Amazon molly (Poecilia formosa) females reproduce asexually, but they need sperm to initiate the process. Such gynogenetic reproduction can be called sperm parasitism since the DNA in the sperm is not used. Since all offspring of asexually reproducing females are females, they can locally outcompete sexually reproducing ones, but their persistence is threatened by the lack of males. Therefore, the existence of Amazon mollies is puzzling. A metapopulation structure has been suggested to enable the coexistence of gynogenetic and sexual species. Previously only Levins-type metapopulation models have been used to investigate this question, but they are not defined on the individual level. Therefore we investigate the evolution of sperm parasitism in a structured metapopulation model, which incorporates both realistic local population dynamics and individual-level dispersal. If the reproduction strategy is freely evolving in a large well-mixed population or in the structured metapopulation model, strong discrimination of asexually reproducing females by males results in evolution to full sexuality, whereas mild discrimination leads to too small probability of sexual reproduction, so that the lack of males causes the extinction of the evolving population, resulting in evolutionary suicide. This classification remains the same also when both sexual reproduction and dispersal are freely evolving. Sexual and asexual behaviour can be observed at the same time in this model in the presence of a trade-off between the reproduction and dispersal traits. However, we do not observe disruptive selection resulting in the evolutionarily stable coexistence of fully sexual and fully asexual females. Instead, the presence of sexual and asexual behaviour is due to females with a mixed reproduction trait.