Page 89 - Introduction to Paleobiology and The Fossil Record

76 INTRODUCTION TO PALEOBIOLOGY AND THE FOSSIL RECORD
numbers of localities remain constant, or even
rise, suggests that the fossil signal is robust
(Wignall & Benton 1999; Benton et al.
2004).
This debate between the preservation bias
and common cause hypotheses only refl ects
the fossils in the rocks, the fossil record as it
is recorded. But paleontologists are concerned
about a deeper question: do the fossils in the
rocks reflect the reality of the past?
Sampling and reality
What are paleontologists doing when they
sample the fossil record? Can they build up
better and better knowledge of the history of
life, or are they simply improving their sam-
pling of a faulty and incomplete record? In a
1994 study, Benton and Storrs showed that
sampling is improving through time. Using a
clade–stratigraphic metric (see Box 3.3), they
compared how paleontological knowledge
changed between 1967 and 1993, and they
found an apparent improvement of 5% in the
26 years (Fig. 3.12). At least, the congruence
between the fossil record as understood in
25
20
1993 data
15
Frequency
10
1967 data
5
0 (meaning 3) because the fossils we have to
–20 0 20 40 60 80 100
RCI value
Figure 3.12 Paleontological knowledge has
improved by about 5% in the 26-year period
between 1967 and 1993. According to 1993
data there is 5% less gap, as assessed by a
relative completeness index (RCI), implied in the
fossil record of tetrapods than in 1967. This
figure was obtained by comparing the order of
branching points in cladograms with the order
of appearance of fossils in the rocks. Will there
be a further 5% shift to the right (i.e. towards
100% completeness) by the year 2019? (Based
on Benton & Storrs 1994.)
1993 was 5% better than 1967 when plotted
against a set of static cladograms. New fossils
were filling the gaps (i.e. reducing the ghost
ranges), rather than adding new gaps (i.e.
increasing the ghost ranges). One conclusion
could be that everything would be known by
about the year 2019, but then there is proba-
bly a “law of diminishing returns”, that ghost
ranges will never entirely disappear, and new
finds will remove ghost ranges less and less
frequently. There is a whole study of ghost
ranges, and their markers, the so-called
Lazarus taxa (Box 3.4).
All these studies are looking at our knowl-
edge of the fossil record. There are three
meanings of the term fossil record:
1 Our current knowledge of the fossils in
the rocks (the usual meaning).
2 Our ultimate knowledge of the fossils in
the rocks (when all fossils have been
collected).
3 What actually lived in the past.
As we have seen, many species never left
fossils of any kind because they had no hard
parts or lived in the wrong place. So, paleon-
tologists can strive to fi ll the gaps in the fossil
record, and that is demonstrably happening
(see Fig. 3.12), but how much closer does that
bring us to an understanding of what actually
lived at any time in the past?
Without supernatural knowledge, that
might seem hard to assess. On a good day,
paleontologists believe the fossil record (mean-
ings 1 or 2) actually does give us a good
outline of the key events in the history of life.
On a bad day, it is easy to despair of ever
really understanding the history of life
hand are such a small remnant of what once
existed.
Nonetheless, paleontologists, and other sci-
entists, mostly accept that the fossil record
(meaning 1) does give us a broadly correct
picture of the history of life (meaning 3). As
evidence for this slightly optimistic view, they
might point to the lack of surprises. If the
fossils were wildly out of kilter with the
history of life, we might expect to fi nd human
fossils in the Jurassic or dinosaur fossils in the
Miocene. We do not (despite Charles Lyell’s
famous expectation in the 1830s that we
might do just that, see p. 13). In fact, new