Wild bees and urban agriculture: assessing pollinator supply and demand across urban landscapes

Growing interest in urban agriculture has increased demand for pollination services. Most studies map pollination supply broadly, and do not consider the impacts of fine-scale urban land-use practices on the dynamics of pollination delivery, leaving a critical gap in our understanding of the pollinator supply-demand balance in urban landscapes. This study demonstrates a spatially-explicit framework, using Iowa City, IA (USA) as the case study region, for assessing the capacity of urban ecosystems to produce pollinator services in support of demand from urban agriculture. We estimate pollinator supply using the InVEST pollination model with detailed land-cover data produced through field survey and Bayesian hierarchical analysis, and we validate modeling results with bee abundance and richness data. We map social demand for pollinators using a kernel density estimation of urban agricultural sites and evaluate supply-demand budgets through spatial overlay analysis. Our results show that incorporating high-thematic-resolution urban land-use data substantially improves the performance of pollination supply modeling. Pollinator supply meets demand in 72% of the city. Surpluses occur in natural areas and heavily-vegetated, established residential neighborhoods, whereas deficits occur in resource-poor lawns. Our mapping framework stresses the key role of humans in modifying resource availability and pollinator services, and demonstrates the effectiveness of using disaggregated socio-economic data in urban land-cover classification for predicting pollinator supply. Our improved ability to identify spatial congruence and disparities in urban pollinator supply and demand can be used to inform pollinator conservation to support sustainable urban agriculture.