Determinants of local and regional communities in intermittent and perennial headwaters of the Bolivian Amazon

SUMMARY

1. The effect of drying events on aquatic biodiversity is still overlooked in wet Neotropical systems. Yet, the responses of local communities and metacommunities in these biodiversity hotspots may differ from what is reported in other areas.
2. We addressed the effect of drying events on local and regional fish and macroinvertebrate communities in the headwaters of the Chipiriri River basin, in the wet Neotropical piedmont of Bolivia. According to current knowledge in intermittent rivers and ephemeral streams (IRES) ecology, we predicted that intermittent (INT) sites would harbour lower a-diversity, but higher b-diversity, than perennial (PER) sites, due to local, negative effect of drying combined with the existence of multiple recovery stages at the network scale.

3. Although habitat variables were similar between INT and PER sites, local and regional diversity patterns differed. Local invertebrate communities were not different between site types as soon as 4–6 weeks after flow resumption. The proximity of colonist sources and frequent rainfall probably enhanced persistence through dry periods and high resilience. In contrast, fish communities were still poorer at INT than PER sites, indicating they were still in the process of recolonising upstream INT reaches.

4. b-diversity analyses confirmed that invertebrate and fish metacommunities were not at the same recovery stage because (i) b-diversity of invertebrates was best explained by physical and environmental distances at both INT and PER sites, whereas that of fish was explained only by physical distances at INT sites; (ii) fish b-diversity was higher at INT than at PER sites, but invertebrate b-diversity was similar; and (iii) physical distances were correlated with the turnover component of invertebrate b-diversity but with the nestedness component for fish.

5. Exploring regional community patterns in IRES and across biota with different dispersal abilities and modes can advance metacommunity theory and improve our ability to predict local community composition in dynamic ecosystems.