Sex-biased sampling may influence Homo naledi tooth size variation

A frequent source of debate in paleoanthropology concerns the taxonomic unity of fossil assemblages, with many hominin samples exhibiting elevated levels of variation that can be interpreted as indicating the presence of multiple species. By contrast, the large assemblage of hominin fossils from the Rising Star cave system, assigned to Homo naledi, exhibits a remarkably low degree of variation for most skeletal elements. Many factors can contribute to low sample variation, including genetic drift, strong natural selection, biased sex ratios, and sampling of closely related individuals. In this study, we tested for potential sex-biased sampling in the Rising Star dental sample. We compared coefficients of variation for the H. naledi teeth to those for eight extant hominoid samples. We used a resampling procedure that generated samples from the extant taxa that matched the sample size of the fossil sample for each possible Rising Star dental sex ratio. We found that variation at four H. naledi tooth positions-I2, M1, P4, M1-is so low that the possibility that one sex is represented by few or no individuals in the sample cannot be excluded. Additional evidence is needed to corroborate this inference, such as ancient DNA or enamel proteome data, and our study design does not address other potential factors that would account for low sample variation. Nevertheless, our results highlight the importance of considering the taphonomic history of a hominin assemblage and suggest that sex-biased sampling is a plausible explanation for the low level of phenotypic variation found in some aspects of the current H. naledi assemblage.