Sources of variation in age at smolting in hatchery-reared Atlantic salmon (Salmo salar)

Abstract

Length-frequency distributions of Atlantic salmon parr are bimodal during the first November post-hatch and the proportion of presumptive 1 + smolts (S1) can be estimated by integrating the area under the upper modal curve. Such proportions were found for a total of 179 full-sib families of approximately 1000 fish per family. Variance between families suggested genetic effects although an unbiased estimate of heritability was not obtainable because of the experimental design. Correlations between length measurements in July and November were significantly (P < 0.01) greater than zero and families, with a mean length advantage in the lower size mode, also tended to produce larger fish in the upper mode. There was a negative correlation (P < 0.07) between mean length in July and the proportion of presumptive 1 + smolts. Thus, while selection for increased size during July would be expected to produce a positive, correlated growth response in both modes in November, a decrease in the number of 1 + smolts would be expected. Key words: Salmon, smolting age, Salmo salar, hatchery