NICHE, MORPHOLOGY AND LOCOMOTION IN LACERTID LIZARDS 
Evolutionary change in structural niche 
Because of the range of habitats occupied by lacertids and the wide 
variation in degree of climbing they exhibit, it is very difficult to 
assign species to a simple set of well defined structural niche states. 
However there are a number of broad categories that can be recog- 
nised, even though there is considerable variation within these and 
some species may be assignable to more than one. 
G_— Ground dwelling in open areas. 
V — Ground dwelling and spending considerable time in situations 
where movement may be restricted, such as dense grassy or twiggy 
vegetation and the analogous interstices of scree etc. 
M -— Climbing regularly in vegetation where the lizard tends to 
progress through or over a matrix of twigs, leaves or grass. 
C — Climbing regularly on more or less continuous largely open 
surfaces, such as rock faces and tree boles. 
The immediate outgroup of the Lacertidae, the Teiioidea, is 
almost entirely ground dwelling and this appears to be the primitive 
situation for the next most closely related outgroups, the 
Scincomorpha and the Anguimorpha. However, while this suggests 
the earliest lacertids may have been ground dwelling too, this is not 
necessarily so for the immediate ancestor of surviving species. 
Unfortunately, the overall history of structural niche within the 
family is difficult to assess because basal relationships within the 
primitive Palaearctic assemblage are not fully resolved. However 
many of the component taxonomic units of this assemblage include 
species that climb to some extent and often, taking these units on 
their own and considering all evidence, it is more parsimonious to 
regard some degree of climbing as the primitive situation relative to 
a more ground-dwelling life mode. This is true for instance in 
Takydromus, Podarcis and its relatives and the Gallotia- 
Psammodromus clade. 
Whether it is assumed ground dwelling or climbing is primitive 
for the surviving members of the family, numerous transitions 
between different kinds of structural habitats have to be postulated. 
Even within Jakydromus there may have been a shift from climbing 
to a more ground dwelling way of life and then two independent 
shifts back to climbing (p. 66). 
When some degree of climbing versus ground-dwelling is plotted 
on the general pattern of relationships assumed here for the primitive 
Palaearctic assemblage, it is more parsimonious to assume some 
degree of climbing as the initial state, with multiple shift to life 
mainly on the ground, either in and around dense vegetation or in 
more open situations. However, this assumption is not very robust, 
as assuming a ground-dwelling ancestry in ZJakydromus rather than 
a climbing one makes the ancestral condition uncertain. 
If a partially climbing ancestry is accepted, there must have been, 
within the primitive Palaearctic assemblage, shifts to more special- 
ised climbing on continuous surfaces (C) in such forms as Lacerta 
oxycephala and L. perspicillata, and to specialised climbing in 
vegetation matrices (M) in TJakydromus. L. vivipara would have 
become ground-dwelling in dense vegetation (V) and this would 
have occurred separately in L. derjugini and L. praticola within the 
L. saxicola group. The L. parva-L. fraasii clade and L. brandtii 
would have separately entered more open ground situations (G), and 
the Psammodromus hispanicus clade occupied often intermediate 
habitats (G and V). 
The history of alteration of structural niche is clearer in the 
Armatured clade where phylogenetic structure is more apparent. 
Here, some climbing on continuous surfaces appears to have been 
the primitive situation. In the Equatorial African group there was one 
shift to specialised open surface climbing (C) in Holaspis, one to 
matrix climbing (M) in canopy in Gastropholis and one to using 
69 
ground vegetation (V) in Adolfus alleni. In the main lineage of the 
Armatured clade, parsimony supports a shift to more extensive 
ground dwelling in the ancestor of the clade made up of Tropidosaura 
and its advanced relatives with subsequent shift to more open 
habitats. At the base of this clade there would have been partial shifts 
to other modes: a reversion to a small degree of climbing in two 
species of Tropidosaura, and to making some use of vegetable 
matrixes in Poromera and perhaps Philochortus. Alternatively, 
Tropidosaura, and Nucras and its advanced relatives may have 
become ground-dwelling independently. 
Overall there may have been a minimum of nine shifts to ground 
dwelling although only about three were into really open situations 
(G), five to climbing in vegetable matrixes and others to specialised 
climbing on continuous surfaces. Among members of the Armatured 
clade, there were multiple shifts on to soft sandy substrata: at least 
one each in Pseuderemias, Meroles and Eremias and perhaps three 
in Acanthodactylus. 
Reversals in structural niche within the Lacertidae are less obvi- 
ous, although morphology suggests this may have happened in 
Takydromus, Acanthodactylus and Meroles. 
MORPHOLOGY 
Body proportions and vertebral number 
Bodies of lacertids vary in their proportions, especially in the extent 
of elongation, and change in number of presacral vertebrae is often 
associated with this. The number shows some individual variation in 
most species and females usually have more presacral vertebrae than 
males (often about one on average but sometimes two). Typically 
there are eight nuchal vertebrae and five sternal vertebrae with ribs 
attached to the sternum, but the number of more posterior presacral 
vertebrae varies considerably. There may be as few as ten in some 
Pseuderemias and Acanthodactylus and as many as twenty in some 
female Nucras lalandei, making the total presacral range for the 
family 23 to 33 vertebrae. 
Fairly elevated presacral counts also occur in Lacerta agilis, 
Lacerta parva and L. fraasii, some members of the L. saxicola 
group, L. andreanszkyi, L. graeca, Adolfus alleni and Gastropholis 
(Arnold, 1973, 1989b). Females of these forms often have a total of 
29 presacral vertebrae while Gastropholis frequently possesses 30. 
Relatively low presacral counts of 24 to 26 in females are usual in the 
more derived members of the Armatured clade including Philo- 
chortus and its sister group; they also occur sporadically elsewhere. 
Presacral vertebral count shows some correlation with habitat. 
Forms where it is high include those that spend time in dense 
vegetation, such as Lacerta agilis, Adolfus alleni, Gastropholis, and 
Nucras lalandei while numbers are particularly low in species 
regularly occurring in open situations. This may be related to the 
amount of body flexibility necessary to negotiate habitats where 
possible paths are often tortuous and ones which are unimpeded. L. 
andreanszkyi which may spend considerable time in the interstices 
of scree also has high counts. However any association between 
vertebral number and the functional demands of habitat is imprecise, 
as high counts also occur in forms that often live in open rocky 
situations, such as Lacerta graeca and members of the L. saxicola 
group. 
Sexual variation in presacral vertebral count is absent in Gallotia, 
and independently lost three times in Acanthodactylus: in A. 
bedriagai, in A. schmidti populations in the United Arab Emirates, 
and in the A. scutellatus group. All these cases appear to involve 
reduction in female presacral number, except A. bedriagai where 
