Estimation of Seasonal Net Carbon Sequestration Under Noncontinuous Flooding in Rice Fields

Non-continuous flooding (NCF) in rice was recently reported to improve the field-scale, seasonal carbon balance. In this response article, we clarify our system boundaries (seasonal, field-scale), address the role of yield carbon and microbial indicators, and add sensitivity checks with expanded data. Across these checks, NCF consistently reduces methane without penalizing yield, supporting our original conclusions. Wang and Zou (2025) have raised concerns regarding our recent contribution to Global Change Biology (Hou et al. 2025). In this work, we conducted a meta-analysis of 1075 data pairs from 72 studies worldwide to quantify the effects of noncontinuous flooding (NCF) on greenhouse gas (GHG) equivalent components and its overall net carbon sequestration (NCS) benefits in rice fields. In particular, they note that our conclusion warrants further scrutiny in terms of methodological assumptions and data representativeness, arguing this may lead to a systematic overestimation of the carbon sequestration potential of NCF in rice systems. In response, we address below the key concerns raised in their letter, focusing on three primary aspects: (1) the calculation framework of NCS, (2) the uncertainties of SOC sequestration estimation, and (3) dataset representativeness. We acknowledge the complexity and ongoing debate regarding the definition and accounting framework of NCS in agricultural systems. In our study, NCS refers to the net carbon balance of the rice cropping system during the growing season, encompassing both carbon uptake and release within system boundaries. Under this definition, PCS, calculated from crop yield, is treated as a component of seasonal carbon uptake. While we agree that PCS does not meet the IPCC's criteria for long-term carbon retention (Ogle et al. 2019), it reflects short-term carbon fixation and system productivity, providing a useful basis for comparing the carbon performance of continuous flooding (CF) and NCF irrigation regimes. Our system boundary excludes downstream emissions resulting from grain consumption, which occur outside the field-scale assessment. Although PCS does not imply long-term storage, it represents a temporally meaningful field-scale carbon sink during the cropping period. This treatment aligns with existing studies emphasizing the role of short-to medium-term carbon fluxes in field-scale assessments (Smith et al. 2020). In addition, we clearly separated PCS from ΔSOC (changes of SOC concentration before and after the growing season) to avoid double counting (see figure 7a in Hou et al. (2025)). PCS accounts only for carbon in harvested rice grain, whereas ΔSOC captures changes in soil carbon associated with nonharvested biomass. As shown in figure 7a in Hou et al. (2025), these two are estimated independently. We acknowledge that there are uncertainties associated with short-term SOC sequestration assessments. While long-term SOC monitoring remains the gold standard (Smith et al. 2020), single-season comparisons are widely used in meta-analysis due to field data limitations (Liu et al. 2024). In our dataset, positive SOC sequestration under CF was observed in systems that incorporated straw, organic fertilizer, and/or biochar application (see supporting information in Hou et al. (2025)), all of which enhance single-season SOC sequestration. Additionally, the reported mean SOC sequestration under CF (~2800 kg CO2-eq ha−1) falls within the range of recent field studies under comparable conditions (table 2 in Liu et al. (2024)). Regarding microbial biomass carbon (MBC), we agree that it is not a direct proxy for microbial activity or SOC mineralization. In our study, MBC was interpreted cautiously—as a reflection of shifts in microbial carbon pools under altered moisture and redox conditions, not as evidence of SOC sequestration loss. The co-occurrence of increased MBC and decreased dissolved organic carbon (DOC) under NCF (figure S7 in Hou et al. (2025)) may suggest enhanced microbial processing of labile substrates, consistent with mechanisms linked to drying-rewetting cycles (Blagodatskaya and Kuzyakov 2013). While we agree that process-based measures such as respiration or enzyme activity would offer stronger support (Blagodatskaya and Kuzyakov 2013), such data were rarely available in the included studies. As discussed in section 4.2 of Hou et al. (2025), we did not treat MBC as a conclusive indicator but rather as part of a broader pattern supporting our interpretation of SOC dynamics. We acknowledge the limitations posed by sample size and dataset completeness. Our study focused on literature published in the recent 5-year period (2019–2023) to guarantee comparability of irrigation practices, as standardized NCF techniques (e.g., alternating wet and dry, AWD, devices, controlled irrigation, CI) have only recently seen broader adoption (Bo et al. 2022). Earlier studies often lacked precise water management descriptions or consistent measurement protocols, which may affect data quality and classification accuracy. These constraints were noted in section 4.4 of Hou et al. (2025). The robustness of the conclusions of Hou et al. (2025) was demonstrated after extensive sensitivity analyses, publication bias tests, and comparisons with previous studies (Bo et al. 2022; Jiang et al. 2019) (Table 1). To address concerns about the temporal coverage, we expanded our analysis to include pre-2019 studies from Bo et al. (2022), yielding a broader dataset for sensitivity checks (Figure 1). While earlier data slightly changed the distribution of the effects of NCF on CH4 emissions and yield, the overall pattern remained consistent: NCF significantly reduces CH4 emissions and insignificantly enhances yield. The larger mitigation effects based on our dataset may also reflect improved irrigation techniques and cultivar advancement in recent years. We appreciate the additional effort by Wang and Zou (2025) in reconciling the data from the Nikolaisen et al. (2023). However, after re-screening 2301 records from Nikolaisen et al. (2023) dataset using our criteria, we found that only 247 observation pairs were eligible, among which 62 overlapped with our post-2019 dataset. This overlap highlights the risk of double-counting when combining datasets without deduplication, potentially inflating sample size and biasing effect estimates. It further emphasizes the importance of carefully tracking study-level duplication in meta-analytic integration. Moreover, the irrigation regimes in the Nikolaisen et al.'s (2023) dataset lacked standardization, complicating the classification of irrigation practices. In contrast, our meta-analysis applied consistent screening at the study level to ensure data traceability and comparability. We emphasize that collecting all the components of NCS remains challenging (section 4.4). Although complete NCS observations remain scarce (n = 23), we conducted component-level meta-analysis for yield, SOC, and non-CO2 GHGs. Importantly, CH4 reduction alone (−4314 kg CO2-eq ha−1) exceeds the combined increases in N2O emissions (+137 kg) and SOC loss (−2856 kg), supporting a robust NCS benefit of NCF. We appreciate the constructive efforts by Jinyang Wang and Jianwen Zou in extending the dataset and for this interesting debate. Yu Hou: formal analysis, methodology, visualization, writing – review and editing. Jingwen Zhang: conceptualization, funding acquisition, methodology, project administration, resources, writing – review and editing. Junjie Guo: conceptualization, methodology, writing – review and editing. Kairong Lin: conceptualization, writing – review and editing. Wang Zhou: conceptualization, writing – review and editing. Ziqi Qin: conceptualization, writing – review and editing. Qingsong Zhu: data curation. Qinxia He: data curation. We acknowledge the support from the National Key R&D Program of China (2023YFC3209400, 2023YFC3209401-03), the National Natural Science Foundation of China (52479034, 52309014), the Fundamental Research Funds for the Central Universities, Sun Yat-sen University (23hytd011), and the Guangdong Basic and Applied Basic Research Foundation (2024A1515010968). The authors declare no conflicts of interest. This article is a Response to a Letter to the Editor by jingyang Wang and Jianwen Zou https://doi.org/10.1111/gcb.70465 regarding Hou et al., https://doi.org/10.1111/gcb.70283. The data used to create Figure 1 in this response come from two publicly archived sources: Hou et al. (2025) dataset [https://doi.org/10.6084/m9.figshare.29117669] and Bo et al. (2022) dataset [https://doi.org/10.6084/m9.figshare.19164893].