Selection among critically endangered landlocked salmon (Salmo salar m. sebago) families in survival and growth traits across early life stages and in different environments.

Endangered wild fish populations are commonly supported by hatchery propagation. However, hatchery-reared fish experience very different selective pressures compared to their wild counterparts, potentially causing genotype-by-environment interactions (G × E) in essential fitness traits. We experimentally studied early selection in a critically endangered landlocked Atlantic salmon population, first from fertilization to the swim-up stage in a common hatchery setting, and thereafter until the age of 5 months in two contrasting rearing environments. Swim-up progeny were moved either to standard indoor hatchery tanks involving conventional husbandry or to seminatural outdoor channels providing only natural food. After the first summer, sampled survivors were assigned to their families by genotyping. Early survival until the swim-up stage was mostly determined by maternal effects, but also involved significant variation due to sires and full-sib families (potential genetic effects). High on-growing survival in hatchery tanks (88.7%) maintained a more even distribution among families (relative share 1.5%-4.2%) than the seminatural environment (0.0%-5.4%). This heterogeneity was mostly maternal, whereas no independent paternal effect occurred. Heritability estimates were high for body size traits in both environments (0.62-0.69). Genetic correlations between the environments were significantly positive for body size traits (0.67-0.69), and high body condition in hatchery was also genetically linked to rapid growth in the seminatural environment (0.54). Additive and phenotypic growth variation increased in the seminatural environment, but scaling effects probably played a less significant role for G × E, compared to re-ranking of genotypes. Our results suggest that not only maternal effects, but also genetic effects, direct selection according to the environmental conditions experienced. Consistently high genetic variation in growth implies that, despite its low overall genetic diversity and long history in captive rearing (>50 years), this landlocked Atlantic salmon population still possesses adaptive potential for response to change from hatchery rearing back to more natural conditions.

Spatial variability of mercury and polyunsaturated fatty acids in the European perch (Perca fluviatilis) - Implications for risk-benefit analyses of fish consumption.

This study evaluated the spatial variability of risks and benefits of consuming fish from humic and clear lakes. Mercury in fish is a potential risk for human health, but risk assessment may be confounded by selenium, which has been suggested to counterbalance mercury toxicity. In addition to the risks, fish are also rich in eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), which are known to be beneficial for cardiovascular health and brain cognitive function in humans. We found that the concentrations of EPA + DHA and mercury in European perch (Perca fluviatilis) vary spatially and are connected with lake water chemistry and catchment characteristics. The highest mercury concentrations and the lowest EPA + DHA concentrations were found in perch from humic lakes with high proportion of peatland (30-50%) in the catchment. In addition, the ratio of selenium to mercury in perch muscle was ≥1 suggesting that selenium may counterbalance mercury toxicity. The observed variation in mercury and EPA + DHA content in perch from different lakes indicate that the risks and benefits of fish consumption vary spatially, and are connected with lake water chemistry and catchment characteristics. In general, consumption of perch from humic lakes exposed humans to greater risks (higher concentrations of mercury), but provided less benefits (lower concentrations of EPA + DHA) than consumption of perch from clear lakes.