Skip to main content

Characterization of Spatial and Temporal Variability in Fishes in Response to Climate Change


Brian Irwin

Tyler Wagner

James Bence

Publication Type:
Year of Publication:
Northeast Climate Science Center


The number of fish collected in routine monitoring surveys often varies from year to year, from lake to lake, and from location to location within a lake. Although some variability in fish catches is expected across factors such as location and season, we know less about how large-scale disturbances like climate change will influence population variability. The Laurentian Great Lakes in North America are the largest group of freshwater lakes in the world, and they have experienced major changes due to fluctuations in pollution and nutrient loadings, exploitation of natural resources, introductions of non-native species, and shifting climatic patterns. In this project, we analyzed established long-term data about important fish populations from across the Great Lakes basin, including from Oneida Lake in NY, Lake Michigan, and the Bay of Quinte in Lake Ontario. Our objective was to evaluate spatial and temporal variation in fish catches from large freshwater lakes that have experienced large-scale changing conditions. We evaluated analytical approaches with the potential to disentangle sources of variability in standardized monitoring data. Specifically, we considered 1) how the decomposition of spatial and temporal variation in fish catches can be used to measure a response to perturbation; 2) how truncation of population age structure can alter population oscillations which may shift how a population is affected by environmental fluctuations; and 3) how the composition of a fish community may respond to a suite of environmental drivers through time. Using long-term gill-net data for walleye, we found that average catch and variance structure differed before and after large-scale perturbations. More generally, our results suggest that fish population responses to changing environments can be complex, but that long-term monitoring combined with modeling approaches can allow for detection of quantifiable changes.