During the last century, the abundance and species composition of Lake Michigan fish assemblages have been greatly influenced by natural and human disturbances, including establishment of exotic species, overfishing, predator stocking programs, and habitat destruction. Since the early 1960s, the Great Lakes Science Center has monitored the abundance of Lake Michigan fishes with an annual survey based on bottom trawl samples from locations distributed around the lake. Center scientists have used the trawl data to calculate the annual biomass of forage fishes in Lake Michigan and provided these estimates to relevant agencies. However, no retrospective analysis of long-term changes in fish abundance has been attempted to date. The objective of this study was to rigorously determine if fish assemblages have statistically changed in Lake Michigan by examining geographic and temporal differences in abundance for 19 species.

Results thus far can be found in Fabrizio et al. 2000 (Using linear models with correlated errors to analyze changes in abundance of Lake Michigan fishes: 1973-1992). The abundances of five major species (bloater, deepwater sculpin, slimy sculpin, alewife, and rainbow smelt) were analyzed based on bottom trawl collections taken at 10 discrete depths (18 to 110 m) off eight fixed ports from 1973 to 1992. Bloaters were the most abundant species of all fish species caught. Mean bloater abundance was low throughout the 1970s, increased during the 1980s, and reached high levels by 1990. Mean abundances of deepwater sculpin and rainbow smelt increased from 1973 to the mid-1980s and declined thereafter. Trends in alewife and slimy sculpin abundances depended on port. The linear model with correlated errors can be readily applied to repeated-measures data from other fixed-station fishery surveys and is appropriate for data exhibiting spatial and temporal autocorrelations.