The purpose of this research was to investigate two of the assumptions made in the assessment of Lake Erie yellow perch (Perca flavescens). In any model, assumptions have to be made in both model structure and data utilization. Testing these assumptions is important to ensure results are not being biased. In Chapter 1, I tested the assumption that each management unit (MU) in Lake Erie has a distinct yellow perch population with no mixing. I investigated the effect this assumption has on assessment and whether it could bias the results if it was being violated. I developed a statistical catch-at-age model that allowed movement between two of the MUs and evaluated how abundance estimates changed for 24 different movement scenarios. The abundance estimates differed between scenarios in unexpected and inconsistent ways, suggesting these models are sensitive to assumptions about movement. In Chapter 2, I investigated the assumption that one of the fishery independent surveys used in the stock assessment model is an unbiased indicator of trends in abundance. If these surveys are being affected by factors besides trends in abundance, the resulting index could be biased if these factors are not taken into account. I used a catch-rate standardization and model selection approach to investigate the effect of temporal, spatial, and environmental factors. Wind was incorporated using a novel approach that combined both wind direction and speed into a single parameter. The patterns seen in the standardized index of abundance from the best-fit model that incorporated other factors versus the non-standardized index were similar for both MUs; however, using a standardized index has the potential to accommodate for future changes in the environment of Lake Erie.
Read
