Résumé
The meta-ecosystem concept provides a theoretical framework to study the effect of local and regional flows of resources on ecosystem dynamics. Meta-ecosystem theory has hitherto been applied to highly abstract landscapes, and meta-ecosystem dynamics in real-world landscapes remain largely unexplored. River networks constitute a prime example of meta-ecosystems, being characterized by directional resource flows from upstream to downstream communities and between the terrestrial and the aquatic realm. These flows have been thoroughly described by the River continuum concept (RCC), a seminal concept in freshwater ecology, stating that observed spatial variations in the relative abundances of invertebrate functional groups reflect systematic shifts in types and locations of food resources, which are in turn determined by the physical attributes of river reaches. Hence, the RCC represents a solid conceptual basis for determining how changes in landscape structure and resource flows will translate into local and regional changes in community composition. Here, we develop and analyse a riverine meta-ecosystem model inspired by the RCC, which builds upon a physically-based landscape model of dendritic river networks. We show that the spatial distributions and regional biomass of invertebrate functional groups observed in stream communities are determined by the spatial structure and scaling attributes of dendritic river networks, as well as by specific rates of resource flows. Neglecting any of these aspects in modelling river meta-ecosystems would result in different community patterns. Moreover, we observed that high rates of resource flow, for example due to river anthropization, have a negative effect on the regional biomass of all functional groups studied, and can lead to cascading extinctions at the meta-ecosystem scale. Our work paves the way for the development of physically-based meta-ecosystem models to understand the structure and functioning of real-world ecosystems.