Biodiversities and ecosystem multifunctionality

Background/Question/Methods

As worldwide biodiversity losses accelerate, a growing body of experiments has documented the consequences of species losses to ecosystem functions and services. Syntheses of these studies show that local species richness (alpha diversity, α) positively affects many ecosystem functions and that these effects become stronger as more functions (ecosystem multifunctionality) are considered simultaneously. However, the small scale and limited number of functions measured in biodiversity experiments do not match the scales at which society usually manages biodiversity or ecosystem functions and services. Previous analyses of plot-scale ecosystem multifunctionality show that the species richness and composition of high-functioning assemblages vary strongly depending on the functions of interest, suggesting that multifunctional landscapes could require a diverse landscape mosaic of different communities (beta diversity, β) or a diverse regional species pool (gamma diversity, γ) as well as high alpha diversity. However, the relative effects of these diversities have never been measured for single or multiple functions. We analyzed data from the longest-running biodiversity-ecosystem functioning manipulation to date at Minnesota’s Cedar Creek Ecosystem Science Reserve to examine the effects of alpha, beta, and gamma diversity on the provision of up to seven ecosystem functions in five different years. We compared the capacities of simulated landscapes comprised of the experiment’s 168 perennial grassland plant communities to provide both single and multiple ecosystem functions.

Results/Conclusions

Using a multifunctionality index (MF) based on the scaled mean minus the standard deviation of all functions, we found that while mean alpha diversity is usually the most important contributor to landscape-level multifunctionality, beta and gamma diversity also make significant contributions. Across all years and functions, analysis of standardized regression weights reveals that all three diversities have significant and independent positive effects on MF (MF = 0.46α + 0.11β + 0.16γ; p<.001, R²=0.32, N=7512 landscapes). Using a thresholds approach to measure multifunctionality, we also found that each diversity metric influences the number of functions achieved above each threshold. Biodiversities explain far less variation in most single functions (mean R² = 0.13±.04) than in multifunctionality (R² = 0.32). Model selection by AIC reveals that mean alpha diversity significantly affects every individual function while beta and gamma diversities become important only when all functions are considered together.  Our findings emphasize that multiple functions must be considered together to capture the full contributions of biodiversity and illustrate the contributions of biodiversities both at and above the patch scale to ecosystem multifunctionality.