COS 156-5 - Urban agroecosystem management and landscape change shift pollinator richness, plant-pollinator networks, and pollination services

Abstract: Given that ecological network structure is tightly linked to community function, network analyses can inform how provisioning of ecosystem services in agroecosystems will be affected by local and landscape management characteristics. Interaction evenness describes the robustness of a community based on its ecological redundancy; nestedness captures potential buffering to temporal changes in abundance of specialist species; and connectance reflects the degree of generalization within a community. Pollination is an essential ecosystem service in urban agroecosystems, but we know relatively little about how local garden and urban landscape management impact pollinator richness, plant-pollinator networks, and pollination services. Our study asked how local garden (e.g., floral abundance, plant richness, ground cover) and landscape management (e.g., natural habitat cover) impact pollinator richness, plant-pollinator networks, and pollination services. We surveyed garden and landscape management for 23 gardens in coastal California in summer 2021. We netted pollinators (e.g. bees, butterflies, flies) for 4h per garden, recorded each visited plant species, and used observation data to construct plant-pollinator networks. We also conducted a pollination limitation experiment with peppers by comparing seed number, seed weight, and fruit weight among closed, open, or open outcross treatments.

We collected 1450 individuals and 98 species of bees, flies, and butterflies visiting 159 plant species from 43 families. Common pollinators included Apis mellifera (32.0% of visits, 99 species), Bombus vosnesenskii (13.8%, 57), and Halictus tripartitus (7.0%; 33). Commonly visited plant families were Asteraceae (29.0% of visits), Lamiaceae (17.9%) and Apiaceae (11.7%) and commonly visited plant species included cilantro (7.9%), lavender (6.3%) tomatillo (6.1%), borage (4.8%), and squash (4.6%). Pollinator richness increased with flowering plant richness (p=0.08), and decreased with mulch cover (p=0.09) and natural habitat (p=0.07). Interaction evenness decreased with vegetation height (p=0.035) and increased with natural habitat (p=0.045). Connectance declined with flowering plant richness (p=0.057) and vegetation height (p=0.067), while nestedness did not vary with any local or landscape factor. Pepper seed number (p=0.02) and fruit weight (p< 0.001) increased with floral abundance, and fruit weight declined with increased pollinator richness (p=0.018). No network metrics predicted pepper pollination. While pollinator richness and network metrics responded to both local and landscape management, pollination responded only to local management and pollinator richness; thus gardeners can manage local habitats to promote both biodiversity and pollination.