Genetic Structure Among Lake Michigan's Lake Whitefish Spawning Aggregates

Abstract

Lake whitefish (Coregonus clupeaformis) have comprised an important commercial fishery on Lake Michigan since the early 1800s. Concerns exist regarding the commercial harvest of potentially shared stocks by Michigan and Wisconsin state-licensed and Tribal commercial fishing operations. Previous studies indicated potential stock structure, however, questions still exist regarding the number, identity, distribution, and discreteness of lake whitefish stocks in Lake Michigan. Stocks represent the basic biological unit of focus for fish management efforts and, subsequently, stock delineation is an integral part of sound, science-based, fisheries management programs. The objectives of the study were to determine the usefulness of lake whitefish microsatellite and mitochondrial DNA genetic diversity to discriminate among spawning stocks of lake whitefish and to describe the genetic population structure of spawning lake whitefish aggregates in northern Lake Michigan and Green Bay in terms of genetic stock identification and degree of stock isolation. I assumed that distinct spawning aggregates represented potential stocks and that differences at molecular markers underlie population differentiation. Twelve resolved microsatellite DNA loci exhibited adequate levels of diversity for population differentiation in terms of their allelic richness and heterozygosity, met Hardy-Weinberg equilibrium expectations and, therefore, the utility of the markers was considered sufficient. Two mitochondrial DNA gene regions (D-loop and ND5) exhibited low sequence diversity (π = 0.0002 and 0.0025, respectively) and low numbers of haplotypes (7 and 5, respectively). The observed diversity did not meet a priori levels of genetic diversity deemed necessary for population differentiation and the molecule was abandoned for further use. Genetic stock identification using the twelve microsatellite loci indicated 5-7 potential genetic stocks were present. The resolved stocks corresponded to geographically proximate populations clustering into genetic groups. Analysis of molecular variance suggested six genetic stocks present in the lake as a significant portion of variation was attributable between groups of populations, but not within groups of populations. Within stock analyses (pairwise Fst) suggested all but one stock delineated by AMOVA was a stable grouping (i.e., no between population differences) with the exception being the Hog Island, Traverse Bay grouping (NOE stock). This putative stock showed significant differences between the two populations indicating gene flow between the NOR stock (Naubinway and Epoufette) and Hog Island but not between NOR and Traverse Bay. Elk Rapids was the most genetically divergent population and actually masked the isolation by distance that exists between all other populations within Lake Michigan. Genetic mixture analysis showed mixed stock analysis using the resolved stocks and molecular markers can be accomplished with high accuracy. The coupling of genetic data from this study with current demographic data and a comprehensive mixed-stock analysis will allow for more efficient and effective management of this economically and socially important resource.