Authors: Cengizhan ÖZBAY, Hasan AKBAYIN
Abstract: Large numbers of specimens of the following 3 species, Polistes dominulus buncharensis Polistes gallicus, Vespula germanica, of Vespidae were collected from different Iocations in the provinces Diyarbakır , Mardin and Şanlıurfa between 1989 and 1991 and identified by conventional identification keys based on variations in coloration and differences in body structure among these species. However, as there is overlap in variability within and between species, it is sometimes rather difficult to identify the specimens. It was our intent to solve promlem by employing discriminant analysis. Therefore, eight measurable budy variables were taken from each collected specimen. Among the specimens, we selected 266 that were closely in accord with the identification keys. Discriminant analysis was applied to this data matrix. Consequently, the discriminant functions obtained were concluded to be more efficient than conventional methods for classifying these species, and discriminant analysis was demonstrated to be an efficient method in systematic studies when there are difficulties in distinguishing species that closely resemble each other .
Keywords: Biometrics, Discriminant analysis, Numerical Taxonomy, Vespidae