Thesis - University Access Only
Master of Science (MS)
Department / School
End-use quality improvement is an important component of winter wheat (Triticum aestivum L.) breeding efforts. A computerized Mixograph has been developed to assist in these efforts, but information is lacking for effective integration of this technology into breeding programs. The objectives of this study were: to explore the application of multivariate analysis to facilitate interpretation and utilization of predictive quality testing using the computerized Mixograph; to identify the association of derived factors with composite baking data using multiple linear regression; and compare two methods of curve analysis, midline and envelope, from the computerized Mixograph. The samples (143 total) used in this study originated from diverse field testing locations in South Dakota from the 1995-96 growing season. Samples were cleaned and tempered to 15.5% moisture before milling. Flour protein concentration was estimated for each sample using the LECO Nitrogen Analyzer. Flour protein quality was evaluated using the computerized Mixograph and a modified sodium dodecyl sulfate (SOS) sedimentation technique. Sample data (39 variables, 17 from envelope analysis and 22 from midline analysis) were subjected to multivariate analysis using an unweighted least squares factor analysis method with varimax rotation. Three primary factors were identified that collectively explained 92% of the total variance. Based on primary loadings on the first factor, eleven variables were identified that collectively appeared to be associated with water absorption. Primary loadings on the second factor identified nine variables that appeared to be associated with mixing tolerance. Six variables that loaded on the third factor appeared to be associated with mixing time. Results from multiple linear regression of factor scores on composite baking data from the samples showed that factors 1 and 3 explained 28% and 42% of the variation in baking absorption and baking mix time, respectively. Factor associations demonstrated that midline analysis is important for mixing time and water absorption whereas envelope analysis is important for mixing tolerance. These results suggest that multivariate analysis and data reduction of computerized Mixograph profiles may be potentially useful tools for prediction of important end-use quality criteria.
Library of Congress Subject Headings
Winter wheat -- Breeding Winter wheat -- Quality Multivariate analysis
Number of Pages
South Dakota State University
Bagci, S. Ahmet, "Multivariate Analysis of Computerized Mixograph Data for End-use Quality Improvement in Winter Wheat" (1998). Electronic Theses and Dissertations. 494.