Analysis by Applying Near-Infrared-Refractography and Statistical Modeling of Sweetpotato Weight Loss, Density and Quality Change Due to Long-Term Storage, Temperature and Water Stress

'Beauregard', 'Covington', 'Evangeline', 'Hatteras' and 'Carolina Rose' sweetpotato (Ipomoea batatas (L) Lam) roots graded U.S. No. 1 were stored in environmentally controlled rooms for 300 days. All the sweetpotato roots used in this study were grown and stored in eastern North Carolina. The effect of relative humidity on root weight loss was experimentally tested by holding temperature constant at 14.1°C +/-0.1°C, and sequentially changing relative humidity (RH) from 85%, to 75% and 65%. Each RH level was held constant for a period of 30 days; which required, a cycle of 90 days to complete one set of treatments. A total of three cycles and an additional period at 85% RH were necessary to complete 300 days of storage. Inside a second room variable temperature conditions were performed, with each temperature cycle lasting 90. Cycles were divided into five periods of 15 days each. At each period temperature was held at constant in the following order: 14.4°C, 17°C, 14.4°C, 19°C, 14.4°C. Simultaneously, 'Covington' sweetpotatoes were stored in commercial storage rooms, with the temperature and relative humidity held constant at 14.4°C +/-1°C and 85% +/- 5%, respectively, for the duration of all tests. Temperature, relative humidity and weight of sweetpotatoes were measured every hour for the entire duration of the study by a data acquisition system designed and built for this specific application. Measuring density during long-term storage of 'Covington' sweetpotato under both variable temperature, and stable commercial storage conditions, was used to characterize the density change over time and varying environmental conditions. Density was also measured on 'Beauregard', 'Carolina Rose', 'Evangeline', and 'Hatteras' sweetpotatoes varieties stored under variable temperature conditions. In addition to density, other parameters such as dry matter, glucose, sucrose, fructose, maltose, Brix degrees, starch, beta-carotene, phenolic, vitamin C and fat in chips after fry-test were analyzed. The nutritional values during the first year were measured by chemical analysis and high pressure liquid chromatography (HPLC) and during the second year by near infrared spectroscopy (NIR) with a prediction curve based on the first year results. Density was measured by simple weight and volume displacement. Advanced statistical modeling was applied to develop mixed effects statistical models to predict weight loss of the five different varieties of sweetpotatoes during long-term storage as a result of temperature and water stresses. The models were developed using the Statistical Analysis System, SAS, version 9.3 for Windows. Linear mixed models were fit using the MIXED procedure.;Relative humidity treatments and variety significantly (P<0.05) affected weight loss of sweetpotatoes. Weight loss rates were lower at high RH because of low vapor pressure deficit (VPD), and increased as RH decreased and VPD increased. 'Beauregard' and 'Covington' had the lowest weight loss rate at any RH treatment when compared to 'Evangeline', 'Hatteras' and 'Carolina Rose'. Different levels of iii environmental RH during long-term storage of sweetpotatoes effected weight loss at different rates; moreover, data showed that industry standard storage conditions of 14°C/85%RH could be improved by increasing RH to 90%. Under variable temperature storage conditions, 'Covington' and 'Beauregard' experience the lowest rates of weight loss relative to the others. 'Carolina Rose' and 'Hatteras' experience the greatest rates of weight loss. The storage conditions significantly (p<0.05) affected the 'Covington' variety density and all nutritional characteristics analyzed in this study. The reduction in density over time in 'Covington' sweetpotatoes constitutes a benchmark standard to assess the storage length and conditions. Consequently, density better serves to describe the change in starch content and fat percentage in fried chips for sweetpotatoes; although, sugars, starch, beta-carotene, phenolic and vitamin C content could also be correlated to density. In the statistical modeling analysis, temperature, RH and variety had significant (P<0.05) effects on estimated values. Statistical models that predict sweetpotato weight loss during long-term storage based on environmental conditions could be used to forecast shrinkage of roots and plan business decisions as well as supply requirements from field production or packing houses.