| The Ants of Africa
CHAPTER 6 - Taxonomy - The Higher Levels of Classification
|Latin Nomenclature||Higher Levels of Classification & Phylogeny||Tribes||Taxonomic Criteria & Glossary|
A curiosity of Bolton's two great works is that both contain dogmatic statements on aspects of the classification, particularly the higher levels, without following normal scientific convention and furnishing clear evidence and argument. Notably, he refers in the second work to changes presented in the first work but, as its title stated, the latter was solely an identification guide with neither evidence nor argument. Especially confusing is the presentation of Tribes (or tribe-rank). Also, Bolton (1995) used the terms Family-Group, Genus-Group and Species-Group, for Family, Genus and Species respectively. This seems to me to be unnecessary and I have not adopted those terms.
This is a complex subject possibly clearly comprehensible only to specialist taxonomists but, from my own struggle to understand the terminology and content of the most recent papers, the following may be useful to the field worker. A brief but easily understood outline of the principles and rules of classification is that written by K.H.L. Key (Key, 1970). Modern taxonomy relies heavily on use of "numerical" or "phenetic" taxonomy. In this the taxonomist estimates the level of overall similarity ("phenetic affinity") between different organisms. Another term used is the "cladistic analysis". This is done using as many characters as possible and avoiding any incorporation of weighting, or deciding which characters have greatest importance. The end result is commonly displayed as a "tree" derived from the most parsimonious characters. Such "character analysis" deviates from the traditional approach of deriving relationships between organisms on the basis of phylogeny, by derivation of what are perceived as "primitive" versus "specialised" characters. Phenetic taxonomy, however, may have its greatest use at the higher levels, rather than at species level. An explanation of a few examples of the terminology may help -
Thus, when seeking to understand how material is classified and how separation into the various taxonomic groupings is achieved, it is worth knowing that highly specialised and very detailed studies often contribute to the process. Examples of this can be seen in the work which has emanated from the laboratories headed by William L. Brown Jr., at Cornell University. These include the study of Myrmicinae related to Solenopsis and Pheidologeton by Ettershank (1966), involving dissection and very detailed examination of body parts (mouthparts, head structure, other structures, and internal organs such as Malpighian tubules); a wider study of mouthparts by Gotwald (1969) and that of the myrmicine sting apparatus by Kugler (1978). Kugler's study, for example, clearly played a part in the synonymy by Bolton (1976, 1980, 1985) of the three genera Tetramorium, Triglyphothrix and Xiphomyrmex to give a single genus Tetramorium. Sadly, it seems that the work at Cornell proceeded in fits and starts, with each of the junior workers completing only one major study.
Following that era, William H. Gotwald Jr., who had moved to Utica College of Syracuse University, and his collaborators, David Barr and Josef van Boven, utilised the newly available computer analysis of multi-character data sets to bring a clearer understanding to the genera Dorylus and Aenictus (e.g. Barr, van Boven & Gotwald, 1985).
The complex issue of the phylogeny and status of the subfamilies has since been tackled, with use of such numerical analysis and very detailed anatomical studies, by a triumvirate of Barry Bolton, still at the Natural History Museum in London, Philip Ward and Steven Shattuck, both the latter at the University of California, Davis. Bolton concentrated on the cerapachyine ants, the subfamily Leptanillinae and gave further attention to the doryline army ants (Bolton, 1990a-c). Philip Ward made an analysis of the Pseudomyrmecinae, in which he utilised 50 morphological characters and measurements in a numerical cladistic analysis (Ward, 1990). Lastly, Shattuck, who moved to join CSIRO in Australia, focused on the subfamilies Aneuretinae, Dolichoderinae and Formicinae (Shattuck, 1992). The process continues, however, as is illustrated by Shattuck's phylogenetic tree which included 14 subfamilies; one, Sphecomyrminae, being known only from fossils. Since then two more living subfamilies have been recognised (Aenictogitoninae and Apomyrminae) (Baroni Urbani, Bolton & Ward, 1992), plus three further fossil subfamilies (Armaniinae, Formiciinae and Paleosminthuridae) (see Bolton, 1995).
|Latin Nomenclature||Higher Levels of Classification & Phylogeny||Tribes||Taxonomic Criteria & Glossary||
©1998, 2003 - Brian Taylor CBiol FSB FRES
11, Grazingfield, Wilford, Nottingham, NG11 7FN, U.K.