Introduction

For more than a century, systematists have debated higher-level relationships among the orders of placental mammals. The order Primates is no

Mark S. Springer • Department of Biology, University of California, Riverside, CA 92521 William J. Murphy, Eduardo Eizirik, and Stephen J. O'Brien • Laboratory of Genomic Diversity, National Cancer Institute, Frederick, MD 21702 Ole Madsen and Wilfried W. de Jong • Department of Biochemistry, University of Nijmegen, Netherlands Mark Scally • Department of Biology, University of California, Riverside, CA 92521; Queen's University of Belfast, Biology and Biochemistry, Belfast, UK Christophe J. Douady • Department of Biology, University of California, Riverside, CA 92521; Department of Biochemistry and Molecular Biology, Dalhousie University, Halifax, Nova Scotia, B3H 4H7 Canada Emma C. Teeling • Department of Biology, University of California, Riverside, CA 92521; Laboratory of Genomic Diversity, National Cancer Institute, Frederick, MD 21702 Michael J. Stanhope • Department of Population Medicine and Diagnostic Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853, USA.

exception. One prominent hypothesis is Archonta that was originally proposed as a superorder by Gregory (1910) to include primates, bats, flying lemurs, and menotyphlan insectivores (i.e., tree shrews, elephant shrews). Minus elephant shrews, the Archonta hypothesis has survived for nearly a century. The bulk of support for this hypothesis derives from modifications of the tarsus (Novacek and Wyss, 1986; Shoshani and McKenna, 1998; Szalay, 1977; Szalay and Drawhorn, 1980; Szalay and Lucas, 1993). Archonta is a recurrent theme in higher-level mammalian classifications (McKenna, 1975; McKenna and Bell, 1997; Szalay, 1977). There are also morphological studies (Cartmill and MacPhee, 1980; Luckett, 1980; Novacek, 1980; Simpson, 1945) that question the monophyly of Archonta. Simpson (1945) suggested that Archonta is "almost surely an unnatural group." Even among studies that advocate Archonta, the possibility that bats are an independently arboreal group from the Archonta has been noted (Szalay and Drawhorn, 1980). In part, this reservation was expressed because bats lack shared, derived tarsal specializations that unite other archontans (Szalay, 1977; Szalay and Drawhorn, 1980). Szalay and Drawhorn (1980) attribute this to the major functional transformation that the chiropteran ankle has undergone in association with the "extreme reorientation of the femoral-acetabular articulation." Given the absence of tarsal modifications that unite bats with other archontans, the primary rationale for including bats in Archonta is the suite of novel features that bats share with flying lemurs (Gregory, 1910; Simmons, 1995; Simmons and Quinn, 1994; Szalay and Drawhorn, 1980).

Aside from whether or not primates belong to a monophyletic Archonta, there are questions pertaining to the sister-group of primates. Several studies resolve archontans into a trichotomy between primates, tree shrews, and Volitantia (i.e., flying lemurs + bats) (Novacek, 1990; Novacek et al., 1988; Novacek and Wyss, 1986; Szalay, 1977). Other studies, some of which support Archonta and others of which do not, support a sister-group relationship between primates (or euprimates) and tree shrews (Martin, 1990; Shoshani and McKenna, 1998; Simpson, 1945; Wible and Covert, 1987; Wible and Novacek, 1988). Beard (1993) has argued for the Primatomorpha hypothesis that postulates a sister-group relationship between flying lemurs and primates. Another alternative is a sister-group relationship between tree shrews and flying lemurs, with this collective group as the sister-group to primates (Sargis, 2001).

Over the last three decades, molecular data have become increasingly important for testing and proposing hypotheses of interordinal relationships. In the mid-1970s, Goodman (1975) summarized immunological and amino acid data bearing on higher-level primate affinities. The latter included parsimony analyses of amino acid sequences for proteins such as myoglobin, and a- and ß-hemoglobins. Based on a consideration of the available molecular evidence, which was not entirely congruent, Goodman (1975) concluded that "the tentative solution adopted from immunodiffusion evidence of grouping Primates and Tupaioidea (also Dermoptera) in the superorder Archonta would seem a valid compromise." Almost two decades later, Stanhope et al. (1993) evaluated higher-level affinities of primates based on nucleotide and amino acid sequences. Addressing the Archonta hypothesis, Stanhope et al. (1993) concluded that there is "marked divergence of Chiroptera from Primates, Scandentia, and Dermoptera" based on analyses of IRBP and e-globin data sets. Instead, Stanhope et al. (1993) suggested that "a more likely primate supraordinal clade consists of Primates, Dermoptera, Lagomorpha, Rodentia, and Scandentia." Adkins and Honeycutt (1993) also concluded that Archonta is not monophyletic based on mitochondrial DNA sequences.

Additional support for the "supraordinal clade" suggested by Stanhope et al. (1993) comes from two analyses of molecular supermatrices (Madsen et al., 2001; Murphy et al., 2001a). In both studies, maximum likelihood analyses resolved placental mammals into the same four major groups: Xenarthra, Afrotheria, Laurasiatheria, and Euarchonta + Glires (= Euarchontoglires of Murphy et al., 2001b,c). The latter group includes Primates, Scandentia, Dermoptera, Lagomorpha, and Rodentia. However, Madsen et al. (2001) and Murphy et al. (2001a) did not provide convincing support for the placement of primates, tree shrews, and flying lemurs relative to each other and to Glires (Lagomorpha, Rodentia). To achieve additional resolution among the orders of placental mammals, including the phylogenetic position of primates, we combined and expanded the molecular data sets of Madsen et al. (2001) and Murphy et al. (2001a). The resulting supermatrix is 16.4 kb in length (comprising 10,059 variable and 7,785 informative characters) and includes segments of 19 nuclear genes and three mitochondrial genes for 44 taxa. Primary analyses of this data set are provided in Murphy et al. (2001b). Here, we provide an expanded set of analyses and suggest a higher-level classification for the living orders of placental mammals based on our molecular results.

0 0

Post a comment