vol . 200 , no . 3 the amer ican natural i st september 2022
Neotropical Birds Respond Innately

to Unfamiliar Acoustic Signals
Luis Sandoval1,* and David R. Wilson2

1. Escuela de Biología, Universidad de Costa Rica, San Pedro, San José, Costa Rica 11501-2060; 2. Department of Psychology, Memorial
University of Newfoundland and Labrador, 232 Elizabeth Avenue, St. John’s, Newfoundland and Labrador A1B 3X9, Canada

Submitted August 31, 2021; Accepted February 22, 2022; Electronically published July 11, 2022

Online enhancements: supplemental PDF.
abstract: Many animals respond to heterospecific signals that
indicate the presence of food or predators. Although the benefits of
responding are clear, the behavioral and cognitive mechanisms under-
lying responses are not. Whether responses are learned, innate, or
an epiphenomenon created by following other species as they re-
spond to signals remains unknown because most studies have in-
volved respondents that are sympatric with their heterospecific signalers
and that have therefore had opportunities to learn their signals. In
this study, we tested the mechanisms underlying avian responses to
heterospecific chick-a-dee calls. All North American parids produce
chick-a-dee calls in response to arousing stimuli, such as food and
predators, and diverse species respond by approaching the caller and
consuming the food or mobbing the predator. We broadcast chick-
a-dee calls plus two control stimuli in Costa Rica, Colombia, and
Brazil, where no parids ever occur. We conducted our trials in the
winter, when Neotropical migrants that might be familiar with chick-
a-dee calls were present, and in the temperate breeding season, when
migrantswere absent.Across 138 trials, 38 resident species from14fam-
ilies and four orders responded to chick-a-dee calls by approaching
to within 5 m of the playback speaker. A phylogenetic logistic regres-
sion showed that whether a species responded was not significantly
associated with the species’ mean body mass or the structural sim-
ilarity between its calls and chick-a-dee calls. Residents were signif-
icantly more likely to approach chick-a-dee calls than either control
stimulus. This pattern was unaffected by the presence of migrants,
thus demonstrating that the observed responses are innate. Our study
shows that learning cannot fully explain responses to heterospecific
chick-a-dee calls and that structural features distinguishing these calls
from other vocalizations are important.

Keywords: chickadee, communication network, eavesdropping, food
call, mobbing, paridae.
* Corresponding author; email: biosandoval@hotmail.com.
ORCIDs: Sandoval, https://orcid.org/0000-0002-0793-6747; Wilson, https://

orcid.org/0000-0002-6558-6415.

American Naturalist, volume 200, number 3, September 2022. q 2022 The University
The American Society of Naturalists. https://doi.org/10.1086/720441
Introduction

Many animals respond to cues and signals produced by
other species (Bradbury and Vehrencamp 2011). In some
cases, the costs of responding are severe (Dawkins and
Krebs 1978). Examples include predators being thwarted
by the startle (e.g., butterflies; Vallin et al. 2005) or decoy
displays (e.g., birds; Humphreys and Ruxton 2020) of prey,
birds starving their offspring by prioritizing brood para-
sites (Soler 2017), and prey such as fireflies (Lloyd 1965),
fish (Pietsch and Grobecker 1978), lizards (Chiszar et al.
1990), and spiders (Wignall and Taylor 2011) being lured
to their death by their predators’ deceptive signals. Re-
sponding to heterospecific cues and signals can also be
beneficial, even if the cues or signals are not intended
for the eavesdropping individual (Bradbury and Vehren-
camp 2011). Many predators and parasites, for example,
acquire their prey and hosts by localizing their scents
and sounds (Zuk and Kolluru 1998), and many prey find
food and avoid predation by responding to heterospe-
cific food calls (e.g., Japanese sika deer [Cevus nippon]
responding to calls of Japanese macaques [Macaca fus-
cata yakui]; Koda 2012) and alarm calls (Magrath et al.
2015). Given the significant costs of responding to some
heterospecific signals and the obvious benefits of respond-
ing to others, selection should favor receivers and sensory
systems that discriminate among heterospecific signals
(Guilford and Dawkins 1991).
Although the costs and benefits of responding to het-

erospecific signals have been well studied, the behav-
ioral and cognitive mechanisms underlying the responses
have not (Magrath et al. 2015). One hypothesis is that
animals appear to respond to heterospecific signals only
because they follow other species that are themselves re-
sponding to signals (Mönkkönen and Forsman 2002; Goo-
dale et al. 2010). A second hypothesis is that animals learn
to associate heterospecific signals with adaptive behavioral
of Chicago. All rights reserved. Published by The University of Chicago Press for

mailto:biosandoval@hotmail.com
https://orcid.org/0000-0002-0793-6747
https://orcid.org/0000-0002-6558-6415
https://orcid.org/0000-0002-6558-6415


000 The American Naturalist
responses, such as freezing or approaching (Griffin 2004),
or with salient environmental features, such as food and
predators (Magrath et al. 2015). Such associative learning
requires an animal to experience the signal (Griffin 2004;
Magrath et al. 2015). Golden-mantled ground squirrels (Sper-
mophilus lateralis) provide a clear example because they
learn to respond with antipredator behavior to previously
unfamiliar sounds that became reliably associated with the
appearance of predators (Shriner 1999). A third hypothesis
is that responses are innate (Magrath et al. 2015). In this
case, even unfamiliar heterospecific signals can elicit re-
sponses if they contain characteristics that are familiar be-
cause of phylogenetic conservation or evolutionary conver-
gence (Marler 1955; Morton 1977; Jurisevic and Sanderson
1998; Johnson et al. 2003; Fallow et al. 2011) or characteris-
tics that broadly stimulate diverse sensory systems (Endler
and Basolo 1998; Owings and Morton 1998; Fitch et al.
2002; Rendall et al. 2009). Of course, these mechanisms are
not mutually exclusive. For example, a species might show
an innate response to an unfamiliar signal but then show
increased responsiveness to the signal after learning that it
predicts the presence of food.
All species of Paridae in North America (chickadees:

Poecile atricapillus, Poecile carolinensis, Poecile cinctus, Poecile
gambeli, Poecile hudsonicus, Poecile rufescens, Poecile sclateri;
titmice:Baeolophus atricristatus,Baeolophus bicolor,Baeolophus
inornatus, Baeolophus ridgwayi, Baeolophus wollweberi) pro-
duce chick-a-dee calls (fig. 1) in response to arousing stimuli,
such as predators, food, and territorial intruders (Dixon
1949; Smith 1972; McLaren 1976; Gaddis 1985; Hailman
1989; Charrier et al. 2004; Bloomfield et al. 2005; Hailman
and Haftorn 2005; Lucas and Freeberg 2007; Owens and
Freeberg 2007; Hoeschele et al. 2009; Moscicki et al. 2010;
Cicero et al. 2020; Nocedal and Ficken 2020). Conspecifics
and diverse heterospecifics approach chick-a-dee calls, where,
upon arriving, they often engage in antipredator, foraging,
or aggressive behaviors appropriate to the stimulus elicit-
ing the calls (Hurd 1996; Templeton et al. 2005; Langham
et al. 2006; Templeton and Greene 2007; Schmidt et al.
2008; Soard and Ritchison 2009; Mahurin and Freeberg
2009; Courter and Ritchison 2010; Wilson and Mennill
2011; Grava et al. 2012; Hetrick and Sieving 2012; Randler
2012; Dutour et al. 2017, 2020; Landsborough et al. 2020).
Heterospecifics responding to chick-a-dee calls potentially
benefit by participating in multispecies mobbing events that
repel predators and deter their return (Pettifor 1990; Flass-
kamp 1994; Pavey and Smyth 1998; Consla and Mumme
2012) or by consuming food they might otherwise fail to
discover (Dolby and Grubb 1998; Wilson and Mennill 2011).
0

2

4

6

8

Fr
eq

ue
nc

y 
(k

H
z)

0 0.5 1 1.5 2

0

1

2

3

4

5

Time (s)

Fr
eq

ue
nc

y 
(k

H
z)

Figure 1: Spectrograms of chick-a-dee call (top) and fee bee song
(bottom) of black-capped chickadee. Chick-a-dee calls begin with
a series of introductory notes (A, B, and C notes) that are short, tonal,
relatively high frequency, and strongly frequency modulated, followed
by a series of dee notes that are relatively long and low frequency,
with harsh, harmonic-like structure and little frequency modula-
tion. The presence and number of each note type varies, but the note
types are always produced in the same order (A, B, C, dee; Lucas and
Freeberg 2007). The example shown here has been standardized for
use as a stimulus call (see “Methods”) and includes two introductory
notes and eight dee notes. Fee bee songs contain two tonal notes, in-
cluding an initial high-frequency fee note that descends in frequency,
followed after a short silence by a relatively low-frequency bee note of
similar duration and little frequency modulation. During a playback,
the selected stimulus was broadcast at a rate of six repetitions per min-
ute for 2 min. Spectrograms were generated with a 512-point fast
Fourier transform with Hamming window and 87.5% overlap. Tem-
poral resolution is 2.9 ms, frequency resolution is 43 Hz, and the
grayscale represents an amplitude range of 45 dB.



Innate Responses to Heterospecific Call 000
The potential benefits of responding to chick-a-dee calls may
thus serve as reinforcement that would motivate hetero-
specifics to learn to respond to calls directly or to follow
other birds as they respond to calls (Griffin 2004). Further
evidence that heterospecific responses to chick-a-dee calls
are learned is that heterospecifics respond less strongly or
not at all to other parid vocalizations, such as songs and
contact calls, which are not reliably associated with preda-
tors, food, or other reinforcing stimuli (Hurd 1996; Schmidt
et al. 2008; Randler 2012). Yet chick-a-dee calls also include
harsh (i.e., spanning broad frequency range at each moment
in time), low-frequency elements (Hailman 1989; Lucas and
Freeberg 2007) that, according tomotivation-structural rules,
should universally reflect high levels of signaler arousal and
hostility (Morton 1977; Owings and Morton 1998) and that
might therefore stimulate diverse sensory systems and elicit
innate responses from receivers. Consistent with the idea
of responses being innate, many birds approach the harsh,
low-frequency, and unfamiliar pishing sounds made by or-
nithologists and birdwatchers (Langham et al. 2006). It is
therefore possible that heterospecific responses to chick-a-
dee calls are innate instead of learned or that they are both.
The behavioral and cognitive mechanisms underlying

heterospecific responses to chick-a-dee calls remain unclear
because many of the relevant studies involved respondents
that are sympatric with the species producing the calls
(Hurd 1996; Templeton and Greene 2007; Schmidt et al.
2008; Wilson and Mennill 2011; Grava et al. 2012; Hetrick
and Sieving 2012; Landsborough et al. 2020), thus conflat-
ing the potential mechanisms involved. A few studies showed
that European birds, including parid and nonparid species,
respond to the unfamiliar chick-a-dee calls of allopatric
chickadees and titmice from North America (Randler 2012;
Dutour et al. 2017, 2020) and that diverse birds in California
respond to the chick-a-dee calls of allopatric parids from
other parts of North America and Europe (Langham et al.
2006). These studies suggest that heterospecific responses
are innate because prior experience with a species’ chick-
a-dee call is not necessary for the call to elicit a response.
However, respondents in these studies were sympatric with
local parids and were probably familiar with their calls. It
is therefore possible that respondents learned to respond
to the chick-a-dee calls of local parids and then general-
ized that response to the acoustically similar but unfamil-
iar calls of allopatric parids (Langham et al. 2006; Randler
2012; Dutour et al. 2017, 2020). A critical test of whether
heterospecific responses to chick-a-dee calls are innate would
be to show that birds that are permanently allopatric to
all Paridae respond to their calls. Yet we know of only one
such study. Nocera et al. (2008) broadcast chick-a-dee calls
from black-capped chickadees (Poecile atricapillus) at sites
in Belize during spring migration and found that migrants—
but not residents—responded (Nocera et al. 2008). The lack
of response by Neotropical residents suggests that chick-
a-dee calls elicit responses only from species that are sym-
patric with parids, though the study cautions that its find-
ings should be replicated at other locations in the tropics
and at other times of the year before generalizing its con-
clusions (Nocera et al. 2008).
In this study, we use an acoustic playback experiment

to test whether heterospecific responses to chick-a-dee calls
are learned, innate, or an epiphenomenon created by fol-
lowing other species. We tested these three hypotheses by
broadcasting chick-a-dee calls of black-capped chickadees
plus two control stimuli in Costa Rica, Colombia, and
Brazil, where no members of Paridae ever occur. We con-
ducted our trials in the winter, when Neotropical migrants
that might be familiar with chick-a-dee calls were present,
and in the temperate breeding season, when migrants were
absent. If heterospecific responses to chick-a-dee calls are
innate, then Neotropical residents should respond more
strongly to chick-a-dee calls than to control stimuli, regard-
less of whether migrants are present. If heterospecifics re-
spond to chick-a-dee calls by following other species that
may themselves be familiar with the calls, then resident
species should respond more strongly to chick-a-dee calls
only when migrants are present. If heterospecifics learn to
respond directly to chick-a-dee calls through experience
and associative learning, then resident species should never
respond to unfamiliar chick-a-dee calls. The study is one
of the few to test the cognitive or behavioral mechanisms
underlying heterospecific communication in animals and
thus provides insight into the development, ecology, and
evolution of this behavior.
Methods

General

We conducted playback trials in Costa Rica, Colombia,
and Brazil. In Costa Rica, we conducted trials during the
temperate breeding season when migrants were away on
their breeding grounds (July 8–13, 2013; N p 30 trials)
and during the preceding (January 7–12, 2013; N p 30
trials) and following winters (December 14–19, 2013; N p
30 trials) when migrants were present. Trials were conducted
at 30 sites that were located at North Heredia (107010N,
847050W, 1,200–1,500-m elevation; N p 60 trials at 20
sites) and the Lankester Botanical Garden (097500N, 837
530W, 1,400-m elevation; N p 30 trials at 10 sites) and that
were the same among seasons. In Colombia and Brazil, we
conducted all trials during the temperate breeding season,
when migrants were away (Colombia: August 11–14, 2013;
Brazil: September 11–16, 2013). In Colombia, we conducted
27 trials at 27 sites in Medellin (067140N, 757340W, 1,550-m
elevation; N p 12 trials) and the Río Claro Natural Reserve



000 The American Naturalist
(057500N, 747520W, 415-m elevation; N p 15 trials). In
Brazil, we conducted 21 trials at 21 sites in Pousada dos
Pirineos (157500S, 467570W, 895-m elevation;N p 11 trials)
and Alto Paraiso (147070S, 477310W, 1,200-m elevation;
N p 10 trials).
Following a randomized complete block design, we broad-

cast three playback treatments in random order during
each trial before moving to the next site and beginning
the next trial. Our study therefore included 138 trials and
three treatments, or 414 trial treatments. Playback sites
at each general location were selected haphazardly and
were located within secondary forest edges, green areas
with trees and bushes, cerrado vegetation, or coffee plan-
tations with live fences. Birds in our study were not color
banded and could not be identified as individuals. We
therefore separated playback sites by at least 100 m to re-
duce the probability of the same individuals responding at
multiple sites. All trials were conducted between 0600 and
1000 hours when diurnal birds are active.
Procedure

After choosing a playback site, we selected a tree that was
devoid of fruits and flowers, attached a loudspeaker (Pana-
sonic, model RP-SP48; frequency range: 140–20,000 Hz) to
a branch 1.5–2.5 m above the ground, oriented the speaker
upward, and connected it to a digital playback device (iPod
Nano Touch) containing our playback stimuli. Four flags
were placed at 907 angles around the speaker at a distance of
5m to assist in estimating the distances between the speaker
and approaching birds. The observer sat on the ground
8.6 m from the speaker, waited until no birds were detected
within 10 m of the speaker for at least 5 min, and began
the first of three playback treatments for that trial.
We broadcast three treatments during each trial (see de-

tails of stimulus construction below). Our experimental
treatment was the chick-a-dee call of the black-capped chick-
adee, which was repeated at a natural rate of 6 calls min21

for 2 min (fig. 1). Our positive control was the fee bee song
of the black-capped chickadee (Ficken et al. 1978), which
was also repeated at 6 songs min21 for 2 min (fig. 1). We
chose the fee bee song because its production is not asso-
ciated with external stimuli such as predators or food that
might be salient to heterospecifics (Ficken et al. 1978). It thus
controlled for heterospecific responses to unfamiliar and
functionally irrelevant biological sounds. Our negative con-
trol was a 2-min period of silence, which controlled for
spontaneous arrivals at our playback apparatus. The order
of treatments was randomized but with the constraint that
it was balanced among treatments across trials. Following
each treatment, we waited until no birds had been seen
within 10 m of the loudspeaker for at least 5 min before pro-
ceeding to the next treatment. We broadcasted the chick-a-
dee call and fee bee song treatments at a sound pressure level
of 80 dB (measured 1 m from the loudspeaker with a Sper
Scientific mini sound level meter, model 840014; 32–130-
dB response range; fast response; C-weighting).
We did not include a familiar positive control, such as

the mobbing call of a local species (as in Nocera et al. 2008),
for three reasons. First, no one species that we know of
produces high-arousal calls that would be familiar to res-
ident birds at all our playback sites. Second, we were con-
cerned that broadcasting two high-arousal calls in our re-
peated measures design could either be too disruptive to birds
or cause them to habituate to playback stimuli. Finally, inter-
preting a potential difference in responses to a familiar con-
trol stimulus and to our unfamiliar experimental stimulus
would be difficult because calls of different species, even
when uniformly familiar or unfamiliar to respondents, may
be unequally evocative. If responses to the familiar control
were stronger than responses to the unfamiliar chick-a-dee
call, it would be impossible to know whether the difference
was due to the difference in familiarity and thus learning
or to differences in the evocativeness of the two stimuli.
Following Sandoval and Wilson (2012), we measured

four response variables in situ during each 2-min treat-
ment: (1) number of species observed within 5 m of the
speaker; (2) maximum number of birds, which was the
sum of the maximum number of individuals of each spe-
cies that could be observed simultaneously within 5 m of
the speaker; (3) latency of the first bird to approach within
5 m of the speaker; and (4) minimum distance of any
bird from the speaker. We selected these measures because
they do not depend on having color-banded individuals.
Times were recorded to the nearest second with a stop-
watch, and distances were estimated along the horizontal
plane to the nearest 0.1 m. As in Sandoval and Wilson
(2012), we used 5 m as the threshold for inclusion because
dense vegetation at our sites made it difficult to detect and
monitor birds beyond this distance. If no birds approached
to within 5 m of the speaker during a 2-min treatment, we
reported zeros for number of species and maximum num-
ber of birds but did not assign values for latency or min-
imum distance. In addition to quantitative measures, we
noted the species of all responding birds and the species
of the first bird to respond. All trials were conducted by
just one of the authors, who was familiar with the avian
communities at the trial locations. Species names follow
the checklists established by the American Ornithologi-
cal Society for North, Middle, and South American birds
(Chesser et al. 2020; Remsen et al. 2021).
Stimuli

We created 30 playback stimuli for the chick-a-dee call
treatment and 30 playback stimuli for the positive control



Innate Responses to Heterospecific Call 000
to minimize problems associated with pseudoreplication
(Hurlbert 1984). Each stimulus was created in Raven Inter-
active Sound Analysis Software (ver. 1.4 Pro; Cornell Lab
of Ornithology Bioacoustics Research Program, Ithaca, NY)
by repeating a single vocalization at a rate of 6 vocaliza-
tions min21 for 2 min (fig. 1). Recordings were from the
authors’ personal collections or the Macaulay Library at
the Cornell Lab of Ornithology. Each stimulus individual
contributed amaximum of one song and one call. Five indi-
viduals contributed both a song and a call, though song and
call stimuli from the same individual were never presented
to the same subject. Vocalizations were selected on the basis
of high signal-to-noise ratios, typical structure, and no
overlapping sounds and were filtered with a high-passfilter
to remove background noise (chick-a-dee calls at 1 kHz; fee
bee songs at 2.7 kHz). As in previous research (Wilson and
Mennill 2011; Scully et al. 2019; Landsborough et al. 2020),
we standardized the note composition of chick-a-dee calls
before constructing the final playback stimuli to minimize
potential effects of note syntax on receiver responses (Tem-
pleton and Greene 2007; Mahurin and Freeberg 2009; Soard
and Ritchison 2009; Courter and Ritchison 2010). For each
call, we removed all but the final two introductory notes
and all but the first D note, and we then repeated the re-
maining D note seven times at a natural rate based on
the original call to create a call with eight D notes. The ampli-
tude of the D notes was adjusted to 27.6 dB, and the peak
amplitude of the two introductory notes was adjusted to
21 dB. This difference reflects the natural amplitude dif-
ference observed among note types (Wilson and Mennill
2011). The fee bee songs were normalized to a peak ampli-
tude of 21 dB.
Migratory Status and Sympatry with Paridae

It was important that subjects were unfamiliar with chick-
adee vocalizations, since prior experience would make it dif-
ficult to determine whether responses to calls were learned
or innate. We ensured that subjects had no prior experi-
ence in two ways. First, we classified each respondent spe-
cies as a Neotropical resident or Neotropical migrant using
the classifications provided in Birds of the World (Billerman
et al. 2020; table S1). Migrants likely overlap with black-
capped chickadees on their temperate breeding grounds or
migration routes and were therefore excluded from subse-
quent statistical analysis. Second, we determined whether
each responding species was sympatric with the black-capped
chickadee or any other parid, since all North American parids
produce some form of the chick-a-dee call. We obtained
digital species distribution maps from BirdLife Interna-
tional and Handbook of the Birds of the World (2020; da-
tum: World Geodetic System 1984 [National Imagery and
Mapping Agency 1997]) and projected them using the Lam-
bert Azimuthal Equal Area projection (latitude at projec-
tion center, 457; longitude at projection center,21007; false
northing, 0 m; false easting, 0 m) in the R package rgdal
(Bivand et al. 2020). Using the R package rgeos (Bivand
and Rundel 2020), we calculated the proportion of each re-
sponding species’ distribution that is sympatric with the
black-capped chickadee and the Paridae (table S1).
As expected, the black-capped chickadee and the Pari-

dae are sympatric with all migrants observed in our study
(table S1). The black-capped chickadee is also sympatric
with one resident species (house wren [Troglodytes aedon])
and the Paridae with 16 resident species (table S1). Although
the area of overlap between Paridae and these 16 resident
species is typically very small (table S1), we nevertheless
investigated whether individuals from the overlap regions
could have traveled to our playback locations. For each of
the 16 resident species, we determined their movement
behavior from their species account in Birds of the World
(Billerman et al. 2020). Eleven of the 16 are described as
sedentary or as wandering locally only along altitudinal
gradients. It is therefore unlikely that individuals from these
species would have experienced chickadee vocalizations
in the region of sympatry and then traveled the minimum
1,500 km to our northernmost playback sites in Costa
Rica. Streaked flycatcher (Myiodynastes maculatus), ver-
milionflycatcher (Pyrocephalus rubinus), orange-billednight-
ingale thrush (Catharsus aurantiirostris), and house wren
(Troglodytes aedon) are considered resident species, but
some of their northernmost populations are known to mi-
grate southward in winter. However, the nightingale thrush
and flycatchers were detected only during summer, and
house wrens migrate south only as far as central Mexico.
Respondents from these species therefore could not have
been from the migratory populations. Finally, boat-billed
flycatcher (Megarynchus pitangua) is known to wander
broadly; however, only 1% of its distribution is sympatric
with Paridae, and it was detected in only one trial. It is there-
fore unlikely that this one individual was familiar with
chick-a-dee calls. Overall, we are confident that all resident
birds responding in our study were unfamiliar with chick-
adee vocalizations.
We updated our response variables such that number

of species and maximum number of birds were exclusively
based on Neotropical resident species. Unfortunately, this
was not possible for latency or minimum distance because
we could not collect those data separately for each species
in the field.
Phylogenetic Context

To understand the taxonomic distribution of respondents,
we identified all avian species that are sympatric with our



000 The American Naturalist
playback locations and thus available to respond. Using
the species maps and spatial analysis techniques described
above, we identified 751 extant specieswith breeding ranges
overlapping at least one of our six playback locations. We
reviewed their habitat descriptions in Birds of the World
(Billerman et al. 2020) and retained the 692 species that in-
habit primarily terrestrial environments, where our play-
backs were conducted (tables 1, S2). Species excluded are
allmembers ofAnatidae, Laridae, Scolopacidae, Ciconiidae,
Alcedinidae,Aramidae, Ralidae (except ocellated crake [Mi-
cropygia schomburgkii]), Cinclidae, Donacobiidae, Ardeidae,
Threskiornithidae, Podicipedidae, Anhingidae, and Phal-
crocoracidae. The phylogenetic relationships of the 692 re-
tained species are shown in figure 2.
We also tested whether any species traits predicted

whether the species responded to chick-a-dee calls. On
the basis of previous research, we assumed that smaller
species (Da Cunha et al. 2017) and species that produce
calls that are structurally similar to chick-a-dee calls (Ju-
risevic and Sanderson 1998; Johnson et al. 2003) would
be more likely to respond. For each species that responded
to chick-a-dee calls in at least one trial, we identified a
closely related species that was sympatric with our study
sites but did not respond (see fig. 2). Mean body mass
values for species that responded and for thematching spe-
cies that did not were obtained from Birds of the World
(Billerman et al. 2020) or vertnet.org. For acoustic similar-
ity, we obtained recordings of vocalizations from five indi-
viduals per species from xeno-canto.org. We used the
library’s metadata to select high-quality recordings of calls
obtained at different locations or in different years, thus
minimizing the probability of sampling the same individ-
ual twice. We reviewed each recording and identified one
call from each with a high signal-to-noise ratio and typical
structure. The call (plus 20 ms of silence before and af-
ter the call) was normalized to a peak amplitude of 0 dB
and exported as a standalone sound clip (WAVE format,
16-bit amplitude encoding, 22.05-kHz sampling rate). We
compared the acoustic structure of each extracted call
with the structure of each of our 30 chick-a-dee call play-
back stimuli using spectrogram cross-correlation in Raven
(settings: 256-point fast Fourier transform, Hamming win-
dow, 87.5% overlap, 0–1-kHz bandstop filter). This tech-
nique compares the overall similarity of two sounds by
sliding them past each other in time and calculating a cor-
relation coefficient (a value between 0 and 1) at each time
offset. The peak correlation coefficient indicates the overall
similarity of the two sounds; a correlation of 0 indicates
that the sounds do not match at all, whereas a value of
1 indicates that the sounds are identical. For each species,
we calculated the median peak correlation between its five
vocalizations and the 30 chick-a-dee calls. We note that
other variables—including local predation pressure, whether
a species forages on the ground, andwhether a species resides
in stable social groups—might also affect their responses
(Sandoval and Wilson 2012; Da Cunha et al. 2017), but
our small sample of respondents and the limited informa-
tion available for many Neotropical residents precluded
the inclusion of these variables.
Statistical Analysis

All analyses were conducted in R (R Development Core
Team 2020), and all data and R code underlying the anal-
yses and figures are in the Dryad Digital Repository
(https://doi.org/10.5061/dryad.vdncjsxwq; Sandoval and
Wilson 2021). Preliminary analyses showed that the num-
ber of resident species and the maximum number of resi-
dent birds responding were highly correlated (Spearman
Table 1: Taxonomic distribution of Neotropical residents
at our study sites
Order
No.
sympatric
families
No.
sympatric
species
No.
responding
species
All species
 58
 692
 38

Accipitriformes
 1
 28 (4.0)
 0 (.0)

Apodiformes
 2
 64 (9.2)
 4 (10.5)

Caprimulgiformes
 1
 12 (1.7)
 0 (.0)

Cariamiformes
 1
 1 (.1)
 0 (.0)

Cathartiformes
 1
 4 (.6)
 0 (.0)

Charadriiformes
 1
 3 (.4)
 0 (.0)

Columbiformes
 1
 24 (3.5)
 0 (.0)

Coraciiformes
 1
 4 (.6)
 1 (2.6)

Cuculiformes
 1
 12 (1.7)
 0 (.0)

Falconiformes
 1
 9 (1.3)
 0 (.0)

Galbuliformes
 2
 8 (1.2)
 0 (.0)

Galliformes
 2
 10 (1.4)
 0 (.0)

Gruiformes
 1
 1 (.1)
 0 (.0)

Nyctibiiformes
 1
 2 (.3)
 0 (.0)

Passeriformes
 30
 422 (61.1)
 32 (84.2)

Piciformes
 4
 31 (4.5)
 1 (2.6)

Psittaciformes
 1
 26 (3.8)
 0 (.0)

Rheiformes
 1
 1 (.1)
 0 (.0)

Steatornithiformes
 1
 1 (.1)
 0 (.0)

Strigiformes
 2
 14 (2.0)
 0 (.0)

Tinamiformes
 1
 8 (1.2)
 0 (.0)

Trogoniformes
 1
 6 (.9)
 0 (.0)
Note: Shown for 22 orders (and for all species combined) are the number
of sympatric families and species plus the number of species that approached
to within 5 m of the playback speaker during the chick-a-dee call treatment
of at least one trial. Values in parentheses show the percent of all sympatric
or responding species. Species were considered sympatric if they inhabited a pri-
marily terrestrial environment and if their breeding range overlapped at least
one of our six playback sites, as determined by species distribution maps from
BirdLife International and Handbook of the Birds of the World (2020). Taxon-
omy follows the American Ornithological Society’s checklists for North, Middle,
and South American birds (Chesser et al. 2020; Remsen et al. 2021). Informa-
tion about individual species is provided in table S2.

http://vertnet.org
https://xeno-canto.org/
https://doi.org/10.5061/dryad.vdncjsxwq


Crypturellus soui

Crypturellus undulatus

Crypturellus parvirostris

Tinamus major
Nothocercus bonapartei
Nothura maculosa

Taoniscus nanus
Rhynchotus rufescens

Rhea americana

Guira guira

Crotophaga sulcirostris

Crotophaga ani

Crotophaga major

Tapera naevia

Dromococcyx phasianellus

Neomorphus geoffroyi

Piaya cayana

Coccyzus minor

Coccyzus melacoryphus

Coccyzus euleri

Coccycua minuta

Micropygia schomburgkii

Uropelia campestrisColumbina squammata

Columbina incaColumbina passerina
Columbina talpacoti

Columbina minuta

Columbina picui

Claravis pretiosa

Paraclaravis mondetoura

Leptotila verreauxi

Leptotila rufaxilla
Zentrygon costaricensis

Zentrygon chiriquensisZenaida macroura

Zenaida auriculata

Zenaida asiatica

Geotrygon montana

Patagioenas cayennensis

Patagioenas picazuro

Patagioenas flavirostris

Patagioenas subvinacea

Patagioenas plumbea

Patagioenas speciosa

Patagioenas fasciata

Nyctiphrynus ocellatus

Antrostomus rufus
Setopagis parvula

Systellura longirostris

Hydropsalis maculicaudus

Hydropsalis torquata

Eleothreptus anomalus

Uropsalis lyra

Nyctidromus albicollis

Chordeiles nacunda

Chordeiles pusillus

Chordeiles acutipennis

Glaucis hirsutus

Threnetes ruckeri

Phaethornis pretrei

Phaethornis ruber

Phaethornis striigularis

Phaethornis syrmatophorus

Phaethornis longirostris

Phaethornis guy

Eutoxeres aquila

M
etallura tyrianthina

Adelomyia m
elanogenys

Lophornis adorabilis

Lophornis m
agnificus

Lophornis helenae

Haplophaedia aureliae

Heliodoxa leadbeateri

Coeligena coeligenaHeliom
aster longirostris

Heliom
aster squam

osus

Lam
pornis calolaem

us

Calliphlox am
ethystina

Calliphlox m
itchellii

Calliphlox bryantae

Chaetocercus m
ulsant

Selasphorus scintilla

Chalybura buffonii

Thalurania furcata

Thalurania colom
bica

Eupherusa exim
ia

Eupherusa nigriventris

Saucerottia saucerottei

C
hionom

esa lactea
C

hionom
esa fim

briata
Saucerottia cyanura

C
hlorestes julie

U
ranom

itra franciae

C
hrysuronia goudoti

Am
azilia tzacatl

Eupetom
ena cirrochloris

Eupetom
ena m

acroura

Cam
pylopterus hem

ileucurus

Cynanthus auriceps

Chlorostilbon gibsoni

Chlorostilbon lucidus

Chlorostilbon m
ellisugus

Colibri coruscans

Colibri delphinae

Colibri thalassinus

Colibri serrirostris

Doryfera ludovicae

Polytmus guainumbi

Chrysolampis mosquitus

Anthracothorax nigricollis

Heliactin bilophus

Tachornis squamata

Panyptila cayennensis

Chaetura brachyura

Chaetura meridionalis

Chaetura vauxi

Cypseloides fumigatus

Cypseloides cryptus

Cypseloides cherriei

Streptoprocne zonaris

Streptoprocne rutila

Nyctibius griseus

Nyctibius grandis

Steatornis caripensis

Car
ac

ar
a 

ch
er

iw
ay

Car
ac

ar
a 

pla
nc

us

M
ilv

ag
o 

ch
im

ac
him

a
Fa

lco
 sp

ar
ve

riu
s

Fa
lco

 fe
m

or
ali

s

Fa
lco

 ru
fig

ula
ris

He
rp

et
ot

he
re

s 
ca

ch
in

na
ns

M
icr

as
tu

r r
uf

ico
llis

M
icr

as
tu

r s
em

ito
rq

ua
tu

s

Pion
us

 m
en

str
uu

s

Pion
us

 se
nil

is

Pion
us

 ch
alc

op
ter

us

Pion
us

 m
ax

im
ilia

ni

Amaz
on

a a
maz

on
ica

Amaz
on

a m
erc

en
ari

us

Amaz
on

a a
utu

mna
lis

Amaz
on

a a
lbi

fro
nsAmazo

na aestiv
a

Amazo
na och

roce
phala

Alip
iop

sit
ta 

xa
nth

op
s

Brot
og

eri
s j

ug
ula

ris

Brot
og

eri
s c

hir
iri

Fo
rp

us
 co

ns
pic

illa
tu

s

Fo
rp

us
 xa

nt
ho

pt
er

yg
ius

Diop
sit

tac
a n

ob
ilis

An
od

or
hy

nc
hu

s h
ya

cin
thi

nu
s

Psit
tac

ara
 le

uc
op

hth
alm

us

Psit
tac

ara
 w

ag
ler

i

Eup
sit

tul
a a

ur
ea

Psit
tac

ar
a f

ins
ch

i

Ara
 ar

ar
au

na

Ara
 se

ve
ru

s

Orth
op

sit
tac

a m
an

ila
tus

Prim
oli

us
 m

ar
ac

an
a

Bo
lbo

rh
yn

ch
us

 lin
eo

la

Cotinga nattereriiProcnias tricarunculatusQuerula purpurataPyroderus scutatusRupicola peruvianus
Ampelion rubrocristatus

Pipreola aureopectus

Chloropipo flavicapilla

Antilophia galeataChiroxiphia linearis

Lepidothrix coronata

Manacus manacus

Ceratopipra erythrocephalaPseudopipra pipraMachaeropterus striolatus

Myiobius atricaudus
Onychorhynchus coronatus

Oxyruncus cristatus

Tityra semifasciataTityra cayanaTityra inquisitor

Xenopsaris albinucha

Pachyramphus polychopterus

Pachyramphus castaneus

Pachyramphus cinnamomeus

Pachyramphus rufus
Pachyramphus viridis

Pachyramphus versicolor
Pachyramphus validus

Schiffornis stenorhyncha

Hirundinea ferruginea

Myiotriccus ornatus

Euscarthmus meloryphus

Euscarthmus rufomarginatus

Zimmerius chrysops

Zimmerius vilissimus

Camptostoma obsoletum
Phyllomyias reiseri

Phyllomyias fasciatus

Phyllomyias griseiceps

Capsiempis flaveola

Phaeomyias murina

Serpophaga subcristata

Serpophaga nigricans

Serpophaga cinerea

Culicivora caudacuta

Mecocerculus leucophrys

Tyrannulus elatus

Myiopagis viridicata
Suiriri suiriri

Suiriri affinis

Elaenia flavogaster

Elaenia spectabilis

Elaenia chiriquensis

Elaenia mesoleuca

Elaenia frantzii
Elaenia sordida

Elaenia cristata

Fluvicola nengeta

Pyrocephalus rubinus

Arundinicola leucocephala

Alectrurus tricolor

Fluvicola albiventer

Sublegatus m
odestus

M
yiophobus flavicans

Ochthoeca fum
icolor

Colonia colonus

Em
pidonax albigularis

Contopus sordidulus

Contopus cinereus

Sayornis nigricans

M
itrephanes phaeocercus

Lathrotriccus euleri

Cnem
otriccus fuscatus

M
yiophobus fasciatus

Knipolegus lophotes

Knipolegus nigerrim
us

Satrapa icterophrys
Xolm

is cinereus

M
yiotheretes fum

igatus

Xolm
is velatus

Legatus leucophaius

Attila spadiceus

Machetornis rixosa

Pitangus lictor

Pitangus sulphuratus

Tyrannus savana

Tyrannus albogularis

Tyrannus m
elancholicus

Empidonomus varius

Megarynchus pitangua

Myiodynastes luteiventris

Myiodynastes maculatus

Myiozetetes similis

Myiozetetes cayanensis

Myiarchus swainsoni

Myiarchus panamensis

Myiarchus tuberculifer

Myiarchus ferox

Myiarchus nuttingi

Myiarchus tyrannulus

Rhytipterna holerythra

Platyrinchus mystaceus

Poecilotriccus ruficeps

Poecilotriccus sylvia

Todirostrum nigriceps

Todirostrum cinereum

Hemitriccus margaritaceiventer

Hemitriccus striaticollis

Atalotriccus pilaris

Hemitriccus granadensis

Oncostoma cinereigulare

Lophotriccus pileatus

Rhynchocyclus olivaceus

Tolmomyias sulphurescens

Mionectes striaticollis

Mionectes olivaceus

Mionectes oleagineus

Leptopogon amaurocephalus

Leptopogon superciliaris

Phylloscartes poecilotis
Corythopis delalandi

Piprites chloris

Poliocrania exsul

Sipia pallia
ta

Hylo
phyla

x n
aevio

ides

Hafferia im
maculata

Cerco
macra

 nigrica
ns

Cerco
macro

ides p
arke

ri

Cerco
macro

ides ty
rannina

Drym
ophila str

iatice
ps

Thamnophilus torquatus

Thamnophilus doliatus

Thamnophilus multistria
tus

Thamnophilus unicolor

Thamnophilus caerulescens

Thamnophilus pelzelni

Thamnophilus nigriceps

Thamnophilus atrinucha

Dysithamnus mentalis

Herpsilochmus atric
apillus

Herpsilochmus longirostris

Taraba major

Cymbilaimus lin
eatus

Myrm
otherula pacifi

ca

Myrm
otherula axill

aris
Form

iciv
ora ru

fa

Form
iciv

ora grise
a

Form
iciv

ora m
elanogaste

r

Micr
orhopias q

uixe
nsis

Euch
repomis c

allin
ota

Conopophaga lineata

Melanopareia torquata

Xenops rutilans

Xenops minutus

Clibanornis rubiginosus

Clibanornis rectirostris
Thripadectes holostictus

Thripadectes rufobrunneus

Thripadectes flammulatus

Anabacerthia striaticollis

Syndactyla dimidiata

Megaxenops parnaguae

Philydor fuscipenne

Premnoplex brunnescens

Furnarius rufus

Cranioleuca pallida
Cranioleuca erythrops

Cranioleuca vulpina

Cranioleuca semicinerea

Synallaxis brachyura

Synallaxis albescens
Synallaxis frontalis

Certhiaxis cinnamomeus
Synallaxis scutata
Synallaxis unirufa

Synallaxis hypospodia

Anumbius annumbi

Phacellodomus ruber

Glyphorynchus spirurus

Dendrocolaptes picumnus

Dendrocolaptes platyrostris

Dendrocolaptes sanctithomae

Xiphocolaptes promeropirhynchus

Lepidocolaptes angustirostris

Lepidocolaptes souleyetii

Lepidocolaptes affinis

Campylorhamphus pusillus

Campylorhamphus trochilirostris

Dendroplex picus

Xiphorhynchus susurrans

Sittasomus griseicapillus

Sclerurus mexicanus

Sclerurus scansor

Geositta poeciloptera

Formicarius analis

Formicarius rufipectus

Grallaricula flavirostris

Grallaricula nana

Hylopezus perspicillatus

Grallaria ruficapilla

Scytalopus spillmanni

Scytalopus atratus

Vireolanius pulchellus

Vireolanius exim
ius

Pachysylvia decurtata

H
ylophilus am

aurocephalus

Pachysylvia sem
ibrunnea

Tunchiornis ochraceiceps

Vireo leucophrys
Vireo flavoviridis

Vireo olivaceus

C
yclarhis gujanensis

C
yanocorax yncas

C
yanocorax cyanopogon

C
yanocorax affinis

Psilorhinus m
orio

C
yanocorax cristatellus

Cyanolyca cucullata

M
im

us gilvus

M
im

us saturninus

C
atharus aurantiirostris

C
atharus fuscater

Turdus leucops

Turdus serranus

Turdus fuscater

Turdus ignobilis

sir
tn

ev
if

ur
su

dr
uT

Turdus grayi

Turdus obsoletus

Turdus leucom
elas

Turdus assim
ilis

Turdus plebejus

M
yadestes ralloides

M
yadestes m

elanops

su
ru

na
le

m
su

ne
ac

oh
p

ma
R

sir
tn

ev
ie

re
ni

c
se

ta
bo

rc
i

M
ae

b
mu

lp
ali

tp
oil

oP
al

oc
i

mu
d

ali
tp

oil
oP

P
he

ug
op

ed
iu

s 
fa

sc
ia

to
ve

nt
ris

P
he

ug
op

ed
iu

s 
ru

til
us

P
he

ug
op

ed
iu

s 
ge

ni
ba

rb
is

P
he

ug
op

ed
iu

s 
m

ys
ta

ca
lissutsedo

m
sulihcrotna

C

sullipacirgin
sulihcrotna

C
sitocuel

sulihcrotna
C H

en
ic

or
hi

na
 le

uc
op

hr
ysatcitsocuel

anihrocine
HC

in
ny

ce
rth

ia
 u

ni
ru

fa nodea
setydolgorT

sisnetalp
surohtotsi

Csu
es

ir
g

su
hc

ny
hr

ol
yp

ma
C

su
ta

no
z

su
hc

ny
hr

ol
yp

ma
C

Ptiliogonys caudatus Eu
ph

on
ia

 c
hl

or
ot

ic
a

Eu
ph

on
ia

 x
an

th
og

as
te

r

Eu
ph

on
ia

 v
io

la
ce

a

Eu
ph

on
ia

 tr
in

ita
tis

Eu
ph

on
ia

 la
ni

iro
st

ris

Eu
ph

on
ia

 fu
lv

ic
ris

sa

Eu
ph

on
ia

 lu
te

ic
ap

illa

Eu
ph

on
ia

 e
le

ga
nt

is
si

m
a

Sp
in

us
 m

ag
el

la
ni

cu
s

Sp
in

us
 p

sa
ltr

ia

Am
bly

ce
rc

us
 h

olo
se

ric
eu

s

Hy
po

py
rrh

us
 py

ro
hy

po
ga

ste
r

Gno
rim

op
sa

r c
ho

pi

Chr
ys

om
us

 ru
fic

ap
illu

s

Chr
ys

om
us

 ic
te

ro
ce

ph
alu

s

Ps
eu

do
lei

ste
s g

uir
ah

ur
o

Molo
thr

us
 ae

ne
us

Molo
thr

us
 bo

na
rie

ns
is

Molo
thr

us
 or

yz
ivo

ru
s

Molo
thr

us
 ru

foa
xil

lar
is

Ag
ela

ius
 ph

oe
nic

eu
s

Quis
ca

lus
 lu

gu
br

is

Quis
ca

lus
 m

ex
ica

nu
s

Di
ve

s d
ive

s

Ict
er

us
 m

es
om

ela
s

Ict
er

us
 ja

m
ac

aii

Ict
er

us
 p

yr
rh

op
te

ru
s

Ic
te

ru
s 

au
ric

ap
illu

s

Ic
te

ru
s 

ch
ry

sa
te

r

Ca
cic

us
 c

el
a

Ca
cic

us
 s

ol
ita

riu
s

Ps
ar

oc
ol

iu
s 

de
cu

m
an

us

Ps
ar

oc
ol

iu
s 

gu
at

im
oz

in
us

Ps
ar

oc
ol

iu
s 

m
on

te
zu

m
a

Le
ist

es
 s

up
er

cil
ia

ris

St
ur

ne
lla

 m
ag

na

Le
ist

es
 m

ilit
ar

isC
hl

or
os

pi
ng

us
 fl

av
op

ec
tu

s
Am

m
od

ra
m

us
 s

av
an

na
ru

m

Am
m

od
ra

m
us

 h
um

er
al

is

Ar
re

m
on

 a
ur

an
tiir

os
tri

s

Ar
re

m
on

 fl
av

iro
st

ris
Ar

re
m

on
 b

ru
nn

ei
nu

ch
a

Ar
re

m
on

 c
ra

ss
iro

st
ris

Ar
re

m
on

 a
ss

im
ilis

At
la

pe
te

s 
la

tin
uc

hu
s

At
la

pe
te

s 
fla

vi
ce

ps
M

el
oz

on
e 

le
uc

ot
is

M
el

oz
on

e 
ca

ba
ni

si

At
la

pe
te

s 
al

bi
nu

ch
a

Zo
no

tri
ch

ia
 c

ap
en

si
s

At
la

pe
te

s 
sc

hi
st

ac
eu

s

Ba
si

le
ut

er
us

 ru
fif

ro
ns

Ba
si

le
ut

er
us

 c
ul

ic
iv

or
us

M
yi

ob
or

us
 m

in
ia

tu
s

M
yi

ob
or

us
 o

rn
at

us

M
yi

ot
hl

yp
is

 le
uc

op
hr

ys

M
yi

ot
hl

yp
is

 fl
av

eo
la

M
yi

ot
hl

yp
is

 lu
te

ov
iri

di
s

M
yi

ot
hl

yp
is

 fu
lv

ic
au

da

Se
to

ph
ag

a 
pe

te
ch

ia

Se
to

ph
ag

a 
pi

tia
yu

m
i

G
eo

th
ly

pi
s 

ae
qu

in
oc

tia
lis

G
eo

th
ly

pi
s 

po
lio

ce
ph

al
a

Phe
uc

tic
us

 tib
ial

is

Cya
no

lox
ia 

cy
an

oid
es

Cya
no

lox
ia 

br
iss

on
ii

Hab
ia 

cri
sta

ta

Hab
ia 

gu
ttu

ral
is

Pira
ng

a b
ide

nta
ta

Pira
ng

a f
lav

a
Mitro

sp
ing

us
 ca

ss
ini

i

Chlo
rop

ho
nia

 py
rrh

op
hry

s

Chlo
rop

ho
nia

 cy
an

ea

Chlo
rop

ho
nia

 ca
llo

ph
rys

Sch
isto

ch
lamys 

rufica
pillu

s

Sch
ist

oc
hla

mys
 m

ela
no

pis

Neo
thr

au
pis

 fa
sc

iat
a

Coereba fla
veola

Asemospiza
 obscu

ra

Tiaris 
oliva

ceus

Tangara gyrola

Tangara inornata

Tangara arthus

Tangara icterocephala

Tangara nigrovirid
is

Tangara dowii

Tangara va
ssorii

Ixothraupis guttata

Stilp
nia cyanicollis

Chalco
thraupis r

ufice
rvix

Stilp
nia cayana

Stilp
nia vit

riolina

Compso
thraupis l

oric
ata

Sporathraupis c
yanocephala

Dubusia
 ta

eniata

Aniso
gnathus la

cry
mosus

Buthraupis m
ontana

Aniso
gnathus s

omptuosus

Thraupis s
aya

ca

Thraupis e
pisc

opus

Thraupis p
alm

arum
Pipr

ae
ide

a m
ela

no
no

ta

Irid
os

orn
is 

po
rph

yro
ce

ph
alu

s

Saltator maximus

Saltator atriceps

Saltator coerulescens
Saltator fu

liginosus

Saltator striatipectus

Saltator similis

Saltatricula atricollis

Hemithraupis guiraHemithraupis flavicollis
Heterospingus xanthopygius

Chlorophanes spiza

Rhopospina caerulescens

Emberizoides herbicola

Microspingus cinereusCypsnagra hirundinacea

Thlypopsis sordidaThlypopsis superciliarisSphenopsis frontalis

Kleinothraupis atropileus

Dacnis lineataDacnis cayana
Tersina viridis

Tachyphonus rufus

Ramphocelus flammigerus

Ramphocelus carbo
Ramphocelus dimidiatus

Trichothraupis melanops

Eucometis penicillataLoriotus luctuosus

Coryphaspiza melanotis

Tachyphonus delatrii

Volatinia jacarina

Conirostrum speciosum

Conirostrum albifrons

Conirostrum leucogenys

Charitospiza eucosma

Diglossa caerulescens
Diglossa cyanea

Sicalis citrinaSicalis columbianaSicalis flaveola

Sporophila intermedia
Sporophila corvina

Sporophila morelletiSporophila plumbeaSporophila leucoptera

Sporophila caerulescens
Sporophila nigricollis
Sporophila luctuosa

Sporophila minuta
Sporophila bouvreuil

Sporophila crassirostris

Sporophila angolensisSporophila funerea

Nemosia pileata

Pa
ss

er
 d

om
es

tic
us

A
nt

hu
s 

lu
te

sc
en

s

Stelgidopteryx ruficollis

Stelgidopteryx serripennis

Progne tapera

Progne chalybea

Pygochelidon cyanoleuca

Tachycineta leucorrhoa

Tachycineta albiventer

Ca
ria

m
a 

cr
ist

at
a

ab
la

ot
yT

air
al

uc
in

uc
en

eh
t

A

munaciratsoc
muidicual

G
G

laucidium
 brasilianum

abilohc
spocsage

M
iikralc

spocsage
MLo

ph
os

tri
x 

cr
is

ta
ta

P
ul

sa
tri

x 
pe

rs
pi

ci
lla

ta

C
ic

ca
ba

 n
ig

ro
lin

ea
ta

C
ic

ca
ba

 v
irg

at
a

C
ic

ca
ba

 a
lb

ita
rs

is sunainigriv
obu

B

suigyts
ois

A
rota

malc
ois

A

Tr
og

on
 c

ur
uc

ui
Tr

og
on

 v
io

la
ce

us
Tr

og
on

 c
hi

on
ur

us
Tr

og
on

 m
as

se
na

Tr
og

on
 m

el
an

ur
us

Tr
og

on
 c

ol
la

ris

N
on

nu
la

 ru
be

cu
la

M
al

ac
op

til
a 

m
ys

ta
ca

lis

N
ys

ta
lu

s 
ch

ac
ur

u
N

ys
ta

lu
s 

m
ac

ul
at

us
N

ys
ta

lu
s 

ra
di

at
us

M
on

as
a 

m
or

ph
oe

us
M

on
as

a 
ni

gr
ifr

on
s

G
al

bu
la

 ru
fic

au
da

R
am

ph
as

to
s 

to
co

R
am

ph
as

to
s 

su
lfu

ra
tu

s

R
am

ph
as

to
s 

vi
te

llin
us

Pt
er

og
lo

ss
us

 to
rq

ua
tu

s

Au
la

co
rh

yn
ch

us
 p

ra
si

nu
s

Au
la

co
rh

yn
ch

us
 h

ae
m

at
op

yg
us

Se
m

no
rn

is
 fr

an
tz

ii
Eu

bu
cc

o 
bo

ur
ci

er
ii

Pi
cu

m
nu

s 
ol

iva
ce

us

Ca
m

pe
ph

ilu
s 

po
lle

ns

Ca
m

pe
ph

ilu
s 

m
el

an
ol

eu
co

s

Ca
m

pe
ph

ilu
s 

gu
at

em
al

en
sis

Ce
le

us
 o

br
ie

ni

Ce
le

us
 fl

av
es

ce
ns

Dr
yo

co
pu

s 
lin

ea
tu

s

Co
la

pt
es

 m
el

an
oc

hl
or

os

Co
la

pt
es

 ri
vo

lii

Co
la

pt
es

 p
un

ct
ig

ul
a

Co
la

pt
es

 c
am

pe
st

ris

Co
la

pt
es

 ru
bi

gi
no

su
s

Pi
cu

lu
s 

lita
e

M
el

an
er

pe
s 

ca
nd

id
us

M
el

an
er

pe
s 

fo
rm

ici
vo

ru
s

M
el

an
er

pe
s 

pu
lc

he
r

M
el

an
er

pe
s 

ru
br

ic
ap

illu
s

M
el

an
er

pe
s 

ho
ffm

an
ni

i

D
ry

ob
at

es
 d

ig
nu

s

D
ry

ob
at

es
 p

as
se

rin
us

D
ry

ob
at

es
 k

irk
ii

D
ry

ob
at

es
 m

ix
tu

s

D
ry

ob
at

es
 fu

m
ig

at
us

Ba
ry

ph
th

en
gu

s 
ru

fic
ap

illu
s

M
om

ot
us

 a
eq

ua
to

ria
lis

M
om

ot
us

 le
ss

on
ii

Sarcoram
phus papa

C
athartes aura

C
athartes burrovianus

C
oragyps atratus

G
am

psonyx sw
ainsonii

Elanus leucurus

C
hondrohierax uncinatus

Leptodon cayanensis
Elanoides forficatus

G
eranospiza caerulescens

M
orphnarchus princeps

R
upornis m

agnirostris
Parabuteo unicinctus

su
dit

in
oe

tu
B

si
sn

ec
ia

ma
j

oe
tu

B
su

ru
yh

ca
rb

oe
tu

B G
eranoaetus albicaudatus

B
uteogallus m

eridionalis

B
uteogallus urubitinga

B
uteogallus coronatus

R
ostrham

us sociabilis
Ictinia plum

bea
Busarellus nigricollis

H
arpagus bidentatus

Accipiter bicolor
C

ircus buffoni
Accipiter striatus
H

arpia harpyja
M

orphnus guianensis
Spizaetus tyrannus
Spizaetus ornatus

C
haradrius collaris

Vanellus cayanus
Vanellus chilensis

Crax fasciolata
Ortalis columbiana

Ortalis cinereiceps

Chamaepetes goudotii

Chamaepetes unicolor

Penelope superciliaris

Colinus cristatus
Odontophorus hyperythrus

Odontophorus guttatus

Dendrortyx leucophrys

Passeriform
es

Rheiformes
Tinamiformes

GalliformesGruiformesCuculiformes
Columbiformes

Steatornithiformes

Nyctibiiformes

Caprimulgiformes

Apodiformes

Charadriiform
es

Cathartiform
es

Accipitriform
es

St
rig

ifo
rm

es

Tr
og

on
ifo

rm
es

C
or

ac
iif

or
m

es

G
al

bu
lifo

rm
es

Pi
cif

or
m

es

Ca
ria

m
ifo

rm
es

Fa
lco

nif
or

m
es

Ps
itta

cif
or

mes

Figure 2: Phylogenetic distribution of extantNeotropical avian species with breeding ranges that overlap at least one of our six playback locations.
The tree is a consensus of 10,000 trees provided by birdtree.org, which provides distributions of trimmed phylogenies based on themethodology of
Jetz et al. (2012), the updated taxonomy of Jetz et al. (2014), and the backbone tree of Hackett et al. (2008). Source trees were combined in the R
package ape (Paradis and Schliep 2019) using the consensus function and a proportion of 0.5 for consensus. Species in red responded to the chick-
a-dee call treatment of at least one trial, whereas species in black or blue never responded. Species in blue were selected as nonrespondents in our
phylogenetic logistic regression. Species in the same taxonomic order are indicated by shading of the same color. Taxonomy follows the American
Ornithological Society’s checklists for South, Central, and North American birds (Chesser et al. 2020; Remsen et al. 2021). Although 692 species
were sympatric with our playback locations, only 689 are shown on the tree because three nonrespondent species recognized by the American
Ornithological Society’s checklists (Arremon atricapillus, Buteo plagiatus,Momotus subrufescens) were not included in the phylogenies provided
by birdtree.org.

http://birdtree.org
http://birdtree.org


000 The American Naturalist
correlation: rho p 0:99,N p 414, P ! :0001), even when
considering only those trial treatments where at least one
resident species responded (rho p 0:82, N p 121, P !

:0001).We assume that this was becausemost resident spe-
cies hold year-round territories (Billerman et al. 2020) and
that for each species, only the individual or breeding pair
that held the territory in which the trial was conducted
would have approached the speaker. To avoid redundancy,
we eliminated the maximum number of birds from fur-
ther consideration. The remaining response variables (num-
ber of resident species, latency, minimum distance) were
either uncorrelated or weakly correlated (all rho ≤ 0:30).
We tested our hypotheses using generalized linear mixed

models implemented in the R package glmmTMB (Brooks
et al. 2017). For our first analysis, we tested whether the
number of resident species that approached to within 5 m
of the speaker was greater in response to unfamiliar chick-
a-dee calls than in response to unfamiliar songs or silence
and whether the pattern of response was affected by the
presence of migrants that might have been familiar with
chickadee vocalizations. Playback treatment (chick-a-dee
call, fee bee song, silence), the status of migrants (present vs.
absent, as determined by season), and their two-way inter-
action were included as fixed factors. We included country
(Costa Rica, Colombia, Brazil) as a fixed factor to account
for unmeasured differences among regions thatmight have
affected responses, such as differences in climate, vegeta-
tion, and the composition of local avian communities.
We included trial number (1–138) nested within playback
site (1–78) as a random effect to account for possible de-
pendencies among responses observed at the same site or
during the same trial. The response was modeled using a
Poisson distribution with log link, which is appropriate for
zero-bounded, positively skewed variables that are based
on count data (Mun 2008).We included all trial treatments
(N p 414 treatments from 138 trials) because trial treat-
ments in which species respond and those in which species
do not respond are both informative.
Our first analysis could not explicitly consider species

identity as a predictor variable, so it is possible that results
would have been based on the same few resident species
responding in all trials. In our second analysis, we ad-
dressed this concern by testing whether a given species’
probability of approaching towithin 5m of the speaker dif-
fered among treatments and as a function of the presence
of migrants. For this analysis, it was necessary to estimate
which resident species were present in the vicinity of the
playback site at the time of the trial and thus available
to respond during a given trial treatment. For each treat-
ment within a trial, we therefore included separate obser-
vations for each resident species that ultimately responded
during any of the three treatments of the trial. After exclud-
ing 42 trials in which no species responded and eight trials
in which only migrants responded, the analysis included
42 resident species observed among 88 trials. The response
variable was whether the species approached to within 5 m
of the speaker during the trial treatment and was modeled
using a binomial distribution and logit link. The binomial
distribution is appropriate for variables that include only
two outcomes (Mun 2008). Treatment, the status of mi-
grants, their two-way interaction, and country were again
included as fixed factors, and species identity and trial
number nested within playback site were included as ran-
dom effects.
For our third analysis, we tested whether the latency

of the first resident species to respond differed among
treatments and as a function of the presence of migrants.
We excluded 293 trial treatments in which no residents
responded and 10 trial treatments in which a migrant
responded before a resident. Our final sample therefore in-
cluded 34 different species and 111 trial treatments. Treat-
ment, the status of migrants, their two-way interaction,
and country were included as fixed factors, and species
identity and trial number nested within playback site were
included as random effects. Latency to respond was re-
scaled to between 0 and 1 and then modeled using a beta
distribution and logit link (Migliorati et al. 2018). The beta
distribution is appropriate for asymmetric doubly bounded
continuous variables, such as latency in this case (Mun
2008; Migliorati et al. 2018).
For our fourth analysis, we tested whether theminimum

distance between the speaker and a resident species dif-
fered among treatments and as a function of the presence
of migrants. We included only trial treatments in which
a single resident species responded, thereby allowing us
to ascertain species identity and include it in the model.
Our final sample included 25 species and 81 trial treat-
ments. Preliminary analyses showed that minimum dis-
tance was skewed to the left and bimodal, which made it
difficult to analyze statistically. We therefore transformed
minimum distance into a dichotomous variable and mod-
eled it using a binomial distribution and logit link. Re-
spondents that approached to a minimum distance from
the speaker of 0–4 m were assigned to one category, and
those that approached to a minimum distance of 4.1–5 m
were assigned to another. We chose a cutoff of 4 because
it separated the two modes of the distribution and split
the sample into groups of similar size (0–4 m, N p 43;
4.1–5 m, N p 38). Treatment, the status of migrants, their
two-way interaction, and country were included as fixed
factors, and species identity and trial number nested within
playback site were included as random effects.
We used the DHARMA package (Hartig 2020) in R to

visualize scaled residuals and assess model fit. Its diagnos-
tic tests and our inspection of residual plots did not identify
any problems associated with the overall distribution of



Innate Responses to Heterospecific Call 000
residuals, over- or underdispersion, frequency of outliers,
or zero inflation, suggesting that all models were adequately
fit. For each model, we used the Anova function in the car
package (Fox and Weisberg 2019) to test the statistical sig-
nificance of each fixed factor. Because we used four separate
models to test the same general hypotheses, we controlled
experimentwise type I error by applying a Bonferroni cor-
rection to alpha. Fixed factors were therefore considered
statistically significant when P ≤ :0125. Where a fixed fac-
tor was significant, we conducted multiple pairwise com-
parisons among its levels using the glht function in themult-
comp package (Hothorn et al. 2008). The P values from the
multiple comparisons of a given factor were adjusted with
the sequential Bonferroni method to control familywise
type I error (Holm 1979).
Finally, we used phylogenetic logistic regression, imple-

mentedwith the binaryPGLMM function (default settings)
in the R package ape (ver. 5.5; Paradis and Schliep 2019),
to test whether phenotypic differences among species pre-
dicted their responses to chick-a-dee calls. For each species,
bodymass (log transformed) and themedian peak correla-
tion between its calls and chick-a-dee calls (square root
transformed) were standardized (mean p 0; SD p 1)
and included as predictor variables. Whether the species
approached to within 5 m of the speaker during the
chick-a-dee call treatment of at least one trial was included
as a binary dependent variable (0p no response; 1p re-
sponse). The phylogeny for the subset of species included
in the analysis was provided by birdtree.org (see details
in fig. 2). To account for phylogenetic uncertainty, we re-
peated the analysis using 10,000 different trees and report
mean coefficients and statistics in the results.
Results

Neotropical species near our playback sites are taxonomi-
cally diverse; 692 species from 58 families and 22 orders
have breeding ranges that overlap at least one of our six
playback sites (tables 1, S2; fig. 2). The species are unevenly
distributed among orders, with 422 (61.1%) in Passeriformes,
64 (9.2%) in Apodiformes, and the remaining 206 (29.7%)
in 20 other orders. Within Passeriformes, the most speciose
families are the tyrant flycatchers (Tyrannidae, N p 93
species), tanagers and allies (Thraupidae, N p 84), oven-
birds and woodcreepers (Furnariidae, N p 42), typical
antbirds (Thamnophilidae, N p 28), and troupials and
allies (Icteridae, N p 27; tables 1, S2; fig. 2).
Across the 138 trials, 38 resident species from 14 avian

families and four orders responded to the chick-a-dee call
treatment of at least one trial by approaching to within 5m
of the playback speaker (tables 1, S1, S2; fig. 2). Our phylo-
genetic logistic regression showed that whether a species
responded to the chick-a-dee call treatment was not sig-
nificantly associated with the species’ mean body mass
(estimate5SE p 0:1250:25, Z p 0:47, P p :6354) or
the structural similarity between its calls and chick-a-dee calls
(estimate5SE p 0:1050:25, Z p 0:39, P p :6993). A
maximum of three resident species responded during any
given trial treatment, and no resident species responded
in 71% of trial treatments (293 of 414). Among the subset
of 60 trials conducted in the winter, 12 migratory species
from four families also responded. Here, a maximum of
six migratory species responded during a given trial treat-
ment, and no migrants responded in 88% (158 of 180) of
trial treatments (table S1).
Experimental treatment affected the number of resident

species approaching the loudspeaker, with significantlymore
species approaching chick-a-dee calls than silence or fee
bee songs and significantly more species approaching fee
bee songs than silence (table 2; fig. 3a). The number of res-
ident species approaching the loudspeaker did not differ
significantly between the temperate breeding season when
migrants were absent and the winter when migrants were
present (table 2; fig. 3a). Furthermore, the pattern of re-
sponses to experimental treatments was not influenced
by the presence or absence of migrants, as revealed by a
nonsignificant interaction between treatment and migrant
status (table 2; fig. 3a). The number of resident species ap-
proaching the speaker was related to the country in which
the playbacks were conducted, with significantly fewer spe-
cies approaching the speaker in Brazil than in Colombia
or Costa Rica (table 2).
Experimental treatment also affected the probability

that a given resident species would approach to within
5 m of the speaker (table 2; fig. 3b). The 42 resident species
that responded across the three treatments were, on aver-
age, significantly more likely to respond to chick-a-dee
calls than to fee bee songs or silence and significantly more
likely to respond to fee bee songs than to silence (table 2;
fig. 3b). Their probability of responding was not related
to the status of migrants (present/absent), the interaction
between treatment and the status of migrants, or the coun-
try in which the playbacks were conducted (table 2; fig. 3b).
Experimental treatment, the presence of migrants, their

two-way interaction, and country were not significantly as-
sociated with residents’ latency to approach the speaker
(N p 111 trial treatments in which a resident species was
first to approach) or with their minimum distance from
the speaker (N p 81 trial treatments inwhich only a single
resident species responded; table 2).
Discussion

Neotropical resident birds observed in our study exhibited
innate responses to unfamiliar heterospecific chick-a-dee
calls. Results are based on a diverse group of 38 resident

http://birdtree.org


000 The American Naturalist
species representing 14 avian families and four orders
and thus add to the growing list of animals that respond
to heterospecific signals (Magrath et al. 2015). More gener-
ally, our study shows that diverse animal species have a
mechanism with which they can respond immediately to
unfamiliar and potentially salient heterospecific signals.
Such a mechanism has broad implications for heterospe-
cific communication. For example, it could facilitate adap-
tive behavioral responses by juveniles that have yet to learn
the sounds of their environment or by invaders that are
leading range expansions into unfamiliar soundscapes. It
also provides a simple mechanism that can help explain
Table 2: Behavioral responses of resident birds during acoustic playback trials
Response and predictor (pairwise comparison)
 Estimate
 SE
 Z
 df
 x2
 P
No. resident species:

Intercept
 1
 48.0
 !.0001

Treatment
 2
 48.0
 !.0001

Call vs. song
 1.18
 .25
 4.7
 !.0001

Call vs. silence
 2.27
 .40
 5.7
 !.0001

Song vs. silence
 1.10
 .44
 2.5
 .0118
Migrants
 1
 1.8
 .1842

Treatment#migrants
 2
 1.0
 .5967

Country
 2
 13.6
 .0011

Colombia vs. Brazil
 .97
 .37
 2.7
 .0161

Costa Rica vs. Brazil
 1.29
 .35
 3.7
 .0007

Colombia vs. Costa Rica
 2.32
 .23
 1.4
 .1650
Whether a species responded (yes/no):

Intercept
 1
 11.3
 .0008

Treatment
 2
 76.7
 !.0001

Call vs. song
 2.98
 .43
 7.0
 !.0001

Call vs. silence
 4.23
 .52
 8.1
 !.0001

Song vs. silence
 1.25
 .47
 2.6
 .0082
Migrants
 1
 .6
 .4229

Treatment#migrants
 2
 4.2
 .1253

Country
 2
 .0
 .9846
Latency to respond (s):

Intercept
 1
 1.3
 .2564

Treatment
 2
 .1
 .9378

Migrants
 1
 .6
 .4479

Treatment#migrants
 2
 .0
 .9893

Country
 2
 8.3
 .0157
Minimum distance (0–4 or 4.1–5 m):

Intercept
 1
 1.5
 .2165

Treatment
 2
 3.4
 .1844

Migrants
 1
 .0
 .8370

Treatment#migrants
 2
 2.8
 .2466

Country
 2
 2.3
 .3092
Note: Responses include the number of resident species approaching to within 5 m of the loudspeaker, whether a given resident species approached to within 5 m,
the latency of the first resident species to approach towithin 5m, and theirminimum approach distance.We broadcast three treatments per trial, including silence and
the fee bee songs and chick-a-dee calls of black-capped chickadees. Trials were conducted in Costa Rica (N p 30), Colombia (N p 27), and Brazil (N p 21) in the
temperate breeding season, when migrants were absent, and in Costa Rica (N p 60) during winter, when migrants were present. Statistically significant results are
in bold. Separate generalized linear mixed models were used to model number of resident species (Poisson distribution, log link), whether a given resident species
responded (binomial distribution, logit link), latency of the first resident species to respond (beta distribution, logit link, overdispersion parameter p 3:28), andmin-
imum approach distance (binomial distribution, logit link). For each model, analysis of deviance was conducted on the predictor variables using type III Wald x2

tests. Where a predictor was statistically significant (a p 0:0125), we conducted multiple pairwise comparisons among its levels (shown in parentheses); estimates
and standard errors (SEs) of the differences (natural log scale for number of resident species, logit scale for whether a given species responded, latency to respond, and
minimumapproach distance),Z statistics, and adjusted P values (a p 0:05; P values adjusted using the sequential Bonferronimethod) are reported. Random effect for
number of resident species: N p 414 trial treatments across 138 trials nested within sites, trial∶site variance5SD p 0:0650:25. Random effects for whether a given
species responded: N p 411 observations from 42 species during 88 trials nested within sites, species variance ! 0:015! 0:01, trial∶site variance ! 0:015! 0:01.
Random effects for latency to respond: N p 111 trial treatments involving 34 species and 82 trials nested within sites, species variance ! 0:015! 0:01, trial∶site
variance p 0:0450:20. Random effects for minimum approach distance: N p 81 trial treatments involving 25 species and 64 trials nested within sites, species
variance p 0:0150:11, trial∶site variance p 0:2850:53.



Innate Responses to Heterospecific Call 000
how diverse animal species interact in large and complex
multispecies communication networks, since participants
would not necessarily need to learn diverse signals and
the environmental features they predict.
Several cognitivemechanisms potentially explain the in-

nate responses of Neotropical residents to unfamiliar het-
erospecific chick-a-dee calls (Magrath et al. 2015; Dutour
et al. 2017). One possibility is that certain acoustic charac-
teristics stimulated diverse sensory systems by exploiting
deeply rooted sensory biases (Endler and Basolo 1998; Ryan
1998; Rendall et al. 2009). A loud bang, for example, does
not necessarily communicate information but nevertheless
causes startle effects in many animals (Davis 1984). Harsh
sounds, such as the dee notes of chick-a-dee calls, may have
similarly widespread effects on arousal and attraction (Fitch
et al. 2002). Information-based explanations are also possi-
ble. For example, unfamiliar chick-a-dee calls may have
characteristics in common with the vocalizations of some
Neotropical residents due to phylogenetic signal or conver-
gent evolution (Marler 1955; Morton 1977; Jurisevic and
Sanderson 1998; Johnson et al. 2003; Dutour et al. 2017).
If such characteristics encode information about arousal
or salient stimuli, as the harsh and low-frequency charac-
teristics of chick-a-dee calls and other diverse predatormob-
bing calls are known to do (Marler 1955; Morton 1977;
Jurisevic and Sanderson 1998; Johnson et al. 2003; Tem-
pleton et al. 2005; Mahurin and Freeberg 2009; Dutour
et al. 2017), then it may be adaptive for resident species
to respond to them. It is important to note that although
the structure of chick-a-dee calls is highly conserved through-
out the Paridae (Hailman 1989; Lucas and Freeberg 2007),
the respondent species in our study are largely allopatric to
the Paridae (table S1), thus limiting the likelihood that
responses evolved in response to chick-a-dee calls specifi-
cally. Furthermore, we do not know of any resident species
near our study sites that produce signals with character-
istics approximating the harsh dee notes of chick-a-dee
calls, but it is possible that responses to such characteristics
remain conserved in resident species even if the charac-
teristics themselves are no longer produced by local birds
(Endler and Basolo 1998; Ryan 1998; Rendall et al. 2009;
Dutour et al. 2017). Future playback experiments could
use natural variants of chick-a-dee calls to identify candi-
date structural characteristics that elicit heterospecific re-
sponses in Neotropical residents, and additional playbacks
Figure 3: Number of resident species approaching to within 5 m

of loudspeaker (N p 414 trial treatments; A) and probability of
any given resident species approaching to within 5 m of loudspeaker
(N p 411 observations from 42 species during 88 trials; B). Playback
treatments included silence and the fee bee songs and chick-a-dee
calls of black-capped chickadees. Trials were conducted in the breed-
ing season, whenmigrants were absent (open circles), and in the win-
ter, when migrants were present (closed circles). Large circles and
error bars show estimated marginal means and their 95% confidence
intervals for each combination of treatment and season; estimates are
derived from generalized linear mixed models (see table 2 and text
for details) in which the effect of country is averaged among the three
countries in which trials were conducted. Raw data are shown as small
circles in A only (N p 234 trial treatments when migrants were ab-
sent, N p 180 trial treatments when migrants were present) and
are jittered vertically and horizontally to reduce overlap and facilitate
viewing. All values are shown untransformed. Treatments with differ-
ent letters are statistically different from each other.



000 The American Naturalist
involving manipulated chick-a-dee calls or synthetic stim-
uli could then test the causal effect of those traits on receiver
responses. Once salient traits are identified, researchers
could test whether they are present and phylogenetically
conserved among local birds. Variation in the structure
(e.g., bandwidth, duration) and number of dee notes pre-
dicts responses to chick-a-dee calls by some parids and
some sympatric nonparids in the temperate zone (Tem-
pleton et al. 2005; Templeton and Greene 2007; Mahurin
and Freeberg 2009; Soard and Ritchison 2009; Courter
and Ritchison 2010; Randler 2012), so we suggest that fu-
ture research begin by focusing on these characteristics.
The taxonomic distribution of the Neotropical residents

that responded to chick-a-dee calls in our study loosely
mirrors the taxonomic distribution of Neotropical resi-
dents that are sympatric with our study sites. Most respon-
dents were Passeriformes (84.2%) and Apodiformes (10.5%),
which comprisemost Neotropical species at our study sites
(Passeriformes: 61.1%; Apodiformes: 9.2%; tables 1, S2;
fig. 2). Within Apodiformes, only the hummingbirds re-
sponded to chick-a-dee calls (family Trochilidae, N p 4),
whereas in Passeriformes, respondents were distributed
widely among Fringillidae (N p 1), Furnariidae (N p 3),
Mitrospingidae (N p 1), Parulidae (N p 1), Passerellidae
(N p 4), Thamnophilidae (N p 1), Thraupidae (N p 8),
Troglodytidae (N p 4), Turdidae (N p 4), and Tyran-
nidae (N p 4; tables S1, S2; fig. 2). These results suggest
that respondents are taxonomically widespread and that
any sensory biases or shared signal characteristics driving
responses are deeply rooted. Our phylogenetic logistic re-
gression found no relationship between whether a species
responded and either their body mass or the similarity be-
tween their calls and chick-a-dee calls, though the small
sample of responding species (N p 38) and the limited in-
formation available for many Neotropical residents pre-
cluded a more comprehensive analysis.
Our results do not support the hypothesis that residents

were simply following migrants that were familiar with
chick-a-dee calls, since residents’ responses were as strong
in the temperate breeding season when migrants were
absent as they were in the winter whenmigrants were pres-
ent. In fact, we were surprised that so few migrants re-
sponded to our playbacks during winter, since many are
sympatric with the Paridae in summer and have been shown
in previous studies to respond to their chick-a-dee calls (ta-
ble S1; Hurd 1996; Nocera et al. 2008). We know that mi-
grants were present near our study sites because we con-
sistently observed them there. Whether migrants responded
to our playbacks may have been related to whether the
migrants overwinter at our playback sites or whether they
were migrating through (Nocera et al. 2008). Using play-
backs of chick-a-dee calls in Belize during the spring mi-
gration, Nocera et al. (2008) showed that migrants passing
through respond strongly to chick-a-dee calls, whereas mi-
grants overwintering in Belize do not. It is therefore possi-
ble that many of the migrants we saw near our study sites
were overwintering there as opposed tomigrating through,
though a detailed survey of migratory species at the time
and location of our playbacks would be needed to establish
which migratory species were present but not responding
(Nocera et al. 2008). Another possibility is that migrants
were present and did respond to our playbacks but did
not approach towithin the 5-m radius required by our pro-
tocol for them to be considered a respondent. In the only
other study that we know of that broadcast chick-a-dee
calls in the tropics, Nocera et al. (2008) purposefully chose
open playback sites that allowed researchers to observe
birds within a 20-m radius of the playback speaker. Dur-
ing 48 trials in which chick-a-dee calls were broadcast in
that study, 48 individuals from 10 migratory species (out
of 24 confirmed to be present at the time and location of
the study) approached to within 20mof the speaker, though
it is unclear how many of those approached to within 5 m
of the speaker (Nocera et al. 2008). Whatever the reason
for the general lack of response by migrants in our study,
the residents responding to unfamiliar chick-a-dee calls
in our study did not simply follow migrants that were fa-
miliar with the calls.
Our findings also do not support the learning hypoth-

esis, which predicts that resident species should never re-
spond to chick-a-dee calls because they have not had the
experience necessary to associate the calls with salient stim-
uli, such as food or predators (Griffin 2004). Consistent
with the learning hypothesis, Nocera et al. (2008) showed
that resident birds in Belize do not respond to chick-a-
dee calls. However, the authors attribute the lack of re-
sponse to the birds’ breeding status, suggesting that active
breeders are more likely than inactive breeders to respond
to mobbing calls (Nocera et al. 2008). In our study, resi-
dents responded strongly to chick-a-dee calls, though we
do not know whether respondents were actively breeding
at the time of our trials. That residents in our study re-
sponded to unfamiliar chick-a-dee calls does notmean that
associative learning is unimportant in governing hetero-
specific responses, only that it was not necessary among
the residents observed in our study. Indeed, there is evidence
that birds can and do learn to respond to heterospecific
vocalizationswhen they are pairedwith salient stimuli (Vieth
et al. 1980; Griffin 2004; Wheatcroft and Price 2013; Ma-
grath et al. 2015). In many avian species, for example, in-
dividuals respond to heterospecific mobbing calls more
strongly in regions where signalers and receivers are sym-
patric versus allopatric (Wheatcroft and Price 2013; Ma-
grath et al. 2015). In other species, naive nestlings do not
respond initially to heterospecific alarm calls, despite re-
sponding to conspecific alarm calls, yet do respond to



Innate Responses to Heterospecific Call 000
heterospecific alarm calls later in life after gaining experi-
ence (Haff andMagrath 2012; Magrath et al. 2015; Dutour
et al. 2019; Carlson et al. 2020). A future study could test
for additive effects of learning by pairing predator stimuli,
such as taxidermic mounts or vocalizations of local preda-
tors, with unfamiliar chick-a-dee calls in one group and
unfamiliar control calls in another group during a series
of learning trials. Afterward, a novel exemplar of a chick-
a-dee call would be broadcast to both groups during a test
trial. If responses to chick-a-dee calls are exclusively innate,
then both groups should respond similarly during the test
trial. If learning also contributes, then birds that experi-
enced chick-a-dee calls paired with predator stimuli during
the learning trials should respond more strongly.
Fewer resident species responded to playbacks conducted

in Brazil than to playbacks conducted in Colombia or Costa
Rica, though the probability of any given species respond-
ing was the same among countries. The intensity of re-
sponse among respondents, as reflected by their latency
to approach and theirminimumdistance from the speaker,
also did not differ among countries. The difference in the
number of species responding cannot be attributed to dif-
ferences in species richness, since our sympatry analysis
indicates that species richness is greater at our two study
locations in Brazil (mean p 298 species) than at our two
study locations in either Colombia (mean p 242 species)
or Costa Rica (mean p 155 species). If structural charac-
teristics of chick-a-dee calls—or the responses of receivers
to those characteristics—are phylogenetically conserved,
then the weaker response in Brazil could be explained by
differences in the phylogenetic relatedness between local
avifauna and the Paridae. Future research could test this
possibility by broadcasting chick-a-dee calls to diverse
tropical species and comparing their phylogenetic related-
ness with the Paridae to whether they respond to their calls.
Unfortunately, we could not test this relationship because
we could not ascertain which species heard our playbacks
but chose not to respond. A different playback approach
in which species were first confirmed to be present be-
fore commencing the playbacks would be needed to test
this hypothesis.
Species that are sympatric with black-capped chickadees

benefit when responding to chick-a-dee calls by detecting
and locating predators earlier than they would without
the calls (Pettifor 1990; Flasskamp 1994; Hurd 1996; Pavey
and Smyth 1998; Consla andMumme2012; Landsborough
et al. 2020) and by accessing food they might otherwise
fail to discover (Mahurin and Freeberg 2009; Wilson and
Mennill 2011). If responses to heterospecific chick-a-dee
calls are at least partially innate, as our results based on
Neotropical residents suggest, then these foraging and anti-
predator benefits could be further enhanced by allowing
birds to respond adaptively to chick-a-dee calls the first
time they hear them and thus earlier in life than if they
had to learn to associate calls with predators or food through
experience (Griffin 2004; Hollén and Radford 2009; Ma-
grath et al. 2015). Although the resident species in our
study will not normally hear chick-a-dee calls, our results
based on 38 species and 14 avian families suggest that
the observed innate response to chick-a-dee calls is taxo-
nomically widespread. Furthermore, nine of the 14 families
that responded to chick-a-dee calls in our study (table S1;
excludingFurnariidae,Mitrospingidae,Momotidae, Tham-
nophilidae, Thraupidae) include at least some species that
are sympatric with the Paridae throughout much of their
range (Billerman et al. 2020) and that therefore could enjoy
the benefits of responding innately to chick-a-dee calls.
Acknowledgments

We thank Vicerrectoría de Investigación, Universidad de
Costa Rica, for supporting this investigation (project B9-
123). We also thank two anonymous reviewers, associate
editor Tony Williams, data editor Bob Montgomerie, and
coeditor Erol Akçay for providing constructive and in-
sightful feedback that improved the manuscript.
Statement of Authorship

L.S. andD.R.W. conceptualized the experiment, developed
themethods, validated the data, and edited themanuscript.
L.S. collected the data and acquired the funds for the exper-
iment. D.R.W. conducted data visualization and statistical
analysis and wrote the initial draft.
Data and Code Availability

The data and R code to reproduce the figures and analyses
have been deposited in theDryadDigital Repository (https://
doi.org/10.5061/dryad.vdncjsxwq; Sandoval and Wilson
2021).
Literature Cited

Billerman, S.M., B. K. Keeney, P. G. Rodewald, andT. S. Schulenberg.
2020. Birds of the world. Cornell Laboratory of Ornithology,
Ithaca, NY.

BirdLife International andHandbook of the Birds of theWorld. 2020.
Bird species distribution maps of the world. Version 2020.1.

Bivand, R., R. Keitt, and B. Rowlingson. 2020. rgdal: bindings for the
“Geospatial” Data Abstraction Library.

Bivand, R., and C. Rundel. 2020. rgeos: interface to Geometry
Engine–Open Source (‘GEOS’).

Bloomfield, L. L., L. S. Phillmore, R. G. Weisman, and C. B. Sturdy.
2005. Note types and coding in parid vocalizations. III. The
chick-a-dee call of the Carolina chickadee (Poecile carolinensis).
Canadian Journal of Zoology 83:820–833.

https://doi.org/10.5061/dryad.vdncjsxwq
https://doi.org/10.5061/dryad.vdncjsxwq


000 The American Naturalist
Bradbury, J. W., and S. L. Vehrencamp. 2011. Principles of animal
communication. 2nd ed. Sinauer, Sunderland, MA.

Brooks,M. E., K. Kristensen, K. J. van Benthem, A.Magnusson, C.W.
Berg, A. Nielsen, H. J. Skaug, et al. 2017. glmmTMB balances speed
and flexibility among packages for zero-inflated generalized linear
mixed modeling. R Journal 9:378–400.

Carlson, N. V., S. D. Healy, and C. N. Templeton. 2020. Wild fledg-
ling tits do not mob in response to conspecific or heterospecific
mobbing calls. Ibis 162:1024–1032.

Charrier, I., L. L. Bloomfield, and C. I. Sturdy. 2004. Note types and
coding in parid vocalizations. I. The chick-a-dee call of the black-
capped chickadee (Poecile atricapillus). Canadian Journal of Zool-
ogy 82:769–779.

Chesser, R. T., S. M. Billerman, K. J. Burns, C. Cicero, J. L. Dunn,
A. W. Kratter, I. J. Lovette, et al. 2020. Check-list of North Amer-
ican birds (online). American Ornithological Society, Chicago.

Chiszar, D., D. Boyer, R. Lee, J. B. Murphy, and C.W. Radcliffe. 1990.
Caudal luring in the southern death adder, Acanthophis antarc-
ticus. Journal of Herpetology 24:253–260.

Cicero, C., P. Pyle, andM. A. Patten. 2020. Juniper titmouse (Baeolo-
phus ridgwayi). Version 1.0. In P. G. Rodewald, ed. Birds of the
world. Cornell Lab of Ornithology, Ithaca, NY.

Consla, D. J., and R. L. Mumme. 2012. Response of captive raptors
to avian mobbing calls: the roles of mobber size and raptor experi-
ence. Ethology 118:1063–1071.

Courter, J. R., and G. Ritchison. 2010. Alarm calls of tufted titmice
convey information about predator size and threat. Behavioral
Ecology 21:936–942.

Da Cunha, F. C. R., J. C. R. Fontenelle, and M. Griessera. 2017.
Predation risk drives the expression of mobbing across bird spe-
cies. Behavioral Ecology 28:1517–1523.

Davis, M. 1984. The mammalian startle response. Pages 287–351
in R. C. Eaton, ed. Neural mechanisms of startle behavior. Plenum,
New York.

Dawkins, R., and J. R. Krebs. 1978. Animal signals: information or
manipulation? Pages 282–309 in N. B. Davies, ed. Behavioural
ecology: an evolutionary approach. Blackwell, Oxford.

Dixon, K. L. 1949. Behavior of the plain titmouse. Condor 51:110–136.
Dolby, A. S., and T. C. Grubb. 1998. Benefits to satellite members in

mixed-species foraging groups: an experimental analysis. Animal
Behaviour 56:501–509.

Dutour, M., J. P. Léna, A. Dumet, V. Gardette, N. Mondy, and
T. Lengagne. 2019. The role of associative learning process on
the response of fledgling great tits (Parus major) to mobbing calls.
Animal Cognition 22:1095–1103.

Dutour, M., J. P. Léna, and T. Lengagne. 2017. Mobbing calls: a sig-
nal transcending species boundaries. Animal Behaviour 131:3–11.

Dutour,M., T. N. Suzuki, andD.Wheatcroft. 2020. Great tit responses
to the calls of an unfamiliar species suggest conserved perception
of call ordering. Behavioral Ecology and Sociobiology 74:37.

Endler, J. A., and A. L. Basolo. 1998. Sensory ecology, receiver biases
and sexual selection. Trends in Ecology and Evolution 13:415–420.

Fallow, P.M., J. L. Gardner, and R. D.Magrath. 2011. Sound familiar?
acoustic similarity provokes responses to unfamiliar heterospecific
alarm calls. Behavioral Ecology 22:401–410.

Ficken, M. S., R. W. Ficken, and S. R. Witkin. 1978. Vocal repertoire
of the black-capped chickadee. Auk 95:34–48.

Fitch, W. T., J. Neubauer, and H. Herzel. 2002. Calls out of chaos:
the adaptive significance of nonlinear phenomena in mamma-
lian vocal production. Animal Behaviour 63:407–418.
Flasskamp, A. 1994. The adaptive significance of avian mobbing.
V. An experimental test of the “move on” hypothesis. Ethology
96:322–333.

Fox, J., and S.Weisberg. 2019. An R companion to applied regression.
3rd ed. Sage, Thousand Oaks, CA.

Gaddis, P. 1985. Structure and variability in the vocal repertoire of the
mountain chickadee. Wilson Bulletin 97:30–46.

Goodale, E., G. Beauchamp, R. D. Magrath, J. C. Nieh, and G. D.
Ruxton. 2010. Interspecific information transfer influences animal
community structure. Trends in Ecology and Evolution 25:354–
361.

Grava, A., T. Grava, and K. A. Otter. 2012. Differential response to
interspecific and intraspecific signals amongst chickadees. Ethol-
ogy 118:711–720.

Griffin, A. S. 2004. Social learning about predators: a review and pro-
spectus. Learning and Behavior 32:131–140.

Guilford, T., and M. S. Dawkins. 1991. Receiver psychology and the
evolution of animal signals. Animal Behaviour 42:1–14.

Hackett, S. J., R. T. Kimball, S. Reddy, R. C. K. Bowie, E. L. Braun,M. J.
Braun, J. L. Chojnowski, et al. 2008. A phylogenomic study of birds
reveals their evolutionary history. Science 320:1763–1768.

Haff, T. M., and R. D. Magrath. 2012. Learning to listen? nestling re-
sponse to heterospecific alarm calls. Animal Behaviour 84:1401–
1410.

Hailman, J. P. 1989. The organization of major vocalizations in the
paridae. Wilson Bulletin 101:305–343.

Hailman, J. P., and S. Haftorn. 2005. Gray-headed chickadee (Poecile
cinctus). Birds of North America Online. Cornell Lab of Ornithol-
ogy, Ithaca, NY.

Hartig, F. 2020. DHARMa: residual diagnostics for hierarchical
(multi-level/mixed) regression models.

Hetrick, S. A., and K. E. Sieving. 2012. Antipredator calls of tufted
titmice and interspecific transfer of encoded threat information.
Behavioral Ecology 23:83–92.

Hoeschele, M., D. E. Gammon, M. K. Moscicki, and C. B. Sturdy.
2009. Note types and coding in parid vocalizations: the chick-a-
dee call of the chestnut-backed chickadee (Poecile rufuscens). Jour-
nal of the Acoustical Society of America 126:2088.

Hollén, L. I., and A. N. Radford. 2009. The development of alarm call
behaviour in mammals and birds. Animal Behaviour 78:791–800.

Holm, S. 1979. A simple sequentially rejective multiple test proce-
dure. Scandinavian Journal of Statistics 6:65–70.

Hothorn, T., F. Bretz, and P. Westfall. 2008. Simultaneous inference
in general parametric models. Biometrical Journal 50:346–363.

Humphreys, R. K., and G. D. Ruxton. 2020. Avian distraction dis-
plays: a review. Ibis 162:1125–1145.

Hurd, C. R. 1996. Interspecific attraction to the mobbing calls of
black-capped chickadees (Parus atricapillus). Behavioral Ecology
and Sociobiology 38:287–292.

Hurlbert, S. H. 1984. Pseudoreplication and the design of ecological
field experiments. Ecological Monographs 54:187–211.

Jetz, W., G. H. Thomas, J. B. Joy, K. Hartmann, and A. O. Mooers.
2012. The global diversity of birds in space and time. Nature
491:444–448.

Jetz, W., G. H. Thomas, J. B. Joy, D. W. Redding, K. Hartmann, and
A. O. Mooers. 2014. Global distribution and conservation of evolu-
tionary distinctness in birds. Current Biology 24:919–930.

Johnson, F. R., E. J. McNaughton, C. D. Shelley, and D. T. Blumstein.
2003. Mechanisms of heterospecific recognition in avian mobbing
calls. Australian Journal of Zoology 51:577–585.



Innate Responses to Heterospecific Call 000
Jurisevic, M. A., and K. J. Sanderson. 1998. A comparative analysis of
distress call structure in Australian passerine and non-passerine
species: influence of size and phylogeny. Journal of Avian Biology
29:61–71.

Koda, H. 2012. Possible use of heterospecific food-associated calls
of macaques by sika deer for foraging efficiency. Behavioural Pro-
cesses 91:30–34.

Landsborough, B., D. R.Wilson, andD. J. Mennill. 2020. Variation in
chick-a-dee call sequences, not in the fine structure of chick-a-dee
calls, influences mobbing behaviour in mixed-species flocks. Be-
havioral Ecology 31:54–62.

Langham, G. M., T. A. Contreras, and K. E. Sieving. 2006. Why
pishing works: titmouse (Paridae) scolds elicit a generalized re-
sponse in bird communities. Ecoscience 13:485–496.

Lloyd, J. E. 1965. Aggressive mimicry in photuris: firefly femmes
fatales. Science 149:653–654.

Lucas, J. R., and T.M. Freeberg. 2007. “Information” and the chick-a-
dee call: communicating with a complex vocal system. Pages 199–
213 in K. A. Otter, ed. Ecology and behavior of chickadees and tit-
mice: an integrated approach. Oxford University Press, New York.

Magrath, R. D., T. M. Haff, P. M. Fallow, and A. N. Radford. 2015.
Eavesdropping on heterospecific alarm calls: from mechanisms
to consequences. Biological Reviews 90:560–586.

Mahurin, E. J., and T.M. Freeberg. 2009. Chick-a-dee call variation in
Carolina chickadees and recruiting flockmates to food. Behavioral
Ecology 20:111–116.

Marler, P. 1955. Characteristics of some animal calls. Nature 176:6–8.
McLaren, M. 1976. Vocalizations of the boreal chickadee. Auk: Orni-

thological Advances 93:451–463.
Migliorati, S., A.M.Di Brisco, andA.Ongaro. 2018. A new regression

model for bounded responses. Bayesian Analysis 13:845–872.
Mönkkönen, M., and J. T. Forsman. 2002. Heterospecific attraction

among forest birds: a review. Ornithological Science 1:41–51.
Morton, E. S. 1977. On the occurrence and significance of motivation-

structural rules in some bird and mammal sounds. American
Naturalist 111:855–869.

Moscicki,M. K.,M.Hoeschele, andC. B. Sturdy. 2010.Note types and
coding in parid vocalizations: the chick-a-dee call of the Mexican
chickadee Poecile sclateri. Acta Ornithologica 45:147–160.

Mun, J. 2008. Understanding and choosing the right probability
distributions. Pages 899–917 in J. Mun, ed. Advanced analytical
models: over 800models and 300 applications from theBasel II Ac-
cord to Wall Street and beyond. Wiley, Hoboken, NJ.

National Imagery and Mapping Agency. 1997. Department of De-
fense World Geodetic System 1984: its definition and relationships
with local geodetic systems. National Geospatial-Intelligence Agency,
Springfield, VA.

Nocedal, J., and M. S. Ficken. 2020. Bridled titmouse (Baeolophus
wollweberi). In A. F. Poole and F. B. Gill, eds. Birds of the world.
Cornell Lab of Ornithology, Ithaca, NY.

Nocera, J. J., P. D. Taylor, and L. M. Ratcliffe. 2008. Inspection of
mob-calls as sources of predator information: response of migrant
and resident birds in the Neotropics. Behavioral Ecology and So-
ciobiology 62:1769–1777.

Owens, J. L., and T. M. Freeberg. 2007. Variation in chick-a-dee
calls of tufted titmice, Baeolophus bicolor: note type and individ-
ual distinctiveness. Journal of the Acoustical Society of America
122:1216–1226.

Owings, D.H., andE. S.Morton. 1998. Animal vocal communication:
a new approach. 1st ed. Cambridge University Press, Cambridge.
Paradis, E., and K. Schliep. 2019. Ape 5.0: an environment for mod-
ern phylogenetics and evolutionary analyses in R. Bioinformatics
35:526–528.

Pavey, C. R., andA.K. Smyth. 1998. Effects of avianmobbing on roost
use and diet of powerful owls, Ninox strenua. Animal Behaviour
55:313–318.

Pettifor, R. A. 1990. The effects of avianmobbing on a potential pred-
ator, the European kestrel, Falco tinnunculus. Animal Behaviour
39:821–827.

Pietsch, T.W., andD. B. Grobecker. 1978. The compleat angler: aggres-
sive mimicry in an antennariid anglerfish. Science 201:369–370.

Randler, C. 2012. A possible phylogenetically conserved urgency re-
sponse of great tits (Parusmajor) towards allopatric mobbing calls.
Behavioral Ecology and Sociobiology 66:675–681.

R Development Core Team. 2020. R: a language and environment
for statistical computing. R Foundation for Statistical Computing,
Vienna.

Remsen, J. V. J., J. I. Areta, E. Bonaccorso, S. Claramunt, A. Jaramillo,
D. F. Lane, J. F. Pacheco, et al. 2021. A classification of the bird spe-
cies of South America. American Ornithological Society, Chicago.

Rendall, D., M. J. Owren, and M. J. Ryan. 2009. What do animal
signals mean? Animal Behaviour 78:233–240.

Ryan, M. J. 1998. Sexual selection, receiver biases, and the evolution
of sex differences. Science 281:1999–2003.

Sandoval, L., and D. R. Wilson. 2012. Local predation pressure
predicts the strength of mobbing responses in tropical birds. Cur-
rent Zoology 58:781–790.

———. 2021. Data from: Neotropical birds respond innately to un-
familiar acoustic signals. American Naturalist, Dryad Digital Re-
pository, https://doi.org/10.5061/dryad.vdncjsxwq.

Schmidt, K. A., E. Lee, R. S. Ostfeld, and K. Sieving. 2008. Eastern
chipmunks increase their perception of predation risk in response
to titmouse alarm calls. Behavioral Ecology 19:759–763.

Scully, E. N., B. C. Schuldhaus, J. V. Congdon, A. H. Hahn, K. A.
Campbell, D. R. Wilson, and C. B. Sturdy. 2019. ZENK expression
in the auditory pathway of black-capped chickadees (Poecile atri-
capillus) as a function of D note number and duty cycle of chick-
a-dee calls. Behavioural Brain Research 356:490–494.

Shriner, W. M. K. 1999. Antipredator responses to a previously neu-
tral sound by free-living adult golden-mantled ground squirrels,
Spermophilus lateralis (Sciuridae). Ethology 105:747–757.

Smith, S. T. 1972. Communication and other social behavior in
Parus carolinensis. 1st ed. Nuttall Ornithological Club, Cambridge,
MA.

Soard, C. M., and G. Ritchison. 2009. “Chick-a-dee” calls of Carolina
chickadees convey information about degree of threat posed by
avian predators. Animal Behaviour 78:1447–1453.

Soler,M. 2017. Avian brood parasitism: behaviour, ecology, evolution
and coevolution. Sinauer, Sunderland, MA.

Templeton, C. N., and E. Greene. 2007. Nuthatches eavesdrop on
variations inheterospecific chickadeemobbingalarmcalls. Proceed-
ings of the National Academy of Sciences of the USA 104:5479–
5482.

Templeton, C. N., E. Greene, and K. Davis. 2005. Behavior: allometry
of alarm calls: black-capped chickadees encode information about
predator size. Science 308:1934–1937.

Vallin, A., S. Jakobsson, J. Lind, and C. Wiklund. 2005. Prey survival
by predator intimidation: an experimental study of peacock butter-
fly defence against blue tits. Proceedings of the Royal Society B
272:1203–1207.

https://doi.org/10.5061/dryad.vdncjsxwq


000 The American Naturalist
Vieth, W., E. Curio, and U. Ernst. 1980. The adaptive significance of
avian mobbing. III. Cultural transmission of enemy recognition in
blackbirds: cross-species tutoring and properties of learning. Ani-
mal Behaviour 28:1217–1229.

Wheatcroft, D., and T. D. Price. 2013. Learning and signal copying
facilitate communication among bird species. Proceedings of the
Royal Society B 280:20123070.

Wignall, A. E., and P. W. Taylor. 2011. Assassin bug uses aggressive
mimicry to lure spider prey. Proceedings of the Royal Society B
278:1427–1433.
“Archæopteryx exhibits naturally a more primitive condition of structu
like.” Figured: “Archæopteryx macroura.” From “Volant Adaptation in
40:537–566).
Wilson, D. R., and D. J. Mennill. 2011. Duty cycle, not signal struc-
ture, explains conspecific and heterospecific responses to the calls
of black-capped chickadees (Poecile atricapillus). Behavioral Ecol-
ogy 22:784–790.

Zuk, M., and G. R. Kolluru. 1998. Exploitation of sexual signals by
predators and parasitoids. Quarterly Review of Biology 73:415–
43/8.

Associate Editor: Tony D. Williams
Editor: Erol Akçay
re than do modern birds, but even here the features are truly bird-
Vertebrates” by Richard S. Lull (The American Naturalist, 1906,