Experimental wildflower additions increase pollination in urban agroecosystems

O_LIThe addition of nectar-rich flower patches in human-modified ecosystems is a common practice to mitigate pollinator declines and boost pollination. However, the benefits of these additions for pollinator communities and pollination services are rarely tested, especially in urban environments. C_LIO_LIIn a city-scale experiment we added floral resources to urban allotments and monitored the effects on bees, hoverflies and moths, and tested for improved seed set in a model crop (tomato, Solanum lycopersicum). C_LIO_LIThe addition of wildflowers did not benefit all insect communities. Only social bee abundance (Bombus and Apis) benefitted from increasing floral resource area whereas other insect taxa showed no changes in abundance potentially due to the divergence in foraging patterns of moths, hoverflies, social bees and solitary bees. The addition of wildflower patches enhanced pollination by supporting a 25.3% increase in tomato seed set, providing evidence that wildflower interventions can improve urban pollination. Seed set was higher in more urban sites, suggesting an "oasis effect" where pollinating insects are concentrated into limited greenspaces. This highlights the precarity of pollination services in highly urban areas. C_LIO_LIOur results suggest that the practice of planting wildflower patches can positively affect pollination services in urban areas. The continued promotion of flower patch addition is likely to benefit some key insect taxa, however, the common wildflower species in seed mixes may not benefit hoverflies and moths compared to bees. The taxon-specific foraging patterns we observed should inform the design and development of pollinator-friendly wildflower seed mixes. C_LI Societal impact statementEnhancing urban greenspaces for pollinator communities by planting wildflower patches is increasingly common, but their efficacy for different groups of insects (bees, hoverflies and moths) is unclear. Our city-scale experiment demonstrated that wildflower patches benefitted the abundance of social bees, but did not increase other groups. Wildflower addition increased pollination services, with an increase in seed-set in our model crop, particularly in more urban areas. Wildflower patches do benefit pollinator communities and in turn humans, through the pollination services, but the species mix may not benefit some insect groups. This suggests there is potential to improve the benefit provided by wildflower patches by redesigning pollinator-friendly seed mixes.