"Cladus" and clade: a taxonomic odyssey.

The fate of "clade," both as concept and word, is reconstructed here beginning with its first appearance in 1866 as "Cladus," in Haeckel's Generelle Morphologie, continuing up to the present. Although central to phylogenetics, the concept of clade is paradoxical since it has been ambiguously understood or even misunderstood by its own promoters. Writings by Ernst Haeckel, Lucien Cuénot, and Julian Huxley, the three authors who discussed the notion of clade at length, are analyzed here in detail as a means of exploring this paradox. First conceived as a rank for a higher-level category, and later as a taxon, the clade is understood today in connection with Hennig's definition of a monophyletic group rather than through Huxley's successful but somehow ambiguous formalization. The inability of these authors to formulate a clear-cut exposition of the concept is considered here within three contexts: firstly, the burden of pre-Darwinian classifications based on similarity; secondly, the underestimation of Darwin's description of a genealogical group; and thirdly, the predominance of thinking in process (vs thinking in pattern), which was the basis of evolutionary systematics in the mid-twentieth century.

Molecular phylogeny of Elephantidae. Extreme divergence of the extant forest African elephant.

A phylogenetic study of the Elephantidae (Proboscidea, Mammalia) is based on the cytochrome b mitochondrial gene: 31 terminals, that is, all known sequences, one non-elephantid proboscidean, the extinct American mastodon, and four outgroups. The data set includes 11 new sequences with the first published sequence of the forest African elephant, L. a. cyclotis. The analyses of extant taxa only and of both extant and extinct taxa show that L. a. cyclotis is highly divergent from L. a. africana. It is as divergent from L. a. africana as Loxodonta is divergent from Elephas. Southern L. a. africana form a clade. The continental subspecies E. m. indicus is paraphyletic with individuals from India and Thailand closer to E. m. maximus (Sri-Lanka). Members of Mammuthus primigenius are more closely related to Loxodonta although they do not form a clade; two specimens of M. primigenius are closer to L. a. africana making the genus Loxodonta paraphyletic. The latter conclusion may be partly due to unequal length of the various polymorphic mammoth sequences.