12 Vertebrate Zoology 70-4-2020 Townsend Et Al

70 (4): 731 – 756
© Senckenberg Gesellschaft für Naturforschung, 2020. 2020
A critically endangered new species of polymorphic

stream frog (Anura: Hylidae: Atlantihyla) from the montane
rainforest of Refugio de Vida Silvestre Texiguat, Honduras
Josiah H. Townsend 1, 2, *, Luis A. Herrera-B. 3, Erich P. Hofmann 1, 4, Ileana R. Luque-
Montes 1, 2, Ayla N. Ross 1, Daniel Dudek, Jr. 1, 5, Catherine Krygeris 1, 6, Joseph E.
Duchamp 1 & Larry David Wilson 2, 7
1
 Department of Biology, Indiana University of Pennsylvania, Indiana, Pennsylvania 15705 – 1081, USA; [email protected] — 2 Centro
Zamorano de Biodiversidad, Escuela Agrícola Panamericana Zamorano, Departamento de Francisco Morazán, Honduras — 3 Departamento
de Biología, Universidad Nacional Autónoma de Honduras en el Valle de Sula, San Pedro Sula, Honduras — 4 Current Address: Science
Department, Cape Fear Community College, Wilmington, North Carolina, 28401, USA — 5 Department of Biology, University of Texas at
Arlington, Arlington, Texas, 76019, USA — 6 Department of Life and Physical Sciences, Mardela High School, 24940 Delmar Road, Mardela
Springs, Maryland 21837, USA — 7 16010 SW 207th Avenue, Miami, Florida 33187 – 1056, USA
Submitted February 25, 2020.

Accepted November 3, 2020.
Published online at www.senckenberg.de/vertebrate-zoology on November 20, 2020.
Published in print Q4/2020.
Editor in charge: Raffael Ernst
Abstract
The Chortís Highlands of Mesoamerica exhibit a high degree of in situ evolutionary diversification, exemplified by numerous endemic
radiations of stream-dwelling treefrogs (Anura: Hylidae: Atlantihyla, Duellmanohyla, and Ptychohyla), which have been a source of
ongoing taxonomic uncertainty. Recent evidence suggests that one species, Atlantihyla spinipollex, may conceal an unrecognized sister
species found in Refugio de Vida Silvestre Texiguat. We applied an iterative integrative taxonomic framework to assess this population
within the context of Chortís Highlands populations of Atlantihyla spinipollex sensu stricto, Duellmanohyla salvadorensis, D. salvavida,
D. soralia, and Ptychohyla hypomykter, using both a single locus (mtDNA: 16S) and multilocus (mtDNA: 12S, 16S; nDNA: POMC,
RAG-1, Rhodopsin) datasets accompanied by distance- and tree-based species delimitation methods to inform our taxonomy. Samples
of A. spinipollex sensu lato formed two deeply divergent monophyletic lineages, suggesting that populations from the central and eastern
Cordillera Nombre de Dios are conspecific, while the population from Refugio de Vida Silvestre Texiguat represents a previously unde-
scribed species. We analyzed morphological and bioacoustic variation within and between the two lineages of A. spinipollex sensu lato and
found support for recognition of two distinct taxa. We restricted the name A. spinipollex to populations in the central and eastern Cordillera
Nombre de Dios, and formally describe the Texiguat population as a new species. We recommend the new species be considered Critically
Endangered due to ongoing habitat loss within what remains of its highly restricted natural distribution. This new species joins 26 other
endemic species of amphibians and reptiles at Texiguat.
Resumen
Las tierras altas Chortís en Centro América presentan un alto grado de diversificación evolutiva in situ, demostrado por un numero de ra-
diaciones endémicas de ranas arborícolas asociadas con ecosistemas lóticos (Anura: Hylidae: Atlantihyla, Duellmanohyla y Ptychohyla), y
previamente han sido fuente de incertidumbre. Evidencia reciente sugiere que una especie, Atlantihyla spinipollex, podría ocultar un espe-
cie hermana no reconocida, en el Refugio de Vida Silvestre Texiguat. Aplicamos un modelo iterativo taxonómico integrativo para evaluar
esta población dentro del contexto de las poblaciones de las Tierras Altas Chortís de Atlantihyla spinipollex sensu stricto, Duellmanohyla
salvadorensis, D. salvavida, D. soralia, y Ptychohyla hypomykter, utilizando ambos, conjuntos de datos con locus génicos individuales
(mtDNA: 16S) y conjuntos de datos con múltiple loci (mtDNA: 12S, 16S; nDNA: POMC, RAG-1, Rhodopsin), acompañados por métodos
de delimitación de especies basados en distancia y arboles filogenéticos. Las muestras de A. spinipollex sensu lato formaron dos linajes
monofiléticos profundamente divergentes, sugiriendo que las poblaciones del centro y Este de la Cordillera Nombre de Dios son conspe-
cíficas, mientras que la población del Refugio de Vida Silvestre Texiguat representa un linaje a nivel de especie no descrito previamente.
Analizamos la variación morfológica y bioacústica dentro y entre los dos linajes de A. spinipollex sensu lato y encontramos un fuerte
ISSN 1864-5755 | eISSN 2625-8498 | DOI: 10.26049/VZ70-4-2020-12 731

Townsend, J. H. et al.: A critically endangered new species of Atlantihyla
soporte estadístico para el reconocimiento de dos taxones distintos. Restringimos el nombre A. spinipollex a las poblaciónes del centro y el
Este de la Cordillera Nombre de Dios, y formalmente describimos la población del Refugio de Vida Silvestre Texiguat como una especie
nueva. Recomendamos que la nueva especies sea considerada En Peligro Critico debido a la amenaza inmediata y continua de pérdida de
hábitat dentro de su distribución natural altamente restringida. Esta nueva especie se une a otras 26 especies endémicas de anfibios y rep-
tiles en Texiguat.
Key words
Atlantihyla melissa sp. nov., Atlantihyla spinipollex, Chortís Block Highlands, Cordillera Nombre de Dios, Ptychohyla hypomykter,
Refugio de Vida Silvestre Texiguat.
Introduction
The Chortís Block biogeographical province of Central Wilson & McCranie, 1989). Duellman (1963, 1970)
America exhibits a high degree of in situ evolution- noted differences between populations from Honduras
ary diversification, generated by a complex and ac- and Guatemala and suggested further work might reveal
tive geological history (Gutiérrez-García & Vásquez- them to represent distinct species, a hypothesis subse-
Domínguez, 2013; Jordan et al., 2008; Rogers et al., quently supported by phylogenetic evidence (Faivovich
2002; Townsend, 2014). Chortís-endemic radiations are et al., 2018).
primarily associated with the highlands and have been Ptychohyla merazi was described by Wilson & Mc
documented across a wide range of organisms, including Cranie (1989) on the basis of morphological differ-
amphibians (Crawford & Smith, 2005; Luque-Montes ences between a collection of stream treefrogs from the
et al., 2018; Rovito et al., 2015; Townsend, 2016). One Quebrada de Oro area on the southeastern side of PN
diverse group of amphibians, the treefrogs (Anura: Hyli- Pico Bonito and Ptychohyla from the rest of Hondu-
dae: Hylini), have multiple endemic radiations across ras, which, at the time, were recognized under a broad
Mesoamerica, including in the Chortís Highlands (Du- concept of P. (= A.) spinipollex. Further evaluation by
ellman, 1970, 2001; Duellman et al., 2016; Faivovich et McCranie & Wilson (1993) revealed that populations
al., 2005; Wiens et al., 2005, 2010). found throughout the Cordillera Nombre de Dios (then
Recently, an expanded phylogeny of Mesoamerican known to include Refugio de Vida Silvestre [RVS] Texi
hylids presented by Faivovich et al. (2018) included guat and PN Capiro y Calentura, in addition to PN Pico
proposed resolutions for long-standing taxonomic uncer- Bonito) were distinct from highland populations from
tainty involving a number of genera, including paraphyly the rest of Honduras, and subsequently 1) restricted the
among stream-associated treefrogs assigned to the gen- name P. (= A.) spinipollex to populations from the Cor-
era Bromeliohyla, Duellmanohyla, and Ptychohyla (see dillera Nombre de Dios, 2) synonymized P. merazi with
Duellman et al., 2016: 19 for a recent summary of these P. (= A.) spinipollex, and 3) described the remaining
issues). Faivovich et al. (2018)’s proposed taxonomic re- populations from Honduras and Guatemala as Ptycho-
visions had implications for a number of species endem- hyla hypomykter (Fig. 1). Following the McCranie &
ic to the Chortís Block. The enigmatic frog Ptychohyla Wilson (1993) concept, A. spinipollex was considered
salvadorensis was reassigned to Duellmanohyla, while to be endemic to three isolated localities in the Cordil-
its former congener P. spinipollex and its putative sister lera Nombre de Dios: RVS Texiguat, PN Pico Bonito,
taxon P. panchoi were placed in a new genus: Atlanti- and PN Capiro y Calentura (McCranie & Wilson, 2002),
hyla. Ptychohyla hypomykter, a species endemic to the and was recently discovered at two additional localities:
Chortís Block, remained a member of the genus Ptycho- Cerro Corre Viento and PN Nombre de Dios (Townsend
hyla, along with P. euthysanota and four other species & Wilson, 2016).
with distributions extralimital to the Chortís Block. Despite decades of confusion regarding the taxo-
The taxon Atlantihyla spinipollex was first described nomic status of Nuclear Central American populations
by Schmidt (1936: 45; as “Hyla spinipollex”) based on a of frogs presently assigned to A. spinipollex and P. hy-
single adult male from the “mountains behind Ceiba, At- pomykter, and long-standing questions as to the mono-
lantida, Honduras,” a locality corresponding to the north- phyly of populations of P. hypomykter found across the
ern slopes of steep mountains south of the coastal city of highlands of Honduras and Guatemala (Duellman, 1963,
La Ceiba that now lie within the limits of Parque Nacional 1970; McCranie & Wilson, 2002; Faivovich et al., 2005,
(PN) Pico Bonito (Fig. 1). This species was proposed for 2018), there has been no phylogenetic analysis to date
inclusion in the genus Ptychohyla by Stuart (1954: 169), that includes comprehensive sampling of populations as-
and subsequent workers considered highland frogs from signed to any of the aforementioned taxa. In an analy-
western Guatemala to northern Nicaragua to represent sis of COI barcodes from amphibians of the Cordillera
a single widespread species, Ptychohyla (=Atlantihyla) Nombre de Dios, Townsend & Wilson (2016) demon-
spinipollex (Duellman, 1963, 1970; Stuart, 1943, 1948; strated that samples assigned to A. spinipollex from
732
VERTEBRATE ZOOLOGY — 70 (4) 2020
RVS Texiguat formed a deeply divergent sister clade to American Society of Ichthyologists and Herpetologists
A. spinipollex from the vicinity of the type locality (PN (Sabaj Peréz, 2016), with the addition of the following
Pico Bonito) — 14.1 – 16.5% divergence for COI — and abbreviations: IRL (Ileana R. Luque-Montes field series),
suggested additional investigation was needed to assess JHT (Josiah H. Townsend field series), JS (Javier Suny-
the taxonomic status of this population. er field series), MMF (Isis Melissa Medina-Flores field
In this paper, we take an iterative approach to assess- series), and N (2006 – 08 Nicaragua field series of Scott
ing the systematics of A. spinipollex sensu lato, in which Travers, Javier Sunyer, Josiah Townsend, and Larry Wil-
we 1) use single-locus (16S rRNA) barcoding to make son). Field-collected tissue samples were preserved in
preliminary taxonomic assignments and identify redun- SED buffer (20% DMSO, 0.25 M EDTA, pH 7.5, NaCl
dant haplotypes for an ingroup consisting of 114 novel saturated; Seutin et al., 1991; Williams, 2007). Forest
samples of Atlantihyla, Duellmanohyla, and Ptychohyla formation definitions follow those of Holdridge (1967)
from the Chortís Highlands; 2) use a subset of the afore- and names of forest types, biogeographic regions, and
mentioned samples supplemented with additional se- mountains ranges that appear capitalized follow those
quence data from the mitochondrial gene 12S rRNA and defined in Townsend (2014).
nuclear genes proopiomelanocortin A (POMC), recombi-
nation activating gene 1 (RAG-1) and rhodopsin (RHO),
and a more broadly sampled comparative dataset, to es- DNA extraction, PCR amplification,
timate the phylogenetic relationships of A. spinipollex and sequencing
sensu lato among; 3) evaluate the morphological system-
atics of populations assigned to A. spinipollex sensu lato, We selected loci for use in phylogenetic analyses to fa-
and 4) formally describe a critically endangered new spe- cilitate comparability of our dataset with previously
cies endemic to the environs of RVS Texiguat, Honduras, published studies (Duellman et al., 2016; Faivovich
an imperiled endemism hotspot in need of active and im- et al., 2005, 2018), choosing to sequence fragments of
mediate conservation measures. the mitochondrial genes 12S rRNA and 16S rRNA, and
the nuclear protein-coding genes proopiomelanocortin
A (POMC), recombination activating gene 1 (RAG-1)
and rhodopsin (RHO). A 16S-only dataset containing all
Materials and methods available samples of Atlantihyla, Duellmanohyla, and
Ptychohyla was used for initial analyses and taxonomic
assignment and identification of redundant mitochondrial
Taxon Sampling haplotypes. Based on prior evidence from the more vari-
able barcode marker COI (Townsend & Wilson, 2016),
Specimens and associated tissue samples used in this we selected the more conservative 16S for use in pre-
study were collected during 2008 – 2018, with fieldwork liminary species delimitation, a marker which has been
completed under a series of research permits issued to much more broadly sampled for Mesoamerican hylids
JHT by the Instituto Nacional de Conservación y Desar- (Faivovich et al., 2005, 2018). Subsequently, a subset of
rollo Forestal, Áreas Protegidas y Vida Silvestre [ICF] representative samples for each taxon was sequenced for
(Resolución GG-MP-055-2006 and Dictamen DAPVS 12S, POMC, RAG-1, and RHO, for inferring the evolu-
0091-2006; Resolución DE-MP-086-2010 and Dicta- tionary relationships of A. spinipollex sensu lato. Tem-
men DVS-ICF-045-2010; Resolución DE-MP-095-2014 plate DNA was extracted from muscle tissue using the
and Dictamen ICF-DVS-112-2014; Resolution DE- Qiagen PureGene DNA Isolation Kit (Valencia, Califor-
MP-065-2018 and Dictamen DVS-ICF-008-2018). Re- nia, USA) following manufacturer’s instructions. Frag-
search involving amphibians was approved under a series ments of the mitochondrial gene 12S were amplified
of protocols from the University of Florida IFAS-Animal using the primers 12SJ-L and 12SK-H (Goebel et al.,
Research Committee (ARC) and Indiana University of 1999) and 16S using 16Sar-L and 16Sbr-H (Palumbi et
Pennsylvania Institutional Animal Use and Care Commit- al., 1991); fragments of the nuclear protein-coding genes
tee (IACUC). Field sampling was primarily carried out at POMC were amplified using POMC-3 and POMC-4
night, with searches focused on stream-associated micro- (Wiens et al., 2005), RAG-1 using R1-GFF and R1-GFR
habitats. Voucher specimens collected were preserved in (Faivovich et al., 2005), and RHO using Rhod1A and
10% formalin and subsequently transferred to 70% etha- Rhod1C (Bossuyt & Milinkovitch, 2000). Amplification
nol for long-term storage. Vouchers were deposited in was carried out in 25µL-volume reactions with the fol-
the Carnegie Museum of Natural History (CM), Florida lowing thermocycling parameters: 12S: 2 min at 94°C,
Museum of Natural History (UF), National Museum of 45s at 50°C, and 2 min at 72°C, followed by 39 cycles
Natural History, Smithsonian Institution (USNM), Mu- of: 94°C for 30s, 50°C for 45s, and 72°C for 90s, with a
seo de Historia Natural, Universidad Nacional Autónoma final extension of 10 min at 72°C; 16S: 3min at 94°C, fol-
de Honduras, Tegucigalpa (UNAH), and Museo de His- lowed by 35 cycles of 45s at 94°C, 45s at 50°C, and 45s
toria Natural, Universidad Nacional Autónoma de Hon- at 72°C, with a final extension of 5 min at 72°C; POMC:
duras en el Valle de Sula (UVS); these and other insti- 2 min at 94°C, followed by 45 cycles of: 94°C for 30s,
tutional abbreviations follow those standardized by the 56°C for 30s, and 72°C for 1 min, with a final extension
733
of 7 min at 72°C; RAG-1: 2 min at 94°C, followed by 45 set of models, and evaluating the best model using
cycles of 30s at 94°C, 30s at 56°C, and 60s at 72°C, with Akaike Information Criterion scores: GTR + I + G was
a final extension of 7 min at 72°C; RHO: 5 min at 95°C, selected for 12S, SYM + I + G for 16S; GTR for POMC
followed by 40 cycles of 30s at 95°C, 40s at 60°C, and codon 1, HKY for POMC codon 2, and GTR for POMC
90s at 72°C, with a final extension of 5 min at 72°C. PCR codon3; JC + I for RAG-1 codon 1, HKY for RAG-1 co-
products were verified using electrophoresis on a 1.5% don 2, HKY + G for RAG-1 codon 3; HKY for RHO co-
agarose gel stained with ethidium bromide. Unincorpo- don 1, JC for RHO codon 2, and F81 for RHO codon 3.
rated nucleotides were removed from PCR products us- Bayesian inference was performed using mrbayes 3.2.6
ing 1 – 2 uL of ExoSAP-IT (USB, Santa Clara, CA, USA) (Huelsenbeck & Ronquist, 2001; Ronquist & Huelsen-
per 10uL of PCR product. Each locus and sample was beck, 2003), and consisted of two parallel runs of four
sequenced for both directions; some 16S sequence data Markov chains (three heated, one cold) run for 10 × 106
were obtained at the Smithsonian Institution Laboratory generations and sampled every 1,000 generations, with
of Analytical Biology (Suitland, Maryland, USA), and a random starting tree and the first 4 × 106 generations
remaining 16S, 12S, POMC, RAG-1, and RHO sequence discarded as burnin.
data were prepared in the Townsend Lab at Indiana Uni- We applied two methods to the single-locus 16S data-
versity of Pennsylvania and sequenced at Eurofins MWG set to aid in species delimitation: Automated Barcode Gap
Operon (Louisville, Kentucky, USA). We obtained refer Discovery (ABGD; Puillandre et al., 2012), a distance-
ence sequences from GenBank (https://www.ncbi.nlm. based approach; and a single threshold general mixed
nih.gov/genbank) and the alignment provided in the Yule coalescent model (GMYC; Pons et al., 2006; Fuji-
supplemental materials of Faivovich et al. (2018). All sawa & Barraclough, 2013), a tree-based method. The
novel sequences generated for this study were deposited ABGD analysis was conducted using the Kimura (K80)
in GenBank (accession numbers MK176937 – 177076, TS/TV substitution model, intraspecific divergence val-
MH177722 – 177747); details of samples used in phylo- ues left at default, 0.001 and 0.1, number of steps equal
genetic analyses are provided in Appendix 1; alignments to 15, and a relative gap width equal to 1.0, all of which
used in phylogenetic analyses are provided in Supporting was implemented on the ABGD web server (http://www
Information. abi.snv.jussieu.fr/public/abgd/abgdweb.html). We then
applied two evaluations of the ABGD results: one using
a conservative assessment (= ABGD-CA) with the initial
Molecular systematics consistent partition as a baseline, to represent the few-
est possible species; and a second moderate assessment
We aligned sequences using clustalw (Thompson et al., (= ABGD-MA) using the recursive partition to repre-
1994) as implemented in the program mega7 (Kumar et sent the most frequently occurring partition and a higher
al., 2016). Maximum likelihood (ML) and Bayesian In- number of species groupings. A single threshold GMYC
ference (BI) phylogenetic analyses were performed on analysis was performed on the Exelixis Lab GMYC
16S-only and mtDNA+nDNA datasets. Analyses of the web server (https://species.h-its.org/gmyc), which im-
16S-only dataset used a final alignment of 134 sequenc- plements the original R code of the GMYC model on a
es with a maximum length of 582 bp, consisting of 114 backend server. The input ultrametric tree was generated
novel sequences of Atlantihyla spinipollex (n = 14), A. sp. using beast2 (Bouckaert et al., 2014) implementing the
nov. (n = 24), Duellmanohyla salvavida (n = 19), D. sora Yule speciation model, strict molecular clock, and the re-
lia (n = 1), P. euthysanota (n = 1), and P. hypomykter versible jump (RB) substitution model. A chain length
(n = 55), along with 20 reference sequences (Appendix of 106, while logging parameter values and trees every
1). For analyses of the concatenated mtDNA + nDNA 1,000 generations, yielded ESS values greater than 200
(12S, 16S, POMC, RAG-1, RHO) dataset, we used a fi- for all parameters.
nal alignment of 34 sequences with a maximum length
of 2,628 bp, consisting of 13 novel samples of A. spini-
pollex (n = 2), A. sp. nov. (n = 4), Duellmanohyla salvavi- Morphological systematics
da (n = 2), D. soralia (n = 1), D. salvadorensis (n = 2), and
P. hypomykter (n = 2), along with 21 reference sequences We examined 76 preserved adults and three metamorphs
(Appendix 1). of Atlantihyla spinipollex sensu lato, representing the
Maximum likelihood was performed in raxml five known populations: RVS Texiguat (23; Dept. Atlán
v8.2.11 (Stamatakis, 2014), consisting of 2,000 pseudor- tida: La Liberación, 1030 m a.s.l.: USNM 578651, 578
eplicates using the ML + rapid bootstrap algorithm under 653, 578665 – 75, 578679 – 81, 578685 – 88; Dept. At-
the GTR + GAMMA + I model of nucleotide substitution, lántida: Cerro El Chino, 1390 – 1430 m a.sensu lato: US
with the dataset partitioned by gene (12S and 16S) and NM 578682 – 83; Dept. Yoro: 2.5 km NNE of La Fortu-
by codon (POMC, RAG-1, RHO). Prior to Bayesian phy- na, 1550 m), PN Pico Bonito (41; Dept. Atlántida: Que-
logenetic analysis, we used partitionfinder v1.1.1 (Lan- brada de Oro, 800 – 1210 m a.s.l.: USNM 514351 – 56,
fear et al., 2012) to select optimal substitution models 514358 – 92, 578664), PN Nombre de Dios (1; Dept. At-
for each gene (12S, 16S) and each codon (POMC, RAG- lántida: trail above Roma, 890 m), PN Capiro y Calentu-
1, RHO), using the “greedy” algorithm and “mrbayes” ra (3 metamorphs; USNM 514394 – 96), and Cerro Corre
734
Fig. 1. The known distribution of focal taxa in the Chortís Block. Filled shapes indicate localities with molecular (at least 16S) sequence
data generated or utilized herein; additional historical records (with no sequence data) are indicated by white shapes. Numbers and gray-
shaded areas refer to relevant localities within the Cordillera Nombre de Dios as follows: 1 = Refugio de Vida Silvestre Texiguat; 2 = Parque
Nacional Pico Bonito; 3 = Parque Nacional Nombre de Dios; 4 = Cerro Corre Viento; 5 = Parque Nacional Capiro y Calentura. TL = type
locality.
Viento (12; Dept. Colón: Cabaceras Río Manatí Creek, inter-orbital distance (IOD; measured at midlength of up-
930 – 1120 m a.s.l.: CM 170519 – 30). per eyelids), tympanum length (TPL), eye length (EL;
Examination and image capture associated with meas- horizontal length of eye at widest point, measured inside
urements, counts of nuptial spines, and other characteri- margin of eyelid), and third finger disc width (DW; meas-
zations were completed using a stereomicroscope and ured ventrally at widest point between outer edges of disc
AMScope MU900 microscope eyepiece camera with covers). Differences between the observed mean values
the software ToupView. Color names and numbers used for measurements from each taxon and sex were calcu-
in the description follow Köhler (2012). Morphologi- lated with a comparison of means test, which calculates
cal measurements were taken using a digital caliper and a significance value (p) representing the probability of
rounded to the nearest 0.1 mm. The following measure- obtaining the observed means if the difference between
ments were taken of each specimen: snout-vent length them is zero.
(SVL), shank length (SHL; hind limb segment between We performed Principal Component Analyses (PCA)
the knee and the heel), foot length (FL; distance from on morphological measurements in R 3.4.1 (R Core Team,
posterior-most portion of inner metatarsal tubercle to tip 2017) to determine if variation in morphology could be
of longest toe), head length (HL; tip of snout to posterior attributed to the genetic lineages. Forty-one adult males
angle of jaw), head width (HW; greatest width), width of and 35 adult females were included in these analyses,
upper eyelid (EW; perpendicular to outer edge of eyelid), and all analyses were conducted on males and females
735
separately to account for sexual size dimorphism. Based Results

on the lineage recovered in our phylogenetic analyses,
specimens from Corre Viento and Pico Bonito were
considered one taxon (A. spinipollex sensu stricto) and Phylogenetic analyses
specimens from the divergent Texiguat lineage as anoth-
er (A. sp. nov.). We first regressed measurements against Analyses of the 16S-only dataset using ABGD (CA and
SVL to account for variation in body size. The data were MA) and GMYC yielded somewhat conflicting results
then scaled to their standard deviation in order to nor- for some taxa, with, perhaps predictably, ABGD-CA
malize their distribution. PCAs were carried out using failing to delimit some closely related nominal species,
‘prcomp’ command in R. The packages ggfortify (Tang and ABGD-MA and GMYC recovering species limits
et al., 2016) and ggplot2 (Wickham, 2016) were used to consistent with the existing taxonomy and delimiting
plot results of analyses. Principle component scores are divergent lineages within recognized taxa into candidate
provided in Appendix 2. species (Fig. 2). All three analyses unambiguously sup-
port delimiting Atlantihyla spinipollex into two or more
species, with GMYC further delineating eastern and cen-
Bioacoustic processing and analysis tral A. spinipollex sensu stricto (as indicated in Fig. 2)
into two species, and ABGD-MA splitting A. spinipollex
An audio recording of the advertisement call for a single sensu lato into five species (Fig. 2). Support for the de-
adult male Atlantihyla sp. nov. (UVS V671) collected limitation of A. panchoi, Duellmanohyla salvavida and
from the Río Jilamito, Refugio de Vida Silvestre Tex- D. soralia was unambiguous across all three analyses.
iguat, Departamento de Atlántida, was recorded at a Within the genus Ptychohyla, the most conservative anal-
laboratory of the Museo de Historia Natural at the Uni- ysis (ABGD-CA) only recognized three species, failing
versidad Nacional Autónoma de Honduras en el Valle to delimit P. euthysanota, P. hypomykter sensu stricto,
de Sula (UNAH-VS) on 7 March 2018, at 1635h, as and P. macrotypanum, or P. leonhardschultzei, P. zopho
it vocalized inside a large terrarium approximately 24 des, and P. sp. (an undescribed species from Guatemala
hours after capture. A video recording of vocalizations first identified in Faivovich et al., 2005; Fig. 2). All three
was taken of an uncollected adult male A. spinipollex analyses recovered P. cf. hypomykter (from central Gua-
sensu stricto from Cerro Corre Viento, Departamento de temala) as a distinct species from P. hypomykter sensu
Colón, on 14 January 2019 at 2239h, as it called from stricto from the Chortís Block (Fig. 2).
among a group of 10+ other males around a pool at the The analysis of the mtDNA+nDNA dataset recovered
base of a ca. 8m tall waterfall. The audio component of a well-supported Atlantihyla clade containing A. pan-
the video recording (M2TS format; 0:51 in total length; choi and A. spinipollex sensu lato as sister to Ptychohyla
29.97 frames/s; audio recorded at 192kbps, sampling (Fig. 3). The majority of our species-level relationships
rate 48,000 Hz) was extracted to an AIF format using were otherwise congruent with those of Faivovich et
Adobe Premiere Pro CC (Adobe Systems Incorporated). al. (2018), which used broader taxonomic and genetic
Temperature data from the time of recording were not sampling, and largely well-supported (PP > 0.99 for all
available. Vocalizations were analyzed using Raven Pro but one node). A notable exception is D. salvadorensis,
1.5 version 64-bit for Windows (Bioacoustics Research which is an unresolved and weakly supported sister to
Program, 2014), at a sampling rate of 48,000 Hz and a the Atlantihyla-Ptychohyla clade. This likely an artifact
resolution of 16 bits. Spectral parameters for calls and of the lack of sampling from the southern clade of Du-
notes were obtained using a Blackman window, DFT ellmanohyla in our phylogeny, and that the phylogenetic
size at 1024 samples, 3 db filter bandwidth 76.9 Hz. relationships and taxonomic assignment of this species
Based on the characteristics of the vocalizations, we has long been problematic, as evidenced by Faivovich
used a note-centered approach to define our termino et al. (2018) devoting a figure and nearly two pages of
logy for calls, notes, pulses, and their associated para text to discussion of this species. Across all analyses, two
meters (Köhler et al., 2017). We recorded the follow- deeply divergent, reciprocally monophyletic lineages
ing spectral and temporal parameters: call duration, note were recovered from samples assigned to A. spinipollex
duration, number of notes per call, number of pulses per from northern Honduras. The first lineage contained
note, pulse duration, inter-pulse interval, pulse repeti- samples from PN Pico Bonito (Dept. Atlántida; near the
tion rate (pulse/second), and dominant frequency (Hz) vicinity of the type locality), PN Nombre de Dios (Dept.
of the call. Given the limitations of the available re- Atlántida), and Cerro Corre Viento (Dept. Colón). Given
cordings and the lack of associated temperature data for that this lineage contains material collected near the type
A. spinipollex sensu stricto, we limited our direct com- locality of A. spinipollex, we consider these samples to
parisons to note duration, pulse count, and dominant fre- represent A. spinipollex sensu stricto. The second lineage
quency, three traits with low intraspecific variability that consisted of samples from RVS Texiguat (Depts. Atlánti-
are minimally influenced by temperature (Köhler et al., da and Yoro), which we refer to as A. sp. nov. Mitochon-
2017). Thus, the resulting comparisons were qualitative drial pairwise distances were relatively high between
and not meant to characterize the full range of variation these lineages and demonstrated a wide gap between
in vocalizations for either population. intraspecific and interspecific distances, ranging from
736
Fig. 2. Bayesian ultrametric tree based on 16S mitochondrial dataset, showing inferred species limits from tree-based (GMYC) and dis-
tance-based (ABGD-MA and ABGD-CA thresholds) analyses. Bayesian posterior probabilities (≥0.9) for nodes corresponding to lineages
delimited in the analyses are shown above corresponding branches. Geographic haplogroups of Ptychohyla hypomykter labeled as follows:
We = Western, Ce = Central, and ES = Eastern-Southern. Photo of paratype (USNM 578672) of A. melissa (© JHT).
737
Fig. 3. Multilocus Bayesian phylogram based on combined mtDNA (12S, 16S) + nDNA (POMC, RAG-1, RHO) dataset; maximum likeli-
hood bootstrap support values (0 – 100) and Bayesian posterior probabilities (0 – 1.0) shown above corresponding branches. Photos from
top to bottom: D. salvavida from Río Jilamito, Dept. Atlántida, Honduras © JHT; D. soralia from Merendón, Dept. Cortés, Honduras ©
JHT; holotype of A. melissa © JHT; A. spinipollex from Cerro Corre Viento, Dept. Colón, Honduras © Jason M. Butler; Ptychohyla hy-
pomykter from Rio Negro de Comayagua, Dept. Comayagua, Honduras (© JHT).
3.8 – 5.2% for 12S (intralineage distances: A. spinipollex t-stat –3.683, p < 0.001; females t-stat = –4.666, p <

s.s 0.0 – 0.7%; A. sp. nov. 0.1%) and 4.1 – 5.0% for 16S 0.0001), relatively longer hind feet (FL/SVL; males 
(A. spinipollex s.s 0.0 – 1.2%; A. sp. nov. 0.0 – 0.4%). t-stat = –3.536, p = 0.0011; females t-stat = –3.056, p <
0.005), and heads that are longer (HL/SVL; males t-stat =
–6.081, p < 0.0001; females t-stat = –7.068, p < 0.0001)
Morphological variation and broader (HW/SVL; males t-stat = –5.902, p < 0.0001,
DF = 39; females t-stat = –5.988, p < 0.0001).
We examined 76 adult specimens and three metamorphs Principal component analyses recovered individuals
of Atlantihyla spinipollex sensu lato, representing the from A. spinipollex sensu stricto and samples from RVS
majority of existing museum specimens and including Texiguat as distinct clusters that overlapped somewhat in
the entire distribution of A. spinipollex: nine males and morphospace (Fig. 5); PCA scores are available in Ap-
15 females from RVS Texiguat, 22 males and 19 fe- pendix 3. The first six principal components explained at
males from PN Pico Bonito, 10 males and one female least 95% of the cumulative variance in both male and fe-
from Cerro Corre Viento, and three metamorphs from PN males. In males, HW, HL, and SHL (all regressed against
Capiro y Calentura (Appendix 2). Based on our phylo SVL) had the strongest positive loadings in PC1 (0.377,
genetic results, we considered the PN Pico Bonito, PN 0.370, 0.362, respectively), which accounted for 66.0%
Nombre de Dios, and Cerro Corre Viento populations to of the variation. In females, the same measurements had
represent a single group, A. spinipollex sensu stricto, and strong negative loadings in PC1 (–0.368, –0.370, –0.360,
compared those samples to samples from RVS Texiguat respectively), which accounted for 66.7% of the vari-
(Table 1, Fig. 4). Both male and female frogs from RVS ation. DW/SVL had strong negative loadings in PC2 in
Texiguat were significantly smaller than A. spinipollex both males (–0.772) and females (–0.724). Additionally, in
sensu stricto (SVL; males: mean difference = 3.05, sd = PC2, males exhibited strong negative loadings in EW/SVL
0.88, t-stat = 3.451, p = 0.0014, DF = 39; females: mean dif (–0.571), while females exhibited strong negative loadings
ference = 5.62, sd = 1.61, t-stat = 3.495, p = 0.0014, DF = in FL/SVL (–0.501). PC2 accounted for 10.9% of the vari-
33), had relatively longer shanks (SHL/SVL; males = ation in males, and 15.5% of the variation in females.
738
Table 1. Comparison of morphological variation between Atlantihyla melissa and A. spinipollex; range followed by mean and standard
error in parentheses.
Atlantihyla spinipollex Atlantihyla melissa

Males (n = 32) Females (n = 20) Males (n = 9) Females (n = 15)
32.9 – 40.5 30.9 – 45.8 30.4 – 38.8 30.9 – 43.9
SVL
(36.24, 2.07) (41.17, 4.57) (33.19, 3.19) (35.55, 4.89)
0.474 – 0.600 0.486 – 0.568 0.521 – 0.666 0.497 – 0.727
SHL/SVL
(0.520, 0.0276) (0.520, 0.0283) (0.564, 0.0440) (0.594, 0.0632)
0.368 – 0.462 0.360 – 0.486 0.425 – 0.568 0.408 – 0.543
FL/SVL
(0.417, 0.0281) (0.418, 0.0323) (0.459, 0.0421) (0.457, 0.0433)
0.312 – 0.392 0.325 – 0.381 0.346 – 0.427 0.367 – 0.477
HL/SVL
(0.348, 0.0187) (0.349, 0.0173) (0.401, 0.0353) (0.409, 0.0324)
0.309 – 0.364 0.313 – 0.370 0.327 – 0.442 0.345 – 0.459
HW/SVL
(0.327, 0.0139) (0.333, 0.0163) (0.370, 0.0327) (0.382, 0.0315)
0.555 – 0.933 0.562 – 0.891 0.576 – 0.745 0.064 – 0.832
EW/IOD
(0.717, 0.1002) (0.730, 0.0868) (0.651, 0.0510) (0.750, 0.0624)
0.375 – 0.521 0.387 – 0.526 0.433 – 0.520 0.425 – 0.554
TPL/EL
(0.460, 0.0485) (0.461, 0.0374) (0.474, 0.0304) (0.486,0.0361)
0.886 – 1.251 0.833 – 1.284 0.915 – 1.246 0.938 – 1.256
IOD/EL
(1.079, 0.1017) (1.048, 0.1209) (1.054, 0.1102) (1.115, 0.0947)
Nuptial 29 – 62 39 – 53
— —
spines (46.0, 9.492; n = 23) (45.75, 4.496; n = 8)
Fig. 4. Direct comparisons of snout-vent length (SVL) and four morphometric ratios between female and male A. spinipollex and A. melis-
sa sp. nov. SHL = shank length; FL = foot length; HL = head length; HW = head width. Asterisks indicate significant comparisons: ** p <
0.001; ** p < 0.0001.
Comparative bioacoustics a dominant frequency of 3421.9 Hz, with each note hav-
ing a duration of 0.47 – 1.3 s (n = 16, x = 0.913, sd = 0.21)
A recording from a single male Atlantihyla sp. nov. from and consisting of 5 – 12 (x = 8.21, sd = 2.16) pulses per
Río Jilamito consisted of a series of four calls, each con- note (Fig. 6). A recording from an intermittently calling
sisting of series of 5 – 6 evenly-spaced pulsed notes with male A. spinipollex sensu stricto from Corre Viento con-
739
Systematic description
The results of our molecular and morphological analyses,

supplemented by bioacoustic and biogeographic data,
provide unambiguous support for recognition of the RVS
Texiguat population as a distinct species, which we name
Atlantihyla melissa sp. nov.

ZooBank urn:lsid:zoobank.org:act:A63E72DA-E7B3-4B96-
9A6D-FAD52411F624
Figs. 7 – 8, 11 – 13
Ptychohyla spinipollex: McCranie et al., 1993: 1060.

Ptychohyla spinipollex (in part): McCranie & Wilson, 1993: 100.
Ptychohyla spinnipolex (sic, in part): McCranie & Solís, 2013:
242.
Atlantihyla spinipollex (in part): Faivovich et al., 2018: 24.
Holotype. USNM 578679 (field number JHT 3114; Figs. 7, 8),

an adult male from alongside the Río Jilamito at La Liberación
(15.5302°N, 87.2939°W, WGS84; Figs. 9, 10a-c), 1,030 m, Refu-
gio de Vida Silvestre Texiguat, Departamento de Atlántida, Hon-
duras; collected 14 June 2010 by Benjamin K. Atkinson, Cesar
A. Cerrato, Luis A. Herrera-B., Mayron McKewy Mejía, Josiah H.
Townsend, and Larry David Wilson. GenBank accession numbers:
12S (MK176937), 16S (MK176946), POMC (MK177061), RAG1
(MW177730), RHO (MK177068).
Paratypes. Twenty-one (21) paratopotypes, including eight adult
males (USNM 578670, 578674 [Fig. 11e], 578675, 578677,
578680 – 81, 578685, 578687) and 13 adult females (USNM
578665 [Fig. 11c], 578666 – 67, 578668 [Fig. 11f], 578669, 578671,
578672 [Fig. 11b], 578673, 578676 [Fig. 11a], 578678, 578682,
Fig. 5. The first two principal components of morphological char- 578686 [Fig. 11d], 578688), collected between 1,015 – 1,090 m el-
acters for males (top) and females (bottom), which accounted for evation in the vicinity of the Río Jilamito at La Liberación, Refugio
de Vida Silvestre Texiguat, Departamento de Atlántida, Honduras;
the majority of variation in the analyses. 10 – 18 June 2010 and 25 – 27 June 2010. Two (2) adult female
paratypes (USNM 578683 – 84) from Cerro El Chino (15.5225°N,
87.2802°W), a peak to the ESE of La Liberación, 1,420 – 1,430 m
sisted of a series of five intermittent pulsed notes with a elevation, Departamento de Atlántida, Honduras, collected 19 June
dominant frequency of 3328.1 Hz, and each note hav- 2010. One (1) adult female paratype (UF 166704) from 2.5 airline
km NNE of La Fortuna (15.4328°N, 87.3093°W; Fig. 10d, 10e),
ing a duration of 0.606 – 0.752 s (x = 0.669, sd = 0.079) 1,610 m elevation, Departamento de Yoro, Honduras, collected 9
and consisting of 6 – 7 pulses (x = 6.4, sd = 0.577). It is April 2008 by Jason M. Butler, John Slapcinsky, Nathaniel Stewart,
not apparent if these vocalizations are representative of Josiah H. Townsend, and Larry David Wilson.
a complete advertisement call for A. spinipollex sensu
stricto, as they were intermittent and do not include notes Referred specimens. Twenty-seven (27) specimens not
in a series forming a call sequence, as seen in A. sp. nov. included in our molecular or morphological analyses are
Due to this difference and the lack of directly comparable referred to this species, including one (1) specimen from
call sequence for A. spinipollex sensu stricto, we limit the 3.8 km east and 1.5 km south of Cerro Cabeza de Negro
conclusions drawn by direct comparison. (UNAH 2708), five (5) specimens from the north slope
Based on these limited samples, we found that A. sp. of Cerro Texiguat (UNAH 1470, 1472, 1493, 1893 – 94),
nov. had a higher dominant frequency than A. spinipol and 14 specimens from Río Jilamito at La Liberación
lex sensu stricto (3421.9 vs. 3328.1 Hz) and the aver (UVS V669 – V682), all localities in Departamento de
age duration of notes was significantly longer than those Atlántida, Honduras; and 12 specimens from approxi-
of A. spinipollex sensu stricto (p = 0.021, t-statistic = mately 2.5 airline km north-northeast of La Fortuna (UF
– 2.505, DF = 20). The pulsed note structures in both 142417 – 21, UNAH 3738, USNM 514397 – 402), Depar-
samples were similar, with notes of A. sp. nov. consist- tamento de Yoro, Honduras.
ing of relatively more pulses (5 – 12, x = 8.21, sd = 2.16)
than those of A. spinipollex sensu stricto (6 – 7, x = 6.4, Diagnosis. The new species is a member of the genus
sd = 0.577), although the means are not significantly dif- Atlantihyla, based on results from the mtDNA+ nDNA
ferent (p = 0.0541, t-statistic =  – 2.04, DF = 21). phylogenetic analyses (Fig. 3) and by possession of a
740
Fig. 6. Advertisement call of an adult male Atlantihyla melissa (UVS V671): (a) waveform (top) and spectrogram (bottom) of a call series
(60 seconds); (b) waveform (top) and spectrogram (bottom) of a single call (15 seconds; third call in series above); and (c) waveform (top)
and spectrogram (bottom) of a single note (1.9 seconds; second note in call above).
741
vertical rostral keel, as well as enlarged dark-colored 0.352; eyelid width 3.0 mm; eye length 4.62 mm; tym-
spine-shaped epidermal projections on the prepollex and a panum diameter 2.1 mm; tympanum/eye diameter 0.446;
pale ventrolateral stripe in males (Faivovich et al., 2018). inter-orbital distance 4.59 mm; inter-orbital distance/
Atlantihyla melissa can be distinguished from its sister eye length 0.994. Snout truncate in lateral view, rounded
species, A. spinipollex (Fig. 4), by having a relatively in dorsal view; weak vertical rostral ridge present; can-
smaller adult size (adult males 30.4 – 38.8 mm [33.19 ± thus distinct and rounded; loreal region concave; nostrils
3.186] and adult females 30.9 – 43.9 mm [35.55 ± 4.89], slightly protuberant; internarial distance 3.2 mm; top of
versus 32.9 – 40.5 mm [36.24 ± 2.07] and 30.9 – 45.8 mm head flat; supratympanic fold well developed, extending
[41.14 ± 4.57] in A. spinipollex), having relatively longer from point 0.4 mm posterior of the eye above tympa-
shanks, a relatively larger head, significantly longer note num to point above insertion of forearm, contacting to
duration in the advertisement call, and relatively more slightly covering upper edge of tympanum; tympanum
pulses per note. The only other member of the genus, distinct, rounded. Pupil horizontally elliptical. Forearm
A. panchoi, differs from A. melissa in having a well-de- somewhat more robust than upper arm, with row of un-
fined white-lined upper lip (upper lip coloration consist- equally elevated ulnar tubercles forming a ridge on the
ent with adjacent areas or slightly paler in A. melissa). ventral portion of the forearm. Prepollex ossified and
Atlantihyla melissa is also well differentiated from other bluntly rounded; finger I of the right hand has 39 dark
taxa on the basis of DNA sequence variation. From its colored, spine-shaped papillary epidermal projections
sister species and closest known relative, A. spinipollex, that are relatively large, well-defined, and juxtaposed
the new species has sequence divergences of 3.8 – 5.2% to one another in a bell-shaped patch, located on the in-
for 12S and 4.1 – 5.0% for 16S, and from A. panchoi, di- terior and ventral surfaces of the prepollical area of the
vergences ranged from 3.8 – 4.2% for 12S and 5.5 – 5.7% digit; subarticular tubercles rounded and globular, distal
for 16S. Townsend and Wilson (2016) also reported COI tubercles on finger III weakly divided and finger IV bifid;
sequence divergence between A. melissa (as Ptychohyla discs on fingers large, discs of fingers II and III slightly
sp.) and A. spinipollex ranging from 14.1 – 16.5%. The smaller than diameter of the tympanum; relative length
distribution of A. melissa is allopatric with respect to of fingers I<II<IV<III; basal webbing between fingers I
A. panchoi and A. spinipollex. and II, webbing formula II 1¾ – 2 ¾ III 2+ – 2 IV; un-
Atlantihyla melissa is known to occur in sympatry or webbed portions of fingers with lateral keels. Heels over-
near-sympatry with four other species of stream-associ- lap when hind limbs adpressed at right angle to body;
ated hylids: Duellmanohyla salvavida, Isthmohyla insol- inner tarsal dermal fold present but weakly defined; outer
ita, Plectrohyla chrysopleura, and P. cf. guatemalensis; tarsal fold absent; heel tubercle present; subarticular tu-
and can be distinguished from D. salvavida in having a bercles on toes rounded, largest on toes I and V; relative
dorsal coloration that can be uniform grayish brown or toe length I<II<III≈V<IV; webbing formula I 1+ – 2- II
with varying degrees of light or dark spots or mottling, 1+ – 2 III 1+ – 2 IV 2 – 1+ V; unwebbed portions of fingers
and large nuptial spines in adult males (dorsum uniform with lateral keels; toes moderately long; toe discs slightly
green or occasionally mottled green with tiny nuptial smaller to about same size as finger discs. Cloacal open-
excrescences in D. salvavida), from I. insolita in having ing directed posteroventrally at midlevel of thighs, with
a relatively smooth dorsal surface, patternless or lightly skin below cloaca featuring several folds radiating out
mottled chin coloration, and large nuptial spines in adult from cloacal opening. Skin on dorsum smooth; skin on
males (dorsal surface tuberculate, chin with well-defined throat, chest, belly and ventral surface of thighs weakly
pale central blotch resembling a broad stripe, and tiny granular, with some weak tubercles present on dorsal
nuptial excrescences in I. insolita), and from P. chrys- surfaces of eyelids. Ventrolateral glands absent. Tongue
opleura and P. cf. guatemalensis in having a prepollex ovoid with faint indication of posterior notch; vomerine
that is only slightly enlarged and lacks a protruding distal teeth, situated on a pair of small, elliptical elevations be-
end (prepollex enlarged and flat, with blunt distal end in tween ovoid internal nares, with three vomerine teeth on
P. chrysopleura, prepollex with two large curved spines left side and two on right side. Vocal slits present on both
that protrude from the skin on the distal end in P. cf. gua- sides, vocal sac single, median, subgular.
temalensis). Although allopatric in distribution, Ptycho-
hyla hypomykter is also known from localities in the De- Measurements of holotype (in mm). SVL 34.44, SHL
partment of Yoro, Honduras, and differs from A. melissa 18.72, FL 15.41, HL 13.22, HW 12.12, EW 3.01, EL
in lacking a pale dorsolateral stripe in adult males and in 4.62, IOD 4.59, TPL 2.06.
having a thick, rounded tarsal fold (irregular pale dorso-
lateral stripe, sometimes incomplete, and thin tarsal fold Colouration of holotype. Colour in life of the adult male
in A. melissa). holotype (USNM 578679, Fig. 7) is as follows: dorsum
of head, body, forelimbs, and hind limbs mottled Light
Description of holotype. An adult male (USNM 578679; Buff (2) and Chamois (84) with Brownish Olive (276)
Figs. 7, 8) having a snout-vent length of 34.4 mm, tibia reticulations and a large, irregular Sepia (286) blotch
length 18.7 mm; tibia length/SVL 0.544; foot length 15.4 covering most of the dorsal surface of the head and
mm; foot length/SVL 0.447; head length 13.2 mm; head roughly one half of the body, resembling an ink drop;
length/SVL 0.384; head width 12.1 mm; head width/SVL upper lip mottled Chamois (84), Flesh Color (249), and
742
Fig. 7. Adult male holotype of Atlantihyla melissa (USNM 578679; SVL = 34.4 mm), shown in life (left) and in dorsal (top right) and
ventral (bottom right) aspects after eight years in preservation (© JHT).
Variation. Atlantihyla melissa exhibits sexual dimor-

phism in terms of size and coloration. Females are larger
than males, with maximum recorded size for males 38.77
mm and females 43.96 mm; the ranges of variation in
measurements and relative ratios for males and females
are summarized in Table 1. All adult male specimens of
A. melissa examined were in breeding condition, with
patches of relatively large, dark-colored, spine-shaped
papillary epidermal projections that are juxtaposed to
one another in a roughly bell-shaped patch on the inte-
rior surface of each thumb. Nuptial spine counts from the
right thumbs of eight adult males number 39 – 53 (45.75
± 4.496).
Male and female A. melissa exhibit a wide range in
variation of coloration and pattern (Figs. 11, 12). Iris
color in males ranges from Salmon Color (58) to Flame
Scarlet (73); in females, iris color is more subdued, rang-
ing from Tawny Olive (17) to Cinnamon Drab (50).
There appears to be no discernable pattern of sexual di-
morphism in relation to dorsal and ventral coloration.
Variation in dorsal pattern ranges from nearly uniform in
Fig. 8. Semi-diagrammatical illustrations of the ventral aspects of coloration (Fig. 12b); extensively mottled (Fig. 11e, 11f,
the hand (left) and foot (right) of the holotype of Atlantihyla melissa 12d); small to medium-sized irregular dark spots (Fig.
(USNM 578679). 11c, 12a); small pale spots that appear glandular (Fig.
12c); large and irregular “ink spot” botching, similar to
that of the holotype (Fig. 12e, 12f); pale irregular blotch-
Vinaceous Pink (245), with no indication of labial stripe; es (Fig. 11d, 12g); or with a pronounced dark middorsal
ventrolateral stripe Pale Buff (1), well-defined but with stripe that can be narrow and well-defined (Fig. 11a, 12j)
somewhat irregular shape, extending from axilla to inser- or broader with irregular borders (Fig. 11b, 12h). Most,
tion of hindlimb, creating sharp distinction between dor- but not all, individuals have an indication of a pale ven-
sal and ventral coloration; ventrolateral stripe bordered trolateral stripe, which can be Pale Buff (1), Cream Yel-
ventrally with Pale Purple (223) and Dark Pinkish Rose low (82), or Spectrum Yellow (79). Ventral coloration of
(220), with scattered Pinkish White (216) granules; ven- paratypes generally falls into two categories: individuals
ter Pinkish White (216), ventral surfaces of limbs Light with venters similar to those of the holotype, featuring a
Lilac (221) to Lilac (222), Light Buff (2) stripe forming gray to purple coloration with purple adjacent to the ven-
a semicircle above and around the cloaca; iris Salmon trolateral stripe; and individuals with ventral coloration
Color (83) with some Sepia (286) reticulations. featuring pale to bright yellow in place of purple (Fig.
743
b
Fig. 9. a) Type locality of Atlantihyla melissa, the Río Jilamito, Departamento de Atlántida, Honduras, 1,030 m a.s.l.; holotype was col-
lected on a palm frond in the vegetation on the left side of the image. b) Adult male A. melissa (not collected) photographed in situ on
vegetation directly overhanging a small tributary of the Río Jilamito, 1,060 m a.s.l. (photographs © JHT).
11a). These yellow-phase individuals can have immacu- Description of tadpole. A tadpole (CM 163366; Fig.
late or Pale Buff venters, with ventral surfaces of front 13A) from the type locality, exhibiting Gosner (1960)
and hindlimbs being Pale Greenish Yellow (86), Sulphur development stage 36, is described as follows: TL 43.5
Yellow (80), Spectrum Yellow (79), or Dark Spectrum mm, BL 13.5 mm, BW 7.1 mm, EL 1.4 mm, IOD 4.1
Yellow (78). The chin and throats of most individuals are mm, IND, TAL 30.1, TMH 4.7 mm, TMW 3.6 mm; body
immaculate; however, occasional individuals have dark somewhat compressed, wider than high; snout broad and
spots that can be scattered or make the chin appear mot- rounded in dorsal aspect, rounded in lateral view; eyes
tled. positioned dorsolaterally, EL/BL 0.104; nostrils small,
744
directed anteriolaterally, and positioned slightly closer to Townsend, Medina-Flores, Wilson, Jadin & Austin, 2013,
eyes than tip of snout; spiracle single, sinistral, directed and the centipede snake Tantilla olympia Townsend, Wil-
posteriorly, opening on midline of body approximately son, Medina-Flores & Herrera-B., 2013; and two sala-
two-thirds distance from tip of snout to posterior end of manders endemic to the Department of Olancho: Nototri-
body; vent tube short, dextral; hind limb buds with sepa- ton mime Townsend, Medina-Flores, Reyes-Calderón &
ration between toes II – III, III – IV, and IV – V; tail more Austin, 2013, and N. picucha Townsend, Medina-Flores,
than twice length of body, TAL/TL 0.692; tail muscula- Murillo, and Austin, 2011. Melissa disappeared without
ture not extending to tip of tail, tail fins relatively shal- a trace on 5 November 2016, after becoming separated
low, tail musculature at midlength of tail slightly taller from her companions while descending from the summit
than height of dorsal or ventral tail fins at same point; of the highest peak in Honduras, Cerro de Las Minas in
oral disk large (Fig. 13b), ventral, with continuous border Parque Nacional Celaque. Despite over a month of con-
of 2 – 3 rows of marginal papillae, with 4 – 5 rows of sub- tinuous searching by military and volunteer rescue teams
marginal papillae present at level lateral to jaw sheath; la- supported by search dogs and aircraft, no evidence of
bial tooth rows 5 – 6/8 – 9, appearing somewhat irregular Melissa’s fate has been found.
with most rows exhibiting some degree of interruption or
fragmentation: row A-1 with four narrow interruptions, Distribution. This species is known to occur between
row A-2 with one narrow lateral interruption, A-3 with 780 – 1,680 m elevation in the western portion of the
two narrow lateral interruptions, A-4 with two narrow Cordillera Nombre de Dios, Departments of Atlántida
and one wide lateral interruptions, A-5 broadly interrupt- and Yoro, Honduras, from localities nearly completely
ed by A-6, which shift anteriorly to take position of A-5 contained within the boundaries of Refugio de Vida Sil-
while A-5 reappears medially in position of A-6, P-1 with vestre Texiguat. The distribution of A. melissa appears to
one narrow medial and one narrow lateral interruption, be centered on the Río Jilamito Valley, where the major-
P-4 with one narrow lateral interruption, P-5 appearing as ity of remaining habitat exists, and extends to the south
a partial row about one-fifth length of P-4, P-7 with two across the ridge and peaks associated with Cerro Tex-
lateral interruptions, P-8 and P-9 highly fragmented (Fig. iguat and into broadleaf riparian zones on the leeward
13b). Coloration in preservative: body mottled grayish- slopes of those peaks.
tan, tail musculature pale tan with brown blotches; tail
fins translucent with scattered dark flecking and a few Natural history. Atlantihyla melissa is found in intact
dark spots on the anterior portion of the dorsal fin. Broadleaf Transitional Cloud Forest and peripherally in
Broadleaf Cloud Forest (Townsend, 2014), correspond-
Advertisement call. The advertisement call of an adult ing to the Premontane Wet Forest formation and Lower
male (UVS V671) from the type locality consists of a Montane Wet Forest formations, respectively, of Hol-
series of 5 – 6 relatively long pulsed notes, presented in dridge (1967). This species occurs in association with
a continuous or nearly continuous call series, which in- fast-flowing rivers and streams with rocky substrates, a
creases in amplitude with each note and the final note of mix of pools and riffles, and abundant riparian vegeta-
each call being the strongest; sometimes followed by a tion (Figs. 9, 10). The locality in the Department of Yoro
single faint trailing note that can lead into the next call on the leeward side of RVS Texiguat (Fig. 10d – e) con-
(Fig. 6a). Spectrum visualization revealed two harmon- sists of a stream and associated broadleaf riparian veg-
ics present in each note (Fig. 6). Dominant frequency of etation flowing through a series of ravines, otherwise
the entire call series was 3421.9 Hz. Calls had the fol- surrounded by Ocotal (pine-oak forest) and Mixed Tran-
lowing quantitative spectral and temporal characteristics sitional Cloud Forest habitat (Townsend, 2014). Indi-
(n = 4): call duration 8.28 – 13.19 s (n = 4, x = 11.16 ± viduals were commonly encountered at night on vegeta-
2.15), 5 – 6 (x = 5.25 ± 0.5) notes per call, note duration near or overhanging the water (Figs. 9b, 12). Males
tion 0.47 – 1.3 s (n = 16, x = 0.913 ± 0.21), note repeti- were actively vocalizing in March and April, including
tion rate 0.405 – 0.604 calls/s (x = 0.481 ± 0.09), interval intermittently from concealed positions during daytime,
between notes 1.04 – 1.79 s (n = 14, x = 1.466 ± 0.211), with vocalizations also being present, but more inter-
5 – 12 (x = 8.21 ± 2.16) pulses per note, pulse duration mittent, during June and July. Females observed during
0.013 – 0.052 s (n = 153, x = 0.027 ± 0.004), pulse repeti- March had well-developed eggs visible in their oviducts.
tion rate 5.669 – 11.086 (x = 9.359 ± 1.375), peak ampli- Emerging and recently emerged metamorphs (Figs. 13c,
tude of each pulse 287 – 2982 U (x = 1288.29 ± 644.22), 13d) were abundant in June and July on rocks and low
inter-pulse interval 0.035 – 0.106 s (n = 134, x = 0.074 ± vegetation adjacent to the Río Jilamito and tributaries,
0.008). while tadpoles were abundant in pools.
At premontane elevations (780 – 1,430 m) in the vi-
Etymology. We name this species in honor of our friend cinity of the Río Jilamito and its tributaries, A. melissa
and collaborator, Isis Melissa Medina-Flores, a field bi- is part of a community of stream-associated anurans that
ologist originally from Mangulile in the Department of includes the endemic treefrogs Duellmanohyla salvavida
Olancho, Honduras. Melissa participated in the discov- and Plectrohyla chrysopleura, the endemic streamside
ery and description of this new species, two other Tex- frog Craugastor aurilegulus, and the glassfrog Terato-
iguat endemics: the palm-pitviper Bothriechis guifarroi hyla pulverata. Townsend et al. (2012: 100) provided a
745
b c
d e
Fig. 10. Habitat of A. melissa: (a) Large pool in the Río Jilamito, just above the type locality; tadpoles of A. melissa. Duellmanohyla
salvavida, and Plectrohyla chrysopleura were found here in April 2017, June 2010, and July 2010. (b) Small tributary of the Río Jilamito
above the type locality, 1,090 m a.s.l., adults and tadpoles of A. melissa and D. salvavida were abundant here in June and July 2010. (c)
Small tributary of the Río Jilamito directly adjacent to the type locality, 1,060 m a.s.l. adults, metamorphs, and tadpoles of A. melissa,
D. salvavida, and P. chrysopleura were abundant here in June and July 2010. (d) Stream habitat approximately 2.5 airline km north-
northeast of La Fortuna, Departamento de Yoro, where A. melissa has been collected alongside endemic anurans Craugastor saltuarius,
C. stadelmani, D. salvavida, Isthmohyla insolita, and Plectrohyla cf. guatemalensis, (e) Same stream as 10(d) showing extensive deforesta-
tion approximately 200 m upstream in April 2008 (photographs © JHT).
detailed overview of the herpetofaunal community pre- Río Jilamito and its tributaries at the site known as La
sent at the type locality of A. melissa, describing popu- Liberacíon. At higher elevations (1,550 – 1,680 m) on the
lations of this species as “extremely abundant” and the leeward slopes of RVS Texiguat, A. melissa is found in
“most commonly encountered amphibian” around the sympatry with the endemic stream-associated treefrogs
746
a b
c d
e f
Fig. 11. Paratypes of Atlantihyla melissa in life: (a) adult female (USNM 578676), (b) adult female (USNM 578672), (c) adult female (USNM
578665), (d) adult female (USNM 578686), (e) adult male (USNM 578674), (f) adult female (USNM 578668) (photographs © JHT).
D. salvavida, Isthmohyla insolita, and Plectrohyla cf. gua that A. spinipollex sensu stricto should remain Endan-
temalensis, and also been documented in near-sympatry gered (B1ab[iii] + 2ab[iii]), and that A. melissa should
with Atelophryniscus chrysophorus, C. saltuarius, and be considered Critically Endangered (B1ab[iii]) due to
C. stadelmani, three species of terrestrial streamside anu- the known distribution being limited to a single threat-
rans also considered endemic to the Cordillera Nombre defined area of highland forest < 25 km2 in total extent,
de Dios (McCranie & Wilson, 2002). which is undergoing a continuous decline in the extent
and quality of remaining habitat due to ongoing defor-
Conservation status. Prior to this study, Atlantihyla spi estation and development.
nipollex sensu lato was considered Endangered (B1ab[iii]
+ 2ab[iii]) on the IUCN Red List due to loss of habi- Remarks. Köhler (2011: 258 – 259) included photo
tat from subsistence agriculture and land-clearing for graphs of the holotype and two paratypes (USNM
timber and pastureland (Cruz et al., 2004). Applica- 578664, 578662) as P. (= A.) spinipollex, and Townsend
tion of the IUCN Red List criteria (IUCN 2012; IUCN et al. (2012: 101) also provided photographs of the hol-
Standards and Petitions Committee, 2019) indicates otype, and one paratype (USNM 578686), and one un-
747
a b c
d e f
g h j
Fig. 12. Variation in color pattern among unvouchered Atlantihyla melissa photographed in situ at night along the Río Jilamito and associ-
ated tributaries at the type locality (1,030 – 1,090 m a.s.l.; photographs © JHT).
b c d
Fig. 13. Semi-diagrammatical illustrations of the (a) larva of Atlantihyla melissa (CM 163366; Gosner stage 36, TL = 43.5 mm) and (b)
oral disk of the same larva, and images of two recently emerged metamorphs (not collected) of A. melissa (c, d) photographed in situ at
the type locality (photographs © JHT).
748
vouchered individual (Fig. 12j) as P. (= A.) spinipollex, Firneno et al., 2020). Clearly, improved genomic sam-
but incorrectly labeled the holotype as an adult female pling, together with more robust hypothesis testing, are
(rather than male) in Fig. 5d. The lowest recorded eleva- needed to better understand the complex evolutionary
tion for A. melissa (780 m) is based on an unvouchered history of these taxa.
individual encountered by JHT at a point where the trail
to La Liberación crossed a tributary of the Río Jilamito
on 13 March 2017. Patterns of diversification in
Ptychohyla hypomykter
In the first comprehensive review of the genus Ptycho-

Discussion hyla, Duellman (1963: 330) noted that specimens of
P. (=Atlantihyla) spinipollex from Honduras, including
the holotype, differed from populations in Guatemala
Biogeography and taxonomy of stream in ventral coloration, and Duellman (1970: 547) reiter-
treefrogs in the Chortís Highlands ated observations of distinctiveness between samples
from Honduras and those from Guatemala, suggesting
that they eventually might be found to represent distinct
The three species of Atlantihyla— A. melissa, A. pan- taxa. While these observations were confounded by the
choi, and A. spinipollex—are distributed across three confusion at that time between A. spinipollex and popu-
highland areas on the Caribbean slope of Nuclear Central lations later described as P. hypomykter (McCranie &
America. The distribution of these species is notable in Wilson, 1993), Faivovich et al. (2018) found samples
that the genus appears to be absent from the intervening representing P. hypomykter sensu stricto from Hondu-
highlands of the Sierra de Omoa and Sierra de Espíritu ras were widely paraphyletic with respect to two avail-
Santo, highlands that otherwise support a rich endemic able samples from Guatemala (UTA A-51040 and UTA
herpetofauna. While the cause for the absence of Atlanti- A-51088) and appeared to represent a distinct taxon
hyla spp. from the highlands of the Guatemala-Honduras (Duellman et al., 2016; Faivovich et al., 2005, 2018).
border area is unclear, this pattern of distribution may be a Our results provide further support for the distinctiveness
reflection of the historical ecophysiography and geology of the central Guatemala samples, and we recovered a
of the Chortís Block, which hypothetically would have well-supported, reciprocally monophyletic Ptychohyla
placed the contemporary Cordillera Nombre de Dios ad- hypomykter sensu stricto clade that included a sample of
jacent to the highlands containing the modern distribu- P. hypomykter (JHT 2855) from the type locality and 55
tion of A. panchoi during the early Miocene (Townsend, samples from 20 other localities across the Chortís Block
2014). A similar pattern of potential “tectonic rafting” in sequenced for 16S (Figs. 1 – 3). The central Guatemala
relation to Chortís Block endemics and their Maya Block samples are from localities to the west and north of the
relatives may also be demonstrated in the salamander ge- Motagua-Polochic Fault system, previously recognized
nus Dendrotriton (Rovito et al., 2012), although hypoth- as a biogeographic barrier for highland amphibians and
eses related to this phenomenon have not been explicitly reptiles (Castoe et al., 2009; Hofmann & Townsend,
tested and the history of geological movement of the 2017; Jadin et al., 2012; Rovito & Parra-Olea, 2016).
Chortís Block remains a matter of some debate. Within Ptychohyla hypomykter sensu stricto, our
Questions remain concerning the deeper evolution- analysis of 56 samples using the mitochondrial locus 16S
ary history and taxonomy of certain stream treefrog recovered three geographically distinct haplotype groups
populations both within and extralimital to the Chortís (Figs. 1, 2). The “western” haplogroup was represented
Highlands. The application of large, multilocus data- by five samples from the Sierra de Omoa, Sierra de Es-
sets to such questions has resolved many species-level píritu Santo, Sierra de Joconal, and Sierra de Merendón
relationships within Hylini (Campbell et al., 2018; Ca- in the departments of Copán, Cortés, Ocotepeque, and
viedes-Solis & Nieto-Montes de Oca, 2018; Duellman Santa Bárbara in western Honduras, and included locali-
et al., 2016; Faivovich et al., 2005, 2018), greatly im- ties along the border with Guatemala. The “central” hap-
proving our understanding of the species within this logroup includes 32 samples from Montaña de Meámbar,
tribe. However, deeper relationships between and within Montaña de Santa Bárbara, Sierra de Comayagua, and
some genera remain uncertain and without satisfactory Sierra de Sulaco in central Honduras, while the “east-
resolution. Challenging groups, such as Duellmanohyla ern-southern” haplogroup includes 19 samples from the
and Sarcohyla, continue to present problems particularly Montaña de Jacaleapa and Sierra de Botaderos in the De-
when molecular and morphological evidence infer con- partment of Olancho in northwestern Honduras and the
tradictory results. These lineages might have previously Cordillera Dariense in the departments of Jinotega and
experienced evolutionary processes—such as reticulate Matagalpa, Nicaragua. Given this structure and relative
evolution or introgression—that are not detectable with- lack of diversification (albeit based on a single, relatively
out genomic scale datasets and which result in confus- conserved mtDNA fragment) across at least 14 isolated
ing morphological overlap (Burbrink & Gehara, 2018; highland areas (Fig. 1)—many of which support their
Grummer et al., 2018; Morales-Briones et al., 2018; own endemic species—we hypothesize that P. hypomyk-
749
ter underwent a relatively recent expansion across the ated endemic community. To paraphrase Edward O. Wil-
Chortís Block, possibly following the last glacial maxi- son (Sheppard, 1990), the loss of such a unique biologi-
mum when isolation among highland areas might have cal community in the name of short-term economic gains
been reduced. While our results provide some initial in- would be akin to burning a Renaissance painting to cook
sight into intraregional patterns of diversification in the a meal. The biological diversity we stand to lose in the
Chortís Block, fine-scale analysis of phylogeographic Río Jilamito Valley cannot be replaced or mitigated, as
patterns, supplemented with data from more variable they exist nowhere else on earth, and our window of op-
loci, is needed to better elucidate patterns and timing of portunity to conserve these unique resources for the next
recent diversification within P. hypomykter. generation is rapidly closing.
Habitat loss and continuing threats to

Acknowledgments
Refugio de Vida Silvestre Texiguat
The importance of Refugio de Vida Silvestre Texiguat We are indebted to the following individuals and organizations for
as a hotspot for biological endemism, and the continu- supporting our work in Refugio de Vida Silvestre Texiguat in 2010:
ally expanding threats to that endemism, have been well- Allan J. Fuentes (PROLANSATE); Adolfo Pagoada-Saybe (Mu-
documented over the past three decades (Holm & Cruz, nicipalidad de Arizona), Saíd Laínez, Iris Acosta, Roberto Down-
1994; McCranie & Castañeda, 2004a, b; McCranie et ing, Andrés Alegría (ICF); guardabosques Efrain Aguilar (San José
al., 1993; Townsend et al., 2012; Wilson et al., 1998). de Texiguat), Arnaldo Contreras (Mezapita), and Alionso Portillo
At least 12 site-restricted endemic species of amphibians (Jilamito Nuevo); and José “Juan” Arita, majordomo at La Lib-
and reptiles are known only from the environs of RVS eración. We thank Benjamin K. Atkinson, César A. Cerrato, Levi N.
Texiguat. The Río Jilamito and its tributaries in the vicin- Gray, Paul House, Mayron McKewy Mejía, Ciro Navarro-Umaña,
ity of La Liberación are also home to the only known ex- Alexander L. Stubbs, and Hermes Vega-Rodríguez for their valued
tant population of the Critically Endangered spikethumb assistance and hard work in the field. We are also grateful to Javier
treefrog Plectrohyla chrysopleura, which was discovered Sunyer and Scott L. Travers for providing some samples for com-
there in 2010 after being feared extinct (Townsend et al., parative genetic analysis. Fieldwork was supported in part by grants
2011). The reserve also protects habitat for at least 14 ad- to JHT from the Critical Ecosystem Partnership Fund, the Indiana
ditional herpetofaunal species that are otherwise endemic University of Pennsylvania (IUP) Department of Biology, IUP Col-
to Honduras (McCranie, 2018; Townsend et al., 2012; lege of Natural Sciences and Mathematics, IUP School of Gradu-
Townsend et al., 2013a), giving RVS Texiguat the unique ate Studies and Research, IUP Faculty Senate, Commonwealth of
opportunity to conserve no fewer than 27 species of en- Pennsylvania University Biologists (CPUB), and a Pennsylvania
demic amphibians and reptiles and stand as a veritable State System of Higher Education Faculty Professional Develop-
“crown jewel” of endemic biodiversity in Honduras and ment Grant. Amy Driskell and Dan Mulcahy (Smithsonian Institu-
the whole of Mesoamerica.  tion Laboratory of Analytical Biology) contributed some of the raw
Despite the strong base of scientific data supporting 16S sequence data, in collaboration with Roy McDiarmid (USNM)
the value of RVS Texiguat, particularly the Jilamito Val- as part of the “Barcoding the Herpetofauna of Eastern Nuclear Cen-
ley that forms the heart of the reserve, destruction and tral America” project. We thank Roy McDiarmid, Jeremy Jacobs,
exploitation of the remaining forest appears to be contin- Steve Gotte, and John Poindexter (USNM) for facilitating and
uing apace. Illegal extraction of timber, and subsequent preparing the loan of comparative materials and accessioning the
illegal land-clearing for coffee farms and livestock have type series, and Steve Rogers, Jennifer Sheridan, Stevie Kennedy-
historically been the primary causes of forest loss within Gold, and Kaylin Martin (CM) for accessioning the tadpole used
the reserve’s boundaries. Additionally, development has to prepare Figure 13 and some comparative specimens. Esbeiry
begun on a project to build a 14-megawatt hydroelec- Cordova-Ortiz and T.J. Firneno, Jr. provided additional assistance
tric dam on the Río Jilamito at the La Liberación site, in obtaining molecular data in the laboratory. Finally, our paper was
just outside the legal boundary of the nuclear zone at the greatly improved by the artistic contributions of Karli Rogers, who
type localities of Atlantihyla melissa, the salamander No- provided hand, foot, tadpole, and oral disk illustrations. This man-
totriton nelsoni (Townsend, 2016), the centipede snake uscript and revisions were completed while JHT was a Fulbright
Tantilla olympia (Townsend et al., 2013a), and the palm- Scholar and Visiting Professor at the Centro Zamorano de Biodi-
pitviper Bothriechis guifarroi (Townsend et al., 2013b). versidad, and JHT would like to thank Narayanaswamy Bharathan,
This project comes as the steady expansion of deforesta- Deanne Snavely, Timothy Moerland, Erika Tenorio, Oliver Komar,
tion on the southwestern, southern, and eastern edges of and Eric Van De Berghe for their support.
RVS Texiguat has begun to reach into the heart of the re-
serve, with the first patches of deforestation appearing in
2014 in the previously undisturbed upper Jilamito Valley.
Without an immediate, sustained, and verifiable effort to
protect the remaining forests of RVS Texiguat, there can
be little doubt that habitat loss will continue until what
remains has lost its ecological functionality and associ-
750
References Fujisawa, T. & Barraclough, T. G. (2013). Delimiting species us-

ing single-locus data and the generalized mixed yule coales-
cent approach: a revised method and evaluation on simulated
Bioacoustics Research Program. (2014). Raven Pro: Interactive data sets. Systematic Biology, 62, 702 – 724.
Sound Analysis Software (Version 1.5) [Computer software]. Goebel, A. M., Donnelly, J. M. & Atz, M.E. (1999). PCR primers
Ithaca, NY: The Cornell Lab of Ornithology. Retrieved from and amplification methods for 12S ribosomal DNA, the control
http://www.birds.cornell.edu/raven region, cytochrome oxidase I, and cytochrome b in Bufonids
Bossuyt, F. & Milinkovitch, M. C. (2000). Convergent adaptive and other frogs, and an overview of PCR primers which have
radiations in Madagascan and Asian ranid frogs reveal covari- amplified DNA in amphibians successfully. Molecular Phylo-
ation between larval and adult traits. Proceedings of the Na- genetics and Evolution, 11, 163 – 199.
tional Academy of Sciences USA, 97, 6585 – 6590. Gosner, K. L. (1960). A simplified table for staging Anuran em-
Bouckaert, R., Heled, J., Kühnert, D., Vaughan, T., Wu, C-H., bryos and larvae with notes on identification. Herpetologica,
Xie, D., Suchard, Ma., Rambaut, A. & Drummond, A. J. 16, 183 – 190.
(2014). beast 2: A software platform for Bayesian evolutionary Grummer, J. A., Morando, M. M., Avila, L. J., Sites Jr., J. W.
analysis. PLoS Computational Biology, 10, e1003537. & Leaché, A. D. (2018). Phylogenomic evidence for a recent
Burbrink, F. T. & Gehara, M. (2018). The biogeography of deep time and rapid radiation of lizards in the Patagonian Liolaemus fitz-
phylogenetic reticulation. Systematic Biology, 67, 743 – 755. ingerii species group. Molecular Phylogenetics and Evolution,
Campbell, J. A., Brodie Jr., E. D., Caviedes-Solis, I. W., Nieto- 125, 243 – 254.
Montes de Oca, A., Luja, V. H., Flores-Villela, O., García- Gutiérrez-García, T. A. & Vásquez-Domínguez, E. (2013). Con-
Vázquez, U. O., Sarker, G. C., Wostl, E. & Smith, E. N. sensus between genes and stones in the biogeographic and evo-
(2018). Systematics of the frogs allocated to Sarcohyla bistinc- lutionary history of Central America. Quaternary Research,
ta sensu lato (Cope, 1877), with description of a new species 79, 311 – 324.
from Western Mexico. Zootaxa, 4422, 366 – 384. Hofmann, E. P. & Townsend, J. H. (2017). Origins and biogeogra-
Castoe, T. A., Daza, J. M., Smith, E. N., Sasa, M. M., Kuch, U., phy of the Anolis crassulus subgroup (Squamata: Dactyloidae)
Campbell, J. A., Chippindale, P. T. & Parkinson, C. L. (2009). in the highlands of Nuclear Central America. BMC Evolution-
Comparative phylogeography of pitvipers suggests a consen- ary Biology, 17, 267.
sus of ancient Middle American highland biogeography. Jour- Holdridge, L. R. (1967). Life Zone Ecology (Revised ed.). San
nal of Biogeography, 36, 88 – 103. José, Costa Rica, Tropical Science Center.
Caviedes-Solis, I. W. & Nieto-Montes de Oca A. (2018). A multi- Huelsenbeck, J. P. & Ronquist, F. (2001). mrbayes: Bayesian infer-
locus phylogeny of the genus Sarcohyla (Anura: Hylidae), and
ence of phylogenetic trees. Bioinformatics, 17, 754 – 755.
an investigation of species boundaries using statistical species
IUCN (2012) IUCN Red List Categories and Criteria: Version 3.1.
delimitation. Molecular Phylogenetics and Evolution, 118, 184 –
Second Edition. Gland, Switzerland and Cambridge, United
193.
Kingdom, IUCN.
Crawford, A. J. & Smith, E. N. (2005). Cenozoic biogeography
IUCN Standards and Petitions Committee. (2019). Guidelines
and evolution in direct-developing frogs of Central America
for Using the IUCN Red List Categories and Criteria. Ver-
(Leptodactylidae: Eleutherodactylus) as inferred from a phylo-
sion 14. Prepared by the Standards and Petitions Committee.
genetic analysis of nuclear and mitochondrial genes. Molecu-
Downloaded 23 Feb 2020 from http://www.iucnredlist.org/
lar Phylogenetics and Evolution, 35, 536 – 555.
documents/RedListGuidelines.pdf.
Cruz, G., Wilson, L. D. & McCranie, J. R. (2004) Ptychohyla
Jadin, R. C., Townsend, J. H., Castoe, T. A. & Campbell, J. A. (2012).
spinipollex. The IUCN Red List of Threatened Species, 2004,
Cryptic diversity in disjunct populations of Middle American
e.T55918A11391648.
Duellman, W. E. (1963). A review of the Middle American tree montane pitvipers: a systematic reassessment of Cerrophidion
frogs of the genus Ptychohyla. University of Kansas Publica- godmani. Zoologica Scripta, 41, 455 – 470.
tions, Museum of Natural History, 15, 297 – 349. Jordan, B. R., Sigurdsson, H. & Carey, S. N. (2008). Ignimbrites
Duellman, W. E. (1970). The Hylid Frogs of Middle America. Law- in Central America and Associated Caribbean Sea Tephra:
rence, KS, Museum of Natural History, University of Kansas. Correlation and Petrogenesis. Saarbrücken, VDM Verlag.
Duellman, W. E. (2001) Hylid Frogs of Middle America (2nd ed.). Köhler, G. (2011). Amphibians of Central America. Offenbach,
Ithaca, New York, Society for the Study of Amphibians and Herpeton.
Reptiles. Köhler, G. (2012). Color Catalogue for Field Biologists. Offen-
Duellman, W. E., Marion, A. B. & Hedges, S. B. (2016). Phyloge- bach, Herpeton.
netics, classification, and biogeography of the treefrogs (Am- Köhler, J., Jansen, M., Rodríguez, A., Kok, P. J. R., Toledo, L.
phibia: Anura: Arboranae). Zootaxa, 4104, 1 – 109. F., Emmrich, M., Glaw, F., Haddad, C. F. B., Rödel, M. O. &
Faivovich, J., Haddad, C. F. B., García, P. C. A., Frost, D. R., Vences, M. (2017). The use of bioacoustics in anuran taxon-
Campbell, J. A. & Wheeler, W. C. (2005). Systematic review omy: theory, terminology, methods and recommendations for
of the frog family Hylidae, with special reference to Hylinae: best practice. Zootaxa, 4251, 1 – 124.
phylogenetic analysis and taxonomic revision. Bulletin of the Kumar, S., Stecher, G. & Tamura, K. (2016). MEGA7: Molecular
American Museum of Natural History, 294, 1 – 240. evolutionary genetics analysis version 7.0 for bigger datasests.
Faivovich, J., Pereyra, M. O., Luna, M. C., Hertz, A., Blotto, B. Molecular Biology and Evolution, 33, 1870 – 1874.
L., Vásquez-Almazán, C. R., McCranie, J. R., Sánchez, D. A., Lanfear, R., Calcott, B., Ho, S. Y. & Guindon, S. (2012). Par-
Baêta, D., Araujo-Vieira, K., Köhler, G., Kubicki, B., Camp- titionfinder: combined selection of partitioning schemes and
bell, J. A., Frost, D. R., Wheeler, W. C. & Haddad, C. F. B. substitution models for phylogenetic analysis. Molecular Bio
(2018). On the monophyly and relationships of several genera logy and Evolution, 29, 1695 – 1701.
of Hylini (Anura: Hylidae: Hylinae), with comments on recent Luque-Montes, I., Austin, J. D., Weinfurther, K. D., Wilson, L.
taxonomic changes in hylids. South American Journal of Her- D., Hofmann, E. P. & Townsend, J. H. (2018). An integrative
petology, 13, 1 – 32. assessment of the taxonomic status of putative hybrid leopard
Firneno, Jr., T. J., J. R. O’Neill, D. M. Portik, A. H. Emery, J. frogs (Anura: Ranidae) from the Chortís Highlands of Central
H. Townsend & M. K. Fujita. (2020). Finding complexity in America, with description of a new species. Systematics and
complexes: assessing the causes of mitonuclear discordance in Biodiversity, 16, 340 – 356.
a problematic species complex of Mesoamerican toads. Mo- McCranie, J. R. (2018). The lizards, crocodiles, and turtles of Hon-
lecular Ecology, 29, 3543 – 3559. duras. Systematics, distribution, and conservation. Bulletin of
751
the Museum of Comparative Zoology, Special Publications Se- Sheppard, R. Z. (1990). Nature: Splendor in the grass. Time Mag-
ries, 2, 1 – 666. azine, 3 September 1990 (retrieved 18 October 2018: http://
McCranie, J. R. & Castañeda, F. E. (2004a). A new species of content.time.com/time/magazine/article/0,9171,971049,00.
snake of the genus Omoadiphas (Reptilia: Squamata: Colubri- html).
dae) from the Cordillera Nombre de Dios in northern Hondu- Stamatakis, A. (2014). RAxML version 8: a tool for phylogenetic
ras. Proceedings of the Biological Society of Washington, 117, analysis and post-analysis of large phylogenies. Bioinformat-
311 – 316. ics, 30, 1312 – 1313.
McCranie, J.R. & Castañeda, F. E. (2004b). Notes on the second Stuart, L. C. (1943). Comments on the herpetofauna of the Sierra
specimens of Geophis damiani Wilson, McCranie, and Wil- de los Cuchumatanes of Guatemala. Occasional Papers of the
liams and Rhadinella tolpanorum Holm and Cruz D. (Colubri- Museum of Zoology, University of Michigan, 471, 1 – 28.
dae). Herpetological Review, 35, 341. Stuart, L. C. (1948). The amphibians and reptiles of Alta Verapaz
McCranie, J. R. & Solís, J. M. (2013). Additions to the amphibians Guatemala. Miscellaneous Publications of the Museum of Zoo
and reptiles of Parque Nacional Pico Bonito, Honduras, with logy, University of Michigan, 69, 1 – 109.
an updated nomenclatural list. Herpetology Notes, 6, 239 – 243. Stuart, L. C. (1954). Descriptions of some new amphibians and
McCranie, J. R & Wilson, L. D. (1993). Taxonomic changes asso- reptiles from Guatemala. Proceedings of the Biological Society
ciated with the names Hyla spinipollex Schmidt and Ptychohyla of Washington, 67, 159 – 178.
merazi Wilson and McCranie (Anura: Hylidae). The Southwest- Tang, Y., Horikoshi, M. & Li, W. (2016). ggfortify: Unified inter-
ern Naturalist, 38, 100 – 104. face to visualize statistical results of popular R packages. The
McCranie, J. R. & Wilson, L. D. (2002). The Amphibians of Hon- R Journal, 8, 474 – 485.
duras. Ithaca, NY, Society for the Study of Amphibians and Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). clustal
Reptiles. W: improving the sensitivity of progressive multiple sequence
McCranie, J. R., Wilson, L. D. & Williams, K. L. (1993). New alignment through sequence weighting, position-specific gap
species of tree frog of the genus Hyla (Anura: Hylidae) from penalties and weight matrix choice. Nucleic Acids Research,
Northern Honduras. Copeia, 1993, 1057 – 1062. 22, 4673 – 4680.
Morales-Briones, D. F., Romoleroux, K., Kolár, F. & Tank, D. C. Townsend, J. H. (2014). Characterizing the Chortís Block biogeo-
(2018). Phylogeny and evolution of the neotropical radiation graphic province: geological, physiographic, and ecological
of Lachemilla (Rosaceae): Uncovering a history of reticulate associations and herpetofaunal diversity. Mesoamerican Her-
evolution and implications for infrageneric classification. Sys- petology, 1, 204 – 251.
tematic Botany, 43, 17 – 34. Townsend, J. H. (2016). Taxonomic revision of Nototriton bar-
Palumbi, S. R., Martin, A., Romano, S., McMillan, W. O., Stice, bouri (Schmidt)(Caudata: Plethodontidae), with description
L. & Grabowski, G. (1991) The Simple Fool’s Guide to PCR. of two new species of moss salamanders from the Cordillera
Honolulu, HI, Department of Zoology, University of Hawaii. Nombre de Dios, Honduras. Zootaxa, 4196, 511 – 528.
Pons, J., Barraclough, T. G., Gomez-Zurita, J., Cardoso, A., Du- Townsend, J. H. & Wilson, L. D. (2016). Amphibians of the Cor-
ran, D. P., Hazell, S., et al. (2006). Sequence-based species dillera Nombre de Dios, Honduras: COI barcoding suggests
delimitation for the DNA taxonomy of undescribed insects. underestimated taxonomic diversity in a threatened endemic
Systematic Biology, 55, 595 – 609. fauna. Mesoamerican Herpetology, 3, 910 – 928.
Puillandre, N., Lambert, A., Brouillet, S. & ACHAZ, G. (2012). Townsend, J. H., Wilson, L. D., Cerrato M., C. A., Atkinson, B.
ABGD, Automatic Barcode Gap Discovery for primary species K., Herrera-B., L. A. & Mejía, M. M. (2011). Discovery of
delimitation. Molecular Ecology, 21, 1864 – 1877. the critically endangered treefrog Plectrohyla chrysopleura in
R Core Team (2017) R: A language and environment for statistical Refugio de Vida Silvestre Texiguat, Honduras. Herpetological
computing. Vienna, R Foundation for Statistical Computing. Bulletin, 115, 22 – 25.
Rogers, R., Karason, H. & van der Hilst, R. (2002). Epeirogenic Townsend, J. H., Wilson, L. D., Medina-Flores, M., Aguilar-Ur
uplift above a detached slab in northern Central America. Geo bina, E., Atkinson, B. K., Cerrato-Mendoza, C. A., Contreras-
logy, 30, 1031 – 1034. Castro, A., Gray, L. N., Herrera-B., L. A., Luque-Montes, I.
Ronquist, F. & Heulsenbeck, J. P. (2003). MrBayes 3: Bayesian R., McKewy-Mejía, M., Portillo-Avilez, A., Stubbs, A. L. &
phylogenetic inference under mixed models. Bioinformatics, Austin, J. D. (2012). A premontane hotspot for herpetological
19, 1572 – 1574. endemism on the windward side of Refugio de Vida Silvestre
Rovito, S.M. & Parra-Olea, G. (2016). Neotropical plethodontid Texiguat, Honduras. Salamandra, 48, 92 – 114.
biogeography: insights from molecular phylogenetics. Copeia, Townsend, J. H., L. D. Wilson, M. Medina-Flores & L. A. Herre-
2016, 222 – 232. ra-B. (2013a). A new species of centipede snake in the Tantilla
Rovito, S. M., Wake, D. B., Papenfuss, T. J., Parra-Olea, G., taeniata group (Squamata: Colubridae) from premontane for-
Muñoz-Alonso, A. & Vásquez-Almazán, C. R. (2012). Spe- est in Refugio de Vida Silvestre Texíguat, Honduras. Journal of
cies formation and geographical range evolution in a genus of Herpetology 47, 191 – 200.
Central American cloud forest salamanders (Dendrotriton). Townsend, J. H., Medina-Flores, M., Wilson, L. D., Jadin, R. C.
Journal of Biogeography, 39, 1251-1265. & Austin, J. D. (2013b). A relict lineage of green palm-pitviper
Rovito, S. M., Parra-Olea, G., Recuero, E. & Wake, D. B. (2015). (Squamata: Viperidae: Bothriechis) from the Chortís Highlands
Diversification and biogeographical history of Neotropical of Mesoamerica. ZooKeys, 298, 77 – 105.
plethodontid salamanders. Zoological Journal of the Linnean Wickham, H. (2016) ggplot2: Elegant graphics for data analysis.
Society, 175, 167 – 188. New York, NY, Springer.
Sabaj Peréz, M. H. (2016) Standard symbolic codes for institutio Wiens, J. J., Fetzner Jr, J. W., Parkinson, C. L. & Reeder, T. W.
nal resource collections in herpetology and ichthyology: An (2005). Hylid frog phylogeny and sampling strategies for spe-
online reference. Version 6.5 (16 August 2016). Washington, ciose clades. Systematic Biology, 54, 778 – 807.
DC, Society of Ichthyologists and Herpetologists. Wiens, J. J., Kuczynski, C. A., Hua, X. & Moen, D. S. (2010). An
Schmidt, K. P. (1936). New amphibians and reptiles from Hondu- expanded phylogeny of treefrogs (Hylidae) based on nuclear
ras in the Museum of Comparative Zoology. Proceedings of the and mitochondrial sequence data. Molecular Phylogenetics
Biological Society of Washington, 49, 43 – 50. and Evolution, 55, 871 – 882.
Seutin, G., White B. N & Boag P. T. (1991) Preservation of avian Williams, S. T. (2007). Safe and legal shipment of tissue samples:
blood and tissue samples for DNA analyses. Canadian Journal Does it affect DNA quality? Journal of Molluscan Studies, 73,
of Zoology, 69, 82 – 90. 416 – 418.
752
Wilson, L. D. & McCranie, J. R. (1989). A new species of Ptycho- Wilson, L.D., McCranie, J. R. & Williams, K. L. (1998). A new
hyla of the euthysanota group from Honduras, with comments species of Geophis of the sieboldi group (Reptilia: Serpentes:
on the status of the genus Ptychohyla Taylor (Anura: Hylidae). Colubridae) from northern Honduras. Proceedings of the Bio-
Herpetologica, 45, 10 – 17. logical Society of Washington, 111, 410 – 417.
ZooBank Registration
at http://zoobank.org
This published work and the nomenclatural acts it contains have

been registered in ZooBank, the online registration system for the
International Commission on Zoological Nomenclature (ICZN).
The ZooBank LSID (Life Science Identifier) can be resolved and
the associated information can be viewed through any standard
web browser by appending the LSID to the prefix http://zoobank.
org. The LSID for this publication is as follows:
urn:lsid:zoobank.org:pub:21E62F29-A8A1-40DD-B96A-
CE1049A05D79
Appendix 1. Taxa, museum voucher or sample ID numbers, and associated GenBank accession numbers for samples used in phylogenetic
analyses; accession numbers MK176937 – 177076 and MW177722 – 177747 represent novel sequences presented in this study.
Taxon Field ID Museum ID POMC RHO RAG1 12S 16S

A. melissa JHT3041 USNM 578665 — — — — MK177009
A. melissa JHT3078 — — — — — MK177020
A. melissa JHT3114 USNM 578679 MK177061 MK177068 MW177730 MK176937 MK176946
A. melissa JHT3116 USNM 578681 MW177743 MW177726 MW177731 MW177722 MK177024
A. melissa JHT3235 USNM 578687 MW177744 MW177727 MW177732 MW177723 MK177029
A. melissa JHT2441 UF 166704 MK177062 MK177069 MW177733 MK176938 MK176947
A. panchoi — USAC 4608 MH004172 — MH004197 MH004056 MH004056
A. spinipollex CAC012 USNM 578664 MK177062 MK177070 MW177734 MK176939 MK176948
A. spinipollex — USNM 509525 AY819138 — — MH004057 MH004057
A. spinipollex — USNM 514381 — AY844735 — AY843748 AY843748
A. spinipollex — UTA A—50561 — — — MH004058 MH004058
A. spinipollex MMF192 CM 163344 — — — — MK177042
A. spinipollex JHT3439 CM 170530 — — — — MK177031
753
Appendix 1 continued.

A. spinipollex JHT3460 CM 170520 MK177063 MK177071 MW177735 MK176940 MK176949
D. salvadorensis JHT2262 UF 166696 MK177064 MK177072 MW177736 MK176941 MK176950
D. salvadorensis JHT3704 CM 170641 MW177745 MK177076 MW177737 MK176945 MK176954
D. salvadorensis — USNM 563959 DQ055810 MH004221 MH004205 MH004067 MH004067
D. salvavida — USNM 559741 — — MH004202 MH004063 MH004064
D. salvavida CAC009 USNM 578645 — — — — MK177043
D. salvavida CAC010 USNM 578646 — — — — MK177044
D. salvavida JHT3107 USNM 578647 MK177067 MK177075 MW177738 MK176944 MK176953
D. salvavida JHT3108 USNM 578648 — — — — MK177045
D. salvavida JHT3328 USNM 578657 MW177746 MW177728 MW177739 MW177724 MK177052
D. salvavida MMF193 CM 163345 — — — — MK177055
D. salvavida JHT3468 CM 163201 — — — — MK177054
D. soralia JHT1585 UF 147627 MW177747 MW177729 MW177740 MW177725 MK177060
D. soralia — USNM 570489 — MH004222 — MH004068 MH004068
D. soralia — UTA A-50812 AY819111 AY844557 AY844378 AY819362 AY819493
P. dendrophasma — UTA A-51838 DQ055790 AY844603 AY844414 AY843623 AY843623
P. euthysanota JS1244 — — — — — MK177008
P. euthysanota — USAC 3629 MH004188 — MH004214 MH004096 MH004096
P. euthysanota — USAC 4613 MH004189 — MH004215 MH004097 MH004097
P. euthysanota — UTA A-54786 — AY844731 AY844509 AY843744 AY843744
P. hypomykter IRL067 — — — — — MK176955
P. hypomykter JHT1622 UF 147625 — — — — MK176963
P. hypomykter JHT2272 — — — — — MK176964
754
Appendix 1 continued.

P. hypomykter JHT2788 — MK177065 MK177073 MW177741 MK176942 MK176951
P. hypomykter JHT2855 — MK177066 MK177074 MW177742 MK176943 MK176952
P. hypomykter JHT2998 USNM 578662 — — — — MK176992
P. hypomykter JHT2999 USNM 578663 — — — — MK176993
P. hypomykter JHT3382 CM 163267 — — — — MK176994
P. hypomykter MMF103 — — — — — MK176997
P. hypomykter N1000 UF 156120 — — — — MK176998
P. hypomykter N293 — — — — — MK177002
P. hypomykter N993 UF 156119 — — — MK177007
P. hypomykter — USNM 563952 — MH004232 — MH004099 MH004099
P. hypomykter — USNM 570496 — MH004233 — MH004100 MH004100
P. cf. hypomykter — UTA A-51040 DQ055809 — — AY819445 MH004098
P. cf. hypomykter — UTA A-51088 — AY844732 — AY843745 AY843745
P. leonhardschultzei — SMF 96519 — — — MH004101 KX423519
P. leonhardschultzei — SMF 96522 — — — MH004103 KX423521
P. leonhardschultzei — UTA A-54782 — AY844733 — AY843746 AY843746
P. macrotympanum — USAC 2475 MH004190 — MH004216 MH004105 MH004105
P. macrotympanum — USAC 2479 MH004191 — MH004217 MH004106 MH004106
P. sp. — JAC21606 — AY844734 AY844511 AY843747 AY843747
P. zophodes — UTA A-54784 — AY844736 AY844513 AY843749 AY843749
755
Appendix 2. Principal component scores for the first ten PCs from male and female morphological measurement analyses; acronyms used:
snout-vent length (SVL), shank length (SHL; hind limb segment between the knee and the heel), foot length (FL; distance from posterior-
most portion of inner metatarsal tubercle to tip of longest toe), head length (HL; tip of snout to posterior angle of jaw), head width (HW;
greatest width), width of upper eyelid (EW; perpendicular to outer edge of eyelid), inter-orbital distance (IOD; measured at midlength of
upper eyelids), tympanum length (TPL), eye length (EL; horizontal length of eye at widest point, measured inside margin of eyelid), and
third finger disc width (DW; measured ventrally at widest point between outer edges of disc covers).
MALES PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 P10
St. Dev. 2.568 1.044 0.898 0.646 0.614 0.504 0.468 0.375 0.241 0.208
Prop. Var. 0.660 0.109 0.081 0.042 0.038 0.025 0.022 0.014 0.006 0.004
Cmlt. Prop. 0.660 0.769 0.849 0.891 0.929 0.954 0.976 0.990 0.996 1.000
SVL – 0.344 0.056 – 0.099 0.435 – 0.118 0.476 – 0.445 0.295 0.386 – 0.071
SHL/SVL 0.362 – 0.014 0.034 – 0.210 – 0.101 – 0.275 – 0.020 0.783 0.205 – 0.285
FL/SVL 0.316 0.041 0.064 0.788 – 0.327 – 0.354 0.091 – 0.120 – 0.081 – 0.110
HL/SVL 0.370 0.034 0.064 – 0.036 0.166 0.245 0.289 – 0.358 0.681 – 0.305
HW/SVL 0.377 0.140 – 0.029 0.004 – 0.047 0.023 – 0.087 0.078 0.261 0.868
EW/SVL 0.140 – 0.571 – 0.793 0.088 0.125 – 0.001 0.046 0.008 – 0.006 0.017
TPL/SVL 0.338 0.156 0.074 0.257 0.529 0.479 0.163 0.255 – 0.434 – 0.031
IOD/SVL 0.330 – 0.015 – 0.072 – 0.243 – 0.694 0.496 – 0.051 – 0.100 – 0.273 – 0.107
EL/SVL 0.344 0.165 – 0.083 – 0.100 0.243 – 0.169 – 0.795 – 0.275 – 0.084 – 0.190
FDW/SVL 0.102 – 0.772 0.577 0.047 0.054 0.087 – 0.200 – 0.012 0.001 0.081
FEMALES PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 P10
St. Dev. 2.583 1.245 0.774 0.620 0.487 0.464 0.373 0.312 0.278 0.170
Prop. Var. 0.667 0.155 0.060 0.038 0.024 0.022 0.014 0.010 0.008 0.003
Cmlt. Prop. 0.667 0.822 0.882 0.920 0.944 0.966 0.980 0.989 0.997 1.000
SVL 0.326 – 0.364 0.072 – 0.136 0.033 – 0.123 0.389 0.710 – 0.201 – 0.160
SHL/SVL – 0.360 – 0.082 – 0.362 – 0.048 – 0.163 0.012 – 0.122 – 0.071 – 0.615 – 0.553
FL/SVL – 0.240 – 0.501 – 0.492 – 0.244 0.378 0.038 0.327 – 0.231 0.262 0.129
HL/SVL – 0.370 – 0.061 – 0.065 0.107 – 0.030 0.001 – 0.369 0.459 0.606 – 0.359
HW/SVL – 0.368 – 0.130 – 0.131 0.209 – 0.028 0.088 – 0.247 0.374 – 0.314 0.692
EW/SVL – 0.313 0.007 0.570 – 0.412 0.471 0.385 – 0.078 0.035 – 0.155 – 0.080
TPL/SVL – 0.342 0.027 0.266 0.044 0.200 – 0.868 0.086 – 0.067 – 0.060 0.019
IOD/SVL – 0.335 0.120 0.094 – 0.514 – 0.671 0.007 0.327 0.033 0.138 0.151
EL/SVL – 0.327 0.224 0.090 0.587 0.060 0.262 0.635 0.044 0.020 – 0.110
FDW/SVL – 0.032 – 0.724 0.433 0.291 – 0.334 0.072 – 0.079 – 0.279 0.020 – 0.050
756

12 Vertebrate Zoology 70-4-2020 Townsend Et Al

Uploaded by

Copyright:

Available Formats

12 Vertebrate Zoology 70-4-2020 Townsend Et Al

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

12 Vertebrate Zoology 70-4-2020 Townsend Et Al

Uploaded by

Copyright:

Available Formats

70 (4): 731 – 756

© Senckenberg Gesellschaft für Naturforschung, 2020. 2020

A critically endangered new species of polymorphic

Submitted February 25, 2020.

ISSN 1864-5755 | eISSN 2625-8498 | DOI: 10.26049/VZ70-4-2020-12 731

separately to account for sexual size dimorphism. Based Results

3.8 – 5.2% for 12S (intralineage distances: A. spinipollex t-stat –3.683, p < 0.001; females t-stat = –4.666, p <

Atlantihyla spinipollex Atlantihyla melissa

The results of our molecular and morphological analyses,

Atlantihyla melissa sp. nov.

Figs. 7 – 8, 11 – 13

Ptychohyla spinipollex: McCranie et al., 1993: 1060.

Holotype. USNM 578679 (field number JHT 3114; Figs. 7, 8),

Variation. Atlantihyla melissa exhibits sexual dimor-

In the first comprehensive review of the genus Ptycho-

Habitat loss and continuing threats to

References Fujisawa, T. & Barraclough, T. G. (2013). Delimiting species us-

This published work and the nomenclatural acts it contains have

Taxon Field ID Museum ID POMC RHO RAG1 12S 16S

Taxon Field ID Museum ID POMC RHO RAG1 12S 16S

Taxon Field ID Museum ID POMC RHO RAG1 12S 16S

You might also like

3.8 – 5.2% for 12S (intralineage distances: A. spinipollex t-stat –3.683, p < 0.001; females t-stat = –4.666, p <

Holotype. USNM 578679 (field number JHT 3114; Figs. 7, 8),