Cluster Analysis in Pubescent Oak Taxa from Series Lanuginosae: A Case Study

Abstract

The main purpose of this study was to compare the different clustering techniques in order to identify the one with the best discriminating power among oak species. By using two groups of trees, one corresponding to pedunculate oak and the other one to pubescent oak, Ward’s Method with Manhattan distances provided the best separation between the two oak species. By contrast, different results were obtained for the two taxa from series Lanuginosae (pubescent oak and Italian oak), due mainly to their similarities in leaf and fruit traits. In conclusion, by using a wide variety of clustering combinations within STATISTICA software, no separation between pubescent oak and Italian oak was achieved.