We also assessed the TMRCA for the rare Sardinia-specific R-M18 haplogroup, and obtained a value of 8.1 (7.8–10.5) KYA.
http://plosone.org/article/info:doi/10.1371/journal.pone.0001430#article1.body1.sec2.p5
I did notice, however, what appears to be a genotyping error in haplogroup R-M18. One of the Cagliari haplotypes appears to have been classified as R-M18 when it should have been classified as R-M269.
This haplotype does not match any of the other seven haplotypes at five of the eight reported STRs, yet matches the majority allele for R-M269 at seven of the eight STRs. In a UPGMA phylogeny constructed using the haplotypes from this paper, it also clustered with R-M269 in all of the trees.
Including this R-M269 haplotype in the R-M18 TMRCA calculation dramatically increases the STR variance, thus producing an errantly large TMRCA calculation.