Background/Question/Methods The importance of competition is often assumed to decrease with increasing disturbance, which, combined with the presence of competition-disturbance tolerance trade-off among species, leads to hump-shaped disturbance-diversity relationships as predicted by the intermediate disturbance hypothesis. However, theoretical evidence suggests that even though the strength of competition may be reduced with high levels of disturbance, its role in regulating community structure may not necessarily be diminished. This arises from the fact that species enduring disturbance attain small population sizes, making them (including disturbance-tolerant species) vulnerable to extinction even at low levels of competition. Under this circumstance, hump-shaped disturbance-diversity patterns may not emerge. We tested this hypothesis with an aquatic microbial microcosm experiment manipulating 11 bacterivorous ciliated protist species and 11 disturbance (imposed by sonication of entire communities) intensity levels in three steps: (i) species were grown in isolation and submitted to the 11 disturbance levels to assess their ability to tolerate disturbance; (ii) all possible two-species combinations were established in the absence of disturbance to assess species competitive abilities; (iii) Communities containing the 11 protist species were subjected to the 11 disturbance levels to evaluate the relationship between disturbance and species diversity.
Results/Conclusions The 11 protist species displayed a clear competition/disturbance tolerance trade-off: when grown in isolation, weak competitors persisted at high levels of disturbance while the best competitors persisted only at low levels of disturbance. However, in the 11-species mixtures, competition increased species extinction rates, especially at high levels of disturbance where the most tolerant species disappeared in the early stage of the experiment. Competition increased the probability of species extinction by reducing the population size of all co-occurring species that increased their vulnerability to extinction over the entire disturbance gradient. Consequently, species diversity declined monotonically with disturbance despite the presence of the competition-disturbance tolerance trade-off. These results demonstrated a predominant role of competition in regulating community structure at high disturbance levels and challenge the common belief among ecologists that the importance of competition declines with disturbance. The important structuring role of competition also provides an explanation for the non-hump-shaped disturbance-diversity relationships commonly reported in empirical studies.