Introduction: Sustainable forage production is critical in arid and semi-arid regions, where water scarcity limits agricultural options. Safflower (Carthamus tinctorius L.), a drought-tolerant, multipurpose crop is a promising forage option, particularly spineless genotypes, for these regions. However, few studies have evaluated genetic diversity for forage-related traits in spineless safflower under arid conditions. This study aimed to assess 36 spineless safflower genotypes to identify superior lines for forage production, analyze key trait relationships, and classify the germplasm to guide future breeding programs.

Materials and Methods: 36 spineless safflower genotypes were evaluated in an Alpha Lattice design with two replications at Saravan University research farm (2023-2024). Data were collected for phenological, agro-morphological, and fresh/dry forage yield traits at the the 5% pod formation stage. Statistical analyses included Analysis of Variance (ANOVA), LSD test for mean comparisons of dry forage yield, Pearson correlation analysis among all traits, cluster analysis (Ward’s method), and Principal Component Analysis (PCA).

Results and Discussion: Analysis of Variance revealed highly significant differences (P<0.01) among safflower genotypes for all evaluated traits, confirming substantial genetic diversity in the germplasm and providing a strong basis for selection. This variation enables the development of improved varieties for challenging environments through breeding programs. Mean comparison for dry forage yield (DFY) identified high-yielding genotypes: Isfahan-4110 (15,671 kg ha⁻¹) and TN79563 (15,235 kg ha⁻¹), followed by CART1, CART162, and Isfahan10. These genotypes stand out as promising candidates for direct use or further breeding efforts. Dry forage yield correlated strongly (P<0.01) with key biomass components: leaf weight (r=0.96), lateral branch weight (r=0.95), and stem weight (r=0.88), and moderately with fresh forage yield (r=0.80) and number of lateral branches (r=0.65). These strong correlations imply that indirect selection based on these more easily assessable traits, particularly biomass components like leaf and branch weight which directly contribute to forage quality and quantity, can be an efficient strategy for indirect improvement of dry forage yield, potentially saving time and resources compared to direct yield measurement. Cluster analysis, using Ward’s method, grouped genotypes into six clusters, with cluster six containing the high-yielded genotypes with superior mean values for forage yield and related biomass traits. The clear distinction of cluster six from low-yielding clusters (such as clusters two and five mentioned in previous discussions) highlights the potential for exploiting heterosis by crossing genotypes from these divergent groups to generate superior F1 hybrids or new segregants in later generations. PCA revealed that the first four PCs explained approximately 82 percent of total genetic variation, primarily representingforage productivity traits. PC1 alone accounted for around 44 percent of the total variance and was strongly associated with traits contributing to overall biomass production and forage yield, essentially representing the primary axis of variation related to forage productivity or ‘general growth’ potential. The biplot analysis provided a clear visual confirmation of these findings. It positioned the elite genotypes of Cluster VI in close association with the vectors for key yield-contributing traits (DFY, LW, SW) along the PC1 axis, confirming their high biomass potential. In contrast, PC2, which captured a secondary dimension of variation, was primarily associated with traits related to plant architecture (e.g., height to first lateral branch and number of leaves), effectively separating genotypes based on their growth habit.

Conclusion the spineless safflower genotypes exhibited significant genetic diversity for forage related traits under arid conditions. The identified superior genotypes (Isfahan-4110, TN79563, CART1, CART162, and Isfahan10) are valuable genetic resources suitable for direct forage cultivation in arid/semi-arid regions or as breeding parents to enhance forage yield and biomass. The strong correlations between DFY and biomass components offer efficient avenues for indirect selection. Multivariate analyses effectively characterized the genetic structure provided a clear strategic path for breeding: crossing the elite genotypes from Cluster VI with members of genetically divergent clusters (e.g., II and V). Such targeted crosses are highlighted as a key strategy for maximizing heterosis and generating superior transgressive segregants. These findings provide critical information for enhancing the efficiency of breeding efforts to develop improved forage safflower varieties adapted to water-limited conditions, supporting sustainable livestock feed production in dry areas. The potential for exploiting genetic distance through targeted crosses between superior and divergent genotypes for maximizing heterosis is also highlighted.