干旱区生态环境知识资源中心

1. The relative importance of contemporary and historical processes is a fundamental question in understanding patterns of biodiversity. We tested the hypothesis that species-sorting into different habitats (limnocrenes, rheocrenes, helocrenes), rather than history, would account for the greatest variation in macroinvertebrate diversity in desert artesian springs of The Bonneville Basin, U.S.A. These springs were isolated at the valley scale c. 9000 years ago by aridity and high salinity. Thus, the valley scale will account for the greatest variation in community composition if history and dispersal limitation are important, whereas niche-based sorting processes will be most important if habitat accounts for the greatest variation in community composition. 2. We identified 302 taxa from 280 sites and used a partial redundancy analysis, additive partitioning and classification strength (CS) to partition the variability in diversity among the springs. The valley scale accounted for more variation in community composition in limnocrene habitats (32.5%) than all other spatial and environmental variables combined. Valleys also accounted for 58% (additive partitioning) and 83% (CS) of the regional variation in diversity in analyses that included all three habitat types. That is, the average community similarity was 25% across the region, but increased to 41% within valleys. By contrast, habitat filtering did not account for significant variation in community composition in any of the analyses. Our study is one of the few suggesting the over-riding importance of neutral processes in determining patterns of diversity (history and dispersal limitation). 3. The expansion hypothesis suggests that the youthful age of a region, combined with slow dispersal by a fauna dominated by generalists, will maximise the imprint of history. These communities appear to exist in a pre-equilibrial state, where the maximum carrying capacity has not been reached and niche space is plentiful. With time, we predict that local richness will increase while beta-diversity decreases as species expand their distribution across the region. Consequently, the importance of niche-based processes may increase with time as the imprint of history fades.