Phylogenetic Analysis and Fossil Plants

Stein Jr., W.E., 1987. Phylogenetic analysis and fossil plants. Rev. Palaeobot. Palynol., 50:31 61. This paper presents a perspective on the relat ionship between cladistic methods and the study of fossil plants. Paleobotany has been slow in evaluat ing the potent ial of these and other explicit methods compared to other branches of systematics. However, the unique or part icular ly difficult problems paleobotanists face make more interesting, ra ther than prohibit, the i r application. A model of a cladistic approach to phylogenetic analysis is introduced in order to evaluate what role fossils, including fossil plants, should play in phylogenetic analysis. Major elements of the model include sets of hypotheses about useful systematic comparison, phylogenetic relationship, evolutionary pat tern and process, and age, linked by recursive directional arguments. Arguments at one level only, those l inking characters with the cladogram, are viewed as being strictly deductive. Character hypotheses at the highest level are dist inguished by the term '¢cladistic character" . When cladistic characters are compatible, estimates of phylogeny suggested by each may be readily combined into a summary result. However, character conflict is common and represents a major challenge to the ul t imate success of cladistic methods. The main advantage of cladistic characters lies in the at tempt to separate out portions of complex phylogenetic hypotheses for independent scrutiny and debate. Under the model of phylogenetic analysis presented here, i t is argued tha t fossils provide morphological information, not obtainable by other means of inquiry, which may have a profound effect on our view of characters and relationship. Fossils also provide estimates of age useful in two dist inct ways. The first involves using age estimates of states, combined with external cri teria such as out-group analysis, in providing defensible proximity and polarity proposals for cladistic characters. The second involves using age estimates of entire taxa to judge historical plausibility of a set of phylogenetic hypotheses generated from a cladogram. Incompatible characters may also be analyzed in this way, allowing evaluat ion of the plausibility of a l ternat ive character formulations. The use of fossils in phylogenetic analysis tends to emphasize the necessity of coming to terms explicitly with the ancestor-descendant relat ionship in cladistic methods. It is proposed that , when properly formulated, the relat ionship is similar in s t ructure to other phylogenetic hypotheses and at least part ly addressed by the hypothetico-deductive format.