A Report on “One Day Symposium on Numerical Cladistics”

Publisher Website
Google Scholar
Abstract
A recent symposium on numerical cladistics held at the American Museum of Natural History in New York City, addressed novel methods for searching tree space, applications of randomizations in cladistic analysis, and data management. One of the major concerns in systematics is that of finding the global optimum in tree length. The space to search is complex because it includes many local optima. It is a difficult task to escape local optima without a great loss in efficiency. The ideal is to search among suboptimal topologies and still obtain an answer in a reasonable amount of time. Nixon presented a new family of methods called "parsimony ratchet," which are successful at escaping local optima. Moilanen presented a new program which may have similar advantages. Two presentations, one by Goloboff and Farris and another by Farris, Goloboff, Källersjö, and Oxelman, introduced modifications to parsimony jackknifing that improved its accuracy when compared to normal heuristic searches. Wheeler discussed the advantages of new methods of analyzing DNA and protein sequence data, which eliminate multiple alignment; the most recent one packs nucleotides into strings which constitute the new characters. Siddall discussed different applications of randomization in cladistics and their logical consistency, finding some more acceptable than others. Nixon and Carpenter presented a new program for managing data. This symposium will probably be a landmark judging from the originality and practicality of the points presented.