2017 Behar et al. Paper on R1a-Y2619 Ashkenazi Levites

Dr. Doron Behar’s paper on R1a1a Ashkenazi Levites, D. Behar et al., The genetic variation in the R1a clade among the Ashkenazi Levites’ Y chromosome, SREP-17-37687 (2017), published in Scientific Reports on November 2, 2017, confirms and expands upon the findings of the 2013 Rootsi & Behar paper on R1a1a Ashkenazi Levites, S. Rootsi & D. Behar, et al., Phylogenetic Applications of Whole Y-Chromosome Sequences and the Near Eastern Origin of Ashkenazi Levites, Nature Communications 4, Article No. 2928 (2013). (Elsewhere on this website, R1a1a Ashkenazi Levites are referred to as R1a-Y2619 Ashkenazi Levites; for historical reasons, this page of the website will use the term R1a1a Ashkenazi Levites as well as the term R1a-Y2619 Ashkenazi Levites to refer to men in that Y-DNA cluster.)

The paper’s findings are also consistent with the findings of the citizen genetic genealogy community concerning R1a1a Ashkenazi Levites, including those of the FTDNA R1a Project, YFull, and this website. (Although I am a co-author of Dr. Behar’s paper, some of the thoughts expressed on this page are my own.)

The paper’s principal findings, based upon the results of whole Y-DNA next generation sequencing testing for 486 men, including 71 men belonging to the R1a1a Ashkenazi Levite cluster, are as follows (some of these findings are implicit in charts and tables, but are not necessarily explicitly discussed in the body of the paper):

The R1a-Y2619 Ashkenazi Levite Cluster

1. All known members of the R1a1a Ashkenazi Levite cluster share the SNP Y2619 (and five other Big Y-tested SNPs at the Y2619 level – Y2740, Y2624, Y2625, Y2627, and Y2745). (The 2013 Rootsi & Behar paper had determined that all R1a1a Ashkenazi Levites share the SNP M582, one level upstream from the Y2619 level.) In keeping with the now-prevalent practice of identifying Y-DNA clusters by haplogroup followed by terminal SNP rather than by ISOGG’s system using alternating letter-and-number combinations, the R1a1a Ashkenazi Levite cluster can now be referred to as the R1a-Y2619 Ashkenazi Levite cluster. (Note that the tables to the Behar paper categorize R1a1a Ashkenazi Levites as belonging to the Y-DNA haplogroup R1a2, in keeping with some recent papers that have designated the Asian R1a-Z93 branch to which R1a-Y2619 Ashkenazi Levites belong as R1a2, distinguishing the branch from the European R1a1 branch.)

2. The men in the R1a-Y2619 Ashkenazi Levite cluster are all descended from a single man (sometimes referred to on this website, but not in the paper, as the R1a-Y2619 progenitor) who lived ~1,743 (1,334 – 2,200) years before present (ybp). (This time frame is several hundred years more distant than the figure that various researchers had previously estimated based upon STR test results. It is possible that the STR-based calculations, which are based upon the total number of STR deviations within a cluster, may have been skewed downwards because the R1a-Y2619 cluster includes a large subcluster with a more recent shared direct male ancestor.)

3. Based upon a current Ashkenazi population of about 4,000,000 males (about 7.9% of whom are R1a-M582 according to the 2013 Rootsi & Behar paper, and therefore are R1a-Y2619 based on the findings of the new paper), as of today there would be about 300,000 Ashkenazi men who are descended on their direct male lines from the R1a-Y2619 progenitor who lived ~1,743 years ago.

4. The R1a-Y2619 progenitor was very likely not only Jewish but also a Levite. (We know this because men on all R1a-Y2619 lines have a tradition of both Jewish and Levite descent, suggesting that their shared ancestor, the R1a-Y2619 progenitor, was not only Jewish but also a Levite.)

5. The R1a-Y2619 Ashkenazi Levite line is, as had previously been believed based upon the current distribution of R1a-Y2619 men, of Ashkenazi origins. Although six of the tested Jewish R1a-Y2619 men had most distant known ancestry in countries that are associated with non-Ashkenazi communities (one R1a-Y2619 subcluster includes two men with direct male ancestry from Turkey, while another R1a-Y2619 subcluster includes a man with direct male ancestry from Algeria and a man with direct male ancestry from Iraq, and there are two men with direct male ancestry from Israel in other subclusters), each of these subclusters also includes men of known Ashkenazi ancestry. This indicates that each subcluster – and, thus, the R1a-Y2619 cluster as a whole – is Ashkenazi in origin, with some migration within each subcluster to non-Ashkenazi regions.

6. The Y-DNA evidence is consistent with the paper genealogies of three known descendants of the Horowitz rabbinical family who trace their ancestry to a member of the Horowitz rabbinical family born in 1615 CE; the Y-DNA evidence also indicates that two other men with the ancestral surname Horowitz but without a paper trail to the Horowitz rabbinical family share a direct male ancestor with the known descendants of the Horowitz rabbinical family who lived ~691 (555 - 852) ybp (but that a sixth R1a-Y2619 man surnamed Horowitz is on a different sub-branch of R1a-Y2619). The members of the Horowitz rabbinical family and their closest matches belong to the R1a-YP268 subcluster of R1a-Y2619. (As previously noted on this website, the R1a-Y2619 men who are descended from the Horowitz rabbinical family – and most men within the R1a-YP268 subcluster to which descendants of the Horowitz rabbinical family belong – are characterized by the STR marker value DYS495=16 (rather than the R1a-Y2619 modal value of DYS495=17).)

Middle Eastern Origins of the R1a-Y2619 Ashkenazi Levite Cluster

1. With regard to the deep origins of R1a-Y2619 Ashkenazi Levites, the paper concludes that “The proposed Middle Eastern origin of the Ashkenazi Levite lineage based on what was previously a relatively limited number of reported samples, can now be considered firmly validated.” The R1a-Y2619 lineage “was likely a minor haplogroup among the Hebrews.”

2. The R1a-M582 cluster, to which the R1a-Y2619 Ashkenazi Levite cluster belongs, has origins in the Middle East as of about 3,000 years ago. The new paper identifies four R1a-M582 subclusters, all of which share a direct male ancestor who lived about 3,000 years ago: (1) the R1a-Y2619 subcluster; (2) a 1,200-year-old subcluster currently including an Iranian Kerman man, an Iranian Azeri man, and a Yazidi man; (3) a 2,700-year-old subcluster currently including an Iranian Azeri man and an Iberian man; and (4) a line, diverging from the other lines about 3,000 years ago, currently including an Iranian Kerman man. The fact that each of the three clusters other than the R1a-Y2619 cluster includes men with geographic origins in Iran suggests strongly that the R1a-M582 progenitor lived in that area. (The 2013 Rootsi & Behar paper had identified Near Eastern origins for R1a1a Ashkenazi Levites based upon the presence of M582 in various R1a1a populations, “with the highest frequency occurring within Iranians collected from the southeastern Kerman population who self-identified as Persians, northwestern Iranian Azeri, and in Cilician Anatolian Kurds.”)

3. The R1a-M582 cluster falls within the R1a-F1345 cluster, which shares a direct male ancestor who lived about 4,600 years ago. The paper identifies five R1a-F1345 subclusters: (1) the R1a-M582 subcluster discussed in the prior paragraph; (2) a 900-year-old subcluster including two Israeli Arab men; (3) a 3,500 year-old subcluster including two North Balkarian men and a North Turkish man; (4) a 2,700-year-old subcluster including an Iranian Azeri man and a Spanish man; and (5) a 3,000-year-old subcluster including a Scottish man and a Polish man.

4. The R1a-Y2619 Ashkenazi Levite cluster does not have origins in Khazaria. Not only did that line have origins in the Middle East as of 3,000 years ago, but the line was Jewish (and Levite) as of 1,750 years ago, more than 500 years before the Khazars (or, in some accounts, Khazarian royalty) are said to have converted to Judaism. (The 2013 Rootsi & Behar paper similarly concluded that the R1a1a Ashkenazi Levite line does not have Khazarian origins. However, the 2003 Behar paper, D. Behar, et al., Multiple Origins of Ashkenazi Levites: Y Chromosome Evidence for Both Near Eastern and European Ancestries, Am. J. Hum. Gen. 73:768-799 (2003), had concluded, based upon the limited genetic analyses available at that time indicating that R1a1a Ashkenazi Levites belonged to the R1a-M198 haplogroup that was prevalent in Eastern Europe, that the R1a1a Ashkenazi Levite cluster might have originated in Eastern Europe. Subsequent studies showed that R1a-M198 is comprised of several clusters, and that the R1a-Z93 cluster to which R1a1a Ashkenazi Levites belong is Asian in origin. Thus, the new paper states that “While the highest frequencies of haplogroup R1a are found in Eastern Europe, our data revealed a rich variation of haplogroup R1a outside of Europe which is phylogenetically separate from the typically European R1a branches.”)

5. “[T]he exact migration pathway of R1a-Y2619 to Europe remains elusive.” Most historical records indicate that there were two major routes of Jewish migration to Europe: (1) Jewish migration via Italy to the Rhine Valley, the migratory pattern of Ashkenazi Jewry; and (2) Jewish migration along with Islamic expansion through North Africa and then to Spain, the migratory pattern of Sephardic Jewry. “Naturally, the strong founding event for R1a-Y2619 among Ashkenazi Jews, coupled with the presence of all known branches of R1a-Y2619 in Ashkenazi Jews, tempts to infer that its migration route from the Levant was directly related to the Ashkenazi founders,” but it is possible that R1a-Y2619 arrived with Sephardic Jews and expanded in the Ashkenazi population. In any event, the R1a-Y2619 line very likely arrived in Europe through a single expansion route, with the expansion of the population occurring after migration to Europe.

6. The currently available data does not allow a determination of how long before the time of the R1a-Y2619 progenitor that his line assumed Levite status. Absent Y-DNA test results for any men whose lines branched off in the 1,400 years between the time of the R1a-Y2619 progenitor (~1,743 years ago) and the time of the R1a-M582 progenitor (~3,143 years ago), this issue may remain unresolved.

The Expansion of the R1a-Y2619 Ashkenazi Levite Cluster in the Context of the General Expansion of the Ashkenazi Population

1. The paper also analyzed the major Y-DNA chromosome haplogroups found among Ashkenazi Jews to determine their origin and expansion times in comparison to those of R1a-Y2619 Ashkenazi Jews. The paper concluded that the expansion of the R1a-Y2619 Ashkenazi Levite cluster was consistent with the expansion of the general Ashkenazi population, as evidenced by Y-DNA.

2. The paper addressed branching in the following Ashkenazi haplogroups: (1) haplogroup J (38% of the Ashkenazi paternal gene pool), with a focus on the Cohen lineage nested within haplogroup J1-P58 but also considering Cohen haplogroups within haplogroups J2-M12 and J-M318; (2) haplogroup E (20.4% of the Ashkenazi paternal gene pool), with a focus on lineages within haplogroups E-Z838, E-PF3780, E-B923, and E-B933; (3) haplogroup G-M537 (9.7% of the Ashkenazi paternal gene pool), refined to haplogroup G-BY764; (4) haplogroup Q-M378 (5.2% of the Ashkenazi paternal gene pool); (5) haplogroup T-M70; and (6) haplogroup R1b-M269.

3. The coalescence age and pattern of expansion of R1a-Y2619 Ashkenazi Levites are similar to those found in certain other major Ashkenazi Y-DNA haplogroups – G-M377 (coalescing ~1,223 ybp), Q-M242 (coalescing ~1,672 ybp), and E-Z838, E-PF3780, E-B923, and E-B933 (coalescing ~1,200 – 1,600 ybp). “This pattern of multiple founding events, not observed among Spanish Jews provides further support that the R1a-Y2619 Ashkenazi Levite ancestor entered Europe through the Ashkenazi route rather than via the Jewish expansion to the Iberian Peninsula.”

4. The J1a-P58 Cohen cluster coalesces ~2,570 years ago, is nested within a Middle Eastern set of samples, and includes both Ashkenazi Cohens and non-Ashkenazi Cohens. Accordingly, the J1a-P58 Cohen cluster, unlike the R1a-Y2619 Ashkenazi Levite cluster, began to expand in the Levant, in pre-Diaspora times.

Additional Comments and Observations

I have the following further comments and observations on the new Behar paper:

1. As discussed in the paper, the R1a-Y2619 Ashkenazi Levite cluster appears to have been bottlenecked from about 3,143 years ago to about 1,743 years ago, i.e., the tested men included no man whose line split off from the R1a-Y2619 line after the split from R1a-M582. Without any man splitting the R1a-Y2619 level, we have no Y-DNA-based ground for making a determination as to when, during that 1,400-year period, the R1a-Y2619 line moved from the Middle East to Europe or entered the Jewish and/or Levite population; put otherwise, based on the Y-DNA evidence taken in isolation, the line could have first become Jewish and/or Levite at any time between 3,143 years ago and about 1,743 years ago.

2. According to a 2014 paper referred to in the Behar paper, S. Carmi et al., Sequencing an Ashkenazi reference panel supports population-targeted personal genomics and illuminates Jewish and European origins: (1) there was a severe autosomal bottleneck in the Ashkenazi population about 700 years ago, as a result of which today’s Ashkenazi population includes autosomal DNA inherited from only 350 Ashkenazi Jews who lived 700 years ago; and (2) the Ashkenazi population expanded massively out of that autosomal bottleneck beginning about 700 years ago. The figures to the Behar paper demonstrate that the autosomally bottlenecked population included multiple R1a-Y2619 lines, and that, after the autosomal bottleneck, those lines experienced the same sort of massive expansion that was experienced in the Ashkenazi population in general. (Note that, as discussed in the paper, the Y-DNA bottlenecks need not have coincided precisely with the autosomal bottleneck, and different Y-DNA clusters had bottlenecks at different times.)

There are three well-defined main branches within the R1a-Y2619 Ashkenazi Levite cluster. The haplogroup R1a phylogeny in the Behar paper includes coalescence dates for those branches and its sub-branches; this page includes an annotated copy of that phylogeny to try to tie the SNPs to those coalescence dates.

The first branch to split off from the main R1a-Y2619 branch was the R1a-FGC18222 branch, splitting off ~1,743 ybp and coalescing ~1,454 ybp. Within R1a-FGC18222 are two sister branches, the R1a-FGC18226 branch, coalescing ~1,246 ybp, and the R1a-YP1074 branch, coalescing ~1,101 ybp. Within the R1a-FGC18226 branch is the R1a-FGC18218 branch, coalescing ~1,166 ybp, and within R1a-FGC18218 is the R1a-FGC18215 branch, coalescing ~694 ybp. Within R1a-YP1074 are two sister branches, the R1a-YP1366 branch, coalescing ~725 ypb, and the R1a-YP4690 branch, coalescing ~635 ybp.

The second branch to split off from the main R1a-Y2619 branch was the R1a-Y2630 branch, splitting off ~1,481 ybp and coalescing ~1,227 years ago. Within R1a-Y2630 are three sister branches, the R1a-YP1386 branch, coalescing ~849 ybp, the R1a-YP264 branch, coalescing ~912 ybp, and an R1a-Y2630* branch that includes numerous sub-branches, most of which seem to have coalesced about 600 to 800 ybp.

The third branch, left behind when the R1a-Y2630 branch split off ~1,481 ybp, was the R1a-Y2619* branch, coalescing ~1,039 ybp. [As of 2017, based upon a single set of Big Y results and about 10 sets of STR marker values, I believed that there was an R1a-Y2619* subcluster, characterized by the STR marker value DYS576=16 – three steps from the R1a-Y2619 modal value of DYS576=19 – that likely split off from the R1a-Y2619 branch at or before the time that the R1a-Y2630 branch split off from the R1a-Y2619 branch. Since that time, additional Big Y testing has shown that: (1) this subcluster is part of a larger R1a-BY29826 cluster, a third major subcluster of R1a-Y2619 Ashkenazi Levites; (2) there is a fourth smaller subcluster, R1a-BY73299; and (3) there are a handful of men who are R1a-Y2619*, i.e., they are R1a-Y2619 Ashkenazi Levites who do not belong to the R1a-Y2630, R1a-FGC18222, R1a-BY29826, or R1a-BY73299 subclusters of R1a-Y2619 Ashkenazi Levites.]

Thus, it appears that there may have been as many as 10 to 15 R1a-Y2619 lines at the time that the Ashkenazi population, according to the autosomal evidence, expanded out of its bottleneck perhaps 700 years ago.

3. Many of the R1a-Y2619 Ashkenazi Levites whose results were analyzed in Dr. Behar’s paper are members of the group associated with this website. For reference by those men and their kit managers, this webpage attaches the following annotated versions of two attachments to Dr. Behar’s paper: (1) a pared-down version of Supplemental Figure 1 that includes (a) the HG R1a spreadsheet, annotated to provide FTDNA kit numbers, surnames, and ancestral names for certain tested men and (b) the portion of the mutation list that includes SNPs for R1a-Y2619 Ashkenazi Levites, adding the names of some SNPs (indicated with yellow highlighting); and (2) four pages from Figure S1, the haplogroup R1a phylogeny, annotated with the kit numbers and surnames of the tested men and SNP names corresponding to coalescence ages for the major clusters and subclusters. A larger version of the annotated spreadsheet is posted toward the bottom of this page.