Itinerant, nomad or invader? A field experiment sheds light on the characteristics of successful dispersers and colonists

Abstract

Abstract Many hypotheses in ecology depend upon identifying species that can disperse across landscapes and colonise new habitat patches. Species differ in dispersal ability, yet some good dispersers reach new locations but fail to colonise them (“itinerants”). Additionally, successful colonists may comprise species that routinely disperse and maintain connectivity among locations (“nomads”) as well as species that can also take advantage of new situations (“invaders”). In a previously reported landscape‐scale, field experiment that boosted detrital resources in Hughes Ck (south‐eastern Australia), species diversity increased at manipulation sites due to successful dispersal and colonisation. The species composition at manipulation sites gradually converged on that seen at upstream sites, implying the latter locations supplied dispersers. Here, we ask (a) Did dispersers arise from within the same creek, and were drift and adult flight both implicated? (b) Were successful colonists (“responders”) clearly better dispersers (e.g., nomads or invaders) than those that did not colonise manipulation sites (“non‐responders”)? (c) Did traits commonly assumed to denote successful dispersal provide a reliable guide to species that were actually successful colonists? Simultaneous with benthic samples, we collected animals in the drift immediately upstream of control and manipulation sites, and trapped winged adults along the banks. We used PERMANOVA to contrast (a) the assemblages of species in the drift with those in the benthos and (b) the assemblages of adults, at different sites and times. For 54 common taxa (26 responders; 28 non‐responders), we used linear regression to test whether benthic densities were significantly related to drift numbers and, if so, whether the relationship was the same at manipulation and control sites (“nomads”) or delivered higher densities at manipulation sites (“invaders”). We compiled information on dispersal traits (presence in drift, voltinism, flight‐capable adults) to assess whether dispersal traits predicted species that successfully colonised manipulation sites. Drift assemblages were more similar to benthic assemblages at manipulation sites than to those at control sites. Adult assemblages did not differ between manipulation sites and controls, but adult assemblages at downstream sites converged on those seen upstream. Many dispersers arose from within Hughes Creek, with ˜60% of common taxa arriving via the drift. Among the common taxa, 4% were poor dispersers (rarely in the drift) and 33% were itinerants (drift and benthos were unrelated), with these categories equally represented among responders and non‐responders. In the remaining taxa, benthic densities were related to drift numbers. Most responders were invaders, whereas most non‐responders were nomads. Responders and non‐responders did not differ in traits that are commonly assumed to reflect a high potential for dispersal and colonisation. Responders were not demonstrably better dispersers than non‐responders but instead were able to colonise manipulation sites. Contrary to common assumptions, these differences in colonisation ability were not captured by dispersal traits. Colonisation ability has rarely been separated from dispersal ability in stream studies, and yet research in other ecosystems demonstrates that differences in capacity to invade local communities are central to understanding community structure across landscapes.