Tempo and mode of evolution in an orthologous Can SINE.

Tempo and mode of nucleotide change were examined in an orthologous carnivoran nuclear repetitive DNA element (Can SINE), and compared with those of the transthyretin intron I (TR-i-I) sequence in which it is embedded, by using a phylogenetic framework. The Can SINE is found in representatives of all living caniform carnivoran families, but no living feliform families. This suggests insertion 40-65 MYA, after the two lineages split, but before the caniform radiation. Despite linkage and a long shared evolutionary history, both parsimony and likelihood analyses showed the Can SINE to be significantly different from TR-i-I in rates of evolution and phylogenetic hypotheses supported. The substitution rate is significantly higher in Can SINE than in TR-i-I, and this is attributable to the tRNA-related region of the insertion. While the incongruence length difference test revealed significant conflict between the Can SINE and TR-i-I partitions, the test was shown to be sensitive to the distribution of homoplasy within partitions. The conflicting phylogenies are likely the result of differences in phylogenetic accuracy (homoplasy distribution) rather than in phylogenetic history (gene trees). The base composition of Can SINE contains a significantly higher GC percentage than TR-i-I. Our results indicate that differences between the two partitions may be the result of homoplasy introduced by an increased substitution rate in the tRNA-related region of Can SINE owing to CpG hypermutability.

Whence the red panda?

The evolutionary history of the red panda (Ailurus fulgens) plays a pivotal role in the higher-level phylogeny of the "bear-like" arctoid carnivoran mammals. Characters from morphology and molecules have provided inconsistent evidence for placement of the red panda. Whereas it certainly is an arctoid, there has been major controversy about whether it should be placed with the bears (ursids), ursids plus pinnipeds (seals, sea lions, walrus), raccoons (procyonids), musteloids (raccoons plus weasels, skunks, otters, and badgers [mustelids]), or as a monotypic lineage of uncertain phylogenetic affinities. Nucleotide sequence data from three mitochondrial genes and one nuclear intron were analyzed, with more complete taxonomic sampling of relevant taxa (arctoids) than previously available in analyses of primary molecular data, to clarify the phylogenetic relationships of the red panda to other arctoid carnivorans. This study provides detailed phylogenetic analyses (both parsimony and maximum-likelihood) of primary character data for arctoid carnivorans, including bootstrap and decay indices for all arctoid nodes, and three statistical tests of alternative phylogenetic hypotheses for the placement of the red panda. Combined phylogenetic analyses reject the hypotheses that the red panda is most closely related to the bears (ursids) or to the raccoons (procyonids). Rather, evidence from nucleotide sequences strongly support placement of the red panda within a broad Musteloidea (sensu lato) clade, including three major lineages (the red panda, the skunks [mephitids], and a clearly monophyletic clade of procyonids plus mustelids [sensu stricto, excluding skunks]). Within the Musteloidea, interrelationships of the three major lineages are unclear and probably are best considered an unresolved trichotomy. These data provide compelling evidence for the relationships of the red panda and demonstrate that small taxonomic sample sizes can result in misleading or possibly erroneous (based on prior modeling, as well as conflict between the results of our analyses of less and more complete data sets) conclusions about phylogenetic relationships and taxonomy.

Evidence from intron 1 of the nuclear transthyretin (Prealbumin) gene for the phylogeny of African mole-rats (Bathyergidae).

A 900- to 1100-bp fragment encompassing intron 1 of the nuclear transthyretin (prealbumin) gene was examined in 12 taxa of Old World hystricognath rodents of the families Bathyergidae, Petromuridae, Thryonomyidae, and Hystricidae. Within the Bathyergidae, Heterocephalus glaber (naked mole-rat) was basal, and the other East African species, Heliophobius argenteocinereus (silvery mole-rat), was sister to a southern African clade containing Bathyergus, Cryptomys, and Georychus (dune, common, and cape mole-rats). These results are congruent with studies using mitochondrial 12S rRNA gene sequences. A combined analysis of transthyretin and 12S rRNA data resulted in a well-supported topology with better resolution than either gene analyzed separately. These data support the findings by M. W. Allard and R. L. Honeycutt (1992, Mol. Biol. Evol. 9: 27-40) and R. L. Honeycutt (1992, Am. Sci. 80: 43-53) that complex social systems evolved independently at least twice, in the common and naked mole-rats.

Phylogeny of the Carnivora (Mammalia): congruence vs incompatibility among multiple data sets.

The purpose of this study was to determine the higher-level phylogenetic relationships among Carnivora, using a conditional data combination (CDC) approach to analyzing multiple data sets. New nucleotide sequences (851 base pairs from intron I of the transthyretin gene) among 22 representatives of the 11 families of Carnivora were generated and analyzed in concert with, and comparison to, other mitochondrial and morphological character data. Conditional data combination analyses of the four independent data sets (transthyretin intron I, cytochrome b, partial 12S rRNA, and morphology) indicate that the phylogenetic results derived from each generally agree, with two exceptions. The first exception, signal heterogeneity in comparisons involving transthyretin and morphology, provides an example where phylogenetic conclusions drawn from total evidence analyses may differ from conclusions drawn from CDC analyses. The second exception demonstrates that while a CDC method may reject the null hypothesis of homogeneity for a particular partition, including that partition in combined analyses, may nevertheless provide an overall increase in phylogenetic signal, in terms of nodal support for most associations, without altering the topology derived from the combined homogeneous data partitions. Phylogenetic reconstruction among the feliform families supports a sister-group relationship between the hyaenas (Hyaenidae) and mongooses (Herpestidae) and places the African palm civet (Nandinia) as basal to all other living Feliformia. Among the caniform families, CDC analyses strongly support the previously enigmatic red panda (Ailurus) as a monotypic lineage that is sister to Musteloidea sensu stricto (mustelids plus procyonids), in addition to pinniped monophyly and a sister-group relationship between the walrus and sea lions.

Mammalian mitochondrial DNA evolution: a comparison of the cytochrome b and cytochrome c oxidase II genes.

The evolution of two mitochondrial genes, cytochrome b and cytochrome c oxidase subunit II, was examined in several eutherian mammal orders, with special emphasis on the orders Artiodactyla and Rodentia. When analyzed using both maximum parsimony, with either equal or unequal character weighting, and neighbor joining, neither gene performed with a high degree of consistency in terms of the phylogenetic hypotheses supported. The phylogenetic inconsistencies observed for both these genes may be the result of several factors including differences in the rate of nucleotide substitution among particular lineages (especially between orders), base composition bias, transition/transversion bias, differences in codon usage, and different constraints and levels of homoplasy associated with first, second, and third codon positions. We discuss the implications of these findings for the molecular systematics of mammals, especially as they relate to recent hypotheses concerning the polyphyly of the order Rodentia, relationships among the Artiodactyla, and various interordinal relationships.

Molecular systematics of hystricognath rodents: evidence from the mitochondrial 12S rRNA gene.

Nucleotide sequence variation among 22 representatives of 14 families of hystricognathid rodents was examined using an 814-bp region of the mitochondrial 12S ribosomal RNA (rRNA) gene composing domains I-III. The purpose of this study was twofold. First, the phylogenetic relationships among Old World phiomorph (primarily African) and New World caviomorph (primarily South American) families were investigated, with a special emphasis on testing hypotheses pertaining to the origin of New World families and the identification of major monophyletic groups. Second, divergence times derived from molecular data were compared to those suggested by the fossil record. The resultant 12S rRNA gene phylogeny, analyzed separately and in combination with other morphological and molecular data, supported a monophyletic Caviomorpha. This finding is counter to the idea of a multiple origin for the South American families. The most strongly supported relationships within the Caviomorpha were a monophyletic Octodontoidea (containing five families) and the placement of New World porcupines (family Erethizontidae) as the most divergent family. Although comparisons to other data were more equivocal, the most parsimonious 12S rRNA trees also supported a monophyletic Phiomorpha that could be subdivided into two major groups, a clade containing the Thryonomyoidea (Thryonomyidae and Petromuridae) plus Bathyergidae and the more divergent Hystricidae (Old World porcupines). No significant differences in rates of 12S rRNA gene divergence were observed for hystricognathids in comparison to other rodent groups. Although time since divergence estimates were influenced by the fossil dates chosen to calibrate absolute rates, the overall divergence times derived from both transversions only and Kimura corrected distances and calibrations using two independent dates revealed a divergence time between Old and New World groups dating in the Eocene.