Prediction of Long-Grain Rice Texture and Pasting Properties From Starch and Protein Fractions

The objective of this study was to determine predictive models of rice functionality from protein and starch chemistry data. One long-grain rice variety (Cocodrie) was harvested from two fields in Arkansas. Rice was harvested from 8 locations in each field. Samples were then stored at 21oC/50% RH until equilibrated to a storage moisture content of 12%. Texture properties were measured using a TPA test on ten cooked rice kernels in conjunction with a Texture Analyzer (TAXT2i, Texture Technologies, Scarsdale, NY) while pasting properties were measured with a Rapid Visco-Analyzer. Rice chemical properties measured included amylose and protein contents, starch molecular weight profiles, and protein molecular weight distribution. Rice harvested from a single rice field was quite variable in physicochemical properties. For example, the protein contents of rice flour samples ranged from 6.62 to 8.44% and peak viscosity of the rice flours ranged from 184.7 to 217 RVU. This implies that small changes in water or nutrient availability can have a significant impact on rice quality. It was also established that functional characteristics can be predicted from both starch and protein chemistry. Peak viscosity (Rc=0.98, Rp=0.72) was well predicted by protein and starch data. Overall, data indicate that starch was more important in predicting peak viscosity than were proteins.