Responses of nitrogen concentrations and pools to multiple environmental change drivers:A meta-analysis across terrestrial ecosystems

Aim: We sought to understand how the individual and combined effects of multiple environmental change drivers differentially influence terrestrial nitrogen (N) concentrations and N pools and whether the interactive effects of these drivers are mainly antagonistic, synergistic or additive. Location: Worldwide. Time period: Contemporary. Major taxa studied: Plants, soil, and soil microbes in terrestrial ecosystems. Methods: We synthesized data from manipulative field studies from 758 published articles to estimate the individual, combined and interactive effects of key environmental change drivers (elevated CO 2 , warming, N addition, phosphorus addition, increased rainfall and drought) on plant, soil, and soil microbe N concentrations and pools using meta-analyses. We assessed the influences of moderator variables on these effects through structural equation modelling. Results: We found that (a) N concentrations and N pools were significantly affected by the individual and combined effects of multiple drivers, with N addition (either alone or in combination with another driver) showing the strongest positive effects; (b) the individual and combined effects of these drivers differed significantly between N concentrations and N pools in plants, but seldom in soils and microbes; (c) additive effects of driver pairs on N concentrations and pools were much more common than synergistic or antagonistic effects across plants, soils and microbes; and (d) environmental and experimental factors were important moderators of the individual, combined and interactive effects of these drivers on terrestrial N. Main conclusions: Our results indicate that terrestrial N concentrations and N pools, especially those of plants, can be significantly affected by the individual and combined effects of environmental change drivers, with the interactive effects of these drivers being mostly additive. Our findings are important because they contribute to the development of models