Little information currently exists on habitat use by subyearling fall Chinook salmon Oncorhynchus tshawytscha rearing in large, main‐stem habitats. We collected habitat use information on subyearlings in the Hanford Reach of the Columbia River during May 1994 and April–May 1995 using point abundance electrofishing. We analyzed measures of physical habitat using logistic regression to predict fish presence and absence in shoreline habitats. The difference between water temperature at the point of sampling and in the main river channel was the most important variable for predicting the presence and absence of subyearlings. Mean water velocities of 45 cm/s or less and habitats with low lateral bank slopes were also associated with a greater likelihood of subyearling presence. Intermediate‐sized gravel and cobble substrates were significant predictors of fish presence, but small (<32‐mm) and boulder‐sized (>256‐mm) substrates were not. Our rearing model was accurate at predicting fish presence and absence using jackknifing (80% correct) and classification of observations from an independent data set (76% correct). The habitat requirements of fall Chinook salmon in the Hanford Reach are similar to those reported for juvenile Chinook salmon in smaller systems but are met in functionally different ways in a large river.

Subyearling fall chinook salmon's Oncorhynchus tshawytscha use of unaltered and riprap habitats in Lake Wallula of the Columbia River was determined with point abundance data collected by electrofishing in May 1994 and 1995. We documented the presence or absence of subyearlings at 277 sample sites and collected physical habitat information at each site. Based on logistic regression, we found that the probability of fish presence was greater in unaltered shoreline habitats than in riprap habitats. Substrate size was the most important factor in determining fish presence, with dominant substrates larger than 256 mm having the lowest probability of fish presence. Water velocity, also included in our model due to its biological importance, was not a significant factor affecting presence or absence (P = 0.1102). The correct prediction rate of fish presence or absence in our sample sites using cross validation was 67%. Our model showed that substrate was the most important factor determining subyearling habitat use, but the model did not include other habitat variables known to be important to subyearlings in more diverse systems. We suggest that resource managers consider alternative methods of bank stabilization that are compatible with the habitat requirements of the fish that use them.