Discriminating Groups of Organisms

A common problem in morphometric studies is to determine whether, and in what ways, two or more previously established groups of organisms differ. Discrimination of predefined groups is a very different problem than trying to characterize the patterns of morphological variation among individuals, and so the kinds of morphometric tools used for these two kinds of questions differ. In this paper I review the basic procedures used for discriminating groups of organisms based on morphological characteristics – measures of size and shape. A critical reading of morphometric discrimination studies of various kinds of organisms in recent years suggests that a review of procedures is warranted, particularly with regard to the kinds of assumptions being made. I will discuss the main concepts and methods used in problems of discrimination, first using conventional morphometric characters (measured distances between putatively homologous landmarks), and then using landmarks directly with geometric morphometric approaches.