Approaches to Modeling Temporal Changes in Age Specific Natural Mortality, Catchability, and Selectivity in an Age-Structured Assessment of Lake Trout in the Main Basin of Lake Huron

We explored approaches to allowing time-varying catchability, selectivity, and natural mortality in statistical catch at age (SCAA) assessment models for lake trout Salvelinus namaycush in the main basin of Lake Huron.  Lake trout was the top native predator in the North America Great Lakes, but was nearly extirpated by the early 1950s due to excessive mortality from sea lamprey Petromyzon marinus predation and fishery harvest.  Sea lamprey control, hatchery stocking, and fishery regulation have been the three primary strategies for lake trout rehabilitation in the Great Lakes.  Despite these efforts, nearly all recruitment to the population over the period modeled (1984-2010) was from releases of hatchery-reared fish.  During this period Lake Huron experienced substantial changes in the food web and the fisheries system.  We conducted SCAA assessments in three regions of the main basin of Lake Huron.  The assessment models allow for time-varying catchability for both of the two fisheries (commercial and recreational), and a fishery independent survey.  The models also allowed selectivity to be time-varying.  We will report on our evaluation of these approaches and alternatives.  Natural mortality in these assessment models consists of mortality due to sea lamprey predation and mortality from other sources.  The rate of mortality due to sea lamprey predation is calculated based on the number of wounds observed on surviving fish and is both age and year dependent.  Here we explore approaches to modeling the other natural mortality within SCAA models as age-specific, based on informative priors.  Age-1 natural mortality is always allowed to vary over time, and this is estimable because of the known number of age-1 recruits from stocking hatchery-reared fish.  We will report our comparison of three options in modeling natural mortality from the sources other than sea lamprey: single constant natural mortality, age-specific natural mortality, and time-varying age-specific natural mortality in response to changes in size-at-age and growth parameters.  Model comparisons are based on patterns of fit to the data, retrospective patterns, and information criteria.