Figure 1.
Recent morphological (A) and combined morphological + molecular (B) hypotheses of artiodactylan phylogeny.
Most cladistic analyses of morphological characters have supported monophyly of extant terrestrial artiodactylans, traditionally called Artiodactyla, as well as the subclades Suiformes and Selenodontia. Note the variable placements of the enigmatic extinct groups †Raoellidae and †Mesonychia in the different topologies. The deeply nested conflict between phylogenetic hypotheses for Artiodactyla is shown very well by these two recent studies: for the major lineages shown, no clades are shared. Lineages that connect extant taxa in the tree are represented by thick gray branches, and wholly extinct lineages are shown as thin black branches. Illustrations are by C. Buell and L. Betti-Nash.
Table 1.
Revised Nomenclature of Artiodactyl Taxa.
Figure 2.
Strict consensus of 20 minimum length trees for the equally-weighted parsimony analysis of the combined data set (57,269 steps).
The contents of 12 taxonomic groups, including the total clades Cetaceamorpha and Cetancodontamorpha are delimited by different colored boxes (‘Hippo’ = Hippopotamidamorpha). Lineages that connect extant taxa in the tree are represented by thick gray branches, and wholly extinct lineages are shown as thin black branches. Estimates of branch support scores are above internodes; given the complexity of the data set, these should be interpreted as maximum estimates. Illustrations are by C. Buell and L. Betti-Nash.
Figure 3.
Comparison of one minimum length tree with agreement subtree superimposed (A) and a topology that is two steps beyond minimum length (B).
Tree A is 57,269 steps; tree B is 57, 271 steps. Tree (A) shows one of twenty minimum length trees. Dashed branches in the minimum length topology connect to nine unstable taxa that were eliminated in the agreement subtree. Disregarding these nine taxa, relationships among the remaining 72 taxa in this tree are found in all 20 minimum length trees supported by the total evidence matrix. Tree (B) is two steps longer than minimum length. Note the highly discrepant positions of †Mesonychia and the †raoellid, †Indohyus, in the two trees. Small red squares at internal nodes mark clades that collapse with the movement of †Mesonychia from outside Artiodactyla (A) to within Cetaceamorpha (B). Taxonomic groups are delimited by colored boxes as in Figure 2. Illustrations are by C. Buell and L. Betti-Nash.
Figure 4.
Stability of phylogenetic results to the exclusion of particular taxa from the total combined data matrix.
Taxa deleted in each parsimony search are indicated above the phylogenetic result for each reanalysis. For simplicity, only the placements of major extant lineages and three critical fossil groups (†Mesonychia, †Creodonta, and †Raoellidae) are shown in the figure. Successive deletion of particular taxa from analysis results in contradictory interpretations of phylogenetic relationships. With the removal of Lipotyphla + †Creodonta, note that two equally parsimonious “islands” of trees are supported (†Mesonychia deep within Artiodactyla or completely outside the clade). †Creodonta is excluded from Artiodactyla in most reanalyses, but with the removal of Lipotyphla and Carnivora, †Creodonta clusters with †Mesonychia in a clade that is the sister group to Cetacea. Illustrations are by C. Buell and L. Betti-Nash.
Figure 5.
Strict consensus of the 48 minimum length trees for the equally-weighted parsimony analysis of 606 characters observable in fossils (3,722 steps).
Note that both Selenodontia (Ruminantia + Camelidae) and Suiformes (Hippopotamidae + Suina) are supported, in contrast to the total evidence analysis (Figure 2). Colored boxes that delimit taxonomic groups are as in Figure 2 (Hippo. = Hippopotamidamorpha).