i analyzed the resistance and resilience of benthic macroinvertebrate and zooplankton communities from 36 trophically similar lakes north of thunder bay, ontario to test predictions of the oksanen et al. (ofan) model of community regulation and the potential for alternative stable states following natural and timber-harvest disturbances. for each disturbance, only a subset of the lakes were previously affected, allowing an undisturbed control group of lakes to be used for comparison. disturbances differed in the manner in which they "displaced" communities and every disturbance, except the impact associated with the actual area of watershed harvested, significantly altered either the zooplankton or macroinvertebrate community from the structure observed in undisturbed lakes (low overall resistance). communities did not converge on the composition in undisturbed lakes (no resilience), suggesting a stable alternative state. these results suggest that resource managers must consider the effects of land-use disturbances, both separately and from a cumulative perspective, to evaluate the potential impacts on lake ecosystems. when potential productivity was augmented by nutrient-addition via cottage inputs, predictions of the ofan model were rejected over the three lower trophic levels analyzed in these four trophic-level lakes. alternative regression approaches supported predictions only at the basal trophic link, similarly rejecting the model. the ofan model cannot account for the pervasive influence of size-structured interactions at upper trophic levels in aquatic communities.