Dispersal is an important factor in metacommunity dynamics and structure. This is especially true in discrete habitats such as small water bodies in which species exist in small habitat patches where they risk local extinction due to stochastic events. Aquatic invertebrates in such environments can passively disperse between local communities using different vectors (wind, overflow or animal), allowing for populations to be recreated and ensuring the regional survival of species. We examined which vector(s) were most efficient at dispersing different invertebrate species in a natural Jamaican rock pool metacommunity. For the study of dispersal we chose 47 pools of diverse proximity to dense vegetation and the open ocean. We fitted these pools with traps designed to catch individuals dispersing via wind, overflow, and animal vectors. Wind and animal traps consisted of replicated water-filled (freshwater or brackish pool water) containers placed on the rocks adjacent to pools or floating within. Overflow traps aimed to capture outflow of pool water (and dispersing animals) following a rain event. These traps consisted of a thin tube moulded to the mouth of the pool outflow channel with a sealed bag to collect water at the other end.

Results/Conclusions

Overall, we intercepted a total of 1,389 aquatic invertebrates dispersing between pools in a six day sampling period. All of these individuals were caught in wind and wind/rain traps but none in the traps collecting animal vectors (such as amphipods, isopods or crabs). Furthermore, the vast majority of individuals dispersed during juvenile life stages (juvenile ostracods and copepod nauplii). Contrary to past studies, overflow was not responsible for dispersing a large number of invertebrates in this study (we collected only 2 individuals) which may have been attributed to the lack of sufficient rainfall required for pool overflow. The greatest numbers of dispersers were ostracods (~72%), followed by copepods (~23%), with the remainder consisting of a small number of both annelids and crab larvae (<5%). These findings suggest that wind is the main vector responsible for dispersal in the Jamaican rock pool metacommunity, and that juvenile forms or eggs are more readily able to disperse via wind; perhaps in part due to their small body size and specific physical structures that allow them to be easily lifted and moved across pools.