Ants and plants as indicators of biodiversity, ecosystem services, and conservation value in constructed grasslands

Grasslands are constructed for soil and wildlife conservation in agricultural landscapes across Europe and North America. Constructed grasslands may mitigate habitat loss for grassland-dependent animals and enhance ecosystem services that are important to agriculture. The responses of animal species richness and abundance to grassland habitat quality are often highly variable, however, and monitoring of multiple taxa is often not feasible. We evaluated whether multiple animal taxa responded to variation in constructed grassland habitats of southwest Ohio, USA, in ways that could be predicted from indicators based on quality assessment indices, Simpson diversity, and the species richness of ants and plants. The quality assessment indices included a widely used Floristic Quality Assessment (FQA) index, and a new Ant Quality Assessment (AntQA) index, both based on habitat specificity and species traits. The ant and plant indicators were used as predictor variables in separate general linear models of four target taxa-bees, beetles, butterflies and birds-with response variables of overall species richness and abundance, and subsets of taxa that included the abundance of ecosystem-service providers and grassland-associated species. Plant Simpson diversity was the best-fitting predictor variable in models of overall bee and beetle abundance, and the abundance of bees classified as ecosystem-service (ES) providers. FQA and plant richness were the best predictors of overall butterfly species richness and abundance. Ant species richness was the best predictor of overall bird species richness and abundance as well as the abundance of ES birds, while the AntQA index was the best predictor for the abundance of grassland bird and butterfly species. Thus, plant Simpson diversity and ant species richness were the most effective indicators for complementary components of grassland animal communities, whereas quality assessment indices were less robust as indicators and require more knowledge on the habitat specificity of individual ant and plant species.