DisplayTitle

Micropycnometer Measurement of Single-Kernel Density of Healthy, Sprouted, and Scab-Damaged Wheats¹

¹Contribution No. 97-386-J of the Kansas Agricultural Experiment Station. ²Research engineer, U.S. Grain Marketing and Production Research Center; assistant professor, Department of Grain Science and Industry; assistant professor, Department of Statistics; and graduate student, Department of Industrial Engineering, Kansas State University, Manhattan, KS. ³Corresponding author. Email: tjh@wheat.ksu.edu

Samples from four market lots of hard red winter and soft red winter wheat containing sprout- and scab-damaged kernels were used to test a prototype single-kernel density micropycnometer. Fifteen kernels for each damage type and an equal number of healthy kernels were weighed to the nearest 0.01 mg, then measured for volume to the nearest 1.0 μL. Volume measurements for all kernels were performed three consecutive times with the micropycnometer, then kernels were evaluated for weight, size, moisture, and hardness using a Single Kernel Characterization System. The structure of the sampling plan and the goals of the study indicated that a mixed-model statistical analysis was needed. The fixed effects were wheat class and type of kernel, and the random effects included lot, the interaction of lot with class and kernel type, kernels within each lot, and repeated measures of single-kernel density. Results indicated that variability of the three measurements per kernel did not depend on type of kernel or class of wheat. The standard deviation for repeated density measurements was 0.0029 g/cm³. Kernel-to-kernel variability changed depending on the type of kernel; healthy and sprout-damaged kernels showed similar variability in density, whereas scab damaged kernels had a variance about four to five times higher. Type of kernel significantly affected mean density; healthy kernels averaged 1.28 g/cm³, sprout-damaged kernels averaged 1.19 g/cm³, and scab-damaged kernels averaged 1.08 g/cm³. Wheat class did not exert a significant influence on singlekernel density. Attempts to predict single-kernel density using kernel weight, size, moisture, and hardness found no relationships of practical importance.