On 9 October 2015 I published my first preliminary findings based on 19 Jamaican AncestryDNA results (see this link). Right now, four years later, I have managed to collect a sample group which is five times greater! Consisting of no less than 100 AncestryDNA results of Jamaican-born or fully Jamaican descended persons.1 Although still limited this data-set already provides a rather robust basis. Allowing for a finer detailed analysis of Jamaican genetics. In the first place with regards to the African regional roots of Jamaicans. But in addition I will also cover the Amerindian, Asian and European admixture scores being reported for Jamaicans on Ancestry. As well as variation in African admixture in general. With a special focus on substructure.
These results were obtained during 2013-2018 but I had not fully processed all the data up till now. Although actually in 2018 I already blogged about this survey group (n=100) in my comparison of various parts of the Afro-Diaspora:
- Update: Afro-Diasporan AncestryDNA Survey (part 1) (2018)
- Update: Afro-Diasporan AncestryDNA Survey (part 2) (2018)
Please keep in mind that AncestryDNA’s ethnicity estimates have been updated several times now! In this blog post I am dealing exclusively with AncestryDNA version 2 which was current between September 2013 and September 2018. All matters being discussed are therefore not pertaining to recently updated results (2018/2019) (unless mentioned so specifically). In my opinion especially version 3 (Sept 2018 – Oct. 2019) has been a downgrade rather than providing any meaningful improvement when wanting to learn more about one’s African origins (see this blogseries). The update which is currently rolled out will be reviewed at a later time.
I will mainly revisit and expand on previous findings. Using new statistics and background information. Furthermore I will present my preliminary 23andme survey findings for 28 Jamaicans. This blog post is also intended to be a prelude for my current research into the African DNA matches being reported for Jamaicans. Which will be featured in a follow-up post to this one. Below an overview of all the topics I will cover:
- African regional breakdown in line with expectations?
- Variation & substructure in African admixture levels
- European breakdown reflecting mostly British ancestry
- Asian admixture: more or less widespread than imagined?
- Traces of Amerindian admixture is proof of enduring Taino legacy?
- Comparison with 23andme results being reported for Jamaicans
- Current update of AncestryDNA’s Ethnicity Estimates
- Screenshots of individual results & Youtube videos
Follow these links for my complete survey data & research methodology:
Table 1 (click to enlarge)
Table 2 (click to enlarge)
Chart 1.1 (click to enlarge)
1) African regional breakdown in line with expectations?
Table 1.1 (click to enlarge)
The text of this first section is a remake of my original assessment in 2015. Based then on only 19 Jamaican AncestryDNA results (see this link). My data-set is still limited but has expanded to a more robust sample size of n=100 in the meanwhile. Which I suspect is actually quite representative already. Especially as I am myself not aware of any published study which aims to perform the same kind of analysis (autosomal & regionally within Africa) on this scale.2 Naturally this blog post is still only intended to be an exploratory exercise. In no way do I pretend to offer any final verdict on the ancestral makeup of the whole Jamaican population! However as the results are randomly picked and from several different parishes within Jamaica they could still be a rough indicator of the scope of African origins in Jamaica as well as its main tendencies.
Jamaica’s African heritage is often associated with Ghana in popular imagination. And more specifically the Ashanti or Akan people. As shown by my AncestryDNA findings above Jamaicans are however characterized by a melange of various regional origins. Rather than just hailing from one particular area within Africa. A reflection of the multi-generational blending of several ethnic groups from Western & Central Africa and to a very minor degree also Southeast Africa. Even if compared with many other parts of the Afro-Diaspora this regional mix appears to be less heterogeneous and more restricted to the Lower Guinea area indeed (see this blogpost).
It is fascinating to contrast my AncestryDNA findings with the estimated regional origins of Jamaicans according to historical, linguistic and mtDNA research. Obviously the regions are not measured in exactly the same way by each separate type of research. But still a rough comparison might be insightful. I will just comment on the most outstanding differences & similarities I am able to pick up on. Otherwise please refer to the blogposts I devoted to these research findings in the past:
- Ethnic Origins of Jamaican Runaway Slaves (Tracing African Roots, 2015)
- Words of African Origin in Jamaican Patois (Tracing African Roots, 2015)
- Jamaican maternal lineages trace back mostly to Ghana? (Tracing African Roots, 2015)
As well as these referenced sources:
- The Links of a Legacy: Figuring the Slave Trade to Jamaica. (Chambers, D., 2007)
- Interdisciplinary approach to the demography of Jamaica (Deason et al., 2012)
- The African Lexis in Jamaican: Its Linguistic and Sociohistorical Significance (Farquharson, 2012)
Trans Atlantic Slave Voyages to Jamaica
Table 1.2 (click to enlarge)
Ethnic origins of Jamaican Runaway Slaves
Table 1.3 (click to enlarge)
Words of African origin in Jamaican Patois
Chart 1.2 (click to enlarge)
mtDNA of Jamaicans according to regional affinity
Table 1.4 (click to enlarge)
1) The percentage of “Mandingo” runaway slaves (12,6%) in table 1.3 seems a bit out of line. When compared to the minimal genetic share from Senegambia according to AncestryDNA: 2.4% “Senegal”, (see table 1.1). But it is striking how much my AncestryDNA findings otherwise correspond with other types of research. Trans Atlantic slave trade from Senegambia being 2.3% (see table 1.2) and the share of Senegambian haplogroups was found to be 4.9% (see table 1.4).
AncestryDNA does report higher “Mali” scores for Jamaicans (7.4% on average, see table 1.1). But this region is more ambivalently defined. It might actually be more correlated with the 9.6% affinity for Sierra Leonean mtDNA (see table 1.4). As well as the “Chamba” captives from presumably northern Ghana/Burkina Faso (see foot note in table 1.3). Probably many of the so-called “Mandingo” runaway slaves hailed from Sierra Leone or Guinea Conakry rather than Senegal or Gambia. Something which might be obscured by using the term “Greater Senegambia”. Senegambian roots in the strictest sense of the word appear to be among the lowest in the Americas for Jamaicans. See also
- How to make more sense of “Senegal” & “Mali” scores (scroll down for it)
2) The combined share of Gold Coast Runaway slaves (“Coromantee” + “Chamba” =16.1%, see table 1.3) seems to understate the large genetic impact from Ghana according to both AncestryDNA & mtDNA research and also implied by slave trade statistics and linguistics. The sample of runaway slaves is however heavily drawn from a later time-period (1775-1817) which might explain this difference. The greatest wave of Gold Coast captives are known to have arrived in Jamaica in the mid-1700’s (see table 2.4). By the time of 1775-1817 many locally born or so-called “Creole” slaves might have had more pronounced Ghanaian origins than reported for the Runaway Slaves. On the other hand it must also be taken into consideration that to a lesser degree “Ivory Coast/Ghana” scores on Ancestry may have been inherited by way of Liberian or Sierra Leonean ancestors. See also:
3) On first sight the “Benin/Togo” scores on AncestryDNA seem to be more prominent than justified by other sources. Especially when comparing with the so-called “Popo” runaway slaves (an umbrella term for mostly Gbe speaking people) who were a mere 2,6% of the sample (see table 1.3). The runaway advertisements were however biased towards later time-periods. While it is known that most Bight of Benin slave trade for Jamaica occurred in the late 1600’s and early 1700’s (see table 2.4). My African AncestryDNA survey has demonstrated that the socalled ‘Benin/Togo” region was in fact also predictive of (eastern) Ghanaian DNA as well as southern Nigerian DNA. Therefore aside from pinpointing genuine Beninese/Togolese lineage this region was also indicative of other types of ethnic origins. Found mainly among non-Akan groups in Ghana and specifically (but not exclusively!) the Ewe. And in addition also southern Nigerian populations such as the Igbo and Yoruba. Which both showed significant amounts of “Benin/Togo” in Ancestry’s 2013-2018 version. See also:
Due to genetic overlap the country name labeling in DNA testing is usually not to be taken too literally. Still it is remarkable how many Jamaicans in my survey group had “Benin/Togo” showing up as primary region, 24/100 see chart 1.1. Also the Jamaican group average for “Benin/Togo” (21.3%) turned out to be among the highest in the Afro-Diaspora. Safe for the Haitians as expected (see this blogpost). In that light it is revealing perhaps that linguistic research (Farquharson, 2012) has been uncovering significant non-Akan influences coming from Ghana as showcased by the percentages of Gbe, Guang and Ga in chart 1.2. Also researchers of Jamaican mtDNA established that the Bight of Benin seemed more important than the Bight of Biafra and Central Africa as a matching region for Jamaican maternal haplogroups. I suppose this may provide a an additional context for seemingly overly prominent “Benin/Togo” results being reported for Jamaicans.
4) There seems to be a striking disagreement about the possible Nigerian ancestral contribution for Jamaicans. According to both mtDNA and linguistic research the Bight of Biafra or Igbo influence was found to be not even secondary. But just quite minor (<10%) compared with the resonating influence being mentioned from Ghana. However my AncestryDNA findings do suggest a prominent autosomal genetic impact of Nigerians. That is genomewide and not just restricted to maternal lineages! It is certainly significant when judging by the high group average of 22.5% “Nigeria” (see table 1.1). As well as the frequency of “Nigeria” being reported as primary region: 26/100 (see chart 1.1). Also when comparing with other parts of the Afro-Diaspora these scores stand out (see this blogpost). Furthermore “Nigeria” has been an imperfect measure of Nigerian DNA (possibly only covering 50-60% in the 2013-2018 version). Which is why it can be assumed that an even more realistic estimate of Nigerian lineage among Jamaicans could be around 35-45%. See also:
On the other hand “Ivory Coast/Ghana” scores were all in all more significant than “Nigeria” scores for my Jamaican survey group. Even when the differences were not that big (take note for example of the median scores in table 1.1). Prominent contributions from both Ghana and (southeast) Nigeria were to be expected based on the Trans Atlantic slave trade statistics (see table 1.2). However according to some scholars the Inter-Colonial slave trade might have been ethnically selective. Disproportionally targeting Igbo captives by re-exporting them from Jamaica to other colonies. The Runaway Slave advertisements provide an interesting twist to this debate. As in fact Igbo (“Eboe”) captives were the most numerously mentioned. Also the so-called “Moco” runaway slaves highlight the significant presence of non-Igbo captives being taken from the Bight of Biafra. Most likely from areas closer to the border with Cameroon or even partially from Cameroon itself. It might very well be that many of their inherited DNA markers were being read as “Cameroon/Congo” rather than “Nigeria” by AncestryDNA’s 2013-2018 version.
5) The Central African legacy for Jamaicans is being reported rather consistently as substantial but clearly secondary to Lower Guinean roots (area in between Liberia and Cameroon). Only the Runaway Slave data seems to be showing a noticeably higher amount of Central African origins (interestingly also specified in both “Congo’s” and “Mungola”!). However as already mentioned this is caused by the time-period bias of these advertisements. As also seen in the West Indian Slave Registers, Central Africa and the Bight of Biafra predominated during the very last decades of British slave trade (before Abolition in 1807). However in earlier periods their presence would have been more subdued.
Documented slave importations from Southeast Africa into Jamaica have been very rare (0.1%!, see table 1.2). Therefore it seems fair to assume that the “Southeast African” category from the mtDNA overview and also the “Southeastern Bantu” scores on AncestryDNA are actually referring to Central African origins from mostly the Congo (DRC and Brazzaville), Gabon and northern Angola. Combining things for the maternal haplogroup findings we then get an estimated proportion of about 18% (8.9% + 9.2%, see table 1.4). Which is very close to the linguistic outcome of 19% (see chart 1.2) and slave trade data of 16.1% (see table 1.2).
It becomes a little bit more tricky however when we contrast this with my AncestryDNA findings. The so-called “Cameroon/Congo” region is after all measuring genetic similarity to reference samples from both Central Africa (Congolese) as well as Cameroon, which would fall under Bight of Biafra. As can be seen from my Nigerian AncestryDNA survey in fact “Cameroon/Congo” is also reported for especially southeastern Nigerians (around 13% for 29 Igbo’s on average with outliers of >30%). A good part of the Jamaican “Cameroon/Congo” scores might therefore be reflecting DNA from Cameroon & eastern Nigeria rather than reflecting Congolese ancestry. In contrast with many other parts of the Afro-Diaspora! See also:
When making a conservative estimate we could deduct 5% from the 17% group average for “Cameroon/Congo”. Ending up with about 12% “genuinely” Central African affinity. Combining this with the much smaller scores for “Southeastern Bantu” (3.3%) and “Pygmy/San” (=“South-Central Hunter Gatherers”) regions (1.1%) we arrive at a rough proxy of about 16% Central African ancestry. In agreement with the other research findings. Although actually it seems quite likely that a larger share of “Cameroon/Congo” is to be attributed to the Bight of Biafra hinterland (incl. Cameroon). Again do notice the considerable 11.6% share of “Moco” captives among Runaway slaves (table 1.3). They are assumed to have been mostly Efik from southeastern Nigeria but also incl. other neighbouring groups from southeast Nigeria (non-Igbo) as well as Cameroon (see this blogpost).
Summing things up we could say that the assumed predominance of origins from Ghana and Nigeria is largely confirmed. Both by my Jamaican AncestryDNA findings (n=100). As well as independently by research performed based on history, linguistics and mtDNA. Albeit that one has to take into account that regional categories in these types of research tend to be inherently blurry. Another way of tackling this topic is by reviewing the maximum regional scores obtained by my Jamaican survey participants.
Maximum regional scores
Figure 1.1 (click to enlarge)
The above compilation is meant to illustrate those regional lineages which seem to have been the least diluted within Jamaica’s genepool. Besides the “average” result the other 7 screenshots feature a maximum regional score I have observed during my Jamaican AncestryDNA survey. See also this page for an actualized overview:
Obviously a much greater variation and a usually more regionally mixed outcome can be observed if you closely examine each and every single individual result (see last section of this page). Also the screenshot I selected as the “average” Jamaican result is naturally just based on the limited samples which were available to me. It happens to show the closest similarity and most likely also closest genetic affinity to the regional averages I have calculated for my 100 Jamaican samples. Especially with regards to the total African amount (80% being the average score for my sample group) and the ranking of the first four regions.
Determining the largest regional components within the African breakdown, on average, for each of my sample groups has been a primary research effort during my AncestryDNA survey (2013-2018). After all these most prominent regional scores might be considered to have the highest reliability and might also be confirmed independently by historical sources. Establishing where each African region is relatively more pronounced or instead more subdued might therefore provide insightful clues into localized ethnogenesis across the Diaspora. For further reading see also:
- Afro-Diaspora AncestryDNA results: A Comparison (2016)
- Update: Afro-Diasporan AncestryDNA Survey (part 1) (2018)
It seems a meaningful finding that among my 100 Jamaican survey participants only three regions produced amounts of above 50%: “Ivory Coast/Ghana” (63%), “Nigeria” (56%) and “Benin/Togo”(53%). Most likely not by coincidence these are also the three primary regions when going by other statistics such as group averages (see table 1.1) and frequency of being the top region (see chart 1.1).
Nonetheless it pays to be cautious with the country name labeling of the AncestryDNA regions. As often in fact border-crossing overlap is rampant and it should be kept in mind that these “ethnic estimates” are not meant to be seen as exact 😉 Furthermore as subsequent updates have demonstrated Ancestry’s African regional framework is far from stable. Sadly the predictive accuracy of several key regions for Jamaicans has been going up and down in wild swings. In particular “Nigeria” and now also “Ghana” (see section 7). Therefore it remains essential to judge each case on its own merits and contrast with plausibility. Something which I have myself always endeavoured to do during my AncestryDNA survey (2013-2018 version). I will revisit this topic in my upcoming blogpost which will explore the possible correlation between African DNA matching patterns and Ancestry’s African breakdown.
- African DNA matches reported by Ancestry for 30 Jamaicans (under preparation)
2) Variation & substructure in African admixture levels
Figure 2.1 (click to enlarge)
Chart 2.1 (click to enlarge)
This section is intended to demonstrate the full extent of African ancestry within my Jamaican survey group. To enable group comparison I have scaled everyone’s African breakdown to 100% within my AncestryDNA survey (see also this page for further details on my methodology). But obviously to put everything in broader perspective the original total African amounts are also still relevant. This kind of data often gets presented in a potentially misleading manner. Only highlighting the mathematical mean of all the results as the average. Which tends to downplay individual variation and obscure possibly skewed distributions. Especially atypical lower scores tend to pull down the mean. As shown in chart 2.1 my Jamaican survey group is clearly showing an asymmetrical admixture distribution. Therefore I have compiled various other statistical measures, see also table 2. In which it can be seen that the median score of 83% (50% cut-off point of all observations) might be a more representative reflection of the group average than the mean of 80.8%.
Furthermore it can be verified from chart 2.1 that the range of 90%-100% African admixture is most common among my Jamaican survey participants. Representing almost 40% of the group (38/100). Still at the same time a majority (62) showed non-African admixture of in between 11%-65%. So generalizations should be avoided as individual variation will always exist.
I like to emphasize that no fictional national averages are implied. Chart 2.1 firstmost reflects the limited number of samples which was available to me. Follow this link for the source data (columns M & O). Even so I do think that my sample size (n=100) is pretty robust already. And therefore my findings might correlate closely to what is to be found within the Jamaican genepool as a whole. As a crude rule of thumb it is often stated that n=30 is a large enough sample size for most standardized types of research. Also my samples have been randomly picked and are – as far as I have been able to tell – from various parishes within Jamaica as well as from various social backgrounds.
Racial composition Jamaicans according to census (1844-1943 & 2011)
Table 2.1 (click to enlarge)
Table 2.2 (click to enlarge)
It is very interesting to compare my findings with what is known about the racial/ethnic composition of Jamaicans according to census. Naturally there are many known shortcomings about racial (self)identification. Which is often not precise at all but based on subjective perception. Also my own survey may have its limitations. For example I have naturally omitted results of Jamaicans without any African admixture at all. As the African breakdown was after all the main focus of my survey. Having said all that I still find it striking how within my survey the share of people of predominant African descent is pretty much the same as according to most recent census. Combining admixture brackets of 60%-100% we arrive at a share of 90% (13+16+23+30+8). While in 2011 a proportion of 92% self-identified Black Jamaicans was reported (see table 2.2).
It is also intriguing to look at the evolution of racial composition of the Jamaican population (see table 2.1 and compare with table 2.2). As it appears that across time the proportion of “Black” Jamaicans has increased (from around 77% before 1943 to 92% in 2011). While that of “Mixed” or “Coloured” Jamaicans has decreased (from around 18% before 1943 to 6% in 2011). However based on documented intermarriage across the generations between Black & non-Black Jamaicans as well as actual DNA test results it may be assumed that minor non-African admixture has become widespread among Black Jamaicans. Even if usually in small amounts of <10%. And in fact Jamaicans of “pure” 100% African descent are also still quite numerous. Actually 100% African was the most frequently reported score during my survey (see modal score in table 2)!
During my survey I have observed 8 results for Jamaicans who are of fully 100% African descent. Which represents a share of 8% (8/100) of my survey group. Again this is of course based on limited sampling. But it seems that this frequency of “pure blooded” African descended persons (8%) is quite notable when compared with most other parts of the Diaspora. During my survey only trumped by Haitians (11/97=11%) and Guyanese (4/21=19%). But considerably higher than for African Americans (10/515=2%).
African Americans actually tend to be quite comparable to Jamaicans. Both groups having an average of around 80% African admixture during my survey. But the distribution of African admixture is more so tending towards the 70-90% bracket for African Americans. Probably this frequency of people with 100% African admixture is mostly correlated with the share of African-born people in the 1800’s. It has been estimated for Jamaica this was about 45% in 1807 and still 37% in 1817 (see this link). While for the USA it would have been much less. Probably more so in South Carolina due to continued Trans-Atlantic slave imports in the early 1800’s. But for Virginia it has been estimated that less than 1% of the slave population was African-born in 1800. See also:
- African Amounts across the Diaspora (spreadsheet)
- African admixture range among 350 African Americans (chart)
- African admixture range among 45 Haitians (chart)
- From African to Creole
Substructure according to admixture level
Table 2.3 (click to enlarge)
One of the most fascinating aspects of my AncestryDNA survey among not only Jamaicans but also other parts of the Diaspora has been that so-called substructure was eventually revealed. Genetic substructure is basically referring to subgroups within greater populations. To be defined along geographical, social, cultural, or even “racial” lines. Despite commonalities various localized factors may still have caused differentiation between two or more subgroups within a given population. In particular pointing towards a distinctive mix of African regional origins. Showing overlap to be sure but still recognizable due to deviating proportions.
In other blog posts I have already explored this theme for African Americans, Puerto Ricans, Dominicans and Haitians. But also for Jamaicans such substructure could be relevant and useful for greater understanding of context-dependent African roots. I have to stress that my findings for Jamaicans are still very much preliminary! And an even greater sample size would be required for the expected patterns to show up more clearly and also in a more robust manner.
In table 2.3 I have calculated the regional averages for two Jamaican sub-groups. Which are defined by their total amount of African ancestry. So we have 38 Jamaican survey participants with total African being greater than 90% and on average 96%. And another subgroup consisting of 23 Jamaican survey participants who are of a more mixed background. Total African being less than 70% and on average 57%. For both groups the African breakdown has been scaled to 100%. The regional differences may appear to be only slight. And naturally individual variation is not to be denied! Still noticeable discrepancies for “Cameroon/Congo” and “Nigeria” can be observed in particular, as well as for “Mali”. Otherwise the differences do not seem really statistically significant.
I made this distinction to test the hypothesis of how early creolization among a sub-segment of Jamaicans might have led to greater retention of African regional roots associated with earlier waves of slave trade. Mainly Gold Coast and Bight of Benin. Or stated in reverse how a relatively high share of African-born ancestors from the late 1700’s/1800’s may have resulted in a larger genetic imprint of African regional roots associated with the last waves of slave trade. Principally Bight of Biafra and Central Africa. As shown below.
Waves of slave trade from specific regions across time
Table 2.4 (click to enlarge)
Table 2.5 (click to enlarge)
Slave trade from Africa to Jamaica took place during a long period and many different coastlines and slave ports were involved. It was also characterized by many fluctuations across time. This may cause a great deal of confusion and may leave behind an impression of prevailing randomness. However if you closely look into the data certain helpful patterns do appear. Indicating clusters of imports from a certain coastline during a specific time period. As demonstrated in the tables above. Basically two main waves can be distinguished based on numbers: involving Ghana and its hinterland (=Gold Coast) during the mid 1700’s. And southeast Nigeria (= Bight of Biafra) during the last 4 decades of slave trade until 1808. However do also take notice of the smaller waves involving for example Bight of Benin during the late 1600’s and early 1700’s! As well as the hike in slave trade from Central Africa in the late 1700’s up till 1808.
Again my survey findings as shown in table 2.3 are not fully in agreement with tables 2.4 and 2.5. Still it is intriguing to already observe a presumably higher genetic contribution from Bight of Biafra and Central Africa among the subgroup of Jamaicans with total African> 90%. As indicated by their higher group averages for “Nigeria” (22.5% vs. 18.1%) and “Cameroon/Congo” (20.2% vs. 14.2%). In fact this gets even more pronounced when only focusing on my 8 Jamaican survey participants who received 100% African scores. Among them 4 people had “Cameroon/Congo” as a primary region and their subgroup average (n=8) for “Cameroon/Congo” is 27%!
Aside from increased slave trade from Central Africa in the late 1700’s this outcome is possibly also linked to the arrival of Central African contract labourers in the mid 1800’s. They are well known for having contributed to the Kumina cultural legacy. And apparently being concentrated especially in St. Thomas parish. However as can be seen in table 4.1 the numbers involved were relatively minor when compared with total Black Jamaican population at that time (293,000-346,000, see table 2.1). Whereas 10,000 African indentured labourers are estimated to have settled in Jamaica in between 1841-1862. Many of them socalled “Congo’s” but actually also incl. Yoruba & other Africans. So Central African contract labourers would have been less than 10,000 even which is also quite minor when compared with the last wave of Central Africans by way of slave trade. Which was 56,000 in between 1792-1803 (see table 2.5). Still an intriguing article in this regard which mentions: “There are several Bongo towns and Congo towns in Jamaica.” :
- Hard-working Congo man remembers (Jamaica Gleaner, 2009)
The reverse expectation of higher genetic impact from the Gold Coast for the subgroup with total African <70% has not really materialized. As actually the average level of “Ivory Coast/Ghana” is quite consistent across my entire Jamaican survey group, irrespective of admixture level. Although not really pronounced there is still a rather elevated “Benin/Togo” group average (20.2% vs. 18.6%) for my Jamaican survey participants with total African <70%. Their “Mali” group average does also stand out somewhat (12.3% vs. 5%). Possibly to be correlated with an earlier Senegambian wave (1676-1725). Or even also with northern Ghanaian captives (Gur speaking “Chamba”!) which would then still connect with an earlier Gold Coast wave! However I suspect that it is actually a few outliers which may have caused much of the difference in “Mali” group averages.
Earlier Creolization among Coloured Jamaicans?
Table 2.5 (click to enlarge)
Basically the argument for substructure among Jamaicans according to admixture level would be similar to what I have described elsewhere for Puerto Ricans already in 2015. That is a genetic “founding effect” derived from earlier African captives from specific regions would be more noticeable among people who can trace most of their African roots to earlier settlements periods. Or at least periods when natural growth would be picking up somewhat. Jamaica’s colonial society obviously being notorious for its continued high mortality and negative reproduction, over all speaking. Still despite all the ordeal people were having children and their surviving descendants formed the nucleus of a steadily growing segment of so-called Creole Jamaicans. That is locally born in Jamaica as opposed to African-born (or European-born in case of White Jamaicans).
I have always argued that the dating of the socalled Creolization process/transition is fundamental for tracing back African ethnic roots. After all a different context will apply when the majority of your African-born ancestors happen to be from the 1800’s. Or just the last phase of slave trade. Because then naturally the slave trade patterns of that particular time period will be most relevant for you. I suppose that also the odds of having higher than average total African admixture might then be deemed greater. As dilution of “pure” African lineage may not have taken place during a shorter interval of generations. Although of course individual genealogies can always account for any kind of introduction of non-African lineage, also in recent generations.
This Creolization process may have occurred most rapidly among people of racially mixed background, so-called Coloured Jamaicans. As can be seen in Table 2.5 they have been documented as a separate part of Jamaican society, at least since the early 1700’s. Researchers have established that this particular segment of Jamaican population also had the highest reproduction rate during colonial times. Both free and enslaved Coloureds! Similar to what took place in the Hispanic Caribbean racial admixture and relative endogamy afterwards may have ensured that earlier regional African origins were preserved among them. Because additional African admixture was not occurring afterwards (or to a much lesser degree) for certain socially/racially defined population segments. And also given that apparently most inter-racial unions occurred between White men and Jamaican-born Creole Black women and not African-born women. See also:
- Genetic consequences of Creolization (scroll down for it)
- Slave Population and Economy in Jamaica, 1807-1834, (B.W. Higman, 1976)
Then again this hypothesis is clearly based on a great deal of generalization. While reality is much more complex. All depending on individual genealogy! A Jamaican with total African being >90% might very well still have a majority of African-born ancestors to be traced back to the mid 1700’s or even earlier. I imagine this would be the case especially for many Maroons! Also a Jamaican with total African <70% may in fact be descended from a relatively recent union between an African contract labourer and an Englishman in the late 1800’s. Just to name one random possibility. Also unlike the Hispanic Caribbean the various waves of African regional slave trade to Jamaica are mostly overlapping and not unique to any given time period. For example Nigerian lineage may be traced back to both the late period of slave trade (probably highest odds) but also the early 1700’s (see table 2.4/2.5). It is certainly not my intention to overlook or downplay this wide array of ancestral options!
Then again I do believe that generally speaking this substructure approach can be fruitful when wanting to find a proper context for your African regional breakdown. But aside from a good grasp of slave trade history it does require a good additional knowledge of your personal family tree as well! In fact another way of looking at substructure among Jamaicans might be to do so along parish or county lines. Jamaica is relatively small and a great deal of migration throughout the island must have taken place across the generations. But still it might be that similar to the USA, Brazil and Haiti certain rural parts of the country do carry a greater genetic imprint from certain African regions due to local differences in slave trade. This is something I aim to pick up in a follow-up blog post to this one:
- African DNA matches reported by Ancestry for 30 Jamaicans (under preparation)
3) European breakdown reflecting mostly British ancestry
Figure 3.1 (click to enlarge)
Chart 3.1 (click to enlarge)
Given the theme of this blog my AncestryDNA survey has naturally been focused on Tracing the African Roots of my survey participants. However most Jamaicans are not exclusively of African descent. Similar to most other Afro-Diasporans they show variable amounts of minor non-African admixture as well. As shown in table 2, this consists mainly out of European and to lesser degree also Asian ancestry for Jamaicans. I fully understand and respect that given the brutal history of the Slavery Period many Afro-Diasporans might not be inclined to learn more about their minor European origins. Even if the possibility of this European ancestry (partially) dating from the Post-Slavery Period cannot be ruled out in advance. Bob Marley’s English father makes for an intriguing example.
- The Influence of Bob Marley’s Absent, White Father (S. Gurtman, 2002)
Still other Afro-Diasporans, incl. Jamaicans, might be more curious about their complete genetic make-up and how this might define them. Despite shared experiences one must be careful to respect the localized context and different historical trajectories across the Afro-Diaspora. In fact there can be several valid reasons to also explore the European origins of Afro-Diasporans in a neutral and unbiased manner. Ironically in the process you might often also acquire more details about African ancestors linked to your European ancestors as well as their biracial offspring. Personal family histories are bound to sometimes deviate from the assumed narrative. It would be self-defeating to allow generalizations about European admixture to determine how you should feel about your own unique DNA makeup. Especially without at least having done any basic genealogical & historical research of your own in advance.3
Unlike the African breakdown actually the European breakdown has been greatly improved after Ancestry’s update in 2018. In particular the detection of British lineage became much more focused through the new region of “England, Wales & Northwestern Europe”. While also detection of Scottish lineage was improved. Even when still combined with Ireland into the new region of “Ireland and Scotland”. Also many additional genetic communities for the UK and Ireland have been introduced which have been appearing especially among the results of Jamaicans with a greater European genetic imprint. In a way this is making the European findings for my Jamaican survey group less relevant. Still given correct interpretation already valuable information was to be obtained.
In Ancestry’s 2013-2018 version many people were confused and often also mislead by their European DNA results. Because they tended to take the country name labeling too literally. Ignoring the geographical & genetic overlap of AncestryDNA’s regions. However in most cases receiving a multitude of European (trace) regions did not imply that you have a confusingly diverse European background! Rather it suggested that your European ancestors were themselves genetically diverse. But still these ancestors could have been from just one or two ethnic groups only.
English & Scottish lineage is historically most plausible
Table 3.1 (click to enlarge)
Table 3.2 (click to enlarge)
“Unlike mainland British North America, where the composition of European migration changed dramatically in the eighteenth century, away from English and to non-English migration, the English continued to dominate migration into Jamaica. Although Scots and Jews formed significant minority communities, the island could still be considered, in ways increasingly impossible in ethnically diverse places such as Pennsylvania, to be an English preserve, at least in regard to the free population. The English who went there tended to come from the metropolitan heart of England.” (Burnard, 1996, p.790)
“we can conclude that the Irish formed a small segment of the white settler society compared to the English and Scottish presence. The majority of migrants arrived from England in the seventeenth century, while the Scottish presence increased in the eighteenth century.” (De Jong, 2017, p.27)
In table 3.1 I have calculated how the scaled European breakdown of my Jamaican survey group compares with the group averages of British, Irish and African American sample groups. The UK shows up as most likely source candidate for European DNA detected among my Jamaican survey participants. As expected, based on known colonial history. Of course this is not to deny or rule out any additional and also post-colonial European migrant lineage from other countries. In individual and documented cases this may be very relevant even. I am however principally concerned with finding more generalized tendencies. Despite obvious sample size limitations I find it remarkable how closely my Jamaican survey group resembles its historically plausible main European source population already. See also this previous blog post for more detailed discussion, incl. also other parts of the Afro-Diaspora:
It must be kept in mind that in AncestryDNA’s 2013-2018 version British ancestry was being described as a composite of “Great Britain”, “Ireland”, “Europe West” and “Scandinavia”. This may be confusing when the regional labeling is taken at face value. However these are all still basically Northwest European regions and given Great Britain’s more ancient history, incl. both Germanic and Celtic influences not really surprising at all. See also:
Establishing the proportion of Scottish and Irish lineage has been more tricky. Because “Ireland” in Ancestry’s 2013-2018 version was not exclusive to Irish DNA but more so indicative of Celtic origins. Also found in Scotland, Wales and in fact England itself as well. Then again as shown in table 3.1 Jamaica’s over all European breakdown clearly deviates from the Irish one. From which it may already be deduced that Irish descent among Jamaicans is quite minor. As also indicated by historical research, see table 3.2.
The research performed by Burnard (1996) is quite impressive. However still some people may have been under-counted as apparently the number of ‘transients’, mostly sailors and traders, has been excluded. Also the European migration during the late 1700’s, 1800’s and 1900’s is obviously still relevant as well. But this is not included in table 3.2! It seems for example that the number of Scottish migrants may have increased also beyond the 1780’s. While also other groups of Europeans are known to have entered Jamaica in smaller numbers during the 1800’s, such as Germans contract labourers.
A high degree of mortality used to be common among White Jamaicans throughout the colonial period. Which also accounts for their rapidly declining proportion in overall Jamaican population. Nowadays less than 1% even! However given their long established presence plentiful occasion for inter-racial unions existed. Even more so due to known gender imbalance in favour of unmarried males. Leading to a steady growth in the number of mixed-race Coloured Jamaicans. As already shown in table 2.5. Who in turn intermarried with Black Jamaicans. Leading to dispersion of European DNA among the general Jamaican population. Further reading:
- European Migration to Jamaica, 1655-1780 (T. Burnard, 1996)
- Jamaica: the country with more Campbells per head of population than Scotland (The Herald, 2015)
- The Irish in Jamaica during the long eighteenth century (1698-1836) (K. De Jong, 2017)
- A failed settler society: marriage and demographic failure in early Jamaica (T. Burnard, 1994)
- White Jamaicans (Wikipedia)
- German contract labourers in Jamaica (Wikipedia)
Minuscule but still detectable Jewish lineage
Figure 3.2 (click to enlarge)
The group average for “European Jewish” was a paltry 0.3% among my Jamaican survey group. Only 19 persons receiving mostly trace amounts of around 1% “European Jewish”. Then again the maximum score was 5%. And for three persons “European Jewish” actually was the highest amount in their otherwise minor European breakdown. See also chart 3.1. and figure 4.1 below. These findings might be deemed to be close to noise-level. But actually it is known that due to their distinctive genetics Jewish admixture can often be reliably detected even at very subdued level. Which also seems to be the case on Ancestry’s 2013-2018 version.
It bears reminding though that the “European Jewish” region is first most describing genetic similarity with Ashkenazi Jews. But based on my observation it seems that Sephardi DNA was described on AncestryDNA more so by a combination of “Europe South”, “Middle East” and usually to a lesser degree “European Jewish”. This is a crucial issue as Jamaica’s Jewish heritage is mostly linked with Portuguese Jews, a.k.a. Sephardi Jews. So it might be that “European Jewish” may have underestimated the real extent of Jewish lineage among my Jamaican survey participants. Further reading:
- History of the Jews in Jamaica (Wikipedia)
- Isaac Mendes Belisario, Jamaican artist in 1800’s (That’s Inked Up, 2013)
- Sean Paul: a Yiddisher Rasta man (Ynet News, 2007)
- East European, Ashkenazi, Gypsy & Finnish AncestryDNA results
It is recommended to take a 23andme test as well in case you wish to cross-check any possible Jewish lineage. As from my experience that company does particularly well at picking up on small amounts of distinctive admixture. As a more robust alternative it is wise to look into your European and/or Jewish DNA matches for further corroboration. Do keep in mind you will have to verify carefully if these matches are truly 100% belonging to the nationality you are searching for. Either through their family trees or the information they are willing to share with you. And even so a myriad of other ancestral options might still be possible as well if you have no certainty how these matches would exactly be related to you or your MRCA. Still if any clear patterns are discernible, conjecture should give way to plausibility eventually 😉 This is something I aim to pick up in a follow-up blog post to this one. As it appears that Jewish matches tend to be plentiful even when only backed up by trace amounts of Jewish admixture. See also:
- African DNA matches reported by Ancestry for 30 Jamaicans (under preparation)
- Scanning & filtering method for zooming into DNA matches with a certain ethnic profile
4) Asian admixture: more or less widespread than imagined?
Figure 4.1 (click to enlarge)
Chart 4.1 (click to enlarge)
Asian admixture scores were only being reported among a minority of my Jamaican survey group. Most (62) did not show any Asian admixture at all. While I suspect many if not most of the 1% scores might have been misreadings on Ancestry’s part. Then again for 19 people Asian admixture was detected above noise level (>2%). And for most of them it was even greater than 10%. This finding represents a stand-out aspect of Jamaican genetics when compared with many other parts of the Afro-Diaspora (excl. the Anglo-Caribbean). For more detailed discussion see also:
The additional so-called “Pacific Islander” scores were merely a misreading of actual Chinese or Indian lineage. “Asia East” usually being combined with so-called “Polynesia”. And “Asia South” usually combined with so-called “Melanesia”. Also in some cases “Asia Central” showed up as an additional component. Again merely reflecting Ancestry’s incapacity at that time to make a valid distinction. After Ancestry’s update in 2018 these outlandish regions have usually disappeared from Jamaican results. And therefore in chart 4.1 I have added these mislabeled minor regional scores to the overall Asian amounts indicating either Chinese or Indian lineage. See also:
- Asian, Pacific & Native American AncestryDNA results
- East Asian & Polynesian AncestryDNA results
- South Asian & Melanesian AncestryDNA results
Indentured Labour Migrants from Asia after 1834
Table 4.1 (click to enlarge)
Jamaica’s national motto is: “Out of Many One People“. Based on the population’s multi-racial roots. As a result of incoming migrations (either voluntary or involuntary) from several parts of the world. This is something which has clearly surfaced during my survey as well. On the other hand some people feel that this characterization is bypassing the greater majority of Black Jamaicans (even when in fact they are also of mixed background within Africa). Such a stance being supported by my findings in regards to most frequent African admixture interval being 90%-100% (see chart 2.1). And it also seems to be qualified by my Asian admixture findings above in chart 4.1. Which shows that a greater majority of my survey participants (81/100) actually did not have any Asian admixture at all or only at noise-level. For an interesting but perhaps provocative article:
- Who Is Jamaica? (New York Times, 2012)
My overall Asian admixture findings are in accordance with Jamaica’s history of Indian contract labourers as well as Chinese migrants arriving after Slavery had been abolished. Their numbers were not really major though. Compared to other West Indian destinations (Trinidad & Guyana) and also when taking into account the population size of Jamaica at that time. Even when they did make a substantial impact on Jamaican society. See table 4.1 above as well as table 2.1. In fact the group averages (see table 1) for South Asian admixture (2.2%) and for Chinese admixture (1.4%) are quite close to the population shares of Indian & Chinese Jamaicans according to census.
Also haplogroup studies from the past have established a similarly minimal but still detectable contribution of Asians within the Jamaican genepool. Deason et al. (2012) found evidence of at most 1.5% (6/400) of Jamaican maternal haplogroups (mtDNA) being of Asian origin. The overwhelming majority being of African origin: 97.5%. Based on Jamaican paternal haplogroups (Y-DNA) Simms et al. (2012) found a somewhat higher proportion: 3.8% of the 159 samples indicative of Chinese lineage and 4.3% indicative of South Asian lineage. Again the majority of male haplogroups being of African origin: 66.7%. But actually combined with a considerable share of European haplogroups: 18.9%. Reflecting how geneflow among Jamaicans has often been sex-biased. Tending towards unions of either European or Asian males with Black Jamaican women.4
Because they were such a small minority and perhaps also because of gender imbalance many of these Asian migrants consequently were absorbed by the mainstream Jamaican population through inter-ethnic unions. Even when the existence of Asian ancestors is not always known by their mixed or even self-identified Black Jamaican descendants. This is something I came across several times during my survey. Whereby people expressed their surprise of their Asian admixture scores as they were not aware of having Asian lineage! See also the Youtube videos posted in section 8. And very intriguingly even ROOTS actor Malachi Kirby was not expecting to receive a 21% score of South Asian admixture when he was tested by 23andme. For further reading:
- Malachi Kirby: ‘Playing Kunta Kinte in Roots reboot allowed me to find my own roots’ (Evening Standard, 2017)
- How my DNA Set me Free (Geoffrey Philp, 2013)
- Finding Samuel Lowe (From Harlem to China) (Book/Documentary, 2015)
Most of the Asian admixture scores in my survey are suggesting dilution across several generations. And often also in line with the expected genetic inheritance of one Asian grandparent (~25%), great-grandparent (12-13%), great-great grandparent (5-6%) or even great-great-great grandparent (2-3%). Which would still all be historically plausible as this goes back to the mid 1800’s. To be confirmed by family tree research in the first place. Although actually also finding your Asian DNA matches can be very helpful! This is something I aim to pick up in a follow-up blog post to this one. See also:
- African DNA matches reported by Ancestry for 30 Jamaicans (under preparation)
- Scanning & filtering method for zooming into DNA matches with a certain ethnic profile
5) Amerindian trace admixture is proof of enduring Taino legacy?
Figure 5.1 (click to enlarge)
Table 5.1 (click to enlarge)
Figure 5.2 (click to enlarge)
The legacy of Jamaica’s indigenous Taino inhabitants is still very evocative to many people. Understandably so given their continued great symbolic meaning (see figure 5.1). But generally speaking the Amerindian genetic inheritance among Jamaicans seems to be statistically quite insignificant. At least according to my AncestryDNA findings. As can be verified from Table 5.1 the group average being only 1/10 of a percent.5 For 84 out of 100 Jamaican survey participants zero % Native American ancestry was detected. The minimal trace amount of “Native American” of <1% was most frequent among the remaining part (11/16).
However given the limited sample size of my survey this still does not rule out the possibility that among selected sections of Jamaicans somewhat more detectable amounts of Amerindian ancestry might yet be found. Especially among those living in isolated and mountainous rural areas where possibly indigenous Taino (a.k.a. Arawak) people and their mixed descendants might have persisted the longest. Even when in fact the highest “Native American” score in my survey (2%) was obtained by a mixed-race Jamaican with a most likely multi-generational urban background.
An important aspect to keep in mind is that due to recombination and dilution over the generations a genetic inheritance from 1 single ancestor only will eventually be “washed” out. This might occur already after 7-10 generations (see this useful graph). Theoretically speaking a Jamaican person who scores zero percent “Native American” on AncestryDNA might therefore still have had one single Taino ancestor living in the 1500’s. However without “replenishment” this ancestral line might already in the 1800’s not longer have been genetically detectable. Unless this Taino ancestor was from a direct maternal or paternal line. In this (fortunate) case uniparental markers (mtDNA or Y-DNA) would have been inherited and the confirmation of Amerindian lineage would be more straightforward. For a comprehensive discussion read the following article:
- Native American DNA Is Just Not That Into You (Roots & Recombinant DNA, 2015)
It is also a good thing to keep in mind that whenever Native American ancestry is reported for Jamaicans this will not automatically trace back to the original inhabitants of Jamaica. Leaving aside the possibility of any recent Hispanic lineage (introducing an external source of Amerindian geneflow). Instead of Taino lineage also Amerindian ancestry from Central America could be indicated by any “Native American” admixture score. Or even from North America due to English slave trading also involving Native Americans! This aspect is often not taken into full consideration. Even when for Jamaica it is known that in particular the Miskito people have been documented to bring along Native American captives. While they themselves or mixed descendants may also have settled freely in Jamaica in a few cases.
The same actually also goes for Hispanic Caribbeans. All too quickly for them an indigenous Taino/Arawak lineage is being assumed. While actually it is historically documented that after the local Taino populations had greatly diminished the Spanish settlers in Hispaniola started to import a large number of enslaved Amerindians from all around the Caribbean, Florida, South Carolina, Mexico and as far south as the Amazon river in Brazil!6 Certainly this history of widespread Native American slave trade also deserves to be told. And therefore such ancestral options must not be dismissed automatically. For further reading see:
- The Earliest Inhabitants: The Dynamics of the Jamaican Taino (L.G. Atkinson, 2006)
- ‘I Am Not Extinct’ – Jamaican Taino Proudly Declares Ancestry (The Gleaner, 2014)
- Mosquito warriors and fighters (Ole Time Sumting, 2017)
- Miskito Slaving and Culture Contact: Ethnicity and Opportunity in an Expanding Population (M. Helms, 1983)
- Colonists shipped Native Americans abroad as slaves (Brown University, 2017)
Taino lineage among Jamaican Maroons
Map 5.1 (click to enlarge)
“for 50 individuals with genealogical ties to Accompong Town. L-type mitochondrial haplogroups were observed in 96% of samples, and the remaining 4% belonged to haplogroup B2. […] Of the two B2 haplotypes, one matched haplotypes throughout the Americas and East Asia and the other matched only in East Asia” (Benn-Torres et al., 2014, p.1)
“Some historians posit that Taino intermarried with European and African colonists, including African Maroon populations thereby disappearing from historical records due to cultural assimilation rather than biological extinction” (Benn-Torres et al., 2014, p.1)
“[…] contemporary Maroons are not genetically isolated along matrilines from the general Jamaican populace. This similarity between Accompong Town Maroons and the general Jamaican populace is reflective of the shared ancestry among enslaved West Africans, and suggests that the Maroon community became less isolated with emancipation in 1834, increasing the possibility of gene flow into and out of Accompong Town.” (Benn-Torres et al.,2014, p.5)
“Fendt (2012) noted that, like Accompong Town Maroons, haplogroup L2a1 was the most commonly detected haplogroup in a sample of nearly 200 Ghanaians. The shared haplogroup L2a1 distribution pattern adds support to the connection between Ghanaians and Accompong Town Maroons.” (Benn-Torres et al. ,2014, p.5)
“However, despite difficulties in assigning the mitochondrial haplotypes to one specific
region within Africa (Ely et al., 2006), the genetic data from the current study in addition to previous genetic work, and historic sources garner support for a West and West Central African origin for Accompong Town Maroons.” (Benn-Torres et al. ,2014, p.5)
The quotations above pretty much speak for themselves. So I will not comment too much. Except for that I wish I also could have included survey participants of fully Maroon descent in my own research! I did see a couple of Jamaican AncestryDNA results for people who mentioned Maroon connections but usually along one family line only. Many of them turned out to be 100% African. Obviously the Maroons are a renowned minority group within Jamaican population. Given greater data collection in the future I might devote a separate blog post on them eventually.
The statements above are taken from a highly interesting study by a team of researchers which also includes Jada Benn-Torres. Originally from Trinidad she has been involved in several studies in relation to Jamaican genetics. And in fact also other parts of the English speaking Caribbean. With a dedicated focus on detecting the genetic legacy of Native Americans within the Caribbean. And she has certainly succeeded in that research effort by producing many insightful findings! Although I do think that her studies based on autosomal analysis have become obsolete due to very LOW resolution admixture analysis. Sadly also her most recent autosomal study (2018) on the genetics of the Maroons suffers from this flaw of not utilizing up-to-date and adequate methodology (see foot note 2). For more details on her fascinating work see these recommended links:
- Origins of Marronage: Mitochondrial Lineages of Jamaica’s Accompong Town Maroons (Benn-Torres et al., 2014)
- Prospecting the past: Genetic perspectives on the extinction and survival of indigenous peoples of the Caribbean (Benn-Torres, 2014)
- Taíno Symposium – Session 2 – Jada Benn Torres (Youtube, 2018)
Figure 5.2 (click to enlarge)
The above research findings are taken from a recent paper which makes for an useful addition to Benn-Torres et al. (2014) discussed above.
- Taino and African maternal heritage in the Greater Antilles (Bukhari et al., 2017)
It does not include the findings among Jamaican Maroons (n=50) which resulted in a higher but still quite subdued Native American haplogroup frequency of 4% (2/50). Instead it relies on the earlier study of Deason et al. (2012) which was based on a greater sample size (n=400) and probably also more reflective of Jamaica as a whole. In this study a Native American haplogroup frequency of 0.5% (2/400) was found. Which is very similar to the average of 0.1% Native American admixture I calculated for my Jamaican survey group, as shown in table 5.1. For another comparison which does include the Jamaican Maroon findings as well as other English speaking Caribbean islands see:
The Bukhari et al. study contains many insightful statements. Just focusing on what’s relevant for Jamaica this quote could be quite crucial (even when it does not explicitly mention the possibility of Taino inter-marriage with Maroons!):
“Since Haiti and Jamaica were under Spanish control from the late 1400s until 1655 and 1664, respectively, it is likely that the Spanish colonists departed with their mestizo families to other nearby Spanish colonies such as Cuba after the territories were ceded to France and England who kept the African slaves as a valuable commodity. This may have effectively removed most of the Taino mtDNA from what is today Haiti and Jamaica, and replaced it with a minority of male French and English migrants and their African slaves.” (Bukhari et al, 2017, p.38)
Leaving aside any petty-minded motivations it is actually worthwhile to compare Jamaicans as well as Haitians with Hispanic Caribbeans. That is in regards to their respective genetic legacy from the Arawak/Taino as well as other Amerindian populations deported to Caribbean. Because this may lead to valuable insights into the demographic evolution of Jamaicans as an ethnic group within its own right. In 2018 I already established that unlike my Jamaican & Haitian sample groups the Hispanic Caribbeans I surveyed did show a considerable Native American genetic component (see this table). Reported consistently above trace level, with no persons scoring zero %. This stands in sharp contrast with the findings I have presented above for my Jamaican survey group.
Intuitively I suspect this difference might be a striking confirmation of how Jamaicans overwhelmingly will trace most of their African ancestry to the 1700’s. When according to historical sources the indigenous Taino may have been extinct for many generations already (at least in the sense of viable communities consisting of “full-blooded” Taino). And in fact also their dwindling mixed descendants as well as deported Amerindians from other places only formed a small fraction of the population. Hence why the occasion for Amerindian admixture within the evolving Jamaican population simply would have been nearly non-existing. At least during the 1700’s into the 1800’s. When it must be remembered a considerable share of Jamaicans was still African-born! This ultimately might explain the very low level of Amerindian ancestry among currentday Jamaicans as detected by AncestryDNA. On the other hand this also underlines the overwhelming primacy of African roots for Jamaicans as a whole.7
Table 5.2 (click to enlarge)
6) Comparison with 23andme results being reported for Jamaicans
Figure 6.1 (click to enlarge)
Figure 6.2 (click to enlarge)
In 2018/2019 23andme rolled out a long awaited update of its African breakdown. Although it does have some (inherent) imperfections I think it is pretty decent. Based on how African DNA testers are getting described that is. I have since then started several surveys among Afro-descendants to find out how predictive the updated African breakdown on 23&me might be for them as well. Incl. also for Jamaicans! As a basic requirement my survey participants naturally need to have 4 Jamaican-born grandparents. Furthermore whenever I was informed I also collected details of their family’s parish/county background. To see if any correlation might exist. In particular in light of the substructure theme I discussed in section 2. Similar to my AncestryDNA survey I scaled everyone’s African breakdown to 100% to make the data inter-comparable. Follow the links below for my main research findings till now:
- Spreadsheet with Jamaican 23andme results
- Speadsheet with Afro-Diasporan 23andme results
- Spreadsheet with African 23andme results
- 23andme’s new African breakdown put to the test (blog post)
Due to an additional update on 23andme in October 2019 I have actually discontinued this survey for now. Although I will restart it again eventually. The current number of Jamaican 23andme results in my survey is 28. Which is of course not a very large sample size but still already quite useful to get some general ideas. And also helpful to compare with my more numerous Jamaican AncestryDNA findings.
First thing to be noted is that even when similarly labeled, ancestral categories on different DNA tests will not be perfect equivalents or measuring the same thing. Basically this is due to differences in reference samples as well as in algorithm applied by each separate DNA test. I will be blogging about this topic in greater detail in the near future. When comparing the African regional framework of AncestryDNA (2013-2018 version) with that of 23andme (2018/2019 version) the following aspects stand out to me right now:
3 West African regions on 23andme versus 5 West African regions on Ancestry
- “Senegambian & Guinean” is quite similar to “Senegal” on Ancestry but also covering Guinea Conakry. Probably somewhat more predictive over all. See also this map.
- “Ghanaian, Liberian & Sierra Leonean” is quite similar to “Ivory Coast/Ghana”. But with a clear shift to Sierra Leone and decreased coverage of Ghanaian DNA. Still pretty accurate for native West Africans from these countries. See also this map.
- “Nigerian” has nearly the exact same labeling as “Nigeria” on Ancestry. However it is clearly much more predictive, going by results for actual Nigerians. Still also covering DNA found in neighbouring countries. Which will be more impactful for especially Afro-Diasporans. See also this map.
- 23andme does not have any regions in place comparable to either “Benin/Togo” or “Mali” on Ancestry. This inevitably results in some shifts within the African breakdown.
- 23andme applies a more conservative approach than Ancestry. Whereby African DNA which cannot be classified reliably (given their limited set-up) is put under either “Broadly West African”, ‘Broadly Congolese & Southern East African”, or “Broadly African”.
3 Central & Southeast African regions on 23andme as well as Ancestry
- “Angolan & Congolese” is quite similar to “Cameroon/Congo” on Ancestry. However it is fortunately much more focused on describing genuine Central African lineage. With hardly any overlap with the Bight of Biafra as was the case for “Cameroon/Congo”. See also this map as well as this blog post.
- “Southern East African” is somewhat similar to “Southeastern Bantu” on Ancestry. But much better defined because 23andme also has 3 separate regions in place for Northeast Africa. Big improvement therefore when wanting to single out such lineage. Even when for Jamaicans this will generally be very subdued. See also this map as well as this blog post.
- “African Hunter-Gatherer” is pretty much the same thing as “South-Central Hunter-Gatherers” on Ancestry. In both cases very minimal for almost all people and hardly relevant as these scores are often to be traced back thousands of years rather than within a genealogically meaningful timeframe.
- Actually 23andme also has 3 additional categories for Northeast Africa as well as 1 for North Africa. But I will not discuss them in more detail as these regions rarely show up for Jamaicans (and if they do nearly always <1%).
Continental breakdown on 23andme
Table 6.1 (click to enlarge)
In table 6.1 above my main findings can be seen when it comes to the continental breakdown of my Jamaican survey group (n=28) on 23andme. Of course very interesting to compare with my more numerous Ancestry findings (n=100). Follow link below:
Actually the differences are quite minor overall speaking. Also keeping in mind that my 23andme survey group is about 4 times smaller. And some of the statistical values might therefore shift somewhat with greater sample size (but from my survey experience over the years not by much after n=30 is reached). Many pundits, even the skeptics, nowadays agree that continental admixture analysis has become pretty accurate in the last few years. I suppose therefore that this outcome already serves as some form of independent corroboration of my main findings. In short:
- African admixture is clearly predominant. The median of 86% on 23andme probably being more representative due to a few low outliers. The most frequent range of 80-90% African (9x) on 23andme was actually closely followed by the number of people in the 90-100% range (7x).
- No 100% African scores in my 23andme survey so far. But one person did get very close (see figure 6.1). Due to its algorithm 23andme offers a finer detection of small admixture amounts than on Ancestry. Usually quite accurate or at least indicative of something distinct. But at times still confused by “noise” admixture as well. Also minor “unassigned” might get in the way. Actually even for native Africans it is not always that easy to appear as “100% African” on 23andme 😉 (99.9% being more common, see this sheet).
- I have not yet kept full track of the European breakdown for Jamaicans. It is clearly mainly Northwest European though. With “British & Irish” almost always in first place. In line with predominantly British origins. Whenever possible I did take notice of the “Recent Ancestral Location” (RAL) mentioned within the results. Based on matching strength and similar to Ancestry’s Migration a.k.a. Genetic Community tool. “United Kingdom” was reported eight times as RAL. While “Ireland” was mentioned twice. Also interestingly one person received a distinctive score of around 3% “Ashkenazi Jewish”. Even when for others it was either absent or very minimal.
- A mentioning of specific Jamaican parishes as RAL also occurred quite often. Specific African RAL’s did not appear though. But for one person with the highest East Asian admixture “Mainland China” was given as RAL (see figure 6.2).
- A slightly higher group average for “South Asian” admixture on 23andme (4.2% versus 2.2% on Ancestry). But this is caused by a few high outliers. On Ancestry the maximum score being 29%. While among my 23andme survey group actually two persons had around 50% South Asian admixture. After 23andme’s latest update in October the resolution for South Asia has greatly improved btw.
- My East Asian findings on 23andme are very similar to the ones on Ancestry. I actually also included Southeast Asian scores. As in most cases (southern) Chinese admixture will be indicated (see figure 6.2 for someone who is confirmed 1/4 Chinese). But I suppose in a few cases also Malagasy lineage might be suggested. Especially when “Filipino & Austronesian” appears on its own. Quite amazing when genuine because for Jamaicans this will usually be traced back to the early 1700’s or even late 1600’s!
- Going by group average (0.3% versus 0.1%) as well as most other statistical measures the very minimal yet still detectable Amerindian admixture scores are again a close reflection of my Ancestry findings. Interestingly though one 23andme survey participant did show a notably elevated level of 3.8% “Native American” admixture! I have received confirmation of this person’s fully Jamaican background. And actually I have also seen his Ancestry results which show no indication of recent non-Jamaican lineage. While actually his Native American score on Ancestry is also a remarkable 4%.
Comparison with the Anglo-Caribbean & other parts of the Diaspora
Table 6.2 (click to enlarge)
Table 6.3 (click to enlarge)
Lots of data on display in tables 6.2 & 6.3 above. Based resp. on my 23andme surveys for populations across the Afro-Diaspora. As well as my AncestryDNA survey (2013-2018) for various parts of the so-called Anglo-Caribbean. These latter results have been collected by me for several years already. But regrettably I was not able to attain a similar sample size as for my Jamaican survey group (n=100). Only the Guyanese and Barbadian survey groups reaching a somewhat decent sample size.8 Still taken all together the comparison should be insightful. I can already say that the robustness and internal coherency of my overall data set seems to be largely confirmed by what is to follow. I will just focus on how it directly relates to my Jamaican research outcomes. As I intend to blog in greater detail about these overviews in the near future. See also :
- Anglo-Caribbean AncestryDNA results (2013-2018) (under preparation)
Take note that for my 23andme findings in table 6.2 I have applied an additional macro-regional framework of my own making. In order to allow some broader patterns to show up more clearly. Similar to what I did during my AncestryDNA survey (Upper Guinea, Lower Guinea, Central/Southeast Africa). But not exactly in the same way due to the fact that 23andme’s West African resolution is less specific (K=3) than on Ancestry (K=5). But on the other hand 23andme does have separate regions for Northeast Africa which were not available on the 2013-2018 version of Ancestry. This part of Africa is pretty much irrelevant though for understanding the genealogical roots of the Trans-Atlantic Afro-Diaspora.
I will just be going along the row when discussing the regional group averages for my Jamaican survey groups on 23andme and Ancestry. Starting with the 2.6% “Senegambian & Guinean” score which is practically equal to the 2.4% “Senegal” on Ancestry! As already discussed in section 1 Jamaica’s strictly Senegambian lineage is minimal going by several measures. This is also confirmed by contrasting with other parts of the Afro-Diapora which are known to have a much greater Upper Guinean genetic imprint. Cape Verde in the first place but also the Hispanic Caribbean as shown in table 6.2.
Very interestingly also within the Anglo-Caribbean AncestryDNA context (table 6.3) this is reflected by the stand-out score of 10.9% “Senegal” among Bahamians. Which is again pretty close to their 9.1% “Senegambian & Guinean” group average on 23andme! Even when the sample size is quite minimal the distinctive African regional origins for the Bahamas are already demonstrated. Going by known slave trade patterns (see this table taken from the TAST database) as well as historical migrations these are more similar to South Carolina than to other parts of the Anglo-Caribbean. See also their Central African scores for example!
Next in line is the 25.3% “Ghanaian, Liberian & Sierra Leonean” group average for Jamaicans on 23andme. Eventhough not exactly measuring the same type of DNA it is still remarkably close to my finding of 24.7% “Ivory Coast/Ghana” for Jamaicans on AncestryDNA! Then again when comparing for example with African Americans it is very likely that for Jamaicans this region will be more so describing Ghanaian DNA than either Liberian or Siera Leonean DNA. While for African Americans it might be more evenhanded. Even when all three are certainly plausible ancestral options for both Jamaicans and African Americans, just in different proportions (going by slave trade patterns). On the other hand Ghanaian DNA might be underestimated to some degree by this category. And also it might include some genetic affinity shared with “Mali” on Ancestry. So all in all perhaps not as clear-cut as it seems at first.
In previous comparisons of Afro-Diasporan AncestryDNA results I have singled out Jamaica as having the highest contribution of Ghanaian DNA. But this was always without taking into account other Anglo-Caribbean populations. As can be seen in both tables 6.2 & 6.3 especially the Barbadians and also Guyanese seem to have a higher degree of Ghanaian lineage. Around 31-33% for both “Ivory Coast/Ghana” as well as for “Ghanaian, Liberian and Sierra Leonean”. Although actually for Guyana also a substantial degree of Liberian and genuine Ivorian connections might be hinted at (due to Dutch slave trading). I hope to clarify this issue by looking closer into the African DNA matching patterns for Jamaicans and other Anglo-Caribbeans eventually.
We now arrive at the largest African component identified by 23andme for Jamaicans: 50.1% “Nigerian”. This might come across as somewhat inflated at first sight. Although a principal ranking of Nigerian lineage in itself is according to expectation (see discussion in section 1). I was myself also a bit taken aback when learning that unlike on Ancestry Jamaicans will almost always have “Nigeria” show up in first place on 23andme and with hefty scores as well. This consistency in primary regional scores seems to contradict the greater variety on Ancestry (see chart 1.1). Which frankly was much more convenient & fitting when wanting to align with Jamaica’s regional roots within Africa.
There are several things to keep in mind though which do help to explain this outcome more or less. First of all it should be remembered that 23andme only has 3 West African regions instead of 5 on Ancestry. Especially “Benin/Togo” is greatly missing in the equation. Secondly “Nigeria” on Ancestry (2013-2018) tended to underestimate Nigerian ancestry. Additional Nigerian DNA being covered by first of all “Benin/Togo” and to a lesser degree also “Cameroon/Congo”. I had therefore already speculated elsewhere that the true proportion of Nigerian ancestry for Jamaicans might be around 30%-40%, on average.
It is essential to be aware that so-called “Nigerian” on 23andme is also including genetic connections with Gbe speakers (incl. the Fon from Benin but also the Ewe from Ghana/Togo). And probably to a lesser degree also genetic ties with (western) Cameroon. The combined shares of “Benin/Togo” and “Nigeria” (21.3% +22.5%) already make for a good approximation of my Jamaican survey group’s 50.1% “Nigerian” score on 23andme. Probably the remaining part (5-10%) is being drawn from the 17% “Cameroon/Congo” Jamaican group average.
Still interesting to also look more closely at the relative position of “Nigeria” scores for Jamaicans when compared with my other survey groups. Instead of having the reputation for greatest degree of Ghanaian ancestry it might be more justified to say that Jamaicans have some of the highest levels of Nigerian DNA across the Diaspora (see also this table). Possibly only trumped by Belizeans in my survey so far. Although actually on 23andme for 12 Belizeans a lower “Nigeria” average of 36% was reached (see my sheet instead of table 6.2).
Finally the 7.3% Central & Southeast African component for Jamaicans seems rather subdued. It is mostly based on a 5.4% “Angolan & Congolese” score. Which is clearly much less than the 17% “Cameroon/Congo” score on Ancestry. The difference arguably being caused by genetic affinity hailing from the Bight of Biafra hinterland (southeast Nigeria & Cameroon). Which actually would prove my suspicion that especially for Jamaicans “Cameroon/Congo” might often indeed be suggestive of genuine Cameroonian lineage. In contrast with the greater part of the Afro-Diaspora.
The minuscule Jamaican group average of 0.7% “South Eastern African” goes very well with equally minimal recorded slave trade from Southeast Africa (0.1%, see table 1.2). The 3% “Southeastern Bantu” average on Ancestry and especially higher outliers of “Southeastern Bantu” among my Jamaican survey group most likely being misreadings. This decrease could therefore be a major improvement. Helpful to distinguish genuine Southeast African lineage, even when it will be very subdued most of the time.
All in all it does appear that Central African lineage might be underestimated for Jamaicans on 23andme (going by expectations of around 16% discussed in section 1). Probably most of the missing part is being described by 23andme as “Broadly African” (3.1%). And perhaps even also under “Broadly West African” (11.5%) some misread genetic affinity with Central Africans might be found. Then again the West African predominance among Jamaicans (89.5%) along with Barbadians (92.4%) and Cape Verdeans (92.8%), is historically very much justified! And also already established during my AncestryDNA survey (2013-2018). Albeit that due to finer West African resolution on Ancestry I was able then to further distinguish this into a Lower Guinean preponderance for Jamaicans as well as Barbadians and a Upper Guinean preponderance for Cape Verdeans.
7) Current update of AncestryDNA’s Ethnicity Estimates
Map 7.1 (click to enlarge)
Starting in October 2019 Ancestry has been rolling out a new update of their Ethnicity Estimates. Because Ancestry has such a huge customer database (>15 million!) it is expected it might take several months before everyone receives their updated results. As I have said before your DNA results are only as good as the next update. So it is best not to get too attached to them 😉 Given scientific advancements and a greater number of relevant reference samples hopefully a greater degree of accuracy may be obtained in the near future. But naturally no guarantees are given that this will indeed be the case. After all Ancestry’s update in 2018 arguably has been a downgrade rather than providing any meaningful improvement. At least in regards to the African breakdown. For more details about the current update:
- Ancestry® Expands Reference Panel to Deliver More Precise Results and New Regions (Ancestry Blog 2019)
- New maps for African regions on AncestryDNA
I intend to do a comprehensive survey among African DNA testers from all over the continent to establish a more solid basis for judgement. So I will refrain from detailed comments for now. However I can already say that most of my suggestions for improvement I blogged about in 2018 still stand… Due to wild fluctuations in just two years it will also be understandable if some people will experience update fatigue. Although for others any update will be welcomed and highly anticipated as some kind of Christmas present 😉 And to be sure for many people (depending on specific background) this update may yet be beneficial. See also these links:
- How to deal with updated DNA test results (Tracing African Roots, 2018)
- Suggestions for improving the African breakdown on AncestryDNA (Tracing African Roots, 2018)
- Ethnicity Update FAQ (Ancestry)
Updated results for 10 Jamaicans compared with 2013-2018 version
Table 7.1 (click to enlarge)
Figure 7.1 (rightclick and open in new tab to enlarge)
My Jamaican survey findings described in this blog post have been based on the 2013-2018 version of AncestryDNA’s Ethnicity Estimates. It remains to be seen how well my present findings will correspond with any newly calculated AncestryDNA results. Will they be rendered completely obsolete or may they still contain lasting insights about the approximate ancestral composition for Jamaicans? I expect that the continental breakdown is bound to remain quite steady. As these scores tend to be most reliable in admixture analysis nowadays. And as shown in section 6 my main outcomes have also been independently corroborated already by my 23andme survey findings.
In order to get a general idea I have kept score of how 10 of my Jamaican survey participants have been impacted by the 2019 update. See screenshots and table 7.1 above. Actually I have also kept score of their results after the 2018 update. Scaling their African breakdown to 100% to make things inter-comparable. As expected there was only slight variation in their continental breakdown. However otherwise there were some wild fluctuations within the African breakdown. See my spreadsheet below for complete details for each of my 10 Jamaican survey participants:
- Spreadsheet with 10 Jamaican AncestryDNA results before and after 2018 & 2019 updates
- Table 7.1, including also impact of 2018 update
The main flaw of the 2018 update was the consistent appearance of heavily inflated “Benin/Togo” and “Cameroon, Congo, & Southern Bantu” scores. Resulting in a dramatic decrease of “Nigeria” amounts. Especially for Jamaicans completely unwarranted and in contradiction with their known African regional roots (see section 1). For my 10 Jamaican survey participants it lead to a ridiculously low group average of barely 2% “Nigeria” after the 2018 update. Even when in the 2013-2018 version it had been a reasonable 23%. And for 4/10 persons “Nigeria” had been the primary region in their African breakdown. As can be seen in table 7.1 the current 2019 update has brought back “Nigeria” but with a heavy kick! Seemingly an over-correction seems to have taken place. Although the group average (n=10) of 55% “Nigeria” does come close to the 50% “Nigerian” average I obtained in my 23andme survey (n=28, see table 6.2).
The new group averages for “Benin/Togo” and “Cameroon, Congo & Southern Bantu” have been more or less restored to their level in the 2013-2018 version. They have even diminished somewhat. Which is probably also more realistic and in line with regional expectations as discussed in section 1. Notice for example the 0% group average for “Eastern Bantu”, while previously there had been a 2.7% group average for “Southeastern Bantu”. Also the “Senegal” and “Mali” amounts are not showing any major changes. Although especially “Mali” has been steadily on the increase, also compared with the 2018 update. Probably due to the renewed addition of Malian samples within Ancestry’s Reference Panel. It was increased from only 16 to 169 during the 2018 update. And now it has again been greatly augmented to a number of 413 (see this link).
But “Ghana” (no longer “Ivory Coast/Ghana”!) has taken a sharp turn for the worse it seems. As can be seen in table 7.1 the group average fell from 33% in the 2013-2018 version to barely 5% after the 2019 update! Actually already after the 2018 update “Ivory Coast/Ghana” amounts practically halved (16.9% group average). And this trend has continued even more so after the 2019 update. Possibly as a consequence of the removal of Ivorian samples within Ancestry’s Reference Panel. As seems to be suggested by the name change. And in fact also the new number of samples for “Ghana” has become less when compared with the 2018 update (decrease from 124 to 109). Probably Ancestry’s algorithm also has something to do with this outcome though.
If you look into the individual results it seems that often the increase of “Nigeria” has come at the price of a steep decrease in “Ghana”. This is especially apparent when reviewing the five results which used to have a primary “Ivory Coast/Ghana” score. A majority of these 10 samples (5/10) which is in line with my overall survey findings (n=100, see chart 1.1). However after the 2019 update “Nigeria” is consistently being reported as primary region (similar to 23andme). In clear contradiction of how Jamaicans are mainly Ghana & Nigeria hybrids based on historical and other clues (see section 1). Possibly Nigerian lineage prevailing somewhat over all. But certainly showing much greater variation than reflected in this newly updated African breakdown. And naturally for many Jamaican individuals Ghanaian lineage could very well be primary, as was reported in Ancestry’s 2013-2018 version.
I am not saying Ancestry’s 2013-2018 version was without its own flaws. Throughout my AncestryDNA survey I have frequently pointed out its many limitations and shortcomings. However compared with this 2019 update it seems that the African breakdown obtained with the 2013-2018 version still offers the best fit for Jamaica’s known regional roots within Africa (see section 1). Nonetheless there are several redeeming aspects about this 2019 update as well. It could very well be that the European breakdown will again show some improvement. Going by these 10 updated Jamaican results I was particularly impressed by the way Scottish lineage is being pinpointed. Even for one person (JAM25) who only had 5% total European admixture!
One might easily get the impression that each update on Ancestry only results in more random changes. But do notice that the macro-regional breakdown has remained pretty much the same. Especially the 79% share of Lower Guinea stayed exactly the same for my 10 Jamaican survey participants! In fact most of the major changes appear to have been an internal reshuffling within the Lower Guinean macro-region. Indicating that Ancestry might not be able yet to finetune between neighbouring and genetically closely related regions. However in the greater scheme of things this less specific macro-regional framework does do justice to what we know about Jamaica’s main African roots hailing from the area in between Ghana and Nigeria. The somewhat decreased proportion of Central African lineage probably also to be counted as an improvement over the 2013-2018 version.
As always it is essential to be fully informed about both strengths and weaknesses for each separate aspect of DNA testing. Therefore I will need more information for a final assessment. Still sadly it seems that this particular update by Ancestry will again not lead to any substantial improvement for understanding the African roots of Jamaicans. But this does not mean that improvement may still be forthcoming, if not on Ancestry than elsewhere! I am specifically referring to admixture analysis a.k.a. ethnicity estimates. As my main discussion of 100 Jamaican AncestryDNA results (2013-2018 version) as well as my previous AncestryDNA survey findings have demonstrated that potentially this tool can be very useful in unlocking the secrets of main African regional lineage for Afro-Diasporans!9
Either way admixture analysis can only get you that far due to inherent limitations. However by conducting careful follow-up research (DNA matches, historical plausibility, genetic genealogy etc.) I do think you can increase the chances of learning more about the likely ethnic backgrounds of your African ancestors. In particular my advise would be to: Look into your African DNA matches for more insight! See also my tutorial below as well as my recently started service which will provide you with a African DNA Matches Report.
In fact this is exactly what I intend to do as a follow-up to this blog post. When I will be analyzing the African DNA matching patterns for 30 Jamaicans. And actually I will also cover South Asian and Jewish matches. Again contrasting the outcomes with historical plausibility as well as my main findings in this blog post. As I will continue aiming for combining insights from various fields. Always looking for correct interpretation. Critical but also staying open-minded and careful not to be dismissive when informational value can still be obtained.
- African DNA matches reported by Ancestry for 30 Jamaicans (under preparation)
8) Screenshots of individual results & Youtube videos
As far as I know and was able to verify all of these screenshots below are from persons with 4 Jamaican-born grandparents. These results form the basis of my survey findings together with results which I retrieved from Youtube and by PM. I like to thank again all the persons who kindly agreed to share their results with me!
Ranked on primary region (rightclick and open in new tab to enlarge)
See also this playlist
1) I like to express my sincere gratitude to all the persons who made my research possible! Most results used in my survey were shared with me directly by invitation by the DNA testers themselves. In addition I have also included several results which were shared on Youtube and social media. In section 8 I have used a few screenshots taken from public websites. As I found them to be of potentially great educational benefit for others. I have asked for prior consent whenever I could but regrettably wasn’t able to do so in all cases. I have naturally taken great care to cut away any name details in order to safeguard everyone’s privacy. Apologies in advance to anyone who recognizes their results and is not comfortable with this blog page featuring them. Please send me a PM and I will remove them right away.
2) In fact many interesting studies have been published on Jamaican genetics over the years. Although usually not reaching the same finer regional within-Africa resolution as available on AncestryDNA. Based both on autosomal analysis and haplogroups (see this link to learn more about the difference). Some of them are however already outdated due to rapid scientific advances in especially autosomal admixture analysis. Which is why I will not be referencing these papers in the main text. I will however compare some of my findings with the haplogroup studies which are still relevant today. Below follows an overview of the studies I am aware of. As well as my personal review. Not meant to be exhaustive.
- Admixture and Population Stratification in African Caribbean Populations (Benn-Torres et al, 2008)
- An anthropological genetic perspective on Creolization in the Anglophone Caribbean (Benn-Torres et al., 2013)
- Investigating the “Taíno” ancestry of the Jamaican Maroons: a new genetic (DNA), historical, and multidisciplinary analysis and case study of the Accompong Town Maroons (Benn-Torres & Fuller, 2018)
- The genetic structure of populations from Haiti and Jamaica reflect divergent demographic histories (Simms et al., 2010)
- Mitochondrial and Y Chromosome Diversity in the English-Speaking Caribbean (Benn-Torres et al., 2007)
- Origins of Marronage: Mitochondrial Lineages of Jamaica’s Accompong Town Maroons (Benn-Torres et al. 2014)
- Interdisciplinary approach to the demography of Jamaica (Deason et al., 2012)
- Y-chromosomal diversity in Haiti and Jamaica: Contrasting levels of sex-biased gene flow (Simms et al., 2012)
Although pioneering at the time I find that the autosomal studies have become obsolete and misleading in regards to their main findings. Which are based on an extremely LOW number of so-called AIM markers. Even for Benn-Torres & Fuller (2018) which is a pretty recent study regrettably very outdated admixture software was used (going back to 1987 or even 1975 apparently!) and only 13 autosomal markers to make the continental distinction (Benn-Torres & Fuller, 2018, p.57). To their credit the authors do concede as much when stating that for future research efforts “newly collected DNA samples should undergo higher-resolution analysis” (Benn-Torres & Fuller, 2018, p.70). One would wish this was already done for the 2018 study which had such a promising research topic!
To put things in perspective commercial DNA testing companies such as Ancestry and 23andme test for around 600-700,000 autosomal locations (SNP’s). And also in other ways their methodology might be more sophisticated than applied by the autosomal studies referenced above. The actual number of AIM’s (a subset) is usually not disclosed as far as I know. However it is bound to be more than just a dozen or so… See also this overview. According to this paper :
“Panels that consider >400 AIMs capture genome ancestry reasonably well, while those containing a few dozen AIMs show a large variability in ancestry estimates“.
And according to Ancestry’s latest White Paper for its Ethnicity Estimates:
“After establishing and validating the reference panel, the next step is to estimate a customer’s ethnicity by comparing over 300,000 single nucleotide polymorphisms (SNPs) from their DNA to those of the reference panel. We assume that an individual’s DNA is a mixture of DNA from the 60 populations represented in the reference panel.”
“At AncestryDNA, we use microarrays to obtain DNA data from customer samples. We look at over 700,000 individual locations on the DNA (SNPs) and determine the nucleotides at each position. For example, we may see an A and a T at position 1, a G and a G at position 2, and so on. We use around 300,000 of these SNPs in the ethnicity estimate.“
Autosomal research findings becoming outdated is unavoidable due to the circumstance that admixture analysis has become a quickly developing field. Allowing for a much finer detailed resolution in genotyping. As well as an ever increasing availability of relevant reference populations. But again these changes are sadly not reflected in the studies referenced above. These developments did however greatly impact the updated DNA results released by AncestryDNA and 23andme over the years. Generally leading to major improvement (although not always 😉 ).
Aside from this aspect I do also have some minor points of criticism in regards to the historical context given in some of these studies. Which at times I find rather lacking, incomplete or even misrepresenting actual history. Although to be honest this is quite common in many DNA studies. See for example the explanation given by Simms et al. (2010, p. 60) on the nature of European admixture (leaving out the possibility of non-consensual unions). Or also their incorrect mentioning of a share of almost 10% slave trade from Southeast Africa into Jamaica (pp. 49-50). When in reality it is pretty much common knowledge among historians that this was less than 1%! See table 1.2, based on the comprehensive and widely referenced TAST database. The haplogroup study by Simms et al. (2012) is still very informational though.
The studies performed by Benn-Torres et al. (2008, 2013 & 2018) are undoubtedly well-intentioned and also using a fresh perspective on the ethno-genesis of the Anglo-Caribbean. In particular with an engaged focus on the genetic survival of the Taino people (see section 5). The first two studies involved Rick Kittles from African Ancestry btw. Still also in these studies some questionable statements are being made. When wanting to explain their low European admixture findings for Barbados for example they mention a relatively subdued European migrant presence on that island (Benn-Torres et al, 2008, p.95). But actually according to several historians Barbados used to have one of the biggest shares of White settlers in the British West Indies! See for example Barry Higman’s Slave Populations of the British Caribbean, 1807-1834 (1984). Also Dominica’s presumed unique social standing of Mulatto or otherwise mixed-race people when compared with other parts of the Anglo-Caribbean seems a bit overstated (Benn-Torres et al., 2013, p.141). At least for Jamaica where racial distinctions such as Mulatto, Sambo, Quadroon and even Mustee were very much current during the Slavery era! And this had significant social consequences, both for enslaved and freed people. See for example Higman’s “Slave Population and Economy in Jamaica, 1807-1834” (1976).
Furthermore it is essential to know that in Benn-Torres et al. (2008 & 2013) a peculiar choice was made not to include any Asian reference samples in their main modeling framework. Inspite of Jamaica’s history of Asian migrant labourers and known inter-ethnic unions! Understandably therefore this omission resulted in greatly overestimated Native American admixture. Completely out of line with admixture analysis performed on a more robust basis. And unnecessarily giving a restricted scope on Jamaica’s genetic diversity. On the other hand even if Asian samples had been included a flawed impression of Jamaica’s continental/racial admixture might still have occurred. Simply due to the rather basic state of admixture analysis at that time.
This tended to be the case for personal DNA testing companies as well in this time period (ca. 2005-2013). This early era of admixture analysis produced many distorted and misleading results. I myself received my first DNA test results with 23andme in 2010. And like almost all Afro-descendants at that time I was given a puzzling Asian score of around 3%. But for some African Americans it could get as high as 10%. Some people interpreted it as Native American DNA. However in subsequent updates by 23andme in 2011 & 2013 this turned out to be merely a misreading of actual African DNA. Principally due to the lack of more diverse African reference sampling (beyond the ubiquitous Yoruba samples).
The same issue also applies to many widely published DNA results of African American celebrities in this pioneering era of DNA testing (prior to 2013). Such as for Snoop Doggy Dog or Oprah Winfrey. Or also British-Jamaican athlete Colin Jackson. Their test results are still being quoted. Even when they were originally taken around 2005-2010 and therefore BOUND to be greatly outdated. Even when sub-continental (regional) admixture analysis might still be very variable. It is generally agreed that continental admixture analysis has now become quite reliable and accurate, but only during the last 5 years or so. If these persons were to take a new DNA test today undoubtedly in many ways their results would look different. And going by what is typical for African Americans, the odds will be quite high that their Native American admixture will be much lower (<2%) while it seems likely that especially Oprah’s European admixture might have been underestimated. Also the 7% “Taino” admixture reported for Colin Jackson in 2006 might then actually turn out to have been (mostly) a misreading.
3) Even when your minor European DNA does (partially) happen to be linked to the Slavery Period, it is prudent to remain open-minded. Beyond a shadow of a doubt slavery itself was a dehumanizing institution which incited cruelty among the slaveholders and their aides. For Jamaica in particular we are well informed about the gruesome details, incl. also massive sexual abuse. See for example the journals of Thomas Thistlewood. However even under these unlikely circumstances history teaches us that there must have been room for human agency as well: individuals who went against the structurally defined norm. And when engaged in inter-racial unions these individuals may have very well have acted out of other motivations than the assumed ones. Including genuine long term mutual affection besides short-lived mutual attraction, opportunism or self-improvement. Especially one might imagine when European indentured labourers were involved.
4) Full references for these haplogroup studies:
- Interdisciplinary approach to the demography of Jamaica (Deason et al., 2012)
- Y-chromosomal diversity in Haiti and Jamaica: Contrasting levels of sex-biased gene flow (Simms et al., 2012)
“The great majority of the profiles observed in Jamaica could be allocated to L sub-Saharan haplogroups (97.5%) […]. There are several lineages most likely belonging to different sub-clades of the Pan-Asian haplogroup D4, and a member of haplogroup M30c1 of presumable Indian origin.” (Deason et al., 2012)
“the current investigation reveals that genetic signatures [i.e., haplogroups H-M69 (0.6%), L-M20 (0.6%), O-M175 (3.8%), and R1a1a*-M198 (3.1%)] of these groups in the Jamaican gene pool are limited (Table 2), likely the result of the relatively low levels of genetic admixture between the Asian immigrants and the local inhabitants.” (Simms et al, p.628)
5) There have been earlier attempts to estimate the Native American admixture among Jamaican sample groups. However due to outdated methodology and the omission of essential Asian reference samples I consider these studies to be obsolete. See also foot note 2 for my assessment.
6) I am in full agreement that the Taino legacy lives on across the Caribbean. Both genetically and culturally I would say. However that does not rule out that *in addition* also other types of Amerindian DNA (non-Taino) have been passed on among especially Hispanic Caribbeans. The enslavement of tens of thousands if not hundreds of thousands of Amerindians from all around the Caribbean Basin (as far north as South Carolina and as far south as the Amazon river in Brazil, but also places like Yucatan, Honduras & Venezuela) has been plentifully documented. Many of them were to be sent to the Hispanic Caribbean during the 1500’s especially. A lot of cruelty and suffering was caused by these forced migrations and circulation of enslaved Amerindians so I see no need to downplay it. Instead I think their very likely genetic inheritance should be acknowledged just as the much more celebrated Taino legacy is. For more details consult chapter 5 of this very recent study:
- Surviving Spanish Conquest: Indian Fight, Flight, and Cultural Transformation in Hispaniola and Puerto Rico (K.F. Anderson-Córdova, 2017)
As well as this dissertation (especially chapter 6):
- Indian Harvest: The Rise of the Indigenous Slave Trade and Diaspora from Española to the Circum-Caribbean, 1492-1542
7) Naturally more research is needed to establish to what extent Taino or otherwise Native American bloodlines have been preserved in Jamaica. Although the current evidence is already greatly indicative I would say. Because of identity politics one must be careful to not use genetics as a tool to divide but rather to enlighten and reveal hidden histories. Going beyond the question of biological extinction and the “numbers game” genetically speaking it seems already incontestable that Amerindian DNA is still present across the Caribbean and also in Jamaica. Although a second crucial question then becomes how much of this could possibly be Taino related and how much of it could be due to Amerindian captives circulating all across the Caribbean?
Proportionality still matters in all of this as well I would argue. Determining the largest regional components, on average, for each Afro-descended nationality has been a primary research effort during my AncestryDNA survey. As they can be considered to have the highest reliability at this stage (even when taken as mere proxies). They are covering a wider span of your ancestral make-up. And might also be confirmed independently by historical sources.
Then again it is human nature to be intrigued by seemingly mysterious details, even at the risk of getting caught up in them. Also it seems to be not just a sign of the times that “exotic” ancestry is often thought to be appealing 😉 Even when Native American DNA of course is very much indigenous and therefore tends to have great emotional significance for people of all backgrounds throughout the Americas! Even when I might have termed my Native American findings as “minor” or even “statistically insignificant”. Of course this is only strictly relatively speaking.
The actual ancestors behind even trace amounts of DNA are not of minor importance themselves! They might still evoke some personal interest for anyone who receives the results. I naturally respect this. Still I also think it will be prudent to at least acknowledge the predominant ancestral components you are made up of. As the people associated with these greater parts of your ancestry will have contributed the most to who you are, at least genetically speaking. This goes even more so when attempting to understand the evolution of Jamaica’s population in its totality and in the bigger scheme of things.
8) Through my interaction over the years with many potential West Indian survey participants I have learnt that my strictly “4 grand parents from the same island” criterium was difficult to achieve. Mainly because of rampant inter-island migrations. Although at times Caribbean inter-island unions also occurred in either North America or the UK. These inter-island migrations are actually a constant of Caribbean history. Not only being very common in our modern age, but also throughout the colonial era. Notoriously of course also involving inter-colonial slave trade. But also at times voluntary migrations and even runaway captives.
This is something I noticed especially for Trinidad and to a lesser degree also Barbados. However for Jamaica it seems much fewer incoming migrations from other islands have taken place. At least in the last 100 years or so. In many cases I was able to verify by either personal confirmation or also public family tree that many Jamaicans happen to have a very long established presence on the island, on all family lines. Aside from Jamaicans also being the biggest Anglo-Caribbean population this enabled me to assemble such a relatively large survey group (n=100).
9) I like to underline that despite inherent limitations my AncestryDNA survey (2013-2018) did produce several potentially insightful findings. Not only for improving the interpretation of the results of Afro-Diasporans, incl. Jamaicans but also for Africans. Despite seemingly contradictory updates on Ancestry in 2018 & 2019 I do firmly believe that regional admixture DOES matter! Basically the differences caused by subsequent updates are a result of changes in the African Reference Panel being used by Ancestry to produce their estimates. Plus also their algorithm which has undergone a major change in 2018. As I argue in this blog post I still believe that the original 2013-2018 version provides the best fit for Jamaica’s known regional origins. Even when again this version also had its own shortcomings.
According to many pundits only continental admixture is to be taken seriously in DNA testing. Sub-continental, a.k.a. ethnicity estimates, a.k.a. regional admixture only being fit for entertainment purposes. I myself have never taken this stance. Preferring to judge each case on its own merits. Attempting to maximize informational value despite imperfections and avoiding source snobbery. Which is why I have conducted my AncestryDNA surveys among both Africans and Afro-descendants in the past. Applying an additional macro-regional framework for extra insight (Upper Guinea, Lower Guinea, Central/Southeast Africa).
Basically combining interrelated & neighbouring genetic regions in order to allow certain regional patterns to show up more clearly. This turned out to be particularly helpful when wanting to explore any rough correlation between aggregated AncestryDNA results for Afro-Diasporans and slave trade patterns. Which I indeed established for the most part. I hope to have demonstrated in this blog post that such an approach can also be beneficial when wanting to reach deeper understanding of Jamaican genetics.
Regional estimates require correct interpretation and each updated version as well as each separate DNA testing company should be judged on its own terms. Then again these admixture results can only take you that far. My advise therefore is to also look into your African DNA matches, as well as historical plausibility and just plain genetic genealogy for greater combined insight.