Frozen Evolution. Or, that’s not the way it is, Mr. Darwin. A Farewell to Selfish Gene.

If we wish to use the molecular clock for absolute dating of events in cladogenesis, i.e. for dating in physical time units, for example, in years, and not only for relative dating, i.e. for determining that a certain event occurred three times as long ago as another event, then it must be possible to calibrate the molecular clock for a particular gene according to absolute time.We must rely on paleontological data when calibrating the molecular clock.However, these data are necessarily accompanied by a number of errors.Less serious errors follow from the fact that the dating of the individual paleontological data need not be completely exact.However, a more serious inadequacy of paleontological data consists in the fact that the oldest known occurrence of certain fossils in the paleontological record allows us only to determine that the relevant species and thus the relevant taxon already existed in a certain period.However, it does not allow us to determine when this species or taxon appeared on the Earth or when it definitively became extinct (see XXII.3.1).Experience in discovery of living fossils (e.g., latimeria) simultaneously indicates that the duration of the period of cryptic existence of a species, i.e. the length of the period during which the given species existed without leaving any visible traces in the paleontological record, can be very long.However, for calibration of the molecular clock, we need to know when a certain taxon originated.Even if very good quality, complete paleontological material were available for the particular period that also contained the first representatives of the relevant taxon, we still cannot be certain that we would actually recognize it as a representative of that taxon.Anagenesis, i.e. the formation of the characteristic morphological traits of the taxon, need not temporally conform with cladogenesis.Following the splitting off of two sister branches, each of which finally leads to the formation of a different taxon, the representatives of the two branches can retain the morphological traits of the original, parent taxon for a long time.

            Generally, very few reliably dated cladogenetic events are available.For example, calibration of the molecular clock used in study of the evolution of metazoans is based on the paleontological estimate of the moment of splitting off of the line leading to birds and to mammals.Some studies are also based on estimation of splitting off of the line leading to primates and to rodents.However, splitting off of primates and rodents was dated using a molecular clock calibrated on the basis of splitting off of the ancestors of birds and mammals.This means that this is not an independent calibration and the accuracy of most molecular dating of cladogenesis is thus basically dependent on the accuracy of the paleontological determination of the splitting off of birds and mammals (Lee 1999).

            Calibration of the molecular clock should be based exclusively on the method of interpolation – i.e. dating of the event should be located within an interval delimited by calibration points.However, this is not possible in practice because sufficiently old calibration points are not available for dating quite fundamental evolutionary events, such as the splitting off of multicellular animals, plants and fungi.In these cases, we must rely on the extrapolation method, i.e. dating of an event using several younger calibration points and assume (or hope) that the rate of molecular evolution inside the interval delimited by the calibration points and outside this interval are not very different.However, this need not be true, so that all the data obtained by the extrapolation method must be accepted with great care.Thus, if analysis performed on 50 genes indicates that fungi, animals and plants diverged approximately 1578 million years ago, with an estimated error of 88 million years, we must be aware that these estimates are based on the assumption that the substitution rate did not vary over this entire period.