Fisheries biologists are concerned that recent declines in wild trout populations in the mountainous region of western North Carolina are due to aquatic habitat degradation resulting from land cover changes.  We describe a process for developing a remote sensing-based land cover classification system that is tailored to include land cover types relatable to trout occurrence.  The classification system is used to analyze land cover in contributing drainage areas for greater than 3,000 wild trout sampling points over three time steps.  By associating landscape composition characteristics to the presence or absence of trout at each sampling site, a robust statistical analysis can be performed that indicates a range of landscape signatures that support trout for each time step. 

Results/Conclusions

The statistical analysis is used to identify thresholds for certain land use types or combinations of land use types that result in impairment to trout habitat and trout populations.  Identification of these thresholds allows for a GIS inquiry to locate watersheds in jeopardy of losing their capability to support trout and those that are candidates for restoration, thus enabling fishery managers to implement effective wild trout preservation and restoration strategies.