Systematics of Huicundomantis, a new subgenus of Pristimantis (Anura, Strabomantidae) with extraordinary cryptic diversity and eleven new species

Nadia B. Páez; Santiago R. Ron

doi:10.3897/zookeys.868.26766

Abstract

Pristimantis is the most diverse genus of tetrapods comprising 532 described species. It contains a large number of morphologically cryptic species that are being discovered with the assistance of genetic evidence. We use molecular, morphological, bioacoustic, and environmental data to assess the phylogenetic relationships and determine the species within an Andean clade of Pristimantis, which is distributed from central Ecuador to northern Peru. We assign to this clade the name Huicundomantis and propose it as a subgenus. Our results show that Huicundomantis is composed of two large clades which we name as the P. phoxocephalus species group and the P. cryptomelas species group. Huicundomantis is composed of 28 species of which 12 have been described and 16 are new. We describe 11 of these undescribed species. The most effective characters to discriminate among species are DNA sequences, qualitative morphology, and advertisement calls. Morphometric and environmental characters are not very useful to define species limits. We clarify the identity of P. riveti and show that populations from southern Ecuador traditionally ascribed to P. riveti are a new species, P. lutzae sp. nov. We also show that P. prometeii is a junior synonym of P. hampatusami. The current diversity and geographic distribution of Huicundomantis are consistent with a model of allopatric speciation. All species have a restricted distribution range (less than 4330 km²) and are assigned to the Red List categories Data Deficient or threatened with extinction. We provide new reasons to increase conservation efforts for these species and their habitat. Taking our results into account, Pristimantis species richness in Ecuador increases from 211 to 221 species, and the number of species endemic to Ecuador from 119 to 129.

Resumen

Pristimantis es el género más diverso de tetrápodos, contando con 532 especies descritas. Contiene un gran número de especies morfológicamente crípticas que están siendo descubiertas con el uso de evidencia genética. En el presente estudio usamos análisis integrativos, incluyendo información molecular, morfológica, bioacústica y ambiental para determinar el contenido de especies de un clado andino de Pristimantis que se distribuye desde el centro del Ecuador hasta el norte de Perú. Asignamos a este clado el nombre de Huicundomantis y lo proponemos con el rango de subgénero. Nuestros resultados indican que Huicundomantis está compuesto por dos grandes clados que nombramos como los grupos de especies P. phoxocephalus y P. cryptomelas. Huicundomantis contiene 28 especies de las cuales 12 están descritas y 16 son nuevas. En este estudio describimos 11 de las especies nuevas. Los caracteres más efectivos para discriminar entre especies de Huicundomantis son secuencias de ADN, morfología cualitativa y cantos de anuncio. Diferencias morfométricas y ambientales entre especies son de poca utilidad para delimitar especies. En este estudio también clarificamos la identidad de P. riveti y determinamos que poblaciones del sur del Ecuador, tradicionalmente consideradas P. riveti, corresponden a la nueva especie P. lutzae sp. nov. Además, reportamos a P. prometeii como sinónimo junior de P. hampatusami. La diversidad y distribución geográfica de Huicundomantis son consistentes con un modelo de especiación alopatrica. Todas las especies tienen un rango de distribución restringido (menos de 4330 km²) y son asignadas a las categorías de Lista Roja de Datos Insuficientes o amenazadas de extinción. Nuestros resultados son un nuevo argumento para aumentar los esfuerzos de conservación de estas especies y su hábitat. Tomando en cuenta nuestros resultados, la riqueza de especies de Pristimantis en el Ecuador aumenta de 211 a 221 especies y su número de especies endémicas, de 119 a 129.

Keywords

Andes, cryptic diversity, integrative taxonomy, Neotropics, Pristimantis phoxocephalus species group, P. cryptomelas species group

Introduction

Pristimantis Jiménez de la Espada, 1870 is the most speciose genus of tetrapods with 532 described species (Frost 2019). It is distributed from Honduras in Central America to northern Argentina and southern Brazil, from the sea level to elevations above 4000 m a.s.l. (Frost 2019). In Ecuador, it comprises 34.3% of the anuran diversity with 204 species, of which 114 are endemic to the country (Ron et al. 2019).

Pristimantis taxonomy has been complex and labile because early studies were based almost exclusively on external morphology (e.g., Lynch and Duellman 1997; Lynch and Duellman 1980; Duellman and Lehr 2009), which was frequently insufficient to infer species limits and phylogenetic relationships. Recent studies have shown that species richness of Pristimantis is underestimated due to the existence of morphologically cryptic species (e.g., Elmer and Cannatella 2009; Padial and de la Riva 2009; Hutter and Guayasamin 2015; Ortega et al. 2015). The use of DNA sequences in combination with morphological and behavioral data has allowed objective species delimitation in a framework known as ‘integrative taxonomy’ (Will et al. 2005; Bickford et al. 2007; Padial et al. 2010). This approach has been increasingly applied to Neotropical amphibians resulting in the discovery of large numbers of cryptic species (e.g., Ron et al. 2006; Fouquet et al. 2007; Elmer and Cannatella 2009; Vieites et al. 2009; Funk et al. 2012; Caminer and Ron 2014; Ortega et al. 2015).

Unfortunately, relatively few groups of Pristimantis have been subject of systematic reviews based on genetic characters. Among the species that await such review are Pristimantis phoxocephalus (Lynch, 1979) and its close relatives Pristimantis riveti (Despax, 1911), P. spinosus (Lynch, 1979), and Pristimantis versicolor (Lynch, 1979). Pristimantis phoxocephalus was described from the Pacific versant of the Andes of central and southern Ecuador; its type locality is Pilaló in Cotopaxi Province (central Ecuador) at an elevation of 2340 m (Lynch 1979). Duellman and Wild (1993) provided two records from the Huancabamba Cordillera in northern Peru. Lynch and Duellman (1997) published a species account and hypothesized a close relationship with P. eugeniae, P. nyctophylax, and P. subsigillatus. They also suggested the inclusion of these species in the P. lacrimosus species group. Duellman and Pramuk (1999) commented on the existence of morphological differences between Ecuadorian and Peruvian populations. Duellman and Lehr (2009) provided a species account, noted additional morphological differences between Ecuadorian and Peruvian populations, and recommended the use of molecular characters to test the conspecificity of Ecuadorian and Peruvian populations. As currently defined, P. phoxocephalus has a wide distribution range, occurring in the Andean slopes from central Ecuador to northern Peru between 1800 and 3400 m a.s.l. (Duellman and Lehr 2009; Ron et al. 2019).

Pristimantis riveti was described from a single specimen collected in Ecuador, El Mirador (Despax 1911). This species name was not associated with any population until Lynch (1969) incorrectly identified populations of P. curtipes as “P. riveti” (Lynch 1979). Subsequently, Lynch (1979) provided a species account including a detailed description of morphological variation, comparisons with similar species, and natural history. The species account was based on specimens collected near Cuenca, Azuay Province, in southern Ecuador. Remarkably, Lynch characterized the distribution range of P. riveti as “both Andean Cordilleras and the connecting nudos at elevations of 2620–3420 m surrounding the Cuenca hoya in southern Ecuador”, a range that excludes the type locality. Subsequent references to the species (e.g., Almendáriz and Orcés 2004; Székely et al. 2016) relied on the morphological characterization provided by Lynch (1979) and ascribed the binomen to populations from Azuay Province and the adjacent Cañar and Loja Provinces. Heinicke et al. (2007) included a sample of “P. riveti” in their phylogeny. The sample was also from Azuay Province. Lynch and Duellman (1980) assigned P. riveti to the P. unistrigatus assemblage, and Hedges et al. (2008) later assigned it to the P. unistrigatus species group. Padial et al. (2014) left it unassigned.

Pristimantis spinosus was described by Lynch (1979) with the type locality “Sapote, Provincia Morona-Santiago, Ecuador, 2470 m”. Lynch (1979) hypothesized a close relationship with P. nigrogriseus and P. cryptomelas. Heinicke et al. (2007) included in their phylogeny a sample from Ecuador, Morona-Santiago, 10.6 km west of Plan de Milagro. Hedges et al. (2008) assigned it to the Pristimantis unistrigatus species group, and Padial et al. (2014) left it unassigned.

Pristimantis versicolor was described by Lynch (1979) from specimens collected at “15 km E Loja, 2800 m” in Loja Province, southern Ecuador. Duellman and Pramuk (1999) provided a species account and reported populations from Cordillera del Cóndor in northern Peru. Heinicke et al. (2007) included a sample of P. versicolor in their phylogeny which showed a close relationship with P. riveti and P. phoxocephalus. Lynch (1979) hypothesized a close relationship with P. cajamarcensis and P. unistrigatus. Lynch and Duellman (1980) assigned it to the P. unistrigatus assemblage, and Hedges et al. (2008) later assigned it to the P. unistrigatus species group. As with P. riveti, Padial et al. (2014) left it unassigned.

Our study aims to determine the content, identity, and phylogenetic relationships of species within P. phoxocephalus and closely related species. Our systematic review includes the description of 11 new species of which 10 belong to the P. phoxocephalus species group and one to the P. cryptomelas species group. Our taxonomic review is based on genetic, morphologic, bioacoustic, and environmental information.

Materials and methods

Species sampling

We focused our analysis in P. phoxocephalus and closely related species (Pyron and Wiens 2011; Padial et al. 2014). We expanded our study group to include morphologically similar species and species shown to be closely related to P. phoxocephalus based on an unpublished phylogeny of Pristimantis obtained by SRR as part of a large-scale review of Ecuadorian Pristimantis.

Phylogenetic analyses

For the molecular phylogenetic analyses, we used sequences of 85 individuals ascribed to the species above and added sequences of 48 individuals as the outgroup (46 Pristimantis spp., Craugastor talamancae, Epipedobates boulengeri). Our final dataset consists of sequences of 133 individuals from 68 localities. Sequences of 117 of these individuals were newly generated and obtained from tissues deposited at the genome bank of the Zoology Museum, Pontificia Universidad Católica del Ecuador (QCAZ), representing 25 described and at least 22 undescribed species of Pristimantis. The remaining sequences, representing 16 species, were downloaded from GenBank (http://www.ncbi.nlm.nih.gov/genbank). Vouchers and GenBank accession numbers are shown in Table 1.

Table 1.

Download as

CSV

XLSX

Genbank accession numbers for DNA sequences used in the phylogenetic analyses.

Species	Voucher	16S	ND1	RAG1	GenSeq Nomenclature
Craugastor talamancae	QCAZ 30675	MK881414	MK881414	MK881322	genseq-3
Epipedobates boulengeri	QCAZ 42251	MK881437	NA	MK881337	genseq-4
P. appendiculatus	QCAZ 16365	MK881401	NA	MK881315	genseq-4
P. atillo sp. nov.	QCAZ 31946	MK881419	NA	MK881324	genseq-2
	QCAZ 40584	MK881435	NA	MK881335	genseq-2
	QCAZ 42485	MK881438	NA	MK881338	genseq-2
	QCAZ 42486	MK881439	NA	MK881339	genseq-2
	QCAZ 42488	MK881440	NA	MK881340	genseq-2
	QCAZ 42489	MK881441	NA	MK881341	genseq-2
	QCAZ 42492	MK881442	NA	MK881342	genseq-2
	QCAZ 42496	MK881443	NA	MK881343	genseq-2
	QCAZ 42498	MK881444	NA	MK881344	genseq-2
	QCAZ 42499	MK881445	NA	MK881345	genseq-2
	QCAZ 42500	MK881446	NA	MK881346	genseq-1
	QCAZ 42501	MK881447	NA	MK881347	genseq-2
	QCAZ 42503	MK881448	NA	MK881348	genseq-2
	QCAZ 42505	MK881449	NA	MK881349	genseq-2
	QCAZ 42506	MK881450	NA	MK881350	genseq-2
	QCAZ 42512	MK881451	NA	MK881351	genseq-2
	QCAZ 42543	MK881452	NA	MK881352	genseq-2
	QCAZ 42548	MK881453	NA	MK881353	genseq-2
P. atratus	QCAZ 45580	MK881471	MK881471	MK881364	genseq-4
P. atratus	QCAZ 45645	MK881473	MK881473	MK881366	genseq-4
P. bicantus	QCAZ 31860	MK881417	NA	NA	genseq-4
P. bicantus	QCAZ 31986	MK881420	NA	MK881325	genseq-4
P. cajamarcensis	KU 217845	EF493663	NA	NA	genseq-4
	QCAZ 17142	MK881403	MK881403	MK881317	genseq-4
	QCAZ 24648	MK881405	NA	NA	genseq-4
	QCAZ 31474	MK881416	NA	NA	genseq-4
P. ceuthospilus	CORBIDI 4306	MK881397	MK881397	MK881311	genseq-4
P. chomskyi sp. nov.	QCAZ 45666	MK881476	MK881476	MK881369	genseq-2
P. chomskyi sp. nov.	QCAZ 45669	MK881477	MK881477	MK881370	genseq-1
P. cryptomelas	QCAZ 45612	MK881472	MK881472	MK881365	genseq-4
P. cryptomelas	QCAZ 45660	MK881475	MK881475	MK881368	genseq-4
P. curtipes	QCAZ 17505	MK881404	MK881404	MK881318	genseq-4
	QCAZ 39568	MK881429	MK881429	NA	genseq-4
	QCAZ 39569	MK881430	MK881430	NA	genseq-4
P. eremitus	QCAZ 43392	MK881460	NA	NA	genseq-4
P. gagliardoi	QCAZ 42575	MK881456	NA	MK881355	genseq-4
P. gagliardoi	QCAZ 46738	MK881480	NA	MK881372	genseq-3
P. glandulosus	QCAZ 40153	MK881431	MK881431	MK881333	genseq-4
P. gloria sp. nov.	KU 218035	EF493348	NA	NA	genseq-4
	QCAZ 16448	MK881402	MK881402	MK881316	genseq-2
	QCAZ 31455	MK881415	NA	NA	genseq-2
	QCAZ 56463	MK881502	NA	NA	genseq-2
	QCAZ 57201	MK881503	NA	NA	genseq-1
P. hampatusami	QCAZ 58042	MK881504	MK881504	MK881387	genseq-3
P. jimenezi sp. nov.	QCAZ 45170	MK881466	MK881466	MK881360	genseq-1
	QCAZ 45178	MK881468	MK881468	MK881362	genseq-2
	QCAZ 46977	MK881481	MK881481	MK881373	genseq-2
	QCAZ 46978	MK881482	MK881482	MK881374	genseq-2
P. kichwarum	QCAZ 25766	JN991458	NA	JQ025196	genseq-4
P. latidiscus	QCAZ 17101	EF493354	NA	EF493440	genseq-4
P. latidiscus	QCAZ 25852	JN991451	NA	JQ025188	genseq-4
P. lividus	QCAZ 54322	MK881497	MK881497	MK881382	genseq-4
P. lutzae sp. nov.	QCAZ 27471	MK881410	NA	NA	genseq-2
	QCAZ 27534	MK881411	NA	NA	genseq-2
	QCAZ 27597	MK881412	NA	NA	genseq-2
	QCAZ 32785	MK881421	MK881421	MK881326	genseq-2
	QCAZ 32786	MK881422	MK881422	MK881327	genseq-2
	QCAZ 32791	MK881424	MK881424	MK881329	genseq-2
	QCAZ 37546	MK881428	NA	NA	genseq-2
	QCAZ 47211	MK881487	NA	NA	genseq-2
	QCAZ 53728	MK881495	NA	NA	genseq-1
P. multicolor sp. nov.	QCAZ 47213	MK881488	NA	NA	genseq-1
P. multicolor sp. nov.	QCAZ 47214	MK881489	NA	NA	genseq-2
P. muscosus	QCAZ 54857	MK881501	MK881501	MK881386	genseq-4
P. nangaritza sp. nov.	QCAZ 41710	MK881436	MK881436	MK881336	genseq-1
P. omeviridis	QCAZ 10564	MK881398	NA	MK881312	genseq-4
P. petrobardus	KU 212293	EF493367	NA	NA	genseq-2
P. philipi	KU 217863	EF493672	NA	NA	genseq-3
	QCAZ 37528	MK881425	MK881425	MK881330	genseq-3
	QCAZ 37537	MK881426	MK881426	MK881331	genseq-3
	QCAZ 37542	MK881427	MK881427	MK881332	genseq-3
P. phoxocephalus	QCAZ 58463	MK881507	MK881507	MK881390	genseq-3
P. pichincha	QCAZ 11673	MK881399	MK881399	MK881313	genseq-4
P. pycnodermis	KU 218028	EF493680	NA	NA	genseq-4
P. pycnodermis	KU 218030	EF493683	NA	NA	genseq-4
P. quaquaversus	QCAZ 25676	JN991463	NA	JQ025201	genseq-4
P. rubicundus	QCAZ 26551	MK881407	MK881407	MK881320	genseq-4
P. sobetes	QCAZ 11737	MK881400	NA	MK881314	genseq-4
P. spinosus	KU 218052	EF493673	NA	NA	genseq-4
P. sternothylax	CORBIDI 13316	MK881393	MK881393	MK881308	genseq-4
P. teslai sp. nov.	QCAZ 46213	MK881478	NA	NA	genseq-1
P. thymelensis	QCAZ 16428	EF493516	NA	EF493442	genseq-4
	QCAZ 62801	MK881508	MK881508	MK881391	genseq-4
	QCAZ 62803	MK881509	MK881509	MK881392	genseq-4
P. tinguichaca	QCAZ 31945	MK881418	MK881418	MK881323	genseq-3
	QCAZ 40581	MK881432	NA	NA	genseq-3
	QCAZ 40582	MK881433	NA	MK881334	genseq-3
	QCAZ 42570	MK881454	NA	NA	genseq-3
	QCAZ 42574	MK881455	NA	MK881354	genseq-3
	QCAZ 42576	MK881457	NA	MK881356	genseq-3
	QCAZ 42578	MK881458	NA	MK881357	genseq-3
	QCAZ 54601	MK881498	NA	MK881383	genseq-3
	QCAZ 54602	MK881499	NA	MK881384	genseq-3
	QCAZ 54603	MK881500	NA	MK881385	genseq-3
P. torresi sp. nov.	QCAZ 47342	MK881490	NA	NA	genseq-1
P. torresi sp. nov.	QCAZ 47397	MK881492	MK881492	MK881380	genseq-2
P. totoroi sp. nov.	KU 218025	EF493349	NA	NA	genseq-4
	QCAZ 25105	MK881406	MK881406	MK881319	genseq-1
	QCAZ 58425	MK881505	MK881505	MK881388	genseq-2
P. unistrigatus	KU 218057	EF493387	NA	EF493444	genseq-4
P. verrucolatus sp. nov.	QCAZ 45174	MK881467	MK881467	MK881361	genseq-2
	QCAZ 45195	MK881469	MK881469	MK881363	genseq-2
	QCAZ 46982	MK881483	MK881483	MK881375	genseq-1
	QCAZ 46992	MK881484	MK881484	MK881376	genseq-2
	QCAZ 46993	MK881485	MK881485	MK881377	genseq-2
P. versicolor	KU 218096	EF493389	NA	EF493431	genseq-4
	QCAZ 45650	MK881474	MK881474	MK881367	genseq-4
	QCAZ 46310	MK881479	MK881479	MK881371	genseq-4
P. wiensi	KU 219796	EF493668	NA	NA	genseq-2
P. yumbo	QCAZ 52241	MK881494	NA	MK881381	genseq-3
P. yumbo	QCAZ 58450	MK881506	MK881506	MK881389	genseq-4
Pristimantis sp. (CCS1)	QCAZ 32790	MK881423	MK881423	MK881328	genseq-4
Pristimantis sp. (CCS2)	QCAZ 45129	MK881462	MK881462	MK881358	genseq-4
Pristimantis sp. (CCS2)	QCAZ 45155	MK881465	MK881465	MK881359	genseq-4
Pristimantis sp. (UCS1)	QCAZ 53999	MK881496	NA	NA	genseq-4
Pristimantis sp. (UCS2)	QCAZ 45029	MK881461	NA	NA	genseq-4
Pristimantis sp. (UCS3)	QCAZ 26642	MK881409	NA	NA	genseq-4
Pristimantis sp.	CORBIDI 14804	MK881394	MK881394	MK881309	genseq-4
	CORBIDI 14805	MK881395	NA	NA	genseq-4
	CORBIDI 2848	MK881396	MK881396	MK881310	genseq-4
	QCAZ 26586	MK881408	NA	MK881321	genseq-4
	QCAZ 29206	MK881413	NA	NA	genseq-4
	QCAZ 40583	MK881434	NA	NA	genseq-4
	QCAZ 43024	MK881459	NA	NA	genseq-4
	QCAZ 45131	MK881463	NA	NA	genseq-4
	QCAZ 45142	MK881464	NA	NA	genseq-4
	QCAZ 45210	MK881470	NA	NA	genseq-4
	QCAZ 47002	MK881486	MK881486	MK881378	genseq-4
	QCAZ 47353	MK881491	MK881491	MK881379	genseq-4
	QCAZ 50732	MK881493	NA	NA	genseq-4

DNA was extracted from liver or muscle tissue preserved in 95% ethanol or tissue storage buffer using a guanidine thiocyanate protocol (M. Fujita, unpublished) with some modifications. Primers used for the PCR amplification of the below-mentioned genes are listed in Suppl. material 1. Amplicons were sequenced by the Macrogen Sequencing Team (Macrogen Inc., Seoul, Korea).

The phylogenetic analyses were based on a 3199 bp dataset containing DNA sequences of the mitochondrial genes 16S (1318 bp, partial sequence), tRNA^Leu (90 bp), NADH dehydrogenase subunit 1 ND1 (961 bp), tRNA^Ile (73 bp), tRNA^Gln (71 bp), and tRNA^Met (31 bp), and the nuclear gene RAG1 (655 bp). The alignment of the sequences was performed in GeneiousPro v. 5.4.6 (GeneMatters Corp.) with the plug-in MAFFT (Katoh et al. 2002) and a posterior manual alignment with Mesquite v. 3.02 (Maddison and Maddison 2014). The aligned matrix is available at http://zenodo.org under https://doi.org/10.5281/zenodo.2555411. The best partitioning scheme and the best-fit model of molecular evolution for each partition were estimated with the software PartitionFinder v. 1.1.1 (Lanfear et al. 2012) under the greedy algorithm and the Bayesian information criterion (BIC).

Phylogenetic analyses were performed under two approaches, Bayesian Inference (BI) and Maximum-likelihood (ML). The Bayesian analysis was carried out with MrBayes v. 3.2.1 (Ronquist et al. 2012) in two independent searches for 3×10⁷ generations; each search had four Monte Carlo Markov Chains. The temperature parameter was set at 0.5. Trees were sampled every 1000 generations. We evaluated the convergence of the chains with software Tracer v. 1.6. (Rambaut et al. 2014) by inspecting plots of parameter values vs generation and confirming stationarity in their values. We also evaluated the Effective Sample Size and considered the search finished when all values were greater than 200. We discarded 25% of the 30000 trees as burn-in. The remaining trees were combined to obtain a 50% majority-rule consensus tree as well as the branch posterior probabilities (BPP).

For the ML analysis, we used Garli v. 2.0 (Zwickl 2006). We run ten replicate searches starting from stepwise trees and 10 from random trees. We finished the search after 100000 generations without topology improvements (genthreshfortopoterm = 100000); we used default values for the remaining parameters. Non-parametric bootstrap values (NPB) were obtained with 500 pseudoreplicates with the same settings of the random search, each with two independent searches. The 50% majority rule consensus for the bootstrap trees was obtained with Mesquite v. 2.75. We used the same settings to run two additional ML analyses, one exclusively for the nuclear gene and one for the mitochondrial genes. Throughout the text, we considered that a node has “strong support” when its bootstrap value is >70 and its Bayesian posterior probability is 0.95 or higher.

In order to identify candidate species, we calculated uncorrected p genetic distances for gene 16S (~1300 bp), within and between clades, with software MEGA v.7.0 (Kumar et al. 2016). See Integrative Analyses section for details.

Morphological analyses

Specimens deposited at the QCAZ collection were preserved in 10% formalin and stored in 70% ethanol. Photographs of preserved specimens were taken while these were submersed in ethanol to avoid reflections. We compared qualitative and quantitative morphological characters between clades. We used only adult specimens for these analyses but, when available, juveniles were considered to describe morphological variation. We examined 750 individuals from central and southern Ecuador ranging from 230 to 4200 m a.s.l. (Figs 1, 2), from collections of the Zoology Museum, Pontificia Universidad Católica del Ecuador (QCAZ), Ecuadorian Museum of Natural Sciences (MECN), Museum of Natural History, University of Kansas, USA (KU), Gustavo Orcés Museum of Natural History, Escuela Politécnica Nacional, Ecuador (MEPN), National Museum of Natural History, Smithsonian Institution, USA (USNM), and French Museum of Natural History of Paris (MNHNP); information for each individual is detailed in Suppl. material 2.

Figure 1.

Distribution of Pristimantis, subgenus Huicundomantis. Records are based on specimens deposited at the Museum of Zoology, Pontificia Universidad Católica del Ecuador (QCAZ). Locality information is shown in Suppl. material 2.

Figure 2.

Distribution of Pristimantis, subgenus Huicundomantis. Records are based on specimens deposited at the Museum of Zoology, Pontificia Universidad Católica del Ecuador (QCAZ). Candidate species numbers correspond to those of the text and Figure 3. Locality information is shown in Suppl. material 2. Abbreviations: CCS = confirmed candidate species, UCS = unconfirmed candidate species.

We determined sex and reproductive condition by looking for vocal sacs, vocal slits, and nuptial pads, and by gonadal inspection. For males, the presence of vocal sac, vocal slits, or nuptial pads was used as indicator of adulthood; when these characters were absent, we examined the gonads and categorized an individual as adult if the testes were swollen and enlarged. For females, we considered the presence of convoluted oviducts and large ovarian eggs as features of adulthood (Duellman and Lehr 2009).

Diagnosis and description of species follow Duellman and Lehr (2009) terminology. We use the term “distinct” for a character that is clearly visible, and “indistinct” for a character that is barely visible. When referring to the tympanum state, we use “prominent” to indicate that it protrudes from the skin surface; when referring to tubercles, we use “prominent” to imply they are conspicuous and elevated. Following Lynch and Duellman (1997), we qualify Toe V as slightly longer than Toe III when the distal end of Toe V does not reach the proximal edge of the distal subarticular tubercle on Toe IV when they are adpressed. Toe V is called longer than Toe III when its distal end reaches the proximal edge of the distal subarticular tubercle in Toe IV and does not extend to its distal edge. Toe V is much longer than Toe III if the distal end of Toe V reaches, or extends beyond, the distal edge of the distal subarticular tubercle on Toe IV. Abbreviations used for quantitative descriptive characters are: snout-vent length (SVL, distance from the tip snout to posterior margin of vent); head width (HW, greatest width of head); head length (HL, distance from the tip of snout to posterior angle of jaw articulation); eye diameter (ED, distance between anterior and posterior borders of eye); interorbital distance (IOD); upper eyelid width (EW); eye-nostril distance (EN); internarial distance (IND); tympanic diameter (TD, distance between external anterior and posterior margins of tympanic annulus); tympanum-eye distance (TED, distance between anterior margin of the tympanum to the closest point of the posterior margin of the eye); tibia length (TL, length of flexed leg from knee to heel); foot length (FL, distance from heel to tip of toe IV). All measurements were taken with a digital caliper with a precision of 0.01 mm.

For morphometric analyses, we measured 232 individuals of the candidate species and P. phoxocephalus and P. versicolor. We performed separate analyses for males and females. As they have been reported to be variable among Pristimantis species (Arroyo et al. 2005), we used SVL, HL, HW, TD, and ED measurements for morphometrics. To assess the degree of morphometric differentiation between candidate species, we performed a Discriminant Function Analysis (DFA) including the variables mentioned above without size correction. For pairwise comparisons, we used the Wilcoxon signed-rank test with corrected measurements. To remove the effect of co-varying body size, we made linear regressions between each variable and SVL. Statistical analyses were performed using software JMP v. 9.0.1 (SAS Institute Inc. 2010).

Bioacoustic analyses

We analyzed recordings of eight individuals belonging to four candidate species of the clade under study, namely Pristimantis jimenezi sp. nov., Pristimantis phoxocephalus, Pristimantis totoroi sp. nov, and Pristimantis verrucolatus sp. nov. (see Results).

Recordings were obtained from the QCAZ audio library and are available in BIOWEB (http://bioweb.bio/faunaweb/amphibiaweb/). Recording temperatures ranged between 8.8 and 12 °C. Recordings of P. phoxocephalus were obtained at its type locality, Pilaló, Cotopaxi Province. We recorded one collected and two non-collected individuals from the same chorus. Their calls had similar properties; however, we did not include the advertisement call of the collected individual in the analysis to avoid accounting for effects of stress on the characteristics of the call given that it was recorded from a plastic bag after the individual’s capture. Advertisement calls for P. jimenezi sp. nov. were recorded from Azuay Province at two localities: at the border of Cajas National Park (QCAZ 45178) and Zadracay River (QCAZ 46977) (see Suppl. material 2 for details on localities); both individuals were sequenced. For P. verrucolatus sp. nov., we had recordings available for two individuals (QCAZ 46984, QCAZ 46981) from Shoupshe, Azuay Province. Finally, recordings for P. totoroi sp. nov. were obtained from Bosque Protector Cashca Totoras, Bolívar Province. We recorded a non-collected and a collected individual (QCAZ 25105), which was subsequently sequenced.

We analyzed calls with the software Raven Pro v. 1.3 (Cornell Lab of Ornithology 2003–2008) using a sampling rate of 44.1 kHz and a frequency resolution of 10.8–11.2 Hz. Temporal data was measured on oscillograms and frequency data on power spectrums (e.g., Caminer and Ron 2014). We randomly chose five advertisement calls from each recording and measured: (1) number of notes per call; (2) number of harmonics; (3) note duration; (4) inter-note interval; (5) peak time of the note (time from beginning of the note to peak energy); (6) dominant frequency; (7) frequency of the second harmonic; (8) initial and (9) final frequency of the note; and (10) frequency change in each note, estimated from the difference in frequency between the beginning and end of a note. Because of the low number of individuals recorded per species, we could not apply statistical comparisons. However, each species had a highly stereotyped call, qualitatively distinct from the others.

Environmental analyses

We considered environmental differentiation as additional evidence to define species boundaries (e.g., Rissler and Apodaca 2007; Ortega et al. 2015). Accordingly, we analyzed differences in bioclimatic and vegetation-related variables between candidate species.

Values at each presence locality were extracted from raster maps using software ArcGIS 10.0 (ESRI). Following Menéndez-Guerrero and Graham (2013), we used eight bioclimatic variables with 30" resolution, downloaded from WorldClim (http://www.worldclim.org). The variables are: (1) annual mean temperature; (2) mean diurnal range; (3) temperature seasonality; (4) maximum temperature of warmest month; (5) minimum temperature of coldest month; (6) annual precipitation; and precipitation of (7) warmest and (8) coldest quarter. We also included three 30 m resolution layers of vegetation characteristics downloaded from Reverb (http://reverb.earthdata.nasa.gov): (1) leaf area index (LAI); (2) normalized difference vegetation index (NDVI); and (3) gross primary production (GPP). We used Principal Component Analysis (PCA) to examine environmental differentiation between candidate species using software JMP v. 9.0.1.

To avoid spatial autocorrelation, we reduced the number of localities using a buffer with a 5 km radius around each point; if two or more localities of the same species were at a distance <10 km from each other, we randomly chose one of them. We obtained a total of 62 localities for the analysis. As species limits in this clade are incorrectly defined, we did not include geographic records from the literature. Locality data are available in Suppl. material 2. A distribution map is presented in Figures 1, 2.

For the description of the habitat and distribution of the species, we assigned natural regions according to Ron et al. (2019); we estimated the Extent of Occurrence and Area of Occupancy of each species with the R package red (Cardoso 2017).

Integrative analyses

To determine species limits, we examined the covariation of independent sets of characters under the framework of integrative taxonomy (Dayrat 2005; Vieites et al. 2009). Our datasets consisted of genetic, morphological, bioacoustic, and environmental characters. We assumed that covariation between independent sets of characters is an indication of evolutionary independence between lineages (Vieites et al. 2009).

Based on the correspondence between morphological and genetic divergence, we set a 0.02 threshold of uncorrected p distances for the gene 16S to identify candidate species. We chose a lower threshold than that proposed by Fouquet et al. (2007) for amphibians (0.03) because there is evidence indicating that sister species in Pristimantis (e.g., Ortega et al. 2015; Padial and de la Riva 2009) and other Neotropical frogs (e.g., Caminer and Ron 2014; Jungfer et al. 2013; Coloma et al. 2012; Funk et al. 2012) often differ by genetic distances <0.03. Candidate species categories follow Vieites et al. (2009). Each candidate species was categorized as: (1) unconfirmed candidate species (UCS) when genetic distance was >0.02 and there were no additional character sets available; (2) confirmed candidate species (CCS) when genetic distance was >0.02 and there was covariation between genetic and morphological, bioacoustic and/or environmental characters; (3) deep conspecific linage (DCL) when genetic distance was >0.02 and there was not covariation between genetic and morphological, bioacoustics, and environmental characters. We followed the species concept proposed by de Queiroz (2007) that defines a species as a separately evolving metapopulation lineage.

Results

Phylogeny and genetic variation

The best partition scheme for the concatenated matrix consisted of the following five partitions and models: 16S, ND1 1^st position (GTR+I+G); ND1 2^nd position (HKY+I+G); ND1 3^rd position (TrN+I+G); RAG1 3^rd position (K80+G); RAG1 1^st and 2^nd position (K81uf+I+G).

Tree topologies under ML and Bayesian inference were similar except for the order of divergence events of P. tinguichaca, P. sp. CCS1, and P. atillo sp. nov. (see Integrative results). The best ML tree based on nuclear and mitochondrial genes is shown in Figure 3; the consensus tree of the Bayesian analysis is available through zenodo.org (http://doi.org/10.5281/zenodo.2556418). We found a strongly supported clade (94 NPB, 1.0 BBP) that has P. philipi diverging basally and sister to a clade composed of two large subclades with strong support (Fig. 3 and Suppl. material 3). We are referring to these clades as the P. phoxocephalus and P. cryptomelas species groups. Figure 3 shows the species limits identified in the integrative analyses (see below). The topology of the ML tree based on mitochondrial genes was congruent with the one based on the complete dataset for branches with strong support (>75% NBP) (Suppl. material 4). The tree based on the nuclear gene (Suppl. material 5) has low resolution and low branch support.

Figure 3.

Phylogenetic relationships of Pristimantis, sugenus Huicundomantis. Maximum likelihood tree for genes 16S, ND1 and RAG1. Bootstrap values (%) followed by Bayesian posterior probabilities are shown under the corresponding branches. Asterisks indicate support values of 100 (bootstrap) or 1 (posterior probabilities); missing values indicate values below 50 (bootstrap) or 0.5 (posterior probability). The collection number, identification, province and locality of the samples are shown next to each terminal; all samples are from Ecuador. Outgroup is not shown. Abbreviations: CCS = confirmed candidate species, NP = national park, PF = protected forest, UCS = unconfirmed candidate species.

Uncorrected p-genetic distances are summarized in Tables 2, 3. Average interspecific genetic distances (16S) for the ingroup range from 2.0 to 12.9% (1.8 to 5.9% between individuals of sister species). Pristimantis gloria sp. nov. was the species with the highest divergence between two populations; the population from Loja Province presented up to 1.3% of genetic distance with populations from Azuay Province (average intraspecific distance: 0.7%). We found 26 clades (i.e., candidate species) with genetic distances >2.0%. Ten of these clades are formerly described species: P. atratus (Lynch, 1979), P. cryptomelas (Lynch, 1979), P. gagliardoi Bustamante & Mendelson, 2008, P. prometeii Székely et al., 2016, P. muscosus (Duellman & Pramuk, 1999), P. philipi, P. phoxocephalus, P. spinosus, P. tinguichaca Brito et al., 2016, and P. versicolor; the remaining 16 clades are candidate species to be confirmed with the following analyses.

Table 2.

Download as

CSV

XLSX

Genetic divergence (Gene 16S) between species of the Pristimantis phoxocephalus species group. Mean uncorrected p distances (%) between groups are shown under the diagonal, standard error estimates above the diagonal. Mean uncorrected p distance within each clade and its standard error are shown on the diagonal. Number of samples is given in brackets after the name of each clade. Standard error estimates were obtained by a bootstrap procedure in MEGA 7.0.

		1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20
1	P. atillo (18)	0	0.9	0.9	0.7	0.9	1.0	0.8	0.9	1.0	0.6	0.8	1.0	0.9	0.8	0.7	0.8	0.8	0.5	0.7	0.9
2	P. atratus (2)	7.3	0	0.7	0.8	1.0	0.7	0.8	0.8	0.8	1.0	1.0	0.9	0.7	0.6	0.8	0.7	0.7	0.9	1.0	1.0
3	P. chomskyi (2)	7.1	5.4	0.2±0.1	0.7	1.0	0.7	0.7	0.8	0.8	1.0	1.0	0.9	0.7	0.7	0.7	0.7	0.7	0.9	1.0	1.0
4	P. gloria (5)	5.1	5.8	6.3	0.7±0.2	0.9	0.7	0.6	0.8	0.8	0.8	0.9	0.9	0.7	0.7	0.7	0.7	0.4	0.8	0.9	0.9
5	P. hampatusami (1)	7.0	11.1	10.9	8.7	NA	1.0	0.9	1.0	1.0	1.0	0.9	1.0	1.0	0.9	0.9	1.0	0.9	1.0	1.1	0.7
6	P. jimenezi (4)	7.7	6.8	7.8	5.8	10.7	0.5±0.2	0.7	1.0	0.9	1.0	1.0	0.9	0.8	0.6	0.7	0.7	0.6	1.0	1.1	1.0
7	P. lutzae (9)	5.0	5.7	5.6	2.9	7.5	5.3	0.2±0.1	0.9	0.8	0.8	0.8	0.9	0.7	0.6	0.7	0.7	0.6	0.8	0.9	0.9
8	P. multicolor (2)	6.7	4.4	4.9	5.5	8.9	6.7	5.5	0	0.9	1.0	1.0	1.0	0.9	0.9	1.0	0.9	0.8	1.0	1.0	1.0
9	P. phoxocephalus (1)	7.8	5.8	6.7	5.4	11.1	6.8	5.8	6.3	NA	1.1	1.0	1.0	0.7	0.8	0.9	0.8	0.9	1.0	1.1	1.0
10	P. teslai (1)	2.6	7.6	7.4	5.9	7.6	7.9	5.2	7.1	8.0	NA	0.9	1.0	1.0	0.8	0.8	0.8	0.9	0.6	0.6	1.0
11	P. tinguichaca (10)	4.7	7.8	7.7	6.1	7.0	7.8	5.3	7.5	7.7	5.7	0	1.0	0.9	0.8	0.8	0.8	0.9	0.8	0.9	0.9
12	P. torresi (2)	8.7	8.7	9.5	7.3	11.1	9.1	7.5	8.3	8.6	9.1	8.6	0.9±0.4	0.9	0.9	0.8	0.9	0.8	1.0	1.1	1.1
13	P. totoroi (3)	7.0	6.0	6.8	6.0	11.1	6.9	5.5	6.4	5.9	7.4	7.2	8.0	1±0.2	0.7	0.7	0.8	0.7	0.9	0.9	0.9
14	P. verrucolatus (5)	4.6	5.4	6.6	4.8	9.7	5.6	4.3	6.0	6.1	4.9	4.9	8.6	5.7	0.2±0.1	0.6	0.5	0.6	0.8	0.8	0.9
15	P. versicolor (3)	3.9	7.2	7.4	6.0	10.2	7.5	5.6	7.9	8.1	4.4	5.8	8.7	7.1	4.7	0.38±0.1	0.6	0.6	0.7	0.8	1.0
16	Pristimantis CCS1 (1)	4.4	6.2	7.2	5.3	10.4	6.4	5.1	6.9	6.8	4.7	4.6	8.6	6.1	3.2	5.0	NA	0.6	0.8	0.9	0.9
17	Pristimantis CCS2 (2)	5.4	5.8	6.0	2.0	7.5	5.8	2.9	5.4	6.0	6.4	5.8	7.6	5.9	4.5	5.7	5.1	0.19±0.1	0.8	0.9	0.9
18	Pristimantis UCS1 (1)	2.5	7.0	7.2	5.4	7.4	7.5	5.1	6.5	7.3	3.0	5.6	8.6	7.4	4.3	2.9	4.8	5.8	NA	0.6	0.9
19	Pristimantis UCS2 (1)	3.2	7.7	7.3	6.3	8.0	8.4	5.9	6.9	8.1	2.5	5.8	9.9	7.4	4.9	4.3	4.8	6.6	2.8	NA	1.0
20	Pristimantis UCS3 (1)	6.9	8.1	7.4	6.6	3.5	7.3	6.2	8.0	7.7	7.1	6.5	8.5	7.7	6.9	7.5	6.7	6.8	7.2	7.7	NA

Table 3.

Download as

CSV

XLSX

Genetic divergence (16S) between species of the Pristimantis cryptomelas species group. Mean uncorrected p distances (%) between groups are shown under the diagonal, standard error estimates above the diagonal. Mean uncorrected p distance within each clade and its standard error are shown on the diagonal. Number of samples is given in brackets after the name of each clade. Standard error estimates were obtained by a bootstrap procedure in MEGA 7.0.

		1	2	3	4	5
1	P. cryptomelas (2)	0	1.1	0.8	0.9	0.6
2	P. gagliardoi (2)	8.8	0.9±0.4	1.1	1.1	1.0
3	P. muscosus (1)	7.6	8.6	NA	1.0	0.7
4	P. nangaritza (1)	8.1	8.3	9.8	NA	0.9
5	P. spinosus (1)	5.9	7.6	7.4	8.1	NA

Morphological analyses

Morphological variation is shown in Figure 4. The ingroup shares the absence of dorsolateral folds (except for P. atratus), presence of a tympanic annulus and membrane (except for P. philipi), dentigerous processes of vomers evident, first finger shorter than the second, presence of lateral fringes on fingers and toes, basal webbing on toes (except for P. philipi and P. prometeii), and small to prominent tubercles on heel. Additionally, all the species of the P. cryptomelas group bear postocular folds and prominent tubercles on eyelids and heels. Because postocular folds and prominent tubercles on eyelids and heels are absent in the P. phoxocephalus group and P. philipi, both characters appear to be synapomorphies for the P. cryptomelas group. Meanwhile, species of the Pristimantis phoxocephalus group have a keel or a papilla at the tip of the snout. The papilla is absent in P. philipi and all species of the P. cryptomelas group, except for P. cryptomelas which has papilla, suggesting that the presence of a papilla is a synapomorphy for the P. phoxocephalus group with an independent origin in P. cryptomelas. Traits that differ the most between species include snout shape, presence or absence of middorsal, lateral and postocular folds, dorsal skin texture, shape and width of discs, and coloration of groins, hidden surfaces of thighs, and iris. Character states for qualitative morphological traits are shown in Table 4.

Figure 4.

Morphological variation within Pristimantis, subgenus Huicundomantis. A Pristimantis philipi. Pristimantis cryptomelas group (*): B Pristimantis gagliardoi C Pristimantis muscosus D Pristimantis cryptomelas. Pristimantis phoxocephalus group (**): E Pristimantis torresi sp. nov. F Pristimantis hampatusami G Pristimantis jimenezi sp. nov. H Pristimantis phoxocephalus I Pristimantis totoroi sp. nov. J Pristimantis atratus K Pristimantis chomskyi sp. nov. L Pristimantis multicolor sp. nov. M Pristimantis sp. (CCS2) N Pristimantis gloria sp. nov. O Pristimantis lutzae sp. nov. P Pristimantis tinguichaca Q Pristimantis verrucolatus sp. nov. R Pristimantis sp. (UCS1) S Pristimantis teslai sp. nov. T Pristimantis versicolor U Pristimantis atillo sp. nov. Red branches are for clades without available photographs of live individuals. Not shown at the same scale.

Table 4.

Download as

CSV

XLSX

Qualitative morphological traits of Pristimantis, subgenus Huicundomantis. Coloration corresponds to live individuals unless otherwise noticed. + is for present; – is for absent.

	Dorsal texture	Cranial crests	Postocular/scapular fold/ridge	Middorsal fold	Tubercles row head	Dorsolateral folds	Lateral fold	Snout shape (dorsal; lateral)	Discs (expansion, shape)	Basal webbing	Iris coloration	Groin coloration
P. atillo sp. nov.	Shagreen with or without scattered small subconical tubercles	–	–	+/–	+	–	+/–	Acuminate; protruding; with a fleshy keel	Expanded to broadly expanded, rounded to elliptical	+	Copper with a faint medial horizontal darker streak	Bright orange, surrounded or not by yellow spots
P. atratus	Shagreen	Low	–	+/–	+/–	+	+	Subacuminate; rounded; with or without a papilla	Broadly expanded, rounded to elliptical	+	Pale yellow to pale bronze, thin brown streak	Black with white or yellow spots
P. balionotus	Tuberculate	–	–	–	–	–	+/–	Subacuminate; rounded; with a papilla	Slightly expanded, elliptical	–	Bronze with a red medial streak	In preservative, pale rusty brown with or without faint brown marbling
P. chomskyi sp. nov.	Shagreen	–	–	–	–	–	–	Subacuminate; rounded; with or without a papilla	Expanded, rounded to elliptical	+	Orange with a faint reddish brown streak	Dark chocolate brown with or without small cream flecks
P. gloria sp. nov.	Tuberculate to warty	–	–	+	–	–	–	Subacuminate; rounded; with or without a papilla	Expanded, truncate to elliptical	+	Light silver to cream with wide black reticulations and a red streak	Pinkish to purplish brown with irregular cream to light brown flecks or spots
P. hampatusami	Shagreen with scattered tubercles	–	Low, W-shaped	+	+	–	+	Subacuminate; rounded; with or without a papilla	Broadly expanded, elliptical to truncate	–	Golden to bronze with a red medial horizontal streak	Reddish brown with yellow spots
P. jimenezi sp. nov.	Shagreen with or without scattered small subconical tubercles	–	–	+	+	–	+	Acuminate; protruding; with a fleshy keel	Expanded to broadly expanded, elliptical to truncate	+	Copper with red medial streak to completely red	Pinkish, purplish or dark brown with small light brown to yellow spots
P. lutzae sp. nov.	Shagreen to tuberculate	–	–	+/–	+/–	–	+/–	Round to subacuminate, with or without a papilla	Expanded, elliptical to truncate	+	Golden to creamy brown with a reddish brown medial streak	Pinkish to reddish brown, suffused or not with orange, with cream or light brown spots
P. multicolor sp. nov.	Shagreen to warty	–	–	–	+/–	–	–	Subacuminate; rounded; with or without a papilla	Expanded to broadly expanded, elliptical to truncate	+	Copper to creamy yellow with or without a red to dark brown streak	Cream, orange, brown, or black with or without cream to yellow flecks or spots
P. phoxocephalus	Shagreen with scattered small tubercles	–	–	+	+	–	+/–	Acuminate; protruding; with a fleshy keel	Broadly expanded, rounded to elliptical	+	Golden with wide black reticulations	In preservative, brown with cream reticulation or spots
P. percultus	Tuberculate	Low	–	–	–	–	–	Subacuminate; acutely rounded; with a keel	Expanded, elliptical	+	Copper with or without a faint red medial streak	Yellow with black reticulations
P. teslai sp. nov.	Tuberculate with prominent and rounded tubercles	–	–	–	+	–	–	Acuminate; protruding; with a fleshy keel	Expanded to broadly expanded, elliptical to truncate	+	Copper	Dark brown with yellow irregular blotches
P. tinguichaca	Smooth	–	–	+	+	–	+	Subacuminate; rounded; with or without a papilla	Broadly expanded, rounded	+	Red with a dark medial streak	Reddish to dark brown
P. torresi sp. nov.	Shagreen	–	–	+	+/–	–	+	Acuminate; protruding; with a fleshy keel	Broadly expanded, elliptical to truncate	+	Golden to beige with red to reddish brown streak	Light purplish brown to brown with or without yellow spots
P. totoroi sp. nov.	Shagreen with or without scattered small tubercles	–	–	+	+	–	+	Acuminate; protruding; with a fleshy keel	Broadly expanded, rounded to elliptical	+	Golden with a medial horizontal red streak	In preservative, brown with or without pale spots
P. verrucolatus sp. nov.	Shagreen with or without scattered small tubercles	–	–	+/–	+/–	–	+	Acuminate; rounded; with a fleshy keel	Broadly expanded, elliptical to truncate	+	Coppery brown	Reddish brown with small light brown, orangey brown or yellow spots
P. versicolor	Shagreen to tuberculate	–	–	+/–	–	–	–	Subacuminate; rounded; with or without a papilla	Expanded, rounded to elliptical	+	Bronze-white with a reddish brown streak	Brown to black with cream to reddish pink flecks or spots
P. philipi	Tuberculate with low tubercles and warts	–	W-shaped ridge	+/–	+	–	–	Rounded	Slightly expanded, truncate	–	Grayish bronze	Black with or without white or yellow streaks
P. cryptomelas	Shagreen	–	Prominent,) (-shaped	+/–	+	–	+	Subacuminate; rounded; with or without a papilla	Broadly expanded, rounded to elliptical	+	Red, orange or cream with a red or orange streak	Black with or without yellow or white spots
P. gagliardoi	Shagreen with scattered tubercles	–	Prominent, W-shaped	+/–	+	–	+	Rounded	Broadly expanded, elliptical to truncate	+	Bronze with or without a brown medial horizontal streak	Pink to orange with or without brown blotches
P. muscosus	Smooth to shagreen	–	Low,) (-shaped	–	+	–	+	Rounded	Broadly expanded, elliptical to truncate	+	Reddish brown	Dark brown with orange, yellow or white spots
P. nangaritza sp. nov.	Finely tuberculate	–	Low,) (-shaped	+/–	+	–	+/–	Subacuminate; rounded	Broadly expanded, rounded to elliptical	+	Unknown	In preservative, light brown with or without pale flecks
P. spinosus	Finely tuberculate	Low	Low,) (-shaped	–	+	–	+	Subacuminate; rounded to truncate	Broadly expanded, rounded	+	Unknown	Black with big white spots

Ranges of quantitative variables widely overlap among species (Fig. 5A, B). However, several species pairs differ significantly from each other according to one or more variables (see Fig. 5A, C for an example); these results will be mentioned when useful to discriminate similar species. The DFA classification assigned the correct species in 77.1% of the specimens in females (64 out of 83), and 68.7% in males (126 out of 182). Although the DFA classification was not perfect, it shows some morphometric differentiation among species. Summarized information of morphometric measurements is shown in Table 5.

Figure 5.

Morphological and environmental comparisons among species of Pristimantis, sugenus Huicundomantis. Intra and interspecific variation in tympanum size, respective to body, in A females and B males. Green horizontal lines represent the mean C Principal components from analysis of eleven environmental variables. See Table 7 for character loadings on each component.

Table 5.

Download as

CSV

XLSX

Descriptive statistics for morphometric variables of Pristimantis, subgenus Huicundomantis. Mean ± SD followed by the range of the measurements are given in each cell. ‘n’ is for the number of samples. Numbers in brackets after the species name refer to the bibliographic source of the data: (1) Lynch 1979, (2) Bustamante and Mendelson 2008, (3) Yánez-Muñoz et al. 2016, (4) Duellman and Pramuk 1999, (5) Lynch and Duellman 1995, (6) Brito et al. 2016; otherwise the source is this study. Abbreviations are: SVL = snout-vent length; HL = head length; HW = head width; TD = tympanum diameter; ED = eye diameter. All measurements are in mm.

Females
Clade	SVL	HL	HW	TD	ED
P. atillo sp. nov. n = 5	31.3 ± 2.3; 29.4–35.3	11.0 ± 0.5; 10.3–11.7	11.5 ± 0.7; 10.9–12.7	1.7 ± 0.2; 1.5–1.8	3.2 ± 0.1; 3.1–3.3
P. atratus (1) n = 10	27.4 ± 2.1; 24.9–29.2	–	–	–	–
P. balionotus (1) n = 7	28.1 ± 0.6; 27.1–29.1	–	–	–	–
P. cryptomelas (1) n = 1	38.6	–	–	–	–
P. gagliardoi (2) n = 5	30.6 ± 3.1; 26.8–33.6	10.7 ± 0.6; 9.8–11.4	12.2 ± 1.0; 10.7–13.3	1.3 ± 0.2; 1.0–1.5	3.8 ± 0.3; 3.5–4.1
P. gloria sp. nov. n = 15	30.1 ± 3.0; 26.7–35.8	11.0 ± 0.7; 10.1–12.4	11.3 ± 1.1; 9.8–13.5	1.6 ± 0.1; 1.4–1.8	3.2 ± 0.2; 2.8–3.6
P. hampatusami (3) n = 11	30.6 ± 2.1; 25.9–34.1	–	–	–	–
P. jimenezi sp. nov. n = 9	35.0 ± 2.1; 31.1–37.4	12.0 ± 0.5; 11.0–12.6	13.1 ± 0.6; 13.0–13.9	1.9 ± 0.1; 1.7–2.1	3.7 ± 0.2; 3.5–4.0
P. lutzae sp. nov. n = 15	31.4 ± 1.3; 29.7–33.9	11.3 ± 0.4; 10.7–12.1	12.2 ± 0.5; 11.4–12.9	1.7 ± 0.1; 1.5–1.8	3.4 ± 0.2; 3.1–3.8
P. multicolor sp. nov. n = 10	35.3 ± 3.5; 29.4–40.5	13.4 ± 1.1; 11.6–15.1	14.4 ± 1.3; 11.9–16.1	2.2 ± 0.2; 1.8–2.5	4.0 ± 0.3; 3.6–4.4
P. muscosus (4) n = 4	37.8; 29.6–46.1	–	–	–	–
P. nangaritza sp. nov. n = 4	29.1 ± 2.7; 25.9–32.4	11.5 ± 0.8; 10.6–12.2	11.2 ± 1.0; 10.1–12.1	1.5 ± 0.1; 1.4–1.7	3.7 ± 0.3; 3.4–4.0
P. percultus (1) n = 1	38.2	–	–	–	–
P. phillipi (5) n = 6	31.2 ± 1.1; 26.5–33.7	–	–	–	–
P. phoxocephalus n = 5	36.9 ± 2.2; 34.2–39.9	12.9 ± 0.9; 11.5–13.8	13.1 ± 1.2; 11.3–14.2	1.9 ± 0.1; 1.8–2.1	3.8 ± 0.3; 3.5–4.1
Pristimantis CCS1 n = 1	37.0	12.5	13.8	2.0	3.5
Pristimantis CCS2 n = 1	38.3	13	14.4	1.7	3.5
Pristimantis UCS1 n = 1	31.2	12.4	12.4	1.7	3.8
P. spinosus (1) n = 29	31.8 ± 1.62; 28.3–34.5	–	–	–	–
P. tinguichaca (6) n = 9	29.7 ± 1.5; 28.1–31.7	10.6 ± 0.4; 10.1–11.0	11.0 ± 0.5; 10.1–11.5	1.4 ± 0.1; 1.3–1.7	3.3 ± 0.3; 2.8–3.7
P. torresi sp. nov. n = 5	34.7 ± 3.7; 30.1–39.5	12.2 ± 1.3 10.4–13.8	13.1 ± 1.2 11.5–14.6	1.9 ± 0.2 1.8–2.2	3.6 ± 0.3 3.3–3.9
P. totoroi sp. nov. n = 7	33.0 ± 0.9; 31.9–34.3	12.1 ± 0.4; 11.5–12.6	12.3 ± 0.3; 11.8–12.6	1.6 ± 0.1; 1.6–1.8	3.3 ± 0.2; 3.1–3.6
P. verrucolatus sp. nov. n = 2	43.6 ± 4.5; 40.4–46.8	15.0 ± 1.1; 14.2–15.8	16.6 ± 1.7; 15.3–17.8	2.2 ± 0.1; 2.1–2.3	4.2 ± 0.0; 4.2–4.3
P. versicolor n = 4	27.8 ± 3.5; 24.9–32.4	10.7 ± 1.6; 9.0–12.4	10.5 ± 1.0; 9.5–11.4	1.7 ± 0.3; 1.4–2.1	3.2 ± 0.4; 2.9–3.7
Males
P. atillo sp. nov. n = 29	24.7 ± 2.5; 17.7–28.1	8.7 ± 0.7; 6.4–9.7	8.8 ± 0.9; 5.9–10.0	1.2 ± 0.1; 0.8–1.4	2.7 ± 0.3; 2.0–3.0
P. atratus (1) n = 19	21.7 ± 3.71; 17.4–24.0	–	–	–	–
P. balionotus (1) n = 2	20.0 ± 0.2; 21.8–22.2	–	–	–	–
P. chomskyi sp. nov. n = 3	28.0 ± 4.2; 24.0–32.4	9.7 ± 1.2; 8.8–11.1	10.6 ± 1.6; 9.4–12.4	1.3 ± 0.2; 1.1–1.5	3.3 ± 0.4; 2.9–3.6
P. cryptomelas (1) n = 4	29.2; 28.2–30.3	–	–	–	–
P. gagliardoi (2) n = 5	22.2 ± 2.0; 19.1–24.3	7.9 ± 0.9; 6.8–9.1	9.0 ± 0.9; 7.7–10.2	0.8 ± 0.0; 0.7–0.8	2.8 ± 0.2; 2.4–3.0
P. gloria sp. nov. n = 24	21.7 ± 2.3; 16.8–24.7	8.6 ± 0.7; 6.7–9.4	8.4 ± 0.8; 6.5–9.5	1.2 ± 0.1; 0.9–1.4	2.6 ± 0.2; 2.1–2.9
P. hampatusami (3) n = 28	21.0 ± 1.8; 17.0–24.9	–	–	–	–
P. jimenezi sp. nov. n = 12	25.5 ± 1.6; 21.8–27.1	8.9 ± 0.6; 7.8–9.6	9.1 ± 0.6; 7.8–9.7	1.4 ± 0.1; 1.3–1.5	3.06 ± 0.24; 2.56–3.37
P. lutzae sp. nov. n = 14	24.6 ± 1.7; 21.4–27.0	8.7 ± 0.5; 7.9–9.4	9.3 ± 0.7; 8.2–10.4	1.3 ± 0.1; 1.1–1.4	2.77 ± 0.19 2.53–3.26
P. multicolor sp. nov. n = 12	26.2 ± 3.5; 19.7–29.7	9.6 ± 1.0; 8.0–11.0	10.1 ± 1.2; 8.0–11.5	1.5 ± 1.2; 1.1–1.8	3.18 ± 0.49; 2.45–3.82
P. nangaritza sp. nov. n = 13	18.8 ± 1.1; 17.4–20.8	7.6 ± 0.6; 7.1–8.8	7.2 ± 0.5; 6.6–8.3	1.0 ± 0.1; 0.8–1.2	2.69 ± 0.17 2.49–3.08
P. percultus (1) n = 1	29.8	–	–	–	–
P. phillipi (5) n = 10	23.0 ± 0.4; 21.1–25.1	–	–	–	–
P. phoxocephalus n = 3	24.7 ± 3.2; 20.8–27.9	9.0 ± 1.36; 7.4–10.6	8.7 ± 1.5; 7.0–10.4	1.2 ± 0.2; 1.0–1.3	2.66 ± 0.25 2.36–2.90
Pristimantis UCS2 n = 2	20.5 ± 1.7; 19.3–21.7	7.6 ± 0.7; 7.1–8.1	7.3 ± 0.4; 7.0–7.6	1.0 ± 0.1; 0.9–1.0	2.43 ± 0.23 2.29–2.62
P. spinosus (1) n = 34	20.1 ± 1.8; 16.1–25.0	–	–	–	–
P. teslai sp. nov. n = 4	25.2 ± 1.8; 23.4–27.3	8.9 ± 0.4; 8.4–9.2	9.0 ± 0.6; 8.2–9.5	1.3 ± 0.1; 1.2–1.5	2.89 ± 0.14 2.76–3.07
P. tinguichaca (6) n = 11	23.4 ± 1.1; 21.2–24.7	8.7 ± 0.3; 8.2–9.2	8.6 ± 0.5; 7.9–9.4	1.3 ± 0.2; 1.1–1.6	2.8 ± 0.4 2.3–3.3
P. torresi sp. nov. n = 18	25.9 ± 2.1; 23.3–30.0	9.2 ± 0.6; 8.4–10.5	9.3 ± 0.7; 8.4–10.5	1.3 ± 0.1; 1.2–1.6	2.98 ± 0.23 2.64–3.47
P. totoroi sp. nov. n = 21	26.8 ± 2.0; 23.2–29.4	9.6 ± 0.6; 8.5–10.5	9.5 ± 0.6; 8.4–10.3	1.3 ± 0.1; 1.1–1.4	2.79 ± 0.25 2.27–3.25
P. verrucolatus sp. nov. n = 15	29.4 ± 2.7; 25.1–34.5	9.7 ± 0.7; 8.5–10.7	10.5 ± 1.0; 8.5–12.1	1.5 ± 0.1; 1.3–1.8	3.21 ± 0.24 2.79–3.64
P. versicolor n = 12	20.5 ± 1.6; 18.1–23.3	8.2 ± 0.5; 7.3–8.9	7.6 ± 0.6; 6.7–8.6	1.3 ± 0.2; 1.0–1.6	2.84 ± 0.32 2.32–3.36

Bioacoustic analyses

Bioacoustic data are summarized in Table 6. The available advertisement calls of P. phoxocephalus and three candidate species have a similar structure: one to several whistles (Fig. 6). Calls from Cashca Totoras (Bolívar Province, Ecuador) were distinct from those from Pilaló (Cotopaxi Province, Ecuador) even though both populations had historically been ascribed to P. phoxocephalus. The clade from high altitudes of Azuay Province shows, by far, the longest duration of the note, the lowest dominant frequency, and the highest variation between the initial and final frequency of the note. The advertisement call of the clade from lower altitudes of Azuay shows the highest frequency of the first and second harmonic. Pristimantis phoxocephalus shows the shortest interval between notes. All analyzed species are separated by large genetic distances (>5.6%). They also have unequivocally distinct advertisement calls (Fig. 6).

Figure 6.

Comparison of advertisement calls of Pristimantis jimenezi sp. nov., P. phoxocephalus, P. totoroi sp. nov. and P. verrucolatus sp. nov. A Pristimantis jimenezi: QCAZ 46977 B Pristimantis phoxocephalus: non-collected individual from Pilaló (type locality). C Pristimantis totoroi: non-collected individual from Cashca Totoras (type locality) D Pristimantis verrucolatus: QCAZ 46981. Oscillograms are shown above spectrograms. X axis for time (s); Y axis for frequency (kHz) in spectrograms.

Table 6.

Download as

CSV

XLSX

Descriptive statistics for bioacoustic variables of Pristimantis jimenezi sp. nov., P. phoxocephalus, P. totoroi sp. nov. and P. verrucolatus sp. nov. Mean ± SD and range in brackets are given in each cell. The number of samples is given in brackets after species name.

	P. jimenezi (2)	P. phoxocephalus (2)	P. totoroi (2)	P. verrucolatus (2)
Notes per call	1–2	3–9	2–6	1
No. harmonics	4	4–5	3–7	4–7
Note duration (s)	0.182 ± 0.035	0.154 ± 0.030	0.125 ± 0.038	0.433 ± 0.037
Note duration (s)	(0.157–0.206)	(0.133–0.175)	(0.098–0.152)	(0.407–0.459)
Interval between notes (s)	0.297 ± 0.088	0.135 ± 0.035	0.339 ± 0.038	NA
Interval between notes (s)	(0.234–0.360)	(0.110–0.159)	(0.312–0.366)
Peak time (s)	0.091 ± 0.018	0.077 ± 0.014	0.063 ± 0.019	0.216 ± 0.018
Peak time (s)	(0.080–0.103)	(0.067–0.087)	(0.049–0.076)	(0.204–0.230)
Dominant frequency (Hz)	2894.58 ± 167.49	2578.12 ± 0	2569.65 ± 21.21	2114.08 ± 265.17
Dominant frequency (Hz)	(2776.15–3013.01)		(2512.22–2627.07)	(1926.58–2301.58)
Initial frequency (Hz)	2789.89 ± 232.02	2452.15 ± 87.01	2550.96 ± 148.20	1837.5 ± 212.13
Initial frequency (Hz)	(2625–2953.13)	(2390.63–2513.67)	(2446.16–2655.75)	(1687.5–1987.5)
Final frequency (Hz)	2976.56 ± 232.02	2586.91 ± 12.43	2619.88 ± 50.73	2231.25 ± 238.65
Final frequency (Hz)	(2812.5–3140.63)	(2578.13–2595.70)	(2584–2655.75)	(2062.5–2400)
Frequency change (Hz)	187.5 ± 0	134.77 ± 99.44	68.92 ± 97.47	393.75 ± 26.52
Frequency change (Hz)		(64.45–205.08)	(0–137.84)	(375–412.5)
2^nd harmonic frequency (Hz)	5373.79 ± 412.25	4532.23 ± 62.15	4714.43 ± 453.59	4171.89 ± 596.64
2^nd harmonic frequency (Hz)	(5446.29–6029.3)	(4488.28–4576.17)	(4393.69–5035.16)	(3750–4593.78)

Environmental analyses

In the PCA, the eleven environmental variables were reduced to four principal components (PCs) with an eigenvalue >1. Together they accounted for 81.58% of the variation. The highest loadings for each PC were: temperature-associated variables for PC I, vegetation-associated variables and precipitation of the coldest quarter for PC II, precipitation-associated variables for PC III, and NDVI and temperature seasonality for PC IV. Percentage of explained variance and variable loadings for each PC are shown in Table 7.

Candidate species widely overlap in environmental space (Fig. 5C). However, we found differences between clades occurring in areas with the lowest altitudes and warmest temperatures ((P. hampatusami = P. prometeii), P. nangaritza sp. nov., P. spinosus) relative to those occurring in paramo (P. gloria sp. nov., P. lutzae sp. nov., P. multicolor sp. nov., P. philipi, P. teslai sp. nov., P. verrucolatus sp. nov., Pristimantis sp. CCS1).

Table 7.

Download as

CSV

XLSX

Character loadings, eigenvalues and percentage of explained variance for Principal Components (PC) I–IV. The analysis was based in eleven environmental variables extracted from presence localities of 26 species of Pristimantis, subgenus Huicundomantis. Bold figures indicate highest loadings.

Variable	PCI	PCII	PCIII	PCIV
Annual mean temperature	0.956	0.014	-0.139	0.153
Mean diurnal temperature range	0.680	-0.365	-0.026	-0.383
Temperature seasonality	-0.257	0.016	0.465	0.642
Max temperature of warmest month	0.966	-0.046	-0.122	0.121
Min temperature of coldest month	0.930	0.050	-0.153	0.201
Annual precipitation	0.382	-0.395	0.777	-0.101
Precipitation of warmest quarter	0.219	0.156	0.709	0.234
Precipitation of coldest quarter	0.428	-0.605	0.210	-0.268
Gross primary production	0.377	0.839	0.128	-0.185
Leaf area index	0.469	0.797	0.146	-0.125
Normalized difference vegetation index	0.339	-0.229	-0.378	0.611
Eigenvalue	4.096	2.074	1.606	1.198
Cumulative variance (%)	37.23	56.09	70.69	81.58

Integrative analyses

By combining evidence from the abovementioned character sets, we conclude that our study group is composed of 28 candidate species. Of them, 25 are confirmed candidate species and 3 are unconfirmed. Of the CCS, 12 have available names and the 13 remaining are new species. We describe 11 of them below. We did not describe two CCS because we did not have enough specimens to characterize species variation. Because of their similar morphology to the examined species, and after a thorough bibliographic analysis, we propose P. balionotus (Lynch, 1979) and P. percultus (Lynch, 1979) as part of our clade of study.

In brief, our ingroup is composed of 27 species divided in two species groups and P. philipi. The P. phoxocephalus species group comprises 22 species: P. atillo sp. nov., P. atratus, P. chomskyi sp. nov., P. gloria sp. nov., P. hampatusami, P. jimenezi sp. nov., P. lutzae sp. nov., P. multicolor sp. nov., P. phoxocephalus, P. teslai sp. nov., P. tinguichaca, P. torresi sp. nov., P. totoroi sp. nov., P. versicolor, P. verrucolatus sp. nov., Pristimantis sp. (CCS1), and Pristimantis sp. (CCS2); and three unconfirmed species: Pristimantis sp. (UCS1), Pristimantis sp. (UCS2), and Pristimantis sp. (UCS3). The P. cryptomelas species group comprises P. cryptomelas, P. gagliardoi, P. muscosus, P. nangaritza sp. nov., and P. spinosus. We present the evidence used to define the limits of each species in the following section.

We identified that the recently described P. hampatusami Yánez-Muñoz et al., 2016 and P. prometeii Székely et al., 2016 belong to the same species. We compared specimens of both species from the QCAZ and INABIO collections. Our comparisons showed that both species are morphologically indistinguishable. The type locality for both species is the same further suggesting that both represent a single species. With this evidence, we propose P. prometeii as a junior synonym for P. hampatusami given the principle of priority.

We found a misidentification of two Genbank sequences with accession numbers KU217863 and KU218025. GenBank identifications are P. cryophilius and P. phoxocephalus, respectively. They actually are P. philipi (KU217863) and P. totoroi sp. nov. (KU218025). Updated identifications of sequences used in this study are shown in Table 1.

In the following section, we present the descriptions of the new species and new taxa identified in this study. We also include an updated diagnosis of P. phoxocephalus sensu stricto, as diagnoses made in its description and later revisions (e.g., Duellman and Pramuk 1993; Lynch and Duellman 1997) included information of non-conspecific populations.

Systematic accounts

Huicundomantis subgenus nov.

http://zoobank.org/A18720A1-FB71-4342-B0B2-8C9C8F06B5F4

Type species

Pristimantis phoxocephalus (Lynch, 1979).

Definition

This clade is strongly supported by genetic evidence (Fig. 3). Morphological synapomorphies are unknown. Members of this clade are characterized by: (i) dorsolateral folds absent (except for P. atratus); (ii) cranial crests absent (except for P. atratus, P. percultus, and P. spinosus); (iii) tympanic membrane and tympanic annulus prominent (both absent in P. philipi); (iv) dentigerous processes of vomer present; (v) small to prominent tubercles on heel; (vi) fingers with lateral fringes; (vii) basal webbing between toes (except for P. balionotus, P. hampatusami, and P. philipi which lack webbing); (viii) Toe V longer or much longer than Toe III (Figs 7–9); (ix) in life, groins and concealed surfaces of thighs with distinctive coloration patterns including flash colors and light or bright colored flecks or spots on a darker background; colors, shapes and sizes of these ornaments are variable among species; (x) SVL females 24.9–46.8 mm; SVL males 16.1–34.5 mm. Species of this clade may bear a fleshy keel or a papilla at the tip of the snout, and lateral, middorsal, or postocular folds.

Figure 7.

Palmar and plantar surfaces of Pristimantis atillo sp. nov., Pristimantis chomskyi sp. nov., Pristimantis gloria sp. nov., and Pristimantis jimenezi sp. nov. Photographs of left hand and foot of the holotypes A Pristimantis atillo sp. nov.: QCAZ 42500, male B Pristimantis chomskyi sp. nov.: QCAZ 47515, male C Pristimantis gloria sp. nov.: QCAZ 57201, female D Pristimantis jimenezi sp. nov.: QCAZ 45170, male. All specimens are shown at the same scale.

Figure 8.

Palmar and plantar surfaces of Pristimantis lutzae sp. nov., Pristimantis multicolor sp. nov., Pristimantis nangaritza sp. nov., and Pristimantis phoxocephalus. Photographs of hand and foot of the holotypes (except for P. phoxocephalus) of the following species: A Pristimantis lutzae sp. nov.: QCAZ 37546, female B Pristimantis multicolor sp. nov.: QCAZ 47213, male C Pristimantis nangaritza sp. nov.: QCAZ 41710, female D Pristimantis phoxocephalus: QCAZ 58463, female from the type locality. All specimens are shown at the same scale.

Figure 9.

Palmar and plantar surfaces of Pristimantis teslai sp. nov., Pristimantis torresi sp. nov., Pristimantis totoroi sp. nov., and Pristimantis verrucolatus sp. nov. Photographs of hand and foot of the holotypes of the following species: A Pristimantis teslai sp. nov.: QCAZ 46213, male B Pristimantis torresi sp. nov.: QCAZ 47342, female C Pristimantis totoroi sp. nov.: QCAZ 25105, male D Pristimantis verrucolatus sp. nov.: QCAZ 46982, male. All specimens are shown at the same scale.

Content

This clade comprises 23 described species (11 of them described below): P. atillo sp. nov., P. atratus, P. balionotus, P. chomskyi sp. nov., P. cryptomelas, P. gagliardoi, P. gloria sp. nov., P. hampatusami, P. jimenezi sp. nov., P. lutzae sp. nov., P. multicolor sp. nov., P. muscosus, P. nangaritza sp. nov., P. percultus, P. philipi, P. phoxocephalus, P. spinosus, P. teslai sp. nov., P. tinguichaca, P. torresi sp. nov., P. totoroi sp. nov., P. versicolor, P. verrucolatus sp. nov. This clade encompasses the P. phoxocephalus and P. cryptomelas species groups.

Distribution

Huicundomantis occurs in Eastern and Western Andean slopes and Inter-Andean valleys of southern and central Ecuador, and Eastern Andean slopes of northern Peru. They inhabit the following Natural Regions: Deciduous Costa Forest, Western Foothill Forest, Western Montane Forest, Paramo, Inter-Andean Shrub, Eastern Montane Forest, and Eastern Foothill Forest, between elevations of 230 and 4200 m a.s.l.

Etymology

We name this clade Huicundomantis because these frogs are frequently found inside bromeliad plants. Huicundo is a word in Quechua, an indigenous South American language, locally used to referring to bromeliads.

Pristimantis phoxocephalus species group new taxon

Definition. The P. phoxocephalus species group is strongly supported in our phylogeny. Members of this group share the following morphological traits: (i) dorsolateral folds absent (except for P. atratus); (ii) snout with a fleshy keel or papilla at the tip; (iii) cranial crests absent (except for P. atratus and P. percultus); (iv) tympanic membrane and tympanic annulus prominent; (v) dentigerous processes of vomer present; (vi) males with vocal slits (except for P. versicolor); (vii) fingers and toes with lateral fringes; (viii) basal webbing between toes (except for P. balionotus and P. hampatusami); (ix) Toe V longer or much longer than Toe III (Fig. 7); (x) in life, groins and concealed surfaces of thighs with distinctive coloration, including flash colors and light or bright colored flecks or spots on a darker background; colors, shapes, and sizes of ornaments variable among species; (xi) SVL females 24.9–46.8 mm; SVL males 16.8–34.5 mm.

Content. Currently, the P. phoxocephalus group comprises 17 described species (10 of them are described below): P. atillo sp. nov., P. atratus, P. balionotus, P. chomskyi sp. nov., P. gloria sp. nov., P. hampatusami, P. jimenezi sp. nov., P. lutzae sp. nov., P. multicolor sp. nov., P. percultus, P. phoxocephalus, P. teslai sp. nov., P. tinguichaca, P. torresi sp. nov., P. totoroi sp. nov., P. versicolor, P. verrucolatus sp. nov.

Distribution. Eastern and Western Andean slopes and Inter-Andean valleys of central and southern Ecuador, in ten provinces: Azuay, Bolívar, Cañar, Chimborazo, Cotopaxi, El Oro, Loja, Morona Santiago, Tungurahua, and Zamora Chinchipe. They inhabit Deciduous Costa Forest, Western Foothill Forest, Western Montane Forest, Paramo, Inter-Andean Shrub, and Eastern Montane Forest Natural Regions, between 230 and 4100 m a.s.l.

Remarks. The P. phoxocephalus species group is sister to the P. cryptomelas species group.

Pristimantis atillo sp. nov.

http://zoobank.org/078D7203-04D4-4509-9769-8DE14FF5B886

Common name

English: Atillo Rain Frog. Spanish: Cutín de Atillo.

Holotype

QCAZ 42500, an adult male from Sangay National Park, Ranger Station, Atillo, Morona Santiago Province, Ecuador (2.1867S, 78.4963W, 3400 m), collected by Elicio Tapia on May 19, 2009. Figures 10A, 11A.

Figure 10.

Holotypes of Pristimantis atillo sp. nov. and P. chomskyi sp. nov. Photographs of preserved holotypes of A P. atillo (QCAZ 42500, male) and B P. chompskyi (QCAZ 47515, male). Dorsal view on the left, ventral view on the right. Specimens are shown at the same scale.

Paratypes

(44: 28 males, 5 females, 11 juveniles). All from Sangay National Park, nearby the type locality. Ecuador: Chimborazo Province: QCAZ 2298–302, QCAZ 2304, adult males, QCAZ 2303, QCAZ 2305–306, juveniles (2.1749S, 78.5047W, 3600 m), collected by Giovanni Onore and Luis Coloma in October 1991; QCAZ 3110, adult female, QCAZ 3111–112, adult males (2.1749S, 78.5047W, 3600 m), collected by Luis Coloma in January 1991; QCAZ 40593, QCAZ 40594, adult females (2.2103S, 78.4991W, 3730 m), collected by Diego Almeida, Galo Díaz, Ítalo Tapia, and Martín Bustamante in May 2003. Morona Santiago Province: QCAZ 31946, adult male (2.1989S, 78.4824W, 3185m), collected by Charles M. Kieswetter and Ítalo Tapia in April 2006; QCAZ 40584, adult male (2.2024S, 78.4697W, 3366 m), collected by Diego Almeida, Galo Díaz, Ítalo Tapia and Martín Bustamante in May 2002; QCAZ 42501, adult female, QCAZ 40897, QCAZ 42483–486, QCAZ 42488–492, QCAZ 42499–500, QCAZ 42502–505, adult males, QCAZ 42493, QCAZ 42496, QCAZ 42498, QCAZ 42506, QCAZ 42512, QCAZ 42543, QCAZ 42548, juveniles (2.1867S, 78.4963W, 3400 m), collected by Elicio Tapia on May 18–19, 2009; QCAZ 59274, adult female, QCAZ 59214, QCAZ 59261, adult males (2.1842S, 78.4972W, 3436 m), collected by Francy Mora, David Velalcázar, Javier Pinto, Luis Tipantiza, Keyko Cruz, Daniel Rivadeneira, Freddy Almeida, Pol Pintanel, and Nadia Páez in March 2015. QCAZ 59228, adult male, QCAZ 59276, juvenile (2.1887S, 78.4823W, 3299 m), collected by Francy Mora, David Velalcázar, Javier Pinto, Luis Tipantiza, Keyko Cruz, Daniel Rivadeneira, Freddy Almeida, Pol Pintanel, and Nadia Páez in March 2015.

Diagnosis

A member of the Pristimantis phoxocephalus group having the following combination of characters: (1) skin on dorsum shagreen with or without scattered small subconical tubercles; middorsal fold ill-defined or absent; head with a middorsal longitudinal row of two or more subconical tubercles; dorsolateral folds absent; skin on venter areolate; discoidal fold present or absent; (2) tympanic membrane and tympanic annulus prominent, its upper and posterior margin concealed by supratympanic fold; (3) snout moderately long, acuminate with a fleshy keel in dorsal view, protruding in profile; (4) upper eyelid with one or more small prominent subconical tubercles surrounded by lower tubercles; cranial crests absent; (5) dentigerous processes of vomers low to prominent, oblique, moderately separated, posteromedial to choanae; (6) males having vocal slits, external vocal sac and white nuptial pads; (7) Finger I shorter than Finger II; discs of digits expanded to broadly expanded, rounded to elliptical; (8) fingers with broad lateral fringes; (9) ulnar tubercles low and rounded, sometimes connected by an ill-defined fold; (10) heel bearing one subconical tubercle surrounded or not by smaller rounded tubercles; outer edge of tarsus bearing low subconical tubercles; inner edge with or without ill-defined tubercles, short inner tarsal fold present; (11) inner metatarsal tubercle ovoid, elevated, five times the size of round outer metatarsal tubercle; supernumerary plantar tubercles numerous; (12) toes with lateral fringes, less conspicuous than those on fingers; basal webbing on feet (Fig. 7A); Toe V longer or much longer than Toe III (disc on Toe III reaches or exceeds distal edge of the penultimate subarticular tubercle on Toe IV, disc on Toe V reaches or exceeds distal edge of the distal subarticular tubercle on Toe IV); toe discs nearly as large as those on fingers; (13) in life, dorsal coloration varies from pale to dark brown; black or brown interorbital stripe, supratympanic stripe, and labial bars present; flanks with black dots and flecks arranged in a diagonal pattern, bordering light reticulations; groins, axils, concealed surfaces of thighs, and shanks orange; venter and throat cream to white; iris copper with a faint medial horizontal darker streak and black reticulations (Fig. 11); (14) average SVL in adult females: 31.3 ± 2.3 mm (29.4–35.3 mm; n = 5); in adult males: 24.7 ± 2.5 mm (17.5–28.1 mm; n = 29).

Figure 11.

Color variation in live individuals of Pristimantis atillo sp. nov. A QCAZ 42500 (holotype, male, SVL 26.3 mm) B QCAZ 42502 (male, SVL 20.6 mm) C QCAZ 59228 (male, SVL 23.7 mm) D QCAZ 59261 (male, SVL 25.3 mm). Dorsolateral view on the left, ventral view on the right.

Comparison with other species

Pristimantis atillo is similar to other species with acuminate and protruding snouts of this group such as P. jimenezi sp. nov., P. phoxocephalus, P. teslai sp. nov., P. torresi sp. nov., P. totoroi sp. nov., and P. verrucolatus sp. nov. The bright orange coloration of its groins and posterior surfaces of thighs distinguishes it from them (brown with small light brown to yellow spots in P. jimenezi sp. nov.; yellow with black reticulations in P. phoxocephalus; dark brown with yellow irregular blotches in P. teslai sp. nov.; brown with or without yellow spots in P. torresi sp. nov.; reddish brown with small light brown to yellow spots in P. verrucolatus sp. nov.). Pristimantis atillo can be further distinguished from P. jimenezi sp. nov. and P. phoxocephalus by the presence of its characteristic black dots in the flanks (black dots absent in P. jimenezi sp. nov. and P. phoxocephalus). Pristimantis atillo differs from P. teslai sp. nov. by the dorsal skin texture (shagreen in P. atillo; tuberculate in P. teslai sp. nov.). The copper coloration of the iris differentiates P. atillo from P. torresi sp. nov. and P. totoroi sp. nov., whose iris is golden with a red streak. Furthermore, tubercles and folds of P. totoroi sp. nov. are more prominent than those of P. atillo, and its head, relative to the body, is longer (males Wilcoxon’s Z = 4.50124, p < 0.001, HL/SVL = 33.5–38.5% in P. atillo, 34.2–38.7% in P. totoroi sp. nov.; females Z = 2.43599, p = 0.0149, HL/SVL = 33.1–36.6% in P. atillo, 35.2–37.6% in P. totoroi sp. nov.). Pristimantis verrucolatus sp. nov. differs from P. atillo in having large tubercles and warts on its flanks. Pristimantis atillo is similar to P. modipeplus (Lynch 1981) in coloration pattern; P. modipeplus has a subacuminate snout in dorsal view and rounded in profile, lacks basal webbing between toes, and large females bear low cranial crests (Lynch 1981); meanwhile, P. atillo has an acuminate and protruding snout with a keel at the tip, presents basal webbing between toes and has no cranial crests.

Description of the holotype

Adult male (QCAZ 42500, SC28232). Measurements (in mm): SVL 26.3; TL 12.0; FL 11.9; HL 8.9; HW 9.1; ED 2.6; TD 1.3; IOD 2.8; EW 2.4; IND 2.1; EN 2.7; TED 1.1. Head narrower than body, slightly wider than long; snout moderately long, acuminate with triangular fleshy keel in dorsal view, protruding in profile; nostrils slightly protuberant, directed laterally with slight dorsal inclination; canthus rostralis concave in dorsal view, rounded in cross section; loreal region slightly concave; cranial crests absent; upper eyelid bearing one small but distinct conical tubercle surrounded by few indistinct smaller tubercles; tympanic membrane and annulus prominent, its upper and posterior margins concealed by supratympanic fold; two enlarged conical postrictal tubercle surrounded by indistinct low tubercles. Choanae median, elliptical, non-concealed by palatal shelf of maxilla; dentigerous processes of vomers prominent, oblique, moderately separated, posteromedial to choanae; each vomer bearing several indistinct teeth; tongue slightly longer than wide, posteriorly notched, posterior two-thirds free; vocal slits slightly curved, located at posterior half of mouth floor in between tongue and margin of jaw; vocal sac present.

Skin on dorsum shagreen; head with a row of two middorsal tubercles; dorsolateral folds absent; flanks with slightly larger tubercles than those on dorsum; skin on throat, chest, belly and ventral surfaces of thighs areolate; discoidal fold absent. Two distinct, low and round ulnar tubercles connected by an ill-defined fold; white nuptial pads present; palmar tubercles prominent, outer palmar tubercle bifid, slightly bigger than ovoid thenar tubercle; subarticular tubercles prominent, rounded; distinct, low supernumerary tubercles at base of fingers; fingers with broad lateral fringes; Finger I shorter than Finger II; discs on fingers expanded and elliptical; ventral pads on fingers surrounded by circumferential grooves (Fig. 7A).

Hindlimbs slender; dorsal surfaces of hindlimbs with scattered low round tubercles; posterior surfaces of thighs smooth, ventral surfaces of thighs areolate; heel bearing one low conical tubercle surrounded by some lower rounded tubercles; outer edge of tarsus bearing low subconical tubercles; inner edge of tarsus with ill-defined tubercles; short inner tarsal fold present; inner metatarsal tubercle ovoid, elevated, five times the size of circular, rounded, outer metatarsal tubercle; plantar surface with numerous indistinct supernumerary tubercles; subarticular tubercles prominent, rounded; toes with broad lateral fringes; small basal webbing between toes; discs nearly as large as those on fingers; discs on toes expanded, elliptical; all toes having ventral pads and circumferential grooves; relative lengths of toes: I<II<III<V<IV; Toe V much longer than Toe III (disc on Toe III reach distal edge of penultimate subarticular tubercle on Toe IV, disc on Toe V reaches distal edge of distal subarticular tubercle on Toe IV; Fig. 7A). Coloration of the holotype in preservative and life is shown in Figures 10, 11.

Variation

This section is based on 45 preserved specimens of the type series and photographs available for 11 individuals. Variation in living and preserved individuals is shown in Figures 11, 12. Coloration in life is mentioned in parenthesis. Dorsum varies from pale to dark brown with or without lighter irregular marks or white spots scattered on dorsum, few individuals have several thin parallel longitudinal stripes or a middorsal band (i.e., longitudinal pattern); the interorbital band can be thin to broad, paler or darker than background and is absent in individuals with longitudinal pattern; interscapular blotch is often present. Flanks have pale oblique reticulations usually surrounded by dark stripes formed by rows of black flecks. Groins, anterior and posterior surfaces of thighs cream (orange, in most cases surrounded by yellow blotches; an individual has yellow groins (QCAZ 42502)). Dorsal surfaces of thighs with brown to black oblique stripes. Venter and throat coloration varies from cream to white (cream to white with different levels of transparency, with or without ill-defined yellow blotches) with or without black flecks. Iris is reddish to orangey copper with a faint medial horizontal darker streak and thin black reticulations; sclera varies from white to light blue.

Figure 12.

Color variation in preserved individuals of Pristimantis atillo sp. nov. A Dorsal view of (from left to right): QCAZ 40593 (female), QCAZ 42501 (female), QCAZ 59274 (female) B dorsal view of: QCAZ 42498 (juvenile female), QCAZ 42503 (male), QCAZ 2302 (male), QCAZ 42548 (juvenile female) C ventral view of specimens in (A) D ventral view of specimens in B. See Suppl. material 2 for locality data. All specimens are shown at the same scale.

Coloration of holotype in preservative. Dorsum brown with lighter markings including W-shaped scapular mark, interscapular blotch, and faint irregular reticulations; head with light brown interorbital band; face with faint canthal and labial bars; dark brown supratympanic bar; flanks light brown with pale oblique reticulations surrounded by rows of black flecks; armpits, groins, anterior and posterior surfaces of thighs cream; dorsal surfaces of hindlimbs brown with lighter transversal bands; limbs with scattered black flecks; ventral surfaces cream; plantar and palmar surfaces dirty cream (Fig. 10A).

Coloration of holotype in life. Based on studio photographs (Fig. 11A). Dorsum brown with cream light markings including a W-shaped scapular mark, interscapular blotch, and irregular reticulations; head with light brown interorbital band; face with faint brown canthal and labial bars; dark brown supratympanic bar; flanks light brown with yellow reticulations surrounded by rows of black flecks; armpits, groins, anterior and posterior surfaces of thighs and shanks orange; dorsal surface of hindlimbs with faint yellowish brown transverse bands and scattered black flecks; venter yellowish white; vocal sac, chest, and ventral surfaces of forelimbs surfaces transparent pinkish cream; ventral surface of thighs yellow suffused with orange; plantar and palmar surfaces dirty cream; iris copper with a faint medial horizontal darker streak and thin black reticulations; light blue sclera.

Distribution, natural history, and conservation status

This species is only known from the type locality, surroundings of Lagunas de Atillo, and nearby locations at Sangay National Park, between 3185 and 3730 m a.s.l (Fig. 1). This corresponds to the Eastern Montane Forest and Paramo biogeographic regions. Most individuals collected at night were found active on low vegetation, from ground level up to 1 m above the ground; an individual was collected from a bromeliad 2.5 m above the ground. During the day, they were found beneath rocks. Most collections were made on roadsides. Calling males have been found in March.

According to available data, this species has a very restricted distribution (Extent of Occurrence 7 km², Area of Occupancy 16 km²). However, less accessible adjacent areas in Sangay National Park are unexplored and represent potential distribution areas for this species. Therefore, we consider this species to be Data Deficient (IUCN 2017).