A Model for Selecting Kiwifruit () Germplasm Resources with Excellent Fruit Quality.

The evaluation of quality traits is an important procedure for kiwifruit breeding and comprehensive utilization. The present study aimed to establish a comprehensive system to assess germplasms by analyzing 22 quality traits on kiwifruit samples collected from a wild population of 236 plants grown in the Jiangxi Province, China. Variability, correlation, principal components, and cluster analyses were carried out using the data collected from fruit quality evaluations. The coefficients of variation (CV) of fruit quality traits ranged from 11.66 to 66.16% (average coefficient = 35.09%), indicating a high level of variation among the 236 plants. In addition, different degrees of correlations were found between the traits, with similar traits demonstrating strong correlations. Principal component analysis (PCA) generated eight comprehensive and independent principal components, accounting for 77.93% of the original fruit quality information. Furthermore, an extensive evaluation from PCA ranked the plants based on cluster analysis and grouped them into seven categories. A stepwise regression analysis generated a prediction model, demonstrating a good fit (0.945) with the principal components of the comprehensive evaluation score. Overall, this study identifies nine quality traits, representing fruit appearance, sweetness, acidity, flavor, and nutritional attributes, as important traits for a comprehensive evaluation of fruits.