Fedorovo culture of Andronovo - Ancestors of Indo-Aryans?

[Andar]

I would like to open a thread about the Fedorovo culture and it possible role in the history of early Indo-Aryans. Pottery of this subgroup of Andronovo was also found in Shortugai a IVC colony in North Afghanistan around 1800 B.C and unlike the Alakul subgroup of Andronovo it practised (partial) cremation which later was typical also for Indo-Aryans. Fedorovo as Andronovo group of the IAMC region was also likely rich in IAMC-ancestry that we later see in Indo-Aryans, Swat_IA and possibly even some Maryannu-related samples. Some Fedorovo samples were published in the last years and it would be interesting to look closer at them.

[Andar]

Y-DNA of Fedorovo samples


I10112 Tomsk_4379 Steppe_Central_MLBA Fedorovo Shoendykol, Russia   -1547 BC
R1a>Z93>Z94>Z2124>Z2122>Y38987 https://www.yfull.com/tree/R-Y38987/ 

I10111 Tomsk_4377 Steppe_Central_MLBA Fedorovo Shoendykol, Russia  -1446 BC
R1a>Z93>Z94>Z2124>Z2125>Z2123>Y934>Y15121>Y210252 https://www.yfull.com/tree/R-Y210252/ 

C1714 L6548 Xinjiang_Tangbalesayi_LBA_Andronovo_Fedorovo   -1493.5 BC
R1a>Z93>Z94>Z2124>Z2125>Z2123>Y15121>FTB5268>Y95238 https://www.yfull.com/tree/R-Y95238/

C4282 Xinjiang_Wutulan_LBA_Andronovo, Russia   -1390.5 BC
R1a>Z93>FGC82884>FGC45861>FT113000>Y188077  https://www.yfull.com/tree/R-Y188077*/

I3389 StPet56, collection 3390, individual 4 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I3394 StPet62, collection 3864, individual 2 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I3392 StPet60, collection 3390, individual 7 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408> https://www.yfull.com/tree/R-FGC56408/

I3393 StPet61, collection 3864, individual 1 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408> https://www.yfull.com/tree/R-FGC56408/

I3396 StPet65, collection 3864, individual 5 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408> https://www.yfull.com/tree/R-FGC56408/

I3391 StPet59, collection 3390, individual 6 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408>YP1456 https://www.yfull.com/tree/R-YP1456*/

I1851 SS7800-24, Ust-Bir IV, grave 28 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1513 BC R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I6718 StPet64, collection 3864, individual 4 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1543 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408> https://www.yfull.com/tree/R-FGC56408/
 
I4790 KZ-DJA-004, Steppe_Central_MLBA Oy_Dzhaylau_MLBA Russia -1483 BC R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>Z2123>Y934>BY30762>Y136333> https://www.yfull.com/tree/R-Y136333/

I4265 KZ-AKT-003, Andronovo Western_Steppe_MLBA Russia -1572 BC
R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I4773 KZ-AKT-002, Andronovo Western_Steppe_MLBA Russia -1558
R1a>YP4141>YP4132 https://www.yfull.com/tree/R-YP4132/

I4774 KZ-AKT-008, Andronovo Western_Steppe_MLBA Russia -1553 BC
R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I10140 KZ-AKT-006b Andronovo Western_Steppe_MLBA Russia -1300
R1a>M459>M735>M198>M417>Z645>Z93>

[PopGenist82]

That’s a nice map; where’s it from ?

[Andar]

That’s a nice map; where’s it from ?

From the first part of a catalog of a german exhibition about archaeology of Kasachstan (p.286-287). The first part can be found online and is mainly about Andronovo culture. The second part is about Iron Age nomads and so on. 

http://archeo-astana.kz/monograph/bochum...cation.pdf

[G-Man]

What about Krasnoyarsk_MLBA?
 

Orak burial place, Russia (n=11)
The site is located near Orak Ulus village in the Krasnoyarsk region of Russia. Most burials of the site are attributed to the Fedorovo type of the Andronovo complex while some are attributed to the Karasuk culture

Narasimhan et al 2019

 
All males are Z2125>S23592

[Andar]

What about Krasnoyarsk_MLBA?
 

Orak burial place, Russia (n=11)
The site is located near Orak Ulus village in the Krasnoyarsk region of Russia. Most burials of the site are attributed to the Fedorovo type of the Andronovo complex while some are attributed to the Karasuk culture

Narasimhan et al 2019

 
All males are Z2125>S23592

Thanks for mentioning. These samples were not labeled as Fedorovo in my spreadsheet so overlooked them. I updated my list. The very curious thing is that R1a-S23592 was very common in (northern) Fedorovo and this is the main line of later Iron Age Iranic Nomads of Siberia and Central Eurasian Steppes.  R1a-S23592 was also found in Alakul linked sites of Central Kazakhstan in the same period but this could be a Fedorovo substrate. 

Still southern and earlier Fedorovo groups seem to be unsampled for now and they likely will show more R1a-Y3 and high R1a-Z2123. Here we also see relict R1a-Y3 and R1a-L657 lines in the Iron Age which likely come from a Pre-Saka Indo-Iranian substrate.

[Andar]

Fedorovo unlike early Iranic-related Alakul practiced partial cremation and had similar burial rituals to early Indo-Aryans



Fedorovo pottery in Shortugai, North Afghanistan around 1800 B.C

[Andar]

Y-DNA of Fedorovo samples


I10112 Tomsk_4379 Steppe_Central_MLBA Fedorovo Shoendykol, Russia   -1547 BC
R1a>Z93>Z94>Z2124>Z2122>Y38987 https://www.yfull.com/tree/R-Y38987/ 

I10111 Tomsk_4377 Steppe_Central_MLBA Fedorovo Shoendykol, Russia  -1446 BC
R1a>Z93>Z94>Z2124>Z2125>Z2123>Y934>Y15121>Y210252 https://www.yfull.com/tree/R-Y210252/ 

C1714 L6548 Xinjiang_Tangbalesayi_LBA_Andronovo_Fedorovo   -1493.5 BC
R1a>Z93>Z94>Z2124>Z2125>Z2123>Y15121>FTB5268>Y95238 https://www.yfull.com/tree/R-Y95238/

C4282 Xinjiang_Wutulan_LBA_Andronovo, Russia   -1390.5 BC
R1a>Z93>FGC82884>FGC45861>FT113000>Y188077  https://www.yfull.com/tree/R-Y188077*/

I3389 StPet56, collection 3390, individual 4 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I3394 StPet62, collection 3864, individual 2 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I3392 StPet60, collection 3390, individual 7 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408> https://www.yfull.com/tree/R-FGC56408/

I3393 StPet61, collection 3864, individual 1 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408> https://www.yfull.com/tree/R-FGC56408/

I3396 StPet65, collection 3864, individual 5 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408> https://www.yfull.com/tree/R-FGC56408/

I3391 StPet59, collection 3390, individual 6 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1650 BC R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408>YP1456 https://www.yfull.com/tree/R-YP1456*/

I1851 SS7800-24, Ust-Bir IV, grave 28 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1513 BC R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I6718 StPet64, collection 3864, individual 4 Steppe_Central_MLBA Krasnoyarsk_MLBA Russia -1543 BC >R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592>FGC56408> https://www.yfull.com/tree/R-FGC56408/
 
I4790 KZ-DJA-004, Steppe_Central_MLBA Oy_Dzhaylau_MLBA Russia -1483 BC R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>Z2123>Y934>BY30762>Y136333> https://www.yfull.com/tree/R-Y136333/

I4265 KZ-AKT-003, Andronovo Western_Steppe_MLBA Russia -1572 BC
R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I4773 KZ-AKT-002, Andronovo Western_Steppe_MLBA Russia -1558
R1a>YP4141>YP4132 https://www.yfull.com/tree/R-YP4132/

I4774 KZ-AKT-008, Andronovo Western_Steppe_MLBA Russia -1553 BC
R1a>M459>M735>M198>M417>Z645>Z93>Z94>Z2124>Z2125>S23592 https://www.yfull.com/tree/R-S23592/

I10140 KZ-AKT-006b Andronovo Western_Steppe_MLBA Russia -1300
R1a>M459>M735>M198>M417>Z645>Z93>

Looking at these samples there is actually no significant Y-DNA match with Indo-Aryans and especially R1a-S23592  (TMRCA 2400 B.C) was later associated with Iron Age Iranics. Instead we have R1a-Y3 (TMRCA 2500 B.C) in Iranic-related Srubnaya-Alakul (70% of Indo-Aryan Y-DNA). But R1a-Y3 and S23592  are clades which predate Proto-Indo-Iranian and itself cant be always linked to a specific Indo-Iranian subgroup in 1500-2200 B.C. Only via later founder effects they became clearly and strongly associated with a specific Indo-Iranian subgroup (R1a-Y3 for Indo-Aryans and S23592 for Iranic Saka).  

Indo-Aryan Y-DNA went through multiple founder effects/bottle necks and the high frequency of R1a-Y3<L657 (TMRCA 2100 B.C) is likely related to a founder effect among IAMC Indo-Iranians and later Proto-Vedic Indo-Aryans of Afghanistan + Indus region. Also Fedorovo considering the wide distribution and relatively long existence of that group would rather represent a chain of (late Proto)-Indo-Iranian groups than a single homogenous Indo-Iranian entity. So we likely have early Indo-Aryan+Nuristani related Indo-Iranians in southern Fedorovo (unsampled Pamir-Tian Shan sites) whereas more northern groups in Kazakhstan and Siberia would be part of some basal Indo-Iranian groups that later became Iranic speaking (R- S23592)

[pegasus]

There is intra variation amongst these groups which needs to be ironed out, archaeologically Fedorovo matches up but the uniparental markers do not match up at all and current Fedorovo and Andronovo samples are no where as good as those Srubnaya Alakul samples which do match up wrt to uniparental markers and on qpAdm they consistently are picked, but archaeologically Alakul does not match up. 

Wrt to Shortugai, it reverts to a more localized Bactrian variant from period II , by period III and IV its very similar to Farkhor, on the opposite side of the Panj river. There is way more ICW ceramics in the Murghab region for whatever reason. 

The steppe (Andronovo type) pottery appear
during Period II (1 fragment) and becomes
more significant, though in small quantities
during Periods III and IV (Francfort et al ,
1989. P. 80. Pl. 58, 11–15; XXVI, 5).

The only pertinent sample from the past 2 years to help resolve their origins , is probably one of the outlier wives from that Alakul grave , who can be modelled as a mix between Steppe MLBA and some IAMC population. My guess she is a very early Indo Aryan woman.

sample: Nepljevski_o2
distance: 1.7797
Sintashta_MLBA: 57
Aigyrzhal_BA: 34
Tarim_EMBA1: 9

She works well with the SPGT Steppe outlier as well

Target: Pakistan_Loebanr_IA_o:I12138
Distance: 1.9457% / 0.01945674 | R4P
51.6 Nepljevski_o2
31.0 Turkmenistan_Gonur_BA_1
17.4 Iran_ShahrISokhta_BA2

[G-Man]

C1714 L6548 Xinjiang_Tangbalesayi_LBA_Andronovo_Fedorovo   -1493.5 BC
R1a>Z93>Z94>Z2124>Z2125>Z2123>Y15121>FTB5268>Y95238 https://www.yfull.com/tree/R-Y95238/

It's unlikely that he has Y95238

15963098 4A 1T
 
YFull and FTDNA placed him at R-Z2123 level

[Andar]

There is intra variation amongst these groups which needs to be ironed out, archaeologically Fedorovo matches up but the uniparental markers do not match up at all and current Fedorovo and Andronovo samples are no where as good as those Srubnaya Alakul samples which do match up wrt to uniparental markers and on qpAdm they consistently are picked, but archaeologically Alakul does not match up. 

Wrt to Shortugai, it reverts to a more localized Bactrian variant from period II , by period III and IV its very similar to Farkhor, on the opposite side of the Panj river. There is way more ICW ceramics in the Murghab region for whatever reason. 

The steppe (Andronovo type) pottery appear
during Period II (1 fragment) and becomes
more significant, though in small quantities
during Periods III and IV (Francfort et al ,
1989. P. 80. Pl. 58, 11–15; XXVI, 5).

The only pertinent sample from the past 2 years to help resolve their origins , is probably one of the outlier wives from that Alakul grave , who can be modelled as a mix between Steppe MLBA and some IAMC population. My guess she is a very early Indo Aryan woman.

sample: Nepljevski_o2
distance: 1.7797
Sintashta_MLBA: 57
Aigyrzhal_BA: 34
Tarim_EMBA1: 9

She works well with the SPGT Steppe outlier as well

Target: Pakistan_Loebanr_IA_o:I12138
Distance: 1.9457% / 0.01945674 | R4P
51.6 Nepljevski_o2
31.0 Turkmenistan_Gonur_BA_1
17.4 Iran_ShahrISokhta_BA2

The current Fedorovo samples are all from places north of Kyrgyzstan and these places way or another would not work as source because the IAMC ancestry would be too northern shifted (even north of KGZ_BA). From what i can see Ror/Jatt and these Megiddo outliners would need a slightly more SC Eneolithic shifted IAMC source than KGZ_BA.  Srubnaya has lot of R1a-Y3+ but R1a-Y3+ itself dont mean because it predates Proto-Indo-Iranians and is technically generic Indo-Iranian. It is older for example than R1a-Z2123 which has a Pan-Indo-Iranian distribution from Kurdistan to South India. Most important is to get R1a-L657 or R-Y27 which is younger .But even R1a-L657 could theoretically be Pan-Indo-Iranian (Yfull underestimates). I mean R1a-Z2123 has the same age for example. But in the case of R1a-L657 it really is lacking anywhere outside IAMC and Indo-Aryans so we have after 2000 B.C a very strong likely specific Proto-Indo-Aryan association.

We have in the Iron Age R-Y27 (KGZ_IA) and R1a-L657 (Illi Valley IA) and both samples dont seem to have recent ancestry from some place south of Tajikistan so it very likely was common in the broader region.

Reading this book (The World of the Oxus civilization: https://www.routledge.com/The-World-of-t...1032570037) gives also interesting hints. Unlike Kuzmina they dont link Bishkent-Vaksh to Andronovo (different pottery, earlier dating than Andronovo, 


Interesting about  Bishkent-Vaksh is that they were seemingly pastoralists with links to Zaman-Baba (Yamnaya-Catacomb similarities) and it pottery was found near Fedorovo-Andronovo pottery in Shortugai. Kuzmina also lacks pottery in Swat to this culture so maybe they are not Indo-Iranian but they are the source of the IAMC ancestry in Indo-Aryans and also Swat_IA (possibly migrated parallel to Indo-Aryans south?)

I also recently just realized that a Petrovka colony was around 1800-2000 B.C present in the Zerafshan valley. It also had Abashevo characteristics which is pretty interesting. So around the period we see Indo-Iranians expanding with the Seima-Turbino network far in the north and east some Indo-Iranians already moved very deep south and opened some distant colonies.

[Andar]

C1714 L6548 Xinjiang_Tangbalesayi_LBA_Andronovo_Fedorovo   -1493.5 BC
R1a>Z93>Z94>Z2124>Z2125>Z2123>Y15121>FTB5268>Y95238 https://www.yfull.com/tree/R-Y95238/

It's unlikely that he has Y95238

15963098 4A 1T
 
YFull and FTDNA placed him at R-Z2123 level

Thanks i will try to correct it. I looked for the Yfull assigment at theYtree which sadly often is wrong if coverage is low

[Desdonas]

There is intra variation amongst these groups which needs to be ironed out, archaeologically Fedorovo matches up but the uniparental markers do not match up at all and current Fedorovo and Andronovo samples are no where as good as those Srubnaya Alakul samples which do match up wrt to uniparental markers and on qpAdm they consistently are picked, but archaeologically Alakul does not match up. 

Wrt to Shortugai, it reverts to a more localized Bactrian variant from period II , by period III and IV its very similar to Farkhor, on the opposite side of the Panj river. There is way more ICW ceramics in the Murghab region for whatever reason. 

The steppe (Andronovo type) pottery appear
during Period II (1 fragment) and becomes
more significant, though in small quantities
during Periods III and IV (Francfort et al ,
1989. P. 80. Pl. 58, 11–15; XXVI, 5).

The only pertinent sample from the past 2 years to help resolve their origins , is probably one of the outlier wives from that Alakul grave , who can be modelled as a mix between Steppe MLBA and some IAMC population. My guess she is a very early Indo Aryan woman.

sample: Nepljevski_o2
distance: 1.7797
Sintashta_MLBA: 57
Aigyrzhal_BA: 34
Tarim_EMBA1: 9

She works well with the SPGT Steppe outlier as well

Target: Pakistan_Loebanr_IA_o:I12138
Distance: 1.9457% / 0.01945674 | R4P
51.6 Nepljevski_o2
31.0 Turkmenistan_Gonur_BA_1
17.4 Iran_ShahrISokhta_BA2

The current Fedorovo samples are all from places north of Kyrgyzstan and these places way or another would not work as source because the IAMC ancestry would be too northern shifted (even north of KGZ_BA). From what i can see Ror/Jatt and these Megiddo outliners would need a slightly more SC Eneolithic shifted IAMC source than KGZ_BA.  Srubnaya has lot of R1a-Y3+ but R1a-Y3+ itself dont mean because it predates Proto-Indo-Iranians and is technically generic Indo-Iranian. It is older for example than R1a-Z2123 which has a Pan-Indo-Iranian distribution from Kurdistan to South India. Most important is to get R1a-L657 or R-Y27 which is younger .But even R1a-L657 could theoretically be Pan-Indo-Iranian (Yfull underestimates). I mean R1a-Z2123 has the same age for example. But in the case of R1a-L657 it really is lacking anywhere outside IAMC and Indo-Aryans so we have after 2000 B.C a very strong likely specific Proto-Indo-Aryan association.

We have in the Iron Age R-Y27 (KGZ_IA) and R1a-L657 (Illi Valley IA) and both samples dont seem to have recent ancestry from some place south of Tajikistan so it very likely was common in the broader region.

Reading this book (The World of the Oxus civilization: https://www.routledge.com/The-World-of-t...1032570037) gives also interesting hints. Unlike Kuzmina they dont link Bishkent-Vaksh to Andronovo (different pottery, earlier dating than Andronovo, 


Interesting about  Bishkent-Vaksh is that they were seemingly pastoralists with links to Zaman-Baba (Yamnaya-Catacomb similarities) and it pottery was found near Fedorovo-Andronovo pottery in Shortugai. Kuzmina also lacks pottery in Swat to this culture so maybe they are not Indo-Iranian but they are the source of the IAMC ancestry in Indo-Aryans and also Swat_IA (possibly migrated parallel to Indo-Aryans south?)

I also recently just realized that a Petrovka colony was around 1800-2000 B.C present in the Zerafshan valley. It also had Abashevo characteristics which is pretty interesting. So around the period we see Indo-Iranians expanding with the Seima-Turbino network far in the north and east some Indo-Iranians already moved very deep south and opened some distant colonies.

OK. But I still have some doubt about the genetic characteristic of Indo-Aryans when they entered South Asia, especially considering the ANF ratio in South Asian populations.

Last year, Davidski mentioned on Eurogenes Blog that most Indians only have a few per cent of ANF ancestry. Also, there is minor excess ANF in India over the expected MLBA-derived amount. But Matt and Vasistha metioned that there may already be some ANF ancestry in Indus periphery. For example, Maier has found 7.7 ± 2.9% Anatolian ancestry in IP group. In all, if the total ANF amount from MLBA and IP is enough to explain the Indian ANF ratio, then Indo-Aryans may mixed with WSHG-rich groups rather than any BMAC-related groups?

[pegasus]

There is intra variation amongst these groups which needs to be ironed out, archaeologically Fedorovo matches up but the uniparental markers do not match up at all and current Fedorovo and Andronovo samples are no where as good as those Srubnaya Alakul samples which do match up wrt to uniparental markers and on qpAdm they consistently are picked, but archaeologically Alakul does not match up. 

Wrt to Shortugai, it reverts to a more localized Bactrian variant from period II , by period III and IV its very similar to Farkhor, on the opposite side of the Panj river. There is way more ICW ceramics in the Murghab region for whatever reason. 

The steppe (Andronovo type) pottery appear
during Period II (1 fragment) and becomes
more significant, though in small quantities
during Periods III and IV (Francfort et al ,
1989. P. 80. Pl. 58, 11–15; XXVI, 5).

The only pertinent sample from the past 2 years to help resolve their origins , is probably one of the outlier wives from that Alakul grave , who can be modelled as a mix between Steppe MLBA and some IAMC population. My guess she is a very early Indo Aryan woman.

sample: Nepljevski_o2
distance: 1.7797
Sintashta_MLBA: 57
Aigyrzhal_BA: 34
Tarim_EMBA1: 9

She works well with the SPGT Steppe outlier as well

Target: Pakistan_Loebanr_IA_o:I12138
Distance: 1.9457% / 0.01945674 | R4P
51.6 Nepljevski_o2
31.0 Turkmenistan_Gonur_BA_1
17.4 Iran_ShahrISokhta_BA2

The current Fedorovo samples are all from places north of Kyrgyzstan and these places way or another would not work as source because the IAMC ancestry would be too northern shifted (even north of KGZ_BA). From what i can see Ror/Jatt and these Megiddo outliners would need a slightly more SC Eneolithic shifted IAMC source than KGZ_BA.  Srubnaya has lot of R1a-Y3+ but R1a-Y3+ itself dont mean because it predates Proto-Indo-Iranians and is technically generic Indo-Iranian. It is older for example than R1a-Z2123 which has a Pan-Indo-Iranian distribution from Kurdistan to South India. Most important is to get R1a-L657 or R-Y27 which is younger .But even R1a-L657 could theoretically be Pan-Indo-Iranian (Yfull underestimates). I mean R1a-Z2123 has the same age for example. But in the case of R1a-L657 it really is lacking anywhere outside IAMC and Indo-Aryans so we have after 2000 B.C a very strong likely specific Proto-Indo-Aryan association.

We have in the Iron Age R-Y27 (KGZ_IA) and R1a-L657 (Illi Valley IA) and both samples dont seem to have recent ancestry from some place south of Tajikistan so it very likely was common in the broader region.

Reading this book (The World of the Oxus civilization: https://www.routledge.com/The-World-of-t...1032570037) gives also interesting hints. Unlike Kuzmina they dont link Bishkent-Vaksh to Andronovo (different pottery, earlier dating than Andronovo, 


Interesting about  Bishkent-Vaksh is that they were seemingly pastoralists with links to Zaman-Baba (Yamnaya-Catacomb similarities) and it pottery was found near Fedorovo-Andronovo pottery in Shortugai. Kuzmina also lacks pottery in Swat to this culture so maybe they are not Indo-Iranian but they are the source of the IAMC ancestry in Indo-Aryans and also Swat_IA (possibly migrated parallel to Indo-Aryans south?)

I also recently just realized that a Petrovka colony was around 1800-2000 B.C present in the Zerafshan valley. It also had Abashevo characteristics which is pretty interesting. So around the period we see Indo-Iranians expanding with the Seima-Turbino network far in the north and east some Indo-Iranians already moved very deep south and opened some distant colonies.

OK. But I still have some doubt about the genetic characteristic of Indo-Aryans when they entered South Asia, especially considering the ANF ratio in South Asian populations.

Last year, Davidski mentioned on Eurogenes Blog that most Indians only have a few per cent of ANF ancestry. Also, there is minor excess ANF in India over the expected MLBA-derived amount. But Matt and Vasistha metioned that there may already be some ANF ancestry in Indus periphery. For example, Maier has found 7.7 ± 2.9% Anatolian ancestry in IP group. In all, if the total ANF amount from MLBA and IP is enough to explain the Indian ANF ratio, then Indo-Aryans may mixed with WSHG-rich groups rather than any BMAC-related groups?

You cannot apply one yard stick across these populations. I have quite a few South Indians samples that have low to nil Steppe but ample ANF and even Levantine related ancestry , compared to Gangetic Indians, who compared to them are depressed for Zagrosian ancestry but have comparable AASI levels.

Also  almost all of the IVCp samples are way too West Eursian shifted to be proxies for the vast majority of Indians. We know IA South Indians are very similar to modern ones with no Steppe ancestry  and some of them are on the Harappan cline, so it make sense using them as a proxy. 

My friends and I have gone through hundreds of models over the years for different populations,  and zeroing on Rors/Haryanvi Jaats they need a lot CHG and WSHG compared to any Indian  group   I have seen

This is their qpAdm model with near perfect tail
https://imgur.com/a/XxZH2ax

the IVC proxy used is actually the GBR groups on the IVCp cline

its the same on G25, I trust G25 more with proximal populations but the proportions are identical. They have substantial Central Asian admixture but it seems much better captured by Bustan_o1-Aigryzhal populations

sample: Ror:Median
distance: 0.9022
Telugu_GBR: 39
Srubnaya_Alakul_MLBA: 35
Bustan_BA_o1: 26


sample: Ror:Median
distance: 1.0129
Telugu_GBR: 39
Srubnaya_Alakul_MLBA: 34
Aigyrzhal_BA: 14.5
Sappali_Tepe_BA: 12.5

Obviously we will need actual genomes but given those samples I posted above and robust models for modern populations, the Vedic IArs definitely admixed with Zerevshan populations rich in Western/Central Siberian ancestry.