Major Clades of Placental Mammals

Our analyses extend studies with the Murphy et al. (2001b) data set and provide robust support for four major clades of placental mammals (Afrotheria, Xenarthra, Euarchontoglires, Laurasiatheria). We also find robust support for a Euarchontoglires + Laurasiatheria clade (i.e., Boreoeutheria; Springer and de Jong, 2001). Sequences for a 1.3 kb segment of the apolipoprotein B gene also recover these clades (Amrine-Madsen et al., 2003). In addition, Waddell et al. (2001) argued for these clades based on analyses of amino acid sequences. In contrast, Arnason et al. (2002) failed to recover monophyletic Laurasiatheria and Euarchontoglires clades in their analysis of mitogenomic sequences. Instead, both Laurasiatheria and Eulipotyphla were diphyletic with Erinaceomorpha (hedgehog, moon rat) as the first placental branch followed by a paraphyletic Euarchontoglires (and Rodentia). In contrast, Lin et al. (2002) found that mitogenomic sequences recover Xenarthra, Afrotheria, Laurasiatheria, and Euarchontoglires in unrooted analyses. This arrangement is in fundamental agreement with nuclear data presented here and elsewhere (Delsuc et al., 2002; Madsen et al., 2001; Murphy et al., 2001a,b; Scally et al., 2001; Waddell et al., 2001). In rooted analyses, however, results similar to Arnason et al. (2002) were recovered with eulipoty-phlan diphyly, rodent paraphyly, and Euarchontoglires paraphyly. As noted by Lin et al. (2002), adding an outgroup should not result in changes within the ingroup if the model of sequence evolution is correct. Lin et al. (2002) concluded that the unrooted mitochondrial tree is correct and that peculiar features of the rooted mitochondrial tree, such as eulipotyphlan diphyly and rodent paraphyly, are the result of inadequate models that do not take into account changes in mutational mechanisms in murid rodents, erinaceo-morphs, and some marsupials. Further, Hudelot et al. (2003) performed rooted analyses with mitochondrial RNA (tRNA and rRNA) gene sequences that incorporated secondary structure information and found support for Xenarthra, Afrotheria, Laurasiatheria, and Euarchontoglires. The monophyly of Euarchontoglires is especially compelling in view of two different deletions in nuclear genes that support this clade (de Jong et al., 2003; Poux et al., 2002).

Morphology agrees with our molecular results in supporting Xenarthra. However, Afrotheria, Euarchontoglires, Laurasiatheria, and Boreoeutheria are all without morphological support. Instead, analyses of morphological characters have placed taxa with similar morphotypes together even though constituent taxa belong to different clades, e.g., paenungulates (Afrotheria) and perissodactyls (Laurasiatheria) are sometimes united together in the superordinal group Altungulata. From a molecular perspective, the occurrence of similar morphotypes in different clades (e.g., ungulates in Afrotheria versus Laurasiatheria) must be regarded as parallel/convergent evolution (Helgen, 2003; Madsen et al., 2001; Scally et al., 2001).

Relationships in Afrotheria

In agreement with most molecular studies and some morphological studies, our analyses support Paenungulata. Resolution of relationships within Paenungulata remains one of the major challenges for future studies of interordinal relationships. Novel hypotheses that emerge from our analyses are Tubulidentata + Macroscelidea + Afrosoricida and Macroscelidea Afrosoricida. The latter clade also receives support from fetal membrane characters. Specifically, afrosoricidans and elephant shrews are the only afrotherians with haemochorial placentas (Carter, 2001).

Relationships Within the Euarchontoglires Clade

Our results agree with morphology in supporting Glires. In contrast, our results disagree with morphological studies that support the Archonta and Volitantia hypotheses. Rather, our analyses support an emended archontan clade that Waddell et al. (1999) dubbed Euarchonta. This hypothesis requires that characters associated with volancy are convergent in bats and flying lemurs. Indeed, advocates of the Archonta hypothesis have long recognized that bats lack tarsal features that occur in other archontan orders (Szalay and Drawhorn, 1980). Within Euarchonta, our analyses favor Scandentia + Dermoptera over competing hypotheses. In his analysis of morphological data that forced Chiroptera outside of Archonta, Sargis (2001) also found support for a sister-group relationship between Scandentia and Dermoptera.

Relationships in Laurasiatheria

Within Laurasiatheria, our analyses suggest a basal split between Eulipotyphla (moles, shrews, hedgehogs) and Variamana (i.e., Chiroptera + Perissodactyla + Cetartiodactyla + Pholidota + Carnivora). Mitochondrial studies are divided between those that are consistent with Variamana (Lin and Penny, 2001) and those that favor a sister-group relationship between bats and Eulipotyphla (represented by a mole) (Nikaido et al., 2001). Within Variamana, there is strong support for a sister-group relationship between carnivores and pangolins. Interestingly, these taxa are unique among living placentals in possessing an osseous tentorium (Shoshani and McKenna, 1998).

0 0

Post a comment