Abstract
Atmospheric carbon dioxide enrichment ($eCO_2$) often increases soil nitrous oxide ($N_2O$) emissions, but the underlying mechanisms are not fully understood. Emerging evidence suggests that $eCO_2$ alters plant N preference in favor of ammonium ($NH_4{^+}-N$) over nitrate ($NO_3{^-}-N$). Yet, whether and how this attributes to the enhancement of $N_2O$ emissions has not been investigated. We examined the effects of $eCO_2$ on soil $N_2O$ emissions in the presence of two N forms ($NH_4{^+}-N$ or $NO_3{^-}-N$), using wheat (Triticum aestivum L.) as a model plant. Our results showed that N forms dominated $eCO_2$ effects on plant and microbial N utilization, and thus soil $N_2O$ emissions. Elevated $CO_2$ significantly increased the rate and the sum of $N_2O$ emissions by three to four folds when $NO_3{^-}-N$, but not $NH_4{^+}-N$, was supplied. Enhanced $N_2O$ emission was related to the reduced plant $NO_3{^-}-N$ uptake in wheat. We propose a new conceptual model in which $eCO_2$-inhibition of plant $NO_3{^-}-N$ uptake and/or $CO_2$-enhancement of soil labile C enhances the N and/or C availability for denitrifiers and increases the intensity and/or the duration of $N_2O$ emissions. Together, these findings suggest that to enhance plant N use efficiency and reduce $N_2O$ emission, crop breeding and management need to consider altered plant preference of N sources under future $CO_2$ scenarios.