Introduction: Saffron (Crocus sativus L.) is one of the most valuable agricultural and medicinal products and holds a special position among industrial and export commodities of Iran. In addition to its use in the food industry, saffron has numerous applications in traditional and modern medicine, including cytoprotective, antioxidant, cardiovascular, respiratory, and digestive effects. Saffron production areas span a wide range of climatic conditions with considerable environmental variation, particularly in temperature. Unlike many crops, saffron has a unique thermal regime, and its growth activity typically begins with the onset of the cold season. Although saffron is generally well adapted to diverse environments, limited studies have examined the influence of cultivation management and production location on its qualitative traits. Establishing geographical indications for saffron quality especially regarding aroma, flavor, and color compounds (secondary metabolites) can provide significant commercial and economic advantages. Therefore, this research aims to support the commercialization of saffron production with an emphasis on quality.

Materials and Methods: Fourteen agronomic ecotypes were collected from Khorasan Razavi and South Khorasan provinces were evaluated using a randomized complete block design with three replications at the Gonabad Saffron and Medicinal Plants Research Station. Ecotypes were planted in late September 2018, and measurements began in the second year after planting. The evaluated traits included flower number, fresh flower weight, dry stigma weight per unit area, and the concentrations of crocin, picrocrocin, and safranal. In addition, hierarchical cluster analysis (HCA) was performed to classify ecotypes based on quantitative flower traits and qualitative secondary metabolites, aiming to identify structural relationships among traits and ecotype groupings.
Results and Discussion: Analysis of variance revealed statistically significant differences among ecotypes for all quantitative and qualitative traits. In the first year, the highest stigma dry weight was observed in the Qain–Shahik ecotype, which was 82% higher than the lowest ecotype (Roshtkhar-Fathabad). In the second year, the highest stigma dry weight was obtained in the Gonabad-Saghi ecotype, which was not significantly different from Qain–Shahik. Comparison of flower number and fresh flower weight between these two ecotypes showed that although Gonabad–Saghi produced more flowers in the second year (266 vs. 235 flowers), its fresh flower weight was lower (76.46 vs. 81.76 g m⁻²). Similar to the first year, Roshtkhar–Fathabad exhibited the lowest stigma dry weight in the second year. Hierarchical cluster analysis of ecotypes based on mean stigma dry weight further demonstrated that ecotypes were separated into two distinct clusters. Ecotypes such as Qain-Hajiabad and Zaveh-Heshmatabad formed the closest pair, showing the smallest linkage distance and the highest similarity in stigma performance. Qain-Binabaj and Roshtkhar-Barakouh were also grouped at low distances, forming a stable core cluster of ecotypes with highly similar stigma yield. In contrast, ecotypes with lower performance were grouped at higher distances, indicating greater divergence in their stigma production profiles. This clustering pattern confirms the presence of meaningful genetic and phenotypic diversity among cultivated saffron ecotypes.

Conclusion: The findings of this study demonstrated substantial quantitative and qualitative diversity among the evaluated saffron ecotypes. Ecotypes such as Qain–Shahik and Gonabad–Saghi consistently exhibited superior stigma dry weight and flower production across both years, highlighting the presence of stable performance‑related genetic variation. Conversely, ecotypes such as Rashtkhar–Fathabad showed consistently low values for key traits, indicating limited productive potential. Hierarchical cluster analysis of ecotypes based on mean stigma dry weight further clarified these differences by separating ecotypes into two distinct groups. Ecotypes including Qain-Hajiabad, Zaveh-Heshmatabad, Qain-Binabaj, and Roshtkhar-Barakouh formed a closely related cluster with minimal linkage distances, indicating highly similar stigma yield profiles and representing the core group of high‑similarity ecotypes. In contrast, ecotypes with lower performance were grouped at greater distances, reflecting higher divergence in their stigma production characteristics. This clustering pattern confirms the presence of meaningful genetic and phenotypic diversity among cultivated saffron ecotypes and underscores the importance of selecting high‑performing and genetically stable ecotypes for cultivation in different environments.