The genetics of mate preferences in hybrids between two young and sympatric Lake Victoria cichlid species

Abstract

The genetic architecture of mate preferences is likely to affect significant evolutionary processes, including speciation and hybridization. Here, we investigate laboratory hybrids between a pair of sympatric Lake Victoria cichlid fish species that appear to have recently evolved from a hybrid population between similar predecessor species. The species demonstrate strong assortative mating in the laboratory, associated with divergent male breeding coloration (red dorsum versus blue). We show in a common garden experiment, using DNA-based paternity testing, that the strong female mate preferences among males of the two species are fully recovered in a large fraction of their F2 hybrid generation. Individual hybrid females often demonstrated consistent preferences in multiple mate choice trials (more than or equal to five) across a year or more. This result suggests that female mate preference is influenced by relatively few major genes or genomic regions. These preferences were not changed by experience of a successful spawning event with a male of the non-preferred species in a no-choice single-male trial. We found no evidence for imprinting in the F2 hybrids, although the F1 hybrid females may have been imprinted on their mothers. We discuss this nearly Mendelian inheritance of consistent innate mate preferences in the context of speciation theory.

Keywords: Pundamilia nyererei; Pundamilia pundamilia; assortative mating; hybridization; sensory drive; speciation-with-gene-flow.

Figure 1.

Individual spawning decisions by the 69 F2 hybrid females. Spawning decisions were determined by microsatellite DNA paternity analyses. Above the line y = 0 is the number of spawning decisions with P. sp. ‘pundamilia-like’, and below the line is the number of spawning decisions with P. sp. ‘nyererei-like’. The order of spawning decisions is arranged with the first spawning on the top, and the last on the bottom with a spawning decision with P. sp. ‘pundamilia-like’, marked in blue and a spawning decision with P. sp. ‘nyererei-like’ marked in red. *p < 0.05, ^a0.05 < p < 0.1.

Figure 2.

(a) Simulated (means and 5–95% error bars) spawning decisions of F2 hybrid females with P. sp. ‘pundamilia-like’ (blue dots), and with P. sp. ‘nyererei-like’ (red dots) based on a repeatability of an individual's spawning decision of R = 0.7. Observed ratio of spawning decisions is shown by the solid black lines. (b) The best fit of the model is with R = 0.7, which minimizes the mean squares (MS) between the observed and simulated spawning distribution. Lower values of R produce a more random spawning pattern, while higher values of R increase the consistency of a females' spawning choices above those observed, which reduced the fit of the model by inflating the MS.