Effects of captivity on endangered populations of Inner Bay of Fundy Atlantic salmon (Salmo salar)

Captive breeding and rearing is increasingly used in efforts to conserve and restore salmonid populations around the world. However, due to the very different environmental conditions, and thus selective pressures, experienced by populations maintained in captivity and those in the wild, genetic change can occur. Assessing to what extent, and how rapidly this genetic change can occur is vital in determining the impact of these conservation programs on captive populations and in predicting captive individuals' potential fitness when released into the wild. This was undertaken in a Live Gene Bank (LGB) captive breeding and rearing program for endangered Inner Bay of Fundy Atlantic salmon (Salmo salar).;Antipredator response, the behavioural reaction to an avian predator, has been shown to decrease after one or more generations in the hatchery environment. Behavioural trials determining antipredator response were conducted on several groups of Atlantic salmon fry. The groups tested included: first- and second-generation fry from the LGB, as well as fully wild and domesticated fry. The second-generation captive fish displayed significantly more risk-taking behaviour before and after exposure to a model predator than first-generation captive fish. These two groups shared rearing conditions and had similar genetic origins, suggesting that rapid evolution possibly due to domestication may be occurring in the LGB. The wild fry showed the greatest antipredator response, while the domesticated showed the least, demonstrating that the methods employed provide meaningful results.;A second study was undertaken that addresses some of the effects of captivity as well as aspects of kin-biased migration. An entire naturally occurring smolt-run on the Big Salmon River, NB, was sampled in 2003, providing a unique opportunity to reconstruct kinship in the context of both run timing and origin (either captive or wild). Significant variation in mn day, length, weight, and condition factor among kin groups was found in both the wild- and the captive-origin smolts. Because the potentially confounding effect of shared environmental variation within families was not present in the LGB-origin group, this among-family variation suggests that these traits contain heritable components. Differences between first and second generation captive smolts in mn day, length and weight, indicate that selection on these traits may be occurring in the Live Gene Banking program. A high ratio of effective to census population size was detected in the captive-origin smolts compared to the wild-origin smolts, suggesting a higher relative effective number of breeders contributing to the captive-origin smolts. This was likely due to effective reduction in variance of reproductive success of captive individuals. Finally, an estimated low incidence of spawning anadromous males was implied from the reconstructed wild pedigree, which was consistent with the hypothesis of a mating structure with numerous mature male pan each mating once with an anadromous female.;The results of the two studies provide useful information for salmonid captive breeding programs, and demonstrate the utility of using molecular markers to address basic and applied biological questions.