Genetic structure of wild and reared Atlantic salmon (Salmo salar L.) populations in Iceland

Abstract

to five (10.4%) in the farmed fish (Isno). Mean polymorphism (P) was 6.5%. Seven variable enzyme loci were observed in this study: AAT-4*, GLUDH-2*, GPI-3*, IDDH-2*, IDHP-3*, MDH-3* and MEP-2*. Average observed heterozygosity (H) varied from 0.022 (s.e.=0.014) to 0.048 (s.e.=0.023) and the mean H was 0.030. Heterogeneity of allele frequencies of sample years and age groups within and between the rivers and drainages in Iceland is presented. Wright’s F,, value was 0.062 and the total gene diversity (Hr) was 0.048. Nei’s genetic distance value (D) between salmon from the 32 rivers varied from 0.000 to 0.006, with a mean value of 0.0017. Genetic distance values between the salmon from the rivers and the farmed salmon varied from 0.000 to 0.008 (mean D=0.0032). The majority of salmon rivers in Iceland contain genetically distinct populations and in larger river systems there may be more than one population. Salmon populations within the same region showed lower values of genetic distance than between populations in different regions. These findings have important management implications. Every effort should be taken to avoid genetic mixing and consequent breakdown of stock differentiation causing destruction of the local adaptations in wild stocks. Salmon enhancement should, therefore, be conducted on a river stock basis, as has been the practice in Iceland for the last 10 to 15 years. Ocean ranching stations should use stocks from the local region, the size of the operation should be carefully considered and measures should be taken to minimize straying. Cage farming of salmon now only occurs in a few locations in Iceland and in most cases these are at a distance from salmon rivers. However, use of local stocks and sterile fish should be seriously considered. 0 1997 International Council for the Exploration of the Sea