San Luis & Delta-Mendota Water Auth. v. Locke

San Luis & Delta-Mendota Water Auth. v. Locke, No. 12–15144, 2014 WL 7240003 (9th Cir. Dec. 22, 2014) (holding that (i) the district court abused its discretion by improperly admitting extra-record declarations and substituting its own analysis for the Agency’s opinion; (ii) National Marine Fisheries Service (“NMFS”) acted within its discretion by using a non-scaled data model to set river flows where it adequately explained its decision and used additional studies to validate its decision; (iii) NMFS did not act arbitrarily or capriciously when determining the State Water Project’s and the Central Valley Project’s (“the Projects”) continued operations were likely to jeopardize viability and essential habitat of species because it demonstrated sufficient research to support its conclusions; and (iv) NMFS’s various reasonable and prudent alternative (“RPA”) recommendations and requirements were not arbitrary or capricious).

Over twenty-five million agricultural and domestic users in California’s arid Central Valley rely on the government’s extraction of water from its rivers. However, this extraction dramatically alters the rivers’ natural states and threatens the viability of the species that depend on the water. To resolve this conflict, the U.S. Department of the Interior’s Bureau of Reclamation (“Reclamation”) ordered the NMFS to evaluate under the Endangered Species Act (“ESA”) how the Projects’ continuing water extraction would impact certain endangered Salmonid species in the rivers.

In a 2009 Biological Opinion (“BiOp”), NMFS found that continuing extraction threatens species and proposed a solution. In response to the proposed remedies, San Luis & Delta–Mendota Water Authority and Westlands Water District (“the Water Districts”) filed suit against the Department of Commerce, the National Oceanic and Atmospheric Administration and NMFS (collectively, “the Federal Defendants”), to challenge whether the 2009 BiOp was developed arbitrarily or capriciously. On summary judgment, the United States District Court for the Eastern District of California (“district court”) found that NMFS violated the Administrative Procedure Act’s (“APA”) arbitrary or capricious standard when developing the BiOp and granted in part and denied in part the claims of both parties. The United States Court of Appeals for the Ninth Circuit (“court”) reviewed this case on appeal.

On appeal, the Federal Defendants asked the court to overturn the components of the BiOp that the district court struck down, and the Water Districts asked the court to overturn the district court’s holdings that were favorable to the BiOp. Before reviewing the 2009 BiOp, the court resolved the initial question of whether the district court erred in its own record review by supplementing the administrative record with extra-record declarations. The court noted that in making its determinations, a district court may only admit extra-record evidence to further understand whether an agency complied with the APA’s arbitrary or capricious standard. However, the district court heavily relied on extra-record scientific opinion to evaluate and question the 2009 BiOp. The court found the district court violated the general rule limiting a court’s review of agency action to the administrative record. Specifically, the court reasoned that the district court erroneously substituted the extra-record declarations for NMFS’s own analysis.

The court next determined whether NMFS complied with the procedural requirements of the APA. Because the ESA does not have its own standard of judicial review, the court evaluated the BiOp under the APA’s deferential arbitrary and capricious standard wherein a court will sustain an agency’s actions if there is a rational connection between the facts and conclusions. In determining the best method to prevent endangered fish species from being caught in a negative flow resulting from pumping the rivers, NMFS used raw salvage data from fish salvage facilities to provide a reasonable and prudent alternative (“RPA”) in its BiOp. The Water Districts challenged the use of raw salvage data instead of data scaled to fish populations and the district court held that using raw salvage data went against the grain of traditional science. Referencing its decision in San Luis & Mendota Water Authority v. Jewell (“Delta Smelt”), where the consulting agency also used raw salvage data, the court held that NMFS’s choice to use raw salvage data was within its substantial discretion. In Delta Smelt, the court determined an agency has substantial discretion to choose whichever available scientific model it wants to use. The court noted that NMFS adequately explained why the loss data usefully assisted it in identifying whether and how fish loss relates to negative flow velocity. Also, NMFS did not base maximum negative flow prescriptions on raw data alone, using other studies to help decide the specific flow requirements imposed. Finally, similar to the Delta Smelt BiOp, the incidental take statement (“ITS”) in this case used population data to scale incidental take, and the RPA used that information to create its flow restrictions. For these reasons, the court held that NMFS acted within its discretion in using non-scaled data models to restrict flows.

Next, the court considered whether NMFS arbitrarily or capriciously determined that the Projects’ continuing activity on the river would jeopardize the viability and the essential habitat of the species. First, the district court determined that NMFS’s designation of the winter-run Chinook as a species at a “high risk” of extinction was unsupported by the record. The court held that NMFS’s explanation of how a 2007 study of the winter-run Chinook influenced its opinion was sufficient to satisfy the requirement that an agency consider all relevant factors and offer an explanation for its conclusions. Second, the district court held that NMFS did not consider all of the relevant factors when it determined the Projects jeopardized orca viability because it failed to address a 2009 Orca BiOp’s contrary finding. The court found that NMFS did in fact discuss the 2009 Orca BiOp and distinguished the two different outcomes as dealing with different time frames and yielding different results. This was sufficient to show that NMFS considered the 2009 Orca BiOp when developing the 2009 Salmonid BiOp. Third, the district court found that NMFS’s conclusions that the Projects’ deviations from CV steelhead preferred spawning water levels significantly reduced spawnable habitat were arbitrary and capricious. The district court stated that NMFS’s use of “maximum habitat” as a benchmark for spawnable area was not a goal of the ESA. However, the record showed that NMFS looked to several studies to determine the point at which the Projects’ restriction of flows on the river would appreciably reduce habitat. The court found that NMFS explained why the studies provided an adequate baseline for developing minimum flows in the river and thus satisfied its obligations under the ESA. Additionally, the record adequately cited studies supporting NMFS’s conclusion that the Projects’ operations negatively impacted spawning gravel quality and quantity. Accordingly, the court held that these findings were not arbitrary or capricious.

Next, the district court held that NMFS failed to articulate the connection between the Projects’ operations, invasive species, and harm to the endangered species. However, the 2009 BiOp found: (i) the Projects’ operations were degrading the environment in the interior delta, making it ill-suited to many native species; (ii) continued Projects operations would cause fish outmigration through the main channels of the delta to divert into intersecting channels that split off from the main rivers; and (iii) 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. Noting that NMFS’s analysis was not perfect, the court held that evidence to support NMFS’s conclusions could be reasonably discerned and that its analysis was therefore not arbitrary or capricious.

Finally, the court determined whether the challenged RPA actions were arbitrary or capricious. The district court reversed and remanded several of the BiOp’s RPA actions because NMFS did not explain how each RPA action was essential to avoiding jeopardy or how each action complied with the ESA’s non-jeopardy factors. Specifically, the district court found that NMFS failed to explain how each RPA action could be implemented in a manner consistent with the intended purpose of the action, consistent with the scope of the Agency’s authority and jurisdiction, and in an economically and technologically feasible way. The court noted that the district court’s analysis was erroneous. In so concluding, the court relied on the holding in Delta Smelt, which explained that the ESA only requires that an agency impose RPAs that are “not likely to jeopardize” the species or its habitat, rather than those that are “essential to avoiding jeopardy.” Applying that standard here, the court found that NMFS was not required to meet all of the non-jeopardy factors but only to conclude that the proposed RPA would not further jeopardize the listed species.

The district court invalidated several RPA actions on the grounds that they were not supported by scientific evidence, were not economically or technologically feasible, or were made arbitrarily or capriciously. Finding that the record showed sufficient evidence or reasonable support for all of the actions, the court reversed all of the district court’s holdings. In so doing, the court reasoned that Action IV.2.1, requiring Reclamation and DWR to implement specific flows on the San Joaquin River to a higher rate to increase survival and abundance, was traceable to the record even if also maximally protective of fish. The court found that Actions IV.2.3 and IV.3, reducing exports from two pumping plants from January through June and November through December, respectively, to mitigate the adverse effects of the negative flows on the Salmonid fish species migrating during those time frames, were also supported by data cited in the record. It upheld Action IV.4.2, requiring the California Department of Water Resources to implement specific measures to reduce pre-salvage fish loss and improve salvage efficiency, because contrary to the district court’s analysis, the ESA did not require NMFS to cite record evidence showing economic and technological feasibility. The court further upheld Action III.1.2, requiring Reclamation to make cold water releases from the New Melones Reservoir to provide more suitable temperatures for the CV steelhead to spawn, as supported by the record. Here, the court noted that Action III.1.2’s exception for when the projected temperatures could not be achieved was sufficiently limited in application because Reclamation must satisfy several procedural requirements before NMFS would grant an exception. The court found that NMFS’s decision to recommend Action III.1.3, requiring Reclamation to operate releases from the East Side Division reservoir to achieve a minimum flow schedule to help sustain the CV steelhead habitat, was properly documented. It reasoned here that the district court erred by failing to defer to the Agency’s interpretation of a scientific study. Finally, the court upheld Action III.2.2, requiring Reclamation to collaboratively develop an operational strategy to achieve floodplain inundation flows that would help restore floodplains and CV steelhead habitat in the Stanislaus River, as within NMFS’s discretion. The court emphasized again that the Agency was not required to explain the Action’s feasibility.

Finally, the court affirmed the portions of the district court’s holdings that upheld the 2009 BiOp. In so doing, the court echoed its holding in Delta Smelt that an agency need not distinguish between discretionary and non-discretionary actions. In holding that the BiOp’s indirect effects on fish mortality were actually direct effects requiring no further elaboration, the court noted that the effects occurred concurrently with the Projects and were therefore direct effects. Finally, the court held that Reclamation was not independently liable under the ESA because the BiOp was legally sound.

The court reversed the district court’s holdings that invalidated the BiOp and affirmed the district court with regard to the issues on cross-appeal. Accordingly, the court remanded for entry of summary judgment in favor of the Federal Defendants.


The title image features an aerial view of the Sacramento Valley in northern Central Valley of California. This image is licensed under the Creative Commons Attribution Share-Alike 2.0 Generic license and the owner does not endorse this blog.