Botocudo Ancient Remains from Brazil

One thing you can always count on in the infant science of population genetics…  whatever you think you know, for sure, for a fact…well….you don’t.  So don’t say too much, too strongly or you’ll wind up having to decide if you’d like catsup with your crow!  Well, not literally, of course.  It’s an exciting adventure that we’re on together and it just keeps getting better and better.  And the times…they are a changin’.

We have some very interesting news to report.  Fortunately, or unfortunately – the news weaves a new, but extremely interesting, mystery.

Ancient Mitochondrial DNA

Back in 2013, a paper, Identification of Polynesian mtdNA haplogroups in remains of Botocudo Amerindians from Brazil, was published that identified both Native American and Polynesian haplogroups in a group of 14 skeletal remains of Botocudo Indians from Brazil whose remains arrived at a Museum in August of 1890 and who, the scientists felt, died in the second half of the 19th century.

Twelve of their mitochondrial haplogroups were the traditional Native haplogroup of C1.

However, two of the skulls carried Polynesian haplogroups, downstream of haplogroup B, specifically B4a1a1a and B4a1a1, that compare to contemporary individuals from Polynesian, Solomon Island and Fijian populations.  These haplotypes had not been found in Native people or previous remains.

Those haplogroups include what is known as the Polynesian motif and are found in Indonesian populations and also in Madagascar, according to the paper, but the time to the most common recent ancestor for that motif was calculated at 9,300 years plus or minus 2000 years.  This suggests that the motif arose after the Asian people who would become the Native Americans had already entered North and South America through Beringia, assuming there were no later migration waves.

The paper discusses several possible scenarios as to how a Polynesian haplotype found its way to central Brazil among a now extinct Native people. Of course, the two options are either pre-Columbian (pre-1500) contact or post-Columbian contact which would infer from the 1500s to current and suggests that the founders who carried the Polynesian motif were perhaps either slaves or sailors.

In the first half of the 1800s, the Botocudo Indians had been pacified and worked side by side with African slaves on plantations.

Beyond that, without full genome sequencing there was no more that could be determined from the remains at that time.  We know they carried a Polynesian motif, were found among Native American remains and at some point in history, intermingled with the Native people because of where they were found.  Initial contact could have been 9,000 years ago or 200.  There was no way to tell.  They did have some exact HVR1 and HVR2 matches, so they could have been “current,” but I’ve also seen HVR1 and HVR2 matches that reach back to a common ancestor thousands of years ago…so an HVR1/HVR2 match is nothing you can take to the bank, certainly not in this case.

Full Genome Sequencing and Y DNA

This week, one on my subscribers, Kalani, mentioned that Felix Immanuel had uploaded another two kits to GedMatch of ancient remains.  Those two kits are indeed two of the Botocudo remains – the two with the Polynesian mitochondrial motif which have now been fully sequenced.  A corresponding paper has been published as well, “Two ancient genomes reveal Polynesian ancestry among the indigenous Botocudos of Brazil” by Malaspinas et al with supplemental information here.

There are two revelations which are absolutely fascinating in this paper and citizen scientist’s subsequent work.

First, their Y haplogroups are C-P3092 and C-Z31878, both equivalent to C-B477 which identifies former haplogroup C1b2.  The Y haplogroups aren’t identified in the paper, but Felix identified them in the raw data files that are available (for those of you who are gluttons for punishment) at the google drive links in Felix’s article Two Ancient DNA from indigenous Botocudos of Brazil.

I’ve never seen haplogroup C1b2 as Native American, but I wanted to be sure I hadn’t missed a bus, so I contacted Ray Banks who is one of the administrators for the main haplogroup C project at Family Tree DNA and also is the coordinator for the haplogroup C portion of the ISOGG tree.

ISOGG y tree

You can see the position of C1b2, C-B477 in yellow on the ISOGG (2015) tree relative to the position of C-P39 in blue, the Native American SNP shown several branches below, both as branches of haplogroup C.

Ray maintains a much more descriptive tree of haplogroup C1 at this link and of C2 at this link.

Ray Banks C1 tree

The branch above is the Polynesian (B477) branch and below, the Native American (P39) branch of haplogroup C.

Ray Banks C2 treeIn addition to confirming the haplogroup that Felix identified, when Ray downloaded the BAM files and analyzed the contents, he found that both samples were also positive for M38 and M208, which moves them downstream two branches from C1b2 (B477).

Furthermore, one of the samples had a mutation at Z32295 which Ray has included as a new branch of the C tree, shown below.

Ray Banks Z32295

Ray indicated that the second sample had a “no read” at Z32295, so we don’t know if he carried this mutation.  Ray mentions that both men are negative for many of the B459 equivalents, which would move them down one more branch.  He also mentioned that about half of the Y DNA sites are missing, meaning they had no calls in the sequence read.  This is common in ancient DNA results.  It would be very interesting to have a Big Y or equivalent test on contemporary individuals with this haplogroup from the Pacific Island region.

Ray notes that all Pacific Islanders may be downstream of Z33295.

Not Admixed

The second interesting aspect of the genomic sequencing is that the remains did not show any evidence of admixture with European, Native American nor African individuals.  More than 97% of their genome fits exactly with the Polynesian motifs.  In other words, they appear to be first generation Polynesians.  They carry Polynesian mitochondrial, Y and autosomal (nuclear) DNA, exclusively.

Botocudo not admixed

In total, 25 Botocudo remains have been analyzed and of those, two have Polynesian ancestry and those two, BOT15 and BOT17, have exclusively Polynesian ancestry as indicated in the graphic above from the paper.

When did they live?  Accelerator mass spectrometry radiocarbon dating with marine correction gives us dates of 1479-1708 AD and 1730-1804 for specimen BOT15 and 1496-1842 for BOT17.

The paper goes on to discuss four possible scenarios for how this situation occurred and the pros and cons of each.

The Polynesian Peru Slave Trade

This occurred between 1862-1864 and can be ruled out because the dates for the skulls predate this trade period, significantly.

The Madagascar-Brazil Slave Trade

The researchers state that Madagascar is known to have been peopled by Southeast Asians and not by Polynesians.  Another factor excluding this option is that it’s known that the Malagasy ancestors admixed with African populations prior to the slave trade.  No such ancestry was detected in the samples, so these individuals were not brought as a result of the Madagascar-Brazil slave trade – contrary to what has been erroneously inferred and concluded.

Voyaging on European Ships as Crew, Passengers or StowAways

Trade on Euroamerican ships in the Pacific only began after 1760 AD and by 1760, Bot15 and Bot17 were already deceased with a probability of .92 and .81, respectively, making this scenario unlikely, but not entirely impossible.

Polynesian Voyaging

Polynesian ancestors originated from East Asia and migrated eastwards, interacting with New Guineans before colonizing the Pacific.  These people did colonize the Pacific, as unlikely as it seems, traveling thousands of miles, reaching New Zealand, Hawaii and Easter Island between 1200 and 1300 AD.  Clearly they did not reach Brazil in this timeframe, at least not as related to these skeletal remains, but that does not preclude a later voyage.

Of the four options, the first two appear to be firmly eliminated which leaves only the second two options.

One of the puzzling aspects of this analysis it the “pure” Polynesian genome, eliminating admixture which precludes earlier arrival.

The second puzzling aspect is how the individuals, and there were at least two, came to find themselves in Minas Gerais, Brazil, and why we have not found this type of DNA on the more likely western coastal areas of South America.

Minas Gerais Brazil

Regardless of how they arrived, they did, and now we know at least a little more of their story.


At GedMatch, it’s interesting to view the results of the one-to-one matching.

Both kits have several matches.  At 5cM and 500 SNPs, kit F999963 has 86 matches.  Of those, the mitochondrial haplogroup distribution is overwhelmingly haplogroup B, specifically B4a1a1 with a couple of interesting haplogroup Ms.

F999963 mito

Y haplogroups are primarily C2, C3 and O.   C3 and O are found exclusively in Asia – meaning they are not Native.

F999963 Y

Kit F999963 matches a couple of people at over 30cM with a generation match estimate just under 5 generations.  Clearly, this isn’t possible given that this person had died by about 1760, according to the paper, which is 255 years or about 8.5-10 generations ago, but it says something about the staying power of DNA segments and probably about endogamy and a very limited gene pool as well.  All matches over 15cM are shown below.

F999963 largest

Kit F999964 matches 97 people, many who are different people that kit F999963 matched.  So these ancient Polynesian people,  F999963 and F999964 don’t appear to be immediate relatives.

F999964 mito

Again, a lot of haplogroup B mitochondrial DNA, but less haplogroup C Y DNA and no haplogroup O individuals.

F999964 Y

Kit F999964 doesn’t match anyone quite as closely as kit F999963 did in terms of total cM, but the largest segment is 12cM, so the generational estimate is still at 4.6,  All matches over 15cM are shown below.

F999964 largest

Who are these individuals that these ancient kits are matching?  Many of these individuals know each other because they are of Hawaiian or Polynesian heritage and have already been working together.  Several of the Hawaiian folks are upwards of 80%, one at 94% and one believed to be 100% Hawaiian.  Some of these matches are to Maori, a Polynesian people from New Zealand, with one believed to be 100% Maori in addition to several admixed Maori.  So obviously, these ancient remains are matching contemporary people with Polynesian ancestry.

The Unasked Question

Sooner or later, we as a community are going to have to face the question of exactly what is Native or aboriginal.  In this case, because we do have the definitive autosomal full genome testing that eliminates admixture, these two individuals are clearly NOT Native.  Without full genomic testing, we would have never known.

But what if they had arrived 200 years earlier, around 1500 AD, one way or another, possibly on an early European ship, and had intermixed with the Native people for 10 generations?  What if they carried a Polynesian mitochondrial (or Y) DNA motif, but they were nearly entirely Native, or so much Native that the Polynesian could no longer be found autosomally?  Are they Native?  Is their mitochondrial or Y DNA now also considered to be Native?  Or is it still Polynesian?  Is it Polynesian if it’s found in the Cook Islands or on Hawaii and Native if found in South America?  How would we differentiate?

What if they arrived, not in 1500 AD, but about the year 500 AD, or 1000 BCE or 2000 BCE or 3000 BCE – after the Native people from Asia arrived but unquestionably before European contact?  Does that make a difference in how we classify their DNA?

We don’t have to answer this yet today, but something tells me that we will, sooner or later…and we might want to start pondering the question.


I want to thank all of the people involved whose individual work makes this type of comparative analysis possible.  After all, the power of genetic genealogy, contemporary or ancient, is in collaboration.  Without sharing, we have nothing. We learn nothing.  We make no progress.

In addition to the various scientists and papers already noted, special thanks to Felix Immanual for preparing and uploading the ancient files.  This is no small task and the files often take a month of prep each.  Thanks to Kalani for bringing this to my attention.  Thanks to Ray Banks for his untiring work with haplogroup C and for maintaining his haplogroup webpage with specifics about where the various subgroups are found.  Thanks to ISOGG’s volunteers for the haplotree.  Thanks to GedMatch for providing this wonderful platform and tools.  Thanks to everyone who uploads their DNA, and that of their relatives and works on specific types of projects – like Hawaiian and Maori.  Thanks to my haplogroup C-P39 co-administrators, Dr. David Pike and Marie Rundquist, for their contributions to this discussion and for working together on the Native American Haplogroup C-P39 Project.  It’s important to have other people who are passionate about the same subjects to bounce things off of and to work with.  This is the perfect example of the power of collaboration!

28 thoughts on “Botocudo Ancient Remains from Brazil

  1. Thank you for having Ray looked into it. I did check the ISOGG tree but I forgot about his other tree. I’m O-P164 (for now until I get more SNPs tested) and was looking to see where my Asian line diverges from (western) Polynesians’ O lineage.

    But now that you pointed out what Ray was able to see from those BAM files, I can direct the project members of the Polynesian DNA project to get those specific SNPs tested to make it more precise.

    Also, it’s no surprise that we get total shared 100+cM to other Polynesians who have been in their own homeland (islands) 4,000 miles away for 8 centuries yet we still get 2nd to 3rd cousin predictions. 15cM is about the average for largest segment even among our own group, i.e. Hawaiians compared to Hawaiians, Maoris compared to Maoris. Which is why it is even more difficult using it for genealogy. All we can do is concentrate solely on the largest segment.

    Thanks so much for all of this!

  2. Roberta,

    By coincidence I was just comparing these results to my families results and that of a Native family that I am connected with. We have many of the same matches that pop up on the one to many for the ancient Brazilian kits. We are all sharing small X matches. I just got off the phone with Carlyle Hinshaw discussing this right before checking my email and seeing your blog. Carlyle has been working on documenting all of our haplogroups and we discussed the haplogruops of the Brazilian kits. I would be happy to show you what I have so far. Thanks.

  3. The answer to how the DNA of of these two individuals ended up in Minas Gerais, Brazil, may well be found in part 4 of “De val van ‘s werelds afgod” written by Jacob Roggeveen (1 February 1659 – 31 January 1729), a Dutch explorer who was sent from the Netherlands to find “Terra Australis,” but he instead came across an island on Easter Sunday which he named Easter Island in 1722. Jacob Roggeveen also encountered other islands before returning to The Netherlands where he wrote the above account of his voyage, probably between 1725 and 1729, when he died. His return trip may have been quite similar to his initial voyage, which took him from Europe first to the Falkland Islands, and then south to the tip of South America where he crossed into the Pacific Ocean, and docked near Valdivia, Chile, then sailed to Juan Fernández Islands, Easter Island (from Easter Sunday, 5 April 1722, until about 12 April, 1722), then the Tuamotu Archipelago, the Society Islands, and Samoa and Batavia. On his return voyage, Roggeveen may well have taken captives to Minas Gerais, which was already known for its gold rush – and need for slaves — during the 18th Century. Thus, in this context the timing makes sense, 1722 – 1724. My bet is on the DNA coming from Easter Island which was in the midst of a social and ecological meltdown at this time. If anyone reading this is fluent in Dutch, take a good look at Roggeveen’s “De val van ‘s werelds afgod” and let us know if it tells much about visiting Portuguese colonies in South America! [Some of this information comes from wikipedia; the intuitive hypotheticals are my own based on my advanced degree in Latin American History].

    • One thing is for sure. If captives were taken aboard a Dutch ship, it would not have been from Samoa or else many of the Samoans on would’ve come up as a match to both of these skulls. Instead, only eastern Polynesians are. Rapa Nui makes sense but I haven’t been able to get a Rapa person to get an atDNA test done just yet.

      • If the following wikipedia quote about a population die-off is correct, it may be next to impossible to obtain a match from a living Rapanui who is descended from the possible captives — unless relatives of the captives survived among those last 36 descendants:
        “Estimates of the pre-European population range from 7–17,000. Easter Island’s all-time low of 111 inhabitants was reported in 1877. Out of these 111 Rapanui, only 36 had descendants, but all of today’s Rapanui claim descent from those 36.”
        Maybe someone could start a crowdfunding project for National Geographic (or FTDNA) to carry out intense testing of the Rapanui!

    • Since you mentioned finding possible relatives of the captives, assuming it was Rapa Nui, we do have a problem with matching each other. At FTDNA generally my Hawaiian mother can share in the 100cM to almost 300cM total shared with Samoans and Tongans. Both are western Polynesians, are older than eastern Polynesians and are genetically much more diverse than eastern Polynesians. My mother can share between 300cM – 690cM with Maoris and other Hawaiians, depending on how admixed they are.

      We come from repeated bottlenecking. Hawaii went through their last known bottlenecking in 1900 (37,635 Hawaiians). In the 2000 US census they reported around 460,000 Native Hawaiians.

      I’ve seen this with Pitcairn island as well starting off with a population of 27 in 1790 (there were only 6 Tahitian women who with 5 men produced offspring), their highest in 1856 with 193 people, reduced to 16 people in 1859 (in 1856 they were all relocated to Norfolk island but 16 returned to Pitcairn), made it to 250 people in 1936 and today around 54 or so. My mother is an autosomal match to someone on Norfolk island.

      So if we all can match to each other with high totals, or even with low total shared with distantly related Polynesians or ancient remains as proven, who knows what that would mean if we were to test ANY Rapa Nui person against these samples.

      The problem is that it is difficult to figure out if we are a close relationship or not. Of course a real close relationship would be indicated by the largest segment size. We can average around 12cM for the largest segment.

  4. I think there may have been many opportunities for Polynesians to embark on European ships prior to 1760. The Wiki article on the Manila Galleons describe the lucrative trade, including in gold, that Spain carried out for hundreds of years, in the process discovering many Pacific islands:

    “Andrés de Urdaneta was the first to sail the Pacific from west to east, establishing the maritime route from Asia to America in 1565, that lasted until 1815. Spanish expeditions discovered Guam, the Marianas, the Carolines and the Philippines in the North Pacific, as well as Tuvalu, the Marquesas, the Solomon Islands and New Guinea in the South Pacific. Spanish navigators also discovered the Pitcairn and Vanuatu archipelagos during their search for Terra Australis in the 17th century.”

    The article goes on to further discuss the possibility that Spain first “discovered” Hawaii but kept it a secret to reduce foreign designs on their new empire. (unsuccessfully in the end)

    As with other research studies, we can’t be sure from where individual remains originated. People have always migrated to distant shores.

    • It was during Gaetano’s expedition from Acapulco to Manila that they say he charted the Hawaiian islands, never landed there but his coordinates were not accurate which allowed the islands to still be isolated until Capt. Cook had the help of a few Tahitian men who knew of the Hawaiian islands based on oral tradition about the land of fire (volcano).

      We do have stories of people going abroad with these Europeans, like chief Kaiana who went to China in 1787 with Capt. Meares, but he returned to Hawaii. Or Henry Obookiah (Opukaha’ia) who went to the east coast of the USA and sparked interest for the missionaries to Hawaii. But these are limited to Polynesians who could have stowed away on a European ship, which if they did would make sense that they lived among the Europeans.

      But what we have here is something that could predate those dates and most importantly somehow they ended up with the other various indian groups of Brazil. The evidence of the kumara [sweet potato] which we call ‘uala is only proof of technological exchange, which we were aware of, but now DNA is showing that it was not limited to just these things but some people actually never came back, just as we’ve always known from our stories of Hema, Kaha’i, Wahieloa, and Laka are some examples. Other Polynesians may have different names for these people, like Tafaki/Tawhaki/Tafa’i/Taha’i, or Vahieroa/Vahieloa and Rata/Lata. The names varied for each island nation but the stories were always the same, that they left and never came back so in the next generation their son left to look for their father and this happened for a few generations.

  5. Pingback: Native American Mitochondrial Haplogroups | DNAeXplained – Genetic Genealogy

  6. This post and the comments which follow discuss possible pre-Columbian voyages between Asia and the Americas. They discuss travel between the Americas and what may have been the great unknown to those who made the trip (Polynesian Islands, i.e. the sweet potato). They discuss legends of those who left their homeland never to return and how future generations went looking for them. All these things are mentioned in the Book of Mormon but, whatever you do, DO NOT read this book. It also contains that dreaded taboo subject — religion. Oops! I said it, didn’t I?

  7. When I first came to Brazil 10 years ago, one of the first things I noticed were the “asian” features of a lot of Brazilian Indians. To me, it’s pretty obvious that they are the descendants of Polynesians. What’s easier if you are a master sailor: keep travelling by boat, or walk over land for thousands and thousands of kilometers?

    I read something a while back about a chicken bone that was found in Brazil matched the same chickens in Polynesia during that time, and the birds aren’t native to Brazil, either.

    • The chicken bone was on the Pacific side of S. America, not Brazil. As for the Asian features, a lot of NA particularly in the Pacific Northwest as well as Alaskan Natives will have this feature, which is understandable. I remember years ago my friend who is Nicaraguan but whose father was from China asked me if I could guess what one of the messenger guys we’d see was. I guessed if not Vietnamese, maybe Filipino. My friend laughed and said he was Mexican, from Mexico and apparently this guy is just indigenous.

      Roberta provided a link to the supplemental information which mentions the testing done and the results of the other Botocudo skulls and how they did not have anything of Polynesian origin. These two samples were the only two unusual samples or skulls that were Polynesian origin. If they had gone to S. America, we’d definitely see it in DNA which is mentioned in the supplemental info and the other link that was provided on the article. Actually the original article questioned that as a possibility but specified how it seemed it was highly improbable.

  8. Pingback: Some Native Americans Had Oceanic Ancestors | DNAeXplained – Genetic Genealogy

  9. I matched both Bot15 and Bot17. The highest was a little over 3 cM. But I’m only half Asian so perhaps mother’s DNA would match much higher. We are haplogroup B5a.

    • Unfortunately lowering the threshold to 1cM/3cM forcing a match does not make it a match. The only true matches to the Botocudo skulls (at the default) are actual Polynesians. Anyone can lower the threshold to 1cM of any ethnic group and may or may not be a match to any of these ancient DNA kits uploaded on

      The mtDNA haplogroups can go back much further. The Polynesian motif (B4a1a1) is estimated to be 17,000 years old. B5a is much, much older than the Polynesian motif. Not to mention the Polynesian motif, with the exception of a low frequency in the Philippines and other places of Islands Southeast Asia, it is pretty much limited to Polynesians (98%).

      The most likely reason why current Polynesians today are matches to these two Botocudo/Polynesian skulls has a lot to do with the fact that Polynesians lack genetic diversity due to the extended period of bottle necking (founder’s effect) which has gone on for 3,500 years. We already can share as high as 500cM for with other Polynesians not of the same geographic area. This is obvious if one were to look at the matches (at the default) of the Botocudo matches and see how these matches’ ONE TO MANY list looks like.

      • Bot15:

        Ch 7
        Start Location 73,206,361
        End Location 76,936,171
        cM 3.3
        SNP’s 310

        So, you are not allowed to lower the SNP’s? Is this what you mean by a “forced match”? I’m not a geneticist. Can you please explain?



      • I found another way to compare one’s results to this data, using Felix’s ancient calculator.

        For example my results ended up as:
        BOT15 7.67%
        BOT17 6.28%

        It’s a fun tool.

      • Apologies. I made an error on the calculation. I had it set to “Compound Segments”. The “Total Shared DNA” is:
        BOT15: 16.67%
        BOT17: 14.78%

        Kalani, I just a read a blog post where I believe you mentioned that the BOT15 and BOT17 ancient DNA samples were actually found out to be from a NZ Museum and Polynesian, not from Brazil. Is that correct? That seems to explain why it matches my mitochondrial dna.

        What I did find puzzling, and I wonder if other people have experienced this, the ancient calculator shows I have 11.27% DNA shared with an Australian Aboriginal sample and yet, as far I know about my family history, we have never lived in Australia. Perhaps these is shared Austronesian or Denisovan genes?

      • Sebastian, I guess you mean the same mtDNA haplotype as one of the Botocudo skulls? I have project members with both B4a1a1 and B4a1a1a which I believe is what they found for the 2 skulls. And also the Y is of C haplogroup (M208), which is my paternal grandfather’s haplotype. But doing a ONE TO MANY with the Botocudo skulls’ sample at the normal threshold pulls up current people and my family members matches both, either one or the other. Definitely not as ancient as they thought.

  10. Pingback: Some Native Americans Had Oceanic Ancestors | Native Heritage Project

  11. Pingback: The Best and Worst of 2015 – Genetic Genealogy Year in Review | DNAeXplained – Genetic Genealogy

  12. Just wondering how i got to share DNA Bot15 and Bot17 (Polynesian DNA)could this have anything to do with mutiny on the bounty? any English person (like myself) will be familiar with the true story in 1784 of how in the crew of the ship called The Bounty had a mutiny and went amongst the islands and had families with the local women. some of the descendants are still on Pitcairn and Trisdan da cunha, Im pretty sure a lot got brought back to England to face justice and then sent off to Australia I imagine if they weren’t hung

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s