The mechanisms responsible for the explosive radiations of haplochromine cichlid fishes
in Africa's great lakes remain controversial. Since species thought to be closely related
often differ most apparently in male breeding colours, I examined patterns of male colour
variation in rock-dwelling cichlids from Lake Malawi to test whether initial divergence
between species is likely to have been caused by adaptation to differing habitats, by
selection against hybridization (reinforcement or character displacement), or by sexual
selection. I found that all significant variation in overall male colour occurs within,
not between, species complexes, in contrast to variation in habitat and behavioural traits,
which differed significantly between complexes. Male colour does not vary significantly
with habitat characteristics such as water depth, and sympatric species differ no more
in colour than allopatric ones. These results fail to provide support for adaptation or
reinforcement as initial causes of colour divergence, and are consistent with the
possibility that divergence may instead have been caused by sexual selection, but a
direct test of sexual selection is not yet possible.