Reproductive isolation in four populations of the striped mouse Rhabdomys

Abstract

Reproductive isolation was investigated in four phylogenetically and/or geographically separated populations of Rhabdomys in South Africa. The four populations, Jonkershoek (Rhabdomys pumilio), Goegap (R. pumilio), Irene (R. dilectus dilectus) and Suikerbosrand (R. d. chakae), represent the two putative species, R. pumilio (Sparrman, 1784) and R. dilectus De Winton (1897), as well as the two sub-species, R. d. dilectus (Wroughton, 1905) and R. d. chakae (Wroughton, 1905), of Rhabdomys. The populations occur > 900 km apart, except for the Irene and Suikerbosrand populations which are approximately 80 km apart. Inter- and intrapopulation breeding experiments and behavioural studies were used to test for pre- and/or postzygotic reproductive barriers. In breeding experiments, most intrapopulation pairs produced offspring. In the interpopulation breeding tests, except for one litter produced by an Irene-Jonkershoek pair, which did not survive, only the Jonkershoek-Goegap pairings produced offspring, which were fertile and had growth rates similar to those of offspring produced in intrapopulation pairings. However, the smaller litter size of the Jonkershoek-Goegap pairings compared to intrapopulation pairings, suggests post-zygotic failure between these two R. pumilio populations. In the behavioural experiments, I tested the responses of females to the soiled bedding of homotype (same population) and heterotype (different population) males. Two experiments were conducted: habituation-discrimination and habituation-generalization tests were used to investigate within- and between-taxon variations in male odour quality and female perception; and choice tests were used to test female preference. The results of the behavioural experiments indicate that there is variation within the R. pumilio (Jonkershoek and Goegap) taxon in odour quality, perception and preference. The Jonkershoek females could discriminate between their own population males and those of Goegap, and preferred their own males, while the Goegap females were unable to distinguish between their own population scent and the Jonkershoek population scent and therefore were unable to display a preference for their own population scent when a Jonkershoek/Goegap scent choice was presented. The two subspecies of R. dilectus perceived the scent of males from their own population as being more similar to each other than to that of R. pumilio, and Irene females perceived the two R. pumilio populations as different. All four populations displayed assortative mate preference and preferred their own population’s scent over all the others, with the exception of the Goegap population. My study indicates that phylogeny, and not geography, appears to be a more parsimonious explanation for the pattern of divergence in these four Rhabdomys populations, although ecological influences cannot be ruled out.