A spectacular new species of Hyloscirtus (Anura: Hylidae) from the Cordillera de Los Llanganates in the eastern Andes of Ecuador

Ethics statement

Our study was authorized under research permits MAE-DNB-CM-2016-0045 and MAE-DNB-CM-2019-0120, issued by the Ministerio del Ambiente del Ecuador. We followed standard guidelines for use of live amphibians and reptiles in field research (Beaupre et al., 2004), compiled by the American Society of Ichthyologists and Herpetologists, the Herpetologists’ League, and the Society for the Study of Amphibians and Reptiles.

Taxon sampling

We examined specimens deposited in the herpetological collections of the Instituto Nacional de Biodiversidad, Quito (DHMECN) and Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogotá (ICN) (Appendix 1). All museum acronyms follow Sabaj (2016). Our taxonomic description employs several lines of evidence, including external morphological characters, genetic divergence, monophyly and preliminary geographic data. Similar approaches have been useful in recognizing and identifying closely related species of anurans in the eastern Andes of Ecuador (Páez-Moscoso, Guayasamin & Yánez-Muñoz, 2011; Reyes-Puig et al., 2019a, 2019b).

The electronic version of this article in Portable Document Format (PDF) will represent a published work according to the International Commission on Zoological Nomenclature (ICZN), and hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved, and the associated information viewed through any standard web browser, by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is:

urn:lsid:zoobank.org:pub:4BF8C735-F06C-41AE-B130-EE41130535CC

The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central and CLOCKSS.

Field work

Two individuals were found fortuitously in the same spot during diurnal walks in botanical expeditions to the summit of Cerro Mayordomo (1.3702 S, 78.2679 W, 2970 m) on 16–20 March 2018 and 18–19 October 2018. Both were collected. A third individual, photographed in situ but not collected, was found in the same spot in December 2019, and a fourth individual was found and collected in the same area in May 2022. Several additional expeditions to the same location failed to find individuals of this species.

Laboratory work

The two original collected individuals of the new species were taken alive, in plastic containers, to INABIO, where they were photographed in life and euthanized with benzocaine. Tissue samples were then taken for DNA sampling. They were subsequently fixed in 10% formalin for twelve hours, and then preserved as voucher specimens in 70% ethanol following the recommendations of Heyer et al. (1994). These specimens were deposited in the herpetological collection (DHMECN) of INABIO as holotype and paratype. The third collected individual, an additional paratype, is being kept alive for observation and analysis at INABIO and will be deposited in the same collection.

External morphological data

Measurements and character descriptions were made according to the specialized literature treating the H. larinopygion group (Coloma et al., 2012; Almendáriz et al., 2014; Ron et al., 2018). Description of webbing formulae of the hands and follow Savage & Heyer (1967) as modified by Myers & Duellman (1982). We obtained morphological measurements of the two specimens preserved in 70% ethanol according to the methodology described in Duellman (1973), using digital calipers (±0.01 mm). The following measurements were taken: snout-vent length (SVL), head length (HL), head width (HW), upper eyelid width (EW), interorbital distance (IOD), inter-nostril distance (IND), eye-nostril distance (END), eye diameter (ED), tympanum diameter (TD), hand length (HAL), tibia length (TL), femur length (FEL), and foot length (FL). Sex was determined by direct examination of gonads.

We also compared qualitative morphological characters between the new species and its closest relatives. Seven characters were evaluated: (1) dorsal coloration; (2) ventral coloration; (3) marks on flanks and hidden surfaces of thighs; (4) iris coloration; (5) prepollex condition; (6) in life, webbing coloration; and (7) cloacal ornamentation. Life coloration was obtained from live specimens and color photographs. Cloacal ornamentation condition was observed on live and preserved specimens.

Osteological data and analysis

The holotype (DHMECN 14416) of the new species, and one specimen of each of five closely related species (DHMECN 12483: Hyloscirtus lindae; DHMECN 12111: H. pacha; DHMECN 6493: H. psarolaimus; DHMECN 3799: H. larinopygion; DHMECN 9686: H. tapichalaca), were scanned using a high-resolution micro-computed tomography (micro-CT) desktop device (Bruker SkyScan 1173; Bruker, Kontich, Belgium) at the Leibniz Institute for the Analysis of Biodiversity Change—Museum Koenig (LIB Bonn, Germany). To avoid movements during scanning, the specimens were placed in a small plastic container and mounted with styrofoam. The scans were conducted over 180 degrees with rotational steps of 0.3‒0.4 degrees, with a source voltage of 35 kV and source current of 150 µA, without the use of a filter, at an image resolution of 39.3‒50.0 µm. Scan duration was 30:01‒45:37 min with an exposure time of 280 ms and average rate of 5 frames per second. The micro-CT datasets were reconstructed using N-Recon software (Bruker MicroCT, Kontich, Belgium) and rendered in three dimensions through the aid of CTVox for Windows 64 bits version 2.6 (Bruker MicroCT, Kontich, Belgium). Additionally, the skull of the holotype of the new species was rendered and segmented to separate and color individual bones in three dimensions using Amira visualization software (FEI, Thermo Fisher Scientific, Waltham, MA, USA). Osteological terminology follows Trueb (1973), Duellman & Trueb (1994), Coloma et al. (2012), Kunisch et al. (2021), Reyes-Puig et al. (2021). For the description of the cranium and the osteology of the hand, we followed the proposal of Coloma et al. (2012). Cartilage structures were omitted from the osteological descriptions, because micro-CT does not render cartilage. To facilitate comparisons among skull bones, we added color to the micro-CT scan images using Adobe Photoshop.

Genetic sampling

We generated two new sequences, one from the individual collected in March 2018 (holotype DHMECN 14416) and the other from the individual collected in October 2018 (paratype DHMECN 14549), for the mitochondrial 16S gene (see Fig. 1), following the primers and protocols described in Guayasamin et al. (2015). The new sequences (GenBank OM293945, OM293945) were aligned with all sequences available for Hyloscirtus in GenBank (http://www.ncbi.nlm.nih.gov/genbank), originally published by Faivovich et al. (2005), Coloma et al. (2012), Almendáriz et al. (2014), Guayasamin et al. (2015) and Ron et al. (2018). GenBank codes of downloaded species are shown in Fig. 1.

Phylogenetic analysis

Sequences were aligned using MAFFT v. 7 (Katoh & Standley, 2013) with the Q-INS-i strategy. Maximum likelihood (ML) trees were estimated using GARLI 2.01 (Genetic Algorithm for Rapid Likelihood Inference; Zwickl, 2006). GARLI uses a genetic algorithm that finds the tree topology, branch lengths and model parameters that maximize ln(L) simultaneously (Zwickl, 2006). In order to determine the outgroups for our analyzes, we conducted preliminary runs in GARLI and selected as outgroups those species that were inferred as most phylogenetically distant to the Hyloscirtus larinopygion group. With this information, we selected species in the H. bogotensis group as outgroups. During the ML analyses, individual solutions were selected after 10,000 generations with no significant improvement in likelihood, with the significant topological improvement level set at 0.01. The final solution was selected when the total improvement in likelihood score was lower than 0.05, compared to the last solution obtained. Default values were used for other GARLI settings, as per recommendations of the developer (Zwickl, 2006). Bootstrap support was assessed via 1,000 pseudoreplicates under the same settings used in tree search. Genetic distances (uncorrected p) between the new species and its closest relatives were calculated using PAUP v.4.0a (Swofford, 2002).

Divergence time estimation

The hybrid time-calibrated phylogeny of Coloma et al. (2012) for the H. larinopygion group, combining external fossil dates and estimated substitution rates, was used to calculate a regression line and correlation coefficient between genetic distances and divergence times for species-pairs in this group. The smallest estimated genetic distance between our new species and the other species in this group can then be inserted into this linear equation to estimate the minimum divergence time between the new species and the other sequenced species in the group. Approximate uncertainties in our estimate of divergence time were based on the uncertainty bars in Fig. 5 of Coloma et al. (2012).

Ecological niche modeling

We use Maxent (version 3.4.2) to obtain a model of the range of ecological niches for the northern clade of the H. larinopygion group. Localities for all species of the group were obtained from literature and museum collections. Recommended default values were used for convergence threshold, maximum number of iterations, and maximum background points; 25% of localities were randomly set aside as test points; regularization was set to 1. Selected format for representation of probabilities for models was logistic. Parametrization was based on WorldClim (version 2.1, Fick & Hijmans, 2017). Statistical analyses of variable contributions for data layers, including jackknife tests and correlation tests, were used to obtain more informative and less correlated variables. Models were evaluated through quantitative and qualitative tests, including threshold-independent test, threshold-dependent test, visual evaluations, and evaluation of variable importance and response curves. A geographical information system was developed based on grids from Maxent with ArcGis Desktop to analyze data and produce relevant maps.

Phylogenetic relationships

Our phylogenetic analysis (Fig. 1) shows that the new species is sister to a clade containing ten Hyloscirtus species: H. criptico (Coloma et al., 2012), H. larinopygion (Duellman, 1973), H. lindae (Duellman & Altig, 1978), H. pacha (Duellman & Hillis, 1990), H. pantostictus (Duellman & Berger, 1982), H. princecharlesi Coloma et al. (2012), H. psarolaimus (Duellman & Hillis, 1990), H. ptychodactylus (Duellman & Hillis, 1990), H. staufferorum (Duellman & Coloma, 1993), and H. tigrinus (Mueses-Cisneros & Anganoy-Criollo, 2008). However, we note that support for the exact topology of this relationship is low (bootstrap = 54%).

Genetic distances (mitochondrial 16S percent differences calculated from uncorrected p values) between the new species and the most closely related Hyloscirtus are given in Table 1. The lowest values of genetic distances between the new species and its relatives were 2.2‒2.9% to H. tigrinus and 2.6‒2.8% to H. ptychodactylus (Table 1).

Most DNA sequences are publicly available (see GenBank codes in Fig. 1). The sequences of the new species are available as Supplemental Files.

Estimation of divergence time

We confirmed our theoretical expectation that the fossil-calibrated divergence times of sequenced species-pairs within the H. larinopygion group (Coloma et al., 2012) were linearly correlated with their genetic distances (Fig. 2), with R² = 0.92 (regression line forced to go through origin) or 0.94 (unconstrained regression line). The latter correlation equation is y = 2.74x + 2.17 where x is the genetic distance (expressed as a percent) and y is the divergence time (expressed in millions of years). Hyloscirtus sethmacfarlanei sp.nov. was not known at the time of that study, but we can use this regression line with our measured genetic distances to estimate the divergence times between H. sethmacfarlanei sp. nov. and its relatives. Based on the mean genetic distance of 2.55 % between the new species and its least-distant relative H. tigrinus, we estimate that the divergence time between the new species and the other sequenced species is approximately 9.1 Mya +/− 4 My.

Systematic account

Hyloscirtus sethmacfarlanei sp. nov.

Proposed standard Spanish name: Rana de torrente de Seth MacFarlane

Proposed standard English name: Seth MacFarlane’s torrent frog

urn:lsid:zoobank.org:pub:4BF8C735-F06C-41AE-B130-EE41130535CC

Holotype (Figs. 3, 4). DHMECN 14416, adult gravid female, collected in the Machay Reserve of Fundacion EcoMinga, Cerro Mayordomo (1.370204 S, 78.267943 W, 2,970 m), Rio Verde parish, Baños township, Tungurahua province, Republic of Ecuador, on 17 March 2018, by Darwin Recalde, Fausto Recalde, Santiago Recalde, and Jordy Salazar.

Paratypes (Figs. 5–8). DHMECN 14549, juvenile, sex undetermined, without differentiation of external and internal sexual characters, collected at the type locality on 19 October 2018, by Fausto Recalde, Santiago Recalde, Darwin Recalde and Jordy Salazar; DHMECN 17554, juvenile sex undetermined, no differentiation of external and internal sexual characters, collected at the type locality on 30 May 2022, by Fausto Recalde, Luis Recalde, and Santiago Recalde.

Generic placement. We assign the new species to the genus Hyloscirtus Peters, 1882, defined according to Faivovich et al. (2005) and Rojas-Runjaic et al. (2018), and to the H. larinopygion group (sensu Duellman & Hillis, 1990; Faivovich et al., 2005), according to its phylogenetic position (Fig. 1) and morphological traits such as wide dermal fringes on fingers and toes, hands and legs with large terminal discs and a reduced membrane, adults characterized by a snout vent length >60 mm, and dark overall skin color contrasting with bright color patterns.

Diagnosis. Hyloscirtus sethmacfarlanei sp. nov. is a member of the Hyloscirtus larinopygion group as diagnosed by Duellman & Hillis (1990), Faivovich et al. (2005) and Weiens et al. (2005), and differs from other members of the group by the following combination of characters: discs of digits narrow; fleshy calcar present cloacal ornamentation with two thick well-defined parallel paracloacal grooves; a well-defined supracloacal fold reaching the vent; skin surrounding cloaca strongly areolate and granular; anterior border of sphenethmoid not in contact with nasal; nasal not in contact with maxilla; frontoparietals rugose; vomers not in medial contact, and with 12‒13 tooth loci; 54‒56 maxillary tooth loci; 10‒11 premaxillary tooth loci; zygomatic ramus of squamosal slightly longer than otic ramus, and otic ramus not in contact with prootic. The adult female, in its reproductive stage with internal sexual characters defined, is further characterized by black ground color covered with large bright red spots on both the dorsal and ventral surfaces, and red tips on all digits.

Description of holotype (Figs. 3, 4). Adult female, SVL 72.0 mm. Body slender, head rounded in dorsal view, longer than wide (head length 113% of head width); width of upper eyelid 72% of the interorbital distance; texture of the dorsal surface of the head rough, including the eyelids; snout truncate in dorsal and lateral views; eye-nostril distance slightly less than the diameter of the eye; canthus rostralis short and slightly rounded, loreal region slightly concave; internarial region flat and slightly depressed; top of head slightly concave; nostrils oval and slightly protuberant, directed laterally; eyes large and protuberant, 25% of head length; interorbital region concave; eye diameter 1.8 times larger than the diameter of the tympanic ring; supratympanic fold well-defined, directed obliquely from the posterior border of the eye, covering the dorsal edge of the tympanum, extending back to the upper shoulder; tympanum and tympanic ring evident and round, 57% of eye diameter, separated from the eye by a distance 1.6 times larger than the diameter of the tympanum.

Anterior and posterior extremities slim. Relative length of fingers I < II < IV < III; fingers with large oval disks slightly wider than finger; subarticular tubercles simple and enlarged, round and prominent; multiple round and oval supernumerary tubercles present; thenar tubercle large and flat, oval and elongated; palmar tubercle asymmetric with a slightly heart-shaped outline; prepollex absent; glandular nuptal pad covering the outer margin of Finger I; fingers long with interdigital webbing basally and with fleshy lateral fringes on all fingers.

Hind limbs long and slender, tibia length 46% of SVL; foot length 46% of SVL; heel tubercles large and round in outline; inner tarsal fold absent; large rounded to slightly oval subarticular tubercles in all fingers, supernumerary foot tubercles rounded, low; toes long, narrower than the disc, discs not expanded; relative lengths of toes I < II < V< III < IV; inner metatarsal tubercle large, oval; outer metatarsal tubercle absent; toes with interdigital membrane, toe membrane formula: I 2-3 II 3- 2 III 3-2 IV 3-2 V (Fig. 4).

Body skin is finely granular, especially on flanks; inguinal glands absent; ventral skin densely areolate, less so towards the throat. Supracloacal flap transversal, well-defined, with supracloacal fold present, reaching the level of the vent, with two paracloacal folds; skin around the cloaca strongly areolate and granular (Figs. 9, 10). Choana large and oval, notably separated from each other and perpendicular to the floor of the mouth; dentigerous processes of vomers transverse, with vomerine teeth numbering 9‒10; tongue wide and rugose, slightly rounded, partially attached to the floor of the mouth.

Coloration of holotype in life (Figs. 5–8, 11). Entire dorsal and ventral surfaces of the head, body, and limbs black with large bright red round to oval spots scattered over the whole body, including the tips of the digits; spots 3‒4 mm in diameter on dorsal surface of body and 5‒10 mm long on ventral surface and throat, more elongated on the extremities and flanks. Iris light grayish with fine dark reticulations, while the nictitating membrane, revealed in defense and at rest, is well-developed, black in color, with irregular red reticulations.

Coloration of holotype in preservative (~70% ethanol) (Fig. 3). Mainly black background; the red spots in life fade to yellowish-white or white; ventral surfaces and throat grayish black with scattered irregular white elongated spots; palms of hands and feet grayish.

Measurements of the holotype (in mm). SVL = 72.0; HL= 22.9; HW = 20.2; EW = 6.0; IOD = 8.3; IND = 5.2; END = 5.4 ; ED = 5.7; TD = 3.2; HAL = 25.2; TL = 33.3; FEL = 26.1; FL = 33.4.

Measurements of the paratypes. See Table 2.

Osteology of the preserved holotype (Figs. 12–15). Coloma et al. (2012) provide a detailed description of the osteology of the H. larinopygion group. In order to avoid redundancy, in the following we describe only those osteological features of the holotype of H. sethmacfarlanei sp. nov. where we found differences from the other species.

Skull (Figs. 11, 12). The anterior border of the sphenethmoid is not in contact with the nasal; the nasal is not in contact with the maxilla; the frontoparietals are rugose; the paired vomers bear 12‒13 tooth loci and are not in contact medially; there are 54‒56 tooth loci on each maxilla and 10‒11 tooth loci on each premaxilla; the zygomatic ramus of the squamosal is slightly longer than the otic ramus, and the latter is not in contact with the prootic.

Posteromedial processes of the hyobranchium (Fig. 15). The posteromedial processes of the hyobranchium are paired ossified structures, longer than broad, the anterior portion with triangular “head-like” shape, and a posterior elongated stem.

Tadpole. Not known.

Advertising call. Not known.

Variation (Figs. 5, 6, 10 and 11). Standard measurements from the three collected individuals are shown in Table 2. The three known juveniles (DHMECN 14549, DHMECN 17554, and the uncollected individual) share the distinctive cloacal ornamentation and skin texture of the holotype, but differ from the female holotype as follows: sexual characters not clearly evident (based on the observation of internal anatomy); in life the dorsal surface with irregular marks mustard yellow heavily stippled with black, especially on flanks and lower back (DHMENC 14549), or a variagated yellow-orange pattern (DHMECN 17554); nictitating membrane dotted with mustard yellow on a gray background (DHMENC 14549) or orange on a black background (DHMECN 17554); extremities orange banded (DHMENC 14549) or spotted (DHMECN 17554), on a grayish black to black ground; flanks black with orange reticulations and irregular spots; throat marbled with irregular orange or yellowish patches with orange tones on a grayish black or black ground; belly and ventral surfaces of the extremities grayish black with irregular sparse diffuse light orange or whitish-yellow patches (DHMENC 14549) or solid orange (DHMENC 17554); palms of hands and feet black with diffuse light orange spots. The uncollected juvenile had a mainly yellow dorsal coloration, with diffuse blackish spots scattered on the flanks and hidden surfaces of the arms and between the fingers, whose tips were yellow. The belly is light cream with diffuse blackish spots. We noted rapid temporal chromatic changes in the juvenile individuals, from dull yellow to orange tones. As observed in other members of the Hyloscirtus larinopygion species group, changes in color pattern may be characteristic of different stages of development and related to ontogenic changes (Coloma et al., 2012). The juveniles all shared the same distinctive cloacal ornamentation as the adult (Fig. 10).

Comparison with similar species (Figs. 8, 9, Table 3). The black and red pattern of the female of the new species is most similar to the patterns of Hyloscirtus pantostictus (Duellman & Berger, 1982), from extreme northeastern Ecuador, and H. princecharlesi Coloma et al. (2012), from the Pacific slopes of the Andes of northwestern Ecuador. The new species differs from these in having both the dorsal and ventral surfaces spotted with red (vs ventral surface without red spots in H. pantostictus and H. princecharlesi, Fig. 8), the cloacal ornamentation (Fig. 9) consisting of a well-defined supracloacal fold reaching next to the vent and the presence of a paracloacal fold (vs reduced supracloacal fold, without paracloacal fold, not contacting the side of the vent, in H. pantostictus, and supracloacal fold defined, reaching the border of the vent, with paracloacal fold thick, in H. princecharlesi), and strongly areolate skin texture (vs smooth in H. pacha, H. staufferorum, and H. larinopygion). The female of the new species also differs from these two species in having red discs on the tips of all digits (vs yellow discs in H. pantostictus and grayish discs in H. princecharlesi).

The new species’ red discs are shared with H. lindae (Duellman & Altig, 1978) from the eastern Andes, but H. lindae does not have red spots on its dorsal surface and does not have a thick supracloacal fold close to the side of the vent (Fig. 9). Juveniles assigned to H. sethmacfarlanei sp. nov. have a pattern similar to those of H. princecharlesi and H. larinopygion (Duellman, 1973) from northwestern slopes of the Andes. They differ from juveniles of both species in having the dorsum mottled and stippled mustard-yellow and black (vs dorsum densely spotted orange-red in H. princecharlesi, and yellowish-brown with distinctive cream bars with black interspaces in H. larinopygion). The supracloacal fold is well-defined and reaches to the vent in H. sethmacfarlanei sp. nov. (vs faintly defined and distant from the side of the vent in H. larinopygion). Hyloscirtus sarampiona (Ruiz-Carranza & Lynch, 1982) from the western slopes of the Colombian Andes has dorsal surfaces orange spotted with pale olive, and further differs from the new species by having hidden areas of the limbs, flanks, palmar, plantar surfaces and discs black.

The skull of H. sethmacfarlanei sp. nov. (Figs. 12, 13) is generally consistent with those of the other species of the H. larinopygion group (Coloma et al., 2012). However, there were a few differences between the new species and the species of the group studied by us or by Coloma et al. (2012). In H. sethmacfarlanei sp. nov. and H. ptychodactylus, the sphenethmoid is not in contact with the nasal, whereas these two bones are in contact in H. criptico and in H. staufferorum, they are anteriorly in contact in H. lindae and H. larinopygion, they are in contact along most of their length in H. psarolaimus, they are in contact along their entire length but with still a visible suture in H. pacha, and they are completely fused without a visible suture in H. tapichalaca. In H. sethmacfarlanei sp. nov., H. lindae, H. pacha, and H. larinopygion the nasal is not in contact with the maxilla, whereas it is in contact with the maxilla in H. criptico, H. pantostictus, H. ptychodactylus, H. staufferorum, and H. tapichalaca. The frontoparietals of H. sethmacfarlanei sp. nov. are comparatively more rugose than in other species of the group (Fig. 12). In contrast to H. pantostictus and H. staufferorum, in H. sethmacfarlanei sp. nov., H. criptico, H. lindae, H. pacha, H. psarolaimus, H. ptychodactylus, H. larinopygion, and H. tapichalaca the otic ramus of the squamosal is not in contact with the prootic. In H. sethmacfarlanei sp. nov. the zygomatic ramus of the squamosal is only slightly longer than otic ramus, whereas it is moderately longer than the otic ramus in H. pacha and H. staufferorum, and distinctly longer than the otic ramus in H. criptico, H. lindae, H. pantostictus, H. ptychodactylus, H. larinopygion, and H. tapichalaca. In contrast to H. criptico, H. larinopygion, and H. tapichalaca, in H. sethmacfarlanei sp. nov., H. lindae, H. pantostictus, H. ptychodactylus, and H. staufferorum the vomers are not in medial contact. Hyloscirtus sethmacfarlanei sp. nov. has 12‒13 vomerine tooth loci, 54‒56 tooth loci on each maxilla, and 10‒11 tooth loci on each premaxilla, whereas we counted 14 vomerian tooth loci, 59‒60 maxillary tooth loci, and 11‒12 premaxillary tooth loci in H. lindae, 14 vomerine tooth loci, 52‒59 maxillary tooth loci, and 9 premaxillary tooth loci in H. pacha, 13‒14 vomerine tooth loci, 52‒53 maxillary tooth loci, and 9 premaxillary tooth loci in H. psarolaimus, 11‒12 vomerine tooth loci, 56 maxillary tooth loci, and 12 premaxillary tooth loci in H. larinopygion, and only 5‒6 vomerine tooth loci, 31‒33 maxillary tooth loci, and 5‒6 premaxillary tooth loci in H. tapichalaca.

In the new species the posteromedial processes of the hyobranchium possess a triangular shaped anterior portion, and a shorter posterior portion compared with the other species shown in Fig. 14, which have an external round border and an internal spine-like border. In H. lindae, H. psarolaimus and H. pacha, the anterior portions have rounded external and internal borders. In H. tapichalaca it is broad and “shell-like” in its anterior border.

There are no relevant differences between the forelimb bones of the new species and those of the other species in the group, with the exception of male specimens of H. tapichalaca, which have a greatly enlarged prepollex (Kizirian, Coloma & Paredes-Recalde, 2003; this study) compared to the other species of the H. larinopygion group.

Distribution (Fig. 16). Hyloscirtus sethmacfarlanei sp. nov. is known at the moment only from the type locality in Fundación EcoMinga’s Machay Reserve, Cerro Mayordomo, 2,970 m altitude, in the eastern cordillera of the central Ecuadorian Andes, in the northern side of the upper Rio Pastaza watershed near the southern border of Llanganates National Park in the province of Tungurahua.

Natural history. The type locality consists of dwarf open mossy forest, covered with bryophytes and epiphytes, and saturated with humidity. All four known individuals of this species were found on a single narrow mountain ridge, in bromeliads of the genus Guzmania growing within 60–90 cm above the ground (Fig. 17). The holotype is an adult gravid female with a mass of eggs in early stage of development in March 2018. Adult male, tadpole and advertisement call remain unknown.

There is some evidence that the striking coloration of the adult female of H. sethmacfarlanei sp. nov. could be aposematic. The frog’s discoverer (Darwin Recalde), after briefly handling the frog, noticed an unpleasant tingling sensation down his arm, not restricted to the area which had contacted the frog; the sensation lasted several hours. Fausto Recalde, who had shorter contact with the frog, developed similar but shorter-lasting symptoms; these reactions were not observed after handling the juveniles. During handling of the holotype specimen in the museum, it emitted a white exudation on dorsal surfaces with a distinctive odor similar to diluted contact cement. Additionally, when tissue was taken from the liver, dark blackish-colored blood was observed.

The bright yellow uncollected juvenile slept during the day, and when disturbed, it adopted a defensive ball-like position, as observed in other species of the H. larinopygion group (Kizirian, Coloma & Paredes-Recalde, 2003; Bejarano-Muñoz, Perez Lara & Brito Molina, 2015). Thus the juvenile coloration may also advertise distastefulness.

Nocturnal surveys done by our team in the habitat of Hyloscirtus sethmacfarlanei sp. nov. revealed three sympatric anuran species: two undescribed Pristimantis species and one species of the Pristimantis buckleyi complex.

Conservation Status. All four known individuals are known only from the same few square meters of ridgeline, but the area is poorly studied and inaccessible because of steep topography. We suggest the IUCN category Data Deficient (DD) for this species.

Etymology. The specific epithet sethmacfarlanei is a patronym in honor of Seth MacFarlane, American writer, director, producer, actor, artist, musician and conservationist, with an outstanding passion for science, biodiversity and the natural world.