Effects of local landscape richness on in-field weed metrics across the Great Britain scale

Weeds sit at the crux of issues of arable biodiversity and productivity. We test whether the effects of a simple local landscape richness score are apparent on in-field data for the predominant weeds, in the seedbank and as standing weeds, across the Great Britain (GB) scale. We test two hypotheses: (H1) that in-field weed species richness is positively related to local landscape richness around arable fields; and (H2) that in-field weed abundance is positively related to local landscape richness. We maintain H1, but find little evidence of a consistent effect of local landscape on the species richness of in-field weeds. Convincing evidence for local landscape richness affecting weed abundance lead us to accept H2 as a working hypothesis for in-field weed abundance. However, we have no mechanism to explain why in-field weed abundance should be linked to off-field landscape features, as expectations for farmer management intensity and local bee abundance affecting weed abundance were not significant. The effects of landscape on weed abundance were consistent across the GB scale dataset despite systematic sources of error due to cropping and regional differences. These results suggest that local landscape might be managed to achieve goals of changing in-field weed abundance at policy-relevant scales. (C) 2012 Elsevier B.V. All rights reserved.