The concept of metacommunity (i.e., a set of local communities linked by dispersal) has gained great popularity among community ecologists. However, metacommunity research mostly addresses questions on spatial patterns of biodiversity at the regional scale, whereas conservation planning requires quantifying temporal variation in those metacommunities and the contributions that individual (local) sites make to regional dynamics. We propose that recent advances in diversity-partitioning methods may allow for a better understanding of metacommunity dynamics and the identification of keystone sites. We used time series of the 2 components of beta diversity (richness and replacement) and the contributions of local sites to these components to examine which sites controlled source-sink dynamics in a highly dynamic model system (an intermittent river). The relative importance of the richness and replacement components of beta diversity fluctuated over time, and sample aggregation led to underestimation of beta diversity by up to 35%. Our literature review revealed that research on intermittent rivers would benefit greatly from examination of beta-diversity components over time. Adequately appraising spatiotemporal variability in community composition and identifying sites that are pivotal for maintaining biodiversity at the landscape scale are key needs for conservation prioritization and planning. Thus, our framework may be used to guide conservation actions in highly dynamic ecosystems when time-series data describing biodiversity across sites connected by dispersal are available.

No abstract

No abstract

Water scarcity is a global threat to freshwater biodiversity, but connecting variation in streamflow to viability of imperiled faunas remains a challenge. Here we combined time‐series modeling techniques on long‐term ecohydrological data to quantify flow–ecology relationships on native and non‐native riverine fish in the American Southwest, and simulate likely fish trajectories and “quasi‐extinction” risks in the near future. Streamflow has been declining conspicuously over the past 30 years in the Colorado and Rio Grande river basins, and year‐to‐year variation in streamflow influences the covariation between native and non‐native fish abundance. Risks of decline are high (>50%) for nearly three‐quarters of the modeled native species, and current trends in streamflow increase quasi‐extinction risk for natives (+8.5%) but reduce this risk for non‐natives (–5.9%). Hydrological changes need to be mitigated if we are to slow down the rapid replacement of native biodiversity with non‐native species in American Southwest rivers.