Development of a rapid screening method for improved breeder popcorn lines
Abstract
For microwaveable popcorn, quality is of great importance. Quality can be improved by developing popcorn breeding lines. However, breeding programs have a small amount of material and need nondestructive analytical methods. The objective was to develop a nondestructive method for selection of breeding material. The following baseline information was determined for 18 popcorn lines: number of kernels per 10g, density, Stenvert hardness, pericarp thickness, moisture loss, and total starch content. Finally, correlations between popping performance and the baseline information were computed. Significant differences were found among samples for popping performance and all sample characteristics. Samples with high expansion volume had generally greater flake sizes, lower number of kernels/10 g, lower kernel density and lower Stenvert hardness values. Pericarp thickness did not appear to be related to popping performance. Moisture loss during microwave heating correlated significantly (r = 0.76) with percentage of unpopped kernels, indicating that the number of unpopped kernels increased as a greater amount of moisture was lost during microwave heating. With regard to total starch content, samples with high expansion volume and large kernel size had the highest starch content. Total starch content had weak but significant correlations with number of kernels/10g (r = -0.64), and flake size (r = 0.59). Number of kernels/10g and flake size were significantly and negatively correlated with each other (r = -0.77). Predictive correlations were not observed between physical and chemical traits and popping performance. It was difficult to find predictive relationships due to the narrow range of the sample traits. Expansion volume was used for developing a model with near-infrared spectroscopy (NIRS). A total of 128 samples were used and models were calibrated with different combinations of the samples. A model calibrated with 84 popcorn samples at equilibrium moisture content (13.5-14%) resulted in a cross validation R 2 = 0.56 and a RMSECV = 2.16. When the model was tested with an analysis of unknown samples, it accurately predicted expansion volume of new samples; especially samples with expansion volume greater than 50 cc/g. However, more work is needed to improve the NIRS model and the accuracy of predicting popcorn expansion volume.
Degree
M.S.
Advisors
Hamaker, Purdue University.
Subject Area
Food Science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.