Spatial Community Ecology, Cannibalism, And The Resilience Of Lake Champlain Forage Fish

Abstract

Nonnative species and changes in the abiotic environment are current drivers of ecological change around the world. Understanding and forecasting the response of fish populations and communities to these types of changes is the focus of the following research conducted in Lake Champlain. Three general themes run throughout this work: adaptation to change, fish early life history, and spatial ecology. In the first section of this work, we compared hatch timing, abundance, growth rates, and subsequent survival of age 0 rainbow smelt Osmerus mordax and alewife Alosa pseudoharengus. Rainbow smelt hatched almost a month earlier than alewife. Later hatching individuals of both species grew faster than those hatching earlier. Because cannibalism by rainbow smelt is likely the mortality source for age-0 Rainbow smelt, early hatching may not be advantageous as the overlap between adult and age-0 rainbow smelt is highest early in the season. However, alewife, first found in Lake Champlain in 2003, may increase age 0 rainbow smelt summer mortality. Increased summer mortality in rainbow smelt should, in turn, favor selection for early hatching. Using stable isotopes of C and N, we found that alewife are now a large component of predator diets, and in particular of the diets of Atlantic salmon and walleye. Walleye and rainbow smelt [delta]15N values changed significantly from prealewife [delta]15N values of the late 1990's. These changes represent a preferential switch to feeding on alewife by Atlantic salmon which was likely caused by the distribution patterns of these species. Distribution patterns of adults and age 0 fish were simulated on a daily basis, and subsequent age 0 mortality rates calculated. Rainbow smelt age 0 mortality rates are highest when rainbow smelt adults are abundant and alewife age 0 mortality rates are highest when alewife adults are abundant, allowing the two species to co-exist. Mortality rates were higher under normal temperature regimes, but late-summer mortality rates were higher in our climate change scenario because of increased overlap of adult and age 0 distributions. Our simulation suggests spatial distribution patterns should be accounted for when forecasting the interacting effects of climate change aquatic nonnative species.