Seasonal N2O emissions respond differently to environmental and microbial factors after fertilization in wheat–maize agroecosystem

Biogeochemical processes regulating cropland soil nitrous oxide (N2O) emissions are complex, and the controlling factors need to be better understood, especially for seasonal variation after fertilization. Seasonal patterns of N2O emissions and abundances of archaeal ammonia monooxygenase (amoA), bacterial amoA, nitrate reductase (narG), nitrite reductase (nirS/nirK), and nitrous oxide reductase (nosZ) genes in long-term fertilized wheat–maize soils have been studied to understand the roles of microbes in N2O emissions. The results showed that fertilization greatly stimulated N2O emission with higher values in pig manure-treated soil (OM, 2.88 kg N ha−1 year−1) than in straw-returned (CRNPK, 0.79 kg N ha−1 year−1) and mineral fertilizer-treated (NPK, 0.90 kg N ha−1 year−1) soils. Most (52.2–88.9%) cumulative N2O emissions occurred within 3 weeks after fertilization. Meanwhile, N2O emissions within 3 weeks after fertilization showed a positive correlation with narG gene copy number and a negative correlation with soil NO3− contents. The abundances of narG and nosZ genes had larger direct effects (1.06) than ammonium oxidizers (0.42) on N2O emissions according to partial least squares path modeling. Stepwise multiple regression also showed that log narG was a predictor variable for N2O emissions. This study suggested that denitrification was the major process responsible for N2O emissions within 3 weeks after fertilization. During the remaining period of crop growth, insufficient N substrate and low temperature became the primary limiting factors for N2O emission according to the results of the regression models.