Abstract
Sexual signals are archetypes of contingent evolution: hyper-diverse across species, often evolving fast and in unpredictable directions. It is unclear to which extent their evolutionary unpredictability weakens deterministic evolution, or takes place bounded by deterministic patterns of trait evolution. We compared the evolution of sound frequency in sexual signals (advertisement songs) and non-sexual social signals (calls) across > 500 genera of the crown songbird families. Contrary to the acoustic adaptation hypothesis, we found no evidence that forest species used lower sound frequencies in songs or calls. Consistent with contingent evolution in song, we found lower phylogenetic signal for the sound frequency of songs than calls, which suggests faster and less predictable evolution, and found unpredictable direction of evolution in lineages with longer songs, which presumably experience stronger sexual selection on song. Nonetheless, the most important deterministic pattern of sound frequency evolution—its negative association with body size—was stronger in songs than calls. This can be explained by songs being longer-range signals than most calls, and thus using sound frequencies that animals of a given size produce best at high amplitude. Results indicate that sexual selection can increase aspects of evolutionary contingency while strengthening, rather than weakening, deterministic patterns of evolution.
Similar content being viewed by others
Introduction
Determinism in evolution refers to lineages under similar selective pressures converging into similar phenotypes, while evolutionary contingency can lead to those lineages diverging phenotypically, for example due to the random appearance and loss of genetic variation, and to stochasticity in evolutionary trajectories1. Evolutionary contingency refers broadly to unpredictability in evolutionary trajectories, even under identical starting points and selective pressures2. It is expected that traits undergoing evolutionary contingency change frequently, because multiple selective optima may exist without the selective landscape needing to change, and that the phenotypes of closely-related lineages diverge easily, resulting in low phylogenetic signal. A related meaning of the term evolutionary contingency is that of historical contingency, which refers to clade-specific constraints or differences in the response to selection3,4, and which causes high rather than low phylogenetic signal. For example, the evolution of species distinctiveness in woodpecker drums follows clade-specific rules and, accordingly, the acoustic structure of drums has high phylogenetic signal5. Here we do not use the term contingency in the sense of historical contingency, but under the more general definition of unpredictable and changing evolutionary trajectories.
We hypothesize that some types of selection, namely sexual selection, may simultaneously increase contingent aspects of evolution (e.g., fast rates of change in unpredictable directions), and also increase the strength of deterministic boundaries within which those contingent effects take place. In such a scenario, increased contingency might not weaken, and could even strengthen, deterministic patterns of trait evolution. Sexually-selected signals and ornaments are notable examples of contingency in phenotypic evolution because they are hyper-diverse across species, they are frequently lost and re-gained evolutionarily6, and they can evolve rapidly, such that related lineages often differ in sexual signals while being phenotypically very similar otherwise (e.g.7,8,9). For example, avian lineages under stronger sexual selection often have faster rates of phenotypic evolution on song10,11 or plumage colors12,13,14, but they do not necessarily evolve towards similar phenotypes10,11. In the case of bird song, behavioral research has even found that diametrically opposed sound frequencies can fulfil identical functions: in some species lower sound frequencies are preferred in courtship or are used to signal aggression, while in other species it is higher frequencies that fulfil those functions (reviewed in15).
Despite sexual signals being archetypes of contingent evolution, they are also subject to deterministic selective pressures12. Continuing with the case of sound frequency, well-known deterministic effects on its evolution are due to body size and habitat type. Acoustic theory predicts a negative linear relation between log-body mass and log-frequency of acoustic signals16,17 because, all else being equal, larger vocal organs and vocal tracts produce and resonate low frequencies more efficiently, and are better at coupling low frequencies to the surrounding medium18. To a lesser extent, forested habitats can also cause evolution of low sound frequencies19,20, mostly because foliage scatters and dampens the transmission of high frequencies21,22,23.
It is unclear to which extent the contingency associated with sexual selection weakens deterministic evolution (i.e., weakening cross-species associations between properties of sexual signals and their morphological or ecological predictors), as opposed to contingent effects taking place bounded by deterministic patterns of trait evolution12. To address this, here we compared deterministic patterns of sound frequency evolution between bird songs, which are an important type of acoustic sexual signal, and, as a control, contact calls, which are acoustic signals that include more general, non-sexually-selected social functions. We tested whether sound frequency in songs shows signs of contingent evolution, and whether or not its evolution became less deterministic than in calls.
The advertisement songs of songbirds are sexual signals typically used in mate attraction and/or competition with rivals24, and differences in their sound frequency are often implicated in female preferences or other sexual functions (reviewed in15). While songs are often sung solely or mostly by adult males25, other types of vocalizations, such as contact calls, are used by all individuals and have mostly non-sexual social functions26. We used the crown families of songbirds (i.e., parvorder Passerida27,28), which comprises ca. 77% and 75% of all songbird species and genera, respectively29, because it is generally straightforward to identify advertisement song and one main type of contact call in these species. We used descriptions in the literature to identify the main call type of each species or, when such descriptions were lacking, we used the most commonly recorded call (excluding alarm calls) as the main, most representative call type of the species. We restricted our study to the Passerida because in clades splitting from more basal nodes in the songbird phylogeny (e.g., Corvoidea or Meliphagoidea) it is common that song is less elaborate or less used, and that birds have various different call types30,31, making songs and contact calls not as easily distinguishable and compared.
We gathered and measured song and call recordings in more than 500 genera of Passerida (Fig. 1), to compare the evolution of sound frequency in songs and calls. Since sexually-selected signals typically evolve in a highly contingent manner, with phenotypic changes in unpredictable or opposed directions6, we expected that the sound frequency of songs should be more evolutionary labile and thus show weaker phylogenetic signal32,33 than that of calls. To ask if species experiencing stronger sexual selection on song tend to evolve sound frequency in one particular direction or in unpredictable directions, we also measured song duration as a proxy for song elaboration and, therefore, for the strength of sexual selection specifically on song. We used song duration because it is an aspect of song motor performance, and because song duration correlates with syllable diversity and repertoire size across species (e.g.,34,35). We did not use more general indexes for the strength of sexual selection, such as color ornamentation or sexual dichromatism, because species differences in ornamentation are generally not associated with song elaboration (reviewed in36,37) and, therefore, these indexes may bear little relation to the strength of sexual selection acting specifically on song. Finally, regarding deterministic evolution, and since body size and habitat type are the two best understood predictors of sound frequency in acoustic signals16,17,19, we evaluated the strength of deterministic evolutionary patterns by the strength of the cross-species association between sound frequency and either body size or habitat type.
Results
Body mass (log10g) and habitat type (proportion of time foraging in forests) had high phylogenetic signal (λ = 0.89 and 0.75, respectively), indicating a good fit between species similarity and phylogenetic relatedness, while song duration (log1016s) had only moderate phylogenetic signal (λ = 0.45), indicating labile evolution with more divergences and convergences across the phylogeny. The phylogenetic signal of song frequency was of intermediate value (λ = 0.59), and lower than that of call frequency (λ = 0.79), indicating more labile evolution of song than call frequencies along the phylogeny.
The sound frequency of songs and calls were negatively related to body size (Fig. 2), but not significantly related to habitat type (phylogenetic generalized least squares [PGLS] for song frequency: partial βst of body size = − 0.51, P < 0.001, partial βst of habitat = − 0.02, P = 0.51, N = 591 spp., model λ = 0.31; PGLS for call frequency: partial βst of body size = − 0.38, P < 0.001, partial βst of habitat = − 0.02, P = 0.58, N = 505 spp., model λ = 0.65). These results, from multiple PGLS regression, are similar to results using simple PGLS regression models of peak frequency on body size alone (songs: βst = − 0.51, P < 0.001, model λ = 0.31; calls: βst = − 0.38, P < 0.001, model λ = 0.65) or of peak frequency on the proportion of time in forested habitat (songs: βst = − 0.00, P = 0.96, model λ = 0.59; calls: βst = 0.03, P = 0.48, model λ = 0.78). Despite the more labile evolution of sound frequency in songs than calls, we found that the association with body size was significantly stronger for songs (βst = − 0.51, Fig. 2A) than calls (βst = − 0.38, Fig. 2B; t-test: Z = − 2.05, two-tailed P = 0.04).
Song duration, used here as a proxy for the strength of sexual selection on song, was not related to the sound frequency of songs (PGLS: partial βst of song duration = 0.03, P = 0.44, partial βst of body size = − 0.52 P < 0.001, and partial βst of habitat = − 0.03 P = 0.49, N = 591 spp., model λ = 0.31; Fig. 3). Inspection of Fig. 3 also shows that the variance of song frequency remains identical across species with short or long songs. To test this, we used residual frequencies from the PGLS regression of sound frequency on body size and habitat, to control their effects on sound frequency, and regressed the absolute values of those residual frequencies on song duration. We found that absolute values of residual frequency were also not related to song duration (PGLS: βst = 0.01, P = 0.37, N = 591 spp., model λ = 0.96), indicating that song duration does not predict the extent of deviations from the frequencies predicted by body size.
Discussion
A degree of contingency is expected in the evolutionary trajectories of sexually-selected traits, with frequent changes in different directions, because of the unstable nature of sexual selection and because of the multiple signals and ornamental phenotypes that it can target6. In agreement with this expectation, we found a lower phylogenetic signal for the sound frequency of songs than calls, indicating more labile evolution in songs, with more frequent divergence of closely-related species and convergence of distantly-related species. Since sound frequency of songs and calls is distributed over the same frequency range across species (Fig. 2), the lower phylogenetic signal of sound frequency in songs also implies that, on average, it evolved at a faster rate than in calls. Our result thus supports previous work concluding that sexual selection causes fast and diversifying phenotypic evolution (e.g.,7,8,9), including work on bird song10,11 and ornamentation12,13,14.
Importantly, despite the more labile evolution of sound frequency in songs than calls, we found a stronger allometry of sound frequency for songs than for calls. The negative-slope allometry of sound frequency across species that we found, whereby larger species use lower frequencies, was documented earlier in many taxa38,39,40, including for bird songs41,42,43,44 and calls45. But ours is the first work analyzing songs and calls in a sufficiently large number of species to compare the strength of their allometries. The stronger allometry of sound frequency in song can be explained by songs being, on average, longer-range signals than contact calls46, and needing to transmit across long distances to attract mates or repel rivals. Therefore, songs should be more strongly selected to use sound frequencies that, for a given body size, animals can best produce at high amplitude. Interestingly, body size had very high phylogenetic signal (λ = 0.89), the highest among all traits that we studied, which makes it more remarkable that the phylogenetic signal of sound frequency was lower in songs than calls. Together, the lower phylogenetic signal and stronger allometry of sound frequency in songs than calls indicate that the greater lability of sound frequency evolution in song took place within stronger deterministic boundaries set by body size, rather than weakening those allometric boundaries.
While body size is one of the best-understood16,17 and strongest correlates of sound frequency across avian species41,42,43,44,47, it has long been known that differences in the acoustic properties between forested and non-forested habitats can also influence the evolution of sound frequency in signals, as per the acoustic adaptation hypothesis21,22,23 (AAH). Nonetheless, our results showed a striking absence of an association between habitat type and the sound frequency of either songs or calls. This may not be unexpected at the broad taxonomic scale of our study, since reviews of the AAH showed that effect sizes of habitat type on avian sound frequencies are generally small and variable across taxonomic groups19,20, so that the effects of the AAH may not be as large or widespread as suggested by early studies21,42,48. Similarly to our results, other large-scale tests of the AAH for the sound frequency of songs, across Neotropical Parrots47 and Passeriformes44 also did not find a significant effect of vegetation density. These studies used different proxies for habitat type (vegetation indexes from satellite imagery) than ours (time foraging from forest canopy to understory) to arrive at identical results, and yet another comparative study, across the large suboscine radiation of ovenbirds (family Furnariidae), even reported an association of forested habitat with higher song minimum frequency, which goes in the opposite direction of the AAH49. Together, these results indicate that, while the AAH may explain differences in sound frequency among some closely-related species (e.g.,50) or across finer habitat differences than those that these large-scale studies can capture (e.g., between different forest strata51, or across Mediterranean-type habitats52), forest living has little predictive power at our broad taxonomic level.
Using song duration as a proxy for the strength of sexual selection on song, we found no evidence that species with longer songs tended to evolve sound frequency in a consistent direction (either high or low frequencies) or frequencies further away from the optimum predicted by body size (in both cases effect sizes were lower than 0.1%). Inasmuch as song duration indicates the strength of sexual selection on song, these results indicate that stronger sexual selection does not create a deterministic pattern of sound frequency evolving in a consistent direction, and it also does not weaken the deterministic constraint placed by body size. Here we used song duration as a proxy to evaluate the strength of sexual selection on song because duration is an aspect of song motor performance (longer songs can place motor and ventilation challenges53), and song motor performance can be sexually selected54. In addition, song duration correlates with syllable diversity and repertoire sizes across species (e.g.,34,35), and sexual selection has often been found to favor long song55, or diverse repertoires56. Also, songs are generally much longer than calls (Fig. S3C,D), indicating that sexual selection often causes the evolution of longer acoustic signals. Therefore, despite sexual selection being able to target several distinct song traits54, when comparing across a very large number of species, it is reasonable to expect that species subject to strong sexual selection on song have, on average, evolved longer songs than species subject to weaker sexual selection on song.
Differences in the strength of sexual selection across bird species are often inferred based on traits unrelated to song, such as ornamentation, sexual dichromatism or sexual size dimorphism (SSD). We did not use proxies based on ornamentation because species differences in ornamentation are generally not associated with song elaboration (reviewed in36,37), and in some cases may be negatively associated57. Therefore, it would be misleading to use color-based indexes for inferring strength of sexual selection on song. Male-biased SSD is also an often-used proxy to infer the strength of sexual selection, especially across non-avian tetrapod species, because the evolution of SSD may be driven by male–male competition (reviewed in58). But many birds interact in the air, where larger male size is not necessarily advantageous for fighting or for performing sexual displays, and where smaller sizes and greater agility may be favored instead59,60,61,62,63,64,65, perhaps explaining why high sound frequencies can signal aggression or be preferred by females in several bird species (reviewed in15). Therefore, it is not straightforward to use male-biased SSD as a proxy for the strength of sexual selection on song, as this proxy may focus on a subset of species not representative of the various ways in which sexual selection can affect bird song. Nonetheless, previous work analyzed the sound frequency of song in relation to SSD, finding only a weak negative association with SSD across the order Passeriformes (with an effect size estimated between 1 and 3%)44, and this effect was not significant for the crown songbirds (Passerida; B. Kempenaers, pers. comm.). This strengthens our conclusion that sexual selection does not set a deterministic pattern of sound frequency evolution in a consistent direction.
In conclusion, our results were consistent with the expectation that sexual selection brings about evolutionary contingency (e.g.,7,8,9,10,11), here in the form of labile evolutionary trajectories, with phenotypic divergences and convergences lowering the phylogenetic signal of song frequency relative to that of call frequency or body size, and also in the form of unpredictable directions of evolution in species with longer and presumably more strongly sexually-selected songs. More importantly, we showed that this enhanced contingency did not weaken the strongest deterministic pattern of song frequency evolution, its negative-slope allometry, compared to that for calls. On the contrary, the allometry of sound frequency was significantly stronger for songs than calls, likely because long-range songs need to use frequencies that animals of a given size can produce at high amplitude. Evolutionary contingency, defined broadly as unpredictability of trajectories, contrasts conceptually with the predictability of deterministic evolution, and the two are often discussed and studied as in direct opposition, for example as when phenotypic evolution is partitioned between deterministic convergence versus among-clade divergence attributable to historical contingency (e.g.,66,67). But contingent and deterministic effects may interact in ways other than weakening each other68, for example by acting at different levels or time-scales. Our results support this more multifaceted view of the interactions between evolutionary contingency and determinism and, for the first time, indicate that sexual selection can simultaneously increase aspects of evolutionary contingency and determinism: the former at a finer time-scale, causing labile and unpredictable evolutionary trajectories, and the latter holding this labile evolution within stronger allometric boundaries.
Methods
Acoustic data
In summary, using the citizen-science repository Xenocanto (www.xeno-canto.org), we could obtain good-quality recordings of songs for 591 species, one species per genus, representing > 76% of Passerida genera, distributed across 62 of the 76 Passerida families (only very small families or families lacking advertisement song were not represented; Table S1). For 505 of these species, we could also obtain recordings of contact calls. When contact calls were not described in the Handbook of the Birds of the World Alive69 (hereafter HBW), we used the most common type of call recorded, avoiding alarm-related calls. By sampling one species per genus, we optimize taxonomic breadth and evenness, contributing to a representative sampling of species diversity in the Passerida (Fig. 1). Within each genus, we prioritized using the species with the most recordings available, to increase the likelihood of finding good quality recordings of both song and calls. While this is not a random choice of species within genera, using well-recorded species, which are likely common species or species easy to record, does not bias choice of species in a manner related to the hypotheses tested here.
We measured peak frequency (frequency with highest cumulative sound amplitude) in up to three recordings per species, and the within-species repeatability of these measurements70 was very high and identical for songs (R = 0.87) and calls (R = 0.86), indicating a similar accuracy when quantifying species differences in either songs or calls. The within-species repeatability for song duration was also high (R = 0.70), although not as high as for frequency, perhaps because birds can often terminate their songs at different points. In total we measured 3056 different recordings, on average 2.90 recordings of songs per species, and 2.29 recordings of calls per species. In the “Supporting information” (Appendix S1) we give detailed methods for obtaining and measuring acoustic recordings. The “Supporting information” (Appendix S1) also shows that further increasing the number of recordings per species does not increase within-species repeatability appreciably, neither for songs or calls, indicating that our sampling was sufficient to evaluate species differences in the sound frequency of songs and calls.
Complete dataset and phylogenies
We took the body mass of each species from a data paper71. Also from this data paper, we quantified the proportion of time foraging in forested habitats as the sum of the proportion of time in all categories that unambiguously refer to forest: understory, mid-high strata, and canopy. Even though some habitat categories in the data paper71 do not unambiguously distinguish forest from non-forest, and the data refer to foraging habitat rather than specifically to the location of singing posts, these data nonetheless provide a better, and more nuanced evaluation of habitat differences between species than a dichotomous categorization of forest vs. non-forest (see also the discussion, for comparison with results using different proxies for habitat type).
Figure S3 in the “Supporting information” shows the distribution of the data across species, both acoustic, body mass and habitat. Peak frequencies (measured in a ratio scale72), log-transformed song duration, and body mass all had approximately bell-shaped distributions (Fig. S3A–E). The proportion of time foraging in forested habitat ranged from 0 to 1 among species, with species that forage more than 50% of the time in forests being only slightly more represented than those foraging less than 50% of the time in forests (“Supporting information”; Fig. S3F). The complete dataset is provided in the “Supporting information” (Tables S2, S3).
We used phylogenetic trees from Birdtree.org73, including all 591 species in the dataset. These phylogenies were constructed on the Hackett backbone74, using available molecular information in a Bayesian framework to obtain a large number of probable trees73. An example phylogenetic tree is shown in Fig. 1. We based all analyses on 1000 different phylogenetic trees to account for phylogenetic uncertainty (as recommended by75).
Analyses
We ran all phylogenetic comparative analyses in R v. 3.5.176, using the package caper (v 1.0.177). First, we estimated the phylogenetic signal of each species trait (song peak frequency and duration, call peak frequency, body mass, and the proportion of time in forested habitat) using Pagel’s λ32, weight-averaged across the 1000 phylogenetic trees. We weighted results based on the Akaike’s Information Criterion (AIC) of each tree’s model output (following78). AIC-weighted averaging takes into account phylogenetic uncertainty and the parsimony of models from each tree75,78.
Separately for songs and for calls, we tested whether peak frequency was associated with body size and habitat type across species, using a phylogenetic generalized least squares (PGLS) regression model of peak frequency on body mass and on the proportion of time in forested habitat. The parameter λ was estimated in all regression models, to adjust the phylogenetic correction to the degree of phylogenetic signal in the model79,80. As before, we used AIC weights for model averaging of results over the 1000 trees, including λ, β, and standard errors. The weighted average of β (partial regression coefficient) was standardized (βst) by multiplying by the standard deviation of the corresponding predictor trait, and dividing by the standard deviation of the dependent trait (peak frequency). We ran a student’s t-test comparing the βst of peak frequency on body mass between songs and calls (following81), and based on the AIC weight-averaged values of βst and associated standard errors from the above multiple PGLS regressions models.
Finally, to test if longer, more elaborate songs, were associated with an increase or a decrease in sound frequency across species, we added song duration as a predictor to the above multiple PGLS regression model of peak song frequency on body mass and proportion of time in forested habitat. There appear to be no issues of multicollinearity in these multiple PGLS regression models, because the three predictors were not associated (PGLS of song duration on mass: βst = 0.05, P = 0.31, model λ = 0.45; PGLS of song duration on habitat: βst = 0.06, P = 0.17 model λ = 0.45; PGLS of mass on habitat: βst = − 0.03, P = 0.39, model λ = 0.89; N = 591 spp. in all cases). To test if longer songs evolved associated with larger, non-directional deviations to the predicted optimum sound frequency for each species body size and habitat, we first computed residual frequency from the multiple PGLS regression model of song peak frequency on body mass and proportion of time in forested habitat; these residuals were weight-averaged across the 1000 trees, as explained before. We then ran a PGLS regression of the absolute values of these residual frequencies (large absolute values of residuals indicating a larger deviation to optimal sound frequency) on song duration. As above, PGLS model results were AIC weight-averaged across 1000 phylogenies.
References
Blount, Z. D., Lenski, R. E. & Losos, J. B. Contingency and determinism in evolution: Replaying life’s tape. Science (80-). 362, eaam5979 (2018).
Losos, J. B. Improbable Destinies (Riverhead Books, 2017).
Labra, A., Pienaar, J. & Hansen, T. F. Evolution of thermal physiology in liolaemus lizards: Adaptation, phylogenetic inertia, and niche tracking. Am. Nat. https://doi.org/10.1086/600088 (2009).
Moen, D. S., Morlon, H. & Wiens, J. J. Testing convergence versus history: Convergence dominates phenotypic evolution for over 150 million years in frogs. Syst. Biol. https://doi.org/10.1093/sysbio/syv073 (2016).
Garcia, M. et al. Evolution of communication signals and information during species radiation. Nat. Commun. https://doi.org/10.1038/s41467-020-18772-3 (2020).
Wiens, J. J. Widespread loss of sexually selected traits: How the peacock lost its spots. Trends Ecol. Evol. 16, 517–523 (2001).
Civetta, A. & Singh, R. S. Sex and speciation: Genetic architecture and evolutionary potential of sexual versus nonsexual traits in the sibling species of the drosophila melanogaster complex. Evolution (N. Y.). 52, 1080 (1998).
Arnegard, M. E. et al. Sexual signal evolution outpaces ecological divergence during electric fish species radiation. Am. Nat. 176, 335–356 (2010).
Pitchers, W., Wolf, J. B., Tregenza, T., Hunt, J. & Dworkin, I. Evolutionary rates for multivariate traits: The role of selection and genetic variation. Philos. Trans. R. Soc. B Biol. Sci. 369, 20130252 (2014).
Snyder, K. T. & Creanza, N. Polygyny is linked to accelerated birdsong evolution but not to larger song repertoires. Nat. Commun. 10, 1–15 (2019).
Price, J. J. Patterns of song evolution and sexual selection in the oropendolas and caciques. Behav. Ecol. 15, 485–497 (2004).
Cooney, C. R. et al. Sexual selection predicts the rate and direction of colour divergence in a large avian radiation. Nat. Commun. 10, 1773 (2019).
Seddon, N. et al. Sexual selection accelerates signal evolution during speciation in birds. Proc. R. Soc. B Biol. Sci. 280, 20131065 (2013).
Gomes, A. C. R., Sorenson, M. D. & Cardoso, G. C. Speciation is associated with changing ornamentation rather than stronger sexual selection. Evolution (N.Y.). 70, 2823–2838 (2016).
Cardoso, G. C. Paradoxical calls: The opposite signaling role of sound frequency across bird species. Behav. Ecol. 23, 237–241 (2012).
Fletcher, N. H. & Fahey, P. Acoustic systems in biology. Phys. Today 46, 79–79 (1993).
Fletcher, N. H. A simple frequency-scaling rule for animal communication. J. Acoust. Soc. Am. 115, 2334–2338 (2004).
Bradbury, J. W. & Vehrencamp, S. L. Principles of Animal Communication (Sinauer Associates, Inc., 2011).
Boncoraglio, G. & Saino, N. Habitat structure and the evolution of bird song: A meta-analysis of the evidence for the acoustic adaptation hypothesis. Funct. Ecol. 21, 134–142 (2007).
Ey, E. & Fischer, J. The ‘acoustic adaptation hypothesis’—A review of the evidence from birds, anurans and mammals. Bioacoustics 19, 21–48 (2009).
Morton, E. S. Ecological sources of selection on avian sounds. Am. Nat. 109, 17–34 (1975).
Graham, B. A., Sandoval, L., Dabelsteen, T. & Mennill, D. J. A test of the Acoustic Adaptation Hypothesis in three types of tropical forest: Degradation of male and female Rufous-and-white Wren songs. Bioacoustics 26, 37–61 (2017).
Blumenrath, S. & Dabelsteen, T. Degradation of great tit (Parus Major) song before and after foliation: Implications for vocal communication in a deciduous forest. Behaviour 141, 935–958 (2004).
Collins, S. Vocal fighting and flirting: The functions of birdsong. In Nature’s Music (ed. Marler, P. & Slabbekoorn, H.) 39–79 (Elsevier, 2004). https://doi.org/10.1016/B978-012473070-0/50005-0
Webb, W. H., Brunton, D. H., Aguirre, J. D., Thomas, D. B., Valcu, M. & Dale, J. Female song occurs in songbirds with more elaborated female coloration and reduced sexual dichromatism. Front. Ecol. Evol. 4, 22 (2016).
Marler, P. Bird calls: A cornucopia for communication. In Nature’s Music: The Science of Birdsong (eds Marler, P. & Slabbekoorn, H.) 132–177 (Elsevier Academic Press, 2004).
Barker, F. K., Cibois, A., Schikler, P., Feinstein, J. & Cracraft, J. Phylogeny and diversification of the largest avian radiation. Proc. Natl. Acad. Sci. 101, 11040–11045 (2004).
Jonsson, K. A. & Fjeldsa, J. A phylogenetic supertree of oscine passerine birds (Aves: Passeri). Zool. Scr. 35, 149–186 (2006).
Gill, F., Donsker, D. & Rasmussen, P. (eds) IOC World Bird List Version 9.2. (IOC 2019). https://doi.org/10.14344/IOC.ML.10.1
Crane, J. M. S., Savage, J. L. & Russell, A. F. Diversity and function of vocalisations in the cooperatively breeding Chestnut-crowned Babbler. Emu 116, 241–253 (2016).
Leighton, G. M. Cooperative breeding influences the number and type of vocalizations in avian lineages. Proc. R. Soc. B Biol. Sci. 284, 20171508 (2017).
Pagel, M. Inferring evolutionary processes from phylogenies. Zool. Scr. 26, 331–348 (1997).
Symonds, M. R. E. & Blomberg, S. P. A primer on phylogenetic generalised least squares. In Modern Phylogenetic Comparative Methods and Their Application in Evolutionary Biology (ed. Garamszegi, L. Z.) 105–130 (Springer Berlin Heidelberg, 2014). https://doi.org/10.1007/978-3-662-43550-2_5.
Cardoso, G. C., Hu, Y. & Mota, P. G. Birdsong, sexual selection, and the flawed taxonomy of canaries, goldfinches and allies. Anim. Behav. 84, 111–119 (2012).
Crouch, N. M. A. & Mason-Gamer, R. J. Identifying ecological drivers of interspecific variation in song complexity in songbirds (Passeriformes, Passeri). J. Avian Biol. 50, jav.02020 (2019).
Mason, N. A., Shultz, A. J. & Burns, K. J. Elaborate visual and acoustic signals evolve independently in a large, phenotypically diverse radiation of songbirds. Proc. R. Soc. B Biol. Sci. 281, 20140967 (2014).
Gomes, A. C. R., Funghi, C., Soma, M., Sorenson, M. D. & Cardoso, G. C. Multimodal signalling in estrildid finches: Song, dance and colour are associated with different ecological and life-history traits. J. Evol. Biol. 30, 1336–1346 (2017).
Martin, K., Tucker, M. A. & Rogers, T. L. Does size matter? Examining the drivers of mammalian vocalizations. Evolution (N. Y.). 71, 249–260 (2017).
Verzijden, M. N. et al. Sounds of male Lake Victoria cichlids vary within and between species and affect female mate preferences. Behav. Ecol. 21, 548–555 (2010).
Gingras, B., Boeckle, M., Herbst, C. T. & Fitch, W. T. Call acoustics reflect body size across four clades of anurans. J. Zool. 289, 143–150 (2013).
Wallschläger, D. Correlation of song frequency and body weight in passerine birds. Experientia 36, 412–412 (1980).
Ryan, M. J. & Brenowitz, E. A. The role of body size, phylogeny, and ambient noise in the evolution of bird song. Am. Nat. 126, 87–100 (1985).
Wiley, R. H. Associations of song properties with habitats for territorial oscine birds of Eastern North America. Am. Nat. 138, 973–993 (1991).
Mikula, P. et al. A global analysis of song frequency in passerines provides no support for the acoustic adaptation hypothesis but suggests a role for sexual selection. Ecol. Lett. https://doi.org/10.1111/ele.13662 (2020).
Jurisevic, M. A. & Sanderson, K. J. A comparative analysis of distress call structure in Australian passerine and non-passerine species: Influence of size and phylogeny. J. Avian Biol. 29, 61–71 (1998).
Dabelsteen, T. Public, private or anonymous? Facilitating and countering eavesdropping. In Animal Communication Networks (ed. McGregor, P. K.) 38–62 (Cambrigde University Press, 2005).
Medina-García, A., Araya-Salas, M. & Wright, T. F. Does vocal learning accelerate acoustic diversification? Evolution of contact calls in Neotropical parrots. J. Evol. Biol. 28, 1782–1792 (2015).
Chapuis, C. Un exemple de l’influence du milieu sur les émissions vocales des oiseaux : l’évolution des chants en forêt équatoriale. La Terre la vie 85, 183–202 (1971).
Derryberry, E. P. et al. Ecological drivers of song evolution in birds: Disentangling the effects of habitat and morphology. Ecol. Evol. 8, 1890–1905 (2018).
Tubaro, P. L. & Lijtmaer, D. A. Environmental correlates of song structure in forest grosbeaks and saltators. Condor 108, 120–129 (2006).
Seddon, N. Ecological adaptation and species recognition drives vocal evolution in neotropical suboscine birds. Evolution (N. Y.). 59, 200–215 (2005).
Cardoso, G. C. & Price, T. D. Community convergence in bird song. Evol. Ecol. https://doi.org/10.1007/s10682-009-9317-1 (2010).
Zollinger, S. A. From brain to song: The vocal organ and vocal tract. In Neuroscience of Birdsong (eds Zeigler, H. P. & Marler, P.) 78–98 (Cambridge University Press, 2008).
Gil, D. & Gahr, M. The honesty of bird song: Multiple constraints for multiple traits. Trends Ecol. Evol. 17, 133–141 (2002).
Neubauer, R. L. Super-normal length song preferences of female zebra finches (Taeniopygia guttata) and a theory of the evolution of bird song. Evol. Ecol. 13, 365–380 (1999).
Soma, M. & Garamszegi, L. Z. Rethinking birdsong evolution: Meta-analysis of the relationship between song complexity and reproductive success. Behav. Ecol. 22, 363–371 (2011).
Shutler, D. Sexual selection: When to expect trade-offs. Biol. Lett. 7, 101–104 (2011).
Littleford-Colquhoun, B. L. et al. How sexual and natural selection shape sexual size dimorphism: Evidence from multiple evolutionary scales. Funct. Ecol. 33, 1446–1458 (2019).
Andersson, M. & Norberg, R. Å. Evolution of reversed sexual size dimorphism and role partitioning among predatory birds, with a size scaling of flight performance. Biol. J. Linn. Soc. 15, 105–130 (1981).
Hedenström, A. & Møller, A. P. Morphological adaptations to song flight in passerine birds: A comparative study. Proc. R. Soc. London. Ser. B Biol. Sci. 247, 183–187 (1992).
Grønstøl, G. B. Ø. Aerobatic components in the song-flight display of male lapwings Vanellus vanellus as cues in female choice. Ardea 84, 45–55 (1996).
Hakkarainen, H. et al. A test of male mating and hunting success in the kestrel: The advantages of smallness?. Behav. Ecol. Sociobiol. 39, 375–380 (1996).
Blomqvist, D., Johansson, O. C., Unger, U., Larsson, M. & Flodin, L. -Å. Male aerial display and reversed sexual size dimorphism in the dunlin. Anim. Behav. 54, 1291–1299 (1997).
Figuerola, J. A comparative study on the evolution of reversed size dimorphism in monogamous waders. Biol. J. Linn. Soc. 67, 1–18 (1999).
Raihani, G., Székely, T., Serrano-Meneses, M. A., Pitra, C. & Goriup, P. The influence of sexual selection and male agility on sexual size dimorphism in bustards (Otididae). Anim. Behav. 71, 833–838 (2006).
Labra, A., Pienaar, J. & Hansen, T. F. Evolution of thermal physiology in liolaemus lizards: Adaptation, phylogenetic inertia, and niche tracking. Am. Nat. 174, 204–220 (2009).
Moen, D. S., Morlon, H. & Wiens, J. J. Testing convergence versus history: Convergence dominates phenotypic evolution for over 150 million years in frogs. Syst. Biol. 65, 146–160 (2016).
Taylor, E. B. & McPhail, J. D. Historical contingency and ecological determinism interact to prime speciation in sticklebacks, Gasterosteus. Proc. R. Soc. London. Ser. B Biol. Sci. 267, 2375–2384 (2000).
del Hoyo, J., Elliott, A., Christie, D. & de Juana, E. (eds) HBW Alive: Handbook of the Birds of the World Alive. (Lynx edition, Barcelona 2018). https://www.hbw.com/. (Accessed 23 September 2018)
Nakagawa, S. & Schielzeth, H. Repeatability for Gaussian and non-Gaussian data: A practical guide for biologists. Biol. Rev. 85, 935–956 (2010).
Wilman, H. et al. EltonTraits 1.0: Species-level foraging attributes of the world’s birds and mammals. Ecology 95, 2027–2027 (2014).
Cardoso, G. C. Using frequency ratios to study vocal communication. Anim. Behav. 85, 1529–1532 (2013).
Jetz, W., Thomas, G. H., Joy, J. B., Hartmann, K. & Mooers, A. O. The global diversity of birds in space and time. Nature 491, 444–448 (2012).
Hackett, S. J. et al. A phylogenomic study of birds reveals their evolutionary history. Science (80-). 320, 1763–1768 (2008).
Rubolini, D., Liker, A., Garamszegi, L. Z., Møller, A. P. & Saino, N. Using the BirdTree.org website to obtain robust phylogenies for avian comparative studies: A primer. Curr. Zool. 61, 959–965 (2015).
R Team. R: A Language and Environment for statistical Computing (R Foundation for Statistical Computing, 2014).
Orme, D. et al. Caper: Comparative Analysis of Phylogenetics and Evolution in R. R Package Version 1.0.1 (https://CRAN.R-project.org/package=caper, 2018).
Garamszegi, L. Z. & Mundry, R. Multimodel-inference in comparative analyses. In Modern Phylogenetic Comparative Methods and Their Application in Evolutionary Biology (ed. Garamszegi, L.) 305–331 (Springer Berlin Heidelberg, 2014). https://doi.org/10.1007/978-3-662-43550-2_12.
Freckleton, R. P., Harvey, P. H. & Pagel, M. Phylogenetic analysis and comparative data: A test and review of evidence. Am. Nat. 160, 712–726 (2002).
Revell, L. J. Phylogenetic signal and linear regression on species data. Methods Ecol. Evol. 1, 319–329 (2010).
Paternoster, R., Brame, R., Mazerolle, P. & Piquero, A. Using the correct statistical test for the equality of regression coefficients. Criminology 36, 859–866 (1998).
Acknowledgements
G.C.C. was supported by Portuguese National Funds through FCT (Fundação para a Ciência e a Tecnologia; ref. DL57/2016/CP1440/CT0011).
Author information
Authors and Affiliations
Contributions
G.C.C. proposed the study, J.I.F. and T.D. refined the study design, J.I.F. gathered, quantified and analyzed the data, and wrote the manuscript with contributions from all authors.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Friis, J.I., Dabelsteen, T. & Cardoso, G.C. Contingency and determinism in the evolution of bird song sound frequency. Sci Rep 11, 11600 (2021). https://doi.org/10.1038/s41598-021-90775-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-021-90775-6
This article is cited by
-
Relative divergence of mobbing calls and songs structures in passerine birds
Journal of Ornithology (2024)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.