Abstract
Rice (Oryza sativa L.) is a major dietary source of arsenic (As) for humans. Understanding the mechanisms of As accumulation in rice is essential for mitigating human exposure. However, the effects of environmental factors on As accumulation in rice have been insufficiently quantified under field conditions. To address these issues, we modeled temporal dynamics of As accumulation in rice plants and grains using a Bayesian state-space model (SSM). In this SSM, As concentrations in flag leaves and rice grains were treated as response variables, whereas four environmental factors, i.e., number of flooding days, temperature, crop evapotranspiration and precipitation, served as explanatory variables. The nonlinear effects of physiological factors were also incorporated into the SSM. The results indicated that among the four environmental factors, flooding days exerted the greatest positive effect on As accumulation in rice plants, with the effect peaking 5-10 days after heading. High temperatures and increased crop evapotranspiration promoted As accumulation, whereas increased precipitation reduced As accumulation. This work is among the first studies to quantify the effects of environmental factors on As accumulation in rice under field conditions, and the findings contribute to the development of region-specific cultivation guidelines for mitigating As exposure through rice.