Reply to Ben-Dor et al.: Oral bacteria of Neanderthals and modern humans exhibit evidence of starch adaptation

We are pleased to see Ben-Dor et al.’s interest (1) in our study on the evolution and changing ecology of the hominid oral microbiome (2), which finds that starch-adapted oral bacteria are characteristic of the oral microbiota of Homo. However, in their critique, Ben-Dor et al. mischaracterize our findings by asserting claims not made in our article, and their conclusion that genetic adaptations in host microbiota have no implication for understanding hominin diets reveals misunderstandings about microbial evolution and the mechanisms of amylase-binding proteins.

Ben-Dor et al. dismiss out of hand our findings of bacterial starch adaptation in the Homo oral microbiome, falling back on unrelated arguments about the archaeological visibility of scavenging and human capacity for gluconeogenesis to promote an extreme vision of early Homo diets as carnivorous. Such arguments are largely theoretical and hypothetical and do not seriously engage with the microbial evidence presented in our study.

Specifically, we do not make quantitative claims about the frequency or amount of starch consumed during early Homo evolution, nor do we assert that starch alone was responsible for encephalization. Rather, we show that the oral microbiome of Homo is distinct from that of chimpanzees, gorillas, and howler monkeys in that it contains a high abundance of starch-adapted oral streptococci and is enriched in carbohydrate-processing genes related to oral streptococci. In particular, human oral streptococci possess genes enabling them to bind and utilize host salivary amylase for nutrient acquisition and dental colonization, a trait that is conferred by multiple independent genes (e.g., abpA, abpB, abpC) in an example of convergent evolution (3). Among these genes, abpA is noteworthy because it is only expressed in the presence of both starch and salivary amylase (4), and thus it only confers a selective advantage to the bacterium in the presence of dietary starches. We find that this apparent microbial coadaptation to dietary starches is not unique to modern humans but is also present in Neanderthals, and thus is consistent with the oral availability of starches in a common ancestor (5–7). We argue that this observation has implications for understanding earlier Homo diets and the energetics of encephalization, and thus warrants further study.

We understand that such findings are inconvenient for models of early Homo carnivory, but we reject Ben-Dor et al.’s argument that such microbial evidence should be ignored. Technical advancements in archaeological science are improving our ability to recover previously understudied and less conspicuous aspects of past diets, and are increasingly uncovering direct evidence of plant material (8–11). As such, our findings contribute to a growing body of evidence that earlier Homo diets may have been more varied than previously appreciated, and we look forward to future evidence-based studies that further enrich our understanding of human evolution.