Sexual selection, conflict, and the evolution of visual mating signals

Many animals use colour signals for sexual signalling. Our interest is in understanding the evolution of the most exaggerated versions of these signals, such as the extremely bright and iridescent structural colours of tropical butterflies and other insects. These colours are often limited to males, or considerably more exaggerated in this sex, and are known to be used in various realms of sexual competition. In the case where such a colour badge is used to indicate mate quality, theory expects them to become highly exaggerated (and, hence, costly), sexually dimorphic, and strongly condition-dependent. Empirical studies across a range of animals have also demonstrated that these traits exhibit higher genetic variances than most other non-sexual traits, probably because their expression is strongly integrated into organismal development and physiology.

We address questions at the forefront of sexual signalling theory by observing and manipulating morphology, behaviour and development, and by using breeding experiments to explore phenotypic and genetic sources of variation. We are also interested in the mechanisms of colour signal production, which we study using reflectance spectrometry and electron microscopy. The male-limited nature of these traits also offers opportunities for studying intralocus sexual conflict and the evolution of sexual dimorphism. Projects along these lines are currently underway using butterflies, harlequin bugs, guppies and lizards.


Predicting the direction of evolution in the wild

We also use animal colouration is to explore whether phenotypes are likely to be inherently favoured as components of sexual signalling systems. The field of colour-based signalling has been pre-occupied with how a select few pigments (mostly carotenoids) give rise to ‘honest’ signals because they must be acquired from the environment (rather than synthesized). However, colour signals generated by a broad range of non-mutually exclusive pigmentary and structural mechanisms that are costly or difficult to express, and hence, potentially highly honest signals. A key question is whether particular mechanisms of colour production are likely to be consistently favoured by evolution due to their information content, or whether certain colours are favoured due to sensory drive/bias, or whether elaboration of colour traits occurs via idiosyncratic species-specific Fisherian runaway methods.

We are presently studying these questions in guppies (Poecilia reticulata) and via ongoing collaborations with Prof. David Reznick (University of California, Riverside). Prof. Reznick heads a major research program whereby guppies from various Trinidadian populations have been transplanted to upstream reaches in which visually-mediated predation is markedly reduced. Sexually selected colouration becomes elaborated in these reduced predation populations, and we can chart the trajectory of such evolution over subsequent years.

Our work on adaptive trajectories also extends to understanding the extent to which evolutionary genetic factors can explain contemporary species distributions, and the potential for evolution under environmental change. This work includes collaboration with Dr. Carla Sgro (Monash University) to test the hypothesis that restricted tropical species may in fact lack adequate amounts of genetic variation for coping with environmental stress. We have studied this using Eurema butterflies, which have rather different life histories than flies, yet show similar distributional patterns and are amenable to quantitative genetic approaches.


Evolutionary/behavioural ecology of host-endosymbiont interactions

Our interest in endosymbiosis stemmed from the serendipitous discovery that one of our model butterfly species (Eurema hecabe) is host to the bacterium Wolbachia. This is a widespread endosymbiont of invertebrates that is transmitted maternally, and sometimes manipulates host reproduction to facilitate its rate of transmission. Wolbachia’s most intriguing manipulation involves “hijacking” its hosts’ sex-determining machinery to cause genetic males to develop into viable female phenotypes. While enhancing bacterial transmission, this disrupts host sex ratios from the Fisherian ideal (50:50), and sets the scene for a co-evolutionary arms race between host and endosymbiont.

We are studying co-evolution in this system by exploring the host counter-adaptations, such as resistance (to feminisation) in infected lines and/or mate discrimination. This work is centred upon E. hecabe, but we are also keen to see whether this strain of Wolbachia has spread horizontally to closely related butterflies or other arthropods that interact with E. hecabe. The presence of feminising Wolbachia offers outstanding possibilities for insights into functional and evolutionary features of sex determination and sex differentiation, including the role played by epigenetics.