Telling the Evolutionary Time: Molecular Clocks and the Fossil Record

The problem

It is of course anthropocentric bias, but the nature of the evolutionary and environmental
events surrounding the origin and early evolution of the phylum Chordata are some of the
most extensively researched problems in evolutionary biology. Theories that have sought
to account for these events are contingent upon shifts in calibration of the geological
timescale, a fossil record that is dynamic both in terms of new discoveries and
reinterpretation of the phylogenetic affinities of old finds and, more recently, the
introduction of molecular clock estimates for the times of divergence of living clades. It is
therefore not surprising that many such hypotheses have fallen purely because events once
thought to be coeval are revealed not to be so. But with so many lines of evidence, many
of which are independent, there remains the possibility that conflict may give way to
consilience, rather than merely to compromise. Recent advances have resulted in a
considerable fleshing out of the early fossil record of chordates (Sansom et al. 2001;
M.P.Smith et al. 2001, 2002), the geological timescale is now more finely calibrated than
at any time in the past (e.g. Remane 2000), and there is an ever increasing database of
molecular sequences for analysing evolutionary relationships and sampling for molecular
clock analyses. With these developments, understanding the events surrounding the origin
and early evolution of the chordate phylum may now prove more tractable than at any time
previously.

The data

The nearest living relatives of the chordates are the echinoderms and hemichordates and,
together, these three phyla comprise the Deuterostomia. Living invertebrate chordates
are a very depauperate group in comparison with their vertebrate relatives, comprising
two or three groups depending upon how the Vertebrata are defined. The most
plesiomorphic groups are the tunicates and cephalochordates, and although there has
historically been a great deal of prevarication surrounding their interrelationships, the
tunicates are now widely recognized as the most basal group of living chordates. The next
most inclusive clade, Craniata, includes only the hagfishes in addition to the vertebrates,
which are in turn composed of the lampreys plus the Gnathostomata (living jawed
vertebrates). Gnathostomes, in turn, are composed of chondrichthyans, and the two most
derived groups, the actinopterygians and sarcopterygians, which includes the lineage
leading to tetrapods.
These taxonomic groups are defined solely on the basis of living taxa, and so it is
possible to provide molecular estimates for the divergence of the various groups without
recourse to the fossil record for anything other than internal and/or external calibration.
However, the divergence of the various lineages does not equate to the origin of the
taxonomic groups, at least not in the sense that most biologists understand these taxa.
This is because most of these groups also include fossil taxa, with varying degrees of
taxonomic diversity and disparity, which are part of the lineage leading to the crown-
group of living taxa, but do not possess the full suite of anatomical characteristics
necessary for inclusion within the crown-group (Jefferies 1979). For instance, the extinct
osteostracans are a group of jawless vertebrates that share a number of derived characters
with gnathostomes that they do not share with lampreys. Hence, osteostracans are

PHILIP C.J.DONOGHUE ET AL. 195