Study on numerical taxonomy of Macadamia spp. germplasm resources

Abstract:【Objective】According to the phenotypic and genetic diversity, the numerical taxonomy of macadamia germplasm resources was studied. The traditional morphological taxonomy is based on one or several botanical characteristics, which may not represent the comprehensive characteristics of each germplasm resource. Morphological numerical taxonomy is based on the similarity of phenotypic traits, and constructs a clustering tree diagram by quantitative comparison and pedigree analysis of biological population, which is more objective and accurate than the traditional morphological taxonomy.【Methods】52 phenotypic traits were observed and described with reference to the methods described in the books of“Description specification and data standard for macadamia germplasm resources”and“Technical specification of identification for macadamia germplasm resources”, and the data were analyzed by Q cluster analysis, R cluster analysis and principal component analysis. After the original data were standardized, the clustering tree diagram of 60 macadamia germplasm resources was constructed by Q cluster analysis using the method of average Euclidean distance and square sum of deviations. 52 phenotypic traits were analyzed by R cluster analysis using the method of correlation coefficient and square sum of deviations. Based on the standardized data, 52 phenotypic traits were analyzed by principal com-ponent analysis, and the numbers of principal components were determined according to the latent root and accumulative contributor ratio, and the phenotypic traits were selected according to the loading value.【Results】The average variation type of 26 descriptive characteristics was up to 3.4, among which the young leaf color, leaf shape and dent size on lateral hilum had the most variation with coefficient 5. The frequency of various types of descriptive characteristics varied greatly. The frequency of 12 descriptive types was more than 60% and the frequency of 5 descriptive types was less than 5%. Among 26 numerical characteristics, the inflorescence length was largest and the coefficient of variation was 21.54%, and the smallest coefficient of variation of the oil content was 1.68%. The variation coefficients of leaf length, leaf width, leaf shape index, leaf petiole length, inflorescence length, floweret number, fresh weight of single fruit, fruit petiole length, dry weight of single seed, shell thickness, dry weight of single kernel and percentage of kernel were 12.53%-21.54%. The leaf shape index was very significantly and positively correlated with leaf length and negatively correlated with leaf width, while inflorescence length was very significantly and positively correlated with floweret number. The fresh weight of single fruit, fruit longitudinal diameter and fruit transverse diameter were very significantly or significantly positively correlated with the dry weight of single seed, seed longitudinal diameter and seed transverse diameter, dry weight of single kernel, kernel longitudinal diameter and kernel transverse diameter. The percentage of kernel was very significantly and positively correlated with the dry weight of single kernel and negatively correlated with the shell thickness. The Grade 1 kernel percentage was very significantly or significantly positively correlated with the dry weight of single seed, dry weight of single kernel and kernel transverse diameter. The oil content was very significantly or significantly positively correlated with the dry weight of single seed, dry weight of single kernel and Grade 1 kernel percentage. The result of Q cluster analysis based on 52 phenotypic characteristics demonstrated that 60 macadamia germplasm resources were divided into 4 groups at the Euclidean distance of 20.83 according to the characteristics of the fresh weight of single fruit, dry weight of single seed, dry weight of single kernel, leaf arrangement, leaf petiole length, young leaf color, floweret color and surface texture of seed, e.g., M. integrifolia with small fruit, M. integrifolia with medium fruit, M. integrifolia with large fruit and M. integrifolia × M. tetraphylla Hybrids with largest fruit. The result of R cluster analysis showed that 52 phenotypic characteristics were significantly clustered into 5 groups with the coefficient of 2.16, which revealed well the correlation between phenotypic characteristics. Principal component analysis indicated 52 phenotypic characteristics were mainly composed of 15 independent principal components, the first and second principal components explained 27.427 2% of total variance, and the results were basically consistent with those of the cluster analysis.【Conclusion】The phenotypic and genetic diversities of plant, leaf, flower and fruit were abundant in macadamia germplasm resources. The main phenotypic characteristics, such as leaf arrangement, leaf petiole length, young leaf color, inflorescence length, floweret number, floweret color, fresh weight of single fruit, dry weight of single seed, dry weight of single kernel, surface texture of seed, percentage of kernel, Grade.1 kernel percentage, shell thickness, fruit abdominal suture, seed abdominal suture, surface speckle distribution of seed, micropyle and hilum, played an important role in the phenotypic diversity of macadamia germplasm resources. 60 macadamia germplasm resources were clustered into four groups according to the characteristics of the weight of single fruit, leaf arrangement, leaf petiole length, young leaf color, floweret color and surface texture of seed. Based on phenotypic characteristics, numerical taxonomy had an applied value for building the taxonomy system for macadamia germplasm resources.