Functional traits have been proposed as a “holy grail” to connect community composition and ecosystem function across scales and time, especially in the face of multiple global change drivers. Much progress has been made on understanding how traits respond to environmental changes and impact function, but it is still not clear how environmental context and interspecific interactions can modify the aggregate trait composition of the community (i.e. the percent nitrogen of all plant tissues combined). Without knowing the magnitude that species interactions or environmental context alter the community’s functional composition, we will not be able to make robust predictions about future community composition or function. Here, we examine how environmental context and interspecific interactions impact the aggregate community trait value. Specifically, we quantify how plant tissue nitrogen, measured on the aboveground biomass of all plant species in a local community, is affected by elevated CO2 and nitrogen fertilization in both monocultures and mixtures of 16 grassland plant species, utilizing data from a 20-year experiment (BioCON) at Cedar Creek Ecosystem Science Reserve, Minnesota, USA. We analyzed our data using Generalized Diversity-Interaction models which allow us to determine how species identity, species interactions, and environmental context affect aboveground tissue nitrogen.
Results/Conclusions
Both species interactions and abiotic environmental context modified community tissue percent nitrogen. The best model included species identity effects, diversity effects, diversity effects modified by environmental context, and year-to-year variation. Specifically, nitrogen addition increased community tissue nitrogen whereas elevated CO2 had little impact. Further, interspecific interactions decreased community tissue nitrogen by almost double the amount that nitrogen addition increased these values. The year-to-year variation ranged from almost no effect to about the same effect as increasing richness and was not readily explained by yearly temperature differences, precipitation differences, or biomass production. Overall, our results suggest that it is necessary to consider both interspecific interactions and environmental context when estimating community trait values. Without considering both of these aspects we may be over- or under-estimating community trait values when we use community weighted means to characterize communities from average traits collected in monocultures and under different environmental conditions.