Implementation of a modified small stream juvenile salmonid survey design

Several modifications to the Hankin and Reeves design (Hankin and Reeves,1988; Dolloff et al., 1993) for estimating abundance of juvenile salmonids in small streams are proposed, field tested, and evaluated. Adaptive Sequential Independent Sampling replaces systematic sampling. The Method of Bounded Counts is utilized to calibrate dive counts when the first phase dive count is small (20 fish), whereas electrofishing with removal method estimation is used at larger population sizes. In addition to shallow pools, riffles, and runs a slightly more complex habitat stratification includes (a) habitat that cannot be effectively sampled, such as cascades and shallow, unproductive riffles, and (b) a new Deep Pool stratum consisting of pool habitat that is too deep to electrofish. Direct observation is eliminated in riffles because previously gathered data (USDA Forest Service, unpublished) shows an inconsistent relationship between dive counts and subsequent removal method estimates in riffles. The modified Hankin and Reeves design was applied in eight different streams during two summer field seasons for a total of 13 applications. Target species were 1+ trout (yearling and older cutthroat trout, Oncorhynchus clarkί, and steelhead, O. mykiss, and 0+ (young of year) coho salmon, O. kisutch. I review problems encountered during field implementation of this modified survey design, I present quantitative results based on data collected from two summer seasons of field work, and I conclude with recommendations for minor changes that should improve upon proposed modifications. This modified protocol shows great promise for use with salmonid stocks listed under the Endangered Species Act of 1973.