Fossil calibration of molecular clocks and the divergence times of geminate species pairs separated by the Isthmus of Panama

Calibration of nucleotide sequence divergence rates provides an important method by which to test many hypotheses of evolution. In the absence of an adequate fossil record, geological events, rather than the first appearances of sister taxa in the geological record, are often used to calibrate molecular clocks. The formation of the Isthmus of Panama, which isolated the tropical western Atlantic and eastern Pacific oceans, is one such event that is frequently used to infer rates of nucleotide sequence divergence. Isthmian calibrations assume that morphologically similar "geminate" species living now on either side of the isthmus were isolated geographically by the latest stages of seaway closure 3.1-3.5 MYA. Here, I have applied calibration dates from the fossil record to cytochrome c oxidase-1 (CO1) and nuclear histone-3 (H3) divergences among six pairs of geminates in the Arcidae to test this hypothesis. Analysis of CO1 first and third positions yield geminate divergences that predate final seaway closure, and on the basis of CO1 first positions, times for all six geminates are significantly greater than 3.5 Myr. H3 sequences produce much more recent geminate divergences, some that are younger than 3.1 Myr. But H3-derived estimates for all arcid geminates are not significantly different from both 0 and 15 Myr. According to CO1, one of the two most divergent pairs, Arca mutabilis and A. imbricata, split more than 30 MYA. This date is compatible with the fossil record, which indicates that these species were morphologically distinct at least 16-21 MYA. Across all CO1 nucleotide sites, divergence rates for arcids are slower than the rates reported for other taxa on the basis of isthmian calibrations, with the exception of rates determined from the least divergent species pair in larger surveys of multiple transisthmian pairs. Rate differences between arcids and some taxa may be real, but these data suggest that divergence rates can be greatly overestimated when dates corresponding to final closure of the Central American Seaway are used to calibrate the molecular clocks of marine organisms.