Understanding Changing Patterns of Size-Specific Marking of Lake Trout by Sea Lamprey

In the Laurentian Great Lakes, parasitic sea lamprey (Petromyzon marinus) attack large-bodied hosts, such as lake trout (Salvelinus namaycush), and cause substantial mortality.  These attacks leave wounds, or marks, on surviving lake trout hosts.  Estimates of the resulting marking rates are used to evaluate the success of a multi-million dollar sea lamprey control program, and as the basis for estimates of sea-lamprey induced mortality in lake trout assessment models.  We estimated the relationship between marking rates and lake trout size, and how this varied spatially and temporally.  We were especially interested in evaluating the behavioral response of sea lamprey host selection to spatial and temporal variation in lake trout population size structures and densities.  We built upon previously published work by fitting marking 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.  We used AIC to compare several candidate models for how the marking rate versus lake trout size function changed over time and space.  For example, foraging theory predicts that a predator will broaden its diet to include less desirable prey in the scarceness of higher quality prey.  We hypothesized that sea lamprey shift their feeding patterns toward larger hosts when available.  To investigate such a behavioral response, we explored alternative assumptions for changes in the inflection point of the logistic function (the size of lake trout at which the marking rates increase most rapidly).  Candidate models allowed logistic function parameters to be constant spatially, temporally, or both, and to vary randomly or in response to the catch per effort of lake trout.   We will discuss the results in terms of (1) the plausibility of the models we developed versus the models currently used in management and (2) the extent to which the responses agree with expectations from foraging theory.