Researchers seek to identify habitat features important to juvenile salmon so that habitat restoration efforts can be directed where they yield the most benefit. We often look for juvenile habitat preferences by comparing the spatial distributions of habitat features and juveniles. However, the initial spatial distribution of juveniles is not determined by habitat preferences, but rather by adult spawning. As time passes, the juvenile spatial distribution is increasingly independent of the spawning distribution and dependent on habitat preferences. We would like to know, when can we interpret the juvenile spatial distribution as indepedent of the adult spawning distribution? To address this, I have simulated juvenile movement away from the spawning sites using a stochastic differential equation model. In the simulation, movement is partly non-directed and partly directed. Non-directed movement is random and normally distributed with increasing variance over time to reflect the increasing swimming speed of growing juveniles. Directed movements are driven by the need to seek out cool waters during hot weather.

Results/Conclusions

I calculate the number of days post-emergence when the juvenile spatial distribution is disassociated from the adult spawning distribution. I show how this time depends on the patchiness of the original spawning distribution, the spatial scale of the analysis, the variance of random juvenile movements, and the propensity to drift in response to changing stream temperatures. I show how these simulation results influence my analysis of spatial data for juveniles and habitat features in the Columbia River Basin.