SEARCH

SEARCH BY CITATION

Keywords:

  • body size;
  • passerine song loudness;
  • song complexity;
  • song syntax

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Songs of passerines are generally complex, long-range acoustic signals, and are highly diverse across species. This diversity must nevertheless be shaped by the capabilities of the avian vocal physiology. For example, within species, loudness has been shown to trade-off with aspects of song complexity. Here, I ask if such trade-offs with loudness influenced the evolutionary diversification of song among passerines. Comparing perceived song loudness across > 140 European and North American species showed that loudness is positively related to body size and to singing with simple trilled syntax, and negatively related to aspects of syllable complexity. Syntax and syllable phonology together explained more variation than body size did, indicating that the acoustic design of songs is an important factor determining loudness. These results show for the first time that loudness covaries with, and possibly limits, song complexity across species, suggesting that a trade-off with loudness shaped the evolutionary diversification of passerine song.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Passerines are characterized by having complex songs, which are used mostly in long-range communication (Catchpole & Slater, 2008). Species differ widely in song phonology (spectral characteristics of sounds) and syntax (ordering of syllables to produce longer songs) (Catchpole & Slater, 2008), and also in loudness (Calder, 1990). What shapes this extraordinary diversity of birdsong across species has largely remained elusive (Garamszegi & Møller, 2004; Boncaraglio & Saino, 2007). Song diversity must to some extent be limited by the capabilities of the avian vocal physiology (Podos et al., 2004). For example, during continuous singing many birds need to take small inspirations (mini-breaths) in between syllables to replenish air volume, and this limits how long syllables or uninterrupted songs can be (Hartley & Suthers, 1989; Forstmeier et al., 2002). Also, many songbird species sing faster syllable repetitions using a narrow frequency bandwidth (Podos, 1997), suggesting that frequency bandwidth limits how fast syllable repetitions can be sung. Knowledge of this type of limitations and trade-offs is necessary to understand the design and diversity of bird song (Podos et al., 2004).

Singing loudly is physiologically demanding because it requires stronger contraction of the respiratory muscles to generate higher air-sac pressures and faster airflow through the syrinx (Suthers et al., 1999). Also, complex frequency or temporal modulation of birdsong requires changes in the configuration of the syrinx and the vocal tract (Suthers et al., 1999), which likely interferes with expiratory muscle contraction or with airflow. Therefore, loudness may trade-off with aspects of song complexity. Suggestive of this is the fact that, in species that sing both full song and soft song, the latter can be more varied and complex (e.g. Titus, 1998; Anderson et al., 2008). The same happens to some extent during the soft and variable sub-song and plastic song stages of song development in oscines, compared to full song (Marler, 1990; Brumm & Hultsch, 2001). More direct evidence for trade-offs with loudness comes from comparisons of sound amplitude of syllables within song bouts of individual birds. These have shown that in some species loudness is negatively related to high sound frequency (Christie et al., 2004), fast rattles (Cardoso et al., 2007), the number of elements in the syllable (Cardoso et al., 2007; Cardoso & Mota, 2009) and, for species that alternate repeated and nonrepeated syntax (i.e. the same syllable repeated several times, and different syllables sung consecutively), syllables are louder in the simpler syntax (syllable repetitions; Cardoso & Mota, 2009).

It is not known if this type of trade-offs with loudness influenced the evolutionary diversification of birdsong across species. This is, if species that sing louder tend to have simpler songs in some way. The greatest difficulty for testing this hypothesis is that, although song loudness differs noticeably among species (Calder, 1990), it is difficult to quantify. Although other song traits can be measured from recordings, comparing loudness across species requires recordings or sound amplitude measurements made at known distances from the birds and in identical conditions (Brumm, 2004; Patricelli et al., 2007; Anderson et al., 2008). Such data do not currently exist that could be compared across a reasonably large number of species (reviewed in Anderson et al., 2008), nor is it foreseeable that it can be obtained in the near future.

To circumvent this problem, I obtained scores of passerine song loudness by panels of experienced field ornithologists from two species-rich regions, Portugal and southern California. I then used this data, together with song measurements from recordings, to test the hypothesis that differences in song loudness among species are related to song phonology and syntax, controlling for phylogeny, body size and vegetation density of the species’ habitat. Body size is controlled for because it likely influences song loudness across species (Calder, 1990; Jurisevic & Sanderson, 1998; but perhaps not within-species, see Brumm, 2009), as larger species have stronger expiratory muscles and can dispose of larger volumes of air. Dense vegetation selects for low frequency and other traits that either transmit better or minimize degradation in reverberant environments (Boncaraglio & Saino, 2007), and we might hypothesize that it also selects for loudness to compensate for reduced visual communication in closed habitats. Therefore, a score of vegetation density is used both to test for an effect on loudness, and to control for possible indirect relations between song traits and loudness mediated by habitat type.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Song loudness scores

I obtained loudness scores by ornithologists with extensive field experience: all had recently been major contributors gathering data for bird atlases in Portugal (ICNB, 2008) and southern California (Unitt, 2004). Three observers from each region independently scored song loudness of the breeding passerines they were familiar with, without being told what the purpose of the study was until completion of the scoring. Each observer was given a list with the common and scientific names of all passerines breeding in Portugal or California in randomized order (excluding those species with no long-range song: Passer spp., Motacilla spp., Cisticola juncidis, Aegithalus caudatus, Psaltriparus minimus, Sitta pygmaea, Bombycilla cedrorum, and the Corvidae inhabiting these regions). The total number of species in the lists was 65 for Portugal and 120 for California. Each observer was asked to score song loudness in the species he or she was familiar with, first in a scale of 3 and then in a scale of 4, and to revise as needed. Only the species scored by all three observers were used in the analysis (60 species for Portugal and 89 for California). Inter-observer repeatability (calculated using mean squares from a one-way anova; Lessels & Boag, 1987) was higher in the scale of 4 (0.66 for the Portuguese sample and 0.56 for the Californian sample). Therefore, I used the average score of the three observers in the scale of 4 as the index of song loudness (Fig. 1).

image

Figure 1.  Scores of song loudness for (a) European and (b) North American species plotted against the index of loudness (average score of all observers). Dot size indicates the number of species. Different observers are represented with different shades of gray, and regression lines with the coloration index are drawn for each (all > 0.80). In each plot, the data from two observers are slightly displaced vertically so that all data are visible; regression lines are not displaced.

Download figure to PowerPoint

To date, the largest set of sound amplitude measurements using comparable methodologies is that compiled in Calder (1990), comprising 14 of the species studied here. For these species, the correlation coefficient between power output (in log mW at 1 m in front of the bird) and the loudness index was 0.61 (= 0.02, = 14). These values of repeatability and of correlation (all ≈ 0.6) are high considering that birds vary the loudness of their song to a certain extent (Cynx et al., 1998; Brumm, 2004; Patricelli et al., 2007; Anderson et al., 2008). Although subjective evaluations of song loudness necessarily carry a degree of inaccuracy, this indicates that they reflect true variation among species.

Song measurements

I measured 11 song traits, which together comprehensively describe variation in song among species, on spectrograms and amplitude spectra of all song recordings (not recordings of other vocalizations) in Perrins (1998) and Cornell Laboratory of Ornithology (1992), using the software avisoft-saslab pro v.4.40 (Avisoft Bioacoustics, Berlin, Germany). The song traits are syntax, song length, length of syllables, length of intervals between syllables, peak frequency, frequency bandwidth, number of elements per syllable, number of frequency inflections per syllable, proportion of each syllable with two voiced sounds, with harmonics, and with rattles.

Syllables were identified as isolated sound elements or groups of elements closely spaced comparatively to other such syllables. Songs were identified as groups of syllables separated from other songs by more than 0.5 s. This threshold was chosen after screening the recordings, as it yielded an intuitive delimitation of songs for all species. Measurements were made for every syllable and then averaged for each species, except for syntax and song length, which were made for every song and then averaged. Syntax is the proportion of trilled syllables in songs, and was calculated as the number of consecutively repeated syllables divided by the total number of syllables (e.g. for the syllable sequence abbbcd, syntax is 3/6 = 0.5).

Peak frequency was measured as the frequency with maximum amplitude in the power spectrum of the syllable. Frequency bandwidth was computed as maximum minus minimum frequencies, where maximum and minimum frequencies are the frequencies at which sound amplitude drops below minus 24 dB relatively to maximum amplitude in the power spectra (e.g. Podos, 1997). The other song measurements were made on spectrograms (resolution 43 Hz and 1.45 ms) following the methods in Cardoso & Mota (2007). Briefly, length of syllables and length of intervals are the durations of syllables and of intervals separating syllables within songs, elements per syllable is the number of temporally separated sounds within each syllable, number of frequency inflections is the number of times a rising frequency modulation is followed by a descending one or vice versa, and two voiced sounds, harmonics, and rattles are the proportion of each syllable’s length that contains, respectively, two voices (two simultaneous sounds produced by the two sides of the syrinx, and modulated in frequency independently of each other), harmonics (octaves of the fundamental frequency), or rattles (or ‘buzzes’: broad-band and harsh sounding fast modulations or repetitions within syllables, e.g. last syllable in figure 4D of Cardoso & Mota, 2007). In four species, all songs were monosyllabic (Anthus campestris, Lanius meridionalis, Ixoreus naevius and Vireo cassinii) and therefore no intervals between syllables could be measured. In these four species, the length of intervals was set to 0.5 s (the threshold to identify distinct songs). Five species did not have or had very noisy song recordings in the above sources, and similar measurements were instead made on the printed spectrograms of Cramp & Perrins (1994) and Brown (1984), using the average of maximum and minimum frequency as a proxy for peak frequency. These species are indicated in Table S1 (see Supporting Information). Sample sizes, values of the loudness index and of all the predictor variables for each species are also given in Table S1.

Vegetation density and body size

Breeding habitats were scored relatively to vegetation density as (1) open (e.g. desert with sparse vegetation, prairie, cultivated fields, rocky habitat), (2) semi-closed with low vegetation (e.g. dense brush, chaparral, marsh, riverine vegetation), (3) semi-closed with high vegetation (e.g. open forest, forest edge, desert yucca or cactuses) and (4) closed (forest) according to the descriptions in Cramp & Perrins (1994) and Poole (2005). Intermediate scores were used when breeding was described in more than one category (Table S1). As a measure of body size, I took body masses from Dunning (2008). When male and female body masses were reported, the male mass was used.

Analysis and phylogeny

I conducted a generalized least squares (GLS) multiple regression (Pagel, 1999) of the loudness index on the above predictors (song measurements, body size and vegetation density). As the loudness index for the European and North American samples was obtained from two different panels of observers, a dichotomous variable (continent) is fixed in the regression model (i.e. it is not removed even if its effect is not significant) to account for differences in scale between the two datasets. Phylogenetic GLS regressions were run with the software bayestraits (M. Pagel & A. Meade, available from http://www.evolution.rdg.ac.uk), each time estimating λ (Pagel, 1999; Freckleton et al., 2002) to adjust the phylogenetic correction to the degree of phylogenetic signal in the data.

Here, I present results using a phylogeny based primarily on Sibley & Ahlquist (1990), on which I added the species that were not originally present there using recent phylogenetic information (see below). I added species to the phylogeny using the convention that within families the distance between genera is 3.4 ΔT50H units, and between congeneric species is 1.1 ΔT50H units (Sibley & Ahlquist, 1990; Bennett & Owens, 2002). The complete phylogeny is given in Fig. S1. Aspects of the Sibley and Ahlquist phylogeny have been criticized, but it remains the only comprehensive phylogeny across the Passerines to include branch lengths. I also ran the analysis using a composite phylogeny based primarily on Barker et al. (2004), and then collating information from various recent phylogenies (see below). This composite phylogeny has no branch lengths because the sources differ in methodology. Therefore, I started with a tree with branch lengths proportional to the number of species in clades (as drawn in Fig. S2; e.g. Grafen, 1989; Garland et al., 1992), and estimated the parameter δ (which scales relative tree length from root to tips to fit the phenotypic data) in the first GLS regression. Results using this composite phylogeny are given in Table S2, and were identical to the ones presented here. Similarly, a multiple regression on simple species values (i.e. not controlling for phylogeny) also gives qualitatively identical results (Table S3), and plots of normality of residuals and of residuals vs. expected values in Fig. S3 show that the data meet the assumptions for multiple regression analysis.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

The full regression model of the loudness index on all predictors explained 36% of the variation among species in the loudness index, and identified body size, song syntax, peak frequency, harmonics and rattles as significantly related to loudness (Table 1). Stepwise removal of the nonsignificant predictors reached a final model that explained 33% of the variation, retaining the same predictors that were significant in the full model, with their partial regression coefficients little changed (Table 1): the loudness index was positively related to body size (partial standardized regression coefficient β = 0.39) and repeated syntax (β = 0.27), and negatively related to peak frequency (β = −0.35), harmonics (β = −0.22), and rattles (β = −0.22). Body size was the trait that explained the greatest amount of variation in loudness (11%, about one third of the explained variation), and the significant song traits explained between 4% and 9% of the variation each (Table 1). These relations are illustrated in Fig. 2.

Table 1.   GLS multiple regression of the loudness index on predictor variables.
 Full modelFinal model
  1. Values are given as β; r2 (P).

  2. β, partial standardized regression coefficient; r2, partial proportion of variation explained; P, two-tailed significance; R2, proportion of variation explained; λ, estimated scaling parameter to adjust the phylogenetic correction according to the fit of the data to phylogenetic distances (phylogeny in Fig. S1). The final model was obtained by step-wise removal of the least significant predictors until all P < 0.05 (except for continent, which is a fixed factor). Residual degrees of freedom are 149 (sample size) minus the number of predictors (14 in the full model and six in the final), minus 1.

Syntax (proportion of repeated syllables)0.234; 0.042 (0.004)0.267; 0.062 (< 0.001)
Song length0.053; 0.002 (0.496)
Length of syllables0.021; < 0.001 (0.821)
Length of intervals between syllables−0.100; 0.007 (0.236)
Peak frequency−0.311; 0.059 (0.001)−0.348; 0.097 (< 0.001)
Frequency bandwidth0.059; 0.002 (0.440)
Elements per syllable0.013; < 0.001 (0.866)
Frequency inflections per syllable0.114; 0.009 (0.169)
Two voiced sounds−0.144; 0.015 (0.081)
Harmonics−0.217; 0.027 (0.020)−0.223; 0.043 (0.003)
Rattles−0.204; 0.035 (0.006)−0.216; 0.052 (0.001)
Body mass0.418; 0.110 (< 0.001)0.388; 0.109 (< 0.001)
Vegetation density0.068; 0.003 (0.408)
Continent0.140; 0.013 (0.105)0.134; 0.014 (0.091)
R20.3580.330
λ0.4800.522
image

Figure 2.  Partial plots of residuals illustrating the relations of loudness with each of the significant predictor variables. Plotted are residuals when each variable is regressed separately on all other predictors, using simple species values. These are for illustrative purposes only, as the main analysis additionally controls for phylogeny. The outlier in the last panel is the largest species in the dataset (Quiscalus mexicanus); all results are qualitatively identical if removing this species.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Across passerine species, an index of song loudness was related to body size, song syntax, and the frequency, harmonics and rattles of songs. The direction of these relations supports the hypothesis that song complexity trades-off with loudness: species singing with simpler syntax (i.e. syllable repetitions), fewer harmonics, and fewer rattles, tended to be louder. These among-species correlations with loudness mirror results of the within-species comparisons done to date (Christie et al., 2004; Cardoso et al., 2007; Cardoso & Mota, 2009). Curiously, there was no evidence for an effect of the number of elements per syllable, which was the trait showing the strongest trade-offs with loudness in some within-species studies (Cardoso et al., 2007; Cardoso & Mota, 2009), possibly because syllables in different species are not homologous and directly comparable. Finally, there was no evidence for an effect of vegetation density on loudness.

Two aspects of the results need to be interpreted with caution. First, loudness scores may be biased if humans perceive certain song traits as louder. Although within human hearing range, high frequency avian vocalizations are less detectable by humans (Schieck, 1997; Simons et al., 2007), presumably because high frequency sound attenuates faster, and this may introduce bias in the scores. This may have contributed to the relation between peak frequency and the loudness index. A negative relation between frequency and loudness was reported before within-species (Christie et al., 2004), and the phenomenon is physiologically sensible because high frequencies involve increased muscle contraction and resistance at the syrinx, which, at a constant air sac pressure, reduces airflow (Goller & Suthers, 1996; Suthers et al., 1999). Nevertheless, the present among-species result must be taken as suggestive only. Apart from this, I am not aware that human perception of loudness introduces directional bias in relation to the song traits studied.

Second, an amount of error inevitably exists in song measurements due to sampling, and also in subjective scores of loudness. As these sources of inaccuracy are random in relation to the hypothesis tested (except regarding peak frequency), the R2 and partial r2 reported are underestimates of the proportions of variance that these predictors explain. A useful alternative to evaluate the importance of the trade-off between song complexity and loudness is to compare the r2 of the song traits to that of body size: even excluding peak frequency, the summed r2 of the significant syntax and phonology traits is larger than that of body size (Table 1), indicating that ‘what they sing’ is at least as important as ‘how big they are’ in explaining differences in song loudness among species.

These results indicate that trade-offs between loudness and aspects of song complexity were important shaping the evolutionary diversification of birdsong, which has implications for our understanding of sexual selection of song and the evolution of song complexity. Sexual selection often favours costly displays, and costly birdsongs are often those that have a combination of demanding vocal traits rather than exaggerate a single trait (Gil & Gahr, 2002; Podos et al., 2004). Therefore, it is predicted that complex song evolves through sexual selection, and that species with more complex songs are under stronger sexual selection (Read & Weary, 1992; Garamszegi & Møller, 2004). However, differences in loudness between species may confound the evaluation of song costs based on complexity alone, which contributes to explain why comparative studies often fail to link indices of sexual selection with song complexity (reviewed in Garamszegi & Møller, 2004).

These trade-offs with loudness also open new possibilities regarding how morphology and ecology may affect song evolution. For example, species that communicate at longer distances, because they have large territories or live at low densities, may need to evolve simpler songs. Or, from a different perspective, species that need to sing loudly relative to their body size may be less prone to evolve elaborate songs by sexual selection. Current knowledge of how ecology affects song evolution pertains mostly to sound transmission properties of forested vs. open habitats, and how this relates to the song frequency of species living in those habitats (reviewed in Boncaraglio & Saino, 2007). This, however, explains a small part of the diversity of song among species (Boncaraglio & Saino, 2007). Exploring the consequences of the loudness-complexity trade-offs, as exemplified in the hypotheses above, may provide a research avenue to further understand the diversity of birdsong among passerines.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

I thank Carlos Pacheco and Philip Unitt for indicating very knowledgeable field ornithologists and, together with Gonçalo Elias, Luís Gordinho, Mary Beth Stowe, and Lori Hargrove, for scoring song loudness.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information
  • Alström, P., Ericson, G.P., Olsson, U. & Sundberg, P. 2006. Phylogeny and classification of the avian superfamily Sylvioidea. Mol. Phylogenet. Evol. 38: 381397.
  • Anderson, R.C., Searcy, W.A., Peters, S. & Nowicki, S. 2008. Soft song in song sparrows: acoustic structure and implications for signal function. Ethology 114: 662676.
  • Araniz-Villena, A., Lowy, E., Ruiz-del-Valle, V., Westerdahl, H., Moscoso, J., Serrano-Vela, J.I., Witzell, H. & Zamora, J. 2007. Evolution of the major histocompatibility complex class I genes in Serinus canaria from the Canary Islands is different from that of Asian and African continental Serinus species. J. Ornithol. 18: S479S484.
  • Badyaev, A.V. 1997. Altitudinal variation in sexual dimorphism: a new pattern and alternative hypotheses. Behav. Ecol. 8: 675690.
  • Barhoum, D.N. & Burns, K.J. 2002. Phylogenetic relationships of the wrentit based on mitochondrial cytochrome b sequences. Condor 104: 740749.
  • Barker, F.K., Cibois, A., Schliker, P., Feinstein, J. & Cracraft, J. 2004. Phylogeny and diversification of the largest avian radiation. Proc. Natl Acad. Sci. USA 101: 1104011045.
  • Bennett, P.M. & Owens, I.P.F. 2002. Evolutionary Ecology of Birds. Oxford Univ. Press, Oxford.
  • Bleiweiss, R. 2007. On the ecological basis of interspecific homoplasy in carotenoid-bearing signals. Evolution 61: 28612878.
  • Blondel, J., Catzeflis, F. & Perret, P. 1996. Molecular phylogeny and the historical biogeography of the warblers of the genus Sylvia (Aves). J. Evol. Biol. 9: 871891.
  • Boncaraglio, G. & Saino, N. 2007. Habitat structure and the evolution of bird song: a meta-analysis of the evidence for the acoustic adaptation hypothesis. Funct. Ecol. 21: 134142.
  • Brown, C.R. 1984. Vocalizations of the purple martin. Condor 86: 433442.
  • Brumm, H. 2004. The impact of environmental noise on song amplitude in a territorial bird. J. Anim. Ecol. 73: 434440.
  • Brumm, H. 2009. Song amplitude and body size in birds. Behav. Ecol. Sociobiol. 63: 11571165.
  • Brumm, H. & Hultsch, H. 2001. Pattern amplitude is related to pattern imitation during the song development of nightingales. Anim. Behav. 61: 747754.
  • Calder, W.A. III 1990. The scaling of sound output and territory size: are they matched? Ecology 71: 18101816.
  • Cardoso, G.C. & Mota, P.G. 2007. Song diversification and complexity in canaries and seedeaters (Serinus spp.). Biol. J. Linn. Soc. 92: 183194.
  • Cardoso, G.C. & Mota, P.G. 2009. Loudness of syllables is related to syntax and phonology in the songs of canaries and seedeaters. Behaviour 146: 16491663.
  • Cardoso, G.C., Atwell, J.W., Ketterson, E.D. & Price, T.D. 2007. Inferring performance in the songs of dark-eyed juncos (Junco hyemalis). Behav. Ecol. 18: 10511057.
  • Carson, R.J. & Spicer, G.S. 2003. A phylogenetic analysis of the emberizid sparrows based on three mitochondrial genes. Mol. Phylogenet. Evol. 29: 4357.
  • Catchpole, C.K. & Slater, P.J.B. 2008. Bird Song: Biological Themes and Variations, 2nd edn. Cambridge University Press, Cambridge.
  • Christie, P.J., Mennill, D.J. & Ratcliffe, L.M. 2004. Pitch shifts and song structure indicate male quality in the dawn chorus of black-capped chickadees. Behav. Ecol. Sociobiol. 55: 341348.
  • Cornell Laboratory of Ornithology. 1992. Peterson Field Guides, Western Bird Songs. Houghton Mifflin, Boston, MA and Cornell Laboratory of Ornithology, Ithaca, NY.
  • Cramp, S. & Perrins, C.M. 1994. The Birds of the Western Paleartic (vol. VIII). Oxford Univ. Press, Oxford.
  • Cynx, J., Lewis, R., Tavel, B. & Tse, H. 1998. Amplitude regulation of vocalizations in noise by a songbird, Taeniopygia guttata. Anim. Behav. 56: 107113.
  • Dunning, J.B. 2008. CRC Handbook of Avian Body Masses. Taylor & Francis, Boca Raton, FL.
  • Forstmeier, W., Kempanaers, B., Meyer, A. & Leisler, B. 2002. A novel song parameter correlates with extra-pair paternity and reflects male longevity. Proc. Biol. Sci. 269: 14791485.
  • Freckleton, R.P., Harvey, P.H. & Pagel, M. 2002. Phylogenetic analysis and comparative data: a test and review of evidence. Am. Nat. 160: 712726.
  • Garamszegi, L.Z. & Møller, A.P. 2004. Extrapair paternity and the evolution of bird song. Behav. Ecol. 15: 508519.
  • Garland, T. Jr, Harvey, P.H. & Ives, A.R. 1992. Procedures for the analysis of comparative data using phylogenetically independent contrasts. Syst. Biol. 41: 1832.
  • Gil, D. & Gahr, M. 2002. The honesty of birdsong: multiple constraints for multiple traits. Trends Ecol. Evol. 17: 133141.
  • Gill, F.B., Slikas, B. & Sheldon, F.H. 2005. Phylogeny of titmice (Paridae): II. Species relationships based on sequences of the mitochondrial cytochrome-b gene. Auk 122: 121143.
  • Goller, F. & Suthers, R.A. 1996. Role of syringeal muscles in controlling the phonology of bird song. J. Neurophysiol. 76: 287300.
  • Grafen, A. 1989. The phylogenetic regression. Philos. Trans. R. Soc. Lond. B Biol. Sci. 326: 119157.
  • Graputto, A., Pliastro, A., Baker, A.J. & Maria, G. 2001. Molecular evidence for phylogenetic relationships among buntings and American sparrows (Emberizidae). J. Avian Biol. 32: 95101.
  • Groth, J.E. 1998. Molecular phylogenetics of finches and sparrows: consequences of character state removal in cytochrome b sequences. Mol. Phylogenet. Evol. 10: 377390.
  • Hartley, R.S. & Suthers, R.A. 1989. Airflow and pressure during canary song: direct evidence for mini-breaths. J. Comp. Physiol. A 165: 526.
  • ICNB 2008. Atlas das aves de Portugal. Instituto para a Conservação da Natureza e Biodiversidade/Assírio e Alvim, Lisbon.
  • Johnson, K.P. & Lanyon, S.M. 1999. Molecular systematics of the grackles and allies, and the effect of additional sequence (cyt b and nd2). Auk 116: 759768.
  • Jurisevic, M.A. & Sanderson, K.J. 1998. A comparative analysis of distress call structure in Australian passerine and non-passerine species: influence of size and phylogeny. J. Avian Biol. 29: 6171.
  • Klicka, J., Voelker, G. & Spellman, G.M. 2005. A molecular phylogenetic analysis of the ‘‘true thrushes’’ (Aves: Turdinae). Mol. Phylogenet. Evol. 34: 486500.
  • Lessels, C.M. & Boag, P.T. 1987. Unrepeatable repeatabilities: a common mistake. Auk 104: 116121.
  • Lovette, I.J. & Bermingham, E. 2002. What is a wood-warbler? molecular characterization of a monophyletic Parulidae. Auk 119: 695714.
  • Lovette, I.J. & Hochachka, W.M. 2006. Simultaneous effects of phylogenetic niche conservatism and competition on avian community structure. Ecology 87: S14S28.
  • Mann, N.I., Barker, F.K., Graves, J.A., Dingess-Mann, K.A. & Slater, P.J.B. 2006. Molecular data delineate four genera of “Thryothorus” wrens. Mol. Phylogenet. Evol. 40: 750759.
  • Marler, P. 1990. Song learning: the interface between behaviour and neuroethology. Philos. Trans. R. Soc. Lond. B Biol. Sci. 329: 109114.
  • Møller, A.P., Merino, S., Brown, C.R. & Robertson, R.J. 2001. Immune defense and host sociality: a comparative study of swallows and martins. Am. Nat. 158: 136145.
  • Møller, A.P., Nielsen, J.T. & Garamzegi, L.Z. 2008. Risk taking by singing males. Behav. Ecol. 19: 4153.
  • Pagel, M. 1999. Inferring the historical patterns of biological evolution. Nature 401: 877884.
  • Patricelli, G.L., Dantzker, M.-S. & Bradbury, J.W. 2007. Differences in acoustic directionality among vocalizations of the male red-winged blackbird (Agelaius pheoniceus) are related to function in communication. Behav. Ecol. Sociobiol. 61: 10991110.
  • Perrins, C.M. (Ed.) 1998. The Complete Birds of the Western Paleartic CD-ROM Version 1.0. Oxford Univ. Press, Oxford.
  • Podos, J. 1997. A performance constrain on the evolution of trilled vocalizations in a songbird family (Passeriformes: Emberizidae). Evolution 51: 537551.
  • Podos, J., Huber, S.K. & Taft, B. 2004. Bird song: the interface of evolution and mechanism. Ann. Rev. Ecol. Evol. Syst. 35: 5587.
  • Poole, A. (Ed.) 2005. The Birds of North America Online. Cornell Laboratory of Ornithology, Ithaca, NY.
  • Read, A.F. & Weary, D.M. 1992. The evolution of bird song: comparative analyses. Philos. Trans. R. Soc. Lond. B Biol. Sci. 338: 165187.
  • Schieck, J. 1997. Biased detection of bird vocalizations affects comparisons of bird abundance among forested habitats. Condor 99: 179190.
  • Sibley, C.G. & Ahlquist, J.E. 1990. Phylogeny and Classification of Birds: A Study in Molecular Evolution. Yale Univ. Press, New Haven, CT.
  • Simons, T.R., Alldredge, M.W., Pollock, K.H. & Wettroth, J.M. 2007. Experimental analysis of the auditory detection process on avian point counts. Auk 124: 986999.
  • Suthers, R.A., Goller, F. & Pytte, C. 1999. The neuromuscular control of birdsong. Philos. Trans. R. Soc. Lond. B Biol. Sci. 354: 927939.
  • Titus, R.C. 1998. Short-range and long-range songs: use of two acoustically distinct song classes by dark-eyed juncos. Auk 115: 386393.
  • Unitt, P. 2004. San Diego County Bird Atlas. San Diego Natural History Museum, San Diego, CA.
  • Valvo, M., Rizzo, M.C., Scarabello, M.P. & Parrinello, N. 1997. Genetic variability and taxonomical considerations about six species of European cardueline finches (Aves, Passeriformes). Comp. Biochem. Physiol. 118B: 771775.
  • Voelker, G. & Spellman, G.M. 2004. Nuclear and mitochondrial DNA evidence of polyphyly in the avian superfamily Muscicapoidea. Mol. Phylogenet. Evol. 30: 386394.
  • Yuri, T. & Mindell, D.P. 2002. Molecular phylogenetic analysis of Fringillidae, ‘‘New World nine-primaried oscines’’ (Aves: Passeriformes). Mol. Phylogenet. Evol. 23: 229243.
  • Zamora, J., Moscoso, J., Ruiz-del-Valle, V., Lowy, E., Serrano-Vela, J.I., Ira-Cachafeiro, J. & Arnaiz-Villena, A. 2006. Conjoint mitochondrial phylogenetic trees for canaries Serinus spp. and goldfinches Carduelis spp. show several specific polytomies. Ardeola 53: 117.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Figure  S1  Phylogeny of the species used in this study, based primarily in the DNA–DNA hybridization tree of Sibley and Ahlquist (1992) and with the remaining species added using the convention that distances between genera in the same family is 3.4 ΔT50H units, and between species in the same genus is 1.1 ΔT50H units.

Figure  S2  Composite phylogeny based primarily in Barker et al. (2004) and collating phylogenetic information from other sources.

Figure  S3  Residual plots for the multiple regression of the loudness index on all predictor variables.

Table  S1  Loudness index, song measurements, body mass, scores of vegetation density, and sample sizes for all species studied (see text for details).

Table  S2  GLS multiple regression of the loudness index on predictor variables using the composite phylogeny in Figure S2.

Table  S3  Multiple regression of the loudness index on predictor variables using species averages.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JEB_1883_sm_TabS1-S3_FigS1-S3.pdf689KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.