Reliable estimates of phylogenetic relationships and divergence times
are a crucial requirement for many evolutionary studies, but are usual
ly difficult because fossils are scarce and their interpretation is of
ten uncertain. Frogs are fresh water animals that generally are unable
to cross salt water barriers (their skin is readily permeable to prov
ide an independent measure of the minimum date of genetic divergence b
etween pairs of such populations. For the genetically well-studied wes
tern Palearctic water frogs (Rana esculenta group), the Aegean region
provides an ideal area for determining the relationship between geneti
c divergence and time of spatial isolation, using a nested set of geol
ogically determined isolation times (12,000 yr, 200,000 yr, 1.8 Myr, 2
-3 Myr, and 5.2 Myr). Using 31 electrophoretic loci for 33 pairs of ne
ighboring frog populations, a linear relationship between geologically
determined isolation time and Hillis' modified Nei genetic distance w
as found: D-Nei = (0.04 +/- 0.01) + (0.10 +/- 0.01) isolation time [M
yr] corresponding to an average divergence rate (''molecular clock'' p
ace) of 0.10 D-Nei/Myr (0.10 D-Nei/Myr). This rate is in the range of
previous estimates reported for protein electrophoretic data; the val
ue is conservative because relatively few of the loci used are ''fast
evolvers'' (13%; sAAT, ALE, EST-5, MPI). Removing these fast evolvers
from the analysis results in 0.08 D-Nei/Myr (0.08 D-Nei/Myr). The con
fidence limits for estimation of the divergence time given the genetic
distance are large, but unusually narrow for this kind of study; they
permit us to estimate divergence times during the Pliocene and Miocen
e. Few previous studies, including sequence analyses, have provided re
asonable estimates of divergence time for the Pliocene. A test using t
he outgroup taxa Rana perezi and Rana saharica (also isolated for 5.2
Myr by the Strait of Gibraltar) fits the calibration well: observed ge
netic Nei distance D-Nei 0.55, expected D-Nei* 0.56. The calculated d
ivergence times, based on this absolute molecular clock, suggest a ser
ies of speciation events after the Messinian (5.2 Myr), possibly trigg
ered by the rapid ecological changes accompanying the desiccation and
refilling of the Mediterranean Basin.