| dc.description.abstract | Rice cultivation represents a major anthropogenic source of methane (CH4) due to anaerobic soil conditions under continuous flooding. In Indonesia, mitigation of CH4 emissions must align with sustained rice productivity. This study evaluated CH4 emissions and yield performance under four cultivation practices: Conventional-Organic (OC), Conventional-Inorganic (IC), SRI-Organic (OS), and SRI-Inorganic (IS). Methane fluxes were quantified using the closed-chamber method and gas chromatography, while soil moisture, temperature, electrical conductivity, redox potential, and pH were monitored. An Artificial Neural Network (ANN) model was applied to predict CH4 flux, followed by polynomial curve fitting to interpret nonlinear responses along a controllable environmental gradient.
Seasonal CH4 emissions were highest under OC (83.91 ton CO2-eq ha?¹ season?¹), followed by IC (70.43) and OS (56.67), with the lowest emissions under IS (41.68), corresponding to an approximate 50% reduction relative to OC. Grain yield reached a maximum under OS (12.05 ton ha?¹ season?¹), while IS maintained stable productivity (11.31 ton ha?¹) alongside substantially lower emissions. Linear correlation and regression analyses showed no statistically significant relationships between individual soil parameters and CH4 emissions (p > 0.05). Redox potential displayed the strongest association with other variables (r = 0.215; p = 0.051), reflecting a meaningful but non-independent role in CH4 dynamics. Soil moisture was selected for post-model evaluation as a manageable proxy for redox-controlled processes.
The ANN model achieved strong predictive performance across treatments (R² = 0.86–0.99). Polynomial analysis of ANN outputs identified a nonlinear CH4 response to soil moisture, with a stationary inflection at ? ˜ 0.33 and a minimum emission range around ? ˜ 0.37–0.38. These findings demonstrate that regulated irrigation under SRI, particularly with inorganic fertilization, supports effective CH4 mitigation while sustaining rice productivity under tropical conditions. | |
