Abstract
Upstream migration rates were assessed for 1801 radio‐tagged adult spring–summer Chinook salmon Oncorhynchus tshawytscha through 12 unimpounded river reaches in the Columbia River basin from 1997 to 2002. Reaches were 36 to 241 km long (mean = 130 km) and included sections of the large Columbia and Snake Rivers and smaller free‐flowing tributaries. Median Chinook salmon migration rates ranged from −1 in the Deschutes and Clearwater Rivers to >35 km day −1 in the Columbia and Snake Rivers. Using multivariate analyses, migration date explained the most variance in Chinook salmon migration rates while river discharge, migration year and migration reach were secondary. Both within and between years, Chinook salmon migrated more rapidly as migration date increased and more slowly when discharge was high. Arrival at high elevation spawning grounds at appropriate times and increased metabolic activity and reproductive maturation may explain the greater power of migration date, relative to river discharge, in predicting migration rates of Columbia basin spring–summer Chinook salmon.