Rice is an important cereal grain crop, providing 20% of calories consumed worldwide. The crop's unique physiology and the environments it is grown in make it susceptible to arsenic uptake and accumulation within its grain, which is elevated compared to other foods. This is due to the seasonal flooding of rice paddies, which create reducing conditions of iron oxides, thereby freeing arsenic into the environment, subject to uptake. The spatial and temporal variations in soil chemistry and hydrology render arsenic levels in grain highly variable and challenging to predict. Arsenic is a naturally occurring metalloid in most soils, and is toxic to both the rice crop as well as the humans that consume it. To understand the environmental factors that govern the accumulation of arsenic in rice grains, we have sampled rice grains, rice leaves, soil, and water from approximately 200 rice paddies across a varied geographical range within Cambodia, a country which relies on rice for over 60% of its annual caloric intake. The traditional farming practices prevalent in the region allow us to isolate and inspect various environmental factors efficiently. Arsenic concentrations within the rice grains and soil were measured using ICP-MS, XRF and soil visible reflectance measurements were made using a visible reflectance spectrophotometer. Here, we develop a conceptual framework linking soil mineralogy, soil chemistry, soil physical properties, and hydrology into arsenic uptake in rice. We find that visible reflectance spectroscopy is correlated with soil iron and arsenic levels, as well as grain arsenic concentrations. We then use our results from soil spectroscopy to characterize soil mineralogy/composition and to identify the potential for arsenic mobilization and uptake into rice for each of our studied paddies. We then connect our characterization of arsenic mobilization and uptake potential to observed hydrologic conditions in order to explain the observed variation in rice grain arsenic concentrations. Our research suggests that the reflectance spectroscopy of soil within the visible light spectrum is a useful tool for forecasting rice grain quality prior to planting the crop, and can have prospective implications for increased food security in Cambodia and other regions where rice is a staple.
Acknowledgements: This research was supported by NSF SiTS grant #2226648 and a NASA CSDA grant. The laboratory analysis was also made possible through NSF MRIs (2018836 and 1229258).