Linking Landscapes and Habitat Suitability Scores for Diadromous Fish Restoration in the Susquehanna River Basin

Abstract

AbstractDams within the Susquehanna River drainage, Pennsylvania, are potential barriers to migration of diadromous fishes, and many are under consideration for removal to facilitate fish passage. To provide useful input for prioritizing dam removal, we examined relations between landscape‐scale factors and habitat suitability indices (HSIs) for native diadromous species of the Susquehanna River. We used two different methods (U.S. Fish and Wildlife Service method: Stier and Crance (1985), Ross et al. (1993a, 1993b, 1997), and Pardue (1983); Pennsylvania State University method: Carline et al. (1994)) to calculate HSIs for several life stages of American shad Alosa sapidissima, alewives Alosa pseudoharengus, and blueback herring Alosa aestivalis and a single HSI for American eels Anguilla rostrata based on habitat variables measured at transects spaced every 5 km on six major Susquehanna River tributaries. Using geographical information systems, we calculated land use and geologic variables upstream from each transect and associated those data with HSIs calculated at each transect. We then performed canonical correlation analysis to determine how HSIs were linked to geologic and land use factors. Canonical correlation analysis identified the proportion of watershed underlain by carbonate rock as a positive correlate of HSIs for all species and life stages except American eels and juvenile blueback herring. We hypothesize that potential mechanisms linking carbonate rock to habitat suitability include increased productivity and buffering capacity. No other consistent patterns of positive or negative correlation between landscape‐scale factors and HSIs were evident. This analysis will be useful for prioritizing removal of dams in the Susquehanna River drainage, because it provides a broad perspective on relationships between habitat suitability for diadromous fishes and easily measured landscape factors. This approach can be applied elsewhere to elucidate relationships between fine‐ and coarse‐scale variables and suitability of habitat for fishes.