Ancient DNA | Colonial Mexico | The African heritage of New Spain
Who were the first Africans of Mexico City?
In 1992, construction of Metro Line 8 uncovered a sixteenth century mass grave at a hospital for the Indigenous people of Mexico City. Three of the men buried there had been born in Africa. Their genomes, their teeth and the pathogens they carried tell part of the story of the first generation of Africans in New Spain. We placed them, and a woman buried in Campeche, in Global25.
Hero image: AI generated illustration of a street beside a Franciscan hospital in Mexico City around 1550, with African, Indigenous and Spanish residents of the new colonial city. Buildings and clothing are approximations, and the scene is not a reconstruction of a documented moment. The figures do not portray the individuals whose genomes are discussed here.
Mexico's African heritage is often called its "third root", alongside its Indigenous and European ancestry. For a long time it was also the least visible. Yet in the sixteenth century, Africans were present in New Spain from the very first years after the conquest, in the mines, on the sugar estates, in the workshops and households of the new capital.
Historical records tell us how many people were brought and from which ports. They rarely tell us who those people were. In 2020, a study of three men buried in Mexico City gave some of them back a place of origin and a life history.
A hospital for the "naturales"
The Hospital Real de San Jose de los Naturales was founded around 1530 in the new Spanish city built over Tenochtitlan. As its name says, it was intended for the "naturales", the Indigenous population. In 1992, rescue excavations for Metro Line 8, near the San Juan de Letran station in the historic centre, uncovered a mass grave in its cemetery. The researchers suggested that the grave may reflect one of the early epidemics of New Spain.
Among the remains, three skulls stood out. The upper front teeth of all three men had been deliberately filed into shapes, a practice recorded among Africans brought to the Americas and still found in some groups in western Africa. It was the first clue that they were not, in fact, "naturales".
Three men born in Africa
In 2020, Rodrigo Barquera, Johannes Krause and colleagues from the Max Planck Institute and the Escuela Nacional de Antropologia e Historia published the genomes, isotopes and pathogens of the three men in Current Biology. They are known as SJN001, SJN002 and SJN003, from burials 150, 214 and 296.
All three died between about 25 and 35 years of age. Strontium isotopes in their teeth show that they were not born in the Americas and most likely spent their childhood in Africa. All three carried a Y-chromosome lineage of haplogroup E-M2, which is very common in West and Central Africa and today is the most frequent paternal lineage among African Americans. Their genomes showed no sign of Indigenous or European admixture, and the authors concluded that their ancestry pointed to either western or southern Africa.
Their skeletons also record hard lives. The researchers describe signs of heavy physical labour, nutritional stress and old injuries, including fractures of the skull and legs. One man still carried five pieces of copper shot in his chest, which had stained the surrounding bones green. All these injuries had healed: they did not cause death. The cause of death of the three men is unknown.
Pathogens from across the Atlantic
From their teeth, the team reconstructed two complete pathogen genomes, both closest to strains found in Africa today. One man carried a hepatitis B virus strain of a type found today in western Africa, the first direct evidence that the virus crossed the Atlantic with the slave trade. Another carried Treponema pallidum pertenue, the bacterium of yaws, a chronic skin and bone infection. Both infections were probably acquired in Africa. The authors found no sign that this hepatitis B strain became established in Mexico.
A note on dates. Radiocarbon dates for the three men span 1436 to 1626. The modelled ranges are 1453 to 1626 for SJN001, 1450 to 1620 for SJN002 and 1436 to 1472 for SJN003. The last range cannot be literally correct, since the hospital did not exist before about 1530. Radiocarbon dates from this period are imprecise and can be shifted by diet, so all three men are best treated as sixteenth century.
Where in Africa?
The trade that brought them was large. About 36,500 enslaved Africans reached New Spain between 1521 and 1594, and more than 50,000 between 1595 and 1622. Historians of the early trade to Spanish America, such as David Wheat, have shown that in the late sixteenth and early seventeenth centuries most captives came from two broad regions: Upper Guinea, from Senegal to Sierra Leone, and West Central Africa, especially Angola and the Kongo region.
Can the genomes say which region each man came from? The 2020 study was cautious. Global25 offers a simple way to test it, with a large panel of African populations, most of them present-day.
Our test: placing four people in Africa
The three men from Mexico City are in the public Global25 files, with coverage of 63 to 74%. We added a fourth person: a woman in her early twenties buried in the early colonial cemetery of Campeche, on the Gulf of Mexico, which was used from the mid sixteenth to the late seventeenth century. In 2022, Vera Tiesler, Jakob Sedig, David Reich and colleagues showed in Antiquity that she had African ancestry with no European or Indigenous admixture, and isotopes indicating that she had probably not grown up in Yucatan.
We compared the four genomes with 24 African reference populations, from Senegambia and Sierra Leone to Nigeria, Cameroon, the Congo, Angola, Namibia, Mozambique and Kenya, including two people buried in the eighteenth century at Kindoki, in the former Kongo kingdom. A single ancient genome of this quality lies about 20 to 25 units from the average of its own population (distances multiplied by 1,000), which sets the scale for what counts as close.
| Person | Closest populations (distance) | Region | Non-African ancestry |
|---|---|---|---|
| SJN001 Mexico City, burial 150 | Mende, Sierra Leone (23); Jola, Gambia (29); Akan, Ghana (30) | Upper Guinea | 0.7% |
| SJN002 Mexico City, burial 214 | Ovambo, Namibia (20); Umbundu, Angola (20); Kongo kingdom, 18th c. (21) | West Central Africa | 0.7% |
| SJN003 Mexico City, burial 296 | Umbundu, Angola (18); Bantu speakers, Cameroon (20); Kongo (20) | West Central Africa | 0.5% |
| Campeche woman Campeche, burial 11-1 | Kwangali, Namibia (26); Kongo kingdom, 18th c. (26); Ovambo, Namibia (27) | West Central Africa | 0.3% |
Distances in Global25 units multiplied by 1,000. Non-African ancestry: Spanish and colonial Maya sources combined, in a four-source model with Umbundu and Mende (condition number 31).
The answer splits the three men. SJN001 is closest to the Mende of Sierra Leone, and to other populations of the Upper Guinea coast. SJN002 and SJN003 are closest to Bantu-speaking populations of Angola, the Congo and northern Namibia, and to the Kongo kingdom of the eighteenth century. They are also close to each other, 26 units apart, while both lie 50 to 57 units from SJN001.
This matches the two main sources of the early trade remarkably well, and it is consistent with the cautious wording of the 2020 study, which pointed to western or southern Africa.
How reliable is the regional assignment?
African populations are close to each other in Global25, so we tested how often a single genome is placed in the right region. We took each reference average, added the noise of a real single ancient genome, and asked which population was closest, 2,000 times for each. The exact population was right in 53 to 100% of runs depending on the population, but the broad region in more than 99%. A simple measure makes the point: we subtracted the distance to the nearest West African population from the distance to the nearest West Central or southern one. In 95% of runs, simulated genomes from Sierra Leone, Gambia and Nigeria scored between +6 and +33, and simulated genomes from Cameroon, the Congo, Angola, Namibia and Mozambique between -50 and -12. SJN001 scores +15, SJN002 and SJN003 score -26 and -24, and the Campeche woman -21.
For comparison, we ran the same test on 18 people buried at Rupert's Valley, on Saint Helena, Africans freed from slave ships intercepted by the Royal Navy between 1840 and 1867. All 18 fall on the West Central and southern side, in line with the published study of these genomes, which traced most of them to the region between northern Angola and Gabon.
No local ancestry
We then tested whether any of the four carried Indigenous Mexican or Spanish ancestry, which would point to a parent or grandparent born in the Americas. Spanish and Maya ancestry are so different from African ancestry in Global25 that even a small share is easy to detect. In simulated African genomes, a true 0% came back at no more than 1.1%, one great-great-grandparent (6.25%) at 5.3 to 7.4%, and one great-grandparent (12.5%) at 11.6 to 13.6%. All four people come out between 0.3 and 0.7%, below the detection floor. Together with the isotopes, this confirms that they were first-generation Africans, born and raised in Africa.
The African root of Mexico today
The Africans of the sixteenth century did not disappear. Over the following centuries, many of their descendants married into Indigenous and European families, and African ancestry became part of the wider Mexican population. In our Global25 model, the average of 131 present-day Mexican mestizos in the Global25 collection carries about 4% African ancestry, against about 11% for Puerto Ricans, 32% for Brazilian pardos and 78% for African Americans. These averages hide large differences between individuals and regions: Afro-Mexican communities, especially on the Costa Chica of Guerrero and Oaxaca and in Veracruz, are not represented in the dataset. In the 2020 census, about 2.5 million people in Mexico identified as Afro-Mexican or of African descent, about 2% of the population.
Two traps in the data
Trap 1: the region, not the people
Global25 places each person in a broad region with confidence, but not in a specific population. SJN001 is closest to the Mende, but the Mende were not necessarily his people: the panel contains only a few populations of the Upper Guinea coast, and the groups that lived there in the sixteenth century were not identical to those of today. The same applies to the Umbundu and the Ovambo for SJN002 and SJN003. The reference populations stand for regions, not for ethnic origins.
Trap 2: present-day references for a sixteenth century trade
Almost all the reference populations are present-day groups, shaped by five centuries of further history, including the slave trade itself. Only the two individuals from Kindoki give a historical reference for West Central Africa. More ancient genomes from Upper Guinea and Angola would make the comparison much sharper.
What this means
Three men buried in a hospital for the Indigenous people of Mexico City were born in Africa, one in the Upper Guinea region and two in West Central Africa, the two great sources of the early trade to New Spain. They spent their childhood in Africa, carried diseases from their homelands, worked hard, survived serious injuries and died young. A woman buried in Campeche, on the Gulf coast, shared the West Central African origin of two of them. Their genomes give a face to the first generation of the African root of Mexico.
Key points
- Three men buried in a sixteenth century mass grave at the Hospital Real de San Jose de los Naturales, in Mexico City, were born in Africa.
- They had filed teeth, Y-chromosome lineages of haplogroup E-M2, and isotopes showing a childhood outside Mexico; all died between about 25 and 35.
- Their teeth gave the earliest evidence of hepatitis B and yaws in human remains from the Americas, both of African type.
- Our Global25 test places one man in Upper Guinea and two in West Central Africa, matching the two main sources of the early trade; a woman buried in early colonial Campeche also falls in West Central Africa.
- None of the four carries detectable Indigenous or Spanish ancestry: they were first-generation Africans.
- About 2.5 million Mexicans identified as Afro-Mexican or of African descent in the 2020 census.
Global25 coordinates used in this article
The model sources, the three men from Mexico City and the woman from Campeche. They can be pasted directly into the Vahaduo tool.
Source_Mende_Sierra_Leone,-0.625524,0.062750,0.021382,0.017168,0.000379,0.011198,-0.024993,0.028314,-0.040938,0.029344,0.005468,-0.004547,0.018353,0.000562,0.011634,-0.008911,0.008554,0.006673,0.000875,-0.001287,0.000247,0.000482,-0.001061,-0.000863,0.000337
Source_Umbundu_Angola,-0.626255,0.063267,0.021741,0.016053,0.001477,0.011658,-0.014512,0.023318,-0.033061,0.016884,0.001965,-0.000300,-0.002728,0.000537,-0.009351,0.006252,-0.005620,0.003003,-0.004582,0.002063,0.000262,0.000779,0.001676,-0.000108,-0.000180
Source_Kongo_Rep_Congo,-0.627734,0.062455,0.022250,0.016958,-0.005539,0.011853,-0.016216,0.025153,-0.035894,0.020957,0.005522,-0.001274,-0.000149,0.000000,-0.007668,0.007624,-0.004889,-0.000190,0.000691,0.003377,-0.002558,-0.000309,-0.000185,-0.000602,-0.008143
Source_Spanish_Andalucia,0.108132,0.146076,0.035707,-0.003434,0.045256,-0.005857,-0.001930,0.002660,0.024715,0.031162,-0.002154,0.007272,-0.013207,-0.011408,0.010243,-0.002498,-0.006176,-0.001100,-0.004261,-0.002133,0.002929,-0.001302,-0.001972,-0.005334,0.001885
Source_Mexico_Campeche_Colonial_Maya_mean,0.051708,-0.316845,0.113567,0.093024,-0.111449,-0.014662,-0.291346,-0.343008,-0.012798,-0.014787,-0.002459,0.001070,0.001678,0.017360,-0.007213,0.007520,0.004917,-0.001412,-0.000485,-0.001840,-0.000321,0.005105,0.000176,0.002341,-0.002241
Mexico_City_SanJose_Naturales_SJN001,-0.621474,0.060932,0.016970,0.015827,-0.007078,0.005857,-0.031491,0.027230,-0.037837,0.032620,0.013965,-0.006744,0.020664,-0.005367,0.006515,-0.009944,0.005215,0.012922,0.007793,-0.007754,-0.002496,0.000742,-0.003574,-0.007712,-0.003113
Mexico_City_SanJose_Naturales_SJN002,-0.623751,0.054839,0.027907,0.015181,0.001231,0.015897,-0.011281,0.018922,-0.035383,0.015308,0.003897,-0.003747,-0.008771,-0.002615,-0.004072,0.009679,0.004172,-0.001014,-0.005782,0.000250,0.001747,-0.001731,-0.001109,0.002651,0.002036
Mexico_City_SanJose_Naturales_SJN003,-0.624889,0.064994,0.026021,0.013243,-0.004001,0.011992,-0.019271,0.018230,-0.041109,0.020957,0.000162,-0.004196,-0.004906,0.004542,-0.009908,0.001856,-0.006258,0.002280,-0.007290,0.003377,-0.002870,-0.002349,-0.004067,-0.003976,-0.000359
Mexico_Campeche_Colonial_African_I10860,-0.624889,0.058901,0.021873,0.016473,0.004616,0.006414,-0.002350,0.014769,-0.031292,0.015855,-0.001461,0.001349,0.010852,0.001101,0.001629,0.016971,-0.005215,0.002660,0.000000,-0.000250,0.000250,-0.003462,0.006655,-0.003735,-0.001437
References
- Ancient DNA Barquera R, Lamnidis TC, Lankapalli AK, Kocher A, Hernandez-Zaragoza DI, Nelson EA, et al., Kuhnert D, Marquez-Morfin L, Krause J (2020). Origin and health status of first-generation Africans from early colonial Mexico. Current Biology 30(11): 2078 to 2091. doi.org/10.1016/j.cub.2020.04.002
- Ancient DNA Tiesler V, Sedig J, Nakatsuka N, Mallick S, Lazaridis I, et al., Price TD, Reich D (2022). Life and death in early colonial Campeche: new insights from ancient DNA. Antiquity. doi.org/10.15184/aqy.2022.79
- History Borucki A, Eltis D, Wheat D (2015). Atlantic history and the slave trade to Spanish America. American Historical Review 120(2): 433 to 461. doi.org/10.1093/ahr/120.2.433
- History Wheat D (2016). Atlantic Africa and the Spanish Caribbean, 1570 to 1640. University of North Carolina Press.
- Ancient DNA Sandoval-Velasco M, et al., Schroeder H (2023). The ancestry and geographical origins of St Helena's liberated Africans. American Journal of Human Genetics. researchprofiles.ku.dk
- Press release Max Planck Institute for the Science of Human History (2020). African skeletons from early colonial Mexico tell the story of first-generation slaves. shh.mpg.de
- Press El Universal (2020). Mexican and German scientists collaborate on genetic research about the first African slaves. eluniversal.com.mx
- Data Global25 PCA coordinates by Davidski, and the Moriopoulos 2026 collection of Global25 coordinates.