18-4 Natural selection for better or worse: Selection analyses as a tool for endangered fish restoration

Tuesday, September 14, 2010: 9:00 AM
403 (Convention Center)
Michael M. Bailey, PhD , Central New England Fishery Resource Office, U.S. Fish and Wildlife Service, Nashua, NH
Michael T. Kinnison, PhD , School of Biology and Ecology, University of Maine, Orono, ME
From an evolutionary standpoint, the significance of connectivity among populations lies in its consequences for adaptive trait variation and mean population fitness.  Such consequences are a product of both the rate and pattern of population connectivity AND the strength and pattern of natural selection across the landscape.  Traditionally, conservation and restoration programs have focused heavily on estimating the former, but have rarely sought to estimate spatiotemporal variation in selection itself.  We suggest that estimation of selection can be a useful tool in conservation in recovery efforts, and provide examples of how natural selection can be studied within the confines of a species recovery program.  Our study system is for endangered Atlantic salmon populations in Maine, USA.  By using natural and induced marks on otoliths, we were able to quantify spatial and temporal patterns of selection on hatchery fry released into the wild.  We found that selection on fry size varies across the protected range of the species and selection is associated with fluvial habitat features.  We also found support for selection on ontogeny that is likely mediated by hatchery practices.  We discuss these findings with respect to the challenges of managing natural and anthropogenic connectivity in an active restoration context.