The Origin and Evolution of Mammals - Moodle

it means that a 200 Ma igneous rock of, say, Upper
Triassic when the first mammals appeared may be
dated at plus or minus 4 Ma, a significant time in
evolutionary terms.
Absolute dating of an actual fossil requires that
the age of the sediment containing it is assessed relative
to available absolutely dated rocks that are
close to it in time and space. This often demands
such further error-prone procedures as measuring
the thickness, and estimating the rate of deposition
of the sediments separating the fossil layer and the
igneous rock exposure.
The International Stratigraphic Chart
The relevant parts of the 2000 version of the
International Stratigraphic Chart for Phanerozoic
time are presented in Fig. 2.1 for reference. It is not
always completely consistent with the text that follows
because a lot of the literature on synapsids has
not been based upon this particular timescale. One
of the most prominent inconsistencies is the frequent
use by synapsid palaeontologists of a twofold
division of the Permian into Early and Late Epochs,
compared to the threefold division now adopted, of
Lower, Middle, and Upper. Another is the difference
between the European-dominated use of Lower,
Middle, and Upper Epochs for the Carboniferous, in
contrast to the North American twofold arrangement
into Mississippian and Pennsylvanian that has
been adopted by the international scheme.
Classification
For all organisms, fossil and living, the first requirement
for practically any evolutionary interpretation
is a cladistic analysis of their relationships, whether
expressed as a cladogram, or as a formal classification
of monophyletic groups. A hypothesis of the
branching pattern that connects species and taxa
allows objective inferences to be made of the pattern
and sequence of evolution of characters, and of the
biogeographic history of the group. The battles of
the 1970s and 1980s over the appropriateness and
significance of cladism have long since been won and
lost, and the focus has turned from the question of
principle to the question of how actually to analyse
the huge amounts of information about characters
and taxa now available, and requiring ever-more
powerful computer and statistical techniques.
Morphological characters
TIME AND CLASSIFICATION 9
Dealing with the morphological characters of fossils
still relies primarily on parsimony, that is to say, the
assumption of a model of evolution based on minimum
evolutionary change. No other model, for
instance one that might make assumptions about different
probabilities of change in different sorts of
characters, is defensible in the absence of any independent
method of measuring such probabilities.
Therefore, minimising the total number of inferred
evolutionary changes in characters generates
the simplest hypothesis of the overall pattern of evolution.
Apart from the obvious incorrectness of the
assumption of equal probability of change in all characters,
morphological analysis suffers from the
intractable problem of what constitutes a unit character.
That different authors can study exactly the
same fossils but then produce different cladograms,
sometimes markedly different ones, is mainly due to
the unit character issue. As an extreme example, a
single mammalian molar tooth could be coded as a
single character state such as ‘hypsodont’, or by
maybe a dozen characters, one for each cusp or loph.
Searching for whether proposed multiple characters
are always correlated with one another, and assuming
in such a case that they should be treated as a single
one, potentially could remove highly informative
evidence of relationships, if the characters were in
fact independently evolved. This and other problems
such as how to assess which characters are homologous
with which, how to treat multistate characters,
what to do about characters not preserved in some of
the specimens, and so on are discussed at length in
the literature, for example, by Kitching et al. (1998)
and will not be pursued here. But the fact remains
that for all the confidence some authors place in their
long lists of morphological characters, their application
of such programmes as PAUP, and the associated
statistical tests for how well the data fits and
how robust the cladogram is, there remain major
outstanding disagreements between them over several
critical parts of the synapsid cladogram.
Molecular taxonomy
The rise of molecular taxonomy based on sequences
of nucleotides in genes has revolutionised the field
of classification of living taxa, and nowhere has there