Multiscale Prediction of Whirling Disease Risk in the Blackfoot River Basin, Montana: a Useful Consideration for Restoration Prioritization?

Habitat restoration for inland trout (family Salmonidae) is common across western North America, but planners rarely consider disease risk when prioritizing restoration sites. Whirling disease is a parasitic infection caused by the invasive myxosporean parasite Myxobolus cerebralis and has been implicated in declines of wild trout populations across western North America. For planners to consider disease, disease risk needs to be predictable across the landscape and influence restoration outcomes. We collated the history of whirling disease infection severity scores on the MacConnell-Baldwin scale from sentinel cage studies for hatchery Rainbow Trout Oncorhynchus mykiss in the Blackfoot River basin from 1998 to 2009. At these same sites, we performed reach-scale geomorphic assessments, derived landscape variables from GIS data layers, and assembled fish composition data. We examined relationships between the severity of infection and several landscape-scale and reach-scale variables for 13 basin-fed streams in the Blackfoot River basin of west-central Montana using classification and regression tree analyses. In our data set, valley slope and forest cover were the best predictors of fine sediment. Both spring creeks and gently sloping alluvial basin-fed tributaries to the Blackfoot River basin with higher proportions of fine sediment (particle size < 0.85 mm) were associated with a high severity (>= grade 3) of infection. Additionally, we explored differences in trout species composition (i.e., susceptible versus resistant species) before and after the whirling disease enzootic using seven basin-fed streams and two spring creeks. We did not detect trout community shifts from susceptible to disease-resistant salmonids in basin-fed disease-positive streams. However, spring creeks showed a negative trend in disease-susceptible salmonids after the whirling disease enzootic. Disease risk appears to be predictable across the landscape and may limit possible restoration outcomes by influencing species composition in spring creeks.