Research, monitoring, and evaluation of emerging issues and measures to recover the Snake River Fall Chinook Salmon ESU: January 2019 - December 2019

Abstract

The portion of the Snake River fall Chinook salmon Oncorhynchus tshawytscha evolutionary significant unit (ESU) that spawns upstream of Lower Granite Dam transitioned from low to high abundance during 19922019 in association with U.S. Endangered Species Act recovery efforts and other federally mandated actions. This annual report focuses on (1) numeric and habitat use responses by natural- and hatchery-origin spawners, (2) phenotypic and numeric responses by natural-origin juveniles, and (3) use of a small unmanned aerial system (sUAS) to search for fall Chinook salmon redds and carcasses. Spawners have located and used most of the available spawning habitat and that habitat is gradually approaching redd capacity. Timing of spawning and fry emergence have been relatively stable, but effects of density dependence are evident in juvenile life stages. Apparent abundance of juvenile fall Chinook salmon has increased and we noted the following responses: parr dispersal from riverine rearing habitat into Lower Granite Reservoir has become earlier; growth rate (g/d) and dispersal size of parr declined; and passage timing of smolts from the two Snake River reaches has become earlier and downstream movement rate faster. These findings coupled with stock-recruitment analyses presented in this report provide evidence for density-dependence in the Snake River reaches and in Lower Granite Reservoir that was influenced by the expansion of the recovery program. The long-term goal is to use this information in a comprehensive modeling effort to conduct action-effectiveness and uncertainty research and to inform Fish Population, Hydrosystem, Harvest, Hatchery, and Predation and Invasive Species Management Research, Monitoring, and Evaluation (RM&E) progams. In 2019, the U.S. Geological Survey (USGS) shifted survey efforts in the Snake River toward deepwater redd searches and fish collection for parentage-based tagging (PBT) analyses because all unmanned aerial system (UAS) activities were suspended by the Department of the Interior two weeks into the spawning season. We counted 81 deepwater redds at 17 of the 29 sites surveyed. Redd depths averaged 3.6 m. We collected a total of 123 live fall Chinook salmon from 16 unique geographic locations that spanned 55 river kilometers. Forty-six fish were recovered at Eureka Bar (rkm 307.1) and Kirby Creek (rkm 352.0), which accounted for 37% of all collected fish in 2019. Most (73 fish) post-spawned salmon were collected from early to mid-November just after peak spawning. A summary of 2019 PBT results can be found in Appendix A.1. In 2019, we PIT tagged subyearling fall Chinook salmon in both the Snake and Clearwater rivers. In the Snake River, we tagged 410 fish with 8-mm tags, 666 fish with 9-mm tags, and 1,082 fish 12-mm tags. During seining, our recapture rate of previously tagged fish was slightly higher in the lower reach at 11.4% than in the upper reach at 10.6%. In an effort to represent more of the population through tagging, we tagged fish as small as 45 mm with 8-mm tags at one site in the upper reach. This allowed us to increase the number of fish tagged in that reach by 17.7%. An additional 10.6% of collected fish could have been tagged in the lower reach had we used 8-mm tags in that reach. In the Clearwater River, we tagged 2,451 subyearlings and recaptured 260 (10.6%) fish in the river and 66 fish (48 tagged by USGS, 18 tagged by the Nez Perce Tribe) at Lower Granite Dam during October to provide information for growth estimation. Within riverine habitats, growth in both length and mass were higher for fish tagged with 8-mm tags than with 9- and 12-mm tags. Estimated growth in length and mass of subyearlings was higher in Lower Granite Reservoir than in riverine habitats. We adapted existing statistical models used to estimate abundance of steelhead and spring/summer Chinook salmon for fall Chinook salmon passing Lower Granite Dam. Run reconstruction efforts to date at Lower Granite Dam for Snake River fall Chinook salmon have provided estimates of the number of returning adults but with no measure of uncertainty about the estimates. The objective of this study was to estimate the abundance, with uncertainty, of marked (coded-wire tagged CWT or adipose clipped) and unmarked fall Chinook salmon past Lower Granite Dam for return years 20032018. Estimating uncertainty is important for informing the state-space life cycle model (Chapter 5), which incorporates both observation and process uncertainty into parameter estimates. The coefficient of variation (CV) for log- abundance was 1.0% or less in all years, whereas the CV for abundance averaged 4.2% and ranged from 1.4% to 10.4% among years. Over the past five years, we have been developing a two-stage state-space life-cycle model for naturally produced fall Chinook salmon in the Snake River basin. Initial efforts focused on generating juvenile and adult abundance estimates, with estimates of uncertainty, for informing the life-cycle model. In this report we 1) describe the statistical life-cycle model and improvements made since our last report to the Independent Scientific Advisory Board (ISAB), 2) estimate the effects of covariates on key demographic parameters, and 3) use the fitted life- cycle model to simulate population trajectories under hydrosystem actions proposed for the NOAA 2020 Biological Opinion (hereafter, the Proposed Action). Major recent advancements to the model include revised juvenile abundance estimates, the ability to estimate smolt-to-adult return rates (SAR) separately for subyearling and yearling juvenile fall Chinook salmon, and improvements in the observation model for estimating age, sex, and outmigration structure in adult returns. We examined the effect of numerous environmental, hydrosystem, and ocean covariates on key demographic parameters but only a few covariates were significant. For the adult-to-juvenile transition, we found the maximum weekly river flows during the winter egg- incubation period had a significant negative effect on the juvenile outmigrant abundance from that brood year. For subyearling outmigrants, percent spill during the summer had a significant positive effect on SAR and the mean winter PDO (Pacific Decadal Oscillation) had a significant negative effect on SAR. For yearling outmigrants, NPGO (North Pacific Gyre Oscillation) had a significant positive effect on SAR. We used the fitted model to simulate population trajectories under the Proposed Action, and the median 10-year geometric mean abundance was 8,222 female spawners (interquartile range: 2,592 26,714). Overall, the probability of quasi- extinction (probability of falling below 50 female spawners for 4 consecutive years) was low, with only 1.6% of all simulations having a quasi-extinction probability >0.95. Although quasi-extinction probability was low, we did not assess the additional effect of climate change, which would be expected to increase quasi-extinction probability.