Dispersal of stream salmonids from nests and stocking sites: patterns, variability, and sampling bias

To reconcile divergent views about the direction and distance that salmonids disperse from nests or release sites, we conducted a systematic search and synthesis of existing data that included 154 data points from 58 papers. After correcting for a sampling bias in the downstream direction, 56.6% of fish dispersed downstream, significantly greater than 50% but much less than expected. The best generalized linear mixed model (GLMM) explaining the percentage of fish moving downstream included a positive effect of the percentage of sampling effort downstream, a negative effect of body size, and differences among species; brown trout (Salmo trutta) tended to disperse upstream more than other species. Dispersal distance increased with body size; median dispersal distance for age 0+ and older fish was 167 m and 908 m, respectively, implying much greater mobility than expected. As predicted, median dispersal distance was greater downstream (715 m) than upstream (602 m). Our analyses indicated that many of our perceptions about salmonid dispersal are based on biased sampling – too much effort that is too close to, and downstream of, the release site.

Methods
We used keywords to search the Web of Science for all are articles with quantitative data on the dispersal of salmonid fishes from nests or stocking sites on October 15, 2022. We identified 247 or which 58 produced 154 data points for our analysis. We recorded information about the direction of fish movement (percentage of fish that moved upstream and downstream from the release site) and the distance of fish movement (median and maximum sampling distance and dispersal distance in each direction).