Relative importance of different dispersal vectors for small aquatic invertebrates in a rock pool metacommunity


  • Bram Vanschoenwinkel,

  • Saïdja Gielen,

  • Hanne Vandewaerde,

  • Maitland Seaman ,

  • Luc Brendonck

B. Vanschoenwinkel (, S. Gielen, H. Vandewaerde and L. Brendonck, Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Univ. Leuven, Ch. Deberiotstraat 32, BE-3000 Leuven, Belgium. – M. Seaman, Centre for Environmental Management, Univ. of the Free State, Mandela Drive, ZA-9300 Bloemfontein, South Africa


The extent and frequency of passive overland dispersal of freshwater invertebrates as well as the relative importance of different dispersal vectors is not well documented. Although anecdotal evidence subscribing the feasibility of individual vectors in various aquatic systems is abundant, dispersal rates have rarely been quantified for different vectors in one study system. Earlier studies also usually investigated dispersal potential rather than actual dispersal rates. In this study we have estimated passive dispersal rates of invertebrate propagules within a cluster of temporary rock pools via water, wind and amphibians in a direct way. Overflows after heavy rains mediated dispersal of a large number of propagules through eroded channels between pools, which were collected in overflow traps. Taking into account model based predictions of overflow frequency, this corresponds with average dispersal rates of 4088 propagules/channel yr−1. Wind dispersal rates as measured by numbers of propagules collected on sticky traps mounted between pool basins were very high (average dispersal rate: 649 propagules m−2 in one month) and were positively related to the proximity of source populations. Finally, invertebrate propagules were also isolated from the faeces of African clawed frogs Xenopus laevis caught from the pools (on average 368 propagules/frog).

The combination of short distance wind and overflow dispersal rates likely explain the dominant species sorting and mass effect patterns observed in the metacommunity in a previous study. Amphibian mediated dispersal was much less important as the Xenopus laevis population was small and migrations very rare.

Based on our own results and available literature we conclude that both vector and propagule properties determine local passive dispersal dynamics of freshwater invertebrates. Accurate knowledge on rates and vectors of dispersal in natural systems are a prerequisite to increase our understanding of the impact of dispersal on ecology (colonisation, community assembly, coexistence) and evolution (gene flow, local adaptation) in fragmented environments.