Serbs are Slavic-speaking people living in Balkan Peninsula, southeastern part of Europe. Majority of Serbs live in Serbia, but also in former Yugoslav republics of Bosnia and Herzegovina, Croatia, Montenegro, and to lower extent Northern Macedonia.
Our sample consists of more than two thousands people, all ethnic Serbs, uniformly distributed across Serbia and other Balkan countries with numerous Serb population. Many of them have been tested in mass actions of Y-DNA testing, during several last years. And here are our results.
Haplogroup I alone comprises about 47% of our statistics.
Haplogroup I2a1 M423> L621> CTS10228 (Slavic-Carpathian) is single largest branch, with estimated age at 500 BC at most, and it comprises about 38% of Serbian paternal lineages. It is important to note that it is pretty uniformly distributed among Serbs across Serbia and other ex-Yugoslav countries with numerous Serb population. It is the most numerous haplogroup among Serbs in every region, in no region with less than 30%, and in some regions up to 47%. Majority of Serbian I2a1-CTS10228 belongs to PH908 (previously known as “Dinaric-South”).
Haplogroup I2a > M223 is represented with about 0.7%, which are spread out across different regions. Some haplotypes we have identified belong to subclades L701 and CTS6433. We have identified several other I2 branches in Serb population, but in very small amounts, all together about 0.3%. It are I2b (L415), I-Y16419 (former I2c2), I2a1>M26, I2a1>L38, and even I2a1>L233.
Paternal Haplogroup E is among Serbs represented predominantly by its subclade V13. It is relatively uniformly distributed across Serb population in different regions, with slightly greater frequency in southern regions, like southern Serbian province of Kosovo and Metohiјa. Highest frequency of V13 in Serb population (over 50%) is recorded in eastern part of Montenegro known as Brda (“Hills”), the home of three largest Serb clans that belong to different subclades of V13: Vasojevici (PH1246>BY14151), Kuci (CTS5856>Z5018>Z16661>BY165837) and Bjelopavlici (CTS5856>Z5017>Z16988>BY155589). Another frequent subclade of V13 among Serbs that also originates from single bloodline from Montenegro is CTS5856>Z5017>Z19851>A18833. However, most diverse subclade of V13 in Serb population is CTS5856>Z5018>L241 which is distributed in all regions settled by Serbs, and unlike previously mentioned subclades, it’s represented by large number of not so numerous bloodlines. Other E subclades like V22, M84 and Z841 are represented in small percentages.
Haplogroup R1a is also a major haplogroup among Serbian population, and it’s distribution is relatively homogenous. Serbs mostly belong to subclades M458 (mostly to downstream L1029, but to smaller degree A11460 and L260 also), and Z280 (downstream subclades Y2613, YP343, L1280 > YP611 and YP4278), It is believed that the most of Serbian R1a is brought to Balkans during Slavic expansion about 1500 years ago.
The frequency of haplogroup I1 among Serbs is about 8%. The majority of it belongs to P109 > FGC22045 subclade, while some other subclades downstream to P109 have also been found (e.g. Y13930). Other major subclades include several subclades of Z63 (e.g. Y53660, Y16434, Y6634 and Y7627), Z58 (Y3560) and M227 (A11380).
Unlike haplogroup I2a1 > CTS10228, haplogroup P109 > FGC22045 is not uniformly distributed, what can be seen here in the chart of haplogroup distribution by regions. In eastern Serbia it is almost nonexistent, whereas it is ubiquitous in other regions populated by Serbs, in some subregions even reaching more than 20%.
Haplogroup R1b among Serbs consists of several distant branches. Most of them (40%) belongs to L23>Z2103>BY611>Z2705 branch, specific to Balkan states. Z2705 is formed 2700 years ago at Balkans (TMRCA 1500 years). It is common to all areas populated by Serbs, except Herzegovina and western Montenegro. Some of them belong to downstream branch Y32147, present among other Balkan nations, but Piperi tribe (Montenegro), Serbs from Bosnia and many others form special branches. Western L23>L51>P312 branch (25%) is a second by presence, and most of them are U152, with minor branches Z36>Z67 (Moračani-Bogićevci tribe), L2>Z49 and L2>FGC13619. Germanic L23>L51>U106 is present by 15% of all R1b haplotypes. About 10% belongs to PF7562 branch, separated from L23 some 6400 years ago.
Haplogroup J2b1-M205 > Y22066/Y22059 comprises about 3.6% of Serbian Y-DNA. It is very young, TMRCA of whole of it is about 1000 years. Despite it’s moderate frequency, it is important since it is well distributed across all regions with Serb population, from the border with Bulgaria to the most western area of Serb settlements (Lika), where it reaches more than 10% of Serbian Y-DNA. Beside, in neighboring peoples (Albanians, Croats, Bulgarians) it is less numerous. It is one of two most intriguing haplogroups in Serbs (other is N2 – P189.2).
J2b2 > M241 is represented with 1.25% among Serbs. Two main subbranches among Serbs are Z638 and PH1602, whereas the first is slightly more numerous. J-M241 among Serbs generally could be found in all regions, but is somewhat more represented in Southern Serbia, in southern Serbian province of Kosovo and Metohiјa and among Serbs in Montenegro.
Haplogroup G among Serbs generally reaches about 3-5%, dependable on the region. Most common G branch among Serbs is L497, which is about 2/3 among tested G Serbs on Serbian DNA Project. M406 branch is found in various regions, but also among two brotherhoods in Montenegro, L14 in Morača’s brotherhood of Selaković and S9591 in Mezgaljani brotherhood in Šekular area.
On the other hand, almost 90% of L497 haplotypes among the Serbs belongs to Z1816>L42>Y11074>YSC33>Z35901>Y128028, a branch which is most common among Serbs and which is found among few tested Bulgarians.
Branch under it, Y60799 is at this moment confirmed among few Serbs and it probably represents characteristic Serbian branch. L42 and it’s branches have highest frequency and diversity in the Alps and probably represent Celtic heritage among the Serbs.
Haplogroup J2a, being very old and very diverse, among Serbs is represented by several subclades. Among the most numerous belongs L70, which is concentrated more in southeastern Serbia. Also numerous is subclade J2a-M92>Z8096>S8230>Z38463, which has more western distribution. Its base are two tribes in western Montenegro, Cuce and Pjesivci. Beside them, subclades SK1336 and S18579 could be found among ethnic Serbs.
Haplogroup N2 > P189.2 > FGC28435, with average frequency among Serbs at 2.8%, is one of two the most intriguing. Outside ethnic Serbs and neighboring Moslems (Bosniaks), it is very rare, few haplotypes of it have been found across Europe and Near East. Even in Serbian population it is not uniformly distributed. Seems that it’s home area is in western Montenegro, near border with Herzegovina, from where it spread to Bosna and Western and Central Serbia. In Eastern Serbia it is almost nonexistent. It’s distribution reminds to distribution of Serbian I1 > P109 > FGC22045. Which is not the big surprise, since both stem from the same region of western Montenegro.
Frequency of haplogroup J1 is about 1.7%. Several subclades are present in Serbian population, the most common is PF7263, which has more western distribution, among Serbs in Herzegovina and Bosnia. FGC11 can be found in southeastern Serbia, and province of Kosovo and Metohija. Also, subclades Z1842 and PH77 can be found, in smaller amounts.
Haplogroup T is represented with 0.38% (8/2086), which are scattered across Serbian population. We have not determined yet deeper subclades of any Serbian T sample with certainty, but it seems some of them belong to T-L131.
Haplogroup Q is also represented with 0.38% (8/2086) in our sample. and it’s branches identified so far among Serbia include Q-L713, Q-Y2209 and Q-L330.
Few L haplotypes have also been found. Two of them belong to L1b-M317, and one to very rare subclade L2-L595.
To haplogroup H1a-M82 belong 0.14% (3/2086) haplotypes of ethnic Serbs of our sample. It marks Roma (Gypsy) admixture among ethnic Serbs.
We also have found very few N1c or C haplotypes, whose frequency is probably less than 0.1% in Serbian population.
Here can be seen the chart of distribution of Y-DNA of ethnic Serbs by regions. Serbs are divided into 13 regions, seven in Republic of Serbia, and six in other former Yugoslav republics. Beside Serbs, for comparison purposes, here are also statistics for ethnic Albanians in three regions, Moslems (Bosniaks) of Bosnia & Herzegovina divided into four regions, and haplotypes from Croatia, divided into five regions (every region with 220 samples, source is “Croatian national reference Y-STR haplotype database”, Mršić et al, 2012), plus collection of haplotypes from Croatian capital Zagreb, Purps et al, 2014.
Haplotypes of Moslems from B&H have been collected from various public sources, like public FTDNA Projects, Bosniak DNA Project, Serbian DNA Project and others. Unfortunately, we do not have great number of haplotypes (three of four regions have less than 100), but we hope more will be available in not so far future.
Haplotypes of Albanians from Albania are from “Shared language, diverging genetic histories: high-resolution analysis of Y-chromosome variability in Calabrian and Sicilian Arbereshe”, by Sarno et al, 2016.
Haplotypes of Albanians from North Macedonia came from work by Jankova et al, 2019.
For statistics of haplogroups of Serbs, we used about 4400 haplotypes from three different sources. That are Serbian DNA Project (poreklo.rs), and haplotype collections of Dr. Ivica Todorovic (Ethnographic Institute, Belgrade), and Dr. Dusan Keckarevic (Faculty of Biology, Belgrade). We also need to explain why we used only about 2100 haplotypes for statistics of Serbs, when we have more than twice that number. The reason is that some regions are overrepresented (for example Herzegovina, and southern Serbia – Province of Kosovo and Metohija, due to targeted research of those two serb subpopulations). On the other side, some regions are underrepresented, for example northern Serbia – Province of Vojvodina, and eastern Serbia. Also, in some regions we have tribes, clans or extended families which are also overrepresented in Serbian DNA Project. For both reasons we have had to exclude some haplotypes to make more balanced statistics, relevant and representative for Serbs as a whole.
We owe special thanks to Nenad Zakula (eonsystem.com), who made this work possible by comprehending and implementing the whole database for Serbian DNA Project, together with whole set of very useful functions, like searches, chart display and so on.
We used NevGen Haplogroup Predictor (nevgen.org) for prediction of haplotypes which were not SNP tested. We plan to update statistics by regions in future, from time to time.