NICHE, MORPHOLOGY AND LOCOMOTION IN LACERTID LIZARDS 
tion to shorter limbs in females rather than increase in limb size in 
males also militates against this explanation. 
It might be thought that the relatively low hindlimb/snout-vent 
ratios of many female lacertids is a result of their usually higher 
average number of presacral vertebrae than males (Arnold, 1973, 
1989a), something that tends to produce a proportionally longer 
body. However, while this may be a partial cause of low ratios, it is 
not a total explanation. In Gallotia atlantica and G. galloti caesaris, 
for instance, where virtually all individuals have 26 presacral verte- 
brae without sexual difference, females still have relatively shorter 
hind legs. 
Relative proportions of femur and tibia 
Measurements of the femur and tibia on dry skeletons and cleared 
and stained preparations of single individuals of a wide range of 
lacertid species show considerable variation. The approximate ratio 
of tibia length to femur length is generally low in members of the 
primitive Palaearctic assemblage and more basal members of the 
Armatured clade where it ranges from about 0.73—0.87. The ratio is 
particularly low, about 0.73—0.77, in such climbing forms as Lacerta 
oxycephala, L. bedriagae, L. horvathi, L. perspicillata, L. mosorensis 
and Holaspis guentheri. 
Generally rather higher ratios, 0.76—-0.87 occur in Takydromus 
septentrionalis, Lacerta agilis, L. vivipara, Psammodromus algirus, 
Lacerta schreiberi, L. pater, L. chlorogaster, L. monticola, L. dugesii, 
Podarcis bocagei, P. muralis, P. sicula, Adolfus jacksoni and Poromera 
fordi. 
Ratios are higher still, 0.88—1.00, in Psammodromus hispanicus, 
Lacerta trilineata, Adolfus alleni and the clade containing more 
advanced members of the Armatured group, namely Nucras and its 
sister group, most of which are largely or entirely ground dwelling 
in open places. Included here are Poromera fordi, Nucras boul- 
engeri, N. lalandei, Philochortus spinalis, Latastia longicaudata, 
Heliobolus lugubris, Ichnotropis squamulosa, Pseuderemias bren- 
neri, Meroles knoxii, M. ctenodactylus, Eremias arguta, Acantho- 
dactylus erythrurus, A. boskianus, Mesalina rubropunctata, 
Ophisops elegans. 
Patterns of limb growth 
Like patterns of tail growth, the way in which the length of hind 
limbs relative to that of the head and body changes during growth 
from hatching to maturity is extremely varied. In such forms as 
Takydromus septentrionalis and Lacerta oxycephala the hindlimbs 
retain much the same relative size, while in Acanthodactylus 
scutellatus andA. schmidti they show distinct reduction, for instance 
growing only 90% as fast as the head and body length in A. schmidti. 
In Podarcis hispanica and P. peloponnesiaca, the relative length of 
the hindlimbs is retained in males but falls substantially in females. 
Lacerta bedriagae, L. laevis, L. danfordi and L. perspicillata appear 
to show some decline in relative rate of limb growth in both sexes, 
perhaps after a slight initial rise, but the decline is more marked in 
females. In cases where relative limb length changes with body size, 
it is important to compare males and females of similar head and 
body length when assessing sexual differences in limb proportions. 
Evolutionary plasticity of limb proportions 
It will be seen from Table | that hindlimb span often varies substan- 
tially among closely related species, for instance within the genus 
Mesalina and within the Lacerta agilis group (L. agilis, L. trilineata, 
L. viridis etc.). This is also sometimes true of forms successively 
derived from a lineage, such as the genera of the Armatured clade. 
Such variation suggests that hind limb proportions relative to the 
body length are quite plastic in evolutionary time, something cor- 
qs 
roborated by the varying amount of sexual dimorphism and the very 
different growth patterns encountered. Lineage effects (Arnold, 
1994b) in the form of phylogenetic, and specifically developmental, 
constraint, consequently do not seem to be important in restricting 
change in relative hind-limb length. 
Although there is a clear tendency among species and sexes for 
increase in relative hind-limb length to be associated with increased 
difference between fore and hindlimbs, this is also unlikely to 
represent a strong developmental constraint as the scatter of points in 
Fig. 3 around the general trend is very substantial. Forms like 
Poromera and Psammodromus algirus have similar relative hind 
limb lengths but differ substantially in forelimb/hindlimb ratios. 
Conversely Latastia longicaudata and Meroles anchietae possess a 
similar forelimb/hindlimb ratio but differ greatly in relative hind 
limb length. Again, although differences between sexes often follow 
the general trend between species, there are cases where this is not 
so. 
It is noteworthy that the primitive Palaearctic assemblage and 
more primitive members of the Armatured clade have quite short 
legs but, given the general plasticity of limb proportions, this seems 
unlikely to represent a developmental constraint and may simply 
reflect the habitats they usually occupy. The limb proportions of 
Psammodromus algirus, which belongs to the primitive Palaearctic 
assemblage but often runs on the ground in open areas, as well as 
climbing, approach those of advanced armatured forms that are 
nearly all found in such situations. 
Functional aspects of differences in limb proportions 
Given the plasticity of limb proportions in lacertid lizards and their 
correlation with kinds of habitats occupied, it would not be surpris- 
ing if differences between species reflected the functional 
problems of locomotion in particular environments and were pro- 
duced by natural selection. A more detailed case for this is given in 
the rest of this paper but likely advantages of different limb propor- 
tions are briefly summarised here. Ground dwelling forms from 
open habitats get most of their forward thrust when running from 
the hind limbs. Such thrust is enhanced by greater general hind- 
limb length relative to the forelimbs, and an extended crus reflected 
in increased tibial length relative to the femur. The openness of the 
habitat allows such long hind limbs to be used effectively and 
probable increase in mass of the caudifemoralis longus muscle 
increases the power of what is a high-gear system of locomotion 
that delivers the high speeds necessary to evade predators in situ- 
ations where cover is sparse. 
In contrast, ground dwelling forms that spend considerable time 
in dense vegetation benefit from generally short limbs which can be 
deployed in the restricted spaces available. Speeds are lower but 
concealment from predators is easier. Possibly the greater relative 
length of the forelimbs reflects greater use in locomotion. Thrust 
from the small hindlimbs may not be optimal for locomotion and the 
flexibility of an often relatively long body may reduce its effective 
transmission. In these circumstances some traction by the forelimbs 
may be advantageous. 
Climbers in vegetation matrixes have similar proportions to those 
just discussed and are likely to encounter similar locomotory prob- 
lems. Another factor favouring short limbs in climbers in general is 
that they give low gearing which is likely to be beneficial when 
moving upwards against the force of gravity. The relatively long 
forelimbs in these forms may also allow them to contribute effec- 
tively to upward locomotion and they are also important in securing 
the foreparts, which are above the centre of gravity of the lizard as a 
whole during vertical climbing and so liable to fall away from the 
surface being climbed if unattached. 
