Native plant communities designed for infiltration swales can tolerate urban heat, flooding, and pollution

• Native grasslands are suitable models to improve communities of urban infiltration swales. • Composition of the seeded grassland communities drives functionality of infiltration swales under stress. • There are no immediate negative effects of heating, flooding, or pollution on the communities. Under global change urban environments need to become resilient to sustainably provide liveable conditions. With growing cities and limited space, the design of Urban Green Infrastructure (UGI) must provide multiple ecosystem services, including microclimate regulation, aesthetic enhancement, biodiversity support, and stormwater management. Infiltration swales are a UGI type designed to manage stormwater by reducing peak runoff, increasing infiltration, and retaining pollutants. Little is known about the respective plant communities that must cope with multiple stressors, including urban heat, periodic flooding, and contamination from surface runoff. Thus, we designed a gradient of seven seed mixtures, each composed of different proportions of 20 species in total, representing a gradient from mesic to wet grasslands. The resulting communities were exposed to heat, flooding, and contamination in a factorial design in a walk-in ecotron for twelve weeks. Total biomass of the established communities increased for higher proportions of mesic species in the seed mix and under flooding. When analysing functional groups, the effect was driven by the mesic legume Lotus corniculatus and partially by the wet grassland legume Tetragonolobus maritimus . Non-legume forbs and grasses had a low proportion of total biomass, while forbs responded positively to heating, and the ratio between functional groups was more even in wet grassland communities. Few functional differences were seen among species, but SLA showed higher plasticity in T. maritimus than in L. corniculatus. Growth rates of individual plants were higher in communities dominated by wet grassland species, while mesic communities produced more total biomass. This might be beneficial to ecosystem services like microclimate regulation. The results indicate that all tested communities tolerated the urban stressors in their first growing season and thus could replace current species-poor grass seed mixes in infiltration swales. Still, long-term studies are needed to validate these results for more mature communities.