Genetic Stock Identification of the Bloater, Coregonus hoyi, from Western Lake Michigan

Abstract

Bloater samples were collected from five western Lake Michigan locations to assess genetic stock structure during the spawning period (February, March) in 1982. During the 1983 spawning period, the same five locations were sampled as well as two additional locations. White muscle, eye, and liver tissue samples from 60 individuals for each collection were subjected to isozyme analysis. Nine enzyme systems, encoded by 17 loci were surveyed to determine levels and sources of genetic variability. Three loci, malic enzyme-A (ME-A), general esterase-B (Est-B), and general esterase-C (Est-C), expressed sufficient allelic variation in white muscle tissue to meet the 95% criterion of polymorphism (P= 17.6%). Mean observed (direct-count) heterozygosity ranged from 0.052 for the 1983 Washington Island collection to 0.072 for the 1982 Inshore Milwaukee collection. No clear cut evidence resulted from this study as to the genetic stock structure of western Lake Michigan bloaters. Tests for conformance with expected Hardy-Weinberg proportions indicate that the bloaters were not in proportions expected for a single, panmictic stock. Yet, F-statistics, tests of allele frequency heterogeneity, identity coefficients and distance indices failed to delineate spatially isolated gene pools. Chi-square contingency tables for allele frequency heterogeneity indicated nonsignificant (P>0.05) variation among the collections for the ME-A locus. Contingency table results for the Est-B and Est-C loci were significant in 1982 and not significant in 1983. Chi-square analyses for conformance of observed genotype frequencies to expected Hardy-Weinberg proportions exhibited significant differences for each polymorphic locus in all collections. These differences may have been due to the presence of more than one stock. However, migration, random genetic drift, or selection can cause significant deviations from Hardy-Weinberg proportions. Wright's (1978) F-statistics indicated that only one percent [F(ST)=0.010] of the observed genetic diversity was due to between site variability in 1982 and two percent in 1983, suggesting that if more than one stock existed at the time of sampling, then they were not spatially discrete. Nei (1975) mean genetic distance and identity coefficients indicated that the collections had less than or equal to three-tenths of a percent difference between any two collection sites within a collection year. Cluster analysis of the Cavalli-Sforza and Edwards chord distance indices produced results that were inconsistent between years. The reason for this discrepancy is not clear.