The estimates are based on partial consensus Y-mutation rate of 0.82×10^-9 mutations/nucleotide/year for a defined Y-sequence area (comBED, 857 regions, 8.4 mbp) using SNP counting (See Adamov et al. 2015, 0.82 also in Poznik et al. 2013, 0.8 in Fu et al. 2014, 0.87 in Helgason et al. 2015). The calibration by Adamov et al. was based mainly on Anzick-1 (Q1a2a1b-Z780, 12632 years BP) with verification trough 41 samples representing 14 genealogies (0.80), Ust-Ishim (0.78, K-M2335, 45000 years BP), Kostenki-K14 (max. 1.08, C1-K29, 37000 years BP). The rate seems consistent for haplogroups ca. 13,000 years old, for genealogical time-frame slightly higher mutation rates might be informative (0.87, 753 males, 274 patrilines, 15.2 Mb Y-sequence, 1,365 meioses, 47,123 years).
The J2 haplogroup evolved in the Middle East, likely in the area from the Glacial Arabo-Persian Gulf Oasis to the Zagros Mountains slopes. It separated from the J1 clade ca. 32,000 ybp and diversified in J2a and J2b ca. 28,000 ybp.
Range of age estimates for Y-chromosome and mtDNA haplogroups connected to estimates for expansion of modern Proto-Eurasians (non-African). See for example (BP = before present):
- YFull Experimental YTree v3.6: CT-M168 formed 86,300 years, TMRCA 68,300 years
- Robb 2012: CDEF-M168 54,000-66,000 years, J 21,000-28,000 years,
- Wei et al. 2013: CDEF-M168 57,000-74,000 years,
- Francalacci et al. 2013: F-R haplogroup ca. 110,000 years
- Iñaki Comas et al. 2013: MTBC age ca. 70,000 years
- S. Gopalakrishnan et al. (unpublished): Eurasian expansion ca. 75,000 years
- Rieux et al. 2014: coalescence of mtDNA L3 72,000 years [54-93 95% HPD]
For a calibration point set at CDEF-M168 (see phylotree.org/Y) with 70 ka (kiloannus = thousend years) ago according to Karafet et al. 2009 the matching Y-Mutation rate is ca. 0.8×10^-9 mutations/nucleotide/year.
If you like to follow/join discussion on J2 haplogroup ages, look at this MolGen Forum thread: age of main J-M304 nodes (J1, J2, J2a, J2b) haplogroup.