C. Barrio-Amorós, A. M. Forero-Cano, R. Stuster, A. Batista, M. Canzoneri & E. Arias
34
ANARTIA
Publicación del Museo de Biología de la Universidad del Zulia
ISSN 1315-642X (impresa) / ISSN 2665-0347 (digital)
https://doi.org/10.5281/zenodo.10516165 / Anartia, 37 (diciembre 2023): 34-46
e presence of Scinax altae (Dunn, 1933) in Costa Rica, with
redescription of its advertisement call
Presencia de Scinax altae (Dunn, 1933) en Costa Rica, con redescripción
de su canto de advertencia
César Barrio-Amorós
1,ǂ,
*, Andrés Mauricio Forero-Cano
2,ǂ
, Remco Stuster
3
, Abel Batista
4
,
Mike Canzoneri
5
& Erick Arias
6,7,8
1
CRWild, San José, Costa Rica (www.crwild.com)
2
Proyecto Naturphilosophie-Conocer para Conservar, Universidad Nacional de Colombia.
3
Dendrobatidae Nederland, Spreeuwenstraat 15, Del, e Netherlands.
4
Fundación Los Naturalistas, David, Chiriquí, 426-01459, Panamá.
5
Finca Mono Alegre, La Florida de Golto, Puntarenas, Costa Rica.
6
Departamento de Zoología, Instituto de Biología, UNAM, AP 70-153 Ciudad Universitaria, CP 04510, Ciudad de México, México.
7
Escuela de Biología, Universidad de Costa Rica, San Pedro, 11501-2060 San José, Costa Rica.
8
Museo de Zoología,
Centro de Inestigaciones en Biodiversidad y Ecología Tropical, Universidad de Costa Rica, San José, Costa Rica.
*Corresponding author: email: cbarrioamoros@crwild.com
ǂ
ese authors contributed equally.
(Received: 30-07-2023 / Accepted: 15-12-2023 / On line: 31-12-2023)
ABSTRACT
Scinax altae is a treefrog hitherto known to occur only in Panama. Herein, we report two populations referred as S. altae
by morphology, bioacoustics and genetics in the Southeastern Pacic region of Costa Rica. We present new maps showing
the geographic distribution and hiatus between S. staueri and S. altae. In addition, we aimed to redescribe the advertise-
ment calls of Scinax altae and Scinax staueri, based on new bioacoustic analysis and interpretation techniques, in order to
evaluate their dierential diagnosis.
Keywords: bioacoustics analysis, biogeography, distribution extension, phylogenetic analyses, treefrog.
RESUMEN
Hasta el presente, la rana arborícola Scinax altae se conocía solo de Panamá; sin embargo el estudio de ejemplares de Scinax
provenientes de la vertiente Pacíca en el Sureste de Costa Rica, a los cuales se le examinó su morfología, bioacústica y genética,
resultaron pertenecer inequívocamente a esta especie. Además de señalar la presencia de S. altae para Costa Rica, presentamos
mapas que muestran la distribución e hiato geográco entre S. staueri y S. altae y analizamos los cantos de ambas especies,
basándonos en nuevas técnicas de análisis e interpretación bioacústica, con el n apoyar la diagnosis de ambas especies.
Palabras clave: ampliación de distribución, análisis bioacústico, análisis logenético, biogeografía, rana arborícola.
INTRODUCTION
e treefrog Scinax altae (Dunn, 1933) (Anura: Hy-
lidae) was described from the Canal Zone in Panama as
Hyla altae without comparing it directly to the most simi-
lar species, S. staueri (Cope, 1865). In a dichotomous
key Dunn (1933) mentions Hyla culex Dunn, 1932 (a
synonym of S. staueri according to Taylor, 1952), which
is dierentiated from S. altae only by coloration (H. altae
with four stripes on body and no barring on legs, vs. no
Scinax altae in Costa Rica
35
body stripes and barred legs on Hyla culex). León (1969)
and Duellman (1970) considered Hyla altae a subspecies
of H. staueri restricted to Panama, while the nominal
subspecies was considered distributed from northwestern
Costa Rica to Mexico (Savage 2002). León (1969) distin-
guished the two proposed subspecies of Hyla staueri as
follows: H. s. staueri has irregular stripes and interrupted
paravertebral stripes on the dorsum, two transverse bars on
shanks and an interorbital bar, while H. s. altae has com-
plete dorsolateral and paravertebral stripes, longitudinal
stripe on shanks and absence of interorbital bar. Fouquette
& Delahoussaye (1977) assigned most species of the Hyla
ruber group, including staueri, to the genus Ololygon
(Fitzinger, 1843), based on sperm morphology, skipping
H. altae in the process, which was considered a subspecies
of staueri. Duellman & Wiens (1992) placed Ololygon
in synomymy of Scinax, including S. altae as a synonym
of S. staueri. Subsequently Duellman (2001) returned
S. altae to species status, justifying this because no more
specimens were collected in the hiatus between both spe-
cies’ distributions. Araujo-Vieira et al. (2023) found a 3%
uncorrected pair distance (UPD) among S. staueri from
Teculután, Zacapa, Guatemala, and S. altae from Los Al-
garrobos, Chiriquí, Panama, placing both species in the S.
cruentomma group. León (1969) had already made a basic
description of the mating calls of S. altae and S. staueri.
No more published information is available about S. altae.
Scinax staueri is distributed from southern Tamau-
lipas and ueretaro in Mexico (with its type locality in
Orizaba, Veracruz, Mexico) to NW Costa Rica, in subhu-
mid and dry forests (Frost 2023). In Costa Rica, Scinax
staueri is restricted to the northwestern Pacic lowlands,
with no reports from the central and southern Pacic
slopes; this leaves a gap of approximately 260 km between
the southernmost locality of S. staueri in Costa Rica and
the closest known locality of S. altae in Panama (Savage
2002). Considering that Costa Rica is one of the best ex-
plored countries for amphibians (Savage 2002, Leenders
2016) and the species is abundant where it occurs, we are
condent that the hiatus is not due to a lack of prospec-
tion. It is indeed important to dig deeper into the distinc-
tiveness among S. altae and S. staueri, especially updating
the known distribution of both species, including locali-
ties on the central and southern Pacic slopes of Costa
Rica. In addition, it is necessary to update data regarding
the advertisement calls, using modern recording equip-
ment and tools for analysis.
We recorded two populations of Scinax, morphologi-
cally referred to as S. altae, along the southeastern Pacic
region of Costa Rica, within the hiatus between S. staueri
and S. altae. Specimens UCR24262-63 fall entirely into
the described morphology and pattern for S. altae (Dunn
1933, León 1969, Duellman 1970, Kohler 2011; Fig 1A,
B), presenting the aforementioned specic characteristics.
Herein, we aim to redescribe the advertisement calls of Sci-
nax altae and Scinax staueri, based on new bioacoustic
analysis and interpretation techniques, in order to evaluate
their dierential diagnosis.
MATERIALS AND METHODS
Fieldwork
In 2015, Edgar Chinchilla informed one of us (CBA)
about a strange frog in Buenos Aires, Puntarenas prov-
ince, Costa Rica (9° 9’41.46”N, 83°19’55.33”W; 372 m
asl; Fig. 2). CBA and RS checked the locality and found a
Figure 1. Scinax altae from Costa Rica. A. Buenos Aires, Puntarenas (day coloration). Photo: Remco Stuster. B. La Florida, near Gol-
to, Puntarenas (night coloration). Photo: Mike Canzoneri.
C. Barrio-Amorós, A. M. Forero-Cano, R. Stuster, A. Batista, M. Canzoneri & E. Arias
36
few individuals (UCR 24262-63; Fig. 1A). However, the
recorded calls were lost and needed to be recorded again.
In the meantime, MC found the species at La Florida, La
Mona, near Golto (8°33’6.18”N, 83° 4’46.25”W; 150 m
asl) and recorded two dierent animals (Fig. 1B).
Distribution
Distribution map generated with QGIS 3.22.6 (QGIS
Development Team, 2023), distribution localities for both
species were recovered from literature (Dunn 1934, León
1969, Duellman 1970, Savage 2002), Herpetology De-
partment, Museo de Zoología, Universidad de Costa Rica
(UCR) database and GBIF species occurrence (GBIF.
org 2023); IUCN geographic ranges were obtained from
IUCN Red List shape les (IUCN 2022); and the ecore-
gions follow Dinerstein et al. (2017).
Amplication and sequencing
We extracted total genomic DNA from the preserved
tissue sample using the phenol-chloroform standard
protocol (Sambrook & Russell 2006). We included one
specimen housed in the UCR (UCR24263) collection
that was previously identied as Scinax altae from Bue-
nos Aires, Puntarenas province, Costa Rica (9°9’41.46”N,
83°19’55.33”W; 372 m asl). Partial sequences of the large
subunit ribosomal RNA (16S) and cytochrome oxidase
I (COI) mitochondrial genes were determined for one
specimen of Scinax from Costa Rica. e DNA extrac-
tion, amplication, sequencing, and editing protocols
follow those of Arias et al. (2018). e sequences are
available under GenBank accession numbers MT176435–
MT176437. e lists of vouchers and GenBank accession
numbers used in this study are provided in Appendix 1.
Figure 2. Map showing the known distribution range for Scinax altae, including two new localities reported in this work (red trian-
gles) and its close relative S. staueri. e inset map highlights the distribution in Costa Rica and western Panama. e distribution
range is courtesy of the IUCN Red List (2022). e ecoregions follow Dinerstein et al. (2017).
Scinax altae in Costa Rica
37
Phylogenetic analyses
e sequences obtained here were compared with those
available in GenBank for 14 specimens of closely related
species to Scinax altae following Araujo-Veira et al. (2023)
and including S. boulengeri (Cope, 1887) and S. elaeo-
chrous (Cope, 1875) because of their presence in Costa
Rica. We used sequences of S. acuminatus (Cope, 1862)
to root all trees based on the results of Araujo-Veira et al.
(2023). e protocols for alignments, selection of the best
partition scheme, selection of the best model sequence
evolution, the maximum likelihood analysis, Bayesian
analysis (MrBayes and BEAST), and the computation of
genetic distances follow those of Arias et al. (2018).
Bioacoustics analysis
We recorded the advertisement call of two males for
each location of the individuals referred to as Scinax altae
in Costa Rica. A total of 55 calls were analyzed. e audio
records were obtained with the digital recorders Zoom
H1 (set at 44.1 kHz sampling rate, 16-bit resolution and
WAVE le format) and with the application WavePad
for Android. We compared, bioacoustically, these popu-
lations on the southwestern Pacic region of Costa Rica
with those of individuals identied as S. altae and S.
staueri. We obtained from repositories and sound collec-
tions 97 calls corresponding to three males of S. altae from
three locations in Panama and 138 calls corresponding to
four male individuals of S. staueri from four locations in
Costa Rica, Nicaragua and Mexico (Table1).
e structural, spectral and temporal variables of the
calls were obtained with the soware Raven Pro 1.6.1 (K.
Lisa Yang Center for Conservation Bioacoustics at the
Cornell Lab of Ornithology, 2023). A pre inspection was
performed with dierent FFT (Fast Fourier Transform)
resolution settings (from low to high FFT resolution) to
indicate the presence of false harmonics and their depen-
dence on the variation of this parameter, and thus dier-
entiate them from true harmonics. It was found that the
best resolution to analyze and visualize the spectrograms
was using the following parameters: Hann window at 90%
overlap, window size of 190 samples; 3dB lter bandwidth
of 334 Hz, hop size of 19 samples, Discrete Fourier Trans-
form (DFT) size of 256 samples, and grid spacing at 172
Hz.
Following Köhler et al. (2017), the following acoustic
parameters were measured: 1) Structural variables: Num-
ber of notes per call, Number of pulses per note, Number
of calls per call series. 2) Temporal variables: Call duration
(=note duration) 90% (s), Intercall interval (s), Pulse Du-
ration 90% (ms), Pulse repetition rate (Pulses /s), Call se-
ries duration, Intercall series interval. 3) Spectral variables:
Frequency 5% (Hz), Frequency 95% (Hz), Bandwidth
90% (Hz), Dominant frequency (Hz), Peak Time (s). e
structural and temporal variables were measured from the
oscillogram and the spectral variables from the spectro-
gram and the power spectrum.
According to Köhler et al. (2017), excluding 10% of
the call frequency and duration reduces subjectivity as
boundary limits are inuenced by brightness and contrast
congurations of the spectrogram. erefore, in order to
reduce the identication biases of the spectral limits, the
“threshold method” on the Power Spectrum described in
Ríos-Chelén et al. (2017) was used. Aer delimiting the
basic boundaries, the automatic measurement tools of the
soware were used: Freq 5% (Minimum frequency at 5%
of the energy) and Freq 95% (Maximum frequency at 95%
of the energy), to obtain the frequency bandwidth, where
90% of the energy of the notes is concentrated. e domi-
nant frequency was calculated from the Power Spectrum,
using the “Peak Frequency” function of the soware. In
the same way, in order to reduce the identication biases
of the temporal limits of the sound units, the “amplitude
peak as a threshold method” on the oscillogram, described
in Köhler et al. (2017) was used. Aer delimiting the ba-
Table 1. Call recording localities of the specimens of Scinax altae and Scinax staueri, used for comparison with the popu-
lations in question from southeastern Costa Rica and to redescribe advertisement calls.
Species Locality Calls (n) Sources
Scinax altae
Chiriquí Viejo River, Hydroelectric project Burica, Bugaba, Panama 38 AB478. Present study
Barro Colorado Nature Monument, Soberania National Park, Panama 20 Ibáñez et al. 1999
Chepo, Panamá Province, Panama 39 ML194038 - Macaulay Library
Scinax staueri
NW Costa Rica 27 UCR. Present study
Amak community, MSB, BOSAWAS, Nicaragua 26 Present study
Telchac, Yucatán, Mexico 51 Present study
Mante, Tamaulipas, Mexico 34 ML208996 - Macaulay Library
C. Barrio-Amorós, A. M. Forero-Cano, R. Stuster, A. Batista, M. Canzoneri & E. Arias
38
sic time boundaries, the automatic measurement tools of
the soware were used: Time 5% (initial time at 5% of the
energy) and Time 95% (nal time at 95% of the energy)
to obtain delta time where 90% of the energy of the sound
units concentrates. e time where the highest energy
or amplitude peak occurs within each note (Peak Time)
was calculated from the oscillogram using the “Peak Time
Relative” function of the soware. e Pulse repetition
rate (pulses/s) was calculated by dividing the number of
counted pulses by the pulse period, as suggested by Köhler
et al. (2017).
e terminology for the classication and description
of calls was based on Köhler et al. (2017) and Emmrich et
al. (2020).
e graphic representation of the spectrograms, oscil-
lograms and power spectra were generated using the See-
wave v 2.2.0 package (Sueur et al. 2008, Sueur et al. 2018)
from RStudio version R 4.3.0 (R Core Team 2023), with
the spectro, oscillo and meanspec functions respectively,
using a Hanning window, 190 points of resolution FFT
and an overlap of 90%.
In order to determine the existence of statistically
signicant dierences and to examine which of the two
species (Scinax altae or Scinax staueri) better t the bio-
acoustic characteristics of the Costa Rican populations in
question, a one-way ANOVA test was applied for 13vari-
ables. aer corroborating the normality and homosce-
dasticity (equality of variances) of the data sets. To assess
normality, a Shapiro-Wilk Test was performed and to
check the homoscedasticity of the data, a Levene Test was
performed. Subsequently, a post hoc Tukey HSD test of
pairwise multiple comparisons was performed to examine
the dierences or similarities between the three data sets.
ese tests were performed using the R packages: rstatix v
0.7.2 (Kassambara 2023), stats (R Core Team 2023) and
car v 3.1-2 (Fox & Weisberg 2019). Once the post hoc
test showed to which species the bioacoustic character-
istics of the populations of Costa Rica in question were
more adjusted, the values of the acoustic variables were
added to this species, and with this data, we made a dier-
ential diagnosis of the bioacoustic characteristics of both
species.
Our sampling unit for the dierent statistical analy-
sis was the recorded male, that is, we calculated for each
male the mean value of the dierent acoustic variables and
call characteristics. Aer clarifying to which species the
populations from southeastern Costa Rica were more ap-
propriate and adding their data to it, a new mean ± stan-
dard deviation (SD) was calculated, followed by the range
of variation (minimum-maximum). A t-student test was
performed for two independent samples of the 13 acous-
tic variables examined, in order to examine the dierences
between the advertisement calls of Scinax altae and Sci-
nax staueri, where Values of p ≤ 0.05 were considered
to be statistically signicant. Finally, to visualize the vari-
ability of the acoustic parameters, box-and-whisker plots
performed using the R package ggplot2 v 3.4.3 (Wickham
2016).
RESULTS
New records
COSTA RICA: Provincia de Puntarenas, Cantón de
Buenos Aires, Distrito de Buenos Aires, Buenos Aires town
(9° 9’41.46”N, 83°19’55.33”W; 372 m asl). Two adult
males (UCR24262–63) 2015 (Fig. 1A). A disjunct local-
ity is: Provincia de Puntarenas, Cantón de Golto, Dis-
trito de Golto, La Florida near La Mona (8°33’6.18”N,
83° 4’46.25”W; 150 masl) in 2023 (Fig. 1B).
Distribution
Following Dinerstein et al. (2017), the presence of S.
altae is expanded to the Isthmian Pacic moist forest (11
of 30 localities, including the 2 new Costa Rican ones),
Panamanian dry forest (6/30), Isthmian Atlantic moist
forest (9/30) which includes the type locality and 4/30
in the Southern Mesoamerican Pacic mangroves (Fig.2).
From North to south, Scinax staueri has 185 locality
records, distributed following Dinerstein et al. (2017) as
follows: 4 from Veracruz moist forests, 37 from Petén-Ve-
racruz moist forests, including the type locality, 1 from Ve-
racruz dry forests, 3 from the Sierra de los Tuxtlas, 3 from
Southern Pacic dry forests, 33 from Central American
pine-oak forests, 5 from Chiapas Depression dry forests,
25 from Yucatan moist forests, 24 from Central American
Atlantic moist forests, 25 from Central American dry for-
ests, 10 from Mesoamerican Gulf-Caribbean mangroves,
13 from Isthmian-Atlantic moist forests and 2 from Costa
Rican seasonal moist forests (Fig. 2).
Molecular phylogenetic
e resulting data matrix had a total sequence length
of 1,226 bp, including gaps; 580 bp for 16S and 646 for
COI. ree partitions were identied with the following
substitution models: SYM+I+G for 16S + COI codon
position 2, HKY for codon position 1 of COI, and F81
for COI codon position 3. e phylogenies from RAxML
and MrBayes were almost identical in topology (Fig. 3).
e individual from the south Pacic slope of Costa Rica
is grouped within S. altae specimens from Panama. e
phylogenetic analysis found Scinax boulengeri as sister
clade to other species. Scinax elaechroa was inferred as sis-
Scinax altae in Costa Rica
39
ter clade to the clade formed by Scinax cf. wandae + [S.
altae + S. staueri]. Genetic distances between the S. altae
and S. staueri are 2.97–3.14 % for 16S.
Bioacoustics analysis
e one-way ANOVA test for 13 acoustic variables of
the three data sets showed signicant dierences for cer-
tain bioacoustic variables. When examining the multiple
comparisons using the Tukey HSD tests, it was found that
the characteristics of the calls of the populations of south-
eastern Costa Rica better t the characteristics of the spe-
cies Scinax altae, since no statistical dierences were found
in comparison with that species; while in contrast to Sci-
nax staueri, signicant dierences were found in several
of the acoustic characteristics (Table 2). Once the identity
of the specimens from southeastern Costa Rica was ascer-
tained, we added them to the Scinax altae data, leaving a
total of 152 calls analyzed for this species. e summary of
call measurements is presented in Table 3.
Using the note-centered approach, both advertisement
calls share a single pulsed-type note, that is, each note is
subdivided into acoustic units (pulses) separated from
each other by modulations of amplitude less than 100%
and duration less than 50 milliseconds (ms) (Fig. 4). ese
Figure 3. Bayesian phylogenetic inference shows the relationships of the Scinax species occurring in Costa Rica based on the 16S and
COI mitochondrial DNA gene fragments. Posterior probabilities (multiplied by 100) from BEAST analysis are shown above bran-
ches. Below the branches are shown bootstrap proportions from RAxML analysis. e scale bar refers to the estimated substitutions
per site. e support values of any node within the species are not shown. e asterisks represent support > 99.
C. Barrio-Amorós, A. M. Forero-Cano, R. Stuster, A. Batista, M. Canzoneri & E. Arias
40
Table 2. One-way ANOVA and Tukey post-hoc test for 13 bioacoustic variables in the three data sets: (1) Populations in
question from southeastern Costa Rica, (2) Scinax altae, (3) Scinax staueri. Statistically signicant results (p ≤ 0.05) are
highlighted in bold.
Acoustic measurements
One-way
ANOVA
Tukey HSD
2-1 3-1 3-2
Call (Note) duration 90% (s)
F (2,6) = 0,218
p = 0,81
η
2
= 0,068
0,848 0,811 0,999
Intercall interval
F (2,6) = 0,514
p = 0,622
η
2
= 0,146
0,959 0,835 0,610
Num. Call/ Call series
F (2,6) = 11,07
p = 0,0097
η
2
= 0,787
0,919 0,019 0,018
Call series duration
F (2,6) = 13,85
p = 0,006
η
2
= 0,822
0,803 0,010 0,012
Intercall series interval
F (2,6) = 12,54
p = 0,019
η
2
= 0,862
0,928 0,026 0,025
Bandwidth 90% (Hz)
F (2,6) = 2,695
p = 0,146
η
2
= 0,473
0,248 0,136 0,905
Frequency 5% (Hz)
F (2,6) = 6,86
p = 0,028
η
2
= 0,696
0,991 0,071 0,037
Frequency 95% (Hz)
F (2,6) = 5,782
p = 0,040
η
2
= 0,658
0,222 0,034 0,324
Dominant frequency (Hz)
F (2,6) = 7,315
p = 0,025
η
2
= 0,709
0,769 0,107 0,025
Peak Time (s)
F (2,6) = 2,382
p = 0,173
η
2
= 0,443
0,870 0,435 0,168
Pulse Duration 90% (ms)
F (2,6) = 0,238
p = 0,796
η
2
= 0,073
0,999 0,843 0,836
Num. of pulses/call-note
F (2,6) = 1,737
p = 0,254
η
2
= 0,367
0,313 0,956 0,317
Pulse repetition rate (Pulses /s)
F (2,6) = 0,579
p = 0,589
η
2
= 0,162
0,824 0,564 0,876
Scinax altae in Costa Rica
41
pulsed notes do not present frequency modulations during
the course of the call. e pulses that make up the notes are
distinct, partly fused, without silent intervals. e major-
ity of calls had an ascendant amplitude modulation from
the rst to the third pulse, until reaching a plateau, which
was sustained throughout the call duration, where the rst
pulse always had a lower intensity (Fig. 5).
In high FFT resolution, the spectrogram is automati-
cally structured into frequency bands (similar to towers
of coins) that should not be confused with harmonics.
According to Jackson (1996) and Gerhardt (1998), these
frequency bands are caused by high rates of emission of
acoustic structures (pulses), and thus reect the pulse rate.
Because each coherent entity of sound emission is sepa-
rated by silent intervals longer than the duration of the
units themselves, we do not consider them as a series of
notes within a song, but as calls (consisting of one note)
that are grouped together in call series with silent inter-
vals between the series. Although the note-centered ap-
proach could consider the call series of songs as the coher-
ent sound unit (call), Köhler et al. (2017) point out that it
is important to compare homologous bioacoustic entities
under a taxonomic perspective and apply the same name
to them. In this way, comparative studies of advertisement
calls of species in the genus Scinax (Carvalho et al. 2015,
Bang & Giaretta 2016, Ferrão et al. 2022) coincide in con-
sidering the advertisement calls consisting of a single note,
grouped into call series (Fig. 5).
e advertisement calls of Scinax altae are struc-
tured in call series composed of 2-8 calls per series (
=
6,079±1,363); their duration varied from 2,217-4,736
s (
=3,550±1,018), and they are separated by time in-
Table 3. Summary and results of statistical comparisons (t Student results) for the 13 bioacoustic characteristics measured
in the calls between Scinax altae and Scinax staueri. Values are reported a mean ± standard deviation, (minimum– maxi-
mum). n = Sample size. Statistically signicant results (p ≤ 0.05) are highlighted in bold.
Acoustic measurements
Scinax altae
(n=152)
Scinax staueri
(n=138)
t-student Test
Num. Notes/call 1 1 -
Call duration 90% (s)
0,123 ± 0,007
(0,115- 0,133)
0,127±0,020
(0,101-0,150)
t=-0,377
p= 0,718
Intercall interval (s)
0,487±0,078
(0,428-0,619)
0,425±0,101
(0,335-0,567)
t=1,047
p= 0,330
Num. Call/ Call series
6,079±1,363
(2-8)
24,333±8,165
(17-35)
t= -4,999
p= 0,002
Call series duration (s)
3,550±1,018
(2,217-4,736)
13,651±4,048
(8,763-17,739)
t= -5,457
p= 0,0009
Intercall series interval (s)
3,358±0,642
(2,633-4,205)
6,140±0,429
(5,837-6,443)
t= -5,490
p= 0,003
Frequency 5% (Hz)
1855,739±32,621
(1815,589-1899,229)
1622,684±127,448
(1528,846-1808,789)
t= 3,993
p= 0,005
Frequency 95% (Hz)
4645,477±397,300
(4086,538-4968,297)
4118,188±189,966
(3937,500-4319,712)
t= 2,418
p= 0,046
Bandwidth 90% (Hz)
2789,739±419,009
(2221,154-3136,017)
2495,504±2426,457
(2338,235-2790,865)
t= 1,279
p= 0,242
Dominant frequency (Hz)
2424,093±431,496
(1981,055-2885,449)
3542,758±421,957
(2918,179-3837,696)
t= -3,901
p= 0,006
Peak Time (s)
0,362±0,153
(0,219-0,540)
0,593±0,154
(0,362-0,680)
t= -2,245
p= 0,060
Pulse Duration 90% (ms)
4,358± 0,624
(3,462- 5,100)
4,652±0,543
(3,890- 5,166)
t= -0,742
p= 0,482
Num. of pulses/call-note
21,591±0,885
(20,727-22,800)
20,998±1,217
(19,444-22,353)
t= 0,851
p= 0,423
Pulse repetition rate (Pulses /s)
131,734±5,336
(125,806-136,539
124,589±16,271
(105,676-145,200)
t= 0,935
p= 0,381
C. Barrio-Amorós, A. M. Forero-Cano, R. Stuster, A. Batista, M. Canzoneri & E. Arias
42
tervals between 2,633-4,205 s ( =3,358±0,642). Call
duration varied from 0,115- 0,133 s (
=0,123 ± 0,007),
with intervals of 0,428-0,619 s (
=0,487±0,078). Calls
had 20,7-22,8 pulses (
=21,591±0,885), emitted at rates
of 125,806-136,539 pulses per second (
=131,734±
5,336). e pulse duration varied from 3,462-5,100 ms
(
=4,358± 0,624). Regarding the spectral characteris-
tics, dominant frequency varied from 1981,055-2885,449
Hz (
=2424,093±431,496); minimum frequency at
5% of the energy (Freq 5%) varied between 1815,589-
1899,229 Hz (
=1855,739±32,621) and maximum fre-
quency at 95% (Freq 95%) between 4086,538-4968,297
Hz (
=4645,477±397,300).
e advertisement calls of Scinax staueri are struc-
tured in call series composed of 17 -35 calls per series ((
=24,333±8,165); their duration varied from 8,763-
17,739 s (
=13,651±4,048), and were separated by time
intervals between 5,837-6,443 s (
=6,140±0,429). Call
duration varied from 0,101-0,150 s (
=0,127±0,020),
with intervals of 0,335-0,567 s (
=0,425±0,101). Calls
had 19,444-22,353 pulses (
=20,998±1,217), emit-
ted at rates of 105,676-145,200 pulses per second
(
=124,589±16,271). e pulse duration varied from
3,890- 5,166 ms (
=4,652±0,543). Regarding the spec-
tral characteristics, dominant frequency varied from
2918,179-3837,696 Hz (
=3542,758±421,957); mini-
mum frequency at 5% of the energy (Freq 5%) varied be-
tween 1528,846-1808,789 Hz (
=1622,684±127,448)
and maximum frequency at 95% (Freq 95%) between
3937,500-4319,712 Hz (
=4118,188±189,966).
Despite the similarity, the acoustic and statistical analy-
ses (t-student test) show that there are signicant dier-
Figure 4. Comparison of the advertisement calls of Scinax altae and Scinax staueri. Spectrogram, oscillogram (A) and mean frequency
spectrum (B) of S. altae. Spectrogram, oscillogram (C) and mean frequency spectrum (B) of S. staueri. Specimen photographs by Abel
Batista and Victor Jiménez Arcos.
Scinax altae in Costa Rica
43
ences between the advertisement calls of these species
for some structural variables (Number of notes per call);
temporal variables (Call series duration, Intercall series in-
terval); and spectral variables (Frequency 5%, Frequency
95%, Dominant frequency). e variability of the acous-
tic parameters can be visualized in box-and-whisker plots
(Fig. 6). e most striking dierence is in Dominant fre-
quency, where S, altae shows 2424,093±431,496 Hz while
S. staueri is much higher (3542,758±421,957 Hz). Like-
wise, it was found that the way in which their calls are or-
ganized is dierent. Series of short calls (6,079±1,363 calls
per series) characterize S. altae, while in S. staueri calls
are organized in series of greater numbers (24,333±8,165
calls per series), showing the importance of examining and
integrating these types of structural characteristics of ad-
vertisement calls, in addition to the traditional temporal
and spectral characteristics.
Previous call analyses of both species (León 1969, Du-
ellman 1970) failed to provide modern statistical data.
Observations on natural history
e Buenos Aires population is in a rapidly reducing
lagoon in the middle of the city. Many species called on
the night of April 30, 2018, aer heavy but short rainfall;
other species calling included Dendropsophus microcepha-
lus (Cope, 1886), D. ebraccatus (Cope, 1874), Scinax el-
aeochrous, Leptodactylus poecilochilus (Cope, 1862). Sci-
nax altae call randomly every several minutes. e call is
striking but dicult to locate among so many other loud
calling species.
In La Florida, only two males were located by MC,
and only once each time, when the photos and recordings
were made. Both individuals were seen around an articial
pond made to attract local amphibians, such as Agalych-
Figure 5. Advertisement calls structure. Oscillogram zooming (through shorter time slices) of Scinax altae and Scinax staueri. Upper
panel: Shown how advertisement calls are organized into call series. Short call series for S. altae, and very long call series for S. staueri.
Upper central panel: An approach to the structure of one of the call series for S. altae and its equivalence in time for a portion of the call
series for S. staueri. Lower central panel: an approach to the structure of the advertisement call, consisting of a single pulsed-type note.
Lower panel: An approach to the structure of the pulses, which are distinct, partly fused, without silent intervals.
C. Barrio-Amorós, A. M. Forero-Cano, R. Stuster, A. Batista, M. Canzoneri & E. Arias
44
nis callidryas, A. spurrelli, Scinax boulengeri, Engystomops
pustulosus, and Leptodactylus savagei, among others. It ap-
pears that the species is a new arrival to the area, as there is
not an apparent established population, but, rather, a few
single males calling and probably dispersing.
DISCUSSION
e phylogenetic evidence suggests that the popula-
tions of Scinax from southwestern Pacic Costa Rica can
be referred to S. altae with high condence. e genetic
distances among S. altae and S. staueri are around the
thresholds of 3% in 16S suggested by Fouquet et al. (2007)
to separate among species or dene candidate species.
However, it is necessary to highlight that the only avail-
able sequence of S. staueri is from Guatemala (Araujo-
Vieira et al. 2023); therefore, it is possible that these rela-
tionships will change with new sequences throughout the
distribution range of this species. It is necessary to include
additional specimens of S. staueri from Northwestern
Costa Rica, to update the distribution range of this species
and further delineate S. altae and S. staueri.
At a bioacoustic level, despite the similarity that it seems
to show, the acoustic and statistical analyses (t-student test)
display that there are signicant dierences between the
advertisement calls of these species, for some structural
variables (number of notes per call); temporal variables
(call series duration,
intercall series interval) and spec-
tral variables (frequency 5%, frequency 95%, dominant
frequency). e variability of the acoustic parameters can
be visualized in box-and-whisker plots (Fig. 6). e most
striking dierence is in Dominant frequency, where S. al-
tae show 2424,093±431,496 Hz while S. staueri is quite
higher (3542,758±421,957 Hz). Furthermore, the way in
which their calls are organized is dierent. Series of short
calls (6,079±1,363 calls per series) characterize S. altae,
while in S. staueri calls are organized in series of greater
numbers (24,333±8,165 calls per series), showing the im-
Figure 6. Box-and-whisker plots for six bioacoustic character measurements of the advertisement calls of Scinax altae and Scinax stau-
eri. that presented statistically signicant results (p ≤ 0.05) from t-student Test. A. Frequency 5%. B. Frequency 95%. C. Dominant
Frequency. D. Number of calls/call series. E. Call series duration. F. Intercall series interval.
Scinax altae in Costa Rica
45
portance of examining and integrating these types of struc-
tural characteristics of advertisement calls, in addition to
the traditional temporal and spectral characteristics.
e record of this acoustically prominent species (Sci-
nax altae) in the Pacic slopes of southern Costa Rica is
highlighted due to the fact that this region has been his-
torically well sampled and explored. is new record joins
recent records of two frogs, Pristimantis taeniatus (Bou-
lenger, 1912) (Arias et al. 2023) and Elachistocleis pearsei
(Ruthven, 1914) (Vargas & Barrio-Amorós 2023) on the
South Pacic versant of Costa Rica. Very little is known
about the natural history and habitat use of these species
in Costa Rica. It is necessary to evaluate whether these
new records correspond to chance nds in unexplored
microhabitats or if it is possible that recent expansion
ranges explain their presence. e distribution ranges of
these three species recently recorded for the Pacic slope
of Costa Rica is limited due to big agricultural expansions
(pineapple and oil palm). However, it seems possible that
these species are dispersed via transportation of fruits or
other agricultural items.
León (1969) and Duellman (1970) described the mat-
ing calls of S. altae and S. staueri as a series of short na-
sals notes “ah-ah-ah-ah”, which present a fundamental fre-
quency of about 106 Hz for S. staueri and around 112
Hz for S. altae.
Currently, the fundamental frequency to describe the
advertisement calls is little used, because it can be rather
dicult to identify which one is the base frequency pro-
duced by the vocal chords, therefore, pinpointing the fun-
damental frequency of secondary importance (Köhler et
al. 2017).
Köhler et al. (2017) and Emmrich et al. (2020) devel-
oped a set of parameters for the description of advertise-
ment calls. According to these authors, in general, the calls
of these two species are better classied as a pulse repeti-
tion sound, since they consist of bursts of sound energy
represented as several acoustic units.
ACKNOWLEDGEMENTS
We thank Edgar Chinchilla for providing the rst no-
tice of the presence of the species in Buenos Aires; Luis
Díaz Gamboa sent calls of Scinax staueri from Yucatan
for comparison, and Gerardo Chaves “Cachí” also sent
us calls recorded by Eberhard Meyer in Costa Rica. Our
gratitude to Mauricio Rivera-Correa, for advice on the
bioacoustic description of the frog calls, as well as to two
anonymous reviewers for their useful comments. is
work is beneted by the corresponding scientic collect-
ing permits (SINAC-SE- GAS-PI-R 59-2015).
REFERENCES
Araujo-Vieira, K., A. C. C. Lourenço, J. V. A. de Lacerda, M.
L. Lyra, B. L. Blotto, S. R. Ron, D. Baldo, M. O. Pereyra, Á.
M. Suárez-Mayorga, D. Baêta, R. B. Ferreira, C. L. Barrio-
Amorós, C. Borteiro, R. A. Brandão, C. A. Brasileiro, M. A.
Donnelly, M. J. M. Dubeux, J. Köhler, F. Kolenc, F. S. F. Leite,
N. M. Maciel, I. Nunes, V. G. D. Orrico, P. L. V. Peloso, T. L.
Pezzuti, S. Reichle, F. J. M. Rojas-Runjaic, H. R. da Silva, M.
J. Sturaro, J. A. Langone, P. C. de A. Garcia, M. T. Rodrigues,
D. R. Frost, W. C. Wheeler, T. Grant, J. P. Pombal, Jr., C.
F. B. Haddad & J. Faivovich. 2023. Treefrog diversity in the
Neotropics: Phylogenetic relationships of Scinaxini (Anura:
Hylidae: Hylinae). South American Journal of Herpetology 27
(SI): 1–143.
Arias, E., G. Chaves & G. Parra-Olea. 2018. A new species of
Craugastor (Anura: Craugastoridae) from the montane rain-
forest of the Cordillera de Talamanca, Costa Rica. Phyllom-
edusa 17(2): 211–232.
Arias, E., C. L. Barrio-Amorós, A. García & G. Chaves. 2023.
Geographic distribution, advertisement call description
and phylogenetic position of Pristimantis taeniatus (Anura:
Craugastoridae). Revista de Herpetologia Latinoamericana
5(4): 1–9.
Bang D. L. & A. A. Giaretta. 2016. Redescription of the adver-
tisement calls of Scinax tigrinus and Scinax maracaya (An-
ura: Hylidae) and an evaluation of their dierential diagno-
sis. Revista Brasileira de Biociências 14(3): 181–186.
Carvalho T. R., B. F. V. Teixeira, W. E. Duellman & A. A. Gi-
aretta. 2015. Scinax cruentommus (Anura: Hylidae) in the
upper Rio Negro drainage, Amazonas state, Brazil, with
the redescription of its advertisement call. Phyllomedusa
14(2):139–146.
Dinerstein, E., D. Olson, A. Joshi, C. Vynne, N. D. Burgess, E.
Wikramanayake, N, Hahn, S. Palminteri, P. Hedao, R. Noss,
M. Hansen, H. Locke, E. C. Ellis, B. Jones, C. V. Barber, R.
Hayes, C. Kormos, V. Martin, E. Crist, W. Sechrest, L. Price,
J. E. M. Baillie, D. Weeden, K. Suckling, C. Davis, N. Sizer, R.
Moore, D. au, T. Birch, P. Potapov, S. Turubanova, A. Tyu-
kavina, N. de Souza, L. Pintea, J. C. Brito, O. A. Llewellyn,
A. G. Miller, A. Patzelt, S. A. Ghazanfar, J. Timberlake, H.
Klōser, Y. Shennan-Farpón, R. Kindt, J-P. Barnekow Lilleso,
P. van Breugel, L. Graudal, M. Voge, K. F. Al-Shammari &
M. Saleem. 2017. An ecoregion-based approach to protect-
ing half the terrestrial realm. BioScience 67(6): 534–545.
Duellman, W. E. 1970. e hylid ogs of Middle America. 2 vol-
umes. Monograph. Museum of Natural History, University
of Kansas: 1–753.
Duellman, W. E. 2001. Hylid ogs of Middle America. Contribu-
tions to Herpetology 18. 2
nd
ed. USA: Society for the Study
of Amphibians and Reptiles, 694 pp.
Duellman, W. E. & J. J. Wiens. 1992. e status of the hylid frog
genus Ololygon and the recognition of Scinax Wagler, 1830.
Occasional Papers of the Museum of Natural History, Univer-
sity of Kansas 151: 1–23.
C. Barrio-Amorós, A. M. Forero-Cano, R. Stuster, A. Batista, M. Canzoneri & E. Arias
46
Dunn, E. R. 1933. A new Hyla from the Panama Canal Zone.
Occasional Papers of the Boston Society of Natural History 8:
61–64.
Emmrich, M., M. Vences, R. Ernst, J. Köhler, M. F. Barej, F.
Glaw, M. Jansen & M.-O. Rödel. 2020. A guild classica-
tion system proposed for anuran advertisement calls. Zoosys-
tematics and Evolution 96(2): 515–525.
Ferrão, M., J. Moravec, A. S. Ferreira, L. Moraes & J. Hanken.
2022. A new snouted treefrog of the genus Scinax (Snura,
Hylidae) from the white-sand forests of Central Amazonia.
Breviora 573: 1–36.
Fouquette, M. J., Jr., & A. J. Delahoussaye. 1977. Sperm mor-
phology in the Hyla rubra group (Amphibia, Anura, Hyli-
dae), and its bearing on generic status. Journal of Herpetology
11: 387–396.
Fouquet, A., A. Gilles, M. Vences, C. Marty, M. Blanc & N.J.
Gemmell. 2007. Underestimation of species richness in Neo-
tropical frogs revealed by mtDNA analyses. Plos One 10 e1 109
Fox J & Weisberg S. 2019. An R companion to applied regression.
3
rd
ed. ousand Oaks, CA: Sage, https://socialsciences.mc-
master.ca/jfox/Books/Companion/
Frost, D. R. 2023. Amphibian Species of the World: an Online
Reference. Version 6.2 (25 September 2023). https://am-
phibiansoheworld.amnh.org/Amphibia/Anura/Hylidae/
Hylinae/Scinax/Scinax-staueri American Museum of Nat-
ural History, New York, USA.
Gerhardt, H. C. 1998. Acoustic signals of animals: recording,
eld measurements, analysis and description. pp. 1–23. In:
Hopp, S. L., M. J. Owren. & C. S. Evans (eds.). Animal acous-
tic communication: Sound analysis and research methods. Ber-
lin: Springer.
Ibáñez, R. D., S. Rand, M. J. Ryan & C. A. Jaramillo. 1999. Voc a-
lizaciones de ranas y sapos del Monumento Natural Barro Co-
lorado, Parque Nacional Soberanía y áreas adyacentes. Costa
Rica: Sony Music Entertainment, audio CD.
Jackson, L. 1996. Sidebands − artefacts or facts?. Bioacoustics 7:
163–164.
K. Lisa Yang Center for Conservation Bioacoustics at the Cor-
nell Lab of Ornithology. 2023. Raven Pro: Interactive Sound
Analysis Soware (Version 1.6.1) [Computer soware].
Ithaca, NY: e Cornell Lab of Ornithology. Available from
https://ravensoundsoware.com/
Kassambara A. 2023. rstatix: Pipe-iendly amework for basic
statistical tests. R package version 0.7.2, https://rpkgs.data-
novia.com/rstatix/
Köhler, G. 2011. Amphibians of Central America. Oenbach,
Germany: Herpeton, 379 pp.
Köhler, J., M. Jansen, A. Rodríguez, P. J. R. Kok, L. F. Toledo,
M. Emmrich, F. Glaw, C. F. B. Haddad, M.-O. Rödel & M.
Vences. 2017. e use of bioacoustics in anuran taxonomy:
theory, terminology, methods and recommendations for best
practice. Zootaxa 4251: 1–124.
Leenders, T. 2016. Amphibians of Costa Rica: A eld guide. Itha-
ca, NY: Cornell University Press, 531 pp.
León, J. R. 1969. e systematics of the frogs of the Hyla rubra
group in Middle America. University of Kansas Publications.
Museum of Natural History 18: 505–545.
R Core Team. 2023. R: A language and enironment for statis-
tical computing. R Foundation for Statistical Computing. Vi-
enna, Austria. www.R-project.org
Ríos-Chelén, A. A., A. N. McDonald, A. Berger, A. C. Perry,
A. H. Krakauer & G. L. Patricelli. 2017. Do birds vocalize
at higher pitch in noise or is it a matter of measurement?.
Behavioral Ecology and Sociobiology 71(1): 10.1007/s00265-
016-2243-7
Sambrook, J. & D.W. Russell. 2006. Purication of nucleic acids
by extraction with phenol: chloroform. Cold Spring Harbor
Protocols, pdb-prot4455.
Savage, J. M. 2002. e amphibians and reptiles of Costa Rica: A
herpetofauna between two continents, between two seas. Chi-
cago, Ill.: e University of Chicago Press, 934 pp.
Sueur, J., T. Aubin & C. Simonis. 2008. Seewave: a free modular
tool for sound analysis and synthesis. Bioacoustics 18: 213–
226.
Sueur J. 2018. Sound analysis and synthesis with R. Berlin:
Springer-Verlag, iv + 637 pp.
Taylor, E. H. 1952. A review of the frogs and toads of Costa
Rica. University of Kansas Science Bulletin 35: 577–942.
Wickham H. 2016. ggplot2: Elegant graphics for data analysis.
New York: Springer-Verlag, https://ggplot2.tidyverse.org.
