Reconsidering the role of 'semi-natural habitat' in agricultural landscape biodiversity: a case study

Semi-natural habitats are considered as the main source of biodiversity in agricultural landscapes. Most landscape ecology research has focused on the amount (relative surface) and spatial organisation of these habitats. However, these two components of landscape heterogeneity, composition and configuration, are often correlated. Also, landscape structure effects on biodiversity were mostly observed locally, while there is a great need for studying landscape-scale gamma diversity. We conducted a mensurative experiment to determine the independent effects of semi-natural habitat amount and configuration on gamma diversity of carabid beetles and vascular plants. The influence of landscape heterogeneity components were tested on species richness, evenness and composition. Local diversity (species richness and composition) was compared across the various cover types to determine their relative contributions. Only carabid species evenness and composition were influenced by semi-natural habitat amount. Carabid and plant species richness and plant species composition remained unaffected by semi-natural habitats. Local diversity analysis showed that three types of habitats can be distinguished in agricultural landscapes: grasslands (temporary and permanent ones), woody habitats (woodlands and hedgerows) and row crops. These results beg for a re-evaluation of the semi-natural covers. Temporary and permanent grasslands are often similar, probably because of comparable farming management. Permanent grasslands and woody habitats are often combined as semi-natural covers, although they support very different communities. The lack of effect of semi-natural habitat amount and configuration on gamma diversity results from a more complex organisation of biodiversity in landscapes and supports the move from semi-natural vs. farmland to habitat mosaic landscape representations.