Seasonal variations and driving mechanisms of CO2 fluxes over a winter-wheat and summer-maize rotation cropland in the North China plain

Abstract

Characterization of crop CO2 fluxes in different growing periods and their corresponding responses to abiotic and biotic factors is essential for the adaption of cropland ecosystem carbon balance to future climate change and management practices. In this study, based on long-term (2003–2018) monitoring of CO2 fluxes, meteorological data and leaf area index (LAI) of a winter wheat-summer maize rotation cropland in the North China Plain, the relationship between CO2 fluxes of different growing periods and the driving factors (i.e., abiotic and biotic factors) was explored to identify the important affecting factors. The Random Forest (RF) algorithm revealed that LAI was the most important driving factor of CO2 fluxes in the wheat season, except for WGPW (from sowing to maturity). Net ecosystem productivity (NEP) had significant positive relationships with LAI, photosynthetically active radiation (PAR) and soil water content (SWC) in four different growing periods. Air temperature (Ta) was positively correlated with ecosystem respiration (ER), and LAI and PAR were positively correlated with gross primary productivity (GPP). In the maize season, the RF algorithm revealed that LAI, PAR and SWC were important drivers of CO2 fluxes. NEP and GPP had significant negative relationships with precipitation (PPT) in RGPM (from tasseling to maturity) and WGPM (from sowing to maturity). Ta was positively correlated with ER, and SWC was positively correlated with NEP and GPP except for VGP2M (from jointing to tasseling). For annual fluxes, the interannual variability (IAV) of NEP, ER and GPP were mainly attributed to IAV of the maximum daily NEP (NEPmax), maximum daily ER (ERmax) and maximum daily GPP (GPPmax), respectively. Our findings provide deep insight into CO2 flux characteristics of wheat-maize rotation systems in different growing periods and their influencing mechanisms, which are crucial for predicting the carbon cycle of cropland ecosystems under climate change.