The Computational Mechanisms of Mate Choice


Sergio Castellano, Università di Torino, Dipartimento di Biologia Animale e dell’Uomo, Via Accademia Albertina, 13, 10123 Torino, Italy.


In many species, mating signals encode information important for both species recognition and mate quality assessment. I investigate how the computational mechanisms used by females to integrate the two sources of information can affect their mating decisions. First, I present a sequential analysis model of decision making based on a two-component signal, in which the first component encodes information important for mate quality assessment and the second component for species recognition. When the components interact additively, the ability of females to discriminate between signals that differ in only the component important for mate-quality assessment is the same independently of the value of the species-recognition component. In contrast, when the components interact multiplicatively, discrimination in mate quality depends also on the species-recognition component, which can either amplify or attenuate the perceived differences in mate quality. In the second part of the paper, I show results of a two-choice phonotaxis experiment on female Italian treefrogs, Hyla intermedia, that confirm the model predictions by showing that directional preferences for call rate (important for mate-quality assessment) are stronger when pulse-rate and fundamental frequency (important for species recognition) are close to the preferred population mean than when they are either higher or lower than the mean.