Nesorohyla, Pinheiro & Kok & Noonan & Means & Haddad & Faivovich, 2019

NESOROHYLA View in CoL GEN. NOV.

Type species: Hyla kanaima Goin & Woodley, 1969 .

Diagnosis: (1) Enlarged prepollex, not modified as a projecting spine; (2) reduced fringes on fingers and toes; (3) nuptial pads present, light-coloured, on inner margin of Finger II (digits numbered as Fabrezi & Alberch, 1996) and inner metacarpal tubercle; (4) two small calcar tubercles; (5) dorsal coloration overall brownish; (6) iris of adults are black under natural light conditions; (7) tadpole oral disc with short, anterior and posterior gaps on marginal papillae; (8) three emarginations on the posterior labium; (10) labial tooth row formula ( LTRF) 2(2)/4[1]; (11) eggs entirely pigmented.

Included species: Nesorohyla kanaima ( Goin & Woodley, 1969) new combination.

Sister taxon: Obtained in alternative, weakly supported (<50% jackknife) positions as the sister taxon of Myersiohyla or as the sister taxon of Hyloscirtus + Bokermannohyla + Aplastodiscus + Boana .

Etymology: The name Nesorohyla is derived from the combination of Greek roots nesos (island) and oros (mountain), and the classical genus Hyla , meaning Hyla from the mountain island. It is an allusion to the geologic formation where the species is found: the tepuis in northern South America, which are usually referred to as altitude islands. The gender is feminine.

Characterization: The only species included in this genus was also characterized by Duellman & Hoogmoed (1992) and MacCulloch & Lathrop (2005).

Comparison with other genera of Cophomantini : From the characters employed in the diagnosis, the only putative autapomorphies of Nesorohyla so far seem to be the completely pigmented ovum ( Duellman & Hoogmoed, 1992; Faivovich et al., 2013) and, possibly, the coloration of the iris in adults (described as black by MacCulloch & Latrop, 2005). Field experience of some of the authors shows that under natural light conditions, the iris of adults appears to be black. This is because it is so densely coloured with melanin that it is not distinguishable from the pupil, unless one shines a bright light on it. For this reason, we refer to the iris of Nesorohyla kanaima as being black. The value of the generic diagnoses that are not based on synapomorphies is extremely limited, as they are of actual use only for species already known to be included in those genera—and on the basis of which the generic diagnosis is made—but have no predictive value for the inclusion of new species. We provide below a comparison of Nesorohyla , based on the diagnostic characters, with all other genera of Cophomantini .

The occurrence of completely black ova and black iris in adults differentiates Nesorohyla from all other genera of Cophomantini . The reduced fringes on fingers and toes differentiate Nesorohyla from all species of Hyloscirtus ( Faivovich et al., 2005; Rivera-Correa & Faivovich, 2013). The absence of a prepollical spine differentiates Nesorohyla from most species in the genera Boana and Bokermannohyla ( Faivovich et al., 2005) . The small, light-coloured nuptial pad differentiates Nesorohyla from Myersiohyla (dark coloured simple or double nuptial pads covering the medial margin of Finger II, including Prepollex and Metacarpal II; Faivovich et al., 2013). The calcar with two small tubercles distinguishes Nesorohyla from Myersiohyla [calcar absent in M. aromatica ( Ayarzagüena & Señaris, 1994) , M. chamaeleo , M. inparquesi ( Ayarzagüena & Señaris, 1994) , M. liliae , and M. neblinaria ; calcar as a transversal ridge on the heel in M. loveridgei ( Rivero, 1961; Ayarzagüena & Señaris, 1994; Kok, 2006; Faivovich et al., 2013)]. The overall dorsal brownish coloration differentiates Nesorohyla from Aplastodiscus and most species in the Hyloscirtus bogotensis group (overall green, with only one brown species in Aplastodiscus ; Guayasamin et al., 2015; Berneck et al., 2016).

The tadpoles of Nesorohyla can be distinguished from those of Myersiohyla by having an oral disc with short anterior and posterior gaps on marginal papillae (gaps absent in M. aromatica , M. chamaeleo , M. inparquesi and M. neblinaria ; Ayarzagüena & Señaris, 1994; Faivovich et al., 2013); three emarginations on the posterior labium: one medial, two posterolateral (emarginations absent in M. aromatica , M. chamaeleo and M. inparquesi ; Ayarzagüena & Señaris, 1994; Faivovich et al., 2013); LTRF 2(2)/4[1] (combined LTRF 4–16/ 7– 21 in M. aromatica , M. chamaeleo , M. inparquesi and M. neblinaria ; Ayarzagüena & Señaris, 1994; Faivovich et al., 2013).

Call description: The call of Nesorohyla kanaima from the Wokomung Massif is normally composed of a single note (N = 6). However, a call of two notes was recorded once. The notes have a pulsatile structure, with 1–5 poorly resolved pulses, and their duration is 46–106 ms (80 ± 17 ms; N = 8). The call composed of two notes has a note interval of 141 ms. Seven calls were emitted in a 29.8 s interval (call rate of 14.08 calls/ min). The dominant frequency band occurs between 1872.2 and 2755.4 Hz, with peak frequency at 2239.5 Hz (N = 5) or 2411.7 Hz (N = 3). There is no modulation in frequency during the call. From the notes with peak frequency of 2411.7 Hz, one has two more harmonics (less energetic): one at 4823.4 Hz and the other at 7235.1 Hz ( Fig. 4). Sometimes an acute groan preceded or was emitted after the call, but unfortunately these notes were not of suitable quality for analysis.

Goin & Woodley (1969) reported the call of Nesorohyla kanaima from the type locality as composed of two notes of two to four pulses, with dominant frequency between 3400 and 3800 Hz. It is unclear if the differences between vocalizations from Mt. Kanaima and Wokomung Massif are due to interpopulation variation, recording artefacts or if there is more than one species under the name Nesorohyla kanaima .

Comparison with other calls of basal Cophomantini : From species of Myersiohyla , the call of Nesorohyla kanaima differs from those of M. aromatica , M. inparquesi and M. neblinaria , by having pulsatile structure with 1–5 poorly resolved pulses (6–11 pulses in those species combined; Ayarzagüena & Señaris, 1994; Faivovich et al., 2013). From M. liliae , it differs by having one to two notes (bout of 214–222 notes in M. liliae ; Kok, 2006). From M. chamaeleo , it differs by having the dominant frequency coincident with the fundamental frequency (dominant frequency at the second harmonic in M. chamaeleo ; Faivovich et al., 2013).

The call of Nesorohyla kanaima can be differentiated from those of Hyloscirtus alytolylax ( Duellman, 1972) , H. colymba ( Dunn, 1931) , H. mashpi ( Guayasamin et al., 2015) , H. phyllognathus ( Melin, 1941) , H. platydactylus ( Boulenger, 1905) , H. simmonsi ( Duellman, 1989) , H. torrenticola ( Duellman & Altig, 1978) , H. criptico Coloma, Carvajal-Endara, Dueñas, Paredes-Recalde, Morales-Mite, Almeida-Reinoso, Tapia, Hutter, Toral-Contreras & Guayasamin, 2012 , H. pacha ( Duellman & Hillis, 1990) , H. staufferorum ( Duellman & Coloma, 1993) and H. tapichalaca ( Kizirian, Coloma & Paredes-Recalde, 2003) by having 1–2 notes (calls emitted in bouts of 2–104 notes in those species combined; Duellman, 1972; Duellman & Altig, 1978; La Marca, 1985; Duellman & Coloma, 1993; Kizirian, Coloma & Paredes-Recalde, 2003; Coloma et al., 2012; Guayasamin et al., 2015). From H. jahni ( Rivero, 1961) , H. palmeri ( Boulenger, 1908) , H. torrenticola , H. antioquia ( Rivera-Correa & Faivovich, 2013) , H. criptico , H. larinopygion ( Duellman, 1973) , H. pantostictus ( Duellman & Berger, 1982) , H. psarolaimus ( Duellman & Hillis, 1990) , H. staufferorum and H. tapichalaca , it differs by the lack of frequency modulation (frequencies with ascending modulation in those species; Duellman & Altig, 1978; La Marca, 1985; Duellman & Coloma, 1993; Kizirian et al., 2003; Coloma et al., 2012; Rivera-Correa & Faivovich, 2014; Rivera-Correa, Vargas-Salinas & Grant, 2017). From H. armatus ( Boulenger, 1902) and H. condor ( Almendáriz, Brito, Batallas & Ron, 2014) , the call of N. kanaima differs by being shorter (note duration 46–106 ms in N. kanaima , 160–240 ms in H. armatus , 286–915 ms in H. condor ; Duellman, De la Riva & Wild, 1997; Almendáriz et al., 2014) and by its peak frequency (2239.5 or 2411.7 Hz in N. kanaima ; 2500 or 4700 Hz in H. armatus ; 770–1270 Hz in H. condor ; Duellman et al., 1997; Almendáriz et al., 2014).

Natural history: Few observations have been made on the biology of Nesorohyla kanaima . Duellman & Hoogmoed (1992) reported the absence of ponds on the slopes of Mt. Roraima (limits between southeastern Venezuela, Guyana and north-eastern State of Roraima, Brazil), and associated this to the large oviducal eggs of females to suggest that probably the species reproduces in rivers and rivulets. MacCulloch & Lathrop (2005) collected tadpoles and juveniles in a lentic riverine pond on Mt. Ayanganna.

On the eastern flanks of Mt. Roraima ( Guyana), adults and a metamorph of Nesorohyla kanaima were observed on arboreal branches at heights below 3 m, close to cascading mountain streams. In the pools of those streams, 1–1.5 m deep, there were plenty of densely black-coloured tadpoles. Frogs were found between 808 (5.267361˚N, 60.691389˚W) and 1244 (5.259207˚N, 60.720227˚W) m a.s.l. in the adjacent woodland habitats. In the same habitats also were found adults of Boana lemai ( Rivero, 1972) , B. roraima ( Duellman & Hoogmoed, 1992) , B. sibleszi , Oreophrynella macconelli Boulenger, 1900 , Otophryne robusta Boulenger, 1900 , Pristimantis cf. saltissimus , Stefania roraimae Duellman & Hoogmoed, 1984 , Stefania sp. , Tepuihyla sp. and an unidentified bufonid.

On the Wokomung Massif (Potaro-Siparuni, Guyana), Nesorohyla kanaima was found to be the most abundant anuran between 708 to 1540 m a.s.l. in primary cloud forests. Expeditions passing through the north-eastern (5.13˚N, 59.81333˚W), north-western (5.110083˚N, 59.820583˚W), central (5.090278˚N, 59.838333˚W) and south-eastern (5.002222˚N, 59.879722˚W) regions of the Massif were conducted during June, July and December of different years. Adults and metamorphs were found perched from low bushes to branches up to three meters high along cascading streams and in adjacent forests. During the mid-year rainy season, dense aggregations of males were found in low vegetation next to, or overhanging, slow-water pools of the streams, or along small, swampy creeks. They were in cacophony, emitting their call composed of one whistled note (see call description above). Both sexes presented noxious smelling skin secretions that tasted bitter. On one occasion, five males of N. kanaima were collected and kept in the same plastic bag as breeding males of Boana lemai , whose skin secretions were perceived to be more pungently smelling and bitter tasting. Two hours later the specimens of N. kanaima were dead, probably due to the skin secretions of B. lemai that had foamed up in the bag. The specimens from the latter species were still alive.

The black tadpoles were found at high density in almost all, small to large, streams visited. They were not fast swimmers and they were not schooling. No fish were observed in those streams, but two species of freshwater crabs were found as potential predators: Kunziana irengis ( Pretzmann, 1971) and Microthelphusa meansi Cumberlidge, 2007 . Other frogs found in the same habitat of Nesorohyla kanaima on the Wokomung Massif were Adenomera lutzi Heyer, 1975 , Anomaloglossus beebei ( Noble, 1923) , A. kaiei ( Kok, Sambhu, Roopsind, Lenglet & Bourne, 2006) , Atelopus hoogmoedi Lescure, 1974 , Boana lemai , B. roraima , B. sibleszi , Oreophrynella macconnelli , Otophryne robusta , O. steyermarki ( Rivero, 1968) , Pristimantis dendrobatoides ( Means & Savage, 2007) , P. marmoratus ( Boulenger, 1900) , P. saltissimus ( Means & Savage, 2007) , Stefania ayangannae ( MacCulloch & Lathrop, 2002) , S. coxi ( MacCulloch & Lathrop, 2002) , S. roraimae , Stefania sp. , Tepuihyla warreni ( Duellman & Hoogmoed, 1992) , T. exophthalma ( Smith & Noonan, 2001) and an unidentified bufonid.

Distribution: Nesorohyla kanaima is known from the Pakaraima Mountains in the Guiana Highlands, in eastern Venezuela (Estado Bolivar) and western Guyana (Cuyuni-Mazaruni District). It has been collected in Mt. Ayanganna ( MacCulloch & Lathrop, 2005), Mt. Kanaima, ( Goin and Woodley, 1969) , Mt. Roraima ( Duellman & Hoogmoed, 1992) ,Wokomung and Maringma-tepui (Kok & Means, pers. obs.).

Remarks: Our study of some adult male specimens [vouchers ROM (Royal Ontario Museum, Toronto, Canada) 39 575–76, 43 861, 43 871, from Mount Ayanganna, western Guyana] did not reveal a mental gland, but this needs to be corroborated histologically. The polarity of most phenotypic diagnostic characters is unclear, with the exception of the entirely pigmented ova and possibly of the black iris in adults (but see Discussion), which are putative autapomorphies of this new genus. Observations of metamorphosing individuals ( Fig. 5A–C) show that the iris is black and speckled with silver, light bronze, and copper (as described in field notes on live specimens from both PJR Kok and DB Means). The pupil is slightly heart-shaped and is delimited by a bronze pupil ring. In adults it turns black to a point where the pupil is difficult to distinguish, unless shining a bright light on it ( Fig. 5D–E).

The tadpoles described by MacCulloch & Lathrop (2005) were tentatively assigned to Nesorohyla kanaima due to the presence of both juveniles and recently metamorphosed individuals, associated with the species by the colour pattern. The identification of these tadpoles was discussed by Faivovich et al. (2013), who raised the possibility of a misidentification. We obtained additional tadpoles from Wokomung Massif ( IRSNB 16763) and the slopes of Maringma-tepui ( IRSNB 16765) in Guyana for which identity with Nesorohyla kanaima has been established by molecular analyses (PJR Kok, unpubl. data). These tadpoles are similar to those described by MacCulloch & Lathrop (2005), leaving no doubt that these authors indeed described the tadpole of Nesorohyla kanaima .

The study of our tadpoles indicates that the first anterior tooth row (A1) does not have a medial gap as described by MacCulloch & Lathrop (2005), but form a medial acute angle that we believe could had led those authors to misinterpret it as a medial gap. The LTRF on studied specimens varied from 2(2)/4 (N = 5) to 2(2)/4(1) (N = 8). Also, MacCulloch & Lathrop (2005) stated that the oral disc of tadpoles from Mount Ayanganna is not emarginated. Our study of tadpoles from Maringma-tepui revealed the presence of three emarginations on the posterior labium: one medial and two posterolateral. The nostrils of studied specimens have a large, medial, fleshy projection, that when pushed against the nostril, almost closes it.