Diversified agroecosystems for biodiversity and ecosystem services : Ecological intensification of faba bean cropping under land use and climate change

Abstract: Loss of natural habitats and lack of continuous floral resources in intensive agricultural landscapes limit the population growth of ecosystem service providers, such as pollinators. Weakened ecosystem service delivery and climate change stressors contribute to crop yield instability. The aim of my thesis was to explore how resource diversification strategies at both field and landscape scale influence arthropod communities and the services they provide, such as pollination and biological pest control. Further, I explored how climate change stressors such as water stress and insect herbivory interact with insect pollination and affect crop yield. To address these aims, I used faba bean cropping as a model system. I found that landscape crop diversity enhanced bumble bee densities and flower strips enhanced ground predators in the crop. Flower strips boosted and honey bee hives counteracted bumble bee queen abundances in the landscape across seasons. Using pollinator exclusion experiments, I showed that insect pollination increased yield. This insect pollination benefit decreased with increasing semi-natural habitat cover in the landscape but was constant irrespective of plant water stress and insect herbivory. I demonstrated that the highest yields are achieved when water stress and insect herbivory are reduced simultaneously. My thesis emphasises that diversified agroecosystems are particularly important in the face of climate change, because they 1) enhance ground predators that can safeguard crops from herbivore outbreaks, and 2) provide pollinators with floral resource continuity, which arguably is particularly important during extremely hot and dry summers, where landscape-wide floral resource bottlenecks are worsened by shortened bloom periods.

  This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.