Why Was Possible African DNA Found at the Cahokia Mounds?

A disputed mitochondrial signal from Mound 72 raises a larger question about scientific method: when evidence is unexpected, should it be discarded before it is independently retested?

Estimated reading time: 6 minutes

An unexpected result is not automatically a discovery. It is also not automatically contamination.

That distinction stands at the center of a largely overlooked passage in Jessica Lou Harrison’s 2021 Indiana University dissertation, An Ancient DNA Perspective on Mound 72, Cahokia. The study revisited mitochondrial DNA from human remains recovered at Mound 72, one of the most significant mortuary locations at the ancient Indigenous city of Cahokia.

Harrison’s dissertation provides valuable criticism of earlier laboratory work. It also exposes a question that deserves more attention: if a genetic signal falls outside the lineages researchers expect to find, does that difference prove contamination, or does it create an obligation to test again?

The honest answer is that the reported signal remains unverified. It should not be promoted as proof of an African population at ancient Cahokia. Yet it should not be treated as scientifically resolved without direct replication.

What the Earlier Cahokia DNA Analysis Reported

Harrison reviewed a 2000 study by Nicole Napier involving DNA extracted from Mound 72 burials. Napier reported usable sequences for ten of twelve individuals. Five reportedly shared a transition at mitochondrial position 16309.

Napier interpreted that shared result as evidence of a high status maternal lineage. Harrison rejected that interpretation. She observed that 16309 was not a defining motif of the commonly recognized American Indian mitochondrial haplogroups used in the analysis and associated the transition with haplogroup L2a1, a lineage widely documented in Africa.

This is the point at which precision becomes essential.

The earlier study did not establish complete L2a1 mitochondrial genomes in five Cahokian individuals. It reported a short, partial sequence with a single shared transition. Full L2a1 identification normally depends upon a combination of mutations across the mitochondrial genome, not one position examined in isolation. Research on African mitochondrial diversity confirms that 16309 occurs within L2a1 patterns, but a partial signal is not equivalent to a fully authenticated haplogroup assignment.

DOCUMENTARY FACT: A 16309 transition was reported in five samples.

RESEARCH CAUTION: That transition alone does not prove that the individuals belonged to haplogroup L2a1.

Why Contamination Was a Serious Concern

Harrison had substantial reasons to criticize the 2000 methods. Her dissertation notes weaknesses in the earlier decontamination procedures, problems with the short DNA fragments, incomplete identifying mutations, and the possibility that endogenous DNA had become mixed with DNA introduced during handling or laboratory work.

Ancient DNA is especially vulnerable to contamination. Remains can be touched during excavation, conservation, storage, and repeated study. Modern human mitochondrial DNA can overwhelm the small amount of surviving ancient material. A result can therefore look real while originating from an excavator, curator, researcher, or laboratory environment.

Harrison’s contamination hypothesis is consequently reasonable. But “reasonable” is not the same as “demonstrated.”

No contamination source was identified for the five matching results in the passage under review. The shared sequence was not shown to match a named researcher or handler. More importantly, Harrison discussed the earlier results as methodologically unreliable, but did not report an independent replication designed specifically to authenticate or disprove the reported L associated signal in those five samples.

The Problem of Filtering by Expectation

The 2021 dissertation explains that five major mitochondrial haplogroups had been identified among sampled Indigenous populations in the Americas: A, B, C, D, and X. It then states that sequences outside those groups were classified as contamination and omitted from further analysis.

That decision may be practical for a narrowly designed population comparison, but it also presents a methodological danger. If “outside the expected groups” becomes the rule for identifying contamination, unexpected ancestry can never survive long enough to be investigated. The assumption begins controlling the evidence.

This does not mean every unusual sequence is authentic. It means the category “unexpected” cannot perform the work of a contamination test.

There is an instructive precedent. In 2007, Ripan Malhi and colleagues reported mitochondrial haplogroup M in two prehistoric North American individuals. Haplogroup M had not previously been reported in ancient or living Indigenous populations in the Americas. Rather than rejecting the result solely because it fell outside the familiar founding lineages, the researchers addressed authentication and argued that ancient American mitochondrial diversity may have been greater than previously recognized.

The lesson is not that the Cahokia result must also be genuine. The lesson is that unusual findings call for stronger testing.

What Harrison’s New Study Actually Established

Harrison assayed 65 people from Mound 72. Only 24 yielded mitochondrial sequences long enough for haplogroup determination. The dissertation reported haplogroups consistent with A, B, C, and D lineages and argued for broad regional population continuity.

Yet Harrison candidly documented the study’s limitations. The sequences were partial. Preservation was poor. Only one individual, Burial 28, was confirmed through a second extraction. Most sequences were not replicated, and later next generation libraries produced very little endogenous human mitochondrial DNA.

That transparency strengthens the dissertation. It also prevents readers from treating it as the final word on every anomalous sequence previously reported from the mound.

FIRST TRIBE ANALYSIS: Harrison presented a credible case that the earlier results were unreliable. She did not establish that every unexpected sequence was definitively contamination. Those are different conclusions.

What Responsible Retesting Would Require

A responsible follow up would not begin by declaring the 16309 signal African, American Indian, or contaminated. It would begin with authentication.

Where consultation, law, descendant authority, and research ethics permit, investigators would need new extracts from the same individuals, preferably from protected internal tooth material. Testing should be independently replicated in a separate ancient DNA laboratory. Researchers should document mitochondrial profiles from everyone who handled the remains or samples, examine molecular damage patterns expected in ancient DNA, use negative controls, and obtain substantially more mitochondrial coverage than the original short fragment.

The study should also test competing explanations. Does the sequence reproduce across separate extracts? Does it appear in laboratory controls? Does it match a known handler? Are the mutations phylogenetically sufficient for L2a1, another lineage, a mixed sequence, or no dependable assignment at all?

Until those questions are answered, the appropriate FTHRI evidence classification is:

UNVERIFIED: A reported genetic observation that has not been sufficiently authenticated or independently reproduced.

Unresolved Does Not Mean Proven, and It Does Not Mean Erased

There are two irresponsible ways to discuss this record. One is to announce that African DNA has been proven at ancient Cahokia. The available evidence does not support that claim. The other is to declare the signal meaningless merely because it falls outside an expected category. The available testing does not fully support that certainty either.

Science requires enough discipline to resist both temptations.

New information should survive skepticism, controls, and replication. It should not receive a free pass because it is exciting. It should also not receive an automatic death sentence because it is unfamiliar.

The courage of scholarship is not found in accepting every anomaly. It is found in asking the next testable question.

At Cahokia, that question remains open.

Source Notes

1. Jessica Lou Harrison, An Ancient DNA Perspective on Mound 72, Cahokia (PhD diss., Indiana University, 2021), especially 56–58, 86–90, and 123. ProQuest open view.

2. Indiana University Department of Anthropology, Spring Graduation 2021, Jessica Harrison.

3. Antonio Salas et al., “The Making of the African mtDNA Landscape,” American Journal of Human Genetics 71, no. 5 (2002): 1082–1111. Full text.

4. Ripan S. Malhi et al., “Mitochondrial Haplogroup M Discovered in Prehistoric North Americans,” Journal of Archaeological Science 34, no. 4 (2007): 642–648. University of Illinois record.

5. Rosa Montiel et al., “Authenticating Ancient Human Mitochondrial DNA,” Human Biology 73, no. 5 (2001): 689–713. PubMed record.