Ecological interactions between crops and wild animals frequently result in increases or declines in crop yield. Yet, positive and negative interactions have mostly been treated independently, owing partly to disciplinary silos in ecological and agricultural sciences. We advocate a new integrated research paradigm that explicitly recognizes cost-benefit trade-offs among animal activities and acknowledges that these activities occur within social-ecological contexts. Support for this paradigm is presented in an evidence-based conceptual model structured around five evidence statements highlighting emerging trends applicable to sustainable agriculture. The full range of benefits and costs associated with animal activities in agroecosystems cannot be quantified by focusing on single species groups, crops, or systems. Management of productive agroecosystems should sustain cycles of ecological interactions between crops and wild animals, not isolate these cycles from the system. Advancing this paradigm will therefore require integrated studies that determine net returns of animal activity in agroecosystems.

Reducing fragmentation and habitat loss by restoring or maintaining connectivity has been promoted as a way to mitigate the negative impacts of human activities on biodiversity. This study is an example of collaboration between spatial researchers and on-ground practitioners, to deliver better informed management options for investment in connectivity and biodiversity outcomes. Using the Border Rivers-Gwydir catchment revegetation programmes in New South Wales, Australia, we describe a fit-for-purpose, cross-scale methodology consisting of multiple-component models, where each component reflected varying spatial scales. The methodology was based on the concepts of metapopulation ecology and landscape ecology and used least-cost paths analyses. At the wider scale, native vegetation extent and condition were used as a surrogate for all biodiversity; at the finer scale, landscape structure and generalised movement parameters related to a focal woodland species group were used to derive least-cost paths. The output from the analyses provided spatially explicit management action zones that were used to prioritise areas for revegetation investment. Combining local priority zones for linking habitat with regional-scale and broad-scale zones should increase access to resources for biota, increase dispersal potential and thereby enhance biodiversity persistence. Promoting connectivity is a global concern. Our approach could be relevant in other geographical settings where the implementation needs of NRM practitioners can be assisted through the application of scientific knowledge.

Bees are known to provide pollination services to a wide range of crops, yet flies are rarely included in estimates of function. As bees and flies differ markedly in their life history characteristics and resource needs, they may be active and hence provide pollination services at different times of the day. Here, we explore the differences in bee and fly diurnal activity patterns and how this may impact upon pollination services provided to 'Brassica rapa', a mass-flowering crop. We observed pollinators at two-hourly intervals from 6:00 to 20:00 h in twelve fields in New Zealand in 2004-2005. Overall, bees were most active in the middle of the day and were more effective pollinators than flies, driven primarily by the high pollinator efficiency of 'Apis mellifera' and 'Bombus terrestris'. Some fly taxa however, visited flowers early and late in the day when there were few bees. The results of this study demonstrate that fine-scale temporal dynamics and the spatial distribution of crop pollinators may directly affect the quantity of pollination services. The maintenance of biodiversity in agro-ecosystems may therefore be critical to ensure pollinator taxa are available under a range of environmental conditions.

'Aim': Pollination services are at risk from land use change and intensification, but responses of individual pollinator species are often variable, making it difficult to detect and understand community-level impacts on pollination. We investigated changes in community composition and functional diversity of insect pollinator communities under land use change in a highly modified landscape. 'Location': Canterbury region, South Island, New Zealand. 'Methods': We trapped insect pollinators every month for 1 year at 24 sites across four land use types of increasing intensity in New Zealand: gardens with native vegetation, blackcurrant orchards, dairy farms, and rotational cropping farms. We investigated changes in pollinator species and functional richness and differences in species and functional composition. 'Results': Under increasing land use intensity, both species and functional richness declined markedly. Changes in functional richness, however, were overall not significantly different than expected based on the observed declines in species richness. Nevertheless, there was a significant trend towards greater than expected functional richness within less-intensive land use types and lower than expected functional richness within intensive land use types. The order of species loss under increasing land use intensity was non-random, as pollinators with a narrow diet breadth, large body size, solitary behaviour and a preference for non-floral larval food resources were lost first. 'Main conclusions': Our study shows that pollinator species bearing particular trait attributes are susceptible to differences in land use. Our study suggests that pollination services may be more vulnerable to environmental changes and disturbances in more intensive land use types as a result of lower pollinator functional richness.