No abstract

Theoretical development in the field of community ecology needs ground proofing with empirical tests. In addition, these tests need to be continuously updated. Cottenie (2005) linked observed metacommunities to theoretical models based on whether environmental and/or spatial effects in the observed metacommunity significantly explain community structure. However, a species-sorting metacommunity with high dispersal and one with limited dispersal cannot be distinguished in this manner; both produce significant environmental and spatial effects. In the present study, we demonstrate a solution to this problem using a zooplankton rock pool metacommunity in Churchill, Manitoba, sampled in August 2006. We established a hierarchy of metacommunities in the Churchill rock bluff system-a large, across-bluff metacommunity, and small, within-bluff metacommunities. Using this spatial hierarchy, it is possible to determine the zooplankton dispersal capability in the rock bluff system and hence to link the metacommunity to its corresponding model. We found the zooplankton rock bluff system to exhibit limited dispersal, meaning that spatial effects were significant at the acrossbluff scale, but depending on the bluff, were significant or insignificant at the within-bluff scale. Environmental effects were significant at both scales. This study demonstrates a novel way to determine dispersal capabilities in species that are cryptic dispersers, and to successfully link observed metacommunities with theoretical models.Keywords Metacommunity Á Dispersal Á Spatial scales Á Rock pools Á Sub-arctic Á Zooplankton Handling editor: S. I. DodsonElectronic supplementary material The online version of this article (