A Phylogenetic Analysis of the Caminalcules. III. Fossils and Classification

Abstract

-The consequences of introducing fossil species into a numerical taxonomic data, set are examined, using the Caminalcules as an example. This group of organisms was generated artificially according to principles believed to resemble those operating in real organisms. Inclusion of fossils into the phenetic classification introduces some nonconvex taxa at higher phenetic levels, but unites phenetically homogeneous groups of mixed Recent and fossil composition. There is good correspondence of phenetics with phylogenetic sequences. The taxonomic relationships of the phenetic classification of Recent plus fossil forms can be explained by the amount of evolutionary change known to have occurred in given internodes. All but one of the mutually closest pairs in the phenogram are ancestor-descendant pairs. Gingerich's stratophenetic method yields a reasonably good estimated cladogram, but is misled by the occurrence of homoplasy in portions of the tree. Adding fossils improves estimates of the true cladogram based on the distance Wagner method but not estimates based on the Wagner parsimony algorithm. Estimated cladograms approach the true cladogram more closely than phenograms do, yet even the best estimate including the fossils has a strict consensus index no higher than 0.667. There are 10 characters that define genera. Their removal from the data base affects phenetic classifications slightly, cladistic classifications somewhat more. A cladistic classification of the Recent Caminalcules unnecessarily raises the ranks of some taxa. The less speciose and more symmetrical taxa emerge at lower taxonomic ranks than more speciose and asymmetrical taxa. Adding fossils to the Caminalcules requires a greater number of ranks for a cladistic than for a phenetic classification. [Phenetic classifications; cladistic classifications; classification of fossils; stratophenetics; Caminalcules; numerical taxonomy.] The availability of the true cladogeny of the Caminalcules, a group of artificially created organisms, has presented an opportunity for examining various principles and practices of current interest to systematists. In an earlier paper (Sokal, 1983a), I presented the images of all 77 Recent and fossil Caminalcules, definitions of their characters and the data matrix employed in this study, as well as a cladogram of their true phylogeny. A new standard phenogram was also featured. In a second paper (Sokal, 1983b), I examined the efficacy of numerical cladistic methods for finding the true cladogram from an assemblage of Recent species and examined the reasons for the departure of numerical phenetic and cladistic dendrograms from the correct genealogy. The current study deals with the consequences of introducing fossil species into a taxonomic data set, using the Caminalcules as a test case. It is of interest to compare the classifications obtained by numerical phenetic and numerical cladistic techniques in such a data set. Does introduction of fossil species simplify or complicate phenetic classifications, will it improve or worsen estimated cladograms, and does it widen or narrow the discrepancy between the two kinds of dendrograms? Knowing the true cladogram, including all fossils and the evolution of characters upon it, permits investigations of some additional taxonomic and evolutionary questions. I have already reported on various types of tree statistics and evolutionary rates (Sokal, 1983a). Here I examine whether a hierarchy of taxonomic characters can be said to exist. Furthermore, I look into the classifications that result when the Caminalcules are classified on the basis of phenograms or of cladograms, with or without the fossils. The Caminalcules are unusual as taxonomic objects in that they include all fossil members of the taxon up to and including their most recent common ancestor. The presence of fossils among the OTUs of a data set create prob-