Figure 9. Distance correlation results using the Gilbert et al. (2020) characters and the Hominidae/Proconsulidae taxon subset. Correlations and distance metrics are shown in the diagram. Filled squares indicate significant, positive distance correlation. Open circles indicate significant, negative correlation. (Hispanopithecus, Kenyapithecus, Pierolapithecus). Co-occurrence of other taxon pairs in the same cluster was considerably less decisive. For example, Turkanapithecus co-occurs most frequently with Simiolus but only in 67% of clustering partitions. Chimpanzee and gorilla nearly always appear in the same cluster, but they do not obviously prefer clustering with other taxa. Likewise, cercopithecoid Presbytis always appears in the same cluster as Victoriapithecus but the next most frequent clustering partner, Aegyptopithecus, appears in only 42% of clusters with Presbytis. DISCUSSION From an evolutionary perspective, conventional taxonomy of Miocene apes has always been challenging, with an array of fossil forms exhibiting blends of character traits that make their phylogenetic analysis difficult. This baraminological analysis appears to be no different in that regard. The only relatively clear result here comes from the gibbons of family Hylobatidae, where records of interspecific hybridization supplemented with mitochondrial DNA similarity link the extant species together in a single monobaramin. Previous studies of fossil hominins (Wood 2016, 2017) have consistently singled out the so-called robust australopithecines of genus Paranthropus as distinct from both humans and other apes, and Wood (2010) suggested this genus could be a holobaramin. Here, our analysis of Martin et al.’s (2021) character matrix yields moderate confirmatory support for this hypothesis. Paranthropus taxa form a cluster in all distance correlation analyses and in the three- and four-cluster medoid partitioning. Because Paranthropus taxa share significant, positive correlation with members of the human holobaramin and because they do not form their own cluster in the two-cluster medoid partition or fuzzy analysis, we judge these results alone to be only moderate support for recognizing them as a holobaramin. Coupled with previous analyses, however, we propose that Paranthropus will continue to exhibit putative discontinuity with other taxa in future analysis, consistent with Paranthropus being a holobaramin. Beyond these two observations, the picture is considerably less clear. No unambiguous clusters are recovered from any of the full character matrices from Pugh (2022), Gilbert et al. (2020), or Rasmussen et al. (2019), except for a group of outgroup taxa. In all three cases, however, reducing the taxon sample did allow a few clusters to be distinguished. For Pugh’s (2022) matrix, by examining only hominids, we found inconsistency in the distance correlation results, and the average silhouette widths for the cluster analyses were quite poor. Reducing the taxon sample to Gilbert et al.’s (2020) Hominidae and Proconsulidae, we found three relatively consistent clusters using distance correlation. The results of the three-cluster medoid partitions generally supported these three clusters, except for the positions of BRUMMEL AND WOOD Preliminary Evaluation of Ape Baramins 2023 ICC 158
RkJQdWJsaXNoZXIy MTM4ODY=