Page 1 of 7 123 ... LastLast
Results 1 to 10 of 64

Thread: Who have more genuine "Arab" ancestry -- Lebanese, or Tunisians/Algerians?

  1. #1
    Registered Users
    Posts
    2,686
    Sex
    Location
    USA
    Nationality
    American
    Y-DNA
    R1a1a
    mtDNA
    H

    United States of America

    Who have more genuine "Arab" ancestry -- Lebanese, or Tunisians/Algerians?

    Since we know Lebanese are very close to the ancient Canaanite samples possibly 90%+, who has more Arab ancestry from the Arab conquest? Tunisians/Algerians, or Lebanese?

    And how do we know?

  2. #2
    Based on Y-DNA: Lebanese.

  3. The Following User Says Thank You to Shamayim For This Useful Post:

     Power77 (12-12-2018)

  4. #3
    Registered Users
    Posts
    249
    Y-DNA
    I-Y16419
    mtDNA
    T2e1

    I used to believe that Arab migrations didn't affect North Africa or the Near East autosomally, but I'll keep an open mind.

    I want to add that North Africa received additional migrations from Arabia in the 11th century, a large movement of tribes connected to Bani Hilal. that I believe was large enough to bring North Africa closer to the Near East genetically, and had an effect on their Arabization.

  5. The Following 3 Users Say Thank You to IronHorse For This Useful Post:

     drobbah (12-12-2018),  pgbk87 (12-12-2018),  Power77 (12-12-2018)

  6. #4
    Registered Users
    Posts
    379
    Sex
    Omitted
    Nationality
    Canadian
    Y-DNA
    E-V32
    mtDNA
    L0a1*

    Somaliland Ethiopia Arab League Adal Sultanate
    Quote Originally Posted by IronHorse View Post
    I used to believe that Arab migrations didn't affect North Africa or the Near East autosomally, but I'll keep an open mind.

    I want to add that North Africa received additional migrations from Arabia in the 11th century, a large movement of tribes connected to Bani Hilal. that I believe was large enough to bring North Africa closer to the Near East genetically, and had an effect on their Arabization.
    Don't the Libyan Arabs have signficant Arabian ancestry? I always thought they were the most 'Arab' in the Maghreb
    My Father's K15
    76.5% Somali + 23.5% Ethiopian_Wolayta @ 2.52
    94.6% Somali + 5.4% Ethiopian_Ari_cultivator @ 2.58


    Mother's K15

    98.2% Somali + 1.8% Lebanese_Druze @ 1.87
    98.1% Somali + 1.9% Tunisian_Jewish @ 1.91
    98.1% Somali + 1.9% Palestinian @ 1.91
    97.8% Somali + 2.2% Egyptian @ 1.92

  7. The Following 2 Users Say Thank You to drobbah For This Useful Post:

     pgbk87 (12-12-2018),  Power77 (12-12-2018)

  8. #5
    Registered Users
    Posts
    94
    Sex
    Ethnicity
    Lebanese
    Y-DNA
    J2a2
    mtDNA
    X2i

    Lebanon Albania United States of America
    Well I think it would be pretty tough to compare the two populations. On one hand, you have the Lebanese who are already a Semitic people sharing some overlap with Arabs. On the other, Algerian/Tunisians who I assume are mainly Berber mixed with not only Arab, but with a large Punic history (that's two Semitic populations having their input). It is definitely easier to differentiate native from foreign for the North Africans, but then you have the question of is it Arab or Phoenician which isn't as easy on such small levels.

    Some Y-DNA for Tunisians; 17% J1 3% J2. Algerians: 22% J1 5% J2. Lebanese can be estimated at around 20% J1 30% J2

    The fact that J1 is native to Lebanon and still at an almost equal frequency shows that more Arab tribes actually migrated to North Africa. That is assuming there was little to no J1 present in North Africa from pre-Arab times. Looking at those numbers I'm guessing there wasn't a huge Punic impact either, or at least way smaller than from Arabs. Thats just looking at the two haplogroups... maybe other studies or autosomal would show something else.
    78.1% Lebanese (Lebanon) + 21.9% Kosovar (Kosovo) @ 3.3

  9. The Following 6 Users Say Thank You to KingofPhoenicia001 For This Useful Post:

     drobbah (12-12-2018),  Erikl86 (12-12-2018),  hartaisarlag (12-12-2018),  pgbk87 (12-12-2018),  Power77 (12-13-2018),  Shamayim (12-12-2018)

  10. #6
    Look at the arabian component in eurogenes k36 for these populations and compare them with Samaritans or the sidon samples. Keep in mind any difference in other components in comparison to the aforementioned groups and how that affects how you interpret the arabian results. That'll help you gauge how significant/insignificant arabian ancestry in these groups are (also keep in mind possible ancestry from egypt which could account for elevated arabian, so keep an eye on how copts score as well). For the record, european jews have quite a handsome amount %-wise when it comes to the arabian component on eurogenes k36. Since they derive at least half of their ancestry or more from the southern levant, that too is something worth keeping in mind for the time being (until we get more recent ancient levantine samples).
    Last edited by notasuckah; 12-12-2018 at 06:24 AM.

  11. The Following 3 Users Say Thank You to notasuckah For This Useful Post:

     palamede (12-12-2018),  pgbk87 (12-12-2018),  Power77 (12-13-2018)

  12. #7
    Gold Member Class
    Posts
    1,563
    Sex
    Ethnicity
    Jewish (Ashkenazi)
    Y-DNA
    Q-YP3924 (Q-M378)
    mtDNA
    K1a1b1a

    Israel
    Since the question of who is an Arab is very fluid, this raises a whole bunch of philosophical questions.

    But even if one is to ignore the linguistic and cultural aspects of who is an Arab, then there's an obvious cline from the BedouinA (which in the Levant itself are considered the most pristine Arabs) all the way to Yemenites, and the problematic issue that both Levantines and Arabs seem to share substantial Neolithic Levantine ancestry. In fact, modern Arabs seem to be most differentiated from modern Levantines in their much lower Iran_ChL admixture, which is Bronze Age Levantine is roughly 50/50, but in some pristine Arab populations it's as low as only ~10% Iran_ChL.

    As was already pointed out, the fact that J1, which is a hallmark Arab uniparental is also native to the Levant, it's going to be a uniparental headache as well. But uniparentals on themselves are not good indicators of admixture proportions, they just show evidence of ancestry from a certain population.

    So to answer this question, I will propose the following models:

    BedouinB
    Mozabite

    For Algerian and Tunisian:

    Code:
    [1] "distance%=1.7162"
    
             Algerian
    
    Mozabite,93.6
    BedouinB,6.4
    Code:
    [1] "distance%=1.8271"
    
             Tunisian
    
    Mozabite,81
    BedouinB,19
    And for Lebanese, I chose Christians, and I used the following populations. I resorted again to use BedouinB as a pristine Arabic population given that we don't have ancient Arabs to use, and given Bedouins own tribal code and low level of intermarriage, as well as the role they played with the spread of Islam. BedouinA would be less adequate as they seem to be quite Levantine shifted - perhaps even originate from some Arabized noamdic Levantine population, who knows. I actually resorted into doing something I don't like, which is mixing ancient and modern populations - but again, we don't have ancient Arabic samples yet.

    So these are the populations I've chosed:

    BedouinB
    Levant_BA_North
    Yamnaya_Samara (remember Lebanese should be differentiated from Bronze Age Levantines by having <10% Steppe-like admixture, according to Haber et al.)

    And here's the result:

    Code:
    [1] "distance%=1.9421"
    
             Lebanese_Christian
    
    Levant_BA_North,91.4
    Yamnaya_Samara,8.6
    Very similar to Haber et al. qpAdm model - but look how the BedouinB sample got kicked out. If I take out the Steppe-like component, this is what I'm left with:

    Code:
    [1] "distance%=3.1951"
    
             Lebanese_Christian
    
    Levant_BA_North,100
    Again, no BedouinB.

    Less but not least, I'll try to model Muslim Lebanese, to see how much Arabic ancestry they have:

    Code:
    [1] "distance%=2.2521"
    
             Lebanese_Muslim
    
    Levant_BA_North,86
    Yamnaya_Samara,14
    Strangely enough, they show up with way too much Steppe-like ancestry, but also the distance is higher than in the case of Christian Lebanese, so I guess there is something missing which skews up the model. I'll try to add some ancient East African ancestry, like Tanzania_Zanzibar_1400BP. It should be a very good model to use when trying to establish SSA admixture in modern Arabs for several reasons:
    1. It's dated to 1,400 years before present - just before the rise of Islam.
    2. It's from Zanzibar - THE place where SSA slaves were imported into the Middle East. In fact, the name itself - Zanzibar - is derived from the Arab term for SSA slaves - Zanj.

    So here goes:

    Code:
    [1] "distance%=1.9018"
    
             Lebanese_Muslim
    
    Levant_BA_North,84.8
    Yamnaya_Samara,13.8
    Tanzania_Zanzibar_1400BP,1.4
    Still elevated Steppe-like admixture compares to Christian Lebanese, which is interesting, but the distance is now lower and similar to the distances we've seen in all the other models (slightly less than 2%). The SSA ancestry also first what is usually estimated for Lebanese Muslims - less than 2%.

    However, in the mean time it seems that as opposed to the Algerians and Tunisians, BedouinB shows no appearance in both Christian or Muslim Lebanese.

    Last but not least, I'll try to also model Lebanese Druze, which is problematic by itself since they are a very endogenic population, but there are conflicting traditions of where they originate - some narrative claim they have Yemenite ancestry, others Anatolian ancestry - so should be interesting to see if they have any BedouinB ancestry:

    Code:
    [1] "distance%=1.9586"
    
             Lebanese_Druze
    
    Levant_BA_North,87.6
    Yamnaya_Samara,12.4
    Seem to also have slightly elevated Steppe-like ancestry compared to the Christian samples, but still - no appearance of the BedouinB sample.

    Just for the heck of it, I'll use the BedouinA instead of B, to see if anything changes. I'll try just the Christian Lebanese sample:

    Code:
    [1] "distance%=1.9421"
    
             Lebanese_Christian
    
    Levant_BA_North,91.4
    Yamnaya_Samara,8.6
    Nothing, despite the fact that BedouinA sample is much closer than BedouinB is in single distance from Lebanese.

    Let's also try to model Algerian with it (Tunisian might be biased because it might have also Punic ancestry which might cause Levant-shifted BedouinA sample to show up in higher proportions):

    Code:
    [1] "distance%=1.4106"
    
             Algerian
    
    Mozabite,89.6
    BedouinA,10.4
    Even higher than BedouinB, and the distance is better, which can point out to two possibilities, IMO:
    1. Algerians also have some Punic ancestry.
    2. Arabs arrived to North Africa after mixing with Levantines slightly - so they haven't arrived as totally pristine Arabs.

    I think in this case, option 2 plays much more plausible part - that is, that Arab tribes mixed with Levantines which can be best represented via BedouinA, are the type of Arabs which settled in North Africa. This also fits nicely with the path Islam spread into North Africa (via the Levant):




    So to summarize this and to answer Sikeliot's question, it seems like Algerians and Tunisians might have mixed more and perhaps were even settled by more Arab tribes than Lebanese and Lebanon.
    Check out my Hidden Content
    My Y-DNA: Q-M242 -> Q-L232 -> Q-L275 -> Q-M378 -> Q-Y2016 -> Q-L245 -> Q-FGC1904 -> Q-Y2209 -> Q-Y2225 -> Q-Y2197 -> Q-Y2750 -> Q-YP1004 -> Q-YP3924;
    My mtDNA: K1a1b1a;

    My dad's mtDNA: K2a2a;

  13. The Following 12 Users Say Thank You to Erikl86 For This Useful Post:

     Просигој (12-13-2018),  alexfritz (12-12-2018),  Bar (12-18-2018),  dany198124 (01-05-2019),  Govan (12-12-2018),  IronHorse (12-12-2018),  kikkk (12-12-2018),  KingofPhoenicia001 (12-12-2018),  Michalis Moriopoulos (12-13-2018),  palamede (12-12-2018),  Power77 (12-13-2018),  Sikeliot (12-12-2018)

  14. #8
    Registered Users
    Posts
    249
    Y-DNA
    I-Y16419
    mtDNA
    T2e1

    Quote Originally Posted by Erikl86 View Post
    Since the question of who is an Arab is very fluid, this raises a whole bunch of philosophical questions.

    But even if one is to ignore the linguistic and cultural aspects of who is an Arab, then there's an obvious cline from the BedouinA (which in the Levant itself are considered the most pristine Arabs) all the way to Yemenites, and the problematic issue that both Levantines and Arabs seem to share substantial Neolithic Levantine ancestry. In fact, modern Arabs seem to be most differentiated from modern Levantines in their much lower Iran_ChL admixture, which is Bronze Age Levantine is roughly 50/50, but in some pristine Arab populations it's as low as only ~10% Iran_ChL.

    As was already pointed out, the fact that J1, which is a hallmark Arab uniparental is also native to the Levant, it's going to be a uniparental headache as well. But uniparentals on themselves are not good indicators of admixture proportions, they just show evidence of ancestry from a certain population.

    So to answer this question, I will propose the following models:

    BedouinB
    Mozabite

    For Algerian and Tunisian:

    Code:
    [1] "distance%=1.7162"
    
             Algerian
    
    Mozabite,93.6
    BedouinB,6.4
    Code:
    [1] "distance%=1.8271"
    
             Tunisian
    
    Mozabite,81
    BedouinB,19
    And for Lebanese, I chose Christians, and I used the following populations. I resorted again to use BedouinB as a pristine Arabic population given that we don't have ancient Arabs to use, and given Bedouins own tribal code and low level of intermarriage, as well as the role they played with the spread of Islam. BedouinA would be less adequate as they seem to be quite Levantine shifted - perhaps even originate from some Arabized noamdic Levantine population, who knows. I actually resorted into doing something I don't like, which is mixing ancient and modern populations - but again, we don't have ancient Arabic samples yet.

    So these are the populations I've chosed:

    BedouinB
    Levant_BA_North
    Yamnaya_Samara (remember Lebanese should be differentiated from Bronze Age Levantines by having <10% Steppe-like admixture, according to Haber et al.)

    And here's the result:

    Code:
    [1] "distance%=1.9421"
    
             Lebanese_Christian
    
    Levant_BA_North,91.4
    Yamnaya_Samara,8.6
    Very similar to Haber et al. qpAdm model - but look how the BedouinB sample got kicked out. If I take out the Steppe-like component, this is what I'm left with:

    Code:
    [1] "distance%=3.1951"
    
             Lebanese_Christian
    
    Levant_BA_North,100
    Again, no BedouinB.

    Less but not least, I'll try to model Muslim Lebanese, to see how much Arabic ancestry they have:

    Code:
    [1] "distance%=2.2521"
    
             Lebanese_Muslim
    
    Levant_BA_North,86
    Yamnaya_Samara,14
    Strangely enough, they show up with way too much Steppe-like ancestry, but also the distance is higher than in the case of Christian Lebanese, so I guess there is something missing which skews up the model. I'll try to add some ancient East African ancestry, like Tanzania_Zanzibar_1400BP. It should be a very good model to use when trying to establish SSA admixture in modern Arabs for several reasons:
    1. It's dated to 1,400 years before present - just before the rise of Islam.
    2. It's from Zanzibar - THE place where SSA slaves were imported into the Middle East. In fact, the name itself - Zanzibar - is derived from the Arab term for SSA slaves - Zanj.

    So here goes:

    Code:
    [1] "distance%=1.9018"
    
             Lebanese_Muslim
    
    Levant_BA_North,84.8
    Yamnaya_Samara,13.8
    Tanzania_Zanzibar_1400BP,1.4
    Still elevated Steppe-like admixture compares to Christian Lebanese, which is interesting, but the distance is now lower and similar to the distances we've seen in all the other models (slightly less than 2%). The SSA ancestry also first what is usually estimated for Lebanese Muslims - less than 2%.

    However, in the mean time it seems that as opposed to the Algerians and Tunisians, BedouinB shows no appearance in both Christian or Muslim Lebanese.

    Last but not least, I'll try to also model Lebanese Druze, which is problematic by itself since they are a very endogenic population, but there are conflicting traditions of where they originate - some narrative claim they have Yemenite ancestry, others Anatolian ancestry - so should be interesting to see if they have any BedouinB ancestry:

    Code:
    [1] "distance%=1.9586"
    
             Lebanese_Druze
    
    Levant_BA_North,87.6
    Yamnaya_Samara,12.4
    Seem to also have slightly elevated Steppe-like ancestry compared to the Christian samples, but still - no appearance of the BedouinB sample.

    Just for the heck of it, I'll use the BedouinA instead of B, to see if anything changes. I'll try just the Christian Lebanese sample:

    Code:
    [1] "distance%=1.9421"
    
             Lebanese_Christian
    
    Levant_BA_North,91.4
    Yamnaya_Samara,8.6
    Nothing, despite the fact that BedouinA sample is much closer than BedouinB is in single distance from Lebanese.

    Let's also try to model Algerian with it (Tunisian might be biased because it might have also Punic ancestry which might cause Levant-shifted BedouinA sample to show up in higher proportions):

    Code:
    [1] "distance%=1.4106"
    
             Algerian
    
    Mozabite,89.6
    BedouinA,10.4
    Even higher than BedouinB, and the distance is better, which can point out to two possibilities, IMO:
    1. Algerians also have some Punic ancestry.
    2. Arabs arrived to North Africa after mixing with Levantines slightly - so they haven't arrived as totally pristine Arabs.

    I think in this case, option 2 plays much more plausible part - that is, that Arab tribes mixed with Levantines which can be best represented via BedouinA, are the type of Arabs which settled in North Africa. This also fits nicely with the path Islam spread into North Africa (via the Levant):




    So to summarize this and to answer Sikeliot's question, it seems like Algerians and Tunisians might have mixed more and perhaps were even settled by more Arab tribes than Lebanese and Lebanon.
    BedouinA is not necesserily Levantine admixed, there are people in Arabia who can be modeled as BedouinA and have no ancestry to Levantine countries, I posted this on the Palestinian thread: https://anthrogenica.com/showthread....l=1#post523241

    me for example, I have no Levantine ancestry, I'm closer to Levantines and Egyptians than to the Saudi or BedouinB samples:

    Bedouin here is BedouinA, dodecad v3.

    # Population (source) Distance
    1 Bedouin (HGDP) 3.24
    2 Egyptans (Behar) 12.91
    3 Jordanians_19 (Behar) 15.24
    4 Yemenese (Behar) 17.56
    5 Syrians (Behar) 17.67
    6 Egypt (Henn) 18.1
    7 Palestinian (HGDP) 18.71
    8 Lebanese (Behar) 20.8
    9 Samaritians (Behar) 22.29
    10 Libya (Henn) 23.09
    11 Georgia_Jews (Behar) 25.38
    12 Iraq_Jews (Behar) 27.01
    13 Iranian_Jews (Behar) 27.37
    14 Druze (HGDP) 28.49
    15 Uzbekistan_Jews (Behar) 28.63
    16 Yemen_Jews (Behar) 29.26
    17 Morocco_Jews (Behar) 29.68
    18 Azerbaijan_Jews (Behar) 30.86
    19 Saudis (Behar) 31.49
    20 Kurd (Xing) 32.31

    and MDLP K16, which is useful for having both BedouinB and Palestinian:

    83% Palestinian (Israel) + 17% BedouinB (Negev) @ 3.66

  15. The Following 3 Users Say Thank You to IronHorse For This Useful Post:

     Erikl86 (12-12-2018),  Govan (12-12-2018),  Power77 (12-13-2018)

  16. #9
    Gold Member Class
    Posts
    1,563
    Sex
    Ethnicity
    Jewish (Ashkenazi)
    Y-DNA
    Q-YP3924 (Q-M378)
    mtDNA
    K1a1b1a

    Israel
    Quote Originally Posted by IronHorse View Post
    BedouinA is not necesserily Levantine admixed, there are people in Arabia who can be modeled as BedouinA and have no ancestry to Levantine countries, I posted this on the Palestinian thread:
    Yep, I've mentioned it myself in my latest post:

    then there's an obvious cline from the BedouinA (which in the Levant itself are considered the most pristine Arabs) all the way to Yemenites, and the problematic issue that both Levantines and Arabs seem to share substantial Neolithic Levantine ancestry. In fact, modern Arabs seem to be most differentiated from modern Levantines in their much lower Iran_ChL admixture, which is Bronze Age Levantine is roughly 50/50, but in some pristine Arab populations it's as low as only ~10% Iran_ChL.
    Basically, the biggest difference between Levantines and Arabs (other than minor Steppe admixture in the former and elevated SSA admixture in the latter) is the much lower Iran_ChL admixture in Arabs than in Levantines. Some Arab populations seem to have gotten more Iran_ChL and thus are closer to Levantines, despite not having any substantial recent (as in - the last few millennia) Levantine ancestry.

    However, considering the geographical proximity between the Southern Levant and Arabia (really the main geographical barrier separating them is the Arabian desert), one cannot but wonder if there wasn't always some sort of diffusion between the South Levant and more northern Arab populations. For example, I'll be shocked if ancient Nabbatean samples, should they be discovered, won't show much more affinity to BA Levant than modern day Yemenites.

    Using the models I've shown above, it seems like the Northern Levant got almost no Arab admixture, while North Africa seem to have gotten more of the Levant-like Arab admixture (similar to BedouinA) than to the less-Levant-like BedouinB.
    Check out my Hidden Content
    My Y-DNA: Q-M242 -> Q-L232 -> Q-L275 -> Q-M378 -> Q-Y2016 -> Q-L245 -> Q-FGC1904 -> Q-Y2209 -> Q-Y2225 -> Q-Y2197 -> Q-Y2750 -> Q-YP1004 -> Q-YP3924;
    My mtDNA: K1a1b1a;

    My dad's mtDNA: K2a2a;

  17. The Following 4 Users Say Thank You to Erikl86 For This Useful Post:

     IronHorse (12-13-2018),  kikkk (12-12-2018),  Michalis Moriopoulos (12-13-2018),  Power77 (12-13-2018)

  18. #10
    Registered Users
    Posts
    249
    Y-DNA
    I-Y16419
    mtDNA
    T2e1

    How can BedouinB be modeled anyway? or Saudis.

  19. The Following 2 Users Say Thank You to IronHorse For This Useful Post:

     Govan (12-12-2018),  Power77 (12-13-2018)

Page 1 of 7 123 ... LastLast

Similar Threads

  1. Replies: 0
    Last Post: 06-24-2018, 08:50 AM
  2. Replies: 9
    Last Post: 05-30-2017, 02:12 PM
  3. Replies: 6
    Last Post: 05-16-2017, 05:36 PM
  4. Replies: 15
    Last Post: 01-30-2017, 09:51 PM
  5. "Indus Valley" and "Kalash" admixtures in Khvalynsk?
    By Tomenable in forum Ancient (aDNA)
    Replies: 12
    Last Post: 08-12-2016, 10:46 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •