Lake Malawi cichlid mating systems: Factors that influence mate selection
The cichlid fishes of the Great Lakes of Africa are the most extreme case of adaptive radiation among vertebrates. This astounding speciation is exemplified in Lake Malawi where 450+ species have arisen within the last 2 million years. Many have argued that sexual selection by female choice could not only bring about speciation, but do so at this rapid pace. The details necessary to evaluate models of sexual selection, however, have remained elusive. My research was designed to evaluate critical aspects of mating in Lake Malawi shallow-water, mouthbrooding cichlids.
I determined the genotypes at two microsatellite loci for the progeny of 13 females, representing seven species of rock- and sand-dwelling cichlids. Multiple paternity (2 to 6 males) was clearly demonstrated for all but two broods. Heterozygosity of the two loci was 89 and 83%, and the expected average exclusion probability was 0.968, suggesting that loci variation was sufficient for quantifying mate choice.
Microsatellite loci were used to evaluate the level of polyandry and intraspecific brood mixing in a paedophagous cichlid that inhabits the rock/sand interface. Broods were fertilized by 1 to 3 males, and intraspecific brood-mixing was found in 4 out of the 6 broods examined. Proportions of foreign fry ranged from 6 to 65%.
I studied mate choice in a lekking, sand-dwelling cichlid and found increased bower (sand structures constructed by males) height and higher numbers of conspecific neighbors to be positively, significantly correlated with male mating success. Experimental manipulation confirmed that increased bower height resulted in increased male success.
I investigated the role of body coloration as a mating criterion in rock-dwelling
cichlids by giving females the choice between three differently colored sympatric
species (1 conspecific and 2 heterospecifics). Using live males, the duration of
indirect interactions with the conspecific was significantly different than what
would be expected by chance. Using this behavior as an indicator of choice, live
males were replaced with animated models. Animation experiments yielded ambiguous
results with only 3 out of 12 females displaying nonrandom indirect interactions.
Body coloration may be insufficient to induce choice, or inadequate animation
models may have influenced the results.