Aspects of flow variability and spatial context predict temporal beta diversity in river metacommunities

River catchments are dynamic networks that contain multiple levels of spatial and temporal complexity. Benthic macroinvertebrates are key indicator taxa throughout catchments, and beta diversity has been used as a metric to explore determinants of community composition at the catchment scale. Although factors influencing spatial beta diversity have been explored, the determinants of temporal beta diversity have been understudied. In this study, we aimed to understand how shifts in benthic macroinvertebrate community composition over multiple years are related to local and regional variables. We used beta regression with a large, publicly available biomonitoring dataset from river networks in California, USA, to model the effects of local environmental variables, flow variability, and spatial network context on benthic macroinvertebrate temporal beta diversity. The spatial variables channel slope, drainage density, and upstream catchment area had the strongest relationships with macroinvertebrate beta diversity. Channel slope was negatively related to temporal beta diversity, whereas drainage density and upstream catchment area were positively related to temporal beta diversity. Temporal beta diversity was also higher when the rate and magnitude of rises and falls in flow were higher in the hydrograph, as well as when the number of zero-flow days and the duration of flow rises and falls were higher. Overall, our results indicate that temporal beta diversity of freshwater benthic macroinvertebrates is shaped by both long-term hydrological context and spatial context and that these factors may serve as better predictors of long-term community variability than variability in point estimates of environmental measurements. Our study supports the need for biomonitoring efforts at long spatial and temporal timescales and highlights the need to consider metacommunity change in the management of freshwater systems.