Resident‐immigrant dichotomy matters for classifying wetland site groups and metacommunities

The fact that species have resident (autochthonous) or immigrant (allochthonous) status at any given locality may have strong implications for ecological analysis. We used wetlands and adult odonates as a model system to evaluate the resident‐immigrant dichotomy for two modes of community analysis: (1) grouping sites based on species compositional variation and (2) identifying metacommunity structure. We tested a hypothesis of gradient‐structured (non‐random) resident occurrence versus unstructured (random) immigrant occurrence in the metacommunity context and predicted the resident occurrence would more effectively partition community variation and produce stronger site groupings than total (resident + immigrant) occurrence. Site group classification after fractioning out resident occurrence consistently and in some cases dramatically outperformed total occurrence. Resident damselflies produced the strongest classifications, which we attribute to greater dispersal limitation, environmental sorting or both. As predicted only the resident occurrence led to identifiable metacommunity structures, primarily Clementsian‐style turnover. This suggests the resident occurrence is gradient‐driven with species responding similarly to abiotic filters, whereas immigrant occurrence is more opportunistic and random. The resident‐immigrant dichotomy appears to have strong influence on quantitative classification of sites and metacommunities, and species composition of resident adult damselflies is potentially useful for differentiating and indicating site groups of non‐forested freshwater wetlands.