CLASSIFYING SALVIA SPECIES BY HIERARCHICAL CLUSTER ANALYSIS OF THEIR SEED FEATURES

This research was to develop a comprehensive seed morphological, colorimetric, physiological and anatomical database for classifying eleven Salvia species into some statistically homogeneous groups, and to compare these groups with those stated in the Flora of Iran based on the morphological characters of the same plant species. For this purpose, a total of 51 seed features, including the 8 morphological, 6 colorimetric, 34 physiological, and 3 anatomical seed traits, were selected and measured for all the Salvia species. Hierarchical Cluster Analysis (HCA), conducted based on Squared Euclidean Distance and Ward's method, was able to classify the Salvia species into four and six groups depending upon the seed features used in the classification process. It is worth mentioning that the classification obtained in six groups based on the combination of seed morphological and colorimetric features were in agreement and very close to the classification of the same plant species mentioned in the Flora of Iran; while the four-group classifications resulted based on the seed physiological or anatomical features, or the combination of all the features together, were different than what stated in the Flora of Iran.

Automated summary

The study developed a comprehensive seed database for classifying Salvia species based on various seed features, and found that using a combination of morphological and colorimetric features led to classification results consistent with the Flora of Iran, while using physiological or anatomical features resulted in different classifications.