Restoration of Native Biodiversity in Altered Environments: Reintroduction of Atlantic salmon into Lake Ontario

Less than a quarter of reintroduction programs have succeeded in re-establishing a self-sustaining population of an extirpated species. Optimal source population selection, based on an evolutionary perspective, could increase the fitness of translocated individuals, thereby improving the success rate of restoring extirpated populations. Here, using three source populations of Atlantic salmon, Salmo salar (LaHave River, Sebago Lake, and Lac Saint-Jean), that are being used for reintroduction efforts into Lake Ontario, I examined two optimal source population selection approaches: environment matching and adaptive potential. For environment matching, source populations from locations containing similar key environment features as the reintroduction location should contain adaptations to these features. For adaptive potential, source populations with high heritable genetic variation should have the potential to adapt to new selection pressures, such as the key environment features in the reintroduction location. I tested environment matching using experimental settings by exposing the three source populations to two key environment features that are likely impediments to a successful reintroduction of Atlantic salmon into Lake Ontario: the presence of non-native salmonids and a high thiaminase diet that can lead to a thiamine (vitamin B1) deficiency. I also quantified the amount of within-population heritable (additive) genetic variation for early-life history traits to assess the adaptive potential of the source populations. Although the average amount of heritable genetic variation was the highest for early-life history traits of the Sebago population, the amount was low, suggesting that the traits have a limited potential to adapt to any new selection pressures in Lake Ontario. Overall, the Sebago population (a match to both key environment features) had the highest performance, followed by the Saint-Jean population (match to a high thiaminase diet but not non-native salmonids), and finally the LaHave population (not a match to either feature). The pattern of overall performance and the low amount of heritable genetic variation of the three source populations generally supports environment matching over adaptive potential; however, further population comparisons are required over the entire life-cycle and in a fully natural setting to make more robust recommendations for large scale reintroduction efforts of Atlantic salmon into Lake Ontario