Design and analysis of salmonid tagging studies in the Columbia Basin. Volume 13. Appraisal of system-wide survival estimation of Snake River yearling chinook salmon released in 1997-98, using PIT-Tags recovered from Caspian Tern and Double-Crested Cormorant breeding colonies on Rice Island

摘要

PIT-tags recovered from tern and cormorant breeding colonies at Rice Island and observations from the interrogation systems at John Day and Bonneville Dams were incorporated into survival analyses. Whether the estimates for the upper reaches of the system, between Lower Granite and McNary Dams were as expected (with weighted averages S(sub LGR-LGS) = 0.996, S(sub LGS-LMN) = 0.837, and S(sub LMN-McN) = 0.941), those for the lower reaches, between John Day and Bonneville Dams, appeared positively biased with survival estimates typically greater than 1. Their weighted averages were S(sub McN-JDA) = 0.707 and S(sub JDA-BON) = 1.792 for 1997 releases. For the 1998 releases, they were S(sub McN-JDA) = 0.795 and S(sub JDA-BON) = 1.312. If the estimates for the lower reaches were biased, the estimates for the whole project would also be biased (S(sub LGR-BON) = 0.819). We determined that bias could have arisen if the terns and cormorants of Rice Island fished for salmon yearlings in waters of the BON-Rice reach at low rates (M(sub BON-Rice) (le) 0.2), and the rates of tag-deposition and tag-detection were low (R(sub D) x R(sub R) (le) 0.4). Moreover, unknown levels of uncensored post-detection mortality and scavenging of previously dead salmon yearlings may have also added to the bias.