Social Memory Formation Rapidly and Differentially Affects the Motivation and Performance of Vocal Communication Signals in the Bengalese Finch (Lonchura striata var. domestica)

Abstract

Cognitive processes like the formation of social memories can shape the nature of social interactions between conspecifics. Male songbirds use vocal signals during courtship interactions with females, but the degree to which social memory and familiarity influences the likelihood and structure of male courtship song remains largely unknown. Using a habituation-dishabituation paradigm, we found that a single, brief (<30 s) exposure to a female led to the formation of a short-term memory for that female: adult male Bengalese finches were significantly less likely to produce courtship song to an individual female when re-exposed to her 5 min later (i.e., habituation). Familiarity also rapidly decreased the duration of courtship songs but did not affect other measures of song performance (e.g., song tempo and the stereotypy of syllable structure and sequencing). Consistent with a contribution of social memory to the decrease in courtship song with repeated exposures to the same female, the likelihood that male Bengalese finches produced courtship song increased when they were exposed to a different female (i.e., dishabituation). Three consecutive exposures to individual females also led to the formation of a longer-term memory that persisted over days. Specifically, when courtship song production was assessed 2 days after initial exposures to females, males produced fewer and shorter courtship songs to familiar females than to unfamiliar females. Measures of song performance, however, were not different between courtship songs produced to familiar and unfamiliar females. The formation of a longer-term memory for individual females seemed to require at least three exposures because males did not differentially produce courtship song to unfamiliar females and females that they had been exposed to only once or twice. Taken together, these data indicate that brief exposures to individual females led to the rapid formation and persistence of social memories and support the existence of distinct mechanisms underlying the motivation to produce and the performance of courtship song.

Keywords: birdsong; courtship; directed song; songbird; vocal performance.

Figure 1

Experimental designs for 3X, 2X and 1X tests. For all tests, males were briefly (<30 s) exposed to individual females at 4–5 min intervals. Within each type of test, different symbols refer to different females.

Figure 2

Familiarity rapidly affects the motivation to produce courtship song. During 3X tests, the probability that a male produced courtship song (i.e., proportion of presentations in which a male produced courtship song) significantly decreased from the 1st to 2nd and 2nd to 3rd exposure to a female (n = 16 males). *p < 0.05.

Figure 3

Changes to song structure as a function of social context and familiarity. (A) Sequence durations (“seq. dur.”) were significantly shorter, song durations (“song dur.”) were significantly longer, mean fundamental frequency (FF; “mean FF”) and the number of introductory notes (“intro. notes”) produced before song were significantly higher, and the variability of FF (“CV of FF”) was significantly lower for courtship songs than for non-courtship songs. To facilitate data visualization and feature comparisons, we present the percent changes (from non-courtship song to courtship song) for all features except transition entropy. For transition entropy, the change (in bits) from non-courtship song to courtship song is plotted. (B) To assess the effect of familiarity on song structure, we compared the structure of courtship songs produced during the first exposure to a female (i.e., when males are more likely to produce courtship song to a female) to the structure of courtship songs produced during subsequent exposures to a female (i.e., when males are less like to produce courtship song to a female). Only courtship song duration significantly decreased from the first exposure to subsequent exposures. Plotted is the change in transition entropy and the percent changes for all other features from courtship songs produced on a male’s first exposure to individual females to courtship songs produced on subsequent exposures to females. For both (A,B), filled circles and open diamonds scale to the left y-axis, whereas filled squares scale to the right y-axis. Introductory notes and sequence and song durations were analyzed per bird, whereas the mean and CV of FF and transition entropy were analyzed per syllable or sequence (see “Results” Section). “*” Indicates p < 0.05.

Figure 4

Decreases in the production of courtship song reflect the formation of short-term memories. (A) Plotted are the proportion of male Bengalese finches (n = 16) that produced courtship song across each exposure to the first six stimulus females during 3X tests (range: 6–10 individual females). Dashed lines indicate when the identity of the stimulus female was changed. Numbers in parentheses indicate the number of males producing courtship song for each presentation. (B) The proportion of presentations in which males produced courtship song changed in a systematic manner across exposures (3X tests). Plotted is the mean change (±SEM) in the likelihood of courtship song (i.e., change in the proportion of presentations that an individual male produced courtship song) from the first to second exposure to a female (“1–2”), from the second to third exposure to a female (“2–3”), and from the third exposure to a female to the first exposure to a different female (“3–1”). “*” Indicates significantly less than zero (p < 0.05), and “#” indicates significantly greater than zero (p < 0.05) and different than “1–2” and “2–3”. (C) The proportion of males producing courtship song to individual females gradually decreased over presentations even when the identity of the female was changed after a single exposure (1X tests; 12 stimulus females). Numbers in parentheses indicate the number of males producing courtship song for each presentation. (D) We compared the rates of change in the proportion of males producing courtship song during 3X tests, in which males were consecutively exposed to the same female (presentations 1–3, 4–6, 7–9, and 10–12; black lines), to the rates of change across the same presentations during 1X tests, in which males were consecutively exposed to different females (gray lines). We found that the proportion of males producing courtship song decreased at a faster rate when males were exposed to the same female across presentations than when they were exposed to different females across presentations. For this analysis, only data for males that were administered both 1X and 3X tests were examined (n = 13 males).

Figure 5

Repeated exposures to females are important for the persistence of social memories over days (longer-term memory tests). (A) When tested 2 days after 3X tests, male Bengalese finches were significantly less likely to produce courtship song to familiar females than to unfamiliar females (n = 13 males; Table 2). Plotted are the average (±SEM) proportions of presentations with courtship song. (B) Two days after 3X tests, males produced shorter courtship songs to familiar females than to unfamiliar females (n = 7 males; i.e., only males that produced at least three courtship songs to familiar females and at least three courtship songs to unfamiliar females). (C) Two days after 1X tests, males did not differentially produce courtship song to unfamiliar and familiar females (n = 7 males). (D) Two days after 1X tests, males produced courtship songs to familiar and to unfamiliar females that were of similar duration (n = 6 males; i.e., only males that produced at least three courtship songs to familiar females and at least three courtship songs to unfamiliar females). (E) Two days after 2X tests, males did not differentially produce courtship song to unfamiliar and familiar females (n = 10 males). (F) Two days after 2X tests, males produced courtship songs to familiar and to unfamiliar females that were of similar duration (n = 8 males; i.e., only males that produced at least three courtship songs to familiar females and at least three courtship songs to unfamiliar females). For all panels, “*” indicates p < 0.05.