Disentangling the interrelated abiotic and biotic pathways linking landscape composition and crop production.
Landscape composition and its related functional agrobiodiversity (FAB) was severely simplified during the last decades. As landscape composition is expected to influence the interrelated microclimate and arthropod community at different scales, this simplification might have led to a decline in multiple agro-ecosystem services, with potential impacts for the growth of crops with different demands from the environment. To study landscape-scale effects on multicrop yield and herbivory in combination with its potential drivers, including functional arthropod community and microclimate, we used standardised 1 m2 mini-gardens planted with ten different vegetable crops as phytometers. The gardens were installed at locations varying in surrounding landscape composition and the vegetables were selected to have contrasting requirements in terms of growing conditions, pollination services and susceptibility to pests. Monitoring of the mini-gardens was performed in 2018 and 2019 in Flanders (Belgium). We found no relationship between landscape composition in a 500 m radius and crop yield. Considering possible mechanisms, we found that an increasing share of woody vegetation (>3 m high) in the surrounding landscape is important to buffer the temperature and soil moisture variation in the mini-gardens. The share of agricultural land use and urban green (<3 m high) is, respectively, positively and negatively related with the activity-density of natural enemies and pollinators. The growth of different crops responded differently to higher temperature ranges and we found no relation between natural enemies and herbivory. In conclusion, these abiotic and biotic pathways did not cause an overall relationship between landscape composition and yield. Synthesis and applications. These outcomes indicate that woody vegetation in the rural landscape buffers soil moisture and temperature variation, which will become even more important due to climate change. Furthermore, pollinators and natural enemies were most active in rural environments and less active in areas with residential green and built-up surfaces. Local, flower rich habitat, such as road sides, can further support natural enemies and buffer soil moisture variation. Yet, high arthropod activity is no assurance for arthropod mediated ecosystem service delivery nor is microclimate buffering an assurance for higher crop yields.