been exhaustively examined for embryophyte remains. It can be argued that the latter is
also unlikely, based on the assumption that in order to reproduce and survive successfully,
the earliest subaerial embryophytes would have produced abundant, small, sporopollenin-
coated spores that were easily dispersed over large distances. One would expect such plant
populations to spread rapidly as founder populations were established by widely dispersed
spores (bear in mind these spores are believed to have been able to self-fertilize; Gray,
1985). Consequently, it is anticipated that the earliest embryophytes would have rapidly
colonized new areas soon after their inception and quickly become widespread. Hence we
might expect the appearance of their spores in the fossil record soon after they had first
evolved.
In conclusion, I consider that the Llanvirn benchmark for the origin of embryophytes is
the most reliable provided to date. This is based primarily on the following assumptions: (i)
plants could not have invaded the land until they had evolved spores with a resistant
sporopollenin wall that were produced in vast numbers and easily transported large
distances; (ii) due to the large transportation potential of their spores, palaeogeographical
spread of the earliest land plants would have been rapid; and (iii) the spores would have
left a rich fossil record, which should have been identified owing to the intensity of
palynological work undertaken on Precambrian and Lower Palaeozoic deposits.
Environmental evidenceIt is generally accepted that the invasion of the land by plants would have had a profound
effect on the environment of planet Earth. Scientists have identified a number of potential
ways in which such global change could be picked up in the fossil/ stratigraphical record
(summarized in Retallack 2000), and these signals serve as a proxy for the appearance of
significant terrestrial vegetation. Most of these methods rely on the assumption that
increased weathering, associated with the development of a substantial vegetation, would
fertilize the oceans with nutrients, and hence drive up isotopic values such as ∂^13 C, ∂^18 O
and^87 Sr/^86 Sr. There are, however, a number of potential pitfalls associated with these
methods. These essentially relate to the fact that a number of independent factors can
cause shifts in isotopic values, and these can be difficult to disentangle. The nature of^87 Sr/
(^86) Sr ratios provides a good example of such problems.
It has been suggested that the^87 Sr/^86 Sr ratio in marine carbonate rocks acts as a proxy
for continental silicate weathering (e.g. Edmond 1992). This model assumes that^87 Sr is
mainly derived from the weathering of silicate minerals in continental rocks, and that it
accumulates in marine carbonate rocks that are the major sink for strontium. Because it is
assumed that continental weathering will have increased rapidly following the
development of a substantial terrestrial flora, it follows that this event may be recorded in
the^87 Sr/^86 Sr record. In fact^87 Sr/^86 Sr values begin to rise quite significantly from a low in
the Middle Cambrian, and it has been suggested that this might be related to the
appearance of a significant flora comprising land plants (summarized in Retallack 2000).
However, a number of problems have been identified that render interpretation of
continental silicate weathering based on^87 Sr/^86 Sr ratios dubious (e.g. Blum 1997;
Edmond and Huh 1997; Quade et al. 1997; Blum et al. 1998; Broecker and Sanyal 1998;
Boucot and Gray 2001). Chief among these is the possibility that the strontium content of
DATING THE ORIGIN OF LAND PLANTS 137