ArticlePDF Available

Miniaturization in Direct-Developing Frogs from Mexico with the Description of Six New Species

Authors:

Figures

Content may be subject to copyright.
Miniaturization in Direct-Developing Frogs from Mexico
with the Description of Six New Species
Authors: Jameson, Tom J.M., Streicher, Jeffrey W., Manuelli, Luigi,
Head, Jason J., and Smith, Eric N.
Source: Herpetological Monographs, 36(1) : 1-48
Published By: The Herpetologists' League
URL: https://doi.org/10.1655/0733-1347-36.1.1
BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles
in the biological, ecological, and environmental sciences published by nonprofit societies, associations,
museums, institutions, and presses.
Your use of this PDF, the BioOne Complete website, and all posted and associated content indicates your
acceptance of BioOne’s Terms of Use, available at www.bioone.org/terms-of-use.
Usage of BioOne Complete content is strictly limited to personal, educational, and non - commercial use.
Commercial inquiries or rights and permissions requests should be directed to the individual publisher as
copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit
publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to
critical research.
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Herpetological Monographs, 36, 2022, 1–48
Ó2022 by The Herpetologists’ League, Inc.
Miniaturization in Direct-Developing Frogs from Mexico with the Description
of Six New Species
TOM J.M. JAMESON
1,6
,JEFFREY W. STREICHER
2,3,6
,LUIGI MANUELLI
3,4,5
,JASON J. HEAD
1
,AND ERIC N. SMITH
2
1
Department of Zoology and University Museum of Zoology, University of Cambridge, Downing Street, Cambridge CB2 3EJ, UK
2
Amphibian and Reptile Diversity Research Center, Department of Biology, The University of Texas at Arlington, 701 S. Nedderman Drive,
Arlington, TX 76019, USA
3
Department of Life Sciences, The Natural History Museum, Cromwell Road, South Kensington, London SW7 5BD, UK
4
Department of Life Sciences, Imperial College London, 180 Queen’s Gate, South Kensington, London SW7 2AZ, UK
5
Department of Genetics and Evolution, University of Geneva, 4 D’Yvoy Boulevard, Geneva 1205, Switzerland
ABSTRACT: The Craugastor mexicanus series (Anura: Craugastoridae) includes six species of direct-developing frogs that occur in Mexico and
Guatemala. Notably, two of these species have small adult body sizes (,18 mm snout to vent length) and several have intraspecific polymorphism
in color pattern. Using a geographic sampling focused on eastern Mexico (the location of most type localities), we conducted a molecular
phylogenetic analysis of two mitochondrial (12S, 16S) and two nuclear (RAG1, TYR) gene fragments. This analysis revealed two widespread
species, C. mexicanus and C. pygmaeus, along with evidence of multiple undescribed taxa from the states of Oaxaca, Mexico, Guerrero, and
Jalisco. Interestingly, the widespread species have stratified geographic distributions with the larger bodied clade restricted to high elevations and
the smaller bodied clade to low elevations. We also identify regions of Guerrero and Oaxaca where multiple species co-occur. To reevaluate the
quality of characters that have been previously used to diagnose species, we tested for heterochrony and sexual dimorphism using microcomputed
tomography and linear measurements. We found evidence for paedomorphosis as the mechanism of miniaturization in small-bodied taxa. Linear
measurements confirmed that tympanum and body size are sexually dimorphic traits in both small- and large-bodied species. We used this
enhanced understanding of morphological variation in the group to describe six new species. Despite this progress, we suspect that additional
species await discovery, particularly in western Mexico and east of the Isthmus of Tehuantepec where our sampling efforts were limited.
Key words: Brachycephaloidea; Craugastor bitonium sp. nov.;Craugastor candelariensis sp. nov.;Craugastor cueyatl sp. nov.;Craugastor
polaclavus sp. nov.;Craugastor portilloensis sp. nov.;Craugastor rubinus sp. nov.; Terraranae
THE EXTENSIVE flora and fauna of Mesoamerica make it
one of the world’s most ecologically diverse regions. Among
amphibians, direct-developing frogs of the genus Craugastor
Cope 1862, are particularly abundant but have also
experienced putative declines in recent times (Crawford
2003; Scheele et al. 2019). Despite the potential threat to
these frogs, several groups of Craugastor remain poorly
studied and in need of systematic revision. One such group
of amphibians is a predominately Mexican radiation, the C.
mexicanus series (sensu Hedges et al. 2008; Streicher et al.
2014). The series contains six described species: C.
hobartsmithi (Taylor 1936), C. mexicanus (Brocchi 1877),
C. montanus (Taylor 1942), C. omiltemanus (G ¨unther
1900a), C. pygmaeus (Taylor 1936), and C. saltator (Taylor
1941). These direct-developing frogs are denizens of the leaf
litter and are distributed throughout subtropical and tropical
Mexico, barely extending into western Guatemala. They
occur in diverse habitats including highland pine–oak
(Pinus–Quercus spp.) forest and tropical lowland deciduous
forest.
Notably, two species have small adult body sizes (C.
hobartsmithi, males ~14 mm [Taylor 1940]; C. pygmaeus,
males 16 mm, females 19 mm [Taylor 1936]). These small
species often occur syntopically with larger bodied species of
Craugastor creating a situation where they may be confused
for juveniles of larger species. Thus, given (1) the expanse
and ecological diversity of Mexico and (2) that miniaturized
anurans are ‘‘exceptionally prone to taxonomic underesti-
mation’’ (Scherz et al. 2019:1), it is quite likely that many
small-bodied species of the C. mexicanus series are yet to be
discovered.
In this study we revised the taxonomy of the C. mexicanus
series aided by molecular phylogenetic analyses, morpho-
metrics, microcomputed tomography, analysis of ossification
patterns, and qualitative assessments of morphology. Col-
lectively, our findings allowed us to describe six new species
and report that heterochrony likely has contributed to
miniaturization in C. hobartsmithi, C. pygmaeus, and several
of the new species. We set the stage for our revision with a
summary of the taxonomic history underlying the C.
mexicanus series.
MATERIAL AND METHODS
Taxonomic History
Brocchi (1887) described Leiuperus mexicanus from
‘‘Mexique.’’ This locality was later restricted to Cerro San
Felipe in Oax aca by Smith and Taylor (1950). G ¨unther
(1900a,b) described Syrrhaphus omiltemanus and Hylodes
calcitrans both from the locality of Omilteme (Omiltemi) in
Guerrero. A small burst of species descriptions occurred
during the 1930s and 1940s based solely on the works of
E.H. Taylor. Taylor (1937) described both Eleutherodactylus
hobartsmithi and E. pygmaeus with holotypes from Uruapan,
Michoaca
´n, and from Rodr´
ıguez Clara, Veracruz, respec-
tively. He also described paratypes of E. pygmaeus from
Chilpancingo, Guerrero. Taylor (1940) described three
species that would later be referred to the C. mexicanus
series. The first was Microbatrachylus albolabris based on a
holotype specimen from C ´
ordoba, Veracruz, and paratypes
from Portrero Viejo, Veracruz, and from San Juan Gracia,
6
CORRESPONDENCE: email, j.streicher@nhm.ac.uk and tjmj3@cam.ac.
uk. These authors contributed equally to this work.
1
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Veracruz. The second species described was M. oaxacae
from a holotype and paratypes collected near Cerro San
Felipe, Oaxaca. Third, M. minimus was described based on a
holotype and paratypes from Agua del Obispo, Guerrero,
and a paratype from Mazatla
´n, Guerrero. In Taylor (1940),
E. pygmaeus was also transferred to the genus Micro-
batrachylus. Taylor (1941) described two species; M.
lineatissimus from a holotype and paratypes originating from
Cerro San Felipe, Oaxaca; and E. saltator from a holotype
and paratypes originating from Omiltemi, Guerrero. Taylor
(1942) described his final two species referable to the C.
mexicanus series; M. montanus designating a holotype from
Mount Ovando, Chiapas, and paratypes from La Esperanza,
Chiapas, Las Nubes, Chiapas, and Salto de Agua, Chiapas;
M. imitator designating a holotype and from La Esperanza,
Chiapas, and paratypes from Colonia Hidalgo (8 km north of
La Esperanza), Chiapas. The most recent species description
was that of Davis and Dixon (1957), who described M.
fuscatus based on a holotype and paratype from Tulancingo,
Hidalgo.
By the early 1960s there was a growing consensus that the
genera Microbatrachylus (typically species with small adult
body sizes) and Eleutherodactylus (species with larger adult
body sizes) were congeneric. This understanding set the
stage for revisionary work that was conducted by W.E.
Duellman and J.D. Lynch. Duellman (1961) synonymized
M. albolabris, M. minimus, and M. imitator with ‘‘ M.
pygmaeus.’’ The rationale for these synonymies was based
on overlapping variation that he observed across multiple
characteristics (including color pattern, relative length of the
hind limb, presence and position of dorsal dermal folds or
pustules, relative size of inner and outer metatarsal
tubercles, and number of palmar tubercles) rendering the
diagnostic features of Taylor (1940, 1942) unreliable. Lynch
(1965) synonymized E. fuscatus with E. mexicanus based on
an evaluation of Davis and Dixon’s (1957) diagnostic
characters of E. fuscatus in some individuals of E. mexicanus.
Lynch (1965) also moved all species of Microbatrachylus to
the genus Eleutherodactylus. He also introduced a nomen
novum, Eleutherodactylus sartori, for E. montanus because
moving it from Microbatrachylus to Eleutherodactylus
created a junior homonym (Eleutherodactylus montanus
had already been described by Schmidt 1919). Lynch (1970)
placed E. lineatissimus and E. oaxacae in the synonomy of E.
mexicanus citing that the type specimens of the latter two
taxa were identical to E. mexicanus. Lynch (2000) placed E.
saltator in the synonymy of E. mexicanus.
At the beginning of the 21st century, relationships and
species boundaries among members of the C. mexicanus
series were still largely uncertain. Lynch (2000) provided a
helpful review of the complexities and confusion associated
with two groups he described: the Eleutherodactylus
rhodopis and E. omiltemanus groups. The first comprehen-
sive molecular assessment of phylogenetic relationships in
these groups was that of Crawford and Smith (2005). Their
sampling of the genus Eleutherodactylus included six of our
focal species (E. mexicanus, E. omiltemanus, E. pygmaeus,
‘‘E. saltator,’’ E. sartori [¼C. montanus], and an unde-
scribed species). The results from Crawford and Smith’s
(2005) analysis supported (1) moving all these species to the
genus Craugastor and (2) creating a new grouping (that they
called the ‘‘C. mexicanus group’’ ) to recognize the mono-
phyly of our focal taxa. Moving E. sartori to the genus
Craugastor allowed for the combination C. montanus to be
used once more. Soon after, Hedges et al. (2008) included
two species in their phylogenetic analysis but redefined the
C. mexicanus group of Crawford and Smith (2005) as the C.
mexicanus series to include seven species. One of these
species was C. saltator, which they noted should be removed
from the synonymy of C. mexicanus based on the earlier
findings of Crawford and Smith (2005). Streicher et al.
(2014) determined that one of the species included by
Hedges et al. (2008) in the C. mexicanus series, C.
occidentalis (Taylor 1941), belonged instead to the C.
rhodopis series of Hedges et al. (2008). Collectively, this
taxonomic history has resulted in six species assigned to the
C. mexicanus series: C. hobartsmithi, C. mexicanus, C.
montanus, C. omiltemanus, C. pygmaeus, and C. saltator
(Fig. 1).
Geographical and Taxonomic Sampling
We examined 461 specimens from Mexico and Guatemala
including the Mexican states of Oaxaca, Guerrero, Nayarit,
Sinaloa, Hidalgo, Jalisco, Veracruz, Morelos, Michoaca
´n,
Mexico, Colima, and Puebla (Appendix I). We examined
specimens referable to all focal species: C. hobartsmithi þC.
cf. hobartsmithi (n¼21), C. mexicanus (n¼225), C.
montanus (n¼8), C. omiltemanus (n¼36), C. pygmaeus (n
¼118), and C. saltator (n¼22). This included type material
corresponding to all available names except Microbatrachy-
lus fuscatus; however, we did examine a near-topotypic
specimen from Hidalgo (UTA A-66138) for both morphology
and DNA. We also examined 31 specimens with uncertain
taxonomic affinities. To examine the geographic distribution
of our focal taxa, we augmented our sampling localities using
VertNet distributional data from several museums in North
America. We made maps using QGIS and data layers
available from DIVA-GIS. Throughout the text, GPS
coordinates follow WGS84 in all cases.
We extracted DNA from tissues of 59 individuals of the C.
mexicanus series (Appendix II). This sampling included
representatives of C. cf. hobartsmithi, C. mexicanus, C.
omiltemanus, C. pygmaeus, and C. saltator. Most samples
were collected on field expeditions by the University of Texas
at Arlington (UTA) and the National Autonomous University
of Mexico (UNAM) over the past two decades (Streicher
2012) and one by Jacobo Reyes-Velasco from Colima (JRV
field catalogue). We used representatives of the subgenus
Craugastor (C. longirostris [KU 177803] from the Craugas-
tor fitzingeri series and C. podiciferus [mtDNA—UCR
16361, nDNA—MVZFC 13463] from the Craugastor
rhodopis series) and the subgenus Hylactophryne (C. uno
[AMCC118080]; Hedges et al. 2008; Streicher et al. 2011) as
outgroup taxa, sequences downloaded from the National
Center for Biotechnology Information (NCBI) GenBank
(Benson et al. 2004). We included several samples from
Hedges et al. (2008) and resampled several individuals used
in Crawford and Smith (2005). All novel DNA sequences
were submitted to NCBI GenBank (Appendix II).
Molecular Phylogenetic Analysis
We sequenced fragments of the mitochondrial (mtDNA)
ribosomal RNA 12S and 16S genes, and nuclear (nDNA)
RAG1 and Tyrosinase (Tyr) genes. Sequences were obtained
2Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
for 58 samples for 12S from Streicher (2012) and for 35
samples from 16S and Tyr from Manuelli (2017). We
amplified and sequenced a further 3 samples for 12S, 26
samples for 16S and TYR, and 47 samples for RAG1
specifically for this study. We amplified DNA fragments
using published primers sets (Table 1) and performed
Polymerase Chain Reaction (PCR; Saiki et al. 1988)
amplification using GoTaqt(Promega). We carried out
amplifications in 14.25-lL volumes containing 2 lLof
template DNA and 12.25 lL of MasterMix (5.8 lL diethyl
pyrocarbonate [DEPC]-treated H
2
O, 0.1 lL forward primer,
0.1 lL reverse primer, and 6.5 lLGoTaqt). These
amplifications were carried out by thermal cycling per-
formed on a TechnetPrime Elite (Techne). We amplified
the 12S and 16S gene fragments using the following thermal
cycling parameters: an initial cycle of 948C (4 min) followed
by 35 cycles of 948C (30 s) denaturing, 508C (30 s) annealing,
and a 728C (2 min) extension. A final phase of 728C (7 min)
FIG. 1.—Male holotype of Eleutherodactylus hobartsmithi Taylor (A, FMNH 100114, SVL ¼14.4 mm); female holotype of Leiuperus mexicanus Brocchi
(B, MNHNP 6318, SVL ~40 mm); male holotype of Microbatrachylus oaxacae Taylor (C, FMNH 100001, SVL ¼18.1 mm); male holotype of
Microbatrachylus lineatissimus Taylor (D, FMNH 100036, SVL ¼20.0 mm); female holotype of Microbatrachylus montanus Taylor (E, USNM 115507, SVL
¼27.0 mm); female holotype of Eleutherodactylus pygmaeus Taylor (F, UIMNH 16125, SVL ¼17.0 mm); female holotype of Microbatrachylus albolabris
Taylor (G, FMNH 100071, SVL ¼16.5 mm); male holotype of Microbatrachylus minimus Taylor (H, FMNH 100323, SVL ¼15.0 mm); female holotype of
Microbatrachylus imitator Taylor (I, USNM 115508, SVL ¼14.2 mm); male lectotype of Syrrhaphus omiltemanus unther (J, BMNH 1947.2.16.62, SVL ¼
19.8 mm); female lectotype of Hylodes calcitrans unther (K, BMNH 1947.2.16.47, SVL ¼34.9 mm); female holotype of Eleutherodactylus saltator Taylor
(L, FMNH 100166, SVL ¼44.0 mm). Symbols correspond to current taxonomic assignments (square ¼C. hobartsmithi, triangle ¼C. mexicanus, crossed-
square ¼C. montanus, pentagon ¼C. omiltemanus, circle ¼C. pygmaeus, rotated-triangle ¼C. saltator). A color version of this figure is available online.
TABLE 1.—Primers used to amplify the fragments of the mitochondrial (mtDNA) and nuclear (nDNA) genes examined in this study.
Locus Primer Sequence (50–30) Reference
12S 12SF AAACTGGGATTAGATACCCCACTAT Bossuyt and Milinkovitch (2000)
12S 12SR ACACACCGCCCGTCACCCTC Liu et al. (2000)
16S 16SAR CGCCTGTTTAYCAAAAACAT Kessing et al. (1989)
16S 16SBR CCGGTCTGAACTCAGATCACGT Kessing et al. (1989)
RAG1 R182 GCCATAACTGCTGGAGCATYAT Heinicke et al. (2007)
RAG1 R270 AGYAGATGTTGCCTGGGTCTTC Heinicke et al. (2007)
TYR Tyr1C GGCAGAGGAWCRTGCCAAGATGT Bossuyt and Milinkovitch (2000)
TYR Tyr1G TGCTGGGCRTCTCTCCARTCCCA Bossuyt and Milinkovitch (2000)
3
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
followed. We amplified the RAG1 and TYR genes using
touchdown PCR (Don et al. 1991) with the following
parameters: an initial cycle of 958C (5 min) followed by 40
cycles of 958C (30 s) denaturing, 588C (30 s) annealing, and
728C (1 min) extension, followed by 3 touchdown cycles of
958C (30 s) denaturing, 588C(18C per cycle; 30 s)
annealing, and 728C (1 min) extension. The touchdown
cycles were followed by a final set of 40 cycles of 958C (30 s)
denaturing, 558C (30 s) annealing, and 728C(1min)
extension, finishing with a final phase of 728C (10 min). All
PCR experiments contained a negative control. All PCR
products were quantified on 1% TAE agrose gel and
successful reactions were submitted to the Natural History
Museum (London) DNA Sequencing Facility for cleaning
and sequencing on an Applied Biosystems 3730 DNA
Anaylzer (ThermoFisher Scientific).
We assembled forward and reverse chromatographs in
Geneious v8.0.4 (Biomatters, Auckland, New Zealand) and
constructed multiple sequence alignments on each data set
(12S, 446 base pairs; 16S, 601 base pairs; RAG1, 662 base
pairs; TYR, 521 base pairs) using the Geneious alignment
algorithms (Kearse et al. 2012). We produced a concatenated
data set (12S þ16S þRAG1 þTYR, 2230 base pairs) from
the aligned sequences using the Geneious concatenation
tool. In the protein coding gene alignments (RAG1 and Tyr),
we confirmed reading frames were open by visual inspection
of alignments. In the ribosomal subunit gene alignments
(12S and 16S), we identified hypervariable loop regions as
regions of ambiguity in alignment with high densities of
indels and removed them to avoid ambiguous alignment.
Following alignment, we generated a matrix of genetic
distances between and within species using MEGA X
(Kumar et al. 2018).
We used Bayesian Markov chain Monte Carlo (MCMC;
Yang and Rannala 1997) and Maximum Likelihood (ML)
analysis on the concatenated, 12S, 16S, RAG1, and TYR data
sets for phylogenetic reconstructions. We partitioned the
data set by gene (12S and 16S) and codon position in
protein-coding genes (RAG1 and Tyr). We used Partition-
Finder v2.1.1 (Lanfear et al. 2016) to select best-fit models of
molecular evolution for each partition employing a Bayesian
Information Criterion (Table 2). We conducted Bayesian
MCMC analysis in MrBayes v3.2.6 (Huelsenbeck and
Ronquist 2001; Ronquist and Huelsenbeck 2003) with
sampling occurring every 1000 generations for 10 million
generations, and the first 25% of generated trees discarded
as burn-in. We used standard deviation split frequencies to
assess the convergence (,0.01). We also carried out ML
analyses in MEGA v7.0.26 (Kumar et al. 2016) for each data
set. Nodal support was assessed via 100 bootstrap pseudo-
replicates. Gaps and missing data were treated as a partial
deletion with a site coverage cut-off of 80%. We visualized
phylogenetic trees in FigTree v1.4.3 (Rambaut 2006).
Linear Morphometrics
We collected linear measurements of major body axes
from all specimens using digital calipers (accurate to the
nearest 0.1 mm). We took all measurements from full-body
alcohol-preserved specimens. All measurements were taken
by the same person (TJ). We used the terminology for the
external anatomy of amphibians from Walker (1980). We
took 15 measurements from each specimen: (1) snout to vent
length (SVL); (2) head width at commissure of the jaw; (3)
head length, measured on right side from snout tip to
commissure of jaw; (4) maximum tympanum width; (5)
maximum eye width; (6) interorbital distance; (7) naris to
snout distance, measured on right side from anterior corner
of naris to anterior corner of snout; (8) eye to naris distance,
measured on right side from anterior corner of eye to
posterior corner of naris; (9) brachial length, measured on
right side from ventral attachment to body to center of
elbow; (10) antebrachial length, measured on right side from
center of elbow to proximal surface thenar tubercle; (11)
manus length, measured on right side from proximal surface
of thenar tubercle to tip of Finger III; (12) femur length,
measured on right side from ventral attachment to body to
center of knee; (13) crus length, measured on right side from
center of knee to center of ankle; (14) pes length, measured
on right side from center of ankle to proximal surface of
inner metatarsal tubercle; (15) toe length, measured on right
side from proximal surface of inner metatarsal tubercle to tip
of Toe IV. All measurements were log-transformed to
linearize allometries and equalize variances (Hammer and
Harper 2006; Sidlauskas et al. 2011) before input into PAST
3 (Hammer et al. 2001) for morphometric analysis via
principal component analysis (PCA) to identify axes of
maximal variance (Hammer and Harper 2006).
In PCA of morphology at least one principal component
(PC) typically is explained by allometric scaling, as identified
by a significant correlation with log centroid size (geometric
mean of logs of all measures per specimen; Sidlauskas et al.
2011). We tested for this by correlating scores of each PC
with log centroid size. For PCs where a significant
correlation with log centroid size was found, we performed
allometric correction, calculating the residuals of the PC
regression with log centroid size to produce a size-
standardized morphospace (Sidlauskas et al. 2011).
The Craugastor mexicanus samples encompassed a size
range of 10.9–37.3 mm SVL, including juvenile individuals.
To control for the effect of putative juveniles we repeated
the PCA excluding all juvenile specimens. The gonads of a
subset of the sample were examined to see whether mature
gonads could be used to separate adult and juvenile
specimens; however, size of ova or testes varied dramatically
between individuals of the same size (suggesting seasonal
variation). No reliable method of separating adult and
juvenile specimens was found; in place of this, the lower
size limit of specimens in this data set identified as adults by
past workers was used to define minimum adult size and all
specimens below this size were excluded (Brocchi 1877;
TABLE 2.—Best-fit models selected for Bayesian phylogenetic analysis in
Mr. Bayes v3.2.1 by Partition Finder under both Bayesian Information
Criteria (BIC) and Akaike Information Criteria (AIC).
Partition
Best-fit
model BIC AIC
c
ln likelihood
12S GTRþG 5347.349 4334.924 –2038.647
16S GTRþG 6011.974 4971.775 –2359.327
RAG1, 1st position K2þG 2286.308 1514.438 –649.080
RAG1, 2nd position T92 1946.933 1175.064 –479.392
RAG1, 3rd position K2þG 3345.163 2573.763 –1178.737
TYR, 1st position JCþG 1769.698 1009.132 –395.191
TYR, 2nd position K2þI 2910.614 2143.574 –961.380
TYR, 3rd position JCþG 1721.681 961.627 –371.432
4Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Boulenger 1882; G ¨unther 1900 a,b; Parker 1927; Taylor
1936, 1940, 1941, 1942; Lynch 1965, 1970; Hedges 1989;
Hedges et al. 2008). For example, we excluded specimens
,20.2 mm SVL in C. mexicanus; ,10.8 mm SVL in C.
pygmaeus; ,25.9 mm SVL C. saltator.
For a priori groups where multiple specimens were
available of multiple different sizes, we tested for differences
in ontogenetic trajectory on size-correlated shape axes by
comparing slopes and elevation of reduced-major axis
regression lines of size correlated PCs (PCAs carried out
on each group individually) with log centroid size (Sidlauskas
et al. 2011) using ANCOVA tests in PAST 3 (Hammer et al.
2001). In all statistical reporting we provide degrees of
freedom as subscripts of statistics.
Testing for Heterochrony
Heterochrony is a change in the timing and/or rates of
processes underlying the ontogenetic formation of morpho-
logical traits (Alberch et al. 1979; Keller and Lloyd 1992;
Gould 2002). The most prevalent form of heterochrony in
anurans is paedomorphic miniaturization, driven by decel-
eration, hypomorphosis, and/or postdisplacement (Reilly et
al. 1997; Yeh 2002). Craugastor hobartsmithi and C.
pygmaeus, have small adult body sizes (,18 mm), so we
tested for the signature of miniaturization in their develop-
ment. Miniaturization is defined as size reduction beyond a
threshold at which dramatic changes in morphology,
physiology, and ecology occur (Hanken and Wake 1993;
Yeh 2002). These reductions relative to ancestral body size
are often accompanied by the loss of skeletal elements (de Sa
´
et al. 2019; Scherz et al. 2019).
We scanned 63 alcohol-preserved specimens of Craugas-
tor at the University of Cambridge Biotomographic Centre
using a Nikon Metrology XT ST High Resolution CT
Scanner at 133 kV, 214 lA, with a scan resolution of 0.01–
0.07 mm. These specimens were selected to maximize size
and phylogenetic variation represented. Reconstructions of
skeletal anatomy in 3D were rendered using AMIRA-AVIZO
v9.4 (ThermoFisher Scientific, Waltham, MA). Importantly,
these scans included type material for all available names
except Eleutherodactylus fuscatus (a junior synonym of C.
mexicanus). We uploaded 3D renders of skeletal anatomy of
all specimens investigated in this study to MorphoSource in
2020 (available at http://MorphoSource.org; Appendix III).
We investigated osteology of CT-scanned specimens by
visual inspection of 3D skeletal reconstructions in AVIZO.
We identified 19 osteological features that varied among
scanned specimen using the osteological terminology of
Duellman and Treub (1986): (1) ossification of calcaneum or
astragalus epiphyses; (2) ossification of epicorocoid; (3)
ossification of the exocipitalprootic; (4) fusion of exocipital-
prootic; (5) ossification of femur epiphyses; (6) fusion of
frontoparietals; (7) presence of posterolateral projection of
the frontoparietal; (8) posterior offset of frontoparietal–
prootic suture; (9) ossification of humeral exterior epiphyses;
(10) ossification of humeral interior epiphyses; (11) ossifica-
tion of hyoid; (12) ossification of prehallux; (13) ossification
of prepollex; (14) ossification of palentines; (15) ossification
of radioulnar exterior epiphyses; (16), ossification of
radioulnar interior epiphyses; (17) ossification of spheneth-
moid; (18) ossification of tibiofibular epiphyses; and, (19)
ossification of vomers. We scored these skeletal characters in
terms of presence or absence. Most of these characters
appear in a consistent order throughout ontogeny (based on
log centroid size), suggesting that there is a distinct
ontogenetic sequence for the ossification of skeletal elements
in the C. mexicanus series. We used these data to generate a
six-stage ontogenetic sequence (Table 3) from which a total
ossification score was generated for each individual.
Presence of posterolateral projections of the frontoparietals
were not included in the ontogenetic sequence because they
did not have a clear association with ontogeny in all species.
We plotted ossification score against log centroid size for
each species to allow the relationships between size or
ontogeny and ossification to be compared among species
(Alberch et al. 1979; Mitteroecker et al. 2005; Piras et al.
2011; Scanferla 2016; Esquerr´
e et al. 2017; Hipsley and
uller 2017; Da Silva et al. 2018). We compared the
relationships between size or ontogeny and ossification in
more detail for groups with sufficiently large sample sizes (C.
mexicanus, C. pygmaeus, C. saltator, and C. hobartsmithi)
by comparing slopes and elevation of reduced-major axis
regression lines of ossification score against log centroid size.
We used ANOVA to test whether slopes of ontogeny versus
ossification differed significantly from zero and between-
species pairwise v
2
tests of reduced-major axis regression
slopes to test whether the rates of ossification versus
ontogeny differed between species. We carried out all
statistical analysis in PAST 3 (Hammer et al. 2001). In all
statistical reporting we provide degrees of freedom as
subscripts of statistics.
Geometric Morphometrics of the Skull
We undertook geometric morphometric analysis on the
skulls of 56 of the CT-scanned specimens using 14
landmarks that were placed on the 3D reconstructions of
TABLE 3.—Six-stage ontogenetic sequence for skeletal development used to diagnose development in the Craugastor mexicanus series.
Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Stage 6
Exoccipital-prootic
ossified
Palatines ossified Sphenethmoid ossified Exoccipital-prootic fusion Prepollex present Frontoparietals fused
Vomers ossified Humeral interior
epiphyses ossified
Frontoparietal-prootic
suture offset posteriorly
Prehallux present Epicorocoid ossified
Hyoid ossified Radioulnar exterior
epiphyses ossified
Humeral exterior
epiphyses ossified
Femur epiphyses ossified
Radioulnar interior
epiphyses ossified
Tibiofibular epiphyses
ossified
Calcaneum/astragalus
epiphyses ossified
5
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
skulls using the landmarking function in AVIZO. Our
landmarks were based on those used on anuran skulls by
Simon et al. (2016): (1) anterior midline tip premaxilla; (2)
maxilla premaxillary suture; (3) anterior tip nasal; (4)
posterior lateral tip frontoparietal; (5) posterior dorsal tip
frontoparietal; (6) posterior dorsal tip squamosal (7) anterior
frontoparietal prootic suture; (8) lateral tip nasal; (9)
squamosal maxillary suture; (10) posterior tip maxilla; (11)
anterior tip parasphenoid; (12) posterior tip parasphenoid;
(13) lateral tip parasphenoid; and, (14) pterygoid maxillary
suture (Fig. 2).
We aligned landmark configurations by Procrustes
superimposition (Hammer and Harper 2006) and performed
PCA in the Program R package Geomorph v3.0.6 (Adams et
al. 2018). We visualized shape change across principal
components in MorphoJ v1.8 (Klingenberg 2011).
Testing for Sexual Dimorphism
Little is known regarding the reproductive biology of our
focal species, but sexual dimorphism in tympanum size and
pigmented gonads have been reported in some species
(Taylor 1936, 1940). We compared SVL to tympanum width
in three species where males have been reported to have
larger tympana than do females: C. mexicanus (n¼149), C.
omiltemanus (n¼27), and C. pygmaeus (n¼86). We
compared correlation coefficients (using nonparametric
Spearman’s qtests) between tympanum width and body
size (SVL) to those between putatively non–sexually
dimorphic characters and SVL. We expected that if tympana
are sexually dimorphic, they would have a lower correlation
coefficient with SVL than nondimorphic characters (which
should scale more linearly with body size). For non-
dimorphic characters, we used eye width in C. pygmaeus
and C. omiltemanus and crus length in C. mexicanus.We
were unable to use the same nondimorphic characters across
all species because of logistical issues; however, we had all
measurements for C. omiltemanus and analyzed both eye
width and crus length to confirm that these characters have
similar scaling relationships with SVL (see Results). In C.
mexicanus, we were able to dissect 27 individuals and
confirm sex. We used this analytical framework instead of
direct statistical comparisons between males and females
because we were not able to dissect most specimens to
confirm sex. All statistical analyses of sexual dimorphism
were performed in R v3.6.2 (R Core Team 2019) and plots
made in SYSTAT v13.2 (Systat Software, San Jose, CA). In
statistical reporting we provide the correlation coefficient
(r
s
), sum of all squared rank differences (S), probability of
the null hypothesis that qis equal to 0 (P), and sample sizes
(n).
Criteria for Species Recognition
We used similar philosophical criteria to those described
by Meik et al. (2018) to determine species boundaries. First,
FIG. 2.—Skull of Craugastor mexicanus (USNM 47905) in anterior, dorsal, lateral, and ventral views showing placements of landmarks. Landmarks are (1)
anterior midline tip premaxilla, (2) maxilla–premaxilla suture, (3) anterior tip nasal, (4) anterior tip frontoparietal, (5) posterior lateral tip frontoparietal, (6)
posterior dorsal tip squamosal, (7) anterior frontoparietal prootic suture, (8) lateral tip nasal, (9) squamosal maxillary suture, (10) posterior tip maxilla, (11)
anterior tip parasphenoid (12) posterior tip parasphenoid, (13) lateral tip parasphenoid, and, (14) pterygoid maxillary suture. A color version of this figure is
available online.
6Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
we considered the evolutionary history of putative lineages
using both concatenated and standalone analyses of mito-
chondrial and nuclear DNA sequences. Specifically, we
surveyed for evidence that species were (1) monophyletic (if
there was more than one individual sampled) and (2) had
proportionally similar branch lengths distinguishing them
from other taxa in both mitochondrial and nuclear phyloge-
netic reconstructions. We viewed similar divergence in both
mtDNA and nDNA data sets as strong evidence of
organismal divergence, and thus a reasonable proxy for
speciation patterns. Second, we made comparisons with two
taxa that are closely related but easily differentiated by
morphology (C. mexicanus and C. omiltemanus), to calibrate
a minimum threshold of genetic divergence (4.9%; Table 4)
that we would expect between most species. Third, we
compared external morphology and skull osteology among
species to identify differences among species. In summary,
we recognized a new species if they met two or more of the
following criteria: (1) phylogenetic distinctiveness in both
mitochondrial and nuclear DNA, (2) large genetic distances
from other species, and, (3) morphological differentiation
from other species.
RESULTS
Phylogenetic Analyses and Genetic Distances
Overall, maximum likelihood analysis recovered lower
branch support than did Bayesian analysis (Fig. 3). We
discuss phylogenetic relationships within species in the
taxonomic accounts below. In this analysis, C. montanus was
found with limited support to be the sister taxon of all other
species in the C. mexicanus series (66 ML; 0.99 BAYES).
Craugastor omiltemanus and C. saltator were weakly
supported as sister taxa (44 ML; 0.74 BAYES). Together
these two species were supported as the sister taxon to C.
mexicanus in the concatenated and mtDNA-only analyses
(Figs. 3 and 4). Craugastor pygmaeus and C. hobartsmithi
were more closely related to each other than any other
species; however, they belong to a clade that also includes six
undescribed species (all of which were found to have levels
of mtDNA and nDNA divergence similar to recognized
species; Figs. 4 and 5). Thus, our phylogenetic results
supported the recognition of six currently recognized species
in the C. mexicanus series (C. hobartsmithi, C. mexicanus, C.
montanus, C. omiltemanus, C. pygmaeus, C. saltator) and six
undescribed species. These six new species are described
herein as C. bitonium, C. candelariensis, C. cueyatl, C.
polaclavus, C. portilloensis, C. rubinus.
The phylogenetic results revealed several interesting
geographical patterns. Two species, C. mexicanus and C.
pygmaeus, are widely distributed throughout Mexico and
appear to have nonoverlapping elevational distributions, with
C. mexicanus being restricted to the highlands of the Sierra
Madre Oriental and Sierra Madre del Sur (Fig. 6). We
observed a difference in elevational distribution using
locality data from museum specimens and confirmed that
C. mexicanus and C. pygmaeus have largely nonoverlapping
elevational distributions (Fig. 7). Another widely distributed
species is C. hobartsmithi þC. cf. hobartsmithi, which
putatively ranges from the state of Guerrero in the east to
the state of Sinaloa in the northwest (Fig. 8). Although we
had relatively few individuals represented, our results
suggest that C. omiltemanus ranges more widely in the
Sierra Madre del Sur than previously thought, with a
distribution extending into Oaxaca. The remaining species
in the C. mexicanus series (C. bitonium, C. candelariensis, C.
cueyatl, C. montanus, C. polaclavus, C. portilloensis, C.
rubinus, and C. saltator) appear to have generally restricted
geographical distributions (Figs. 6 and 8) and four species
(C. candelariensis, C. portilloensis, C. polaclavus, and C.
pygmaeus) occur in near sympatry in the region of
Candelaria, Loxicha in Oaxaca.
Genetic distance analysis of the concatenated alignment
(Table 4) revealed that between-species divergence ranged
from 3.4% (C. hobartsmithi and C. rubinus) to 10% (C.
polaclavus to C. saltator). Within-species genetic distances
ranged from 0.3% (C. candelariensis) to 2.2% (C. mexica-
nus).
Morphology and Development
Multivariate analyses of linear measurements suggested
that there were two main morphological groups in our data
set (Fig. 9). The first group consisted of mostly small-bodied
species (C. bitonium, C. candelariensis, C. cueyatl, C.
hobartsmithi, C. polaclavus, C. portilloensis, C. pygmaeus,
and C. rubinus). The second group consisted of large-bodied
species (C. mexicanus, C. montanus, C. omiltemanus, and C.
saltator). The small-bodied morphological group corre-
sponds to a clade in our phylogenetic analyses (Figs. 3–5).
This may suggest that small adult body sizes evolved once
within the C. mexicanus series.
TABLE 4.—Genetic P-distances derived from the concatenated alignment 2230 base pairs of 12S and 16S (mtDNA) þRAG1 and TYR (nDNA) for species
contained within the Craugastor mexicanus series. Bolded distances on diagonal are within species P-distances (if more than a single individual was used to
estimate mean genetic distance).
bitonium candelariensis cueyatl hobartsmithi mexicanus montanus omiltemanus polaclavus portilloensis pygmaeus rubinus saltator
bitonium N/A
candelariensis 0.069 0.003
cueyatl 0.063 0.070 N/A
hobartsmithi 0.074 0.093 0.081 N/A
mexicanus 0.060 0.078 0.065 0.067 0.022
montanus 0.075 0.080 0.073 0.077 0.061 N/A
omiltemanus 0.070 0.080 0.069 0.085 0.049 0.073 0.012
polaclavus 0.059 0.064 0.067 0.068 0.060 0.066 0.067 0.013
portilloensis 0.065 0.070 0.073 0.077 0.056 0.057 0.065 0.058 N/A
pygmaeus 0.047 0.064 0.059 0.070 0.056 0.068 0.067 0.061 0.061 0.009
rubinus 0.072 0.083 0.076 0.034 0.066 0.080 0.079 0.066 0.067 0.066 0.005
saltator 0.097 0.070 0.081 0.089 0.051 0.088 0.056 0.100 0.063 0.073 0.088 N/A
7
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
FIG. 3.—Concatenated Maximum Likelihood analyses of two mitochondrial (12S and 16S) and two nuclear (RAG1 and TYR) gene fragments sequenced
from the Craugastor mexicanus series (total of 2230 base pairs). Node values correspond to bootstrap support from Maximum Likelihood analysis and
posterior probabilities from a Bayesian analysis of the same data set, respectively. Support values are not reported for nodes that had ,50/0.50. NS indicates
no support in the Bayesian analysis for a depicted relationship. Locality abbreviations are as follows: COL ¼Colima, HID ¼Hidalgo, GRO ¼Guerrero, JAL
¼Jalisco, MEX ¼Estado de M´
exico, OAX ¼Oaxaca, PUE ¼Puebla, and VER ¼Veracruz.
8Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Despite the difference in adult body sizes, the develop-
mental trajectories (log centroid size vs. PC1 scores) among
well-sampled species (C. hobartsmithi, C. mexicanus, C.
omiltemanus, and C. pygmaeus) had mostly indistinguishable
slopes (Fig. 10). Although the slope of C. hobartsmithi was
distinct from the other taxa (ANCOVA, F
4,93
¼11.600, P,
0.001), this may be related to small sample sizes. In contrast,
the slopes of C. mexicanus, C. omiltemanus, C. pygmaeus,
FIG. 4.—Maximum Likelihood analyses of mitochondrial DNA (mtDNA) markers sequenced from the Craugastor mexicanus series (total of 1047 base
pairs). Black circles indicate nodes receiving bootstrap support values .90 from the Maximum Likelihood analysis and 0.90 posterior probabilities in
corresponding Bayesian analyses. Where posterior probabilities were .0.90 but Maximum Likelihood bootstrap support was ,90, a number indicates the
bootstrap support. NS indicates no support in the Bayesian analysis for a depicted relationship. Locality abbreviations are as follows: COL ¼Colima, HID ¼
Hidalgo, GRO ¼Guerrero, JAL ¼Jalisco, MEX ¼Estado de M´
exico, OAX ¼Oaxaca, PUE ¼Puebla, and VER ¼Veracruz.
9
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
and C. saltator were not significantly different (ANCOVA,
F
3,91
¼1.072, P¼0.365), suggesting that most taxa,
regardless of body size, have similar ontogenetic trajectories.
The similarity in ontogenetic trajectory between the large-
bodied taxa (C. mexicanus, C. omiltemanus, and C. saltator)
and the small-bodied taxon C. pygmaeus suggests that
miniaturization of C. pygmaeus (and potentially other
members of the small-bodied clade) is due to a halt in body
size increase at an earlier stage of development in small-
bodied taxa relative to large-bodied taxa. This suggests that
the mechanism for miniaturization is hypomorphic paedo-
morphosis (Reilly et al. 1997).
The results of our ossification staging analysis (Fig. 11)
revealed that multiple small-bodied species (C. bitonium, C.
candelariensis, C. hobartsmithi, and C. pygmaeus) complete
the same ontogenetic sequence as their large-bodied
FIG. 5.—Maximum Likelihood analysis of nuclear DNA (nDNA) markers sequenced from the Craugastor mexicanus series (total of 1183 base pairs).
Black circles indicate nodes receiving bootstrap support values .90 from the Maximum Likelihood analysis and 0.90 posterior probabilities in corresponding
Bayesian analyses. Where posterior probabilities were .0.90 but Maximum Likelihood bootstrap support was ,90, a number indicates the bootstrap
support. NS indicates no support in the Bayesian analysis for a depicted relationship. Locality abbreviations are as follows: COL ¼Colima, GRO ¼
Guerrero, JAL ¼Jalisco, MEX ¼Estado de M´
exico, OAX ¼Oaxaca, PUE ¼Puebla, and VER ¼Veracruz.
10 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
relatives (C. mexicanus, C. omiltemanus, and C. saltator).
We found a correlation between centroid size and ossifica-
tion score for the whole sample (ANOVA, F
1,64
¼4.784, P¼
0.032) and each species individually (P,0.050) for which
enough taxa were available for statistical analysis. C.
pygmaeus had a steeper slope of ontogeny versus ossification
than C. mexicanus (v
21,32
¼10.069, P¼0.002; Fig. 11). In
other words, C. pygmaeus reaches the same level of
ossification as C. mexicanus at an earlier stage of ontogeny
or smaller size. Two specimens of C. mexicanus had
unexpected placements, including the examined paratype
of ‘‘M. lineatissmus,’’ which had a unique body size for
having a relatively well-ossified skeleton (similar to C.
montanus), and a small heavily ossified individual.
Completing the same ontogenetic sequence as large-
bodied species suggests that although C. pygmaeus and at
least three other small-bodied species (C. bitonium, C.
candelariensis, and C. hobartsmithi) may be miniaturized
through a process of heterochronic hypomorphosis, hetero-
chronic processes have not affected ossification. Alternative-
ly, other small-bodied species (C. ceuyatl, C. polaclavus, C.
portilloensis, and C. rubinus) had low levels of ossification
but putatively mature gonads. This suggests that these
species are paedomorphic both in terms of size and level of
ossification (Reilly et al. 1997; also see Discussion).
Skull Morphometrics and Sexual Dimorphism
Geometric morphometrics of 3D skulls revealed that
members of the Craugastor mexicanus series fall into two
groups based on skull shape. These two groups are mainly
defined by variation in the shape of the frontoparietals and
parasphenoid, as described by PC2 (Fig. 12). These skull
shape groups largely overlap with the large- and small-
bodied clusters identified using linear measurements (Fig.
9). The skull shape data suggest clades within the C.
mexicanus series can be recognized morphologically by skull
shape. The smaller bodied species (C. bitonium, C.
candelariensis, C. cueyatl, C. hobartsmithi, C. pygmaeus,
and C. rubinus) typically have a more recessed anterior
frontoparietal–prootic suture and a more anterior tip of the
parasphenoid, compared with the larger bodied species that
form a monophyletic clade (C. mexicanus, C. omiltemanus,
C. saltator). Craugastor montanus has an intermediate skull
morphology between these groups. Craugastor polaclavus
and C. portilloensis are distinct within this framework, for
possessing small body sizes but skull morphologies that are
similar to those of larger bodied species. One specimen of C.
mexicanus had unexpected placement; the examined para-
type of ‘‘M. lineatissimus,’’ which overlapped more with the
small-bodied cluster.
In C. mexicanus, tympanum width was less associated
with body size (r
s
¼0.604, S¼218,318, P,0.001, n¼149)
than crus length was with body size (r
s
¼0.977, S¼12,653,
FIG. 6.—Distribution of Craugastor mexicanus,C. montanus, C.
pygmaeus, and C. saltator in Mexico and Guatemala. Dots inside symbols
indicate individuals used in the molecular analyses in this study. Relevant
type localities are indicated by text and arrows.
FIG. 7.—Box plots depicting elevational distribution of Craugastor
mexicanus and C. pygmaeus inferred from locality data associated with
museum specimens. Results are shown for both specimens used in the
molecular analysis (Molecules þMorphology) and for specimens assigned to
species using only morphology. Asterisks represent potential outliers, which
are .1.5 times the interquartile range (¼range inside the box).
FIG. 8.—Distribution of Craugastor hobartsmithi, C. cf. hobartsmithi, C.
omiltemanus and six new species from Mexico. Dots inside symbols indicate
individuals used in the molecular analyses in this study. Inset depicts region
of Oaxaca where three of the new species exist in near sympatry. A question
mark indicates uncertain georeferencing. Relevant type localities are
indicated by text and arrows.
11
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
P,0.001, n¼149). In C. pygmaeus, tympanum width was
less associated with body size (r
s
¼0.462, S¼56,959, P,
0.001, n¼86) than eye width was with body size (r
s
¼0.776,
S¼23,690, P,0.001, n¼86). In C. omiltemanus,
tympanum width was less associated with body size (r
s
¼
0.603, S¼1300, P¼0.001, n¼27) than eye width (r
s
¼
0.933, S¼220, P,0.001, n¼27) or crus length (r
s
¼0.862,
S¼451, P,0.001, n¼27). These results are consistent
with sexual dimorphism in tympanum size in C. mexicanus,
C. omiltemanus, and C. pygmaeus. Using the specimens of
C. mexicanus with confirmed sexes revealed that males have
larger tympanum sizes than do females (Fig. 13). A handful
of C. omiltemanus and C. pygmaeus specimens that we were
able to dissect also followed this pattern, with males having
larger tympana. Thus, as in other species of anuran (e.g.,
Werner et al. 2009) multiple species of the C. mexicanus
series have males with larger tympana than do females,
suggesting that bioacoustic communication or signaling is
important to these frogs.
FIG. 9.—Body size and shape variation within the Craugastor mexicanus series as depicted by principal components analysis of 15 linear measurements.
Principal component scores for the first two components (left) are depicted along with their residuals (right). Large-bodied species are those with an adult
body size .20 mm SVL (C. mexicanus, C. montanus, C. omiltemanus, C. saltator), and small-bodied species are those with adult body sizes ,20 mm (C.
bitonium, C. candelariensis, C. cueyatl, C. hobartsmithi, C. polacalvus,C. portilloensis, C. pygmaeus, and C. rubinus).
FIG. 10.—Difference in ontogenetic trajectories between select large-
bodied (Craugastor mexicanus, C. omiltemanus, C. saltator) and small-
bodied species (C. hobartsmithi, C. pygmaeus) of the C. mexicanus series as
evidenced by reduced-major axis regression of size-correlated shape axes
with log-centroid size following Sidlauskas et al. (2011).
FIG. 11.—Ossification levels within the Craugastor mexicanus series
compared with body size (estimated as log centroid size). Ossification scores
are based on the six-stage ontogenetic sequence described in Table 3.
12 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Taxonomic Revision
Considering our phylogenetic and morphological analy-
ses, we provide updated taxonomic accounts for the C.
mexicanus series. Our taxonomic revision includes rede-
scriptions of six species and the description of six new
species. Each account includes a description, diagnosis,
commentary on phylogenetic and osteological variation, and
available natural history information. Diagnoses and com-
parisons are summarized in Table 4 (intraspecific and
interspecific genetic distances), Table 5 (sex-specific body
size and gonad pigmentation), Table 6 (diagnostic character
comparisons), and Figs. 14–21 (body, head, hand, and foot
comparisons). We also provide a dichotomous key for the
group. Morphological diagnoses are intended to be used
with adult specimens (unless otherwise noted).
TAXONOMIC ACCOUNTS
Craugastor bitonium sp. nov.
Holotype.—UTA A-64254 (field ID: JAC 22117; Fig.
22A), adult female from road between Yerba Santa and
Yextla (HWY 196), Guerrero, Mexico, 17.526668N,
99.95798W, 2071 m, collected by J.A. Campbell and
colleagues on 10 June 2002.
Paratypes (5).—MZFC-HE-35600–01 and UTA A-
66117–18 adult females, and UTA A-66119 adult male
(Fig. 22B–D), all same collection data as holotype.
Diagnosis.—A species of Craugastor distinguished by the
following combination of characters: (1) small adult size
(maximum SVL ¼16.7 mm); (2) full ossification of skeletal
elements in adults; (3) absence of posterolateral projection of
frontoparietal; (4) absence of vomerine odontophores; (5)
presence of raised tubercles on eyelids; (6) supratympanic
fold absent or poorly developed; (7) face flank barred or with
supralabial pale stripe, and with or without dark canthal
stripe; (8) single postrictal tubercle; (9) gular region
peppered with melanocytes; (10) dorsal surface two-toned,
usually with a dark suprascapular shape, or almost
unicolored; (11) pale or ground color middorsal ridge; (12)
scattered fine tubercles on dorsum; (13) body flank barred
darker anteriorly, slightly shagreened to smooth; (14)
inguinal glands present and axillary glands absent in adults;
(15) when leg adpressed to body, heel reaches middle of eye
to slightly beyond snout; (16) outer tarsal ridge with 1–6
small mostly round tubercles, no raised fringe; (17) finger
FIG. 12.—Skull shape variation within the Craugastor mexicanus series as
depicted by principal components analysis of 14 geometric morphometric
landmarks (Fig. 10). The first 7 principal components explained ~75% of the
variance. The second and third principal components are depicted because
they maximized differences between large-bodied and small bodied species.
FIG. 13.—Evidence of sexual dimorphism of tympanum width in Craugastor mexicanus (left) and C. pygmaeus (right). Sex was determined by directly
examining gonads for select specimens of C. mexicanus and is implied by scaling patterns in those specimens where sex was not directly determined.
13
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
and toe tips round, slightly lanceolated, slightly expanded;
(18) inner metatarsal tubercle larger than outer metatarsal
tubercle.
Comparisons.Craugastor bitonium can be differenti-
ated from C. mexicanus, C. montanus, C. omiltemanus, and
C. saltator by their large adult body sizes of SVL .20 mm
(,20 mm in C. bitonium), the presence of vomerine
odontophores (absent in C. bitonium), and the presence of
three palmar tubercles (one palmar tubercle in C. bitonium).
Craugastor bitonium can be differentiated from C. cande-
lariensis, C. cueyatl, C. hobartsmithi, and C. portilloensis by
the presence of metatarsal tubercles of similar sizes
(different sizes in C. bitonium). Craugastor bitonium can
be differentiated from C. rubinus by the presence of
posterolateral projections of the frontoparietal (absent in
C. bitonium). Craugastor bitonium can be differentiated
from C. polaclavus by its shorter eye–nostril distance with an
eye–nostril distance 9–10% SVL in C. bitonium and 10–13%
SVL in C. polaclavus.Craugastor bitonium is most similar to
C. pygmaeus (in morphology, osteology, and genetic
distance), but may be differentiated from this taxon by the
condition of the outer tarsal ridge; 1–6 small tubercles in C.
bitonium versus no tubercles (¼smooth) in C. pygmaeus.
Holotype description.—Holotype small female with
unpigmented developing ova (SVL ¼15.8 mm); snout
rounded and short (0.9 mm naris–snout; 6% SVL); short
eye–nostril distance (1.42 mm; 9% SVL); tympanum 1.4 mm
(8.9% SVL); small supratympanic fold terminating in
shoulder tubercle; finger length formula III ,IV ,II ,
I; single palmar tubercle; single prepollical tubercle;
subarticular tubercles present on all fingers; no supernu-
merary tubercles present on hands; toe length formula IV ,
III ,V,II ,I; inner metatarsal tubercle larger than outer
metatarsal tubercle; subarticular tubercles present on all
toes; supernumerary tubercles present on plantar surface;
small dark supracloacal fold present; white lip bar in life
(Fig. 22A), still evident in preservative (Fig. 19A); 2–3
incomplete bands on each arm; one leg removed for genetic
analysis and 4 bands on thigh of remaining leg; dorsum
brown and mottled on head and anterior-most third of body,
TABLE 5.—Snout–vent length (SVL in mm) and gonad pigmentation
condition of male and female specimens of species in the Craugastor
mexicanus series. Number of specimens examined is indicated in
parentheses.
Species Male SVL Female SVL
Male gonad
pigmentation
Female gonad
pigmentation
bitonium 12.3 (1) 15.2–16.9 (5) Yes (1) No (5)
candelariensis 12.4–14.3 (3) 18.6 (1) Yes (3) No (1)
cueyatl 11.7–12.3 (2) 15.7 (1) Yes (1) Slightly (1)
hobartsmithi 11.3–15.2 (5) 16.68 (1) Yes (3) Unknown
mexicanus 17.2–28.5 (8) 28.3–40.5 (7) Yes (12) Variable (14)
montanus 19.6–21.8 (3) 24.9–25.8 (2) Yes (2) Yes (1)
omiltemanus 18.0–21.8 (3) 30.2–37.7 (6) Yes (3) No (6)
polaclavus 11.6–12.3 (2) 15.3 (1) Yes (2) No (1)
portilloensis 8.4–12.6 (3,
subadult/
juvenile)
11.4–12.1 (2,
subadult)
Yes (3) No (2)
pygmaeus 11.6–14.9 (8) 14.3–17.5 (7) Variable (8) Variable (7)
rubinus 10.8–12.6 (3) Unknown Yes (3) Unknown
saltator 16.7–22.1 (4) 32.6–36.9 (4) Yes (4) Rarely (4,
sometimes
very slightly
pigmented)
TABLE 6.—Select characteristics of species in the Craugastor mexicanus series. X indicates presence, O indicates absence, and ? indicates uncertainty. If greater than one, number of specimens examined is
indicated in parentheses.
bitonium candelariensis cueyatl hobartsmithi mexicanus montanus omiltemanus polaclavus portilloensis pygmaeus rubinus saltator
Eyelid raised tubercles X O X X X/O X/O X/O X X/O X/O X X/O
Well-developed tympanic fold O O O O X X/O X O O O O X
Postrictal tubercles 1 2 1 1–2 1 (2 fused) 1 (2 fused) 1–2 1–2 1–2 1 (2 fused) 1 1–2
Posterolateral projection
of frontoparietal
O O (2) X O (3) X/O (27) X X (21) X X O (21) X (2) X (3)
Vomerine odontophores O X (2) O O (3) X (27) X/O X (21) O O O (21) O (2) X (3)
Axillary glands in adults O O O O X/O X O O O O O X
Heel reach past snout O O O O X X O O X/O X/O O X
Finger lengths III .IV
.II .I
X/O (II ¼IV) X X X (3) X/O (I II,
rare)
X X/O (I II) X X X/O (rarely
II ¼Ior
II ¼IV)
O (II ¼IV) X
Metacarpal tubercles
(including thenar)
2222332322223
Toe lengths IV .III .V
.II .I
XXX/O(V¼III) X/O (V ¼III) X/O (V ¼III) X/O (V ¼III) X X/O (V ¼III) X/O (V ¼III) X/O (III .
V, rarely
V¼III)
X X/O (V ¼
III, rarely
III .V)
Outer/inner metatarsal
tubercle
0.40–0.75 (6) 0.75–1.00(4) 0.56–0.89(2) 0.62–0.67 (3) 0.50–0.75 (12) 0.50–0.75 (4) 0.40–0.55% (14) 0.50–0.83 (6) 0.67–1.25 (5) 0.35–0.67 0.50–0.55 (3) 0.36–0.58 (10,
0 in one
individual,
one side)
Eye–nostril distance ratio 9–10% (6) 9–13% (4) 7–11% (2) 8–10% (5) 8–13% (42) 8–11% (5) 8–12% (39) 10–13% (6) 10–11% (5) 7–12% (48) 10–11% (3) 10–12% (16)
Crus ratio 50–55% (6) 51–58% (4) 51–54% (2) 53–61% (5) 53–69% (191) 53–59% (5) 32–65% (39) 0.50–0.58 (6) 51–61% (5) 50–66% (48) 50–58% (3) 62–73% (16)
14 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
eclipsed by lighter coloration on posterior two-thirds of
body; ventral surface lightly colored in preservative; skull of
holotype lacks vomerine odontophores (although vomers
present).
Variations in paratypes.—Body sizes (SVL) 15.8 mm
(MZFC-HE-35600), 15.2 mm (MZFC-HE-35601), 16.9 mm
(UTA A-66117), 16.7 mm (UTA A-66118), 12.3 mm (UTA A-
66119); eye–nostril distance 9–10% SVL; tympanic ratios 7–
9%; dorsal color patterns variable, often with two distinctive
patches of differing ground coloration ranging from orange
to tan.
Etymology.—The specific epithet is a combination of the
Latin prefix bi- meaning two and tonium meaning tone. It is
a reference to the two distinctive patches of color found on
the holotype and several paratypes that create the appear-
ance of a ‘‘two-tone’’ dorsal coloration.
Distribution.—This species is known only from the
Sierra Madre del Sur of central Guerrero (~2071 m). The
FIG. 14.—Dorsal body surface in representative specimens of the Craugastor mexicanus series. Female holotype of C. bitonium (A, UTA A-64254, SVL ¼
15.8 mm); male holotype of C. candelariensis (B, UTA A-64253, SVL ¼13.3 mm); male holotype of C. cueyatl (C, UTA A-62348, SVL ¼13.0 mm); male C.
hobartsmithi (D, UMMZ 94231, SVL ~11 mm, photo by J. David Curlis); male C. mexicanus (E, UTA A-6907, SVL ¼22.3 mm); male C. montanus (F,
UMMZ 88002, SVL ¼21.8 mm); female C. omiltemanus (G, UTA A-66140, SVL ¼30.9 mm); female holotype of C. polaclavus (H, UTA A-62392, SVL ¼
14.7 mm); female holotype of C. portilloensis (I, UTA A-62393, SVL ¼11.4 mm); female C. pygmaeus (J, UTA A-54809, SVL ¼17.3 mm); male holotype of
C. rubinus (K, UTA A-62345, SVL ¼12.6 mm); female C. saltator (L, UTA A-55239, SVL ¼38.2 mm). A color version of this figure is available online.
15
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
closest named places to the type locality are Izotepec to the
north and Los Bajos to the southwest. The habitat at the type
locality is montane pine–oak forest.
Diet.—CT-scan of the holotype revealed the presence of
a small millipede (Diplopoda) in the stomach. We also noted
the presence of a small red ant (Formicidae) in the mouth of
the holotype.
Phylogenetics.Craugastor bitonium was inferred to be
the sister taxon of C. pygmaeus, with high support in the
concatenated analyses (ML ¼99; BAYES ¼1.0; Fig. 3). This
sister relationship was also recovered in both mtDNA and
nDNA analyses, although with lower support in the nDNA-
only analyses (ML ¼54, BAYES ¼0.67; Figs. 4 and 5).
Craugastor bitonium is separated from C. pygmaeus by a P-
distance of 4.7% (Table 4).
Remarks.—The skull of C. bitonium is similar to C.
hobartsmithi, C. montanus, and C. pygmaeus.Craugastor
bitonium displays a developmental pattern similar to C.
FIG. 15.—Ventral body surface in representative specimens of the Craugastor mexicanus series. Female holotype of C. bitonium (A, UTA A-64254, SVL ¼
15.8 mm); male holotype of C. candelariensis (B, UTA A-64253, SVL ¼13.3 mm); male holotype of C. cueyatl (C, UTA A-62348, SVL ¼13.0 mm); female C.
hobartsmithi (D, UMMZ 94230, SVL ~12 mm, photo by J. David Curlis); female C. mexicanus (E, UTA A-28754, SVL ¼35.9 mm); male C. montanus (F,
UMMZ 88002, SVL ¼21.8 mm); female C. omiltemanus (G, UTA A-66140, SVL ¼30.9 mm); female holotype of C. polaclavus (H, UTA A-62392, SVL ¼
14.7 mm); female holotype of C. portilloensis (I, UTA A-62393, SVL ¼11.4 mm); female C. pygmaeus (J, UTA A-54809, SVL ¼17.3 mm); male holotype of
C. rubinus (K, UTA A-62345, SVL ¼12.6 mm); female C. saltator (L, UTA A-55239, SVL ¼38.2 mm). A color version of this figure is available online.
16 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
pygmaeus (with high levels of ossification at small sizes),
suggesting small adult body sizes (Fig. 11). This species likely
co-occurs with C. pygmaeus, C. omiltemanus, and C. saltator
in central Guerrero (Figs. 6 and 8). The specimen UTA A-
66132 is referred with some hesitation because it was
collected from a lower elevation than the type locality and
has a Finger I length nearing C. pygmaeus (which also
occurs in Guerrero). Six of the specimens were adult females
containing unpigmented ovaries with yolked eggs and thick
oviducts, the seventh (UTA A-66119; Fig. 22C, far left) was
an adult male with pigmented testes.
Craugastor candelariensis sp. nov.
Holotype.—UTA A-64253 (field ID: JAC 21885), male
collected by E.N. Smith and colleagues N of Candelaria on
the road to Oaxaca; Sierra Madre del Sur, Oaxaca, Mexico,
15.949608N, 96.471108W, 668 m, on 21 January 2002
FIG. 16.—Lateral body surface in representative specimens of the Craugastor mexicanus series. Female paratype of C. bitonium (A, UTA A-66118, SVL ¼
16.7 mm); female paratype of C. candelariensis (B, UTA A-55247, SVL ¼18.6 mm); female paratype of C. cueyatl (C, MZFC-HE-35614. SVL ¼17.2 mm);
male C. hobartsmithi (D, UMMZ 94231, SVL ~11 mm, photo by J. David Curlis); male C. mexicanus (E, UTA A-6907, SVL ¼22.3 mm); female C.
montanus (F, UMMZ 87970, SVL ¼25.8 mm); male C. omiltemanus (G, UTA A-66139, SVL ¼18.1 mm); female holotype of C. polaclavus (H, UTA A-
62392, SVL ¼14.7 mm); female holotype of C. portilloensis (I, UTA A-62393, SVL ¼11.4 mm); male C. pygmaeus (J, UTA A-64414, SVL ¼10.3 mm); male
holotype of C. rubinus (K, UTA A-62345, SVL ¼12.6 mm); male C. saltator (L, UTA A-54931, SVL ¼18.4 mm). A color version of this figure is available
online.
17
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
between 1130 and 1200 h, near stream bordering coffee
plantation and secondary forest.
Paratypes (3).—MZFC-HE-35617 (formerly UTA A-
64252; field ID: JAC 21873; Fig. 23), male with heavily
pigmented testes, same data as holotype except collected 1.2
mi on rough road toward Pluma Hidalgo on the Candelaria–
Portillo road, 15.956108N, 96.449308W, 1051 m, on 21
January 2002 at 1000 h in leaf litter of coffee plantation. UTA
A-66116 (Field ID: JAC 21851), male with pigmented testes
collected by E.N. Smith and colleagues from San Gabriel
Mixtepec, Puente de Hamaca, Oaxaca, Mexico, 16.105108N,
97.063108W, 710 m, on 20 January 2002 at 1520 h on forest
floor. UTA A-55247 (Field ID: ENS 9698), female with
unpigmented gonads and extended oviducts collected by
Karin S. Castaneda along the Carretera San Gabriel
Mixtepec–Miahuatla
´n of the Sierra Madre del Sur, Oaxaca,
Mexico, 16.1605568N, 97.001118W, 1270–1350 m, on 15
March 1998 at 1630 h from pine forest habitat.
FIG. 17.—Dorsal surface of the head in representative specimens of the Craugastor mexicanus series. Female paratype of C. bitonium (A, UTA A-66118,
SVL ¼16.7 mm); female paratype of C. candelariensis (B, UTA A-55247, SVL ¼18.6 mm); female paratype of C. cueyatl (C, MZFC-HE-35614. SVL ¼17.2
mm); female C. hobartsmithi (D, UMMZ 94230, SVL ~12 mm, photo by J. David Curlis); female C. mexicanus (E, UTA A-28754, SVL ¼35.9 mm); female
C. montanus (F, UMMZ 87970, SVL ¼25.8 mm); male C. omiltemanus (G, UTA A-66139, SVL ¼18.1 mm); female holotype of C. polaclavus (H, UTA A-
62392, SVL ¼14.7 mm); female holotype of C. portilloensis (I, UTA A-62393, SVL ¼11.4 mm); female C. pygmaeus (J, UTA A-54809, SVL ¼17.3 mm);
male holotype of C. rubinus (K, UTA A-62345, SVL ¼12.6 mm); male C. saltator (L, UTA A-54931, SVL ¼18.4 mm). A color version of this figure is
available online.
18 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Diagnosis.—A species of Craugastor distinguished by the
following combination of characters: (1) small adult size
(maximum SVL ¼18.6 mm); (2) full ossification of most
skeletal elements in adults, lacking ossification only of Stage
6 (Table 3); (3) absence of posterolateral projection of
frontoparietal; (4) presence of vomerine odontophores; (5)
absence of raised tubercles on eyelids; (6) supratympanic
fold absent or poorly developed; (7) face flank with nostril–
canthal–supratympanic stripe, lips colored as dorsum; (8)
two postrictal tubercles; (9) gular region uniformly pale to
slightly evenly peppered with melanocytes; (10) dorsal
surface unicolored pale; (11) pale middorsal ridge, some-
times with few tiny spots; (12) evenly fine tubercles on
dorsum; (13) body flank unicolored pale, shagreened with
fine tuberculation; (14) inguinal glands present and axillary
glands absent in adults; (15) when leg adpressed to body,
heel reaches between eye and tip of snout; (16) outer tarsal
ridge with 3–8 tiny and pointed tubercles on slightly raised
FIG. 18.—Ventral surface of the head in representative specimens of the Craugastor mexicanus series. Female paratype of C. bitonium (A, UTA A-66118,
SVL ¼16.7 mm); female paratype of C. candelariensis (B, UTA A-55247, SVL ¼18.6 mm); female paratype of C. cueyatl (C, MZFC-HE-35614. SVL ¼17.2
mm); male C. hobartsmithi (D, UMMZ 94231, SVL ~11 mm, photo by J. David Curlis); male C. mexicanus (E, UTA A-6907, SVL ¼22.3 mm); female C.
montanus (F, UMMZ 87970, SVL ¼25.8 mm); male C. omiltemanus (G, UTA A-66139, SVL ¼18.1 mm); female holotype of C. polaclavus (H, UTA A-
62392, SVL ¼14.7 mm); female holotype of C. portilloensis (I, UTA A-62393, SVL ¼11.4 mm); male C. pygmaeus (J, UTA A-64414, SVL ¼10.3 mm); male
holotype of C. rubinus (K, UTA A-62345, SVL ¼12.6 mm); male C. saltator (L, UTA A-54931, SVL ¼18.4 mm). A color version of this figure is available
online.
19
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
fringe; (17) finger and toe tips lanceolate to mucronate (toes
and outer two fingers); (18) similar sizes of inner and outer
metatarsal tubercles.
Comparisons.Craugastor candelariensis can be differ-
entiated from C. bitonium, C. mexicanus, C. montanus, C.
omiltemanus, C. polaclavus, C. pygmaeus, C. rubinus, and C.
saltator by a larger inner metatarsal tubercle (inner and
outer metatarsal tubercles are similar sizes in C. candelar-
iensis). Craugastor candelariensis can be differentiated from
C. cueyatl and C. hobartsmithi by the absence of vomerine
odontophores (present in C. candelariensis). It can be
differentiated from C. portilloensis by the presence of
posterolateral projections of the frontoparietal (absent in
C. candelariensis).
Description of holotype.—Holotype small male (SVL ¼
13.3 mm); snout rounded and short (0.5 mm naris–snout; 4%
SVL); long eye–nostril distance (1.7 mm; 13% SVL);
tympanum 1.2 mm (7.6% SVL); no supratympanic fold and
FIG. 19.—Lateral surface of the head in representative specimens of the Craugastor mexicanus series. Female holotype of C. bitonium (A, UTA A-64254,
SVL ¼15.8 mm); female paratype of C. candelariensis (B, UTA A-55247, SVL ¼18.6 mm); male holotype of C. cueyatl (C, UTA A-62348, SVL ¼13.0 mm);
female C. hobartsmithi (D, UMMZ 94230, SVL ~12 mm, photo by J. David Curlis); female C. mexicanus (E, UTA A-28754, SVL ¼35.9 mm); male C.
montanus (F, UMMZ 88002, SVL ¼21.8 mm); female C. omiltemanus (G, UTA A-66140, SVL ¼30.9 mm); female holotype of C. polaclavus (H, UTA A-
62392, SVL ¼14.7 mm); female holotype of C. portilloensis (I, UTA A-62393, SVL ¼11.4 mm); female C. pygmaeus (J, UTA A-54809, SVL ¼17.3 mm);
male holotype of C. rubinus (K, UTA A-62345, SVL ¼12.6 mm); male C. saltator (L, UTA A-54931, SVL ¼18.4 mm). A color version of this figure is
available online.
20 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
no shoulder tubercle; finger length formula III ,IV ,II ¼
I; single palmar tubercle; single prepollical tubercle;
subarticular tubercles present on all fingers; supernumerary
tubercles present on Finger III; toe length formula IV ,III
,V,II ,I; inner metatarsal tubercle and outer metatarsal
tubercle equal size; subarticular tubercles present on all toes;
supernumerary tubercles present on plantar surface; unable
to verify supracloacal fold state because posterior end
damaged when removing leg for genetic analysis; entire
body lightly colored in preservative (appears some saponi-
fication may have occurred).
Variations in paratypes.—Body sizes (SVL) 12.4 mm
(MZFC-HE-35617), 14.3 mm (UTA A-66116), 18.6 mm
(UTA A-55247); eye–nostril distance 10–13% SVL (males),
9% SVL (female); tympanic ratios 7–10%.
Etymology.—The name is an abbreviated allusion to the
municipality of Candelaria Loxicha (near the type locality)
and the Latin suffix -ensis meaning place. It is simultaneously
FIG. 20.—Ventral surface of the hand in representative specimens of the Craugastor mexicanus series. Female paratype of C. bitonium (A, UTA A-66118,
SVL ¼16.7 mm); female paratype of C. candelariensis (B, UTA A-55247, SVL ¼18.6 mm); male holotype of C. cueyatl (C, UTA A-62348, SVL ¼13.0 mm);
female C. hobartsmithi (D, UMMZ 94230, SVL ~12 mm, photo by J. David Curlis); female C. mexicanus (E, UTA A-28754, SVL ¼35.9 mm); male
Craugastor montanus (F, UMMZ 88002, SVL ¼21.8 mm); female C. omiltemanus (G, UTA A-66140, SVL ¼30.9 mm); female C. polaclavus (H, UTA A-
66097, SVL ¼16.0 mm); female holotype of C. portilloensis (I, UTA A-62393, SVL ¼11.4 mm); female C. pygmaeus (J, UTA A-54809, SVL ¼17.3 mm);
male holotype of C. rubinus (K, UTA A-62345, SVL ¼12.6 mm); male C. saltator (L, UTA A-54931, SVL ¼18.4 mm). A color version of this figure is
available online.
21
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
a reference to the Latin noun cand ¯
ela meaning a fire or light
made of wax, given the translucent yellow appearance of
several type specimens in preservative, as if someone were
shining a candle through them.
Distribution.—This species is known from intermediate
elevations of southern Oaxaca (668–1350 m), an area that
mostly consists of Sierra Madre del Sur pine–oak forest
habitat.
Phylogenetics.Craugastor candelariensis was strongly
supported as monophyletic in the concatenated analysis (ML
¼100; BAYES ¼1.0; Fig. 3). In this analysis, the sister taxon
of C. candelariensis was inferred to be C. polaclavus (ML ¼
80; BAYES ¼0.99). We also observed this sister relationship
in the nDNA-only analysis (Fig. 5); however, in the mtDNA-
only analysis C. candelariensis was inferred to be the sister
taxon of a clade containing C. bitonium þC. pygmaeus (Fig.
4). In terms of genetic distances, Craugastor candelariensis
FIG. 21.—Ventral surface of the foot in representative specimens of the Craugastor mexicanus series. Female paratype of C. bitonium (A, UTA A-66118,
SVL ¼16.7 mm); female paratype of C. candelariensis (B, UTA A-55247, SVL ¼18.6 mm); female paratype of C. cueyatl (C, MZFC-HE-35614. SVL ¼17.2
mm); male C. hobartsmithi (D, UMMZ 94231, SVL ~11 mm, photo by J. David Curlis); male C. mexicanus (E, UTA A-6907, SVL ¼22.3 mm); male
Craugastor montanus (F, UMMZ 88002, SVL ¼21.8 mm); male C. omiltemanus (G, UTA A-66139, SVL ¼18.1 mm); female holotype of C. polaclavus (H,
UTA A-62392, SVL ¼14.7 mm); female paratype of C. portilloensis (I, MZFC-HE-35581, SVL ¼12.1 mm); female C. pygmaeus (J, UTA A-54809, SVL ¼
17.3 mm); male paratype of C. rubinus (K, MZFC-HE-35616, SVL ¼10.8 mm); female C. saltator (L, UTA A-55239, SVL ¼38.2 mm). A color version of
this figure is available online.
22 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
is most similar to C. polaclavus and C. pygmaeus (both 6.4%;
Table 4).
Remarks.—The skull of C. candelariensis is similar to C.
bitonium, C. hobartsmithi, C. montanus, and C. pygmaeus,
with more posteriorly placed anterior suture of the
frontoparietal and prootic than in other species. Two type
specimens appear white-yellowish in preservative (possibly
having been saponified). This species likely co-occurs with C.
pygmaeus, C. polaclavus, and C. portilloensis in southcentral
Oaxaca (Figs. 6 and 8). In terms of body size and ossification
level it is the smallest member of the C. mexicanus series to
complete Stage 5 of our ontogenetic sequence.
Craugastor cueyatl sp. nov.
Microbatrachylus hobartsmithi: Duellman 1961:33 (in part,
based on field IDs EHT-HMS Nos. 18292–18352,
includes CAS 87816). [Misidentification].
Eleutherodactylus hobartsmithi: Castro-Franco et al.
2006:107. [Misidentification].
Holotype.—UTA A-62348 (field ID: JAC 27244; Fig.
24B), male collected by J.W. Streicher, C.L. Cox, J. Reyes-
Velasco, G. Weatherman, and C.M. Sheehy, III, on the road
from Avandaro to El Manzano, East of Cerro Gordo, Estado
de M´
exico, Mexico, 19.122098N, 100.139698W, 2311 m on
18 June 2008.
Paratypes (2).—MZFC-HE-35614 (Fig. 24A), female
with developed ova covered in mildly pigmented connective
tissue, same data as holotype, except 19.117358N,
100.139408W, 2282 m. AMNH A-57809 (Fig. 24C), male
from Tepozteco, Morelos, Mexico (Aztec archaeological site,
19.000798N, 99.101568W, 2000 m).
FIG. 22.—Female holotype of Craugastor bitonium (A, UTA A-64254, SVL ¼15.8 mm); female paratype (B, UTA A-66117, SVL ¼16.9 mm); (B); one
male and two female paratypes (C, left to right, UTA A-66119, SVL ¼12.3 mm; MZFC-HE-35600, SVL ¼15.8 mm; MZFC-HE-35601, SVL ¼15.2 mm);
female paratype (D, UTA A-66118, SVL ¼16.7 mm). Note two-tone dorsal color pattern in A, B, and C. All collected from the road between Yextla and
Vuelta del Sur, Guerrero, Mexico, 2071 m. A color version of this figure is available online.
FIG. 23.—Male paratype of Craugastor candelariensis (MZFC-HE-
35617, SVL ¼12.4 mm) from Sierra Madre del Sur, 1.2 mi on the road
between Candelaria and Portillo del Rayo, Pluma Hidalgo, Oaxaca, Mexico,
1051 m. A color version of this figure is available online.
23
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Referred specimens (4).—USNM 122054–55, and
USNM 139380 from Estado de M´
exico and Morelos,
respectively. MZFC 1089, from Tepoztla
´n, Barrio de
Ixcaltepec, central Morelos.
Diagnosis.—A species of Craugastor distinguished by the
following combination of characters: (1) small adult size
(maximum SVL ¼15.7 mm); (2) reduced ossification of the
skeleton relative to other members of the series, lacking
ossification of any skeletal elements beyond Stage 2 (Table
3); (3) presence of posterolateral projection of frontoparietal;
(4) absence of vomerine odontophores; (5) presence of
raised tubercles on eyelids; (6) supratympanic fold absent or
poorly developed; (7) face flank barred, with snout–nostril–
canthal–supratympanic stripe; (8) on postrictal tubercle; (9)
gular region lightly pigmented; (10) dorsal surface unicol-
ored dark; (11) dark middorsal ridge; (12) evenly tubercular
dorsum; (13) body flank unicolored pale with or without a
dark supra- or postaxillary band, shagreened; (14) inguinal
glands present and axillary glands absent in adults; (15) when
leg adpressed to body, heel reaches anterior corner of eye;
(16) outer tarsal ridge with 3–7 rounded tubercles, no raised
fringe; (17) finger tips round and not expanded, toe tips
slightly lanceolate and barely expanded; (18) similar sizes of
inner and outer metatarsal tubercles.
Comparisons.Craugastor cueyatl can be differentiated
from C. candelariensis, C. mexicanus, C. montanus, C.
omiltemanus, C. portilloensis, and C. saltator by the
presence of vomerine odontophores (absent in C. cueyatl).
It can be differentiated from C. bitonium, C. hobartsmithi,
C. polaclavus, and C. pygmaeus by the absence of
posterolateral projections of the frontoparietal (present in
C. cueyatl). It can be differentiated from C. rubinus by a
larger inner metatarsal tubercle than outer metatarsal
tubercle (inner and outer metatarsal tubercles are similar
sizes in C. cueyatl).
Description of the holotype.—Holotype small male
(SVL ¼12.3 mm); highly pigmented testes; snout rounded
and short (0.6 mm naris–snout; 4% SVL); medium eye–
nostril distance (1.4 mm; 11% SVL); tympanum 1.2 mm
(10% SVL); supratympanic fold terminating in small
shoulder tubercle; finger length formula III ,IV ,II ,
I; single palmar tubercle; single prepollical tubercle;
FIG. 24.—Female paratype (A, MZFC-HE-35614, SVL ¼15.8 mm) and male holotype (B, UTA A-62348, SVL ¼12.3 mm) of Craugastor cueyatl from
type locality; male paratype (C, AMNH A-57809, SVL ¼11.7 mm) from Tepozteco, Morelos, Mexico, ~2000 m (C). Leaf litter habitat near the type locality
east of Cerro Gordo, Estado de M ´
exico, Mexico, 2282 m (D).
24 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
subarticular tubercles present on all fingers; supernumerary
tubercles not present on hand; left arm removed for genetic
analysis; toe length formula IV ,III ,V,II ,I; inner
metatarsal tubercle and outer metatarsal tubercle equal in
size; subarticular tubercles present on all toes; supernumer-
ary tubercles present on plantar surface; small supracloacal
fold; 1–2 bands of melanocytes on remaining arm; bands on
legs evident in life barely visible in preserved specimen.
Variations in paratypes.—Body sizes (SVL) 15.7 mm
(MZFC-HE-35614), 11.7 mm (AMNH A-57809); eye–nostril
distance 7% SVL (female), unavailable (male); tympanic
ratios 8% (female), 11% (male); crus ratio 51% (female),
54% (male).
Etymology.—The specific epithet is taken from the word
for frog in Nahuatl, an Aztec language that has been spoken
in the Valley of Mexico since the 7th century, a region that
includes the type locality of C. cueyatl.
Distribution.—The new species occurs throughout the
central region of the Trans Volcanic Mexican Cordilleras
slopes, those facing the R´
ıo Balsas, in the states of Mexico
and Morelos (1670–2264 m), an area of mesic pine–oak
forest habitat. Castro-Franco et al. (2006) associated C.
cueyatl (as E. hobartsmithi) with dry tropical forests and
cultivated areas.
Diet.—The holotype’s stomach was found to contain a
small spider of unknown taxonomy.
Phylogenetics.Craugastor cueyatl was inferred to be
the sister taxon of C. bitonium þC. pygmaeus with strong
support in the concatenated analysis (87 ML; 0.99 BAYES;
Fig. 3). The placement of C. cueyatl was less certain in the
mtDNA and nDNA-only analyses (Figs. 4 and 5). In terms of
genetic distances, Craugastor cueyatl is most similar to C.
pygmaeus (5.9%; Table 4).
Remarks.—The skull of C. cueyatl is similar to that of C.
hobartsmithi, C. montanus, and C. pygmaeus with a more
posteriorly placed anterior suture of the frontoparietal and
prootic than in other species. Several of the referred
specimens are based on geographic occurrence and general
gestalt and should be further examined. One referred
specimen of C. cueyatl (USNM 139380) is reported to be
collected from Distrito Federal (¼Mexico City). We
georeferenced this locality as being in the city (Fig. 8), but
it is most likely the specimen was collected from outside the
metropolis. Very rarely do species of vertebrate occurring
near or in major metropolitan areas remain hidden from
science; the late discovery of this new species is likely
explained by the diminutive size of C. cueyatl and previous
confusion with C. hobartsmithi and C. pygmaeus.
Craugastor hobartsmithi (Taylor 1936)
Eleutherodactylus hobartsmithi Taylor 1936:355. Holotype
male (FMNH 100114) from ‘‘Uruapan, Michoaca
´n,
Mexico.’’ [Examined].
Microbatrachylus hobartsmithi (Taylor): Taylor 1940:501.
Craugastor hobartsmithi (Taylor): Crawford and Smith
2005:536.
Microbatrachylus pygmaeus Duellman 1961:34. [Misidenti-
fication].
Craugastor pygmaeus Ahumada-Carrillo et al. 2013:1338.
[Misidentification].
Diagnosis.—Based on six specimens. A species of
Craugastor distinguished by the following combination of
FIG. 25.—View from the type locality of Craugastor hobartsmithi, Uruapan, Michoaca
´n, looking toward Cerro Tancitaro (A). Craugastor cf. hobartsmithi;
female (B, UTA A-66134, SVL ¼13.2 mm); male (C, MZFC-HE-35613, SVL ¼9.6 mm); female (D, UTA A-66133, SVL ¼18.3 mm); all from near Las
Playitas (Carretera Playitas–Torre de Microndas), Michoaca
´n, Mexico, 1565 m. Craugastor cf. hobartsmithi from Montitlan, Colima, Mexico photographed
13 April 2020 (E, F, not collected, photo by J. Reyes-Velasco). A color version of this figure is available online.
25
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
characters: (1) small adult sizes (males 11.3–15.2 mm,
females ~16.7 mm; Table 5); (2) full ossification of most
skeletal elements in adults, lacking ossification beyond Stage
4 (Table 3); (3) absence of vomerine odontophores; (4)
absence of posterolateral projection of frontoparietal; (5)
presence of a row of 4–6 rounded tubercles along outer
edge, and 1–3 in mid upper eyelid; (6) supratympanic fold
poorly developed; (7) face flank barred with or without
distinctive canthal stripe; (8) one or two postrictal tubercles;
(9) gular region peppered with melanocytes; (10) dorsal
surfaces finely blotched, usually with dark interorbital bar
and suprascapular -shape, some individuals with pale
dorsal color and four stripes, paravertebral and lateral,
originating at corners of eyes and ending above groin (lateral
more prominent); (11) middorsal ridge (dark or background
color); (12) mostly smooth dorsum or with just fine tubercles
or folds toward back; (11) body flank darker anteriorly,
around axilla, slightly tubercular; (14) inguinal glands
present and axillary glands absent in adults; (15) when leg
adpressed to body, heel reaches between anterior corner of
eye and snout; (16) outer tarsal ridge with 4–6 rounded to
slightly pointed tubercles, no raised fringe; (17) finger and
toe tips round, finger tips slightly expanded, toe tips
expanded; (18) similar sizes of inner and outer metatarsal
tubercles.
Comparisons.Craugastor hobartsmithi can be differ-
entiated from C. bitonium, C. mexicanus, C. montanus, C.
omiltemanus, C. polaclavus, C. pygmaeus, and C. rubinus by
an inner metatarsal tubercle that is twice the size of the outer
(these are similar sizes in C. hobartsmithi). It can be
differentiated from C. candelariensis and C. saltator by the
presence of vomerine odontophores (absent in C. hobarts-
mithi). It can be differentiated from C. cueyatl and C.
portilloensis by the presence of posterolateral projections of
the frontoparietal (absent in C. hobartsmithi).
Description.—In previous literature, C. hobartsmithi has
been described as small-bodied with pigmented gonads
(Taylor 1936, 1940); presence of tubercles on the tarsus
(Duellman 1961); two palmar tubercles (Lynch 1965); tarsus
bearing a row of tubercles along its outer edge (Lynch 1970).
Holotype (FMNH 100114) small male (13.5 mm); snout
rounded and short (0.9 mm naris–snout; 6% SVL); short
eye–nostril distance (1.18 mm; 8.7% SVL); tympanum 1.9
mm (14% SVL). We further examined two specimens of C.
cf. hobartsmithi from coastal Michoaca
´n (UTA A-66133–34;
Fig. 25B and C) and noted the following characteristics:
supratympanic fold terminating in two postrictal tubercles;
finger length formula III ,IV ¼II ,I; toe length formula
IV ,III ,V,II ,I; inner metatarsal tubercle and outer
metatarsal tubercle equal size.
Distribution.Craugastor hobartsmithi occurs in the
pine–oak forest of Michoaca
´n. Craugastor cf. hobartsmithi
occurs throughout western Mexico in low to intermediate
habitats of Jalisco, Nayarit, Michoaca
´n, Guerrero, and
Sinaloa (Fig. 8; Hardy and McDiarmid 1969). Flores-
Cobarrubias et al. (2012) reported C. hobartsmithi from
Hostotipaquillo, Jalisco. Garc´
ıa and Ceballos (1994) reported
C. hobartsmithi from coastal Jalisco. The records of C.
pygmaeus reported in Duellman (1961) and Ahumada-
Carrillo et al. (2013) are all likely C. hobartsmithi or C. cf.
hobartsmithi because our molecular results indicate that C.
pygmaeus does not occur west of Guerrero. Similarly, many
iNaturalist (https://www.inautralist.org, accessed June 2019)
accounts of C.cf.hobartsmithi are listed under C.
pygmaeus—these accounts also suggest that C. cf. hobarts-
FIG. 26.—Skulls of ‘Microbatrachylus lineatissimus’ (A, paratype, FMNH 104548), Craugastor mexicanus (B, UTA A-56558) and C. pygmaeus (C, UTA A-
64272) in dorsal and lateral views. Posterolateral projection of the frontoparietal on C. mexicanus (B) is indicated by an arrow and grey shading. A color
version of this figure is available online.
26 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
mithi is much more widely distributed in western Mexico
than museum collections indicate.
Phylogenetics.Craugastor cf. hobartsmithi was found
to be the sister taxon of C. rubinus with strong support (ML
¼100; BAYES ¼1.0) in both the concatenated and separate
mtDNA and nDNA analyses (Figs. 3, 4, and 5). The pairwise
P-distances between C. cf. hobartsmithi and C. rubinus is
3.4% (Table 4); this is the smallest genetic distance between
any species of the C. mexicanus series, suggesting recent
divergence.
Remarks.—The skull of C. hobartsmithi is similar to C.
montanus and C. pygmaeus, with more posteriorly placed
anterior suture of frontoparietal and prootic than in other
species. We tentatively referred several museum collections
to C. hobartsmithi (Fig. 8) as C. cf. hobartsmithi, but these
should be further examined. The specimens of C. pygmaeus
reported by Duellman (1961) from Arteaga, Michoaca
´n, are
referred to C. cf. hobartsmithi because we examined several
C. pygmaeus from Oaxaca with tubercles on the outer edge
of the tarsus rendering Duellman’s (1961) apomorphic
character for C. hobartsmithi unreliable. Craugastor ho-
bartsmithi may co-occur with C. pygmaeus in southcentral
Guerrero (Figs. 6 and 8). The tissues of C. cf. hobartsmithi
used in our phylogenetic analysis originated from Colima.
Although we lack a voucher specimen for the tissue, the
collector of the tissue (J. Reyes-Velasco) provided us with
photographs of C. cf. hobartsmithi from Montitlan, near
where the tissue was collected (Fig. 25E and F). One female
specimen of C. hobartsmithi from near Uruapan (UMMZ
94230) had several intradermal trombiculid mites on its
venter (Fig. 15D).
Craugastor mexicanus (Brocchi 1877)
Leiuperus mexicanus Brocchi 1877:184. Holotype unsexed
(MNHNP 6218) from ‘‘Mexico’’ (¼Cerro San Felipe,
Oaxaca, Mexico [Smith and Taylor 1950]). [Examined].
Paludicola mexicana (Brocchi): Boulenger 1882:237; Neiden
1923:513.
Pleurodema mexicana (Brocchi): Parker 1927:475.
Microbatrachylus oaxacae Taylor 1940:505. Holotype male
(FMNH 100001) from ‘‘Cerro San Flipe, near Oaxaca,
Oaxaca, Mexico.’’ [Examined].
Microbatrachylus lineatissimus Taylor 1941:87. Holotype
male (FMNH 1000036) from ‘‘Cerro San Flipe, near
Oaxaca, Oaxaca, Mexico.’’ [Examined].
Microbatrachylus fuscatus Davis and Dixon 1957:146.
Holotype female (TCWC 12171) from ‘‘20 miles east
of Tulancingo, Hidalgo, Mexico.’’ [Not examined; near-
topotypic specimen from Hidalgo examined (UTA A-
66138)].
Eleutherodactylus oaxacae (Taylor): Lynch 1965:3.
Eleutherodactylus lineatissimus (Taylor): Lynch 1965:3.
Eleutherodactylus mexicanus (Brocchi): Gorham 1966:86.
Craugastor mexicanus (Brocchi): Crawford and Smith
2005:536.
Diagnosis.—Based on 26 specimens. A species of
Craugastor distinguished by the following combination of
characters: (1) large adult size (maximum SVL ¼40.5 mm);
(2) full ossification of the skeleton in adults; (3) presence of
posterolateral projection of frontoparietal (Fig. 26B); (4)
presence of vomerine odontophores (in larger individuals);
(5) presence or absence of raised tubercles on eyelids, ,4
smooth to round and only slightly protruding tubercles; (6)
supratympanic fold developed; (7) face flank, labium barred
with or without distinctive canthal stripe; (8) one (or two
fused) postrital tubercles; (9) gular region peppered with
melanocytes; (10) dorsal surface extremely variable, ranging
from dark or light stripes, dark hourglass, dark or pale dorsal
stripe of different widths, being unicolored dark or pale
brown, to bird-dropping coloration (black peppered grayish
dorsum with a dirty white medial streak and bright-white
heels), all color morphologies have variable presence of
interocular band and suprascapular or rump blotching or
stripping; in most snout colored as rest of body but
sometimes pale; (11) variable middorsal ridge; (12) dorsal
skin smooth or tubercular and may have hourglass and/or
vertebral and/or paravertebral ridges; (13) body flank
unicolored pale, slightly darker anteriorly, or spotted
posteriorly and or anteriorly, supratympanic dark coloration
sometimes reaching posterior axillary area; smooth to
shagreened; (14) inguinal gland present and axillary gland
sometimes present in adults; (15) when leg adpressed to
body, heel reaches snout tip or beyond; (16) outer tarsal
ridge with 0–5 rounded tubercles; smooth or with thick but
only slightly raised fringe; (17) finger and toe tips round and
expanded (rarely slightly spatulate or barely expanded and
somewhat pointed); (18) inner metatarsal tubercle larger
than outer metatarsal tubercle.
Comparisons.Craugastor mexicanus can be differenti-
ated from C. candelariensis, C. cueyatl, C. hobartsmithi, and
C. portilloensis by the equal sizes of the inner and outer
metatarsal tubercles (unequal sizes in C. mexicanus). It can
be differentiated from C. bitonium and C. pygmaeus by the
absence of a posterolateral projection of the frontoparietal
(present in C. mexicanus). It can be differentiated from C.
polaclavus and C. rubinus by the absence of vomerine
odontophores (present in C. mexicanus). It can be dif-
ferentiated from C. montanus by the condition of supra-
tympanic folds in adults; developed in C. mexicanus versus
moderate to poorly developed in C. montanus. It can be
differentiated from C. omiltemanus by ventral skin texture in
life; smooth to granular in C. mexicanus versus areolate in C.
omiltemanus.Craugastor mexicanus is most similar to C.
saltator. We were unable to identify any reliable morpho-
logical characters to differentiate C. mexicanus and C.
saltator; however, they have nonoverlapping geographic
distributions, with C. saltator only occurring in western
Guerrero (Fig. 6).
Description.—In previous literature, described as large-
bodied, long-legged with row of small tubercles on outer
edge of the tarsus (Taylor 1941); variable palmar tubercle
arrangements (palmar tubercle divided, single, or evaginat-
ed; Lynch 1965); inner metatarsal tubercle larger than outer
metatarsal tubercle; lack of tarsal tubercles (Lynch 2000).
Holotype (MNHNP 6318) is large (~40 mm SVL); owing
to poor preservation, columella of holotype partially
ruptured the tympana on both sides (Fig. 1B); finger lengths
III .IV .II .I; toe length IV .V¼III .II .I.
Distribution.—This species is widespread throughout
eastern Mexico in high-elevation habitats (1554–2700 m) of
Oaxaca, Puebla, Hidalgo, and Veracruz (Fig. 6). These
habitats span the Sierra Madre Oriental and Sierra Madre
del Sur. Canseco Ma
´rquez and Guti´
errez May´
en (2010)
27
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
report that C. mexicanus occurs in the forests adjacent to the
Tehuaca
´n-Cuicatla
´n Valley in Puebla and Oaxaca. It is likely
this species occurs in Guerrero; however, most specimens
resembling C. mexicanus we examined from Guerrero are
referrable to C. saltator.
Phylogenetics.—The concatenated data set placed C.
mexicanus as the sister taxon to a clade of C. omiltemanus þ
C. saltator with high support in the BAYES analysis (0.98)
but moderate support in the ML analysis (75; Fig. 3). This
appears to be a relationship supported exclusively by the
mtDNA data set because separate analysis of the nDNA
markers did not confidently infer a sister taxon of C.
mexicanus (Figs. 4 and 5). Although BAYES analyses
strongly supported the monophyly of C. mexicanus
(.0.90), the ML analyses recovered variable support
ranging from 79 (concatenated analysis) to 54 (nDNA-only
analysis). In terms of genetic distances (Table 4), C.
mexicanus was most similar to C. omiltemanus (4.9%),
followed by similarity to C. saltator (5.1%).
Intraspecific variation.—We examined over 220 speci-
mens of C. mexicanus for this revision (Appendix I), and
briefly provide some patterns of intraspecific variation that
we observed. Many populations have substantial color-
pattern polymorphism (Fig. 27), similar to what is observed
in the C. rhodopis series (Lynch 1966; Streicher et al. 2014).
In some northern populations (Hidalgo and Puebla), the
canthal mask is broken into a black spot posterior to the
tympanum (often with a thin yet distinctive white margin).
Specimens throughout Oaxaca varied in whether they had a
single or divided palmar tubercle (as reported by Lynch
1965, see his Fig. 2); in six specimens we examined there was
asymmetry with a single palmar tubercle on one hand and
divided on the other.
Patterns of dorsal ridging and coloration often coincide,
with observations of the following morphs: (1) straight raised
ridges that are tubercular and darker in coloration than the
rest of the dorsum; (2) straight raised dorsal ridges that are
tubercular but of the same color as the rest of the dorsum;
(3) straight lines of color that are dark but with no tubercular
raised ridges; (4) ridges that are not straight but form an
hourglass shape and are tubercular; (5) ridges that are not
straight but form an hourglass shape and are not tubercular;
(6) scattered tubercles on the dorsum, which are sometimes
dark and sometimes same color as the background; (7) pale
or dark unicolored dorsum; (8) a unicolored dorsum with a
pale median stripe that can be ridged or smooth; (9) a wide
dark band on the dorsum, which usually co-occurs with a
pale upper labium; and, (10) white hills of coloration and a
peppered grayish dorsum sometimes with a diffused lighter
cream color, a pattern that may mimic bird droppings
(appearing superficially as a smear of uric acid and digestive
waste).
FIG. 27.—Examples of intraspecific polymorphism in Craugastor mexicanus from Mexico; female (A, UTA A-56558, SVL ¼37.3 mm) from Putla de
Guerrero municipality; Oaxaca, female (B, UTA A-64415, SVL ¼28.3 mm) from Carretera Sola de Vega–Juquila, Oaxaca; subadult (C, UTA A-64271, SVL ¼
11.9 mm) from Carretera Sola de Vega–Juquila, Oaxaca; subadult and adult (D, UTA A-64258, SVL ¼13.6 mm and MZFC-HE-35584, SVL ¼unavailable)
both from 8.1 mi south of Sola de Vega, Oaxaca; adult (E, UTA A-66107, SVL ¼unavailable) from Carretera Sola de Vega–Juquila, Oaxaca; subadult (F,
UTA A-64260, SVL ¼16.6 mm) from Carretera Sola de Vega–Juquila, Oaxaca; female (G, UTA A-64413, SVL ¼23.5) from Sierra Miahuatla
´n, Oaxaca;
female (H, UTA A-56579, SVL ¼21.3 mm) from Sierra Negra, Puebla; female (I, UTA A-56559, SVL ¼35.4 mm) from Sierra Mazateca, Puerto Soledad,
Oaxaca; subadult (J, UTA A-64257, SVL ¼17.6 mm) from Sierra Mixe, Carretera Ayutla–Zacatepec, Oaxaca; subadult (K, UTA A-64259, SVL ¼10.9 mm)
from 6.1 mi south of San Miguel Suchixtepec, Oaxaca; subadult (L, UTA A-66138, SVL ¼unavailable) from 20 mi east of Tulancingo, Hidalgo. A color
version of this figure is available online.
28 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Remarks.—The skull of C. mexicanus is similar to that of
C. omiltemanus and C. saltator, with more anteriorly placed
anterior suture of the frontoparietal and prootic than in other
species. We noticed that the ‘‘M. lineatissimus’’ morphotype
(raised and parallel ridges on the dorsum) occurs widely
throughout the range of C. mexicanus. Interestingly, the
examined paratype of ‘‘M. lineatissimus’’ has a unique skull
shape for C. mexicanus. The shape is similar to that of C.
hobartsmithi, C. montanus, and C. pygmaeus, with a more
posteriorly placed anterior suture of the frontoparietal and
prootic than in other species, coupled with a narrower back
of the skull than in any other species (Fig. 27A). This unique
condition likely explains the specimen as an outlier in several
of our statistical analyses (Figs. 9, 11, and 12). However, we
were unable to CT-scan other individuals with the ‘‘ M.
lineatissimus’’ morphotype (or the holotype of M. lineatiss-
mus), so future investigation is necessary to determine
whether the examined paratype (FMNH 104548) is an
aberrant individual of C. mexicanus or represents a valid
taxon.
Lynch (1965) reports that ‘‘E. oaxacae’’ has parotoid
glands. However, we saw no evidence of these glands in the
two type specimens that we examined (FMNH 100001 and
FMNH 126638), nor are we aware of any species of
Craugastor that possess parotoid glands (¼large poison
glands on the nuchal region). Craugastor mexicanus likely
co-occurs with C. omiltemanus at high-elevation localities of
central Oaxaca (Figs. 6 and 8). At intermediate elevation
localities in Veracruz and Oaxaca, it may overlap with C.
pygmaeus (Fig. 7). Male C. mexicanus have significantly
larger tympana than do female C. mexicanus (Fig. 13).
Craugastor montanus (Taylor 1942)
Microbatrachylus montanus Taylor 1942:67. Holotype fe-
male (USNM 115507) from ‘‘ Mount Ovando, Chiapas,
Mexico.’’ [Examined].
Eleutherodactylus sartori Lynch 1965:10. [Replacement
name].
Craugastor montanus (Taylor): Crawford and Smith
2005:536.
Diagnosis.—Based on holotype (Fig. 1E). A species of
Craugastor distinguished by the following combination of
characters: (1) moderate adult size (holotype, SVL ¼24.5
mm); (2) ossification of most of skeleton in adults; (3)
presence of posterolateral projection of frontoparietal; (4)
presence of vomerine odontophores; (5) presence or absence
of raised tubercles on eyelids; (6) supratympanic fold
moderate to poorly developed; (7) face flank barred with
or without canthal stripe, 1–2 particularly dark bars below
eye; (8) one (or two fused) postrictal tubercles; (9) gular
region with pale spotting; (10) dorsal surface blotched or
unicolored pale; diffuse interorbital bar, small suprascapular
spots; sometimes with two dark rump spots (11) middorsal
ridge present; (12) dorsum smooth with only few fine
tubercles; (13) body flank darker anteriorly (post axillary),
slightly shagreened to smooth; (14) inguinal gland present
and axillary gland present in adults; (15) when leg adpressed
to body, heel reaches snout tip or beyond; (16) outer tarsal
ridge with 0–2 small and round tubercles close to heel, no
raised fringe; (17) finger and toe tips round, finger tips
slightly or not expanded, toe tips expanded; (18) inner
metatarsal tubercle larger than outer metatarsal tubercle.
Comparisons.Craugastor montanus can be differenti-
ated from C. candelariensis, C. cueyatl, C. hobartsmithi, and
C. portilloensis by equal sizes of the inner and outer
metatarsal tubercles (unequal sizes in C. montanus). It can
be differentiated from C. bitonium and C. pygmaeus by the
absence of a posterolateral projection of the frontoparietal
(present in C. montanus). It can be differentiated from C.
polaclavus and C. rubinus by the absence of vomerine
odontophores (present in C. montanus). It can be differen-
tiated from C. mexicanus, C. omiltemanus, and C. saltator by
the general shape of its skull (Fig. 12). It can be
differentiated from C. mexicanus by the condition of
supratympanic folds in adults; moderate to poorly developed
in C. montanus versus developed in C. mexicanus. It can be
differentiated from C. omiltemanus by ventral skin texture in
life; smooth to granular in C. montanus versus areolate in C.
omiltemanus. It can be differentiated from C. saltator by
relative leg length; crus 53–59% SVL in C. montanus versus
62–73% SVL in C. saltator.
Description.—In previous literature, described as mod-
erately sized (males average 16.2 mm SVL, females average
24.0 mm SVL); Finger I shorter than Finger II; three palmar
tubercles; testes black; inner metatarsal tubercle larger than
outer metatarsal tubercle (Lynch 2000). Lynch (2000:133)
redescribed C. montanus (as E. sartori) owing to ‘‘ errors in
Taylor’s (1942) original description that were repeated by
Lynch (1965, 1970)’’.
Distribution.—This species is known from intermediate
to high elevations (~2000 m) of the Sierra Madre de Chiapas
in the state of Chiapas, Mexico (Lynch 2000), and adjacent
regions of the Department of San Marcos, Guatemala
(Crawford and Smith 2005). This region contains a complex
mixture of dry forests, mixed forests, cloud forests, and pine–
oak forests.
Phylogenetics.—In the concatenated analysis, C. mon-
tanus was recovered as the sister taxon of all other members
of the C. mexicanus series (ML ¼66, BAYES ¼0.99; Fig. 3).
It also had this placement in the mtDNA-only analysis (Fig.
4), but not in the nDNA analysis where it was found with
weak support to be the sister taxon of a clade containing all
taxa except C. mexicanus and C. omiltemanus (ML ¼33,
BAYES ¼0.73; Fig. 5). This differs from the phylogenetic
placement of C. montanus (as E. sartori) in the nDNA-only
analysis of Crawford and Smith (2005). In terms of genetic
distances (Table 4), C. montanus was most similar to C.
polaclavus (5.7%) followed by similarity with C. mexicanus
(6.1%).
Remarks.—The skull of C. montanus is similar to C.
hobartsmithi and C. pygmaeus, with more posteriorly placed
anterior suture of the frontoparietal and prootic than in other
species. The skull of C. montanus was also described by
Lynch (2000) as E. sartori. Lynch (1965) created the
neonym, Eleutherodactylus sartori, because Eleutherodacty-
lus montanus was preoccupied by a West Indian species.
This is the most southernly distributed species in the C.
mexicanus series. Craugastor montanus likely co-occurs with
C. pygmaeus.
The type locality of C. greggi (Bumzahem 1955) is Volcan
Tajumulco in San Marcos, Guatemala near where we
sampled C. montanus for our molecular analysis (Fig. 6).
29
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
Although C. greggi was placed in the C. laticeps series of
Hedges et al. (2008), it is allied to the C. mexicanus series by
having Finger I shorter than Finger II. This affinity was
noted in the original description: ‘‘ Eleutherodactylus greggi
seems to agree most closely with the member of the
Eleutherodactylus mexicanus group. ..’’ (Bumzahem
1955:119). However, C. greggi was differentiated from C.
montanus (¼E. sartori) by Lynch (2000) on the basis of
fusion between the last presacral vertebrae and sacrum (not
fused in Cmontanus). Nonetheless, our preliminary
examinations of the holotype of C. greggi (Fig. 28) and
collections from the Sierra de Chiapas suggest that future
research is needed to confirm C. greggi can be differentiated
from C. montanus because the taxa are united by multiple
characters including (but not limited to) adult body size,
presence of vomerine odonotophores, presence of postero-
lateral projection of frontoparietal, gular region with pale
spotting, finger lengths, toe lengths, unequal sizes of the
inner and outer metatarsal tubercles, and geographic
distribution.
Craugastor omiltemanus (G ¨unther 1900a)
Syrrhaphus omiltemanus unther 1900a:213. Holotype
unsexed (BMNH 1901.12.19.7) from ‘‘ Omilteme, Guer-
rero, Mexico.’’ [Examined].
Hylodes calcitrans unther 1900b:229. Holotype unsexed
(BMNH 1901.12.19.25) from ‘‘ Omilteme, Guerrero,
Mexico.’’ [Examined].
Eleutherodactylus omiltemanus (G ¨unther): Lynch 1970:175.
Craugastor omiltemanus (G ¨unther): Crawford and Smith
2005:536.
Diagnosis.—Based on Hylodes calcitrans type series of
25 individuals and 1 additional specimen. A species of
Craugastor distinguished by the following combination of
characters: (1) large adult size (maximum SVL ¼38.8 mm);
(2) reduced ossification of skeleton in adults relative to other
members of series, lacking ossification of any elements
beyond Stage 2 (Table 3) except for the sphenethmoid; (3)
presence of posterolateral projection of frontoparietal; (4)
presence of vomerine odontophore; (5) presence or absence
of raised tubercles on eyelids, smooth, six round and only
slightly protruding tubercles, sometimes a few aligned at the
outer edge; (6) supratympanic fold developed; (7) face flank
barred or dark; canthus dark, pale, or spotted; canthus with
or without a stripe (complete or broken); (8) one or two
postrictal tubercles; (9) gular region from evenly scattered
fine pigmentation, to densely pigmented with a mid-pale
stripe; (10) dorsal surface unicolored pale brown or grey;
unicolored to lightly spotted above the scapular and/or rump
areas, or tubercles and ridges (if present), interorbital bar;
(11) with or without a middorsal ridge; (12) dorsum smooth
with only few fine tubercles toward flanks and urostyle or
with ridges forming hourglass patterns and medium
tuberculation; (13) body flank darker anteriorly (postaxillary)
due to posterolateral expansion of supratympanic stripe;
sometimes with contrasting white and dark blotching
inguinal area, otherwise pale colored; shagreened; (14)
inguinal gland present and axillary gland present in adults;
(15) when leg adpressed to body, heel reaches middle of eye
to mid-canthal area; (16) outer tarsal ridge 0–5 rounded and
only slightly raised tubercles, no raised fringe; (17) finger and
toe pads round, fingertips slightly or not expanded, toe tips
expanded; (18) inner metatarsal tubercle larger than outer
metatarsal tubercle.
Comparisons.Craugastor omiltemanus can be differ-
entiated from all other members of the C. mexicanus series
by the combination of rough (and often raised) areolate skin
on its venter and a massive inner metatarsal tubercle (Figs.
21G and 29; Lynch 1970, 2000).
Description.—In previous literature described as large-
bodied, short-legged, with Toe III ,Toe V; inner metatarsal
tubercle up to five times larger than outer metatarsal
tubercle (Fig. 29A); subarticular tubercles conical; vomerine
odontophores; few supernumerary plantar tubercles (Taylor
1941; Lynch 2000).
There are two syntypes of C. omiltemanus, BMNH
1901.12.19.7 (reregistered as BMNH 1947.2.16.62; Fig. 1J)
and BMNH 1901.12.19.8 (reregistered as BMNH
1947.2.16.63). We designate the former as the lectotype
FIG. 28.—Female holotype of Craugastor greggi (FMNH 20876, SVL ¼36.0 mm) of the C. laticeps series (sensu Hedges et al. 2008) from Volcan
Tajumulco, near San Marcos, Guatemala, in dorsal (left) and ventral (right) views. Images courtesy of the Field Museum of Natural History. A color version
of this figure is available online.
30 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
and the latter as the paralectotype of this species. Both
specimens appear to have partially desiccated at some point
in their history, whereas the type series of Hylodes calcitrans
is in much better condition (Fig. 1K; Fig. 23A and B). We
designate BMNH 1901.12.19.29 (reregistered as BMNH
1947.2.16.47) as the lectotype of H. calcitrans. Relative
finger lengths of the types are III .IV .II .I and relative
toe lengths are IV ,III ,V,II ,I. Dorsal skin texture
raised, often with an X pattern of sebaceous glands in life
(Fig. 30A–D).
Distribution.—Prior to our study, C. omiltemanus was
known only from Guerrero (G ¨unther 1900a; Crawford and
Smith 2005). Our discovery of a specimen from Oaxaca
(UTA A-64264) that is both phylogenetically and morpho-
logically assignable with C. omiltemanus extends the
distribution of this species eastward. This species occurs in
high-elevation pine–oak forest habitats of the Sierra Madre
del Sur in the states of Guerrero and Oaxaca (Fig. 8; Fig.
30E and F).
Phylogenetics.—In all analyses, C. omiltemanus was
found to be monophyletic with strong support (ML .90;
BAYES .0.90; Figs. 3 and 4). In the concatenated analysis
C. omiltemanus was found to be the sister taxon of C.
saltator, although with limited support (ML ¼44; BAYES ¼
0.74; Fig. 3). There was less support for the sister
relationship with C. saltator in the mtDNA-only analysis
(ML ¼33; BAYES ¼0.82; Fig. 4). In the nDNA-only
analysis (which did not include C. saltator), C. omiltemanus
was placed in basal polytomy with two other clades: (1) C.
mexicanus and (2) all other species of the C. mexicanus
series (Fig. 5). In terms of genetic distances (Table 4), C.
omiltemanus was most similar to C. mexicanus (4.9%),
followed by similarity to C. saltator (5.6%).
Remarks.—The skull of C. omiltemanus is similar to C.
mexicanus and C. saltator with a more anteriorly placed
anterior suture of frontoparietal and prootic than that in
other species. Skull examination also was conducted on a
smaller subadult individual (UTA A-55240, SVL ¼16.5 mm)
that had only Stage 1 of the ontogenetic sequence complete
and only two features of Stage 2 present. Unlike the adult
specimen, the nasals are completely unossified and the
frontoparietal greatly reduced, and while vomerine odonto-
phores are present the posterolateral projection of the
frontoparietal is absent. Craugastor omiltemanus likely
shares a similar distribution with other Craugastor species
endemic to the Sierra Madre del Sur (e.g., C. uno; Streicher
et al. 2011). Lynch (2000) reports that C. omiltemanus has
white testes, but we observed pigmented testes in this
species (Table 5). Throughout its range, C. omiltemanus
likely co-occurs with C. bitonium, C. mexicanus, C.
FIG. 29.—Ventral coloration and areolate skin texture of Craugastor omiltemanus. Venter in preservative, an arrow indicates the large inner metatarsal
tubercle that is characteristic of the species (A, BMNH 1947.2.16.44, SVL ¼36.5 mm, photo by W.-Y. Cooper); venter in preservative (B, BMNH
1947.2.16.47, SVL ¼39.6 mm, photo by W.-Y. Cooper); venter in life (C, UTA Field ID: JAC 27695); venter in life (D, UTA Field ID: JAC 27694). A color
version of this figure is available online.
31
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
pygmaeus, and C. saltator. Male C. omiltemanus have
significantly larger tympana than do female C. omiltemanus
(Lynch 2000; this study).
Craugastor polaclavus sp. nov.
Holotype.—UTA A-62392 (field ID: JAC 21230; Fig. 31),
female collected by E.N. Smith and colleagues in Portillo del
Rayo, Distrito San Pedro Pochutla, Sierra Madre del Sur,
Oaxaca, Mexico, 15.973038N, 96.997118W, 1550–1585 m, 24
September 2001.
Paratypes (4).—UTA A-55246, a recent hatchling
collected by E.N. Smith and Jos ´
eLu
´
ıs Camarillo Rangel
from R´
ıo Salado, Sierra Madre del Sur, Oaxaca, Mexico,
16.1941678N, 97.09758W, 1245 m, 26 September 1997. UTA
A-66098 and MZFC-HE-35582–83, adult or subadult
specimens all collected with the holotype.
Referred specimen (1).—UTA A-66097, female, same
data as holotype.
Diagnosis.—A species of Craugastor distinguished by the
following combination of characters: (1) small adult size
(maximum SVL ¼14.7 mm); (2) highly reduced ossification
of skeleton in adults relative to other members of series; (3)
presence of posterolateral projection of the frontoparietal;
(4) absence of vomerine odontophores; (5) presence of
raised tubercles on eyelids; (6) supratympanic fold absent or
poorly developed; (7) face flank barred with no distinctive
canthal stripe, 1–2 particularly dark bars below eye; (8) one
or two postrictal tubercles; (9) gular region with pale
spotting; (10) dorsal surface blotched; suprascapular chev-
ron, interorbital bar; (11) pale or as background middorsal
ridge; (12) dorsum smooth with only large scattered
tubercles; (13) body flank darker anteriorly, no sharp
delineation of color change, smooth to shagreened; (14)
inguinal glands present and axillary glands absent in adults;
(15) when leg adpressed to body, heel reaches nostril; (16)
outer tarsal ridge with 0–4 extremely small, flat, and round
tubercles, no raised fringe; (17) finger and toe pads round,
finger tips not expanded, toe tips slightly expanded; (18)
FIG. 30.—Adult Craugastor omiltemanus (A, MZFC-HE-35699 [field ID: TJD 780], SVL ¼unavailable, photo by Thomas Devitt) from near Mazatla
´n,
Guerrero; female (B, UTA A-66140, SVL ¼30.5 mm) from habitat along the road between Chichihualco and Chilpancingo, Guerrero, 2221 m; male (C,
UTA A-64264, SVL ¼12.9 mm) from near Ayutla, Oaxaca, 1900 m; female (D, UTA A-60796, SVL ¼27.5 mm) from Sierra de Malinaltepec (Carretera San
Luis Acatla
´n–Tlapa de Comonfort), 2298 m; habitat found 13.2 km west of Mazatla
´n, Sierra Madre del Sur (E, photo by Thomas Devitt); habitat found in the
Sierra Madre del Sur west of Chilpancingo, Guerrero (F). A color version of this figure is available online.
FIG. 31.—Subadult female holotype of Craugastor polaclavus (UTA A-
62392, SVL ¼14.7 mm) from San Pedro Pochutla district, Portillo del Rayo,
Oaxaca, Mexico, 1550–1585 m. A color version of this figure is available
online.
32 Herpetological Monographs 36, 2022
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
inner metatarsal tubercle larger than outer metatarsal
tubercle.
Comparisons.Craugastor polaclavus can be differenti-
ated from C. candelariensis, C. mexicanus, C. montanus, C.
omiltemanus, and C. saltator by the presence of vomerine
odontophores (absent in C. polaclavus). It can be differen-
tiated from C. bitonium, and C. cueyatl by the condition of
adpressed leg where the heel does not reach the nostril
(reaches nostril in C. polaclavus). It can be differentiated
from C. hobartsmithi and C. pygmaeus by the absence of a
posterolateral projection of the frontoparietal (present in C.
polaclavus). It can be differentiated from C. portilloensis by
metatarsal tubercles of equal size (unequal in C. polaclavus).
It can be differentiated from C. rubinus by relative finger
lengths of IV ¼II (IV .II in C. polaclavus).
Description of holotype.—Holotype small female (SVL
¼14.7 mm); snout rounded and short (0.8 mm naris–snout;
5% SVL); short eye–nostril distance (1.4 mm; 7% SVL);
tympanum 1.4 mm (10% SVL); mild supratympanic fold
terminating in small shoulder tubercle; finger length formula
III ,IV ¼II ,I; single palmar tubercle; single prepollical
tubercle; subarticular tubercles present on all fingers;
supernumerary tubercles not present on hand; toe length
formula IV ,III ,V,II ,I; inner metatarsal tubercle
larger than outer metatarsal tubercle; subarticular tubercles
present on all toes; supernumerary tubercles present on
plantar surface; small supracloacal fold; dark supratympanic
fold in life with orange shoulder tubercle (Fig. 25), shoulder
tubercle color not visible in preservative; many bands
present on arms and legs; left leg removed for genetic
analysis.
Variations in paratypes.—Body sizes in SVL 8.2 mm
(UTA A-55246), 12.3 mm (UTA A-66098), 11.6 mm (MZFC-
HE-35582), 13.5 mm (MZFC-HE-35583); eye–nostril dis-
tance 11–13% SVL; tympanic ratios 7–8%.
Distribution.—Intermediate elevations in the foothills of
the Sierra Madre del Sur in Oaxaca 1245–1585 m (Fig. 7).
These habitats consist of mixed tropical dry and temperate
sierra forests.
Etymology.—The specific epithet is a combination of the
Latin pola meaning small and clavus meaning wart. The
name is an allusion to the small size and rugose appearance
of several individuals in the type series.
Phylogenetics.Craugastor polaclavus was inferred to
be the sister taxon of C. candelariensis with strong support in
the concatenated analysis (90 ML; 0.99 BAYES; Fig. 3) and
nDNA-only analysis (ML .90, BAYES .0.90; Fig. 5). The
placement of C. polaclavus was less certain in the mtDNA,
where it was found to be the sister taxon of a clade
containing C. cf. hobartsmithi þC. rubinus (ML ¼51,
BAYES ¼0.74; Fig. 4). In terms of genetic distances (Table
4), C. polaclavus was most similar to C. portilloensis (5.8%),
followed by similarity to C. bitonium (5.9%).
Remarks.—The skull of C. polaclavus is similar to C.
mexicanus, C. omiltemanus, and C. saltator, with more
anteriorly placed anterior suture of the frontoparietal and
prootic than in other species. Specimens were dissected;
three (UTA A-62392, 66098, and MZFC-HE-35582) seem to
be subadult females with unpigmented ovaries and thin
undeveloped oviducts, UTA A-66097 is an adult female with
a thickened oviduct and also unpigmented ovaries containing
yolked eggs, UTA A-66098 adult and MZFC-HE-35582
subadult are males with pigmented testes (smaller on second
specimen), and the hatchling (8.2 mm SVL) was not
dissected. This species likely co-occurs with C. candelar-
iensis, C. portilloensis, and C. pygmaeus in southcentral
Oaxaca (Figs. 6 and 8). It was collected in sympatry with C.
portilloensis at Portillo del Rayo, Oaxaca, Mexico (Fig. 8).
The hatchling paratype specimen (UTA A-55246) had no
evidence of skeletal ossification (Fig. 11).
Craugastor portilloensis sp. nov.
Holotype.—UTA A-62393 (field ID: JAC 21431; Fig. 32),
subadult female collected by E.N. Smith and colleagues in
Portillo del Rayo, Distrito San Pedro Pochutla, Sierra Madre
del Sur, Oaxaca, Mexico, 15.9794448N, 96.5166678W, 1550
m, 1 October 2001.
Paratypes (4).—MZFC-HE-35580 and UTA A-66095,
juvenile males. MZFC-HE-35581 subadult female, and UTA
A-66096 subadult male. All collected at the type locality
between 1550 and 1585 m on 24 September 2001 by E.N.
Smith and colleagues.
Diagnosis.—A species of Craugastor distinguished by the
following combination of characters: (1) small adult size
(maximum SVL ¼11.4 mm); (2) reduced ossification of
skeleton in adults relative to other members of the series,
lacking ossification of any skeletal elements beyond Stage 2
(Table 3); (3) presence of posterolateral projection of
frontoparietal; (4) absence of vomerine odontophores; (5)
presence or absence of raised tubercles on eyelids; (6)
supratympanic fold absent or poorly developed; (7) face flank
darker than dorsum, faintly barred lips, snout–nostril–
canthal–supratympanic stripe; (8) one or two postrictal
tubercles; (9) gular region with pale spotting; (10) dorsal
surface unicolored pale, diffuse interorbital bar, sometimes
with two dark rump spots; (11) with or without a middorsal
ridge; (12) dorsum smooth or with scattered fine tubercles;
(13) body flank has dark supratympanic stripe extending
toward lower mid-flank, smooth, very few small tubercles;
(14) inguinal glands present and axillary glands absent in
adults; (15) when leg adpressed to body, heel reaches
FIG. 32.—Female holotype of Craugastor portilloensis (UTA A-62393,
SVL ¼11.4 mm) from Portillo del Rayo, Oaxaca, Mexico, 1550 m. The
specimen’s leg was removed for genetic analysis. A color version of this
figure is available online.
33
JAMESON ET AL.—NEW SPECIES OF CRAUGASTOR FROM MEXICO
Downloaded From: https://bioone.org/journals/Herpetological-Monographs on 06 Apr 2022
Terms of Use: https://bioone.org/terms-of-useAccess provided by Natural History Museum London
between eye and slightly past snout; (16) outer tarsal ridge
smooth or with 1–3 extremely small, flat, and round
tubercles, no raised fringe; (17) finger and toe pads round,
finger tips not expanded, toe tips slightly expanded; (18)
similar sizes of inner and outer metatarsal tubercles.
Comparisons.Craugastor portilloensis can be differen-
tiated from C. candelariensis, C. mexicanus, C. montanus, C.
omiltemanus, and C. saltator by the presence of vomerine
odontophores (absent in C. portilloensis). It can be
differentiated from C. bitonium, C. hobartsmithi, and C.
pygameus by the absence of a posterolateral projection of the
frontoparietal (present in C. portilloensis). It can be further
differentiated from C. pygmaeus by the absence of a canthal
mask (present in C. portilloensis). It can be differentiated
from C. cueyatl and C. polaclavus by metatarsal tubercles of
unequal size (equal in C. portilloensis). It can be differen-
tiated from C. rubinus by relative finger lengths of IV ¼II
(IV .II in C. portilloensis).
Description of holotype.—Holotype small female (SVL
¼11.4 mm); snout rounded and short (0.8 mm naris–snout;
6% SVL); short eye–nostril distance (1.1 mm; 9.8% SVL);
tympanum 0.8 mm (6.8% SVL); no supratympanic fold;
small shoulder tubercle; finger length formula III ,IV ,II
,I; single flat palmar tubercle; single flat prepollical
tubercle; subarticular tubercles present on all fingers;