Contrasting rice-dryland crop rotation regimes drive divergent greenhouse gas emissions via labile organic carbon modulation, with trade-offs for crop yield

Context Optimizing crop rotations is essential for synergistically increasing yield and reducing environmental impacts. However, in rice-dryland crop rotation systems, the underlying mechanisms through which different dry-season crops influence subsequent rice yield and greenhouse gas (GHG) emissions remain to be elucidated. Objective This study aims to clarify the mechanisms by which dry-season crops influence the productivity and GHG emissions of the subsequent rice season, to identify agronomic strategies that balance grain yield and environmental effects. Methods Over three consecutive years (2021–2023), a field experiment was conducted to assess how different dry-season crops affected soil properties, rice productivity, and GHG emissions in the subsequent rice season. Results Four rotation treatments were compared: fallow-rice rotation (FR), wheat-rice rotation (W R ), garlic-rice rotation (GR), and rapeseed-rice rotation (RR). Compared with the FR system, all rice-dryland crop rotation treatments enhanced rice yield by 10.74%–16.97% but also elevated global warming potential (GWP) by 13.96%–35.88%. Among the three rice-dryland crop rotation treatments, W R system exhibited the lowest rice yield (11.00 Mg ha⁻¹), GWP (3247.11 kg CO 2 -eq ha −1 ), and greenhouse gas intensity (GHGI, 0.29 kg CO 2 -eq kg −1 ). In contrast, the GR treatment achieved the highest rice yield (11.80 Mg ha −1 ), GWP (4009.53 kg CO 2 -eq ha −1 ), and GHGI (0.34 kg CO 2 -eq kg −1 ). The RR treatment showed intermediate values, with a yield of 11.77Mg ha −1 , GWP of 3722.73 kg CO 2 -eq ha −1 , and GHGI of 0.32. Methane (CH 4 ) was the predominant GHG, accounting for 95.80% of total GWP. Conclusions The increased labile organic carbon (LOC) in topsoil under GR and RR treatments stimulated methanogenic activity by providing abundant substrates, thereby significantly enhancing CH 4 emissions. In contrast, the W R treatment maintained significantly lower levels of labile organic carbon, effectively suppressing CH 4 emissions. Moreover, the higher nutrient content and organic matter in GR and RR treatments enhanced rice productivity. Implications These findings elucidate the underlying mechanisms through which different dry-season crops modulate soil organic carbon fractions, thereby influencing subsequent rice yield and GHG emissions. Specifically, the W R treatment effectively reduced GHG emissions but at the expense of yield, underscoring a trade-off between environmental sustainability and agricultural productivity.