A new species of nurse-frog (Aromobatidae, Allobates) from the Madeira River basin with a small geographic range

Abstract

I describe the seventh species of nurse-frog (Allobates) from the Madeira River basin in Brazilian Amazonia. The new species is distinguished from similar congeneric species by its small body size (snout-to-vent length ranging between 14.0-14.7 mm in adult males and between 14.7-14.9 mm in adult females), by the absence of dark brown regular shapes (e.g. hourglass, "X" or polygon-like marks) on the dorsum, by the absence of transverse dark bars on the dorsal surface of the thigh, and by the light gray to white ventral surfaces, light to dark gray only on throat in live male and female specimens. Males have a distinctive advertisement call characterized by the emission of long (7-11 s) trills of short notes (0.04 s in average) with dominant frequency at 5.9-6.3 kHz and emission rate ranging between 6.7-8.7 notes/s. DNA barcode analyses based on a fragment of the 16S rDNA mitochondrial gene provides additional support to the recognition of the new taxon, which is probably distributed on the east riverbank of the Madeira River, in the interfluve between the Aripuanã and Ji-Paraná rivers.

Full text

1
1
2
3
4
5
6
7
8
This is a personal unedited copy of the article
9
10
11
A new species of nurse-frog (Aromobatidae, Allobates) from
12
the Madeira River basin with a small geographic range.
13
14
15
Later published in Zootaxa 4083 (4): 501–525
16
17
18
Please find the original publication at:
19
20
http://www.mapress.com/j/zt/
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35

2
A new species of nurse-frog (Aromobatidae, Allobates) from the Madeira River
36
basin with a small geographic range.
37
38
PEDRO IVO SIMÕES
1,2
39
1
Laboratório de Sistemática de Vertebrados, Faculdade de Biociências, Pontifícia
40
Universidade Católica do Rio Grande do Sul, Av. Ipiranga 6681, Prédio 40, sala 110,
41
90619-900, Porto Alegre, RS, Brazil. E-mail: pedro.simoes@pucrs.br
42
2
Coordenação de Pesquisas em Biodiversidade, Instituto Nacional de Pesquisas da
43
Amazônia, Av. André Araujo 2936, 69011-970, Manaus, Amazonas, Brazil.
44
45
46
Abstract
47
48
I describe the seventh species of nurse-frog (Allobates) from the Madeira River basin in
49
Brazilian Amazonia. The new species is distinguished from similar congeneric species by
50
its small body size (snout-to-vent length ranging between 14.0–14.7 mm in adult males
51
and between 14.7–14.9 mm in adult females), by the absence of dark brown regular
52
shapes (e.g. hourglass, “X” or polygon-like marks) on the dorsum, by the absence of
53
transverse dark bars on the dorsal surface of the thigh, and by the light gray to white
54
ventral surfaces, light to dark gray only on throat in live male and female specimens.
55
Males have a distinctive advertisement call characterized by the emission of long (7–11
56
s) trills of short notes (0.04 s in average) with dominant frequency at 5.9–6.3 kHz and
57
emission rate ranging between 6.7–8.7 notes/s. DNA barcode analyses based on a
58
fragment of the 16S rDNA mitochondrial gene provides additional support to the
59
recognition of the new taxon, which is probably distributed on the east riverbank of the
60
Madeira River, in the interfluve between the Aripuanã and Ji-Paraná rivers.
61
62
Keywords: Advertisement calls, Amazonia, Dendrobatoidea, endemism, mtDNA.
63
64
65
Resumo
66
67
Descrevo a sétima espécie de rã-cuidadora (Allobates) da bacia do rio Madeira, na
68
Amazônia brasileira. A nova espécie se distingue de espécies similares pertencentes ao
69
mesmo gênero por seu tamanho corporal pequeno (comprimento rostro-uróstilo variando
70
entre 14.0–14.7 mm em machos adultos e entre 14.7–14.9 mm em fêmeas adultas), pela
71
ausência de formas regulares marrom-escuras (e.g. em formato de ampulheta, “X” ou
72
polígonos) no dorso, pela ausência de barras transversais escuras na superfície dorsal da
73
coxa e pelas superfícies ventrais cinza-claras a brancas, cinza-claras a cinza-escuras
74
apenas na região da garganta de machos e fêmeas em vida. Os machos têm uma
75
vocalização de anúncio distintiva, caracterizada pela emissão de séries longas (7–11 s) de
76
notas curtas (0.04 s em média), com frequência dominante entre 5.9–6.3 kHz e taxa de
77
emissão variando entre 6.7–8.7 notas/s. Análises de DNA barcode baseadas no gene
78
mitocondrial 16S rDNA apoiam o reconhecimento do novo táxon, cuja distribuição
79
geográfica provavelmente abrange a margem leste do Madeira, no interflúvio entre os
80
rios Aripuanã e Ji-Paraná.
81

3
82
Palavras-chave: Vocalizações de anúncio, Amazônia, Dendrobatoidea, endemismo,
83
mtDNA.
84
85
86
Introduction
87
88
Nurse-frogs (Allobates Zimmermann & Zimmermann 1988) are a diverse group
89
of ground-dwelling diurnal species, which are common inhabitants of the leaf litter of
90
forests across the Amazonian lowlands. Their taxonomy is challenging due to widespread
91
conservatism of external morphology and color patterns among most species, as well as
92
due to intraspecific variation in phenotypic traits traditionally considered as diagnostic
93
characters (Grant & Rodríguez 2001; Grant et al. 2006; Simões et al. 2013a). However,
94
the inclusion in recent species descriptions of different combinations of data such as
95
those based on advertisement calls, color in life, reproductive traits and mitochondrial
96
DNA has allowed more accurate estimates of the species diversity contained within
97
Allobates and more reliable approximations of their geographic distributions (e.g. Kok et
98
al. 2013; Lima et al. 2014; 2015).
99
The Madeira River is the largest southern tributary of the Amazon River and a
100
well known barrier between many terra-firme taxa, limiting both biogeographic
101
provinces and major areas of endemism in the Amazonian lowlands (Ron 2000; Silva et
102
al. 2005; Morrone 2006). These areas are traditionally recognized as Inambari (west of
103
the Madeira) and Rondônia (east of the Madeira). Recent phylogeographic and
104
phylogenetic studies suggest that further structuring of species distributions occurs within
105
these major areas of endemism, often associated with the Madeira´s largest tributaries or
106
with pronounced environmental gradients along continuous interfluves (e.g. Fernandes et
107
al. 2012, Dias-Terceiro et al. 2015).
108
Knowledge about the diversity and distribution of nurse-frogs across forests
109
within the Madeira River drainage has been steadily accumulating, especially in the last
110
two decades. From an initial assessment suggesting the occurrence of two species of
111
nurse-frogs widely distributed in the area (the brightly colored Allobates femoralis and a
112
second unknown species misidentified as Hyloxalus [Colostethus] peruvianus – Heyer
113
1977), six species have been detected and described from more recent fieldwork:
114
Allobates caeruleodactylus (Lima & Caldwell 2001), A. nidicola (Caldwell & Lima
115
2003), A. paleovarzensis (Lima et al. 2010), A. hodli (Simões et al. 2010), A. grillisimilis
116
(Simões et al. 2013b) and A. flaviventris (Melo-Sampaio et al. 2013). Additional species,
117
readily diagnosable by their distinctive advertisement calls and color patterns, occur in
118
forests along the riverbanks of the Madeira on its upper course and await for formal
119
taxonomic description (Dias-Terceiro et al. 2015).
120
Here I describe a new species of small and cryptically colored nurse-frog with
121
known geographic range restricted to the interfluve between rivers Aripuanã and Ji-
122
Paraná, on the east bank of the Madeira River, at its middle course. In addition to the
123
description of morphological traits, I provide data on the species distinctive
124
advertisement calls and mtDNA barcode analyses. Lastly, I discuss the species
125
distribution and conservation status.
126
127

4
128
Methods
129
130
Specimens and data used in the following species description proceed from
131
fieldwork carried out between 02–05 December 2007, 6.5 km south of the city of
132
Manicoré, in small forest remnants along a dirt road locally known as Estrada do Miriti
133
(05°52'05" S, 61°17'13" W).
134
Recording of advertisement calls of 11 males was conducted using a Sony WM-
135
D6C tape recorder (2004, Sony Corr., Japan) and an AKG 568 EB directional
136
microphone (2003, AKG acoustics GMBH, Austria). The microphone was positioned
137
approximately 1 m away from each focal male. All recordings were done in the morning,
138
between 07:10–11:30 h. Air temperature at the time of each recording was measured with
139
a digital thermometer, and ranged from 23.9 to 27.4°C (average 24.9°C). Nine recorded
140
males were captured manually after recording procedures, and their snout-to-vent length
141
ranged from 14.0 to 14.7 mm (average 14.3 mm). Recordings were digitized from tapes
142
at a sample rate of 22050 Hz and 16 bits sample format in Raven 1.2 (Charif et al. 2004).
143
Advertisement calls of the new species are formed by trills of short notes (see
144
Call description section below). For the description of spectral and temporal parameters
145
of calls, I selected the three note trills with best recording quality from the recording
146
length of each male (n = 33 trills). From each note trill, I selected 20 notes uniformly
147
distributed along its length (n = 660 notes). The duration of note trills, note duration and
148
duration of silent intervals between notes were measured from oscillograms. The number
149
of notes in each note trill were counted from sonograms. Peak (i.e. dominant), upper and
150
lower frequencies of notes were measured from power spectrum graphs. Spectral
151
analyses were conducted with frequency resolution of 82 Hz and 2048 points, using
152
Blackman window type. Additionally, I calculated the rate of note emission in each trill
153
(notes/s). Upper and lower frequencies of notes were measured 20 dB below peak
154
frequency intensity, avoiding overlap with background noise. Values of acoustic
155
parameters are presented as the averages among the mean values for the 11 males
156
recorded.
157
The nine males captured after recording and four additional specimens (two males
158
and two females collected opportunistically) were transported to an improvised
159
laboratory in the field. All specimens were anesthetized and killed with topical
160
benzocaine solution (50mg/g), fixed in 10% formalin solution and preserved in 70%
161
ethanol. Description of color in life was based in photographs and field notes.
162
Preserved specimens were examined and measured under a stereo microscope
163
with graduated lenses or using a digital caliper (to the nearest 0.1 mm). Terminology
164
followed Grant et al. (2006). Morphometric measurements and diagnostic characters
165
followed Lima et al. (2007) and Barrio-Amorós & Santos (2009). Morphometric
166
measurements were taken as follows: snout to-vent length (SVL); head length from tip of
167
snout to posterior edge of maxilla articulation (HL); head width at the level of maxilla
168
articulation (HW); snout length from tip of snout to the center of nostril (SL); eye-to-
169
nostril distance from anterior corner of the eye to the center of nostril (EN); internarial
170
distance (IN); eye length from anterior to posterior corner (EL); interorbital distance (IO);
171
maximum diameter of tympanum (TYM); forearm length from proximal edge of palmar
172
tubercle to outer edge of flexed elbow (FAL); upper arm length from anterior corner of
173

5
arm insertion to the outer edge of flexed elbow (UAL); lengths from proximal edge of
174
palmar tubercle to tips of fingers I, II, III, and IV (respectively HAND I, HAND II,
175
HAND III, HAND IV); width of disc on Finger III (WFD); width of Finger III’s third
176
phalanx (WPF); diameter of palmar tubercle (DPT); leg length from the posterior
177
extremity of the urostyle region to the outer edge of flexed knee (LL); tibia length from
178
outer edge of flexed knee to heel (TL); foot length from proximal edge of outer
179
metatarsal tubercle to tip of Toe IV (FL); width of disc on Toe IV (WTD) and width of
180
thenar tubercle (WTT). All voucher specimens are housed at the herpetology section of
181
the zoological collection of Instituto Nacional de Pesquisas da Amazônia, in Manaus,
182
Brazil (accession numbers INPA-H 35397–35409).
183
Before fixation in formalin, samples of muscle tissue of six specimens were
184
collected and preserved in absolute ethanol. I extracted total genomic DNA from tissue
185
samples using a modified cetyltrimethylammonium bromide (CTAB - Doyle & Doyle
186
1987). We used primers 16Sar and 16Sbr (Palumbi, 1996) to amplify a 502 bp fragment
187
of the 16S rRNA mitochondrial gene. Amplification via polymerase chain reaction (PCR)
188
used a mix with final volume of 16 µL, containing 6.7 µL ddH
2
O, 2.0 µL of 25mM
189
MgCl
2
, 1.5 µL of 10 mM dNTPs, 1.5 µL of 10X amplification buffer (75mM Tris HCl,
190
50 mM KCl, 20 mM (NH4)2SO4), 1.5 µL of a 2 µM solution of each primer, 0.3 µL of
191
Taq DNA polimerase 5 U/µL (Biotools, Spain) and 1.0 µL of DNA (~ 30 ng/ µL).
192
Reaction conditions had a pre-heating step at 92°C for 60 s, 35 cycles of denaturation at
193
92°C for 60 s, primer annealing at 50°C for 50 s, and primer extension at 72°C for 90 s,
194
followed by final extension step of ten minutes at 72°C. Sequencing reactions were
195
carried out after PCR product purification with exonuclease and alkaline phosphatase
196
(Fermentas Life Sciences, Canada) and followed ABI BigDye Terminator Cycle
197
Sequencing Kit protocols, as indicated by the manufacturer. The forward primer was used
198
in sequencing reactions, with annealing temperature of 50°C. The resulting single-
199
stranded products were resolved in an ABI 3130xl automatic sequencer. Sequences were
200
checked by eye against their original electropherograms in BioEdit 7.2.5 (Hall 1999,
201
version update in November 2013).
202
I obtained additional sequences from the National Center for Biotechnology
203
Information´s (NCBI) GenBank. Sequence selection was restricted to samples belonging
204
to cryptically colored Allobates collected at localities in Brazil and across cis-Andean
205
northern South America, most of which proceeded from topotypic voucher specimens
206
(Appendix I). This sequence selection strategy is necessary given that some available
207
sequences are improperly assigned to nominal taxa and that the actual geographic range
208
of most Allobates species is unclear. In order to detect terminals potentially conspecific to
209
the new species, I submitted original sequences of the new species to a query in NCBI´s
210
nucleotide database using the Basic Local Alignment Search Tool (BLASTN 2.2.32 –
211
Zhang et al. 2000). This search turned out no similar sequences (highest percent identity
212
> 90%) belonging to unidentified Allobates voucher specimens proceeding from Brazil or
213
from cis-Andean South America.
214
I aligned the sequence set using MUSCLE algorithm (Edgar 2004), implemented
215
in MEGA 6.06 (Tamura et al. 2013). Due to the presence of short sequences among the
216
16S rDNA database available in Genbank, all aligned sequences were pruned to 481 bp
217
for the analyses described below. I used a phylogenetic tree and genetic distances among
218
nominal taxa inferred from the 16S rDNA sequence fragments in order to estimate
219

6
genetic divergence between the new species and specimens of other Allobates.I generated
220
a maximum likelihood tree using MEGA 6.06. A general time reversible model with a
221
gamma distribution of rate variation (GTR+G) was selected as the best DNA evolution
222
model for the alignment by Bayesian Information Criterion (BIC) in MEGA 6.06. Clade
223
support was inferred from 5000 bootstrap replicates. I used the same program to estimate
224
two interspecific genetic distances commonly used in Allobates taxonomy: uncorrected-
225
pairwise and Kimura-2-parameter (K2P; Kimura 1980).
226
227
228
Allobates bacurau sp. nov.
229
Figures 1–9
230
231
Holotype. INPA-H 35398 (original field number APL 12647). An adult male collected
232
by P.I. Simões on 4th of December 2007, at Estrada do Miriti, in the outskirts of the city
233
of Manicoré, on the right bank of the Madeira River, State of Amazonas, Brazil
234
(05°52'05" S, 61°17'13" W). Advertisement calls of holotype recorded at 11:30 h and
235
25.2°C.
236
237
Paratypes. INPA-H 35397, 35399–35409 (original field numbers APL 12644–12646,
238
12648–12656), ten males, two females. Collected by P.I. Simões on 4th–5th of December
239
2007, proceeding from the same locality.
240
241
Etymology. The specific epithet is in reference to the Portuguese word “bacurau”
242
(modified from the original native Tupi word “wakura’wa”), which designates several
243
species of nighthawks (Family Caprimulgidade). Inhabitants of Manicoré have
244
historically entitled themselves “Povo Bacurau” (“The Nighthawk People”). The origin
245
of this habit is disputed, but it possibly relates to the nickname of a notable citizen, Mr.
246
Francisco A. Vieira Nunes (1939–1997), a writer and politician who gained prestige as a
247
social worker and activist for Hansen’s disease awareness. The epithet is applied as a
248
noun in apposition.
249
250
Diagnosis. Allobates bacurau is characterized by: (1) skin texture of dorsum granular,
251
flat granules scattered from the level of tympanum to the urostyle region, more evident
252
from mid to posterior dorsum; (2) paired dorsal digital scutes present; (3) distal tubercle
253
absent on Finger IV; (4) discs on fingers II and III unexpanded, discs of fingers I and IV
254
weakly expanded; (5) dermal lateral fringes and basal webbing absent on fingers; (6)
255
metacarpal ridge absent; (7) Finger III swollen in male specimens, width of swelling
256
variable among males; Finger III not swollen in females; (8) carpal pad absent; (9) male
257
excrescences absent on thumbs; (10) thenar tubercle conspicuous; (11) black gland absent
258
on arm; (12) tarsal keel present, tubercle-like, strongly curved; (13) disc of Toe I weakly
259
expanded; discs of toes II, III and IV moderately expanded; (14) basal webbing present
260
between toes III and IV; (15) metartarsal fold absent; (16) cryptic external coloration;
261
background color of dorsum tan brown, with darker brown spots scattered from the level
262
of tympanum to posterior dorsum; large diamond, “X” or hourglass-shaped marks absent
263
on dorsum; dorsal surface of thigh grayish brown in life, light brown in preserved
264
specimens, with no transverse bars; dorsal surface of arm yellow to dark copper in life,
265

7
cream in specimens preserved for at least five years; pale dorsolateral stripe present,well
266
defined, variable in width among specimens; a dark brown lateral stripe surrounds the
267
whole body, reaching leg-body insertion; pale oblique lateral stripe absent; pale
268
ventrolateral stripe present, partially diffuse, extending from behind the eyes to groin,
269
cream in preserved specimens, iridescent white in live individuals; small, brown irregular
270
blotches or marbling present ventrolaterally from snout to groin; pale paracloacal mark
271
present; (17) dark throat-collar absent; (18) color of throat generally cream in preserved
272
specimens, appearing light brown in specimens with higher concentration of
273
melanophores on throat skin; density of melanophores on throat variable among
274
specimens; throat light to dark gray in life; (19) central abdominal region immaculate,
275
white to translucent in live male and female specimens; throat region with a variable
276
number of melanophores in both sexes; chest and posterior third of throat light to dark
277
gray in live males, white to translucent in live females, with parietal peritoneum
278
projecting anteriorly from the level of chest and visible through skin; (20) iris metallic
279
gold with tiny black flecks and a light gold pupil ring; (21) large intestine unpigmented;
280
(22) testis unpigmented; (23) mature oocytes pigmented, pigments dark brown; (24)
281
median lingual process absent; (25) tympanum inconspicuous; (26) vocal sac single; (27)
282
maxillary teeth absent or inconspicuous; (27) habit is diurnal, males calling during
283
daytime; (28) advertisement calls characterized by distinctively long trills (7–11 s)
284
formed by approximately 60–80 short notes, emitted at an average rate of seven notes per
285
second; (29) very small snout-to-vent length (SVL), adult males 14.4 mm on average
286
(range 14.0–14.7 mm), adult females 14.8 mm on average (range 14.7–14.9 mm;
287
The new species is assigned to genus Allobates by the combination of characters
288
14, 17, 22, 24, 27 and 29.
289
290
Species comparisons. The new species has a small geographic distribution in central
291
Brazilian Amazonia (see Geographic distribution section below), located at least 600 km
292
from the nearest country border. Thus, we restrict comparisons to Allobates species
293
occurring in Brazil and to three additional species distributed outside the country, but
294
which closely resemble the new species in external morphology.
295
Allobates bacurau is distinguished from A. femoralis (Boulenger 1884), A. myersi
296
(Pyburn 1981) and A. hodli Simões, Lima & Farias 2010 by its smaller body size and by
297
lacking bright aposematic colors on body and dorsal surface of thighs (maximum SVL of
298
adult males = 14.8 mm in A. bacurau, minimum SVL of adult males of remaining species
299
= 22.2 mm, in A. hodli; yellow, orange or bright-red flash marks present on thighs of
300
these large species).
301
Allobates bacurau is distinguished from A. brunneus (Cope 1887), A. crombiei
302
(Morales 2002), A. gasconi (Morales 2002), A. flaviventris Melo-Sampaio, Souza &
303
Peloso 2013, A. goianus (Bokermann 1975) and A. magnussoni Lima, Simões & Kaefer
304
2014 by lacking a wavy-edged or hourglass-shaped dark brown pattern on dorsum.
305
Maximum body size of A. bacurau also smaller than minimum body size registered
306
among specimens of A. crombiei, A. flaviventris and A. magnussoni (maximum SVL of
307
adult males = 14.8 mm in A. bacurau, minimum SVL of males of the latter three species
308
= 16.0 mm, reported for Allobates magnussoni). Most specimens of A. olfersioides (Lutz
309
1925) with X-shaped dark-brown patterns on dorsum (dorsum solid tan-brown with tiny
310

8
darker brown spots in A. bacurau); A. olfersioides distributed along the Brazilian Atlantic
311
forest, not reaching the Amazon basin.
312
Allobates bacurau is distinguished from A. masniger (Morales 2002), A. nidicola
313
(Caldwell and Lima 2003) and A. paleovarzensis Lima, Caldwell, Biavati & Montanarin
314
2010 by its smaller body size (minimum SVL of adult males among the three species =
315
17.9 mm reported for A. masniger; maximum SVL of adult males = 14.8 mm in A.
316
bacurau). Pronounced sexual dimorphism in color of throat and chest of live specimens
317
of A. masniger, A. nidicola and A. paeovarzensis. Throat and chest of Allobates masniger
318
and A. nidicola black to gray in males and uniformly yellow in females, fading to pale in
319
preserved females. Throat of A. paleovarzensis grayish violet in live males, yellowish in
320
live females (throat light gray to dark gray in live male and female A. bacurau). A.
321
bacurau is distinguished from A. fuscellus (Morales 2002) by its smaller body size
322
(minimum SVL in males of A. fuscellus = 15.8 mm) and by uniformly pale to translucent
323
color of ventral surfaces of preserved specimens, darker only on distal parts of limbs and
324
throat (ventral surfaces uniformly dark gray in preserved specimens of A. fuscellus).
325
Minimum SVL of adult male A. vanzolinius (Morales 2002) 21.5 mm, much larger than
326
maximum SVL of A. bacurau males (14.8 mm).
327
Considering smaller species, Allobates bacurau is distinguished from A.
328
caeruleodactylus (Lima & Caldwell 2001) by lacking sky-blue colors on fingers of live
329
specimens and by the presence of defined pale dorsolateral and ventrolateral stripes
330
(fingers sky-blue at least distally, pale dorsolateral and ventrolateral stripes absent in A.
331
caeruleodactylus). A. bacurau is distinguished from A. conspicuus (Morales 2002) by
332
lacking transverse bars on dorsal surface of thigh and by the presence of a partially
333
diffuse pale ventrolateral stripe (thigh with transverse dark brown bars and continuous
334
pale ventrolateral stripe present in A. conspicuus). A. bacurau is distinguished from A.
335
grillisimilis Simões, Sturaro, Peloso & Lima 2013 by lacking transversal dark brown bars
336
on the dorsal surface of thigh (dark brown bars absent or present in variable number on
337
dorsal surface of thigh in A. grillisimilis), by upper arm golden yellow in live specimens
338
(upper arm olive brown in A. grillisimilis) and by throat in variable shades of gray in live
339
male specimens (throat white to translucent, with a few melanophores distributed only on
340
chin in male A. grillisimilis).
341
Allobates bacurau is distinguished from A. marchesianus (Melin 1941) by lacking
342
basal webbing between Toes II and III and by the absence of strong sexual dimorphism in
343
color of ventral surfaces of preserved and live specimens (basal webbing present between
344
Toes II and III in A. marchesianus; throat and chest of males strongly pigmented,
345
appearing dark gray; ventral surfaces of preserved females uniformly cream, bright
346
yellow in life).
347
Allobates bacurau is distinguished from A. subfolionidificans (Lima, Sanchez &
348
Souza 2007) by the presence of melanophores on throat and by the presence of a pale and
349
conspicuous dorsolateral stripe in preserved specimens (throat unpigmented, uniformly
350
cream or white, pale dorsolateral stripe absent in A. subfolionidificans). Preserved
351
specimens of A. bacurau are distinguished from A. tapajos (Lima, Simões & Kaefer
352
2015) by a solid dark brown lateral stripe surrounding the body, by a partially diffuse
353
pale ventrolateral line and pale marbling on the ventrolateral surface of body (dark brown
354
lateral stripe with a diffuse lower edge; pale ventrolateral stripe and pale lateral marbling
355

9
absent in A. tapajos). Live males of A. tapajos also with golden yellow throat and chest
356
(throat and chest gray in A. bacurau).
357
Preserved specimens of Allobates bacurau are almost identical to specimens of
358
Allobates sumtuosus (Morales 2002), but A. sumtuosus lacks dark pigments
359
(melanophores) on throat, which appear uniformly white, light gray or translucent (throat
360
always pigmented A. bacurau, appearing lighter or darker gray depending on
361
melanophore coverage). The two species are also distinguished by their color in life.
362
Throat, chest and abdominal surfaces of female A. sumtuosus uniformly light yellow; the
363
same surfaces are uniformly light gray to translucent in male A. sumtuosus (throat gray,
364
chest and abdomen light gray to translucent in live male and female A. bacurau). Dorsal
365
surface of thigh uniformly bluish gray in live A. sumtuosus, occasionally with a few
366
brown irregular spots (dorsal surface of thigh grayish brown, sometimes with dark brown
367
blotches extending from the lateral surfaces in A. bacurau).
368
Three Allobates species distributed in forested areas of the Guiana Shield and
369
eastern Andean Piedmont of Cordillera de Merida and Cordillera Oriental de Colombia
370
outside Brazil resemble A. bacurau morphologically. A. algorei Barrio-Amorós & Santos
371
2009 is distinguished from A. bacurau by the presence of a dark brown transverse stripe
372
on thigh (absent in A. bacurau), by a dark brown lateral stripe fading posteriorly (dark
373
brown lateral stripe solid around whole body in A bacurau) and by a light yellow
374
posterior abdomen in live specimens (posterior abdomen light gray to translucent in A.
375
bacurau). A. amissibilis Kok, Hölting & Ernst 2013, from Guyana, is distinguished from
376
A. bacurau by the slightly larger size of males (minimum SVL of male A. amissibilis 16.3
377
mm), by an absent or thin and diffuse pale dorsolateral stripe (pale dorsolateral stripe
378
unbroken and well defined in live and preserved A. bacurau), by the presence of a broad
379
oblique lateral stripe extending from groin to mid-body (oblique lateral stripe absent in A.
380
bacurau), and by color of throat in live specimens, pinkish gray in males, cream to
381
yellow in females (throat gray in live male and female specimens of A. bacurau). A.
382
granti (Kok, MacCulloch, Gaucher, Poelman, Bourne, Lathrop & Langlet 2006) from
383
French Guiana differ from A. bacurau in lacking a pale dorsolateral line (present in A.
384
bacurau) and by the ventral color of live females (throat of female A. granti pale, free of
385
melanophores, abdomen yellowish; throat of female A. bacurau gray, with scattered
386
melanophores, abdomen white to translucent).
387
Allobates bacurau can be distinguished from other species of cryptically colored
388
Allobates by means of its advertisement calls, formed by long trills of notes. See “Call
389
description and comparisons” section below for comparisons with advertisement calls of
390
similar congeneric species.
391
392
Description of holotype. Measurements of the holotype are presented in Table 1. Body
393
is robust, head wider than long (HL/HW = 0.84), head length 0.29 times the SVL (Fig1A,
394
B; Fig. 2A). Eye diameter larger than distance from anterior corner of the eye to nostril
395
(EN/EL = 0.63). Nares located posterolaterally to tip of snout, directed laterally, visible
396
in dorsal, ventral, lateral, and anterior views. Distance between nostrils 0.82 times that of
397
HW. Canthus rostralis slightly convex from tip of snout to nostril, straight from nostril to
398
anterior corner of the eye. Loreal region vertical. Tympanum round, 0.37 times the
399
maximum diameter of the eye. Margins of tympanum indistinct to the naked eye, visible
400
under a dissecting microscope, more conspicuous anteroventrally. Maxillary teeth absent,
401

10
not visually detectable under 40X magnification or when a wire probe is moved along the
402
maxillary surface. Tongue conspicuously longer than wide, with anterior tip attached to
403
the mouth floor. Median lingual process absent. Choanae round, positioned
404
anterodorsally to eye bulge. Vocal sac single, corresponding to most of the area of the
405
medial and posterior subgular region. Lateral vocal slits conspicuous.
406
Palmar tubercle elliptical, with the longer width aligned towards the base of
407
Finger II. Thenar tubercle conspicuous, oval, evident in ventral view and in hand profile.
408
Maximum diameter of thenar tubercle 68% of maximum diameter of palmar tubercle
409
(Fig. 3A). Subarticular tubercles of Fingers II, III and IV round, small, never exceeding
410
the width of phalanges. Subarticular tubercle on Finger I elliptical, protuberant, 1.2 times
411
larger than thenar tubercle in maximum diameter. Distal subarticular tubercle absent on
412
Finger IV. Supernumerary tubercles absent. Carpal and metacarpal ridges absent. Fringes
413
or webbings absent on fingers. Length of Finger II equivalent to approximately 85% of
414
Finger I´s length. Tip of Finger IV does not reach distal subarticular tubercle of Finger III
415
when fingers are juxtaposed. Relative lengths of fingers: IV = II < I < III. Finger III
416
swollen; swelling preaxial. Discs of fingers II and III with about the same width of that of
417
underlying phalanges. Discs of fingers I and IV weakly expanded, width of discs
418
corresponding to 1.1 and 1.2 times the width of adjacent phalanx (Fig. 3A).
419
Tibia length corresponding to approximately half the SVL (TL/SVL = 0.52).
420
Tarsal keel present, tubercle-like, strongly curved at its proximal end, flattening and
421
straightening towards metatarsal tubercle (Fig. 4). Metatarsal fold absent. Preaxial edge
422
of tarsus smooth, with no fringe. Basal webbing present between toes III and IV. Less
423
conspicuous basal webbing is present between toes II and III. Relative lengths of toes: I <
424
II < V < III < IV. Disc of Toe I weakly expanded, width 1.2 times the width of adjacent
425
phalanx. Discs of Toes II, III, IV and V moderately expanded, width of discs 1.5, 1.5, 1.7
426
and 1.4 times the width of adjacent phalanx, respectively (Fig. 4).
427
Skin on dorsum is moderately granular, with relatively low granules scattered
428
from the urostyle region to about the level of tympana (granules are evident by
429
aggregations of melanophores on their tips). Skin is smooth ventrally and laterally.
430
Dermal flap absent above cloaca (Fig 1A, B).
431
432
Color in alcohol of holotype. Dorsal surface of body uniformly tan brown, darker only
433
above the orbits and on tips of larger skin granules (Fig 1A). A pale dorsolateral stripe is
434
present, characterized by a narrow line (~ 0.8 mm wide at mid-abdomen level) adjacent
435
to the darker lateral surface of body where density of melanophores is reduced; inner
436
edge of pale dorsolateral stripe well marked, not diffuse (Fig. 1A). Lateral surface of
437
body characterized by a solid dark brown stripe, extending from tip of snout to groin.
438
Pale ventrolateral line present, cream colored, extending from snout to groin, diffuse
439
anteriorly, wider and more conspicuous from arm-insertion level to groin. Irregular
440
projections of ventrolateral line form a marbling pattern on the ventrolateral region
441
towards the abdomen, on light brown background (Fig. 2A). Throat, gular, and anterior
442
pectoral regions light brown, with evenly scattered melanophores. Posterior pectoral
443
region, abdomen and ventral surface of thigh uniformly cream (Fig. 1B).Tongue is
444
cream-colored.
445
Arms cream to pale brown in dorsal view, melanophores concentrated in small
446
and irregular light brown blotches on upper arm, forearm and hand. Tip of fingers light
447

11
brown. Paired scutes on finger discs cream. Upper arm and forearm cream to translucent
448
in ventral view, continuous with color pattern of abdomen. Outer lateral edge of forearm
449
tan brown. Carpal and metarcarpal regions tan brown in ventral view (Fig. 1A, B; Fig.
450
2A).
451
Area immediately around vent solid dark brown, flanked by an unpigmented, C-
452
shaped transverse band, corresponding to the pale paracloacal mark. Thigh light brown in
453
dorsal view, with a few dark brown irregular spots. Inner and outer dorsolateral surfaces
454
of thigh dark brown. Dorsal surface of shank same color as thigh, with irregular dark
455
brown blotches, more frequent on the outer dorsolateral edge. Outer dorsal surface of
456
tarsal region lighter than overall pattern of legs, with scattered dark brown blotches. Toes
457
light brown, with irregularly distributed melanophores. Ventral surface of shank and
458
tarsal region predominantly cream to translucent. Dark brown marbling appears along
459
inner and outer ventrolateral edges and on knee. Ventral surface of metatarsal regions
460
uniformly dark brown. Toes and toe discs light brown in ventral view (Fig. 1A, B; Fig.
461
2A).
462
463
Variation in type-series. Variation in morphometric measurements is presented in Table
464
1. Specimens in type-series resemble the overall external morphology of the holotype
465
(Fig. 1C, D; Fig. 2B; Fig. 5). Females are slightly larger than males in average, but SVL
466
of largest male and smallest female paratypes overlap. Head proportions and relative
467
length of tibia also equivalent between sexes, head slightly wider than long in male and
468
female specimens, and tibia length about half the SVL (average HL/SVL = 0.30 among
469
male paratypes and among female paratypes; average HW/SVL = 0.35 among male
470
paratypes, 0.34 among female paratypes; average TL/SVL = 0.52 in both sexes). Preaxial
471
swelling of Finger III variable among specimens in the type-series: not swollen in the two
472
female (Fig. 3B) and one male paratype. Finger III swollen in the remaining male
473
paratypes. Finger III swelling is strong in preserved males with more evident vocal slits;
474
weak in specimens with less conspicuous vocal slits. Width of pale dorsolateral stripe
475
variable among type specimens (Fig. 5), ranging from 0.59 to 0.95 mm. Pigmentation on
476
throat variable among male and female specimens, general throat color varying from
477
predominantly cream (with melanophores densely scattered only on chin) to uniformly
478
light brown (melanophores densely scattered on most of the throat surface).
479
480
Color in life. Dorsal surface of body tan brown, with scattered darker brown spots on tips
481
of skin granules that are more evident from mid to posterior dorsum. Pale dorsolateral
482
stripe evident, bright tan brown (Figs. 6, 7) Lateral surface of body surrounded by a dark
483
brown stripe from tip of snout to groin. A few faint spots may be present post-laterally on
484
the dark brown stripe about the inguinal region, but never forming a pale oblique lateral
485
line. Iridescent white ventrolateral stripe present along the lower margin of dark brown
486
flanks, from tip of snout to groin, often interrupted, frequently at the level of arm
487
insertion and below the eye. Pale iridescent marbling, same color as of ventrolateral
488
stripe, present below ventrolateral stripe, towards the abdomen, over unpigmented
489
background (Figs. 6, 7). Color of throat can vary from light gray to dark gray in male and
490
female individuals (Fig.7). Color of throat is sexually dimorphic only posteriorly among
491
live specimens, where the parietal peritoneum projects anteriorly past chest level in
492
females and is visible through skin, rendering a white color, instead of the light to dark
493

12
gray posterior throat in males (Fig. 7A). Pectoral region and abdomen white to
494
translucent, with scattered brown pigmentation appearing only ventrolaterally in some
495
specimens (Fig. 7C). Vocal sac of males light gray to gray when inflated.
496
Dorsal surface of upper and forearm predominantly light tan brown, with golden
497
flecks, light gray to translucent only around arm-body insertion. A few dark brown
498
scattered blotches may be present laterally on upper arm and forearm. Upper arm light
499
gray to translucent, same color as chest in ventral view. Forearm, carpal and metacarpal
500
regions gray to dark brown in ventral view. Fingers gray in ventral and dorsal views,
501
darker in ventral view. Paired scutes on finger discs iridescent white.
502
Surfaces immediately adjacent to vent solid dark brown, flanked by a light cream
503
paracloacal mark. Distal edge of paracloacal mark merging shortly with darker brown
504
shades on thigh. Dorsal surface of thigh grayish brown, with irregular dark brown
505
blotches extending from lateral surfaces. Dorsal surface of shank same color as thigh,
506
with a variable number of dark brown blotches. Ventral surface of thigh and shank
507
generally light gray to translucent, darker than abdomen (Fig. 7A), with dark brown
508
marbling along inner and outer edges. Tarsal region light gray in ventral view, brown
509
with dark brown flecks in dorsal view. Toes with gray and dark brown patterning. Paired
510
scutes on toe discs are iridescent white.
511
512
Call description and comparisons. Calls of Allobates bacurau are constituted by the
513
emission of short notes arranged in trills (Fig. 8A). The average duration of note trills
514
among all individuals recorded was 9.72 ± 1.19 (6.92–11.07) seconds. The average rate
515
of note emission was 7.21 ± 0.64 (6.70–8.75) notes/s, and each note trill contained an
516
average of 69 ± 6 (60–81) notes. Average note duration was 0.040 ± 0.006 (0.029–0.045)
517
seconds. Notes have an ascending frequency modulation (Fig. 8B) Average peak
518
frequency was 6.12 ± 0.15 (5.89–6.34) kHz. Lower frequency of notes was around 5.49 ±
519
0.14 (5.26–5.76) kHz on average and average upper frequency reached 6.74 ± 0.18
520
(6.53–7.06) kHz. Average duration of silent intervals between notes was 0.101 ± 0.015
521
(0.071–0.129) seconds. The fundamental frequency was observed between 2.75–3.55
522
kHz, only in sonograms of recordings with very low background noise (including the
523
holotype´s – Fig. 8).
524
Advertisement calls of Allobates algorei, A. caeruleodactylus, A. goianus, A.
525
magnussoni, A. masniger, A. nidicola and A. subfolionidificans are constituted by a single
526
note, emitted continuously between regular (A. goianus, A. magnussoni, A. nidicola, A.
527
masniger) or irregular (A. algorei, A. caeruleodactylus, A. subfolionidificans) inter-note
528
silent intervals, not arranged in discrete note trills (Lima & Caldwell 2001; Caldwell &
529
Lima 2003; Lima et al. 2007; 2014; Barrio-Amorós & Santos 2009; Bastos et al. 2011;
530
Kaefer et al. 2012). Some Allobates species alternate between two types of advertisement
531
calls, emitting notes continuously or arranged in bouts or trills. However, note trills
532
emitted by these species are shorter in length and formed by a smaller number of notes
533
when compared to trills emitted by A. bacurau: Allobates brunneus (1.70–4.20 s, 6–11
534
notes), A. marchesianus (3.39–4.40 s, 21–24 notes), A. sumtuosus (1.95–5.8 s, 15–35
535
notes) and A. tapajos (2.46–3.37 s, 10–14 notes) (Caldwell et al. 2002; Simões et al.
536
2013a; Lima et al. 2009; 2015).
537
Advertisement calls of some Allobates species are constituted exclusively by note
538
trills. However, when compared to those of A. bacurau, note trills are shorter and formed
539

13
by a smaller number of notes in A. amissibilis (0.02–7.97 s, 1–19 notes), A. crombiei
540
(1.91–4.53 s, 25–59 notes), A. grillissimilis (0.12–0.30 s, 3–15 notes), A. paleovarzensis
541
(0.72–3.02 s, 3–21 notes) (Lima et al. 2010;2012; Kok et al. 2013; Simões et al. 2013b).
542
Advertisement calls of A. flaviventris and A. granti are emitted as trills of note-pairs or
543
couplets, not trills of single notes (Kok et al. 2006; Kok & Ernst 2007; Melo-Sampaio et
544
al. 2014; Lima et al. 2014).
545
Considering the overlap of advertisement calls in spectral space, A. goianus, A.
546
magnussoni, A. masniger, A. nidicola and A. paleovarzensis emit calls at a frequency
547
bandwidth lower than 5.3 kHz, below the lower frequency threshold of A. bacurau
548
(Caldwell & Lima 2003; Lima et al. 2010; 2014; Bastos et al. 2011; Kaefer et al. 2012).
549
Importantly, the advertisement calls of A. bacurau can be sonographically
550
distinguished from calls of all cryptically colored Allobates species distributed in the east
551
bank of the middle and lower Madeira River (Fig. 9).
552
553
Genetic distances and evolutionary relationships. Based on a phylogenetic analysis of
554
a 481 bp fragment of the 16S rRNA mitochondrial gene, the six Allobates bacurau
555
paratypes form a well-supported clade, monophyletic in relation to all species of
556
cryptically-colored Allobates occurring in Brazil and in other regions of cis-Andean
557
South America for which similar sequence samples were available (Fig. 10). Sequences
558
proceeding from topotypic specimens of A. sumtuosus form the sister clade to A.
559
bacurau, but their evolutionary relationship is weakly supported (i.e. the sister clade may
560
also contain sequences of A. sumtuosus occurring in Manaus, out of A. sumtuosus type
561
locality). Average uncorrected pairwise and Kimura-2-Parameter (K2P) genetic distances
562
between A. bacurau and A. sumtuosus are 4% (Table 2).
563
Average genetic distances between A. bacurau and the remaining samples
564
analyzed are never lower than 5%. Average uncorrected pairwise genetic distances
565
estimated between A. bacurau and other cryptically colored Allobates distributed on the
566
east bank of the middle and lower Madeira River (A. caeruleodactylus, A. grillisimilis, A.
567
masniger) are no lower than 9% (10% considering K2P genetic distances). Two
568
additional species are distributed on the east bank along the upper river course, A.
569
gasconi and A. flaviventris, but minimum average pairwise genetic distance estimated
570
between A. bacurau and the available specimens sampled is 7% (8% considering K2P
571
distances between A. bacurau and A. gasconi)(Table 2).
572
573
Geographic distribution. Allobates bacurau is known only from the type locality, in the
574
the Municipality of Manicoré, on the right bank of the Madeira River, Amazonas State
575
(Fig. 11). This region is bounded north by the Aripuanã and south by the Ji-Paraná River.
576
Both are large blackwater tributaries of the Madeira River, and have been frequently
577
considered important geographic barriers to vertebrate species ranges and genetic clusters
578
(see Discussion).
579
I sampled one location on the immediately opposite bank of the Madeira River
580
(Village of Democracia, 05°48'21" S, 61°26'43" W – Fig. 11) for diurnal frogs for four
581
days prior to fieldwork in Manicoré, but no A. bacurau specimens were found. Many
582
locations on the eastern bank of the Madeira River, upstream and downstream of
583
Manicoré, have also been systematically sampled for aromobatid frogs in the recent years
584

14
(e.g. Kaefer et al. 2013; Simões et al. 2013b; Dias-Terceiro et al. 2015) and the new
585
species has never been detected in any of them.
586
At the type locality, A. bacurau is syntopic with at least two larger congeneric
587
species: the aposematically colored A. femoralis, and the cryptically colored A. masniger.
588
Three additional Allobates species occur north of the Aripuanã River, in forested sites
589
near the cities of Novo Aripuanã, Borba and Nova Olinda do Norte (Fig. 11): A.
590
caeruleodactylus, A. grillisimilis and a third, undescribed species, which can be easily
591
distinguished from all mentioned species by the bright yellow vocal sac of males (Fig.
592
12).
593
594
Discussion
595
596
Allobates bacurau is the seventh species of nurse-frog described from forests of
597
the Madeira River basin, highlighting a steady trend in species discoveries and hinting at
598
a potentially diverse anuran fauna inhabiting this relatively unexplored Amazonian
599
region. Although the new species can be distinguished from all its congeners by the
600
combination of external morphological characters and color pattern, it was diagnosed in
601
the field for the first time due to its advertisement calls, formed by unusually long note
602
trills. This single acoustic trait (length of note trills) can be used to accurately distinguish
603
A. bacurau from all Allobates with known advertisement calls which could potentially
604
occur syntopically with the new species, reinforcing earlier statements about the
605
advantages of incorporating detailed characterizations of call repertoire (and its variation)
606
in taxonomic descriptions of new nurse-frog species (Peloso 2010; Simões 2013a,b). In
607
this sense, I encourage authors of future descriptions to include complementary species
608
comparisons based on acoustic data, as the one presented in the “Call description and
609
comparisons” section above.
610
Although calling males were locally common, I found no tadpoles in nearby rain
611
puddles or in water accumulated on crevices on the vegetation or on the axillae of leaves,
612
all of which are common sites for Allobates larval development. Jelly nests containing
613
eggs or tadpole clutches were also not found on the leaf litter. The absence of nests or
614
tadpoles next to calling males, as well as the presence of mature oocytes in the oviducts
615
of female voucher specimens, suggest that fieldwork was carried out at the outset of the
616
species reproductive season. Further data about the species reproductive mode, behavior
617
and larval morphology are currently lacking.
618
Regarding the morphological characterization of the new species, I call attention
619
to the existence of intra-sexual variation in morphological traits commonly applied to
620
Allobates taxonomy. Throat color and swelling of Finger III are both variable among
621
reproductive males collected at the same site, most of which were engaged in calling
622
activity. Throat color is also variable among the two female vouchers. Color of throat is
623
sexually dimorphic only posteriorly, the white parietal peritoneum extending anteriorly
624
from chest and visible underneath the skin in living females. In this case, the simple
625
categorization of sexual dichromatism of the throat either as present (throat white
626
posteriorly only in live females) or absent (throat generally light gray to dark gray in both
627
live males and females) is too restrictive, and would mask the actual intraspecific
628
variation. Thus, I opted for describing throat color in the species diagnosis rather than
629
using the more traditional, dichotomic characterization (see Morales 2002).
630

15
In past phylogenetic studies, swelling of Finger III of Allobates males has been
631
coded both as a single, two-state character (absent/present – Morales 2002), or broken
632
into two complementary characters (presence of Finger III swelling and morphology of
633
Finger III swelling, with four possible states according to position and relative width of
634
swelling along finger length – Grant et al. 2006). Among male paratypes of A. bacurau,
635
preaxial Finger III swelling could be coded as weak or strong (states 1 and 2 of character
636
21 of Grant et al. 2006). The finding that there is intraspecific variation in swelling of
637
Finger III in males is not a new one, and has been reported for at least some
638
Dendrobatoidea (Silverstone 1976; Myers et al. 1991; Grant & Rodríguez 2001).
639
Morphology of finger swelling can be also subject to different interpretations by different
640
authors (Simões et al. 2013a). Thus, I recommend that this character is coded based on its
641
presence among a large series (>10 reproductive male specimens), while noting that
642
information is lacking whether variation in Allobates Finger III morphology has an
643
underlying genetic basis or it is under hormonal control, therefore varying seasonally.
644
An additional concern related to the morphological characterization of the new
645
species was the detection of dermal keels or short fringes along the inner lateral surfaces
646
of fingers I, II and III upon a recent inspection of voucher specimens (A.P. Lima,
647
September 2015, personal communication). These were not conspicuous in 2012, when
648
morphological characterization and photographs (Figs. 3, 4) of these specimens were
649
obtained. As dermal expansions on fingers are often produced as artifacts of dissection or
650
preservation, I follow the conservative definition by Grant et al. (2006), according to
651
which finger fringes are characterized as “the skin that stands from the dorsal surface
652
extends ventrad and appears to fold over the side of the digit… In ventral (palmar) view
653
the folding over can be seen to create a deep longitudinal crease or groove”. Therefore, in
654
the absence of dermal folds or grooves, fingers of A. bacurau were considered as lacking
655
dermal fringes.
656
DNA barcode analyses based on fragments of the 16S rRNA gene have been
657
successfully applied in confirming the discovery of new taxa among morphologically
658
conserved genera of South American diurnal frogs (e.g. Lötters et al. 2009; Rojas et al.
659
2014; Lima et al. 2015). Genetic distance estimates between A. bacurau and A.
660
sumtuosus (4%) are only slightly above threshold values suggested as effective in
661
detecting different species of congeneric Neotropical and Malagasy frogs (~3% - Fouquet
662
et al. 2007; Vieites et al. 2009). However, the two species can be unequivocally
663
distinguished by their advertisement calls or color pattern. Importantly, genetic distances
664
among described Allobates species with cryptic coloration distributed on the east bank of
665
the Madeira River are never shorter than 7% when considering the mitochondrial 16S
666
rDNA, rendering this marker readily applicable to the identification of voucher
667
specimens proceeding from that region, including tadpole samples.
668
The area comprised by the Madeira River drainage is recognized for its increased
669
biodiversity and by environmental gradients often associated with the structure of species
670
assemblages across vertebrate taxa, of which the river´s main channel is the most obvious
671
boundary to anuran species distributions (Ron 2000; Dias-Terceiro et al. 2015).
672
Nevertheless, there is accumulating evidence that the eastern tributaries of the Madeira
673
also act as important boundaries between distinct evolutionary lineages of tetrapods such
674
as titi monkeys (Roosmalen et al. 2002), diurnal geckos (Geurgas & Rodrigues 2010) and
675
antbirds (Fernandes et al. 2012). For instance, historical fluctuations in water discharge
676

16
and repositioning of the main channels of the Aripuanã and Ji-Paraná rivers are well
677
documented and both sub-basins seem to constitute extensive megafan regions with
678
headwaters fixed on the cratonic basement and highly variable main channel orientation
679
(Latrubesse 2003; Wilkinson et al. 2010). Such dynamic history of river courses could be
680
partially related to the evolution of vertebrate lineages and to their restricted geographic
681
ranges within the east bank of the Madeira River, rendering the Rondônia area of
682
endemism a mosaic of smaller areas, containing each subsets of exclusive vertebrate taxa.
683
Considering the 2001 IUCN Red List´s classification and criteria for risk
684
assessment, at present A. bacurau should be categorized as Data Deficient, as thorough
685
evaluations of its abundance and distribution have not been carried out. However, the
686
species is known only from its type locality, in the outskirts of the city of Manicoré, in
687
lands not subject to environmental protection. The closest conservation unit is Parque
688
Nacional do Juruena, a national forest preserve. Its northern boundary is 260 km
689
southeast of the species type locality. Additionally, natural environments in the Madeira
690
River basin face many challenges, with potential impacts arising from upstream river
691
damming (causing the breakdown of natural hydrological cycles), roadways and from
692
widespread change in land use, which includes the expansion of deforestation (Fearnside
693
& Graça 2006; Trancoso et al. 2009; Fearnside 2014). Despite its remoteness and
694
difficult access by land, eventual peaks in deforestation activity have been reported for
695
Manicoré, which were probably associated with recently established agricultural
696
settlements (e.g. in October 2012 – Martins et al. 2012). The village of Democracia,
697
across the river from Manicoré, is also an important pathway for timber outflow, as its
698
small harbor can be accessed from BR-319 (a roadway connecting the cities of Manaus
699
and Porto Velho) in the drier months of the year. Due to its small known geographic
700
range and to the evidence of deforestation outbursts in Manicoré, studies should be
701
carried out aiming to evaluate population trends and to search for the species in protected
702
areas in order to assess the species conservation status.
703
704
705
Acknowledgments
706
707
I am grateful to Mr. Francisco Gomes for field assistance in Manicoré and
708
Democracia. Albertina P. Lima, Stefan Lötters, Evan Twomey and one anonymous
709
reviewer provided valuable comments to the manuscript. DNA analyses were conducted
710
at Laboratório de Evolução e Genética Animal of Universidade Federal do Amazonas
711
(LEGAL-UFAM) and I thank Izeni P. Farias for access to lab resources and extended
712
supervision at that time. Conselho Nacional de Desenvolvimento Tecnológico (CNPq)
713
provided funding for field and lab work (CTAmazônia/CT-Energia No. 13/2006;
714
470811/2006-Ed 02/2006 Universal; CNPq/CTAmazônia 575603/2008-9; CNPq
715
401327/2012-4–Prog. Ciência Sem Fronteiras). RAN/ICMBio provided collecting
716
permits (131/04-RAN; 037/2007). I received a doctoral fellowship between 2006–2010
717
and a research fellowship between 2011-2012 (PCI/CNPq-INPA, Proc. 300975/2011-2)
718
from CNPq, while conducting field and lab work relating to this study. I currently receive
719
a post-doctoral fellowship from PNPD-CAPES through Programa de Pós-Graduação em
720
Zoologia of PUCRS.
721
722

17
723
References
724
725
Barrio-Amorós, C.L. & Santos, J.C. (2009) Description of a new Allobates (Anura,
726
Dendrobatidae) from the eastern Andean piedmont, Venezuela. Phyllomedusa, 8, 89–
727
104.
728
Bastos, R.P., Signorelli, L., Morais, A.R., Costa, T.B., Lima, L. & Pombal Jr., J.P. (2011)
729
Advertisement calls of three anuran species (Amphibia) from the Cerrado, Central
730
Brazil. South American Journal of Herpetology, 6, 67–72.
731
Bokermann, W.C.A. (1975) Uma nova espécie de Colostethus do Brasil Central (Anura,
732
Dendrobatidae). Iheringia Série Zoologia, 46, 13–18.
733
Boulenger, G.A. (1884) On a collection of frogs from Yurimaguas, Huallaga River,
734
Northern Peru. Proceedings of the Zoological Society of London, 1883, 635–638.
735
Caldwell, J.P., Lima, A.P. & Keller, C. (2002) Redescription of Colostethus
736
marchesianus (Melin, 1941) from its type locality. Copeia, 2002, 157–165.
737
Caldwell, J.P. & Lima, A.P. (2003) A new Amazonian species of Colostethus (Anura:
738
Dendrobatidae) with a nidicolous tadpole. Herpetologica, 59, 219–234.
739
Charif, R.A., Clark, C.W. & Fristrup, K.M. (2004) Raven 1.2 User’s Manual. Cornell
740
Laboratory of Ornithology, Ithaca, NY.
741
Cope, E.D. (1887) Synopsis of the Batrachia and Reptilia obtained by H. H. Smith, in the
742
Province of Mato Grosso, Brazil. Proceedings of the American Philosophical
743
Society, 24, 44–60.
744
Dias-Terceiro, R.G., Kaefer, I.L., Fraga, R., Araújo, M.C., Simões, P.I. & Lima, A.P.
745
(2015) A matter of scale: Historical and environmental factors structure anuran
746
assemblages from the Upper Madeira River, Amazonia. Biotropica, 47, 259–266.
747
Doyle, J.J. & Doyle, J.L. (1987) A rapid DNA isolation procedure for small quantities of
748
fresh leaf tissue. Phytochemical Bulletin, 19, 11–15.
749
Edgar, R.C. (2004) MUSCLE: multiple sequence alignment with high accuracy and high
750
throughput. Nucleic Acids Research, 32, 1792–1797.
751
Fearnside, P.M. (2014) Impacts of Brazil’s Madeira River dams: Unlearned lessons for
752
hydroelectric development in Amazonia. Environmental Science & Policy, 38, 164–
753
172.
754
Fearnside P.M. & Graça, P.M.L.A. (2006) BR-319: Brazil’s Manaus-Porto Velho
755
highway and the potential impact of linking the arc of deforestation to central
756
Amazonia. Environmental Management 38, 705–716.
757
Fernandes, A.M., Wink, M. & Aleixo, A. (2012) Phylogeography of the chestnut-tailed
758
antbird (Myrmeciza hemimelaena) clarifies the role of rivers in Amazonian
759
biogeography. Journal of Biogeography, 39, 1524–1535.
760
Fouquet, A., Gilles, A., Vences, M., Marty, C., Blanc, M. & Gemmell N.J. (2007)
761
Underestimation of species richness in Neotropical frogs revealed by mtDNA
762
analyses. Plos One 10, 1–10.
763
Geurgas, S.R. & Rodrigues, M.T. (2010) The hidden diversity of Coleodactylus
764
amazonicus (Sphaerodactylinae, Gekkota) revealed by molecular data. Molecular
765
Phylogenetics and Evolution, 54, 583–593.
766

18
Grant, T. & Rodríguez, L.O. (2001) Two New Species of Frogs of the Genus Colostethus
767
(Dendrobatidae) from Peru and a Redescription of C. trilineatus (Boulenger, 1883).
768
American Museum Novitates, 3355, 1–24.
769
Grant, T., Frost, D.R., Caldwell, J.P., Gagliardo, R., Haddad, C.F.B., Kok, P.J.R., Means,
770
B.D., Noonan, B.P., Schargel, W. & Wheeler, W.C. (2006) Phylogenetic systematics
771
of dart-poison frogs and their relatives (Anura: Athesphatanura: Dendrobatidae).
772
Bulletin of the American Museum of Natural History, 299, 1–262.
773
Hall, T.A. (1999) BioEdit: a user-friendly biological sequence alignment editor and
774
analysis program for Windows 65/98/NT. Nucleic Acids Symposium Series, 41, 95–
775
98.
776
Heyer, W.R. (1977) Taxonomic notes on frogs from the Madeira and Purus rivers, Brasil.
777
Papéis Avulsos de Zoologia, 31, 141–162.
778
Kaefer, I.L & Tsuji-Nishikido, B.M. & Lima, A.P. (2012) Beyond the river: underlying
779
determinants of population acoustic signal variability in Amazonian direct-
780
developing Allobates (Anura: Dendrobatoidea). Acta Ethologica, 15, 187–194.
781
Kaefer, I.L., Tsuji-Nishikido, B.M., Mota, E.P., Farias, I.P. & Lima, A.P. (2013) The
782
early stages of speciation in Amazonian forest frogs: Phenotypic conservatism despite
783
strong genetic structure. Evolutionary Biology, 40, 228–245.
784
Kimura, M. (1980) A simple method for estimating evolutionary rate of base
785
substitutions through comparative studies of nucleotide sequences. Journal of
786
Molecular Evolution, 16, 111–120.
787
Kok, P.J.R. & Ernst, R. (2007) A new species of Allobates (Anura: Aromobatidae:
788
Allobatinae) exhibiting a novel reproductive behaviour. Zootaxa, 1555, 21–38.
789
Kok, P.J.R., MacCulloch, R.D., Gaucher, P., Poelman, E.H., Bourne, G. R., Lathrop, A. &
790
Lenglet, G.L. (2006) A new species of Colostethus (Anura, Dendrobatidae) from
791
French Guiana with a redescription of Colostethus beebei (Noble, 1923) from its type
792
locality. Phyllomedusa, 5, 43–66.
793
Kok, P.J.R, Hölting, M. & Ernst, R. (2013) A third microendemic to the Iwokrama
794
Mountains of central Guyana: a new “cryptic” species of Allobates Zimmerman and
795
Zimmerman, 1988 (Anura: Aromobatidae). Organisms, Diversity & Evolution, 13,
796
621–638.
797
Latubresse, E.M. (2003) The Late-Quaternary Palaeohydrology of large South American
798
fluvial systems. Palaeohydrology: Understanding Global Change (ed. by K. J.
799
Gregory and G. Benito), pp. 193–212. John Wiley & Sons Ltd, Hoboken, USA.
800
Lima, A.P. & Caldwell, J.P. (2001) A new Amazonian species of Colostethus with sky
801
blue digits. Herpetologica, 57, 133–138.
802
Lima, A.P., Caldwell, J.P., Biavati, G. & Montanarin, A. (2010) A new species of Allobates
803
(Anura: Aromobatidae) from Paleovárzea Forest in Amazonas, Brazil. Zootaxa, 2337,
804
1–17.
805
Lima, A.P., Sanchez, D.E.A. & Souza, J.R.D. (2007) A new Amazonian species of the
806
frog genus Colostethus (Dendrobatidae) that lays its eggs on undersides of leaves.
807
Copeia, 2007, 114–122.
808
Lima, A.P., Caldwell, J.P. & Strussmann, C. (2009) Redescription of Allobates brunneus
809
(Cope) 1887 (Anura: Aromobatidae: Allobatinae), with a description of the tadpole,
810
call, and reproductive behavior. Zootaxa, 1988, 1–16.
811

19
Lima, A.P., Simões, P.I. & Kaefer, I.L. (2014) A new species of Allobates (Anura:
812
Aromobatidae) from the Tapajós River basin, Pará State, Brazil. Zootaxa, 3889, 355–
813
387.
814
Lima, A.P., Simões, P.I. & Kaefer, I.L. (2015) A new species of Allobates (Anura:
815
Aromobatidae) from Parque Nacional da Amazônia, Pará State, Brazil. Zootaxa, 3980,
816
501–525.
817
Lötters, S., Schmitz, A., Reichle, S., Rödder, D. & Quennet, V. (2009) Another case of
818
cryptic diversity in poison frogs (Dendrobatidae: Ameerega) - description of a new
819
species from Bolivia. Zootaxa, 2028, 20–30.
820
Lutz, A. (1925) Batraciens du Brésil. Comptes Rendus et Mémoires Hebdomadaires des
821
Séances de la Société de Biologie et des ses Filiales. Paris 93, 1925, 137–139.
822
Martins, H., Fonseca, A., Souza Jr., C., Sales, M., & Veríssimo, A. (2012) Boletim
823
Transparência Florestal da Amazônia Legal (outubro de 2012). Belém, Imazon.
824
Available at: http://imazon.org.br/boletim-do-desmatamento-sad-outubro-de-2012/.
825
Melo-Sampaio, P.R., Souza, M.B. & Peloso, P.L.V. (2013) A new, riparian, species of
826
Allobates Zimmermann and Zimmermann, 1988 (Anura: Aromobatidae) from
827
southwestern Amazonia. Zootaxa, 3716, 336–348.
828
Morales, V.C. (2002) Sistemática y biogeografía del grupo trilineatus (Amphibia, Anura,
829
Dendrobatidae, Colostethus), con descripcíon de once espécies nuevas. Publicaciones
830
de La Asociacíon de Amigos de Doñana, 13, 1–59.
831
Morrone, J.J. (2006) Biogeographic areas and transition zones of Latin America and the
832
Caribbean Islands based on panbiogeographic and cladistic analyses of the
833
entomofauna. Annual Review of Entomology, 51, 467–94.
834
Myers, C.W., Paolillo O., A. & Daly, J.W. (1991) Discovery of a defensively malodorous
835
and nocturnal frog in the family Dendrobatidae: phylogenetic significance of a new
836
genus and species from the Venezuelan Andes. American Museum Novitates, 3002,
837
1–33.
838
Palumbi, S.R. (1996) Nucleic acids II: the polymerase chain reaction. In: Hillis, D.M.,
839
Moritz, C. & Mable, B.K. (Eds) Molecular Systematics. Sinauer & Associates Inc.,
840
Sunderland, Massachusetts, pp. 205–247.
841
Peloso, P.L.V. (2010) A safe place for amphibians? A cautionary tale on the taxonomy
842
and conservation of frogs, caecilians and salamanders in the Brazilian Amazonia.
843
Zoologia, 27, 667–673.
844
Pyburn, W.F. (1981) A new poison-dart frog (Anura: Dendrobatidae) from the forest of
845
southeastern Colombia. Proceedings of the Biological Society of
846
Washington, 94, 67–75.
847
Rojas, R.R., Carvalho, V.T., Gordo, M., Ávila, R., Farias, I.P. & Hrbek, T. (2014) A new
848
species of Amazophrynella (Anura:Bufonidae) from the southwestern part of the
849
Brazilian Guiana Shield. Zootaxa, 3753, 79–95.
850
Ron, S. (2000) Biogeographic area relationships of lowland Neotropical rainforest based
851
on raw distributions of vertebrate groups. Biological Journal of the Linnean Society,
852
71, 379–402.
853
Roosmalen, M.G.M. van, Roosmalen, T. van & Mittermeier, R.A. (2002) A taxonomic
854
review of the Titi Monkeys, genus Callicebus, Tomas, 1903, with the description of
855
two new species, Callicebus bernhardi and Callicebus stephennashi, from Brazilian
856
Amazonia. Neotropical Primates, 10 (Suppl.), 1–52.
857

20
Silva, J.M.C., Rylands, A.B. & Fonseca, G.A.B. (2005) The fate of the Amazonian areas
858
of endemism. Conservation Biology, 19, 689–694.
859
Silverstone, P.A. (1976) A revision of the poison-arrow frogs of the genus Phyllobates
860
Bibron in Sagra (family Dendrobatidae). Natural History Museum of Los Angeles
861
County Science Bulletin, 27, 1–53.
862
Simões, P.I., Lima, A.P. & Farias, I.P. (2010) The description of a cryptic species related
863
to the pan-Amazonian frog Allobates femoralis (Boulenger 1883) (Anura:
864
Aromobatidae). Zootaxa, 2406, 1–28.
865
Simões, P.I., Kaefer, I.L., Farias, I.P. & Lima, A.P. (2013a) An integrative appraisal of
866
the diagnosis and distribution of Allobates sumtuosus (Morales, 2002) (Anura,
867
Aromobatidae). Zootaxa, 3746, 401–421.
868
Simões, P.I., Sturaro, M.J., Peloso, P.L.V. & Lima, A.P. (2013b) A new diminutive
869
species of Allobates Zimmermann and Zimmermann, 1988 (Anura, Aromobatidae)
870
from the northwestern Rio Madeira/Rio Tapajós interfluve, Amazonas, Brazil.
871
Zootaxa, 3609, 251–273.
872
Tamura, K., Stecher, G., Peterson, D., Filipski, A. & Kumar, S. (2013) MEGA6:
873
Molecular Evolutionary Genetics Analysis version 6.0. Molecular Biology and
874
Evolution, 30, 2725–2729.
875
Trancoso, R., Carneiro Filho, A., Tomasella, J., Schietti, J., Forsberg, B.R. & Miller, R.P.
876
(2009) Deforestation and conservation in major watersheds of the Brazilian
877
Amazon. Environmental Conservation, 36, 277–288.
878
Vieites, D.R., Wollenberg, K.C., Andreone, F., Köhler, J., Glaw, F. & Vences, M. (2009)
879
Vast underestimation of Madagascar’s biodiversity evidenced by an integrative
880
amphibian inventory. PNAS, 106, 8267–8272.
881
Wilkinson, M.J., Marshall, L.G., Lundberg, J.G. & Kreslavsky, M.H. (2010). Megafan
882
environments in northern South America and their impact on Amazon Neogene
883
aquatic ecosystems. Amazonia, landscape and species evolution (ed. by C. Hoorn, C.
884
and F.P. Wesselingh), pp. 162–184. Blackwell Publishing, Oxford.
885
Zhang, Z. Schwartz, S., Wagner, L. & Miller, W. (2000) A greedy algorithm for aligning
886
DNA sequences. Journal of Computational Biology, 7, 203–214.
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902

21
TABLE 1. Measurements (in mm) of Allobates bacurau holotype and male and female
903
paratypes. Values in type-series columns correspond to mean ± one standard deviation
904
(minimum value recorded–maximum value recorded). See text for a description of
905
morphometric measurements.
906
Measurement
Holotype
Males (n = 10)
Females (n = 2)
SVL
14.1
14.4 ± 0.2 (14.0–14.7)
14.8 ± 0.1 (14.7–14.9)
HL
4.1
4.4 ± 0.1 (4.3–4.5)
4.5 ± 0.0 (4.5–4.5)
HW
4.9
5.0 ± 0.1 (4.9–5.2)
5.0 ± 0.0 (5.0–5.0)
SL
4.0
4.5 ± 0.2 (4.1–4.9)
4.9 ± 0.0 (4.9–4.9)
EN
2.9
2.9 ± 0.2 (2.5–3.1)
3.1 ± 0.1 (3.1–3.2)
IN
4.0
4.3 ± 0.1 (4.1–4.5)
4.5 ± 0.3 (4.3–4.7)
EL
4.6
4.7 ± 0.2 (4.5–5.0)
4.4 ± 0.3 (4.2–4.7)
IO
4.3
4.4 ± 0.1 (4.1–4.6)
4.5 ± 0.1 (4.4–4.6)
TYM
1.7
1.8 ± 0.2 (1.5–2.1)
1.7 ± 0.3 (1.5–2.0)
FAL
3.0
3.4 ± 0.1 (3.2–3.6)
3.5 ± 0.0 (3.5–3.5)
UAL
4.1
4.3 ± 0.2 (3.9–4.5)
4.2 ± 0.3 (4.0–4.5)
HANDI
2.7
2.8 ± 0.2 (2.5–3.1)
2.8 ± 0.3 (2.6–3.0)
HAND II
2.3
2.5 ± 0.1 (2.3–2.7)
2.5 ± 0.0 (2.5–2.5)
HAND III
3.2
3.5 ± 0.2 (3.2–3.7)
3.3 ± 0.2 (3.2–3.5)
HAND IV
2.3
2.4 ± 0.1 (2.3–2.5)
2.2 ± 0.1 (2.1–2.3)
WFD
0.9
0.8 ± 0.1 (0.7–1.0)
0.9 ± 0.1 (0.8–1.0)
WPF
0.9
0.8 ± 0.1 (0.6–0.9)
0.6 ± 0.1 (0.6–0.7)
DPT
0.8
0.7 ± 0.1 (0.6–0.9)
0.7 ± 0.1 (0.6–0.8)
LL
7.1
7.2 ± 0.2 (7.0–7.5)
7.3 ± 0.0 (7.3–7.3)
TL
7.4
7.5 ± 0.2 (7.1–7.7)
7.6 ± 0.3 (7.3–8.0)
FL
6.1
6.4 ± 0.1 (6.1–6.6)
6.4 ± 0.2 (6.3–6.6)
WTD
1.4
1.1 ± 0.1 (1.0–1.3)
1.2 ± 0.1 (1.2–1.3)
WTT
1.0
1.1 ± 0.1 (0.9–1.4)
1.1 ± 0.1 (1.0–1.2)
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926

22
TABLE 2. Average uncorrected pairwise (lower left) and Kimura-2-Parameter (upper right)
927
genetic distances between Allobates bacurau and other cryptically colored taxa of Allobates
928
distributed in Brazil and cis-Andean South America. Distances were based in a 481 bp
929
fragment of the mitochondria 16S rDNA. Refer to Fig. 10 and Appendix I for the number of
930
sequence samples grouped in each taxon for the estimation of average genetic distances.
931
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1. bacurau
0.09
0.10
0.10
0.08
0.09
0.08
0.09
0.10
0.09
0.10
0.09
0.16
0.11
0.05
0.04
0.11
0.09
0.08
0.11
2. algorei
0.08
0.08
0.08
0.09
0.10
0.09
0.09
0.11
0.09
0.10
0.10
0.17
0.10
0.08
0.07
0.09
0.09
0.08
0.14
3. caeruleodactylus
0.09
0.08
0.10
0.09
0.08
0.08
0.10
0.10
0.08
0.10
0.10
0.16
0.10
0.08
0.10
0.10
0.10
0.09
0.13
4. conspicuus
0.09
0.07
0.09
0.10
0.09
0.09
0.09
0.13
0.09
0.10
0.10
0.15
0.10
0.08
0.09
0.02
0.11
0.08
0.14
5. crombiei
0.07
0.08
0.08
0.09
0.08
0.08
0.10
0.10
0.07
0.09
0.08
0.17
0.11
0.07
0.07
0.10
0.10
0.09
0.12
6. flaviventris
0.08
0.09
0.07
0.08
0.07
0.08
0.10
0.10
0.05
0.08
0.07
0.17
0.10
0.07
0.08
0.10
0.09
0.08
0.12
7. gasconi
0.07
0.08
0.08
0.08
0.07
0.07
0.09
0.10
0.08
0.08
0.08
0.17
0.08
0.07
0.07
0.09
0.07
0.05
0.14
8. granti
0.08
0.08
0.09
0.08
0.08
0.09
0.08
0.12
0.08
0.11
0.11
0.15
0.11
0.08
0.08
0.10
0.10
0.09
0.13
9. grillisimilis
0.09
0.09
0.09
0.10
0.08
0.08
0.09
0.10
0.12
0.10
0.11
0.18
0.13
0.10
0.09
0.14
0.11
0.10
0.12
10. magnussoni
0.08
0.08
0.07
0.08
0.07
0.05
0.07
0.07
0.10
0.07
0.06
0.15
0.10
0.07
0.08
0.09
0.09
0.09
0.11
11. masniger
0.09
0.09
0.08
0.09
0.07
0.07
0.07
0.10
0.09
0.06
0.05
0.17
0.11
0.09
0.09
0.10
0.10
0.09
0.11
12. nidicola
0.08
0.09
0.08
0.08
0.07
0.07
0.07
0.09
0.09
0.06
0.04
0.15
0.08
0.07
0.07
0.09
0.08
0.09
0.10
13. olfersioides
0.13
0.14
0.13
0.13
0.14
0.13
0.14
0.12
0.14
0.13
0.14
0.13
0.17
0.15
0.16
0.16
0.18
0.16
0.14
14. ornatus
0.09
0.09
0.09
0.08
0.09
0.09
0.07
0.09
0.11
0.09
0.09
0.07
0.14
0.09
0.10
0.10
0.10
0.10
0.16
15. paleovarzensis
0.05
0.07
0.07
0.07
0.07
0.07
0.06
0.07
0.08
0.07
0.08
0.07
0.13
0.08
0.04
0.08
0.07
0.07
0.11
16. sumtuosus
0.04
0.07
0.08
0.08
0.06
0.07
0.06
0.07
0.08
0.07
0.08
0.06
0.13
0.08
0.04
0.09
0.08
0.07
0.12
17. subfolionidificans
0.09
0.08
0.09
0.02
0.09
0.08
0.08
0.08
0.11
0.08
0.09
0.08
0.13
0.08
0.07
0.08
0.11
0.08
0.14
18. tapajos
0.08
0.08
0.08
0.09
0.09
0.08
0.06
0.09
0.10
0.08
0.09
0.07
0.14
0.09
0.07
0.07
0.09
0.08
0.13
19. trilineatus
0.07
0.07
0.08
0.07
0.08
0.07
0.05
0.08
0.09
0.08
0.08
0.08
0.13
0.08
0.06
0.06
0.07
0.07
0.13
20. undulatus
0.10
0.11
0.11
0.11
0.10
0.10
0.11
0.11
0.10
0.09
0.09
0.09
0.11
0.13
0.09
0.10
0.11
0.11
0.11
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958

23
FIGURE 1. (A) Dorsal and (B) ventral views of Allobates bacurau holotype (INPA-H
959
35398 – original field number APL 12647), an adult male. (C) Dorsal and (D) ventral
960
views of a female paratype (INPA-H 35397 – original field number APL 12656)
961
Scratches on skin occurred post-preservation.
962
963
964
965
966
967
968
969
970
971
972

24
FIGURE 2. (A) Lateral view of Allobates bacurau holotype (INPA-H 35398). (B) Lateral
973
view of a female paratype (INPA-H 35397).
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998

25
FIGURE 3. (A) Hand of Allobates bacurau holotype (INPA-H 35398). (B) Hand of a
999
female paratype (INPA-H 35397). Edges of some tubercles have been digitally highlighted
1000
for improved clarity. Variation in swelling of Finger III is not exclusively sexually
1001
dimorphic. Some male specimens have Finger III more similar to (B).
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025

26
FIGURE 4. Foot of Allobates bacurau holotype (INPA-H 35398). White arrow indicates
1026
basal webbing between Toes III and IV.
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056

27
FIGURE 5. Variation in dorsal color pattern among six male paratypes of Allobates
1057
bacurau. Clockwise from upper left: INPA-H 35405, 35403, 35406, 35409, 35401 and
1058
35407. Note presence of a pale dorsolateral stripe and the absence of dark transverse bars
1059
on thighs of all specimens. Scratches on skin occurred post-preservation.
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080

28
FIGURE 6. Color in life of Allobates bacurau photographed at the species type locality
1081
in Manicoré, on the eastern bank of the Madeira River, in Amazonas State, Brazil. (A)
1082
Dorsolateral view of a male specimen. (B) Lateral view of a second male specimen. (C)
1083
Calling male. Note gray-colored vocal-sac. Also note variation in coverage of iridescent-
1084
white reticulation on the ventrolateral region among the three specimens.
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096

29
FIGURE 7. (A) Dorsal and ventral views of a live female paratype of Allobates bacurau
1097
(INPA-H 35408). Note that throat is pigmented anteriorly and white posteriorly. (B)
1098
Dorsolateral view of a second live female paratype (INPA-H 35397). (C) Variation in
1099
throat pigmentation density among live male paratypes of A. bacurau (from left to right:
1100
INPA-H 35398, 35402, 35401).
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120

30
FIGURE 8. Advertisement calls of Allobates bacurau´s holotype (INPA-H 35398),
1121
recorded at 25.2°C. Advertisement calls of A. bacurau are characterized by the emission
1122
of short notes is long trills. (A) Waveform (upper graph) and sonogram (lower graph) of a
1123
12 s long note trill. (B) Detailed view of a 1 s long section of the same recording. Note
1124
ascending frequency modulation of notes.
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135

31
FIGURE 9. Waveforms (upper graphs) and sonograms (lower graphs) of advertisement
1136
calls of Allobates bacurau and three cryptically colored Allobates distributed in forests on
1137
the east bank of the middle and lower course of the Madeira River, Amazonas State,
1138
Brazil. In comparison with advertisement calls of A. bacurau, A. grillisimilis emits
1139
shorter (< 0.5 s long) note trills; emission of notes by A. caeruleodactylus is continuous
1140
(i.e. notes are not arranged in discrete trills) and silent intervals between notes are longer
1141
than the ones in the calls of A. bacurau; A. masniger also emits notes continuously, but at
1142
a lower frequency bandwidth. Calls of A. grillisimilis, A. caeruleodactylus and A.
1143
masniger recorded in 2008 in Borba (04°26'03" S, 59°37'25" W), north of Manicoré.
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164

32
FIGURE 10. Unrooted maximum likelihood phylogenetic tree reconstructed from a 481
1165
bp fragment of the mitochondrial 16S rDNA sampled from six Allobates bacurau
1166
paratypes and other cryptically colored species of Allobates distributed in Brazil and
1167
across cis-Andean South America. Clade labels indicate bootstrap support values
1168
estimated from 5000 bootstrap replicates (only support values >80% are shown). Basal
1169
clades with low bootstrap support should not be considered to reflect the true
1170
evolutionary history between taxa. Locations described in parentheses indicate sequence
1171
samples obtained from voucher specimens that did not proceed from the species type
1172
locality.
1173
1174
1175

33
FIGURE 11. (A) Relative location of middle and lower course of the Madeira River in
1176
South America. Black rectangle delimits the area shown in detail in (B). (B) Sampling
1177
sites discussed in text (see Geographic Distribution). 1- Manicoré; 2- Village of
1178
Democracia; 3- Novo Aripuanã; 4- Borba; 5- Nova Olinda do Norte. (C) LandSat image
1179
of the city of Manicoré and surroundings. White square: central area of the city of
1180
Manicoré. Yellow star: type locality of Allobates bacurau on Estrada do Miriti, a dirt
1181
road south of the city. White areas correspond to open habitat; gray shades indicate
1182
forested areas; black colors correspond to water bodies.
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192

34
FIGURE 12. Cryptically colored Allobates species occurring in forested areas on the
1193
eastern bank of the middle and lower course of the Madeira River, in Amazonas State,
1194
Brazil. (A) Allobates bacurau photographed at the species type locality in Manicoré.
1195
Note golden yellow color of upper arm. (B) Calling male of A. bacurau photographed in
1196
the same area. (C) Male Allobates cearuleodactylus photographed in Borba, 280 km
1197
downstream of Manicoré. Note the sky-blue fingers. (D) Dorsolateral view of a male
1198
Allobates grillisimilis, photographed at the species type locality in Borba. Note olive
1199
brown color of upper arm. (E) Calling male of A. grillisimilis photographed in Nova
1200
Olinda do Norte, 370 km downstream of Manicoré. (F) Male and (G) female Allobates
1201
masniger, photographed at A. bacurau type locality in Manicoré. Color of throat and
1202
chest surfaces in A. masniger are gray to black in males and light yellow in females. (H)
1203
and (I) an undescribed species of Allobates, photographed in Nova Olinda do Norte and
1204
Borba, respectively. Note the bright yellow vocal-sac and conspicuous solid white
1205
ventrolateral stripe extending from groin to the upper lip. Photos by P.I. Simões, except
1206
(D), by A.P. Lima.
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216

35
Appendix I
1217
1218
Origin and accession numbers of 16S rDNA sequences of Allobates species occurring in
1219
Brazil and across cis-Andean South America used in this study.
1220
Species
Locality
Coordinates
16S
Reference
bacurau sp. nov.
Manicoré, Amazonas,
Brazil
05°52'05" S,
61°17'13" W
KU195698
This study
bacurau sp. nov.
Manicoré, Amazonas,
Brazil
05°52'05" S,
61°17'13" W
KU195699
This study
bacurau sp. nov.
Manicoré, Amazonas,
Brazil
05°52'05" S,
61°17'13" W
KU195700
This study
bacurau sp. nov.
Manicoré, Amazonas,
Brazil
05°52'05" S,
61°17'13" W
KU195701
This study
bacurau sp. nov.
Manicoré, Amazonas,
Brazil
05°52'05" S,
61°17'13" W
KU195702
This study
bacurau sp. nov.
Manicoré, Amazonas,
Brazil
05°52'05" S,
61°17'13" W
KU195703
This study
algorei
Road to Rio Negro,
Tachira, Venezuela
10°19’17” S,
64°33’47” W
HQ290950
Santos &
Cannatella 2011
caeruleodactylus
Castanho, Amazonas,
Brazil
03°37’10” S,
59°86’78” W
EU342541
Santos et al.
2009
caeruleodactylus
Castanho, Amazonas,
Brazil
03°37’10” S,
59°86’78” W
EU342542
Santos et al.
2009
conspicuus
Porto Walter, Rio
Juruá, Acre, Brazil
08°15’31” S,
72°46’37” W
DQ502135
Grant et al. 2006
conspicuus
Porto Walter, Rio
Juruá, Acre, Brazil
08°15’31” S,
72°46’37” W
DQ502134
Grant et al. 2006
crombiei
Cachoeira do Espelho,
Altamira, Pará, Brazil
03°39’00” S,
52°22’33” W
KF250508
Simões et al.
2013a
crombiei
Cachoeira do Espelho,
Altamira, Pará, Brazil
03°39’00” S,
52°22’33” W
KF250509
Simões et al.
2013a
cf. flaviventris
Jaci-Paraná, Rondônia,
Brazil
09°12’16” S,
64°21’43” W
KJ747337
Lima et al. 2014
cf. flaviventris
Cachoeira do Jirau,
Rondônia, Brazil
09°19’27” S,
64°42’39” W
KJ747336
Lima et al. 2014
cf. flaviventris
Guajará-Mirim,
Rondônia, Brazil
10°19’17” S,
64°33’47” W
EU342521
Santos et al.
2009
gasconi
Porto Walter, Rio
Juruá, Acre, Brazil
08°15’31” S,
72°46’37” W
EU342561
Santos et al.
2009
gasconi
Porto Walter, Rio
Juruá, Acre, Brazil
08°15’31” S,
72°46’37” W
EU342562
Santos et al.
2009
gasconi
Eirunepé, Rio Juruá,
Amazonas, Brazil
06°45’50” S
70°09’49” W
KJ747333
Lima et al. 2014
gasconi
Eirunepé, Rio Juruá,
Amazonas, Brazil
06°45’50” S
70°09’49” W
KJ747334
Lima et al. 2014
gasconi
Eirunepé, Rio Juruá,
Amazonas, Brazil
06°45’50” S
70°09’49” W
KJ747335
Lima et al. 2014
granti
Lely Mountain,
Suriname
JN690923
Fouquet et al.
2012
granti
Lely Mountain,
Suriname
JN690924
Fouquet et al.
2012
granti
Saul, French Guiana
JN690927
Fouquet et al.
2012
granti
Haut Marwini, French
Guiana
JN690928
Fouquet et al.
2012

36
Appendix I (continued)
1221
1222
1223
1224
Species
Locality
Coordinates
16S
Reference
grillisimilis
Nova Olinda do Norte,
Amazonas, Brazil
05°52’05” S,
61°17’13” W
KF250506
Simões et al.
2013a
grillisimilis
Nova Olinda do Norte,
Amazonas, Brazil
05°52’05” S,
61°17’13” W
KF250507
Simões et al.
2013a
grillisimilis
Borba, Amazonas, Brazil
04°26’03” S,
59°37’25” W
KF250504
Simões et al.
2013a
grillisimilis
Borba, Amazonas, Brazil
04°26’03” S,
59°37’25” W
KF250505
Simões et al.
2013a
magnussoni
Parque Nacional da
Amazônia, Itaituba, Pará,
Brazil.
04°33’12” S;
56°18’00” W
KJ747328
Lima et al.
2014
magnussoni
Parque Nacional da
Amazônia, Itaituba, Pará,
Brazil.
04°33’12” S;
56°18’00” W
KJ747329
Lima et al.
2014
masniger
Borba, Amazonas, Brazil
04°26’03” S,
59°37’25” W
JQ966875
Kaefer et al.
2013
masniger
Parque Nacional da
Amazônia, Pará, Brazil
04°32’51” S,
56°18’13” W
JQ966888
Kaefer et al.
2013
nidicola
Autazes Road Km 12,
Amazonas, Brazil
03°28’02” S,
59°49’09” W
EU342518
Santos et al.
2009
nidicola
Autazes Road Km 12,
Amazonas, Brazil
03°28’02” S,
59°49’09” W
EU342519
Santos et al.
2009
olfersioides
São José da Vitória, Bahia,
Brazil
15°28’ S, 39°18’ W
DQ502120
Grant et al.
2006
ornatus
Tarapoto, San Martin, Peru
EU342549
Santos et al.
2009
ornatus
Tarapoto, San Martin, Peru
EU342550
Santos et al.
2009
paleovarzensis
Careiro, Amazonas, Brazil
03°22’26” S,
59°52’06” W
JQ966835
Kaefer et al.
2013
subfolionidificans
Rio Branco, Acre, Brazil
09°57’ S, 67°57’ W
KF250492
Simões et al.
2013a
subfolionidificans
Rio Branco, Acre, Brazil
09°57’ S, 67°57’ W
KF250493
Simões et al.
2013a
sumtuosus
REBio Trombetas, Pará,
Brazil
01°22’12” S,
56°51’08” W
KF250499
Simões et al.
2013a
sumtuosus
REBio Trombetas, Pará,
Brazil
01°22’12” S,
56°51’08” W
KF250498
Simões et al.
2013a
sumtuosus
Reserva Ducke, Manaus,
Amazonas, Brazil
02
°
55’–03
o
01’S,
59
°
53’–59
o
59’W
KF250501
Simões et al.
2013a
sumtuosus
Reserva Ducke, Manaus,
Amazonas, Brazil
02
°
55’–03
o
01’S,
59
°
53’–59
o
59’W
KF250502
Simões et al.
2013a

37
Appendix I (continued)
1225
1226
1227
Species
Locality
Coordinates
16S
Reference
tapajos
Fazenda Treviso, near river Curuá-
Una, Pará, Brazil
03°08’44” S,
54°50’33” W
EU342545
Santos et al.
2009
tapajos
Parque Nacional da Amazônia,
Itaituba, Pará, Brazil.
04°33’12” S;
56°18’00” W
KR047027
Lima et al.
2015
tapajos
Parque Nacional da Amazônia,
Itaituba, Pará, Brazil.
04°33’12” S;
56°18’00” W
KR047028
Lima et al.
2015
trilineatus
Panguana, Rio Llullapichis,
Huanuco, Peru
DQ502118
Grant et al.
2006
undulatus
Cerro Yutaje, Amazonas, Venezuela
05°46’ N, 66°08’ W
DQ502028
Grant et al.
2006
undulatus
Cerro Yutajé, Amazonas, Venezuela
05°46’ N, 66°08’ W
DQ502029
Grant et al.
2006
undulatus
Cerro Yutajé, Amazonas, Venezuela
05°46’ N, 66°08’ W
DQ283044
Grant et al.
2006