Effects of crop management at the landscape scale on biodiversity : a review and research perspectives

International audienceAgricultural activity, through farming practices and landscape changes, is an important driver ofbiodiversity and ecosystem service provision. While the relationship between land-use and biodiversityis widely studied, few studies consider the "hidden" heterogeneity of agricultural landscapes representedby crop management. Indeed, most species living in agroecosystems are affected by the intensity andspatial distribution of farming practices, which can be as—or even more—influential as the diversity ofcrops and non-crop habitats. The objective of our review is to assess 1) how crop management wasdescribed at the landscape scale and 2) the estimated effects on biodiversity and potential of ecosystemservice provision. According to a Web of Science request, we retrieved 133 original studies dealing withthe broad notion of crop management at the landscape scale and rejected studies only considering landcover and land-use metrics. Three components of crop management are studied in analyzed papers:effect of agricultural system (i.e., organic vs. conventional farming), effect of crop rotation, and effect ofindividual farming practices (e.g., soil tillage, pesticide or nitrogen use). Natural enemies, insect pests,pollinators, birds, and weeds are the primary organisms for which the effect of crop management at thelandscape scale has been studied. Overall, preliminary results show that many studies examine theeffect of a single component of crop management (in most cases, the effect of organic farming) and donot compare the effects of crop management at the local and landscape scales. Extensiveness of cropmanagement in the landscape seems to benefit biodiversity and the provision of ecosystem services.While expansion of organic farming in the landscape benefits to pollinators and weeds, diversificationof rotation has positive effect on bird. Individual farming practices have shown species and practicesdependent effects, suggesting that the direction of effect is not obvious and deserves to be studied. Ourfindings suggest that the interest for crop management in the landscape needs to be reinforced inecological studies. A better knowledge of their impact on biodiversity would allow intensifying the jointlandscape-scale management of both heterogeneities of land use and crop management