Estimating nitrogen and phosphorus losses from lowland paddy rice fields during cropping seasons and its application for life cycle assessment

Estimating nitrogen and phosphorus losses into water bodies is crucial for evaluating the environmental impacts of rice cultivation. In this study, published data (N = 80) were used to estimate the total nitrogen (TN) and total phosphorus (TP) that could be lost through leaching and runoff during cropping seasons, taking into account factors that influence these losses. According to second-order Akaike's Information Criterion (AICc, one of the model selection methods), a model with soil temperature, TN input, inflow water, the incorporation of rice straw residue, puddling, and the use of slow-release fertilizer (P < 0.05) was chosen to estimate TN losses. A model with soil temperature, no application of phosphate fertilizer, medium- and coarse-textured soils (P < 0.05) and inflow water was chosen to estimate TP losses. The root mean square error was reduced by 34.5% for TN and by 21.8% for TP when these models were used compared to when the average proportions of outflowing nutrients were used (e.g., 17.0% for TN and 7.3% for TP in this study). The multiple regression model, taking into account site-specific conditions, was proven to be a readily available method to obtain knowledge of nutrient loading. Once process-based models are readily available, they will improve the estimates of eutrophication potential in life cycle assessments. Further improvement of the method is required, especially for the winter fallow period.