Multi-Trait Composite Performance Index for the Identification of Elite Pistachio ( <i>Pistacia vera</i> L.) Genotypes with Superior Kernel Efficiency

Abstract Identifying high-efficiency pistachio genotypes through integrative selection tools is essential for accelerating breeding programmes and enhancing commercial nut quality. In this study, therefore, 169 Pistacia vera L. genotypes were comprehensively evaluated using a Composite Performance Index (CPI) integrating kernel yield, shell thickness, nut size, split ratio, and associated biometric traits. CPI values ranged from –1.9 to 1.4, enabling precise ranking and discrimination of genotypes based on multidimensional performance. Nineteen elite genotypes representing the top 11.2% of the population were selected, with KS-168, KS-103, and KS-132 ranking highest due to superior kernel yield (up to 58.80%), ultra-thin shells (0.50– 0.53 mm), and commercially favorable split ratios (up to 96.7%). Standardized z-score analysis revealed that 16 of these 19 genotypes exhibited negative shell thickness deviation, confirming effective selection pressure towards thin-shelled phenotypes. ANOVA detected significant variability across all pomological traits ( p &lt; 0.05), and Duncan’s test separated the genotypes into statistically distinct performance groups. Hierarchical clustering further classified selections into high-yielding, efficient-shell, and balanced phenotypic classes. Pearson correlation analysis indicated a moderate association between nut size and kernel yield (r = 0.48), while other traits showed weak or negligible relationships, highlighting opportunities for simultaneous improvement without major trade-offs. These findings demonstrate that CPI-based integrative selection is a robust and scalable approach for identifying elite genetic material in pistachio breeding programmes.