Diversification in colour is a widespread and striking phenomenon that characterises many of the classic adaptive radiations in nature. The overall aim of this thesis is to examine the processes that underpin the origin and maintenance of such patterns of colour variation. To do this, I used a combination of field experiments, laboratory-based trials and visual models. I tested three types of explanation for adaptive variation in colour patterns at the population and species levels, within the Ctenophorus decresii species complex: 1) natural selection for crypsis, 2) sexual selection and 3) species recognition. The Ctenophorus decresii species complex (C. decresii, C. Jionni, C. rufescens, C. tjantjalka and C. vadnappa) represents an ideal system with which to address such questions because, although they are extremely similar in most aspects of ecology and morphology, they exhibit marked colour diversification, both within and between populations and are highly sexually dichromatic. Subsequently, I used phylogenetic comparative analyses to test whether the micro-evolutionary processes that contribute to colour differentiation at the population and species level are also associated with non-random patterns of species richness at the macro-evolutionary scale.
To examine the role of natural selection for crypsis, I used two approaches. First, I used a model of the avian visual system to show that 'vivid' and 'dull' males of C decresii and C. vadnappa differ in their conspicuousness to avian predators. Field-based predation experiments using plaster replicas confirmed that there is a predation cost associated with conspicuous male coloration. Second, I quantified geographic and sex differences in crypsis among six populations belonging to the three species that display the most colour variation within the group: C. decresii, C. fionni and C vadnappa. Populations and the sexes differed in their extent of crypsis. Body regions exposed to predators were more cryptic than hidden body regions in females, but not males. There was some evidence for local adaptation in crypsis. However, several populations were no more cryptic against their own background than against the background of neighbouring populations. Taken together, these results suggest that in addition to natural selection for crypsis, selection for signalling is likely to be important in the evolution of colour variation in this group.
To investigate the role of sexual selection, I conducted laboratory based male contest and female association trials for Ctenophorus decresii and C. vadnappa. In C. decresii, the hue of the flank markings predicted male contest success whereas chest patch brightness predicted female association. In C. vadnappa, on the other hand, there was little evidence that male colour traits were important in either male contests or female choice. Overall, results of the sexual selection experiments provide some support for the hypothesis that conspicuous male colour patterns may be sexually selected, but again indicate that sexual selection is not the only salient mechanism.
I also used laboratory based male contests to examine to role of colour in species recognition. Here, I measured the level of inter-specific aggression relative to intra-specific aggression in Ctenophorus decresii and C. vadnappa. I found equally high levels of intra- and inter-specific aggression among males of both species. Male C. vadnappa exhibited higher aggression towards heterospecifics in sympatric than allopatric populations, whereas they exhibited similar levels of aggression toward conspecifics. This supports the hypothesis that high interspecific aggression is an adaptive response to interspecific resource competition, rather than the result of species misidentification. Males do recognise members of different species, though results from trials in which colour was manipulated suggest that colour alone may not be sufficient for species recognition in close-range encounters between males.
Finally, to explore the factors that may affect the distribution of species richness among agamid genera, I first tested whether observed patterns of species richness could result from purely stochastic processes. I show that some agamid lineages have many more species, and others many fewer, than expected by chance. I then use phylogenetic comparative methods to test six types of deterministic explanation for variation in species richness: body size, life history, sexual selection, ecological generalism, range size and latitude. I found that members of more species rich groups tended to have reduced sexual size dimorphism, but greater sexual dichromatism. These results provide further support for the idea that selection for signalling is important in phenotypic differentiation, and ultimately speciation in agamid lizards.
This study is novel in its combination of field and lab based experimental approaches, visual models and phylogenetic comparative approaches to the evolution of colour variation and speciation in agamid lizards. Overall, results of this study are consistent with the view that colour patterns represent a balance between the need for conspicuousness for signalling and the need for crypsis to avoid predation. Results also support the view that the micro-evolutionary processes that contribute to colour differentiation can result in non-random patterns of species richness at the macro-evolutionary scale.