A comparative study of Gaussian mating preference functions: a key element of sympatric speciation models

Abstract

Simulating the evolution of reproductive isolation under sympatric speciation scenarios is a complex process that requires modelling several phases, including evolution of phenotypes, demography, migration, fitness components and mating preference. The last has been shown to be a key parameter in several simulation studies, allowing the incorporation of assortative mating (premating isolation). Mating preference can be modelled by different mathematical functions but, as far as we know, a formal comparison of those functions has not yet been undertaken. In this work, we briefly review the main functions used in the literature and suggest a new one. In doing so, we also define three basic properties (monotonicity, proportionality and symmetry) that an ideal function should satisfy when generating assortative mating. We simulated several scenarios to compare how all these functions perform based on these properties. We also draw attention to the fact that the existing functions are affected distinctly by changing the scale of the preferred trait value. Some functions remain unaffected by scaling the trait, while in others assortative mating increases proportionally to the trait value. Most of the functions tested did not fulfil all the properties studied, and we find certain flaws in some of them that should be considered before being used in future studies. We provide some general recommendations for using the preference functions in simulation studies, and suggest that an unnoticed scaling effect could have underestimated the chance to obtain speciation under certain scenarios. © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 642–657.