Blomberg | WNF 7321 A | Arquivos de Zoologia - University of Richmond

ISSN 0066-7870
Arquivos de Zoologia
VARIATION AND TAXONOMIC
CLARIFICATION OF THE LARGE SPECIES
OF THE LEPTODACTYLUS PENTADACTYLUS
SPECIES GROUP (AMPHIBIA:
LEPTODACTYLIDAE) FROM MIDDLE
AMERICA, NORTHERN SOUTH AMERICA,
AND AMAZONIA
W. Ronald Heyer
VOLUME 37, FASCÍCULO 3
MUSEU DE ZOOLOGIA DA UNIVERSIDADE DE SÃO PAULO
SÃO PAULO, 2005
UNIVERSIDADE DE SÃO PAULO
Prof. Dr. Adolpho José Melfi
Reitor
MUSEU DE ZOOLOGIA
Prof. Dr. Carlos Roberto Ferreira Brandão
Diretor
Prof. Dr. Hélio Nogueira da Cruz
Vice-Reitor
Profª. Drª. Eliana Marques Cancello
Vice-Diretora
Prof. Dr. Adilson Avansi de Abreu
Pró-Reitor de Cultura e Extensão Universitária
COMISSÃO EDITORIAL
Hussam El Dine Zaher (editor)
EDITORES ASSOCIADOS
Antonio Carlos Marques (Universidade de São Paulo)
Carlos José Einicker Lamas (Universidade de São Paulo)
Mário de Pinna (Universidade de São Paulo)
Sérgio Antonio Vanin (Universidade de São Paulo)
SERVIÇO DE BIBLIOTECA E DOCUMENTAÇÃO
Teresa Beatriz Nunes Guimarães (bibliotecária)
Adriana Nascimento Flamino (bibliotecária)
Airton de Almeida Cruz (arte-finalista)
MUSEU DE ZOOLOGIA DA UNIVERSIDADE DE SÃO PAULO
Av. Nazaré, 481, Ipiranga, CEP 04263-000, São Paulo, SP, Brasil
Arquivos de Zoologia. v.15- 1967Museu de Zoologia da Universidade de São Paulo.
v. ilust.
ISSN 0066-7870
Continuação de: Arquivos de Zoologia do Estado de São Paulo,
v.1-14; 1960-66.
1. Zoologia. I. Museu de Zoologia, USP.
As instruções aos autores são publicadas em inglês e português no fascículo introdutório de cada
volume, assim como a lista dos consultores que assessoram a Comissão Editorial.
Arquivos de Zoologia
MUSEU DE ZOOLOGIA DA UNIVERSIDADE DE SÃO PAULO
CONTEÚDO
W. RONALD HEYER VARIATION AND TAXONOMIC CLARIFICATION OF THE LARGE SPECIES
OF THE LEPTODACTYLUS PENTADACTYLUS SPECIES GROUP (AMPHIBIA:
LEPTODACTYLIDAE) FROM MIDDLE AMERICA, NORTHERN SOUTH AMERICA, AND
AMAZONIA ....................................................................................................................................... 269
ISSN 0066-7870
Arq. Zool., S. Paulo
São Paulo
v. 37
n. 3
p. 269-348
2005
Contents
Abstract ............................................................................................................................................... 269
Introduction ........................................................................................................................................ 270
Materials and Methods ....................................................................................................................... 270
Discussion ........................................................................................................................................... 337
Acknowledgments .............................................................................................................................. 338
References .......................................................................................................................................... 339
Appendix. Specimens examined. ........................................................................................................ 342
Arquivos de Zoologia
MUSEU DE ZOOLOGIA DA UNIVERSIDADE DE SÃO PAULO
ISSN 0066-7870
ARQ. ZOOL. S. PAULO 37(3):269-348
30.11.2005
VARIATION AND TAXONOMIC CLARIFICATION OF THE LARGE
SPECIES OF THE LEPTODACTYLUS PENTADACTYLUS SPECIES GROUP
(AMPHIBIA: LEPTODACTYLIDAE) FROM MIDDLE AMERICA,
NORTHERN SOUTH AMERICA, AND AMAZONIA
W. RONALD HEYER1
ABSTRACT
Variation of external morphology features and advertisement calls are analyzed for species
currently identified as Leptodactylus knudseni, L. labyrinthicus, L. myersi, and L. pentadactylus from
Middle America, the Pacific versants of Colombia and Ecuador, northern South America, greater
Amazonia, and the corridor of open formations from Argentina to northeastern Brazil. Although there
is noticeable geographic variation between eastern and western samples of L. knudseni, the variation
is considered to be intraspecific in nature. Geographic variation within L. labyrinthicus is more
pronounced and most consistent with recognizing three species: L. labyrinthicus (Spix, 1824),
L. turimiquensis new species, and L. vastus A. Lutz, 1930. No new data are available for variation
within L. myersi, which had previously been noted as possibly containing two species. Variation within
L. pentadactylus is also pronounced and most consistent with recognizing four species: L. pentadactylus
(Laurenti, 1768), L. peritoaktites new species, L. rhodomerus new species, and L. savagei new species.
Some specimens that had been identified in collections as either L. knudseni or L. labyrinthicus from
the Brazilian State of Pará are considered to represent an undescribed species, herein described as
L. paraensis new species.
Standard simple statistical tests of significance for sexual dimorphism in members of the study
group may not indicate biological significance. Adult morphological and advertisement call
differentiation patterns among the species recognized in this paper do not provide completely reliable
information for identifying the species involved, suggesting a different pattern of differentiation than
occurs in the Leptodactylus fuscus species group. Larval morphological variation and habitat
differentiation may be important in the evolution of species differentiation in the taxa dealt with in this
paper.
Key words: Variation, taxonomy, Leptodactylus, Middle America, northern South America, Amazonia.
Amphibians and Reptiles, Department of Vertebrate Zoology, National Museum of Natural History, PO Box 37012, Smithsonian
Institution, Washington, DC 20013-7012, USA. E-mail: heyerr@si.edu
Recebido para publicação em 26.01.2004 e aceito em 12.11.2004.
1
270
Arquivos de Zoologia
INTRODUCTION
Ulisses Galatti (unpublished manuscript)
undertook comparative studies of the reproductive
biology of Leptodactylus pentadactylus just north
of Manaus in Brazilian Amazonia and central
Panama. He found striking differences in his studies
of the organisms at the two sites and convincingly
concluded that the two populations studied should
be recognized as distinct species. This conclusion
is supported by the available genetic differentiation
data (Maxson and Heyer, 1988). After receiving a
draft of Galatti’s unpublished manuscript for
comment, I re-examined the data I had taken on
specimens of L. pentadactylus (Heyer, 1979) to
determine whether they were adequate to separate
the specimens I had examined into two taxa; they
were not.
At the same time, examination of certain
specimens of large Leptodactylus from the State of
Pará in Brazil indicated that they were neither
L. knudseni nor L. labyrinthicus. I therefore began
taking more extensive data on large species of
Leptodactylus in an attempt to define the taxa
involved and to determine their geographic
distributions. The taxa included in this study involve
the specimens that were or would be identified as
Leptodactylus knudseni, labyrinthicus, myersi, or
pentadactylus in my previous studies (Heyer, 1979,
1995). The large species of the Leptodactylus
pentadactylus group not included in this study are
L. fallax, flavopictus, and laticeps, because there
is no evidence to suggest that there are taxonomic
problems associated with them and they have
distributions largely or entirely apart from the taxa
included in this study.
MATERIALS AND METHODS
Evaluating variation to discern species limits
in the frogs included in this study is made difficult
by two problems, both likely related to the large
size of the adults. First, due to space constraints in
the Division of Amphibians & Reptiles at the
Smithsonian Institution, I could not borrow all of
the museum specimens even if the curators of the
collections involved were willing to loan the
specimens to me for study. Second, over much of
the geographic ranges of the taxa involved, sample
sizes of adults are characteristically small, in general
due to problems of storage of large specimens in
museum collections.
The limited storage space available did not
allow me to evaluate variation in the way I have
found works most effectively. My optimum method
is to borrow all available materials so that as datataking and analysis proceed, specimens can be reexamined when questions arise. For example,
during data-taking, pattern standards are continually
added and specimens evaluated early in the study
often should be re-inspected against the newer
standards.
The generally small sample sizes of adults
available for most localities do not allow a rigorous
analysis of variation of specimens from single
localities, a critical measuring stick for evaluating
inter-population variation.
Given the nature of the above two problems,
the following approach was used.
Variation in Leptodactylus myersi has been
evaluated relatively recently (Heyer, 1995);
consequently data for this species are used only to
compare with data from the other taxa. As noted
previously (Heyer, 1995), there is significant
variation within L. myersi that may conform better
to recognition of two species. There are no new
data to address this issue.
Data were taken anew for the following
characters of transformed juveniles and adults: sex;
dorsal pattern; belly pattern; lip pattern; posterior
thigh pattern; upper shank pattern; body folds; male
arm hypertrophy; male thumb spines; male chest
spines; male chin, throat, and chest tubercles; upper
shank texture; outer tarsal texture; foot texture;
snout-vent length (SVL); head length; head width;
eye-nostril distance; maximum tympanum diameter
including annulus; thigh length; shank length; foot
length.
Data were accumulated during the years
1994-2003.
Museum abbreviations follow Leviton et al.,
1985 and Leviton and Gibbs, 1988 (for the two
Swedish collections involved), with the exception
of OMNH, herein used for the Sam Noble
Oklahoma Museum of Natural History and the
addition of CBF = Colección Boliviana de Fauna,
La Paz, CV-ULA = Collection of Vertebrates,
Universidad de Los Andes, Mérida, FHGO =
Laboratorio de Anfibios & Reptiles, Universidad
San Francisco de Quito, Fundación Herpetológica
“G. Orces,” Quito, MECN = Museo Ecuatoriano
Vol. 37(3), 2005
de Ciencias Naturales, Quito, and QCAZ = Museo
de Zoología de la Pontificia Universidad Católica
del Ecuador, Quito. Data on most specimens of the
taxa involved were taken from the MZUSP, OMNH,
and USNM collections. In addition, I took data on
specimens from Colombia and Venezuela during
visits to several collections in those countries. Based
on my previous work on these taxa (Heyer, 1979),
the data from these materials should be sufficient
for the purpose of this study.
Five categories were used for sex: (1) adult
females were determined by having curly oviducts
or well-developed ova; (2) juvenile females were
determined by having straight oviducts – only larger
individuals were dissected to make this
determination; (3) adult males were determined as
having vocal slits; (4) juvenile males were
determined as having testes or other secondary
characteristics such as a small thumb spine, but
lacking vocal slits – only larger individuals were
dissected, if needed, to make this determination;
(5) juveniles were small individuals, clearly not
adults, and were not dissected.
For all the pattern characters, outline
drawings of the body area involved were filled in
with patterns used previously or with new sketches.
Each pattern was uniquely labeled.
Body folds are unambiguous in wellpreserved specimens, but may be difficult to
evaluate in less well-preserved specimens. At the
beginning of the data-taking phase, two folds were
evaluated, namely (1) a dorsolateral fold, which
(when present) extends from behind the eye towards
or past where the sacrum articulates with the ilial
shaft; and (2) a flank fold, extending from the
supratympanic fold and curving downward on the
flank, ending about mid-flank. If the dorsolateral
fold is present, it is usually highlighted by a stripe
of pigment which is darker than that which occurs
on the area surrounding the dorsolateral fold. This
dark stripe is used as a proxy for the fold itself where
the state of preservation of the animal makes it
unclear whether a fold is present or not. The flank
fold may or may not be darkly highlighted, so its
condition is much more difficult to evaluate in many
less than well-preserved animals. Towards the end
of the data-taking phase, I noticed that some
particularly well-preserved individuals also had
lateral body folds arising together from the same
origin as the flank fold (from the supratympanic
fold), but extending more or less in a straight line
271
to the groin region. These lateral folds are not
highlighted with darker pigment and thus much
more difficult to evaluate in many of the specimens.
Information is now available that allows a
better understanding of the nature of male
secondary characteristics involving arm
hypertrophy and thumb and chest spines. Jaslow
(1985) indicated that the dark keratinized thumb
and chest spines were shed by Leptodactylus
pentadactylus from Panama during the nonbreeding season. He reported that the male arm,
while larger than the female arm due to humeral
hypertrophy, also had additional muscle growth and
diminution on a seasonal basis. Observations in
captive breeding colonies support and extend
Jaslow’s findings. Richard Gibson (pers. comm.)
observed that captive male L. fallax shed “… the
black keratinous sheath from their thumb spines. A
white hard cone is left in place.” (Leptodactylus
fallax lack chest spines.) James Ellis (pers. comm.)
found that specimens identified by St. Louis zoo
personnel as L. pentadactylus shed both thumb and
chest spines on a regular annual cycle, but a marked
diminution of the arm when the spines were shed
was not observed in the captive colony. A smaller
white core was still visible on the thumbs after the
keratinized sheath was shed, but the entire chest
spine was shed such that no raised protuberance
was seen from where the chest spines had been shed.
Under the dissecting microscope, a white basal
platform from where the keratinized chest spines
will be regrown can usually be seen (pers. observ.).
Thus, size alone does not correlate with
development of these particular secondary
characteristics of male Leptodactylus. There also
appears to be variation related to preservation. In
some individuals thumb spines or chest spines are
well developed but are white, presumably due to
the keratin layer having fallen off. Thus, white and
black spines are not different character states in the
sense of character coding for phylogenetic analysis.
Measurements were taken with dial calipers
recorded to the nearest 0.1 mm following Heyer
et al. (1990) with the addition of tympanum
diameter (described above) and eye-nostril
distance, measured from the middle of the nares to
the anterior corner of the eye. SVL of specimens
that exceeded the ability of measurement with
calipers was determined using a ruler. Following
the recommendation and protocol of Hayek et al.
(2001), an adult male (MZUSP 131849) and an
272
Arquivos de Zoologia
adult female (MZUSP 64198) were each
remeasured 20 times.
Advertisement calls were analyzed using
Canary 1.2 (Charif et al., 1995) following the
terminology in Heyer et al. (1990) unless indicated
otherwise. The calls were digitized at a sample rate
of 22050 Hz, sample size of 17 bits, at an input
speed of 1x. Call duration was measured from the
wave form. Calls were examined using the
audiospectrogram display (spectrogram in Canary
terminology) to determine whether the recordings
would benefit from filtering around the
advertisement calls. Most calls were filtered
because they had considerable signal below and/or
above the advertisement call frequencies. The
audiospectrogram analyses used settings of analysis
resolution filter bandwidth 349.70 Hz and frame
length 256 points; grid resolution of time 5.805 ms,
overlap 50%, frequency 86.13 Hz, FFT size 256
points; window function hamming; amplitude
logarithmic; clipping level –80 dB; display style
smooth. Dominant frequency was determined using
the spectrum analysis of Canary with settings of
analysis resolution filter bandwidth 349.70 Hz, FFT
size 256 points; window function hamming;
amplitude logarithmic; clipping level –79.79 dB.
For beginning and ending call frequencies,
selections of the smallest value equal to or greater
than 25 ms were made at the beginning and end of
the call and analyzed using the same spectrum
analysis settings. Frequency sweep is the difference
of the ending frequency and beginning frequency.
Call amplitude modulation and harmonics were
evaluated from visual inspection of expanded wave
form displays, audiospectrogram displays, and
power spectrum displays.
Most larval samples are identified as
belonging to the species included in this study by
their unique facultative carnivore morphotype
(Heyer et al., 1975). However, few larvae are
unambiguously associated with their adult parent(s).
Larval characters examined are discussed in the
Inter-Unit Analysis section.
Only data for juveniles and adults are
(marginally) sufficient to analyze for
interpopulation variation; in no cases are data
sufficient for statistical comparisons. I exercised
my experience with variational studies in
Leptodactylus to form samples for analysis of
geographic variation and interpreting the variation
observed. In spite of the too strong an influence of
experience guiding decisions rather than having
adequate data for the variation to be understandable
on their own, I have tried to report the process and
results so that other workers could repeat the study
with the same materials.
The first experience-guided decision was in
choosing the basic Units for analysis of variation
throughout the geographic ranges involved. Four
Units were chosen to analyze individually, all based
on my identifications at the outset of the study: (1)
L. knudseni; (2) L. pentadactylus; (3)
L. labyrinthicus; and (4) a previously apparently
unrecognized species from the State of Pará, Brazil,
called the Pará Unit. Five specimens (MZUSP
24947-8 from Jacearacanga, Pará; USNM 202518
from Barreira do Matupiri, Rio Madeira,
Amazonas; MZUSP 131123, USNM 303909 from
Alto Paraiso, Rondônia, all in Brazil) could not be
confidently placed in these four Units. They are
evaluated in the section dealing with species limits.
As indicated above, Leptodactylus myersi was not
included in this part of the study, as it has already
been analyzed.
For each of the four Units, all the localities
for which I had specimen data were plotted on an
appropriately scaled map. Based on clustering
patterns of localities and presence of at least some
adult male individuals, certain near-by localities
were grouped together into areas, with an attempt
to have the areas cover the range of the total
distribution. Data were summarized for all
individuals within each area. In at least some of the
taxa involved, juvenile color patterns either differ
from or are more distinct than adult patterns. The
pattern data were summarized for all individuals
within each area separately by juvenile, male, and
female categories. The pattern data were resummarized in subsequent analyses (Inter-Unit
Analysis and Species Accounts), so the word
descriptions of character states differ. In addition,
some of the patterns are most usefully visualized
in the Inter-Unit Analysis section and the Figures
involved are located there. However, as appropriate,
those Figures are cited in the Intra-Unit Analyses
as well.
Juveniles for the pattern analyses were
defined after examination of the data for all
specimens for each of the four Units and a breakoff point of 90 mm SVL was established below
which all specimens were considered to be
juveniles. The data sheets used to record the
Vol. 37(3), 2005
pattern data used individual identifying symbols
for the patterns (a combination of letters and
numbers). For purposes of analysis, I translated
the various individual patterns into more general
categories described by words rather than
analyzing the alphanumeric symbols which
represent individual pattern standards. Pattern
standards and data sheets are archived at the
Smithsonian Institution.
Variation among area samples was evaluated
to determine if there was geographic variation. Once
the among-area variation was summarized and
evaluated, the data for all other specimens from the
same Unit were compared with the data for their
nearest geographical area to determine whether
these additional specimens demonstrated variation
beyond that observed in the most approximate area
samples.
The patterns of morphological variation
served as one primary piece of evidence in
determining the degrees of differentiation within
each of the four Units. These data, in combination
with available advertisement call data were used to
define differentiated Units for further analysis.
The terms distinct and distinctive have
different meanings as used to describe
differentiation in this study. A character state is
distinct when it occurs in one sample but not in one
or more other samples being compared. A character
state is distinctive when the relative frequency of
occurrence differs markedly between or among
samples being compared.
Summary and analytical procedures are
described at appropriate places in the text.
Remeasurement Data
For both the male and female data,
measurement variation ranges from 3.1-3.6 mm for
SVL and head length, 1.4-2.2 mm for head width,
0.7-0.9 mm for eye-nostril distance and tympanum
diameter, 0.5-0.8 mm for shank length, and
1.6-2.0 mm for foot length. The measurement range
for female thigh length is 1.1 mm and for male
2.3 mm (Table 1).
The coefficients of variation are lower for
SVL, head width, thigh length, shank length, and
foot length than for head length, eye-nostril
distance, and tympanum diameter for both sexes
(Table 1).
273
Table 1. Data for one female (MZUSP 64198) and one male
(MZUSP 131849), each measured 20 times.
♀ SVL
♀ Head length
♀ Head width
♀ Eye-nostril distance
♀ Tympanum diameter
♀ Thigh length
♀ Shank length
♀ Foot length
♂ SVL
♂ Head length
♂ Head width
♂ Eye-nostril distance
♂ Tympanum diameter
♂ Thigh length
♂ Shank length
♂ Foot length
Minimum
124.7
44.8
47.4
10.7
8.3
49.6
53.4
54.7
124.6
42.2
45.3
11.1
8.2
45.4
51.1
53.8
Maximum
128.2
48.4
49.6
11.4
9.0
50.7
54.2
56.7
127.7
45.6
46.7
12.0
8.9
47.7
51.6
55.4
Coefficient
of Variation
0.006
0.019
0.010
0.020
0.023
0.007
0.004
0.010
0.005
0.025
0.007
0.020
0.025
0.012
0.002
0.008
Intra-Unit Analyses
Pará Unit
Geographic Areas – Three Areas containing at least
two adult males and two adult females are situated
in the Belém region and on the lower and middle
portion of the Rio Xingu (Figure 1).
Variation – Dorsal Pattern: There is no noticeable
variation between juveniles and adults or among
areas for adults. Dorsal patterns include almost
uniform, dark irregularly spotted, dark regularly
spotted, and dark transversely elongate bars that
are distinct or slightly in contact with each other.
One non-Area specimen from Tucuruí (MZUSP
62554) has alternating darker and lighter transverse
bands. The other non-Area specimens do not differ
from the geographic Area specimens.
Belly Pattern: Adults in Areas A, B, C have darkly
vermiculated bellies (reverse of pattern of Figure 13B)
or a dark- and light-spotted belly (one female)
(intermediate between Figures 13B and E). The
juveniles have a very distinct to less distinct
labyrinthine belly pattern (Figure 13D). There is no
variation among Areas. The non-Area specimen from
CEMEX (OMNH 34762) has a dark belly with a few
light irregular spots (intermediate between Figures
13B and E), differing in detail from geographic Area
specimens. All other non-Area specimens do not
differ from the geographic Area specimens.
274
Arquivos de Zoologia
Figure 1. Geographic areas for Pará Unit. A = Belém region, B = lower, and C = middle portion of the Rio Xingu.
Lip Pattern: A pattern common to adult specimens
from Areas A, B, C is one of dark triangular bars
on the upper lip with apices extending to the eye/
canthal region. One of the dark bars reaches the
eye in two adults from Area B (Figure 12A). The
three juvenile specimens from Area C are distinct
from the adults in having dark upper lips with
narrow, light, angled bars from the lip to the eye/
canthal region (Figure 12B). The non-Area
specimen from CEMEX (OMNH 34762) has
narrow dark stripes to the eye and loreal region
(Figure 12C). All other non-Area specimens do not
differ from the geographic Area specimens.
Thigh Pattern: Specimens from Areas A, B, C
share a pattern of dark posterior thighs with light
vermiculations (Figure 14A), with or without a
light extension of the light dorsal thigh bars on
the upper portion of the posterior thighs. Two
adults from Area A have a greater development of
light markings on the thigh than the rest of the
adults. One juvenile has a uniform upper posterior
thigh and a darker lower posterior thigh with small,
light, irregular spots. A non-Area female from
Serra do Cachimbo, Pará (MNRJ 2567) has more
extensive light areas on the thighs than the
geographic Area specimens. All other non-Area
specimens do not differ from Area A, B, C
specimens.
Shank Pattern: All individuals have dark, welldefined, transverse bars on the upper shank.
The non-Area specimen from CEMEX (OMNH
34762) has interrupted transverse bars. All other
non-Area specimens do not differ from Area A, B,
C specimens.
Body Folds: All individuals range from
having a series of warts where dorsolateral folds
occur in other species extending from the eye to
the sacrum to having only one or two such warts.
Some individuals within each Area have lateral
folds extending as far as the mid-body, with others
apparently lacking lateral folds. Individuals within
each Area either have no indication of a flank fold
or have a series of 1-4 dark warty spots in the region
where the posterior extent of the flank fold occurs
in other taxa. All non-Area specimens do not differ
from Area A, B, C specimens.
Male Sexual Secondary Characteristics:
There is no apparent variation among Areas for any
male sexual secondary characteristic. The arms
range from having no hypertrophy to extreme
hypertrophy. Males either have small, black
tubercles on the throat, chest, anterior belly and
ventral surface of the upper arms or lack them.
Males have one thumb spine, ranging from small
to moderate size, either black or white. All males
Vol. 37(3), 2005
except for one (118 mm SVL) have a pair of chest
spines; the spines are black or white and of small
or moderate size. There is only one adult male from
a non-Area locality (OMNH 34359) and it does
not differ from the Area A, B, C males.
Upper Shank Texture: Most individuals
from Areas A, B, C have very few to many white or
black tubercles. One individual each from Areas B
and C has, in addition, a weakly developed shagreen
texture. Two individuals from Area C have granular
shanks in addition to the tubercles. The non-Area
specimen from CEMEX (OMNH 34762) has a
shagreen and scattered black tubercle texture. All
other non-Area specimens do not differ from Area
A, B, C specimens.
Outer Tarsal Texture: Most individuals from Areas
A, B, C have few to several white or black tubercles.
Two individuals from Area A and one from Area B
have no or very few white tubercles on the outer
tarsal region. All non-Area specimens do not differ
from Area A, B, C specimens.
Foot Texture: Most individuals from Areas A, B, C
have no tubercles on the sole of the foot. A few
individuals from each Area have very few white or
black tubercles on the sole of the foot. All nonArea specimens do not differ from Area A, B, C
specimens.
275
Adult Size: Given the small sample sizes
and ranges of sizes observed (Table 2), it would
seem that the three Area samples could be drawn
from the same general population. All adult nonArea specimens do not differ from Area A, B, C
specimens.
Measurement Data: Sample sizes are much
smaller than desired to perform discriminant
function analyses. Both male and female results
(Figure 2) indicate that measurement error is as
great as any difference among Area samples.
Life Colors: Field notes for OMNH 34762,
a 105 mm juvenile, are kindly made available by
Janalee Caldwell: “Dorsum dark brown with gray
crossbars. Limbs with light and dark brown stripes.
Posterior surfaces of thighs black with small light
orange spots. Venter gray with tiny white
reticulations. Eye bicolored: upper half golden,
lower copper.”
Advertisement Calls: None available.
Table 2. Adult sizes (SVL in mm), Pará Unit Geographic Areas.
For the definition of geographic areas, see Figure 1.
Geographic Area
A
B
C
Males
104-120
(N = 2)
112-127
(N = 5)
99-129
(N = 4)
Females
122-140
(N = 2)
111-126
(N = 2)
118-136
(N = 3)
Figure 2. Discriminant function plot for male (A) and female (B) measurement data, Pará Unit geographic areas. Sample confidence
ellipses p:0.683. A-C symbols in legend on figure represent areas A-C; the symbol for R represents remeasurement data for a
single specimen (see text for further explanation).
276
Arquivos de Zoologia
Evaluation of Differentiation: The variation
observed is consistent with being entirely
intraspecific with no convincing evidence of
geographic differentiation.
Leptodactylus knudseni Unit
Geographic Areas – Eight geographic areas each
containing at least two adult males are distributed
broadly throughout the total range of the localities
sampled for this study (Figure 3).
Variation – Dorsal Pattern: There are five general
pattern types into which most specimens fit (Table
3). Juvenile and adult patterns are distinctive from
one another (but not as distinctive as for the belly
pattern). In the Area B sample, juveniles have
distinctly dark outlined cross-bars and the adults
are either uniformly dark or with more irregular,
not as distinctly outlined transverse bars as in the
juveniles. In the Area G sample, juveniles are
distinct from adults in having a series of alternating
dark and light sharply defined cross bars. Some
adults have uniform dorsal patterns whereas all
juveniles have some type of distinct pattern. All
non-Area specimens do not differ from the
specimens in their geographically closest Area
samples.
Belly Pattern: In all Areas that have both adults
and juveniles represented, the belly patterns of the
juveniles are distinctive from the adult patterns. In
addition, the juvenile patterns are different between
Areas B and G. In Area B, a common pattern is a
dark belly with distinct light spots (Figure 13E), a
pattern not found in juveniles in Area G. Conversely,
a common pattern in Area G is no pattern (no
melanophores on belly), a pattern not found in any
juveniles from Area B. The differences observed
between juvenile patterns from Areas B and G could
be due to relatively small sample sizes (12 juveniles
from Area B, 15 for Area G), but the juvenile
Figure 3. Geographic areas for Leptodactylus knudseni Unit. See text for explanation.
Vol. 37(3), 2005
patterns from Areas F and G are similar to each
other. Adults from areas D, E, F, G, H have uniform
to indistinctly mottled bellies (Figure 13A). The
only exception is a female from Area F with a light
belly with distinct dark vermiculations (reverse of
pattern shown in Figure 13B). Some adults from
Areas B and C, in addition to having indistinctly
mottled bellies, have dark bellies with distinct light
spots (Figure 13E). The adult belly patterns from
Area A can not be placed readily into any of the
patterns described for the other Areas. The bellies
are indistinctly to distinctly mottled. ZMB 6759,
8534 from Suriname have bellies with more distinct
mottling than specimens from proximate Area D.
All other non-Area specimens do not differ from
the specimens in their geographically closest Area
samples.
Lip Pattern: There are four general pattern types
into which most individuals (both adult and
juvenile) fall (Table 4). In the two Area samples
with the most juveniles, there do appear to be
differences between juvenile and adult patterns. In
Area B, the juvenile lip patterns have well defined
triangular marks with apices on the edge of the
upper lip and with one of the triangular marks
extending as far as the eye (Figure 12A) or not.
The adults from Area B have either uniform dark
lips or indistinct bars. In Area G, most juveniles
have two dark bars that extend from the lip to under
the eye (Figure 12D), whereas the adults do not
have this pattern and have much more uniform lips.
In addition to the patterns indicated in Table 4, one
individual from Area B has an indistinctly barred
277
lip and one individual from Area G has an unusual
pattern combining elements of one versus two
complete dark bars from the lip to the eye. Area G,
which has the largest sample size, contains all
patterns, suggesting that the variation observed
among Areas may be due to inadequate sample
sizes. Non-Area specimen MZUSP 15907 from
Alto Rio Machado, Rondônia, Brazil lacks a lip
pattern, differing from specimens from proximate
Area H. All other non-Area specimens do not differ
from the specimens in their geographically closest
Area samples.
Thigh Pattern: There seems to be a difference
between juvenile and adult thigh patterns in the Area
B sample. All juveniles have a uniformly dark
posterior thigh surface, while one adult male has a
finely mottled thigh and two adult males have boldly
mottled thighs. The thighs among specimens from
Area G do not show as much contrast as for Area
B. The juveniles from Area G have mottled thighs
that are less boldly mottled than some of the adults,
although most of the adults were scored as having
the same thigh patterns as the juveniles from this
Area. Other than the differences discussed above,
there seems to be no variation among specimens
from Areas A through H. The adult thighs are
mottled, though several individuals from Area H
show less distinct mottling than found in the rest of
the adult specimens from Areas A through G. NonArea specimens from Cachoerinha, Rio Madeira
(MZUSP 202517) and Nova Aripuanã, Amazonas,
Brazil have less distinctly patterned thighs than
specimens from proximate Areas F and G. MZUSP
Table 3. Dorsal patterns, Leptodactylus knudseni Unit. For the definition of geographic areas, see Figure 3.
Patterns
Uniform
Distinct, narrow cross bands
Distinct, moderately broad cross bands
Broad cross bars
Irregular cross bars
Blotched
Coalesced bars anteriorly, coalesced blotches posteriorly
A
X
X
X
B
X
X
X
C
X
X
X
Geographic Areas
D
E
F
X
X
X
X
X
X
X
G
X
H
X
X
X
X
X
X
X
X
Table 4. Lip patterns, Leptodactylus knudseni Unit. For the definition of geographic areas, see Figure 3.
Patterns
Two complete dark bars from lip to eye
One dark bar from lip to eye
Upper lip marks only
Uniform
A
B
X
X
X
X
C
X
X
X
Geographic Areas
D
E
F
X
X
X
X
X
X
X
G
X
X
X
X
H
X
X
278
Arquivos de Zoologia
25169 from Forte Príncipe da Beira, Rondônia,
Brazil also has less distinctly patterned thighs than
specimens from proximate Area G. OMNH 34796
from CEMEX, Pará, Brazil has light spots on the
upper thigh, a pattern not found in other individuals
from proximate Area F. All other non-Area
specimens resemble the specimens in their
geographically closest Area samples.
Shank Pattern: Area G individuals demonstrate
variation found in all other specimens from the other
Areas; there is no observed difference between
juvenile and adult patterns.
Body Folds: There is little variation among
Area samples in terms of development of the
dorsolateral fold. The fold is continuous from
behind the eye to the sacrum in most specimens.
Variations include: the fold is interrupted; the fold
extends only 3/4 distance from the eye to the
sacrum; or there is a short break behind the eye
before the fold begins. Specimens from Area H have
less well-developed dorsolateral folds in general
than specimens from the other Areas, but this could
well be an artifact due to preservation differences
and/or small sample sizes. The lateral fold occurs
in at least some individuals from Areas A, B, C,
and D. Most individuals from Areas E, F, G, H were
scored as lacking lateral folds, but a few individuals
from Areas E, F, H were scored as having barely
recognizable lateral folds. There is no indication
of a flank fold in specimens from Area B and only
weak indications of a flank fold in some individuals
from Areas A and C. Most individuals from Areas
D, E, F, G, H have a dark elongate wart in the region
where the flank fold occurs in other members of
this species cluster. EBRG 2390, 2444 have
complete flank folds, differing from specimens in
proximate Area E. OMNH 34360-34361 from Alter
do Chão, Pará, Brazil have slightly shorter
dorsolateral folds (2/3 to 3/4 distance from eye to
sacrum) than any specimen from proximate Area F
(minimum of 7/8 distance from eye to sacrum). All
other non-Area specimens do not differ from the
specimens in their geographically closest Area
samples.
Male Sexual Secondary Characteristics: At
least some males from each Area have very
extensively developed male arms, except for Area
F (N = 2). At least some males from each Area have
a field of black tubercles on the throat, chest, and
anterior belly, except for Area F. For all Areas
except G, males that have a well-developed black
thumb spine also have a pair of black chest spine
patches. In Area G, there are only two males that
have black thumb spines, neither of which have
chest spines. The other males from Area G have
white thumb spines. The lack of chest spines in Area
G could be due to at least two causes: (1) adult
males from Area G are small (see size, below) and
lack chest spines; or (2) males from Area G are as
large as males from other Areas but were not
sampled, and these larger, unsampled males would
have both black thumb and chest spines. The present
data can not distinguish between these alternatives.
Non-Area specimens USNM 288740-288741 from
Parque Nacional de Amazonia, Rio Tapajós, Pará,
Brazil have a secondary proximal bump on the male
thumb, differing from specimens from
geographically proximate Area F. All other nonArea specimens do not differ from the specimens
in their geographically closest Area samples.
Upper Shank Texture: The upper shanks
usually have a few to several scattered black and
white tubercles. A few specimens from Areas B and
F also have a shagreened surface. The only Area
from which a noticeable number of individuals lack
any indication of shagreen or tubercles is Area C.
Non-Area specimens do not differ from the
characterization for Area specimens.
Outer Tarsal Texture: The outer tarsal surface
usually has a few to several scattered black or white
tubercles. Lack of tubercles is unusual in the
samples as is presence of a shagreened surface. Area
G is distinctive in having several individuals with
many scattered black or white tubercles. Four nonArea specimens lack tubercles on the outer tarsal
surface and differ from specimens from proximate
geographic Areas (ROM 28451-28452, USNM
535773 – Area E; OMNH 34360 – Area F). All
other non-Area specimens do not differ from the
specimens in their geographically closest Area
samples.
Foot Texture: Specimens from Areas A, B, C, D
lack tubercles on the sole of the foot. Most
individuals from Areas E, F, G, H also lack a
shagreen or tubercles, but one to three individuals
from each of these Areas have a few white tubercles
scattered on the sole of the foot and one individual
from Area F has a shagreened surface. All non-Area
Vol. 37(3), 2005
specimens do not differ from the specimens in their
geographically closest Area samples.
Adult Size: The most instructive sample for
understanding variation in adult size is that from
Area H, where there is almost one order of
magnitude of difference between the smallest and
largest adult male (Table 5). When using the ranges
of adult size of Area H as a standard for evaluating
the other Area samples, only the sample from Area
G appears unusual. It contains the largest number
of males. The maximum size of these Area G males
is 128 mm SVL. These Area G males may have
just reached sexual maturity and if so, would be
expected to grow to a larger size and develop
hypertrophied arms, large thumb spines, and large
chest spines. Non-Area specimen ZMB 8534 from
Suriname is larger (158 mm SVL) than males from
proximate Area E. The following non-Area males
are larger than males in proximate Area F: MZUSP
54667, 136 mm SVL; MZUSP 54668, 146 mm
SVL; OMNH 34360, 137 mm SVL; OMNH 34361,
147 mm SVL; OMNH 34796, 125 mm SVL. All
other non-Area specimens do not differ from the
specimens in their geographically closest Area
samples.
Measurement Data: Sample sizes for most
Area samples are too small for discriminant
function analyses, and are minimally adequate for
the male data. There is broad overlap in the plot of
the first two canonical variables (Figure 4) when
using the complete data option in Systat 10
(Engelman, 2000) for many of the Area samples.
279
Using the remeasurement data as a frame of
reference, Areas B, C, and G are distinct from one
another. It is interesting that these three Areas are
geographic neighbors among the samples (Figure
3). The Area B and G samples differ only along the
first canonical axis, which is typically size related.
The first axis accounts for 47% of the dispersion.
The Area C sample differs on the second canonical
axis, however. The second axis accounts for 29%
of the dispersion and is most heavily influenced by
foot length and eye-nostril distance (based on
standardized by within-variance values).
Life Colors: Ecuador. In juveniles, the head
is mostly yellowish green, the dorsum has greenish
yellow bands enclosing brownish-green areas that
are black bordered, the chin has yellow marks along
the edge, the posterior thigh is jet black, the belly
is gray with lighter punctations, the iris is gold
yellow above and rusty gold below (Heyer, 1972:5).
There is only one adult female in the study sample
with color information; it had a green head in life
(Coloma and Ron, 2001, fig. 50, QCAZ 16183).
Peru. Color notes by Robert P. Reynolds for
80 mm juvenile, USNM 206636: “Dorsum reddish
brown with green transverse cross bands on snout,
between eyes and on back. Dark brown canthal
stripe from snout to behind tympanum. Side of face
green with black bars on lips. Chin, throat, chest,
and belly dark brown with white spots. Orange spots
on flanks in front of insertion of rear legs. Arms,
Table 5. Adult sizes (SVL in mm), Leptodactylus knudseni
Unit geographic areas. For the definition of geographic areas,
see Figure 3.
Geographic Area
A
B
C
D
E
F
G
H
Males
131-156
(N = 4)
143-159
(N = 5)
117-148
(N = 5)
132-147
(N = 3)
117-144
(N = 6)
101-115
(N = 2)
94-128
(N = 13)
96-152
(N = 7)
Females
120-129
(N = 2)
(N = 0)
121-133
(N = 3)
154
(N = 1)
134
(N = 1)
140
(N =1)
103-128
(N = 7)
134
(N =1)
Figure 4. Discriminant function plot for male measurement
data, Leptodactylus knudseni Unit geographic areas. Sample
confidence ellipses p:0.683. A-H symbols in legend on figure
represent Areas A-H; the symbol for R represents
remeasurement data for a single specimen (see text for further
explanation).
280
Arquivos de Zoologia
legs, and flanks dark brown.” A color photo of
USNM 342999, a 28 mm specimen, shows dorsal
crossbands that are tan, brown, and almost black.
The side of the head is tan and brown. The flanks
are dark, but not as dark as the juveniles from
Ecuador. There is no indication of green. A color
photo of USNM 343244, a 83 mm specimen, shows
the dorsum with alternating green and brown
transverse bars, the side of the head is tan (not
green) and the flanks are yellow-tan. An
unvouchered photo of an adult from Tambopata,
Madre de Dios (Roy W. McDiarmid photo) shows
the dorsum indistinctly marked with different
shades of dark browns, the side of the head is tan/
brown and the lower flanks are golden tan. William
E. Duellman provided the following description
from a manuscript on the Cuzco Amazonico, Madre
de Dios herpetofauna (pers. comm.): “In life, the
dorsum in adults is tan to reddish brown with dark
brown markings consisting of an irregularly shaped
spot on top of the head anterior to the orbits, a
transverse bar between the orbits, three-to five
broad transverse marks on the body to the level of
the sacrum, and several small spots postsacrally.
All markings on the body are between the orangebrown dorsolateral folds. The dorsal surfaces of the
limbs are tan to reddish brown with narrow
transverse brown bars. The flanks are dull orange,
and a pale orange-brown longitudinal stripe is
present on the upper surface of the forearm (Fig.).
The upper lip is tan with a brown margin and brown
spots, at least posteriorly, and a dark brown canthal
stripe is present. The venter is cream with diffuse
brown mottling, except for the throat, which is dark
brown with creamy white flecks. Notable
ontogenetic changes occur in the color pattern.
Juveniles tend to have a greenish yellow dorsum
and gray flanks, and the markings on the face and
body are distinct; in the largest individuals, the
pattern is barely discernible. Also, in juveniles, the
posterior surfaces of the thighs are black, whereas
in large adults they are black to dark brown with a
few creamy white spots dorsally.” The juvenile
color pattern is illustrated by a W. Lamar
photograph from Peru, Loreto (Rodríguez and
Duellman, 1994, Plate 11, fig. D, data for photo
provided by W.E. Duellman, pers. comm.). This
distinctive green and black juvenile color pattern
occurs in many but not all Peruvian juveniles.
Bolivia. De la Riva et al. (2000) have a color
photo of presumably an adult individual with a
brown dorsum (no evidence of green), and the
posterior thigh has tan markings on a black
background.
Venezuela. A color photograph from the
Llanos of Venezuela (Kornacker and Dederichs,
1998:72, fig. 12, as L. pentadactylus) shows a
brown dorsum with striking narrow chartreuse green
cross stripes; the side of the face is predominantly
lemon yellow and flanks are uniform orange-tan.
No size is given for the animal photographed and it
is not possible to tell from the photograph whether
the individual is a juvenile or adult. A color photo
of AMNH 131088 from Neblina shows a dorsum
with alternating lighter and darker brown transverse
bars, a brown side of the head, and golden tan lower
flanks. A color photo of an unvouchered specimen
(juvenile judging by size of frog relative to leaves
on ground in photo) from Neblina (Roy W.
McDiarmid photo) shows the dorsum with
alternating greenish-yellow (not vivid) and brown
transverse bars; the side of the head and flanks are
tan. Gorzula and Señaris (1998: color photograph
49) show an adult (no locality information, in
contrast to other photo captions) with a brown and
tan dorsum (no evidence of green).
French Guiana. Individual specimens have
reddish maroon backs with broad transverse bands
of yellow chestnut bordered by black, posterior
thighs black crossed by orange bars or orange red
(Hoogmoed and Avila Pires, 1991; Lescure and
Marty, 2000 [presumably based on French Guiana
adults]); anterior part of flank red, a large orange
spot on the posterior part of the flanks on a glandular
area just in front of the groin, throat dark grey with
white spots (Hoogmoed and Avila Pires, 1991); pale
gray belly with yellow marbling, red maroon iris
below, golden above (Lescure and Marty, 2000).
Brazil. Specimens have tan/brown on the
dorsum and sides of head and flanks with a large
central orange-tan spot (Rodríguez and Duellman,
1994, Plate 11, fig. C, W. Hödl photo of adult from
Manaus, fide W.E. Duellman, pers. comm.).
Advertisement Calls: Although there are
several recordings available for the Leptodactylus
knudseni Unit throughout its distributional range,
many of the calls lack associated voucher specimens
(Table 6) and half of them have considerable
background noise in the same broadcast channel
that the frogs are using. The best quality recordings
have calls that rise moderately quickly in intensity
and maintain relatively high intensity until the final
687
679
238
173
515
38.1
12-14
0.32-0.36
48
9
228, 1, Brazil, Pará1, +
–
516-687
690
432
591
285
479
319
170
516
540
35.6
31.0
8-10
6-7
0.16-0.20
0.26-0.30
60
66
7
227, 2, Brazil, Pará1, +
25
10
227, 1, Brazil, Pará1, –
25
515
344+515
461
499
177
212
174
177
344
346
31.0
29.0
10-11
13
0.42-0.43
0.36-0.38
25
20
–
French Guiana CD, –
–
4
4
213, 13, Brazil, Roraima2, +
344
517
526
498
139
198
168
175
341
346
36.5
29.7
9-10
11-13
0.34-0.38
0.32-0.34
16
20
3
209, 3, Brazil, Roraima1, –
–
4
224, 4, Brazil, Amazonas2, –
–
514
342
798
606
150
156
82
171
343
260
26.0
35.8
9-10
8-9
0.31-0.34
0.25-0.27
18
20
4
21, 1, Brazil, Amazonas1, –
–
2
267, 3, Peru, M. de Dios1, –
26
514
435
293
29
487
27.7
36.8
8-10
6-7
0.20-0.24
0.21-0.25
29
28
265, 10, Peru, M. de Dios1, –
–
5
6
105, 1, Peru, Amazonas, +
23
Frequency modulation, Hz
Mean dominant
Entire call
Mid-call pulse
frequency, Hz
Mean, start Mean, range
Mean, start Mean, range
343
174
146
447
342
Mean pulse
rate per s
# pulses
per call
Call rate/ Call duration
minute
range in s
Air temp, C
N
Tape & Cut,
Country, State, Vouchered
Table 6. Advertisement call data for Leptodactylus knudseni Unit. Numbers associated with States indicate whether the data are for the same or different specific localities in the same State.
Vol. 37(3), 2005
281
1/3 to 1/4 of the call, which diminishes in intensity
(Figures 5, 6). In the noisiest recordings, the final
portion of the call is apparently indistinguishable
from the background noise, which would result in
data that are truncated in terms of call duration and
pulses per call. The noisy recordings are USNM
recordings 105 cut 1, 265 cut 10, 21 cut 1, 209 cut
3, and 227 cut 1. Of the four recordings with
museum vouchers, I have examined two for this
study (USNM recording 105 cut 1, USNM 566002;
USNM recording 213 cut 13, MZUSP 65675).
General features of all calls include: (1) each
call consists of a single note which is either almost
entirely or partially pulsed (Figures 5, 6); (2) most
calls have modest, but noticeable increasing
dominant frequencies from the beginning to the end
of the call (USNM recordings 265, cut 10 and 267
cut 3 are exceptions with little or no modulation of
the entire call); (3) within each call, individual
pulses are frequency modulated in a complex
fashion – the best recordings indicate both a sharp
rise and fall of frequency within each pulse (Figure
6); and (4) the dominant frequencies are all low
pitched (Table 6).
Understanding the variation among calls is
not straightforward.
Given the ranges in adult sizes (Table 5),
one would expect there to be a correlation between
call duration and carrier frequency with size, as
amply demonstrated in other studies (Duellman and
Trueb, 1986). The available data are inadequate for
such an evaluation, however.
Two recordings from Tambopata, Madre
de Dios, Peru (USNM recordings 265 cut 10, 267
cut 3) and three recordings from Usina Kararahô,
Altamira, Pará, Brazil (USNM recordings 227 cut
1, 227 cut 2*, 228 cut 1) provide data for
evaluating intrapopulation call variation. Variation
of calls within each of these localities encompasses
the total range of all variation observed in call
duration and number of pulses per call. The lowest
values for these two parameters from both
localities are from noisy recordings. It is likely
that the actual call values for these two recordings
approximate the values for cleaner recordings at
each locality.
* A.J. Cardoso associated specimen ZUEC 7232,
Leptodactylus pentadactylus, as the voucher specimen for
this recording. The specimen is a 106.5 mm SVL juvenile
female, however, and could not be the source of the recording.
282
Arquivos de Zoologia
There are some features unique to given
recordings. USNM recording 213 cut 13 has two
equal but distinct dominant frequencies for the
Figure 5. Wave form and audiospectrogram of advertisement
call of Leptodactylus knudseni Unit from Peru (USNM
recording 267, cut 3).
Figure 6. Wave form and audiospectrogram of advertisement
call of Leptodactylus knudseni Unit from Pará, Brazil (USNM
recording 228, cut 1).
entire call (Table 6). USNM recording 227 cut 1
has a dominant frequency band extending from 516
to 687 Hz.
Most of the rest of the variation is interpreted
as individual variation with two exceptions.
1) Two of the three recordings from Usina
Kararahô, Pará, Brazil have more pronounced
pulsatile notes than observed in all other
recordings (contrast Figures 5 and 6).
2) The call rate for the recordings from Usina
Kararahô is much greater than for all other
recordings. The voice tags on USNM recording
227 sound as though the tape was recorded at a
slightly lower speed than normal, because the
human voice seems somewhat high in pitch.
However, the USNM recording 228 cut 1 voice
tag appears normal and the call rate for this
individual specimen is still much greater than
for recordings from other localities. The few
recordings with temperature data indicate that
this call rate difference is not due to temperature
differences.
Given the nature of the data, the calls from
Usina Kararahô, Pará, Brazil and Tambopata,
Madre de Dios, Peru are considered distinctive from
each other and from all other recordings.
Evaluation of Differentiation: There is
unquestionably geographic variation within the
L. knudseni Unit, both in terms of morphology and
advertisement calls.
The dorsal pattern, belly pattern and flank
fold data indicate that the Ecuador and Peru samples
are similar to each other and distinctive from all
other samples. The measurement data indicate that
the Area B and C samples are distinctive from each
other. The advertisement call data indicate that the
Pará, Brazil and Madre de Dios, Peru samples are
distinct from each other and all other samples.
Only the call rate sharply defines the Pará
sample from all other samples.
The life color patterns of juveniles from
Ecuador (Area B) and southern Peru (Area C) are
similar. There is an imbalance of males and females
between the samples of Ecuadorian and Peruvian
specimens that were included in Area B. Only one
adult female and 12 juveniles from Ecuador and 5
adult males from Peru comprise the Area B sample.
The data for the Peruvian Area B specimens do not
closely match either the data for the Ecuadorian
Vol. 37(3), 2005
Area B specimens or those for Peruvian Area C.
Additional data were obtained for the Inter Unit
analysis section.
The variation within the Leptodactylus
knudseni Unit is interpreted to mean that there is
one differentiated Unit for combined Area A, D-H
specimens and either two or three differentiated
Units for Area B and C individuals.
Leptodactylus pentadactylus Unit
Geographic Areas – Nine geographic areas are
distributed broadly throughout the total range of
the localities sampled for this study (Figure 7). Due
to small sample sizes, many of the geographic areas
are relatively large and one (Area I) does not contain
any adult male individuals.
Variation – Dorsal Pattern: There is no indication
of differences in juvenile and adult patterns in the
available data. This finding is surprising. I
remember collecting juvenile Leptodactylus
pentadactylus in Costa Rica during my first tropical
field experience (1964) and thinking that they were
a different species from adult L. pentadactylus from
the same locality. The life coloration is more vivid
283
in juveniles (see below), but the patterns that remain
after preservation are similar in juveniles and adults.
There is no indication of variation between adult
sexes or among geographic areas. Dorsal patterns
range from uniform; one narrow transverse bar in
the scapular region; two narrow transverse bars with
the second in the sacral region; a series of irregular
dark-outlined transversely elongate rectangles
alternating dark and light and the same pattern
except the irregular rectangles are larger and
coalesce and may be uniform in intensity rather than
alternating dark and light. One non-Area specimen
(La Salle 272 from Florencia, La Vega, Caquetá,
Colombia) has a completely variegated dorsal
pattern. All other non-Area specimens match dorsal
patterns found in their closest geographic area.
Belly Pattern: There is a suggestion that juveniles
from Areas B, C, E, F, I more commonly have a
pattern of a dark belly with rather distinct, small,
light spots than adults have. Otherwise, there is no
indication of juvenile belly patterns being different
from adult patterns. At least some individuals in all
Areas except Area E have dark bellies with distinct,
relatively small light vermiculations (Figure 13B).
The majority of specimens from Areas A, B, D, F
Figure 7. Geographic areas for Leptodactylus pentadactylus Unit. See text for explanation.
284
Arquivos de Zoologia
have a visually striking pattern of dark bellies with
distinct, large and regular (Figure 13E) to irregular,
spots on the belly. A few individuals from Areas B,
C, D, E, G have much more indistinctly mottled
bellies (Figure 13A) than the rest of the individuals
from those Areas. All non-Area individuals match
patterns with the closest geographic sample
specimens.
Lip Pattern: Juveniles do not appear to have
different lip patterns from adults. There is as much
variation among individuals within the geographic
area samples as among samples. Most individuals
have dark inverted triangular markings on the upper
lip, one of which often extends to the eye (Figure
12A). Two individuals from Area A and one
individual from Area C were recorded as having
no lip pattern. All non-Area individuals match
patterns with the closest geographic sample
specimens.
Thigh Pattern: Juveniles do not appear to have
different thigh patterns from adults. Most
individuals exhibit one of four primary kinds of
thigh patterns, three of which vary among
geographic area samples for the specimens analyzed
(Table 7). In addition to these four patterns, (Table
7), a few individuals from Areas C and E have
extended horizontal light stripes connecting light
vertical markings; one individual each from Areas
F and I were scored as having indistinctly mottled
to uniformly patterned thighs; one individual each
from Areas F and G were scored as having a boldly
patterned thigh with extensive, irregular, more-orless horizontally oriented light stripes; and several
individuals from Area H were scored as
demonstrating a distinct to less distinct pattern of
large dark blotches outlined by a contrasting light
border (Figure 14C). One non-Area specimen
(UNSM 521020 from Aguas Negras, Río Lagarto
Cocha, Loreto, Peru) is closest geographically to
Area G but has the distinctive pattern described for
Area H. All other non-Area individuals match
patterns found in the closest geographic samples.
Shank Pattern: Juvenile upper shank patterns do
not differ from adult patterns. There is as much
variation among individuals within geographic
areas as among areas. The upper shanks are either
uniformly dark or with dark transverse bands on a
lighter background. For patterns with dark bands,
they either traverse the entire upper shank surface
or fall short distally. There was no additional
variation observed in the non-Area specimens.
Body Folds: Lateral folds are difficult to
determine in most of the specimens. The range
includes no indication of a lateral fold, an
interrupted lateral fold, and a weak but distinct
entire lateral fold. There is considerable variation
in the extent of expression of dorsolateral and flank
folds, with variation in degree of development
among areas (Tables 8, 9). Two non-Area specimens
that are nearest to Area C (USNM 1976 from San
Juan del Norte, Río San Juan, Nicaragua; USNM
29953 from Turrialba, Cartago, Costa Rica) have
dorsolateral folds that are continuous to the sacrum.
Two specimens that are closest to Area G differ by
either having the dorsolateral fold fall just short of
the sacrum on one side but past the sacrum on the
other (La Salle 272 from Florencia, La Vega,
Caquetá, Colombia) or having no indication of a
flank fold (La Salle 293, same locality as La Salle
272). All other non-Area specimens match
conditions found in specimens from the nearest
geographic area samples.
Male Sexual Secondary Characteristics:
The arms are not hypertrophied in most of the
males from Areas A, B, F, G, H. Three males from
Area B and one male from Area F show a modest
degree of hypertrophy. At least some males in Areas
C, D, E have extensively hypertrophied arms. One
male from Area G has very weakly hypertrophied
arms.
All males from Areas A and H lack tubercles on the
chin, throat, and chest. Data were not recorded for
this feature from Area E individuals. One male from
Area B has tubercles on the throat. Most males from
Areas C and D have tubercles on the throat or on
the throat and chest. One male from Area F has
tubercles on the throat and chest. One individual
from Area G has tiny white tubercles limited to the
side of the throat.
All males from Areas A, F, G, H lack chest spines.
At least some, if not most, individuals from Areas
B, C, D, E have a well-developed pair of black chest
spines.
All males except from Area D have one spine per
thumb (medially placed). The thumb spines from
Vol. 37(3), 2005
285
Table 7. Posterior thigh patterns, Leptodactylus pentadactylus Unit. For the definition of geographic areas, see Figure 7.
Major pattern types
1. Dark thighs with small to medium sized light vermiculations (Fig. 14A)
2. Boldly patterned thighs with large light regular or irregular spots
3. Light dorsal transverse bars ending in light spots on posterior thigh (Fig. 14 B)
4. Extensive light areas (Fig. 14 D)
A
X
X
X
B
X
X
X
X
Geographic Areas
C D E F G
X X X X X
X X
X X
X X
H
X
I
X
Table 8. Dorsolateral fold conditions, Leptodactylus pentadactylus Unit. For the definition of geographic areas, see Figure 7.
Major character states
1. Folds continuous to just short of sacrum
2. Folds continuous to sacrum
3. Folds continuous to sacrum, interrupted from sacrum to groin
4. Folds complete from eye to groin, but interrupted
5. Folds complete from eye to groin
A
X
X
B
X
X
X
X
X
Geographic Areas
C D E F G
X
X
X
X
X
X
X
H
I
X
X
X
X
X
X
X
X
X
X
X
X
X
H
I
X
X
X
X
X
X
Table 9. Flank fold conditions, Leptodactylus pentadactylus Unit. For the definition of geographic areas, see Figure 7.
Major character states
A
B
1. Very short, lying somewhere in flank fold field
2. Interrupted from tympanum to posterior arm insertion area
3. Complete from tympanum to posterior arm insertion area
4. Interrupted from tympanum to mid-body
5. Complete from tympanum to mid-body
X
X
X
X
X
X
Area A males range from tiny to moderate size
white; Area B males have small to large white
spines; all Area C males have a large black spine
on each thumb; Area D male variation includes
small to large white spines, large black spines, one
large white and an additional small white spine at
the base of the thumb, and one large black spine
and one tiny white spine at the base of the thumb;
Area E males have small, white to large, black
spines; Area F male spines range from tiny white
to large black; Area G males have tiny or small white
spines; Area H males have no spines or tiny white
spines.
One non-Area male (Berlin 30972) from Rio Gutaki
(probably Jutahy = Jutaí), Amazonas, Brazil differs
from Area G and H males in having just slightly
more hypertrophied arms and having a larger white
thumb spine. All other non-Area males match
conditions found in specimens from the nearest
geographic area samples.
Upper Shank Texture: There is as much
within as among Area variation. The conditions
include smooth, with or without a shagreen, and
none to many white or black tubercles. There is no
additional variation in non-Area specimens.
Geographic Areas
C D E F G
X X
X X
X X X X X
X X X X X
X X X X X
Outer Tarsal Texture: Most individuals from Area
A have smooth surfaces; two individuals have
shagreened surfaces. Individuals from the other
geographic areas range from having smooth,
shagreened, or several white or black tubercles.
There is no additional variation in non-Area
specimens.
Foot Texture: The within Area sample variation is
as great as among Area sample variation. The foot
texture is smooth in most specimens; a few
individuals have no more than a few white or black
tubercles. There is no additional variation in nonArea specimens.
Adult Size: The ranges of adult size within
areas combined with generally small sample sizes
may be masking real size differences. The most
suggestive size variation is that the females from
Areas G, H, I may be larger than females from the
other Areas (Table 10). One non-Area female
(USNM 192577 from 10 miles S Turrialba,
Cartago, Costa Rica) is smaller (110.2 mm SVL)
than females from its nearest Area sample (Area
C). One male (MZUSP 40303 from São José
(Jacaré), Rio Solimões, Amazonas, Brasil) is
considerably larger than any male in the Area G
286
Arquivos de Zoologia
Table 10. Adult sizes (SVL in mm), Leptodactylus
pentadactylus Unit. For the definition of geographic areas, see
Figure 7.
Geographic Area
H
Males
128-151
(N = 3)
118-142
(N = 7)
134-148
(N = 6)
113-154
(N = 8)
119-144
(N = 6)
112-146
(N = 6)
121-161
(N = 4)
100-150
(N = 4)
I
(N = 0)
A
B
C
D
E
F
G
Females
134-149
(N = 12)
119-148
(N = 5)
122-145
(N = 4)
118-152
(N = 10)
121-164
(N = 6)
115-158
(N = 7)
142-174
(N = 5)
146-155
(N = 4)
135-162
(N = 5)
and H samples (195 mm SVL). This large male has
very weakly hypertrophied arms and a small white
spine on each thumb.
Measurement Data: Sample sizes are
minimal for discriminant function analyses (note
that no males are available for Area I). Results using
the complete model and forward step options of
discriminant function analyses of the measurement
data for males and females are quite similar using
Systat 10 software (Engelman, 2000). The forward
step model does slightly better in terms of
classification; scores for Wilks’ lambda, Pillai’s
trace, Lawley-Hotelling trace; and cumulative
proportion of total dispersion of the canonical
functions. Therefore, the forward step procedure
is the model used for the L. pentadactylus Unit
measurement data.
Both the male and female discriminant
function analyses perform rather poorly in
separating the geographic area data. For the male
data, 63% of the sample was correctly identified in
the posterior classification matrix, with only 42%
of the sample correctly identified using the posterior
jackknifed classification procedure. Forty four
percent of the total dispersion is accounted for with
the first canonical function, 82% by the first and
second, and 100% is accounted for with the first
through fourth canonical functions. For the female
data, 49% of the sample is correctly identified in
the posterior classification matrix, with 38% of the
sample correctly identified using the posterior
jackknifed classification procedure. Sixty four
percent of the total dispersion is accounted for with
the first canonical function, 89% with the first and
second, and 100% is accounted for with the first
through fifth canonical functions.
The plot of the first against second canonical
variables (Figure 8) indicates that the female data
perform slightly better in discriminating among
Area samples because the envelope for the
Figure 8. Discriminant function plot for measurement data, Leptodactylus pentadactylus Unit geographic areas. A = male data;
B = female data. Sample confidence ellipses p:0.683. A-I symbols in legend on figure represent Areas A-I; the symbol for R
represents remeasurement data for a single specimen (see text for further explanation).
Vol. 37(3), 2005
remeasurement data is smaller in the female plot
(Figure 8B) than for the male data (Figure 8A).
There is no meaningful separation of Area groups
for the male data (Figure 8A). There is modest
separation of Areas G and H from the rest of the
Area samples in the female data along the first axis,
which is typically size-related (Figure 8B).
However, the raw size data (Table 10) would
indicate that Area samples G, H, I would cluster
together if based only on size. Area F demonstrates
some distinctiveness from Areas A and C on the
second axis (Figure 8B), although the size of the
remeasurement envelope suggests that this
separation as well as the separation on the first axis,
are not robust and the discrimination among area
results would likely change considerably if new data
were added to the analyses.
Life Colors: Middle America. Honduras.
“Color in life of a female (USNM 319943): dorsum
of head and body Drab (27) with slightly darker
brown spots; dorsolateral ridges Drab (27) with
slightly darker brown pigment along lower edges;
supratympanic folds Drab (27), outlined below by
dark chocolate brown; dorsal surface of forelimbs
mottled Drab (27) and slightly darker brown; dorsal
surface of hind limbs Drab (27) with darker brown
crossbars on thighs and shanks; upper lip with dark
brown bars; flanks with large Peach Red (94) ovoid
blotch; flank glands Pale Horn Color (92); ventral
surfaces mottled Drab (27) and Buff (124); iris pale
brown with dark brown reticulations on upper half,
lower half dark brown. Color in life of a juvenile
(LSUMZ 33648; SVL 45.0 mm): dorsal surface of
body between dorsolateral folds rust-red with dark
brown markings; dorsal surface of head rust-red
with dark brown-edged interorbital bar; dorsal
surfaces of limbs dark grayish green with dark
crossbars; side of head pale copper with dark brown
markings; flanks grayish green; venter dark gray
with numerous small white spots; iris dark copper
with fine black reticulations (McCranie and Wilson,
2002:452).”
Nicaragua. “All of the specimens but the
largest one have the ventral surface dark brown
spotted with white – the characteristic coloration
of the immature. In all of the specimens there was,
in life, a considerable amount of red on the thighs
(Noble, 1918).”
Costa Rica. Roy W. McDiarmid
unvouchered photos from the Osa Peninsula show:
(1) a juvenile with a mostly chocolate brown
287
dorsum that is sharply demarcated from the gray
flanks by the dorsolateral folds; the side of the head
is tan and black (posterior thighs not visible); (2)
an adult with a dorsum of lighter and darker brown
markings, the side of the head is tan/brown, there
is a light cream post-commissural stripe, the flanks
are brown (ventral-most flanks and posterior thighs
not visible; these same color patterns are
demonstrated in two color photos in Savage (2002,
Plates 86,87), which also do not show the ventralmost flanks or posterior thighs.
Panama. “Groin and anterior surfaces of
thighs with red orange color; … throat grayish
brown with white spots; chest and venter white to
pale yellow and mottled with gray; posterior
surfaces of thighs red orange with black
reticulations, in the inferior part these reticulations
enclose white spots (Ibáñez et al., 1999).”
West Coast South America. Colombia.
Dorsally, a broad cream band from tip of snout to
vent, delimited on head by broad dark brown band,
the line that extends to posterior extent of tympanum
is without border. Large broad band-like spots dark
brown, rest of dorsum reddish brown. Anterior
surface of thigh brown, posterior red with
continuous irregular dark brown spots. Shank and
tarsus olive brown. Concealed surfaces of shank
and tarsus red. Ventrally: throat and chest brown
with cream spots. Flanks pink. Iris yellow bordered
above with bronze, bordered below by copper color;
these colors separated by a reddish brown spot.
(Color notes for ICN 16663 provided and translated
by John D. Lynch.)
Ecuador. A photo of MCZ A 91719 from
Pichincha Province shows a dorsum of darker and
lighter browns, the side of the head is tan with bold
black markings, the flanks are tan and brown
(posterior thighs not visible). Posterior thighs with
bright red marks (Rafael de Sá color photo, QCAZ
17056).
Amazonia. Peru. There are 10 photos/color
descriptions available. The variation is encompassed
with the following three characterizations. USNM
538196: “Dorsum light gray brown with darker
transverse brown bands. Laterally opalescent gray
with red tinge. Iris gold. Black band from snout
through eye to behind tympanum. Legs gray with
brown cross bars. Venter brown with white spots
and white mottling. Lips with black bars.” Photo of
USNM 317518: Dorsum olive drab and brown,
lower lip tan-brown, flanks brown above and yellow-
686
1030
439
424
460
793
342
347
342
690
-
49.5
-
0.18-0.19
0.24-0.28
23
15
251, 17, Brazil, Amazonas 2, –
26.0
3
4
254, 3, Brazil, Amazonas 1, –
26.0
1031
1029
454
456
787
796
169
170
860
861
64.6
43.5
16
13-18
0.18-0.29
0.38
12
15
2
104, 18, Peru, M. de Dios 1, –
–
3
72, 7, Peru, M. de Dios 1, –
–
691
863
450
508
438
705
343
369
442
690
50
-
12
0.24
0.20-0.23
37
4
–
64, 7, Peru, M. de Dios 1, –
23.2
7
1
115, 5, Colombia, Meta, –
863
860
415
420
613
629
312
173
686
549
26.8
22.8
5-8
6-7
0.20-0.24
0.20-0.30
34
–
37
–
5
103, 10, Ecuador, Pichincha 1, – 4
98, 5, Ecuador, Pichincha 1, –
514
340
459
466
248
142
214
171
343
301
31.1
45.7
5-6
11-13
0.34-0.42
0.24-0.28
49
40
–
8
10
89, 6, Costa Rica, Puntarenas, –
BCI CD, Panama, –
36.6
8-10
0.26-0.28
40
7
106, 1, Costa Rica, San José, –
–
Mean pulse
rate per s
# pulses
per call
Call rate/ Call duration
minute
range in s
Air temp, C
N
Tape & Cut,
Country, State, Vouchered
brown ventralmost, bits of exposed posterior thighs
are black with white markings. USNM 346146:
“Greenish brown-tan dorsum. Crossbands between
eyes and along back. Arms and legs gray with black
crossbars. Dorsolateral ridges edged with black.
Belly mottled tan and black.”
Brazil. “Dorsum reddish-brown, sides
grayish red, demarcated by dark brown dorsolateral
lines. Limbs gray with fine black lines. Very dark
brown line from snout through eye to arm insertion.
Sides of head below canthus reddish-gray. Lips with
dark brown and white stripes. All undersurfaces
white with gray reticulations. Eye copper.” (Janalee
P. Caldwell field notes for JPC 12294, a juvenile
female.)
French Guiana. “Iris red-copper below,
golden above. … Dorsal surface reddish-brown
with 5 or 6 transverse bands, those of juveniles are
white bordered on a grayer background. …
Posterior face of the thighs mottled white and black.
Venter marbled cream-white and black, especially
posteriorly.” (Crude translation of Lescure and
Marty, 2000, presumably based on specimens from
French Guiana.)
Advertisement Calls: None of the calls
analyzed has an associated voucher specimen. Most
of the recordings are both noisy (other signals in
the same broadcast channel as the frog calls) and
recorded from afar. Only two recordings have a
good signal to noise ratio, the BCI CD recording
from Panama and USNM recording 254 cut 3 from
Brazil. Certain call parameters are difficult to
measure accurately in most of the recordings. Only
three recordings have associated temperature data
(Table 11).
Interpreting the variation observed among
the available recordings is difficult. For example,
given the range in size of calling males
(100-195 mm SVL), one would expect there to be
a strong and significant correlation of carrier
frequency with size. Thus, the total variation
observed in carrier frequency might be accounted
for by size differences in calling males.
Common features of all calls include: (1)
each call consists of a single note; (2) calls have
increasing carrier frequency, especially in the first
half of the call; and (3) there appears to be frequency
modulation within calls associated with pulse
structure when pulses are distinct.
The temporal structure of the calls is
variable and complex. All of the calls except one
Frequency modulation, Hz
Mean dominant
Entire call
Mid-call pulse
frequency, Hz
Mean, start Mean, range
Mean, start Mean, range
342
171
162
437
346+515
Arquivos de Zoologia
Table 11. Advertisement call data for Leptodactylus pentadactylus Unit. Numbers associated with States indicate whether the data are for the same or different specific localities in the same State.
288
Vol. 37(3), 2005
(see next paragraph) have at least pulsatile structure.
The rate of pulse or pulsatile structure among the
calls spans the threshold at which human ears detect
sounds as either continuous or pulsed (~ 50 pulses
per s). As noted by Hero and Galatti (1990), the
calls of the two large Leptodactylus that occur
sympatrically in the central Amazon, sound
completely different to the human ear. However,
the frogs themselves do not necessarily make the
same kind of distinctions as humans do, because
frogs have much better temporal resolution than
humans have (Walkowiak, 1988:279, showed that
frogs can process discrete signals given at a rate of
least 100 times per s, whereas Greenewalt,
1968:102, indicated that humans can differentiate
a trill or vibrato if the rate does not exceed 30 times
per s). Frogs could be expected to perceive the calls
as different in degree, not in kind.
The calls Schneider et al. (1988) described
and figured from Reserva Ducke and near Manaus,
Amazonas, Brazil are different from all other calls
analyzed in lacking pulse or pulsatile structure
(Tables 11, 12, see also Schneider et al., 1988
figs. 7-8, p. 84). Schneider et al. (1988) also
indicated that the calls they analyzed started at
170 Hz, but their figure of the audiospectrogram
(fig. 8) exhibits no energy below about 300 Hz. The
Schneider et al. (1988) calls contrast with the calls
of Leptodactylus pentadactylus reported by
Zimmerman and Bogart (1984) recorded from
Reserva Ducke and the INPA-WWF Reserves north
of Manaus, Amazonas, Brazil. The calls and data
analyzed by Zimmerman and Bogart (1984) match
the variation observed in the other recording data
(Tables 11, 12) and contrast with the call described
by Schneider et al. (1988) in terms of pulse/pulsatile
structure. The seemingly obvious explanation is that
the Schneider et al. (1988) recording represents a
different species from those represented by all other
known recordings of the L. pentadactylus Unit.
289
There is no voucher specimen for the recording
published by Schneider et al. (1988). No specimen
was collected. Walter Hödl (pers. comm.) only saw
one individual calling from 2 m down in a large
burrow. The identification was based on information
from W. Magnusson and J.-M. Hero that calls
obtained from such burrows and lacking the audibly
pulsed component of L. knudseni were from
L. pentadactylus. It is impossible to address this
problem by comparing specimens. Therefore, the
Schneider et al. (1988) recording is considered to
be an anomalous recording, probably representing
a different species than those represented by the
Zimmerman and Bogart (1984) recordings.
Call temporal organization varies among the
calls as follows (refer to Table 11 for call recording
associations): (A) the two calls from Costa Rica
are pulsatile from beginning to end; (B) the call
from Panama starts weakly partially pulsed
followed by two pulses followed by either (i) one
long unpulsed portion or (ii) a weakly partially
pulsed long pulse, followed by a terminal pulse that
is itself weakly partially pulsed; (C) the two calls
from Pichincha Province, Ecuador are pulsed
throughout; (D) the call from Colombia has a
mixture of pulsed and partially pulsed structure at
the beginning and end of the calls with a middle
unpulsed portion; (E) the three recordings from Peru
range from calls with a mixture of pulses and partial
pulses throughout to pulsed at the beginning and
end with a partially pulsed central portion; (F) calls
of USNM recording 254 cut 3 from Brazil are
essentially unpulsed from the beginning until the
very end, which is partially pulsed; (G) calls of
USNM recording 215 cut 17 from Brazil are
partially to not pulsed at the beginning and partially
to completely pulsed for the rest of the calls.
The pulse rates for calls were determined
for all pulses after the initial pulse of each call
except for the pulse rate in the Panama recording
Table 12. Published advertisement call data for Leptodactylus pentadactylus Unit. The Brazilian data are from the same localities
in part. Asterisk indicates mean value.
Publication; Country,
State; Vouchered
Schlüter, 1980; Peru,
Huanuco; ?
Zimmerman & Bogart,
1984; Brazil, Amazonas; ?
Schneider et al., 1988;
Brazil, Amazonas; –
N individuals/calls
Temperature
Call rate/
Call
# pulses
minute duration per call
range in s.
26
0.3-0.4
?
?
26
7/59
?
4-33
0.16-0.25
?/many
Includes
22.8-23.8
28-34
0.16-0.19*
Beginning
call
frequency
200-300?
Ending
call
frequency
700-800?
Dominant
frequency
?
700-830
880-1020
770-920
1
170
(~300?)
880
?
800?
290
Arquivos de Zoologia
which was based only on the two distinct pulses
following the initial call pulse. All calls from
Amazonia sound continuous to the human ear; all
the other calls sound pulsed.
Sympatric species of the Leptodactylus
fuscus group are characterized as differing by at
least an order of magnitude in either a temporal or
frequency component (e.g. Heyer, 1978, Straughan
and Heyer, 1976). Even though Galatti (pers.
comm.) has demonstrated that the Panamanian and
Amazonian populations currently identified as
L. pentadactylus represent two distinct species, the
recordings for the Middle American and Amazonian
populations do not differ from each other by an
order of magnitude in any feature analyzed (Tables
11, 12, with exception of the Schneider et al. 1988
data noted previously).
Taking the kinds of differences among calls
from Middle America and Amazonia as guidelines,
the recordings from Costa Rica differ somewhat
from the Panamanian recording. The Middle
American recordings are different from the western
Ecuador recordings; the western Ecuador recordings
are quite distinctive from all other recordings. The
three recordings from Peru and USNM recording
254 cut 3 from Brazil appear to be similar to each
other and differ somewhat from the other Amazonian
calls from Colombia and Brazil (USNM recording
251 cut 17), while the latter two recordings are
somewhat similar to each other.
The published data by Schlüter (1980) and
Zimmerman and Bogart (1984) for Amazonian calls
from Peru and Brazil respectively differ somewhat
from the other recordings analyzed (compare Tables
11 and 12). The differences could be due to different
analytic techniques. Given that possibility, it is best
to conclude that the calls are those of the same
species represented by the other Peruvian and
Brazilian specimens analyzed (Table 11).
Finally, the Middle American calls of the
L. pentadactylus Unit are extremely similar to the
calls of the L. knudseni Unit (compare Tables 6 and
11).
Evaluation of Differentiation: There is
considerable variation within the L. pentadactylus
Unit in terms of size, patterns, male secondary
sexual characteristics, and advertisement calls. The
relationship between observed variation and species
limits is not clear, however.
The variation among specimens from
Amazonia is interpreted as intraspecific (with the
possible exception of the specimens represented by
the recording of Schneider et al., 1988). Small
sample sizes preclude a firm understanding of the
nature of the geographic variation involved, but the
data do indicate that a geographic component exists.
There is also variation among the Middle
American samples of the L. pentadactylus Unit. The
differences between Area A and B individuals could
be due to sampling problems. For example, there
is only one large male from Area A and the fourth
pattern type of the posterior thigh (Table 7) is
represented by only one individual from Area B.
The most striking difference between the Area C
and D samples is that more individuals from Area
D have dark bellies with rather discrete, regular light
spots than found in the Area C sample. There are
no discrete differences either between Area C and
D samples or among Area A, B, C, D samples,
which is consistent with the hypothesis that
variation in the Middle American samples
represents intraspecific geographic variation and
is not indicative of species level differentiation.
Thus, the Middle American samples and
Amazonian samples are considered to each
represent a single species distinct from one another.
The Area E and F samples are more similar
to the Middle American species than the Amazonian
species, especially with respect to advertisement
calls (based on the few recordings analyzed).
However, there is variation within Areas E and F
that needs to be evaluated at a finer scale than
analyzed above.
The more detailed analysis of Area E and F
samples was done in three steps.
The first step was to combine the data for all
specimens from Costa Rica and Panama (except
for the two Panamanian samples that were included
in the Area E sample) as one sample. A second
sample consisted of the specimens from two
localities from the Departamento de Magdalena,
Colombia. The third sample comprised specimens
from coastal Ecuador.
The second step was to compare the data among
these three samples to determine whether any two
of them were identical (when taking sample sizes
into account). Each of the three samples differ from
each other, but it is not clear from the data whether
the differences represent sampling problems or
geographic variation, particularly between the
Vol. 37(3), 2005
Costa Rica/Panama and Magdalena, Colombia
samples.
The third step consisted of comparing the data for
all other locality samples from Area F with the three
samples defined above, starting with the samples
closest to the coastal Ecuador sample and
proceeding northward.
The results are modestly surprising but
geographically coherent. There is a morphological
break between the coastal Ecuador sample and the
specimens from the west coast of Colombia. Many
of the West Coast Colombia specimens have
extensive light areas on the posterior thighs, a
pattern not found in any specimen from the Coastal
Ecuador sample. All males from West Coast
Colombia have either tiny or small white or tan
thumb spines while the two males from Coastal
Ecuador have a large white or black spine on each
thumb. The data on specimens from near Valdivia,
Antioquia, Colombia (ICN 9934, 9936) match the
data for the Magdalena, Colombia sample. The
specimen data from the two Panamanian localities
previously included in Area E and the data for the
specimen from near Turbo, Antioquia, Colombia
bridge the differences between the Costa Rica/
Panama and Magdalena, Colombia samples. The
combined data for the specimens from Costa Rica/
Panama/Antioquia, Colombia/Magdalena,
Colombia in turn are distinct from data from the
West Coast Colombia specimens. The thigh pattern
difference described for the Coastal Ecuador and
West Coast Colombia specimens exists between the
more northern specimens and the West Coast
Colombia specimens. The more northern specimens
have well developed chest spines, throat and chest
tubercles, and a large thumb spine on each thumb,
differing from the conditions observed in the West
Coast Colombia specimens.
Thus, there are three geographically
coherent Units involved: (1) Specimens from
Honduras, Nicaragua, Costa Rica, Panama, and
northern Colombia in the general Isthmus of
Panama region; (2) specimens from west coast
Colombia; and (3) specimens from coastal Ecuador.
The specimens from coastal Ecuador differ
somewhat in advertisement calls from the Middle
American and northern Colombia Unit as discussed
previously. In addition, one male from coastal
Ecuador (USNM 196745) has throat and chest
291
tubercles but no chest spines, indicating that this
Unit in fact lacks chest spines, while the Middle
American and northern Colombia specimens do
have chest spines. After this analysis was
completed, specimens examined in Ecuadorian
collections (EPN, QCAZ) in March 2002 extended
the distribution of the west coast Colombia Unit
into adjacent Ecuador. The additional data do not
add information to the morphological analyses and
are not added to the data for the Inter-Unit analysis
in the next section. The measurement data are
included in the Inter-Unit analysis.
The data are most consistent with
recognizing four differentiated population
groupings: (1) Middle America and northern
Colombia; (2) West Coast Colombia; (3) West
Coast Ecuador; and (4) Amazonia.
Leptodactylus labyrinthicus Unit
Geographic Areas – Eight geographic areas are
distributed broadly throughout the total range of
localities sampled for this study (Figure 9).
Variation – Dorsal Pattern: The available data do
not demonstrate any differences between juveniles
and adults or between males and females. There is
considerable variation in dorsal pattern, but the
entire range of variation occurs within the Area A
sample. There is no obvious variation in dorsal
pattern among the Area samples. Dorsal patterns
include uniform light, uniform dark, a few small
scattered marks, indistinctly defined dark blotches,
broad dark almost coalescing bands anteriorly with
large dark blotches posteriorly, and a broad dark
irregular interorbital bar and two large dark spots
in the shoulder and sacral regions either light
outlined or not.
Belly Pattern: The only specimens lacking belly
patterns from Areas F and H are juveniles and one
male from Area A lacks a belly pattern. Otherwise,
the available data do not demonstrate any
differences between juveniles and adults or males
and females. There is notable variation in belly
patterns within Area samples, but there is some
variation among Area samples that is not likely
explained by relatively small sample sizes (Table
13). All non-Area specimens do not differ from the
specimens in the most proximate geographic Area
with the exception of MZUSP 24936 from Serra
292
Arquivos de Zoologia
Figure 9. Geographic areas for Leptodactylus labyrinthicus Unit. See text for explanation.
de Parima, Roraima, Brazil, which has a distinct
labyrinth pattern (Figure 13D) not found in
specimens from Area A, but found in all other Areas.
Lip Pattern: The available data do not demonstrate
any differences between juveniles and adults or males
and females. A pattern of the upper lip with broad
alternating dark and light vertical bands, distinct
above the lip and fading as they reach the eye level
occurs in some specimens from all Areas. Beyond
that, there is variation both within and among Area
samples (Table 14). Non-Area specimen USNM
146521 from near Buena Vista, Santa Cruz, Bolivia
has a dark lip with two distinct, narrow light stripes
from the lip to the anterior and posterior eye (Figure
12B), a pattern not found in specimens from the
nearest geographic Area B. Several specimens from
Cachimbo, Pará, Brazil have the same
aforementioned lip pattern, which was not found in
specimens from proximate Area C. Specimens from
Araguatins and São Domingos, Goiás, Brazil, have
an upper lip with broad alternating dark and light
bands from the lip to the eye level, a pattern not found
in specimens from proximate Area G.
Thigh Pattern: The only two individuals with finely
mottled or indistinctly mottled thigh patterns are
small juveniles (23-30 mm SVL). Late Gosner stage
larvae have indistinct thigh patterns, suggesting that
post-metamorphic individuals retain the larval
pattern before developing the patterns found in
larger juveniles and adults. Otherwise, the available
data do not demonstrate any differences between
juveniles and adults or males and females. Some
specimens from all Areas have boldly mottled
thighs. There is also variation within and among
Area samples for additional thigh patterns (Table
15). Non-Area specimen MZUSP 21746 has a
darkly mottled thigh with distinct large light spots
extending from light dorsal bands and the thigh of
MZUSP 21747 exhibits very large contrasting dark
and light markings with a large central dark spot
that resembles a large eye. Both of these specimens
are from São Luis, Maranhão, Brazil and the
patterns do not occur in proximate Area G.
Specimens from Santo Amaro das Brotas, Sergipe
and São Miguel dos Campos, Alagoas, Brazil have
the pattern described for MZUSP 21746 (above),
which pattern was not encountered in specimens
from proximate Areas G and H.
Shank Pattern: The available data do not
demonstrate any differences between juveniles and
adults or males and females. The variation within
Area samples is comparable to inter-Area variation.
Patterns include uniform dark, a few small dark
Vol. 37(3), 2005
293
Table 13. Belly patterns, Leptodactylus labyrinthicus Unit. For the definition of geographic areas, see Figure 9.
Major pattern types
1. Virtually no pattern
2. Mottled (Fig. 13A)
3. Light with scattered dark irregular marks
4. Dark with large light irregular marks (Fig. 13 C)
5. Distinct labyrinth pattern (Fig. 13D)
A
X
X
B
C
X
X
X
X
X
X
X
X
Geographic Areas
D
E
F
X
X
X
X
X
X
X
G
X
X
X
X
H
X
X
X
X
Table 14. Lip patterns, Leptodactylus labyrinthicus Unit. For the definition of geographic areas, see Figure 9.
Major pattern types
1. No light or dark stripes/bars from upper lip to eye
2. Upper lip dark with two distinct, narrow light stripes
from lip to anterior and posterior eye (Fig. 12B)
3. Upper lip with broad alternating dark and light bands
from lip to eye (Fig. 12D)
4. Upper lip with broad alternating dark and light bands,
distinct above lip, fading as reaching eye level
5. Upper lip with narrow dark bands from lip to eye
level (Fig. 12C)
A
X
X
X
B
X
C
X
Geographic Areas
D
E
F
X
X
X
X
X
X
X
X
X
X
X
G
X
X
H
X
X
X
X
X
H
X
X
Table 15. Thigh patterns, Leptodactylus labyrinthicus Unit. For the definition of geographic areas, see Figure 9.
Major pattern types
1. Bold mottle
2. Dark mottle with large light spots extending from
darker dorsal bands (Fig. 14 B)
3. Mostly (and obviously) light, with or without a few
dark irregular marks (Fig. 14 D)
4. Distinct labyrinth pattern
A
X
X
dots, or (usually) dark broad transverse bands that
are either sharply or roughly defined and extend
the entire width of the upper shank surface or about
two-thirds the distance starting from the outer edge.
Body Folds: Lateral fold conditions were
recorded for relatively few specimens. The most
data are available for Area G individuals (N = 7)
and the conditions range from none visible,
interrupted, to continuous and complete.
Most specimens have interrupted dorsolateral folds
that extend from the eye to at least 1/2 the distance
toward the sacrum to the full distance toward the
sacrum. There is some additional variation in
dorsolateral fold development that does vary among
geographic areas (Table 16). Non-Area specimens
do not differ from dorsolateral fold conditions found
in the proximate Area sample(s).
There is considerable variation in the extent of
expression of flank folds, with variation in degree
B
X
C
X
Geographic Areas
D
E
F
X
X
X
X
G
X
X
X
X
X
X
X
X
of development among Areas (Table 17). All
specimens (N = 4) from Alto Palmar, Cochabamba
and near Buena Vista, Santa Cruz, Bolivia have no
discernible flank folds, differing from proximate
Area B specimens. All other non-Area specimens
match flank fold conditions found in proximate Area
sample(s).
Male Sexual Secondary Characteristics:
Three large males from Area C have weakly
hyptertrophied arms. At least some males from all
other Areas have pronounced arm hypertrophy and
some males from Areas A, D, E, F, and H have arms
as extensively hypertrophied as can occur in
Leptodactylus.
Data were not recorded for development of
tubercles for males from Area C. At least some
males in all other Area samples have a field of black
or tan-tipped tubercles on the throat, chest, and
anterior belly. One male from Area F (USNM
121284) has black tubercles only on the throat, as
294
Arquivos de Zoologia
Table 16. Dorsolateral fold conditions, Leptodactylus labyrinthicus Unit. Parentheses indicate interrupted folds; no parentheses
indicate continuous folds. For the definition of geographic areas, see Figure 9.
Major character states
1. None
2. Extend from 1/4 to less than 1/2 the distance from
eye to sacrum
3. Extend from 1/2 to the full distance from eye to
the sacrum
4. Extend from the eye to beyond the sacrum
A
X
B
C
(X)
(X)
Geographic Areas
D
E
F
X
G
H
X, (X)
(X)
X
(X)
(X)
(X)
X, (X) X, (X)
(X)
Table 17. Flank folds, Leptodactylus labyrinthicus Unit. For the definition of geographic areas, see Figure 9.
Degree of development
1. None
2. Elongate wart in flank fold field
3. Interrupted, complete
4. Continuous, complete
A
X
X
does non-Area specimen USNM 146507 from Alto
Palmar, Cochabamba, Bolivia.
Some adult males lack chest spines (presumably
the males were collected during the nonreproductive season). Most Area males have a pair
of well-developed black (usually), tan-tipped, or
white chest spines.
One or two males from Areas C, D, E, F lack thumb
spines, which, as with chest spines, indicates the
specimens could have been collected in the nonreproductive season. Most males from all Areas
have a single white or black spine on each thumb.
Three males from Area A have an additional black
keratin tip on an expanded prepollex, forming a
smaller second spine on each thumb. Non-Area
specimen USNM 146507 from Alto Palmar,
Cochabamba, Bolivia has slightly larger prepollical
spines in addition to the medial thumb spines found
in three of the males from Area A. There are no
other differences between non-Area male thumb
spines and males from proximate Areas.
Shank Texture: The Area E sample (the
largest sample size), has no specimens that were
scored as lacking tubercles. The Area A, G, H
samples each have two to four specimens scored
as lacking tubercles. All Area samples have
individuals with few to many black and/or white
tubercles on the dorsal surface of the shank. Some
individuals from Areas A, C, D, E were scored as
also having a shagreened surface. Non-Area
specimen USNM 146507 from Alto Palmar,
B
X
X
C
X
X
X
X
Geographic Areas
D
E
F
X
X
X
X
X
X
G
X
X
X
H
X
X
Cochabamba, Bolivia was scored as lacking
tubercles, differing from the proximate sample from
Area B. There are no other noteworthy differences
between non-Area specimens with proximate Area
specimens.
Outer Tarsal Texture: Most individuals from Area
A and many individuals from Area H have smooth
outer tarsal textures; single individuals from Areas
D, F were scored as lacking texture. All Area
samples have individuals with few to several black
and/or white tubercles. Some individuals from
Areas B, D, E, G, H have many black and/or white
tubercles. Only one individual each from Area C
and E has a shagreened surface. Non-Area specimen
USNM 146507 from Alto Palmar, Cochabamba,
Bolivia was scored as lacking tubercles, differing
from specimens from proximate Area B. USNM
96978 from Belo Horizonte, Minas Gerais, Brazil
was scored as having many white tubercles,
differing from specimens from proximate Area F.
All other non-Area specimens do not differ from
proximate Area specimens.
Foot Texture: Most individuals from all Areas lack
tubercles on the sole of the foot. All Areas have a
few individuals with few to scattered black or white
tubercles. No individual was scored as having a
shagreened surface. All non-Area specimens have
conditions found in specimens from proximate
Areas.
Adult Size: Small sample sizes preclude
definitive statements regarding size variation (Table
Vol. 37(3), 2005
18). The following two points are suggested by the
available data. First, there may be some variation
in size among Areas as indicated by size differences
(albeit overlapping) between males from Areas A
and H, although males from Area E demonstrate as
much variation as occurs among all Areas (Table
18). Second, males apparently reach larger sizes
than females (Table 18). Non-Area specimen
USNM 21746 from São Luiz, Maranhão, Brazil is
an adult female of 120 mm SVL with large ova,
just smaller than any female from Areas A-H.
Measurement Data: Sample sizes are
minimally adequate only for male data from all
Areas. The forward step discriminant function
analysis of the male data did not have any variables
which meet the default F-value to enter (0.15) in
SYSTAT10 (Engelman, 2000), indicating that the
variation among Areas is not dramatically different
than the variation within Areas. The backward step
model only used two variables in the final model.
There are no clear indications that the backward
step model performs better than the complete data
model and vice versa. The complete data model
analysis is used for discussion.
The discriminant function analysis performs
rather poorly in separating the geographic area data.
Sixty-nine percent of the individuals are correctly
identified in the posterior classification matrix with
only 51% of the sample correctly identified using
the posterior jackknifed classification procedure.
Fifty-six percent of the total dispersion is accounted
for by the first canonical function, 75% by the first
and second canonical functions, 88% by the first,
295
second, and third canonical functions. All variation
is accounted for with the first through seventh
canonical functions.
Two variables have much higher loading
values on the first canonical function than the other
variables: head width and shank length. The two
variables with the highest loading values on the
second canonical variable are head width and eyenostril distance. The plot of the first against second
canonical variable (Figure 10) indicates essentially
no discrimination for all but one Area sample. The
males from Area F are distinctive in the analysis.
Life Colors: Lutz (1926, Plate 30, figs. 1,
2) and Bokermann and Sazima (1974) have color
illustrations of specimens with crimson red groins
and posterior thigh marks, but there are no locality
data for the illustrations. De la Riva et al. (2000)
and Gorzula and Señaris (1998) have color photos
of specimens from Bolivia and north coast
Venezuela respectively, but the aspects
photographed do not show any areas where flash
colors occur.
USNM 207674, an adult female from Assis, São
Paulo, Brazil (Area F) had the following colors in
life (WRH field notes): “Throat and chest white.
Belly and limb venters yellow. Thigh spots rust
brown dorsally and yellow ventrally. Dorsum rust
brown.”
Table 18. Adult sizes (SVL in mm), Leptodactylus
labyrinthicus Unit geographic areas. For the definition of
geographic areas, see Figure 9.
Geographic Area
A
B
C
D
E
F
G
H
Males
127-160
(N = 23)
122-170
(N = 9)
129-170
(N = 4)
110-182
(N = 5)
115-188
(N = 13)
118-163
(N = 5)
140-167
(N = 5)
135-180
(N = 10)
Females
122-128
(N = 2)
147-149
(N = 3)
124-155
(N = 5)
150-163
(N = 2)
132-166
(N = 7)
127-139
(N = 2)
127-167
(N = 9)
121-162
(N = 3)
Figure 10. Discriminant function plot for male measurement
data, Leptodactylus labyrinthicus Unit geographic areas.
Sample confidence ellipses p:0.683. A-H symbols in legend on
figure represent Areas A-H; the symbol for R represents
remeasurement data for a single specimen (see text for further
explanation).
430
273
309
180
258
23
4-5
0.17-0.19
54
10
233, 3, Brazil, Paraíba, photo
22.0
430
276
326
214
258
28
5-6
0.18-0.19
61
10
233, 2, Brazil, Paraíba, photo
21.0
429
277
320
207
257
31
4-6
0.14-0.17
56
10
233, 1, Brazil, Paraíba, photo
21.0
428
NA
NA
215
292
NA
1
0.19-0.21
50
10
229, 2, Brazil, São Paulo, –
23.0
429
NA
NA
171
0.18-0.21
35
10
229, 1, Brazil, Minas Gerais, –
24.0
1?
0.14-0.18
41
2
321, 3, Brazil, Mato Grosso, –
21.5
1
NA
258
?
?
?
130
258
386
4
Bolivia, Santa Cruz, ?
–
54
0.15-0.24
3-5
20
386
98
310
154
400
100
400
27?
9?
0.33
1
Venezuela, Sucre, ?
–
Mean pulse
rate per s
# pulses
per call
Call rate/ Call duration
minute
range in s
Air temp, C
The most striking differences in the calls are in
terms of call modulation. Six of the calls are
amplitude and/or frequency modulated to such a
degree that the calls sound distinctly pulsed to the
human ear; two of the calls lack amplitude
modulation within the call and they sound unpulsed
to the human ear (Haddad et al., 1988:16; Figure
11). Differences of this magnitude correlate with
species limits in other Leptodactylus species (e.g.,
Heyer et al., 1996). However, Marcos Gridi Papp
(personal communication, 12 December 2001)
informed me that individuals of Leptodactylus
ocellatus in the State of São Paulo, Brazil, give
pulsed and unpulsed calls at the same site. He did
not know if an individual gives both kinds of calls,
however. Clearly, this is an interesting situation that
begs for detailed study. It would seem as though
two alternatives are most likely: (1) two species
currently called L. ocellatus occur at some
localities in the State of São Paulo; or (2) there is
individual or intra-population variation in calls
including pulsed and non-pulsed calls. If the latter
is correct, that might apply to the observed
variation in Leptodactylus labyrinthicus as well.
The two recordings with unpulsed calls are from
N
Advertisement Calls: None of the calls
analyzed has an associated voucher specimen that
I have examined. Three of the recordings have
photo vouchers (not found in photo collection at
ZUEC) (Table 19). All recordings (for which data
elements are available) are similar in terms of call
rate, dominant frequency, and beginning frequency
(Table 19). The recordings of three individuals from
the same locality (Paraíba) demonstrate little
variation among themselves (Table 19).
Tape & Cut,
Country, State, Vouchered
Hödl (1993, fig. 3) has a color photo of a specimen
from João Pessoa, Brazil (Area H) showing a pale
yellow and dark gray reticulum on the hidden
portions of the shanks. Lutz (1926, Plate 30, figs. 3,
4) has a color illustration of a specimen from near
Independencia (now Guarabira, Paraiba fide P.E.
Vanzolini, pers. comm.), Brazil (Area H) with bright
yellow markings on the backs of the thighs and a
faint yellow wash only on the posterior belly and
hidden portions of the shanks. Lutz (1926)
considered this yellow-thighed frog to represent a
species (L. gigas) distinct from the rest of the
Brazilian material with red marked thighs that he
considered to be L. labyrinthicus.
Frequency modulation, Hz
Mean dominant
Entire call
Mid-call pulse
frequency, Hz
Mean, start Mean, range
Mean, start Mean, range
Arquivos de Zoologia
Table 19. Advertisement call data for Leptodactylus labyrinthicus Unit. Recordings from Paraíba are from the same locality. Bolivian data are from same recording reported on by Marquez
et al., 1995. Venezuelan call data are from Rivero and Esteves, 1969. NA = not applicable.
296
Vol. 37(3), 2005
297
Figure 11. Advertisement calls of Leptodactylus labyrinthicus Unit showing wave form and audiospectrogram. A = USNM
recording 229, cut 1 from Minas Gerais, Brazil; B = USNM recording 233, cut 3 from Paraíba, Brazil.
Area F. The pulsed recordings are from Areas A
and H and from a locality that is closest to Area B.
The recording from Area B is equivocal relative
to pulse structure due to the poor quality of the
recording.
In addition to the striking pulse differences
among calls, there is additional variation. The call
from Area A (Pico del Mango, Sucre, Venezuela,
Rivero and Esteves, 1969) is notably longer (and
probably has more pulses as a consequence) than
all other recordings of calls except for the recording
from Puerto Almacén, Santa Cruz, Bolivia. The
frequency sweep of the recordings associated with
Areas A and B is notably less than that seen in the
other recordings (Table 19).
Evaluation of Differentiation: There is
considerable variation within the Leptodactylus
labyrinthicus Unit in terms of patterns,
measurement data, and advertisement calls.
The advertisement call data are consistent with
recognizing the specimens from Area F as distinct
from all other specimens (although see previous
section). In addition, a particular lip pattern is found
only in some individuals from Area F and Area F
male measurement data are distinctive from all other
male measurement data in a discriminant function
analysis (Figure 10).
The Area A sample demonstrates some
differentiation in terms of belly pattern, male thumb
spine development, and perhaps advertisement
calls. The Area A sample appears to represent a
geographic isolate. As indicated previously, (Heyer,
1979), the Leptodactylus labyrinthicus Unit occurs
in open formations. However, the Area A sample
occurs in humid forest or partially cleared humid
forest habitats (Péfaur and Sierra, 1995).
Interestingly, however, there are no Leptodactylus
labyrinthicus Unit members known from the
Venezuelan llanos (Péfaur and Sierra, 1995). The
only other Area samples demonstrating regional
differentiation are the Area G and H samples. These
samples have individuals that share a thigh pattern
unobserved elsewhere.
298
Arquivos de Zoologia
The available data are most consistent with
recognizing four differentiated population systems:
(1) Leptodactylus labyrinthicus Area A; (2)
L. labyrinthicus Areas B-E; (3) L. labyrinthicus
Area F; and (4) L. labyrinthicus Areas G-H.
Inter-Unit Analysis
Morphological Data
The morphological data represent the most
extensive type of data available for the frogs in this
study. These data are used as a first step in analyzing
the nature of variation among the differentiated
Units recognized in the previous section together
with data for Leptodactylus myersi.
The intra-unit analyses resulted in most
variation among differentiated Units occurring in
lip pattern, belly pattern, thigh pattern, dorsolateral
folds, flank folds, male thumb spines, and male
chest spines. These characters were re-summarized
for the differentiation Units recognized in the
previous section and for L. myersi. Additional data
were obtained for L. knudseni from Area B in March
2002 and are included in the Inter-Unit Analysis.
Character state distributions were
summarized for each differentiated Unit. In
addition, character states were summarized for
specimens that are from localities between the
differentiated Units. In all cases but one, the
distribution of character states for specimens from
localities between differentiated Units
unambiguously associated with one of the
differentiated Units and not the other. After
additional data for Ecuadorian L. knudseni were
taken in March 2002, the specimens from northern
Amazonian Peru could not unambiguously be
associated with the samples from (more
geographically proximate) Ecuador or southern
Amazonian Peru. The data for these north
Amazonian Peru L. knudseni are omitted from the
Inter-Unit Analysis.
Upon completion of the summaries of
character state distribution, the range of variation
was assessed to recognize major character states,
described as:
Lip Patterns. (1) Uniform light, no pattern,
(2) dark triangular marks on edge of upper lip only,
(3) dark triangular marks on edge of upper lip, one
or two elongate and approaching or reaching the
lower eye border (Figure 12A), (4) elongate dark
triangular marks on upper lip edge, fading towards
eye, two dark marks under eye, (5) dark with two
narrow, light chevrons from lip entering eye (Figure
12B), (6) alternating broad light bands and narrow
dark vertical stripes, 1-3 dark stripes entering eye
(Figure 12C), (7) broad irregular darker and lighter
bands, two dark bands entering eye (Figure 12D),
(8) broad light stripe, regular above, regular or
irregularly defined below (Figure 12E), (9) uniform
dark.
Belly Patterns. (1) Light belly, without
pattern, (2) belly light with lateral and anterior
mottling, (3) belly mottled or uniform dark (Figure
13A), (4) belly dark with small light vermiculations
(Figure 13B), (5) belly dark with large light
vermiculations (Figure 13C), (6) belly light with
dark vermiculations, (7) labyrinthine pattern (Figure
13D), (8) dark with small light discrete spots, (9)
belly dark with large light discrete spots (Figure
13E).
Posterior Thigh Patterns. (1) Very dark and
uniform, (2) relatively uniformly dark or indistinctly
mottled, (3) dark with small distinct light
vermiculations (Figure 14A), (4) dark with large
light vermiculations, some coalescing of
vermiculations, (5) labyrinthine (same pattern as
for Figure 13D) (6) dark with distinct, discrete light
spots, (7) dark with contrasting large light irregular
blotches/spots extending from light dorsal
transverse bars, rest of thigh dark with light
vermiculations (Figure 14B), (8) very dark irregular
blotches highlighted by light pin stripe on dark
background (Figure 14C), (9) mostly very distinctly
light with few irregular dark marks (Figure 14D).
Dorsolateral Folds. (1) None, (2) entire or
interrupted fold extending from eye to 1/4 distance
to sacrum, (3) interrupted fold extending from eye
to more than 1/4 distance to 1/2 distance to sacrum,
(4) interrupted fold extending more than 1/2
distance to sacrum or to the sacrum, (5) entire fold
extending from eye to at least 1/2 distance to sacrum
or to sacrum, (6) interrupted fold from eye to past
sacrum, (7) entire fold extending from eye to past
sacrum but not to leg, (8) entire fold extending from
eye to leg.
Flank Folds. (1) None, (2) elongate dark
spot in flank fold field, (3) interrupted fold to
shoulder region, (4) entire fold to shoulder region,
(5) interrupted complete fold, (6) entire, complete
fold.
Vol. 37(3), 2005
Figure 12. Lip patterns. A – Dark triangular marks, one or two
elongate approaching or entering lower eye; B – Dark with
two narrow light chevrons from lip entering eye; C – Alternating
broad light bands and narrow dark vertical stripes, 1-3 dark
stripes entering eye; D – Broad irregular darker and lighter
bands, 2 dark bands entering eye; E – Broad light stripe regular
above, regularly or irregularly defined below.
Figure 13. Belly patterns. A – Mottled; B – Dark with small
light vermiculations; C – Dark with large light vermiculations;
D – Labyrinthine; E – Dark with large, light, discrete spots.
Figure 14. Posterior thigh patterns. A – Dark with small, distinct
light vermiculations or vermiculations and spots; B – Dark with
contrasting large light irregular blotches/spots extending from
light dorsal transverse bars, rest of thigh dark with light
vermiculations; C – Very dark irregular blotches highlighted
by light pin stripe on dark background; D – Mostly very
distinctly light with few irregular dark marks.
299
Male Thumb Spines. (1) Absent, (2) one tiny
to small, (3) one large white or black, (4) one large
black and second smaller bump.
Male Chest Spines. Male chest spines are
completely deciduous seasonally and may leave no
trace of their former presence in preservation, so
the state deemed to be characteristic of sexually
active males is defined for each differentiated Unit
in terms of whether chest spines are present or
absent.
The frequency distribution of states for all
characters except chest spines are presented in
Tables 20 and 21. Presence and absence of chest
spines is presented in Table 21. There is
considerable variation in relative and absolute
differences in distribution of states among the
differentiated Units.
The focus of the remainder of this section is
to evaluate differentiation in terms of delineating
species boundaries.
The first step is to evaluate the
morphological data (Tables 20, 21) in terms of
levels of differentiation among the differentiated
Units. The data from Tables 20 and 21 were
compared pairwise to determine the number of
characters that completely diagnose the pair from
one another and the number of characters for which
some states are unique to one or the other
differentiated Unit being compared. For these latter
distinctive, but not completely diagnostic
characters, an evaluation of whether the unique
states could be due to small sample sizes was
assessed by visual examination of the data. For
example, if Sample A of 40 specimens has State 1
in 39 individuals and State 2 in one individual and
Sample B of 3 specimens has only State 1, the
differences are considered likely to be due to the
small sample size of Sample B. However, if the 40
specimens of Sample A all have State 1 and Sample
B contains two specimens with State 1 and one
specimen with State 2, these differences are
considered to be meaningful as they are probably
not due to sampling problems.
Table 22 shows the results of the numbers
of diagnostic and distinctive characters of pairwise
comparisons for the differentiated Units. The total
number of characters involved in the comparisons
is seven. Two conclusions can be drawn from this
analysis: (1) there is some level of distinctiveness
among all differentiated pairs for the morphological
data involved, and (2) no differentiated Unit is
Lip pattern
Patternless, uniform light
Dark triangular marks on edge of upper lip only
Dark triangular marks, one or two elongate approaching or
entering lower eye
Elongate triangular marks dark on lip edge, fading towards eye,
two dark marks under eye
Dark with two narrow light chevrons from lip entering eye
Alternating broad light bands and narrow dark vertical stripes,
1-3 dark stripes entering eye
Broad irregular darker and lighter bands, 2 dark bands entering eye
Broad light stripe regular above, regularly or irregularly defined below
Uniform dark
Belly Pattern
Patternless (light)
Lateral and anterior mottling
Mottled or uniform dark
Dark with small light vermiculations
Dark with large light vermiculations
Light with dark vermiculations
Labyrinthine
Dark with small, light, discrete spots
Dark with large, light, discrete spots
Posterior Thigh Pattern
Distinctly uniform dark
Relatively uniform dark or indistinctly mottled
Dark with small, distinct, light vermiculations or vermiculations
and spots
Dark with large, distinct, light vermiculations, some coalescing
of vermiculations
Labyrinthine
Dark with distinct, discrete light spots
Dark with contrasting large light irregular blotches/spots
extending from light dorsal transverse bars, rest of thigh dark
with light vermiculations
Very dark irregular blotches highlighted by light pin stripe on
dark background
Mostly very distinctly light with few irregular dark marks
1
5
2
1
4
29
3
8
10
2
8
5
14
17
2
3
1
5
1
4
3
2
52 14
3
9
24
2
7
2
5
4
9
11
2
4
4
1
6
3
25
2
1 15
2
3
7
9
1
1
1
8
1
2 12
1
3
2
1
3
2
8
1
1
6
1
3
2
3
knud
B
J A
2
7 22
1
A
J
J
A
para
myersi
6
1
4
4
1
3
2
3
9 21
8
8 52
1 4
8 4
5 2
2
1
8
4
2
12
7 2 6
4 20 60
2
1
1
2
1 1
6 13
1 10 24
2 6
8 13 15 45
1
4
4
1
2
4
3
8
1
2
17
21
1
5
7
12
3
knud knud
labyr
C
A, D-H
A
J A J A J A
3
1
8
4
3 24
2
9 49
5
1
6
2
2
5
2
4
1 12
2
4 22
4 15
5
1 1
5 7
9 36
2
7
1
7 20
1 2
7 19
10
2 11
14 50
1
3
1
3
pent
1
J A
2
4
3 39
1
15
9 19
5 4
1 24
2
1
8
3 30
3 8
1
3
14
2 13 62
labyr
G+H
J A
10
1
1
6
5
labyr
F
J A
7
1
9 34
3
4 30
1
9
1
labry
B-E
J A
1
3
1
1
1
1
4
5
1
1
1
1
2
7
9
pent
2
J A
1
1
1
2
1
3
pent
4
J A
3
2
6
30 29
4 12
2
5
1
4
1
5
1
1
2
1
1 32 43
6
1
1 2 16
6 36 33
pent
3
J A
Table 20. Lip, belly, and posterior thigh pattern character state distribution for Inter-Unit analysis. J = juvenile data, A = adult data. myersi = Leptodactylus myersi, para = Pará Unit, knud
B = L. knudseni Unit from Ecuador, knud C = L. knudseni Unit from central and southern Peru, knud A, D-H = rest of L. knudseni, labyr A = L. labyrinthicus Unit from coastal Venezuela, labyr
B-E = L. labyrinthicus Unit from Areas B-E, labyr F = L. labyrinthicus Unit from Area F, labyr G+H = L. labyrinthicus Unit from Areas G+H, pent 1 = L. pentadactylus Unit from Middle
America and adjacent Colombia, pent 2 = L. pentadactylus Unit from Pacific coastal Colombia and adjacent Ecuador, pent 3 = L. pentadactylus Unit from central coastal Ecuador, pent
4 = L. pentadactylus Unit from Amazonia.
300
Arquivos de Zoologia
13
17
Entire from at least 1/4 to full distance from eye to sacrum
Interrupted to at least between sacrum and some distance to groin
1
1
2
+
+
+
–
5
Male chest spines
9
2
16
1 + prepollical bump
1
1 tiny to small
None
Male thumb spines
1
1
22
3
96
9
+
5
46
10
4
1
1
33
92
27
knud
A, D-H
Entire from tympanum to lower flank
1
12
5
7
1
16
5
knud
C
1
9
10
6
2
32
knud
B
Interrupted from tympanum to lower flank
Entire from tympanum to shoulder
Interrupted from tympanum to shoulder
Dark spot/wart in area where fold would be between tympanum
and shoulder
None
Flank folds
Entire from eye to groin
3
40
Interrupted from at least 1/2 to full distance from eye to sacrum
Entire to at least between sacrum and some distance to groin
6
2
Interrupted from at least 1/4 distance to 1/2 distance from eye
to sacrum
18
2
9
para
Interrupted or entire to 1/4 distance from eye to sacrum
None
Dorsolateral folds
myersi
+
3
19
13
5
1
8
6
3
7
labyr
A
+
1
29
5
3
5
49
5
2
3
72
14
1
labry
B-E
+
5
1
1
12
1
12
2
labyr
F
+
22
1
3
33
13
1
1
51
10
labyr
G+H
+
2
22
6
36
20
28
1
25
40
7
19
pent
1
–
4
7
4
1
7
5
pent
2
–
2
1
7
2
1
8
1
1
pent
3
–
1
12
2
43
32
15
1
76
6
5
4
pent
4
Table 21. Dorsolateral fold, flank fold, male thumb spine, and male chest spine character state distributions for Inter-Unit analysis. myersi = Leptodactylus myersi, para = Pará Unit, knud
B = L. knudseni Unit from Ecuador, knud C = L. knudseni Unit from central and southern Peru, knud A, D-H = rest of L. knudseni, labyr A = L. labyrinthicus Unit from coastal Venezuela, labyr
B-E = L. labyrinthicus Unit from Areas B-E, labyr F = L. labyrinthicus Unit from Area F, labyr G+H = L. labyrinthicus Unit from Areas G+H, pent 1 = L pentadactylus Unit from Middle
America and adjacent Colombia, pent 2 = L. pentadactylus Unit from Pacific coastal Colombia and adjacent Ecuador, pent 3 = L. pentadactylus Unit from central coastal Ecuador, pent
4 = L. pentadactylus Unit from Amazonia.
Vol. 37(3), 2005
301
myersi
*
para
1/4+1
*
knud B
1/5+0
0/5+1
*
knud C
1/5+1
0/5+1
0/1+4
*
knud A, D-H
1/3+2
0/3+3
0/5+1
0/3+3
*
labyr A
1/5+1
0/4+2
0/4+2
0/3+2
0/3+3
*
labry B-E
1/3+2
0/3+3
0/5+0
0/4+2
0/4+1
0/5+1
*
labyr F
1/5+0
0/3+2
2/4+1
1/1+4
0/4+2
1/4+1
0/3+3
*
labyr G+H
1/5+0
0/4+1
0/5+0
0/5+1
0/4+2
0/4+2
0/2+2
0/2+2
*
pent 1
1/5+1
1/3+2
0/3+2
0/3+3
0/6+0
0/5+1
0/5+0
1/4+1
0/5+1
*
pent 2
0/5+1
2/5+0
1/4+2
1/5+1
2/2+3
1/6+0
2/4+1
3/3+0
2/5+0
1/5+1
*
pent 3
0/4+1
3/3+1
1/4+2
2/2+3
1/4+2
2/4+1
1/5+1
4/3+1
1/5+1
1/0+5
0/3+2
*
pent 4
0/5+0
2/5+0
1/5+1
1/5+1
1/5+1
1/6+0
1/6+0
2/5+0
1/6+0
1/4+1
0/2+4
0/3+1
*
Arquivos de Zoologia
myersi
para
knud B
knud C
knud A, D-H
labyr A
labyr B-E
labyr F
labyr G+H
pent 1
pent 2
pent 3
pent 4
Table 22. Differentiation summary of morphological characters from Inter-Unit analysis. Value to left of slash is number of diagnostic characters (those that completely differentiate all
specimens of the two Units being compared). Values to right of slash are number of characters that differentiate Units (the number of characters for which some states only occur in one of the
two Units being compared); the first value is the number of characters for which sample sizes are substantial, the second value is the number of characters which might be impacted by small
sample sizes. myersi = Leptodactylus myersi, para = Pará Unit, knud B = L. knudseni Unit from Ecuador, knud C = L. knudseni Unit from central and southern Peru, knud A, D-H = rest of
L. knudseni, labyr A = L. labyrinthicus Unit from coastal Venezuela, labyr B-E = L. labyrinthicus Unit from Areas B-E, labyr F = L. labyrinthicus Unit from Area F, labyr G+H = L. labyrinthicus
Unit from Areas G+H, pent 1 = L. pentadactylus Unit from Middle America and adjacent Colombia, pent 2 = L. pentadactylus Unit from Pacific coastal Colombia and adjacent Ecuador, pent
3 = L. pentadactylus Unit from central coastal Ecuador, pent 4 = L. pentadactylus Unit from Amazonia.
302
entirely diagnostic from all other differentiated
Units.
The following combination of diagnostic
and distinctive characters is conservatively
considered to represent species-level
differentiation: at least one diagnostic character and
at least three distinctive characters not likely due
to small sample sizes. Thirty-nine of the total 78
comparisons fit this conservative criterion.
The remaining 39 comparisons were ranked
from most differentiated to least differentiated to
serve as a template to add information from
additional characters (Table 23). In cases where
more than one pair had the same formula in Table
22 (e.g., -/5 + 1), ranking was based on the total
number of individual differentiated states not
impacted by small sample sizes for the lip, belly,
and thigh patterns. The one comparison that may
be misplaced in the relative ranking is that of Middle
American L. pentadactylus – West Coast Ecuador
L. pentadactylus, because the placement is high in
the ranking order based on having a diagnostic
difference (chest spines), but there are no other
characters that differentiate this pair except those
that could be due to small sample sizes.
Morphometric Data
Measurement data were analyzed using
discriminant function analysis (SYSTAT 10,
Engelman, 2000). In general, discriminant function
analyses are more robust with larger data sets. As
specimens examined for the previous study (Heyer,
1979) did not have tympanum diameter data, two
data sets were evaluated: (1) specimens without
tympanum data (this data set is larger in terms of
number of individuals); and (2) specimens with
tympanum data (this data set is larger in terms of
number of variables). Any specimen with a missing
value (other than tympanum diameter) was not
included in the analyses. The conservative approach
of analyzing male and female data separately as
recommended by Hayek et al. (2001) was followed,
resulting in four basic data sets.
Each of the four basic data sets was analyzed
using the complete data model and the backward
step model, using the default options in SYSTAT
10.
The backward step procedure results were
identical for three of the four data sets (that is, the
Vol. 37(3), 2005
303
Table 23. Ranking of morphological differentiation of pairs of differentiated Units that may or may not represent distinct species
based on morphological characters. The most differentiated Units are at the top of the list. Morphometrics data: m or f indicates
male or female data; lower case indicates diagnostic either in data set without tympanum diameter or data set with tympanum
diameter; upper case indicates diagnostic for both data sets; parentheses indicate that N<10 for both Units. Juvenile and adult life
colors: + indicates support, – indicates no support to diagnose the Unit pairs. Tadpoles: IOA indicates distinctive internal oral
anatomy features; + or – indicate support or lack thereof based on larval tooth row formulae. Advertisement calls: the number of
features that differ between the calls of the Units being compared (also see Table 24). See text for evaluation of characters.
Differentiated Unit Pairs
knudseni C – labyrinthicus F
MA pentadactylus – WCE pentadactylus
knudseni A, D-H – MA pentadactylus
knudseni B – knudseni A, D-H
knudseni C – labyrinthicus G+H
labyrinthicus G+H – MA pentadactylus
labyrinthicus A – MA pentadactylus
labyrinthicus A – labyrinthicus B-E
para – knudseni B
para – knudseni C
myersi – WCC pentadactylus
myersi – Amazonian pentadactylus
knudseni B – labyrinthicus G+H
labyrinthicus B-E – MA pentadactylus
knudseni B – labyrinthicus A
knudseni C – labyrinthicus B-E
knudseni A, D-H – labyrinthicus F
knudseni A, D-H – labyrinthicus G+H
labyrinthicus A – labyrinthicus G+H
para – labyrinthicus A
myersi – WCE pentadactylus
para – labyrinthicus G+H
knudseni A, D-H – labyrinthicus B-E
knudseni C – knudseni A, D-H
knudseni C – MA pentadactylus
labyrinthicus B-E – labyrinthicus F
para – labyrinthicus B-E
para – knudseni A, D-H
knudseni A, D-H – labyrinthicus A
knudseni B – MA pentadactylus
knudseni C – labyrinthicus A
para – labyrinthicus F
WCC pentadactylus – WCE pentadactylus
WCE pentadactylus – Amazonian pentadactylus
WCC pentadactylus – Amazonian pentadactylus
labyrinthicus F – labyrinthicus G+H
labyrinthicus B-E – labyrinthicus G+H
knudseni B – knudseni C
Morphometrics
(F)
–
–
–
M, f
–
F
–
(F)
M, (F)
M, (F)
M, F
M, f
–
(F)
m, F
m, F
M, f
f
(f)
M, (F)
–
f
f
f
–
–
–
–
–
M, (F)
m, (F)
(m)
F
m
M, F
–
(m), (f)
backward step analysis used all variables in the final
model). The female data set including tympanum
diameter data omitted SVL and head length in the
final backward step model.
The results of the five different analyses
were evaluated for Unit pairwise diagnostic
information as follows. Both the classification
matrix using all data and the jackknifed
classification matrix were examined for whether any
specimens were incorrectly identified in the
posterior classification results. For example, for the
Life Colors
J
A
+
–
–
Tadpoles
Advertisement Calls
+
1
2
0
IOA, –
+
+
1
2
0
2
+
+
+
–
2
–
+
2
+
+
+
+
+
2
1
0
2
+
+
–
+
+
+
4
+
–
–
–
1
0
0
0
1?
+
0
–
+
+
–
+
2
1
1
–
first cell in Table 23 [value of (F) under
Morphometric column], no specimens of
L. knudseni C were posteriorly identified as
L. labyrinthicus F in any results; 1 individual of
L. labyrinthicus F was posteriorly identified as
L. knudseni C in both male data sets including or
excluding tympanum diameter in the posterior
classification using all data; 2 individuals of
L. labyrinthicus F were posteriorly identified as
L. knudseni C in both male data sets including or
excluding tympanum diameter in the jackknife
304
Arquivos de Zoologia
matrix; and no L. labyrinthicus F individuals were
posteriorly identified as L. knudseni C individuals
in the female data sets including or excluding
tympanum diameter for all results using the
complete data posterior classification model and
the jackknife model.
The posterior classification results are
summarized as follows. All values for each posterior
classification must be zero for the data set to be
considered diagnostic. That is, using the first cell
in Table 23 as an example, the male data without
tympanum, the male data with tympanum, and the
female data without tympanum measurements, no
L. knudseni C individuals could be posteriorly
classified as L. labyrinthicus F individuals and no
L. labyrinthicus F individuals could be posteriorly
classified as L. knudseni C individuals in both the
complete data posterior classification matrix and
the jackknife matrix for that data set to be
considered diagnostic. For the female data with
tympanum diameter data, in addition to the previous
conditions, all values had to be zero for both the
complete data discriminant function model and the
backward step model for the female data set with
tympanum measurements to be considered
diagnostic. If only one of the two data sets for each
sex was diagnostic, a lower-case letter (either m or
f) is used; if both data sets for each sex are
diagnostic, a capital letter is used. As discriminant
function analyses can be unrealiable for cells with
small sample sizes, parentheses are used to indicate
that the sample sizes for both Units being compared
are less than 10 individuals.
Life Colors
Many aspects of life colors are (apparently)
undocumented for the Units being evaluated. There
are three aspects of life colors that appear to vary
markedly among the Units: juvenile dorsal colors
of greenish-yellow versus brown or brownish-red,
posterior thighs black or with red/orange markings;
and adult posterior thighs with white, yellow, or
orange/red markings. Additional information on life
colors will certainly help diagnose several of the
Units involved. For example, it is very likely that
all L. labyrinthicus Units can be differentiated from
L. knudseni B and L. knudseni C Units by juvenile
life color patterns, but at present, I find no
documented life colors for the critical features for
any juveniles from any L. labyrinthicus Units. The
available comparisons are indicated in Table 23.
For juvenile colors, a + indicates a difference in
either dorsal or posterior thigh colors.
Tadpoles
There is not a lot of descriptive published
data available on tadpoles of the Leptodactylus
treated herein.
Wassersug and Heyer (1988: Table 1) found
several differences between the larvae of
L. knudseni (from the Area A, D-H Unit) and
L. pentadactylus (Middle American Unit) in terms
of internal oral features. The single larvae they
examined for each species were about the same size,
but the L. knudseni was developmentally more
advanced (Gosner, 1960, stage 39 versus 34 for
the L. pentadactylus larva). Although many of the
differences categorized in Table 1 of Wassersug and
Heyer (1988) could possibly be related to different
developmental stages, at least three of them are not
(number of filter rows per plate on ceratobranchial
IV, folding pattern of filter rows, number of
postnarial papillae). The differences of internal oral
features are diagnostic for these two Units.
Larson and de Sá (1998) found just one
variable character in chondrocranial morphology
among L. knudseni, labyrinthicus, and pentadactylus
larvae they examined: chondrocranial width to length
ratio. For purposes of this exercise, the level of
variation observed is not considered diagnostic
between any of the specimens of L. knudseni,
labyrinthicus, and pentadactylus they studied.
External morphologies have been described
for four Units being evaluated in this section. All
larvae share a unique morphology within
Leptodactylus related to being at least facultative
carnivores. The tadpoles have long muscular tails,
the oral disk is smaller and much more anteriorly
oriented than in typical pond-dwelling surfacescraping feeder tadpoles, and the larvae grow to
quite large sizes (at least 70 mm total length).
The only consistent variation found among
the known larvae of the taxa in this study is in the
number of tooth rows.
Hero (1990) found that larval
L. pentadactylus (Amazonian Unit) completed their
entire development within burrows where the foam
nest was laid and had a tooth row formula of 1/2(1);
Vol. 37(3), 2005
whereas syntopic L. knudseni (Area A, D-H Unit)
had post-hatching free-living aquatic larvae with a
tooth row formula of 2(2)/2(1). I think that all freeliving aquatic larvae of the L. pentadactylus type
that have been collected in Amazonia belong to
L. knudseni Units. I examined Amazonian larval
samples of L. knudseni Units (from Area C Unit
and Area A, D-H Unit) and found that in most larvae
there is an abbreviated entire third tooth row on
the lower labium. There is variation in presence or
absence of the third posterior tooth row in larvae
from the Biological Dynamics of Forest Fragments
sites north of Manaus, Amazonas, Brazil (USNM
313497-313499).
Middle American L. pentadactylus Unit
larvae have been illustrated and described by
several authors from different localities throughout
much of its distribution (Breder, 1946; Heyer, 1970;
McCranie and Wilson, 2002; Villa, 1972). The tooth
row formula for the Middle American
L. pentadactylus Unit is 2(2)/3(1).
Vizotto (1967) described the larva of
L. labyrinthicus (Area F) as having a tooth row
formula of 1/2(1). USNM specimens examined
from Paraguay (USNM 253610, outside of Area
B-E and F Units) and the Area B-E Unit (USNM
507875) also have the tooth row formula 1/2(1).
Karyotypes
The few specimens that pertain to this study
that have been karyotyped have a diploid number
of 22. I find the available data uninformative with
respect to diagnosing any Units for which
individuals have been reported (see individual
species accounts for literature involved).
Advertisement Calls
The summary data for advertisement calls
from Tables 6, 11, 12, and 19 were combined by
Units with data for Leptodactylus myersi taken from
the compact disk recording, Sound Guide to the
Tailless Amphibians of French Guiana by C. Marty
and P. Gaucher. The variation in call rate within
the L. knudseni Area A, D-H Unit almost
encompassed the total variation. The call rate data
were not used to determine differentiation of
advertisement calls among Units.
305
Advertisement calls between Units are
considered distinctive if there is (1) no overlap of
ranges of values for call duration, number of pulses
per call, pulse rate in seconds; (2) at least a 300 Hz
difference between ranges of values for the
beginning frequency; and (3) at least a 100 Hz
difference between ranges of values for call range
(none of the Units differ by this criterion).
There are no experimental data on large
Leptodactylus species to know whether the above
thresholds used for assessing distinctiveness in call
features are sufficient for the frogs to recognize the
calls as different. Generally, frogs can discriminate
amplitude/timing aspects of calls very well. If the
basilar papilla is the organ used in hearing
advertisement call frequencies, generally calls need
to differ by at least 400 Hz, beyond which the frog
essentially does not hear the sound (see
experimental data in Zakon and Wilczynski, 1988
and empirical data in Blair and Littlejohn, 1960).
If the amphibian papilla is the hearing organ used
(for around 1000 Hz and less, Zakon and
Wilczynski, 1988:147), discrimination is much
better within the range of frequencies the organ can
process. The frogs in this study should be using
their amphibian papillae to process advertisement
calls. A difference in any of the parameters included
in this analysis could be sufficient for the frogs to
perceive them as different. A difference in two
parameters should have a high degree of likelihood
that the frogs perceive the calls as different (Table
24). Of course, the frogs could be able to
discriminate smaller differences than those used as
threshold values in this exercise.
Species Limits
The preceding data, as summarized in Table
23, is now evaluated in terms of species-level
recognition among the Units.
The level of differentiation of Leptodactylus
myersi from all other Units is at a level consistent
with recognizing it as a distinct species.
The Pará Unit is very distinct from all
L. pentadactylus Units at a level considered to
represent species-level differentiation. The Pará
Unit is equivocally differentiated from all the
L. knudseni and L. labyrinthicus Units. There are
no data for larvae or advertisement calls for the
Pará Unit. Larval data should differentiate the Pará
306
Arquivos de Zoologia
Table 24. Call differentiation among Units. 1 = no overlap in call duration; 2 = no overlap in number of pulses/call; 3 = no
overlap in pulse rate/s; 4 = at least 300 Hz difference between beginning frequencies; 5 = at least 300 Hz difference between
dominant frequencies.
knudseni C
knudseni A, D-H
MA pentadactylus
WCE pentadactylus
Amazonian pentadactylus
labyrinthicus A
labyrinthicus B-E
labyrinthicus F
labyrinthicus G+H
L. myersi
kC
–
k A, D-H
0
–
MA p WCE p
0
4, 5
0
0
–
3, 5
–
Unit from either all L. knudseni Units or all
L. labyrinthicus Units. At present, the available data
(morphological) are equivocal. My subjective
impression and current conclusion is that the Pará
Unit is specifically distinct from all L. knudseni
Units and all L. labyrinthicus Units.
The Leptodactylus knudseni Units are
distinct from the L. labyrinthicus B-E and F Units
at a level consistent with species level
differentiation. The L. knudseni C Unit has the same
magnitude of differentiation from the
L. labyrinthicus G+H Unit, but the L. knudseni B
and A, D-H Units do not have the same level of
differentiation from the L. labyrinthicus G+H Unit.
Advertisement calls do not differ at the previously
defined levels associated with species limits
between the L. knudseni C – L. labyrinthicus A
Units and the L. knudseni A, D-H – L. labyrinthicus
G+H Units; thus, currently additional call data may
not aid resolution of species limits among the
equivocal Units under discussion. However, life
colors and larval characters will probably provide
helpful data to robustly define species limits among
at least some of the Unit pairs under discussion.
With the available data (primarily morphological),
my subjective evaluation and current conclusion is
that all of the L. knudseni Units differ from all of
the L. labyrinthicus Units at the species level.
All Leptodactylus knudseni Units are
distinct from all L. pentadactylus Units at a level
consistent with species level differentiation except
when comparing the L. knudseni B and C Units with
the Middle American L. pentadactylus Unit.
Advertisement calls do not differ between the
L. knudseni A, D-H and C Units with the Middle
American L. pentadactylus Unit, therefore calls
may not differ between the L. knudseni B Unit and
the Middle American L. pentadactylus Unit when
Amaz p
2, 3
3
5
2, 3
–
lA
0
3?
0
1, 5
2, 3
–
l B-E
1, 2
2
1, 2
1, 2
2
1, 2
–
lF
2
2
1, 2
2, 5
2
1, 2
0
–
l G+H
1
0
1
1, 5
2, 3
1, 2
2
2
–
L. myersi
2, 3
2, 3
2, 3
1, 2, 3, 4
2, 3
2, 3
1, 2
1, 2
1, 2, 3
–
calls of the L. knudseni B Unit become available. I
find the level of differentiation of larval internal
oral features to be surprisingly great between the
L. knudseni A, D-H Unit and the Middle American
L. pentadactylus Unit tadpoles. I anticipate that
when larvae of the L. knudseni B and C Units are
examined for internal oral features, they will also
demonstrate significant variation relative to the
Middle American L. pentadactylus Unit. I think that
the L. knudseni Units and the Middle American
L. pentadactylus Unit are separate species, and that
the Middle American L. pentadactylus Unit is the
sister-group to the L. knudseni Units rather than to
the Amazonian L. pentadactylus Unit.
All Leptodactylus labyrinthicus Units differ
from all L. pentadactylus Units at a level consistent
with species level differentiation except for the
L. labyrinthicus A Unit and the Middle American
L. pentadactylus Unit. The advertisement calls do
not differ between the L. labyrinthicus A and
Middle American L. pentadactylus Units at the
previously defined levels for species level
differentiation. Larvae are not known for the
L. labyrinthicus A Unit. The available data are
equivocal whether the L. labyrinthicus A Unit and
the Middle American L. pentadactylus Unit
represent distinct species. My subjective impression
is that they do; thus all L. labyrinthicus Units are
considered to be distinct at the species level from
all L. pentadactylus Units.
The Leptodactylus knudseni Units either do
not differ from each other at levels expected for
species differentiation (L. knudseni B – L. knudseni
A, D-H Units) or at equivocal levels of
differentiation. The L. knudseni B and C Units were
not combined after the intra-Unit analysis because
the only available juvenile color information was for
one L. knudseni C Unit specimen, which lacked the
Vol. 37(3), 2005
distinctive greenish yellow colors and black posterior
thigh colors of L. knudseni B Unit juveniles.
Subsequently, life colors were obtained for additional
L. knudseni C specimens, and they are identical to
those of the distinctively colored L. knudseni B Unit
juveniles. There is minimal variation between the
L. knudseni B and C Units which raises the question
of whether the L. knudseni B+C Unit represents a
different species from the L. knudseni A, D-H Unit.
There is weak morphological and measurement
differentiation; some life color differentiation; and
no larval or advertisement call differentiation
between the L. knudseni B+C and A, D-H Units. I
interpret the available information to be most
consistent with recognizing a single species for all
L. knudseni Units.
Variation among the Leptodactylus
labyrinthicus Units is consistent with species level
differentiation except for the L. labyrinthicus B-E
and F Unit comparisons. The differentiation
between these two Units is primarily yellow versus
red posterior thigh color in life. There is also some
morphological differentiation, but the larvae and
advertisement calls apparently do not differ. Call
recordings from Area F consist of a single pulse,
and calls of the single poor quality Area B-E
recording sounds the same as recordings from Area
F. The two Units may eventually prove to represent
distinct species, but available data indicate that the
variation is intraspecific in nature. The following
are recognized as distinct species: the
L. labyrinthicus A Unit; the L. labyrinthicus B-F
Unit; and the L. labyrinthicus G+H Unit. Two
307
geographic samples can not be assigned with
certainty to the three species as recognized: (1) the
recording from Santa Cruz, Bolivia is clearly
pulsed, differing from the unpulsed calls of the
geographically proximate L. labyrinthicus B-F
species; (2) the locality of Serra de Parima,
Roraima, Brazil (MZUSP 24936) is a geographic
outlier and cannot be associated with any of the
three recognized species. For present purposes, all
Bolivian specimens exhibiting the general
L. labyrinthicus morphology and the specimen from
Roraima, Brazil are treated as unknown species.
Diagnostic characters among Leptodactylus
pentadactylus Units provide clear as well as
ambiguous interpretations for species level
differentiation (Table 25). The level of
differentiation between the pairs Middle American
Unit – West Coast Colombia Unit, Middle
American Unit – Amazonian Unit, and West Coast
Ecuador Unit – Amazonian Unit strongly supports
species level differentiation. There is moderate
support for recognizing the Middle American Unit
and West Coast Ecuador Unit as distinct species.
Equivocal support exists for species level
differentiation between the West Coast Colombia
Unit – West Coast Ecuador Unit and the West Coast
Colombia Unit – Amazonian Unit. The following
example illustrates the difficulty in interpretation
of the moderate and equivocal levels of support for
species level differentiation. If the data were
interpreted to mean that the West Coast Colombia
Unit represented the same species as the Amazonian
Unit and the West Coast Colombia Unit represented
Table 25. Diagnostic characters among Leptodactylus pentadactylus Units. Bold indicates characters considered to be robustly
diagnostic in terms of species level differentiation.
West Coast Colombia Unit
Chest spines
Thumb spines
Belly pattern
Thigh pattern
Dorsolateral folds
Flank folds
West Coast Ecuador Unit
Call
Chest spines
West Coast Colombia Unit
X
West Coast Ecuador Unit
X
Thumb spines
Belly pattern
Thigh pattern
X
Middle American Unit
Amazonian Unit
Call
Tadpoles
Life colors
Chest spines
Thumb spines
Belly pattern
Lip pattern
Thigh pattern
Life colors
Thigh pattern
Flank fold
Call
Life colors
Thumb spines
Female measurements
Belly pattern
Thigh pattern
308
Arquivos de Zoologia
the same species as the West Coast Ecuador Unit,
then one would expect there to be much less
differentiation between the West Coast Ecuador
Unit and Amazonian Unit than is represented in the
data summarized in Table 25. There are four
alternative interpretations of the equivocal and
moderate levels of differentiated Units (Table 25):
(1) the West Coast Colombia Unit represents the
same species as the West Coast Ecuador Unit; (2)
the West Coast Colombia Unit represents the same
species as the Amazonian Unit; (3) the Middle
American Unit represents the same species as the
West Coast Ecuador Unit; and (4) each Unit
represents a distinct species. The first two
alternatives would be easier to evaluate if
advertisement calls and tadpole data were available
for the West Coast Colombia Unit. If alternative
(2) were correct, one would predict that the West
Coast Colombia Unit would have a tadpole that
completes development within the terrestrial
burrow where the foam nest was deposited and that
the larvae would have a tooth row formula of 1/2(1).
My evaluation of the available data is that they best
support the fourth alternative. Acceptance of the
fourth alternative has the consequence of not being
able to determine whether the specimen from La
Tola, Esmeraldas, Ecuador (QCAZ 19859) belongs
to the West Coast Colombia or West Coast Ecuador
Unit, as the female does not have sufficient
diagnostic characteristics. The available albumin
microcomplement fixation data support recognition
of the Amazonian Unit, the Middle American Unit,
and the West Coast Ecuador Units as distinct species
(Maxson and Heyer, 1988).
In summary, the following are recognized
as distinct species:
•
•
•
•
•
•
•
•
•
•
Leptodactylus myersi
the Pará Unit
Leptodactylus knudseni (B+C+A, D-H Unit)
Leptodactylus labyrinthicus Unit A
Leptodactylus labyrinthicus B-F Unit
Leptodactylus labyrinthicus G+H Unit
Leptodactylus pentadactylus Middle American
Unit
Leptodactylus pentadactylus Amazonian Unit
Leptodactylus pentadactylus West Coast
Colombia Unit (extending geographically into
adjacent Ecuador)
Leptodactylus pentadactylus West Coast
Ecuador Unit.
Problematic Specimens. – Earlier, certain
individuals were identified as exhibiting features
that did not allow their assignment to any of the
groupings being analyzed. These distinctive
samples are discussed relative to the species as
recognized herein.
Specimen OMNH 37583. The specimen is
a 95 mm SVL male with a moderate-sized black
spine on each thumb. It fails to match almost all
L. knudseni males in lacking chest spines. Other
small male L. knudseni also have a thumb spine
and lack chest spines, such as MZUSP 80660,
96 mm SVL, from Aripuanã, Mato Grosso, Brazil.
Larger males from Aripuanã have chest spines in
addition to thumb spines. OMNH 37583 is
considered to be a small male L. knudseni.
Specimens MZUSP 131123, USNM
303909. The two males, 129 and 110 mm SVL
respectively, are from Alto Paraiso, Rondônia,
Brazil. I did not associate these specimens with the
OTUs (Operational Taxonomic Units) in this study
because both males have a single thumb spine but
lack chest spines and the larger male with
extensively hypertrophied arms has black tubercles
on the throat, chest, and anterior belly, indicating
that the specimens from this locality do lack chest
spines. After this manuscript was completed, I
discovered that there were an additional three males
and two small juveniles from the same locality
(AMNH 124825-124829), collected at the same
time as MZUSP 131123 and USNM 303909.
AMNH 124825 is a 132 mm SVL adult male with
extensively hypertrophied arms that has one large
black spine on each thumb, one pair of chest spines,
and a field of black tubercles on the throat, chest,
and anterior belly. The combined morphological
data for the AMNH, MZUSP, and USNM
specimens match those for either L. knudseni or the
L. labyrinthicus B-F Unit. Discriminant function
analyses using measurement data for males of the
species recognized herein result in poor
discrimination among species, with only 47-55%
of the specimens being correctly assigned
posteriorly in either the standard classification or
jackknife classification matrices using data sets with
and without the tympanum diameter variable
(results not shown). The data from the Rondônia
adult males were not used to develop the
discriminant function models but were included to
determine with which species the Rondônia adult
Vol. 37(3), 2005
male data were most consistent. The discriminant
function model not including tympanum diameter
data predicts that three of the specimens are
L. knudseni and two specimens belong to the Pará
Unit. The discriminant function model including
tympanum diameter data predicts that two
specimens are L. knudseni, two specimens belong
to the Pará Unit, and one specimen belongs to the
L. pentadactylus Middle American Unit.
Ronald I. Crombie recorded the following for
MZUSP 131123: “Light brown with reddish-brown
darker transverse markings. Upper thighs very
reddish. Bright yellow on concealed thighs, tibiae
and in groin. Venter dirty white, suffused with ivory
to dull yellow reticulum, particularly on limbs and
edges of jaws. Soles of feet purplish. Iris gold with
median lateral reddish marks.” Charles W. Myers
recorded the following for the AMNH specimens:
“Adults (18406-07) [= AMNH 124825-124826]
light gray over head and upper arms, turning light
grayish brown over back and hindquarters. Dull
yellow and black mottling on rear thigh. Ventral
surfaces whitish. Iris golden bronze with variable
black venation and bright orange-brown butterfly
mark. Juveniles (18408-09) [= AMNH
124827-124828] pale greenish brown with dark
gray-brown crossbands. Rear thighs spotted with
pale gray. Iris pale orange above, pale gray below
(no butterfly mark).”
309
thumb spines, but lack chest spines. Both specimens
have small black chest tubercles, indicating that the
lack of chest spines is real and not due to seasonal
shedding of the spines (but note the problem of
interpretation of chest spine presence/absence in
small sample sizes in immediately preceding
paragraphs). When the two Jacareacanga specimens
were posteriorly classified using the same models
described for the preceding Rondônia specimens,
MZUSP 24947 was classified as L. knudseni in the
model that excluded tympanum diameter data and
as Middle American L. pentadactylus in the model
that included tympanum diameter data. MZUSP
24948 was posteriorly classified as L. knudseni in
both models that included and excluded tympanum
diameter data. I know of no data on life colors or
habitat for these specimens. The data are equivocal
whether the specimens represent L. knudseni or a
new species. For present purposes, they are
considered unidentifiable to species.
Specimen USNM 202518 (Figure 20). The
adult female is most similar to L. pentadactylus with
the exception that it does not have any bars on the
lip as found in all individuals examined of
L. pentadactylus. Pending further data, I consider
this difference to be of enough significance to retain
it as unidentifiable to species and not significant
enough to consider it a distinct species.
Nomenclature
The combined data for these specimens is consistent
only with those of L. knudseni, although the
posterior thigh colors of the adults are not the same
as those for which descriptions are available (but
are variations of the same pigment cell type) and
the juveniles are not as distinctively colored as those
from Ecuador and Peru (see color descriptions in
L. knudseni Unit portion of Intra-unit Analyses
section). These specimens are considered to
represent the same species as the L. knudseni Unit
species. Morphological and measurement data for
these specimens are not included in the Species
Accounts section for L. knudseni.
Specimens MZUSP 24947-24948. The two
male specimens, 150 and 152 mm SVL
respectively, are from Jacareacanga, Pará, Brazil
(Figure 17). The specimens share the most character
states and have most similarity with L. knudseni and
L. myersi. Both specimens have well-developed
The information previously provided on
nomenclature of this group of frogs under study
(Heyer, 1979) is not repeated here.
Rana pentadactyla Laurenti, 1768. I
previously designated the name bearer to be the
specimen illustrated in Seba, 1734, Plate 75, Figure
1 (Heyer, 1979:13). This figure could represent any
of the four species currently identified as
Leptodactylus pentadactylus. The available
evidence indicates that Seba’s frog came from
Suriname, South America (Bauer, 2002:9; Heyer,
1979:13).
Seba’s collections have had a complicated
history (Boeseman, 1970); those specimens still
surviving are in a number of collections. To date,
no one has identified and published that the type of
Rana pentadactyla is extant in their collection.
Boeseman (1970) indicates nine collections have
Seba specimens: (1) Academy of Sciences, St.
310
Arquivos de Zoologia
Petersburg (ZIL), (2) Copenhagen Museum of
Natural History (ZMUC), (3) Museum National
d’Histoire Naturelle, Paris (MNHN), (4)
Naturhistoriska Riksmuseet, Stockholm (NRM), (5)
The Natural History Museum, Leiden (RMNH), (6)
The Natural History Museum. London (BMNH),
(7) Uberseemuseum, Bremen (UMB), (8)
Universität Humboldt, Berlin (ZMB), and (9)
Universitein van Amsterdam (ZMA).
Each of the nine collections Boeseman
(1970) indicated as obtaining Seba specimens was
contacted to determine whether any Seba specimens
that Laurenti included in materials described as
Rana pentadactyla are extant.
The following curators reported that they
did not have a Seba specimen that could be a
potential type of Rana pentadactyla (numbering
from paragraph above).
2) Dr. Jens Rasmussen, ZMUC, in an e-mail
message dated 14 October 2002 stated: “To the best
of my knowledge we do not possess any herptiles
originating from the Seba collection. All the
specimens (a dozen) which we possess of the
requested species seem to have been caught in
nature by well known Danish collectors.
Accordingly, I do not believe the type is here.”
3) Dr. Annemarie Ohler, MNHN, in an e-mail
message dated 5 November 2003 wrote: “I checked
our files and could not find any specimen that might
come from the Seba collection. All specimens have
French collectors and seem to be more recent than
Seba. There are no frogs from Leiden “exchange”
(the usual origin of Seba specimens).”
5) Dr. Marinus Hoogmoed (e-mail message
dated 24 July 2002) informed me that the RMNH,
does have three Seba specimens of L. pentadactylus
from Suriname, but none of the three match Seba’s
figure. Dr. Hoogmoed stated, “Thus, I can assure
you we do not have the depicted Seba specimen
here.”
6) Dr. Mark Wilkinson, BMNH, in an e-mail
message of 3 November 2003 wrote in response to
my query of any Leptodactylus that came from
Seba’s second collection: “I’ve just searched
through our catalogues and can find no specimen
with any indication of a linkage to Seba … .”
7) Dr. Peter Renè Becker, UMB, responded
in an e-mail message of 9 January 2004: “I am sorry
to say that we hold neither Leptodactylus
pentadactylus nor any Seba specimens in our
collections.”
8) Dr. Rainer Günther, ZMB, in an e-mail
message dated 15 October 2002, stated: “None of
our specimens of Leptodactylus pentadactylus
could be identified as a ‘Seba-type.’ We also do
not have indications in our catalogues that such a
type should be stored here.”
9) Dr. Axel Groenveld, ZMA, in an e-mail
message dated 8 September 2003, stated: “I could
find only five specimens of Leptodactylus
pentadactylus in our collection … . Although three
of them are from doubtful origin, none of them refer
to the Seba collection. … .”
I examined the holdings of Leptodactylus
at the NRM (4 in paragraph above) in August 2003
and did not find any specimen that could be a Seba
specimen of Rana pentadactyla.
Dr. Natalia Ananjeva and Konstantin Milto,
ZIL (1 in paragraph above) informed me that a Seba
specimen identified as Leptodactylus pentadactylus
was in their collection. I examined the specimen,
ZIL 274, in August 2003. The specimen is an adult
male Leptodactylus knudseni. Seba’s illustration
(1734, Plate 75, Figure 1), although somewhat
stylized (it has five full fingers on each hand, which
no known extant frog has), is very clear in showing
a pair of complete dorsolateral folds extending from
the eye to the groin. ZIL 274 has a pair of
dorsolateral folds that are distinct only from the
eye to the sacral region. Hence ZIL 274 could not
represent the specimen illustrated in Seba and is
not a potential syntype of Rana pentadactyla
Laurenti, 1768.
Given the morphological similarities among
the species to which the name Rana pentadactyla
Laurenti, 1768 has been applied, designation of a
neotype is necessary to associate the name with a
specimen to achieve nomenclatural stability. As
Seba’s specimens came from Suriname, I hereby
designate RMNH 29559 as the neotype of Rana
pentadactyla Laurenti, 1768. RMNH 29559 is a
134 mm adult male (with vocal slits and a tiny
white spine on each thumb) from Suriname,
Marowijne, Lelygebergte, Suralcokamp V,
collected on 15 August 1975 by Marinus S.
Hoogmoed.
Rana pentadactyla Laurenti, 1768 is the
oldest available name for the species analyzed
herein as the Amazonian L. pentadactylus Unit.
Rana gigas Spix, 1824. Hoogmoed and
Gruber (1983:355) did not find the type in the
collections in Munich or Leiden. Peters (1873) and
Vol. 37(3), 2005
Heyer (1979) provided evidence that the name
applied to the Amazonian L. pentadactylus Unit as
defined in this paper. Smith et al. (1977) pointed
out that Rana gigas Spix is a preoccupied name
and not available for any species of Leptodactylus.
Since Rana gigas Spix is a synonym of
Leptodactylus pentadactylus (Laurenti, 1768), there
is no need to propose a replacement name for Rana
gigas Spix.
Rana coriacea Spix, 1824. Hoogmoed and
Gruber (1983:355) reported the type was lost.
Peters (1873) and Heyer (1979) concluded that
Rana coriacea Spix was a member of the
Amazonian L. pentadactylus Unit as defined in this
paper. Thus, Rana coriacea Spix is considered to
be a synonym of Leptodactylus pentadactylus
(Laurenti, 1768) and there is no present need to
designate a neotype for Rana coriacea Spix.
Rana labyrinthica Spix, 1824. Heyer (1979)
indicated that this name applies to the Area F
L. labyrinthicus Unit as analyzed in this paper and
is thus the oldest name for the Leptodactylus
labyrinthicus Area B-F Unit species.
Leptodactylus goliath Jiménez de la Espada,
1875. Heyer (1979) indicated that the lectotype is
a member of the Amazonian L. pentadactylus Unit
as analyzed in this paper. Hence, Rana goliath
Jiménez de la Espada is a synonym of Leptodactylus
pentadactylus (Laurenti, 1768).
Leptodactylus wuchereri Jiménez de la
Espada, 1875. Heyer (1969) examined the type and
transferred the name from the synonymy of
L. mystacinus to L. pentadactylus sensu lato. Later
(Heyer, 1979:22), I placed L. wuchereri in the
synonymy of L. labyrinthicus. The type specimen
was collected in Argentina, somewhere between
Montevideo, Uruguay and Santiago, Chile, to the
south of parallel 31°S, passing through the known
distribution of L. labyrinthicus in Misiones and
northern Corrientes Provinces, Argentina (De la
Riva, 2000). Leptodactylus wuchereri is considered
to be a synonym of L. labyrinthicus (Spix, 1824).
De la Riva (2000) suggested that L. wuchereri could
either be a synonym of L. labyrinthicus or
L. mystacinus, based on the itinerary of Sr. Amor,
the collector of the type of L. wuchereri, and thus
needed further evaluation. Based on examination
of the holotype and previous discussion (Heyer,
1969), the holotype could refer only to
L. labyrinthicus in Argentina and specifically not
to L. mystacinus.
311
Leptodactylus bufo Andersson, 1911. Heyer
(1979:15) indicated that the name was a synonym
of Leptodactylus labyrinthicus. The type locality
(Ponta Grossa, Paraná, Brazil) is closest
geographically to the Area F L. labyrinthicus Unit
as analyzed in this paper. Leptodactylus bufo
Andersson is herein considered to be a synonym of
Leptodactylus labyrinthicus (Spix, 1824).
Leptodactylus macroblepharus MirandaRibeiro, 1926. Heyer (1979:16) indicated that the
type represented the Amazonian L. pentadactylus
Unit as defined in this paper. Hence,
L. macroblepharus Miranda-Ribeiro is a synonym
of L. pentadactylus (Laurenti, 1768).
Heyer (1979:16) incorrectly stated that
MZUSP 377 was the holotype of
L. macroblepharus. Miranda-Ribeiro described the
species based on three specimens, which were given
the same museum catalogue number. There were
actually four specimens that bore MZUSP number
377. Three of the specimens were recatalogued. I
examined the specimens in January 2003. The
specimen that retains MZUSP 377 is a 114 mm SVL
adult male with testes in the body cavity, but the
other internal organs are removed. MZUSP 56719
is a 46 mm SVL juvenile with all internal organs
removed; MZUSP 56720 is a 96 mm SVL male
with all internal organs removed; MZUSP 56721
is a 72 mm SVL juvenile with all internal organs
removed. The specimen that retains MZUSP 377
is without doubt the specimen that Miranda-Ribeiro
described. MZUSP 377 is hereby designated as the
lectotype of Leptodactylus macroblepharus
Miranda-Ribeiro.
Leptodactylus vastus A. Lutz, 1930. Heyer
(1979:16) indicated that this name applies to the
Area G+H L. labyrinthicus Unit as defined in this
paper. Hence, L. vastus A. Lutz is the oldest
available name for the Area G+H L. labyrinthicus
Unit species.
Adolpho Lutz (1930) proposed the name
vastus for three specimens that he had previously
reported and figured (as L. ? gigas) from
Independencia (= Guarabira), Paraíba, Brazil (A.
Lutz, 1926). When the Adolpho Lutz collection was
transferred from the Instituto Oswaldo Cruz to the
Museu Nacional collection, there was but a single
specimen labeled as the type of L. vastus, AL-MN
70 (Ulisses Caramaschi, pers. comm.). This
specimen, a 170 mm SVL adult male, is hereby
designated as the lectotype of L. vastus A. Lutz.
312
Arquivos de Zoologia
Leptodactylus pentadactylus dengleri
Melin, 1941. Heyer (1979:16) indicated the type
from Amazonian Peru belonged to the Amazonian
L. pentadactylus Unit as defined in this paper.
Taylor (1952:649-650) concluded that his Costa
Rican materials best matched the description Melin
provided for L. pentadactylus dengleri and used
that name for the Costa Rican materials he
examined. The attribution of Middle American
specimens to L. pentadactylus dengleri, following
Taylor, has been used by certain other authors, most
notably Cei and collaborators in a series of papers
on skin secretion biochemistry (e.g., Erspamer
et al., 1964). Leptodactylus pentadactylus dengleri
Melin is herein considered to be a synonym of
L. pentadactylus (Laurenti, 1768).
Melin based his description on two specimens. No
museum numbers were given for either specimen
and there is no indication in the description that
Melin designated one of the specimens as the
holotype. I previously (Heyer, 1979:16) designated
the larger of the two specimens as the lectotype. At
the time I made that designation, I had not
communicated with the curatorial staff at the GNM.
I recently asked Dr. Göran Nilson if he could
provide the GNM number for the lectotype. Dr.
Nilson responded in an e-mail message of 4
September 2003: “The two types of Leptodactylus
pentadactylus dengleri have the museum numbers
GNM Ba.ex. 496 (the smaller specimen, which is
the ‘type’ according to the original label, and listed
as holotype in our museum catalogue) and GNM
Ba.ex. 497 (the bigger specimen, not listed as ‘type’
on the original label).” In retrospect, it would have
been more appropriate to designate the smaller
specimen as the lectotype, as that seemed to be
Melin’s intent. However, a specimen label and
catalogue indication that the smaller specimen is
the type does not meet the requirements of Article
74 of the International Code of Zoological
Nomenclature (International Commission on
Zoological Nomenclature, 1999). Therefore, my
previous designation of the larger specimen as
lectotype stands.
Leptodactylus pentadactylus rubidoides
Andersson, 1945. Heyer (1979:16) indicated that
this name applies to the Amazonian
L. pentadactylus Unit as defined in this paper.
Leptodactylus pentadactylus rubidoides Andersson
is thus a synonym of L. pentadactylus (Laurenti,
1768).
Andersson (1945:49-51) identified two individuals
representing his new form rubidoides, the specimen
he described and figured in the NRM collection
and a specimen he did not personally examine
described by Boulenger (1882:243) as an unusual
specimen of L. rubido from Canelos, Ecuador.
These two specimens clearly constitute the type
series of L. pentadactylus rubidoides according to
Article 72.4.1 of the International Code of
Zoological Nomenclature (International
Commission on Zoological Nomenclature, 1999).
I hereby designate NRM 1928 as the lectotype of
Leptodactylus pentadactylus rubidoides.
Leptodactylus
pentadactylus
mattogrossensis Schmidt and Inger, 1951. Heyer
(1979:16-17) indicated that this name applies to
the Area B L. labyrinthicus Unit as analyzed in this
paper. Thus, L. pentadactylus mattogrossensis
Schmidt and Inger is a synonym of L. labyrinthicus
(Spix, 1824).
Leptodactylus knudseni Heyer, 1972. Heyer
(1979:17) indicated that this name applies to the
Area B L. knudseni Unit as analyzed in this paper.
Thus, L. knudseni is the oldest available name for
the L. knudseni (Area B+C+A, D-H Units) species.
No names are available for the following
Units recognized as distinct species: Pará Unit;
L. labyrinthicus Unit A; Middle American
L. pentadactylus Unit; West Coast Colombia
(extending into adjacent Ecuador) L. pentadactylus
Unit; and the West Coast Ecuador L. pentadactylus
Unit. These are described as new in the following
section.
Species Accounts
Dorsal patterns were not treated in the InterUnit analysis section. Dorsal patterns were
categorized anew for inclusion in the adult
characteristic sections of the species accounts. Only
those character states that occur at a frequency >
5% in any species are included in the adult
characteristic descriptions (Table 26).
The character states analyzed in the InterUnit Analysis section are included in the adult
characteristic sections of the species accounts only
16
24
104
70
31
18
25
7
11
11
25
28
186
21
17
1
4
14
31
140
1
1
31
62
1
4
93
1
1
55
1
30
12
Venezuela
L.
Middle
species pentadactylus America
species
N
25
0
15
1
1
3
32
27
L.
vastus
4
106
32
8
0
39
0
37
8
6
9
L.
L.
knudseni labyrinthicus
Variegated
Dark with light interorbital bar and scattered,
irregular light marks
4
2
Series of irregularly spaced, fuzzy small blotches
Single broad dark chevron posterior to interorbital
bar followed by series of fairly regularly spaced
large dark spots
17
Series of regularly spaced, well-defined small blotches
Broad, irregular, dark longitudinal band
4
6
More than 2 dark, broad transverse bands of equal
intensity in addition to interorbital band,
confluent laterally or not
3
60
Well developed dark interorbital band/chevron and
2 moderate to large equally intense dark chevrons,
second chevron in sacral region, chevrons
confluent or not
Irregular quadrangular or rectangular markings,
of equal intensity or alternating lighter/darker,
confluent laterally or not
3
Uniform with 1 or 2 well-defined, narrow
transverse bars
4
4
2
Series of regularly or irregularly placed small,
dark spots
12
Pará
species
Two broad dark transverse bands between
interorbital bar and arm insertion area, rest of
dorsum with large spots
4
Uniform light or dark
L.
myersi
Table 26. Dorsal pattern characterization among species. N = number of individuals. Other values are percentages.
15
33
67
16
62
6
31
West Coast West Coast
Colombia
Ecuador
+ species
species
Vol. 37(3), 2005
313
314
Arquivos de Zoologia
if they occur at frequencies > 5% in any species.
Pattern data are based on both juveniles and adults;
measurement data, unless otherwise noted, are
based on adults only. Use of the word “usually”
means the state involved occurs in more than 50%
of the sample; “often” means an occurrence from
10-50%; and “rare” an occurrence of less than 10%.
The larval materials examined for this study
were those in the USNM collection, none of which
can be unambiguously associated with adults. For
this reason, only published larval information is
summarized in the following species accounts.
There are virtually no morphological or
advertisement call character states that will
consistently diagnose adults of all the taxa in this
study. To attempt to diagnose a given species from
all others would result in very long statements that
would defeat the purpose of a diagnosis. The
approach taken here is to limit the diagnoses to taxa
that occur in the same geographic areas and major
habitat types and/or that are most similar to each
other and most likely to be confused.
Measurement data were evaluated for sexual
dimorphism using the separate variance t-test in
SYSTAT 10 (Anonymous, 2000). The ratio
variables were arcsine transformed prior to
evaluation with t-tests. A P-value of 0.05 or less is
considered significant. Initial sample sizes pertain
for all variables unless specified otherwise. Mean
values are indicated by the letter “m.”
Leptodactylus knudseni Heyer, 1972
Leptodactylus knudseni Heyer, 1972:3. Type
locality: Limoncocha, Napo, Ecuador, 00°25’S,
76°38’W. Holotype: LACM 72117, juvenile
female.
Suggested English name – Knudsen’s thintoed frog.
Suggested Portuguese name – Rã-de-dedosdelgados-de-Knudsen.
Suggested Spanish name – Rana de dedos
delgados de Knudsen.
Diagnosis – Leptodactylus knudseni has a
broad Amazonian distribution, occurring in the
same general areas as L. labyrinthicus, L. myersi,
L. paraensis, and L. pentadactylus. Juvenile
L. knudseni often have green dorsal coloration in
life and solid black posterior thighs; juvenile
L. labyrinthicus, L. myersi, L. paraensis, and
L. pentadactylus are not green in life nor do they
have solid black posterior thigh patterns. The
advertisement calls of L. knudseni are pulsed, the
advertisement calls of L. labyrinthicus are not
pulsed. The dorsolateral folds of L. knudseni are
usually entire, the dorsolateral folds in L. paraensis
are interrupted and are often interrupted in
L. labyrinthicus. Large, sexually active male
L. knudseni have a pair of chest spines; chest spines
are lacking in L. myersi and L. pentadactylus.
Adult Characteristics – Lip pattern usually
with dark triangular marks, 1 or 2 elongate,
approaching or entering lower eye (Figure 12A).
Lip pattern often of dark triangular marks on edge
of upper lip only. Lip pattern rarely uniform light;
or dark elongate triangular marks on lip edge, fading
towards eye, 2 dark marks under eye; or dark with
2 narrow light chevrons from lip entering eye
(Figure 12B); or alternating broad light bands and
narrow dark vertical stripes, 1-3 dark stripes
entering eye (Figure 12C); or broad irregular darker
and lighter bands, 2 dark bands entering eye (Figure
12D); or uniform dark.
Dorsal pattern often with more than 2 dark, broad
transverse bands of equal intensity in addition to
interorbital band, confluent laterally or not; or
irregular quadrangular or rectangular markings, of
equal or alternating lighter/darker intensity,
confluent laterally or not. Dorsal pattern rarely
uniform light; or uniform with 1 or 2 well-defined,
narrow transverse bands; or well developed dark
interorbital band/chevron and 2 moderate to large
equally intense dark chevrons, second chevron in
sacral region, chevrons confluent or not; or 2 broad
dark transverse bands between interorbital bar and
arm insertion area, rest of dorsum with large spots;
or dorsum with single broad dark chevron posterior
to interorbital bar followed by a series of fairly
regularly spaced large dark spots.
Belly pattern often light (no pattern); or mottled
(Figure 13A) or uniform dark; or dark with small
light vermiculations (Figure 13B); or dark with
small light discrete spots. Belly pattern rarely with
lateral and anterior mottling only; or dark with
large light vermiculations (Figure 13C); or light
with dark vermiculations; or labyrinthine (Figure
13D); or dark with large light discrete spots
(Figure 13E).
Vol. 37(3), 2005
315
Figure 15. Distribution map for Leptodactylus knudseni.
Posterior thigh pattern often distinctly uniform dark;
or dark with small distinct light vermiculations or
spots (Figure 14A); or dark with large distinct light
vermiculations, some coalescing of vermiculations;
or dark with contrasting large light irregular
blotches/spots extending from light dorsal transverse
bars, rest of thigh dark with light vermiculations
(Figure 14B). Thigh rarely relatively uniform dark
or indistinctly mottled; or labyrinthine; or dark with
distinct, discrete light spots; or very dark, irregular
blotches highlighted by light pin-stripe on dark
background (Figure 14C); or mostly very distinctly
light with few irregular dark marks (Figure 14D).
Dorsolateral folds usually entire from at
least 1/4 to full distance from eye to sacrum.
Dorsolateral folds often interrupted from at least
1/2 to full distance from eye to sacrum. Dorsolateral
folds rarely entire to at least between sacrum and
some distance to groin.
Flank folds usually a dark spot/wart in area
where fold would be between tympanum and
shoulder. Flank folds often absent. Flank folds
rarely entire from tympanum to shoulder; or
interrupted from tympanum to lower flank; or entire
from tympanum to lower flank.
Male thumb usually with 1 large spine. Male
thumb often with 1 tiny to small spine; or 1 large
spine and a prepollical bump.
Large breeding males with chest spines.
Female (N = 37) SVL 102.7-154.0 mm
(m = 132.0), male (N = 78) SVL 94.0-170.0 mm
(m = 131.4), not sexually dimorphic (t = 0.253,
df = 104.4, P = 0.801). Female head length/SVL
ratio 0.32-0.46 (m = 0.354), male head length/SVL
ratio 0.32-0.40 (m = 0.357), not sexually dimorphic
(t = -0.724, df = 55.3, P = 0.472). Female head
width/SVL ratio 0.34-0.42 (m = 0.371), male head
width/SVL ratio 0.34-0.44 (m = 0.380), sexually
dimorphic (t = -2.832, df = 90.6, P = 0.006).
Female eye-nostril distance/SVL ratio 0.08-0.11
(m = 0.093), male eye-nostril distance/SVL ratio
0.08-0.10 (m = 0.094), not sexually dimorphic
(t = -0.832, df = 63.2, P = 0.409). Female (N = 32)
tympanum diameter/SVL ratio 0.06-0.09
(m = 0.072), male (N = 77) tympanum diameter/
SVL ratio 0.05-0.08 (m = 0.071), not sexually
dimorphic (t = 0.890, df = 57.7, P = 0.377). Female
thigh/SVL ratio 0.35-0.45 (m = 0.405), male thigh/
SVL ratio 0.35-0.46 (m = 0.414), not sexually
316
Arquivos de Zoologia
dimorphic (t = -1.780, df = 67.3, P = 0.080).
Female shank/SVL ratio 0.38-0.46 (m = 0.419),
male shank/SVL ratio 0.38-0.48 (m = 0.424), not
sexually dimorphic (t = -1.368, df = 80.9,
P = 0.175). Female foot/SVL ratio 0.40-0.50
(m = 0.444), male (N = 76) foot/SVL ratio
0.38-0.51 (m = 0.444), not sexually dimorphic
(t = -0.050, df = 73.8, P = 0.960).
Larvae – At least the later stage larvae live
and feed in streamside and isolated forest ponds in
the central Amazon (Hero, 1990). Larvae elongate;
spiracle sinistral; vent median; oral disk entire,
almost terminal, with single row of marginal
papillae with broad anterior gap; tooth row formula
2(2)/2-3(1); maximum total length, stage 40, 69 mm
(Hero, 1990).
Advertisement Call – Calls of single notes,
given at rates of 16-66 calls/min; call duration
0.16-0.43 s; calls pulsed, 6-14 pulses/call, mean
pulse rates among individuals 26-38 pulses/s; call
frequency modulated, rising whoop, mean initial
frequency among individuals 260-540 Hz;
dominant frequency 340-690 Hz (Table 6, Figures
5, 6).
Karyotype – Diploid number 22,
fundamental number 44, commonly with 3 pairs of
metacentrics, 4 pairs of submetacentrics, and 4 pairs
of subtelocentrics (Heyer, 1972).
Habitat and Distribution – Leptodactylus
knudseni occurs in primary rain forest, secondary
forests, and open habitats throughout the
Amazonian Morphoclimatic Domain as defined by
Ab’Sáber (1977) (Appendix, Figure 15).
“Urucum de Corumba, Matto Grosso, Brazil,”
(locality now in Mato Grosso do Sul) 19°10’S,
57°39’W. Holotype: FMNH 9240, adult female.
Leptodactylus labyrinthicus (Spix, 1824)
English name – Pepper frog.
Portuguese name – Rã-pimenta.
Spanish name – Rana pimienta.
Diagnosis – Leptodactylus labyrinthicus
occurs in open formation habitats, including cerrado
enclaves in Amazonia. The species most likely to
be confused with L. labyrinthicus are L. knudseni,
L. paraensis, and L. vastus. Most adult
L. labyrinthicus have a distinctive labyrinthine belly
pattern; adult L. knudseni lack this pattern. The
advertisement call of L. labyrinthicus is unpulsed;
the advertisement call of L. knudseni is pulsed.
Leptodactylus paraensis is documented only from
closed canopy rain forest. There is no consistent
morphological feature that completely diagnoses
L. labyrinthicus from L. paraensis. Leptodactylus
labyrinthicus is somewhat larger (males
117-188 mm SVL, females 124-166 mm SVL) than
L. paraensis (males 94-170 mm SVL, females
102-154 mm SVL).
Adult Characteristics – Lip pattern often
dark elongate triangular marks on lip edge, fading
towards eye, 2 dark marks under eye; or broad
irregular darker and lighter bands, 2 dark bands
entering eye (Figure 12D). Lip pattern rarely
uniform light; or dark triangular marks on edge of
upper lip only; or dark triangular marks, 1 or 2
elongate approaching or entering lower eye (Figure
12A); or dark with two narrow light chevrons from
lip entering eye (Figure 12B); or alternating broad
light bands and narrow dark vertical stripes, 1-3
dark stripes entering eye (Figure 12C).
Rana labyrinthica Spix, 1824:31. Type locality: Rio
de Janeiro (State), Brazil. Holotype: ZSM
2501/0, lost (Hoogmoed and Gruber, 1983).
Leptodactylus wuchereri Jiménez de la Espada,
1875:68. Type locality: “Republica Argentina,”
somewhere between Montevideo, Uruguay and
Santiago, Chile. Holotype: MNCN 1694,
juvenile.
Leptodactylus bufo Andersson, 1911:1. Type
locality: “Ponta Grossa, Paraná, Brazil,”
25°06’S, 50°10’W. Holotype: NRM 1495,
male.
Leptodactylus pentadactylus mattogrossensis
Schmidt and Inger, 1951:444. Type locality:
Dorsal pattern often uniform light or dark; or with
a series of regularly or irregularly placed small, dark
spots; or with series of irregularly spaced, illdefined small blotches; or with single broad dark
chevron posterior to interorbital bar followed by a
series of fairly regularly spaced large dark spots.
Dorsal pattern rarely uniform light with 1-2 welldefined, narrow transverse bands; or well developed
dark interorbital band/chevron and 2 moderate to
large equally intense dark chevrons, second chevron
in sacral region, chevrons confluent or not; or two
broad dark transverse bands between interorbital
bar and arm insertion area, rest of dorsum with large
spots.
Vol. 37(3), 2005
317
Figure 16. Distribution map for Leptodactylus labyrinthicus (dots), L. turimiquensis (triangles), and L. vastus (squares). Question
marks indicate specimen localities from which individuals are currently unidentifiable to species.
Belly usually labyrinthine (Figure 13D). Belly often
light with dark vermiculations. Belly rarely light
(no pattern); or with lateral and anterior mottling
only; or mottled (Figure 13A) or uniform dark; or
dark with large light vermiculations (Figure 13C).
Posterior thigh pattern usually dark with large light
vermiculations, some coalescing of vermiculations.
Thigh often dark with small distinct light
vermiculations or spots (Figure 14A); or mostly
very distinctly light with few irregular dark marks
(Figure 14D). Thigh rarely dark with distinct,
discrete light spots; or dark with contrasting large
light irregular blotches/spots extending from light
dorsal transverse bars, rest of thigh dark with light
vermiculations (Figure 14B).
Dorsolateral folds usually interrupted from at
least 1/2 to full distance from eye to sacrum.
Dorsolateral folds often interrupted from at least 1/4
distance to 1/2 distance from eye to sacrum.
Dorsolateral folds rarely absent; or entire from at least
1/4 to full distance from eye to sacrum; or interrupted
to at least between sacrum and some distance to groin.
Flank folds usually with dark spot/wart in area
where fold would be between tympanum and
shoulder. Flank folds often absent. Flank folds
rarely interrupted from tympanum to lower flank;
or entire from tympanum to lower flank.
Male thumb usually with one large spine.
Male thumb often with one tiny to small spine. Male
thumb rarely with one large spine and a prepollical
bump.
Large breeding males with chest spines.
Female (N = 16) SVL 124.0-166.0 mm
(m = 141.6), male (N = 40) SVL 117.1-188.0 mm
(m = 149.6), not sexually dimorphic (t = -1.892,
df = 42.6, P = 0.065). Female head length/SVL
ratio 0.34-0.39 (m = 0.364), male head length/SVL
ratio 0.33-0.40 (m = 0.369), not sexually dimorphic
(t = -1.062, df = 25.9, P = 0.298. Female head
width/SVL ratio 0.36-0.45 (m = 0.389), male head
width/SVL ratio 0.35-0.45 (m = 0.392), not
sexually dimorphic (t = -0.480, df = 24.8,
P = 0.635). Female eye-nostril distance/SVL ratio
0.08-0.10 (m = 0.90), male eye-nostril distance/
SVL ratio 0.08-0.10 (m = 0.091), not sexually
dimorphic (t = -0.828, df = 40.0, P = 0.412).
Female tympanum diameter/SVL ratio 0.06-0.08
(m = 0.069), male tympanum diameter/SVL ratio
318
Arquivos de Zoologia
0.06-0.08 (m = 0.070), not sexually dimorphic
(t = -0.328, df = 30.8, P = 0.745). Female thigh/
SVL ratio 0.40-0.47 (m = 0.435), male thigh/SVL
ratio 0.33-0.48 (m = 0.425), not sexually dimorphic
(t = 1.395, df = 44.5, P = 0.170). Female shank/
SVL ratio 0.40-0.47 (m = 0.438), male shank/SVL
ratio 0.39-0.48 (m = 0.436), not sexually dimorphic
(t = 0.292, df = 32.9, P = 0.772). Female foot/SVL
ratio 0.39-0.47 (m = 0.434), male (N = 39) foot/
SVL ratio 0.39-0.50 (m = 0.439), not sexually
dimorphic (t = -0.636, df = 34.1, P = 0.529).
Larvae – Larvae occur in temporary ponds.
Larvae elongate; spiracle sinistral; vent median; oral
disk almost terminal, entire, with single row of
marginal papillae with broad anterior gap; tooth row
formula 1/2(1); maximum total length stage 40,
80 mm (Vizotto, 1967).
Advertisement Call – Calls of single notes,
given at rates of 35-50 calls/min; call duration
0.14-0.21 s; calls unpulsed; call frequency
modulated, rising whoop, mean initial frequency
among individuals 258-292 Hz; dominant
frequency about 430 Hz (Table 19, Figure 11A).
Karyotype – Diploid number 22, 3 pairs of
metacentrics, 4 pairs of submetacentrics, and 4 pairs
of small metacentrics or submetacentrics; nucleolus
organizer region on short arms of chromosome pair
8 (Denaro, 1972; Silva et al., 2000).
Habitat and Distribution – Leptodactylus
labyrinthicus occurs in subtropical and tropical
open formations, including much of the Cerrado
Morphoclimatic Domain as defined by Ab’Sáber
(1977) and cerrado-like enclaves in tropical rain
forests. The species occurrence suggests that it is a
good colonizer of man-made open habitats in
formerly closed-forest habitats (Appendix, Figure
16).
Leptodactylus myersi Heyer, 1995
Diagnosis – Leptodactylus myersi is most
likely to be confused with L. knudseni. Leptodactylus
myersi occurs in the same geographic region as
L. knudseni, but L. myersi is typically limited to rocky
outcrops and L. knudseni is not. Large reproductively
active males of L. myersi lack chest spines; such
males of L. knudseni have chest spines. Juvenile
L. myersi have brilliant red on their venters and
posterior thigh surfaces; juvenile L. knudseni do not
have such red coloration, but often have obvious
dorsal green coloration not found in L. myersi.
Adult Characteristics – Lip pattern often
with dark triangular marks on edge of upper lip only;
or dark triangular marks, 1 or 2 elongate
approaching or entering lower eye (Figure 12A);
or dark elongate triangular marks on lip edge, fading
towards eye, 2 dark marks under eye; or broad
irregular darker and lighter bands, 2 dark bands
entering eye (Figure 12D); or broad light stripe,
regular above, regularly or irregularly defined
below (Figure 12E). Lip pattern rarely uniform
light; or dark with two narrow light chevrons from
lip entering eye (Figure 12B).
Dorsal pattern usually with well developed dark
interorbital band/chevron and 2 moderate to large
equally intense dark chevrons, second chevron in
sacral region, chevrons confluent or not. Dorsal
pattern often with series of regularly spaced, welldefined small blotches. Dorsal pattern rarely
uniform light or dark; or with a series of regularly
or irregularly placed small, dark spots; or uniform
light with 1 or 2 well-defined, narrow transverse
bands; or more than 2 dark, broad transverse bands
of equal intensity in addition to interorbital band,
confluent laterally or not; or with irregular
quadrangular or rectangular markings, of equal or
alternating lighter/darker intensity, confluent
laterally or not; or with series of irregularly spaced,
ill-defined small blotches.
frog.
Belly pattern often dark with small light
vermiculations (Figure 13B); or dark with large
light vermiculations (Figure 13C); or dark with
small light discrete spots. Belly pattern rarely light
(no pattern); or with lateral and anterior mottling
only; or mottled (Figure 13A) or uniformly dark.
Suggested Portuguese name – Rã-de-dedosdelgados-de-Myers.
Suggested Spanish name – Rana de dedos
delgados de Myers.
Posterior thigh pattern usually dark with contrasting
large, light irregular blotches/spots extending from
light dorsal transverse bars, rest of thigh dark with
Leptodactylus myersi Heyer, 1995:712. Type
locality: Mucajaí, Roraima, Brazil, 2°25’N,
60°55’W. Holotype: MZUSP 66089, male.
Suggested English name – Myers’ thin-toed
Vol. 37(3), 2005
319
Figure 17. Distribution map for Leptodactylus myersi (dots) and MZUSP 24947-24948 (question mark), currently unidentifiable
to species.
light vermiculations (Figure 14B). Thigh rarely
distinctly uniform dark; or relatively uniform dark or
indistinctly mottled; or dark with small distinct light
vermiculations or spots (Figure 14A); or dark with
large distinct light vermiculations, some coalescing
of vermiculations; or labyrinthine; or dark with
distinct, discrete light spots; or mostly very distinctly
light with few irregular dark marks (Figure 14D).
Dorsolateral folds often absent; or interrupted
from at least 1/2 to full distance from eye to sacrum;
or entire from at least 1/4 to full distance from eye to
sacrum; or interrupted to at least between sacrum
and some distance to groin. Dorsolateral folds rarely
interrupted from at least 1/4 distance to 1/2 distance
from eye to sacrum; or entire to at least between
sacrum and some distance to groin.
Flank folds often absent; or interrupted from
tympanum to shoulder; or entire from tympanum
to shoulder. Flank folds rarely entire from
tympanum to lower flank.
Male thumb usually with one large spine.
Male thumb rarely with one tiny to small spine.
Large breeding males without chest spines.
Female (N = 5) SVL 78.9-112.9 mm
(m = 103.2), male (N = 20) SVL 74.2-123.4 mm
(m = 101.0), not sexually dimorphic (t = 0.298,
df = 6.9, P = 0.774). Female head length/SVL ratio
0.37-0.38 (m = 0.378), male head length/SVL ratio
0.33-0.40 (m = 0.376), not sexually dimorphic
(t = 0.431, df = 20.0, P = 0.671). Female head
width/SVL ratio 0.39-0.42 (m = 0.401), male head
width/SVL ratio 0.36-0.44 (m = 0.399), not
sexually dimorphic (t = 0.181, df = 12.8,
P = 0.859). Female eye-nostril distance/SVL ratio
0.07-0.09 (m = 0.087), male eye-nostril distance/
SVL ratio 0.07-0.10 (m = 0.091), not sexually
dimorphic (t = -1.049, df = 5.5, P = 0.338). Female
(N = 4) tympanum diameter/SVL ratio 0.07-0.08
(m = 0.077), male (N = 14) tympanum diameter/
SVL ratio 0.07-0.09 (m = 0.077), not sexually
dimorphic (t = -0.168, df = 4.8, P = 0.874). Female
thigh/SVL ratio 0.43-0.46 (m = 0.441), male
(N = 16) thigh/SVL ratio 0.40-0.45 (m = 0.431),
not sexually dimorphic (t = 1.864, df = 9.2,
P = 0.095). Female shank/SVL ratio 0.41-0.44
(m = 0.424), male (N = 16) shank/SVL ratio
0.38-0.44 (m = 0.413), not sexually dimorphic
(t = 1.502, df = 6.8, P = 0.178). Female foot/SVL
ratio 0.42-0.47 (m = 0.444), male (N = 16) foot/
SVL ratio 0.40-0.48 (m = 0.438), not sexually
dimorphic (t = 0.621, df = 6.8, P = 0.555).
Larvae – Unknown.
Advertisement Call – Calls of single notes,
given at a rate of 36 calls/min; call duration
0.33-0.36 s; calls pulsed, 2-3 pulses/call, pulse rates
5-9 pulses/s; call frequency modulated, rising
whoop, initial frequency about 190 Hz, dominant
320
Arquivos de Zoologia
frequency 600-690 Hz (data from compact disk
recording, Guide Sonore des Amphibiens Anoures
de Guyane by C. Marty and P. Gaucher).
Karyotype – Unknown.
Habitat and Distribution – Leptodactylus
myersi is restricted to rocky outcrop habitats in
French Guiana, Suriname, and northern Brazil
(Acre, Pará, Roraima) (Appendix, Figure 17).
Leptodactylus paraensis sp. nov.
Figure 18
Holotype. – MZUSP 69321, an adult male
from Brazil; Pará, Serra de Kukoinhokren, 07°46’S,
51°57’W. Collected by Miguel T. Rodrigues and
Barbara Zimmerman on 17 November 1992.
Paratopotypes. – MZUSP 69318, 69320,
69322, USNM 559809 (one female and three males
respectively), collected by Miguel T. Rodrigues and
Barbara Zimmerman on 17 November 1992; 70365
(juvenile), collected by Andrew A. Chek on 1
December 1993; MZUSP 70549 (female), collected
by Miguel T. Rodrigues on 2 November 1994;
MZUSP 70918 (juvenile), collected by Andrew A.
Chek, 22 February – 12 March 1995.
Figure 18. Holotype of Leptodactylus paraensis sp. nov.
Suggested English name – Pará thin-toed
frog.
Suggested Portuguese name – Rã-de-dedosdelgados-do-Pará.
Suggested Spanish name – Rana de dedos
delgados de Pará.
Diagnosis – Leptodactylus paraensis has an
eastern Amazonian distribution, occurring in the
same general region with L. knudseni, L. myersi,
and L. pentadactylus. Leptodactylus paraensis is
morphologically most similar to L. labyrinthicus,
L. turimiquensis, and L. vastus, of which only
L. labyrinthicus has a possibly overlapping
distribution. The only habitat data for L. paraensis
indicate that the species is limited to primary rain
forests as is L. pentadactylus, whereas
L. labyrinthicus is only found in open formations
and L. myersi is restricted to rocky outcrops. The
dorsolateral folds in L. paraensis are interrupted
and extend no further from the eye than to the
sacrum; the dorsolateral folds of L. pentadactylus
are either continuous or, if interrupted, extend
beyond the sacrum. Large, sexually active male
L. paraensis have a pair of black chest spines;
L. myersi lacks chest spines. Leptodactylus
paraensis is smaller (males 99-129 mm SVL,
Vol. 37(3), 2005
females 110-140 mm SVL) than both L. knudseni
(males 94-170 mm SVL, females 102-154 mm
SVL) and L. labyrinthicus (males 117-188 mm
SVL, females 124-166 mm SVL). Most L. knudseni
have continuous dorsolateral folds. Juvenile
L. paraensis do not have any green color in life;
juvenile L. knudseni often have green color in life
on the dorsum.
Description of Holotype – Snout nearly
rounded in dorsal and profile views. Canthus rostralis
indistinct. Loreal obtusely weakly concave.
Tympanum distinct, greatest diameter about 7/8 eye
diameter. Vomerine teeth in 2 strongly arched series,
between and extending posterior to choanae,
narrowly separated medially. Vocal slits elongate,
parallel to lower jaw. Vocal sac not visible externally.
Finger lengths in increasing order II < IV < I < III.
Inner surfaces of fingers II and III strongly ridged,
otherwise fingers smooth. Metacarpal tubercles
large; inner triangular-rounded, almost as large as
cordiform outer tubercle. Arms hypertrophied. Each
thumb with 1 large sharp black spine. One pair of
well-developed black chest spines, each with 3-4
cusps. Throat, chest, ventral arm surfaces, and
anterior belly with many small black tubercles.
Dorsum with scattered pustules, often capped with
a small tubercle. Tympanic fold well-developed from
eye to posterior 1/2 of vertical diameter of
tympanum, then weakly developed to above arm
insertion area. Two very short ridges, 1 behind eye,
1 above arm insertion area in dorsolateral fold field.
Lateral fold weak, interrupted. No flank fold. Flanks
glandular appearing. Venter smooth except for small
tubercles described above and ventral-posterior thigh
surfaces granular. Belly disk fold only weakly defined
posteriorly. Toe tips rounded or very slightly swollen.
Toes with very weak ridges on inner sides of toes I
and II, otherwise smooth. Subarticular tubercles
moderately pungent, ovoid. Outer metatarsal tubercle
ovoid, about 2/3 size of ovoid inner metatarsal
tubercle. Tarsal fold well-developed, straight,
extending about 3/4 length of tarsus. No metatarsal
fold. Upper shank with a few scattered pustules.
Outer tarsus with a few scattered light spots, but spots
not raised into distinct tubercles. Sole of foot with 3
low, rounded whitish tubercles.
Upper lip with 3 dark triangular marks,
fading dorsally, middle mark extending to anterior
portion of lower eye. Dorsum with light tan
interorbital bar followed by darker, broader
brownish-gray interorbital band; elongate
321
transverse dark brownish-gray spot above arm
insertion area followed by series of large irregular
brownish-gray spots on a lighter background,
pattern most distinct past sacrum. Flanks coarsely
marbled. Upper arms with indistinct cross-bands,
upper legs with well-defined, regular dark cross
bands. Throat dark with indistinct lighter marks.
Chest finely mottled. Anterior belly with large light
vermiculations on a slightly darker background,
posterior belly with faint but distinct labyrinthine
pattern. Ventral limb surfaces with indistinct bold
mottling, most distinct on distal portions of thighs.
Posterior thigh surfaces dark with scattered light
vermiculations, some light vermiculations
extending from dark dorsal transverse bands.
Measurements (mm): SVL 128.7, head
length 48.1, head width 49.9, eye-nostril distance
11.6, greatest tympanum diameter 9.6, thigh length
56.8, shank length 54.0, foot length 55.2.
Adult Characteristics – Lip pattern often
with 1 or 2 dark elongate triangular marks
approaching or entering the lower eye (Figure 12A);
or dark elongate triangular marks on lip edge, fading
towards eye, 2 dark marks under eye; or dark with
two narrow light chevrons from lip entering eye
(Figure 12B). Lip pattern rarely patternless; or
alternating broad light bands and narrow dark
vertical stripes, 1-3 dark stripes entering eye (Figure
12C); or broad light stripe, regular above, regularly
or irregularly defined below (Figure 12E).
Dorsal pattern often uniform light or dark; or well
developed dark interorbital band/chevron and 2
moderate to large equally intense dark chevrons,
second chevron in sacral region, chevrons confluent
or not; or more than 2 dark, broad transverse bands
of equal intensity in addition to interorbital band,
confluent laterally or not; or single broad dark
chevron posterior to interorbital bar followed by a
series of fairly regularly spaced large dark spots.
Dorsal pattern rarely with series of regularly or
irregularly placed small, dark spots; or 2 broad dark
transverse bands between interorbital bar and arm
insertion area, rest of dorsum with large spots; or
irregular quadrangular or rectangular markings, of
equal or alternating lighter/darker intensity,
confluent laterally or not; or series of irregularly
spaced, ill-defined small blotches.
Belly pattern often mottled (Figure 13A) or uniform
dark; or dark with small light vermiculations (Figure
322
Arquivos de Zoologia
13B); or dark with large light vermiculations
(Figure 13C); or labyrinthine (Figure 13D). Belly
pattern rarely light with dark vermiculations; or dark
with small light discrete spots.
Posterior thigh pattern usually dark with small
distinct light vermiculations or spots (Figure 14A).
Thigh pattern often dark with large distinct light
vermiculations, some coalescing of vermiculations;
or dark with contrasting large light irregular
blotches/spots extending from light dorsal
transverse bars, rest of thigh dark with light
vermiculations (Figure 14B). Thigh rarely relatively
uniform dark or indistinctly mottled.
Dorsolateral folds usually interrupted from
at least 1/2 to full distance from eye to sacrum.
Dorsolateral folds often interrupted from at least
1/4 to 1/2 distance from eye to sacrum. Dorsolateral
folds rarely interrupted or entire to 1/4 distance from
eye to sacrum.
Flank folds often absent; or dark spot/wart in area
where fold would be between tympanum and
shoulder; or interrupted from tympanum to
shoulder.
Male thumb usually with 1 large spine. Male
thumb rarely with 1 tiny to small spine.
Large breeding males with chest spines.
Female (N = 7) SVL 110.8-139.8 mm
(m = 126.6), male (N = 12) SVL 99.1-128.7
(m = 117.3), not sexually dimorphic (t = 1.986,
df = 10.9, P = 0.073). Female head length/SVL
ratio 0.36-0.39 (m = 0.373), male head length/SVL
ratio 0.35-0.40 (m = 0.379), not sexually dimorphic
(t = -0.925, df = 16.8, P = 0.368). Female head
width/SVL ratio 0.36-0.40 (m = 0.378), male head
width/SVL ratio 0.37-0.47 (m = 0.398), sexually
dimorphic (t = -2.213, df = 16.7, P = 0.041).
Female eye-nostril distance/SVL ratio 0.08-0.09
(m = 0.088), male eye-nostril distance/SVL ratio
0.08-0.10 (m = 0.91), not sexually dimorphic
(t = -0.895, df = 10.6, P = 0.391). Female
tympanum diameter/SVL ratio 0.07-0.08
(m = 0.072), male tympanum diameter/SVL ratio
0.06-0.08 (m = 0.074), not sexually dimorphic
(t = -1.042, df = 16.9, P = 0.312). Female thigh/
SVL ratio 0.39-0.46 (m = 0.414), male thigh/SVL
ratio 0.37-0.46 (m = 0.418), not sexually dimorphic
(t = -0.376, df = 13.7, P = 0.713). Female shank/
SVL ratio 0.40-0.45 (m = 0.418), male shank/SVL
ratio 0.39-0.45 (m = 0.424), not sexually dimorphic
(t = -0.755, df = 13.0, P = 0.463). Female foot/SVL
ratio 0.40-0.47 (m = 0.432), male foot/SVL ratio
0.42-0.49 (m = 0.445), not sexually dimorphic
(t = -1.108, df = 10.3, P = 0.293).
Etymology – The species is named after the
Brazilian State of Pará, from which all known
specimens occur.
Larvae – Unknown.
Advertisement Call – Unknown.
Karyotype – Unknown.
Habitat and Distribution – Leptodactylus
paraensis is known only from rain forest habitats
in eastern Amazonia in the Brazilian State of Pará
(Appendix, Figure 1).
Leptodactylus pentadactylus (Laurenti, 1768)
Rana pentadactyla Laurenti, 1768:32. Type
locality: “Indiis,” corrected to Suriname by
Müller (1927:276). Neotype: RMNH 29559,
adult male.
Rana gigas Spix, 1824:25. Type locality: “in locis
paludosis fluminis Amazonum,” Brazil. Type:
ZSM 89/1921, now destroyed (Frost, 2002).
Rana coriacea Spix, 1824:29. Type locality: “aquis
lacustribus fluvii Amazonum,” Brazil. Type:
ZSM 2502/0, now destroyed (Frost, 2002).
Leptodactylus goliath Jiménez de la Espada,
1875:57. Type locality: “Archidona (Oriente del
Ecuador),” 00°55’S, 77°48’W. Lectotype:
MNCN 1691, adult female.
Leptodactylus macroblepharus Miranda Ribeiro,
1926:144. Type locality: “Manáos –
Amazonas,” Brazil, 03°04’S, 60°00’W.
Lectotype: MZUSP 377, adult male.
Leptodactylus pentadactylus dengleri Melin,
1941:51. Type locality: “Roque, Peru,”
06°24’S, 76°48’W. Lectotype: GNM 497.
Leptodactylus pentadactylus rubidoides
Andersson, 1945:47. Type locality: “Rio
Pastaza,” Ecuador. Holotype: NRM 1928,
juvenile female.
English name – Smoky jungle frog.
Portuguese name – Rã-defumada-da-selva.
Spanish name – Rana ahumado de la selva.
Vol. 37(3), 2005
Diagnosis – Leptodactylus pentadactylus
has an Amazonian distribution, overlapping
distributions with L. knudseni, L. labyrinthicus,
L. myersi, and L. paraensis. Leptodactylus
pentadactylus occurs in the tropical wet forests
themselves; L. labyrinthicus occurs only in open
formations (the forest canopy is not closed) within
Amazonia, L. myersi is limited to rocky outcrops,
L. knudseni occurs in primary rain forest, secondary
forest, and open habitats, the few data for
L. paraensis indicate that it is also limited to
primary rain forest habitat. Most L. pentadactylus
have a well-developed continuous pair of
dorsolateral folds from the eye to at least the
sacrum; almost no L. labyrinthicus and no
L. paraensis have such extensively developed folds.
Large, sexually active male L. pentadactylus
usually do not have a large black spine on each
thumb; large, sexually active male L. knudseni,
L. labyrinthicus, L. myersi, and L. paraensis have
a large black spine on each thumb. Large, sexually
active male L. pentadactylus lack chest spines;
large, sexually active L. knudseni, L. labyrinthicus,
and L. paraensis have a pair of black chest spines.
Adult Characteristics – Lip pattern usually
dark triangular marks, 1 or 2 elongate approaching
or entering lower eye (Figure 12A). Lip pattern
often dark triangular marks on edge of upper lip
only.
Dorsal pattern usually uniform light with 1 or 2
well-defined, narrow transverse bands. Dorsal
pattern often with irregular quadrangular or
rectangular markings, or equal or alternating lighter/
darker intensity, confluent laterally or not. Dorsal
pattern rarely uniform light or dark; or with a series
of regularly or irregularly placed small, dark spots.
Belly usually dark with small light vermiculations
(Figure 13B). Belly often dark with large light
vermiculations (Figure 13C). Belly rarely mottled
(Figure 13A) or uniform dark; or light with dark
reticulations; or dark with small light discrete spots;
or dark with large light discrete spots (Figure 13E).
Posterior thigh pattern usually dark with small
distinct light vermiculations or spots (Figure 14A).
Thigh pattern often with very dark irregular blotches
highlighted by light pin-stripe on dark background
(Figure 14C). Thigh pattern rarely uniform dark or
indistinctly mottled; or dark with large distinct light
323
vermiculations, some coalescing of vermiculations;
or labyrinthine; or dark with contrasting large, light
irregular blotches/spots extending from light dorsal
transverse bars, rest of thigh dark with light
vermiculations (Figure 14B).
Dorsolateral folds usually entire from eye
to groin. Dorsolateral folds rarely entire from at
least 1/4 to full distance from eye to sacrum; or
interrupted to at least between sacrum and some
distance to groin; or entire to at least between
sacrum and some distance to groin.
Flank folds often entire from tympanum to shoulder;
or interrupted from tympanum to lower flank; or
entire from tympanum to lower flank. Flank folds
rarely absent.
Male thumb usually with one tiny to small
spine. Male thumb often lacking spines. Male thumb
rarely with one spine.
Males lacking chest spines.
Female (N = 26) SVL 135.0-174.2 mm
(m = 154.5), male (N = 26) SVL 100.2-195.0
(m = 140.8), sexually dimorphic (t = 3.081,
df = 41.2, P = 0.004). Female head length/SVL
ratio 0.33-0.40 (m = 0.363), male head length/SVL
ratio 0.34-0.42 (m = 0.370), not sexually dimorphic
(t = -1.270, df = 49.8, P = 0.210). Female head
width/SVL ratio 0.36-0.42 (m = 0.394), male head
width/SVL ratio 0.36-0.46 (m = 0.402), not
sexually dimorphic (t = -1.421, df = 47.1,
P = 0.162). Female eye-nostril distance/SVL ratio
0.08-0.10 (m = 0.090), male eye-nostril distance/
SVL ratio 0.08-0.11 (m = 0.093), sexually
dimorphic (t = -2.161, df = 50.0, P = 0.036).
Female tympanum diameter/SVL ratio 0.06-0.07
(m = 0.064), male tympanum diameter/SVL ratio
0.06-0.08 (m = 0.067), not sexually dimorphic
(t = -1.565, df = 44.3, P = 0.125). Female thigh/
SVL ratio 0.38-0.48 (m = 0.435), male thigh/SVL
ratio 0.37-0.50 (m = 0.436), not sexually dimorphic
(t = -0.192, df = 48.1, P = 0.849). Female shank/
SVL ratio 0.41-0.50 (m = 0.449), male shank/SVL
ratio 0.39-0.54 (m = 0.455), not sexually dimorphic
(t = -0.804, df = 46.6, P = 0.426). Female foot/SVL
ratio 0.42-0.50 (m = 0.454), male foot/SVL ratio
0.42-0.55 (m = 0.459), not sexually dimorphic
(t = 0.684, df = 46.4, P = 0.497).
324
Arquivos de Zoologia
Larvae – Tadpoles complete their
development through metamorphosis in terrestrial
burrows (Hero, 1990). Larvae elongate; spiracle
sinistral; vent median; oral disk terminal, entire,
with single row of marginal papillae with broad
anterior gap; tooth row formula 1/2(1); maximum
total length, stage 39, 81 mm (Hero, 1990).
Advertisement Call – Calls of single notes,
given at rates of 4-37 calls/min; call duration
0.18-0.40 s; calls pulsed, 12-18 pulses/call, mean
pulse rate among individuals 44-65 pulses/s; call
frequency modulated, rising whoop, mean initial
frequency among individuals about 340-860 Hz;
mean dominant frequency among individuals about
680-1030 Hz (Tables 11, 12, Figure 19).
Karyotype – Diploid number 22, 2 pairs of
metacentrics, 6 pairs of submetacentrics, 3 pairs of
subtelocentrics (Heyer and Diment, 1974). The
karyotype data reported for L. pentadactylus by
Bogart (1974) are based on specimens of
L. flavopictus from Boracéia, São Paulo, Brazil and
specimens, presumably of L. pentadactylus, from
Peru (James P. Bogart, pers. comm.).
Habitat and Distribution – Leptodactylus
pentadactylus occurs in closed canopied rain forest
habitat throughout the Amazonian Morphoclimatic
Domain as defined by Ab’Sáber (1977) (Appendix,
Figure 20).
Figure 19. Wave form and audiospectrogram of advertisement
call of Leptodactylus pentadactylus, USNM recording 254, cut
3 from the Rio Juruá, Amazonas, Brazil.
Figure 20. Distribution map for Leptodactylus pentadactylus. Question mark indicates specimen from this locality questionably
identifiable to L. pentadactylus.
Vol. 37(3), 2005
Leptodactylus peritoaktites sp. nov.
Figure 21
Holotype – USNM 196739, an adult male
from Hacienda Equinox, 38 km NNW of Santo
Domingo de los Colorados, Esmeraldas, Ecuador,
1000’, 00°03’S, 79°20’W. Collected by James A.
Peters on 22 June 1954.
Suggested English name – Coastal Ecuador
smoky jungle frog.
Suggested Portuguese name – Rãdefumada-de-selva-costeira-de-Ecuador.
Suggested Spanish name – Rana ahumada
de la selva costera de Ecuador.
Diagnosis – Leptodactylus peritoaktites has
a parapatric distribution with L. rhodomerus;
L. peritoaktites occurs in the more southern wet
tropical forest region of Pacific coastal South
America than L. rhodomerus. Large sexually active
male L. peritoaktites have a single large spine on
each thumb; large sexually active male
L. rhodomerus either lack thumb spines or have
only a small spine on each thumb. Most
L. rhodomerus have dark with large light
Figure 21. Holotype of Leptodactylus peritoaktites sp. nov.
325
vermiculation belly patterns; no L. peritoaktites
have this pattern.
Description of Holotype – Snout nearly
rounded from above, rounded-obtuse in profile.
Canthus rostralis indistinct. Loreal region weakly
concave-obtuse. Tympanum distinct, greatest
diameter about 3/4 eye diameter. Vomerine teeth
in two strongly arched series, between and
extending posteriorly to choanae, narrowly
separated medially. Vocal slits elongate, parallel
to lower jaw. Vocal sac single, median. Finger
lengths in increasing order II ~ IV < I ~ III. Inner
sides of fingers II and III ridged, other lateral
surfaces smooth. Metacarpal tubercles large,
triangular-rounded inner more distinct and just
larger than ovoid bifid outer tubercle. Arms not
hypertrophied. Each thumb with 1 medium large
white spine. No other secondary sexual
characters. Dorsal texture smooth. Tympanic fold
distinct from eye to shoulder. Dorsolateral fold
entire from eye to groin. Lateral fold not
indicated. Flank fold extending from tympanic
fold to above posterior arm insertion region, then
interrupted to lower flank. Commissural glands
326
Arquivos de Zoologia
distinct. Flanks and posterior thighs with
extensive dark brown glands, gland in groin
squarish-oblong, about 20 x 40 mm. Ventral
texture smooth, posterior portion of ventral
surfaces of thighs weakly granular. Belly disk fold
weakly indicated posteriorly. Toe tips rounded,
not expanded. Sides of toes weakly ridged.
Subarticular tubercles weakly pungent, ovoid.
Outer metatarsal tubercle weakly developed,
rounded, about 1/4 size of distinct, elongate-ovate
inner metatarsal tubercle. No metatarsal fold.
Upper shank with several keratinized tubercles.
Outer tarsal surface smooth. Sole of foot smooth.
Upper lip with dark brown marks, 1
elongate, extending almost to eye. Dorsum brown
with 2 faintly outlined irregular transverse bands
in addition to faintly outlined interorbital bar.
Dorsolateral, tympanic, flank folds dark brown
highlighted. Upper arms with faint darker brown
cross bands, upper shanks almost uniform dark
brown, rest of upper legs with indistinct to distinct
darker brown transverse bands. Throat uniform
brown, rest of venter brown with moderate sized,
distinct dirty cream spots. Dark brown glands on
posterior surfaces of thighs in a wide swath from
vent sweeping down and laterally across thighs, rest
of posterior thighs with bold dirty cream mottle on
brown background.
Measurements (mm): SVL 146.3, head
length 52.7, head width 55.4, eye-nostril distance
13.1, greatest tympanum diameter 8.7, thigh length
60.0, shank length 62.6, foot length 64.4.
Adult Characteristics – Lip pattern usually
with dark triangular marks, 1 or 2 elongate
approaching or entering lower eye (Figure 12A).
Lip pattern often with dark triangular marks on edge
of upper lip only.
Dorsal pattern usually with irregular quadrangular
or rectangular markings, of equal or alternating
intensity, confluent laterally or not. Dorsal pattern
often uniform light with 1 or 2 well-defined, narrow
transverse bands. Dorsal pattern rarely with more
than 2 dark, broad transverse bands of equal
intensity in addition to interorbital band, confluent
laterally or not.
Belly usually dark with large light discrete spots
(Figure 13E). Belly often mottled (Figure 13A) or
uniform dark; or dark with small light
vermiculations (Figure 13B).
Posterior thigh pattern usually dark with contrasting
large light irregular blotches/spots extending from
light dorsal transverse bars, rest of thigh dark with
light vermiculations (Figure 14B). Thigh pattern
often relatively uniform dark or indistinctly mottled;
or dark with small distinct light vermiculations or
spots (Figure 14A); or dark with large distinct light
vermiculations, some coalescing or vermiculations;
or dark with distinct, discrete light spots.
Dorsolateral folds usually entire from eye
to groin. Dorsolateral folds often entire from at least
1/4 to full distance from eye to sacrum; or entire to
at least between sacrum and some distance to groin.
Flank folds usually entire from tympanum to lower
flank. Flank folds often entire from tympanum to
shoulder; or interrupted from tympanum to lower
flank.
Male thumb usually with 1 moderate to large
spine; male thumb often lacking spines.
Males without chest spines.
Female (N = 5) SVL 115.3-133.1 mm
(m = 121.0), male (N = 3) SVL 124.0-146.3
(m = 132.4), not sexually dimorphic (t = -1.488,
df = 2.8, P = 0.238). Female head length/SVL ratio
0.34-0.38 (m = 0.364), male head length/SVL ratio
0.36-0.39 (0 = 0.379), not sexually dirmorphic
(t = -1.297, df = 4.0, P = 0.265). Female head
width/SVL ratio 0.36-0.38 (m = 0.372), male head
width/SVL ratio 0.38-0.40 (m = 0.389), not
sexually dimorphic (t = -3.024, df = 3.0,
P = 0.057). Female eye-nostril distance/SVL ratio
0.10 (m = 0.098), male eye-nostril distance/SVL
ratio 0.09-0.10 (m = 0.098), not sexually dimorphic
(t = -0.067, df = 2.5, P = 0.952). Female tympanum
diameter/SVL ratio 0.06-0.07 (m = 0.066), male
tympanum diameter/SVL ratio 0.06-0.07
(m = 0.066), not sexually dimorphic (t = 0.084,
df = 4.1, P = 0.937). Female thigh/SVL ratio
0.40-0.44 (m = 0.426), male thigh/SVL ratio
0.41-0.47 (m = 0.440), not sexually dimorphic
(t = -0.701, df = 3.1, P = 0.532). Female shank/
SVL ratio 0.46-0.47 (m = 0.462), male shank/SVL
ratio 0.43-0.47 (m = 0.455). Female foot/SVL ratio
0.46-0.49 (m = 0.474), male foot/SVL ratio
0.44-0.48 (m = 0.462), not sexually dimorphic
(t = 0.944, df = 3.3, P = 0.409).
Vol. 37(3), 2005
Etymology – From the Greek peritos – west,
and aktites – coast dweller, in allusion to the
geographic distribution of the species.
Larvae – Unknown.
Advertisement Call – Calls of single notes,
given at rates of 34-37 calls/min; call duration
0.20-0.30 s; calls pulsed, 5-8 pulses/call, mean
pulse rate among individuals 23-27 pulses/s; call
frequency modulated, rising whoop, mean initial
frequency among individuals about 550-690 Hz;
mean dominant frequency about 860 Hz (Table
11).
Habitat and Distribution – Leptodactylus
peritoaktites occurs inside rain forests and in
previously rainforested habitats in the middle and
southern rain forests of coastal Ecuador (Appendix,
Figure 22).
Leptodactylus rhodomerus sp. nov.
Figure 23
Holotype – ICN 13322, an adult male from
campamento Chancos, Vereda Campo Alegre,
Municipio de Restrepo, Valle de Cauca, Colombia,
327
460 m, 3°58’N, 76°44’W. Collected by John D.
Lynch on 3 June 1983.
Paratopotypes – ICN 13320-13321
(juveniles), exact same locality as for holotype,
collected by John D. Lynch on 23 and 27 May 1983
respectively; 13323 (female), taken slightly up road
from locality as for holotype at Quebrada La Mula,
collected by Juan Manuel Renjifo from 4-15
February 1984.
Suggested English name – Red-thighed thintoed frog.
Suggested Portuguese name – Rã-de-dedosdelgados-de-coxas-vermelhas.
Suggested Spanish name – Rana de dedos
delgados de muslos rojos.
Diagnosis – Leptodactylus rhodomerus is
most similar morphologically to L. pentadactylus,
which has an allopatric distribution with
L. rhodomerus; L. rhodomerus occurs in the wet
tropical forest regions of western Colombia and
adjacent Ecuador, L. pentadactylus occurs in the
Amazonian wet tropical forests. Leptodactylus
rhodomerus has bright red markings on the
posterior thigh surfaces in life; L. pentadactylus
does not have red on the thighs in life.
Figure 22. Distribution map for Leptodactylus peritoaktites (dots) and L. rhodomerus (triangles).
328
Arquivos de Zoologia
Figure 23. Holotype of Leptodactylus rhodomerus sp. nov.
Leptodactylus rhodomerus has a parapatric
distribution with L. savagei to the north and
L. peritoaktites to the south along Pacific coastal
South America. Large sexually active male
L. rhodomerus have either no thumb spines or a
single small spine on each thumb and lack chest
spines; large sexually active male L. savagei have
a single large black spine on each thumb and have
a pair of black chest spines. Leptodactylus
rhodomerus specimens often have very extensive
distinctly light areas (bright red in life) on the
posterior thigh surfaces; L. savagei individuals only
rarely have this pattern. Large sexually active male
L. peritoaktites have a single large black spine on
each thumb, contrasting with the condition that
occurs in L. rhodomerus described above. Most
L. rhodomerus have dark with large light
vermiculated belly patterns; no L. peritoaktites have
this pattern.
Description of Holotype – Snout nearly
rounded from above, rounded in profile. Canthus
rostralis indistinct. Loreal region obtuse-concave.
Tympanum distinct, greatest diameter about 2/3 eye
diameter. Vomerine teeth in strongly arched series,
between and extending posterior to choanae,
separated medially by about 1/4 length of single
vomerine tooth row. Vocal slits elongate, parallel
to lower jaw. Single vocal sac not expressed
externally. Finger lengths in increasing order
II ~ IV < I < III. Fingers with strong ridges
medially, weak ridges laterally. Inner metacarpal
tubercle protuberant, triangular-rounded in shape,
just smaller than weakly developed, rounded, heartshaped outer metacarpal tubercle. Arms not
hypertrophied. Thumb with a single tiny tan-tipped
spine; no other male secondary sexual
characteristics. Dorsal texture weakly shagreened
with small, scattered keratinized-tipped tubercles
in the post-sacral region. Tympanic fold welldeveloped from eye to above arm insertion.
Dorsolateral fold weak, but outlined with light and
dark striping from just behind eye complete to
above mid-thigh insertion. No lateral fold. Flank
fold distinct from supratympanic fold to lower
flanks about 3/5 distance posteriorly between limb
insertions; folds very glandular; right fold entire,
left fold with very short interruption above upper
arm insertion. Commissural glands distinct; very
extensive irregular-shaped glands in groin
extending anteriorly about 3/4 distance on flanks.
Ventral texture smooth except for areolate ventralposterior thighs. Belly fold weakly defined
anteriorly, well-defined posteriorly. Toe tips
rounded, just broader than toes immediately behind
tips. Toes with strong ridges medially, weak ridges
laterally; very vestigial webbing best developed
Vol. 37(3), 2005
between toes II-III-IV. Subarticular tubercles
moderately pungent, ovate. Low relief outer
metatarsal tubercle rounded, diameter about 2/3
length of ovate, well-developed, inner metatarsal
tubercle. Tarsal fold well-developed, slightly
curved, extending from medial surface of inner
metatarsal tubercle to about 3/4 length of tarsus.
No metatarsal fold. Upper shank scattered with tantipped tubercles. Outer tarsus profused with very
small white tubercles, among which, many large,
white, tan-tipped tubercles. Sole of foot with 3-6
moderate-sized white, tan-tipped tubercles.
Upper lip with 3 well-developed dark
triangular-shaped bars, middle triangle extending
to just short of lower eye. Dorsum generally almost
uniform brown with a well-defined dark-outlined
transverse interocular band and a moderately
defined narrow transverse band mid-dorsally above
arm insertion area; dorsolateral folds with welldefined lighter medial stripes and dark-outlined
stripes laterally. Upper limbs dark brown with
narrow darker brown incomplete or complete
transverse bands. Ventral surfaces dark with small,
well-defined light spots and vermiculations.
Posterior thigh with extensive light area with dark
brown irregular markings encroaching the light area
dorsally and ventrally.
Measurements (mm): SVL 134.1, head
length 50.4, head width 50.7, eye-nostril distance
13.1, greatest tympanum diameter 8.9, femur length
64.0, shank length 66.4, foot length 71.9.
Adult Characteristics – Lip pattern with
dark triangular marks, 1 or 2 elongate approaching
or entering lower eye (Figure 12A).
Dorsal pattern usually uniform light with one or
two well-defined, narrow transverse bands. Dorsal
pattern often with irregular quadrangular or
rectangular markings, of equal or alternating lighter/
darker intensity, confluent laterally or not.
Belly usually dark with large light vermiculations
(Figure 13C). Belly often dark with large light
discrete spots (Figure 13E). Belly rarely dark with
small light vermiculations (Figure 13B); or dark
with small light discrete spots.
Posterior thigh pattern often relatively uniform dark
or indistinctly mottled; or mostly very distinctly
light with few irregular dark marks (Figure 14D).
Thigh pattern rarely dark with small distinct light
329
vermiculations or spots (Figure 14A); or dark with
large distinct light vermiculations, some coalescing
of vermiculations; or dark with contrasting large
light irregular blotches/spots extending from light
dorsal transverse bars, rest of thigh dark with light
vermiculations (Figure 14B).
Dorsolateral folds usually entire from eye
to groin. Dorsolateral folds often entire to at least
between sacrum and some distance to groin.
Flank folds usually entire from tympanum to lower
flank. Flank folds often interrupted from tympanum
to lower flank. Flank fold rarely with dark spot/wart
in area where fold occurring in other specimens.
Male thumb with one tiny to small spine.
Males without chest spines.
Female (N = 5) SVL 133.5-157.8 mm
(m = 140.4), male (N = 6) SVL 112.2-143.8
(m = 132.0), not sexually dimorphic (t = 1.317,
df = 8.9, P = 0.221). Female head length/SVL ratio
0.36-0.40 (m = 0.381), male head length/SVL ratio
0.36-0.41 (m = 0.378), not sexually dimorphic
(t = 0.317, df = 8.8, P = 0.759). Female head width/
SVL ratio 0.38-0.40 (m = 0.384), male head width/
SVL ratio 0.35-0.39 (m = 0.378), not sexually
dimorphic (t = 0.737, df = 8.5, P = 0.481). Female
eye-nostril distance/SVL ratio 0.09-0.11
(m = 0.100), male eye-nostril distance/SVL ratio
0.10-0.11 (m = 0.101), not sexually dimorphic
(t = -0.252, df = 8.5, P = 0.807). Female tympanum
diameter/SVL ratio 0.06-0.07 (m = 0.064), male
tympanum diameter/SVL ratio 0.06-0.07
(m = 0.065), not sexually dimorphic (t = -0.680,
df = 7.9, P = 0.516). Female thigh/SVL ratio
0.46-0.50 (m = 0.478), male thigh/SVL ratio
0.39-0.48 (m = 0.455), not sexually dimorphic
(t = 1.454, df = 8.0, P = 0.184). Female shank/SVL
ratio 0.48-0.52 (m = 0.487), male shank/SVL ratio
0.36-0.50 (m = 0.460, not sexually dimorphic
(t = 1.236, df = 6.4, P = 0.260). Female foot/SVL
ratio 0.48-0.53 (m = 0.497), male foot/SVL ratio
0.44-0.54 (m = 0.496), not sexually dimorphic
(t = 0.096, df = 8.4, P = 0.926).
Etymology – From the Greek rhodon – rose,
and meros – thigh, in reference to the bright red
coloration on the posterior thighs in life in most
individuals.
330
Arquivos de Zoologia
Larvae – Unknown.
Advertisement Call – Unknown.
Karyotype – Unknown.
Habitat and Distribution – Leptodactylus
rhodomerus occurs in the Colombian Choco and
adjacent rain forests of Ecuador (Appendix, Figure 22).
Leptodactylus savagei sp. nov.
Figure 24
Holotype – USNM 227652, an adult male
from Rincon de Osa, Puntarenas, Costa Rica,
08°42’N, 83°29’W. Collected by Miriam H. and
W. Ronald Heyer on 11 June 1973.
Paratopotypes – USNM 219538, juvenile,
collected by Roy W. McDiarmid, 13 July 1973,
USNM 227645-227646, 227649, females,
collected by Heyers on 9 June 1973, USNM
227647-227648, 227650-227651, males, collected
by Heyers on 9 June 1973, USNM 227653-227654,
juveniles, collected by Heyers on 13 June 1973,
USNM 227655, juvenile, collected by Heyers on
15 June 1973, USNM 227656, female, collected
by Heyers on 19 June 1973.
Figure 24. Holotype of Leptodactylus savagei sp. nov.
Suggested English name – Savage’s thintoed frog.
Suggested Portuguese name – Rã-de-dedosdelgados-de-Savage.
Suggested Spanish name – Rana de dedos
delgados de Savage.
Diagnosis – Leptodactylus savagei is
morphologically very similar to L. knudseni. The
two species are allopatric in their distributions, with
L. savagei occurring only in northern Colombia
adjacent to Panama in South America, whereas
L. knudseni has an Amazonian distribution. There
are no adult morphological or advertisement call
characters that consistently diagnose these two
species from each other. Larval L. savagei have nine
filter rows per plate on ceratobranchial IV;
L. knudseni larvae have seven rows. Leptodactylus
savagei and L. rhodomerus are the only two species
of the L. pentadactylus group that occur in coastal
Pacific and northern Colombia. The known locality
data indicate parapatric distributions of these two
species, with L. savagei occurring from Colombia
adjacent to Panama to Honduras and L. rhodomerus
occurring in the Colombian Choco and just
extending into neighboring Ecuador. Large sexually
Vol. 37(3), 2005
active male L. savagei have a single large black
thumb spine and a pair of black chest spines; large
sexually active male L. rhodomerus have either no
thumb spines or a single small spine on each thumb
and lack chest spines. Leptodactylus rhodomerus
specimens often have very extensive distinctly light
areas (bright red in life) on the posterior thighs with
a few irregular dark markings; L. savagei specimens
rarely have this pattern.
Description of Holotype – Snout nearly
rounded from above, rounded-obtuse in profile.
Canthus rostralis indistinct. Loreal region obtusely
flared. Tympanum distinct, greatest diameter about
4/5 eye diameter. Vomerine teeth in two strongly
arched series, between and extending posteriorly
to choanae, narrowly separated medially. Vocal slits
elongate, parallel to lower jaw. Vocal sac not visible
externally. Finger lengths in increasing order
II ~ IV < I < III. Inner lateral surfaces of fingers I,
II, III strongly ridged, other lateral surfaces very
weakly ridged or smooth. Metacarpal tubercles low
and large, about equal in size; inner triangularrounded, outer strongly bifid. Arms moderately
hypertrophied. Each thumb with one large black
spine. Chest with pair of small black spines of a
single cusp. Throat, chest, under arms, and flanks
profused with small black tubercles. Dorsal texture
smooth. Tympanic fold well developed from eye
to shoulder. Dorsolateral folds well-developed,
continuous from eye to groin. Lateral fold glandular,
complete from supratympanic fold to groin. Flank
fold glandular, diverging from lateral fold extending
to lower flank. Flanks glandular with pronounced
glands, especially in groin region. Ventral texture
smooth except for small black tubercles described
above and posterior ventral thighs granular. Belly
disk fold weakly indicated posteriorly. Toe tips just
larger than toe widths immediately behind tips,
rounded. Toes ridged laterally, strongest on toes
I-IV. Subarticular tubercles weakly pungent, ovoid.
Outer metatarsal tubercle low, rounded-ovoid,
about 1/4 size of elongate ovoid inner metatarsal
tubercle. Tarsal fold distinct, straight, extending
about 5/6 length of tarsus. No metatarsal fold. Upper
shank with weak shagreen and scattered white
tubercles. Outer tarsus scattered with low, white
tubercles. Sole of foot smooth.
Upper lip with series of dark brown marks
next to jaw, bordered above by tan then becoming
darker brown. Commissural gland cream-tan.
Dorsum with a faint pattern of complex,
331
coalescing transverse bars/rectangles beginning
with interorbital chevron. Dorsolateral folds
weakly highlighted by light and dark pin-striping,
darker highlights most pronounced posteriorly.
Upper arm with one transverse darker brown
band. Upper legs with series of distinct dark
brown transverse bands on a lighter brown
background. Throat almost uniform gray, rest of
venter mottled gray/brown and dirty cream.
Posterior thigh surfaces almost black with a few
sinuous light marks, some extending from light
dorsal transverse bands and a few irregular light
spots, especially ventrally.
Measurements (mm): SVL 144.4, head
length 56.3, head width 60.3, eye-nostril distance
13.8, greatest tympanum diameter 10.8, thigh length
63.5, shank length 62.8, foot length 64.6.
Adult Characteristics – Lip pattern
usually dark triangular marks, one or two
elongate approaching or entering lower eye
(Figure 12A). Lip pattern often dark triangular
marks on edge of upper lip only. Lip pattern
rarely uniform light.
Dorsal pattern usually with irregular quadrangular
or rectangular markings, of equal or alternating
lighter/darker intensity, confluent laterally or not.
Dorsal pattern often uniform light or dark; or
uniform light with one or two well-defined, narrow
transverse bands. Dorsal pattern rarely with more
than two dark broad transverse bands of equal
intensity in addition to interorbital band, confluent
laterally or not.
Belly often dark with small light vermiculations
(Figure 13B); or dark with large light
vermiculations (Figure 13C); or dark with small
light discrete spots; or dark with large light discrete
spots (Figure 13E).
Posterior thigh pattern often dark with small distinct
light vermiculations or spots (Figure 14A); or dark
with large distinct light vermiculations, some
coalescing of vermiculations; or dark with
contrasting large light irregular blotches/spots
extending from light dorsal transverse bars, rest of
thigh dark with light vermiculations (Figure 14B).
Thigh pattern rarely labyrinthine; or dark with
distinct, discrete light spots; or mostly very
distinctly light with few irregular dark marks (Figure
14D).
332
Arquivos de Zoologia
Dorsolateral folds often entire from at least
1/4 to full distance from eye to sacrum; or entire to
at least between sacrum and some distance to groin;
or entire from eye to groin. Dorsolateral folds rarely
interrupted to at least between sacrum and some
distance to groin.
Flank folds often entire from tympanum to shoulder;
or interrupted from tympanum to lower flanks; or
entire from tympanum to lower flank. Flank folds
rarely with dark spot/wart in area where fold would
be between tympanum and shoulder.
Male thumb usually with one large spine.
Male thumb often with one tiny to small spine. Male
thumb rarely with one large spine and a prepollical
bump.
Large breeding males with chest spines.
Female (N = 74) SVL 110.2-164.1 mm
(m = 137.1), male (N = 75) SVL 106.0-156.3 mm
(m = 133.2), sexually dimorphic (t = 2.165,
df = 146.1, P = 0.032). Female head length/SVL
ratio 0.34-0.40 (m = 0.371), male head length/SVL
ratio 0.33-0.40 (m = 0.377), sexually dimorphic
(t = -3.049, df = 146.8, P = 0.003). Female head
width/SVL ratio 0.33-0.42 (m = 0.383), male head
width/SVL ratio 0.36-0.44 (m = 0.392), sexually
dimorphic (t = -3.880, df = 145.3, P = 0.000).
Female eye-nostril distance/SVL ratio 0.08-0.11
(m = 0.094), male eye-nostril distance/SVL ratio
0.08-0.10 (m = 0.094), not sexually dimorphic
(t = -0.269, df = 143.9, P = 0.789). Female
(N = 44) tympanum diameter/SVL ratio 0.06-0.08
(m = 0.069), male (N = 30) tympanum diameter/
SVL ratio 0.06-0.08 (m = 0.069), not sexually
dimorphic (t = -0.052, df = 50.1, P = 0.959).
Female thigh/SVL ratio 0.37-0.46 (m = 0.420),
male thigh/SVL ratio 0.37-0.47 (m = 0.425), not
sexually dimorphic (t = -1.129, df = 145.0,
P = 0.261). Female shank/SVL ratio 0.39-0.48
(m = 0.441), male shank/SVL ratio 0.40-0.49
(m = 0.444), not sexually dimorphic (t = -0.850,
df = 136.3, P = 0.397). Female (N = 71) foot/SVL
ratio 0.40-0.51 (m = 0.457), male (N = 74) foot/
SVL ratio 0.38-0.51 (m = 0.458), not sexually
dimorphic (t = -0.139), df = 139.8, P = 0.890).
Etymology – The species is named in honor
of Jay M. Savage for his substantial contributions
to furthering biological research in the Neotropics
in general and those of the Middle American
herpetofauna in particular.
Larvae – Larvae either live in temporary
ponds or complete development in burrows that the
foam nest was laid in and are facultative carnivores
(Heyer et al., 1975; Muedeking and Heyer, 1976).
Larvae elongate; spiracle sinistral; vent median; oral
disk almost terminal, entire; broad anterior gap in
oral disk lacking marginal papillae, followed by
single row of papillae laterally, double row
ventrolaterally, and single row ventrally; tooth row
formula 2(2)/3(1); maximum total length, stage 40,
83 mm (Heyer, 1970).
Advertisement Call – Calls of single notes,
given at rates of 40-49 calls/min; call duration
0.24-0.42 s, calls pulsed, 5-13 pulses/call, mean
pulse rate among individuals 31-46 pulses/s; call
frequency modulated, rising whoop, mean initial
frequency among individuals about 300-345 Hz;
mean dominant frequency among individuals about
350-520 Hz (Table 11, Figure 25).
Karyotype – Unknown.
Habitat and Distribution – Leptodactylus
savagei occurs in primary and secondary forests,
forest edges, and deforested areas from Honduras
(McCranie and Wilson, 2002:453) to northern
Colombia adjacent to Panama (Appendix, Figure
26).
Figure 25. Wave form and audiospectrogram of advertisement
call of holotype of Leptodactylus savagei, USNM recording
106, cut 1.
Vol. 37(3), 2005
333
Figure 26. Distribution map for Leptodactylus savagei.
Leptodactylus turimiquensis sp. nov.
Figure 27
Holotype – AMNH 70667, an adult male
from Caripito, Monagas, Venezuela, ~ 100 m,
10°08’N, 63°06’W. Collected by William Beebe
in 1942.
Paratopotype – AMNH 70668, adult male,
same data as for holotype.
Suggested English name – Calf frog.
Suggested Portuguese name – Rã-vitelo.
Suggested Spanish name – Rana ternero.
Diagnosis – Leptodactylus turimiquensis is
known from and proximate to the State of Sucre in
Venezuela and is the only large species of the
L. pentadactylus group that occurs in the region.
Leptodactylus turimiquensis is most similar to
L. labyrinthicus and L. vastus. The available data
indicate a substantial distributional hiatus between
L. turimiquensis and L. labyrinthicus and an
extensive hiatus between L. turimiquensis and
L. vastus. Leptodactylus turimiquensis occurs in
forested or previously forested habitats;
L. labyrinthicus and L. vastus occur exclusively in
open formation habitats. Leptodactylus
turimiquensis has a pulsed advertisement call; the
call of L. labyrinthicus is unpulsed. Leptodactylus
turimiquensis specimens rarely have a labyrinthine
belly pattern; most L. vastus have this pattern.
Description of Holotype – Snout rounded
from above, rounded-obtuse in profile. Canthus
rostralis indistinct. Loreal region concave-obtuse.
Tympanum distinct, greatest diameter about 2/3 eye
diameter. Vomerine teeth in strongly arched series,
between and extending posterior to choanae,
narrowly separated medially. Vocal slits elongate,
at obtuse angle to lower jaw. Vocal sac not expressed
externally. Finger lengths in increasing order
II ~ IV < I < III. Strong ridges on outer side of
finger I and inner sides of fingers II and III, other
sides smooth. Inner metacarpal tubercle
pronounced, triangular-rounded, about same size
as flat, ovoid, bifid outer metacarpal tubercle. Arms
greatly hypertrophied. Thumb with one large black
spine and a bump on the prepollex. Chest with one
large dark four-cusped spine (right) and one large
white three-cusped spine (left). Small dark tubercles
on finger ridges and inner dorsal surfaces of fingers
II and III, chest, anterior flanks and belly; throat
with many small white tubercles and a few brown
tubercles. Dorsum with scattered warts and short
ridges, more pronounced posteriorly. Tympanic fold
well developed from eye to above arm insertion.
No dorsolateral fold; a few small warts in dorsal
fold area. Lateral fold weak from tympanic fold to
posterior arm insertion area then a disconnected
series of warty glands to groin. No flank fold.
Commissural gland moderately developed, flanks
glandular. Ventral texture smooth except for small
tubercles described above and granular ventralposterior thighs. Belly disk fold moderately defined
posteriorly only. Toe tips rounded, just broader than
334
Arquivos de Zoologia
Figure 27. Holotype of Leptodactylus turimiquensis sp. nov.
toes immediately behind tips. Toes smooth laterally
except for barely discernible ridge on inner side of
toe II; barest trace of vestigial web between toes II
and III. Subarticular tubercles weakly pungent,
ovoid. Outer metatarsal tubercle low, round, about
1/3 size of elongate ovoid inner metatarsal tubercle.
Tarsal fold distinct, straight, about 2/3 length of
tarsus. No metatarsal fold. Upper shank with several
scattered white tubercles. Outer tarsus with a few
scattered white tubercles. Sole of foot smooth.
Upper lip with a series of ill-defined
vertical bars, just darker than background, 2 bars
extending to lower eye. Dorsum brown with
scattered small dark spots associated with warts/
short ridges. Upper arm with irregular, just darker
brown than background, transverse bands. Upper
legs with well defined broad darker brown
transverse bands on a lighter brown background.
Throat uniform dark brown. Chest light brown
with faint pattern of bold mottling. Belly and
ventral limb surfaces brown with large dirty cream
vermiculations. Posterior thigh with light narrow
elongate extensions from light dorsal bands, rest
of thigh dark brown with scattered large light dirty
cream vermiculations.
Measurements (mm): SVL 142.6, head
length 51.0, head width 59.2, eye-nostril distance
12.3, greatest tympanum diameter 9.3, thigh length
57.9, shank length 59.9, foot length 60.7.
Adult Characteristics – Lip pattern often
with dark triangular marks, one or two elongate
approaching or entering lower eye (Figure 12A);
or dark elongate triangular marks on lip edge, fading
towards eye, two dark marks under eye; or dark
with two narrow light chevrons from lip entering
eye (Figure 12B); or uniform dark.
Dorsal pattern often uniform light or dark; or with
series of regularly or irregularly placed small, dark
spots; or well-developed dark interorbital band/
chevron and two moderate to large equally intense
dark chevrons, second chevron in sacral region,
chevrons confluent or not; or two broad dark
transverse bands between interorbital bar and arm
insertion area, rest of dorsum with large spots; or
with single broad dark chevron posterior to
Vol. 37(3), 2005
interorbital bar followed by a series of fairly
regularly spaced large dark spots. Dorsal pattern
rarely with more than two dark, broad transverse
bands of equal intensity in addition to interorbital
band, confluent laterally or not.
Belly usually mottled (Figure 13A) or uniform dark.
Belly rarely dark with small light vermiculations
(Figure 13B).
Posterior thigh pattern usually dark with small
distinct light vermiculations or spots (Figure 14A).
Thigh pattern often dark with contrasting large light
irregular blotches/spots extending from light dorsal
transverse bars, rest of thigh dark with light
vermiculations (Figure 14B). Thigh pattern rarely
dark with large distinct light vermiculations, some
coalescing of vermiculations; or labyrinthine.
Dorsolateral folds often absent; or
interrupted or entire to 1/4 distance from eye to
sacrum; or interrupted from at least 1/4 distance to
1/2 distance from eye to sacrum; or interrupted from
at least 1/2 to full distance from eye to sacrum.
Dorsolateral folds rarely entire from eye to groin.
Flank folds usually with a dark spot/wart in
area where fold would be between tympanum and
shoulder. Flank folds often absent.
Male thumb usually with one spine. Male
thumb often with one spine and a prepollical bump.
Large breeding males with chest spines.
Female (N = 2) SVL 122.4-128.0 mm, male
(N = 23) SVL 127.2-160.0 mm (m = 144.0),
sexually dimorphic (t= -5.492, df = 2.2, P = 0.025).
Female head length/SVL ratio 0.35-0.36, male head
length/SVL ratio 0.34-0.39 (m = 0.370), not
sexually dimorphic (t = -1.860, df = 1.6,
P = 0.231). Female head width/SVL ratio
0.39-0.40, male head width/SVL ratio 0.39-0.46
(m = 0.418), not sexually dimorphic (t = -2.344,
df = 1.4, P = 0.198). Female eye-nostril distance/
SVL ratio 0.09-0.10, male eye-nostril distance/SVL
ratio 0.09-0.10 (m = 0.093), not sexually dimorphic
(t = -0.197, df = 1.5, P = 0.867). Female tympanum
diameter/SVL ratio 0.07, male (N = 17) tympanum
diameter/SVL ratio 0.06-0.08 (m = 0.070), not
sexually dimorphic (t = -0.183, df = 5.2,
P = 0.862). Female thigh/SVL ratio 0.42-0.43, male
thigh/SVL ratio 0.38-0.47 (m = 0.421), not sexually
335
dimorphic (t = 0.698, df = 22.7, P = 0.492). Female
shank/SVL ratio 0.40-0.42, male shank/SVL ratio
0.39-0.47 (m = 0.430), not sexually dimorphic
(t = -1.409, df = 1.3, P = 0.352). Female foot/SVL
ratio 0.42-0.44, male foot/SVL ratio 0.38-0.47
(m = 0.433), not sexually dimorphic (t = -0.458,
df = 1.5, P = 0.704).
Etymology – Jaime E. Péfaur, at my request,
kindly suggested naming this species
L. turimiquensis after the Serranía de Turimiquire,
which encompasses the known distribution of the
species (see Péfaur and Sierra, 1995: fig.2). Spanish
and English authors have transliterated the
indigenous name for the mountain range involved
as Turimiquire and Turumiquire. Although some
of the botanical and ornithological literature uses
Turumiquire, the name used in official papers of
the Republic of Venezuela is Turimiquire (J.E.
Péfaur, e-mail message of 5 August 2003).
Larvae – Unknown.
Advertisement Call – Calls of single notes;
call duration 0.33 s; calls pulsed, about 9 pulses/
call, mean pulse rate about 27 pulses/s; call weakly
frequency modulated; initial frequency about
400 Hz; dominant frequency about 400 Hz (Table
19, data extrapolated from Rivero and Esteves,
1969).
Karyotype – Unknown.
Habitat and Distribution – Leptodactylus
turimiquensis occurs in and proximate to the State
of Sucre, Venezuela and is found in relatively
undisturbed rain forests, second growth, and
agricultural clearings of previously forested habitats
(Péfaur and Sierra, 1995) (Appendix, Figure 17).
Leptodactylus vastus A. Lutz, 1930
Leptodactylus vastus A. Lutz, 1930:14. Type
locality: “Independencia,” now Guarabira,
Paraiba, Brazil, 06°51’S, 35°29’W. Lectotype:
AL-MN 70, adult male.
Suggested English name – Northeastern
pepper frog.
Suggested Portuguese name – Rã-pimentado-Nordeste.
Suggested Spanish name – Rana pimienta
del nordeste.
Diagnosis – Leptodactylus vastus is known
from northeast Brazil and is the only large species
336
Arquivos de Zoologia
of the L. pentadactylus group that occurs in the
region it occupies. The available data indicate a
modest hiatus in the distributions of L. vastus and
L. labyrinthicus and an extensive hiatus between
L. vastus and L. turimiquensis. Leptodactylus
vastus and L. labyrinthicus occur in open
formations; L. turimiquensis occurs in wet forest
or former wet forest habitats. Leptodactylus vastus
has a pulsed advertisement call; L. labyrinthicus
has an unpulsed advertisement call.
Adult Characteristics – Lip pattern often
dark elongate triangular marks on lip edge, fading
towards eye, two dark marks under eye (Figure
12B); or dark with two narrow light chevrons from
lip entering eye; or broad irregular darker and
lighter bands, two dark bands entering eye (Figure
12D). Lip pattern rarely dark triangular marks, one
or two elongate approaching or entering eye (Figure
12A); or alternating broad light bands and narrow
dark vertical stripes, one to three dark stripes
entering eye (Figure 12C); or uniform dark.
Dorsal pattern often uniform light or dark; or with a
series of regularly or irregularly placed small, dark
spots; or with series of irregularly spaced, ill-defined
small blotches; or with single broad dark chevron
posterior to interorbital bar followed by a series of
fairly regularly spaced large dark spots. Dorsal
pattern rarely uniform light with one or two welldefined, narrow transverse bands; or well developed
dark interorbital band/chevron and two moderate
to large equally intense dark chevrons, second
chevron in sacral region, chevrons confluent or not.
Belly usually labyrinthine (Figure 13D). Belly often
mottled (Figure 13A) or uniform dark; or light with
dark vermiculations. Belly rarely dark with light
vermiculations (Figure 13B, C).
Posterior thigh pattern often dark with small distinct
light vermiculations or spots (Figure 14A); or dark
with large distinct light vermiculations, some
coalescing of vermiculations; or labyrinthine; or
mostly very distinctly light with few irregular dark
marks (Figure 14D). Thigh pattern rarely dark with
contrasting large light irregular blotches/spots
extending from light dorsal transverse bars, rest of
thigh dark with light vermiculations (Figure 14B).
Dorsolateral folds usually interrupted from
at least 1/2 to full distance from eye to sacrum.
Dorsolateral folds often interrupted from at least
1/4 distance to 1/2 distance from eye to sacrum.
Dorsolateral folds rarely entire from at least 1/4 to
full distance from eye to sacrum; or interrupted to
at least between sacrum and some distance to groin.
Flank folds usually a dark spot/wart in area where
fold would be between tympanum and shoulder.
Flank folds often absent. Flank folds rarely entire
from tympanum to lower flank.
Male thumb usually with one large spine.
Male thumb rarely with one tiny to small spine.
Large breeding males with chest spines.
Female (N = 27) SVL 120.4-167.0
(m = 151.1), male (N = 28) SVL 110.6-180.3
(m = 156.0), not sexually dimorphic (t = -1.293,
df = 50.2, P = 0.202). Female head length/SVL
ratio 0.35-0.40 (m = 0.374), male head length/SVL
ratio 0.34-0.41 (m = 0.378), not sexually dimorphic
(t = -1.047, df = 52.9, P = 0.300). Female head
width/SVL ratio 0.36-0.42 (m = 0.398), male head
width/SVL ratio 0.38-0.45 (m = 0.410), sexually
dimorphic (t = -2.857, df = 52.9, P = 0.006).
Female eye-nostril distance/SVL ratio 0.08-0.10
(m = 0.089), male eye-nostril distance/SVL ratio
0.08-0.10 (m = 0.089), not sexually dimorphic
(t = -0.586, df = 51.9, P = 0.560). Female
tympanum diameter/SVL ratio 0.06-0.08
(m = 0.068), male tympanum diameter/SVL ratio
0.06-0.08 (m = 0.070), not sexually dimorphic
(t = -0.843, df = 46.2, P = 0.404). Female thigh/
SVL ratio 0.34-0.46 (m = 0.416), male thigh/SVL
ratio 0.39-0.48 (m = 0.435), sexually dimorphic
(t = -3.182, df = 50.7, P = 0.003). Female shank/
SVL ratio 0.37-0.46 (m = 0.413), male shank/SVL
ratio 0.40-0.46 (m = 0.423), not sexually dimorphic
(t = -1.853, df = 49.8, P = 0.070). Female foot/SVL
ratio 0.38-0.47 (m = 0.423), male foot/SVL ratio
0.37-0.47 (m = 0.432), not sexually dimorphic
(t = -1.538, df = 52.7, P = 0.130).
Larvae – Unknown.
Advertisement Call – Calls of single notes,
given at rates of 54-61 calls/min; call duration
0.14-0.19 s; calls pulsed, 4-6 pulses/call, mean
pulse rate among individuals 23-31 pulses/s; call
frequency modulated, rising whoop, initial
frequency about 260 Hz; dominant frequency about
430 Hz (Table 19, Figure 11B).
Vol. 37(3), 2005
Karyotype – Unknown.
Habitat and Distribution – Leptodactylus
vastus occurs in tropical open formations, including
the Caatinga Morphoclimatic Domain and adjacent
northern portion of the Cerrado Morphoclimatic
Domain as defined by Ab’Sáber (1977) (Figure 17).
DISCUSSION
Relationships and zoogeography are being
treated separately (de Sá and Heyer, in preparation).
Sexual Dimorphism
Standard statistical tests with standard
significance levels indicate that there are several
sexually dimorphic features in size and shape in the
species included in this study. However, determining
whether the statistically significant sexually
dimorphic features are biologically significant is not
straightforward. The examples in the following two
paragraphs describe the problems involved.
Males and females are sexually dimorphic
in size, with females being larger in most species
of frogs (Duellman and Trueb, 1986:54). Some
feature of reproductive biology is typically
associated with situations in which there is no sexual
dimorphism or the male is larger, such as vigorous
male territorial defense or male parental care
(Shine, 1974). Thus, the statistical results of no
dimorphism in L. knudseni and L. vastus and larger
males in L. labyrinthicus, for which sample sizes
are quite reasonable, are unusual and suggest that
some interesting aspect of their reproductive
biology is involved. The L. pentadactylus t- test
results for SVL indicate that females are statistically
larger than males. However, there is one male that
is much larger than any female measured (195 mm
SVL, 174 mm SVL, respectively). This finding
suggests two interrelated possibilities: (1) males are
biologically significantly larger than females; and/
or (2) there might be a strong sampling bias in terms
of field collecting of adult L. pentadactylus. For
example, suitable burrows for foam nest placement
and larval development may be very limited and
only the largest males can defend these sites.
Younger and smaller males may be prevented from
having access to burrows and spend much more of
their time on the forest floor where they are more
available to collecting than large males residing in
337
burrows. There is often a collecting bias against
large individuals of frogs, as well. Typically, field
storage specimen containers limit the total volume
of material that can be collected and preserved. I
usually collected no more than one or two large
specimens of L. pentadactylus cluster species from
any locality, while I was much more unbiased in
collecting and preserving juvenile specimens.
Parenthetically, it is worth pointing out that the
MZUSP collections are unique in being much more
unbiased relative to specimen size than other
collections with which I have familiarity. Sampling
is often biased, posing problems for standard
statistics, which assume unbiased sampling.
Head width in frogs determines the upper
size limit of their prey. Thus, sexual dimorphism in
head width should indicate that males and females
are differentially capable of capturing and
swallowing the largest size prey. Typically, one large
prey item of equal weight to two smaller prey items
provides the frog more nutrition than the two
smaller-sized prey because of the surface to volume
relationships of the digestible portions of the prey.
For head width, the typical statistical significance
convention of 0.05 may be too strict in biological
terms. Why wouldn’t a 0.20 level be biologically
meaningful in this case? If the 0.20 level were
accepted for head width significance, then 7 species
differ significantly in head width ratio and in all
cases, the male has the wider head. The biological
implications are actually different for the different
species depending on whether there is variation in
sexual dimorphism in size as well. For example, in
L. savagei the females are larger than males, and
male head widths are wider, suggesting that adults
are actually eating the same-sized prey. Conversely,
available data indicate that male L. turimiquensis
are larger than females and have proportionally
broader heads, suggesting that males and females
are not eating the same largest sized prey.
The problems of sampling bias, biological
versus statistical significance, and appropriate
statistical models for evaluating sexual dimorphism
in Leptodactylus are being explored elsewhere
(Hayek and Heyer, 2005).
Fuzzy Morphological and Bioacoustic Species
Differentiation
Biologists in the field have no problems
in recognizing the minimum number of species
338
Arquivos de Zoologia
of the taxa in this study at any given site.
Biologists may err in recognizing too many
species at a site in thinking that juveniles of some
species represent a species different from the
adults.
Determining whether samples collected
from different sites represent the same species is
not at all straightforward, however. In many
instances, there are neither features of external
adult form morphology nor advertisement call that
completely and consistently diagnose the species
recognized herein (the reason that no artificial key
to the species is attempted in this paper). The
limited data suggest that habitat and larval
morphology differentiation have been at least as
important historically for the species of this study
as adult form morphology and advertisement call
differentiation. For example, there are three known
distinct larval morphology and larval habitat
associations involved: (1) tooth row formula of
1/2(1) pond-dwellers; (2) tooth row 2(2)/2-3[1][2]
pond or terrestrial burrow dwellers; and (3) tooth
row formula 1/2(1) obligate terrestrial burrow
dwellers. Leptodactylus myersi is restricted to
rocky outcrops, a habitat not used by the other
species. The available data indicate that
L. turimiquensis occurs only in closed canopy
rainforests and human-modified rainforest habitats,
whereas the morphologically similar
L. labyrinthicus and L. vastus are restricted to open
formation habitats. This topic is explored further
in Heyer et al. (2005).
ACKNOWLEDGMENTS
Many individuals have freely responded to
what must seem like innumerable requests from me
over the last 10 years relative to this project.
Without their help and support, this study would
have been impossible.
James Ellis – formerly of the St. Louis Zoo,
USA and Richard Gibson – Durrell Wildlife
Conservation Trust, United Kingdom, provided
unpublished information on seasonal variation in
male secondary structures for Leptodactylus
“pentadactylus” and fallax respectively.
James P. Bogart – University of Guelph,
Canada, José A. Langone – MHNM, and Jean
Lescure – MNHN, responded to literature
clarifications and queries.
The following provided access to collections
and/or responded to collection queries: Erik
Ahlander – NRM, Ana Almendáriz – EPN, Natalia
Ananjeva – ZIL, James E. Aparicio – CBF, Márcio
Silva Araujo – ZUEC, J. W. Arntzen – RMNH,
Janalee P. Caldwell – OMNH, David Cannatella –
TNHC, Ulisses Caramaschi – MNRJ, Diego
Francisco Cisneros-Heredia – FHGO, Luis A.
Coloma – QCAZ, Ignacio De la Riva – MNCN,
William E. Duellman – KU, Amy Estep – OMNH,
Linda S. Ford – AMNH, W. Chris Funk – University
of Montana, USA, José E. González Fernández –
MNCN, Axel Groenveld – ZMA, Rainer Günther
– ZMB, Walter Hödl – University of Vienna,
Austria, Sven O. Kullander – NRM, John D. Lynch
– ICN, Rafael Marquez – MNCN, Carolina Mello
– MZUSP, Konstantin Milto – ZIL, Charles W.
Myers – AMNH, Göran Nilson – GNM, Annemarie
Ohler – MNHN, Jaime E. Péfaur – CV-ULA, José
P. Pombal Jr. – MNRJ, Jens Bødtker Rasmussen –
ZMUC, Alan Resetar – FMNH, Juan Francisco
Rivadeneiro-Romero – MECN, Juan A. Rivero –
University of Puerto Rico, Mayagüez, Puerto Rico,
Santiago Ron – QCAZ, Ivan Sazima – ZUEC, Celsa
Señaris – SCN, Anders Silfvergrip – NRM, John
E. Simmons – KU, P.E. Vanzolini – MZUSP.
The following loaned specimens from their
respective collections: Janalee P. Caldwell –
OMNH, the late Adão J. Cardoso – ZUEC, Luis
A. Coloma – QCAZ, Amy Estep – OMNH, Linda
S. Ford – AMNH, Carolina Mello – MZUSP,
Jaime E. Péfaur – CV-ULA, P.E. Vanzolini –
MZUSP.
Rafael O. de Sá, University of Richmond,
USA and P.E. Vanzolini, MZUSP, graciously
provided critical reviews of the manuscript. Miriam
Muedeking Heyer improved the clarity of the text.
James A. Poindexter II, Biological
Resources Division, U. S. Geological Service
stationed at USNM, took the digital photographs
for Figures 18, 21, 23, 24, and 27. Addison Wynn,
USNM, produced Figures 3, 7, 9, 15, 16, 17, 20,
22, and 26.
The Neotropical Lowland Research
Program, Smithsonian Institution, USA (Richard
P. Vari, Principal Investigator) and the National
Science Foundation, USA (award 9815787 to R.
de Sá and W. R. Heyer) provided financial support
during the tenure of this study.
I express my heartfelt thanks to the above
individuals and institutions for their help.
Vol. 37(3), 2005
REFERENCES
Ab’Sáber, A.N. 1977. Os domínios morfoclimáticos na América
do Sul. Primeira aproximação. Geomorfologia,
52:1-23.
Andersson, L.G. 1911. A new Leptodactylus and a new
Nototrema from Brazil. Arkiv för Zoologi, 7(17):1-6.
Andersson, L.G. 1945. Batrachians from east Ecuador collected
1937, 1938 by Wm. Clarke-Macintyre and Rolf
Blomberg. Arkiv för Zoologi, 37A(2):1-88.
Anonymous. 2000. T tests in SYSTAT. In: SYSTAT ® 10.
Statistics II. SPSS Inc., Chicago, p.II577-II590.
Bauer, A.M. 2002. [Book review] Albertus Seba, cabinet of natural
curiosities. The complete plates in colour, 1734-1765.
2001. Taschen, Köln. 587p., 6 fold-out plates. Newsletter
and Bulletin of the International Society for the History
and Bibliography of Herpetology, 3:8-15.
Blair, W.F. & Littlejohn, M.J. 1960. Stage of speciation of two
allopatric populations of chorus frogs (Pseudacris).
Evolution, 14:82-87.
Boeseman, M. 1970. The vicissitudes and dispersal of Albertus
Seba’s zoological specimens. Zoologische
Mededelingen, 44:177-206.
Bogart, J.P. 1974. A karyosystematic study of frogs in the genus
Leptodactylus (Anura: Leptodactylidae). Copeia,
(3):728-737.
Bokermann, W.C.A. & Sazima, I. 1974. Os sapos. Sua Boa
Estrela, 38:4-11.
Boulenger, G.A. 1882. Catalogue of the Batrachia Salientia s.
Ecaudata in the collection of the British Museum.
2.ed. Taylor and Francis, London.
Breder, C.M., Jr. 1946. Amphibians and reptiles of the Rio
Chucunaque drainage, Darien, Panama, with notes on
their life histories and habits. Bulletin of the American
Museum of Natural History, 86:375-436.
Charif, R.A.; Mitchell, S. & Clark, C.W. 1995. Canary 1.12
user’s manual. Cornell Laboratory of Ornithology,
Ithaca.
Coloma, L.A. & Ron, S.R. 2001. Ecuador megadiverso.
Megadiverse Ecuador. Anfibios, reptiles, aves,
mamíferos. Amphibians, reptiles, birds and mammals.
Serie de Divulgación, 1. Museo de Zoología, Pontificia
Universidad Católica del Ecuador, Quito.
De la Riva, I. 2000. La obra herpetológica de Jiménez de la
Espada: Su relevancia y validez después de un siglo.
In: López-Ocón, L.& Pérez-Montes, C. (Eds.), Marcos
Jiménez de la Espada (1831-1898). Tras la Senda de
un Explorador. CSIC, Madrid, p.76-90.
De la Riva, I.; Köhler, J.; Lötters, S. & Reichle, S. 2000. Ten
years of research on Bolivian amphibians: Updated
checklist, distribution, taxonomic problems, literature
and iconography. Revista Española de Herpetología,
14:19-164.
Denaro, L. 1972. Karyotypes of Leptodactylidae anurans.
Journal of Herpetology, 6:71-74.
Duellman, W.E. & Trueb, L. 1986. Biology of Amphibians.
McGraw-Hill Book Co., New York.
Engelman, L. 2000. Discriminant analysis. In: SYSTAT® 10.
Statistics I. SPSS Inc., Chicago, p.I275-I326.
Erspamer, V.; Roseghini, M. & Cei, J.M. 1964. Indole-,
imidazole-, and phenyl-alkylamines in the skin of
thirteen Leptodactylus species. Biochemical
Pharmacology, 13:1083-1093.
339
Frost, D. 2002. Amphibian species of the world, version 2.21.
Available at: http://research.amnh.org/herpetology/
amphibia/index.html.
Gorzula, S. & Señaris, J.C. 1998. Contribution to the
herpetofauna of the Venezuelan Guayana I. A data base.
Scientia Guaianae, (8):1-269.
Gosner, K.L. 1960. A simplified table for staging anuran
embryos and larvae with notes on identification.
Herpetologica, 16:183-190.
Greenewalt, C.H. 1968. Bird song: acoustics and physiology.
Smithsonian Institution Press, Washington, D.C.
Haddad, C.F.B.; Andrade, G.V. & Cardoso, A.J. 1988. Anfíbios
anuros no Parque Nacional da Serra da Canastra,
Estado de Minas Gerais. Brasil Florestal, (64):9-20.
Hayek, L.-A.C. & Heyer, W.R. 2005. Determining sexual
dimorphism in frog measurement data: integration of
statistical significance, measurement error, effect size,
and biological significance. Anais da Academia
Brasileira de Ciências, 77:45-76.
Hayek, L.-A.C.; Heyer, W.R. & Gascon, C. 2001. Frog
morphometrics: a cautionary tale. Alytes, 18:153-177.
Hero, J.-M. 1990. An illustrated key to tadpoles occurring in
the central Amazon rainforest, Manaus, Amazonas,
Brasil. Amazoniana, 11:201-262.
Hero, J.-M. & Galatti, U. 1990. Characteristics distinguishing
Leptodactylus pentadactylus and L. knudseni in the
central Amazon rainforest. Journal of Herpetology,
24:226-228.
Heyer, W.R. 1969. Studies on frogs of the genus Leptodactylus
(Amphibia, Leptodactylidae). V. Taxonomic notes on
L. latinasus, rhodonotus, romani, and wuchereri.
Herpetologica, 25:1-8.
Heyer, W.R. 1970. Studies on the genus Leptodactylus
(Amphibia: Leptodactylidae). II. Diagnosis and
distribution of the Leptodactylus of Costa Rica.
Revista de Biología Tropical, 16:171-205.
Heyer, W.R. 1972. The status of Leptodactylus pumilio
Boulenger (Amphibia, Leptodactylidae) and the
description of a new species of Leptodactylus from
Ecuador. Contributions in Science. Natural History
Museum of Los Angeles County, (231):1-8.
Heyer, W.R. 1978. Systematics of the fuscus group of the frog
genus Leptodactylus (Amphibia, Leptodactylidae).
Science Bulletin. Natural History Museum of Los
Angeles County, (29):1-85.
Heyer, W.R. 1979. Systematics of the pentadactylus species
group of the frog genus Leptodactylus (Amphibia:
Leptodactylidae). Smithsonian Contributions to
Zoology, (301):1-43.
Heyer, W.R. 1995. South American rocky habitat Leptodactylus
(Amphibia: Anura: Leptodactylidae) with description
of two new species. Proceedings of the Biological
Society of Washington, 108:695-716.
Heyer, W.R.; Sá, R.O. de & Rettig, A. 2005. Sibling species,
advertisement calls, and reproductive isolation in frogs
of the Leptodactylus pentadactylus species cluster
(Amphibia, Leptodactylidae). In: Proceedings of the
12th Ordinary General Meeting of Societas Europaea
Herpetologica. In press.
Heyer, W.R. & Diment, M.J. 1974. The karyotype of Vanzolinius
discodactylus and comments on usefulness of
karyotypes in determining relationships in the
Leptodactylus-complex (Amphibia, Leptodactylidae).
340
Arquivos de Zoologia
Proceedings of the Biological Society of Washington,
87:327-336.
Heyer, W.R.; García-Lopez, J.M. & Cardoso, A.J. 1996.
Advertisement call variation in the Leptodactylus
mystaceus species complex (Amphibia:
Leptodactylidae) with a description of a new sibling
species. Amphibia-Reptilia, 17:7-31.
Heyer, W.R.; McDiarmid, R.W. & Weigmann, D.L. 1975.
Tadpoles, predation and pond habitats in the tropics.
Biotropica, 7:100-111.
Heyer, W.R.; Rand. A.S.; Cruz, C.A.G.; Peixoto, O.L. & Nelson,
C.E. 1990. Frogs of Boracéia. Arquivos de Zoologia,
31:231-410.
Hödl, W. 1993. Amazonien aus der Froschperspektive. Kataloge
des OÖ. Landesmuseums, Neue Folge, (61):499-546.
Hoogmoed, M.S. & Avila-Pires, T.C.S. 1991. Annotated
checklist of the herpetofauna of Petit Saut, Sinnamary
River, French Guiana. Zoologische Mededelingen,
65:53-88.
Hoogmoed, M.S. & Gruber, U. 1983. Spix and Wagler type
specimens of reptiles and amphibians in the Natural
History Musea in Munich (Germany) and Leiden
(The Netherlands). Spixiana, Supplement,
9:319-415.
Ibáñez D., R.; Rand, A.S. & Jaramillo A., C.A. 1999. Los
anfibios del Monumento Natural Barro Colorado,
Parque Nacional Soberania y areas adyacentes. The
amphibians of Barro Colorado Nature Monument,
Soberania National Park and adjacent areas.
Editorial Mizrachi/Pujol, S.A., Panamá.
International Commission on Zoological Nomenclature. 1999.
International code of zoological nomenclature.
International Trust for Zoological Nomenclature,
London.
Jaslow, A.P. 1985. Variation in secondary sexual characters in
Leptodactylus pentadactylus (Anura: Leptodactylidae)
[abstract]. American Zoologist, 25:130A.
Jiménez de la Espada, M. 1875. Vertebrados del viaje al
Pacífico verificado de 1862 a 1865 por una comisión
de naturalistas enviada por el gobierno español.
Batracios. Miguel Ginesta, Madrid.
Kornacker, P.M. & Dederichs, U. 1998. Herpetologische
Eindrücke einer Venezuelareise – Teil 2: Die Llanos.
Elaphe, 6:68-73.
Larson, P.M. & Sá, R.O. de. 1998. Chondrocranial morphology
of Leptodactylus larvae (Leptodactylidae:
Leptodactylinae): its utility in phylogenetic
reconstruction. Journal of Morphology, 238:287-305.
Laurenti, J.N. 1768. Specimen medicum, exhibens synopsin
reptilium emendatam cum experimentis circa venena
et antidota reptilium austriacorum. Trattnern, Vienna.
Lescure, J. & Marty, C. 2000. Atlas des amphibiens de Guyane.
Patrimoines Naturels, 45:1-388.
Leviton, A.E. & Gibbs, R.H., Jr.1988. Standards in herpetology
and ichthyology. Standard symbolic codes for
institution resource collections in herpetology and
ichthyology. Supplement No. 1: Additions and
corrections. Copeia, (1):280-282.
Leviton, A.E.; Gibbs, R.H., Jr.; Heal, E. & Dawson, C.E. 1985.
Standards in herpetology and ichthyology: Part I.
Standard symbolic codes for institutional resource
collections in herpetology and ichthyology. Copeia,
(3):802-832.
Lutz, A. 1926. Observações sobre batrachios brasileiros. Parte
I: O gênero Leptodactylus Fitzinger. Memórias do
Instituto Oswaldo Cruz, 19:139-157.
Lutz, A. 1930. Segunda memória sobre espécies brasileiras do
gênero Leptodactylus, incluindo outras alliadas.
Memórias do Instituto Oswaldo Cruz, 23:1-20.
Marquez, R.; De la Riva, I. & Bosch, J. 1995. Advertisement
calls of Bolivian Leptodactylidae (Amphibia, Anura).
Journal of Zoology, 237:313-336.
McCranie, J.R. & Wilson, L.D. 2002. The amphibians of
Honduras. Society for the Study of Amphibians and
Reptiles, Contributions to Herpetology,
(19):1-625.
Maxson, L.R. & Heyer, W.R. 1988. Molecular systematics of
the frog genus Leptodactylus (Amphibia:
Leptodactylidae). Fieldiana, Zoology, New Series,
(41):1-13.
Melin, D. 1941. Contributions to the knowledge of the
Amphibia of South America. Göteborgs Kungliga
Vetenskaps – och Vitterhets – Samhälles Handlingar,
följden 6, series B, 1:1-71.
Miranda-Ribeiro, A. 1926. Notas para servirem ao estudo dos
gymnobatrachios (Anura) brasileiros. Tomo primeiro.
Archivos do Museu Nacional do Rio de Janeiro,
27:1-227.
Muedeking, M.H. & Heyer, W.R. 1976. Descriptions of eggs
and reproductive patterns of Leptodactylus
pentadactylus (Amphibia: Leptodactylidae).
Herpetologica, 32:137-139.
Müller, L. 1927. Amphibien und Reptilien der Ausbeute Prof.
Bresslau’s in Brasilien 1913-1914. Abhandlungen der
Senckenbergischen Naturforschenden Gesellschaft,
40:259-304.
Noble, G.K. 1918. The amphibians collected by the American
Museum Expedition to Nicaragua in 1916. Bulletin
of the American Museum of Natural History,
38:311-347.
Péfaur, J.E. & Sierra, N.M. 1995. Status of Leptodactylus
labyrinthicus (calf frog, rana ternero) in Venezuela.
Herpetological Review, 26:124-127.
Peters, W. 1873. Über die von Spix in Brasilien gesammelten
Batrachier des Königl. Naturalienkabinets zu
München. Monatsberichte der Königlichen
Preussische Akademie der Wissenschaften Berlin,
1872:196-227.
Rivero, J.A. & Esteves, A.E. 1969. Observations on the
agonistic and breeding behavior of Leptodactylus
pentadactylus and other amphibian species in
Venezuela. Breviora, (321):1-14.
Rodríguez, L.O. & Duellman, W.E. 1994. Guide to the frogs of
the Iquitos region, Amazonian Peru. The University
of Kansas Natural History Museum Special
Publication, (22):1-80.
Savage, J.M. 2002. The Amphibians and Reptiles of Costa Rica.
A Herpetofauna between Two Continents, between
Two Seas. The University of Chicago Press, Chicago.
Schmidt, K.P. & Inger, R.F. 1951. Amphibians and reptiles of
the Hopkins-Branner Expedition to Brazil. Fieldiana,
Zoology, 31:439-465.
Schneider, H.; Joermann, G. & Hödl, W. 1988. Calling and
antiphonal calling in four Neotropical anuran species
of the Family Leptodactylidae. Zoologische
Jahrbücher, Physiologie, 92:77-103.
Vol. 37(3), 2005
Schlüter, A. 1980. Bio-akustische Untersuchungen an
Leptodactyliden in einem begrenzten Gebiet des
tropischen Regenwaldes von Peru (Amphibia:
Salientia: Leptodactylidae). Salamandra, 16:227-247.
Seba, A. 1734. Locupletissimi Rerum Naturalium Thesauri
Accurata Descriptio, et Iconibus Artificiosissimis
Experessio, per Universam Physices Historiam. Opus,
cui, in hoc Rerum Genere, nullum par Exctitit. Ex
toto Terrarum orbe Collegit, Digessit, Descripsit, et
Depingendum Curavit Albertus Seba, Etzela
Oostfrisius, Academiae Caesareae Leopoldino
Carolinae Naturae Curiosorum Collega Xenocrates
Dictus; Societatis Regiae Anglicanae, et Instituti
Bononiensis, Sodalis. Tomus I. JanssonioWaesbergios/J. Wetstenium/Gul. Smith.
Shine, R. 1979. Sexual selection and sexual dimorphism in the
Amphibia. Copeia, (2):297-306.
Silva, A.P.Z.; Haddad, C.F.B. & Kasahara, S. 2000.
Chromosomal studies on five species of the genus
Leptodactylus Fitzinger, 1826 (Amphibia, Anura)
using differential staining. Cytobios, 103:25-38.
Smith, H.M.; Schneider, T. & Smith, R.B. 1977. An overlooked
synonym of the giant toad Bufo marinus (Linnaeus)
(Amphibia, Anura, Bufonidae). Journal of
Herpetology, 11:423-425.
Spix, J. 1824. Animalia nova sive species novae testudinum et
ranarum, quas in itinere per Brasiliam annis
MDCCCXVII-MDCCCXX jussu et auspiciis
Maximiliani Josephi I. Bavariae Regis. Franc. Seraph.
Hübschmanni, Monachii.
341
Straughan, I.R. & Heyer, W.R. 1976. A functional analysis of
the mating calls of the Neotropical frog genera of the
Leptodactylus complex (Amphibia, Leptodactylidae).
Papéis Avulsos de Zoologia, 29:221-245.
Taylor, E.H. 1952. The frogs and toads of Costa Rica. University
of Kansas Science Bulletin, 35(1):577-942.
Villa, J. 1972. Anfibios de Nicaragua. Instituto Geográfico
Nacional/Banco Central de Nicaragua, Managua.
Vizotto, L.D. 1967. Desenvolvimento de anuros da região
norte-ocidental do Estado de São Paulo. Faculdade
de Filosofia, Ciências e Letras, São José do Rio Preto,
SP.
Walkowiak, W. 1988. Central temporal encoding. In: Fritzsch,
B.; Ryan, M.J.; Wilczynski, W.; Hetherington, T.E. &
Walkowiak, W. (Eds.), The Evolution of the Amphibian
Auditory System. John Wiley & Sons, New York.
p.275-294.
Wassersug, R.J. & Heyer, W.R. 1988. A survey of internal oral
features of leptodactyloid larvae (Amphibia: Anura).
Smithsonian Contributions to Zoology, (457):1-99.
Zakon, H.H. & Wilczynski, W. 1988. The physiology of the
anuran eighth nerve. In: Fritzsch, B.; Ryan, M.J.;
Wilczynski, W.; Hetherington, T.E. & Walkowiak, W.
(Eds.), The Evolution of the Amphibian Auditory
System. John Wiley & Sons, New York. p.125-155.
Zimmerman, B.L. & Bogart, J.P. 1984. Vocalizations of primary
forest frog species in the central Amazon. Acta
Amazonica, 14(2):473-519.
342
Arquivos de Zoologia
APPENDIX. SPECIMENS EXAMINED.
Unassignable Specimens
BOLIVIA. COCHABAMBA. Chapare: lake at Alto Palmar, on road from Cochabamba to Villa Tunari,
2500’, USNM 146507. SANTA CRUZ. Chiquitos: El Pailón, 350 m, CM 36170; El Portón, 550 m, MCZ
30135. German Busch: El Carmen, CM 336166. Ichilo: Buenavista, 400 m, CM 3808; 5 km W Buenavista,
400 m, USNM 146519-146521. Sara: Río Colorado, CM 4296. Velasco: Flor de Oro, Parque Nacional
Noel Kempff Mercado, CBF 2305. All of these specimens are most similar to Leptodactylus labyrinthicus,
turimiquensis, and vastus.
BRAZIL. AMAZONAS. Barreira do Matupirí, USNM 202518. This specimen is most similar to
L. pentadactylus. BAHIA. Fazenda Cana Brava, mun. Maracás, MNRJ 30947; Ilhéus, MNRJ 703. The
Bahia specimens are either L. labyrinthicus or vastus. PARÁ. Jacareacanga, MZUSP 24947-24948. These
specimens are most similar to L. knudseni. RORAIMA. Serra do Parima, MZUSP 24936. This specimen
is most similar to L. labyrinthicus, turimiquensis, and vastus.
ECUADOR. ESMERALDAS. La Tola, QCAZ 19859; San Miguel and environs, EPN 7861-7868. These
specimens are either L. peritoaktites or L. rhodomerus.
Leptodactylus knudseni
BOLIVIA. BENI. Moxos: Areruta, CBF 01491. LA PAZ. Iturralde: Rurrenabaque, CBF 3673-3674,
UMMZ 108594-108596. Sud Yungas: Canton Cotapata, CBF 3086.
BRAZIL. STATE NOT KNOWN. Lower Amazonia, USNM 28966. AMAPÁ. Serra do Navio, Rio
Amapari, MZUSP 10140. AMAZONAS. Cachoeirinha, Rio Madeira, MZUSP 56592, USNM 202517;
Biological Dynamics of Forest Fragments Project sites, north of Manaus, MZUSP 60077, 60132, 60134;
Manaus, MZUSP 53743-53744; 40 km S Manaus at km 12 on road to Autazes, OMNH 37583; Novo
Aripuanã, MZUSP 56593-56594; Reserva Ducke, MZUSP 53745-53747; Rio Ituixi, Scheffer Madeireira,
OMNH 36854; Tefé, MCZ 1294. MATO GROSSO. Apiacás, MZUSP 80865-80869; Aripuanã, MZUSP
80655-80662; Claudia (Fazenda Iracema), MZUSP 83175-83176; Rio Teles Pires, MZUSP 71187. PARÁ.
Alter do Chão, OMNH 34360-34361; CEMEX, 101 km S, 15 km E Santarém, OMNH 34796; Parque
Nacional da Amazônia, Rio Tapajós, MZUSP 54667-54668, 58218, USNM 288734-288741; Reserva
Biológica Rio Trombetas, MZUSP 56624-56639, USNM 289039-289047; Rio Mapuera, at equator,
AMNH 49484. RONDÔNIA. Alto Paraiso, AMNH 124825-124829, MZUSP 131123, USNM 303909;
Cachoeira de Nazaré, Rio Machado, MZUSP 63826; Calama, Rio Madeira, USNM 202516; Forte Príncipe
da Beira, MZUSP 25169; Porto Novo, MZUSP 62179; Porto Velho, MZUSP 16658-16664, 16667,
16670, 16672-16674, 16676-16680, 16684; Rio Maçangana, fl. Jamari, prx. Alto Paraiso, MZUSP 60404;
alto Rio Machado, MZUSP 15907; RO-399, km 21, MZUSP 62210; Santa Barbara, MZUSP 62033-62034,
62036-62038; Santa Cruz da Serra, MZUSP 61556. RORAIMA. Colonia Apiaú, MZUSP 65890, 65963,
66087-66088; Igarapé Cocal, MZUSP 67086-67087, USNM 302415; Ilha de Maracá, MZUSP 63294,
65675; Santa Maria do Boiaçu, MZUSP 67331, 67354, 68293, 68295.
COLOMBIA. AMAZONAS. Río Apaporis, USNM 144847. META. Menegua, E of Puerto López, upper
Río Meta, USNM 147272; Villavicencio, ICN 2359, 14094, MLS 482; Villavicencio, Pozo Azul, ICN
13941-13943. VAUPÉS. Mitú, ICN 03283.
ECUADOR. FRANCISCO ORELLANA. Parque Nacional Yasuní, QCAZ 6670, 8182, 8314, 9676,
13077, 13079, 13094, 13244, 16183, QCAZ 24323-24324, 24326. MORONA-SANTIAGO. San José,
Vol. 37(3), 2005
343
USNM 283829-283831; Sucúa, 2 miles E of on trail from Sucúa to Río Upano, 2700’, USNM 196722.
NAPO. Archidona, QCAZ 429, 670; Estrellayacu, QCAZ 4502; Yampuna, QCAZ 2257. PASTAZA.
Cachiyacu, 138.5 km S of Coca (= Francisco de Orellana), 288 m, USNM 320985; Río Conambo, USNM
196723; Río Danta, 500 m, EPN 3103; Río Pucayacu, Río Bobonaza, USNM 196726; alto Río Pucayacu,
Río Bobonaza, USNM 196724, 527975-527977; Río Shyona, mouth of, in Conambo River, USNM
196725. SUCUMBÍOS. Cuyabeno, EPN 7209; Reserva Faunistica Cuyabeno, QCAZ 5862; Lago Agrio,
EPN 4064; La Selva Hostría, QCAZ 8496; Pozo Sacha, QCAZ 6319; Comunidad Quichua Singüe, QCAZ
16324.
GUYANA. DISTRICT UNKNOWN. Marudi, AMNH 49264. EAST BERBICE. Shudi-kar-wau (River),
AMNH 70117; EAST DEMERARA. Georgetown, AMNH 39636. MAZARUNI-POTARO. Kamakusa,
AMNH 21406. NORTH WEST. Arakaka, UMMZ 66782; Baramita, USNM 535773. POTAROSIPIRUNI. Paramakatoi, ROM 28451-28452. RUPUNUNI. Iwokrama Forest Reserve, USNM
531512-531520, IWK 282, 392, 602.
PERU. AMAZONAS. Huampami, on Río Cenepa, USNM 317514-317516; La Poza, USNM 566002;
Shaim, Río Cenepa, USNM 560371. CUZCO. San Martín, ca 5 km N Camisea River, 474 m, USNM
538184. HUANUCO. Tingo María, USNM 193875-193883. MADRE DE DIOS. Lago Sandoval, 200 m,
KU 215130-215131; Pakitza, USNM 298922, 345285; Tambopata Reserve, Explorer’s Inn, 280 m, USNM
247356-247357, 268968-268969, 342999, 343244. PASCO. Tsioventeni, 4200’, USNM 205553.
SURINAME. BROKOPONDO. Bergen Tal (= Bergendaal), ZMB 7321. NO OTHER DATA. ZMB
6759, 8534.
VENEZUELA. AMAZONAS. Apepada, alto Río Ventuari, EBRG 1907; Atabapo, Río Puruname, EBRG
1155; Brazo Casiquiare, Capibara, 106 km SW of Esmeralda, USNM 216784-216785; Esmeralda, 108 km
SSE of, Río Mavaca, 140 m, USNM 216786, 216794; Maraquita, alto Río Orinoco, EBRG 2915; Mavaca,
alto Río Orinoco, EBRG 2934; Misión Coromoto-Atures, FLSM 1193; Paso del Diablo, AMNH 23164;
Puerto Ayacucho, Rincones de Chacorro, Atures, EBRG 775; carretera Puerto Ayacucho-Gavilán, 18 km
de Planta Inos, Atures, EBRG 1274; Río Negro, frente a Isla Cigarrón, EBRG 76; Río Negro, Solano,
upstream from Cano Manu, tributary of Río Casiquiare, 250 m, USNM 248008; Stanford Zent, alto Río
Cuao, EBRG 1869. BOLÍVAR. Puente Cuyuni, 13 km S and 1 km E, KU 166495-166497; Reserva
Forestal de Imataca, 40 km ENE de Tumeremo, 180 m, EBRG 2271, 2390, 2442-2444; Vaso de Guri,
quebrada de San Luis, FLSM 11114.
Leptodactylus labyrinthicus
BRAZIL. BAHIA. Barreiras, MNRJ 1087-1088. GOIÁS. Aruanã, MZUSP 4989-4990, 25306-25308;
Cachoeira Alta, MZUSP 10429-10433; Cavalcante (Fazenda Santo Antonio), MZUSP 66565; Jataí
(Fazenda Nova Orlandia), MZUSP 25298-25301, (Fazenda Santa Adelia), MZUSP 20988-20990,
25298-25301; Lagoa Formosa, Cabeceiras, MZUSP 25296; Minaçu, Serra da Mesa, MNRJ 30941-30942;
Monte Alegre de Goiás (Córrego Riacho Dagua), MZUSP 66388; Rio Verde (Fazenda Transvaal), MZUSP
12506-12507, 24538-24540, 25329-25339; Santa Rita do Araguaia (Fazenda Babilonia), MZUSP 66681;
complexo das cavernas São Mateus – São Domingos, MZUSP 58676; UHE Serra da Mesa, MNRJ 20292,
20296, MZUSP 71532-71534, 71792-71796, 72153-72154, 72578. MATO GROSSO. Barra do Tapirapés,
MZUSP 24541-24542, 25250-25253; Chapada dos Guimarães, Burití, MZUSP 37333-37343; Chapada
dos Guimarães, Salgadeira, USNM 507904; Dumbá, MZUSP 4311; Porto Esperidião, MZUSP 59732,
56595; São Domingos, Rio das Mortes, MZUSP 1777, 1779, 4301-4306; São Felix, MZUSP 25327-25328;
Utiariti, MZUSP 24543, 25203. MATO GROSSO DO SUL. Santa Luzia, MZUSP 28548-28549. MINAS
GERAIS. Belo Horizonte, USNM 96978-96980; Lassance, Santa Rita, USNM 98786-98787. PARÁ.
344
Arquivos de Zoologia
Cachimbo, MZUSP 21734, 21862-21864, 21870, 21872, 25950-25951. PARANÁ. Ponta Grossa, USNM
125504. RIO DE JANEIRO. No other data, USNM 70593. SÃO PAULO. Assis, km 433, USNM 207674;
Bertioga, USNM 123393; Jurumirim, USNM 121286; Luis Antonio, ca 5 km S of, Fazenda Jataí, USNM
303175-303178; Rechã, USNM 121284; near Ribeirão Preto, farm Schmidt, railway Mogyana, USNM
100970; São Paulo, USNM 100971, 121285.
PARAGUAY. AMAMBAY. Parque Nacional Cerro Cora, ca 32 km WSW of Pedro Juan Caballero, ca
500 m, USNM 253115.
Leptodactylus myersi
BRAZIL. AMAZONAS. Rio Aracá (Serrinha), MZUSP 59016, 59018, 59026-59028. PARÁ. Igarapé
Jaramacaru, Campos do Ariramba, MZUSP 28405, 54110-54114. RORAIMA. Colonia Apiaú, MZUSP
65949, USNM 302267; Mucajaí, AMNH 128021-128031, MZUSP 66089 (Holotype), 70976-70986,
USNM 302190-302206.
FRENCH GUIANA. No further data, MNHN 1982•81; Massif des Emerillons, MNHN 1982•82;
Montagne des Trois Pitons, MNHN 1982•153; Montagne St. Marcel, MNHN 1982•73-1982•80; Peolaue
(Ht. Oyapock), MNHN 1982•83; Trois Saut, MNHN 1982•84.
SURINAME. BROKOPONDO. Tafelberg, MCZ 97259-97261, 97303, 97306, 97308, RMNH 23912,
23919-23930, 23951-23961. NICKERIE. Amotopo, RMNH 23964-23968; Blanche Marie-Vall, RMNH
23910-23911. SARAMACCA. Voltzberg, RMNH 23974-23976; Raleighvallen-Voltzberg Nature Preserve,
MCZ 92363.
Leptodactylus paraensis
BRAZIL. PARÁ. Aldeia Aukre, MZUSP 70023; Alegre, MZUSP 24997, 25949; Altamira, MZUSP 63337;
Alter do Chão, OMNH 34359; BR 010, km 93, Belém-Brasilia, MZUSP 24945; Canindé, MZUSP 25009;
CEMEX, 101 km S, 15 km E Santarém, OMNH 34762; Juruá, Rio Xingu, MZUSP 64198; Rio Vermelho,
MZUSP 70075; Serra do Cachimbo, MNRJ 2567, 11721; Serra de Kukoinhokren, MZUSP 69318,
69320-69322 (69321 = Holotype), 70365, 70549, 70918, USNM 559809; Tucuruí, MZUSP 62554,
75618, 85170, USNM 523765; Xingó dam site, MZUSP 131849-131853.
Leptodactylus pentadactylus
BOLIVIA. LA PAZ. Iturralde: Serranía de Eslabon, CBF 3876, 23925. PANDO. General Federico Roman:
Río Negro, ca 150 m, USNM 336179. Manuripi: 8 km SO Santa Rosa, CBF 1303.
BRAZIL. NO STATE. Pará to Manaus, lower Amazonia, USNM 28929. ACRE. Recordação, Rio Moa,
MZUSP 51552; Rio Branco, MZUSP 70974. AMAZONAS. Igarapé Belém, Rio Solimões, MZUSP
24897; Lago Amanã, MZUSP 58527; Manaus, MZUSP 377, 56719-56721; Reserva Ducke, MZUSP
53748; Reservas BDFF, north of Manaus, MZUSP 57359, 60087, 68200-68201; Rio Cuieiras, MZUSP
65406; Rio Gutaki (probably Jutahi = Jutaí), ZMB 30972; Rio Ituxí, Scheffer Madeireira, OMNH 36851;
São José (Jacaré), Rio Solimões, MZUSP 40303; Serrinha, Rio Japurá, MZUSP 56777-56779. MATO
GROSSO. Claudia (Fazenda Iracema), MZUSP 83181-83183. PARÁ. Altamira, ca 50 km (airline) S of,
near Cachoeira do Espelho, rocky island in Rio Xingu, USNM 303466; Altamira, Usina Kararahô, km 10
do acesso do acampamento Juruá, ZUEC 7232; Altamira, Usina Kararahô, km 27 do acesso do
Vol. 37(3), 2005
345
acampamento Juruá, ZUEC 7233; Canindé, Rio Gurupi, MZUSP 25010-25011; Curuá-Una, MZUSP
58437; Furo do Panaquara, MZUSP 35677; Juruá, Rio Xingu, MZUSP 64195-64197, 64199-64201,
64253-64254; Monte Cristo, Rio Tapajós, MZUSP 38956; Óbidos, MZUSP 22126; Reserva Biologica
Rio Trombetas, at junction of Igarapé Jacaré and Rio Trombetas, USNM 289058; Serra de Kukoinhokren,
MZUSP 69331, 70364, 70917; Vai-Quem-Quer, MZUSP 69625.
COLOMBIA. AMAZONAS. Leticia, USNM 147053; Leticia, vereda Caña Brava, 200 m, ICN 35884.
CAQUETÁ. Florencia, vereda Santa Elena, 25.9-27.5 km, 980 m, ICN 24174-24176; Florencia, La Vega,
MLS 272, 293. META. Río Guayabero, Angostura No. 1, USNM 150490.
ECUADOR. FRANCISCO DE ORELLANA. No other data, QCAZ 15953; Estación Científica Yasuní,
240 m, QCAZ 5236, 16659; Llinta, Aguarico, EPN 6637-6638; Nuevo Rocafuerte, Aguarico, PCSA-2,
212 m, EPN 6572; Parque Nacional Yasuní, Aguarico, EPN 2541, 2678, 2812, 6511, QCAZ 24321-24322;
Pozo Yampuna, Orellana, EPN 2813-2814. NAPO. Parque Nacional Sumaco Napo Galeras, QCAZ 16134;
San José de Payamino, Loreto, EPN 1723; San Pablo de Kontesya, 300 m, MECN 275, 367-371.
PASTAZA. Arajuno, Curaray, Villano, EPN 6861, 7409; Canelos, USNM 196742; Pozo Misión, 240 m,
EPN 1036-1042; Puyo, Pastaza, EPN 1848; Puyo, 2 km E of on trail to Veracruz, USNM 196741, 226350;
Río Arajuno, cabaceras del, tributary of Río Napo, USNM 196744; Río Pindo, USNM 196747; Río
Rutuno, tributary of Río Bobonaza, USNM 196746; Sarayacu, MZUSP 12652; Tiguino, 130 km S of
Coca (now Francisco de Orellana), 300 m, USNM 320986-320987. SUCUMBÍOS. Cuyabeno, MECN
344; La Selva Hostría, QCAZ 8495; Reserva de Producción Faunistica Cuyabeno, QCAZ 312, 2121,
2052, 5861-5863, 8013; Santa Cecilia, QCAZ 290. ZAMORA-CHINCHIPE. Destacamento Militar Miasi,
EPN 4132-4135.
FRENCH GUIANA. CAYENNE. Arataye River, south bank of, ca 5 km below Saut Parare, 20-50 m,
USNM 287753-287754.
PERU. AMAZONAS. San Antonio, vicinity of, on the Río Cenepa, USNM 317519; Shaim, vicinity
of, on the Río Alto Comaina (tributary of the Río Cenepa), USNM 317517; Shaim, Río Cenepa,
USNM 560372; Tseasim, on the upper Río Huampami (tributary of the Río Cenepa), USNM 317518.
CUZCO. Cashiriari-2 (Armihuari), ca 4 km S of the Camisea River, 579 m, USNM 538202; Pagoreni
on the Camisea River, 465 m, USNM 538203-538204; Paucartambo, 84 km (by road) NE of, puente
Quita Calzon (= km 164 on Paucartambo-Atalaya road), 1180 m, USNM 346146; San Martín, ca
5 km N of the Camisea River, 474 m, USNM 538196-538201. JUNÍN. Chonkareni, near Otica, Río
Tambo, USNM 234000. LORETO. Río Lagarto Cocha, Aguas Negras, USNM 521020. MADRE DE
DIOS. Pakitza, Reserve Zone, ca 57 km (airline) NW mouth of Río Manu on Río Manu, 350 m,
USNM 298902, 334139-334140, 342641-342642, 342856, 345286; Tambopata Reserve, Explorer’s
Inn, 30 km (airline) SSW Puerto Maldonado, 280 m, USNM 222283-222284, 247368-247369,
268971-268972; Zona Reservada Tambopata-Candamo, Colpa de Guacamayo, W bank of Río
Tambopata, USNM 332460.
SURINAME. Albina, ZMB 5806.
Leptodactylus peritoaktites
ECUADOR. AZUAY. Tamarindo, FHGO 069. COTOPAXI. Sigchos, USNM 196743. ESMERALDAS.
Hacienda Equinox, 38 km NW of Santo Domingo de Los Colorados, 1000’, USNM 196739 (Holotype).
PICHINCHA. Bosque Protector la Perla, QCAZ 4530; Centro Científico Río Palenque, 47 km S of
Santo Domingo de Los Colorados, 150-220 m, USNM 285391-285392; Ramsey Farm, 18 km W Santo
Domingo de Los Colorados, km 19 Chone road, USNM 196740, 196745, 527992.
346
Arquivos de Zoologia
Leptodactylus rhodomerus
COLOMBIA. ANTIOQUIA. Parque Regional Natural Los Orquideas, vereda Venados, Municipio de
Frontino, 850-950 m, ICN 35540. CHOCÓ. km 39 from Carmen de Atrato to Quibdó, 700 m, ICN 17035;
corrego Guayabal, ca 8 km de Quibdó, ICN 10070; Lloró, Granja Experimental “CEMA”, ICN 16663.
NARIÑO. Vereda Gualcalá, Municipio de Barbacoas, 360 m, ICN 13936. VALLE DE CAUCA. Near
Buenaventura, Virology Field Station, Río Raposo, USNM 151460-151461; Daqua, Queremal, 1100 m,
ICN 32700; Río Calima, USNM 145094, 150756; vereda Campo Alegre, Municipio de Restrepo, 460 m,
ICN 13320-13323 (13322 = Holotype).
ECUADOR. ESMERALDAS. Alto Tambo, 253 m, QCAZ 17056; Luis Vargas Torres, 8.5 km SE de la
población de Selva Alegre, EPN 7869.
Leptodactylus savagei
COLOMBIA. ANTIOQUIA. Chicorodó, near Turbo, USNM 153914; 5 km S Valdivia, ICN 9934, 9936.
MAGDALENA. Parque Nacional Tayrona, trail between Cañaveral and El Pueblito, USNM 200376;
Parque Nacional Tayrona, Santa Marta, 360-400 m, ICN 13655-13658, 13668, 21995, 23255; Parque
Nacional Tayrona, Santa Marta, localidad El Cedro, ICN 20674-20675; Santa Marta, road to Altamira,
Sierra Nevada de Santa Marta, ca 680 m, ICN 20335.
COSTA RICA. NO OTHER DATA. USNM 19600. CARTAGO. La Suiza, KU 25713, 28188; Moravia
de Turrialba, KU 30407, 65709, UTA 1395; near Peralta, Tunnel Camp, KU 33165, 33167; 2.7 miles NE
Río Reventazon bridge, Peralta road, UMMZ 117274-117276; Turrialba, AMNH 62251, FMNH 57532,
101800, 101802, KU 30408-30411, 65707-65708, USNM 29953-29954; 10 miles S Turrialba, USNM
192577. GUANACASTE. Estación Experimental Enrique Jiménez Nuñez, 13.6 km SW of Cañas, 20 m,
USNM 219777. HEREDIA. 7 km NE Puerto Viejo, 300 m, KU 33135-33139, 33144-33152, 65710-65712.
LIMÓN. La Lola, KU 34967, UMMZ 117277; Los Diamantes, FMNH 101797, 101799, KU 25716-25718,
30405-30406, 65706; Puerto Viejo, KU 35924; Suretka, UMMZ 129019, 135387, 135392, KU 35927;
Tortuguero, AMNH 75098, MCZ 29134. PUNTARENAS. Agua Buena, KU 35928-35929, 34965; 33
miles N Canoas, Río Barrow pit, 150’, USNM 148524; Golfito, KU 33153, 34968-34969, 65715-65717,
TCWC 19305-19306; Osa Tropical Science Center, ca 2.5 km SW Rincon de Osa, 30 m, LACM 116317,
USNM 219538, 227645-227656 (227652 = Holotype); Villa Neily, 75 m, KU 65713-65714,
100338-100339, 100354. SAN JOSÉ. El General, KU 25715; 3 miles SSE San Isidro del General, FMNH
101801; 13 miles SSW San Isidro del General on Dominical road, 710 m, KU 34966, 35925-35926,
LACM 114334.
HONDURAS. COLÓN. Belfate, AMNH 45704; Quebrada Machín, 540 m, USNM 534216-534220;
Salamá, USNM 242000-242014. EL PARAÍSO. Ca 1 km E Arenales, 390 m, USNM 524346. OLANCHO.
El Torno, 180 m, USNM 514579; Quebrada de Las Marías, ca 12 km NNE La Colonia, 660 m, USNM
523764, 524347; Nueva Esperanza, 710 m, USNM 534215; Quebrada El Guásimo, 140 m, USNM 535865;
near Quebrada El Mono, 100 m, USNM 538627; Río Kosmako, 130 m, USNM 538626; confluence Río
Wampú and Quebrada Siksatara, 95 m, USNM 319944; confluence of Río Yanguay and Río Wampú,
100 m, USNM 319943.
NICARAGUA. BLUEFIELDS. 6 km W Rama, 50’, TCWC 19307. GRENADA. Grenada, LACM 37870.
MATAGALPA. Finca Tepeyac, 10.5 km N, 9 km E Matagalpa, 960 m, KU 85146; Greytown (= San Juan
del Norte), USNM 19765. ZELAYA. Bonanza, KU 85147-85148, 101168; Camp Corozo, Río Huahuashan
(= Río Wawashan), AMNH 54980-54981; Camp Santa Ana, Río Huahuashan (= Río Wawashan), AMNH
54999; Cara de Mono, 50 m, KU 112703-112704; El Recreo, 25 km W Rama, KU 112666-112672,
Vol. 37(3), 2005
347
LACM 13945, 20475; Masawas, Waspuk River, AMNH 58435; Recreo, 10 miles above Río Mico, UMMZ
79751-79752.
PANAMA. BOCAS DEL TORO. Almirante, KU 79992, USNM 142334; Boca del Drago, USNM 142318;
Cayo Nancy, USNM 338479-338480; Isla Colón, La Gruta, USNM 338121-338124; Isla Cristobal,
Bocatorito camp, USNM 348134-348135; Isla Popa, 1 km SE Deer Island channel, USNM
298079-298080; S end of Isla Popa, 1 km E of Sumaco channel, USNM 347152-347156; Laguna de
Tierra Oscura, 3.7 km S of Tiger Key, USNM 348432-348436; Peninsula Valiente, Punta Alegre, USNM
338606; Punta de Peña, USNM 38714. CHIRIQUÍ. Progreso, UMMZ 58221. COCLÉ. El Valle, AMNH
59590, KU 76573, 107229, 116825. COLÓN. Frijoles, USNM 196303. DARIÉN. Camp Creek, Camp
Townsend, AMNH 40786-40788, 41061; Casita (Tacarcuna Casita), USNM 141783; near Jaqué, at jct
of Río Jaqué and Río Imamado, USNM 161215; Laguna, 820 m, KU 76570-76572; Río Canglón, UMMZ
123158-123161; Río Chucunaque, ca 7 km above Río Mortí, 150 m, KU 107230; Río Silugandí, UMMZ
124018; Río Tuira at Río Mono, KU 115308; Río Ucurgantí, ca 7 km above mouth, KU 97014-97017,
107231-107232; Tacarcuna, 550 m, KU 76513-76514. LOS SANTOS. Guánico Arriba, 60 m, KU 107228.
PANAMÁ. Altos de Maje, AMNH 88743; Barro Colorado Island, AMNH 69728, FMNH 175986, KU
76568, MCZ 15266, UMMZ 63594, USNM 161151; Cerro Campana, KU 76569; Chili Brillo, Cave A,
AMNH 62338; Juan Mina, Chagres River, USNM 129908; near Madden Dam, AMNH 87143, KU
115306, UMMZ 78481; Tapia, AMNH 18922-18923. SAN BLAS. Armila, USNM 150090; Armila,
Quebrada Venado, USNM 150089; Camp Sasardí, 12 m, KU 108688-108689. VERAGUAS. Mojara,
AMNH 123320; mouth of Río Concepción, KU 115307.
Leptodactylus turimiquensis
VENEZUELA. ANZOÁTEGUI. Cueva del Agua, near Puerto La Cruz, CV-ULA iv•1180, USNM 559810;
Pekin abajo, Río Neveri, SCN 2294-2295; Puerto La Cruz, SCN 12359. MONAGAS. Camp. MARNR,
Río Guarapiche, EBRG 2360; Caripito, AMNH 70667-70668 (70667 = Holotype); El Caliche, Juasjuillar,
Dtto. Caripe, EBRG 2085. SUCRE. Casa Inparques, Los Mangos, Paria, SCN 12323-12326; Cumanacoa,
CM 9065; Guaraúnos, EBRG 422, KU 166492-166493; carretera via Hacienda Buena Vista, Península
de Paria, SCN 12327-12328; La Fragua, Sierra del Turimiquire, CV-ULA iv•5541, iv•5620-5622; near
Latal, Hacienda Mirasol, CM 9098; La Yaguara, 4 km E Río Salado, Paria, EBRG 423; Parare, ZMFK
36063; Parque Nacional Península de Paria, Hacienda Solís, 150 m, EBRG 2561-2562; bajada a San
Juan de los Galdonas, SCN 10705.
Leptodactylus vastus
BRAZIL. ALAGOAS. Fazenda do Prata, Mun. São Miguel dos Campos, MNRJ 9925; Murici, MNRJ
9767; São Miguel dos Campos, MZUSP 9244. CEARÁ. Açude Atalhos, Mun. Brejo Santo, MNRJ 30957;
estrada para o Açude Atalhos, Mun. Brejo Santo, MNRJ 30958-30959; Açude dos Prazeres, MNRJ
30951, 30955-30956; tanques de piscicultura abaixo do Açude Quixabinha, MNRJ 30933, 30962; Arajara,
MZUSP 56603-56604; Aurora, MNRJ 30960; Baturité (Açudinho), MZUSP 25151-25152; Crato, MNRJ
431, 1085; Fortaleza, Macuripe, MNRJ 2589, 12035; Itapipoca, MZUSP 25110-25113; Lima Campos,
MZUSP 24534; Maranguape, USNM 109148-109151; Santana do Cariri, MZUSP 54753-54754, USNM
216072-216079. GOIÁS. Araguatins, MZUSP 25309; São Domingos, MZUSP 66620, 66622.
MARANHÃO. São Luiz, MZUSP 21746-21747. PARAÍBA. Campina Grande, USNM 109144; Coremas,
MZUSP 22907-22908; Elembuzeiro (= Umbuzeiro), USNM 109143; Gurinhém (Fazenda Salgado),
MZUSP 65367; Junco do Seridó, MZUSP 52286-52287, 60352; Mamanguape, MZUSP 22865-22867.
PERNAMBUCO. No further locality data, USNM 57719; near Caruaru, on way to Serra dos Cavalos,
USNM 284552; Exu and vicinity, MZUSP 54526, 56596-56598, 56605-56606; Igarassu, MZUSP 25030;
348
Arquivos de Zoologia
Ponta de Pedras, MZUSP 34317-34320; Recife (Dois Irmãos), MZUSP 4461; Serra dos Cavalos, 13 km
ESE São Caetano, MZUSP 63167. PIAUÍ. Valença, MZUSP 50186-50187; 25 km N Valença, MZUSP
50199. RIO GRANDE DO NORTE. Ceará Mirim, MZUSP 10812; Cruzeta, USNM 109145-109147;
Natal, USNM 81131. SERGIPE. Areia Branca, MZUSP 37815-37820; Fazenda Capivara, Brejo Grande,
MNRJ 30944; Fazenda Cruzeiro, Cristianópolis, MNRJ 30726; Santo Amaro das Brotas, MZUSP
56599-56602.
Download PDF