Male mate preferences in mutual mate choice: finches modulate their songs across and within male-female interactions

Abstract

Male songbirds use song to advertise their attractiveness as potential mates, and the properties of those songs have a powerful influence on female mate preferences. One idea is that males may exert themselves maximally in each song performance, consistent with female evaluation and formation of mate preferences being the primary contributors to mate choice. Alternatively, males may modulate their song behaviour to different degrees in the presence of different females, consistent with both male and female mate preferences contributing to mutual mate choice. Here we consider whether male Bengalese finches, Lonchura striata domestica, express mate preferences at the level of individual females, and whether those preferences are manifest as changes in song behaviour that are sufficient to influence female mate choice. We tested this idea by recording songs performed by individual unmated males during a series of 1 h interactions with each of many unmated females. Across recording sessions, males systematically varied both the quantity and the quality of the songs that they performed to different females. Males also varied their song properties throughout the course of each interaction, and behavioural tests using female birds revealed that songs performed at the onset of each interaction were significantly more attractive than songs performed by the same male later during the same interaction. This demonstration of context-specific variation in the properties of male reproductive signals and a role for that variation in shaping female mate preference reveals that male mate preferences play an important role in mutual mate choice in this species. Because these birds thrive so well in the laboratory and are so amenable to observation and experimentation across generations, these results yield a new model system that may prove especially advantageous in disentangling the role of male and female mate preferences in shaping mutual mate choice and its long-term benefits or consequences.

Keywords: Bengalese finch; Lonchura striata domestica; female mate preference; male mate preference; mutual mate choice; song entropy.

Figure 1

(a) Bengalese finch song consists of a series of notes, shown as a spectrogram: amplitude is reflected in the darkness of each note; grey boxes indicate song portions shown in (b) and (c). (b–c) The sequence in which notes are sung can be stereotyped in some portions and variable in other portions. For example, note 6 is always followed by note 7 in this and every song that we recorded from this bird, whereas note 3 can be followed by either note 4 or note 5. (d) We quantified the variability of note sequence in songs performed by each male throughout a series of recordings when the male was singing alone (undirected song, UNDIR) and when the male was singing to each of six different females (directed song, DIR).

Figure 2

Summary of note transitions performed by male BF39 during a 1 h recording in the (a) undirected state (UNDIR_ALL) and (b) directed state (DIR_ALL, one interaction between this male and one female bird). Grey boxes indicate transitions in this male’s song (all note transitions, including stereotyped sequences, were included in this analysis). (c) Comparison of note transition entropy of songs performed in the undirected and directed contexts for all males and all male–female interactions. (d) Songs performed at the onset of each male–female interaction (DIR_EARLY) were typically quite different from those performed in the undirected context. This panel depicts a DIR_EARLY song performed by the same male depicted in (a) and (b). As elaborated in the text, note transitions present in DIR_EARLY songs were more stereotyped than those in UNDIR_ALL and DIR_ALL songs (this bird is one of 15 males quantified in (c)).

Figure 3

Comparison of note transition entropy for songs performed at the onset of each male–female interaction (DIR_EARLY) and (a) throughout the undirected context (UNDIR_ALL) and (b) by the same birds later during each 1 h male–female interaction (DIR_LATE). Dashed line indicates the line of identity in each panel.

Figure 4

Entropy values across (a) DIR_EARLY and (b) DIR_LATE songs performed by each male to each of several different females (all females that passed the criterion for inclusion in our data set are shown for each male). Dotted line indicates no change from the entropy detected in undirected songs; each point represents one male–female interaction. (c) Difference in entropy values of DIR_EARLY and DIR_LATE songs performed to each female during each 1 h interaction (each point represents one male–female interaction; X axis is normalized by the maximum number of songs that each male performed in any male–female interaction; solid line represents linear regression). (d) Mean ± SE percentage of normalized entropy of directed songs during the early and late portion of each male–female interaction. Dotted line indicates no change from the entropy detected in undirected songs. *P <0.05. Results for the difference in entropy values between the DIR_LATE and UNDIR_ALL conditions are given in Supplementary Table S2.

Figure 5

Mean ± SE differences in the degree of song modulation between DIR_EARLY and DIR_{LATE_3} stimuli. Filled symbols: females that expressed a significant preference; open symbols: females that expressed no significant preference.