Real genomic representation for topopatric speciation
The evolutionary theory for speciation has produced several models to explain the diversity of life. Four of these models are called alopatric, peripatric, parapatric and sympatric, named after the predominant kind of genetic flux disturbance acting among the population. Physical barriers and ecological interactions are the two usual factors.
Recently, a new kind of speciation was proposed [Aguiar, 2009]. The model is based in selective mating determined by genetic affinity and spatial proximity. The model does not include any kind of geographical barrier, ecological interaction or natural selection. This new kind of speciation is named topopatric, as it emphasize the role of the spatial auto organization of the species origin and distribution. The usual approach is to represent a specimen as a binary string, and defying usual genetic algorithm with crossover and mutation operators, but without the selection phase. Differences between individuals reproductive rates are randomly attributed at reproduction time, rather than being due to any special ability.
We present a modified model with a real-coded genetic algorithm, where the specimens are represented by real numbers, with modified crossover and mutation operators. The reproductive rates are similar to the previous binary model. We studied the number of new species, the abundance of the species, and the distribution of the species over space in function of the genomic distance tolerance, the searchable radius for mating, the migration and the random reproductive rate. All simulations start with an initially uniform population, with the same real coded genetic algorithm crossover and mutation operators, and were performed in torus and ring like spatial setups. The results resemble the ones already reported in the literature, but are not strictly equivalent. We attribute this difference mainly to topological differences in the space of the genome representation.
Aguiar, M. A. M., et al. "Global patterns of speciation and diversity." Nature 460.7253 (2009): 384-387.