LAUSR

Abstract Rice paddy fields are one of the world’s largest anthropogenic sources of methane (CH4). With the largest area of rice paddy fields, China is experiencing a rapid shift from conventional seedling-transplanted rice (TPR) to direct-seeded rice (DSR) due to efforts to introduce labor-saving practices. However, the potential effect of this change on agricultural ecosystem CH4 flux (FCH4) are less studied and remain poorly understood. Here, we analyze FCH4 measured with the eddy covariance technique over a rice paddy where TPR was applied in 2013 and DSR in 2016. Meteorological conditions (i.e., friction velocity, radiation and temperature) between the two growing seasons were similar. However, compared to the TPR system, cumulative CH4 emissions in the DSR system were 25% higher (610.5 ± 73.3 vs 488.8 ± 56.2 kg CH4 ha−1). The increase in CH4 emissions mainly occurred during the flooding periods (i.e. DOY 173–203 and 222–260). After eliminating the effect of differences in weather conditions and water management practices between the TPR and DSR systems, daily CH4 emissions in the DSR system remained significantly higher than in the TPR system. Gross ecosystem productivity (GEP) and rice density were higher in the DSR system than in the TPR system. Cross correlation and wavelet coherence analyses showed that FCH4 were significantly correlated to GEP. Thus, increased CH4 emissions in the DSR system are most likely due to greater GEP, which was associated with higher rice plant density. With the rapid development of DSR, a scientifically sound reduced seeding density could be a promising strategy to reduce CH4 emissions.