Dawn song of male blue tits as a predictor of competitiveness in midmorning singing interactions

acta ethol (2004) 6: 65 71 DOI 10.1007/s10211-004-0086-0 ORIGINAL ARTICLE Angelika Poesel. Torben Dabelsteen. Simon Boel Pedersen Dawn song of male blue tits as a predictor of competitiveness in midmorning singing interactions Received: 3 June 2003 / Revised: 26 January 2004 / Accepted: 26 January 2004 / Published online: 22 April 2004 # Springer-Verlag and ISPA 2004 Abstract During the dawn chorus, territorial male songbirds vocalise intensively within signalling range of several conspecific males and can therefore be considered members of a busy communication network. The more or less continuous singing over a long period of time under standardised stimulus conditions makes the dawn song a potentially important information source both for simple receivers and for eavesdroppers. Male blue tits (Parus caeruleus) vary in features of their dawn song, e.g. older males sing longer strophes, and females choose males that sing longer strophes as extra-pair partners. However, so far, dawn song in the blue tit has been investigated separately from other singing behaviour of the same males. In this study, we investigate aspects of blue tit male quality, reflected in dawn song characteristics, and their predictive value for how males behave during singing interactions later in the morning. We acted as simple receivers by recording the singing activity of one male at a time at dawn and compared features of its dawn song, such as onset before sunrise, repertoire size, mean bout length, strophe length and percentage performance time to responses of the same male to a territory intrusion simulated by playback of synthesised songs later during the same morning. We assume that an aggressive response towards an intruder will involve a fast approach to the loudspeaker broadcasting strophes of blue tit song, searching for the intruder (flying around), and a high amount of counter singing and overlapping of the intruder s songs. Aspects of vigour of response to the simulated intrusion could be predicted from all five investigated dawn song parameters as well as male age. Communicated by P.K. McGregor A. Poesel (*). T. Dabelsteen. S. B. Pedersen Department of Animal Behaviour & Centre for Sound Communication, Zoological Institute, University of Copenhagen, Tagensvej 16, 2200 Copenhagen N, Denmark e-mail: apoesel@zi.ku.dk Tel.: +45-3532-1346 Fax: +45-3532-1299 This is, to our knowledge, the first indication that a simple receiver could extract reliable information from a male s dawn singing behaviour about its competitiveness later in the day. Keywords Dawn song. Parus caeruleus. Singing interaction Introduction The breeding season of many temperate zone songbirds is characterised by a particularly high singing activity at dawn giving rise to the so-called dawn chorus (Catchpole and Slater 1995). The functions of this dawn chorus have been speculated on extensively, and hypotheses range from environmental pressures, hormone-related cycles, to social functions (review in Staicer 1996). Social functions can be split into two groups following the generally accepted dual function of song, territory defence and mate attraction: Males might sing at dawn because it is crucial to advertise territory ownership after a night of silence (territory defence hypothesis in Staicer 1996). On the other hand, dawn song might be a reliable signal of male quality (e.g. black-capped chickadee (Parus atricapillus): Otter et al. 1997; blue tit (P. caeruleus): Poesel et al. 2001) and relate to male attributes such as parasite load or territory quality and function in attracting or stimulating a mate (e.g. great tit (P. major): Mace 1986; Slagsvold et al. 1994; willow tit (P. monatus): Welling et al. 1995). Receivers that focus on physical characteristics of the song of one male at a time (simple receiving) may gain information about the male s willingness to escalate a fight (e.g. blackbird (Turdus merula): Dabelsteen and Pedersen 1990, 1992), its previous experience or age (e.g. great tit: Lambrechts and Dhondt 1986; willow tit: Welling et al. 1995), current body condition (e.g. robin (Erithacus rubecula): Godfrey and Bryant 2000) and social rank in a network of territorial individuals (e.g. black-capped chickadee: Otter et al. 1997). Several studies have found that aspects of song complexity and song output seem to

66 predict male quality as they influence female mate choice (e.g. song sparrow (Zonotrichia melodia): Hill et al., submitted; common whitethroat (Sylvia communis): Balsby 2000; reviewed in Searcy andyasukawa 1996). Two such aspects, repertoire size and song length, may be especially important in predicting male quality (e.g. song sparrow and field sparrow (Spizella pusilla): Searcy 1983; Stoddard et al. 1988; coal tit: Adhikerana and Slater 1993). Given the high singing rate and the relatively long duration of each individual s singing period at dawn, males can be expected to cycle through a large part of their song repertoire at dawn (blue tit: Doutrelant et al. 2000) making repertoire size potentially easier assessable at dawn than later in the day. The new communication network approach considers that an individual can gather reliable information about other individuals both by simple receiving and by eavesdropping, where eavesdropping allows third parties to gather relative information about other individuals engaged in a signalling interaction (McGregor and Dabelsteen 1996). Recent studies have shown that both males and females can eavesdrop on male singing interactions during the day and utilise the males singing pattern to gather relative information about their dominance relationship which is used in successive decisionmaking (e.g. great tit: Otter et al. 1999; Peake et al. 2001, 2002; black-capped chickadee: Mennill et al. 2002). In these close range singing interactions an overlapping pattern, i.e. one singer starts its song before the counterpart has terminated its, signals more aggressive behaviour or readiness to escalate a conflict than alternating songs (Dabelsteen et al. 1996; great tit: Langemann et al. 2000; robin: Dabelsteen et al. 1997; nightingale (Luscina megarhynchos): Naguib and Todt 1997). Since the dawn chorus may be considered a busy communication network encompassing temporary singing interactions among several territory holders (Burt et al. 2001), gaining relative information by eavesdropping should probably be added to the options for gathering reliable information from the dawn chorus (Poesel & Dabelsteen 2001). Recent experiments with great tits indicate that such a network environment might also include females inside a nestbox as they may receive songs from their own mate as well as from neighbouring males (Blumenrath et al. accepted). If features of dawn song characteristics or singing patterns provide reliable information about the singer s quality, condition and ensuing behaviour, other individuals should utilise this information in decision-making, as has been suggested for females, and the features should predict how the singers themselves behave later in the day, for instance in defending their territory against an intruder. We test this prediction as regards information conveyed by dawn song characteristics in the blue tit. In this species, strophe length (Kempenaers et al. 1997) and percentage performance time at dawn (Poesel et al. 2001) have previously been shown to reflect aspects of male quality. Furthermore, age, as a measure of surviving ability or male quality, seems to play a role in female choice processes in the blue tit (Kempenaers et al. 1992). We can thus predict that male performance at dawn and during a territorial intrusion later in the morning will be positively correlated, probably influenced by male experience or age. Besides song, we use additional measures of vigour of response to an intruder to compare with dawn singing performance and test whether a male s dawn song is a reliable predictor of its performance in later midmorning agonistic encounters. Methods Study population The study was made in 2002 with a population of blue tits breeding in 104 nestboxes in mixed deciduous woodland at Kolbeterberg, Vienna, Austria (48 13 N, 16 20 E). Nestboxes, each 40 m apart, were checked regularly to determine onset of nest building, egg laying, hatching and fledging as part of a long-term study on blue tit reproduction (Foerster 2002). Individuals were colour-banded, sexed and aged (first-year or older bird based on plumage characteristics; Svensson 1992). Dawn song recording and analysis Apart from individual characteristics, external and internal factors may also contribute to the measured variation in song characteristics among males. We reduced the influence of external and internal factors by (1) recording male blue tits under relatively standardised context at dawn from the first strophe sung until the female left the nestbox, and (2) by recording males during the early egg laying period of their mates (between egg 2 and 8 of median 12 eggs) when females were fertile. We recorded dawn song of 15 first-year and 12 older males between 24 March and 15 April. In 16 cases the recording of dawn singing stopped when the female left the nest box and copulated with the focal male which always ends dawn song (dawn song duration (mean±se): 38.5 min±2.2). In the other 11 cases the recording of dawn singing finished when both members of the pair were observed in the territory, i.e. we did not actually see the female leave the nestbox, but singing had stopped (dawn song duration (mean±se): 34.8 min±3.8). There was no significant difference in the duration of the dawn chorus recordings between these two groups (t-test: t=0.74, df=25, P=0.47), or in any of the other dawn song variables (all P>0.5). All of the recordings were made with a Sennheiser MKH 816 T directional microphone connected to a Sony TCD-D8 DAT recorder. Sonographic analysis was performed with the software package Avisoft-SASLab Pro 4.10 (FFT 256, overlap 50%, bandwidth 112 Hz, frequency resolution 86 Hz, time resolution 5.8 ms). At dawn, male blue tits sang continuously with short inter-strophe pauses. On average, 80% of the strophes recorded at dawn could be classified as song, 9% as calls and 11% as scolding (see Bijnens and Dhondt 1984 for definitions of song types, calls and scolding). Calls and scolding are emitted in bouts like song strophes, but scolding is usually used only in response to an external stimulus and its intensity is likely to vary with the strength of the stimulus that it is elicited by. Since we were interested in dawn song features reflecting male quality and did not have control over these stimuli we omitted scolding from the analysis of dawn song and pooled data for songs and calls following Poesel et al (2001). The males switched between 3 and 10 song and call types, which were emitted in 4 17 bouts. Dawn song was quantified by five independent measures of complexity and output obtained from all strophes of songs and calls: (1) onset of dawn singing in minutes before sunrise; (2) repertoire of song and call types; (3) mean duration of bouts of different vocaliation types (in seconds); (4) mean strophe length of

67 all songs and calls (in seconds); and (5) mean percentage performance time (PPT) of all songs and calls, i.e. the ratio of strophe length to strophe length plus successive inter-strophe pause. Playback and experimental procedure We synthesised strophes of the most common song type in the population using an application for MathCAD 8.0 (written by SB Pedersen), which allowed us to enter exact values for frequency and duration measured from strophes of different previously recorded male blue tits (treatment males). These strophe templates were then transferred to EXBIRD, a customised Windows based software that allows playback from a computer (written by SB Pedersen). In the field, synthesised strophes of a single treatment male were played to a subject and treatment males were varied between subjects to avoid pseudoreplication. Strophes (strophe lengths: 1.59±0.01 s.) were played in a 2-min loop with randomly varied inter-strophe pauses (2.8 3.8 s) resulting in a predetermined mean (±SE) pause duration of 3.3±0.02 s and a PPT of 32.6±0.1% as observed in natural singing bouts of blue tits. Playback was accomplished from a portable computer through a 35 W amplifier (Denon DC-600) connected to a VIFA 1 Neodymium tweeter (Larsen and Dabelsteen 1997). The output sound pressure level (SPL) was adjusted to the natural maximum level of blue tit songs of about 64 db(a) at 10 m. The SPL for blue tit song was determined from measurements made with a Brüel and Kjaer SPL meter type 2236 (fast setting, A-filter) at different distances (12 36 m) from five males singing in the area in late March (SPL [db(a)=86.4 24.7 log (distance)]. To standardise the perceived level of threat from the intrusion we placed the speaker 2 3 m above the ground in a small tree pointing towards the entrance hole of the nestbox at a distance of 10 m. Males challenged by playback had their dawn song recorded on the same day. Before an experiment started, males were located within the territory and at least 5 min had passed since they had last vocalised. Playback experiments were carried out between 0830 and 1100 hours, at least 2 h after the dawn chorus had stopped. We challenged 27 males with a simulated intrusion of a 2-min loop playback. When the subject did not respond, we waited for 15 min before we repeated the loop playback. In four cases we still could not elicit a response and abandoned the experiment. Response measures to playback During the 2-min playback treatment, the vocal response of the focal male was recorded while the male s non-vocal behaviour was spoken onto a tape by another observer placed at the edge of the territory. Vocalisations varied and could be classified as songs, calls or scolding according to the definitions by Bijnens and Dhondt (1984). In the analysis of the vocal responses, scolding was considered as strophes on equal terms with calls and songs because all vocalisations had been elicited by the playback. From the recordings of non-vocal behaviour we extracted two approach measures: (1) latency to approach within 5 m of the speaker, and (2) time spent within 5 of the speaker, and one agitation measure: (3) total number of flights during playback. From the sound recordings, we extracted one output measure: (4) number of strophes sung in response to playback, and two measures of singing pattern: (5) the delays from the start of playback strophes to the subject s strophes, and (6) the mean extent of the subject s song overlapping each playback strophe. A highly aggressive response of the territory owner towards the intruder would correspond to a short latency of approach, long time spent close, high number of flights, many strophes, short delay of subject s strophes, and large extent of subject s song overlapping. However, not all values might be maximised as there could be a trade-off between different behaviour patterns. A correlation table of the six response measures revealed relatively low and non-significant correlations among all variables (correlation coefficients r <0.3), but significantly higher correlation among two pairs of variables: The latency to approach within 5 m of the speaker was correlated with the time spent within 5 m of the speaker (Pearson correlation: r P = 0.58, n=23, P=0.003). Similarly, the delay to playback strophes was correlated with the extent of the subject s song overlapping each playback strophe (Pearson correlation: r P = 0.42, n=23, P=0.04). These pairs of correlated variables could be interpreted as measures of approach and singing pattern, respectively. To reduce the number of variables (from originally six to four; see Table 1) and to obtain independent response measures to playback we subjected each pair of these correlated variables to a separate principal component analysis. A single component (PC) was extracted from each pair, which explained 79% [PC (approach)] and 69% [PC (singing pattern)] of the variation in the original variables. Statistical analysis All statistical analyses are based on Sokal and Rohlf (1995) and were performed using SAS 8.0. If data deviated from normal distribution according to Shapiro Wilk statistics, we approached normality and homogeneity of variance of data by applying logtransformation (bout length, number of flights and number of strophes) and square root transformation (repertoire size and PPT). We applied standard Bonferroni technique for multiple testing (Rice 1989). We investigated relationships between dawn chorus variables and playback response measures for each of the four response measures in a separate general linear model. All models initially included one single playback response measure as dependent variable and six main factors [five independent dawn song parameters (continuous variables) and age (two groups)] and first order interactions between each dawn song parameter and age. Models were then refined in a stepwise procedure eliminating variables with non-significant influence. Table 1 Four response measures of male blue tits (Parus caeruleus) to playback simulated intrusions. For the two principal components [PC(approach) and PC(singing pattern)] the loadings on the original variables are given: A high value of PC(approach) Measure of response to playback indicated a fast approach and long time spent close to the speaker. A high value of PC(singing pattern) indicated a short delay after the playback strophe and a high proportion of the subject s song overlapping the playback strophe. Behavioural aspect PC(approach): latency to approach within 5 m 0.89 Approach Time spent within 5 m 0.89 Number of flights agitation Number of strophes sung song output PC(singing pattern): delay after playback strophe 0.83 singing pattern Proportion of subject s song overlapping 0.83

68 Results Response to playback simulated intrusion One first-year and three older males did not respond to the playback in two trials and were thus excluded from further analyses. These four males had sung with lower PPT (ttest: t= 2.4, df=26, P=0.02) and with a smaller repertoire (t-test: t= 2.1, df=26, P=0.04) at dawn than the 23 males that did respond to the simulated intrusion, but they did not differ in onset of dawn song, strophe length or mean bout length (all P>0.5). All 23 responding males approached the loudspeaker within the first minute of playback and vocalised during at least part of the 2-min. experiment. Nineteen (82.6%) males responded with song to the playback, singing either song types they had used at dawn or similar types with trill. Five (21.7%) males matched exactly the song type of the simulated intruder, while three of these males had not used this song type at dawn. Two 1-year-old males (8.7%) responded with scolding only. Subject males responded with lower PPT than at dawn (paired t-test: t=3.45, df=22, P=0.002) but with similarly long strophes (paired t-test: t=1.65, df=22, P=0.11). Dawn song features and response to playback The five dawn song parameters were not correlated (all r <0.4; all P>0.1) and could, in combination with male age, predict the following aspects of a test male s response to the simulated playback intrusion during midmorning (Table 2 and Fig. 1). PC (approach) was best predicted by mean bout length at dawn. Males singing shorter bouts at dawn responded quicker to the intruder and stayed close for a longer time. None of the other variables was included in the final model. Age had a strong influence on how much the playback agitated territory owners, i.e. how much males flew around. In combination with male age onset of dawn singing, bout length, strophe length and PPT could all predict the number of flights during the simulated intrusion. In particular, older males with early onset of singing at dawn and long strophes flew around more during the playback. Repertoire size did not have a significant influence on the number of flights and was thus excluded from the final model. The two-song response measures (number of strophes and singing pattern) could also be predicted from some dawn song parameters. The number of strophes sung in response to the simulated intruder was negatively related to strophe length, PPT and interactions between male age and these two parameters. Particularly older males that had sung with higher PPT at dawn, responded with fewer vocalisations to the simulated intruder. The PC (singing pattern) was predicted by repertoire size, males that used many different song and call types at dawn responded with long delays after the playback strophes and overlapped strophes to a lesser extent. None of the other dawn song parameters or male age could explain significantly more variation in this parameter. Table 2 Results of four general linear models with (1) PC age. Full models were refined in a stepwise order by eliminating (approach), (2) number of flights, (3) number of strophes and (4) PC (singing pattern) as the dependent variable, age as factor (first-year and older males), onset of dawn song, repertoire size, bout length, variables with non-significant influence. Only independent variables included in the final model are shown; parameter estimates (β values) for each independent variable are given strophe length and PPT as covariates and first order interactions with Dependent variable df R 2 F P ß PC(approach) Model 1 0.22 5.97 0.023 Bout length 1 5.97 0.023 3.95 Number of flights Model 4 0.66 3.41 0.022 Begin age 1 4.25 0.036 0.04 Bout length age 1 7.85 0.005 0.002 Strophe length age 1 6.45 0.010 0.61 PPT age 1 10.09 0.002 0.02 Number of strophes Model 4 0.53 5.06 0.007 a Strophe length 1 7.15 0.015 0.08 PPT 1 5.21 0.035 0.28 Strophe length age 1 5.58 0.030 0.76 PPT age 1 6.94 0.017 0.18 PC(singing pattern) Model 1 0.27 7.76 0.011 a Repertoire 1 7.76 0.011 1.78 a Still significant after Bonferroni correction for repeated testing in four models (α =0.013)

69 Fig. 1 Comparisons between dawn song features and responses to a playback-simulated territory intrusion by blue tits (Parus caeruleus) during midmorning. Shown are relations between the four response measures [PC(approach), number of flights, number of strophes and PC(singing pattern)] and independent variables included in the final linear models presented in Table 2. The upper two graphs show regression lines of the pooled data, while data are split by the two age classes in the bottom four graphs. Filled circles and solid lines represent more than 1-year-old males while open circles and dashed lines represent 1-year-old males in cases where the model indicated a significant interaction between age and an independent variable

70 Discussion All five investigated dawn song parameters in combination with male age could predict aspects of male blue tit vigour of response to a simulated territory intrusion later in the same day. However, song responses to the simulated intrusion were better predicted than approach or agitation behaviours. Overall, our results indicate that male and female blue tits acting as simple receivers could potentially extract reliable information from dawn song about a male s willingness to approach and escalate eventual fights with daytime intruders. Strophe length and PPT at dawn had previously been shown to correlate with breeding behaviour in the blue tit, i.e. males with longer strophes were chosen for EPCs (Kempenaers et al. 1997) while males with higher PPT were paired to females that laid their clutches earlier (Poesel et al. 2001). Both studies indicate that features of dawn song in male blue tits might advertise aspects of male quality. Our results are in accordance with this quality advertisement function of dawn song in the blue tit as they indicate that strophe length and PPT can predict the amount of song with which a territorial male responds to an intruder later in the day. Potentially, dawn song features could even predict whether a male will respond to an intruder at all since the four non-responding males had on average lower PPT and smaller repertoire size. However, this would need further investigation. A strong response to an intruder can be reflected in different ways. One possibility is that a territorial male starts countersinging the moment the intruder can be heard and overlaps as many and as large an extent of the intruder s strophes as possible. Alternatively, but possibly perceived as similarly aggressive, a male can approach an intruder closely and, in case it cannot locate it immediately, fly around and search for it. Males might face a trade-off between these different behaviours, for example a male searching for an intruder might not be able to sing at a high rate or a male mate-guarding might not be able to advertise its quality simultaneously (Johnsen et al. 1997). An individual might apply different response strategies depending on the context, for example the perceived threat, although each individual may predominantly use one strategy (e.g. Dabelsteen and McGregor 1996). Recent evidence for individual differences in response has started to accumulate (e.g. Nowicki et al. 2002). Mennill and Ratcliffe (2004), for example, could document significant differences in playback responses between high- and lowranking male black-capped chickadees. Similarly in our study, the response of territorial male blue tits to a simulated intruder was highly influenced by the age of the test subject. Males of lower quality as possibly indicated by smaller repertoire, shorter strophes and lower PPT in the blue tit (Bijnens 1988; Kempenaers et al. 1997; Poesel et al. 2001) responded with a higher amount of countersinging and short delays after the playback strophe and thus overlapped with a large extent of their own strophes. This might indicate a strategy to overlap the intruder s song by simultaneously avoiding to be overlapped in return as song stops shortly after the overlapped strophe (e.g. Dabelsteen et al. 1996). The rather unexpected direction of difference between high and low quality males in response to the simulated intruder could be explained by the fact that all males faced a similar average male as intruder which meant that the relative quality differences between subject and playback were different and could have affected a male s response. Males which assess competitiveness of potential opponents at dawn might have an advantage in singing interactions later in the day. Compared to day singing, dawn song has the potential for providing more reliable and comparable information because all males sing simultaneously under relatively standardised conditions as regards motivating factors; for instance, all males sing after a night without social activity (social dynamics in Staicer 1996). Assessing potential opponents at dawn might reduce the number of time- and energy-consuming fights wasted on low-ranking opponents, which actually might not present an immediate threat of losing a territory or paternity. This theoretical background, together with the correlations shown in this study, opens up a wide range of possibilities of information gain from dawn song. These possibilities are opposed by a lack of studies showing whether, for how long and of how many individuals male blue tits can remember song performance, and whether at all they can gather information from the dawn singing of rivals while engaged in intense dawn singing themselves, either by simple receiving or by eavesdropping on singing interactions. Furthermore, the social environment of a male territory owner might have a strong influence on its response towards an intruder, as the frequency of natural intrusions will vary with population density (e.g. Silverin 1998). Acknowledgements We thank Hans Winkler from the Konrad Lorenz Institute for Comparative Ethology, Vienna for continuous support during this study. We are grateful to Bart Kempenaers from the Research Centre for Ornithology for logistic support and Agnes Tuerk for invaluable help in the field. Raphael-Thomas Klumpp and Alfred Fojt from the institute of Silviculture, Vienna, provided access to their facilities in the study area. We thank Christian Schlögel for excellent help with song recordings and Ronald Smetacek for technical assistance. We are grateful to Thorsten Balsby for fruitful discussions, statistical advice and comments on the manuscript and to two anonymous referees whose comments improved the manuscript. The Danish National Research Foundation and the Centre for Sound Communication supported this study. The research was conducted in accordance with the laws in Austria and Denmark. References Adhikerana AS, Slater PJB (1993) Singing interactions in coal tits, Parus ater: An experimental approach. Anim Behav 46:1205 1211 Balsby TJ (2000) Song activity and variability in relation to male quality and female choice in whitethroats Sylvia communis. J Avian Biol 31:56 62

71 Bijnens L (1988) Blue Tit Parus Caeruleus Song in relation to survival reproduction and biometry. Bird Study 35:61 68 Bijnens L, Dhondt AA (1984) Vocalizations in a Belgian blue tit, Parus c. caeruleus, population. Gerfaut 74:243 269 Blumenrath S, Dabelsteen T, Pedersen SB (accepted) Do locations inside nest boxes complicate the receiving conditions for great tit females? Bioacoustics Burt JM, Trillo PA, Vehrencamp SL (2001) Two-way and mulit-way vocal interactions in a territoral songbird. XXVII International Ethological Conference, Tübingen, Germany Catchpole CK, Slater PJB (1995) Bird song: biological themes and variations. Cambridge University Press, Cambridge Dabelsteen T, McGregor PK (1996) Dynamic acoustic communication and interactive playback. In: Kroodsma DE, Miller EH (eds) Ecology and evolution of acoustic communication in birds Cornell University Press, Ithaca, N.Y., pp 398 408 Dabelsteen T, Pedersen SB (1990) Song and information about aggressive responses of blackbirds, Turdus merula: evidence from interactive playback experiments with territory owners. Anim Behav 40:1158 1168 Dabelsteen T, Pedersen SB (1992) Song features essential for species discrimination and behaviour assessment by male blackbirds. Behaviour 121:259 287 Dabelsteen T, McGregor PK, Shepherd M, Whittaker X, Pedersen SB (1996) Is the signal value of overlapping different from that of alternating during matched singing in great tits? J Avian Biol 27:189 194 Dabelsteen T, McGregor PK, Holland J, Tobias JA, Pedersen SB (1997) The signal function of overlapping singing in male robins. Anim Behav 53:249 256 Doutrelant C, Blondel J, Perret P, Lambrechts MM (2000) Blue tit song repertoire size, male quality and interspecific competition. J Avian Biol 31:360 366 Foerster K (2002) Extra-pair paternity in the blue tit Parus caeruleus Doctoral thesis, Zoological Institute, Univ. Vienna Godfrey JD, Bryant DM (2000) State-dependent behaviour and energy expenditure: an experimental study of European robins on winter territories. J Anim Ecol 69:301 313 Hill CE, Campbell SE, Beecher MD (subm) Does femal choice select for male song repertoires? Female choice of extra-pair mates and male song characteristics in the song sparrow. Behavioural Ecology Johnsen A, Lifjeld JT, Rohde PA (1997) Coloured leg bands affect male mate-guarding behaviour in the bluethroat. Anim Behav 54:121 130 Kempenaers B, Verheyen GR, Van Der Broeck M, Burke T, Van Broeckhoven C, Dhondt AA (1992) Extra-pair paternity results from female preference for high-quality males in the blue tit. Nature 357:494 496 Kempenaers B, Verheyen GR, Dhondt AA (1997) Extrapair paternity in the blue tit (Parus caeruleus): female choice, male characteristics, and offspring quality. Behav Ecol 8:481 492 Lambrechts M, Dhondt AA (1986) Male quality reproduction and survival in the great tit Parus major. Behav Ecol Sociobiol 19:57 64 Langemann U, Tavares IP, Peake TM, McGregor PK (2000) Response of great tits to escalating patterns of playback. Behaviour 137:451 471 Larsen ON, Dabelsteen T (1997) The VIFA 1 Neodymium tweeter: a versatile speaker for playback experiments. Bioacoustics 8:323 326 Mace R (1986) Importance of female behaviour in the dawn chorus. Anim Behav 34:621 622 McGregor P, Dabelsteen T (1996) Communication networks. In: Kroodsma DE, Miller EH (eds) Ecology and evolution of acoustic communication in birds Cornell University Press, Ithaca, N.Y., pp 409-425 McGregor PK, Dablesteen T, Shepherd M, Pedersen SB (1992) The signal value of matched singing in great tits: evidence from interactive playback experiments. Anim Behav 43:987 998 Mennill DJ, Ratcliffe L (2004) Overlapping and matching in the song contests of black-capped chickadees. Anim Behav (in press) Mennill DJ, Ratcliffe L, Boag P (2002) Female eavesdropping on male song contests in songbirds. Science 296:872 873 Naguib M, Todt D (1997) Effects of dyadic vocal interaction on additional conspecific receivers in nightingales. Anim Behav 54:1535 1543 Nowicki S, Searcy WA, Krueger T, Hughes M (2002) Individual variation in response to simulated territorial challenge among territory-holding song sparrows. J Avian Biol 33:253 259 Otter K, Chruszcz B, Ratcliffe L (1997) Honest advertisement and song output during the dawn chorus of black-capped chickadees. Behav Ecol 8:167 173 Otter K, McGregor PK, Terry AMR, Burford FRL, Peake TM, Dabelsteen T (1999) Do female great tits (Parus major) assess males by eavesdropping? A field study using interactive song playback. Proc R Soc Lond B 266:1305 1309 Peake TM, Terry AMR, McGregor PK, Dabelsteen T (2001) Male great tits eavesdrop on simulated male-male vocal interactions. Proc R Soc Biol Sci B 268:1183 1187 Peake TM, Terry AMR, McGregor PK, Dabelsteen T (2002) Do great tits assess rivals by combining direct experience with information gathered by eavesdropping? Proc R Soc Lond B 269, 1925 1929 Poesel A, Dabelsteen T (2001) Variation in blue tit down song - A consequence of male quality and female eavesdropping? Advances in Ethology 36:240 Poesel A, Foerster K, Kempenaers B (2001) The dawn song of the blue tit Parus caeruleusand its role in sexual selection. Ethology 107:521 531 Rice WR (1989) Analyzing tables of statistical tests. Evolution 43:223 225 Searcy WA. (1983) Response to multiple song types in male song sparrows and field sparrows. Anim Behav 31:948 949 Searcy WA, Yasukawa K (1996) Song and female choice. In: Kroodsma DE, Miller EH (eds) Ecology and evolution of acoustic communication in birds. Cornell University Press, Ithaca, N.Y., pp 454 473 Silverin B (1998) Territorial behaviour and hormones of the pied flycatcher in optimal and suboptimal habitats. Anim Behav 56:811 818 Slagsvold T, Dale S, Sætre G-P (1994) Dawn singing in the great tit (Parus major): mate attraction, mate guarding, or territorial defence?behaviour 131:115 138 Sokal RR, Rohlf FJ (1995) Biometry. The principles and practice of statistics in biological research. Freeman, New York Staicer CA (1996) Honest advertisement of pairing status evidence from a tropical resident wood-warbler. Anim Behav 51:375 390 Stoddard PK, Beecher Michael D, Willis M (1988) Response of territorial male song sparrows to song types and variations. Behav Ecol Sociobiol 22:125 130 Svensson E (1992) Identification guide to European passerines. Fingraf, Sweden Welling P, Koivula K, Lahti K (1995) The dawn chorus is linked with female fertility in the willow tit Parus montanus. J Avian Biol 26:241 246