Using Local and Catchment-Level Variables to Explain Differences in Bia-Predicted Body Composition of Brook Trout in Headwater Streams of West Virginia, USA

Body condition is a driver of both individual and population characteristics such as behavior, survival, and reproduction.  We used bioelectrical impedance analysis (BIA) along with mark-recapture procedures on populations of Brook Trout to assess factors influencing body composition both spatially and temporally in headwater streams of West Virginia, USA.  We hypothesized proximate body composition (measured as percent dry mass) would vary across watershed characteristics such as elevation, geology, and drainage area and local stream characteristics such as proportion of pools, spawning area, and woody debris density.  We found no statistical evidence of watershed level characteristics influencing percent dry mass with the exception of a weak direct relationship with elevation (p=0.07).  However, a regression model including elevation along with local stream variables for proportion of pool habitat, maximum pool depth, and spawning area showed a significant (p = 0.01, adj. R² = 0.75) relationship.  Across time, we found recaptured fish increasing in both length and weight, but changes in body composition were not always correlated with changes in size.  Finally, body composition of adult trout in fall 2013 was not statistically correlated with juvenile abundance in fall 2014, suggesting evidence for environmental drivers of recruitment and spawning success in these streams.