Time spent close to a sexual partner as a measure of female mate preference in a sex-role-reversed population of the blenny Salaria pavo (Risso) (Pisces: Blenniidae)

In most species females do not exhibit conspicuous sexual behaviours and female mate preferences are often measured by means of the time spent close to males. In spite of its widespread use, in only a few studies has this measure been validated as a reliable indicator of female mate preference. Sex-role-reversed species offer a better opportunity to test female preferences, as females usually court males. We tested in a sex-role-reversed population of the blenny Salaria pavo (Risso) if the time spent by females close to males related to the number of courtship displays directed towards males, and whether these measures of female preference reliably predicted matings. Females were simultaneously presented with two males behind a glass partition and the time spent close to each male was measured. We then allowed females to have access to the males' compartments and measured the courtship behaviours performed by females towards each male and recorded with which of the males females spawned. Females spent more time close to the male that subsequently received more courtship displays, and the preferred male had a more developed head crest. However, both measures of female preference failed to predict matings. Females were often attacked by males and probably had, in some occasions, to spawn with the less preferred male. Females that spawned with the previously preferred male had more swollen bellies than females that spawned with the less preferred male, suggesting a male preference towards more ripe females. These results validate the use of "time spent close to a sexual partner" as a measure of female preference in S. pavo. We argue that matings may not always reliably indicate mate preferences, particularly in the less choosy sex.