Thanks for your reply, as you see in Table 1, these haplogroups have the highest frequency in Iran:
1. R1a-M198
2. R1b-L23
3. J2a-M530
4. J1c3
Is it true that J1 and J2 relate to Semitic, Elamite, Hurrian, ... people but R1a and R1b relate to Indo-Europeans? Who were these Indo-European people? Were they Indo-Iranians? Or Celtic people?
Haplogroup J1 and J2 are interesting haplogroups.
The oldest finding of J1 comes from the Satsurblia Cave in Georgia, which is estimated to date to 13,200 BCE. Certain J1 branches are associated with Semitic groups, notably J-P58 (J1c3 in the Iranian Y-chromosome study you cited, the SNP at that time was referred to as PAGE08). Right now the theory for J-P58 appears to to be that it expanded from the more southerly portions of the Levant across the Arabian peninsula in the Bronze Age, some specific clades of J-P58 such as J-Z18297 and J-ZS227 (this one includes the haplotype most commonly believed to be Cohanim), there is another SNP known as FGC12, which seems to be linked to the expansion of Arabic populations in medieval period. There are some haplogroups which are suspected to be linked with Phoenicians, however we don't quite have enough evidence to solidify these links. Now a haplogroup can't exactly be Semitic, Celtic, Slavic, etc. it can merely be associated with these various ethnolinguistic groups. J1 haplogroups do show up in Europe in lower frequencies.
J2 is a much more seemingly widespread haplogroup with an estimated origin somewhere in the Middle East sometime 15,000-22,000 years ago. The oldest sample of J2 belongs to J2a in remains in the Hotu Cave in Iran, this sample dates to 9100-8600 BC (Lazaridis et al reported this in 2016) and another sample found Georgia in Kotias Klde dates from 7940-7600 BC (Broushaki et al reported this in 2016). J2 has an obviously strong presence in Western Asia, however it is also found in Europe and is the most commonly associated lineage with the spread of cereal farming from the Fertile Crescent through to Anatolia and eventually Europe. However this lineage isn't the only likely farmer lineage. Other farmer lineages include G2a, E1b1b and T1a.
However, surprisingly, in 2015, Mathieson et al tested 13 Early Neolithic farmers, specifically their Y-DNA in NW Anatolia these samples were dated to 6500-6200 BC) and only 1 sample belonged to J2a. Lazaridis in 2016 tested 44 ancient NE samples only 2 belonged to J2, same goes for 100 samples tested in Neolithic European sites and yet again only two J2 samples were found. J2 is seemingly absent from Chalcolithic and Bronze Age Indo-European cultures. Interestingly J2 samples have been found if I recall correctly, Minoan sites, which is quite fascinating.
Are you referring to R1b and R1a in Western Asia and specifically Iran? Haplogroups of R1b such as R-PF7562 (estimated to have formed in the Early Bronze Age) are found in the Balkans, Turkey and Armenia, a branch of R-L23 (EBA in Pontic Steppe), specifically R-Z2103 (Bronze Age) is found in Eastern Europe and West Asia. We have to remember that R1b is a HUGE haplogroup, infact we would more accurately refer to it as a paragroup. We have to realize that R1b split into many different haplogroups, one very early branch of R1b, R-V88 is found more commonly in Africa. We can only attach certain haplogroups and their dispersion to PIEans, others are explained by different independent lineages that parted ways with their relatives who would later go on to the PIE homeland.
The map for R1b from the Paleolithic to the Bronze Age may be of interest to you as it places M343 somewhat north in the Iranian region, which then a descending branch of P25 heads further west where it splits into P297 and V88, the former heading north and the latter heading south through the Levant. The branch that heads north splits into M269 (which heads to the Pontic Caspian steppe) and M73 heads east. M269 eventually migrates toward the Balkans and a descending clade of L23 fans out in multiple directions, with a migration into Anatolia (this branch is labelled R-Z2103), obviously some very old lineages of R1b that may be very private (meaning not widespread) may exist in Iran or the Iranian diaspora.
View attachment 11088
The same thing can be said of the R1a paragroup, it has a WIDE dispersion and it has many branches to its tree, again showing that some branches are of different geographic origins. I recommend you take a look at the R1a haplogroup page here on Eupedia's mainpage, and more specifically the Indo-Iranian section of the page. It explains that the Proto-Indo-Iranian speakers seemed to have originated in the Sintashta-Petrovka culture (from 2100-1750 BC) east of the Ural Mountains, with an ancient sample of R-Z93 being found quite east in the range of the most likely PIE homeland it lines up with this proposed theory. The lineages which appear in Iranian locations are extremely old, with M343 estimated to have formed 22,800 years before present with a TMRCA (diversification) of 20,400 years before present.
An interesting read that mentions some ancient R1b Y-DNA from the Iranian Neolithic as well as basal R-M343* finds in Kazakhstan and Iran, all of which support a southern migration route through Iran
https://indo-european.info/ie/R1b-M343_and_West_Hunter-Gatherers
https://www.eupedia.com/europe/Haplogroup_R1a_Y-DNA.shtml#Indo-Iranian, there is even an interesting entry for both the Tarim mummies and Turkic speakers and their connection to R1a.
The people who brought R-Z93 (and associated, notably the descending branch of Z94) lineages to Iran would obviously be the Proto-Indo-Iranian people in the case of Iran, other IE groups made their way elsewhere. This is why the map on Eupedia's R1a page for the Neolithic to Bronze Age migration is the "most accurate" we have (at least on this site) right now because it takes into account the ancient samples and the timing of the formation of the branches. Obviously some branches of R1a could have trickled in to Iran through entirely different routes. We need ancient DNA from Iran to better build up the information and help improve things. Now, referring to the R1a map again, obviously it isn't a hardline map, obviously various lineage groups can go off in their own direction, so groups in the Balkans could have jumped across to Anatolia, Caucasus populations with R1a could have had minor migrations into the Middle East and elsewhere. Certain branches of R1b (excluding the old basal ones) may have hitched a ride with the Iranian speakers as they made their way into Iran (which on the R1b map would look like they took a U-turn.)
In regards to these maps, they are not set in stone, they are subject to change and they will change with time as more data comes in. As Ygorcs says aDNA is proving to be a bit more solid than Y-DNA, obviously so as Y-DNA is uniparental and all it takes is one merchant, soldier of fortune, refugee, etc for a lineage to end up somewhere and expand. So far the aDNA that supports the more current accepted theories.
The Celtic end of the Indo-European tree is more likely nestled in central Europe in terms of its origin, this is seen in archaeological record of Bronze Age Europe. Obviously there was a later migration of Celts into Anatolia after some of these Gaulish tribes had been exhausted from warring with Greek kingdoms.