The ability to design plant communities to optimize particular ecosystem functions and thus more effectively provide ecosystem services would improve all types of ecosystem management, including agriculture. We propose a novel quantitative multi-step method for selecting mixtures of plant species to meet ecosystem objectives: collecting trait data; identifying suites of traits related to relevant community processes; relating those processes to the ecosystem functions of interest; optimizing the planted mixture; and evaluating trade-offs. This approach was tested using planted mixtures of 14 grassland species common in managed pastures of the northeastern United States. Ordination of trait data from greenhouse and small plot studies was used to relate species traits to community processes, followed by generalized additive modeling to relate community-weighted mean process scores to field estimates of biomass production and resistance to invasion. Multi-criteria optimization identified seed mixtures that maximized or minimized selected ecosystem functions, facilitating quantification of trade-offs in processual ability within species and between ecosystem functions at the community level. In this study, predicted seasonal and annual biomass values were within expected ranges, but the optimal mixtures differed from forage mixtures traditionally planted in the northeastern United States. Perennial ryegrass was the most commonly-selected grass. Legumes were represented less than expected, possibly due to attributes such as longevity and forage quality that are important in pastures but were not directly included in the optimization. Ecosystem function was not linearly related to species richness. While further research on quantitatively relating species traits and community processes is needed, multi-criteria optimization offers a new path for linking ecosystem function to ecosystem management.
All Science Journal Classification (ASJC) codes
- Ecology, Evolution, Behavior and Systematics