Patch size matters more than dispersal distance in a mainland-island metacommunity.

Micro-arthropods in moss patches have been used as a model system to investigate the effects of habitat destruction and fragmentation on population viability and ecosystem functioning. Previous assessments of the sensitivity to fragmentation and the effectiveness of mitigating landscape structures have to some extent been contradictory, one possible reason being a lack of knowledge of the realised dispersal distances of the species involved. We investigated the dispersal capabilities of oribatid mites (Acari: Oribatida) and springtails (Collembola) in an experimentally fragmented system consisting of bryophytes on a bare rock surface. We used defaunated patches that were recolonized from populated patches nearby as well as from a mainland surrounding the experimental arena, during 10 weeks in summer. We measured within-mainland, mainland-to-island, and island-to-island dispersal, and found that: (1) Oribatid mites were severely dispersal limited within the time frame of the experiment, even at isolation distances of only 5 cm; (2) springtails did not show any dispersal limitation over distances as far as 300 cm; (3) despite the observed dispersal limitation, the mainland had a relatively large influence on microarthropod occurrence, even at 300 cm distance; and (4) the dispersal rates were high enough for both species sorting and-in the case of collembolans-mass effects processes to occur. Our results indicate that fragmentation can strongly influence species occurrence and abundance in natural systems that are limited by dispersal. They also show that the presence of a distant mainland can override the influence of nearby habitat patches on local diversity and abundance.