Estimating wounding of lake trout by sea lamprey in the upper Great Lakes: Allowing for changing size-specific patterns

Abstract

A primary fishery concern in the Laurentian Great Lakes is mitigating the persistent negative impact of parasitic sea lamprey (Petromyzon marinus) on native lake trout (Salvelinus namaycush). Wounds observed on surviving lake trout are commonly used by managers to assess damages associated with sea lamprey predation. We estimated the relationship between wounding rates and lake trout size, and how this varied spatially and temporally. We built upon previously published work by fitting wound rates as a logistic function of lake trout size. By using longer time series and data from three Great Lakes, our analysis harnessed substantially more contrast in host populations than previous work, and we also employed software advances for nonlinear mixed-effect models. Candidate models allowed logistic function parameters to be constant or to vary spatially, temporally, or both. Temporal effects were modeled as random walk processes. We also considered models that assumed either Poisson or negative binomial distributions for the number of wounds per fish at a given length. Models that allowed for both spatial and temporal effects in the shaping parameters and assumed a negative binomial wound distribution resulted in the best fit as indicated by Akaike's Information Criterion. Wounding rate estimates from models selected for each lake in this analysis are contrasted with those of wounding rate models currently used as components of lake trout population assessments. Although model fit was improved substantially, differences in wounding rates estimates are modest and estimates follow very similar temporal trends. However, as time series of wound data continue to grow, models that incorporate temporal variability in parameter estimation are expected to be increasingly favored. This research provides managers with an updated tool to obtain more reliable estimates of sea lamprey wounding.