Behavioural Processes 70 (2005) 271–281

Call-based species recognition in black-capped chickadees Isabelle Charrier ∗ , Christopher B. Sturdy Department of Psychology, University of Alberta, Edmonton, Alta., Canada T6G 2E9 Received 16 March 2005; received in revised form 28 July 2005; accepted 28 July 2005

Abstract Species recognition is essential for efficient communication between conspecifics. For this to occur, species information must be unambiguously encoded in the repertoire of each species’ vocalizations. Until now, the study of species recognition in songbirds has been focused mainly on male songs and male territorial behaviour. Species recognition of other learned vocalizations, such as calls, have not been explored, and could prove useful as calls are used in a wider range of contexts. Here, we present an experimental field study investigating the coding of species information in a learned vocalization, the ‘chick-a-dee’ call of the black-capped chickadee (Poecile atricapillus). By modifying natural calls in both temporal and spectral domains and by observing the vocal responses of black-capped chickadees following the playback of these modified calls, we demonstrate that species recognition in chickadees relies on several acoustic features including syntax, frequency modulation, amplitude modulation, and to a lesser extent, call rhythmicity and frequency range. © 2005 Elsevier B.V. All rights reserved. Keywords: Acoustic communication; Coding–decoding; Species recognition; Chickadee

1. Introduction Animals use acoustic signals in a wide range of contexts including territorial defense, mate selection, maintenance of social bonds, individual recognition, foraging and alarm. Animals mainly direct their acoustic signals towards conspecifics (Marler, 1957; Kroodsma and Miller, 1996). Species recognition is essential to allow animals to effectively communicate with conspecifics and establish social links. Because ∗ Corresponding author. Tel.: +33 1 69 15 68 26; fax: +33 1 69 15 77 26. E-mail address: [email protected] (I. Charrier).

0376-6357/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.beproc.2005.07.007

of this necessity, information must be unambiguously encoded in a species’ acoustic signals, such that each species develops distinctive, species-specific characters in their vocal repertoire. Species recognition has been well-studied in songbirds. Most research has focused on male songs as they are often used in species-specific interactions, and are involved in clear functions such as territorial defense and mate attraction (for reviews see Becker, 1982; Falls, 1992; Catchpole and Slater, 1995). Furthermore, most field playback studies involve observing the territorial behaviour of males upon the presentation of conspecific song playback because males respond strongly to other males’ songs whereas females’ responsive-

272

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

ness is relatively weak (for review, see Becker, 1982). Species recognition via calls has not been as rigorously studied despite their importance in mediating social behaviours (Catchpole and Slater, 1995) and despite the fact that calls can be complex in acoustic structure. Indeed, in songbirds calls are typically innate, shorter and less complex than songs and used in a wider range of contexts. In contrast, songs are typically learned, composed of different elements and used in two main functions of territory defence and breeding (for review, see Marler, 2004). Here, we investigate chick-a-dee call-based species recognition in black-capped chickadees (Poecile atricapillus). The black-capped chickadee is a small North American songbird that possesses a large, well-studied vocal repertoire (for review, Hailman and Ficken, 1996). Two of their most well-studied vocalizations, the ‘feebee’ song and the ‘chick-a-dee’ call, are both learned (Ficken et al., 1978; Ratcliffe and Weisman, 1985; Hughes et al., 1998; Hailman and Ficken, 1996). The song, mainly produced by males, is used to attract and retain a mate and to defend a territory during the breeding season. The call (Ficken et al., 1978; Hailman et al., 1985), produced by both males and females throughout the year, is thought to serve a variety of functions such as raising mild alarm, maintaining contact between mates and coordinating flock activities. In contrast to many songbird species, black-capped chickadee song is significantly less acoustically complex than their ‘chick-a-dee’ call. The fee-bee song consists of two whistled notes that are sung at a constant pitch interval relative to each other (Weisman et al., 1990; Kroodsma et al., 1999). By comparison, chick-a-dee calls contain four note types, A, B, C and D, sung in a fixed order (A → B → C → D), but note types can be repeated or omitted, producing chick-adee calls with seemingly infinite combinations of notes (e.g., AACDDD, ABDDD; Ficken et al., 1978). Moreover, the number of notes and note type occurrence in chick-a-dee calls change with context, potentially signifying different “meanings” for calls with different note composition (Hailman, 1989). Analyses of chicka-dee calls have revealed that the different note types potentially convey information regarding the individual identity and gender of the caller (Charrier et al., 2004). Furthermore, previous playback studies have shown that chickadees are able to discriminate the calls of their own flock from the calls of foreign flocks, with

birds calling more in response to playback of foreign flocks’ calls (Nowicki, 1983). This flock discrimination is apparently mediated by D-notes (Mammen and Nowicki, 1981), whereas C notes may indicate the location or availability of food sources (Freeberg and Lucas, 2002). The coding of species recognition has been investigated in black-capped chickadee fee-bee songs (Ratcliffe and Weisman, 1985; Weisman and Ratcliffe, 1989; Weisman et al., 1990). These studies have shown that the structure and syntax of song elements are important, but the pitch interval between ‘fee’ and ‘bee’ is essential for species recognition. To date no experimental study has been performed on the species coding in chick-a-dee calls in spite of the fact that this vocalization is used throughout the year and in a wider range of social interactions than is the fee-bee song. Since chick-a-dee calls have an acoustic structure that is markedly different than fee-bee song (as described above) the coding of species recognition in these two classes of signals will likely rely on different acoustic features. In the present study, we determined some of the acoustic features in the chick-a-dee call that are used by black-capped chickadees in recognizing their species. On the basis that black-capped chickadees respond more strongly to conspecific calls than to heterospecific calls (Kershner and Bollinger, 1999), and respond more strongly to foreign flocks’ calls than to calls of their own flock (Nowicki, 1983), we broadcast both normal and modified chick-a-dee calls in which one given acoustic parameter was modified. We then compared the vocal responses obtained with modified calls to those obtained with natural calls. From this, we determined which acoustic features are used in the species recognition process. Finally, we discuss the effectiveness of the species’ vocal signature in relation to the ecological and environmental constraints of black-capped chickadees.

2. Materials and methods 2.1. Study location This study was carried out on black-capped chickadees at two different locations in the River Valley forest in Edmonton (Alta., Canada, 53◦ 34 N–113◦ 25 W) between 26 November and 19 December 2003,

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

from 830 to 1130 MST. Within each location we chose three sites away from permanent feeders and with each site being separated from the next nearest site by at least 250 m to maximize the independence of studied sites (Freeberg and Lucas, 2002; Clucas et al., 2004). 2.2. Recordings and signal acquisition We selected six natural chick-a-dee calls recorded from six different birds (three males and three females) captured during the winter of 2002, approximately 18 months prior to the current experiments, in a location adjacent to the area in which playback experiments were performed. All calls contained each of the four note types (i.e., A–D notes) found in black-capped chickadee calls to minimize any potential response biases based on note composition (Clucas et al., 2004). Birds were recorded following the methods described in Charrier et al. (2004). Gray-crowned rosy finch (Leucosticte tephrocotis) calls were recorded in California in May 1999 with a Sennheiser ME66 short shotgun microphone connected to a Sony SME Modified TCM-5000EV Bird Version cassette recorder (extended high-frequency response, Saul Mineroff Electronics Inc., USA). Calls were digitized at 44.1 kHz 16-bit samples per second using Canary v1.2 (Cornell Lab of Ornithology, Ithaca, NY, USA). All natural calls were band-pass filtered (1.150– 10 kHz) to remove background noise, and amplitudes among calls were equalised with respect to peak values (average RMS = −15dB, peak value = −1.9 dB). 2.3. Acoustic stimuli 2.3.1. Control signals Our positive control signals were six different black-capped chickadee chick-a-dee calls, consisting of between five and seven notes (i.e., the average number of notes in chick-a-dee calls; Fig. 1A). Our negative control signals were a series of high quality calls from three different gray-crowned rosy finches, as described above, with each series composed of five to eight ‘chew’ calls. We chose the gray-crowned rosy finch as a negative control because its calls are acoustically distinct from those of the black-capped chickadee (Fig. 2). Our rationale was that responses elicited from rosy finch calls would provide a good

273

basis for comparison with responses to chickadee calls as a test of the species-specificity of the black-capped chick-a-dee call. Moreover, black-capped chickadees at this location are not familiar with calls of gray-crown rosy finch (the latter being found in alpine regions). Because of this, gray-crowned rosy finches, therefore, are not competitors and do not represent a known threat to chickadees. We specifically chose gray-crowned rosy finches’ chew calls as negative controls to avoid using calls from either a known competitor (such as a nuthatch) or a known predator (such as a magpie) as using such signals may have confounded our results with naturally-occurring aggressive responses to heterospecific signals. 2.3.2. Experimental signals Using the six natural chick-a-dee calls, we created 11 different types of experimental signals by either modifying one of their characteristics (syntax, frequency or temporal features), or by synthesizing novel signals (Fig. 1). Natural calls were modified using SIGNAL (Engineering Design, 2003), SYNTANA (Aubin, 1994) and GOLDWAVE (Craig, 1996) for a total of 66 experimental signals. 2.3.3. Modifications in the frequency domain We created two types of noise signals, one broadband noise signal (bns) (Fig. 1B) and one filtered noise signal (fns) (Fig. 1C). To create the bns, we synthesized white noise that was band-pass filtered between 1150–10,000 Hz in the same manner as the natural stimuli. We then edited ‘notes’ from the white noise equal in duration to the notes found in the six natural calls that we were emulating. We then added silence between the noise ‘notes’ that equalled the silent internote intervals found in the natural calls. These broadband noise signals did not contain frequency modulation or temporal cues (such as FM, AM), but did contain a natural call rhythmicity. To create the fns signals, we used the broadband noise signals (described above) and digitally filtered each noise ‘note’ so that its frequency bandwidth matched the original frequency range of each note found in the natural call upon which the fns signal was based. These manipulations resulted in the fns retaining a natural call rhythmicity and frequency range for each note of the emulated call but without any frequency or amplitude modulation cues.

274

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

Fig. 1. Spectrograms of the different experimental signals used in the playback experiments (FFT window size = 512 pts; frequency resolution = 86 Hz). (A) Positive control (call composition: AABCDDD), (B) broadband noise signal, bns, (C) filtered noise signal, fns, (D) call shifted down by 1 kHz, −1000, (E) call shifted up by 1 kHz, +1000, (F) call with reversed syntax, dcba, (G) call with time-reversed note, revnote, (H) call without AM, wo AM, (I) call with inter-note silence shortened by 2, sil div2, (J) call with inter-note silence lengthened by 2sil, mult2.

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

Fig. 2. Spectrogram of two typical calls produced by the graycrowned rosy finch.

We created four additional types of experimental signals using SYNTANA in which the entire frequency spectrum was linearly shifted up or down: (1) +500, signal shifted up by 500 Hz, (2) +1000, signal shifted up by 1 kHz (Fig. 1E), (3) −500, signal shifted down by 500 Hz, (4) −1000, signal shifted down by 1 kHz (Fig. 1D). These manipulations were carried out on the entire signal by selecting a data record using a square window, applying short-term overlapping (50%) fast Fourier transform (FFT), followed by a linear shift (+ or −) of each spectrum and a short-term inverse fast Fourier transform (FFT−1 ; Randall and Tech, 1987). In each case, we used a window size of 4096 points (frequency precision = 10 Hz). In these experimental signals both natural amplitude and temporal structure (AM and FM) remained unchanged. Our linear shift values were chosen on the basis of our previous bioacoustic analyses on black-capped chickadee calls (Charrier et al., 2004). Measurements of start, peak and end frequencies on different notes types revealed that these features had standard deviations ranging from 120 to 820 Hz across all four note types. Based on these extensive bioacoustics measurements made on our studied population, we selected a mean value for frequency shifts that was well within this range (500 Hz) and one that was well outside of this range (1000 Hz). 2.3.4. Modifications in the time domain We created one type of experimental signal (dcba) in which call syntax was modified by reversing the note order within the call, thus modifying the rhythmicity of the call while maintaining frequency and temporal characteristics (Fig. 1F). We also created two other types of experimental signals: one in which we modified the local pattern of frequency and amplitude modulations, note-

275

by-note (revnote) (Fig. 1G), and the other in which the amplitude modulation was removed without modification of frequency modulation (wo am) (Fig. 1H). To create revnote experimental signals, each note in each call was time-reversed, without altering the natural syntax of the call. To create wo am experimental signals, we removed amplitude modulation from calls by using the analytical signal calculation (MbuNyamsi et al., 1994) performed in SYNTANA. This results in a signal with a normal FM and duration but without AM. The final two types of signals were created by modifying the natural rhythmicity of the call. To do this, we either shortened (Fig. 1I) or lengthened (Fig. 1J) the silent, inter-note intervals by a factor of 2 (sil div 2 = inter-note intervals divided by two; sil mult 2 = inter-note intervals multiplied by two). Both manipulations created calls with inter-note intervals that are outside the species-typical range of our population (46 ± 15 ms (S.D.); analysis made on 100 calls from 10 birds, 580 silences; Charrier unpublished data). 2.4. Playback procedure Signals were broadcast with a Sony D-SJ301 CD player connected to an Audix PH3s powered speaker (25W; frequency response = 0.1–20 kHz ± 10 dB). The speaker was placed in a tree at about 2 m from the ground, and the experimenter observed and recorded vocal responses a few meters from the speaker. All signals were played back at a natural sound pressure level (SPL = 82 ± 2 dB at 1 m measured with a Radioshack digital sound level meter, A weighting, slow response, TX, USA). This intensity level is similar to that of a call recorded at 1 m in the laboratory. Each experimental series was composed of modified or synthetic chick-a-dee calls, whereas the control series were composed of non-modified chick-a-dee calls from black-capped chickadee (positive control) and natural chew calls from gray-crowned rosy finches (negative control). Experimental sessions started with a 90-s preplayback observation period (baseline) followed by a 90-s playback test consisting of the playback of one series. Each series, either experimental or control, was composed of three ‘calls’ (one call every 10 s) followed by 1 min of silence, so that each series lasted a total of

276

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

90 s. To allow birds to return to a baseline vocal activity, we waited for at least 4 min following each playback test before conducting the next pre-playback observation and playback test. Each experimental session consisted of presenting six different series, four experimental series and two control series. One of each control series (positive and negative) was always played back during each session to ensure that responses were species-specific. The logic for this is as follows: if species recognition occurs, birds should strongly respond to the positive control and weakly respond to the negative control, as found in a previous playback study involving chickadees (Kershner and Bollinger, 1999). If the birds did not respond at a level above baseline to the positive control signal, the session was excluded from our analysis. This did not occur in any of our playbacks. Sites were visited no more than five times with at least 2 days (range = 2–10 days) between each visit and we used a different positive control series for each visit. Each type of experimental signal was presented 10 times across the six sites resulting in any given experimental signal being played either once or twice on each particular site. If the experimental signal type was played twice at a particular site, two different versions were used. To avoid habituation and to reduce any potential order effects, the presentation order of the six different series was randomly selected. The testing order of the three sites was also changed between days. Considering all these different precautions, and the fact that birds are highly territorial and very active during winter mornings in search of food, it was unlikely that a same flock has been tested at more than one site. 2.5. Behavioural response measures Since the study was conducted on a non-banded population, we tallied only the vocal behaviour of the birds to quantify the birds’ response. Total vocal production is a behavioural measurement that has previously been used in playback experiments on chickadees (Nowicki, 1983; Freeberg and Lucas, 2002; Clucas et al., 2004). Typically, only one given winter flock was tested at a time, with tallied calls produced by, on average, four to six birds, with flock size being stable over winter (Dhondt and Lowe, 1995). We counted every audible call produced by the particular chickadee flock present

during the 90 s pre-playback period (baseline) and during the 90 s playback period, and used this as a measure of responsiveness to a particular signal. 2.6. Statistical analyses We analyzed the difference in call rate between playback and pre-playback (baseline) sessions for playbacks of natural chick-a-dee calls (positive control) and each of the other playback types (negative control and all experimental types) using a multivariate analysis of variance (MANOVA) integrating the type of experimental series and the location site. This was followed by a post hoc two-tailed Dunnett’s test (Atil and Unver, 2001; Rafter et al., 2002) to detect differences in response between positive control and experimental signals (n = 6 for each comparison). We performed the MANOVA on responses averaged for each site and for each experimental signal (n = 6, each experimental signal was tested once or twice on each site) since an eventual non-independence of some playback sessions may occur. Statistical analyses were made using Statistica Version 6 (Statsoft Inc., 1984–2003).

3. Results A total of 28 playback sessions were performed across the six different sites. The results of the omnibus MANOVA on averaged responses showed that birds’ vocal behaviour was significantly different following playback of the 13 different types of stimuli (i.e., 11 experimental signals with 10 presentation each, 2 controls with 28 presentations each; F = 3.036, d.f. = 12, p < 0.002), and that there was no effect of location site (F = 0.905, d.f. = 5, p = 0.484). Following this, we conducted post hoc analyses (Dunnett’s test) to determine the source of these differences. 3.1. Control signals Black-capped chickadees responded strongly to their own species’ calls whereas they responded very weakly to gray-crowned rosy finch calls. The playback of black-capped chickadee chick-a-dee calls elicited an increase in their calling rate from baseline (positive control: 7.2 ± 5.2 calls/min, ranging from −1.3 to 22.7 calls/min, n = 28, Fig. 2) whereas the playback of gray-

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

277

Fig. 3. Modification of the calling rate (calls/min) obtained with the control signal and with the different experimental signals. To assess the modification of the birds’ calling rate, we subtracted the calling rate observed during the baseline from the calling rate obtained during the playback session, and we compared this baseline-corrected calling rate to those obtained with control signals. For abbreviations, for Fig. 1 legend.

crowned rosy finch calls led to a decrease in their calling rate from their baseline (negative control: −0.05 ± 5.8 calls/min, ranging from −17.3 to 7.3 calls/min, n = 28 (Fig. 3); Dunnett’s test, n = 6, p < 0.02). 3.2. Modifications in the frequency domain Vocal activity in response to both types of noise signals (i.e., bns and fns) was significantly different from elicited by the positive control signals (n = 6, ps < 0.02). The presentation of either type of noise signals rarely elicited calls, and in fact, often significantly decreased the calling rate below baseline levels (bns signal = −15 to 5 calls/min; fns = −12 to 0 calls/min; see Fig. 3). Interestingly, black-capped chickadees respond differentially to signals that are either positively or negatively shifted in frequency (Fig. 3). Indeed, responses to both the +500 and the +1000 Hz signals were not significantly different from responses to positive con-

trol signals (n = 6, p = 0.27 and p = 0.53, respectively) in spite of the fact that positively shifted signals led to an increase above baseline in calling rate (mean ± S.D.; 3.3 ± 6.6 and 3.3 ± 5.5 calls/min for +500 and +1000, respectively). In contrast, responses obtained following the presentation of the −500 and −1000 Hz signals were weaker than responses to the positively shifted signals (3.2 ± 5.5 and 1.3 ± 5.6 calls/min for the −500 and −1000, respectively). Only responses to the −1000 Hz signal were significantly different (less calling) from those elicited by the positive control signals (n = 6, p = 0.044) whereas responses to the −500 Hz signal were not significantly different from those to the positive control signals (n = 6, p = 0.153). This suggests that the allowable frequency deviations (i.e., those that will result in a bird strongly responding to the call) in black-capped chickadees are asymmetrical, and that similar decreases and increases in call pitch are not perceived as functionally equivalent.

278

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

3.3. Modifications in the time domain Calls in which the syntax was reversed (i.e., dcba) induced significantly less calling behaviour than positive control signals (n = 6, p < 0.01). Chickadees rarely responded to these syntax-reversed signals, and even tended to decrease their vocal output relative to baseline levels (calling rate range = −8 to 4 calls/min). The removal of amplitude modulation or the modification of frequency modulation patterns also significantly altered species recognition (Fig. 3). Vocal responses to both types of experimental signal were significantly lower than to positive control signals (n = 6, ps < 0.003 for both amplitude and frequency modulation), with birds significantly decreasing their call production following the playback of these signal types (calling rate range = −10.7 to 4.7 calls/min for revnote and −5.3 to 4.7 calls/min for wo am). Finally, black-capped chickadees responded differentially to the two modifications in call rhythmicity. Chickadees did not respond in a statistically different manner to calls in which the rhythmicity was accelerated (calling rate increased compared to the baseline, 2.2 ± 3.6 calls/min) in comparison to that observed in response to positive control signals (sil div2, n = 6, p = 0.599). In contrast, the vocal response was significantly lower when call rhythmicity was decelerated (sil mult2, n = 6, p < 0.01), with birds decreasing their call production relative to the baseline period following the presentation this signal (−0.13 ± 7.2 calls/min; range = −17.3 to 9.3, see Fig. 2). In addition, we found that all experimental series that elicited significantly less responding than the positive control were not significantly different from those obtained with the negative control (Dunnett’s test, pvalues ranging from 0.83 to 1).

4. Discussion Here we show that black-capped chickadees respond more vigorously to calls of their own species than to calls of gray-crowned rosy finches, demonstrating that they are able to discriminate between calls of their own and other species. This is consistent with other field studies showing similar species recognition in black-capped chickadees (Smith, 1993; Kershner and Bollinger, 1999). By playing back modified calls to

black-capped chickadees, we are able to extend these initial findings and determine which acoustic features play a role in species recognition. 4.1. Importance of call rhythmicity to species recognition Neither signal composed of white noise (i.e., bns or fns) elicited species recognition in spite of the fact that several parameters were similar to normal black-capped chickadee calls. This indicates that call rhythmicity alone, or call rhythmicity with the natural frequency range of the different note types, are not sufficient to elicit species recognition, suggesting that more information is necessary than these two parameters. This is not to say that rhythmicity does not play any role in species recognition. Birds respond strongly to calls with an accelerated rhythmicity (sil div2) and respond weakly to calls with a decelerated rhythmicity (sil mult2). Thus, birds are differentially-sensitive to the call rhythmicity, depending on the kind of modification. Indeed, if slower call rhythmicity is considered unnatural in black-capped chickadees, a faster rhythmicity is considered more natural. In bird species, note duration or inter-note intervals does not encode species information (for review see Becker, 1982; Mathevon and Aubin, 2001), whereas in other species, this feature is relatively important (Becker, 1982; Dabelsteen and Pedersen, 1992, 1993; Slabbekoorn and ten Cate, 1999). Signals in which the inter-note intervals were shortened elicited a stronger behavioural response than those in which the silences were lengthened. This may be explained by the fact that inter-note intervals measured in black-capped chickadee chick-a-dee calls are, on average, 46 ± 15 ms. Therefore, halving the silences results in modified calls that are closer to the natural inter-note interval duration range compared to doubling the silences, which shifts the modified calls out of this natural range. In practical terms, if the first note in a call is not quickly followed by the second note, call rhythmicity is sufficiently disrupted. This may cause birds to not consider all subsequent notes as components of that call, but rather, as separate elements, ultimately impairing species recognition. A more likely explanation is that accelerated calls are often considered as ‘super-stimuli’ that increase excitement and have been shown to induce more reaction (Becker, 1982).

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

4.2. Importance of call syntax to species recognition Syntax is very important for species recognition in black-capped chickadees. Indeed, when calls with reversed note orders are played (i.e., D → C → B → A), birds respond very weakly. This is similar to previous studies showing that a signal with a modified syntax does not elicit species recognition (Becker, 1982; Ratcliffe and Weisman, 1985; Kreutzer and Br´emond, 1986; Holland et al., 2000; Nelson, 1988) although the necessity of syntax to species recognition is not universal among songbirds (Beletsky et al., 1980; Searcy, 1990; Dabelsteen and Pedersen, 1993). Even if there are a great number of chick-a-dee call variants, including call combinations in which some note types are omitted, the normal note order of A → B → C → D (Hailman et al., 1985) is essential for accurate species recognition. It remains untested, and thus unclear, whether more subtle deviations in syntax (e.g., A → C → B → D) will similarly disrupt species recognition. 4.3. Importance of call pitch to species recognition Black-capped chickadees respond strongly to calls shifted upwards in pitch by either 500 or 1000 Hz, but respond less strongly to calls reduced in pitch by the same values. The effects of the downward frequency shifts are consistent with studies performed on female blackbirds (Dabelsteen and Pedersen, 1993) and field sparrows (Nelson, 1989). In a previous study (Charrier et al., 2004), we measured the different frequency values for A, B and C notes, as well as the value of the fundamental frequency for D notes. On the basis of this analysis, if a chick-a-dee call is shifted up or down by 500 Hz (the highest S.D. for these frequency features was 820 Hz), most of the frequency features are still within the natural frequency range. This is not the case when shifting up or down by 1000 Hz. However, black-capped chickadees were more responsive to higher pitched calls than lower pitched calls, so this asymmetry might be due to the limits of the audible frequency range in black-capped chickadees, assuming their audibility curves are similar to those available for other Paridae species (Dooling, 1982; Langemann et al., 1998). We could explain this difference by the fact that very low-pitched calls are considered as pro-

279

duced by a “super-dominant” chickadee male. However, because of their small size and the constraints of anatomy on sound production (Kelemen, 1963), chickadees may not be able to produce such low-pitched calls, which may explain why these calls are not treated as natural calls by black-capped chickadees. Still a third explanation for this asymmetry in responding to linear pitch modification could involve the incidental modification of relative pitch cues. Alterations in relative pitch within a signal are a by-product that accompanies linear frequency shifts, and relative pitch is a feature known to be important in song-based species recognition in black-capped chickadees (Weisman and Ratcliffe, 1989). It remains possible that the influence of our linear shifts on relative pitch cues contributed to our results, and could be investigated in future studies. 4.4. Importance of FM and AM to species recognition Finally, our playback experiments showed that signals with modified amplitude and frequency modulation pattern or with no amplitude modulation were not recognised by chickadees as natural calls, thus implicating both frequency and amplitude modulation in species recognition. This is consistent with previous studies which have shown that frequency modulation is often used to encode species or individual identity (Becker, 1982; Mathevon and Aubin, 2001; Aubin and Jouventin, 2002; Charrier et al., 2001). However, our results are inconsistent with those examining the role of amplitude modulation in species recognition. Amplitude modulation cues are highly degraded during propagation through dense vegetation (Wiley and Richards, 1982), and because of this, are usually only used to locate the emitter and to estimate distances (Mathevon, 1998; Saberi et al., 1999; Nelson, 2000; Naguib and Wiley, 2001). That said, our results with the amplitude-modulated signals are consistent with a study on blackbirds showing that, in blackbirds as in black-capped chickadees, AM is a very important feature in species recognition (Dabelsteen and Pedersen, 1992, 1993). Based upon the results presented here, the acoustic features most essential to species recognition in blackcapped chickadees appear to be syntax, frequency and amplitude modulation, and to a lesser extent, call rhythmicity and frequency range. The use of multiple acous-

280

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281

tic features in species recognition has been shown in several species (Becker, 1982; Nelson, 1988; Holland et al., 2000), and may aid chickadees’ species recognition within the environmental constraints of their habitat. Degradations of their calls during propagation and vocal overlap with others birds may impair the perception of some call features. If species information is encoded in several parameters, and one or more parameters are degraded under some conditions, then the other un-degraded parameters may provide birds with a reliable species recognition mechanism. An examination of the influence that social context has on species recognition in black-capped chickadees, recently shown to be critical in the social zebra finch (Taeniopygia guttata; Vignal et al., 2004), could add to our understanding of how social animals regulate their overt behaviour (i.e., vocal output) as a function of their social context. Indeed, we could investigate the importance of call note composition to vocal responsiveness. In which social context (i.e., foraging, mild alarm, etc.) do black-capped chickadees will be more sensitive to a particular type of calls (i.e., calls mainly composed of A and D notes, calls with a lot C notes, etc.).

Acknowledgements CBS was supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery grant, an Alberta Ingenuity Fund (AIF) New Faculty Grant, a Canada Foundation for Innovation (CFI) New Opportunities Grant, along with start-up and CFI Partner Funding from the University of Alberta. IC is funded by an Izaak Walton Killam Memorial Trust Postdoctoral Fellowship and an Alberta Ingenuity Postdoctoral Fellowship. This research was approved by the University of Alberta Biological Sciences Animal Care Committee (protocol number 351201) and the University of Calgary Life and Environmental Sciences Animal Care Committee (protocol number BI2001-028). Birds were captured under an Environment Canada Canadian Wildlife Service Scientific Permit (permit number WSA-1-02) and Alberta Fish and Wildlife Research Permits (numbers 4619 GP, 4621 GP and 8734 GP) and Collection Licenses (numbers 088 CN, 089 CN, and 147 CN). We warmly thank Scott A. MacDougall-Shackleton for providing us gray-crowned rosy finch calls. Special thanks to Lau-

rie Bloomfield, Andrew Iwaniuk, Tiffany Lee and three anonymous referees for their helpful comments on the manuscript.

References Atil, H., Unver, Y., 2001. Multiple comparisons. J. Biol. Sci. 1, 723–727. Aubin, T., 1994. Syntana: a software for the synthesis and analysis of animal sounds. Bioacoustics 6, 80–81. Aubin, T., Jouventin, P., 2002. How to vocally identify kin in a crowd: the penguin model. Adv. Study Behav. 31, 243–277. Becker, P.H., 1982. The coding of species-specific characteristics in bird sounds. In: Kroodsma, D.E., Miller, E.H. (Eds.), Acoustic Communication in Birds, vol. 1. Academic Press, New York, pp. 213–252. Beletsky, L.D., Chao, S., Smith, D.G., 1980. An investigation of song-based species recognition in the red-winged blackbird (Agelaius phoeniceus). Behaviour 73, 189–203. Catchpole, C.K., Slater, P.J.B., 1995. Bird Song: Biological Themes and Variations. Cambridge University Press, Cambridge. Charrier, I., Bloomfield, L.L., Sturdy, C.B., 2004. Note types and coding in Parid vocalizations I: the chick-a-dee call of the black-capped chickadee (Poecile atricapilla). Can. J. Zool. 86, 769–779. Charrier, I., Mathevon, N., Jouventin, P., Aubin, T., 2001. Acoustic communication in a Black-headed Gull colony, how chicks identify their parents? Ethology 107, 964–974. Clucas, B.A., Freeberg, T.M., Lucas, J.R., 2004. Chick-a-dee call syntax, social context, and season affect vocal responses of Carolina chickadees (Poecile carolinensis). Behav. Ecol. Sociobiol. 57, 187–196. Craig, C.S., 1996. Goldwave Version 3.22. Dabelsteen, T., Pedersen, S.B., 1992. Song features essential for species recognition and behaviour assessment by male blackbirds (Turdus merula). Behaviour 121, 259–287. Dabelsteen, T., Pedersen, S.B., 1993. Song-based species discrimination and behaviour assessment by female blackbirds, Turdus merula. Anim. Behav. 45, 759–771. Dhondt, A.A., Lowe, J.D., 1995. Variation in black-capped chickadee group size. Birdscope 9 (1). Dooling, R.J., 1982. Auditory perception in birds. In: Kroodsma, D.E., Miller, E.H. (Eds.), Acoustic Communication in Birds, vol. 1. Academic Press, New York, pp. 95–130. Engineering Design, 2003. SIGNAL, Engineering Design, Belmont, MA, USA. Falls, J.B., 1992. Playback: a historical perspective. In: McGregor, P.K. (Ed.), Playback and Studies of Animal Communication. Plenum, New York, pp. 11–33. Ficken, M.S., Ficken, R.W., Witkin, S.R., 1978. Vocal repertoire of the black-capped chickadee. Auk 95, 34–48. Freeberg, T.M., Lucas, J.R., 2002. Receivers respond differently to chick-a-dee calls varying in note composition in Carolina chickadees (Poecile carolinensis). Anim. Behav. 63, 837– 845.

I. Charrier, C.B. Sturdy / Behavioural Processes 70 (2005) 271–281 Hailman, J.P., 1989. The organization of the major vocalizations in the Paridae. Wilson Bull. 101, 305–343. Hailman, J.P., Ficken, M.S., 1996. Combinational animal communication with computable syntax: chick-a-dee calling qualifies as ‘language’ by structural linguistics. Anim. Behav. 34, 1899–1901. Hailman, J.P., Ficken, M.S., Ficken, R.W., 1985. The ‘chick-a-dee’ calls of Parus atricapillus: a recombinant system of animal communication compared with written English. Semiotica 56, 191–224. Holland, J., Dabelsteen, T., Paris, A.L., 2000. Coding in the song of the wren: importance of rhythmicity, syntax and element structure. Anim. Behav. 60, 463–470. Hughes, M., Nowicki, S., Lohr, B., 1998. Call learning in blackcapped chickadees (Parus atricapillus): the role of experience in the development of ‘chick-a-dee’ calls. Ethology 104, 232–249. Kelemen, G., 1963. Comparative anatomy and performance of the vocal organ in vertebrates. In: Busnel, R.G. (Ed.), Acoustic Behaviour of Animals. Elsevier, Amsterdam, pp. 489–521. Kershner, E.L., Bollinger, E.K., 1999. Aggressive response of chickadees towards black-capped and Carolina chickadee calls in central Illinois. Wilson Bull. 111, 363–367. Kreutzer, M., Br´emond, J-C., 1986. Les effets additifs de la syntaxe et de la forme des syllabes lors de la reconnaissance sp´ecifique chez le Troglodyte (Troglodytes troglodytes). Can. J. Zool. 64, 1241–1244. Kroodsma, D.E., Byers, B.E., Halkin, S.L., Hill, C., Minis, D., Bolsinger, J.R., Dawson, J., Donelan, E., Farrington, J., Gill, F.B., Houlihan, P., Innes, D., Keller, G., Macaulay, L., Marantz, C., Ortiz, J., Stoddard, P.K., Wilda, K., 1999. Geographic variation in black-capped chickadee songs and singing behaviour. Auk 116, 387–402. Kroodsma, D.E., Miller, E.H., 1996. Ecology and Evolution of Acoustic Communication in Birds. Cornell University Press, Ithaca. Langemann, U., Gauger, B., Klump, G.M., 1998. Auditory sensitivity in the great tit: perception of signals in the presence and absence of noise. Anim. Behav. 56, 763–769. Mammen, D.L., Nowicki, S., 1981. Individual differences and within-flock convergence in chickadee calls. Behav. Ecol. Sociobiol. 9, 179–186. Marler, P., 1957. Specific distinctiveness in the communication signals of birds. Behaviour 11, 13–39. Marler, P., 2004. Bird calls: their potential for behavioural neurobiology. In: Marler, P., Zeigler, P. (Eds.), Behavioural Neurobiology of Birdsong. New York Academy of Sciences, New York, pp. 31–44. Mathevon, N., 1998. Degraded temporal sound features as a function of distance and potential as cues for ranging in birds. Bioacoustics 9, 17–33.

281

Mathevon, N., Aubin, T., 2001. Sound-based species-specific recognition in blackcap Sylvia atricapilla shows high tolerance to signal modifications. Behaviour 138, 511–524. Mbu-Nyamsi, R.G., Aubin, T., Br´emond, J.C., 1994. On the extraction of some time dependent parameters of an acoustic signal by means of the analytic signal concept. Its application to animal sound study. Bioacoustics 5, 187–203. Naguib, M., Wiley, R.H., 2001. Estimating the distance to a source of sound: mechanisms and adaptations for long-range communication. Anim. Behav. 62, 825–837. Nelson, D.A., 1988. Feature weighting in species song recognition by the field sparrow, Spizella pusilla. Behaviour 106, 158– 182. Nelson, D.A., 1989. Song frequency as a cue for recognition of species and individuals in the field sparrow. J. Comp. Psychol. 103, 171–176. Nelson, D.A., 2000. Avian dependence on sound pressure level as an auditory distance cue. Anim. Behav. 59, 57–67. Nowicki, S., 1983. Flock-specific recognition of chickadee calls. Behav. Ecol. Sociobiol. 12, 317–320. Rafter, J.A., Abel, M.L., Braselton, J.P., 2002. Multiple comparison methods for means. Soc. Ind. Appl. Math. Rev. 44, 259–278. Randall, R.B., Tech, B., 1987. Frequency Analysis. Br¨uel and Kjaer, Naerum. Ratcliffe, L.M., Weisman, R.G., 1985. Frequency shifted in the feebee song of the black-capped chickadee. Condor 87, 555–556. Saberi, K., Takahashi, Y., Farahbod, H., Konishi, M., 1999. Neural bases of an auditory illusion and its elimination in owls. Nat. Neurosci. 2, 581–583. Searcy, W.A., 1990. Species recognition of song by female redwinged blackbirds. Anim. Behav. 40, 1119–1127. Slabbekoorn, H., ten Cate, C., 1999. Collared dove response to playback. Slaves to the rhythm. Ethology 105, 377–391. Smith, S.M., 1993. Black-capped chickadee (Poecile atricapilla). In: Poole, A., Stettenheim, P., Gill, F., (Eds.), The Birds of North America. No. 39, Academy of Natural Sciences, Philadelphia, and American Ornithologists’ Union, Washington, pp. 1–19. Vignal, C., Mathevon, N., Mottin, S., 2004. Audience drives male songbird response to partner’s voice. Nature 431, 448–451. Weisman, R.G., Ratcliffe, L.M., 1989. Absolute and relative pitch processing in black-capped chickadees (Parus atricapillus). Anim. Behav. 38, 685–692. Weisman, R.G., Ratcliffe, L.M., Johnsrude, I., Hurley, T.A., 1990. Absolute and relative pitch production in the song of the blackcapped chickadee. Condor 92, 118–124. Wiley, R.H., Richards, D.G., 1982. Adaptation for acoustic communication in birds. sound transmission and signal detection. In: Kroodsma, D.E., Miller, E.H., Ouellet, H. (Eds.), Acoustic Communication in Birds. Academic Press, New York, pp. 131– 181.