J Ornithol (2007) 148:123–128 DOI 10.1007/s10336-006-0116-y ORIGINAL ARTICLE Implications of conspecific background noise for features of blue tit, Cyanistes caeruleus, communication networks at dawn Angelika Poesel Æ Torben Dabelsteen Æ Simon Boel Pedersen Received: 3 January 2006 / Revised: 8 November 2006 / Accepted: 8 November 2006 / Published online: 5 December 2006 Ó Dt. Ornithologen-Gesellschaft e.V. 2006 Abstract Communication among animals often com- may potentially constrain the perception of singing prises several signallers and receivers within the sig- patterns and may constitute costs for eavesdroppers. nal’s transmission range. In such communication On the other hand, signallers may position themselves networks, individuals can extract information about strategically and privatise their interactions. differences in relative performance of conspecifics by eavesdropping on their signalling interactions. In Keywords Communication network Á Cyanistes songbirds, information can be encoded in the timing of caeruleus Á Singing interaction signals, which either alternate or overlap, and both male and female receivers may utilise this information when engaging in territorial interactions or making Introduction reproductive decisions, respectively. We investigated how conspecific background noise at dawn may overlay Territorial animals form neighbourhoods of several and possibly constrain the perception of such singing individuals within signalling and receiving range (see patterns. We simulated a small communication net- examples in McGregor 2005). In order to defend ter- work of blue tits, Cyanistes caeruleus, at dawn in ritorial space successfully, it may be advantageous to spring. Two loudspeakers simulated a singing interac- gather information about competitors and thus be able tion which was recorded from four different receiver to adjust responses to an opponent’s strength. Such positions simulating potential eavesdroppers. During information can be obtained by attending to signals the recordings, resident blue tit males were vocalising and signalling strategies reflecting aspects of individual and created natural conspecific background noise. male quality. For example, the fundamental frequency Levels of conspecific background noise were high and of a red deer, Cervus elaphus, stag’s roar reflects age varied among positions of potential eavesdroppers. We and weight of the signaller (Reby and McComb 2003). conclude that conspecific background vocalisations Animals may also eavesdrop on signalling interactions and in this way gain valuable relative information Communicated by F. Bairlein. about the interacting individuals without engaging in the interaction themselves (McGregor and Dabelsteen A. Poesel (&) Á T. Dabelsteen Á S. B. Pedersen 1996). Animal Behaviour Group, Biological Institute, Experiments using playback-simulated interactants University of Copenhagen, Tagensvej 16, on nightingales, Luscinia megarhynchos, (Naguib and 2200 Copenhagen N, Denmark e-mail: [email protected] Todt 1997; Naguib et al. 1999) and great tits, Parus major, (Peake et al. 2001, 2002) have shown that males Present Address: extract information from asymmetries in the timing of A. Poesel songs. Individuals that alternate their signals with those Borror Laboratory of Bioacousatics, The Ohio State University, EEOB, 1315 Kinnear Road, of an opponent seem to be perceived as less dominating Columbus, OH 43212, USA or aggressive than individuals that overlap an 123 124 J Ornithol (2007) 148:123–128 opponent’s signals. Having gained relative information cific vocalisations have a potential for providing con- about individuals, eavesdroppers may respond more siderable background noise at dawn in which signals vigorously to a more dominant singer (Peake et al. 2002). from particular individuals need to be detected. In female birds, long term effects of such eavesdropping This study aimed at quantifying how often and to have been documented in black-capped chickadees, what extent conspecific vocalisations may overlay Poecile atricapillus, which seem to use information essential parts of temporal interaction patterns in the gathered by eavesdropping on male–male interactions blue tit in a natural habitat at dawn. We simulated four in reproductive decisions (Mennill et al. 2002). Eaves- receivers and two males that engaged in a singing dropping seems to occur across a variety of taxa (re- interaction in three different communication networks viewed in McGregor 2005). Thus, eavesdropping at dawn. Vocalisations from resident blue tits provided appears to be a general pattern of information gathering conspecific background noise that may interfere with in animal communication (see also Johnstone 2001). the simulated singing interaction. We quantified the Many singing interactions take place at dawn in spring frequency of overlay of beginnings and ends of the when most songbird species show a peak of singing vocalisations, which determine the temporal singing activity (Catchpole and Slater 1995; review by Staicer pattern of the interaction, and calculated the duration et al. 1996). The function of this dawn song can be ex- of conspecific background noise vocalisations in rela- plained by the general dual function of song: territorial tion to the duration of sound and pauses of the play- males may sing after a night of silence to announce their back. presence to conspecifics, and also engage in singing interactions to re-establish dominance relations with neighbouring males. For example, in night-migrating species, newly-arrived males may try to establish terri- Methods tories on the breeding ground at dawn, and local males thus need to defend their territories by singing (Amrhein In mixed deciduous woodland in Vienna, Austria et al. 2004). Experimental evidence for dawn song con- (Kolbeterberg; 48°13¢N, 16°20¢E), we simulated singing taining information about male resource-holding po- interactions between two male blue tits at dawn, using tential comes from a study on the blue tit, Cyanistes playback from two loudspeakers. We recorded the caeruleus, showing that dawn song features relate to a singing interaction and background vocalisations, which male’s ability to defend the territory later in the morning were added by natural and thus uncontrolled conspecific (Poesel et al. 2004). Dawn song may also be directed to males with an array of four microphones. To obtain females: territorial males may sing to attract a mating songs for playback, we recorded male blue tits at dawn partner, to stimulate their own female, or to attract fe- and synthezised songs using an application (written by males for extra-pair copulations (e.g. Double and S.B. Pedersen) for MathCAD 8.0. The application al- Cockburn 2000). Evidence for a female-directed func- lowed us to enter exact values for frequency and dura- tion comes from studies showing a correlation between tion measured from songs of different previously dawn song features and paternity of males (e.g. Kemp- recorded male blue tits. These song templates were then enaers et al. 1997; Poesel et al. 2006; Hasselquist et al. transferred to EXBIRD, a customised Windows based 1996; Møller et al. 1998). software (written by S.B. Pedersen) that allows playback Communication range is limited by the distance over from a computer. Songs were played in three different which a receiver can perceive a signal over a background temporal patterns. One pattern was overlapping where of noise in a particular habitat (e.g. Klump 1996). song from the second loudspeaker was played while song Acoustic signals get attenuated and degraded, and also from the first loudspeaker had not yet finished. The overlaid by more or less stationary background noise of second pattern was alternating, where song from the the environment, especially at dawn (e.g. Forrest 1994; second loudspeaker was played during the pauses be- Dabelsteen and Mathevon 2002). Masking effects on tween songs from the first loudspeaker. In the third single signals by overlaying background noise have been pattern, we played songs from both loudspeakers ran- investigated both in the laboratory (e.g. Lohr et al. 2003; domly, i.e. without any consistent time pattern between Langemann and Klump 2001; Langeman et al. 1998; the songs from the two loudspeakers. Playback was Klump and Langemann 1995) and in the field (e.g. accomplished from two loudspeakers of the type VIFA Forrest 1994), but it is unknown how noise may affect 1’’ Neodymium tweeter with transmission properties signalling interactions and how it may restrict informa- similar to live birds (Larsen and Dabelsteen 1997). tion that can be gained from time patterns therein, such The loudspeakers were placed 2.5 m above ground as alternating or overlapping signal patterns. Conspe- on trees in the centre of two neighbouring blue tit 123 J Ornithol (2007) 148:123–128 125 territories at a distance of 40 m from each other, and provided ample masking potential for the simulated facing each other. The density of breeding territories singing pattern. The recording positions of the four was relatively high with 79 breeding pairs in the 35-ha microphones can be interpreted to either represent fe- study area (2.3 pairs/ha; personal observation). For male blue tits, which were still roosting on the nest, or each of the three different temporal patterns, we neighbouring males on adjacent territories, both po- played three different song types which commonly tential receivers and eavesdroppers. occur in the population, one with slow trill and one In