examination of post-extinction intervals in order to distinguish between the various
possibilities that could account for the temporary loss of many taxa within them.
Phylogeny as a test of stratigraphyWhilst much discussion has recently been generated by the possibility of stratigraphic
evidence being employed in phylogenetic reconstruction (e.g. Wills 2001), in this
instance,the opposite direction of confirmation may also be employed. If bilaterian origins
were genuinely many tens or hundreds of millions of years before the base of the
Cambrian, then the time of appearance of particular clades in the Cambrian record should
be a matter of chance, and not reflect phylogeny. Alternatively, if the order of appearance
does reflect known or postulated phylogeny, then this agreement is strong —indeed,
almost overwhelming—evidence that the true time of origin of the clades in question was
close to the appearance in the record. Naturally enough, even on the most optimistic
reading of the fossil record, there will be a gap between the true appearance of a clade,
and its entry in the fossil record. If these gaps are in general small, then one should expect
that they will not perturb the order of appearance in the record, because the time gaps
between successive nodes of a phylogeny will be large relative to the lags between
appearance and true origin (Figure 9.1A). Conversely, if the lag times are large relative to
the temporal gaps between nodes, then one would not expect the true order of
appearance to be reflected in the fossil record (Figure 9.1B). The degree of
correspondence between stratigraphic order of appearance and phylogeny must therefore,
in principle, place important constraints on the average lag times, and should be
considered to be a critical test of the hypothesis that true divergence times are greatly
underestimated by the fossil record (for the development of a model essentially similar to
this test, albeit one applied to phylogenetic and not molecular clock studies, see
Huelsenbeck 1994). As with so many issues, therefore, resolution of the molecular clock/
fossil record discrepancy could revolve around obtaining and applying the results from
accurate phylogenies. In this particular case, it is important to exclude the timing of fossil
origins from phylogenetic reconstruction, because of the unacceptable circularity this
could introduce.
This sort of test has largely been ignored, at least for the Cambrian explosion, partly
because of the common assumption that all phyla appear more or less at once; that is,
there is no temporal resolution of order of appearance. However, this assumption is
manifestly untrue. Both trace fossils and body fossils appear progressively through the
latest terminal Proterozoic and Early-Middle Cambrian (Budd and Jensen 2000; Budd in
press; Jensen in press). The question is whether or not the perceived order is compatible
with a reasonable phylogeny or not. Although a satisfactory answer to this question is not
yet forthcoming, both because of inadequacies in Cambrian systematics and dating, in
broad terms it does seem compatible. This can be seen in two ways: first, by the broad
succession of the faunas, and second, by the relative appearance of stem- and crown-group
representatives of living clades.
GRAHAM E.BUDD AND SÖREN JENSEN 181