smelt and alewife, and collapse of the lake char likely contributed to the population growth of both of these forage fish. Expansion of rainbow smelt and alewife populations corresponded with the collapsing stocks of native lake herring, and the causal mechanisms of these changes in forage fish communities continue to be debated. After 1960, smelt populations declined, whereas alewife populations boomed, culminating in the infamous die-offs that littered Lake Michigan beaches in the late 1960s.
Control of the sea lamprey was followed by highly successful stocking of exotic coho and chinook salmon in Lake Michigan. By the 1980s, stocked salmonids formed the basis of a sport fishery valued in excess of a billion dollars per annum (Kitchell and Crowder, 1986). By 1978, careful analyses of salmonid diets and bioenergetic requirements indicated that heavy predation was likely to trigger a collapse of the alewife stock (Stewart et al., 1981). By 1983, it was evident that a severe decline in alewife abundance was under way (Kitchell and Crowder, 1986). It is ironic that ''Save the Alewife" tee-shirts could be purchased in Milwaukee less than 20 years after massive die-offs fouled water intakes and beaches.
Lake Michigan cannot be viewed as a pristine, natural system. Ecosystem dynamics are determined mainly by nonnative species and decisions made by managers. At present, the food web's keystone species are exotic fish whose population dynamics are determined by management policies and are uncoupled from typical ecological feedbacks (Figure A.3). A substantial share of the variability in lower trophic levels is determined by the predatory effects of these fish (Kitchell and Crowder, 1986). Though the species composition of the community is dramatically different from the ancestral one, the extent to which ecosystem functions and trophic structure resemble those that existed prior to disturbance remains an open question. At present, fish biomass at all trophic levels is around twice as large as it was before collapse of the native stocks (J. F. Kitchell, Center for Limnology, University of Wisconsin, personal communication, June 1990).
The management of Lake Michigan's fish stocks must be judged a success by several criteria. An extremely successful sport fishery has created and sustained public interest in the resource while controlling the nuisance alewife. However, the ecosystem is an artificial one. The exotic salmonids are susceptible to outbreaks of disease, such as the current epidemic of bacterial kidney disease, exacerbated by complete dependence on hatcheries. Restoration of native species has not occured and in many cases seems unlikely. It has proved very difficult to establish reproducing populations of lake char. Ironically,