Behavioral Ecology and Sociobiology

, Volume 61, Issue 8, pp 1293–1302

Female preferences for multiple attributes in the acoustic signals of the Italian treefrog, Hyla intermedia


    • Dipartimento di Biologia Animale e dell’Uomo
  • Alessandra Rosso
    • Dipartimento di Biologia Animale e dell’Uomo
Original Paper

DOI: 10.1007/s00265-007-0360-z

Cite this article as:
Castellano, S. & Rosso, A. Behav Ecol Sociobiol (2007) 61: 1293. doi:10.1007/s00265-007-0360-z


The advertisement call of frogs and toads is an example of multiple message signal because different acoustic properties encode different kinds of biologically significant information. In the Italian treefrog, Hyla intermedia, pulse rate and frequency have been found to be under stabilizing female preferences and to encode information important for mate recognition, whereas the number of calls per call group have been found to be under directional preferences and, thus, to be important for mate quality assessment. In this study, we investigate preferences for calls that differ simultaneously in frequency, pulse rate, and number of calls per call group, and we ask how these properties interact with each other in influencing female mating decisions. Results of two-choice phonotaxis experiments provide no evidence to support the hypothesis that females process multi-attribute signals in a hierarchical way. In contrast, the pattern of preferences is consistent with the ‘preference function’ hypothesis, that is, with the hypothesis that females rank signals along an ordinal scale of values and choose accordingly. Pulse rate and frequency influence mating preferences more than does the number of calls per call group. The interaction between pulse rate and frequency is not additive but multiplicative: small differences in either pulse rate or frequency that, alone, have no effects on female choice, interact synergistically so that their combination has strong influence on female preferences. A preference repeatability test shows strong among-female differences in preference for multi-attribute signals. We suggest that this result reveals not only a variation in attribute values among females, but also a variation in the way females weight and combine attribute values into a single preference score.


Sexual selectionPlaybackFemale choiceMultiple messages

Copyright information

© Springer-Verlag 2007