Tooth size apportionment, Bayesian inference, and the phylogeny of Homo naledi

This study has three main objectives--two methodological and one summative, namely, further characterization of Homo naledi ([~]335-236 ka) to more firmly establish its evolutionary history. Using mathematically-corrected mesiodistal and buccolingual crown dimensions, the species was compared with samples of Pan troglodytes, Australopithecus africanus, A. afarensis, Paranthropus robustus, P. boisei, H. habilis, H. ergaster, H. erectus, H. heidelbergensis, H. neanderthalensis, and H. sapiens; the correction yields equivalently scaled samples unaffected by significant interspecific size differences. After initial cluster analysis, the data were used in tooth size apportionment analysis to determine how size is distributed relatively in each species dentition, while visualizing this variation in a sample scatterplot. The first main objective then, after quantitative coding, is evaluating the utility of these characters to estimate phylogenetic relationships, here using Bayesian inference with an Mkv model. The second objective, for the first time in paleoanthropological study, is estimating relationships using continuous characters, i.e., the scaled data, through Bayesian inference under a Brownian-motion model. This strategy facilitates maximum reception of potential phylogenetic signal. The final objective based on all analyses, though principally continuous Bayesian inference, is to elucidate the phylogeny of H. naledi. Relationships are largely congruent across methods and, with markedly higher node support, most of those inferred in prior systematic studies using qualitatively discretized traits. The present results place H. naledi as a sister taxon to H. habilis (node support [~]70-99%), with a plesiomorphic pattern of relative tooth size. It is nested within a clade comprising australopiths and early Homo dating 3.3 Ma to [~]800 ka, distinct from younger H. erectus through H. sapiens. This suggests that H. naledi originated well before the geological date range associated with the Dinaledi Chamber, from which the remains in this study were recovered, to represent a long-lived side branch in the genus.