Muskellunge Recruitment Dynamics in Northern Wisconsin Lakes

Abstract

The stock-recruit relationship describes the ability of animal populations to replace themselves with offspring, so is one of the most important topics of study in fishery science. The stock-recruit relationship is difficult to quantify because parental and recruit abundances are both measured with error and because recruitment is influenced by stochastic variation related to environmental conditions. The muskellunge Esox masquinongy is an important recreational fish species in North America and especially in Wisconsin. Some populations of muskellunge are in decline despite a reduction in harvest by anglers, due largely to a growing catch-and-release ethic. My objectives were to: (1) determine if muskellunge stocking enhanced natural recruitment and (2) determine if biotic and abiotic factors muskellunge recruitment. To accomplish my first objective, I used Ricker stock-recruitment models to determine if stocking enhanced muskellunge abundance in Wisconsin lakes. I constructed two models using data from lakes in northern Wisconsin that were either stocked or not stocked with hatchery reared juvenile muskellunge. Analysis of covariance (ANCOVA) was used to test for significant differences between alpha and beta parameters of the two Ricker models. Recruitment did not differ among category 1, 2, and 3 lakes in years when stocking did not occur, so all three lake types were combined in the assessment of stocking success. Density dependence in recruitment rates (beta parameters) did not differ significantly between stocked and non-stocked lakes, whereas the recruitment rate (alpha parameters) was significantly higher in stocked lakes than in non-stocked lakes. I conclude that muskellunge stocking significantly increased recruitment in Wisconsin lakes and that stocking should continue to be used to supplement muskellunge populations with insufficient natural recruitment. To address my second objective, I incorporated environmental variables into stock-recruit models to explain residual variation in muskellunge recruitment in Wisconsin lakes. Candidate models were compared using Akaike’s Information Criterion (AIC) to select the most parsimonious model. I found that several abiotic and biotic variables influenced muskellunge recruitment. For non-stocked populations, adult stock density, and the ratio of watershed area to lake area explained 47% of the variability in muskellunge recruitment. For stocked populations, the top model included adult stock density, density of stocked muskellunge and average spring temperature and explained 53% percent of the variation in muskellunge recruitment. The model representing environmental variation also showed some support and included abundance of adult muskellunge, density of stocked muskellunge, average spring temperature, the coefficient of variation in spring temperature, and the average spring wind speed. This model explained 34% of the variation in muskellunge recruitment. My results indicate that muskellunge recruitment in northern Wisconsin lakes is regulated by the watershed area to lake area ratio, average spring air temperature, spring air temperature variation, and spring wind speed.