How did the researchers read the microbiomes?

The team used a technique called deep metagenomic sequencing. In plain terms, the method reads all the DNA letters in a stool sample, generating millions of small DNA sequences. Overlapping stretches of letters show where the short fragments match up into longer sequences, which researchers then compare against databases of known microbial genomes to identify exactly which organisms are present.

The Tsimane Health and Life History Project team, whose members collaborated on the study, had collected voluntarily provided stool samples from the Tsimane. Sonnenburg and colleagues had previously performed metagenomic sequencing on Hadza samples, reported in a 2023 Cell study. That earlier work showed the average Hadza individual carries about 750 microbial species, while the average Californian carries just 250.

For the new study, the researchers performed the first deep sequencing of the Tsimane samples, which had earlier been sequenced only at low resolution in a 2020 Nature Communications study. The effort produced a comprehensive census of microbes, including hard-to-capture, low-abundance species. Overall, the sampled Tsimane individuals hosted about 1,400 different species, and they shared 1,231 of those—about 90%—with the Hadza.

"We were really surprised to see that the vast majority of the species in the Tsimane's microbiome correspond with the Hadza's," Sonnenburg said.

Different diets, overlapping microbes

The surprise is real. The Hadza hunt mammals such as impala and porcupine, along with birds and fish, and forage for tubers, berries, fruits and vegetables. The Tsimane grow much of their own food—plantains, rice, manioc root and corn—and eat lean meat from fish, peccary and other forest animals.

Two groups with such different diets, separated for tens of thousands of years, still carry largely the same microbial species.

"That made us wonder if there really is a core set of bacterial and other species that traveled with the ancestors of the Tsimane as they migrated around the globe," said Benjamin Good, Ph.D., the study's senior co-author and an assistant professor of applied physics who studies the evolutionary dynamics and population genetics of the human gut microbiome.

"Could these bugs have been in continual interaction with us since long before the ancestors of today's Hadza and Tsimane became geographically separated?" Sonnenburg asked. "To answer that, we turned to Ben's team, and what they found blew our minds."

How do you date a bacterium's family tree?

Reconstructing the evolutionary history of a microbiome is difficult. Microbes evolve quickly, and a single person's microbiome can shift from season to season with changing food availability. Bacteria also complicate matters because they reproduce by cloning but frequently swap DNA with other strains—a process called horizontal gene transfer.

To get around these obstacles, Good's group looked for several independent genomic signatures that could distinguish deep shared ancestry from recent microbial exchange. They also examined mutation rates in vertically inherited DNA. Because mutations accumulate at a steady pace, like ticks of a clock, they serve as a rough timer for when species diverged from a common ancestor.

The analysis indicated that many bacterial species shared between the two nonindustrialized groups have evolutionary histories stretching back thousands of years, with time estimates consistent with the major human migrations.

"We see evidence that these bacteria were present in the ancestors of both the Tsimane and the Hadza, and therefore presumably the ancestors of all of us," Good said.

What does this mean for health?

The findings sharpen an important hypothesis: if humans interacted with many of these organisms for tens of thousands of years, what happens when industrialization eliminates a large portion of them in just a few generations? Previous work in the field links microbiome changes to many disease states.

The researchers point to dramatic shifts in our resident microbial populations attributed to antibiotic exposure; low-fiber, high-fat, high-sugar processed diets; highly sanitized living conditions; and other factors. These changes, they argue, could be incompatible with the biology encoded in our human genomes. The biological consequences of this recent biodiversity loss remain an open question and a target for future study.

"If these microbes hosted by the Hadza and Tsimane really are part of our evolutionary biology, then those of us in industrialized countries are missing a huge part of what our human genome has potentially adapted to over a vast time period," Sonnenburg said. "We're interested in more fully exploring these long-standing relationships between humans and our gut microbes."

The Stanford team plans to continue investigating the history of our microbiome and its links to health and well-being.