Long-term trends of bloater (Coregonus hoyi) recruitment in Lake Michigan: evidence for the effect of sex ratio

Long-term population trends are generally explained by factors extrinsic (e.g., climate, predation) rather than intrinsic (e.g., genetics, maternal effects) to the population. We sought to understand the long-term population dynamics of an important native Lake Michigan prey fish, the bloater Coregonus hoyi. Over a 38-year time series, three 10- to 15-year phases occurred (poor, excellent, and then poor recruitment) without high interannual variability within a particular phase. We used dynamic linear models to determine whether extrinsic (winter and spring temperature, alewife predator densities) or intrinsic factors (population egg production, adult condition, adult sex ratio) explained variation in recruitment. Models that included population egg production, sex ratio, winter and spring temperature, and adult bloater condition explained the most variation. Of these variables, sex ratio, which ranged from 47% to 97% female across the time series, consistently had the greatest effect: recruitment declined with female predominance. Including biomass of adult alewife predators in the models did not explain additional variation. Overall our results indicated that bloater recruitment is linked to its sex ratio, but understanding the underlying mechanisms will require additional efforts.

Modeling time-varying growth using a generalized von Bertalanffy model with application to bloater (Coregonus hoyi) growth dynamics in Lake Michigan

A concurrent increase in lakewide abundance and decrease in size-at-age of bloater (Coregonus hoyi) in Lake Michigan have suggested density-dependent growth regulation. We investigated these temporal patterns by fitting a dynamic von Bertalanffy model and length–weight relationship with time-varying parameters to mean length- and weight-at-ages (ages 1–7) from annual surveys (1965–1999). We modeled yearling length, asymptotic size (L[Formula: see text]), and the parameters of a power relationship between mean weight and mean length (α and β) as changing slowly over time using a random walk model. The Brody growth coefficient (k) was modeled as a linear function of L[Formula: see text] with year-specific random deviations. Our results support a positive relationship between L[Formula: see text] and k, indicating that under conditions supporting larger asymptotic lengths, individuals approach the asymptote more rapidly. We explored the relationship between year-specific growth parameters and indices of lakewide bloater abundance and found evidence of density-dependent growth. However, in the most recent years, L[Formula: see text] and yearling length have remained low in Lake Michigan despite low bloater abundances, suggesting the occurrence of a fundamental shift in the food web.