Biodiversity and ecosystem functioning contribute to human well-being by providing ecosystem services. While local and landscape management have been assumed to directly impact biodiversity and indirectly impact ecosystem services, the specific relationship between habitat management, biodiversity, and service provision has rarely been evaluated simultaneously for multiple taxa across landscape contexts, especially in the increasingly widespread urban agroecosystem. Further, few studies have examined whether trade-offs or synergies may exist between services given distinct local and landscape management practices. In this research we ask: 1) How do local management factors and landscape characteristics affect biodiversity and ecosystem services in urban agroecosystems? 2) Are there trade-offs or synergies among factors that affect distinct services? and 3) Do the impacts of local factors on ecosystem services depend on landscape context? We used data collected over 5 years in 25 urban community gardens in the California central coast to systematically examine how a suite of local features (mulch cover, herbaceous plant richness, floral abundance, and garden size) and landscape features (natural cover) influence ecosystem service providers (e.g. natural enemies, pollinators) and services (e.g., pest control, climate mitigation).
Results/Conclusions
First, we show that in many cases, the relationship between the putative ecosystem service provider and ecosystem service is not strong, and that local habitat factors may be more predictive of service level. Second, we reveal that multiple trade-offs exist in the management of local habitat for multiple ecosystem services. For example, while increasing local tree density and ground cover significantly increases climate regulation services, these two local habitat factors negatively impact pest-control services and pollination services, respectively. At the landscape scale, increasing natural habitat availability positively affects climate regulation and pollination services, but negatively affects pest-control. Finally, we show that nearly half of all ecosystem services are best predicted by a local habitat and landscape interaction. For instance, the impacts of local management (e.g., tree density) on ecosystem service provision (e.g., pollination) critically depend on natural habitat availability; this is likely driven by the large number of mobile ecosystem service providers that only colonize gardens given a minimum threshold of landscape-level habitat availability. Overall, these results provide unique insight into how gardeners can manage local and landscape abiotic and biotic factors to better support biodiversity and maximize service synergies.