Relative effects of local management and landscape heterogeneity on weed richness, density, biomass and seed rain at the country-wide level, Great Britain

Weed management is a critical issue as it is faced with a daunting set of challenges linked to crop productivity and farmland biodiversity. Developing new strategies for weed management requires a clear understanding of the relative role of local management and landscape heterogeneity on weeds. Yet few studies have investigated the combined effect of these factors on a variety of weed metrics that reconcile agronomical and ecological aims. Here, we analyzed the relative role of local management intensity and landscape heterogeneity according to the distance to field margin on four weed metrics (species richness, density, biomass and seed rain) and community composition in 257 fields (beet, maize, spring and winter rape fields) across Great Britain. Generalized mixed effect models followed by a model averaging procedure were applied on weed metrics and permutational multivariate analyses of variance were applied on weed species composition. Our analysis confirmed the overriding role of the distance to field margin on weeds. Although weed density and seed rain negatively responded to local management intensity, they did not respond in the same way to landscape configurational heterogeneity, namely the field size. We found interactions between management intensity and landscape heterogeneity but only in relation to weed biomass in beet and spring rape fields and to seed rain in beet fields. The relative importance of local management intensity and landscape heterogeneity varied depending on the distance to field margin, which can be attributed to spatial heterogeneity in management practices. We recommend that not only species richness but also a wide range of metrics should be considered in weed studies as they did not responded in the same way to local and landscape factors. We conclude that weed management strategies should be thought by integrating a multi-level approach as the combined effects of local management and landscape heterogeneity are likely to both reduce weed infestation whilst enhancing biodiversity.