Abstract

Natural selection theory predicts that similar ecological pressures can result in similar
phenotypes evolving in distantly related species through a process known as convergent
evolution. Although there are many described cases of species adaptively evolving similar
phenotypes in response to the same habitats (i.e., “ecomorphs”), this is not a universal
outcome of natural selection. Liolaemus lizards embody an exceptional example of
adaptive radiation where, thus far, studies have not identified any evidence of
ecomorphological evolution. Why some lineages display such signals of convergent
evolution while others do not remains an open question. Here, shape was more accurately
quantified by using geometric morphometrics (as opposed to the linear measures of
morphology used previously) to address the question of whether Liolaemus lizards have
undergone convergent evolution in body morphology and to conduct phylogenetic
modelling to quantify patterns and rates of multivariate morphological evolution in this
radiation. Further to this, the effect of diet and parity mode on morphology were explored.
The morphometric analysis produced three principal component (PC) axes that explained
over 10% of variance. Each axis captures different changes in shape; PC1 shows a
reduction in relative head size when the body becomes wider and more elongated and
vice versa. PC2 shows changes in the width of the body, with little change in the length
of the body and of changes in the head. PC3 shows lengthening in the posterior of the
body while there is a pinching towards the front of the body and an increase in relative
head size and vice versa. Results generally showed no significant separation in body
shape between microhabitat groups, and thus suggest ecomorphs are not present. Diet and
parity groups were significantly separated, however, not when phylogeny was included
in the analyses. Phylogenetic macroevolutionary analyses were used to develop an
understanding of the evolution of biodiversity. OU models of evolution were found to be
the best fitting evolutionary models in explanation of the evolution of shape within
Liolaemus when BM, OU, δ and κ models were fit. Phenotypic rate variation showed
some increases in rate of evolution on PC1 whereas PC2 and PC3 showed decreases.
PGLS showed a significant relationship between SVL and PC1 and for SVL and PC2
when microhabitat, parity and diet were included. Pairwise comparisons showed a
significant difference between insectivorous and herbivorous species for SVL and PC1
open-ground shrubs and open ground species for SVL and PC2. The best fitting model
9
for PC1 and PC2 included microhabitat and diet, but not parity, for PC3 the best fitting
model included diet and parity, but not microhabitat. Results suggest the lack of an
ecomorphological relationship could be a consequence of other ecological factors
exerting a stronger pressure on the evolution of morphology or that the adaptations are
present but methodologies are not appropriate to identify them.