| Rice is one of the main crops in the world. Being a major source of dietary nutrients, the germplasm with better quality traits would be useful for improving the rice quality in rice breeding. In this study, NIRS technology was used to determine the rice quality traits which were amylose content, gel consistency, alkaline spread value, protein content, total of amino acid content, different amino acid contents including asparagine, threonine, serine, glutamine, glycine, alanine, cysteine, valine, methionine, isoleucine, leucine, tyrosine, phenylanine, lysine, histidine, arginine, proline and rapid viscosity analyzer traits including peak viscosity, hot paste viscosity, breakdown viscosity and setback viscosity for 990 rice varieties/lines. The predicted genotypic values were unbiased predicted using mixed linear model approach and used to measure the genetic similarities among rice varieties/lines by mahalanobis distance method. Eight hierarchical clustering methods including single linkage method, complete linkage method, median linkage method, centroid method, unweighted pair-group average method, Ward's method, flexible-beta method and weighted pair-group average method were combined with three sampling strategies, random, preferred and deviation sampling to develop 24-core collections. Four statistical parameters were used to evaluate and appraise the core collection in relation to the initial collection: (1) the percentage of significant difference (MD%, a = 0.05) between core collection and initial collection for means of traits, (2) the percentage of significant difference (VD%, a = 0.05) between core collection and initial collection for variance of traits, (3) coincidence rate (CR%) and (4) variable rate (VR%). Major results obtained are as follows:1. Rice germplasm genetic resource potentialThere exist large genetic potential for rice quality traits studied. The diversities of 990 varieties/lines studied for quality traits offer promise for need to carry evaluation procedures to identify materials that can be incorporated in the breeding programs focused in this area. All these resources if properly utilized can help in improvement of rice breeding programs. The analysis of genetic variation for quality traits among these rice varieties/lines studied could provide information about genetic diversity, development of core collection and is a platform for stratified sampling of breeding populations. The improvement for quality traits of rice could be achieved for identifying diverse parental combinations to create segregating progenies with maximum genetic variability for further selection, and introgressing desirable genes from diverse germplasm into available genetic base.2. The effects on core collection development with different sample proportionsDuring core collection development a difficult task comes when attempting to reduce the sample size of a large collection without losing its diversity. In this study, in order to find suitable sample proportion which could be effective to develop core collection for quality traits of rice, the samples were subjected to analysis at sample proportion 10%, 15% and 20%, respectively. In general, the needs for developing suitable core collection in which MD% is no more than 20%, CR% is no less than 80%, with the higher VD% and VR% value. The results revealed that the sample proportion 15% was considered better because it met all the requirements. Hence, this method was able to maintain the diversity of initial collection at 15% than at any other proportion and could be used in developing collection of rice quality traits.3. Evaluation of different cluster methods and sampling strategies... |
