Stacy M. Philpott, Gabriella L. Pardee, Jenny Hsu, Andy J. Kulikowski, Peter Bichier, Heidi Liere, Brenda B. Lin, Shalene Jha
Species richness, a fundamental driver of agroecosystem function, is shaped by habitat management and landscape surroundings. Yet, delivery of ecosystem services, like pollination, in agroecosystems is dependent on ecological interactions, not just species richness. Documenting changes in functional groups and ecological networks can inform expected shifts in mutualistic dynamics and impacts on ecosystem services. We recorded pollinator visitation to plants in 23 urban agroecosystems in California during July 2021 and used generalized linear models to examine how local management and landscape changes influenced abundance and species richness of three pollinator functional groups (bees, flies, butterflies/moths) and plant-pollinator networks. We observed 99 pollinator species visiting 157 plant species. Together, three pollinator species accounted for >52 % of the pollinator visits. Five plant species hosted nearly 30 % of pollinator visits. Pollinator species richness increased with higher floral richness, but decreased with more natural habitat in the landscape, potentially due to a dilution effect. Bee richness increased with higher bee abundance, higher floral richness, smaller mean flower size, and larger garden size, but bee abundance and richness declined with more natural habitat cover. Fly richness increased with higher fly abundance and decreased with larger garden size. Butterfly/moth richness increased with higher butterfly/moth abundance, while butterfly/moth abundance declined with more natural habitat cover. Network specialization increased while nestedness, interaction evenness, and connectance all decreased with higher herbaceous plant species richness, but no other local or landscape changes affected network metrics. Individual pollinator and plant species differed in degree of specialization. Pollinator functional groups selectively visited different plants with bees more likely to visit ornamentals, flies more likely to visit crops, and butterflies more likely to visit weeds. We show that urban agroecosystems can support pollinators, and that changes to floral richness, flower volume, and natural habitat cover are the most important for shaping ecological interactions between pollinators and plants.