Mixed-species shoals and the maintenance of a sexual–asexual mating system in mollies

The mating system of the sexual–asexual complex of sailfin mollies, Poecilia latipinna, and Amazon mollies, P. formosa, depends on the close proximity of individuals in aggregations or shoals because P. formosa is gynogenetic (that is, sperm only initiate embryogenesis but do not contribute genetic material) and must be inseminated by male P. latipinna. Furthermore, female P. latipinna copy the mate choice of female P. formosa with the result that male P. latipinna become more attractive to conspecific females when they mate with P. formosa. Mate copying of heterospecifics has not been reported in other systems, and in mollies it could result from the inability of females to discriminate species. In experiments both P. formosa and P. latipinna females preferred to associate with a conspecific than a heterospecific, but preferred a larger heterospecific group to a smaller conspecific one. Females can thus discriminate between conspecific and heterospecific females, so that heterospecific mate copying is not a result of mistaken identity. Females preferred to associate with conspecifics, but this preference was overridden by preference for larger groups although the threshold differed between the two species. ? 1996 The Association for the Study of Animal Behaviour The Amazon molly, Poecilia formosa (Girard 1859), is an all-female species of fish of hybrid origin (Hubbs & Hubbs 1932; Schartl et al. 1995b) which relies on insemination with sperm from males of closely related sexual species (Turner 1982; Balsano et al. 1989). It reproduces by gynogenesis, that is the sperm normally only trigger embryogenesis without a genetic contribution (but see Schartl et al. 1995a). The geographical range of P. formosa encompasses two sexual species, P. mexicana Steindachner 1863, which is widely distributed throughout the Atlantic regions of Mexico, and P. latipinna (Le Sueur 1821), which is found in northeastern Mexico, southeastern Texas, and in an introduced population in the San Marcos River in central Texas (Courtenay & Meffe 1989). In Mexico P. formosa generally live in the same habitat and in mixed shoals with P. mexicana, and in Texas with P. latipinna. Poecilia mexicana is known to be the maternal ancestor of this hybrid (Avise et al. 1991), and recent research (Schartl et al. 1995b) shows that either an unknown population of P. latipinna or an extinct form from the P. latipinna ancestry was the paternal form. This group has received considerable interest from evolutionary biologists because the maintenance of this mating complex depends upon mixed-species associations and mating between heterospecifics. It is difficult to understand why host males mate with female P. formosa since this should involve some costs and no apparent benefits for the males. It is not due to the males being unable to discriminate. Several studies have shown that both P. mexicana males (Schlupp et al. 1991) and P. latipinna males (Hubbs 1964; Ryan et al., in press; I. Schlupp, M. Schartl & J. Parzefall, unpublished data) can do so. Mate copying (e.g. Dugatkin 1992) might offer a resolution (Schlupp et al. 1994). Female P. latipinna copy the mate choice of P. formosa. Thus male P. latipinna Correspondence and present address: I. Schlupp, Universität Würzburg, Theodor-Boveri-Institut für Biowissenschaften, Physiologische Chemie I Am Hubland, D-97074 Würzburg, Germany (email: [email protected]). M. J. Ryan is at the Department of Zoology, University of Texas at Austin, Austin, TX 78712, U.S.A. 0003–3472/96/110885+06 $25.00/0 ? 1996 The Association for the Study of Animal Behaviour