“…However, for our divergence date estimation, we modified their 'method 2' by replacing 'I ij ' (=fraction of shared nucleotides between two sequences i and j) by '1 À (mean p-distance between lineages i and j)' and accordingly, we used average nucleotide frequencies of lineage i, and j, respectively, to infer the asymptotic identity for each of the two lineages. For divergence dating in the 16S gene, we applied 'clock rates' of 0.53%/Ma (Stillman and Reeb, 2001; calibrated by the divergence of porcelain crabs, genus Petrolisthes, at 3 Ma BP), of 0.02-0.04%/Ma, 0.06-0.12%/Ma (inferred from coral genera and families, respectively, see Romano and Palumbi, 1997) and of 0.05-0.2 and 0.07%/Ma (clocks calibrated by the divergence between Drosophila species, and between Aedes and Drosophila, respectively, as presented in Romano and Palumbi, 1997, referring to data from DeSalle et al, 1987).…”