heretical ‘neutralist’ theories of both ecologists and palaeontologists. The ecological
theory claims that diversity and phenotypic divergence is largely the result of selective
pressures acting within communities. Indeed, the recognition that selection could be
divergent in its effect is an important scientific discovery of the last century.
When applied to the Cambrian explosion, the results of this view would be to claim
that the numerous important adaptive features that appear in fossils at this time, such as
hard parts, are the result of ecological interactions (cf. Bengtson 2002) that have arisen
through the sorts of population processes that can be seen ongoing today. Why, then, is
the event so problematic? These difficulties can be divided into both methodological
problems and peculiarities about the event itself. It is only in recent years that rigorous
efforts to understand the phylogenetic patterns displayed by Cambrian organisms have
been made (Smith 1988; Briggs and Fortey 1989; Budd 1993, 2002; Conway Morris and
Peel 1995), and before this time it was easy to make claims about the nature of the
radiation that are in fact contrary to analysis (e.g. that few patterns of relationship could
be seen in the Cambrian fauna). These analyses have increasingly shown that the origins of
the phyla that have been studied are of a progressive and adaptive nature, more fitting
with the ‘correlated progression’ view (e.g. Kemp 1982) of macroevolution than the
more fashionable ‘key innovation’ one (e.g. Hunter 1998). Naturally enough, the highly
fragmentary record, and difficulties of understanding the more problematic taxa, have
greatly impeded this effort.
Second, there are problems associated with the event itself. As no one believes today
that the pattern presented by the fossil record is entirely artefactual, and it is generally
accepted that that there was a genuine ecological radiation affecting plants and animals
during this time, then one is entitled to ask what sorts of ecological events were involved.
Given the likelihood that the macrobenthic realm was being exploited by highly mobile
bilaterians for the first time, with no obvious competitors for resources, the initial
expansion of bilaterians is likely to have more clearly approximated to an expansion in
vacuo than to any other major evolutionary expansion. Such a situation is (on such a large
scale) a non-actualistic one, but perhaps could be compared with ecological expansions
after mass extinctions (e.g. Conway Morris 1998b). Even then, the situation is not truly
comparable. Ecosystems in general seem to remain highly fragile after major extinction
events, perhaps partly because of the destruction of the supporting communities that taxa
evolved to live in. With the exception of opportunistic pioneer species, the expansion of
other species from refugia may be impeded until suitable community structures can be re-
established (Erwin 2001). Such considerations may not apply to the earliest stages of
bilaterian expansion and, if not, early bilaterian expansion may have been less impeded by
the lack of pre-existing community structures to which organisms were adapted.
Another even more important factor would be the lack of biogeographical diversity and
invasion resistance. Modern studies are increasingly showing that successful invasion of a
new region by a particular species is in general impeded by high biodiversity in the target
region (e.g. Stachowicz et al. 1999). This resistance would be negligible in the early stages
of the bilaterian expansion. The world that the crown-group bilaterians expanded into
would be one populated by sponges and cnidarians, together with the stem-group
bilaterians (some of which may be presented by the Ediacaran taxa). The essentially sessile
nature of all benthic members of these groups would imply a very different ecological
178 DATING THE ORIGIN OF BILATERIA