Content uploaded by Gustavo Brant Paterno
Author content
All content in this area was uploaded by Gustavo Brant Paterno on Jan 30, 2017
Content may be subject to copyright.
ORIGINAL ARTICLE
Background noise as a selective pressure:
stream-breeding anurans call at higher frequencies
David Lucas Röhr
1
&Gustavo Brant Paterno
1
&Felipe Camurugi
2
&Flora Acuña Juncá
3
&
Adrian Antonio Garda
1,4
Received: 18 June 2015 / Accepted: 30 November 2015
#Gesellschaft für Biologische Systematik 2015
Abstract Acoustic signals are an important part in the behav-
iour of many species and may play a key role in speciation.
However, little is known about the importance of natural se-
lection on the evolution of such signals. Acoustics signals are
the main communication channel for most anuran species, and
background noise from streams is a constant source of
masking interference for species reproducing in these environ-
ments. Herein, we test if the noise of flowing water habitats
has favoured advertisement calls with higher dominant fre-
quencies in frogs. Phylogenetic generalized least square mod-
el analysis revealed a significant influence of reproductive
environment and body size on dominant frequency, with no
significant interaction between habitat and body size. While
stream breeders call at higher dominant frequencies, this
acoustic parameter is inversely correlated with body size in
both environments. We discuss the biologicalconsequences of
long-term adaptive shift in this acoustic parameter and possi-
ble trade-offs with other evolutionary forces.
Keywords Acoustic communication .Advertisement call .
Comparative methods .Evolution .Masking interference
Introduction
Acoustic signals are a fundamental part in the communication
system of many species, having evolved independently several
times in different clades (Gerhardt and Huber 2002). Stochastic
processes, pleiotropic effects, and sexual and natural selection
may drive the evolution of sound communication in animals
(Wilkins et al. 2013). The role of stochastic evolution has been
quantified contrasting molecular and acoustic variation with
the aid of recent phylogenetic hypotheses (Goicoechea et al.
2010). Morphological and physiological constrains affect
signal characteristics, and selective pressures on these may lead
to pleiotropic signal divergence (Podos 2001). Sexual selection
is the best-studied evolutionary force shaping acoustic signal
evolution. Females may show strong preference for specific
acoustic parameters, resulting in differential mating success
(Ritchie 1996) and ultimately leading to species divergence.
In contrast, much less is known about the role of natural selec-
tion on acoustic signal evolution (Wilkins et al. 2013).
Background noise is one of the main constraints on acous-
tic communication, limiting the active space of every natural
communication system (Brumm 2013). Short duration noise
is circumvented by plastic responses, whereas more predict-
able and constant noise should result in long-term adaptive
processes (Brumm 2013). While short-term plastic responses
to background noise have been fairly well studied, less is
known about long-term adaptive processes in constantly noisy
environments (Brumm and Slabbekoorn 2005).
Electronic supplementary material The online version of this article
(doi:10.1007/s13127-015-0256-0) contains supplementary material,
which is available to authorized users.
*David Lucas Röhr
davidlucasr@yahoo.com.br
1
Programa de Pós-graduação em Ecologia, Universidade Federal do
Rio Grande do Norte, Lagoa Nova, 59072-970 Natal, RN, Brazil
2
Programa de Pós-Graduação em Ciências Biológicas (Zoologia),
Departamento de Sistemática e Ecologia, Centro de Ciências Exatas e
da Natureza, Universidade Federal da Paraíba, João
Pessoa 58059-900, PB, Brazil
3
Departamento de Ciências Biológicas, Universidade Estadual de
Feira de Santana, BR 116, Km 03, Campus Universitário,
44031-460 Feira de Santana, BA, Brazil
4
Departamento de Botânica e Zoologia, Centro de Biociências,
Universidade Federal do Rio Grande do Norte, Campus
Universitário, Lagoa Nova, 59078-900 Natal, RN, Brazil
Org Divers Evol
DOI 10.1007/s13127-015-0256-0
Stream noise is characterized by constancy, often high inten-
sities, and emphasized energy in low frequency bands (Goutte
et al. 2013) that potentially overlap with low frequency anuran
vocalizations (Wells 2010). Hence, higher frequencies should
be favoured in stream breeders by reducing the energy expen-
diture needed to diminish interference by increasing intensity.
Indeed, flowing water noise pressure level is one of the best
predictors of anuran community composition in the vicinity of
streams (Goutte et al. 2013) and communicating in high
frequencies near streams improves signal detection and
discrimination (Boonman and Kurniati 2011). Still, the role of
flowing water noise as a selectivepressureonanuranadvertise-
ment calls is contentious (Boeckle et al. 2009;Hoskinetal.
2009). One analysis using 110 species in five families found
that stream species use slightly higher dominant frequencies,
but this trend vanished in analyses controlling for body size and
phylogeny (Vargas-Salinas and Amézquita 2014).
Herein, we test if masking interference from low frequency
noise of streams has favoured advertisement calls with higher
dominant frequencies in frogs reproducing near these environ-
ments. To do so, we gathered information on calls of 509
species from 31 frog families in all biogeographic realms.
We test this hypothesis using phylogenetic comparative
methods controlling for adult male body sizes.
Materials and methods
We constructed a database composed of mean advertisement
call dominant frequencies (the frequency band with the
greatest amount of energy) and maximum male body sizes
(snout-vent length (SVL)) reported for each species from data
available inthe literature. We searched for species reproducing
exclusively in flowing or still waters (leaving out species that
use both habitats and species which reproduce independently
from water bodies) and included in the phylogeny proposed
by Pyron and Wiens (2011). In order to achieve a large data
set, we used practical and pre-established criteria for data in-
clusion. For multiple literature hits on the same species, we
included the most recently reported mean dominant frequency
and the overall largest male body size. However, to ensure that
these criteria do not include a bias in the analyses, we tested a
random subset of the data demonstrating that there is no sig-
nificant difference in dominant frequency between older and
recent publications and that maximum and mean SVL are
highly correlated (more than 98 %) (Supporting Information
Sects. 2.4 and 2.5). For a few species where authors reported a
dominant frequency range, we used the average between max-
imum and minimum values. In rare cases where the publica-
tion did not report values for dominant frequency but included
a spectrogram with a straight and clearly identifiable empha-
sized spectrum, we included a visual estimation of this
parameter.
We evaluated the phylogenetic signal of our data set using
Blomberg’s K, which varies from zero to infinity and indicates
the strength of phylogenetic signal under Brownian motion
model of evolution (Blomberg et al. 2003). Next, we used a
phylogenetic generalized least square model (PGLS), which
takes into account the nonindependence of observations due
to phylogeny and assumes a Brownian motion model of evo-
lution (Freckleton et al. 2002). We used the dominant frequen-
cy as the response variable and reproduction habitat (still/
flowing) and SVL as the explanatory variables to test if dom-
inant frequency was affected by reproduction environment.
Dominant frequencies and body sizes were log transformed
(natural logarithm) before the analysis. To optimize branch
length transformation, the lambda value was set by maximum
likelihood (Orme et al. 2012). All statistical analyses were
performed in R 3.1.2 using the packages Caper (Orme et al.
2012) and Picante (Kembel et al. 2010).
Results
We complied a dataset of 509 species representing 31 of the 54
currently recognized anuran families (see Supporting
Information Sect. 2 for phylogenetic tree and dataset; see
Appendix 1for complete table with references). Stream-
reproducing species (N= 177) have a mean dominant frequen-
cy of 3.37 ± 2.04 kHz (range 0.42–15.97) and a mean SVL of
41.7 ± 20 mm (range 20–138), while still water reproducing
species (N= 332) average dominant frequency and SVL were
2.18 ± 1.26 kHz (range 0.18–9.17) and 51.2 ± 29.3 mm (range
15–245), respectively.
Dominant frequency (K= 0.37, Z variance = −4.16,
p<0.001) and SVL (K= 0.44, Z variance = −4.45, p<
0.001; Table 1) presented a significant phylogenetic signal.
PGLS analysis revealed a strong influence of reproductive
environment and body size on dominant frequency (R
2
=
0.38), with no significant interaction between habitat and
body size (Table 2;Fig.1; see Sect. 4.5 in Supporting
Information for model diagnostic). While stream breeders call
with higher frequencies than still water species, dominant fre-
quency decreases with increasing body size in both environ-
ments (β=−0.874, standard error = 0.052). Model residuals
Tabl e 1 Phylogenetic signal for dominant frequency (DF) and body
size (SVL) calculated through Blomberg’sK
Source K PIC.mean PIC.rdn.mean pvalue
logDF 0.3660 0.00754 0.02918 0.001
logSVL 0.4387 0.00321 0.01501 0.001
Residuals 0.1002 0.00013 0.00016 0.173
From Blomberg et al. 2003
D.L. Röhr et al.
showed a nonsignificant phylogenetic signal (K= 0.055,
Z variance = 1.301, p=0.888).
Discussion
Frogs reproducing in streams use higher dominant frequen-
cies, suggesting that advertisement calls have evolved to di-
minish masking interference from flowing waters background
noise by reducing spectral overlap. This hypothesis was pre-
viously corroborated in studies on a single community
(Preininger et al. 2007) and one specific genus (Boeckle
et al. 2009) but contradicted in another analysis (Hoskin
et al. 2009). Nevertheless, all these studies used a limited
taxonomic sampling and did not account for phylogeny.
Conversely, an analysis including a wider taxonomic sam-
pling found that, although stream breeders call with higher
average dominant frequencies, this difference is not signifi-
cant when controlling for phylogeny and body size (Vargas-
Salinas and Amézquita 2014). The authors found that stream
breeders were significantly smaller and attributed this differ-
ence to habitat filtering (size selected for higher dominant
frequency), natural selection favouring small size in those
habitats (reduced pressure from evaporative loss), or both
(Vargas-Salinas and Amézquita 2014). The dataset used by
the authors was one fifth of the one used in our present anal-
ysis (110 species), included only five anuran families (and
hence a significantly smaller number of phylogenetic con-
trasts), and was restricted to New World frogs, mostly from
the Amazon region. Indeed, while few species reproduce in
flowing waters in the Amazon region, about half of the species
in southeast Asia, for example, are riparian and develop in
streams (Zimmerman and Simberloff 1996), a biogeographic
bias that is likely to influence the results. The current study
encompasses a much larger interspecific variation in body size
(230 mm vs 123 mm in their study) and a larger overlap in
body size between the two categories of reproductive environ-
ment (118 mm vs 47 mm). Therefore, the families and species
chosen in Vargas-Salinas and Amézquita (2014)surveydonot
adequately represent size differences among reproductive
habitats for all anurans, making it difficult to disentangle the
effects of body size and environment.
Although environmental influence on dominant frequency
is highly significant in our analysis, its effect is small when
compared to body size (Table 2). This is expected considering
the inverse relationship between vocal apparatus mass and call
frequency, which makes the variation in this acoustic param-
eter limited by morphological constraints. Indeed, our com-
plete PGLS model accounts for about 40 % of the variation in
this parameter, and other selective forces might be important
(see below). Moreover, the importance of environmental
noise as a selective pressure can vary among different anuran
clades, and future studies should focus on more restrictive
groups (such as a single family) with better representation
of its species and accurate measures of sound pressure levels.
Even though the environment is not the main driver of dom-
inant frequency variation, there is a mean difference of nearly
1200 Hz between environments. Thus, considering the impor-
tance of this parameter for anuran reproduction (Gerhardt
and Huber 2002), its biological relevance should not be
overlooked.
Tabl e 2 ANOVA table for the phylogenetic generalized least square
model {log(DF) −habitat × log(SLV)} evaluating the effects of body size
(SVL) and habitat on advertisement call dominant frequencies
Source df SQ MSQ Fpvalue
Habitat 1 0.0764 0.0764 33.0 <0.001
SVL 1 0.6461 0.6461 279.4 <0.001
Habitat × SVL 1 0.0008 0.0008 0.3 0.5674
Residuals 505 1.1675 0.0023 ––
2.5 2.7 2.9 3.1 3.3 3.5 3.7 3.9 4.1 4.3 4.5 4.7 4.9 5.1 5.3 5.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
Lo
g
snout−vent len
g
th (lnSVL)
Log dominant frequency (lnDF)
Flowing water (C = 11.05)
Still water (C = 10.86)
Y = −0.87x + C
Fig. 1 Relationship between
dominant frequency and body
size (SVL) for species calling in
still and flowing waters (n=509).
Lines represent PGLS regressions
Stream background noise and anuran communication
In large species, even an evolutionary increase in dominant
frequency, within physical constraints on sound production
mechanisms, may still not overcome the emphasized spectrum
from background stream noise to obtain release from masking
interference. Furthermore, because advertisement calls are
crucial for anuran reproductive behaviour, other evolutionary
forces besides pleiotropic effects of body size and noise inter-
ference might be involved in the establishment of dominant
frequency differences. For instance, some species/clades may
evolve different strategies to cope with such interference, such
as visual communication (Starnberger et al. 2014;Hödland
Amézquita 2001).
Several complex trade-offs among selective pressures may
be involved in shaping anuran vocalization. Call attractiveness
to females and detectability may be selected by opposing
forces in streams. Females may show increased phonotaxis
for low or median values of dominant frequency, leading to
directional or stabilizing selection (Gerhardt (1991), but see
Gerhardt and Schwartz (2001) for further discussion on fe-
male preference for dominant frequency). During aggressive
acoustic encounters, dominant frequency might be determi-
nant for the outcome (Davies and Halliday 1978), and territo-
rial males may lower call frequency in the presence of in-
truders (Wagner 1989; Bee and Bowling 2002). Hence, males
near streams may face a trade-off between the need to increase
call frequency to enhance signal detection at the expense of
reducing attractiveness and overall recognition.
Additionally, a trade-off between sound propagation and
detectability in forested stream environments is also expected.
Low-frequency calls are more efficient in habitats with many
physical barriers compared to higher frequencies (Ey and
Fischer 2009). Thus, species reproducing in forest streams
should face opposite selective pressures, where low dominant
frequencies suffer less attenuation and degradation, but high
dominant frequencies experience less masking interference.
Furthermore, community composition may promote addition-
al limits and selective pressures by driving the evolution of
anuran advertisement call dominant frequencies in two dis-
tinct manners. First, the presence of sympatric phylogenetical-
ly related taxa with similar vocalizations may lead to sexual
character displacement to decrease hybridization probability
(Lemmon 2009). Second, in highly diverse acoustic habitats,
calls may evolve to fill different acoustic niches and spectral
silent windows should be favoured (Chek et al. 2003). In
either case, dominant frequency changes favoured by these
scenarios could reinforce or counterbalance the selective
forces of flowing water masking interference.
Although background noise is probably the main differ-
ence in the acoustic scenario between still and flowing water
habitats, these environments also vary in a myriad of other
factors that might affect its acoustic community and should
be considered in future studies, such as (1) community of
sound-guided predators (Ryan and Tuttle 1983), (2)
vegetation coverage which might act as propagation barriers
(Ey and Fischer 2009), (3) sympatric species with prominent
acoustic signals (Chek et al. 2003), and (4) tadpole develop-
ment environment leading to differences in adult body size
and steroid hormones (Wells 2010).
Considering all the different evolutionary processes acting
upon anuran advertisement calls, the importance of stream back-
ground noise as a selective pressure is expected to vary between
clades, especially considering that stream colonization took
place many times independently and the evolutionary time un-
der this condition varies. For example, when only the three most
representative families with species from both habitats were
tested separately, the habitat effect was not significant for one
family, while the effect of size was very different between the
two others (Supporting Information Sect. 4.7.2). Therefore, our
study reveals a general pattern for anurans (broad phylogenetic
scale), while selection by background noise might vary between
clades with contrasting evolutionary histories.
Finally, even with this complex evolutionary scenario, we
found a significant trend for anuran species calling near
streams to use higher dominant frequencies. Other advertise-
ment call characteristics may respond similarly. For example,
sound intensity and call rate are expected to be higher in frogs
reproducing in stream habitats and using calls with dominant
frequencies similar to surrounding noise. Patterns for other
variables, such as call duration and complexity, are less clear.
Testing predictions for these variables, however, is much
harder because of the lack of appropriate descriptions in the
literature. Background noise from streams is clearly determi-
nant for the evolution of anuran advertisement calls, and fu-
ture work should explore the generality of these results for
other groups of animals.
Acknowledgments We thank Carlos Roberto Fonseca, Alex Pyron,
Pablo Martinez, Marcelo Gehara, and Frank Burbrink for suggestions
on the manuscript and fruitful discussions. AAG and FAJ thank National
Counsel of Technological and Scientific Development - CNPq for finan-
cial support (Universal # 473503/2012-3 and #305704/2013-3, respec-
tively).
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
Ethical approval This article does not contain any studies with human
participants or animals performed by any of the authors.
References
Bee, M. A., & Bowling, A. C. (2002). Socially mediated pitch alteration
by territorial male Bullfrogs, Rana catesbeiana. Journal of
Herpetology, 36(1), 140–143.
Blomberg,S.P.,Garland,T.,&Ives,A.R.(2003).Testingfor
phylogenetic signal in comparative data: behavioral traits are
D.L. Röhr et al.
more labile. Evolution, 57(4), 717–745. doi:10.1111/j.0014-
3820.2003.tb00285.x.
Boeckle, M., Preininger, D., & Hödl, W. (2009). Communication in noisy
environments I: acoustic signals of Staurois latopalmatus Boulenger
1887. Herpetologica, 65(2), 154–165. doi:10.1655/07-071r1.1.
Boonman, A., & Kurniati, H. (2011). Evolution of high-frequency com-
munication in frogs. Evolutionary Ecology Research, 13(2), 197–
207.
Brumm, H. (2013). Animal Communication and Noise (Vol. 2).
Heidelberg: Springer.
Brumm, H., & Slabbekoorn, H. (2005). Acoustic communication in
noise. Advances in the Study of Behavior, 35,151–209.
Chek, A. A., Bogart, J. P., & Lougheed, S. C. (2003). Mating signal
partitioning in multi-species assemblages: a null model test using
frogs. Ecology Letters, 6(3), 235–247. doi:10.1046/j.1461-0248.
2003.00420.x.
Davies, N. B., & Halliday, T. R. (1978). Deep croaks and fighting assess-
ment in toads Bufo bufo. Nature, 274(5672), 683–685. doi:10.1038/
274683a0.
Ey, E., & Fischer, J. (2009). The Bacoustic adaptation hypothesis^-a
review of the evidence from birds, anurans and mammals.
Bioacoustics, 19(1-2), 21–48.
Freckleton, R. P., Harvey, P. H., & Pagel, M. (2002). Phylogenetic anal-
ysis and comparative data: a test and review of evidence. American
Naturalist, 160(6), 712–726. doi:10.1086/343873.
Gerhardt, H. C. (1991). Female mate choise in treefrogs: static and dy-
namic acoustic criteria. Animal Behaviour, 42,615–635.
Gerhardt, H. C., & Huber, F. (2002). Acoustic communication in insects
and anurans: common problems and diverse solutions.Chicago:
University of Chicago Press.
Gerhardt, H. C., & Schwartz, J. J. (2001). Auditory tuning and frequency
preferences in anurans. In M. J. Ryan (Ed.), Anuran communication
(pp. 73–85). Washington: Smithsonian Institution Press.
Goicoechea, N., De La Riva, I., & Padial, J. M. (2010). Recovering
phylogenetic signal from frog mating calls. Zoologica Scripta,
39(2), 141–154.
Goutte, S.,Dubois, A., & Legendre, F. (2013). The importance of ambient
sound level to characterise anuran habitat. Plos One, 8(10), e78020.
doi:10.1371/journal.pone.0078020.
Hödl, W., & Amézquita, A. (2001). Visual signaling in anuran amphib-
ians. In M. J. Ryan (Ed.), Anuran Communication (pp. 121–141).
Washington: Smithsonian Institution Press.
Hoskin, C. J., James, S., & Grigg, G. C. (2009). Ecology and taxonomy-
driven deviations in the frog call-body size relationship across the
diverse Australian frog fauna. Journal of Zoology, 278(1), 36–41.
doi:10.1111/j.1469-7998.2009.00550.x.
Kembel, S. W., Cowan, P. D., Helmus, M. R., Cornwell, W. K., Morlon,
H., Ackerly, D. D., et al. (2010). Picante: R tools for integrating
phylogenies and ecology. Bioinformatics, 26(11), 1463–1464. doi:
10.1093/bioinformatics/btq166.
Lemmon, E. M. (2009). Diversification of conspecific signals in sympat-
ry: geographic overlap drives multidimensional reproductive char-
acter displacement in frogs. Evolution, 63(5), 1155–1170.
Orme, C. D. L., Freckleton, R. P., Thomas, G. H., Petzoldt, T., Fritz, S. A.,
Isaac, N., et al. (2012). Caper: comparative analyses of phyloge-
netics and evolution in R. R package version 0.5.
Podos, J. (2001). Correlated evolution of morphologyand vocal signal
structure in Darwin’s Fnches. Nature, 409,185–188.
Preininger, D., Boeckle, M., & Hödl, W. (2007). Comparison of anuran
acoustic communities of two habitat types in the Danum Valley
Conservation Area, Sabah, Malaysia. Salamandra, 43(3), 129–138.
Pyron, R. A., & Wiens, J. J. (2011). A large-scale phylogeny of Amphibia
including over 2800 species, and a revised classification of extant
frogs, salamanders, and caecilians. Molecular Phylogenetics and
Evolution, 61(2), 543–583. doi:10.1016/j.ympev.2011.06.012.
Ritchie, M. G. (1996). The shape of female mating preferences.
Proceedings of the National Academy of Sciences, 93, 14628–
14631.
Ryan, M. J., & Tuttle, M. D. (1983). The ability of the frog-eating bat to
discriminate among novel and potentially poisonous frog species
using acoustic cues. Animal Behaviour, 31(3), 827–833.
Starnberger, I., Preininger, D., & Hödl, W. (2014). From uni- to
multimodality: towards an integrative view on anuran communica-
tion. Journal of Comparative Physiology A, 200,777–787.
Vargas-Salinas, F., & Amézquita, A. (2014). Abiotic noise, call frequency
and stream-breeding anuran assemblages. Evolutionary Ecology,
28(2), 341–359. doi:10.1007/s10682-013-9675-6.
Wagner, W. E., Jr. (1989). Fighting, assessment, and frequency alteration
in Blanchard’s cricket frog. Behavioral Ecology & Sociobiology,
25(6), 429–436.
Wells, K. D. (2010). The ecology and behavior of amphibians. Chicago:
University of Chicago Press.
Wilkins, M. R., Seddon, N., & Safran, R. J. (2013). Evolutionary diver-
gence in acoustic signals: causes and consequences. Trends in
Ecology & Evolution, 28(3), 156–166. doi:10.1016/j.tree.2012.10.
002.
Zimmerman, B. L., & Simberloff, D. (1996). An historical interpretation
of habitat use by frogs in a Central Amazonian Forest. Journal of
Biogeography, 23,27–46.
Stream background noise and anuran communication
