Early life history attributes and run composition of PIT-tagged wild subyearling chinook salmon recaptured after migrating downstream past lower granite dam

Abstract

Seaward migration timing of Snake River fall chinook salmon (Oncorhynchus tshawytscha) smolts is indexed using subyearling chinook salmon passage data collected at Lower Granite Dam. However, not all of the subyearlings are fall chinook salmon. For six years, we recaptured wild subyearling chinook salmon smolts, which had been previously PIT tagged in the Snake River, to genetically determine if the fish were offspring of spring and summer (hereafter, spring/summer), or fall chinook salmon. Springfall chinook salmon comprised over 10% of the samples of recaptured smolts in five of six years. For these five years, we used discriminant analysis to determine run membership of PIT-tagged smolts that were not recaptured (i.e., not sampled for genetic identification). Accuracy of the discriminant analysis models, based on genetically identified smolts, varied between 75 and 85%. After using discriminant analysis to classify run membership for each PIT-tagged smolt that was not genetically identified, we compared early life history attributes between fall and spring/summer chinook salmon and calculated annual run composition. The life history attributes we studied overlapped, but spring/summer chinook salmon reared along the shoreline of the free-flowing Snake River earlier, were larger, and began seaward migration earlier than fall chinook salmon. Spring/summer chinook salmon made up from 15.1 to 44.4% of the tagged subyearling smolts that were detected passing Lower Granite Dam. As a result, the presence of spring/summer chinook salmon makes migration timing for the fall chinook salmon seem earlier and more protracted than is the case. If wild subyearling spring/summer chinook salmon smolts are not considered, fall chinook salmon abundance at Lower Granite Dam will be overestimated