Enhancing landscape planning: Vegetation-mediated ecosystem services predicted by plant traits

Abstract

Although well recognised in ecology, the concept of ecosystem services is still not widely applied in practical environmental planning. Environmental planning often has to evaluate whole landscapes in a spatially explicit way, including marginal landscape elements for which data on provisional, regulatory, and cultural ecosystem properties and services are not readily available or are costly to provide. On the other hand, many planning schemes include a habitat- or vegetation survey, raising the question whether ecosystem services can be predicted from vegetation properties alone. I used a large data set of herbaceous plant communities collected across Germany, The Netherlands, and Denmark, as well as additional data, to predict (i) the probability of obtaining a high marginal income from forage production, (ii) the probability of a high conservation value based on the occurrence of regionally rare plants, and (iii) carbon stocks in mineral soils (SOC). The predictions are based on 22 plant traits of 123 common species of reeds and of dry, mesic, wet, and ruderal grasslands. The individual traits were aggregated to trait modules representing major biological functions that affect the provision of services. High marginal income from forage production was predicted for plant communities functionally composed of intermediate levels of above-ground size and allocation, and an acquisitive leaf economy on the one hand, but little lateral expansion, generative reproduction and root mass on the other. The probability of finding regionally rare plants increased in communities with small-to-intermediate above-ground size, a conservative leaf economy, low C:N ratios, few and large seeds, and little lateral expansion. Communities displaying pronounced lateral expansion and above-ground size were linked to high soil organic-carbon stocks. The models predict trade-offs between these ES that are relevant to environmental planning. Where income from forage production is high, regionally rare plants will not be found, and SOC will be low. Where SOC is high, income from forage production will be low and rare plants infrequent. Where many rare plants are found, neither income from forage production nor SOC will be high. Using the predictive functions and the trait data provided in the Appendix S3, users can evaluate the supply of the three services for many perennial herbaceous communities in northwestern Europe.