A framework to identify indicator species for ecosystem services in agricultural landscapes

Abstract

Improving our understanding of the relationships between biodiversity and the delivery of ecosystem services is crucial for the development of sustainable agriculture. We introduce a novel framework that is based on the identification of indicator species for single or multiple ecosystem services across taxonomic groups based on indicator species analyses. We utilize multi-species community data (unlike previous single species approaches) without giving up information about the identity of species in our framework (unlike previous species richness approaches). We compiled a comprehensive community dataset including abundances of 683 invertebrate, vertebrate and plant species to identify indicator species that were either positively or negatively related to biological control, diversity of red-listed species or crop yield in agricultural landscapes in southern Sweden. Our results demonstrate that some taxonomic groups include significantly higher percentages of indicator species for these ecosystem services. Spider communities for example included a higher percentage of significant positive indicator species for biological control than ground or rove beetle communities. Bundles of indicator species for the analysed ecosystem service potentials usually included species that could be linked to the respective ecosystem service based on their functional role in local communities. Several of these species are conspicuous enough to be monitored by trained amateurs and could be used in bundles that are either crucial for the provision of individual ecosystem services or indicate agricultural landscapes with high value for red-listed species or crop yields. The use of bundles of characteristic indicator species for the simultaneous assessment of ecosystem services may reduce the amount of labour, time and cost in future assessments. In addition, future analysis using our framework in other ecosystems or with other subsets of ecosystem services and taxonomic groups will improve our understanding of service-providing species in local communities. In any case, expert knowledge is needed to select species from the identified subsets of significant indicator species and these species should be validated by existing data or additional sampling prior to being used for ecosystem service monitoring.