Classification study on quantitative characteristics of pomelo germplasm resources for DUS test

Abstract: 【Objective】The DUS (Distinctness, Uniformity and Stability) test refers to the process of testing the specificity, consistency and stability of new plant varieties. The DUS test guideline is the technical basis and standard for DUS test of tested varieties, and has become the technical basis for variety approval, variety registration and variety protection in China. At present, the DUS testing guideline of Citrus in China (NY/T 2435—2013) covers the entire Citrus genus. However, there are great differences in some characteristics between the different species in the same genus . Therefore, this study aimed to explore the adaptability of quantitative characteristics in the DUS testing guidelines for Citrus in China to pomelo varieties, and to provide reference for more objective and scientific description of quantitative characteristics of pomelo and revision of DUS test guidelines of Citrus.【Methods】In this study, twenty- five quantitative characteristics of the 117 pomelo germplasm resources in the National Citrus Germplasm Resource collection were observed according to the DUS testing guideline of Citrus in China. The data of these twenty- five quantitative characteristics were collected by measurement method. The normality of the data was comprehensively judged according to the significant test valueof the K-S test for each quantitative characteristics, the fitting of the frequency distribution diagram and the normality curve, and the deviation between the measured value of the Q-Q diagram and the predicted normal value. The least significant difference method was used to grade the quantitative characteristics that conformed to the normal distribution, and the range method was used to grade the quantitative characteristics that did not conform to the normal distribution.【Results】The intra-variety variation coefficient of the twenty- five quantitative characteristics was between 2.96% and 26.96%, and the intervariety variation coefficient was between 10.33% and 62.90%. Most quantitative characteristics showed great polymorphism and stability. According to the results of normality test, the petal length, petal length/width, leaf blade length, leaf blade length/ width, petiole length, petiole wing width, stylar scar diameter, core diameter, fruit rind thickness, number of segments per fruit, number of seeds, fruit juiciness, total soluble solids, fruit juice acidity and edible rate conformed to normal distribution; The length of spring shoot internode, petal width, stamens number per flower and leaf blade width conformed to the approximate normal distribution, calyx diameter, fruit lengt, fruit diameter, fruit length/diameter, juice vesicles length and single- fruit weight did not conform to normal distribution. Among the six quantitative characteristics that did not follow a normal distribution, the frequency distribution of the fruit weight, fruit length and fruit diameter showed a clear bimodal distribution, which was mainly caused by the significant differences between pomelo and grapefruit species. Grapefruit and pomelo showed significant differences in these three characteristics, leading to a bimodal distribution. Therefore, if pomelo and grapefruit were uniformly graded on these quantitative characteristics, it would lead to a wide grading range and misjudge varieties with differences as no differences, thereby increase the probability of errors occurring. This study developed different grading standards for pomelo and grapefruit based on these three characteristics. After separating the data of the two species based on the three quantitative characteristics mentioned above, it was found that except for the pomelo group whose data on fruit diameter did not follow a normal distribution, the rest basically followed a normal distribution or approximate normal distribution. The least significant difference method was used to grade the quantitative characteristics that conformed to the normal distribution, and the range method was used to grade the quantitative characteristics that did not conformed to the normal distribution. The classification standards of the twenty-five quantitative characteristics were obtained. Compared with the grading numbers in the DUS testing guideline of Citrus in China, the calyx diameter, fruit shape index and juice cell length increased from 3 levels to 5 levels, the petal length, petal width, leaf length, leaf width and pericarp thickness decreased from 9 levels to 7 levels, and the petal index, leaf shape index, central column size, capsule number, juice content and soluble solids decreased from 9 levels to 5 levels. For fruit weight, fruit length and fruit diameter, the grading standards of pomelo and grapefruit were formulated respectively. The Shannon diversity index of each characteristics under this grading standard ranged from 0.878 to 1.954, and the Shannon diversity index of fruit juice acidity was relatively large, indicating that this characteristics had more phenotypes and was evenly distributed in different grades. The correlation coefficient of the twenty-five quantitative characteristics ranged from -0.69 to 0.91, and the correlation coefficient between fruit diameter and fruit weight was the highest, reaching 0.91, showing a very significant positive correlation. Six principal components were obtained by principal component analysis, with a cumulative contribution rate of 72.76% . The first principal component contained the most information, which was mainly determined by the 11 characteristics, including the fruit diameter, fruit weight, fruit length, petal width, calyx diameter, petal length, core diameter, stamens number per flower, juice vesicles length, leaf blade width and number of seeds.【Conclusion】This study estab-lished a preliminary classification index for the twenty-five quantitative characteristics of pomelo germplasm resources, which would have certain guiding significance for the evaluation, utilization, and breeding of pomelo germplasm resources. At the same time, it would provide reference for the identification of pomelo germplasm and the revision of pomelo DUS testing guideline in China. In addition, the classification standard of this study was based on the growth performance of pomelo in Chongqing, and the specific classification range might not be fully applicable to other ecological regions. In the process of DUS testing, researchers should make corresponding adjustments according to the expression status of standard varieties and combine with the field conditions, so as to ensure accurate evaluation and comparison of the character differences of the varieties in different regions.