Figure 1.1 A super-tree of relationships among early (fossil) tetrapods. To aid in interpreting the structure of the tree, we have color-coded major groups that are discussed in the text. In addition, color-coded lines indicate clades from which extant

1
Figure 1.1 A super-tree of relationships among
early (fossil) tetrapods. To aid in interpreting
the structure of the tree, we have color-coded
major groups that are discussed in the text. In
addition, color-coded lines indicate clades from
which extant ampibians (orange) and extant
reptiles (red and blue) arose. Orange lines
indicate the Lissamphibia, the group from which
all extant amphibians originated. Blue lines
indicate the Parareptilia, the group from which
turtles were once believed to have originated.
Although modern turtles have historically been
placed in the Parareptilia based on their anapsid
skull, recent molecular data indicate that they
are nested within the Eureptilia. Red lines
indicate the Eureptilia, the group from which all
modern reptiles originated. It is useful to refer
back to this graphic as you read through the
history of tetrapod evolution in order to tie
group or fossil names with appropriate
evolutionary groups. Adapted from Ruta and
Coates, 2003, and Ruta et al., 2003.
2
Figure 1.2 A sampling of adult body forms in
living amphibians.
3
Figure 1.3 A sampling of adult body forms in
living reptiles.
4
Figure 1.4 Relationships, body forms, and limb
structure of the seven key fossil vertebrates
used to recover the evolution of supportive limbs
in tetrapods. Glyptolepis is the outgroup.
Adapted from Ahlberg and Clack, 2006 Clack 2006
Daeschler et al., 2006 and Schubin et al., 2006.
5
Figure 1.5 Air-breathing cycle of the longnosed
gar (Lepisosteus osseus). As the gar approaches
the surface at an angle, it drops its buccal
floor and opens its glottis so air can escape
from the lungs (bottom center, clockwise). By
depressing the buccal floor, the gar flushes
additional air from the opercular chamber. Once
flushed, the gar extends its snout further out of
the water, opens its mouth, depresses the buccal
floor drawing air into the buccal cavity, and
shuts the opercula. The mouth remains open and
the floor is depressed further then closing its
mouth, the gar sinks below the surface. Air is
pumped into the lungs by elevating the buccal
floor. Adapted from Smatresk, 1994.
6
Figure 1.6 Fin and limb skeletons of some
representative fishes and tetrapods. Top to
bottom, ray-finned or actinopterygian fin,
osteolepiform lobed fin, actinistian lobed fin,
porolepiform lobed fin, lungfish or dipnoan lobed
fin, and a tetrapod limb. Adapted from Schultze,
1991.
7
Figure 1.7 Comparison of the skulls of an early
amphibian Edops and an early reptile Paleothyris.
Scale bar 1 cm. Reproduced, with permission,
from Museum of Comparative Zoology, Harvard
University
8
Figure 1.8 A branching diagram of the evolution
within the Tetrapoda, based on sister group
relationships. The diagram has no time axis, and
each name represents a formal clade-group name.
After Clack, 1998 Gauthier et al., 1988a,b,
1989 and Lombard and Sumida, 1992 a strikingly
different pattern is suggested by Laurin and
Reisz, 1997.
9
Figure 1.9 A branching diagram of the evolution
of basal Amniota and early reptiles, based on
sister group relationships. The diagram has no
time axis, and each capitalized name represents a
formal clade-group name. Opinion varies on
whether the mesosaurs are members of the Reptilia
clade or the sister group of Reptilia. If the
latter hypothesis is accepted, the Mesosauria and
Reptilia comprise the Sauropsida. Turtles
(Testudines) are shown here as nested within the
Parareptilia based on morphology. More recent
molecular analyses indicate that they are nested
in the Eureptilia (see Chapter 18). After
Gauthier et al., 1989 Laurin and Reisz, 1995
and Lee, 1997 a strikingly different pattern is
suggested by deBraga and Rieppel, 1997.
10
Figure 1.10 Presence of the amnion defines the
Amniota. Viviparity is not necessarily associated
with presence of an amnion. This distribution of
egg-retention based on extant species does not
permit the identification of the condition in
basal amniotes. The origin of terrestrial
amniotic eggs as an intermediate stage is equally
parsimonious with the evolution of amniotic eggs
within the oviduct to facilitate extended egg
retention. After Laurin and Reisz, 1997.
11
Figure 1.11 A branching diagram of the evolution
of basal reptile clades, based on sister-group
relationships. The diagram has no time axis, and
each capitalized name represents a formal
clade-group name. Plesiosaurs is used as a
vernacular name and is equivalent to Storrs's
(1993) Nothosauriformes. After Caldwell, M.,
1996 Gauthier et al., 1989. .
12
Figure 1.12 A branching diagram of the evolution
within the Archosauromorpha, based on
sister-group relationships. The diagram has no
time axis numerous clades and branching events
are excluded and each capitalized name
represents a formal clade-group name. After
Benton and Clark, 1988 Gauthier et al., 1989
Gower and Wilkinson, 1996 .
13
Figure 1.13 A branching diagram of the evolution
within the Lepidosauromorpha, based on
sister-group relationships. The diagram has no
time axis numerous clades and branching events
are excluded and each capitalized name
represents a formal clade-group name. After
Gauthier et al., 1989 Rieppel, 1994 Caldwell
(1996) and deBraga and Rieppel (1997) provide
different interpretations of lepidosauromorph
relationships.
14
Figure 1.14 Linnean taxonomy places organisms in
categories based on overall similarity.
Evolutionary taxonomy places organisms in clades
based on relatedness (homologies), which has a
clear time component. A dendogram based on
Linnean taxonomy (a) contains many polytomies
because categories are discreet, (b) can contain
some "species" (AF and GK) that are "equal" in
rank with similar hierarchical organization to
the subfamily level and others (L in particular)
that contain this structure only in name, and (c)
has no time component. Thus species L is in L
subfamily. Dashed lines indicate where the
taxonomic categories would occur for species L. A
dendogram of evolutionary relationships has no
clear genus, subfamily, or family structure but
presents a relatively accurate hypothesis of
known relationships and relative divergence
times. Species are endpoints of divergences.
Because of the implicit lack of a time element,
individual taxonomic groups in the Linnean system
often do not have comparable evolutionary
histories across taxa. For example, a family of
scorpions might have a much deeper (older)
evolutionary history than a family of snakes.
15
Figure 1.15 In evolutionary taxonomy, names of
evolutionary groups of organisms (clades) can be
confusing. Node-based clades are defined as the
most recent common ancestor (the black circle)
and all descendents. For example, Anura is the
most recent common ancestor of Ascaphus and
Leiopelmatidae. Stem-based clades are defined as
those species sharing a more recent common
ancestor with a particular organism (the stem)
than with another. Thus Salientia is all taxa (in
this case Ascaphus and Leiopelmatidae) more
closely related to Anura than to Caudata.
Apomorphy-based clades share a particular unique
character (the bar in the graphic on the right).
Thus Anura would be the clade stemming from the
first amphibian to have a urostyle (a skeletal
feature unique to frogs).
16
Figure 1.16 An abbreviated cladogram of tetrapods
illustrating monophyly, paraphyly, and polyphyly.
The heavier lines and capitalized group names
depict the monophyletic groups of Amphibia and
Reptilia recognized in the text. The boxes define
earlier concepts of Amphibia (polyphyletic) and
Reptilia (paraphyletic).
17
Figure 1.17 An electrophoretic gel (zygogram) of
an esterase stain for numerous individuals of the
salamanders Plethodon cinereus and P. shenandoah.
Each vertical bar is an individual salamander,
whereas each dark crossbar is a stained enzyme.
Courtesy of A.Wynn.
18
Figure 1.18 The production of phylogenetic trees
from gene sequence data is a relatively easy
process, at least conceptually. Gene sequences
are assembled from the organisms of interest (A).
These can be obtained from animals collected,
tissues borrowed, or sequences already available
(see GenBank). Typically, at least one outgroup
(distantly related taxon) is included to root the
tree (determine oldest nodes within the group of
interest). Homologous sequences are then
assembled from the various samples (B). All
sample sequences are then aligned (homologous
nucleotides in columns) to identify insertions
and deletions (different nucleotides than
expected based on homology) (C). These indicate
evolutionary change for a particular sample
sequence. Models of sequence evolution for
analyses are then chosen (D) based on data
available and model complexity. Traditional
analyses and/or Bayesian analyses are then
applied to data to reconstruct evolutionary trees
from the data (E). A number of traditional
approaches exist (Table 1.8) that are based on
analyses of bootstrapped data (a subsample of
data used to define models to test with remaining
data) (E). The relatively newly applied Bayesian
approaches use a Markov chain Monte Carlo (MCMC)
analysis, a randomization procedure that has much
stricter rules (E, and see Holder and Lewis,
2003). Both of these produce numerous trees that
differ slightly in structure. A "best" tree is
selected based on a set of criteria, or in some
cases, several "best" trees are reported if the
analyses provide support for more than one (F).
Because all phylogenetic trees are hypotheses,
they can then be tested with additional data (G).
19
Figure 1.19 Construction of branching diagrams by
two methods phenetics and cladistics. The OTU x
Character matrix (upper left) contains five OTUs
(AE) and six characters (16). Each character
has two states, 0 or 1 (e.g., absent or present,
small or large, etc.). Pairwise comparison of
OTUs creates an OTU x OTU matrix. The distance
values are the sums of the absolute difference
between states for all six characters. Zeros fill
the diagonal because each OTU is compared to
itself only half of the matrix is filled with
the results of a single analysis because the two
halves are mirror images of one another. An
unweighted pair-group method (UPGM) clustering
protocol produces a phenetic dendrogram
(phenogram, middle left) in UPGM, the most
similar OTUs are linked sequentially with a
recalculation (middle right) of the OTU x OTU
matrix after each linkage. The cladogram (lower
left) derives directly from the OTU x Character
matrix. The solid bars denote a shared-derived
(synapomorphic) character state, the open bars an
evolutionarily reversed state, and the character
numbers. For comparison with the UPGM phenogram,
the cladogram is present in a different style
without the depiction of character state
information.