<![CDATA[Abstract. Sa’diyah H, Waluyo B, Rusdiana RY, Sari VK, Hadi AF. 2026. Evaluating fruit morphological and physicochemical characteristics in different banana (Musa spp.) cultivars using multivariate analysis. Biodiversitas 27 (4): d270435. https://doi.org/10.13057/biodiv/d270435. Banana (Musa spp.) are an important tropical fruit in Indonesia, widely cultivated and contributing significantly to local food production. Indonesia hosts numerous banana cultivars with diverse morphological and physicochemical characteristics; however, comprehensive trait-based evaluations of local cultivars, particularly those integrating multiple traits within a single analytical framework, remain limited. This study aimed to assess the morphological and physicochemical characteristics of 20 banana cultivars collected from traditional markets in East Java and to explore their variability using a multivariate statistical approach. Sixteen quantitative traits, including fruit biometric attributes and nutrient composition, were measured at a standardized ripening stage to enable meaningful cultivar comparisons at market-ready maturity. Cluster analysis using Euclidean distance with complete linkage and Principal Component Analysis (PCA) was applied to examine trait distributions and similarities among cultivars. The results showed clear phenotypic and nutritional variation among cultivars. Several cultivars were separated from others, each exhibiting a distinct profile. These cultivars had high levels of several quality-related traits, including high Fruit Firmness (FF), Total Soluble Solids (TSS), and nutrient content, indicating potential value for improving fruit quality. Meanwhile, several cultivars formed clusters characterized by more moderate trait values. Berlin was identified as an outlier with a distinctive profile, specifically in its moisture, ash, and carbohydrate content. The clustering patterns observed in the PCA biplot were consistent with the cluster heatmap, supporting the use of the methods for integrated characterization. Overall, this study demonstrates that combining morphological and physicochemical traits can uncover complex patterns of variability that may not be apparent through univariate assessment. These findings provide useful information for cultivar differentiation, utilization, and the conservation of local banana germplasm, while also highlighting the need for future studies integrating molecular approaches to better disentangle genetic and environmental influences on key quality traits and to support precision breeding for improved nutrition and stress adaptation.]]>
