Survival of Summer Steelhead Caught and Released from an Experimental Seine Fishery in the Lower Columbia River

Categories:

Published: July 2016

Pages: 62

Publication number: FPT 16-10

Author(s): Dan Rawding, Ann Stepheseon, Josua Holowatz, Ben Warren and Mara Zimmerman


Executive Summary

A study to determine the survival of adult summer steelhead released from commercial fishing gear occurred from 2011 to 2013. Steelhead were incidentally captured in an experimental Columbia River salmon beach and purse seine fishery (fishery) below Bonneville Dam (BON). Survival estimates were divided into short-term (between the fishery and BON), intermediate-term (between Bonneville and McNary (MCN) dams including tributaries and fisheries between BON and MCN), and cumulative (between the fishery and MCN, including tributaries and fisheries between BON and MCN). Steelhead survival was estimated based on a Ricker-Two-Release (R2R) study design. Summer steelhead were tagged with Passive Integrated Transponder (PIT) tags prior to release from the seines and tag "recoveries" were detections of the PIT tags at mainstem Columbia River dams, tributary interrogation sites, and in fisheries above BON. A control group of fish was captured at BON, PIT tagged, trucked downstream, and released into the fishery area.

Based on R2R model selection, the most parsimonious model for short-term and intermediate-term survival included a constant survival probability across years and seine type, and a capture probability that varied by year. For summer steelhead released from seines, the short-term survival was estimated to be 97.8% (95% CI: 96.4%-99.2%). Two alternate methods were used to estimate short-term survival. The first method used adult steelhead captured in seines and at BON that were previously tagged as juveniles and analyzed using the same Ricker-Two-Release design, which resulted in a survival estimate of 97.1%. The second method estimated the survival of seine caught steelhead, without a control group, by adjusting survival based on tag effects and tag detection efficiency at BON. The pooled survival estimate using this method was 96.8%. The weight of evidence from the three different analyses provides very strong support that the short-term survival of steelhead is greater than 96%. Based on the best model the intermediate-term survival was 99.8% (95% CI: 98.4%-100%). A second intermediate-term survival estimate calculated using previously tagged juvenile was 96.2% (95% CI: 77.4%-100%). Based on the overlap of the 95% CI there was no difference in the survival estimates using both approaches. This suggests that almost all the fishery mortality occurred between the fishery and BON, which represents 21 km of river and typically takes steelhead 1.9 days to navigate. The cumulative survival of steelhead from the fishery to MCN including the fishery and tributary PIT tag detections was 97.5% (95% CI: 95.7%-99.0%).

Logistic regression was used to explore variables that may explain the recovery of steelhead from the fishery to BON. We considered tagging location (control vs. treatment), origin (natural vs. hatchery), group (A vs. B steelhead), and water temperature as possible covariates with year modeled as a random effect. The most important covariates in the model were the random effect of year and tagging location. The best model, which included the intercept and year, had a 0.95 posterior model probability. Analysis of contingency tables indicated very strong support for association between recovery at BON and injury. During seining, fish that were classified as wedged/gilled or tangled in the seines were 11.7 and 1.9 times less likely to be recovered than fish that were uninjured. The condition of each fish was also classified at both capture and release. Fish that were classified as lethargic at the time of capture or release were 2.3 times less likely to be recovered than those fish classified as vigorous. These results suggest that fish injury and condition are associated with the recovery of steelhead at BON.

Our adult steelhead survival results are similar to those reported from seining on the Rogue River (96%) but generally much higher than those reported in the literature for salmon caught in seines or tangle nets. We believe the low number of steelhead (mean=1.24) and salmon (mean=11.49) caught per set, short set times, and the strict fish handling guidelines contributed to the high survival rates observed in our study. Due to a combination of fish behavior, gear type, and soak time, 96.5% of all captured steelhead showed no signs of injury and only 1.5% of fish had injuries associated with high mortality (e.g. gilled or wedged). In addition, most fish were individually removed from the seine with rubber nets, which limited the typical seine mortalities resulting from fish being crushed or compressed by hoisting large numbers of fish from the water. The high survival rate of summer steelhead caught and released through seines determined through this analysis provides evidence that seines, when employing best management practices, can be an effective management tool in minimizing steelhead bycatch mortality in seine fisheries targeting salmon.

Suggested citation