Improvement In The Quality In Morphological Data

Before the 1970s, most studies of Mesozoic mammals concentrated on dental morphology, partly because teeth offer a surprisingly rich source of morphological information, but more importantly because little else was available for study. For nearly a century, it was generally held that the major features of early mammalian evolution were those documented by changes in molar pattern (Osborn, 1907; Patterson, 1956; Kermack, 1967b)—a perspective that was inevitable, given that the fossil record had yielded mainly teeth and tooth-bearing jaws. Recovery of Meso-zoic mammals by screen washing, which blossomed in the 1950s and 1960s, greatly expanded our knowledge of tax-onomic diversity and geographic distribution, but morphological data other than teeth and jaws remained stagnant. The late Alfred Sherwood Romer (1968) observed this problem with a sense of humor: "So great has been this concentration on dentitions that I often accuse my 'mammalian' colleagues, not without some degree of justice, of conceiving of mammals as consisting solely of molar teeth and of considering that mammalian evolution consisted of parent molar teeth giving birth to filial molar teeth and so on down through the ages." Before the 1980s, there were few examples in which cranial data were used to infer phylogeny of Mesozoic mammals, except for the well-known case of the lateral wall of the braincase, which was used to support the grouping of "prototherians" (Kermack, 1963; Hopson, 1964).

Remedy for this conspicuous paucity of cranial characteristics for phylogeny began in the 1980s, when skull anatomy became known for several key taxa of early mammals. By the early 1990s, enough basicranial features were known for major clades of Mesozoic mammals that it became possible for some workers to attempt phylogenetic analyses on the basis of this character complex alone (e.g., Wible and Hopson, 1993; Wible et al., 1995; Rougier, Wible, and Hopson, 1996). Embryological studies of cranial features have also provided important insight into how to interpret the orbitotemporal and rostral structures of stem mammals of the Mesozoic (e.g., Presley, 1981; Kuhn and Zeller, 1987b; Maier, 1987, 1989, 1999; Wible,

1987; Zeller, 1989a, 1993; Wible et al., 1990; Hopson and Rougier, 1993; Hillenius, 2000).

Although the skeletal anatomy of cynodonts and some stem mammals was well described in the 1970s (e.g., Jenkins, 1971b; Jenkins and Parrington, 1976), incorporation of skeletal characteristics into phylogenetic inference only began with two later studies by Kemp (1983) and Rowe (1988). These were followed by several rounds of reevaluation and further expansion of data based on skeletal characteristics, as more complete skeletons belonging to additional clades of Mesozoic mammals and relevant cynodonts were published (Sues, 1985b; Jenkins and Schaff, 1988; Kielan-Jaworowska and Gambaryan, 1994; Sereno and McKenna, 1995; Hu et al., 1997; Ji et al., 1999, 2002; Luo et al., 2002).

By the late 1990s, comparative anatomical datasets for parsimony-based phylogenetic analysis of Mesozoic mammals had been expanded to include many informative features of the dentition, mandible, basicranium and other cranial structures, and postcranial skeleton. Inevitably, various discrepancies and inconsistencies have arisen among existing data matrices. We expect that these issues will be resolved, as explicitly coded characters can be reexamined and clarified, reinterpreted, or corrected as needed, and that the currently used character matrices (e.g., Rowe, 1988; Wible and Hopson, 1993; Rougier, Wible, and Hopson, 1996; Rougier et al., 1998; Hu et al., 1997; Ji et al. 1999, 2002; Luo, Kielan-Jaworowska, and Cifelli, 2001; Luo, Crompton, and Sun, 2001; Luo et al., 2002) will continue to be revised, improved, and replenished. Given the availability of such a wealth of data, it is no longer excusable to propose hypotheses of relationships among Mesozoic mammals on the basis of some single-character complexes: such issues must be framed and tested in the context of all available morphological evidence from all relevant taxa.

0 0

Post a comment