Weed metacommunity dynamics in intensive farmland landscapes reveal complex interactions between local disturbance patterns and spatial dispersal

EA GESTAD INRA EA GESTAD INRAEAGESTAD INRAMetacommunity theory represents an exciting framework to identify the ecological mechanisms shaping the spatio-temporal organization of species assembly. Among them, dispersal and niche-based processes have been proposed as major drivers of metacommunity dynamics, whose relative role has led to the emergence of distinct paradigms. Mass-effect emphasizes a greater influence of dispersal whereas patch quality is of primary importance in the species-sorting perspective. Weeds represent the basic trophic component in agricultural food webs, but may induce crop yield loss. Many weed species occur in both crop and non-crop areas such as field edges. Recent studies have revealed the importance of spatial dispersal in shaping weed community diversity. However, up to know, no studies have explored the relative contributions of dispersal versus local processes, such as disturbances induced by agricultural practices. In this study, we quantified the relative importance of local environment, geographic distances, and landscape heterogeneity in shaping weed community. Results revealed that the local environmental conditions, landscape heterogeneity and geographic distances predicted weed diversity, but their relative contributions differed according to spatial scales. Moran’s Eigenvector Maps showed that spatial distances over 3 km strongly contributed to weed richness (up to 30%), but for less than 5% to weed composition, suggesting a weak influence of dispersal on weed assembly at this scale. Using metacommunity theory to predict weed dynamics may help and provide useful insights to develop innovative management and conservation strategies able to limit the general decline of weeds and foster their prominent role in supporting ecological services