The interaction between the effects of habitat heterogeneity and isolation on beta diversity suggest the nature of community assembly in river networks

Background/Question/Methods

Community ecologists are shifting their view of species assemblages from isolated, locally-interacting species to those occurring on a landscape connected via dispersal.  This expanded view of how species co-occur has led to exciting insights into broad scale patterns in local and regional diversity, and species turnover in space and time (beta diversity). With low dispersal, beta diversity is hypothesized to develop as habitat heterogeneity increases, owing to interspecific variation in habitat suitability.  However, dispersal constraints should interact with habitat heterogeneity to generate patterns in beta diversity. Two mechanisms invoked to explain this include differential assemblage histories and stochastic colonization. If assemblage histories, or priority effects, are important, beta diversity in common species should be higher in well-connected metacommunities as habitat heterogeneity increases.  However, if stochastic colonization is driving assembly in well-connected metacommunities, beta diversity in rare species should increase. To test these predictions, we partitioned invertebrate diversity in 42 river drainages across the state of Maryland (USA) separately for small, headwater streams and larger, mainstem streams. We analyzed beta diversity as a function of basin-scale habitat heterogeneity, measured as spatial variation in impervious land cover, stream size, and rare vs. common taxa. 

Results/Conclusions

For both headwaters and mainstems, beta diversity increased with basin-scale habitat heterogeneity.  However, the magnitude of this increase differed depending on rarity. In general, beta diversity in common species assemblages were more weakly associated with habitat heterogeneity than rare species assemblages, suggesting that stochastic colonization is playing a large role in shaping invertebrate community structure. In smaller streams, though, the overall level of beta diversity was much higher for rare species compared with mainstems, supporting previous findings that isolation in this spatially structured environment is important to explaining patterns in species turnover. We did observe a slightly more positive relationship between beta diversity of common species assemblages and habitat heterogeneity in mainstems compared with headwaters. While a relatively small difference, this suggests the potential for priority effects as important in larger rivers. We conclude that different assembly mechanisms are driving patterns in species turnover in river networks, emphasizing the need to consider differential dispersal and/or habitat connectivity when partitioning biodiversity in spatially-structured ecosystems.