SPATIOTEMPORAL PATTERNS OF GENETIC STOCK MIXING AND THEIR APPLICATION IN MIXTURE-INFORMED CATCH-AT-AGE MODELS FOR LAKE WHITEFISH IN LAKE MICHIGAN

Abstract

Lake Michigan lake whitefish (Coregonus clupeaformis) support the lake’s largest commercial fishery and is comprised of multiple stocks managed across a series of geopolitical management zones (MZs). Limited movement among MZs is an underlying assumption of current lake whitefish management and lake whitefish stock assessment models, yet previous research suggests stock mixing occurs to varying degrees across MZs. This can create challenges when estimating stock-specific harvest and dynamic rates in stock assessment models. In Lake Michigan, statistical catch-at-age models (SCAA) are used to estimate recruitment, abundance, and mortality for lake whitefish MZs to inform harvest quotas and management. Understanding stock compositions of the mixed-stock lake whitefish fishery can aid in allocating fishing effort on spatial and temporal scales to avoid overharvest of vulnerable stocks evaluated by the SCAA models. My research objectives were to: 1) determine spatiotemporal patterns of lake whitefish stock mixing throughout Lake Michigan, 2) determine appropriate sample sizes required to accurately conduct mixed stock analyses, and to 3) evaluate model performance of a newly developed mixture-informed SCAA model that incorporates genetic stock identification mixture compositions for the purpose of understating changes in lake whitefish recruitment. I applied a newly developed genotyping-in-thousands (GT-seq) panel containing 472 single nucleotide polymorphisms to conduct genetic stock identification (GSI) of Lake Michigan lake whitefish. Using GSI, I quantified contributions from each of five genetically distinct reporting units for 39 mixed stock lake whitefish samples collected throughout Lake Michigan from 1977-2022 and evaluated the sensitivity of SCAA model recruitment estimates adjusted to the determined mixture compositions. GSI results demonstrated that the lake whitefish fishery in Lake Michigan is a mixed stock fishery throughout much of the year and that stock compositions vary across seasons and, to a lesser extent, years. In general, the mixture-informed SCAA model was able to converge upon an estimated set of parameters and demonstrated good fit to observed gill and trap net harvest in most MZ models. This thesis provides information on the presence of genetic stocks in different MZs throughout the non-spawning period, further delineation on the genetic structure of lake whitefish in Lake Michigan, as well as initial developments of a mixture-informed SCAA model which inform recruitment estimates valuable to the management of lake whitefish.