degree of bioturbation in the Proterozoic and earliest Cambrian, both in terms of intensity
and depth, arguably resulted in sediment properties that were particularly conducive to
preservation of very shallow tiers (Droser et al. 2002).
Nevertheless, it could be argued that, although advantageous, the trace fossil record
does suffer from some less favourable secular variation. For example, it might be argued
that the transition from the microbial mat-dominated sediment surfaces of the late
Proterozoic to more bioturbated soft sediment bottoms of the Phanerozoic (Seilacher
1999) might bias the record in favour of Phanerozoic traces. While it has been argued that
mats may have been important in structuring early benthic communities (Seilacher 1999;
Bottjer et al. 2000), these mats can hardly be seen as leathery, impenetrable objects: many
lifestyles of macro-organisms are thought to have involved constant disruption and
utilization of mats. This would include ‘mat-stickers’ penetrating the mats, and ‘under-
mat miners’ burrowing just beneath the mats (Seilacher 1999). If so, the presence of the
mats might actually enhance trace fossil preservation (Budd and Jensen 2000); although it
should be again stressed that many Precambrian bedding planes show no signs of having
once been covered by biomats (e.g. they have rippled surfaces).
Two candidates for extremely old trace fossils have recently been described. The first
(Seilacher et al. 1998), from India, was reported as being some 1 Ga, a date that at the time
of description coincided well with molecular estimates of the divergence of the Bilateria,
and is now considered to date from more than 1.6 Ga (Rasmussen et al. 2002a; Ray et al.
2002). This description has not, it is fair to say, been widely accepted. The sediments
from which it was described are full of sedimentary structures such as mud cracks and, as
detailed in Budd and Jensen (2000), these traces show some suspicious features such as an
irregular, crinkly appearance and tapering terminations; all of which are suggestive of
inorganic origin. The other candidate is from the Stirling Range of South Australia
(Rasmussen et al. 2002b). These millimetric structures are from rocks that are dated at
between 1.2 and 2 Ga. They consist of parallel-sided ridges preserved in sandstones,
sometimes showing a distinct narrowing towards a rounded termination. The authors
interpret these structures, with some hesitation, as the surficial traces made by a probable
soft-bodied worm, citing an earlier study (Collins et al. 2000) as evidence that secreted
mucus can act as a glue to stick displaced sediment grains together. The problem with this
interpretation, apart from any doubts about the preserved morphology (e.g. the
narrowing of the terminations would be highly unusual in a trail), is primarily a taphonomic
one. By their nature, these traces would have been made on a sandy surface, and whether
or not binding mucus was produced, preservation of this level in the fossil record must be
considered to be doubtful in general. Indeed, a rigourist school of thought would argue
that all locomotion trace fossils preserved in the marine fossil record are essentially
undertracks—impressions made on an interface within the sediment, not on its surface
(Seilacher 1957). On these grounds alone, the interpretation of these structures as made
by a bilaterally symmetrical worm moving across the surface of the sediment must be
strongly questioned. Even more problematic is how the traces as found are preserved.
The authors describe them as being preserved in sharp positive hyporelief on the base of
sandstones (i.e. as positive protrusions from the underside of the sandstone). This is a very
common mode of preservation of trace fossils, where an overlying sediment type casts
impressions in an underlying one. However, the structures under discussion are meant to
GRAHAM E.BUDD AND SÖREN JENSEN 173