Simulated emergence of cyclic sexual-asexual reproduction

Motivated by the cyclic pattern of reproductive regimes observed in some species of green flies (" aphids "), we simulate the evolution of a population enduring harsh seasonal conditions for survival. The reproductive regime of each female is also seasonal in principle and genetically acquired, and can mutate for each newborn with some small probability. The results show a sharp transition at a critical value of the survival probability in the winter, between a reproductive regime in the fall that is predominantly sexual, for low values of this probability, or asexual, for high values. The question of why sexual reproduction prevails in the immense majority of metazoan species still defies a complete and satisfactory answer-for a recent review we direct the reader to Reference [1]. A number of different theories have been put forth to try to explain this puzzle, but all have so far fallen short of becoming a paradigm in the field. One of the major difficulties one finds in this research is the lack of convincing observational data with which to compare one's theoretical effort. This is especially critical when addressing the onset of sexual reproduction itself, and in this case we can only speculate about this origin, but cannot test these speculations. A more promising line of research lies in analyzing the maintenance of sexual reproduction against the establishment of an apparently more efficient asexual variety. This efficiency stems from the absence of the necessity to generate males, which by itself alone should give the asexual variety the upper hand against a competing sexual alternative [2,3]. Explaining the evolutionary choice favoring sexual reproduction in spite of its shortcomings has been the long-term goal for a number of researchers in this area. It is not surprising that computer simulations have become a major tool in this field over the last few years. With the introduction of a number of simple simulational models, theoretical ideas began to be put to a " live " test on virtual populations. These models encompass dynamic rules that mimic the action of natural evolution in a variety of ways, and create a manageable representation of the conflict between selection and mutation, believed to be the driving evolutionary force. Among the models that have been used, we make special mention of the bit-string class, using Monte Carlo techniques [4–7]. They seem particularly well suited to represent the microscopic dynamics of genetic evolution and …