There are three territorial salmonids in our study site—brook trout, brown trout, and Atlantic salmon. They generally remain in a small area, and yet they are capable of long-distance movement. Such movements are known to be triggered by smolting (in the case of salmon) and by spawning behavior which involves search for new habitat in all three species. Additional triggers for long-distance migration such as infectious disease and failure to establish a territory due to competition are suggested in the literature. The importance of these secondary processes could be key to pathogen population dynamics, long-range gene-flow, and the reseeding of extinct populations. We are interested in accounting for the known patterns and inferring what other processes can trigger long-distance movement behavior. Our data set is a record of captures and recaptures of marked individuals using electrofishing for quarterly physical recaptures. We also have ongoing records from PIT-tag antennas. We use generalized linear mixed models to describe the probability, direction, and distance of movement based on habitat quality, discharge, stream temperature and biomass density data. 
Results/Conclusions

Our models reproduce the seasonal, species-specific, movement patterns which follow known smolting and spawning behavior. In addition, high discharge levels present in spring also relate to downstream movement in the youngest year class, but they do not appear to affect larger individuals. After the species-specific movement patterns are controlled for, the strongest effect is the previous movement state of the individual—settled or seeking a territory. We are currently exploring possible explanations for the ongoing search behavior—specifically habitat quality and biomass density, as well as factors which may generate long-distance migration out of the study site.