The advent of metagenomic analysis of microbial communities has led to some unexpected insights about human biology. These techniques have quietly been leading to new discoveries from old archaeological contexts. One example is Alan Cooper's work demonstrating long-term changes in oral microbiota from ancient dental calculus ("Tracing teeth troubles with fossil bacteria").
Another is a recent paper from Cecil Lewis' lab, "Insights from characterizing extinct human gut microbiomes."
Coprolites may seem simple, but each represents a unique history of deposition and subsequent preservation. The microbial community may shift during the early stages of this history, and subsequent DNA damage may shift estimates of microbial abundances away from their true values. They found one of their sites appeared to preserve a good signal, while the others were degraded:
Most striking, both Rio Zape coprolites exhibited a gut microbiome signature with similarities to the children from a rural African village with the exclusion of a sample of U.S. modern adult gut microbiomes (see Figure S4 for a heat map of these data and Figure S5 for the variability in the source proportion estimates). ZA04 also harbored similarities to non-human primate gut. The coprolites from Caserones and Hinds Cave showed little similarity to a gut microbiome environment. A portion of Caserones coprolite microbial community was similar to compost, which may be explained by the post-mortem gut serving as an organic bioreactor filled with carbon and nitrogen from decaying food detritus. The microbial community assignment for Hinds Cave failed to assign well to any source environment.
From this, we can see that any interpretation of data from a sample of ancient coprolites must be cautious. We're generally interested in how microbial communities may have changed in ancient populations, particularly in response to other factors such as shifts in diet. But as yet it's not very clear what kinds of changes we should predict in association with diet or other changes. That makes it hard to develop a convincing test of a hypothesis.
This paper is more of a proof of principle. And in its discussion, Tito and colleagues present different ways to explain the kinds of differences that they found in the ancient coprolite microbiota. To me, the most provocative hypothesis is that changes may have more to do with parasite load than diet:
Information from Rio Zape also supports a current hypothesis about the composition of human microbiomes in traditional communities, potentially revealing an important aspect of the ancestral human microbiome. Spirochaetes are atypical of gut microbiomes in cosmopolitan communities. However, Treponema was reported by Filippo et al.  in their comparative study of modern microbiota in children from Europe and rural Africa. In their study, Treponema was observed in the rural African children but was absent in the European children. They hypothesized that the Treponema may enhance the hosts ability to extract nutrients from fibrous foods and may provide anti-inflammatory capability. They raise the hypothesis that microbiota coevolved with ancient diets and that changes in food production greatly impacted the intestinal microbiota. Treponema was also observed in the published rural data for Malawi and Venezuela . The results from Rio Zape provide further support for Treponema as part of the rural human microbiome. Specifically, Treponema now is observed in four rural communities from different continents, three extant communities and one community that has been extinct for over a thousand years.
As we uncover more comparative data from living people, we will begin to have a better picture of the covariates of microbial community structure. Today's oral bacterial populations in "cosmopolitan" post-industrial peoples are uncharacteristic of past variation. The gut microbiota of cosmopolitan peoples may be just as uncharacteristic. The diversity may have had great importance to ancient health, especially at key times when pathogens were spreading through post-agricultural populations.