Genetic Variation Assessment of Rainbow Smelt in Green Bay and Western Lake Michigan
Cahill, Joseph R.
University of Wisconsin-Stevens Point, College of Natural Resources
MetadataShow full item record
Collections of 60 or 22 rainbow smelt, Osmerus mordax, were obtained from each of six spawning sites in Green Bay and western Lake Michigan during the spawning season, April, 1991 and April, 1992. Smelt were also collected from Crystal Lake, Michigan in June, 1990, and Lake Erie and Green Bay in summer, 1992. The smelt were assessed with isozyme electrophoresis and restriction fragment length polymorphisims (RFLP) of mitochondrial DNA (mtDNA) to determine the degree of genetic differentiation among the collection sites. Evidence of discrete stocks among locations was inconsistent and inconclusive Analysis of the products.of 4 isozyme loci from the 1990 and 1991 sites provided some evidence of genetic differentiation. Evidence consistent with the existence of different stocks among sites was: 1) contingency chi-square tests of allele frequency heterogeneity among all loci indicated that the IDH-2 locus was significantly different; 2) subsequent contingency chi-square tests of allele frequency heterogeneity at the IDH-2 locus revealed significant differences among·several collections; 3) the Little Bay de Noc sample was different at the IDH-2 locus from three other Green Bay sites, Days River, Turtle Creek and Green Island; 4) the Green Island collection was different from one other sample, Little Bay de Noc; 5) the Green Island collection had low genetic identity and high genetic distance when compared to each of the other collections; and 6) consequently, the Green Island collection was not clustered on a dendrogram with any of the other five sites. Evidence that did not indicate different stocks among sites was: 1) the mean number of alleles and average heterozygosity per locus were similar among all collections; 2) the highest F(ST) value (IDH-2 locus) indicated only 2.5% of the total genetic variation was due to differences among sites; 3) coefficients of genetic similarity and distance did not indicate existence of discrete genetic stocks among collections; and 4) unweighted pair group method (UPGM) clustering of genetic distances revealed no geographic relation among collections. Restriction fragment length polymorphisims (RFLP) analyses of mtDNA from the 1992 samples also provided inconsistent and inconclusive evidence of genetic differentiation among the collections. Some evidence for differentiation at the Green Island site was found. Green Island clonal line frequencies differed significantly from those of Turtle Creek and Little Bay de Noc; however, generally low levels of genetic variation were found both within and among collections. Unweighted pair group method clustering of mitochondrial DNA genotypes revealed two major clonal lines separated by a mean sequence divergence of only 1.405%. Nucleon diversity index values were low both among collections and for pooled data from all collections, suggesting that levels of genetic differentiation among collections were low. With the exception of Green Island, contingency chi-square tests of clonal line frequency heterogeneity also indicated no genetic difference among collections. No geographical relation among clonal line frequencies was evident among collections. Analyses of pooled data did not indicate genetic difference between Green Bay and Lake Michigan. Green Bay and Lake Michigan allele frequencies did not differ significantly, nor did Green Bay and Lake Michigan clonal line frequencies. Percentage of loci polymorphic and average heterozygosity at all loci were similar between Green Bay and Lake Michigan. Coefficients of genetic identity and distance between Green Bay and Lake Michigan indicated little difference. The mean F(ST) value indicated that ot the total variation observed at all loci, only 0.2% was due to differences between Green Bay and Lake Michigan.