Data from: A fresh approach reveals how dispersal shapes metacommunity structure in a human-altered landscape

1. To understand species losses from disturbed landscapes, it is important to distinguish the effects of degraded environmental conditions from those caused by barriers to dispersal between habitat patches. To assess the relative importance of these effects, we developed a new approach using permutation and association tests applied to rank abundance data, using the invertebrate fauna of two rivers in two seasons. 2. Our study streams were Hughes Creek and Seven Creeks, in south-eastern Australia, which have both been degraded by agriculture in downstream sections. We collected benthic invertebrates and also dispersing individuals (drift, terrestrial adults) during two seasons in 2007–2008. Study sites spanned strong environmental gradients as well as the main dispersal route (up- and down-channel). Environmental data were analysed to set up permutation tests on rank abundances. Survey and disperser data were contrasted using contingency table analyses. 3. The results suggest dispersal plays a strong role in community structure. Environmental effects were evident and strongest upstream, but evidence of environmental effects was weak over much of the gradient. Many species had different distributions in different data sets or dispersers that were abundant at locations distant from centres of benthic distribution. 4. Our results differ from many studies, but few have been able to evaluate dispersal effects directly. Our method provides a practical approach for evaluating the role dispersal plays in driving species abundance patterns across landscapes, thus bridging a gap between theory and practice. 5. Synthesis and applications. Managers typically use indices of ecosystem health that assume environmental conditions largely determine species diversity and abundance. Dispersal between habitat patches is known to be important, but there are no reliable methods to assess the role dispersal may play. We provide an approach that allows both dispersal and environmental effects on species distributions to be evaluated from survey data. This may open the way for dispersal information to be incorporated into management actions. Additionally, the approach should allow improved siting of restoration projects that depend greatly on successful dispersal of individuals for successful outcomes