World Families Forums - Decyphering the origins of the western y-dna of Latin America

Welcome, Guest. Please login or register.
December 18, 2014, 12:21:10 AM
Home Help Search Login Register

+  World Families Forums
|-+  General Forums - Note: You must Be Logged In to post. Anyone can browse.
| |-+  R1b General (Moderator: rms2)
| | |-+  Decyphering the origins of the western y-dna of Latin America
« previous next »
Pages: [1] Go Down Print
Author Topic: Decyphering the origins of the western y-dna of Latin America  (Read 4229 times)
argiedude
Old Hand
****
Offline Offline

Posts: 146


« on: August 04, 2010, 06:36:54 PM »

I started this a week ago in rootsweb, which I titled Decyphering the Sephardic ancestry of Mexico. I noted that Latin Americans' western y-dna (no Q or E1b1a) was way too low to justify a Spanish origin for it, so I looked further at it and discovered they also had an unsually high rate of J1. The answer was obvious, they must have a significant minority Sephardic Jewish ancestry. I then looked at J2 and also discovered that Latin Americans have an extra amount of J2, compared with Spain, and which would fit perfectly with Sephardic ancestry, because the extra J1 and J2 amounts were similar in size, as they are in Jews, Sephardic or not.

But now I've changed my mind. I won't spill it out immediately, but I'll be finished in a couple of days. And what I'm going to conclude involves the main topic of this forum very, very much. I think this is going to be very interesting.

Here's the main map to highlight the problem of Latin Americans' (western only) ancestry (if it's too small click on the link beneath it):



http://img508.imageshack.us/img508/5464/decypheringthewesternyd.gif

These results were estimated using thousands of samples, so they're very trustworthy. Spain has 65% R1b1b2, but Spanish speaking Latin Americans have only 55% R1b1b2 (of their western y-dna only, no E1b1a or Q). Given that the sample size in either population was in the thousands, a difference of 10% is absolutely not due to random variance in the results themselves, it's signaling clearly that there's something a little more complex about Latin's western ancestry then just Spanish ancestry. Notice that Brazil, on the other hand, has 57% R1b1b2, exactly as much as Portugal. This reenforces the conclusion, since Portugal obviously operated under their own separate rules of who was allowed to go to the Americas. By the way, it should be kept in mind that the modern y-dna ancestry of Latin Americans is almost exclusively the product of the first couple of centuries of colonialism, and quite honestly, probably just the first 50 years after 1492. Since then, I wouldn't be surprised if the addition of African y-dna has been more significant than the addition of subsequent European y-dna (African y-dna in Spanish Latin Americans is about 3%).

I already explained my (initial) thoughts about Sephardic ancestry, in trying to decypher who caused this anomaly in Latin Americans' y-dna. But that changed after I looked at a new haplogroup, in the E branch. That's what I'll write about tomorrow. The evidence from this haplogroup is very suggestive. The ultimate implications have everything to do with the main theme of our forum: R.

Sincerely,
argiedude
Logged

y-dna: R1b L21
mtdna: U5
argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #1 on: August 04, 2010, 06:48:57 PM »

Notice in the map that the 3 most deviant results for Latin Americans, 60% and 63% in blue and 44% for Peru, all have well below average sample sizes. All of them are probably outliers, and the real frequencies of R1b1b2 in their regions are probably not different than elsewhere in Latin America.

And notice that east-west cline in Iberia! As alan was just pointing out today in the Iberian L21 thread, about how L21 is starting to give the appearnace of an east-west cline.
« Last Edit: August 04, 2010, 06:54:00 PM by argiedude » Logged

y-dna: R1b L21
mtdna: U5
IALEM
Old Hand
****
Offline Offline

Posts: 267


« Reply #2 on: August 05, 2010, 05:28:40 AM »


 By the way, it should be kept in mind that the modern y-dna ancestry of Latin Americans is almost exclusively the product of the first couple of centuries of colonialism, and quite honestly, probably just the first 50 years after 1492.


Where did you get that idea? Migration to Latin America in the 18th, and especially 19th and early 20th century has been quite heavy.
Logged

Y-DNA L21+


MDKA Lope de Arriçabalaga, born c. 1390 in Azcoitia, Basque Country

argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #3 on: August 06, 2010, 06:07:05 PM »

My first impression was that Latin Americans' lower rate of R1b1b2 (considering only their western y-dna, not Q or E1b1a) was due to Sephardic ancestry. This was backed up by the fact that J1 and J2, typical Jewish haplogroups, was higher in Latin Americans than in Spain, by 2,5% in either case. But 2 problems arose. Judging from the drop in R1b1b2, the Sephardic ancestry in Latin Americans would have to be 18%, and since J1 and J2 were 2,5% each, this meant those Sephardic people had just 28% J1 + J2, while genetic tests of Sephardic Jews today, and of almost any group of Jews, shows that J1 and J2 add up to between 40% and 50%. The second problem is simply that 18% seems too high. That's subjective, of course. So I tried to look for another haplogroup that could clarify the picture. I found two.

The first one is E1b1b1b (M81). This haplogroup, unlike most in Iberia, has a very clinal distribution. As can be seen in the following map, which I made 2 years ago, the western half of Iberia has 5% or 6% M81, while the eastern half has 1% to 3%. Note that since I made this map two more sample sets were published for southeast Spain, the most undersampled region of Iberia, one for the Murcia region (look in the map for the city of Murcia) with 2% M81, and another for Valencia with 3% M81.



http://img708.imageshack.us/img708/2013/iberianm817201samples.gif

The frequency of M81 amongst Latin Americans is 3,6%. [Remember, this is calculated over their western y-dna only, not Q or E1b1a, and I'm always referring to Spanish Latin America, not Brazil.] The frequency in Spain is 4,3%. This suggests greater ancestry from eastern Spain, but if there's any significant amount of Sephardic (or other western) ancestry in Latin Americans then this lower frequency would just be reflecting that non-Iberian ancestry. So I analysed instead the internal structure of M81. The following map shows the variance of M81 in several regions of Iberia, and in Latin America. It also shows the presence of 2 notable subclusters of M81 which are concentrated in western Iberia.



http://img230.imageshack.us/img230/5464/decypheringthewesternyd.gif

Brazil and Portugal both have a variance of 0,12. But Latin Americans have 0,11 while Spain has 0,08. In Iberia, there's a strip of coastal land from south Portugal to Galicia which has a high variance. The rest of Iberia drops to a much lower variance. M81's variance in Latin Americans is suggesting a disproportionately high ancestry from the west.

Even more curious, about 4% of Latin American M81 samples belong to two clusters of M81 found in west Iberia. These clusters make up 9% of Portuguese M81, 4% of Galician/Asturian M81, and 0% of the M81 of the rest of Spain (0 out of 105 M81 samples). A uniformly pan-Spanish origin for Latin Americans' western y-dna would have resulted in less than 1%, not 4%. That's a 5-fold difference in frequency, and it once again suggests that Latin Americans' Iberian ancestry comes disproportionately from the west of the peninsula. I would say specifically from Galicia, but I've just recently learned that Seville was the administrative capital in charge of the Latin American colonies, and of course this could make a good case that the mass of sailors and soldiers that molded Latin Americans' y-dna came disproportionately from the Andalucia region. Whatever the case, the evidence from M81 paints a picture of greatly disproportionately western Iberian origin for Latin Americans' y-dna.

If that's the case, then all my speculations on Sephardic ancestry are way off. Not only does western Iberia/Spain have a notably lower frequency of R1b1b2, which would explain to a great extent Latin Americans' also lower frequency of R1b1b2, but it also has both a higher frequency of J1 and J2! But there's still a problem with the Latin American R1b1b2. Even in the extreme case that all of Latin Americans' Spanish y-dna came from west Spain, that would still result in Latin Americans having 58% R1b1b2. But they have instead 55%. That's a much smaller difference then before, but it's still probelmatic, given that these figures were calculated using gigantic sample sizes, and thus have a very small margin of error, of maybe just 1%. The difference shouldn't be that great, even if it doesn't seem too much.

A combination of western Spanish, eastern Spanish, and Sephardic ancestry seems the best fit to me. I'm thinking 70% west Spanish origin, 20% east Spanish origin, and 10% Sephardic origin. I didn't calculate those fractions using any formula, it's just an educated guess considering everything I've observed so far. In this way, everything more or less fits together, including the drop in R1b1b2, the higher rate of J1 and J2, the internal structure of M81, and other minor details such as the R1b1* cluster in Puerto Rico.

And now for the other haplogroup: R1b1b2.

R1b1b2 has a moderately clinal distribution, forming an east-west cline, but 2 of its subhaplogroups have a hugely distorted distribution, much worse than M81, the most lopsided haplogroup distribution in Iberia. They're SRY2627 and M153. Making it even better, they're rather quite common, making up 9% of Iberia's y-dna, and probably 10% or 11% of Spain's alone. Here's another map I made some time ago showing their distribution in Iberia:



http://img413.imageshack.us/img413/8766/iberiasry2627m153.gif

These 2 haplogroups are perfect. Their combined rate in northwest Spain is 2%, in the southwest it's 3%, but in the northeast it's 22% and in the southeast it's 10%. A huge difference, much greater than for M81, and with even bigger absolute frequencies. The only problem? Geneticists haven't studied these SNPs in Latin Americans, and the samples from commercial databases only identify the samples that tested positive for these SNPs but, crucially, they don't show how many tested negative, without which we can't make an estimate of their frequency. FTDNA has the answer. What would be needed is to know the SRY2627 and M153 results of several hundred R1b samples with origin in Latin America (excluding Brazil). I wonder if we could convince them. It would be awesome to dilucidate the specific regional origin within Spain of the western ancestry of Latin Americans, more so if in fact it comes vastly disproportionately from some specific region, instead of being simply a pan-Spanish origin.
« Last Edit: August 06, 2010, 06:20:32 PM by argiedude » Logged

y-dna: R1b L21
mtdna: U5
Mike Walsh
Guru
*****
Offline Offline

Posts: 2964


WWW
« Reply #4 on: August 07, 2010, 12:38:01 AM »

My first impression was that Latin Americans' lower rate of R1b1b2 (considering only their western y-dna, not Q or E1b1a) was due to Sephardic ancestry. ....
A combination of western Spanish, eastern Spanish, and Sephardic ancestry seems the best fit to me. I'm thinking 70% west Spanish origin, 20% east Spanish origin, and 10% Sephardic origin. ..
Argiedude, thank you for studying this. You mention "that Latin Americans' Iberian ancestry comes disproportionately from the west of the peninsula."
You also say that "A combination of western Spanish, eastern Spanish, and Sephardic ancestry seems the best fit to me."

I'm not sure if you are including Portugal with western Spanish.  Are you?

Have you considered the effects of the staging of the Spanish Inquisition?

Quote from: wikipedia
The Jews of the kingdom of Castile emigrated mainly to Portugal (from where they were expelled in 1497) and to Morocco.
http://en.wikipedia.org/wiki/Spanish_Inquisition
http://en.wikipedia.org/wiki/Portuguese_Inquisition

A western Iberia orientation for immigration to Latin America could still mean a heavy Shephardic and Moorish influence.  The Inquisition was initiated in Spain and then decades later came to Portugal.  Portugal was a temporary "safe haven."
« Last Edit: August 07, 2010, 12:39:28 AM by Mikewww » Logged

R1b-L21>L513(DF1)>S6365>L705.2(&CTS11744,CTS6621)
argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #5 on: August 07, 2010, 11:57:05 AM »

I made some new calculations and they refute everything I concluded. Grrrr...

There are several Iberian R1b1b2 haplogroups that have a high rate of 437=14. I think P312* is one, and also M153. I calculated what percentage of the R1b1b2 samples had 437=14, and found that there's a difference in Iberia, and once again the difference follows an east-west cline:


Region            R1b1b2 samples       % with 437=14
Northwest                 500                21%
Southwest                 400                18%
Northeast                 400                27%
Southeast                 150                29%
Baleares                  100                42%

Latin America (Spanish)  2000                29%

Brazil                    400                17%


Northwest includes north Portugal, Southwest includes central & south Portugal.

Absolutely terrible. Latin America (Spanish) is indicating in this particular comparison that it would be completely descended from the eastern half of Spain!

Also notice how Brazil, yet again, is a mirror image of Portugal.

I'm not sure if you are including Portugal with western Spanish.  Are you?

No, though I'm not so sure anymore that Portuguese would have been totally excluded by the Spanish crown from participating in the conquest.
« Last Edit: August 07, 2010, 12:27:07 PM by argiedude » Logged

y-dna: R1b L21
mtdna: U5
argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #6 on: August 07, 2010, 04:55:58 PM »

Another piece of evidence refuting a predominantly western Spanish origin for Spanish Latin Americans' western y-dna. I calculated the frequency of the modal haplotype of R1b1b2-M65 as a percentage of the total R1b1b2 samples in each region.


Region            R1b1b2 samples       % with GD<=2 from M65 modal
Northwest                 600                0,0%
Southwest                 600                1,0%
Northeast                 500                3,5%
Southeast                 300                1,5%
Baleares                  200                0,5%
Portugal only             900                0,2%

Latin America (Spanish)  2500                1,6%

Brazil                    500                0,2%


Northwest includes north Portugal, Southwest includes central & south Portugal.

Absolutely terrible, again. This comparison points to a pan-Spanish origin for Spanish Latin Americans, or to a predominantly southwest Spanish origin.

Brazil, as usual, is identical to Portugal. That's very interesting. In every test I made between Brazil and Portugal, their y-dna is identical. But comparing Spanish Latin Americans' western only y-dna with Spain's y-dna is a disaster. I have no idea what to make of Spanish Latin Americans' western y-dna. It's a mystery. If only FTDNA would allow us to view the results of a couple hundred random Latin American R1b samples, to see how many tested positive and negative for SRY2627 and M153. That would be by far the most significant test to decypher this mystery. So far, things tend to point somewhat to a predominantly southwest Spanish origin for Latin Americans, but it's all very iffy. Not Brazil, of course, their western y-dna is a remarkably flawless mirror image of Portuguese y-dna.
« Last Edit: August 07, 2010, 05:04:00 PM by argiedude » Logged

y-dna: R1b L21
mtdna: U5
Maliclavelli
Guru
*****
Offline Offline

Posts: 2157


« Reply #7 on: August 09, 2010, 04:41:40 PM »

Investigating on the possible Italian origin of a Cavazos who is on Ysearch  
and who could be an R1b1b2a with DYS393=13 and DYS461=12 like some individuals I am studying
about the SNP L277, I found this:

"Muchos italianos llegaron a la Nueva España desde los inicios de la conquista,
ya sea como misioneros, marineros y colonizadores. Preferentemente procedían del
Reino de Sicilia y Nápoles que pertenecían a los Habsburgos (but during the first
centuries after 1492 South Italy was under Catalunya and Spain, and this could have
facilitate a massive migration to Americas), aunque también llegaron de las regiones
norteñas de Lombardía y del Véneto, que estaban bajo dominio de los Austrias.

En el siglo XVII arribaron pobladores al Nuevo Reino de León con probable ascendencia
italiana: Juan Cavazos, quien llegó en 1630 y que estaba casado con Elena de la Garza,
firmaba como Cavasso o Cavassos. Aunque era de Castilla la Vieja, probablemente sus
orígenes están en Italia pues existe un pueblo llamado Cavazzo Carnico en la provincia
de Urdina. Hacia 1636 llegó José de Cantú y ese apellido abunda mucho en la provincia de
Como (this was the thinking also of Rich, but not mine). Otro de los pobladores que
llegaron es Juan de Espíndola, quien declara en 1643 que es hijo de Alejandro de Spíndola
de origen genovés. Se dedicó al comercio y fue de los que introdujeron el culto a la
Virgen de San Juan de los Lagos" etc.

I am not certain that Cavazos (Cavasso) is of Italian extraction, although in Italy there is all
the line of linked surnames (Cavasso, Cavassa, Cavassi (and Cabassi), Cavassin etc.).
Only a paper trail can document a theory, and less certain I am that is of Italian extraction
the surname Trivino, like the generous writer, in spite of his surname (Antonio Guerrero
Aguilar Jueves) is.

The temptation to find a formula that solves all our questions is strong, but dangerous: see
the ridicolous paper of two Brazilian scholars on DYS464 I discussed in another thread.
I spent thirty years of patient researches to arrive to my ancestor Signorino del Badia at the
beginning of 14th century and know very well how many mirages are along the street.

Solve which is the link between you and these Callaways I spoke on another thread and work on
archives, and perhaps you will remove to Rich also the unique R-L21 found in Italy.
« Last Edit: August 09, 2010, 05:05:01 PM by Maliclavelli » Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

Mike Walsh
Guru
*****
Offline Offline

Posts: 2964


WWW
« Reply #8 on: August 10, 2010, 09:36:01 AM »

...... So far, things tend to point somewhat to a predominantly southwest Spanish origin for Latin Americans, but it's all very iffy. ...
All of Columbus' voyages left from the Bay of Cadiz and I believe many of the other voyages did as well so it makes sense that their might be a southwest Spain bias.
Logged

R1b-L21>L513(DF1)>S6365>L705.2(&CTS11744,CTS6621)
IALEM
Old Hand
****
Offline Offline

Posts: 267


« Reply #9 on: August 10, 2010, 02:18:46 PM »

Argiedude
Are you sure the Latin American samples are of Spanish origin only? As Macliavelli points, there was a strong Italian migartion during the Colonial period, and after that in the 19th century. In the 20th century there is also a sizeable migration of Christian Syrians and Lebanese, as well as Askenazi Jews. Even as ealry as the 17th century a relation of the population of the Virreinato del Perú mentions that less than half the white population of Perú was Castillian, the rest being Italians, Irish, Greeks and Flemish.
Logged

Y-DNA L21+


MDKA Lope de Arriçabalaga, born c. 1390 in Azcoitia, Basque Country

argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #10 on: August 10, 2010, 07:19:29 PM »

Argiedude
Are you sure the Latin American samples are of Spanish origin only?

Iberia has the lowest frequency of R1a in Europe, at exactly 1,3%, calculated using only truly Iberian-origin samples. Italy has 3,5%. The western samples from Latin America numbered around 5,000, and their frequency of R1a was slightly less than 1,3%. In contrast, the Argentine and Brazilian samples from yhrd that are labeled "European" have about 3% R1a (of their western y-dna only). Oh, yeah, I forgot to clarify, those 5,000 samples (actually 7,000 but 5,000 of them are western y-dna) are specifically labeled as mestizo.

Also, virtually none of the Iberian R1a belongs to the 2 distinctive M458+ clusters that make up a huge chunk of East European R1a. Likewise, not one of the many R1a samples from the Latin American mestizos belonged to either of these 2 clusters. In contrast, the R1a from the European samples of Argentina and Brazil, in yhrd, had quite a few that belonged to either cluster of M458.

Quote
As Macliavelli points, there was a strong Italian migartion during the Colonial period, and after that in the 19th century.

This is subjective, but I believe that the mestizo population, which makes up almost the entire population of Latin America, was "forged" in the first 50 years after 1492. What happened in the centuries after was relatively very minor, and I think it's possible that the addition of sub-Saharan y-dna was greater than the addition of European y-dna; sub-Saharan y-dna makes up about 3% of mestizos' y-dna. Europeans that went to Latin America between 1600 and 1800 became part of the high classes, who made up and make up 1% of the population. Lastly, between 1870 and 1920, as part of the same phenomenon occuring back then in the USA, Canada, and Australia, several million people European people moved into Argentina and south Brazil, and almost nowehere else in Latin America. They mostly didn't become mixed with the mestizos. I excluded samples from Argentina precisely because of this problem. From Chile to Mexico, there was hardly any European people moving in, and the few that did overwhelmingly became part of the high classes, so even those didn't affect mestizos' y-dna.

Quote
Even as ealry as the 17th century a relation of the population of the Virreinato del Perú mentions that less than half the white population of Perú was Castillian, the rest being Italians, Irish, Greeks and Flemish.

But the white population could have been just 1% of the total. The Mexican high classes today are so genetically distinct from the other 95% or more of the country they might as well be 2 different nations. And yet both have been living side by side for 400+ years. And though I've seen a few non-Spanish names that indicate an origin in Mexico, it looks like 98% or more are Spanish surnames.
Logged

y-dna: R1b L21
mtdna: U5
Maliclavelli
Guru
*****
Offline Offline

Posts: 2157


« Reply #11 on: August 20, 2010, 08:55:59 AM »

Not so long ago we discussed about the possible Italian origin of Cavazos (one Mexican tested
from Spanish origin: Ysearch 44QRU). Now, with the last update of SMGF, I have updated (and completed) Rosaio
(Ysearch BKKB4). He is very close to another Rosaio tested by FTDNA (Ysearch 2ZZR3) and also to
Manzino (Ysearch A8QUA): both are from Giusvalla, near Savona, the town of Rosaio (BKKB4).
Cavazos matches closely all them, but, being their values very close to the modal, it isn't
strictly probative. We would need of the five fast mutating markers from DYS607 (which is the
less mutating) to CDYb, in fact Rosaio (2ZZR3) and Manzino (A8QUA) (15,18,17,35,39 / 15,18,17,35,39)
are very closely related and far from an apparent close match like McCaulley (2PVKN) who is
14,18,17,36,37.

This said, and waiting for a deep exam of Cavazos, an Italian origin of him is yet possible, remembering that
surnames in -asso are typical Ligurian (the famous Spanish painter Picasso takes his surname from
a Ligurian ancestress of his). One of the Rosaio tested (2ZZR3) is R1b1b2a1b and probably are all
the others, then the haplogroup isn't so recent and, even though the markers values are in these
cases around the modal, others have had the possibility to have gone for the tangent, like these
fast mutating ones. A10=14 isn't so diffused: it is mine, who am R1b1b2a.
Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

adriano
New Member
*
Offline Offline

Posts: 1


« Reply #12 on: August 29, 2010, 07:31:23 AM »

Investigating on the possible Italian origin of a Cavazos who is on Ysearch  
and who could be an R1b1b2a with DYS393=13 and DYS461=12 like some individuals I am studying
about the SNP L277, I found this:

"....orígenes están en Italia pues existe un pueblo llamado Cavazzo Carnico en la provincia
de Urdina....
Hi,
I'm from Cavazzo Carnico and I have seen lots of genealogical records from this small town (almost all the available sources from 1500 to 1900). I have never seen any relation to Spain, South America or whatever might be related to Cavazos.
Regards,
Adriano
Logged
Maliclavelli
Guru
*****
Offline Offline

Posts: 2157


« Reply #13 on: August 29, 2010, 07:59:45 AM »

Caro Squecco,
who looks at Italy from far, doesn't understand anything of Italy and of Italians. I had understood soon that Cavasso hadn't anything to do with your town, and in fact I have given other explications, that remain all valid, I think.
As you have written and I can contact you, have you seen this "bomb" of R-U152 in North Italy, but also Central, South, Islands? I think that nobody will speak more of Celts, etc.
If we Italians aren't united, you can see how we are treated by this presumed "Celts", etc. I avoid to answer, because they are painful, in the heart and in the mind.

Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

Pages: [1] Go Up Print 
« previous next »
Jump to:  


SEO light theme by © Mustang forums. Powered by SMF 1.1.13 | SMF © 2006-2011, Simple Machines LLC

Page created in 0.081 seconds with 18 queries.