Factors affecting lake whitefish recruitment in the Laurentian Great Lakes
Abstract
Recent ecological changes in the upper Laurentian Great Lakes have led to reduced growth and condition of several lake whitefish stocks. These changes have raised concerns over age estimate reliability and the role of intrinsic biotic factors in reproduction and subsequent recruitment. To address these issues, I: (1) examined variation and efficacy of three age estimation methods for the Bailey's Harbor, Naubinway, and Saugatuck stocks; (2) investigated patterns of female condition and egg quality in relation to reproductive life-history strategies for eight stocks; (3) tested the prediction that spawner condition drives recruitment via energy allocation dynamics that affect juvenile physiological condition (i.e., parental effects hypothesis); and (4) conducted a theoretical assessment of recruitment dynamics generated by eight life history based recruitment mechanisms. I found that scales provided less precise and systematically lower age estimates compared to both fin rays and sagittal otoliths. A decision-matrix analysis showed that fin rays were more precise and economical, and were therefore the best aging method. With regards to the second objective, my results supported the reproductive trade-off hypothesis, whereby females maintain gamete quality through trade-offs among somatic growth, composition, and fecundity. Overall, only 8% of the female predictor variables in multiple regression models contributed significantly to explaining variation in egg properties. With regards to the third objective, a linear model which included sampling site, female condition, and egg quality explained 39% of the variation in juvenile physiological condition; however, sampling site alone explained 23% of the variation. Neither females (1.4%) nor eggs (2.7%) explained much variation in juvenile physiological condition. To address the fourth objective, eight life-history based recruitment models were fit to observed larval and age-0 juvenile pre-recruit data. A zooplankton availability model best fit the observed data. The Ricker Model failed to predict the kind of recruitment dynamics that are typically observed for lake whitefish in the Great Lakes. By contrast, the other seven models examined generated recruitment curves that better represented the observed data and predicted dynamic behavior that was more consistent with that observed for lake whitefish in the Great Lakes.
Degree
Ph.D.
Advisors
Sutton, Purdue University.
Subject Area
Ecology|Aquatic sciences
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.