The Unz Review - Mobile
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
Email This Page to Someone

 Remember My Information

Authors Filter?
Razib Khan
Nothing found
 TeasersGene Expression Blog

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New Reply
🔊 Listen RSS

Today there was a short article in Discover on a paper published last spring on the models for the settling of Madagascar. I didn’t pay too much attention when the paper came out for two reasons. First, it focused on Y and mtDNA, and I’ve been playing with Malagasy autosomes. Second, it seemed a ridiculously brutal computational attack on a question which seems to have a straightforward intuitive explanation: yes, Madagascar was settled by a small founding group. With hindsight I may have spoken too soon, or passed judgement too hastily. Looking at the paper the explicit model building of demography does still seem like overkill, but they obtain some important precision here. The phylogenetics and the archaeology align nicely.

Though the authors of the article talk about future directions, I think we will find that the Malagasy originate from a small group of Malayo-Polynesians who did find themselves stranded on Madagascar (later to absorb African admixture). This is not controversial. Rather, when I came to this position with enough solidity I began to look at the cultural anthropology of Madagascar. In particular, what do the Malagasy remember of about their own past in Southeast Asia? From what I could tell (the literature on Madagascar is not too rich in English) the Malagasy don’t recall much. This is important, because it tells us just how fragile oral memory can be when you have a major geographical and demographic rupture. The influence of Sanskrit is apparently evident within Malagasy, attesting to the early period if Indic influence in Southeast Asia. But the Malagasy are not part of Dharmic or Islamic civilization. They are the people forgotten by time. I think what little we know about the Malagasy can shed on light memories and legends preserved by peoples who we suspect were migrants into the only homelands they knew (e.g., how could the Aryans be exogenous if they didn’t record any memory of lands before India?).

🔊 Listen RSS

A month ago I posted the genetic results of a Malagasy individual of Merina identity. Today I post those for someone of Betsileo heritage. All the technical details are the same. You can find all the ADMIXTURE and PCA files here.

This genotyping was paid for by readers. I’ll update the post with the names of those who contributed below the fold later. If you contributed but don’t want to be named, email me at contactgnxp -at- gmail -dot- com, and I’ll leave you off the list.

This analysis was made possible by the generous donations of the following individuals:

Åse Kvist Innes-Ker
J B Massey
Matt Reusswig
Paul Conroy
Dr. Joseph Pickrell
G H Chinoy
Ian Ramjohn
Loganayagam Ramalingam
April Shepherd
Paul Gomes III
Robert Beckstead

• Category: Science • Tags: Malagasy, Personal Genomics 
🔊 Listen RSS

I’m going to address two points in this post. The next possible target for getting an undersampled population, and the Malagasy results.

First, lots of great submissions in regards to populations which are undersampled. Some of them are actually already in the data sets. For example, the Burusho and Kalash are in the HGDP. There has been a major dump of data from the Americans recently as well. Zack Ajmal at HAP has the most systematic description online about where to find these that I know of. Additionally, I’m looking for stuff which is interesting where N = 1 would make a difference. I think that was the case for the Tutsi sample, as well as the Malagasy. When you have no prior information, adding one data point is notable. Obviously I can’t afford the money, time, and energy, required to get a good representative sample from a given region. Though I hope researchers who have a gusher of grant money might look at the above thread for ideas.

I think the next population to look for is someone with Ainu ancestry. This is easier said that done, so I need to think about it (both because of dilution and the language barrier). But then again, the Tutsi and Malagasy requests had a much more positive and faster turnaround than I had expected. So I’m not going to get all down about the likelihood.

Second, we’ve got the issue of the Malagasy results. As with the Tutsi genotype there are lots of impassioned responses, some of which I don’t post. After reading through the Malagasy related comments I’m starting to wonder if people have a natural cognitive bias to over-complicate their ancestral narrative because it makes it more memorable and salient. I think I didn’t appreciate this because in the New World in admixed populations there is often a tendency to obscure mixed origins of groups due to social norms (e.g., black ancestry among Mexicans). But this may be atypical. People are very worried about issues of representativeness in regards to the Merina I analyzed, and that’s a legitimate issue. That being said, the previous run of someone who was 33% Sakalava seemed to give the same qualitative results (though this is complicated by that individual’s extreme admixture). I’m pretty sure that the Betsileo sample which I’ll get in ~1 month is going to be similar to the Merina one. If not, all the more exciting! We might not be able to answer questions with novel and unexpected results, but we’ll proliferate the questions.

The major aspect which I’m now confident of in regard to the Malagasy is that it seems unlikely that they have exotic African ancestry, whether it be Pygmy, Khoisan, or something else. An N = 2 is a lot to base this on, but there are plenty of other circumstantial elements which confirm the likely Bantu origin of the African component of Malagasy ancestry. One question though would be whether this element arrived in successive waves due to slave trading, or whether the Malagasy ethnogenesis occurred on the mainland, and they arrived a hybrid population.

In any case, please don’t leave impassioned comments about Madagascar’s ethnic politics anymore. I understand this is serious stuff for some people, but not for me. I don’t care about that much, though I know the broad outlines.

🔊 Listen RSS

Last week I begged for a Malagasy genotype. I didn’t quite get that, but I got the second best thing: a part Malagasy-genotype. I decided to take it for a spin.

But first some preliminaries. Here’s what we know about this individual (or what this individual knows):

- 25% French (paternal grandfather)
- 18.75% West African? 6.25% French? (paternal grandmother French Antilles)
- 19% Indian Muslim Bohras from Bombay + 6.25% Malagas, Sakalava tribe, royal family of Mahajanga (maternal grand -father)
- 25% Malagasy (Sakalava, maternal grandmother mtDNA haplogroup M23)

This is a very mixed individual in terms of ancestry. As for the Malagasy people, we know both a lot and a little about them. They’re a hybrid population, more or less, of Austronesians with a very close connection to the to the Dayaks of southern Borneo. I have hypothesized that these Austronesians were part of a circum-Indian ocean trading network which was marginalized by the rise of Islam in the second half of the first millennium. Such an early date would explain why the Malagasy seem to have been only lightly touched by Indic cultural influences, let alone Islamic ones. There is also the African component to their ancestry, which is more prominent in the lowland populations to the west of the island of Madagascar. The Sakalava are a somewhat more African group (as opposed to the Merina of the eastern highlands, who are more Austronesian).

Below are some results from ADMIXTURE and PCA generated with EIGENSOFT. Most of the PCA plots were not too useful, because I didn’t fine-tune the populations ahead of time too much (this is a first pass), so I didn’t post them. The ADMIXTURE runs are those which seem highly informative to me. There were three data sets into which I merged the part-Malagasy individual:

- #1, A Southeast Asian focused one, using mostly the Pan-Asian Consortium populations

- #2, An Asian focused data set which used the HGDP

- #3, An African focused data set which used the Henn et al. populations as well as some HGDP ones

#1 is plagued by a thin marker set. The Southeast Asian groups had ~56,000 markers, but the part-Malagasy individual only shared ~22,000 with them. Still, I made a go of it. I probably overcompensated in #2, as I used ~590,000 markers (the HGDP has a pretty good overlap with the 23andMe raw data). Finally, #3 had ~180,000 markers, which I feel to be very sufficient for this sort of exploratory endeavor.

Population K1 K2 K3 K4 K5 K6 K7 K8 K9 K10 K11 K12 K13
Taiwan Aborigine 0% 0% 0% 2% 0% 0% 5% 23% 69% 0% 0% 0% 0%
Hmong 0% 0% 3% 78% 0% 0% 7% 4% 3% 2% 0% 0% 2%
Jinuo 0% 0% 73% 2% 0% 0% 9% 1% 1% 3% 1% 0% 9%
Wa 0% 0% 5% 4% 0% 0% 17% 2% 1% 7% 0% 0% 63%
Malagasy 31% 0% 2% 0% 6% 26% 0% 0% 0% 2% 31% 0% 3%
Alorese 0% 83% 0% 0% 1% 2% 0% 3% 2% 0% 0% 8% 0%
Javanese 0% 4% 1% 1% 0% 0% 1% 25% 8% 22% 1% 5% 32%
Lamaholot 0% 61% 0% 1% 1% 3% 1% 13% 7% 2% 0% 8% 2%
Mentawai 0% 0% 0% 0% 0% 0% 0% 98% 0% 1% 0% 0% 0%
West Javanese 0% 4% 1% 1% 0% 2% 0% 26% 9% 22% 1% 4% 31%
Toraja 0% 13% 0% 1% 1% 1% 3% 44% 21% 7% 0% 2% 5%
Indian 1% 1% 0% 0% 2% 59% 0% 0% 0% 1% 25% 1% 11%
Japanese 1% 1% 2% 3% 0% 1% 87% 1% 1% 1% 0% 0% 1%
Kensiu Negrito 0% 0% 0% 0% 0% 0% 1% 1% 1% 2% 0% 90% 4%
Utah, white 1% 0% 0% 0% 2% 20% 0% 0% 0% 0% 74% 0% 3%
Luhya 78% 0% 1% 0% 2% 18% 0% 0% 0% 0% 1% 0% 0%
Mamanwa 0% 1% 0% 0% 83% 0% 0% 9% 3% 0% 0% 2% 0%
West Mindanao 0% 6% 1% 4% 3% 3% 11% 36% 24% 5% 2% 1% 4%
West Luzon 0% 4% 1% 5% 2% 2% 16% 36% 24% 4% 1% 1% 4%
Karen 9% 1% 6% 6% 0% 0% 18% 2% 1% 8% 0% 1% 50%
Plang 1% 1% 8% 5% 0% 0% 12% 6% 2% 18% 1% 2% 43%
HTin 1% 0% 1% 1% 0% 0% 0% 2% 0% 84% 1% 0% 10%
Han Taiwan 0% 0% 6% 16% 0% 0% 45% 10% 13% 3% 0% 0% 7%

Population K1 K2 K3 K4 K5 K6 K7 K8 K9 K10
Mbuti Pygmies 0% 100% 0% 0% 0% 0% 0% 0% 0% 0%
Biaka Pygmies 0% 9% 6% 0% 0% 0% 0% 59% 0% 25%
French 0% 0% 100% 0% 0% 0% 0% 0% 0% 0%
Papuan 1% 0% 0% 99% 0% 0% 0% 0% 0% 0%
Cambodians 75% 0% 0% 2% 10% 5% 8% 0% 1% 0%
Japanese 0% 0% 0% 0% 0% 0% 98% 0% 1% 0%
Han 28% 0% 0% 0% 0% 1% 70% 0% 1% 0%
Mandenka 0% 0% 0% 0% 0% 0% 0% 0% 0% 100%
Yakut 0% 0% 3% 0% 0% 0% 1% 0% 95% 0%
San 0% 0% 0% 0% 0% 0% 0% 100% 0% 0%
Bant S Africa 0% 4% 0% 0% 0% 0% 0% 23% 0% 73%
Tujia 34% 0% 0% 0% 0% 3% 63% 0% 0% 0%
Yizu 12% 1% 0% 1% 0% 9% 78% 0% 0% 0%
Miaozu 48% 0% 0% 0% 0% 1% 51% 0% 0% 0%
Hezhen 1% 0% 0% 0% 0% 1% 69% 0% 29% 0%
Xibo 4% 0% 2% 0% 1% 2% 73% 0% 18% 0%
Dai 88% 0% 0% 0% 0% 2% 10% 0% 0% 0%
Lahu 11% 0% 0% 0% 0% 82% 7% 0% 0% 0%
She 49% 0% 0% 0% 0% 0% 51% 0% 0% 0%
Naxi 4% 0% 0% 1% 0% 9% 85% 0% 1% 0%
Tu 8% 0% 4% 0% 3% 3% 74% 0% 7% 0%
Bantu Kenya 0% 5% 2% 0% 1% 0% 0% 7% 0% 84%
Malagasy 5% 1% 37% 0% 11% 3% 0% 4% 4% 36%
Indian 0% 0% 0% 0% 99% 0% 0% 0% 0% 0%

Population K1 K2 K3 K4 K5 K6 K7 K8 K9 K10
Hadza 16% 0% 76% 1% 5% 1% 1% 1% 0% 0%
Yemen Jews 0% 0% 0% 84% 0% 0% 0% 0% 0% 15%
Ethiopian 19% 0% 4% 52% 25% 0% 1% 0% 0% 0%
Sandawe 6% 0% 1% 2% 90% 1% 0% 1% 0% 0%
Biaka Pygmies 0% 0% 0% 0% 0% 100% 0% 0% 0% 0%
Mbuti Pygmies 0% 0% 0% 0% 0% 0% 100% 0% 0% 0%
French 0% 0% 0% 20% 0% 0% 0% 0% 1% 80%
Cambodians 0% 100% 0% 0% 0% 0% 0% 0% 0% 0%
Mandenka 98% 0% 0% 0% 0% 1% 0% 0% 0% 0%
Yoruba 96% 0% 0% 0% 0% 4% 0% 0% 0% 0%
Bant S Africa 72% 0% 0% 0% 0% 9% 1% 17% 0% 0%
Bantu Kenya 77% 0% 1% 2% 12% 5% 2% 0% 0% 0%
Malagasy 32% 13% 0% 3% 5% 4% 1% 0% 7% 36%
Luhya 77% 0% 2% 1% 12% 5% 3% 0% 0% 0%
Indian 0% 1% 0% 0% 0% 0% 0% 0% 99% 0%
San 8% 2% 0% 1% 0% 1% 0% 79% 3% 6%

I don’t really trust the proportions for the Pan-Asian focused data set. But I figured I should report them. No idea why the Malagasy shows so much Yakut. Could be an artifact from the hybridization? As for the rest, it seems that the African ancestry of this individual isn’t too atypical for an East African Bantu.

🔊 Listen RSS

I would like to throw out the word that I am looking for a person with Malagasy ancestry for the African Ancestry Project. To my knowledge there are no thick marker autosomal analyses of the Malagasy people. After my recent exploration of Southeast Asian genetics I think even one individual would be highly informative.

As usual I would guarantee that these data are entirely private, and I do not share it with anyone. But in this case I would like to make an exception and stipulate that Joseph K. Pickrell, a graduate student at the University of Chicago, would also be very interested in access to a Malagasy genotype for the purposes of research. Since this is an undersampled population the marginal returns to a Malagasy genotype would be enormous for science, a public good rather than just a private gain.

Also, I am still looking for a Tutsi genotype so that I can ascertain the origin of this population.

Please contact me at africanancestryproject -at- gmail -dot- com.


Je recherche une personne d’origine malgache dans le cadre du projet “l’African Ancestry Project”.

A ma connaissance, il n’existe à ce jour aucune analyse des marqueurs autosomiques du peuple malgache. Ma récente exploration de la génétique d’Asie du Sud me laisse penser que même l’étude d’une seule personne resterait très instructive.

Je garantis que ces données resteront confidentielles et ne seront pas divulguées à des tiers, à l’exception de Joseph K. Pickrell, un étudiant en thèse à l’Université de Chicago, également très intéressé par l’accès à un génotype malgache pour des fins de recherche.

Cette population étant sous-échantillonnées, les rendements marginaux d’un génotype malgache seraient énormes pour le science et d’intérêt général.

Par ailleurs, je suis toujours à la recherche d’un génotype Tutsi afin de déterminer l’origine de cette population.

Merci de bien vouloir me contacter à africanancestryproject-at-gmail-point-com.

🔊 Listen RSS

In my post on Empires of the Word I observed that quite often the written record is silent on many matters which only language or genes tell us must have occurred. The Indo-Aryan character of the dominant language on the island of Sri Lanka seems to be a geographical anomaly in the least, but perhaps most strange of all is the existence of a language and ethnic group of clear Southeast Asian provenance on the island of Madagascar. To my knowledge Arab, Persian and South Asian sources do not record the existence of a prominent Southeast Asian maritime diaspora which spanned the Indian ocean in the years before 1000 A.D., but we know that it did exist. A new paper on the genetics of the island of Comoros fleshes out another piece of the puzzle, Genetic diversity on the Comoros Islands shows early seafaring as major determinant of human biocultural evolution in the Western Indian Ocean:

The Comoros Islands are situated off the coast of East Africa, at the northern entrance of the channel of Mozambique. Contemporary Comoros society displays linguistic, cultural and religious features that are indicators of interactions between African, Middle Eastern and Southeast Asian (SEA) populations. Influences came from the north, brought by the Arab and Persian traders whose maritime routes extended to Madagascar by 700–900 AD. Influences also came from the Far East, with the long-distance colonisation by Austronesian seafarers that reached Madagascar 1500 years ago. Indeed, strong genetic evidence for a SEA, but not a Middle Eastern, contribution has been found on Madagascar, but no genetic trace of either migration has been shown to exist in mainland Africa. Studying genetic diversity on the Comoros Islands could therefore provide new insights into human movement in the Indian Ocean. Here, we describe Y chromosomal and mitochondrial genetic variation in 577 Comorian islanders. We have defined 28 Y chromosomal and 9 mitochondrial lineages. We show the Comoros population to be a genetic mosaic, the result of tripartite gene flow from Africa, the Middle East and Southeast Asia. A distinctive profile of African haplogroups, shared with Madagascar, may be characteristic of coastal sub-Saharan East Africa. Finally, the absence of any maternal contribution from Western Eurasia strongly implicates male-dominated trade and religion as the drivers of gene flow from the North. The Comoros provides a first view of the genetic makeup of coastal East Africa.

In the paper they note that ~6% of the Y chromosomal lineages were Southeast Asian, while ~15% of mtDNA lineages were. That indicates that the Southeast Asian presence on the Indian ocean was a case of folk migration, men, women and children on the move. The data from Madagascar indicate something similar, both male and female lineages show Southeast Asian imprint among the highland Malagasy (I don’t make much of the proportional difference because this is just one sample). In contrast, they show in this paper that there’s a substantial West Eurasian (probably Arab, Indian and Persian) Y chromosomal gene flow into the population of Comoros, but no West Eurasian mtDNA. So in this case you have a clear contrast with that of the Southeast Asian seafarers, the Muslim merchants who settled on the Comoros did not bring their children or womenfolk. It was not a folk migration, but a mercantile network. Because of the nature of the sources, and the cultural influence of the West Asians, we know of their presence from the historical record. In contrast, the arguably more substantial folk migration of Southeast Asian seafarers from Borneo is hidden in the text. They may have been of no concern or beneath mention from the perspective of the Muslim merchant princes, but the fact that they were no longer on the high seas by the time the Portuguese arrive may also indicate that they were driven off by the same Muslim merchant princes in the years after 1000. If the latter is the case the silence may be due to the inclination to forget an unpleasant rivalry.

All this goes to show that history’s reliance on text can mislead and obscure real dynamics. Even social and economic history which attempts to tunnel-down to the level of the populace is still heavily reliant on written records. In the case of seafarers it seems likely that even archaeologists would be unable to detect their movements because of the liminal nature of their settlements. The linguistic and cultural influences in Madagascar and in East Africa indicate a sojourn by Austronesians in that coast, but there is no physical or textual record. There is the “dark history” which we ignore because of current ideological preferences, and then there is the dark history which has fallen outside of our methodological window.

Dienekes has more on this paper.

🔊 Listen RSS

ResearchBlogging.orgYesterday I pointed to a new paper, Plasmodium vivax clinical malaria is commonly observed in Duffy-negative Malagasy people. P. vivax is the least virulent of the malaria inducing pathogens, and it is presumably responsible for the fact that the Duffy antigen locus is one of the more ancestrally informative ones in the human genome. In most of Eurasia the the Duffy negative null allele* is present at very low frequencies, less than 5%, and often simply absent. In contrast, in Sub-Saharan Africa the Duffy negative variant reaches frequencies as high as 95% in West Africa, and and 90% in many other regions. In North Africa and the Middle East the frequencies are intermediate, likely due to the necessity for local adaptation to malaria in many regions, and the historical introduction of the Duffy negative allele via the slave trade.

_47495404__47060392_rajoelina_afp-1Before genomics, looking at the Duffy locus was one simple way that geneticists ascertained the proportion of white admixture in the African American population. The Duffy negative allele was nearly absent in Europeans, and present in frequencies of ~95% in West Africa. Therefore, the ~70% frequency in African Americans indicates what we know from other sources, a substantial minority European contribution to their ancestry. The people of Madagascar are similar insofar as they are a byproduct of admixture between African and non-African populations. The source of the non-African ancestry is rather easy to determine, unlike most African countries Madagascar has one language, Malagasy, and it is of the Barito family of languages. Aside from Malagasy the Barito languages are spoke only in a small region of southern Borneo in Indonesia. There are other aspects of the Malagasy culture which make their Southeast Asian provenance clear. The photo above is of Andry Rajoelina, the current President of Madagascar. Two aspects of his visage are salient, his youth (he used to be a disk jockey!), and the fact that his features do not seem typical Sub-Saharan African. Many of the leaders of Madagascar, including the former royal family, are from the highlands where Asiatic features and folkways are more prevalent.

But there is also a clear African component to the Malagasy, more obvious among coastal populations, but also possibly dominant in a genetic sense in terms of proportion to the Asian according to research using uniparental markers. An analysis of Y lineage Fst genetic distances suggests that the Malagasy are, on the whole, somewhat closer to East Africans than to people from Borneo. I stipulate on the whole because as implied above there seems to be regional variation, which Southeast Asian ancestry and culture least hybridized with a Sub-Saharan African in the central highlands, likely for ecological reasons.

malagas1If the Duffy negative allele was viewed purely as a neutral locus, and so ancestrally informative, one would assume that the Malagasy were mostly African. In the figure to the left the red tinted portions represent Duffy negative proportions, the green Duffy positive, and the darker shade P. vivax positivity. The green star indicates a site where P. vivax positivity was only found among the Duffy positive, while at the sites with red stars it was found among both antigen state groups. As you can see at none of the sites was the Duffy positive allele modal, and at Andapa the frequency of Duffy negative was typical of much of Sub-Saharan Africa. In the total data set 72% of the individuals were Duffy negative. Going by the previous cited work this would underestimate Asian ancestry, which seems likely to be near parity, if not quite.

Two points come to mind:

1) It seems clear that the Duffy locus is not neutral. It is subject to natural selection, as even though the malaria caused by P. vivax is relatively mild, it t does reduce fitness. Natural selection should result in an increase in frequency of the negative allele in regions where malaria caused by P. vivax is endemic. In the American South malaria was not as extreme of a problem, nor does Duffy negative status have a strong side effect (e.g., sickle cell), so it was a neutral locus and appropriate to inform ancestry.

2) Modern African populations may not be an accurate representation of the allele frequencies of Duffy in the ancestral groups which contributed to the ancestry of the Malagasy. More plainly, the Africans who intermarried with the Barito speakers may have had much higher frequencies of Duffy positive alleles because natural selection had not proceeded so that the null allele was driven to near fixation.

To assess the plausibility of #2, one needs to know how the Malagasy, or more accurately, the speakers of the Barito language which became Malagasy, got where they are. Unfortunately, no one really knows, and the hypotheses are controversial because of their speculative nature. It seems likely that the Southeast Asian mariners initially arrived in the western Indian ocean region ~2,000 years ago, but widespread settlement of Madagascar’s interior may not have been occurring until ~1,000 years ago. By the 13th century there was a large Muslim city in the north of Madagascar integrated into the Indian ocean trade network, so Madagascar is on the fringes of written history at that point. The anthropological evidence seems to point to a sojourn on the coast of East Africa by Southeast Asians, as there are aspects of Malagasy culture which seem related to Bantu groups in that area. Additionally, there some genetic data which point to an African contribution on the mtDNA from populations further north on the coast, toward Kenya, and Y DNA which suggests a connection with the adjacent region of the continent in Mozambique. A model of how this could occur is that the initial colonists in East Africa picked up local wives along the northern coast, and eventually resettled in Madagascar. After this settlement there were periodic migration of Africans from nearby regions, either voluntary or forced through slavery, which added the later diversity. The fact that this component is male-biased would point to slavery of the sort practiced in the New World, whereby Africans were forced to work in agriculture and male robustness was prized (this is in contrast with much of the Middle East, where female African domestic servants were the primary driver of slavery).

mapmapOne of the mysterious aspects of the arrival of the Malagasy is that there aren’t records by the literate polities which fringed the Indian ocean of their movements. But why should there be? Open ocean traders were generally marginal to these states, who simply extracted rents from the activities of the merchants and migrants. It seems entirely plausible that many populations have been on the move throughout history, their impact in particular regions slowly being ablated by time. There is one aspect of Africa which makes it entirely plausible that the Barito presence would disappear or be marginal: the local populations seem biologically very well adapted to the pathogens on the continent. It is notable for example that the Arab and Persian cultural influence in East Africa never spread inland beyond the Indian ocean littoral. And yet these groups were present on the East African coast from the time of the Romans on. It seems likely to me that Africa is relatively resistant to “back-migration” from Eurasia on ecological grounds. North Africa is part of the Palearctic ecozone, while the highlands of Ethiopia are also ecologically distinct. Both these regions are strongly shaped genetically by populations with Eurasian connections, in the former case predominantly so, but both they are exceptions which prove the rule.

The maps to the left show topography and population density respectively. In Madagascar in the highlands Southeast Asians could transfer wet rice agriculture, and also escape the most baleful influences of African diseases (which would naturally be introduced with African populations). It is also where there is the greatest population density. In contrast the coastal regions are more lightly populated and have more African influence. Like South Africa or the Kenyan highlands I believe that Madagascar was one region of Sub-Saharan Africa which was open to the settlement of outsiders who lacked biological defenses because of its ecology. Granted, it seems to have been unsettled before the Malagasy arrived, but if its pathogen environment was equivalent to that of the mainland I suspect that African genes and culture would have replaced the Malagasy component rather rapidly. The Malagasy are just one of many populations which made some sort of great trek. Most of them disappear, get absorbed or become extinct. But in a few rare cases, such as in that of Iberians in the 16th century, or Polynesians 2,000 years ago, and the Malagasy, these travelers encountered territory which they were able to settle easily. And so we have concrete evidence of their past existence, their present existence. You couldn’t plausibly invent the cultural makeup of Madagascar, because our model of history and human population movement is simplified, and all the outliers and rough edges have been hidden or consciously removed.

Though the highlands of Madagascar allowed the Southeast Asian settlers a refuge for endogenous population growth, which allowed them to perpetuate their culture and leave a stamp on the island, Madagascar is African, and much of the island is clearly suited for malaria. The evolutionary dynamics may be contingent on the peculiarities of the island’s demographic history, but they will still proceed nonetheless. It is noted in these results that though varieties of P. vivax seem to have moved from the Duffy positive to the Duffy negative segment of the population, it is still much more virulent in those who are Duffy positive. There were 15 times as many full blown cases of P. vivax induced malaria (as opposed to positive infection status) among those who were Duffy positive than among those who were negative. Nevertheless, the emergence of strains able to infect Duffy negative blood cells opens up the possibility for more virulent strains in the future which could result in many more cases of full blown malaria within this population.

Let me jump to the conclusion:

Our observations in Madagascar showing conclusive evidence that P. vivax is capable of causing blood-stage infection and disease in Duffy-negative people illustrate that in some conditions P. vivax exhibits a capacity for infecting human erythrocytes without the Duffy antigen. The data assembled in this study suggest that conditions needed to clear the barrier of Duffy negativity may include an optimal human admixture. In Madagascar with significant numbers of Duffy-positive people and full susceptibility of hepatocytes in Duffy negatives, P. vivax may have sufficient exposure to Duffynegative erythrocytes, allowing more opportunities for de novo selection or optimization of an otherwise cryptic invasion pathway that nevertheless seems less efficient than the Duffy-dependent pathway.

There are several issues that I’ve glossed over in this paper, and one of them is that there are other populations which have a mix of negative and positive individuals. Implicitly the American South is one. But malaria is not endemic in most of the South. But in Brazil there is a similar racial mixture, and its climate is conducive to tropical diseases. It seems there are issues with detecting the P. vivax pathogen within blood cells, and so earlier studies as to the possibility of the infection of those who were Duffy negative were often muddled or inconclusive. In this study they established the existence of this group rather clearly, but is it due to the peculiarities of Madagascar’s population mixture and history? True, Brazil also has an admixed population whose Duffy allele frequencies are interchangeable with that of Madagascar, but Brazil has been settled for only the past ~300 years or so, with much of the population being of more recent origin (Brazil had the highest slave attrition rate on the American mainland, which explains the African nature of Afro-Brazilian culture. Many of the slaves were from Africa, or first generation, at emancipation). A lower bound for Madagascar is ~1,000 years, and the coexistence of Barito and African populations is likely closer to ~2,000 years. So the existence of P. vivax lines which can penetrate the negative allele population may be a function of the longer time given to the emergence of adaptive strategies.

I suspect the fact that there is a component of what ecologists term “patchiness” in the settlement patterns of various populations and ecology in Madagascar might have aided in the persistence of the Duffy positive allele. It seems that in much of the rest of Africa once agriculture became common and the conditions for the mosquito which carries P. vivax improved the Duffy negative allele swept to fixation. At this point the P. vivax infection rates were so low that natural selection became less of an issue (the extant variation was reduced, and only a small proportion of the population would have been subject to selection). It is on marginal areas where fixation did not occur that you’d have the diversity which might allow for the emergence of different P. vivax lineages. Another place to look besides Madagascar would be the margins of Ethiopia, as well as South Africa, where Bantu farmers came up against a very different ecologies and populations which they could not assimilate, or did so only partly.

* Duffy is really the the antigen itself, so “Duffy negative” means lacking the antigen. But I’m going to use the shorthand Duffy negative to point to the alleles which confer this state, which have names such as FY*A and FY*B. The gene itself is DARC.

Citation: Ménard D, Barnadas C, Bouchier C, Henry-Halldin C, Gray LR, Ratsimbasoa A, Thonier V, Carod JF, Domarle O, Colin Y, Bertrand O, Picot J, King CL, Grimberg BT, Mercereau-Puijalon O, & Zimmerman PA (2010). Plasmodium vivax clinical malaria is commonly observed in Duffy-negative Malagasy people. Proceedings of the National Academy of Sciences of the United States of America, 107 (13), 5967-71 PMID: 20231434

Image credit: BBC, Wikipedia

Razib Khan
About Razib Khan

"I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. If you want to know more, see the links at"