Sexual selection and speciation in Lake Malawi cichlids  

Several hundred species of cichlid fish have arisen independently within the three Great Lakes of Eastern Africa within a remarkably short time period. Despite international scientific interest, the mechanisms underlying this evolutionary phenomenon are as yet poorly understood. One theory is that sexual selection, specifically female choice for particular mates, could bring about rapid changes in male breeding colours, leading very quickly, on an evolutionary time scale, to reproductive isolation. Unfortunately our knowledge of most aspects of the biology and ecology of these cichlids has to date been inadequate for such theoretically modelled hypotheses to be either supported or refuted.

The presented thesis focuses on three members of the Pseudotropheus (Maylandia) complex: shallow-dwelling rock-frequenting cichlids of the mbuna group. The abundance, accessibility and widespread distribution of these taxa throughout Lake Malawi, make them ideal model species. These species differ primarily in male breeding coloration. Experiments were conducted on these taxa to address one of the key questions in relation to evolution via sexual selection: what maintains reproductive isolation between closely related sympatric dwelling forms? It was demonstrated that males cannot visually distinguish conspecific females but a study determining parentage using microsatellite loci found that reproductive isolation between the three taxa is nonetheless maintained without exception in the artificial confines of aquaria. This provides indirect evidence for reproductive isolation by direct mate choice and suggests that females may play a more important role in mate recognition than males. An analysis of morphometric data also identified potential male secondarily sexually selected traits. This information is likely to be of use in future studies investigating the cues females may be using to choose spawning partners.

Within-population dispersal, another important parameter on which no data have hitherto been available, was addressed here using microsatellite molecular markers to statistically estimate pairwise relatedness, which was in turn correlated with pairwise geographic distance. The results of this study were highly suggestive of male-biased dispersal. Male-biased dispersal is more conducive to speciation under recently proposed sexual selection models than is female-biased dispersal. Further, knowledge of small-scale dispersal takes the field closer to a comprehensive understanding of the population dynamics of these species.