieger, LLP,
Riverside, California; Paeter E. Garcia, Best Best & Krieger
LLP, Los Angeles, California, for Plaintiffs-Appellees State
Water Contractors.
Alexis K. Galbraith, Karna Elizabeth Harrigfeld, Jennifer
Lynn Spaletta, and Jeanne M. Zolezzi, Herum Crabtree,
Stockton, California, for Plaintiff-Appellee Stockton East
Water District.
Christopher J. Carr and William M. Sloan (argued), Morrison
& Foerster LLP, San Francisco, California; Linus
Masouredis, Chief Deputy General Counsel, Sacramento,
California, for Plaintiff-Appellee Metropolitan Water District
of Southern California.
Michael M. Edson, Allison Goldsmith, Daniel S. Harris,
Clifford T. Lee (argued), Deputy Attorneys General—Office
of the California Attorney General, San Francisco, California,
for Intervenor-Plaintiff–Appellee California Department of
Water Resources.
SAN LUIS V. LOCKE 15
Ellen J. Durkee (argued) and Bridget McNeil, United States
Department of Justice, Washington, D.C., for Defendants-
Appellants/Cross-Appellees United States Department of
Commerce, National Oceanic and Atmospheric
Administration, National Marine Fisheries Service, United
States Department of the Interior, and United States Bureau
of Reclamation.
Trent W. Orr and George Matthew Torgun, Earthjustice, San
Francisco, California, for Intervenor-Defendants–Appellants/
Cross-Appellees Pacific Coast Federation of Fisherman’s
Associations, Institute for Fisheries Research, The Bay
Institute, California Trout, Friends of the River, Northern
California Counsel of the Federation of Fly Fishers, San
Francisco Baykeeper, Sacramento River Preservation Trust,
Winnemem Wintu Tribe.
Katherine S. Poole (argued) and Douglas A. Obegi, San
Francisco, California, for Intervenor-Defendant/Appellant
Natural Resources Defense Counsel.
16 SAN LUIS V. LOCKE
TABLE OF CONTENTS
I. FACTS AND PROCEEDINGS BELOW. . . . . . . . . . 23
A. Background. . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1. Factual Background—The Sacramento-
San Joaquin Delta. . . . . . . . . . . . . . . . 23
a. The Central Valley and the River
Systems. . . . . . . . . . . . . . . . . . 23
b. The State Water Project and the
Central Valley Project. . . . . . . 25
c. Threatened and Endangered
Species in the Delta. . . . . . . . . 28
2. Legal Background—The Endangered
Species Act.. . . . . . . . . . . . . . . . . . . . . 30
B. Proceedings Leading to the Current Controversy
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1. The 2009 Salmonid Biological Opinion
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
a. The Consultation Request. . . . 32
b. The Jeopardy Opinion. . . . . . . 33
c. The Reasonable and Prudent
Alternatives. . . . . . . . . . . . . . . 34
SAN LUIS V. LOCKE 17
2. The Present Case. . . . . . . . . . . . . . . . . 35
II. STANDARDS OF REVIEW. . . . . . . . . . . . . . . . . . . . 39
III. THE RECORD ON REVIEW. . . . . . . . . . . . . . . . 40
IV. THE LEGAL FRAMEWORK.. . . . . . . . . . . . . . . 44
A. The APA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
B. The ESA.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
V. THE MERITS OF THE BIOLOGICAL OPINION. . . 47
A. We Defer To the Agency’s Choice To Use Raw
Salvage Figures. . . . . . . . . . . . . . . . . . . . . . . . 49
B. The Challenged Jeopardy Opinion Components
Are Not Arbitrary or Capricious. . . . . . . . . . . 52
1. Winter-Run Chinook. . . . . . . . . . . . . . 52
2. Southern Resident Orca. . . . . . . . . . . . 53
3. Steelhead Critical Habitat. . . . . . . . . . 55
4. Indirect Mortality Factors. . . . . . . . . . 57
C. The Challenged RPA Actions Are Not Arbitrary
or Capricious. . . . . . . . . . . . . . . . . . . . . . . . . . 60
1. The Legal Requirements for an RPA
Action.. . . . . . . . . . . . . . . . . . . . . . . . . 60
18 SAN LUIS V. LOCKE
a. The ESA Does Not Require
NMFS To Explain How Each
RPA Action Is Essential To Avoid
Jeopardy. . . . . . . . . . . . . . . . . . 62
b. The ESA Does Not Require
NMFS To Articulate Compliance
with the Non-Jeopardy Factors
. . . . . . . . . . . . . . . . . . . . . . . . . 63
2. Challenged RPA Actions. . . . . . . . . . . 64
a. Action IV.2.1. . . . . . . . . . . . . . 64
b. Action IV.2.3 and Action IV.3
. . . . . . . . . . . . . . . . . . . . . . . . . 66
c. Action IV.4.2. . . . . . . . . . . . . . 68
d. Action III.1.2. . . . . . . . . . . . . . 69
e. Action III.1.3. . . . . . . . . . . . . . 71
f. Action III.2.2. . . . . . . . . . . . . . 73
VI. CROSS-APPEAL. . . . . . . . . . . . . . . . . . . . . . . . . 74
A. NMFS Need Not Distinguish Discretionary and
Non-Discretionary Actions. . . . . . . . . . . . . . . 74
B. The Biological Opinion’s Indirect Mortality
Factors Are Direct Effects Under the ESA. . . 75
SAN LUIS V. LOCKE 19
C. Reclamation Is Not Independently Liable Under
the ESA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
VII. CONCLUSIONS. . . . . . . . . . . . . . . . . . . . . . . . . . 77
GLOSSARY OF TERMS. . . . . . . . . . . . . . . . . . . . . . . . . 79
20 SAN LUIS V. LOCKE
OPINION
TALLMAN, Circuit Judge:
And then the dry years would come, and
sometimes there would be only seven or eight
inches of rain. The land dried up and the
grasses headed out miserably a few inches
high and great bare scabby places appeared in
the valley. The live oaks got a crusty look
and the sage-brush was gray. The land
cracked and the sprigs dried up and the cattle
listlessly nibbled dry twigs. Then the farmers
and the ranchers would be filled with disgust
for the Salinas Valley. The cows would grow
thin and sometimes starve to death. People
would have to haul water in barrels to their
farms just for drinking.
John Steinbeck, East of Eden 5 (Penguin Books 2002) (1952).
Although John Steinbeck wrote about California’s Salinas
Valley, the same can be said for California’s Central Valley.
Like the Salinas Valley, the Central Valley is rich and fertile.
It is home to some of California’s most productive
agriculture, and food grown in the Valley sits on the tables in
most American homes. But the Central Valley is also
naturally dry. The Valley floor receives an average of five to
sixteen inches of rainfall per year; the United States
Geological Service considers it to be arid or semi-arid. In its
natural state, the Valley could not sustain the level of
agriculture that the country demands from it.
SAN LUIS V. LOCKE 21
To remedy this problem, the federal and state
governments have invested enormous sums of money
developing infrastructure to pump water out of the rivers that
crisscross the Valley’s floor, store it, and deliver it to
agricultural and domestic consumers in California. This
water is essential to the continuing vitality of agriculture in
the Central Valley, and some 25 million Californians depend
on it for daily living. But that water is also an important
habitat for thousands of river and anadromous fish, many of
which are endangered.
And therein lies the conflict: If the governments did not
extract water from the Central Valley’s rivers, the Valley
could not support the farms that feed, the dams that power,
and the canals that hydrate millions of Americans. But by
extracting the water, people dramatically alter the rivers’
natural state and threaten the viability of the species that
depend on them. People need water, but so do fish.
This case is about the competing demands for these
limited water resources. In 2006 the Department of Interior’s
Bureau of Reclamation (“Reclamation”), the federal agency
that oversees water resources in the West, asked the
Commerce Department’s National Marine Fisheries Service
(“NMFS”) to evaluate under the Endangered Species Act
(“ESA”) the impact of continuing water extraction in the
Central Valley on certain threatened and endangered
Salmonid species that live there. In response, NMFS
developed a Biological Opinion (“BiOp”) in which it
determined that Reclamation’s proposed project would
jeopardize some of the Delta’s endangered Salmonids. See
generally 2009 Salmonid BiOp at 574–75. To remedy this
problem, NMFS required Reclamation to change the way it
pumps water out of the Valley’s rivers. See id. at ch. 11. A
22 SAN LUIS V. LOCKE
number of groups that depend on the Central Valley’s water
sued to halt this change. On summary judgment, the district
court found that NMFS had violated the Administrative
Procedure Act’s (“APA”) arbitrary or capricious standard
when developing much of the BiOp. See generally In re
Consolidated Salmonid Cases, 791 F. Supp. 2d 802, 955–59
(E.D. Cal. 2011); 5 U.S.C. § 706(2)(a) (2012). Defendants—
joined by environmental groups—appealed, and once again
we enter the fray.1
We hold that the district court did not give NMFS the
substantial deference it is due under the APA. On
independent record review, we find that the components of
the BiOp invalidated by the district court are reasonable and
supported by the record. As a result, we uphold the BiOp in
its entirety. We, therefore, REVERSE and REMAND for
entry of summary judgment in favor of Defendants.
1
This is not the first time we have addressed this conflict, nor is it likely
to be the last. We recently addressed the conflict between Delta irrigation
and a small threatened fish known as the Delta Smelt. See San Luis &
Delta Mendota Water Auth. v. Jewell (Delta Smelt), 747 F.3d 581 (9th Cir.
2014). In Delta Smelt we reversed the district court and upheld a 2008
BiOp in which the Fish and Wildlife Service (“FWS”) concludes that
continued water extraction from the Central Valley’s rivers would
jeopardize the Delta Smelt and offers reasonable and prudent alternatives
that Reclamation should take to ameliorate this impact. See id. at 593–92.
Our opinion in Delta Smelt informs much of our analysis here.
SAN LUIS V. LOCKE 23
I. FACTS AND PROCEEDINGS BELOW
A. Background
1. Factual Background—The Sacramento-San
Joaquin Delta
a. The Central Valley and the River Systems
The Central Valley is a flat-bottom basin covering 22,500
square miles in inland California. The walls of the basin are
created by several mountain ranges: the Sierra Nevada and
Cascade Mountains to the east, the Klamath Mountains to the
north, the Coast Range to the west, and the Tehachapi
Mountains to the south. See infra, Fig. A. The Valley is long
and narrow. It stretches from Bakersfield in the south to
Redding in the north (about 450 miles) and is between 40 and
60 miles wide from east to west.
The Central Valley contains several major river systems.
Those systems are comprised of the San Joaquin River
(which flows west from the Cathedral Range of the Sierra
Nevada Mountains, turns north around Fresno, and enters the
San Francisco Bay north of Berkeley), the Sacramento River
(which flows south from the Salmon and Trinity Mountains
around Redding and passes through Sacramento before
joining the San Joaquin River), and their tributaries. The
Sacramento and San Joaquin Rivers converge around
Antioch, about thirty miles northeast of Oakland, where they
form the San Joaquin River Delta. The water from the Delta
flows past Chipps Island, into the Suisun Bay, through Bulls
Head Channel, and into the San Francisco Bay. The water
passes the city of San Francisco and flows under the Golden
Gate Bridge where it finally enters the Pacific Ocean. See
24 SAN LUIS V. LOCKE
infra, Fig. A. The brackish body of water through which the
rivers flow on their way to the Pacific Ocean is called the
“Bay-Delta.” The river delta is called simply “the Delta.”
Fig. A. Delta Map.2
2
California Water Science Center, U.S. Geological Survey,
http://ca.water.usgs.gov/gama/Provs/CenVly.htm (last visited Oct. 21,
2014, 9:09 a.m.).
SAN LUIS V. LOCKE 25
b. The State Water Project and the Central
Valley Project
Since the early part of the twentieth century, land owners,
local irrigation districts, and the federal and California state
governments have pumped fresh water out of the San Joaquin
and Sacramento Rivers (and their tributaries) to irrigate the
agricultural lands of the Central Valley and to provide
drinking water to the people of California. See Cent. Delta
Water Agency v. United States, 306 F.3d 938, 943 (9th Cir.
2002). California governs this pumping through the State
Water Project (“SWP”) and the federal government does so
through the Central Valley Project (“CVP”) (collectively,
“the Projects”).
The SWP is the largest state-built water project in the
United States. Delta Smelt, 747 F.3d at 594. It consists of
“21 dams and reservoirs, . . . five power plants, 16 pumping
plants, and 662 miles of aqueduct.” Id. (internal citations
omitted). The California Department of Water Resources
(“DWR”)—Plaintiff-in-Intervention here—oversees
operations of the SWP. Id.
The CVP is “the largest federal water management project
in the United States.” Cent. Delta Water Agency, 306 F.3d at
943. Congress initially authorized it in the Rivers and
Harbors Act of 1935. Id. It comprises a series of dams, “21
reservoirs, 11 hydropower plants, and 500 miles of canals and
aqueducts.” Delta Smelt, 747 F.3d at 594. Reclamation
oversees operations of the CVP. The CVP is partially
governed by the Central Valley Project Improvement Act
(“CVPIA”), id. at 594, which Congress passed in 1992 to
“achieve a reasonable balance among competing demands for
use of Central Valley Project water, including the
26 SAN LUIS V. LOCKE
requirements of fish and wildlife, agricultural, municipal and
industrial and power contractors.” Central Valley Project
Improvement Act, Pub L. No. 102–575, 106 Stat. 4706
(1992).
Together, the Projects provide water to more than 25
million agricultural and domestic consumers in central and
southern California. They do so, in part, by pumping fresh
water out of the Delta using the Harvey O. Banks Pumping
Plant (“Banks pumping plant”) and the C.W. “Bill” Jones
Pumping Plant (“Jones pumping plant”), both of which are
located near Tracy, California.3 The Banks pumping plant is
capable of pumping water at the rate of 10,300 cubic feet per
second (“cfs”), but it generally operates closer to 6,680 cfs.
See OCAP BA at 2-2. The Jones pumping plant has a
maximum pumping capability of 4,600 cfs. See id. The
plants operate by lifting water from the Delta using motor-
generated pumps.4 They pump the water into pipes that
deliver it into the California Aqueduct or the Delta-Mendota
Canal, respectively. See Jones & Banks Pumping Facilities.
From there, the Projects deliver the water to agricultural users
in the Central Valley and domestic users in central and
southern California. See id.; see also Fig. B.
3
See Biological Assessment on the Continued Long-term Operations of
the Central Valley Project and the State Water Project, U.S. Bureau of
Reclamation, http://www.usbr.gov/mp/cvo/ocapBA_051608.html 2-1 (last
visited Oct. 20, 2014) [hereinafter “OCAP BA”]; see also Fig. A.
4
Central Valley Project’s C.W. “Bill” Jones Pumping Plant and Tracy
Fish Collection Facility, U.S. Bureau of Reclamation (July 2012),
http://www.usbr.gov/mp/PA/docs/fact_sheets/Jones_Pumping_Plant.pdf
[hereinafter “Jones and Banks Pumping Facilities”].
SAN LUIS V. LOCKE 27
The Projects also control the volume of water flowing
through the Central Valley’s rivers by prescribing releases
from upstream reservoirs, which operate as water storage
facilities. Releases from CVP/SWP reservoirs cool water
temperatures, reduce the salinity of the Delta, provide flood
control, improve volume for fish habitat and migration, and
supply additional water for agricultural use. See OCAP BA
at 2-5.
Fig. B. CVP and SWP Map.5
5
Central Valley Project, U.S. Bureau of Reclamation,
http://www.usbr.gov/projects/Project.jsp?proj_Name=Central+Valley+
Project (last visited Oct. 21, 2014 8:55 a.m.); James Nickles et al.,
California’s BAY-DELTA: USGS Science Supports Decision Making,
http://pubs.usgs.gov/fs/2010/3032/ (last visited Oct. 21, 2014, 9:00 a.m.).
28 SAN LUIS V. LOCKE
c. Threatened and Endangered Species in the Delta
Although the Projects provide substantial benefits to
people and to state agriculture, they arguably harm species
native to the Delta by modifying those species’ natural
habitats. Five such species are at issue in this case: (1) the
endangered Sacramento River winter-run Chinook salmon
(“winter-run Chinook”); (2) the threatened Central Valley
spring-run Chinook salmon (“spring-run Chinook”); (3) the
threatened Central Valley steelhead (“CV steelhead”); (4) the
threatened Southern Distinct Population Segment of North
American green sturgeon (“green sturgeon”); and the
endangered Southern Resident orca whale (“Southern
Resident orca”). See 2009 Salmonid BiOp at 30.
SAN LUIS V. LOCKE 29
The four Salmonid species (the first four listed) are
anadromous fish, and Southern Resident orca are marine
mammals. Anadromous fish live most of their lives in salt
water.6 Nevertheless, they are born, mature, lay eggs, and
often die in inland freshwater lakes and rivers. After they
grow from fry (baby fish) to smolts (juvenile fish) in fresh
water, anadromous salmon outmigrate through rivers and
deltas into the oceans and seas where they will spend most of
their adult lives. When it is time to reproduce, these salmon
migrate back through the deltas to the rivers and lakes in
which they were born to lay eggs. During this migration,
salmon must pass impediments in inland rivers such as locks,
dams, channels, and pumps.
The San Francisco Bay-Delta is an essential conduit for
anadromous fish that return to California’s inland rivers and
lakes to reproduce. Nevertheless, human interactions with the
Delta and California’s inland rivers over the past century
have significantly altered them, threatening their ability to
serve as salmonid habitats. SWP and CVP operations
increase pollution, encourage the growth of non-native
species, and create water shortages in the Delta that harm
salmon by exposing them to unnatural stressors. See 2009
Salmonid BiOp at 374–82. Migrating salmon can also be
caught in, and killed by, the large water pumps that serve the
Projects. Finally, CVP/SWP operations that limit cold water
releases from dams upstream of traditional spawning sites
potentially impact critical spawning habitat by making the
rivers less conducive to reproduction.
6
See, e.g., Chinook Salmon (Oncorhynchus tshawytscha), NOAA
Fisheries Office of Protected Resources, http://www.nmfs.noaa.gov/
pr/species/fish/chinooksalmon.htm (last updated May 15, 2014).
30 SAN LUIS V. LOCKE
2. Legal Background—The Endangered Species Act
We must review NMFS’s formal opinion as to how
Reclamation and DWR should operate the Projects to avoid
jeopardizing endangered Salmonid species. Before further
discussing the relationship between the Projects and the
species, we briefly review the legal framework for that
opinion.
The federal government protects listed and threatened
species, such as the five at issue here, primarily through the
ESA. See Endangered Species Act, 16 U.S.C. §§ 1531–1544
(2012). When Congress passed the ESA in 1973, it sought to
bring about the “better safeguarding, for the benefit of all
citizens, [of] the Nation’s heritage in fish, wildlife, and
plants.” Id. § 1531(a)(5).
Section 7 of the ESA “addresses the obligations of federal
agencies with respect to conservation and protection of
species listed as either endangered or threatened under the
ESA.” Lawrence R. Liebesman & Rafe Petersen,
Endangered Species Deskbook 39 (2d ed. 2010). ESA section
7 prohibits a federal agency from taking any action that is
“likely to jeopardize the continued existence” of any listed or
threatened species or “result in the destruction or adverse
modification” of those species’ critical habitat. 16 U.S.C.
§ 1536(a)(2).
Section 7 requires an agency proposing a project that
might harm listed or threatened species to consult with either
NMFS or the Interior Department’s Fish and Wildlife Service
SAN LUIS V. LOCKE 31
(“FWS”)7 about the proposed action. 16 U.S.C. § 1536(a)(2).
As part of this consultation, the action agency prepares an
initial assessment of the project in which it evaluates the
project’s impact on any listed or endangered species. This is
called a Biological Assessment (“BA”). 50 C.F.R. § 402.02
(2009). The appropriate consultation agency reviews the
action agency’s BA and uses it to prepare a Biological
Opinion (“BiOp”) in which it ultimately determines whether
the proposed agency action is likely to adversely impact
endangered or listed species, or negatively modify their
critical habitats. See 16 U.S.C. § 1536(a)(2).8
If the agency concludes that the proposed action will
jeopardize species or critical habitats, “the Biological
Opinion must outline any ‘reasonable and prudent
alternatives’ that the [agency] . . . believes will avoid that
consequence.” Bennett v. Spear, 520 U.S. 154, 158, 117 S.
Ct. 1154, 1159 (1997) (quoting 16 U.S.C. § 1536(b)(3)(A)).
Reasonable and prudent alternatives (“RPAs”) are
alternative actions identified during formal
consultation 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
7
Whether an agency consults with NMFS or FWS depends on the
species for which it is seeking consultation. NMFS consults on marine
and anadromous species. See Liebesman & Petersen, supra, at 40.
8
The consultation agency’s determination with regard to whether the
proposed project is likely to jeopardize listed species is called the
“jeopardy opinion” component of the BiOp. See U.S. Fish & Wildlife
Serv. & Nat’ l Marine Fisheries Serv., ESA Section 7 Consultation
Handbook 4-37 (Mar. 1998) [hereinafter “Handbook”].
32 SAN LUIS V. LOCKE
agency’s legal authority and jurisdiction, that
is [sic] economically and technologically
feasible, and that the Director believes would
avoid the likelihood of jeopardizing the
continued existence of listed species or
resulting in the destruction or adverse
modification of critical habitat.
50 C.F.R. § 402.02. The consulting agency may also
issue—with the BiOp—an incidental take statement (“ITS”)
that permits the action agency to harm listed species when
implementing the RPAs without violating the ESA. See id.
B. Proceedings Leading to the Current Controversy
1. The 2009 Salmonid Biological Opinion
a. The Consultation Request
In 2006, Reclamation asked NMFS to prepare a BiOp
assessing the impact of continued and future CVP/SWP
operations on Delta Salmonid species.9 This request was
motivated by the listing of new endangered species and the
designation of new critical habitats. 2009 Salmonid BiOp at
31. In it, Reclamation asked NMFS to evaluate the effect of
Reclamation and DWR’s proposal to continue to operate the
Projects “to divert, store, and convey Project water . . . ,”
OCAP BA at 2-1, on winter-run Chinook, spring-run
Chinook, CV steelhead, CCC steelhead (a fish not at issue
9
Although the SWP is a state project, it is subject to federal consultation
along with the CVP because of a 1986 agreement between the federal
government and DWR. 2009 Salmonid BiOp at 31.
SAN LUIS V. LOCKE 33
here), green sturgeon, and Southern Resident orca, see 2009
Salmonid BiOp at 30.
Reclamation developed a BA that could provide the basis
for such a consultation in the fall of 2008. Id. at 32. Using
the data Reclamation provided in the BA, supplementing it
with its own research, NMFS developed a draft Salmonid
BiOp that it released to Reclamation and DWR for comment
in the winter of 2008. Id. at 33. Reclamation and DWR
reviewed and commented on the document. NMFS also
requested and received peer review from the CALFED Bay-
Delta Program and the Center for Independent Experts. See
id. Based on the comments received, NMFS published a final
560-page BiOp on June 4, 2009. Reclamation provisionally
accepted the BiOp that same day.
Plaintiffs contest the legality of the 2009 BiOp here,
arguing—for various reasons—that parts of it are arbitrary or
capricious in violation of the APA. Before discussing
Plaintiffs’ specific challenges and the issues on appeal, we
briefly review the portions of the BiOp that are relevant.
b. The Jeopardy Opinion
In the first part of the BiOp, NMFS concludes that “the
long-term operations of the CVP and SWP are likely to
jeopardize the continued existence of the” winter-run
Chinook, the spring-run Chinook, the CV steelhead, the green
sturgeon, and the Southern Resident orca. Id. at 575.
Similarly, “[t]he long-term operations of the CVP and SWP
are likely to destroy or adversely modify critical habitat for”
winter-run Chinook, spring-run Chinook, CV steelhead, and
green sturgeon. Id.
34 SAN LUIS V. LOCKE
c. The Reasonable and Prudent Alternatives
Because NMFS concludes that ongoing CVP/SWP
operations would threaten listed species, it issued over
seventy RPAs that Reclamation is supposed to implement to
avoid jeopardy. See generally id. at 574–724. The proposed
RPAs fall into five operational categories—(I) Sacramento
River Division, (II) American River Division, (III) East Side
Division, (IV) Delta Division, and (V) Fish Passage Program.
See id. at 19. On appeal, the parties challenge provisions of
the RPAs falling into categories III and IV.
Actions in category III relate to CVP/SWP operations on
the Stanislaus River, which provides critical spawning and
smolting grounds for the CV steelhead. See id. at 619–20.
Prior to the construction of the New Melones Dam on the
Stanislaus River, CV steelhead spawned in the cold
tributaries upstream of where the New Melones Reservoir is
now located. See id. at 107–08, 619. Now, “[t]he steelhead
population on the Stanislaus River is precariously small and
limited to habitat areas below the [Goodwin and New
Melones] Dams that historically were unsuitable owing to
high summer temperatures.” Id. at 619. In RPA category III,
NMFS prescribes certain volumes of releases from the
Goodwin and New Melones Dams that, according to NMFS,
will cool the rivers enough to facilitate steelhead spawning.
See id. at 620. The flows will also rejuvenate the gravel that
is essential to steelhead spawning habitat and provide
migratory cues to adult and juvenile fish.
Actions in category IV relate to operations in the Delta.
See id. at 628–30 (describing Delta division action). NMFS
concludes that the proposed actions for the Projects, which
“include continued diversion of water from the Delta at the
SAN LUIS V. LOCKE 35
project’s export facilities, with increased export levels,” “will
increase the level of stressors in the Delta,” further degrading
it as a habitat. Id. at 629. The category IV RPA Actions
address this jeopardy finding by imposing flow-to-export
ratios for the Old and Middle Rivers, see id. at 643–44,
prescribing maximum negative flow rates for the Old and
Middle Rivers, id. at 648, and requiring a certain salvage
efficiency at major fish salvage stations, id. at 655.
2. The Present Case
On June 15, 2009, Plaintiffs San Luis & Delta-Mendota
Water Authority and Westlands Water District challenged the
legality of the 2009 BiOp by filing suit against the
Department of Commerce, the National Oceanic and
Atmospheric Administration, and NMFS (collectively
“Federal Defendants”)10 in the Eastern District of California.
See Compl. at 1, ECF No. 1. The district court consolidated
that case with several other cases in which state water
districts challenged the 2009 BiOp.11 DWR intervened as a
plaintiff. See Joinder by Calif. Dep’t of Water Res., ECF No.
137. And several environmental and fishing groups
10
Reclamation and the United States Department of the Interior were
later joined. They are included under the umbrella of “Federal
Defendants.” See In re Consolidated Salmonid Cases, 791 F. Supp. 2d at
813.
11
The plaintiffs fall into three separate groups. The Export Plaintiffs
are San Luis & Delta Mendota Water Authority and Westlands Water
District; State Water Contractors; Kern County Water Agency and
Coalition for a Sustainable Delta; and Metropolitan Water District of
Southern California. The Stanislaus River Plaintiffs (or “SR Plaintiffs”)
are Stockton East Water District, Oakdale Irrigation District, and South
San Joaquin Irrigation District. The DWR Plaintiff in Intervention is the
California Department of Water Resources.
36 SAN LUIS V. LOCKE
intervened as defendants.12 See In re Consolidated Salmonid
Cases, 791 F. Supp. 2d at 813.
On August 6, 2010, several Plaintiffs moved for summary
judgment on their claim that the 2009 BiOp violates the ESA
and the APA. Id. The Stanislaus River Plaintiffs and DWR
filed separate motions for summary judgment. Id. The
Federal Defendants and Defendant-Intervenors responded
with cross-motions for summary judgment. Id. “These cross-
motions, which included over 700 pages of briefing and
thousands of pages of supporting declarations and exhibits,
came on for hearing on December 16 and 17, 2010.” Id. On
September 20, 2011, the district court filed a 157-page
opinion granting in part and denying in part Plaintiffs’ claims,
and granting in part and denying in part Defendants’ claims.
Id. at 955–59.
The district court made dozens of conclusions relating to
almost every component of the BiOp when rendering this
complex and lengthy opinion. We briefly review the
conclusions at issue in this appeal.
The Defendants ask us to overturn the following of the
district court’s holdings in which it struck down components
of the BiOp:
• NMFS acted unlawfully by relying on raw salvage data to
set negative flow thresholds for the Old and Middle
12
Those Defendant-Intervenors are The Bay Institute; California Trout;
Friends of the River; Natural Resources Defense Council; Northern
California Council of the Federation of Fly Fishers; San Francisco
Baykeeper; Sacramento River Preservation Trust; Winnemem Wintu
Tribe; and Pacific Coast Federation of Fishermen’s Associations, Inc.
SAN LUIS V. LOCKE 37
Rivers. Basic scientific principles require the agency to
use data scaled to population to determine the impact of
exports on fish survival. Id. at 827.
• NMFS erred by failing to provide sufficient support for
its classification of the winter-run Chinook as “high risk”
rather than the less serious classification of “not viable.”
Id. at 864.
• NMFS erred by failing to reconcile the 2009 Salmonid
BiOp’s jeopardy determination relating to the Southern
Resident orca with an apparently contradictory conclusion
in a different 2009 BiOp (“2009 Orca BiOp”). Id. at 866.
• NMFS failed to adequately explain how continued
operation of the Projects will adversely modify the CV
steelhead’s critical habitat by reducing spawnable area
and degrading gravel quality and quantity. Id. at 936.
• Although NMFS sufficiently established that delta
hydrologic conditions—as altered in part by the
Projects—are favorable to invasive species, the BiOp
does not support the conclusion that continued CVP/SWP
operations promote invasive species, which in turn
threaten listed species. Id. at 870. Nor does the BiOp
sufficiently explain “how the projects influence
contaminants or cause food limitations.” Id.
• NMFS provided no support for its decision to use
“maximum steelhead habitat” as a benchmark for
evaluating the effect of East Side Division operations on
listed species in the Stanislaus River. Because the
modeling related to the New Melones Dam flows are
38 SAN LUIS V. LOCKE
based on the “maximum habitat” benchmark, the New
Melones Dam flow limits violate the APA. Id. at 938.
• NMFS’s modeling assumptions relating to the Stanislaus
River are flawed because NMFS set its goal as
“doubling” CV steelhead habitat. Id. at 950.
• NMFS failed to establish how each RPA Action complies
with 50 C.F.R. § 402.02’s non-jeopardy factors.
Specifically, NMFS did not establish how each RPA
Action complies with the many purposes of the CVPIA,
id. at 918, nor did it consider how each Action is feasible,
id. at 919. NMFS erred particularly by failing to show
how Delta Action IV.4.2 is feasible.
• NMFS erred by failing to explain how certain RPA
Actions are “essential” to avoid jeopardy of the listed
species or adverse modification of their habitats. Id. at
897 & n.26. Those actions are Delta Division RPAs
IV.2.1, IV.2.3, and IV.3 and East Side Division RPAs
III.1.2, III.1.3, and III.2.2.
Plaintiffs ask us to overturn the following of the district
court’s holdings that were favorable to the BiOp:
• Reclamation did not violate its obligations under section
7 of the ESA when it accepted the 2009 Salmonid BiOp.
Id. at 955.
• NMFS did not need to segregate discretionary and non-
discretionary agency activities when constituting the
environmental baseline. Id. at 852.
 SAN LUIS V. LOCKE 39
• NMFS did not err by classifying indirect mortality factors
as a direct effect of the continuing Projects. Id. at
870–71.
The district court entered its final judgment on December
12, 2011. See Final Judgment, ECF No. 655. The parties
timely cross-appealed. This court has jurisdiction under
28 U.S.C. § 1291 (2012).
II. STANDARDS OF REVIEW
We review the district court’s summary judgment rulings
de novo. McFarland v. Kempthorne, 545 F.3d 1106, 1110
(9th Cir. 2008) (internal citations omitted). Summary
judgment is appropriate when the pleadings and record
demonstrate that “there is no genuine dispute as to any
material fact and the movant is entitled to judgment as a
matter of law.” Fed. R. Civ. P. 56(a). This court also reviews
de novo the district court’s evaluations of an agency’s
actions. Sierra Club v. Babbit, 65 F.3d 1502, 1507 (9th Cir.
1995) (“De novo review of a district court judgment
concerning a decision of an administrative agency means we
view the case from the same position as the district court.”).
We evaluate a district court’s decision to admit extra-record
evidence for abuse of discretion. Lands Council v. Powell,
395 F.3d 1019, 1030 n.11 (9th Cir. 2004).
This is a record review case, so we will conduct our own
review of the administrative record and, if necessary, “direct
that summary judgment be granted to either party . . . .” Id.
at 1026.
40 SAN LUIS V. LOCKE
III. THE RECORD ON REVIEW
Before reviewing the merits of the 2009 Salmonid BiOp,
we must resolve an initial evidentiary question: Did the
district court err in its own record review by supplementing
the administrative record with dozens of extra-record
declarations? The district court relied on extra-record
declarations comprising thousands of pages of scientific
opinion, to evaluate and—in some circumstances—call into
question the BiOp. See In re Consolidated Salmonid Cases,
791 F. Supp. 2d at 813 (describing how the parties’ cross
motions for summary judgment “included . . . thousands of
pages of supporting declarations and exhibits”). It did so
under our holding in Lands Council, 395 F.3d at 1030, which
permits district courts to supplement an administrative record
in a few limited circumstances. Id.; see Tr. of Proceeding
Mot. to Admit Expert Test. vol. 1, at 12, ECF No. 695
[hereinafter Expert Tr. vol. 1]; id. at 14–15 (describing the
Lands Council exceptions). The question here is whether the
district court properly applied Lands Council, or whether it
went beyond Lands Council to improperly question NMFS’s
scientific determinations. We hold, based in part on our
opinion in Delta Smelt, 747 F.3d at 602–04, that the district
court went beyond the Lands Council exceptions when it
admitted extra-record declarations and substituted the
analysis in those declarations for that provided by NMFS.
In general, a court reviewing agency action under the
APA must limit its review to the administrative record. See
Camp v. Pitts, 411 U.S. 138, 142, 93 S. Ct. 1231, 1244
(1973). We have applied this rule many times, in many
different contexts. See, e.g., Delta Smelt, 747 F.3d at 602–04
(stating the rule and applying it to strike extra-record
declarations admitted by the district court); Fence Creek
SAN LUIS V. LOCKE 41
Cattle Co. v. U.S.F.S., 602 F.3d 1125, 1131 (9th Cir. 2010)
(“Generally, judicial review of an agency decision is limited
to the administrative record on which the agency based the
challenged decision.”); Sw. Ctr. for Biological Diversity v.
U.S.F.S., 100 F.3d 1443, 1450 (9th Cir. 1996) (“Judicial
review of an agency decision typically focuses on the
administrative record in existence at the time of the decision
and does not encompass any part of the record that is made
initially in the reviewing court.”); Asarco, Inc. v. E.P.A.,
616 F.2d 1153, 1159 (9th Cir. 1980) (“[A]gency action must
be examined by scrutinizing the administrative record at the
time the agency made its decision.”).
This rule ensures that the reviewing court affords
sufficient deference to the agency’s action. The APA gives
an agency substantial discretion “to rely on the reasonable
opinions of its own qualified experts even if, as an original
matter, a court might find contrary views more persuasive.”
Marsh v. Or. Natural Res. Def. Council, 490 U.S. 360, 378,
109 S. Ct. 1851, 1861 (1989). “When a reviewing court
considers evidence that was not before the agency, it
inevitably leads the reviewing court to substitute its judgment
for that of the agency.” Asarco, 616 F.2d at 1160. In so
imposing its judgment, the reviewing court effectively
conducts a de novo review of the agency’s action rather than
limiting itself to the deferential procedural review that the
APA’s arbitrary or capricious standard permits. See River
Runners for Wilderness v. Martin, 593 F.3d 1064, 1070 (9th
Cir. 2010) (per curiam).
But we have also recognized several exceptions to this
rule. Under Lands Council, a reviewing court may consider
extra-record evidence where admission of that evidence (1) is
necessary to determine “‘whether the agency has considered
42 SAN LUIS V. LOCKE
all relevant factors and has explained its decision,’” (2) is
necessary to determine whether “‘the agency has relied on
documents not in the record,’ (3) ‘when supplementing the
record is necessary to explain technical terms or complex
subject matter,’ or (4) ‘when plaintiffs make a showing of
agency bad faith.’” 395 F.3d at 1030 (quoting Sw. Ctr. for
Biological Diversity v. U.S.F.S., 100 F.3d at 1450). These
exceptions are to be narrowly construed, and the party
seeking to admit extra-record evidence initially bears the
burden of demonstrating that a relevant exception applies.
See Fence Creek, 602 F.3d at 1131.
The first Lands Council exception—the “relevant factors”
exception—is the most difficult to apply, so we pause here to
further examine it. Although the relevant factors exception
permits a district court to consider extra-record evidence to
develop a background against which it can evaluate the
integrity of the agency’s analysis, the exception does not
permit district courts to use extra-record evidence to judge the
wisdom of the agency’s action. Asarco, 616 F.2d at 1160.
This distinction is a fine, but important, one. Reviewing
courts may admit evidence under this exception only to help
the court understand whether the agency complied with the
APA’s requirement that the agency’s decision be neither
arbitrary nor capricious. See id. at 1159; see also Motor
Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto Ins.
Co., 463 U.S. 29, 103 S. Ct. 2856 (1983) (further describing
the APA’s standards). But reviewing courts may not look to
this evidence as a basis for questioning the agency’s scientific
analyses or conclusions. Asarco, 616 F.2d at 1160–61.
We most recently considered the scope of the Lands
Council exceptions in Delta Smelt. See Delta Smelt, 747 F.3d
at 602–04. There, like here, the district court admitted
SAN LUIS V. LOCKE 43
“multiple declarations from multiple experts” to augment the
administrative record. Id. at 603. We held that, in doing so,
the district court violated the APA for two reasons. First, the
court admitted more than forty expert declarations under
Lands Council in addition to certifying four expert
declarations under Federal Rule of Evidence 706. Id. at 599
n.13 (citing Fed. R. Evid. 706), 603. We questioned whether
the district court needed the extra-record declarations to
explain the technical language in the BiOp or provide
background material because the Rule 706 court-appointed
experts served those purposes. See id. at 603. Thus, we were
critical of the district court opening the administrative record
as a forum for the experts to debate the merits of the BiOp.
Id. at 603–04.
Second, we held in Delta Smelt that the district court
erred when it used the extra-record declarations as a basis for
judging the wisdom of the agency’s scientific analysis.
747 F.3d at 604. Even if a reviewing court properly admits
extra-record evidence under Lands Council, it may not use
the admitted extra-record evidence “to determine the
correctness or wisdom of the agency’s decision.” Asarco,
616 F.2d at 1160. Such use is never permitted.
Here too, the district court violated Delta Smelt’s holding
when it used several extra-record declarations to question
NMFS’s scientific judgments. As in Delta Smelt, the district
court here “relied . . . on the declarations of the parties’
experts-as-advocates as the basis for rejecting the BiOp.”
747 F.3d at 604. In this way, the district court overstepped
the bounds of Lands Council by opening the administrative
record as a forum for the experts to debate the merits of the
BiOp. The district court employed extra-record declarations
at the following points for this impermissible purpose: In re
44 SAN LUIS V. LOCKE
Consolidated Salmonid Cases, 791 F. Supp. 2d at 827, 852
(Deriso Decl., ECF No. 440); id. at 829, 832, 834, 841
(Burnham Decl., ECF No. 439); id. at 840, 841 (Hilborn
Reply Decl., ECF No. 493); id. at 863 (Cramer Decl., ECF
No. 448); id. at 880 (Cummings Decl., ECF No. 445); id. at
884, 889–90, 893 (Cavallo Decl., ECF No. 446–1); id. at
942–43 (Dotan Decl., ECF No. 442). By admitting these
declarations and relying on them to question the wisdom of
NMFS’s judgments, the district court abused its discretion
under Lands Council.
IV. THE LEGAL FRAMEWORK
A. The APA
The ESA does not provide its own standard of judicial
review, so we evaluate the BiOp under the APA’s arbitrary or
capricious standard. See Bennett, 520 U.S. at 174–77; Delta
Smelt, 747 F.3d at 601. Section 706(2)(A) of the APA
requires a reviewing court to uphold agency action unless it
is “arbitrary, capricious, an abuse of discretion, or otherwise
not in accordance with law.” 5 U.S.C. § 706(2)(A). Under
this standard, we will “sustain an agency action if the agency
has articulated a rational connection between the facts found
and the conclusions made.” Pac. Coast Fed’n of Fishermen’s
Ass’ns v. U.S. Bureau of Reclamation, 426 F.3d 1082, 1090
(9th Cir. 2005).
The arbitrary or capricious standard is a deferential
standard of review under which the agency’s action carries a
presumption of regularity. See Citizens to Preserve Overton
Park, Inc. v. Volpe, 401 U.S. 402, 415–16, 91 S. Ct. 814
(1971), abrogated in part on other grounds as recognized in
Califano v. Sanders, 430 U.S. 99, 97 S. Ct. 980 (1977); Kern
SAN LUIS V. LOCKE 45
Cnty. Farm Bureau v. Allen, 450 F.3d 1072, 1076 (9th Cir.
2006). Although the court’s inquiry must be “searching and
careful, . . . the ultimate standard of review is a narrow one.”
Marsh, 490 U.S. at 378 (internal citations omitted). Thus,
“[e]ven when an agency explains its decision with ‘less than
ideal clarity,’ a reviewing court will not upset the decision on
that account ‘if the agency’s path may be reasonably
discerned.’” Ala. Dep’t of Envt’l Conservation v. E.P.A.,
540 U.S. 461, 497, 124 S. Ct. 983 (2004) (quoting Bowman
Transp. v. Ark.—Best Freight Sys., Inc., 419 U.S. 281, 286,
95 S. Ct. 438 (1974)). It is not the reviewing court’s task to
“make its own judgment about” the appropriate outcome.
River Runners for Wilderness, 593 F.3d at 1070. “Congress
has delegated that responsibility to” the agency. Id. “The
court’s responsibility is narrower: to determine whether the”
agency complied with the procedural requirements of the
APA. Id.
This traditional deference to the agency is at its highest
where a court is reviewing an agency action that required a
high level of technical expertise. Marsh, 490 U.S. at 377; see
also Baltimore Gas & Elec. Co. v. Natural Res. Def. Council,
Inc., 462 U.S. 87, 103, 103 S. Ct. 2246 (1983) (“When
examining this kind of scientific determination . . . a
reviewing court must generally be at its most deferential.”).
As part of this deference, we afford the agency discretion to
choose among scientific models; we “reject an agency’s
choice of a scientific model only when the model bears no
rational relationship to the characteristics of the data to which
it is applied.” Delta Smelt, 747 F.3d at 621 (internal citations
omitted).
Nevertheless, the deference we owe an agency is not
unlimited. We may not automatically defer to an agency’s
46 SAN LUIS V. LOCKE
conclusions, even when those conclusions are scientific. See
Marsh, 490 U.S. at 378. Rather, our review must be
sufficiently probing to ensure that the agency has not
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.
State Farm, 463 U.S. at 43. A different approach “would not
simply render judicial review generally meaningless, but
would be contrary to the demand that courts ensure that
agency decisions are founded on a reasoned evaluation of the
relevant factors.” Marsh, 490 U.S. at 378 (internal citations
omitted).
B. The ESA
The ESA requires an agency to use “the best scientific
and commercial data available” when formulating a BiOp.
16 U.S.C. § 1536(a)(2); 50 C.F.R. § 402.14(g)(8). An
agency’s failure to do so violates the APA. See 5 U.S.C.
§ 706(2)(A); Pac. Coast Fed’n v. Gutierrez, 606 F. Supp. 2d
1195, 1244 (E.D. Cal. 2008).
The purpose of the best available science standard is to
prevent an agency from basing its action on speculation and
surmise. Bennett, 520 U.S. at 176. Under this standard, an
agency must not “‘disregard[] available scientific evidence
that is in some way better than the evidence [it] relies on.’”
SAN LUIS V. LOCKE 47
Kern Cnty., 450 F.3d at 1080 (quoting Sw. Ctr. for Biological
Diversity v. Babbitt, 215 F.3d 58, 60 (D.C. Cir. 2000)). The
standard does not, however, require an agency to conduct
new tests or make decisions on data that does not yet exist.
See Am. Wildlands v. Kempthorne, 530 F.3d 991, 998–99
(D.C. Cir. 2008) (holding that an agency’s use of available
data and test methods was reasonable even though better test
methods existed because those test methods had not yet been
used on the species in question). Moreover, if the only
available data is “‘weak,’ and thus not dispositive,” an
agency’s reliance on such data “does not render the agency’s
determination ‘arbitrary and capricious.’” Greenpeace Action
v. Franklin, 14 F.3d 1324, 1336 (9th Cir. 1992) (quoting Stop
H-3 Ass’n v. Dole, 740 F.2d 1442, 1460 (9th Cir. 1984)). An
agency complies with the best available science standard so
long as it does not ignore available studies, even if it
disagrees with or discredits them. See Kern Cnty., 450 F.3d
at 1081 (rejecting Kern’s argument that FWS violated the
best available science standard when it cited but allegedly
misinterpreted three studies).
Finally, what constitutes the best scientific and
commercial data available is itself a scientific determination
deserving of deference. Miccosukee Tribe of Indians of Fla.
v. United States, 566 F.3d 1257, 1265 (11th Cir. 2009) (citing
Marsh, 490 U.S. at 377–78). For that reason “[a] court
should be especially wary of overturning such a
determination on review.” In re Consolidated Salmonid
Cases, 791 F. Supp. 2d at 821.
V. THE MERITS OF THE BIOLOGICAL OPINION
With these standards in mind, we evaluate each BiOp
component that is challenged on appeal.
48 SAN LUIS V. LOCKE
Plaintiffs originally challenged dozens of specific
components of the 2009 Salmonid BiOp. At summary
judgment, the district court upheld as valid many of them, and
determined that several others were arbitrary or capricious.
See id. at 955–59. Defendants appeal each part of the district
court opinion in which the court found the BiOp unlawful.
Plaintiffs cross-appeal several portions of the opinion in
which the district court upheld the BiOp.
We discuss each challenge to the BiOp in turn. First, we
address the district court’s objections to NMFS’s use of raw
salvage data. Second, we discuss the challenges to NMFS’s
jeopardy opinions, including the portion of the BiOp dealing
with indirect mortality factors. Third, we review the
challenges to NMFS’s RPAs, clarifying what the ESA and its
implementing regulations require from the agency when it is
developing and defining RPAs. We then evaluate the
challenged Actions in light of those requirements. Finally,
we discuss the three cross-appeal issues.
As our record review will show, the district court—in
many instances—did not afford the agency proper deference
under the APA. Rather than evaluating the agency’s
decision-making process and deferring to the agency’s
scientific conclusions when those conclusions are fairly
traceable to the record, the district court engaged in an in-
depth substantive review of the science supporting the BiOp
and substituted its own opinions, and those of the parties’
experts, for the opinions of NMFS. As a result, the district
court invalidated much of the BiOp under a quasi de novo
review. But the APA does not permit such an in-depth
review, particularly where, like here, the conclusions
implicate agency expertise. See Marsh, 490 U.S. at 375–77.
SAN LUIS V. LOCKE 49
We correct the district court’s errors in our own review;
and as a result, we uphold the BiOp in its entirety. After
reviewing the record as a whole, we are satisfied that, when
developing each component of the BiOp, NMFS relied on the
factors that Congress intended it to consider, considered all
important aspects of the problem, and offered explanations
for its decisions that are in line with the evidence. See State
Farm, 463 U.S. at 43. We are also satisfied that, in doing so,
NMFS used the best scientific data available, even if that
science was not always perfect. Cf. Greenpeace Action,
14 F.3d at 1336.
A. We Defer To the Agency’s Choice To Use Raw
Salvage Figures
The Projects pump fresh water out of the Old and Middle
Rivers in volumes sufficient to reverse the rivers’ traditional
flow. Delta Smelt, 747 F.3d at 606. Absent pumping, the
rivers would flow north into the Delta. Under pumping
operations, the rivers flow south to the Jones and Banks
pumping plants. Listed species—particularly juveniles—are
caught in the negative current and drawn towards the
pumping facilities. See 2009 Salmonid BiOp at 651. Some
of these fish are salvaged at the pumps, meaning they are
diverted from the fatal pumping plants to fish salvage
facilities and into tanks where they are counted, measured,
loaded into trucks, driven north, and dumped back into the
Delta.13 But even if salvaged, fish that are drawn towards the
pumps in the Rivers’ negative flow have a lower likelihood
13
See Fish Facilities Unit Monitoring and Operations Projects, Cal.
Dep’t of Wildlife, http://www.dfg.ca.gov/delta/data/salvage/
salvageoverview.asp (last visited Oct. 20, 2014, 4:02 p.m.).
50 SAN LUIS V. LOCKE
of surviving outmigration than their counterpoints that are
lucky enough to avoid entrainment.14
NMFS concluded in the BiOp that as negative flow of the
Old and Middle Rivers increases, fish are more likely to be
diverted out of the main Delta and towards the pumping
facilities. 2009 Salmonid BiOp at 651. To counter this effect
and enhance the likelihood of salmonids successfully exiting
the Delta, several of the RPA Actions regulate negative flows
and limit exports when fish numbers are high or are likely to
be high. NMFS developed these RPA Actions, in part, by
considering the raw number of fish salvaged at certain
volumes of negative flow. See id. at 361–62 (Figs. 65 & 66).
Plaintiffs argue that NMFS violated the ESA by using raw
salvage data instead of data scaled to fish population. They
assert that the number of fish salvaged every month could be
related to the number of fish in the Delta rather than to the
volume of negative flows in the Old and Middle Rivers. The
district court agreed, concluding that it goes against the grain
of traditional science to use raw instead of scaled salvage
numbers. In re Consolidated Salmonid Cases, 791 F. Supp.
2d at 827. And because “[t]he agency is required to apply
generally recognized and accepted biostatistical principles,
which constitute the best available science, in reaching its
decisions,” NMFS’s use of raw salvage data was arbitrary or
capricious. Id. Defendants appeal that holding here.
14
A fish is “entrained” when it follows diverted water rather than the
natural course of a river, stream, pond, or lake. The danger with
entrainment is that fish can become stranded in irrigation canals or killed
when they are trapped in pumps.
SAN LUIS V. LOCKE 51
This issue is almost entirely controlled by our holding in
Delta Smelt. There, the consulting agency—FWS—also used
raw salvage data to set maximum negative flows for the Old
and Middle Rivers. See Delta Smelt, 747 F.3d at 606–07;
2009 Delta Smelt BiOp at 349–50. We determined that the
choice to use raw salvage data was appropriate for three
reasons. First, the agency has substantial discretion to choose
between available scientific models, provided that it explains
its choice. See Delta Smelt, 747 F.3d at 610 (citing Nw. Coal.
for Alts. to Pesticides v. E.P.A., 544 F.3d 1043, 1050 (9th Cir.
2008)). Second, other studies helped inform the specific flow
requirements imposed. Finally, the flow limits “work in
tandem with the incidental take statement (“ITS”), which
accounts for population-level impacts.” Id. at 608.
All three factors are present here. First, the agency
adequately explained why the loss data, although un-scaled
to population, usefully assisted NMFS in identifying whether
and how fish loss relates to negative flow velocity. See 2009
Salmonid BiOp at 360–62; OCAP BA at 13-43–45.
Second, NMFS—like FWS—did not base its maximum
negative flow prescriptions on raw salvage data alone.
Rather, it used the same particle tracking models FWS used
in the Delta Smelt BiOp to evaluate the effect of heightened
exports on naturally buoyant particles. 2009 Salmonid BiOp
at 362–63. It also relied on studies, such as a 2008 study by
Wim J. Kimmerer, to support its conclusion that there exists
a positive relationship between the volume of water exported
from project pumping plants and juvenile salmonid
entrainment at those plants. See, e.g., id. at 361; Delta Smelt,
747 F.3d at 612 (describing how FWS used the same study).
52 SAN LUIS V. LOCKE
Finally, here—like in the Delta Smelt BiOp—the ITS
uses population data to scale incidental take, and the RPA
uses data generated from incidental take to introduce more
restrictive flows in the Old and Middle Rivers. Like the Delta
Smelt BiOp, the Salmonid BiOp sets a range of acceptable
negative flow in the Old and Middle Rivers and requires the
action agency to use population-based data generated from
incidental take to scale in more permissive or restrictive
flows, with a minimum flow of -5,000 cfs. 2009 Salmonid
BiOp at 650.
For these three reasons, the agency acted within its
substantial discretion when it used a non-scaled data model
to set flows in the Old and Middle Rivers.
B. The Challenged Jeopardy Opinion Components Are
Not Arbitrary or Capricious
NMFS determined that the proposed continuing
operations of the Projects are likely to jeopardize the viability
and essential habitat of the listed species. Id. at 575. The
district court invalidated several specific components of this
provision as arbitrary or capricious. See In re Consolidated
Salmonid Cases, 791 F. Supp. 2d at 955–59. Defendants
appeal the district court’s holdings, so we review them here.
1. Winter-Run Chinook
According to the district court, NMFS based its finding
that ongoing CVP/SWP operations jeopardize winter-run
Chinook in part on its determination that winter-run Chinook
is at a “high risk” of extinction. Id. at 864. The court
concluded that NMFS’s “high risk” designation was
“completely unsupported by the record.” Id. As a result, the
SAN LUIS V. LOCKE 53
district court determined that this aspect of the BiOp must be
remanded and fixed.
The district court was incorrect in so concluding. NMFS
did not characterize winter-run Chinook as being at “high risk
of extinction” instead of characterizing the species as being
“not viable.” See id. at 864. Rather, NMFS informed its
designation of winter-run Chinook as “not viable” by
considering Dr. Lindley’s 2007 study, in which he suggests
that winter-run Chinook is at a “high risk” of extinction in
several categories. See 2009 Salmonid BiOp at 85–88.
NMFS discussed the limitations of Lindley’s categories and
explained how it made up for these limitations by relying on
other studies. See id. (citing McElhany et al. (2000),
Liermann and Hilborn (2001), and others). In doing so,
NMFS adequately explained how its various descriptions of
winter-run Chinook as “high risk” influenced its ultimate
jeopardy opinion. Such an explanation is sufficient to satisfy
State Farm’s requirement that the agency consider all
relevant factors and offer an explanation for its conclusion
that is grounded in the evidence. See 463 U.S. at 43. Thus,
this part of the BiOp need not be remanded and fixed.
2. Southern Resident Orca
NMFS concludes in the BiOp that continued CVP/SWP
operations are likely to jeopardize the viability of the
Southern Resident orca. 2009 Salmonid BiOp at 573–74.
The logic supporting this conclusion is relatively simple. The
orca population at issue has fewer than ninety members, and
so NMFS felt compelled to scrutinize “even small effects on
the fitness of individuals that increase the risk of mortality or
decrease the chances of successful reproduction.” Id. at 573.
Winter-run and spring-run Chinook are a critical prey base
54 SAN LUIS V. LOCKE
for the Southern Resident orca. Id. According to NMFS,
reduction in populations of this prey-base jeopardize the
Southern Resident orca because, for example, less food
requires whales to spend too much energy foraging and
“insufficient prey could cause whales to rely upon their fat
stores, which contain high contaminant levels.” Id. NMFS
concluded that continued CVP/SWP operations threaten the
viability of winter-run and spring-run Chinook. Id. at
574–75. This determination led it to also conclude that these
operations jeopardize the Southern Resident orca. Id.
The district court reversed and remanded this conclusion.
It held that NMFS did not consider all relevant factors of the
problem because it failed to discuss a seemingly contrary
finding it made, in a BiOp issued on May 5, 2009, that
commercial ocean “harvest of salmon would not jeopardize
the Southern Resident Killer Whales.” See In re
Consolidated Salmonid Cases, 791 F. Supp. 2d at 864–65.
The district court’s conclusion is incorrect because NMFS
did discuss the 2009 Orca BiOp in the Salmonid BiOp,
showing that it did not “entirely fail[] to consider” an aspect
of the problem. See State Farm, 463 U.S. at 43. NMFS
discussed the 2009 Orca BiOp as part of its baseline analysis.
See 2009 Salmonid BiOp at 218–21. NMFS clarified that the
2009 Orca BiOp—unlike the 2009 Salmonid BiOp—does not
consider the long-term health of Chinook on the continued
viability of the Southern Resident orca, but rather analyzes
the year-to-year impact of commercial harvest on the whales’
short-term food supply. Id. at 218. In this way, NMFS
distinguished the two BiOps as dealing with different time
frames. NMFS’s discussion of how findings in the 2009 Orca
BiOp relate to findings in the 2009 Salmonid BiOp, although
brief, is sufficient to show that NMFS considered the 2009
SAN LUIS V. LOCKE 55
Orca BiOp when developing the 2009 Salmonid BiOp. This
consideration satisfies NMFS’s obligations under State Farm.
See 463 U.S. at 43.
3. Steelhead Critical Habitat
The BiOp makes two relevant conclusions regarding how
the proposed action will adversely modify CV steelhead
critical habitat in the Stanislaus River. First, NMFS
concludes that CV steelhead prefer to spawn when water is
flowing at 200 cfs; proposed deviations from that flow could
reduce spawnable habitat as much as ninety-five percent in
some years. 2009 Salmonid BiOp at 306, 311. Second,
NMFS concludes that continued CVP/SWP operations,
specifically those that dictate flows from the New Melones
and Goodwin Dams, will degrade spawning gravel below the
Goodwin Dam, thereby undermining replenishment efforts.
Id.
The district court found these conclusions to be arbitrary
or capricious. Specifically, with regard to spawnable area,
the district court found that NMFS used “maximum habitat”
as a benchmark for evaluating the Projects’ impacts. That
benchmark was improper because “maximizing” habitat is
not a goal of the ESA. In re Consolidated Salmonid Cases,
791 F. Supp. 2d at 935. The district court also found that no
record evidence supported NMFS’s conclusion that the
CVP/SWP operations cause the recorded gravel loss. Id. at
936.
We side with the agency on both issues. First, NMFS did
not misapply the ESA by relying on a study that sets a goal of
56 SAN LUIS V. LOCKE
“maximizing” habitat.15 The record shows that NMFS looked
to Aceituno (1993) and other studies to determine the point at
which the Projects’ restriction of flows in the Stanislaus River
would “appreciabl[y] reduce[]” habitat, see 2009 Salmonid
BiOp at 42 (citing 50 C.F.R. § 402.02). The record does not
show that NMFS abandoned the ESA’s prescription to “avoid
jeopardy” in favor of Aceiunto’s goal of “maximizing
habitat,” see id. (discussing jeopardy requirement). Rather,
NMFS explained why Aceiunto’s 1993 study provided an
adequate baseline for developing minimum and pulse flows
in the Stanislaus River.16 In providing this explanation,
NMFS satisfied its obligations under the ESA and State
Farm. 463 U.S. at 43.
Second, the record provides adequate support (grounded
in best available science) for NMFS’s conclusion that
CVP/SWP operations negatively impact spawning gravel
quantity and quality. Before construction of dams, channel
forming flows of 8,000 cfs and mobilizing flows of 5,000 to
8,000 cfs created channels—outside of traditional gravel
spawning grounds—in which the river deposited fine
sediment. 2009 Salmonid BiOp at 308 (citing Mesick (2001);
15
It is to be expected that the language of the studies on which an
agency relies will not always track the statutory language of the ESA. Not
all studies are conducted to serve as a basis for section 7 consultation.
Thus, the mere fact that Aceituno’s study seeks to “maximize” CV
steelhead habitat does not require NMFS to disregard it.
16
See Memorandum from Rhonda Reed, Section 7 Biologist, on The
Development of the Reasonable and Prudent Alternatives (RPA) to Avoid
Jeopardy to CV Steelhead in the Stanislaus River, Specifically as it
Relates to Flow and Temperature 2–9 (May 31, 2009) (NMFS biologist
Rhonda Reed describes how NMFS used Aceituno’s suggested minimum
flows as a starting point but altered those flows based on discussions with
agency and stakeholder scientists).
SAN LUIS V. LOCKE 57
Kondolf et al. (2001)). But CVP/SWP operations have all but
halted these flows in recent years. Id. Thus, fine sediment
collects in CV steelhead gravel spawning ground, degrading
the quality of spawning areas. Id. According to Dr.
Kondolf’s 2001 study (upon which NMFS bases much of this
part of the jeopardy opinion), “poor quality of spawning
gravels due to deposition of sand and fine sediment” is one of
four primary factors limiting salmon survival in the
Stanislaus River.17 The specific component of the BiOp
challenged here essentially adopts this conclusion. See 2009
Salmonid BiOp at 308 (citing Kondolf et al. (2001)).
Although NMFS could have done a better job making the
connection between CVP/SWP operations and the quantity of
gravel suitable for CV steelhead rearing, that connection is
fairly discernable from a review of the “whole record.” See
5 U.S.C. § 706; Bowman Transp., 419 U.S. at 286. The
conclusion is, thus, not arbitrary or capricious.
4. Indirect Mortality Factors
The BiOp evaluates the impact of both direct and indirect
mortality factors on listed species. Direct mortality factors,
such as entrainment, are those project components that
directly harm or kill listed species. See 50 C.F.R. § 402.02.
Indirect mortality factors are those caused by continued
operations that do not directly cause the death of listed
species, but lead to it. Those indirect mortality factors
include predation, harm inflicted on native species by
non-native species, pollution, and food limitations. See 2009
Salmonid BiOp at 374. NMFS concludes that CVP/SWP
17
G.M. Kondolf, et al., Reconnaissance-Level Assessment of Channel
Change and Spawning Habitat on the Stanislaus River Below Goodwin
Dam, Rpt. to the U.S. Fish and Wildlife Service 1 (Mar. 22, 2001).
58 SAN LUIS V. LOCKE
operations cause indirect mortality for listed species by
creating conditions in the Delta that favor non-native species,
species that prey on listed Salmonids. CVP/SWP operations
also negatively influence the listed species by lengthening the
time members remain in the interior delta—where they are
exposed to pollution and other indirect mortality
factors—before outmigrating to the ocean. See id.
Plaintiffs challenged this finding at summary judgment.
See In re Consolidated Salmonid Cases, 791 F. Supp. 2d at
869–71. The district court mostly agreed with them, holding
that although NMFS sufficiently established that Delta
hydrologic conditions—as altered in part by the Projects—are
favorable to invasive species, NMFS failed to articulate the
connection among continuing Projects operations, invasive
species, and harm to listed species. See id. at 870 (posing the
following questions: “What effect do these exotics have on
the Listed Species? To what extent does the contribution of
the Projects to the continued presence of these exotics
contribute to the jeopardy finding?”). That failure, according
to the district court, rendered the indirect mortality analysis
arbitrary and capricious. Id. at 870–71.
We disagree. NMFS adequately connected indirect
mortality factors to CVP/SWP operations, thus satisfying its
obligations under the APA and ESA. NMFS’s conclusion
that the Projects’ operations exacerbate Salmonid indirect
mortality proceeds in three steps. First, NMFS explains how,
over the past half century, the Projects’ operations have
worked to degrade the environment in the interior delta,
converting a thriving river system into an unnatural inland
lake-like habitat ill-suited to many native species. This
statement is uncontested. See id. at 870 (“Plaintiffs do not
directly contest the conclusion that the altered hydrologic
SAN LUIS V. LOCKE 59
conditions are favorable for invasive species. Nor do
Plaintiffs challenge the BiOp’s conclusion that CVP and SWP
operations contribute to this ecosystem alteration.”). Second,
NMFS concludes that continued CVP/SWP operations
(specifically pumping from the Jones and Banks facilities)
cause fish outmigrating through the main channels of the
Delta to divert into intersecting channels that split off from
the main rivers and lead towards the inner delta. 2009
Salmonid BiOp at 374. The Projects’ operations cause this
diversion by, among other things, reversing the flows of the
Old and Middle Rivers. Id. at 651 (citing Vogel (2004) to
support the conclusion that “fish chose channels leading south
more frequently when exports were elevated, than when
exports were lower”). Third, fish that are drawn through
intersecting channels and into the inner Delta have a lower
survival rate than fish that remain in the main Delta. Id. at
375. Not all of these fish are killed in pumping plants; many
are eaten by non-native predators, trapped by non-native
plants, or fall prey to pollution in the inner Delta. Id. at
374–81.
The second step provides the critical causal link between
the Projects’ operations and indirect mortality factors that the
district court found lacking. We find that NMFS cited
enough scientific evidence to support its conclusions that high
levels of pumping from the Jones and Banks facilities
influence fish to swim towards the inner Delta where they fall
prey to indirect mortality factors. See id. at 651 (citing Vogel
(2004), SJRGA (2006), SJRGA (2007), SJRGA (2008)).
Although the agency’s analysis is not perfect, it may
60 SAN LUIS V. LOCKE
reasonably be discerned, see Bowman Transp., 419 U.S. at
286, and is thus not arbitrary or capricious.18
C. The Challenged RPA Actions Are Not Arbitrary or
Capricious
We now consider the RPA Actions invalidated by the
district court. Before wading into the specific Actions, we
clarify what the ESA and its implementing regulations require
from an agency when the agency is developing RPAs as part
of a BiOp.
1. The Legal Requirements for an RPA Action
ESA section 7 provides that “[i]f jeopardy or adverse
modification is found [during consultation], the Secretary
shall suggest those reasonable and prudent alternatives which
he believes would not . . . ,” 16 U.S.C. § 1536(b)(3)(A),
18
Nevertheless, we can see where the district court got derailed into
thinking that NMFS blamed continuing CVP/SWP operations for exotics
in the Delta. See In re Consolidated Salmonid Cases, 791 F. Supp. 2d at
870. NMFS essentially makes this statement, without any record support,
in its summary of the indirect mortality component of the BiOp. 2009
Salmonid BiOp at 382. The district court properly questioned this
conclusion: NMFS did not support the assertion that continuing
CVP/SWP operations cause that level of environmental decline. Although
NMFS seemed to say as much on page 382 of the BiOp, the crux of its
indirect mortality argument is in the pages preceding the summary on page
382. NMFS makes clear that the question is not whether “altered project
operations reduce [or exacerbate] the presence of exotics?” but rather
“whether altered project operations could keep more fish in the main delta
where they are less likely to come into contact with exotic species and
die?” As explained, NMFS believes the answer to this question is yes,
and it supported its conclusion by relying on the best available science.
See, e.g., id. at 651.
SAN LUIS V. LOCKE 61
“jeopardize the continued existence of any endangered [or
threatened] species . . .” or result in adverse modification of
critical habitat, id. § 1536 (a)(2). Reasonable and prudent
alternatives are alternative actions identified during formal
consultation that (1) “can be implemented in a manner
consistent with the intended purpose of the action,” (2) “can
be implemented consistent with the scope of the Federal
agency’s legal authority and jurisdiction,” (3) are
“economically and technologically feasible,” and (4) “the
Director believes would avoid the likelihood of jeopardizing
the continued existence of listed species or resulting in the
destruction or adverse modification of critical habitat.” 50
C.F.R. § 402.02. The first three of these factors are the
non-jeopardy factors developed by the agency. The final is
the jeopardy factor, and it is taken from ESA section 7. See
Delta Smelt, 747 F.3d at 634.
Under these provisions, the district court reversed and
remanded several RPA Actions because the agency did not
(1) explain how each RPA Action is “essential to avoid
jeopardy,” In re Consolidated Salmonid Cases, 791 F. Supp.
2d at 922; see also id. at 897 n.26, or (2) explain how each
RPA Action complies with § 402.02’s three non-jeopardy
factors. We recently held in Delta Smelt that these are not the
correct legal standards under which to evaluate an RPA
Action. As we further clarify below, neither section 7 nor
§ 402.02 require NMFS to explain why each Action is
“essential” or to fully elucidate the non-jeopardy factors.
62 SAN LUIS V. LOCKE
a. The ESA Does Not Require NMFS To
Explain How Each RPA Action Is Essential
To Avoid Jeopardy
The district court held that § 402.02 requires NMFS to
show how each RPA Action is essential to avoid jeopardy.19
The effect of this holding was to impose an onerous, highly
precise standard on NMFS under which the district court
invalidated RPA Actions anytime NMFS did not explain why
the Action was necessary, over all others, to preserve the
species. See, e.g., In re Consolidated Salmonid Cases, 791 F.
Supp. 2d at 898.
As we explained in Delta Smelt, neither the ESA nor its
implementing regulations require this level of precision from
the agency. The ESA requires only that the agency impose
RPAs that are “not likely to jeopardize” the species or its
habitat. See 16 U.S.C. § 1536(a)(2), (b)(3)(B). The
regulations interpret this section as requiring the agency to
develop RPAs “that the Director believes” would avoid
jeopardy. 50 C.F.R. § 402.02. This moderate and deferential
language is a far cry from that which would impose a strictly
essential requirement. Rather, this language imposes a
“flexible standard for the consulting agency” that does not
require the Secretary “to explain why he chose one RPA over
another . . . .” Delta Smelt, 747 F.3d at 624 (citing Sw. Ctr.
for Biological Diversity v. U.S. Bureau of Reclamation,
19
The district court articulated its holding as requiring NMFS to explain
how each RPA Action is an “essential component of an overall RPA
designed to avoid jeopardy.” In re Consolidated Salmonid Cases, 791 F.
Supp. 2d at 897 n.26. But in practice, the district court invalidated BiOp
provisions when NMFS failed to explain how they were “essential to
avoid jeopardy . . . .” Id. at 922. That is the holding we review.
SAN LUIS V. LOCKE 63
143 F.3d 515, 523 (9th Cir. 1998)). Under this deferential
standard, the agency need not pick the best RPA or the one
most likely to avoid jeopardy. Id. Rather, we give the
agency flexibility to choose among several appropriate
alternatives. We will uphold that choice so long as it is
reasonably supported based on a review of the record as a
whole. See Sw. Ctr. for Biological Diversity v. U.S. Bureau
of Reclamation, 143 F.3d at 523.
b. The ESA Does Not Require NMFS To
Articulate Compliance with the Non-Jeopardy
Factors
The district court also held that Agency regulations
require NMFS to describe how each RPA Action meets
§ 402.02’s non-jeopardy factors. See In re Consolidated
Salmonid Cases, 791 F. Supp. 2d at 917. It invalidated
several RPA Actions, including Action IV.4.2, for failing to
establish compliance with these factors. We hold that the
district court erred in interpreting § 402.02.
Again, this issue is largely controlled by Delta Smelt. We
said in Delta Smelt that, “[n]othing in § 402.02 obligates the
[consultation agency] . . . to address the non jeopardy factors
when it proposes RPAs. Section 402.02 is a definitional
section; it is defining what constitutes an RPA, not setting out
hoops that the [consultation agency] . . . must jump through.”
747 F.3d at 635. Thus, while “a ‘thorough’ documentation of
jeopardy/adverse modification in the BiOp is always
required, . . . documentation of the non jeopardy factors is
only required when the RPA fails to meet a non jeopardy
factor.” Id. at 635–36. Based on this conclusion, we rejected
the district court’s finding that the agency acted arbitrarily or
capriciously by failing to include “some exposition in the
64 SAN LUIS V. LOCKE
record of why the agency concluded (if it did so at all) that all
four regulatory requirements for a valid RPA were satisfied.”
San Luis & Delta-Mendota Water Auth. v. Salazar, 760 F.
Supp. 2d 855, 957 (E.D. Cal. 2010), aff’d in part rev’d in part
by Delta Smelt, 747 F.3d 581.
This holding applies with equal force here. NMFS is not
required to document its compliance with § 402.02’s non-
jeopardy factors. Rather, it needed only to fairly conclude—
based on the record—that the proposed RPAs do not further
jeopardize the listed species or adversely affect critical
habitats. See Delta Smelt, 747 F.3d at 635. We evaluate
whether it did so below.
2. Challenged RPA Actions
a. Action IV.2.1
The district court invalidated several RPA Actions related
to the San Joaquin Delta. The first, Action IV.2.1, prescribes
San Joaquin River inflow to export ratios between April 1 and
May 31. 2009 Salmonid BiOp at 641. After a brief
adjustment period, Action IV.2.1 requires Reclamation and
DWR to implement specific flow to combined export ratios
on the San Joaquin River (measured at Vernalis, California).
Id. at 643. Those ratios are:
San Joaquin Valley Vernalis flow (cfs): CVP/SWP
Classification combined export ratio
Critically dry 1:1
Dry 2:1
SAN LUIS V. LOCKE 65
Below normal 3:1
Above normal 4:1
Wet 4:1
2009 Salmonid BiOp at 643–44.
The district court invalidated the 4:1 flow-to-export ratio
as arbitrary and capricious. Although it concluded that record
evidence provided support for some flow-to-export ratio, the
district court determined that the agency did not provide
sufficient support for the specific 4:1 flow-to-export ratio.
See generally In re Consolidated Salmonid Cases, 791 F.
Supp. 2d at 894–98.
We disagree with the district court and hold that the
record supports NMFS’s decision to impose the 4:1 ratio.
NMFS bases its decision to impose a 4:1 flow-to-export ratio
primarily on Vernalis Adaptive Management Plan (“VAMP”)
studies of Chinook salmon smolts. 2009 Salmonid BiOp at
644–45. VAMP has tested Salmonid survival based on a 2:1
ratio, but not a 4:1 ratio. Drawing on VAMP and other data
showing a positive correlation between a high-flow-to-low-
export ratio and successful salmonid outmigration,20 NMFS
concluded that “flow to export ratios should be at least 2:1
and preferably higher to increase survival and abundance.”
Stuart 4:1 Memo., supra, at 22. NMFS settled on the 4:1
20
Memorandum from Jeffrey Stuart, NMFS Fisheries Biologist, on The
San Joaquin River “4:1 Flow to Export ratio” Reasonable and Prudent
Alternative (RPA) for the formal section 7 consultation regarding the
Long-Term Operations of the Central Valley Project and State Water
Project 20–21 (June 2, 2009) [hereinafter Stuart 4:1 Memo.].
66 SAN LUIS V. LOCKE
ratio as a high ratio (appropriate in above-normal
precipitation years) by studying historic monthly average
flows at Vernalis. Id. at 16. “This data shows that
approximately 6,000 cfs of flow is available at Vernalis in 50
percent of the wet and above normal water years.” Id. at 17.
Being that the minimum export level to maintain health and
safety is 1,500 cfs, id. at 22, a 4:1 export ratio in wet and
above normal years—although maximally protective of
fish—is traceable to the record. It is within the agency’s
discretion to choose a conservative threshold that will afford
maximum protection to the species so long as that threshold
is fairly supported, which it is. Tenn. Valley Auth. v. Hill,
437 U.S. 153, 184–85, 98 S. Ct. 2279 (1978).
b. Action IV.2.3 and Action IV.3
Actions IV.2.3 and IV.3 specify river flow management
strategies for the Old and Middle Rivers. Although the Old
and Middle Rivers typically flow north, CVP and SWP
pumping reverses that flow, drawing the water south to the
Jones and Banks pumping plants. See Delta Smelt, 747 F.3d
at 606. According to Particle Tracking Model (“PTM”) and
fish tagging studies cited by NMFS, listed fish outmigrating
through the San Joaquin River are vulnerable to diversion
into the channels that lead to the export facilities when
pumping is high and the flow of the Old and Middle Rivers
is very negative. 2009 Salmonid BiOp at 651. These
diverted fish have a lower rate of survival than their
counterparts that bypass the inner Delta and migrate directly
through the outer Delta to the San Francisco Bay.
Actions IV.2.3 and IV.3 seek to mitigate these effects by
imposing negative flow restrictions on the Old and Middle
Rivers. Action IV.2.3 requires the Projects to reduce exports
SAN LUIS V. LOCKE 67
from the Jones and Banks pumping plants between January 1
and June 15 such that the negative flow of the Old and
Middle Rivers is limited to -2,500 to -5,000 cfs, depending on
the presence of salmonids. Id. at 648–52. Action IV.3
requires the Projects to reduce exports between November 1
and December 31 when fish salvage numbers (the numbers of
fish caught at the pumps) meet certain triggers. Id. at 652–53.
The district court invalidated both Actions. It held, with
regard to Action IV.2.3, that the agency did not adequately
explain how imposition of the specific flow requirements in
the Action are “essential to avoid jeopardy.” In re
Consolidated Salmonid Cases, 791 F. Supp. 2d at 909
(citation omitted). It invalidated Action IV.3 because NMFS
based the specific triggers on raw salvage data and “failed to
provide any record explanation for why the specific triggers
were chosen.” Id. at 911.
We again reverse the district court and find that the record
supports NMFS’s decision to impose both Actions. The
record fairly supports NMFS’s imposition of the particular
flow restrictions in Action IV.2.3. PTM modeling cited by
NMFS supports the conclusion that risk of fish entrainment
at pumping facilities increases substantially between -2,500
and -5,000 cfs. 2009 Salmonid BiOp at 652. That same data
shows that the risk of entrainment increases at an even greater
rate with flow restrictions more negative than -5,000 cfs. Id.
Thus, it is reasonable for NMFS to impose the -2,500 to
-5,000 cfs range as a minimum negative flow during times
when salmonids are likely to pass channel openings. The raw
data salvage numbers bolster this conclusion. According to
that data, “[l]oss of older juveniles at the CVP and SWP fish
collection facilities increases sharply at Old and Middle River
flows of approximately -5,000 cfs . . . .” Id. at 361. NMFS
68 SAN LUIS V. LOCKE
explained its rationale for imposing the specific flow
restrictions in Action IV.2.3, and supports that rationale with
what it has determined is the best available science—PTM
studies and raw salvage data.21 It has, thus, satisfied its
procedural and substantive obligations under the APA and
ESA.
Action IV.3 is also fairly traceable to the 2009 Salmonid
BiOp and accompanying studies. NMFS explains that the
triggers imposed by Action IV.3 are developed from previous
work done by DWR, Reclamation, NMFS, and FWS.22 The
specific triggers in Action IV.3 (eight fish/thousand acre feet
or fifteen fish/thousand acre feet) come from data compiled
by NMFS tending to show that when salvage exceeds those
levels, there is a pulse of fish in the system. See Stuart PTM
Memo., supra, at 28 (Fig. 15). The agency’s decision to set
these as particular triggers is based on its own data generated
over nine years, data that is well documented in the BiOp and
supporting memoranda. See id. For that reason, these
particular triggers are not arbitrary or capricious.
c. Action IV.4.2
Action IV.4.2 requires DWR to implement specific
measures to (1) reduce pre-salvage fish loss and (2) improve
21
We have already held, consistent with our opinion in Delta Smelt, that
NMFS acted within its considerable discretion when it elected to use raw
salvage data as a guide for setting certain RPA Actions.
22
Memorandum from Jeffrey Stuart, NMFS Fisheries Biologist, on
Particle Tracking Model results for Old and Middle River flow
manipulation (June 3, 2009) [hereinafter Stuart PTM Memo.] (describing
how the agencies have used a salmon “decision tree” based on salvage
data).
SAN LUIS V. LOCKE 69
salvage efficiency. 2009 Salmonid BiOp at 655. To reduce
pre-salvage loss, the Action requires DWR to “commence
studies to develop predator control methods for Clifton Court
Forebay,” the body of water the fish cross before reaching the
Tracy and Skinner Fish Collection Facilities. Id. at 656. The
Action also sets a specific benchmark for salvage efficiency
at the facilities, requiring DWR to “achieve a minimum 75
percent salvage efficiency for CV salmon, steelhead, . . . and
green sturgeon” at the Skinner Fish Collection Facility. Id. at
655.
Plaintiffs argued that this Action is not technologically or
economically feasible and that the agency thus violated
§ 402.02 by requiring it. The district court agreed. It
concluded that NMFS failed to “cite any record evidence
indicating that the efficiency improvement, albeit a minor
one, is economically or technologically feasible.” In re
Consolidated Salmonid Cases, 791 F. Supp. 2d at 926.
Delta Smelt made clear that the ESA does not require
NMFS to cite record evidence showing that each RPA Action
is economically and technologically feasible. Delta Smelt,
747 F.3d at 635. Thus, NMFS’s failure to cite such evidence
here was not arbitrary or capricious.
d. Action III.1.2
The remaining actions challenged by Plaintiffs,
invalidated by the district court, and challenged here, relate
to CVP/SWP operations on the Stanislaus River, in the east
side of the Central Valley.
Action III.1.2 pertains to the temperature of the Stanislaus
River. According to NMFS, increased temperature in the
70 SAN LUIS V. LOCKE
Stanislaus River threatens the critical habitat of the CV
steelhead. 2009 Salmonid BiOp at 619–20. To remedy this
problem and achieve desired temperatures, Action III.1.2
requires Reclamation to “make cold water releases from New
Melones Reservoir to provide suitable temperatures for CV
steelhead rearing, spawning, egg incubation smoltification,
and adult migration in the Stanislaus River downstream of
Goodwin Dam . . . .” Id. at 620–21. Action III.1.2 includes
an exception to this requirement when the projected
temperatures cannot be achieved. Id. at 621 (describing the
process that Reclamation should use to apply for an exception
to the temperature requirements).
The district court remanded this action to the agency after
determining that the agency did not sufficiently document
“the extent to which this RPA is ‘essential’ to avoiding
jeopardy . . . .” In re Consolidated Salmonid Cases, 791 F.
Supp. 2d at 947–49. More specifically, the court determined
that because the Action includes an exception with “no
limitations” it necessarily does not avoid jeopardy. Id. at 947.
The record does not support the district court’s conclusion
that the “Federal Defendants describe an exception that ‘has
no limitations.’” Id. at 947. NMFS will consider granting an
exception to the temperature requirements only when
Reclamation demonstrates that “after taking all actions within
its authorities, it is unlikely to meet” the temperature
requirements. 2009 Salmonid BiOp at 621. If that happens,
Reclamation must convene the Stanislaus Operations Group
(“SOG”)23 to obtain recommendations on how to proceed.
23
“Reclamation created a Stanislaus Operations Group (SOG) to provide
a forum for real-time operational flexibility and implementation of the
alternative actions defined in the RPA.” NOAA Fisheries, Stanislaus
SAN LUIS V. LOCKE 71
See id. at 621. If the SOG cannot come to a consensus,
NMFS will make recommendations. Reclamation must
satisfy several procedural requirements before NMFS will
grant an exception under Action III.1.2, leading us to
conclude that application of the exception is limited. For this
reason, the record supports NMFS’s conclusion that
imposition of Action III.1.2, notwithstanding its exception, is
likely to avoid jeopardy.
e. Action III.1.3
Action III.1.3 also relates to how CVP/SWP operations on
the Stanislaus River impact the CV steelhead. CV steelhead
adults respond to certain flows in the Stanislaus River as a
natural cue for fall migration. Juveniles depend on a
particular volume of spring flows to assist them in migrating
out of the River to the Delta and eventually to the Pacific
Ocean. Id. at 625. Pulse flows in the Stanislaus River also
benefit CV steelhead habitat by maintaining gravel quality,
promoting channel formation, and enhancing access to varied
rearing habitats. Id. at 624. To better provide these essential
cues and to sustain CV steelhead habitat, Action III.1.3
requires Reclamation to “operate releases from the East Side
Division reservoir to achieve a minimum flow schedule as
prescribed” in the RPA. Id. at 623. The minimum flow
schedule incorporates short periods of high volume flows in
October (fall attraction flows), several times in March and
April (outmigration cue flows), and in May (outmigration
flows). Id.
Operations Group, http://www.westcoast.fisheries.noaa.gov/
central_valley/water_operations/sog.html (last visited Oct. 20, 2014,
4:58 p.m.).
72 SAN LUIS V. LOCKE
The district court invalidated this Action because NMFS
failed to explain why the pulse flows would maintain gravel
quality in the Stanislaus River. In re Consolidated Salmonid
Cases, 791 F. Supp. 2d at 950. It remanded for further
explanation on this point, noting that “[p]articularly in light
of the potentially high water costs of these pulse flows, the
rationale for Action III.1.3 must be lawfully explained and
justified on remand.” Id.
We hold that the district court erred by failing to defer to
the Agency’s interpretation of a scientific study. NMFS
based Action III.1.3’s flow numbers on a 2001 study
conducted by Dr. Kondolf, et al. In that study, Dr. Kondolf,
et al. conclude that “flows around 5,000 to 8,000 cfs are
necessary” to mobilize the channel bed material. Kondolf et
al., supra, at 36. NMFS determined, after weighing the
relevant interests, that implementing pulse flows at the low-
end of Kondolf’s flow range would achieve the appropriate
balance between habitat protection and maintaining water
reserves in the East Side Division Reservoir. See Reed,
supra, at 7–8. In doing so, NMFS balanced Kondolf’s pulse
flow suggestions against Reclamation’s conclusion that
prolonged flows exceeding 1,500 cfs would cause flooding.
See id. Congress delegated this type of balancing to
administrative agencies when it passed the APA and ESA.
See River Runners for Wilderness, 593 F.3d at 1070. As long
as the agency’s decision is properly documented, as it is here,
we will not overturn it. See State Farm, 463 U.S. at 43.24
24
Nor do we overturn NMFS’s choice to use the SJR salmon model to
help prescribe pulse flows on the Stanislaus River. Although it is true that
the SJR model determines flows needed to double salmon population,
NMFS explains why this model was a helpful guide for developing this
RPA. Also, NMFS did not rely exclusively on this model to prescribe
SAN LUIS V. LOCKE 73
f. Action III.2.2
Finally, the district court invalidated RPA Action III.2.2,
which relates to floodplain restoration and innundation flows
in the Stanislaus River. Prior to the construction of the New
Melones Dam in the late 1970s, snow melt from the Sierra
Nevada Mountains created pulse flows in the Stanislaus River
that formed new and scoured existing channels in the riverbed
and surrounding floodplains. 2009 Salmonid BiOp at 627.
CV steelhead juveniles used (and continue to use) these
channels as a rearing habitat. Id. However, the floodplain
habitats that were “inundated before operation of the New
Melones Dam have become fossilized with fine material and
thick riparian vegetation that is never rejuvenated by
scouring,” id., because pulse flows from New Melones Dam
are infrequent. Thus “[f]loodplain juvenile rearing habitat
and connectivity will continue to be degraded by New
Melones operations, as proposed.” Id. To remedy this
impact, Action III.2.2 requires Reclamation to “seek advice
from SOG to develop an operational strategy to achieve
floodplain innundation flows that inundate CV steelhead
juvenile rearing habitat on a one- to three-year return
schedule.” Id. The district court found this action arbitrary
or capricious, holding that because it defines no action per se,
NMFS did not—and in fact could not—perform a feasibility
analysis as required by 50 C.F.R. § 402.02. In re
Consolidated Salmonid Cases, 791 F. Supp. 2d at 952.
Stanislaus River flows. See Reed, supra, at 5–7 (citing, in addition to the
SJR salmon model, Aceituno (1993) and Cramer Fish Sciences (2009)).
Thus, the record does not support the district court’s conclusion that
“[n]othing in the record explains why it is appropriate to use a model
designed to double the existing salmon population to set numeric flow
targets to avoid jeopardy to the CV steelhead.” In re Consolidated
Salmonid Cases, 791 F. Supp. 2d at 950.
74 SAN LUIS V. LOCKE
The district court erred in invaliding this action. We held
in Delta Smelt that § 402.02 does not require the consultation
agency to explain how each Action is feasible. And neither
Plaintiffs nor the district court provide any reason why the
SOG would recommend an action that Reclamation and
DWR could not adopt. See Sw. Ctr. for Biological Diversity
v. U.S. Bureau of Reclamation, 143 F.3d at 523–24 (noting
that feasibility is examined from the perspective of the
agency). Thus, this court has no reason to declare that Action
III.2.2 violates § 402.02’s feasibility factor.
VI. CROSS-APPEAL
Plaintiffs cross-appeal several components of the district
court opinion in which the district court upheld the BiOp. We
affirm the district court on all three cross-appeal issues.
A. NMFS Need Not Distinguish Discretionary and Non-
Discretionary Actions
ESA section 7 provides that, after an agency seeks
consultation on a potential project, the agency providing
consultation shall write a BiOp “detailing how the agency
action affects the species or its critical habitat.” 16 U.S.C.
§ 1536(b)(3)(A). To determine how agency action affects
listed species, the consulting agency must analyze the action
in relation to the “environmental baseline.” 50 C.F.R.
§ 402.02. “This baseline is intended to form a basic
‘snapshot’ of the status of the species at a particular moment
in time before the action is taken.” Liebesman & Petersen,
supra, at 46.
Plaintiffs argue that NMFS must separate discretionary
aspects of the Projects from non-discretionary aspects of the
SAN LUIS V. LOCKE 75
Projects to define the environmental baseline. The district
court disagreed, holding that “[n]othing in the law requires
NMFS to segregate discretionary aspects of coordinated
Project operations from non-discretionary ones in the manner
Export Plaintiffs demand.” In re Consolidated Salmonid
Cases, 791 F. Supp. 2d at 852.
Plaintiffs’ discretionary/non-discretionary argument is the
same argument that we entertained and rejected in Delta
Smelt. See 747 F.3d at 638–40. We again reject these
arguments and affirm the district court on this point.
B. The Biological Opinion’s Indirect Mortality Factors
Are Direct Effects Under the ESA
For the purposes of ESA section 7 consultation, the
“effect” of a proposed action includes both direct and indirect
effects. 50 C.F.R. § 402.02. To show that something is an
indirect effect of the proposed action, an agency must
demonstrate (1) that it is caused by the action, (2) that it is
later in time than the action, and (3) that it is reasonably
likely to occur. Handbook, supra, at 4-27 (citing 50 C.F.R.
§ 402.02). Whether NMFS needed to make these findings
with regard to “indirect mortality factors” identified in the
BiOp, see 2009 Salmonid BiOp at 374, is a key issue on
cross-appeal.
NMFS concludes in the BiOp that CVP/SWP operations
subject listed species to indirect mortality factors—such as
predation and exposure to toxins—in the inner Delta. See
generally id. at 374–82. The district court determined that
“the indirect mortality findings challenged by Plaintiffs do
not constitute ‘indirect effects’” within the meaning of 50
C.F.R. § 402.02 because they are “caused by the action
76 SAN LUIS V. LOCKE
subject to consultation, not by some other action . . . .” In re
Consolidated Salmonid Cases, 791 F. Supp. 3d at 868 (citing
Handbook, supra).
We agree with the district court. Indirect effects are
typically more attenuated than those described in the 2009
BiOp. National Wildlife Federation v. Coleman provides a
clear, oft-cited example of an “indirect effect.” 529 F.2d 359,
373 (5th Cir. 1976). There, the Fifth Circuit held that the
Department of Transportation must consider the residential
and commercial development “that can be expected to result
from the construction of the highway” as an indirect effect of
highway construction. Id. NMFS and FWS provide another
example of an indirect effect in the consultation handbook.
See Handbook, supra, at 4-29. This example is a little bit
closer to home:
A very complex example of indirect effects
arose in determining effects of renewing water
services contracts . . . in the San Joaquin
Basin . . . . Upon checking with other Federal
and State agencies, the FWS determined that
the distribution of water for agricultural use
on the higher east side of the Valley provided
a hydrologic head maintaining the
groundwater table on the west side of the
Valley at a level making it economical to
pump.
Id. As a result, residents could use the pumped water to
convert the land to agriculture. But the conversion of the land
to agriculture destroyed the habitat of several listed species.
Id. FWS considered this an indirect effect of renewing the
water services contracts. Id. These two examples show that
SAN LUIS V. LOCKE 77
an indirect effect—as envisioned by 50 C.F.R. § 402.02—is
one that the action makes possible (or indeed, more
probable), but does not directly cause.
The indirect mortality factors described in the BiOp are
direct effects. According to NMFS, CVP/SWP operations
draw listed fish into the inner Delta by reversing the flows of
the Old and Middle Rivers. See 2009 Salmonid BiOp at
361–62. NMFS concludes that the interior Delta is a
dangerous place for migrating salmonids partially because of
Project operations. See id. at 374–75, 433. These effects
occur concurrently with the Projects; they are not future
“indirect” actions “reasonably certain” to occur. See
50 C.F.R. § 402.02.
C. Reclamation Is Not Independently Liable Under the
ESA
Plaintiffs’ argument that Reclamation is independently
liable under the ESA is predicated on a finding that the BiOp
is legally flawed. See Pyramid Lake Paiute Tribe of Indians
v. U.S. Dep’t of the Navy, 898 F.2d 1410, 1415 (9th Cir.
1990) (compliance with a BiOp satisfies an action agency’s
procedural obligations under the ESA, but it does not satisfy
the agency’s substantive obligation to comply with section 7).
Because we hold that the BiOp is legally sound, we dismiss
Plaintiffs’ argument.
VII. CONCLUSIONS
Based on the foregoing, we REVERSE the components of
the district court’s opinion in which it invalidated the BiOp
and AFFIRM the district court with regard to the three issues
78 SAN LUIS V. LOCKE
on cross-appeal. We REMAND for entry of summary
judgment in favor of defendants.
Each party shall bear its own costs.
REVERSED IN PART, AFFIRMED IN PART, AND
REMANDED.
SAN LUIS V. LOCKE 79
GLOSSARY OF TERMS
anadromous fish fish that ascend rivers from the sea
for breeding
APA Administrative Procedure Act
BA Biological Assessment
Bay-Delta San Francisco Bay and Sacramento-
San Joaquin Delta
BiOp 2009 Salmonid Biological Opinion
cfs cubic feet per second
CVP Central Valley Project
CVPIA Central Valley Project Improvement
Act
DWR California Department of Water
Resources
ESA Environmental Species Act
IFIM incremental flow instream
methodology
ITS Incidental Take Statement
80 SAN LUIS V. LOCKE
listed species (1) the Sacramento River winter-run
Chinook salmon (“winter-run
Chinook”); (2) the Central Valley
spring-run Chinook salmon (“spring-
run Chinook”); (3) the Central
Valley steelhead (“CV steelhead”);
(4) the threatened Southern Distinct
Population Segment of North
American green sturgeon (“green
sturgeon”); and the Southern
Resident killer whales (“Southern
Resident orcas”).
NMFS National Marine Fisheries Service
the Projects Central Valley Project and State
Water Project
PTM Particle Tracking Model
RPA reasonable and prudent alternatives
Reclamation U.S. Bureau of Reclamation
SOG Stanislaus Operations Group
SWP State Water Project
VAMP Vernalis Adaptive Management Plan