Application of functional tree classification (FTs) for detection of body shape variation within and among two species of Aphanius (Teleostei, Cyprinodontidae)

Morphometric investigation was conducted on two species of Aphanius (Aphanius dispar (Rüppell, ۱۸۲۸), and A. ginaonis (Holly, ۱۹۲۹). A. ginaonis was sampled from Geno hot spring in Hormozgan province, Iran while Aphanius dispar, was from two locations in  Bushehr proviance (Boneh Jabery drainage and Mir Ahmad Hot spring). Fourteen characters   were measured based on the conventional morphometric concept. In this paper, we introduce an innovative data mining technique (functional trees classification) to assess the pattern of variation within and among the population of cyprinodontidae and to classify them in the northern part of Persian Gulf coast. In the first model, A. dispar population was classified as one class and A. ginaonsis population as another class using different parameters. To do so, +۱ was set as class of fishes from A. dispar and -۱ as class of fishes from A. ginaonis. The speices of A. ginaosis from Geno hot spring were discriminated from A. disparfrom from Mirahmad hot spring and Boneh Jaberi driage. The result also showed that two populations of A. dispar from Bushehr province were well discriminated from each other using functional tree  technique. The derived model of the two populations of A. dispar and one population of A. ginaonsis in this study could be explained as the result of the isolation of the populations and species.

Keywords:

Aphanius , conventional morphometric , functional tree modeling , data mining , Iran