Abstract
The alpha taxonomy of dinosaurs encounters difficulties from the inability to apply biological species criteria to fossil material, and the prevalence of small sample sizes and of incomplete specimens. These enforce a quasi-typological praxis, in which differences in form must be used to distinguish species. The problem is recognizing taxonomically significant features that distinguish closely related species. Certain modern Australian lepidosaur and marsupial species would be indistinguishable if known only from fossil material available in small samples. Ecological arguments that one or few conspecific herbivores will occupy a given area and that the larger the body size of a terrestrial herbivore the less likely the existence of a sympatric closely related herbivore species, are not compelling. Ontogenetic variation in large theropods involves predominantly changes in proportion of skeletal structures, such as limbs, and some changes in proportions of individual bones. Individual variation involves largely changes in proportions of individual bones and also in the number of serial elements (teeth).