Flow-related disturbance creates a gradient of metacommunity types within stream networks

Abstract

Context: Metacommunities are sets of local communities linked by dispersal. Their characteristics are defined by both large-scale spatial processes such as dispersal, and local environmental processes, although which factors are likely to predominate in a given situation is poorly understood. Objectives: We investigated whether flow regime at the network-scale helped explain the relative importance of spatial and local-environmental processes in structuring stream metacommunities. Methods: Spatial sampling of stream macroinvertebrates was carried out in stream networks in New Zealand. Local environmental variables were also measured throughout the stream networks, while hydrographs were modelled and calibrated with field measurements. Results: Significant associations with both spatial and local-environmental predictor variables were found, consistent with several metacommunity types. In particular, two measures of flow regime were associated with different metacommunity types. Thus, stream networks characterised by a period of stability just before sampling, and networks sampled after a long period of instability, had more significant spatial structuring of metacommunities than those of intermediate flow stability. The importance of spatial processes in structuring the network metacommunities also increased with time since the last community-resetting flow. Our results therefore suggested that metacommunity type depended on the flow regime. Dispersal traits and network topology also helped explain some of the variation among the metacommunities. Conclusions: Overall, our findings conform to theoretical predictions related to dispersal limitation and topology, and indicate that metacommunity models need to be dynamic to capture processes in both space and time.