Igbo and African Ancestry

Roseline Siguret's tabulation of African slave "nations" in the Quarter of Jacmel (from "Esclaves d'indigoteries et de caféières au quartier de Jacmel (1757-1791)"

After reading FonteFelipe's recent blog on African ancestry and matches of his sample of 30 Haitians, we decided to reexamine our own African matches on Ancestry. Unfortunately, we could only find 2 obvious African matches, both Igbo-descended individuals. With one of them we share 23 cM of DNA. Our Haitian parent, who, unsurprisingly, has more African ancestry, shared 29 cM with the same Igbo individual. In addition, they also had African matches who, based on their surnames and publicly posted genealogies, hailed from Mali, Congo, Nigeria, and what appears to be Sierra Leone. The known ethnic breakdown is Igbo, Yoruba, and Mandinka. We could not easily determine the ethnic background of the other Africans. Needless to say, our Haitian parent's African matches were usually of a low degree of shared cM of DNA, 8-16. Her ethnicity estimates assigned 1% Yorubaland, 2% Nigeria-North Central and 2% Nigerian Woodlands (in addition to Mali, Senegal, Ivory Coast & Ghana, Benin & Togo, Cameroon, Western Bantu Peoples). She received a whopping 27% Benin & Togo and 21% Nigeria in the latest update. While there are many problems with her results in the update (her European ancestry is assigned mostly to Spain, even though her previous results assigned the largest European percentage to France), looking at her African matches does suggest some generally useful information on Haitian ancestry. The preponderance of Nigerian matches is likely a reflection of the greater number of Nigerian customers of Ancestry as well as the deep ancestry of Haitians in Nigeria, Benin & Togo and neighboring areas that supplied many of the African captives enslaved in Saint-Domingue. Here is a table breakdown illustrating her matches:

After her match with one Igbo person at 29 cM shared DNA, the next closest match was at 16 cM with someone bearing a Yoruba name. As suggested by FonteFelipe and the work of scholars such as Geggus, there clearly was a larger presence of Igbo and people from the Bight of Biafra in Saint-Domingue's south (which included Jacmel, in the calculations of Geggus). This noticeable Igbo presence is attested to by plantation inventories, testaments, and other contracts mentioning the enslaved population in 18th century Bainet. After Creoles, Igbos were one of the most common groups. The Yoruba ancestry and related Benin/Togo is possibly, in part, represented by the large Creole population among the slaves in Bainet during the second half of the 18th century. Many of the Creoles were likely the children of people who were purchased from the Slave Coast, which would have included various groups represented in Saint-Domingue (like the "Aradas"). This also could have accounted for the "Senegalese" results in our parent's ethnicity estimates. After all, if Senegal was important early on in the French slave trade, then some of the Creoles may have harbored ancestry from Senegambia. As for Mali connection, we assume this to be a sign of ancestry from "Mandingues" and, perhaps, Bambara, in Saint-Domingue. The Bambara would have reached Saint-Domingue through slave trading networks in Senegal, as was likely the case for "Mandingues" who may have not always been clearly distinguished from the Bambara. Interestingly, the Bamana were present among the enslaved in Bainet and one of them played a prominent role during the Haitian Revolution in the Sud (Gilles Bambara). 

Sadly, without more data from African customers or better ethnicity estimate information on African reference panel groups (such as directly telling us which ones we share DNA with), we are still left with only a general picture. Nonetheless, the Igbo contribution to our African ancestry seems very likely based on our shared DNA with one Igbo family. Furthermore, it seems our African ancestry probably includes Mandinka or Malinke (and Bambara?) ancestry as well as Yoruba. To some extent there was undoubtedly a Congolese or Central African contribution, though "Congos" in Bainet could have included Mondongues, Loango, Kongo, and others. As for "Senegal" and areas like Nigeria-North Central, we assume that could be Hausa and Wolof or Fulbe ancestry. Perhaps the "Senegal" and Mali ancestry is slightly inflated by Poulard or Fula, while Nigeria-North Central is from the Hausa or another another northern Nigerian population whose name was not recognized or used in Saint-Domingue. There definitely was a small flow of captives from northern Nigeria, but usually of Hausa or Nupe origin in Saint-Domingue's nomenclature for African "nations." We can only hope that subsequent updates see improvements in the area of Nigeria and West Africa as a whole.