In the 1950s, burrowing mayflies, Hexagenia spp. (H. limbata and H. rigida), were virtually eliminated from the western basin of Lake Erie (a 3300 km² area) because of eutrophication and pollution. We develop and present a deterministic model for the recolonization of the western basin by Hexagenia to pre-1953 densities. The model was based on the logistic equation describing the population growth of Hexagenia and a presumed competitor, Chironomus (dipteran larvae). Other parameters (immigration, low oxygen, toxic sediments, competition with Chironomus, and fish predation) were then individually added to the logistic model to determine their effect at different growth rates.

The logistic model alone predicts 10–41 yr for Hexagenia to recolonize western Lake Erie. Immigration reduced the recolonization time by 2–17 yr. One low-oxygen event during the first 20 yr increased recovery time by 5–17 yr. Contaminated sediments added 5–11 yr to the recolonization time. Competition with Chironomus added 8–19 yr to recovery. Fish predators added 4–47 yr to the time required for recolonization.

The full model predicted 48–81 yr for Hexagenia to reach a carrying capacity of ≈350 nymphs/m², or not until around the year 2038 if the model is started in 1990. The model was verified by changing model parameters to those present in 1970, beginning the model in 1970 and running it through 1990. Predicted densities overlapped almost completely with actual estimated densities of Hexagenia nymphs present in the western basin in Lake Erie in 1990. The model suggests that recovery of large aquatic ecosystems may lag substantially behind remediation efforts.