Reassessing hominoid phylogeny: evaluating congruence in the morphological and temporal data

Paleobiology, Fall 2004 by Finarelli, John A, Clyde, William C

The FAE of Sivapithecus is documented within the Chron C5Ar.1, making it stratigraphically lower than the latest possible estimate for the LAE of Griphopithecus. A range-through assumption to the end of the MN 6 would place the LAE of Griphopithecus locally at the Chron C5Ar.1r (Steininger et al. 1996). However, coding an overlap in this case would create a character state based on the correlation of the well-documented FAE of Sivapithecus to the terminus of a faunal zone for which a range-through assumption must be made. Conservatively, such an assumption of contemporaneity is not warranted. Similarly, Oreopithecus overlaps with the terminal portion of the Sivapithecus range. The V2 horizon at Baccinello is correlated to earliest MN 13 (Rook et al. 2000). If a similar range-through assumption to the end of MN 13 were made, Oreopithecus's LAE would extend beyond Sivapithecus. Again, this correlation compares a well-documented stratigraphic datum to the terminus of a faunal zone for which a range-through assumption is made, and coding of a distinct character state is unwarranted.

Cladistic Analysis of the Hominoidea

The Morphologically Most Parsimonious Cladograms

A cladistic analysis using the branch and bound algorithm in PAUP* (version 4.0b10 [Swofford 2002]) recovered four most parsimonious cladograms with respect to the morphological data. Each had a tree length of 447 steps and a Retention Index (RI) of 0.69 (Fig. 2A). This value for the RI is somewhat low when compared to other published RIs for studies that include fossil taxa. For example, Clyde and Fisher (1997: Table 2) compared RI values across 29 studies of varied taxonomic resolution. These studies had a median RI of 0.80, indicating that the degree of homoplasy observed in the Hominoidea is somewhat higher than for other taxonomic groups.

The strict consensus cladogram for the morphologically most parsimonious cladograms (hereafter referred to as the MMPC) highlights topological features common to all of the cladograms as well as areas of ambiguity (Fig. 2B). Proconsul is the sister taxon to all other ingroup taxa. Unfortunately, this does not resolve the debate over the phyletic position of Proconsul, because this position is consistent with its being either a basal hominoid (Andrews and Martin 1987a; Andrews 1992) or an undifferentiated Miocene catarrhine (Harrison 1987; Harrison and Sanders 1999). The most notable feature of the MMPC is the distinction between a postcranially primitive clade of early to middle Miocene hominoids from East Africa and a postcranially derived hominoid clade that includes both late Miocene Eurasian forms and all of the extant hominoid genera. Morotopithecus is joined to the base of the derived clade, supporting MacLatchy et al. (2000), who hypothesized the existence of two evolutionary distinct hominoid lineages in East Africa by 20 Ma on the basis of several derived features of the femur and axial skeleton. However, it is not the inclusion of Morotopithecus in this phylogenetic analysis that is responsible for the division of the hominoids into these two separate clades. The phylogenetic analysis was repeated on the data set while excluding Morotopithecus, and a set of eight cladograms was recovered, each distinguishing a monophyletic radiation of early to middle Miocene hominoids seen in the MMPC. The topologies of each of these eight cladograms were identical to the MMPC, except that the relative position of Proconsul varied across the set creating a polytomy between Proconsul and the two hominoid clades. Thus, early Miocene hominoids from East Africa form a distinct clade in the MMPC, and the early appearing Morotopithecus is allied with the postcranially derived clade of hominoids.