Male response to intruders is related to song ... - Springer Link

J Ethol (2010) 28:371–377 DOI 10.1007/s10164-009-0198-0

ARTICLE

Male response to intruders is related to song characteristics in Darwin’s small tree finch (Camarhynchus parvulus) Rebekah Christensen • Jeremy Robertson Sonia Kleindorfer

•

Received: 3 March 2009 / Accepted: 2 December 2009 / Published online: 22 December 2009 Ó Japan Ethological Society and Springer 2009

Abstract Bird song functions in mate choice and species recognition, and hence variation in song can contribute to divergence and reproductive isolation. We used playback experiments to examine male response to conspecific song in Darwin’s small tree finch. Song is a reliable signal of bill morphology in this species, and individuals displayed stronger response to songs of males with similar bill size. These findings suggest that, in the context of territorial defence, males discriminate between intruders on the basis of song characteristics. Given that male response to song may be examined as a proxy for female response, this study also implies that females could discriminate between males on the basis of song. The findings suggest that: (1) perceived threat of intruders is related to reproductive competition and not fighting assessment, and (2) geographical isolation is not required for biologically meaningful song variation in Darwin’s finches. Keywords Mating signal Playback Reproductive isolation Small tree finch Song

Introduction Mating signals can act as behavioural barriers to gene flow and contribute to the development of reproductive isolation, even among closely related species (Mayr 1963; West-Eberhard 1983; Price 1998). Mating signals are also

R. Christensen J. Robertson S. Kleindorfer (&) School of Biological Sciences, Flinders University, GPO Box 2100, Adelaide, SA 5001, Australia e-mail: [email protected]

relevant for species discrimination, which has been shown across a range of taxa (reviewed in Andersson 1994). In birds, song is an important signal for mate choice and species recognition that can contribute to premating isolation. Geographic variation in bird song has been linked with reproductive isolation across populations (Irwin et al. 2001; Podos 2007; Seddon and Tobias 2007), while changes in song over time can also contribute to the rise of behavioural barriers across populations (Derryberry 2007). Within the Darwin’s finch group, song is an important signal for mate attraction and territorial defence (Grant 1999). Song is also important for reproductive isolation between species, and between geographically distinct but conspecific populations (Ratcliffe and Grant 1985; Grant and Grant 2002; Podos 2007). Song can be influenced by cultural factors such as drift or imprinting, or by habitat characteristics (Bowman 1983; Gibbs 1990; Grant and Grant 1996). In addition, recent studies have demonstrated that song characteristics are influenced by bill morphology at both the inter- and intra-specific level (Podos 2001; Christensen et al. 2006; Huber and Podos 2006). In Darwin’s small tree finch, Camarhynchus parvulus, song is correlated with bill morphology on both Santa Cruz and Floreana Islands (Christensen et al. 2006; Christensen and Kleindorfer, in preparation). On Santa Cruz Island, birds with larger bills sing songs with slower trill rate, broader frequency bandwidth, and higher frequency, while on Floreana Island birds with smaller bills sing songs with those features (Christensen et al. 2006; Christensen and Kleindorfer, in preparation). Thus, within each island population, song has the potential to function as a reliable indicator of bill morphology. Given that song functions as a mating signal in this species, correlations between song and morphology could be significant for mate choice. For example, females could use song to select mates on the

123

372

basis of bill size. This suggestion is supported by the finding of assortative mating for bill size in the Santa Cruz population of small tree finch, and in other species of Darwin’s finch (Christensen and Kleindorfer 2007; Huber et al. 2007). Playback experiments are commonly used to examine response to variations in song, and hence to examine the biological significance of song variation. It is often impractical to undertake playback experiments on female subjects, and male response is instead tested as a possible proxy for female response (Irwin 2000; Balakrishnan and Sorenson 2006; Derryberry 2007; Podos 2007). In order to examine the role of song as a mating signal in the small tree finch, we examine male response to playback (as a proxy for female response) on both Floreana and Santa Cruz Islands. In particular, we use intraspecific playback tests to assess whether males differentiate their playback response based on song characteristics (that correlate with bill size) of the caller (cf. Ratcliffe and Grant 1985; Grant and Grant 2002; Podos 2007). Given assortative mating for bill size in the small tree finch, and the finding that song is a reliable indicator of bill morphology in this species, both males and females could use song to assess bill size of potential competitors or mates. In the context of assortative mating on bill size and competition for mates, males of similar bill sizes present a greater threat than males of different bill sizes. If song acts as a reliable indicator of bill size, there is also potential for males to use song in fighting assessment (Huntingford and Turner 1987; Andersson 1994). The ability to accurately assess the fighting ability or resourceholding potential of a competitor is advantageous for an individual, as they can avoid entering into contests in which they are likely to lose or incur significant losses (Maynard Smith 1982; Enquist and Leimar 1983; Bee et al. 1999). Fighting assessment has been observed across a wide range of taxa, and studies have also documented the use of acoustic signals for fighting assessment (Davies and Halliday 1978; Andersson 1994). For this reason, we also examine the potential for song to function in fighting assessment in the small tree finch. Thus, we tested the following predictions: (1) if perceived threat of conspecific intruders is linked to mating competition, then there will be stronger male response to playback of same-sized versus different-sized callers, and weaker response in paired versus unpaired males to all playback types; (2) if perceived threat of conspecific intruders is linked to fighting assessment or direct competition between males, then there will be directional male response to playback, with strongest response to smallerbilled playback, and weakest response to larger-billed playback (Parker 1974; Maynard Smith 1982).

123

J Ethol (2010) 28:371–377

Materials and methods Study species and sites Darwin’s small tree finch is a socially monogamous species that occurs primarily in the moist highland forest of the Galapagos Archipelago (Lack 1947; Grant 1999; Kleindorfer 2007). On each island our study sites were approximately 2 ha in size and located within the highland forest at Los Gemelos (0°370 S, 90°210 W, Santa Cruz) and Cerro Pajas (01°170 S, 090°270 W, Floreana). Breeding typically occurs between December and April, and the onset of breeding is dependent upon sufficient rainfall (Lack 1950). At the onset of breeding, males establish territories, build display nests and sing to attract females. Playback trials were carried out during the breeding seasons of 2005 and 2006. Trials were undertaken on Santa Cruz Island from 20 to 30 January 2005 and from 16 March to 1 April 2006 on Floreana Island. Mist-netting and location of males for testing Birds were mist-netted and marked with a numbered aluminium band and unique combination of plastic colour bands. Bill length (culmen length from anterior edge of nare opening), bill depth and bill width were measured to the nearest 0.1 mm using dial callipers. We systematically searched our study sites for colour-banded males and their nests. Once located, each nest was numbered and the location was recorded using a Garmin 12XC GPS. Male behaviour at the nest was observed to determine pairing status as unpaired (nest-building alone and singing) or paired (nest-building with female, feeding female, incubation, feeding nestlings). We conducted playback trials on a total of 13 males on Santa Cruz and 13 males on Floreana. Bill size factor Song in both island populations of small tree finch was correlated with overall bill size (a combination of length, depth and width) (Christensen et al. 2006; Christensen and Kleindorfer, in preparation). Thus, we created a bill size factor for our test males that incorporated these three dimensions, calculated by multiplying length 9 depth 9 width for each male. The bill size factor ranged from 296.5 to 429.4 on Santa Cruz, and from 292.3 to 408.0 on Floreana. Song recordings for playback trials Song recordings of colour-banded males were obtained during the breeding seasons of 2002 and 2004 for Santa

J Ethol (2010) 28:371–377

Cruz (Christensen et al. 2006), and 2004 and 2006 for Floreana (Christensen and Kleindorfer 2009). Recordings were made using a Sony DCD-100 DAT recorder or Sony WMD6 cassette recorder with Sennheiser ME 80 directional microphone (Christensen et al. 2006; Christensen and Kleindorfer 2009). Songs were transferred to Macintosh G4 Powerbook using M-Audio Firewire transfer and we selected the clearest field recordings using Peak 4.11 for Mac OSX (Bias Inc. Software). Songs from each island were used for the respective playbacks on that island. In total, we used song recordings of 20 males in the Santa Cruz playback trials and 13 males in the Floreana trials. Tracks for playback trials were created using Peak 4.11. Each playback track consisted of an initial 5 min of silence (pre-trial observation period) followed by a 5-min trial period with song. The first, third and fifth minutes of the trial period contained stimulus, while the second and fourth minutes were silent. Each stimulus minute consisted of a song followed by 6.5 s of silence (to mimic natural singing behaviour). This was repeated for the duration of the 1-min stimulus period. Playback trial tracks were adjusted to a consistent amplitude before being transferred to compact disc (CD), and were played at constant volume through a portable CD player and speaker in the field. The speaker was placed within a 2–5 m radius of the male’s nest and was positioned at least 2 m off the ground. Playback trials All males were tested six times in total. Each male was tested twice with songs of: (1) two individuals of larger bill size (bill size factors at least 20 units larger than that of the focal male), (2) two individuals of smaller bill size (bill size factors at least 20 units smaller) and (3) two individuals of similar bill size (bill size factors within 20 units of the focal male). Males were only tested with song from their own island population. In one case, it was not possible to provide ‘Small’ stimuli to one individual as there were no songs available from males with smaller bill size than the focal male. This male was tested with ‘Large’ and ‘Same’ stimuli and only included in statistical comparisons of ‘Same’ and ‘Different’ (see below). All playback tests were carried out between 06:30 and 10:30 hours. Males were only tested once daily, and tests of each playback type (Small, Large, Same) were carried out in random order. Trials were not started until the male was observed within a 15 m radius of his nest, and males were never tested with song of direct neighbours. The small tree finch shows high site fidelity across years (Christensen and Kleindorfer, personal observation), meaning there is limited chance that neighbours change across years.

373

Response variables A single observer recorded male behaviour within 15 m of the nest. The observer was unaware of the type of playback stimulus being presented in each trial. Both acoustic and physical response to playback trials was recorded using three separate response variables. The acoustic response variable was SongCount, the number of songs during the trial period. A high value of SongCount indicates high singing activity and hence strong response to playback. There are two physical response variables: (1) latency to first response (s) (Latency), and (2) minimum approach (m) to the speaker during the trial period (MinApp). First response was defined as the first song following the end of the first stimulus song, or first direct movement towards the speaker following the first stimulus song. Prior to the playback trial commencing, two 10-m ropes were positioned at right angles below the target nest to aid in ocular estimation of MinApp. Small values of Latency and MinApp indicate strong response to playback. Statistical analysis SongCount, Latency and MinApp were all log-transformed prior to analyses in order to meet assumptions of the tests used. We used Kruskal–Wallis tests to compare the three response variables across individual playback stimuli songs. Repeated-measures analysis of variance (ANOVA) was used to compare the three response variables across playback trials. We pooled data from both islands and included ‘Island’ as a between-subjects factor to test for island variation. We found no effect of Island in any comparisons reported below. Male pairing status was also included as a between-subjects factor to test for variation in response across paired and unpaired males. All analyses were carried out using SPSS 11 for Mac OSX.

Results Comparison of trial songs Due to the limited sample of songs available for use in playback tracks, it was not possible to present each test male with a completely unique set of stimuli, and some males were tested using the same stimulus songs. To examine for possible effects of individual songs (e.g. consistently stronger response to a particular song across all individuals) we used Kruskal–Wallis tests to compare response across each individual stimulus song. The results showed no significant differences in response across stimuli songs [Santa Cruz (df = 19): Latency H = 26.81, P = 0.11; MinApp H = 23.56, P = 0.21; SongCount

123

374

J Ethol (2010) 28:371–377

H = 14.24, P = 0.77; Floreana (df = 12): Latency H = 20.00, P = 0.07; MinApp H = 18.29, P = 0.11; SongCount H = 11.82, P = 0.46]. Comparison of three trial types We calculated the means of response variables for the Small and Large trials for each individual, and also the overall mean response to Small and Large trials combined for each individual to assess response to Different size males. We also calculated the mean response of the two Same trials for each individual. We found no significant differences in SongCount or Latency when comparing response across Small, Same and Large trials (Table 1). However, there was a significant difference in MinApp across trial types (Table 1). Inspection of the mean values of MinApp for each trial type showed the closest approach in Same trials, followed by Large then Small trials (Table 1). Individual post hoc comparisons across each pairing of playback trial type (Large–Small, Large–Same, Small–Same) showed significant differences between Same and both Large (F1,22 = 4.20, R = 0.40, P = 0.05) and Small (F1,21 = 9.50, R = 0.56, P = 0.006) trials. However, there was no significant difference between Small and Large trials (F1,21 = 0.354, R = 0.13, P = 0.56). Comparison of Same and Different trials We found no significant differences in SongCount or Latency when comparing Same and Different playback trials (Table 1). There was a significant difference in MinApp across trials, with significantly closer approach in Same trials in comparison with Different trials (Table 1).

Male pairing status and response to playback We found a significant difference in SongCount between paired and unpaired males, which held in both the comparison of three trial types and the comparison of Same and Different trials (Table 2). Unpaired males sang significantly more during playback trials than paired males (Table 2). We also found a significant difference in Latency across paired and unpaired males, with unpaired males taking significantly less time to respond to playback trials (Table 2). This significant difference was only evident in the comparison of three playback trials (Table 2). There was no difference in MinApp across paired and unpaired males (Table 2).

Discussion We used playback experiments to examine male response to conspecific song in the small tree finch. We undertook experiments on Santa Cruz and Floreana Islands, and measured male response by recording three variables: Latency, MinApp and SongCount. On both islands, individuals made significantly closer approaches to the playback speaker when they were played the songs of a male with similar bill size. There were no significant differences in either Latency or SongCount across playback trials. However, there was a consistent trend for males to show stronger response to the songs of males with similar bill size (see mean response to playback trials, Table 1). Male small tree finch song functions in territorial defence and mate attraction (Christensen et al. 2006). Our findings suggest that, in the context of territorial defence, males discriminate between intruders on the basis of song.

Table 1 Results of within-subjects contrasts calculated in repeated-measures ANOVA to test for variation in response variables across playback trials Response variable

Large

Small

Same

Different

Test

F

R

P

SongCount

12.04 ± 9.31

12.18 ± 9.39

11.44 ± 8.82

12.11 ± 9.25

Three-trial comparison Same versus Different comparison

0.39 0.74

0.13 0.18

0.54a 0.40a

Latency (s)

40.06 ± 49.24

62.24 ± 66.28

43.25 ± 58.48

50.93 ± 58.70

Three-trial comparison

0.30

0.12

0.59a

Same versus Different comparison

1.30

0.23

0.26

Minimum approach (m)

3.77 ± 2.59

4.27 ± 2.92

2.77 ± 2.09

4.01 ± 2.74


4.17

0.40

0.05


9.14

0.53

0.006

Mean ± standard deviation (SD) is shown for each response variable across playback trials. ‘Three-trial comparison’ compared response across Large, Small and Same trials (df = 2,22). ‘Same versus Different comparison’ compared response across Same and Different trials (df = 1,24). Data from Santa Cruz and Floreana Islands are combined; Island was included as a between-subjects factor, but no Island effects were found. Bold values indicate statistically significant difference (P \ 0.05) in male response across trials. Male pairing status was also included as a between-subjects factor; effects of pairing status on response are shown in Table 2 a

Significant effect of pairing status, see Table 2 for statistics

123

J Ethol (2010) 28:371–377

375

Table 2 Effect of male pairing status on response to playback Response variable SongCount Latency (s) Minimum approach (m)

Paired 8.286 ± 8.64

Unpaired 12.02 ± 9.69

110.93 ± 114.38

71.13 ± 93.85

6.07 ± 5.16

5.02 ± 3.92

Test

F

R

P 0.03


5.22

0.43


5.63

0.44

0.03


4.96

0.42

0.04


3.20

0.34

0.09


0.01

0.02

0.91


0.02

0.03

0.90

Results of between-subjects contrasts calculated by repeated-measures ANOVA to test for variation in response variables across playback trials. Mean ± SD is shown for each response variable across male pairing status. ‘Three-trial comparison’ compared response across Large, Small and Same trials (df = 2,22). ‘Same versus Different comparison’ compared response across Same and Different trials (df = 1,24). Data from Santa Cruz and Floreana Islands are combined; Island was included as a between-subjects factor, but no Island effects were found. Bold values indicate statistically significant difference (P \ 0.05) in response across paired and unpaired males

Because song is related to bill morphology in the small tree finch, and males responded more strongly to song of males of similar bill size, our findings imply that males may be able to use song to assess bill size. In contrast to many studies that examine fighting assessment, our study suggests that intruder threat in the small tree finch is not related to the physical threat an intruder poses. Instead, threat may be related to the competitive threat an intruder poses in competing for the same ‘subset’ of mates. Assortative mating for bill size has been documented in the small tree finch on Santa Cruz Island (Christensen and Kleindorfer 2007), but data is not available to examine assortative mating on Floreana Island. Where assortative mating occurs, males of similar bill size will compete for the same subset of potential mates and therefore pose greater threats to each other. This would justify stronger response to intruders of similar bill size, as we have observed. In a situation where males fight to maintain territory ownership or access to females, strong response to males of similar bill size is predicted due to equal matching of fighting abilities (assuming that bill size is related to fighting ability) (Parker 1974; Maynard Smith 1982). Stronger response is also predicted to smaller-billed individuals, as the territory owner is more likely to succeed in challenging a smaller individual (Parker 1974; Maynard Smith 1982). While we observed strongest response to playback of song from males of similar bill size, we did not observe different responses to playback of larger- or smaller-billed individuals (Table 1). This does not support the hypothesis that males use song in fighting assessment of competitors. In Darwin’s ground finches, song type is transmitted from father to son through an imprinting process during early development (Grant and Grant 1996, 1997). While no study to date has tested for a learning or imprinting process in the tree finches, it is likely that cultural inheritance of song also occurs in this group. Because bill morphology is

also highly heritable across all Darwin’s finches (Boag and Grant 1978; Boag 1983), males with similar bill size are likely to have more similar songs due to imprinting on similar tutors. Hence, male response to the song of males with similar-sized bills may also be influenced by the process of imprinting. However, we currently have no data available to test for this possibility. Response to playback trials also varied across paired and unpaired males. Unpaired males sang more during playback trials than paired males, and had shorter latency to response (Table 2). Song activity naturally varies across paired and unpaired males, as paired males often reduce song output once they have successfully obtained a mate (Christensen and Kleindorfer, personal observation). This natural pattern likely explains the variation in response to playback trials that we observed. However, the shorter latency to response observed in unpaired males suggests that unpaired males may show consistently stronger response to intruders than do paired males. We have not examined extra-pair copulations in the small tree finch, but molecular analysis of parentage in other species of Darwin’s finches has shown only low levels of extra-pair copulation (Grant and Grant 1989; Petren et al. 1999). This suggests that, after a male successfully pairs with a mate, there is only a small likelihood of intruding males posing a competitive threat. Therefore, paired males may have reduced response to intruders due to the reduced reproductive threat they pose. In contrast, unpaired males are still competing for a mate and may have greater cause to defend their territory. Previous studies of Darwin’s finches have demonstrated an important role for song in maintaining reproductive isolation between species, or between geographically distinct conspecific populations (Ratcliffe and Grant 1985; Grant and Grant 2002; Podos 2007). In contrast to this prior research, our study examined the role of song as a mating signal at the intrapopulation level. Our findings suggest that song could be used by male small tree finches to assess the

123

376

bill size of competitors. Male response to playback trials may be viewed as a surrogate or proxy for female response (Irwin 2000; Balakrishnan and Sorenson 2006; Derryberry 2007; Podos 2007). Thus, our findings may also suggest that female small tree finches are able to use song to assess male bill size. This hypothesis is supported by the findings of assortative mating for bill morphology in the small tree finch on Santa Cruz Island (Christensen and Kleindorfer 2007). Because the bill is important for resource use in Darwin’s finches and is under strong natural selection (see Grant 1999; Grant and Grant 2008), any non-random mating linked to bill morphology could contribute to rapid divergence in Darwin’s finches (Podos 2001; Podos and Nowicki 2004; Huber and Podos 2006). Models of speciation in Darwin’s finches highlight the importance of geographical isolation between divergent lines for the build-up of divergence in morphological traits and mating signals (Grant 1999; Grant and Grant 2008). Our findings show that geographical isolation may not be necessary for the build-up of biologically meaningful variation in song in Darwin’s finches. We found varying responses to conspecific song relating to the bill morphology of the singer, suggesting that variation in song within populations can be biologically relevant in Darwin’s finches. Acknowledgments We are grateful to the Charles Darwin Research Station and Galapagos National Park Service for the opportunity to work on the Galapagos, and for logistical support. This work was generously supported by Flinders University (Research Establishment Grant to S.K., Overseas Field Trip Grant to R.C.), Conservation International and the American Bird Conservancy with awards to S.K., the Australian Federation of University Women SA with a Barbara Crase Bursary to R.C., and TAME Airlines who provided reduced airfares. Many thanks to R. Dudaniec and J. O’Connor for field assistance, and to F. Sulloway and two anonymous reviewers for insightful comments on earlier versions of this manuscript.

References Andersson M (1994) Sexual selection. Princeton University Press, Princeton Balakrishnan CN, Sorenson MD (2006) Song discrimination suggests premating isolation among sympatric indigobird species and host races. Behav Ecol 17:473–478 Bee MA, Perrill SA, Owen PC (1999) Size assessment in simulated territorial encounters between male green frogs (Rana clamitans). Behav Ecol Sociobiol 45:177–184 Boag PT (1983) The heritability of external morphology in Darwin’s ground finches (Geospiza) on Isla Daphne Major, Galapagos. Evolution 37:877–894 Boag PT, Grant PR (1978) Heritability of external morphology in Darwin’s finches. Nature 274:793–794 Bowman RI (1983) The evolution of song in Darwin’s finches. In: Bowman RI, Berson M, Leviton AE (eds) Patterns of evolution in Galapagos organisms. American Association for the Advancement of Science, San Francisco, pp 237–537

123

J Ethol (2010) 28:371–377 Christensen R, Kleindorfer S (2009) Bill morphology does not influence vocal performance in Darwin’s small tree finch on Floreana Island. Zool Res 30:423–428 Christensen R, Kleindorfer S (2007) Assortative pairing and divergent evolution in Darwin’s small tree finch, Camarhynchus parvulus. J Ornithol 148:463–470 Christensen R, Kleindorfer S, Robertson JGM (2006) Song is a reliable signal of bill morphology in Darwin’s small tree finch, Camarhynchus parvulus, and vocal performance predicts male pairing success. J Avian Biol 37:617–624 Davies NB, Halliday TR (1978) Deep croaks and fighting assessment in toads Bufo bufo. Nature 274:683–685 Derryberry EP (2007) Evolution of bird song affects signal efficacy: an experimental test using historical and current signals. Evolution 61:1938–1945 Enquist M, Leimar O (1983) Evolution of fighting behaviour: the effect of variation in resource value. J Theor Biol 127:187–205 Gibbs HL (1990) Cultural evolution of male song types in Darwin’s medium ground finches, Geospiza fortis. Anim Behav 39:253– 263 Grant PR (1999) Ecology and evolution of Darwin’s finches. Princeton University Press, Princeton (Reprint Edition) Grant BR, Grant PR (1989) Evolutionary dynamics of a natural population: the large Cactus Finch of the Galapagos. The University of Chicago Press, Chicago Grant BR, Grant PR (1996) Cultural inheritance of song and its role in the evolution of Darwin’s finches. Evolution 50:2471–2487 Grant PR, Grant BR (1997) Hybridization, sexual imprinting, and mate choice. Am Nat 149:1–28 Grant BR, Grant PR (2002) Simulating secondary contact in allopatric speciation: an empirical test of premating isolation. Biol J Linn Soc 76:545–556 Grant PR, Grant BR (2008) How and why species multiply: the radiation of Darwin’s finches. Princeton University Press, Princeton Huber SK, Podos J (2006) Beak morphology and song features covary in a population of Darwin’s finches (Geospiza fortis). Biol J Linn Soc 88:489–498 Huber SK, de Leon LF, Hendry AP, Bermingham E, Podos J (2007) Reproductive isolation of sympatric morphs in a population of Darwin’s finches. Proc R Soc Lond B 274:1709–1714 Huntingford F, Turner A (1987) Animal conflict. Chapman and Hall, London Irwin DE (2000) Song variation in an avian ring species. Evolution 54:998–1010 Irwin DE, Bensch S, Price TD (2001) Speciation in a ring. Nature 409:333–337 Kleindorfer S (2007) Nesting success in Darwin’s small tree finch, Camarhynchus parvulus: evidence of female preference for older males and more concealed nests. Anim Behav 74:795–804 Lack D (1947) Darwin’s finches. Cambridge University Press, Cambridge Lack D (1950) Breeding seasons in the Galapagos. Ibis 92:268–278 Maynard Smith J (1982) Evolution and the theory of games. Cambridge University Press, Cambridge Mayr E (1963) Animal species and evolution. Harvard University Press, Cambridge Parker GA (1974) Assessment strategy and the evolution of fighting behaviour. J Theor Biol 47:223–243 Petren K, Grant RB, Grant PR (1999) Low extrapair paternity in the Cactus Finch (Geospiza scandens). Auk 116:252–256 Podos J (2001) Correlated evolution of morphology and vocal signal structure in Darwin’s finches. Nature 409:185–188 Podos J (2007) Discrimination of geographical song variants by Darwin’s finches. Anim Behav 73:833–844

J Ethol (2010) 28:371–377 Podos J, Nowicki S (2004) Beaks, adaptation, and vocal evolution in Darwin’s finches. Bioscience 54:501–510 Price TD (1998) Sexual selection and natural selection in bird speciation. Phil Trans R Soc Lond B 353:251–260 Ratcliffe LM, Grant PR (1985) Species recognition in Darwin’s finches (Geospiza Gould). III. Male responses to playback of different song types, dialects and heterospecific songs. Anim Behav 33:290–307

377 Seddon N, Tobias JA (2007) Song divergence at the edge of Amazonia: an empirical test of the peripatric speciation model. Biol J Linn Soc 90:173–188 West-Eberhard M (1983) Sexual selection, social competition, and speciation. Q Rev Biol 58:155–183

123