Construction Of Core Germplasm Of Camellia

Camellia originating in China, is rare flowers on China and even the world, pivotal in flowerproduction. China has rich camellia germplasm resources, for indepth study of camellia uniqueresources and good gennes to provide adequate basic conditions, but bring great difficulties to saveresources and high cost, causing huge loss of resources, and also limits the further system researchand development and utilization of resources. Construction of core germplasm repository is theimportant method to solve the problem of current situation. In this study, the investigation of theoriginal resource data Camellia camellia strains strains and assignment processing. Use differentcombinations of different clustering methods and sampling methods were built Core Collection. Andeffectiveness indicators computing analysis, identification and evaluation of different strategies tobuild the core collection. The main results are as follows:1, Follow this research requirement, screening and classification traits selected46(coveringflower, leaf, bud, such as camellia main ornamental parts), and set up1075strain of Camelliavarieties of germplasm resources database table. Statistics shows high heterogeneity, camellia itselfhas significant variability.2, Adopting the processing mode of standardization and standardization to assignment theestablish of germplasm resources database. Build1075strain Camellia varieties correspondinggermplasm resources data assignment table.3, Apply different genetic distances (Mahalanobis, Euclidean), sampling methods (completelyrandom, multiple clustering random, multiple clustering difference degree) and clustering methods(the shortest distance, center distance, the longest distance, center of gravity method, not weightedaverage method, the weighted arithmetic mean and deviation method) combination, respectively tobuild15core germplasm repository. Indicates that the shortest distance with gravity method ofcombination plan, more suitable for smaller strain aggregate construct genetic differences.4, Data results show that the Mahalanobis genetic distance is better than Euclidean; multipleclustering difference degree sampling method is superior to other two methods; the shortest distanceideal and gravity method, followed by the method of center distance.The research results show that using Mahalanobis genetic distance, multiple clustering differencedegree, the shortest distance, or center of gravity method is used to build the core germplasm is themost ideal effect. That fixed genetic distance range and heterogeneous degree of big data, the basis ofseveral times with cluster sampling difference degree effect is best; and for genetic distance betweenthe unknown or range were once large data, the basis of the need to consider multiple clusteringsampling priority to build processing. Core germplasm build needs to take into account the basiccharacteristics and data processing, construction strategy scheme, evaluation appraisal index ofuniformity among the selected match.