U.S. Department of Commerce


Date of this Version



Conservation Biology Volume 4, No. 2, June 1990


Recent reductions in the abundance of all Pacific salmon species (Oncorhynchus spp.), coupled with large in- creases in artificial production, demand that careful attention be paid to genetic changes occurring in both wild and cultured populations. Analysis of electrophoretic data for chinook salmon (O. tshawytscha) from the Pacific coast of Oregon revealed substantial allele frequency changes over 2-4 years in hatchery, but not wild, populations. Unfortunately, our understanding of the causes of this result is hampered by a lack of theoretical models designed for organisms with life history features like those of Pacific salmon. We used computer simulations to provide a context for under- standing genetic changes observed in the hatchery populations.

Simulation results indicated that annual fluctuations in population allele frequencies due to genetic drift can typically be expected to be several percent, with the absolute magnitude determined primarily by the effective number of spawners each year rather than the age structure. Changes over 10- to 25-year periods were only slightly greater than short-term changes (1-5 years). The magnitude of allele frequency change over time was different for juvenile and adult samples. The probability of a significant test statistic com- paring allele frequencies in temporally spaced samples in- creased with the ratio of sample size to effective number of breeders per year. This is an important consideration for conservation biologists, who typically are concerned with populations of small effective size.

Simulation results indicate that it is necessary to postulate unrealistically large selection coefficients to explain the genetic changes in the hatchery populations by natural selection. The changes observed are consistent with a pure drift model provided that the effective number of breeders was as small as about 25-50 per year. Analysis of brood stock in- formation for the hatcheries indicates that the effective population number may indeed have been this low, although the number of returning adults was often much larger. This conclusion underlines the importance of monitoring the genetic consequences of the large-scale artificial propagation pro- grams involving Pacific salmon.