Polar Biol (2007) 30:1027–1033 DOI 10.1007/s00300-007-0262-6

O RI G I NAL PAPE R

Acoustic communication in the Kittiwake Rissa tridactyla: potential cues for sexual and individual signatures in long calls Thierry Aubin · Nicolas Mathevon · Vincent Staszewski · Thierry Boulinier

Received: 21 September 2006 / Revised: 21 January 2007 / Accepted: 5 February 2007 / Published online: 22 February 2007 © Springer-Verlag 2007

Abstract Sex and individual recognition systems vary among species and can have various functions in diVerent contexts. In order to determine the basis of identiWcation by voice in the Kittiwake (Rissa tridactyla), the greeting calls of 32 individuals (18 males and 14 females) were recorded in May–June 2004 on the Kittiwake colony of Hornøya island (Barents sea) and analysed. On the basis of coeYcient of variation calculations and discriminant analyses, we show (1) that calls are sexually dimorphic and that the dimorphism is mainly based on the value of the fundamental frequency, and (2) that calls are individually distinct, individuality being due to a complex of temporal and frequency parameters located in diVerent parts of the signal. This coding strategy is discussed in the context of the colonial breeding habitat of the species. Keywords Acoustic communication · Vocal signature · Coding strategy · Gull · Rissa tridactyla

T. Aubin (&) · N. Mathevon Equipe ‘Communications Acoustiques’, Université Paris-Sud, NAMC CNRS UMR 8620, Bât 446, 91405 Orsay, France e-mail: [email protected] N. Mathevon Université Jean Monnet, ENES EA 3988, 23 rue Michelon, 42023 Saint-Etienne cedex 2, France V. Staszewski · T. Boulinier CEFE, CNRS UMR 5175, 1919 Route de Mende, 34293 Montpellier, France

Introduction In most seabird and marine mammal species, reproductive adults aggregate on collective breeding areas where tens to thousands of individuals can breed next to each other. Such colonies are usually very noisy, with acoustic signals being widely used for social interactions like pair formation and/or reunions between mates and between parents and young. In this context, acoustic recognition of the other sex during mate choice processes can be facilitated by a sexual dimorphism between male and female voices that has been reported in some colonial seabirds (e.g. Brooke 1978). In addition, the acoustic channel often represents the main mode for individual recognition between mates or between parents and their oVspring. A large set of experimental investigations has shown that animals have developed speciWc acoustic recognition processes allowing them to ensure eYcient meetings in spite of the noisy and jamming environment of the colony (e.g. in penguins: see Aubin and Jouventin 1998; in seals: Charrier et al. 2002). Interestingly, some convergence have been observed: both colonial seabirds and marine mammals use complex broadband signals that are highly redundant with regard to the coding process and consequently strongly resistant to masking eVects. In all cases studied to date, and somewhat unsurprisingly, acoustic signals used for mates or parent-oVspring recognition show a well-marked individual signature. However, encoding of individual information diVers among species and this seems to be in accordance with the diYculty for making a reunion. Thus, penguins, who raise their chicks in crèches show a more sophisticated and reliable acoustic recognition system compared with species which build an individual nest and

123

1028

experience less problems in Wnding their oVspring (Aubin and Jouventin 2002; Jouventin and Aubin 2002). As attempts to Wnd evolutionary convergence, other studies have started to focus on colonial bird species with analogue constraints, such as Xamingos (Mathevon 1997) and gulls (Charrier et al. 2001; Mathevon et al. 2003). Gulls breed in colonies and have developed eYcient acoustic communication systems for individual recognition. For instance, the recognition of parental voices by oVspring is eYcient in the Black-headed gull Larus ridibundus (Charrier et al. 2001). Interestingly, as for penguins, the complexity of individual signature encoding seems to depend also on the diYculty of ensuring reliable meetings between protagonists. Thus, the potential for individuality coding is more important in a nidifugous gull species, such as the Slender-billed gull Larus genei, than in a nidicolous one such as the Black-headed gull (Mathevon et al. 2003). Depending on the biology of the species, one could also expect individual recognition to vary as a function of age. For instance, in some gull species, there is little evidence of parent-oVspring recognition before nestlings have left the nest, but the recognition of parental voice could be realised by chicks once they have Xedged from the nest and Wrst come back on it to be fed (Evans 1970; Storey et al. 1992). To pursue the research of evolutionary convergences among such species, the present paper aims to investigate the potentiality for sexual and individual vocal recognition in a gull species, the Black-legged Kittiwake Rissa tridactyla. This species breeds in large and dense colonies of several hundreds to thousands of pairs on vertical cliVs along seacoasts (Cramp and Simmons 1983; Danchin and Monnat 1992). Outside the period of colony attendance, the Kittiwake is pelagic (i.e. from October to March-April, Cramp and Simmons 1983). The age of Wrst breeding is 4–5 years, male and female are usually paired for life and the life span is more than 20 years (Cramp and Simmons 1983). Neighbour nests can be very close to each others. However, these birds are strictly territorial and both sexes defend the nest and a surrounding small area. Colonies are very noisy: the most typical call is the long call (or “kittiwake” call), which is uttered in bouts of several renditions ¡3 to 10 calls- by both sexes (Wooller 1978). This call is used most often in mates’ meeting-ceremony (“vocal greeting display”), but also during aggressive interactions with neighbours and arrival of other birds close by (Cramp and Simmons 1983; Danchin 1991). On the basis of temporal measures done on spectrograms and of qualitative visual comparison of frequency spectra, Wooller (1978) suggested that an individual signature is likely to be encoded in the

123

Polar Biol (2007) 30:1027–1033

“kittiwake” call. Conversely, no diVerence was found between male and female calls. The aim of the present paper is to assess, by more formal quantitative acoustic analyses, the potential for sexual and individual signature in the call of the Kittiwake.

Methods Study location and animals This study was carried out on the Kittiwake colony of Hornøya, an island of the Barents sea (Norway, 70°21⬘N, 31°02⬘E), from May 27th to June 6th 2004, i.e. during the egg brooding period of the breeding season. More than 10,000 pairs of Kittiwakes are present on the island (Barrett and Tertitski 2000) during the breeding period, the colony being deserted outside this period. Portions of the colony have been used for observational and experimental studies using individually marked birds (Erikstad et al. 1995; Boulinier et al. 2002). To ensure sexual and individual identiWcation of the recorded birds, we focused on such ringed individuals that bred on monitored plots and that had been sexed (sexing was done using morphometry conWrmed by DNA methods; Gasparini et al. 2002). Recordings and signal acquisition Kittiwakes use various calls, among which the ‘kittiwake’ call which can notably be emitted by a bird in Xight just before it lands on its breeding site (Danchin 1987). This long call is also emitted by pairs of individuals when they are greeting each other on their nest (Danchin 1991). We recorded the «kittiwake» calls emitted by adults with a SENNHEISER ME 64 microphone (frequency response 100–5,000 Hz, §4 dB) equipped with a 60 cm TELLINGA parabola connected to a MARANTZ PMD670 digital recorder (sampling frequency 48 kHz). During the recordings, the distance between the emitting bird and the microphone was 10–20 m. Signal analysis The «kittiwake» call (Fig. 1) is basically composed of three syllables: two short notes («ki-ti») followed by a longer one («wake»). Each syllable is a complex sound, composed of a fundamental frequency and its harmonics. For many birds, a frequency shift of the fundamental frequency approximately occurs at the middle part of the “wake”. This change in the “wake” structure, probably due to a sharp change of the tension of the

Polar Biol (2007) 30:1027–1033

1029

Fig. 1 Oscillogram (a), spectrogram (b) and power spectra (c) of a long call of adult Kittiwake (Rissa tridactyla) and some acoustic features taken in account in the analysis (see text)

tympaniform membranes of the syrinx (Brackenbury 1982), lead us to distinguish, for the birds concerned by this phenomenon, two parts in the “wake” syllable. We analysed 136 calls from 32 individuals (18 males and 14 females) using the Avisoft-Pro software (version 4.1). To characterise the acoustic structure of the calls, we measured the following parameters (the parameters marked with a * were not measurable in all the calls): In the temporal domain (envelope measurements, measurement error: §0.7 ms):

• the largest value between Fex-Fdep and Fex-FWn (DeltaF), • the average value of the fundamental (F0), • the frequency of maximum amplitude (MaxFre), • the frequency value at the upper limit of the Wrst quartile of energy (Quarter) • the frequency value at the upper limit of the second quartile (Half) • the frequency value at the upper limit of the third quartile (Threequarter)

• the duration of the «ki» part (Tki)*, • the duration of the silence between «ki» and «ti» (Ts1)*, • the duration of the «ti» part (Tti), • the duration of the silence between «ti» and «wake» (Ts2), • the duration of the «wake»‘s Wrst part (Tdebwake)*, • the duration of the «wake» part (Twake), • the duration between the beginning of a «wake» and the beginning of the next «wake» (Tinterwake).

If the second part of the «wake» was present, we also measured:

In the frequency domain (measurements on power spectra and spectrograms, measurement error: §10 Hz): • the frequency of the second harmonic at the beginning of the «wake» (Fdep), • the maximum frequency value of the second harmonic during the «wake» (Fex), • the frequency of the second harmonic at the end of the «wake»‘s Wrst part (FWn),

• the average value of the fundamental during the «wake» second part (F0b) • the frequency of maximum amplitude during the «wake» second part (MaxFreb), • the frequency value at the upper limit of the Wrst quartile of energy during the «wake» second part (Quarterb) • the frequency value at the upper limit of the second quartile during the «wake» second part (Halfb) • the frequency value at the upper limit of the third quartile during the «wake» second part (Threequarterb). Finally, the “kittiwake” call is uttered in bouts of 3–10 renditions: to know if these characteristics could constitute an index of sexual dimorphism and/or individual stereotypy, we measured the number of calls emitted by bout (Ncallsperbout).

123

1030

Polar Biol (2007) 30:1027–1033

Statistical analysis Measured data were analysed using Statgraphics Plus 3.1 software (Statistical Graphics Corporation 1994 version). For each call parameter, we computed the between-individual and within-individual coeYcients of variation (CVbi and CVwi) as follows: CV = {100(SD/Xmean)[1 + (1/4n)]}, where SD is standard deviation, Xmean the mean of the sample and n the population sample (Sokal and Rohlf 1995). To assess the potential for the coding of individual identity (Potential of Individual Coding: PIC) for each parameter, we calculated the ratio CVbi/mean CVwi (mean CVwi being the mean value of the CVwi of all individuals) (Scherrer 1984; Robisson et al. 1993). For a given parameter, a PIC value greater than 1 means that this parameter may be used for individual recognition as its intra-individual variability is smaller than its inter-individual variability. The signiWcance of PICs was tested using Kruskal–Wallis tests. In the same way, we calculated the potential of sexual coding (PSC) for each parameter, using the same formula as for PIC except that CVbi and CVwi were replaced respectively by the between-sex and withinsex coeYcients of variation (CVbs and CVis). To assess the signiWcance of PSCs, we performed a univariate

analysis of variance using Kruskal–Wallis tests. To avoid pseudo-replication problems, the data set used in these analyses was composed of only 1 call per individual, chosen at random. Subsequent multivariate analyses were done on the variables which were identiWed as relevant by the univariate analysis of variance. We performed Principal Component Analyses to get independent variables, followed by discriminant analyses using the principal components.

Results Calls are sexually dimorphic The results of Kruskal–Wallis tests (Table 1) and the calculation of PSCs indicate that the most discriminant variables to sex the signals are frequency cues, all being linked to the absolute pitch of the call (frequency of the second harmonic at the beginning of the «wake», maximum frequency value of the second harmonic during the «wake», frequency of the second harmonic at the end of the «wake»‘s Wrst part, average of the fundamental value). For temporal cues, the duration between the beginning of a «wake» and the beginning

Table 1 Analysis and comparison of acoustic parameters between male and female “kittiwake” calls Parameters (in ms or Hz) (number of female calls, number of male calls)

X § SD (min-max) (females)

X § SD (min-max) (males)

Mean CVis

CVbs

PSC

ANOVA Kruskal–Wallis H

Tki (ms) (11,17) Ts1 (ms) (11,17) Tti (ms) (14,18) Ts2 (ms) (14,18) Tdebwake (ms) (14,18) Twake (ms) (14,18) Tinterwake (ms) (12,18) Fdep (Hz) (14,18) Fex (Hz) (14,18) FWn (Hz) (14,18) DeltaF (Hz) (14,18) F0 (Hz) (14,18) MaxFre (Hz) (14,18) Quarter (Hz) (14,18) Half (Hz) (14,18) Threequarter (Hz) (14,18) F0b (Hz) (3,8) MaxFreb (Hz) (3,8) Quarterb (Hz) (3,8) Halfb (Hz) (3,8) Threequarterb (Hz) (3,8) Ncallsperbout (14,18)

50.4 § 35.7 (22–150) 54.8 § 19.8 (14–93) 75.0 § 24.2 (30–119) 133.1 § 103.1 (43–372) 473.8 § 170.2 (221–750) 487.0 § 144.6 (237–752) 971.3 § 85.4 (787–1,065) 1,595 § 147 (1,350–1,820) 1,644 § 142 (1,400–1,820) 1,596 § 163 (1,350–1,920) 68.8 § 41.2 (13–157) 539 § 44 (447–621) 2,194 § 554 (1,004–2,852) 2,025 § 328 (1,412–2,625) 3,038 § 310 (2,534–3,556) 4,887 § 637 (3,870–6,246) 1,645 § 49 (1,594–1,692) 2,180 § 875 (1,660–3,190) 2,190 § 849 (1,680–3,170) 3,900 § 892 (3,360–4,930) 6,563 § 180 (6,360–6,700) 4.7 § 1.4 (3–11)

39.5 § 10.7 (22–68) 46.9 § 17.7 (21–86) 91.2 § 26.2 (36–143) 104.0 § 69.2 (48–335) 406.4 § 116.0 (182–609) 452.6 § 114.2 (293–773) 905.4 § 134.5 (710–1,214) 1,962 § 264 (1,590–2,390) 1,855 § 204 (1,500–2,250) 1,816 § 236 (1,400–2,290) 75.9 § 49.9 (7–173) 632 § 59 (520–738) 2,186 § 884 (632–3,149) 2,108 § 327 (1,555–2,613) 3,034 § 374 (2,206–3,960) 4,680 § 692 (3,102–5,876) 1,541 § 418 (855–1,839) 2,979 § 1,474 (1,652–5,540) 2,333 § 719 (1,664–3,480) 3,624 § 981 (2,540–5,530) 5,303 § 1,345 (3,339–6,980) 5.2 § 1.9 (3–12)

50.53 38.51 31.01 73.21 32.75 27.91 8.97 11.53 9.98 11.77 63.82 8.91 33.36 16.12 11.45 14.14 15.61 47.24 41.99 26.36 14.56 34.37

56.46 39.15 31.46 73.82 33.25 27.45 15.06 15.97 11.80 13.60 63.30 11.93 40.75 15.81 11.36 14.07 23.06 49.87 31.77 25.49 23.04 34.51

0.98 1.02 1.01 1.01 1.02 0.98 1.68 1.39 1.18 1.16 0.99 1.34 1.22 0.98 0.99 1.00 1.48 1.06 0.76 0.97 1.58 1.00

1.12 NS 1.74 NS 2.92 NS 0.09 NS 1.87 NS 0.61 NS 4.30* 13.36*** 8.17** 6.83** 0.04 NS 14.10*** 0.42 NS O.64 NS 0.00 NS 0.58 NS 0.67 NS 1.5 NS 0.67 NS 0.38 NS 2.67 NS 0.08 NS

Abbreviation for the statistical parameters: X mean, SD standard deviation, CVbs and Cvis between-sex and within-sex coeYcients of variation, PSC potential of sexual coding; see text for the description of measured parameters * P < 0.05, ** P < 0.01, *** P < 0.001

123

Polar Biol (2007) 30:1027–1033

of the next «wake» is the only temporal variable which diVers signiWcantly between the sexes. To sum up the call diVerences between the sexes, it appears that females emit signals with a lower fundamental frequency than males and with a lower temporal rhythm (the intervals between successive “kittiwake” calls are longer; conversely the number of calls emitted by bout does not diVer between males and females). Apart these diVerences, female and male calls show great similarities with regards to intra-call temporal variables (duration of the syllables and silences), call modulation frequency, as well as energy distribution within the call spectrum. On the basis of a discriminant analysis using the principal components calculated with the variables showing a signiWcant between-sex variation, we were able to correctly classify 100% of the recorded signals as male or female calls. If the discriminant analysis is done over the signiWcant variables related to frequency alone, the percentage of correctly classiWed calls remains very high (95.65%). Calls are individually distinct The results of Kruskal–Wallis tests (Table 2) and the calculation of PICs indicate that most of the measured variables are individually determined and thus participate to an individual vocal signature. Both temporal and frequency variables are strongly identiWed as potential cues to discriminate between individuals. The only variables which show no signiWcant variation between individuals are the number of calls emitted by bout and those related to the energy distribution among the spectrum of the second part of the «wake». On the basis of a discriminant analysis using the principal components calculated from the temporal variables alone, we were able to correctly classify 58.26% of the calls (analysis done on 23 individuals with a minimum of 3 calls/individual). When using only the variables belonging to the frequency domain, 71.90% of the calls were correctly classiWed. Finally, when both temporal and frequency variables were taken into account, 93.04% of the calls were correctly classiWed.

Discussion The “kittiwake” call is emitted by both sexes, mainly during pre-pairing periods and at the beginning of each reunion period following a separation between partners (Danchin 1987, 1991). On this basis, we can suppose that the signal is susceptible to carry information

1031 Table 2 Comparison between individuals of acoustic parameters measured on “kittiwake” calls Parameters (number of individuals)

Mean CVwi

CVbi

PIC

ANOVA Kruskal– Wallis H

Tki (19) Ts1 (19) Tti (23) Ts2 (23) Tdebwake (23) Twake (23) Tinterwake (23) Fdep (23) Fex (23) FWn (23) DeltaF (23) F0 (23) MaxFre (23) Quarter (23) Half (23) Threequarter (23) F0b (10) MaxFreb (10) Quarterb (10) Halfb (10) Threequarterb (10) Ncallsperbout (23)

26.49 25.93 19.22 39.28 16.56 24.84 11.05 8.65 8.00 7.05 54.95 6.41 38.23 12.77 14.21 11.39 8.20 33.93 26.65 22.07 20.67 56.61

43.55 36.61 28.33 73.29 30.86 32.64 15.39 14.76 11.99 11.99 62.45 12.06 44.75 20.27 16.05 13.90 12.24 42.09 33.33 25.33 20.81 56.68

1.64 1.41 1.47 1.86 1.86 1.31 1.39 1.71 1.50 1.70 1.14 1.88 1.17 1.59 1.13 1.22 1.49 1.24 1.25 1.15 1.00 1.00

61.54*** 40.97** 63.45*** 75.46*** 84.03*** 77.81*** 73.07*** 85.75*** 75.36*** 82.33*** 37.14* 91.68*** 35.43* 80.85*** 41.53** 43.52** 32.54*** 18.92* 16.49 NS 16.01 NS 14.92 NS 24.8 NS

Abbreviation for the statistical parameters: CVwi and Cvbi within and between individuals coeYcients of variation, PIC potential of individual coding; see text for the description of measured parameters * P < 0.05, ** P < 0.01, *** P < 0.001

about the sex and identity of the emitter. In the Kittiwake, the sexes are similar in visual appearance and a sexual dimorphism of voice could play a role in mate choice, territory defence against same-sexed birds and to modulate behavioural responses according to the sex of the emitter (Morton 1996). For instance, Brooke (1978) showed that in the Manx shearwater PuVinus puVinus, incubating males respond preferentially to male calls and incubating females to female calls. Contrary to the statement of Wooller (1978), our acoustic analysis shows that there is a clear distinctive sexual dimorphism of voice in the Kittiwake. The discriminant analysis correctly classiWed 100% of the signals according sex, with the rhythm of call production and above all the frequency value of the fundamental being the most discriminant parameters. The value of the fundamental has been often presented as reXecting the size or weight of individuals in numerous species (Wallschläger 1980; Fletcher 2004). The fundamental frequency being lower for females than for males, this can lead to the assumption that, in kittiwakes, females are greater in size and/or bigger than males. In fact, anatomo-morphological measures provide opposite results (Barrett et al. 1985): compared to males,

123

1032

females are slightly inferior in body size and weight. This underlines the fact that the relationship between body size or weight and fundamental frequency value is only a general tendency that suVers exceptions. In the present study, the “kittiwake” calls have been recorded during a unique breeding season. It is possible that the observed sexual dimorphism would be less sensible at other periods of the year. Moreover, although a preliminary study over a 6 years period (cited in Wooller 1978) showed that the calls’ structure of an individual seems to be stable from year to year, the individual signature might change over years. Further investigations are needed to investigate these questions. The combination of monomorphic adult plumage and vocal greeting displays, which occurs in many colonial species, suggests that sound is likely to play a role in kittiwake mate recognition. In addition, Kittiwakes breed in dense colonies, with nests on narrow edges that they defend against intruders, and call recognition may assure the mate of a landing space free of attack (Falls 1982). This is especially so when considering that the call is notably emitted by an individual in the process of landing on its site (Danchin 1987). Further, in long-lived species such as Black-legged Kittiwakes, mate recognition may allow the same pair to reform from year to year (Wooller 1978; Coulson and Thomas 1983). Wooller (1978) eVectively demonstrated by playback experiments that 82% of the tested birds responded vocally to the call of their partner whereas only 27% of them responded to a non-partner’s call, showing that an individual vocal recognition could occur between mates. At last, another role of the individuality of the “kittiwake” call could be at the end of each breeding season, when chicks just Xedged from their nest may use their parents’ calls to Wnd their nest (Danchin personal communication; Boulinier personal observation), even though there is no apparent parentoVspring recognition at an early age (Storey et al. 1992). Once discrimination by sounds has been proved, it remains to determine which acoustic features are involved. Individual variation at the level of the call structure is a prerequisite for individual recognition. An individually distinctive signal must be stereotyped within each individual and must vary among individuals. Our PIC analysis shows that we are in this case for all the parameters measured in the temporal and the frequency domains, except one: the high frequencies of the Wnal part of the “wake”. In his preliminary study, Wooller (1978) suggested that the individuality of the “kittiwake” call could be due to “the length and the tonal quality of its last part” (i.e. the “wake”). But, as underlined by the author himself, the analysis was imprecise as it was mainly based on a visual inspection

123

Polar Biol (2007) 30:1027–1033

of spectrographic representations. In fact, according to our discriminant analysis, the individuality is supported by a complex of temporal and frequency parameters located not only on the “wake” part, but also on the other parts (“ki” and “ti”). The existence of a redundant and multi-parametric coding of individual information could be seen as a strategy to facilitate the detection of individuals in a crowded colonial context, with a considerable background noise (Aubin and Jouventin 1998). In addition, an identiWcation system based on several parameters may secure the vocal signature and reduce the risks of confusion between birds. A given individual could thus potentially identify any bird in the colony on the basis of an acoustic analysis. But until now, in Kittiwakes, only mate vocal recognition has been proved experimentally and there is still no evidence of recognition of neighbours or other individuals. The potential for individual recognition thus not only opens opportunities for research in the context of mate choice and pair bond, but also in studies on breeding habitat selection in a colonial context, where interactions with conspeciWc neighbours may be important (e.g. Danchin et al. 1998; Boulinier et al. 2002). Acknowledgments We are grateful to Rob Barrett, Kjell Einar Erikstad, Julien Gasparini, Karen McCoy, Audrey Simon and Torkild Tveraa for their help at various stages of the work. Financial support was provided by the French Polar Institute (IPEV, programme 333). We also thank Fylkes mannen i Finnmark and Kystverket for allowing us to carry out work on Hornøya. At last, we thank two anonymous referees for comments and suggestions.

References Aubin T, Jouventin P (1998) Coktail-party eVect in king penguin colonies. Proc R Soc Lond B 265:1665–1673 Aubin T, Jouventin P (2002) How to identify vocally a kin in a crowd? The penguin model. Adv Stud Behav 31:243–277 Barrett RT, Tertitski GM (2000) Black-legged kittiwake Rissa tridactyla. In: Anker-Nilssen Bakken V, Strom H, Golovkin AN, Bianki VV, Tatarinkova IP (eds) The status of marine birds breeding in the Barents sea region. Norsk Polarinstitutt Rapportserie, Tromsø, pp 100–103 Barrett RT, Fieler R, Anker-Nilssen T, Rikardsen F (1985) Measurements and weight changes of Norwegian adult PuYns Fratercula arctica and Kittiwakes Rissa tridactyla during the breeding season. Ringing Migr 6:102–112 Boulinier T, Yoccoz NG, McCoy KD, Erikstad KE, Tveraa T (2002) Testing the eVect of conspeciWc reproductive success on dispersal and recruitment decisions in a colonial bird: design issues. J App Stat 29:509–520 Brackenbury J (1982) The structural basis of voice production and its relationship to sound characteristics. In: Kroodsma DE, Miller EH (eds) Acoustic communication in birds Vol 1. Academic, New York, pp 53–73 Brooke ML (1978) Sexual diVerences in the voice and individual vocal recognition in the Manx shearwater (PuVinus puVinus). Anim Behav 26:622–629

Polar Biol (2007) 30:1027–1033 Charrier I, Mathevon N, Jouventin P, Aubin T (2001) Acoustic communication in a Black-headed Gull colony: How do chicks identify their parents? Ethology 107:961–974 Charrier I, Mathevon N, Jouventin P (2002) How does a fur seal mother recognize the voice of her pup? An experimental study of Arctocephalus tropicalis. J Exp Biol 205:603– 612 Coulson JC, Thomas C (1983) Mate choice in the kittiwake gulls, Rissa tridactyla (L.). In: Bateson P (ed) Mate choice. Cambridge University Press, Cambridge, pp 361–376 Cramp S, Simmons KEL (1983) Handbook of the birds of Europe, the Middle East and North Africa, Vol.3. Oxford University Press, Oxford, pp 913 Danchin E (1987) The behaviour associated with the occupation of breeding site in the kittiwake gull Rissa tridactyla – the social status of landing birds. Anim Behav 35:81–93 Danchin E (1991) Social displays of the kittiwake Rissa tridactyla. Bird Behaviour 9:69–80 Danchin E, Monnat JY (1992) Population dynamics modelling of two neighbouring kittiwake Rissa tridactyla colonies. Ardea 80:171–180 Danchin E, Boulinier T, Massot M (1998) ConspeciWc reproductive success and breeding habitat selection: implications for the study of coloniality. Ecology 79:2415–2428 Erikstad KE, Tveraa T, Barret RT (1995) Adult survival and chick production in long lived seabirds: a 5-year study of the kittiwake Rissa tridactyla. In: Skjoldal HR, Hopkins C, Erikstad KE, Leinaas HP (eds) Ecology of fjords and coastal waters. Elsevier, Amsterdam, pp 471–477 Evans RM (1970) Imprinting and mobility in young ring-billed gulls Larus delawarensis. Anim Behav Monogr 3:195–248 Falls B (1982) Individual recognition by sounds in birds. In: Kroodsma DE, Miller EH (eds) Acoustic communication in birds Vol 2. Academic, New York, pp 237–278

1033 Fletcher NH (2004). A simple frequency-scaling rule for animal communication. J Acous Soc Am 115:2334–2338 Gasparini J, McCoy KD, Tveraa T, Boulinier T (2002) Related concentrations of speciWc immunoglobulins against the Lyme disease agent Borrelia burgdorferi sensu lato in eggs, young and adults of the kittiwake (Rissa tridactyla). Ecol Let 5:519–524 Jouventin P, Aubin T (2002) Acoustic systems are adapted to breeding ecologies: individual recognition in nesting penguins. Anim Behav 64:747–757 Mathevon N (1997) Individuality of contact calls in the Greater Flamingo Phoenicopterus ruber and the problem of background noise in a colony. Ibis 139:513–517 Mathevon N, Charrier I, Jouventin P (2003) Potential of individual recognition in acoustic signals: A comparative study of two gulls with diVerent nesting patterns. C R Biol 326:329–337 Morton ES (1996) A comparison of vocal behaviour among tropical and temperate passerine birds. In: Kroodsma DE, Miller EH (eds) Ecology and evolution of acoustic communication in birds. Cornell University Press, Ithaca, pp 258–281 Robisson P, Aubin T, Brémond JC (1993) Individuality in the voice of emperor penguin Aptenodytes forsteri: adaptation to a noisy environment. Ethology 94:279–290 Scherrer B (1984) Biostatistique. Gaëtan Morin, Chicoutimy, pp 850 Sokal RR, Rohlf FJ (1995) Biometry. W.H. Freeman and Company, New York, pp 887 Storey AE, Anderson RE, Porter JM, MacCharles AM (1992) Absence of parent-young recognition in kittiwakes – a reexamination. Behaviour 120:302–323 Wallschläger D (1980) Correlate of song frequency and body weight in passerine birds. Experientia 36:412 Wooller RD (1978) Individual vocal recognition in the Kittiwake gull, Rissa tridactyla Z für. Tierpsychol 48:68–86

123