Soil nitrogen (N) is one major limiting factor that constrains terrestrial ecosystem productivity and its responses to atmospheric CO2 enrichment (eCO2) and climate warming will, therefore, critically modulate long-term ecosystem feedbacks to climate change. To date, eCO2 or warming effects on N pools and fluxes have been examined largely in single-factor experiments, and it remains unknown whether the results obtained are indicative to the interactive impact of concurrent eCO2 and warming under future climate change scenarios. We re-analyzed the data of 176 publications that quantified the effects of eCO2, warming, and/or their combination on N pools and fluxes to examine whether any unifying patterns exist. We examined the directions and magnitude of eCO2 and warming effects on N pools and fluxes in both single-factor and two-factor experiments, as well as the impacts of plant growth stages, experimental duration and experimental conditions (pot vs. field).
Results/Conclusions
eCO2 or warming alone stimulates N cycling and significantly increases soil NO3- and the emission of potent greenhouse gas N2O. When both factors occur in concert, most of soil N pools and fluxes were not significantly affected. Moreover, eCO2 or warming alone did not significantly affect soil NH4+ but their combined effect was highly significant, particularly in early- to mid- growing season, suggesting an interactive plant-microbial stimulation of mineralization. More surprisingly, eCO2-enhancement of soil NO3- only occurred during the active growing period (middle growing season) when plant N demand is usually high. We propose a unifying framework in which eCO2 increases soil NO3- by reducing plant NO3- uptake and assimilation but warming mitigates this eCO2 effect, reconciling the discrepancy between single- and two-factor experiments. Together, these findings indicate that the emergent effect of eCO2 and warming on plant NO3- acquisition and assimilation is a major determinant of soil N pools and fluxes, and suggest that terrestrial models need to incorporate this effect to predict ecosystem C balance and ecosystem feedbacks to future climate conditions.