Precipitation Pattern Regulates Soil Carbon Flux Responses to Nitrogen Addition in a Temperate Forest

Abstract

Changes in precipitation frequency and intensity are predicted to be more intense and frequent accompanying climate change and may have immediate or potentially prolonged effects on soil CO2 and CH4 fluxes. However, how soil CO2 and CH4 fluxes respond to change in precipitation patterns remains poorly understood, especially under nitrogen (N) addition. In this study, we investigated the fluxes of soil CO2 and CH4 during a two-year field experiment and the effects of long-time precipitation reduction (− 30% of through-fall), short-term precipitation pulse and change of snow thickness on their response to simulated N deposition (50 kg N hm−2 y−1). Soil CO2 flux was greater enhanced by N addition in the relatively wet growing season (2016) than in the relatively dry growing season (2017), whereas soil CH4 flux was more inhibited by N addition in the relatively wet growing season 2016 than in the relatively dry growing season 2017. However, reduced precipitation decreased the positive effect of N addition on soil CO2 flux in the relatively wet growing season 2016, while changing the direction of N addition effect in the relatively dry growing season 2017. These results suggested that the precipitation patterns in the growing season may affect soil CO2 and CH4 fluxes response to N addition. Furthermore, we found that short-term extreme precipitation and thickness of snow cover had a significantly positive effect on soil CH4 flux but had a significantly negative effect on soil CO2 flux. The extreme precipitation, that is, very heavy precipitation, reduced the response magnitudes of soil CO2 and CH4 fluxes to reduced precipitation within 3 days, compared to wet precipitation and heavy precipitation. Our results indicated the significant effects of precipitation pattern changes on responses of soil C flux to N deposition, which should be incorporated into the global-C-cycling models to improve the prediction and reduce the uncertainty of C-climate feedbacks.