Assessing available silicon impact on rice yield and quality through on‐farm experimentations

Abstract Silicon (Si) is recognized as one of the beneficial nutrients to enhance rice ( Oryza sativa L.) growth, yield, and quality. However, the practical knowledge of the needs for Si fertilization in real‐world farming remains unclear. Farmer‐centered experiments focusing on soil‐available Si are essential to improve productivity and profitability. This study aimed to explore factors affecting rice yield and quality using the Bayesian approach along with the economic feasibility of Si fertilization. On‐farm experimentation was conducted over 3 years (2020–2022) across 169 fields with varying nitrogen (N) fertilization rates (48–75 kg N ha −1 ) in Gero City, Japan. The yield and protein content of rice were measured using commercial yield monitors, and soil properties such as mineralizable N and available Si were chemically analyzed. A Bayesian linear regression model was employed to explore the impact of multiple factors on rice yield and quality. The results revealed substantial field‐level yield variability, with expected deviation ranging from −194 to 742 kg ha −1 , and seasonal differences, with the highest yield observed in the third year. However, high N applications reduced rice grain quality by increasing protein content by an average of 0.34% without increasing yield. Furthermore, single‐season Si fertilization showed a consistently low probability of profit, even under favorable conditions. Therefore, it might be necessary to explore the long‐term effect of Si multi‐seasonal applications on yield and profitability. The Bayesian approach provided a probabilistic evaluation of yield variability and economic feasibility, offering valuable insights for optimizing Si fertilization strategies in rice production.