The evolution of ecosystem multi-functionality
The emergent functioning of ecosystems can be quantified across a multitude of functions, from biomass production, to nutrient cycling, to supporting other trophic levels. Through 20 years of research it is well understood that diversity strongly influences ecosystem function in plant assemblages and this relationship strengthens as we increase the number of functions considered. How species contribute to function is, in part, a product of past evolution, and recent work has shown that the evolutionary relationships among species provide strong explanatory power for productivity –one measure of function. The mechanism for this pattern is that distantly related species have a greater propensity to utilize differing resources, thus turning more of the available resources into biomass. However, it cannot be assumed that this complementarity argument holds for other functions. Some functions (e.g., nutrient cycling and pollinator support) may be maximized in highly functioning lineages. Thus I expect that individual ecosystem functions will be maximized in either assemblages with high phylogenetic diversity (evolutionary complementarity) or assemblages that contain species from specific lineages (evolutionary selection). However, if functions that are consistent with evolutionary selection are maximized in different lineages, then multifunctionality can only be maximized when diverse evolutionary lineages are present. I test this prediction using a biodiversity-ecosystem function experiment where the phylogenetic diversity was manipulated in 96 plots. Functions measured included soil nitrogen and carbon, pollinator diversity and abundance, light interception, plant biomass, litter decomposition, pathogen damage, and flower abundance and timing.
Some functions (e.g., biomass, light interception, 1/pathogen damage) were maximized in plots with high phylogenetic diversity, while other functions (e.g., soil nitrogen, decomposition and pollinator abundance) were maximized in plots where certain lineages were present –for example, nitrogen availability was highest in plots with Fabaceae present, and pollinator abundance was highest in plots with Lamiaceae. For these functions there is a phylogenetic signal in species contributions to function, and phylogenetic diversity does not provide the best explanation. However, because function is maximized by different lineages, measures of multifunctionality are highest when phylogenetic diversity is maximized. Thus, maximizing ecosystem multifunctionality requires the presence of diverse plant lineages, arguing that conservation and habitat restoration explicitly consider species relatedness.