Historical records on transatlantic slavery show that an estimated 12.5 million people were forcibly taken from Africa and sent to the Americas between 1515 and 1865. More than one and a half centuries after slavery was abolished, a study led by Steven Micheletti, a Population Geneticist at 23andMe, examines how the slave trade impacted the genetic landscape of the Americas.
During the course of the study, Micheletti, et al, analyzed genotype array data from over 50,000 research participants representing the major slave trading regions of Atlantic Africa, Europe, and the Americas and compared the results with shipping manifests and other historical documents. This comparison allows those who are descendants of slaves to learn more about their roots, but it also illuminates the extent to which slavery still impacts the gene pool today.
Micheletti, et al, primarily focused on participants who were 23andMe customers whose grandparents were born in one of the geographic regions of trans-Atlantic slavery. Using this database allowed the researchers to ask focused questions of the participants and garner data that academics are typically not privy to.
To learn more about the study, we interviewed Steven Micheletti. Below, he shares the inspiration behind the study, challenges that arose, and surprises that were found when comparing historical records to the genetic information gathered.
What factors inspired the origin of this project, and how did you personally become involved?
Our mission is to help people access, understand, and benefit from the human genome. The foundations of this study began almost 13 years ago by Joanna Mountain, senior author on the paper, out of a recognized need to further understand the ancestral history of African Americans. Personally, my interest in population genetics, historical migrations, and the potential of helping people better understand their ancestral origins drew me to this project. Over the years, we eventually acquired enough study participants to have sufficient genetic representation of people from countries across the Atlantic and could make inferences about the genetic landscape that resulted from the transatlantic slave trade.
What were the main challenges with integrating the data from the 23andMe genetic database with the other databases used in the study?
The historical records we used, such as the Slave Voyages, describe events that impacted millions of people up to 500 years ago. Therefore, we carefully included study participants from the Americas that are possible descendants of the people that the historical records describe. This specifically meant selecting a subset of our millions of study participants who had ancestry originating from Atlantic Africa as well as deep historical ties to geographic regions involved in the transatlantic slave trade. In Africa, we included study participants self-identifying from ethnolinguistic groups that were directly impacted by the transatlantic slave trade, as described by historians. After selecting our study participants, it was a matter of determining the genetic connections between people of the Americas and Africa and comparing those results with records describing regional slavery practices. Perhaps not surprising, historical literature describing the atrocities of slavery and inhumane treatment of enslaved people best described resulting genetic patterns in people of African descent today.
Does next-generation sequencing (NGS) offer any advantages, compared to the SNP arrays used in this study?
Higher density genetic data produced by NGS offers the potential advantage of identifying more distant genetic connections by identifying smaller shared genetic segments that were passed down by common ancestors. This would be useful for more distant historical events; however, for the purpose of identifying population structure and genetic connections originating from, at most, 500 years ago, our arrays containing 500,000 – 1.2 million genetic variants suffice.
Given the size of the Slave Voyages database, how did your team approach the problem of finding the relevant data?
With a collection of over 35,000 historical shipping manifests, we had to curate data from Slave Voyages so that they could be directly mapped and aggregated by contemporary geographic locations. This meant determining which country and state each historical port in Africa and the Americas is located in today and determining the number of Africans that were forced between each embarkment and disembarkment port. Similarly, we used geographic locations of study participants’ grandparents to determine the likely ports where their enslaved ancestors disembarked. Once we mapped all of the ports, we aggregated additional variables of interest, such as the proportion of men, women, and children on each voyage. Together, this painted a picture of how many enslaved people were being transported between regions of Africa and the Americas and what composition were women and children.
Additionally, we collaborated with historians and African American studies scholars to put genetic results into historical context.
How did the study’s results compare with the historical records? Were there any unexpected results?
In general, ancestral connections between slave trading regions of Africa and the Americas mirror the proportion of individuals forcibly moved between those regions. People of African descent in the Americas share the most recent common ancestors with people of West Central Africa. This is consistent with records indicating West Central Africa experienced the largest proportion of deportations in the mid-1800s.
However, there were some key differences:
We observed an over-representation of Nigerian ancestry in many parts of the Americas compared to what shipping records show. This most likely reflects the often under-discussed active trade that moved slaves within the Americas, even after the slave trade had been banned by several European nations.
Senegambia appears to be underrepresented given the overall proportion of enslaved people who were deported from this region into the Americas. This may be because Senegambians were transported to rice plantations in the United States which were often rampant with malaria, resulting in an extremely high death rate.
Latin American individuals had strikingly lower African ancestry proportions compared to those from the United States and British Caribbean. The proportion of people with greater than 5% African ancestry was five times lower in Latin America, despite Latin, Central, and South America receiving roughly 70% of all disembarked African slaves.
This could be due to higher mortality and smaller effective population sizes in enslaved people brought to Latin America. Another factor to consider is that many Latin American countries promoted the dilution of African heritage through intermarriage between fair-skinned Europeans and women of African descent.
Confirming other studies, we saw a bias towards African female contributions to the gene pools of the Americas due to generations of rape and exploitation. This is particularly evident in Latin America where we estimated about 15 African women reproduced for each African man in Central America and estimated similar values for South America and the Latin Caribbean. The evidence that African females contributed to the gene pools of the Americas to a much greater extent than did African males is surprising given that the majority of enslaved individuals were male.