Estimating Natural Selection Acting on Stream‐Dwelling Atlantic Salmon: Implications for the Restoration of Extirpated Populations

Abstract

Abstract:Efforts to restore populations to locations from which they have been extirpated may be hampered by maladaptation in the introduced group because they came from a different environment. Estimates of natural selection acting on the new population can be used to deduce maladaptation and tailor efforts to reduce its effects. We estimated natural selection acting on Atlantic salmon ( Salmo salar ) undergoing restoration to the Connecticut River ( U.S.A. ). More than 3500 mark‐recapture records in a small tributary ( West Brook, Massachusetts ) were used to estimate selection acting on body length, body mass, condition factor, and growth. Estimates were obtained for three to four different cohorts, two age classes ( second and third years of life ), and two seasons ( winter and summer ). We found little evidence that any of the traits were subject to strong directional selection ( favoring larger or smaller trait values ). Interestingly, fish that were larger or had higher condition factors rarely survived at higher rates, a result conflicting with the conventional “bigger is better” expectation. We also found little evidence that any of the traits were subject to strong or consistent stabilizing selection ( favoring trait values near the mean ). Our results suggest that the specific traits we examined are not limiting adaptation and are probably not preventing the population from becoming self‐sustaining. Future efforts should concentrate on other potentially limiting traits, such as the timing of smolt migration. Our results also suggest that any additional introductions of exogenous fish need not be concentrated on the size or growth of juveniles in potential source populations.