Simulation of the Effect of Rainbow Trout Introduction in Lake Washington

Abstract

We developed a simulation model based on energetics, habitat selection, feeding selectivity, and population dynamics to examine the effect of introductions of rainbow trout Oncorhynchus mykiss on parr of sockeye salmon O. nerka and longfin smelt Spirinchus thaleichthys resident in Lake Washington (Washington State). We modeled growth and population dynamics of rainbow trout cohorts introduced between 1981 and 1984, and compared results with length-at-age data obtained from marked released fish and with data on diets. Graphical comparisons indicated a reasonable fit to the growth data over the 150–400-mm size range. The model gave a close prediction of the size at which rainbow trout begin to eat fish, these results being sensitive to a size-selective feeding function. Diet data showed that model predictions for the major prey fish (longfin smelt and sockeye salmon) were reasonable. However, prey vulnerability had to be set lower for sockeye salmon than for longfin smelt to produce reasonable rainbow trout diets. Simulation experiments examined the effect ofaltered rainbow trout enhancement, fishing pressure, prey vulnerability, and prey abundance and species composition on prey fish survival. These experiments indicated that the effect of rainbow trout predation on prey fish strongly depended both on the magnitude of rainbow trout introductions and on the prey fish abundance and species composition, which varied from year to year.