The lizard, Ano/is oculatus (Squamata: Iguanidae), is endemic to the Commonwealth of Dominica in the Lesser Antilles. A manipulative experimental approach was adopted to test for the causes of natural selection operating in natural environments. Four large enclosures were constructed in xeric Caribbean coastal woodland in which both control and enviromnentally translocated populations could be maintained and monitored.
Both wet and dry season translocation experiments were canied out, with two
replicate enclosures filled with control populations and two replicate enclosures filled with montane rainforest specimens. The intensity of selection was calculated from multivariate differences between survivors and non-survivors. Partial regressions allowed selection coefficients to be calculated for specific character traits permitting speculation on trait specific directional selection. Results indicate that significant and rapid selection occurs in the montane populations following the selection event.
Due to the pronounced geographic variation in phenotype of Anolis oculatus, shown to be congruent to several environmental parameters, it is suggested that the observed natural selection, is at least partly due to environmentally induced selection from current ecological conditions. To further test this theory, twelve experimental populations, sourced from an environmental gradient from across Dominica, were translocated to the enclosures. The results show a gradient of selection intensity, strongly suggesting an environmental component is significantly present in the natural selection experienced by the experimental
populations.
To investigate an aspect of the evolutionary ecology of Ano/is oculatus, sexual size dimorphism was recorded across the species range and tested against a variety of behavioural, ecological, envirnnmental, and phylogenetic variables. This study reveals that sexual selection, in the form of male-male competition, appears to be the most closely linked trait to the observed patterns of sexual size dimorphism. However, gender specific diet selection is also a significant component in the ecology of sexual size dimorphism.
Finally, a common garden experiment was conducted to test for the presence of
environmentally induced phenotypic plasticity in many of the character traits used to assess natural selection. This was done by incubating, hatching, and raising anoles from nine environmentally distant habitats in a common garden enclosw-e and testing the common garden juvenile phenotypes against those of source population specimens. Results suggest that environmentally induced developmental phenotypic plasticity does not significantly effect the character traits under analysis.