No spillover of flower-visiting insects between community gardens and forest fragments in the city

Abstract Cities are a fragmented matrix of cultivated and semi-natural greenspaces. In rural agroecosystems, croplands attached to forest often have higher pollinator diversity than isolated croplands due to temporary spillover from adjacent forests. A similar dynamic could be at play in cities, where different cultivated and semi-cultivated greenspaces could interact via spillover. We studied spillover of flower-visiting insects between different types of cultivated and semi-natural urban greenspaces; community gardens that were attached to urban forests, isolated community gardens that were at least 1 km from urban forests, and isolated urban forests that were at least 1 km away from community gardens. If spillover was occurring, we expected to see higher insect diversity and abundance in the community gardens that were attached to forests compared to isolated community gardens (and visa-versa). We found no evidence of insect-visitor spillover between community gardens and urban forests for all metrics measured. We did, however, find higher native bee and insect species richness in community gardens as compared with urban forests. Unlike their rural counterparts, community gardens are generally more diverse than monoculture farms and may offer the flowering and habitat needs such that insects do not need to spillover into forest. Further investigation into the relative lack of flower-visiting insect species in urban forests compared with gardens is required as this may have ramifications for native plant-pollinator interactions in cities. Our results suggest that growing gardens can attract pollinators regardless of their placement relative to natural forest within the urban landscape.