Taxonomic differences in deciduous lower first molar crown outlines of Homo sapiens and Homo neanderthalensis

Recent studies have demonstrated that the outline shapes of deciduous upper and lower second molars and the deciduous upper first molar are useful for diagnosing hominin taxa-especially Homo neanderthalensis and Homo sapiens. Building on these studies, we use geometric morphometric methods to assess the taxonomic significance of the crown outline of the lower first deciduous molar (dm₁). We test whether the crown shape of the dm₁ distinguishes H. neanderthalensis from H. sapiens and explore whether dm₁ crown shape can be used to accurately assign individuals to taxa. Our fossil sample includes 3 early H. sapiens, 7 Upper Paleolithic H. sapiens, and 13 H. neanderthalensis individuals. Our recent human sample includes 103 individuals from Africa, Australia, Europe, South America, and South Asia. Our results indicate that H. neanderthalensis dm₁s cluster fairly tightly and separate well from those of Upper Paleolithic H. sapiens. However, we also found that the range of shapes in the recent human sample completely overlaps the ranges of all fossil samples. Consequently, results of the quadratic discriminant analysis based on the first 8 principal components (PCs) representing more than 90% of the variation were mixed. Lower dm1s were correctly classified in 87.3% of the individuals; the combined H. sapiens sample had greater success (90.2%) in assigning individuals than did the H. neanderthalensis sample (61.5%). When the analysis was run removing the highly variable recent human sample, accuracy increased to 84.6% for H. neanderthalensis, and 57.1% of Upper Paleolithic H. sapiens were classified correctly by using the first 4 PCs (70.3%). We conclude that caution is warranted when assigning isolated dm₁ crowns to taxa; while an assignment to H. neanderthalensis has a high probability of being correct, assignment to Upper Paleolithic H. sapiens is less certain.