Hylid sp. nov. a — McDiarmid and Paolillo (1988). Species list.

Hylid sp. C (Neblina) — McDiarmid and Donnelly (2005). Species list.

Holotype: AMNH A-131173 (♂). Venezuela: Departamento Amazonas: Cerro de la Neblina: Camp I, 1820–1880 m, collected by C. J. Cole on February 7, 1984.

Paratopotypes: AMNH A-131174 (♂), USNM 562048 (♂), USNM 562054 (♂), USNM 562062 (♀), USNM 562063 (♂), USNM 562064 (♂).

Paratypes: Venezuela: Departamento Amazonas: Cerro de la Neblina: Camp II, 2085–2100 m: AMNH A-131176 (♀), USNM 562051 (♂). Camp III, 1820 m: AMNH A-131178 (♂, recorded). Camp VII, 1730–1850 m: USNM 562065 (♂, recorded), AMNH A-123709 (♂), USNM 562055 (♂); Camp X, 1690 m: USNM 562057 (♂), USNM 562058 (♂), USNM 562066 (♂), USNM 562067 (♂). Camp XI, approx. 1450 m: USNM 562059 (♂, recorded), USNM 562060 (♂, recorded), USNM 562061 (♂), USNM 562068 (♀).

Referred Specimens: Venezuela: Departamento Amazonas: Cerro Neblina: Camp II, 2085– 2100 m.: USNM 562049–562050, 562052–562053, AMNH A-131175, 131177 (juveniles). Camp VII: USNM 562056 (juvenile).

Etymology: The species name chamaeleo is a noun in apposition. It is derived (like that of the lizard genus Chamaelo) from Latin chamaeleon < Greek chamailen, a kind of lizard—in reference to this frog's similar ability to change coloration.

Diagnosis: A species of Myersiohyla characterized by (1) SVL of male 44.6–49.6; female 46.0–56.9; (2) presence of two nuptial pads; (3) absence of a row of tubercles along the ventrolateral edge of the arm; (4) an externally evident mental gland in males; (5) thighs and shank without longitudinal bars; (6) dorsal color pattern with stellated melanophores over a ground color; (7) presence of a slip of m. depressor mandibulae that originates on the dorsal fascia at the level of the m. levator scapulae; (8) unpigmented eggs; (9) tadpoles with globular body; (10) fins reaching the base of the tail; (11) labial tooth row formula (LTRF) 4/7 to 6/11.

Comparison with other species: The new species differs from Myersiohyla inparquesi, M. kanaima, M. loveridgei, and M. neblinaria, n. sp., by having two nuptial pads, and by its dorsal color pattern (stellated melanophores over a ground color), with yellowish flanks in life, the absence of a row of tubercles along the ventrolateral edge of the arm, and the presence of a slip of m. depressor mandibulae that originates on the dorsal fascia at the level of the m. levator scapulae (absent on M. kanaima, and M. neblinaria; unknown in M. aromatica, M. inparquesi, and M. loveridgei). The color pattern differentiates the new species from M. aromatica (marbled with copper and dark brown; dark brown bars on arms, thighs and shanks). The new species can also be differentiated from H. kanaima because of its larger males (44.6–49.6, vs. 37.0–37.8 mm, = 37.4, n = 5, in M. kanaima), more robust body (slender in M. kanaima), more developed prepollex (proportionally reduced in M. kanaima), and its unpigmented eggs (eggs completely pigmented in M. kanaima). Furthermore, tadpoles of M. chamaeleo have a more reduced labial tooth row formula than those of M. aromatica and M. neblina (maximum 6/11 vs. a minimum of 10/17 in M. aromatica; tadpoles unknown in M. loveridgei), but larger than in M. kanaima (2/4; MacCulloch and Lathrop, 2005, but see discussion).

Description of the holotype (figs. 3A, 4–5): Body robust; head barely longer than wide; as wide as body; head width 34% SVL; head length 36% SVL. Snout in dorsal view rounded, in profile truncate; canthus rostralis rounded; loreal region concave; lips not flared; nares slightly protuberant, directed dorsolaterally, and posterior to anterior margin of lower jaw. Internarial region and top of head slightly depressed. Interorbital distance less than eyelid. Eye prominent; diameter 90% of eye-nostril distance. Tympanum rounded, diameter 63% of eye diameter. Supratympanic fold thin, starting behind the eye and extending to the posterior margin of the insertion of the arm, obscuring dorsal margin of tympanic membrane.

Vomerine teeth in two medially convergent series that contour the posterior margin of the choanae and give a slight S-shape to each series. Each series bears 14 (right) and 15 (left) teeth. Choanae kidney shaped; separated by a distance greater than their maximum diameter. Tongue ovoid, attached overall (narrowly free around lateral and posterior margin), posterior margin notched. Vocal slits present, longitudinal, originating on the sides of the tongue and extending to the corner of the mouth. Vocal sac single and subgular. Macroscopically evident glandular tissue irregularly distributed in the mental area, extending posteriorly up to the beginning of the vocal sac.

Forearm robust and prominent. No row of tubercles along ventrolateral edge of forearm. Fingers long and slender, bearing large, ovoid discs, with its circumferential groove clearly defined by the size difference between the disc and the smaller pad; width of disc on third finger 88% tympanum diameter. Relative lengths of fingers 1<2<4<3. Fingers webbed basally, without dermal fringes; webbing formula of outer fingers II 2–3½ III 3–3 IV. Subarticular tubercles large; the distal tubercle of finger IV ovoid, the others are round. Large supernumerary tubercles, flat and irregularly shaped on the base of fingers III, IV, and V. Outer metacarpal tubercle small, nearly round and bifid, barely noticeable. Inner metacarpal tubercle large, elliptical. Prepollex enlarged, semicircular. Nuptial excrescences in two contiguous, discrete pads, one along the medial margin of the prepollex, and the other medially and dorsally in the area between the distal margin of the prepollex and the subarticular tubercle; they are composed of a thin glandular pad, covered by minute dark epidermal projections.

Hind limbs slender; tibia length 51% of snout-vent length; foot length 36% SVL. Calcar absent; tarsal fold absent; subtle dermal ridge along the margin of the tarsus. Inner metatarsal tubercle large, elliptical; outer metatarsal tubercle round. Toes short, bearing discs smaller than those on fingers; disc of toe I smaller than others; relative length 1<2<3≈5<4; webbing formula I 2-2⁺ II 1⁺-2½ III 1⁺-2½ IV 2-1⁺ V. Subarticular tubercles large, round. Cloacal opening directed posteroventrally, at upper level of thighs; some flat, irregular, whitish tubercles scattered around and below cloaca. Dorsal skin smooth, granular ventrally. Pectoral fold absent.

Measurements of holotype (all in mm): SVL 47.4; HL 17.1; HW 16.1; IND 3.4; IO 5.3; ED 4.3; EN 4.8; TD 2.7; TL 24.3; FL 17.1.

Coloration of holotype in life: From field notes taken by Cole (Feb. 7, 1984): Dorsum reddish brown (as at night when caught) but sometimes changing to light green when awakened in daytime, with numerous scattered dark brown dots. Sides, anterior and posterior surfaces of thighs, and ventral surfaces of lower legs orangish gold, but much duller on legs than on sides (bright). Throat pale green. Ventral surfaces of arms, chest, and abdomen translucent, with pinkish purple showing through. Iris black with an intricate metallic copper reticulum.

Coloration of holotype in preservative: The dorsal ground coloration is light cream. The dorsum of body, forelimbs, and hind limbs are homogenously covered by two types of chromatophores that vary in size. The larger ones are brown and stellate, with creamy white centers. The smaller are round glossy white guanophorelike units. The fact that some of these have white stellated arms raises the possibility that these are early stages in the maturating process of the brown units described above, or considering that this species easily changes color in life, they could be just different physiologic stages of the same pigment unit. Pigment is nearly absent from the margins of the eyelid, where the background coloration is creamy yellow. The large chromatophores are less abundant or absent in the hidden areas of thighs, shanks and feet, as well as on the dorsal surfaces of the hands and feet. They are absent on the flanks of the body, where only small chromatophores are present. The ventral coloration of body, throat, and surfaces of limbs is slightly paler than the dorsal ground color.

Variation among paratypes: Morphological variation is relatively low. On specimens USNM 562051, 562054, 562060, 562062–64, which were fixed with their thumbs slightly deflected, the outer metacarpal tubercle is more obvious, as are several large, flat supernumerary tubercles on the proximal surfaces of the digits. With the exception of USNM 562061 and AMNH A-131176, most paratypes, unlike the holotype, have a median notch on the subarticular tubercle of finger IV. Males show two discrete nuptial pads (fig. 6A), that in a few cases almost coalesce medially because of the proliferation of epidermal projections (fig. 6B) into the space separating the glandular pads.

Variation of the webbing formula can be expressed as: I (1½-2)-(2-2⁺) II (l⁺-1½)-(2⁺-3^-) III (1-1½)-(2-2½) IV (2–2½)-(1⁺-1½) V. Vomerine teeth (n = 17): 10–18 (right, X = 14.3) and 10–18 (left, X = 13.7).

Most evident variation involves the dorsal coloration and pattern (figs. 3, 7). Several paratypes, most of them from Camp I (USNM 562048, 562054, 562064, 562063, AMNH A-131174) but also from camps II (USNM 562051) and X (USNM 562058) have a larger number of dorsal chromatophores, which makes them appear brown. Of these USNM 562048 has dark brown irregular blotches covering the dorsum and dorsal surfaces of limbs. In USNM 562058 only the body is darker; chromatophores are few on dorsal surfaces of thighs, and almost lacking on dorsal surfaces of arms that are cream with the exception of a few isolated chromatophores.

Coloration in life of some paratypes and referred specimens: We describe the variation in coloration and pattern based on descriptions of live specimens taken from field notes of McDiarmid, Myers, and Zweifel. Unless specified otherwise, notes were taken during the day and are presented here by camp number. Camp II. USNM 562049–50: Both went from bright grass green to brownish olive dorsally (brown specks on USNM 562049); sides bright yellow (USNM 562049) to very pale yellow (USNM 562050); yellow line on upper eyelid; blue green in all joints ventrally; center of chin and chest area blue green flesh; opalescent white where peritoneum shows; viscera visible in part through belly; iris bright coppery in bright sun. USNM 562051: Dorsally brown, light tan upper eyelids, bright yellow orange sides and anterior/posterior thighs; tan line along outer finger and toe; white rump patch; venter opalescent yellowish white on chin, pectoral area and belly; fleshy colored on forearms; toe pads whitish ventrally, pale greenish white dorsally; iris bright gold and brown (McDiarmid's field notes, Feb. 16, 1984). AMNH A-131176: The dorsal ground color of the head and body is light tan. The forearms and shanks are the same ground color, whereas there is merely a yellowish brown strip along the upper surface of the thigh. A pattern of small, yellowish, elongate spots is present over the head, body, lower arm and shank, extending along the foot. Webbing of the hind feet is sparse and tanish. The sides of the body, especially posteriorly, are an orange shade, which also extends onto the anterior and posterior surfaces of the thigh, but is less intense there. The chin is patternless with a yellowish tan tinge. The remaining ventral surfaces are grayish tan and translucent, so that the organs are at least dimly visible (Zweifel's field notes, Feb. 17, 1984). Camp VII. AMNH A-123709: Changing from brownish green to green above, speckled with black. Flanks rather bright orange; anterior and posterior faces of thighs yellowish orange. Ventral surfaces pale bluish green, except whitish on chest owing to underlying peritoneum. Iris pale bronze orange, dark banded by a fine venation (Myers' field notes, Dec. 1–2, 1984). Camp XI. USNM 562068: Back light green with brown spots (entire back appeared brown at night); turquoise in axilla, leg joints, groin, and behind eye (drawing in field notes); faint yellow on front and yellow on back of thighs; flanks orange; iris copper (McDiarmid's field notes, Feb. 25, 1985.) USNM 562061: Greenish with orange thighs (McDiarmid's field notes, Feb. 26, 1985.)

Measurements of the paratypes (all in mm, ± 1 standard error): Males (n = 15). SVL 44.6–49.6 (= 46.1 ± 1.4); HL 13.8–17.5 ( = 16.3 ± 1.0); HW 15.7–17.2 ( = 16.2 ± 0.1); IND 3.3–3.8 ( = 3.6 ± 0.0); IO 4.8–6.1 ( = 5.3 ± 0.1); ED 4.2–5.2 ( = 4.6 ± 0.1); EN 4.4–4.8 ( = 4.6 ± 0.0); TD 2.7–3.3 (X = 2.9 ± 0.0); TL 22.3–25.0 (X = 23.4 ± 0.2); FL 16.0–19.3 ( = 17.0 ± 0.2). Females (n = 5). SVL 46.0–56.9 (X = 51.1 ± 2.0); HL 16.2–19.9 ( = 18.0 ± 0.7); HW15.9–19.7 ( = 17.8 ± 0.7); IND 3.7–4.2 ( = 3.8 ± 0.1); IO 5.3–6.8 ( = 5.7 ± 0.3); ED 4.5–5.2 ( = 4.8 ± 0.1); EN 4.4–5.5 ( = 4.8 ± 0.2); TD 2.9–3.4 ( = 3.3 ± 0.1); TL 22.6–29.4 (X = 25.5 ± 1.2); FL 15.9–21.7 ( = 18.5 ± 1.1).

Remarks: The skin on the chin and anterior half of the gular area is glandular in all males when viewed with strong incident light and appears thicker than the loose skin of the vocal sac. On a few specimens the glands are evident externally without magnification (fig. 8). Superficial dissection of the skin of male paratype USNM 562058 reveals macroscopically evident glandular tissue (fig. 9). The prepollex is equally enlarged in males and females (figs. 5, 10). Two specimens partially dissected when tissue samples were removed (USNM 562061 ♀, and USNM 562057 ♂) show a parietal peritoneum partially covered with what seem to be iridophores at the level of flanks and dorsal body wall. The urinary bladder of the male is densely covered with the same putative iridophores (removed from the female when tissues were sampled), and the testes are ovoid, somewhat flattened, and unpigmented (left testis length 6.9 mm; SVL 46.6 mm). The female (SVL 51.2 mm) has large, unpigmented, ovarian eggs. The dissected left ovary (the right one was left intact) contained 54 eggs with largest diameter 2.9–3.1 mm ( = 3.02; s = 0.07; n = 15). Females are generally larger than males (SVL 46.0–56.9 mm; males 44.6–49.6 mm), with two (USNM 562066 and 562068) of the five females having SVL within the SVL range of the males. Superficial dissections in male USNM 562058 reveal the m. intermandibularis with a large diamond-shaped aponeurosis, through which the m. genioglossus is visible by transparency; no fibers of the m. intermandibularis contact the oval and araphic m. submentalis. Supplementary elements of the m. intermandibularis absent. Muscle depressor mandibulae with a differentiated slip that originates on the dorsal fascia at the level of the m. levator scapulae; tympanic origin contiguous with the belly that originates from the otic ramus of the squamosal.

We consider that the description of blue-green coloration described in the joints, and mental and ventral areas of living specimens, is very likely an indication of the occurrence of physiological chlorosis (Barrio, 1965a).

It is worth noting that the partial 16S sequences of juvenile referred specimen USNM 562056, and tadpole USNM 562723 from Camp VII are 1.1% different from those of specimens from the almost adjacent Camp XI (2.5 km SSE) and the more distant camps I and X (table 1). Interestingly, the vocalizations of USNM 562065 (USNM tape 68), also from Camp VII are indistinguishable from those of USNM 562060 (Camp XI); differences in vocalizations with specimens USNM 562054 (Camp I), and AMNH A-131178 (Camp III) are discussed below. Morphologically, all these adult specimens appear indistinguishable.

Ecology and natural history: Most specimens of Myersiohyla chamaeleo were found along or near streams. During the day, frogs were found in two types of plants, the leaves of which formed rosettes and held considerable water; at night they were found on vegetation and rocks along streams or in surrounding forest. Near Camp I, a male (USNM 562048) was found during the day hidden among the leaves of Bonnetia maguireorum growing along a stream. A snake (Leptophis cupreus) was photographed the following day (Feb. 11, 1984) as it searched the axils of these plants, presumably foraging for frogs (Albuquerque and McDiarmid, 2010). Two males (AMNH A-131173-74) had been previously located (Feb. 7, 1984) as they called from the leafy part of these plants, and another male (AMNH A-131178) was found 10 days later near Camp III also calling from a Bonnetia plant. Seven juveniles (SVL 28.6–34.0 mm) were collected from the large terrestrial bromeliad, Brocchinia tatei, which grew in large patches near Camp II and Camp VII. Three from Camp II (USNM 562049-50, USNM 562052) were collected during the day from the outer axils of a bromeliad growing along a small stream; both were in a head-up position. A fourth (AMNH A-131175) was hiding in a partly dead, rolled-up leaf and another (AMNH A-131175) was found in a bromeliad. Two other juveniles (USNM 562052-53) were found together in the central axil of a Brocchinia in the same area. A seventh juvenile was found at night (Feb. 12, 1985, 19:30–21:30 hr) on a leaf of a Brocchinia about 1 m above the ground near the heliport of Camp VII. A gravid female (USNM 562062) was found in the back of a cave adjacent to a creek ESE (above) of Camp I at 14:00 hr. Three frogs (USNM 562054, USNM 562063-64) formed a chorus around a 3 m wide and >1 m deep pool in the creek near Camp I (20:30–21:30 hr). One was taken from moss on a rock face 5 cm above the water, and two were calling from vegetation along the shore and about 10 cm above the water.

The five adult female Myersiohyla chamaeleo (SVL 46.0–56.9 mm) collected during the expedition were all found between January 31 and February 26, and all were gravid. Likewise, all adult males (SVL 44.6–49.6 mm, n = 16) had well-developed nuptial pads and were assumed to be in breeding condition. Calling males were heard at night along streams near camps I, III, VII, and XI during the same period.

Vocalizations: Recordings of calling males are available from four camps on opposite sides of the Cañon Grande (fig. 1). Recordings from east of the great canyon (camps VII, XI) are of males calling from along streams, whereas those from west of the canyon (camps I, III) are of frogs calling in camp from plastic bags. These segregate as two call types, one characterized by pulsatile notes (eastern camps), and the other is characterized by nonpulsatile notes (plastic-bag recordings from western camps). These are described separately and then compared.

Calls with Pulsatile Notes: A male was recorded calling on a vertical stem 1.5 m above streamside at Camp VII, at an air temperature of 14° C (USNM 562065 on USNM tape 68). Two males were recorded at Camp XI at an air temperature of 17° C (USNM 562059-562060 on USNM tape 70); the first was calling from the lower leaves of a bromeliad 1.5 m above a pool in the stream, while the second called from the lower, dead leaves of a bromeliad nearly 2 m above streamside.

The call is a short-to-long train of well-spaced, short notes given at a rate of about 1–2 notes/ sec at 14°–17° C. Note duration is about 0.04 sec (32–51 ms, = 44.3 ± 5 ms, n = 29). Each note is composed of several poorly resolved pulses that are indicated on wideband spectrograms (fig. IIB) but which are best visualized from expanded waveforms (fig. 11D). The individual note is a short “beep” composed of two harmonics, including a weak fundamental frequency at about 1100 Hz and a strong dominant frequency at about 2200 Hz.

Calls with Nonpulsatile Notes: One of three males was recorded calling from a plastic bag at Camp I at an air temperature of 17° C (recording of one of USNM 562054, 562063–562064 on USNM tape 67:17). A male collected at Camp III male was recorded in a plastic bag at the lowland Base Camp three days after capture, at an air temperature of 24° C (AMNH A-131178 on AMNH Herpetology Reel 250).

The calls of the frogs in plastic bags are given at rates of about 1.2 notes/sec at 17° C and up to about 4 notes/sec at 24° C. Note duration of the nonpulsatile notes is about 0.02 sec (17–27 ms, = 22.3 ms, n = 18). Most notes in the two recordings were examined by expanding the waveforms and were seen, without exception, to be nonpulsatile. The note is a short “beep” composed of two harmonics at about 950 Hz and 1900 Hz; the first or fundamental frequency is also the dominant frequency. The note is frequency modulated, with an initial increase followed by a steep decrease; owing to the shortness of the note, this modulation is barely indicated in figure 12C (but see fig. 13A′).

Pulsatile and Nonpulsatile Notes Compared: The descriptions above and figures 11–12 show both similarities and differences in the notes. The pulsatile note is longer (about 0.04 sec) and has a weak fundamental frequency at about 1100 Hz, whereas the nonpulsatile (single-pulse) note is half as long (about 0.02 sec) and has a strong fundamental/dominant frequency at about 950 Hz. Each is frequency modulated in a similar way, with a sharp rise and subsequent sharp decline in the second harmonic; slight frequency modulation occurs also in the first harmonic of the nonpulsatile note.

Frequency modulation in the longer pulsatile note is evident in a standard narrowband spectrogram (fig. 11C), but not so obvious in the shorter nonpulsatile note (fig. 12C). However, it becomes evident in the shorter note when an exemplar is graphed at a different analysis size (fig. 13A′).

Clearly two different call types are involved, although their biological significance is uncertain. Slight divergence of the calls may have occurred on opposite sides of the great Cañon Grande of Cerro de la Neblina. On the other hand, perhaps confinement (in plastic bags) induced a call different from the normal advertisement call. In any case, we could find no external morphological differences among the specimens.

Tadpoles: The uncorrected pairwise distances of the 16S sequences (table 1), indicates that the sequence of tadpole USNM 562718 is similar (0%–0.2%) with those of the adult paratypes USNM 562048, 562057, 562061. All these have a 0.9%–1.1% uncorrected p-distance with sequences of a tadpole USNM 562723 and juvenile USNM 562056, these from Camp VII. Conversely, these are 12.5%–13.6% different from the sequences of available specimens of Myersiohyla neblinaria. Considering the small distances or identical sequences, we consider this tadpole and all others with the same general morphology (see below) to be larvae of M. chamaeleo. We describe a typical larva of M. chamaeleo (fig. 14) and then comment on the variation found in other available material.

Description: A tadpole (USNM 562727) in stage 25 from Camp XI (see below) has the following measurements (all in mm): total length 43.0; body length 16.6; basal tail muscle height 6.3; basal tail muscle width 4.3; maximum dorsal fin height 3.6, this point being 12.2 mm away from body; maximum ventral fin height 2.7, this point being 11.7 mm away from body; body width 9.5; body depth 8.3; eye diameter 1.9; pupil diameter 1.4; interorbital distance 4.1; internarial distance 3.7; snout to nostril distance 2.7; snout to eye 4.1; snout to spiracle 12.0; oral disc diameter 8.2.

Oral disc ventral, not emarginate but with posterior fold. Marginal papillae in a single row sometimes slightly offset, without gaps; submarginal papillae absent. LTRF 5(5)/9(1). Gaps in A-5 and P-l small (∼0.1 mm), posterior gap about 40% of anterior one. Labial tooth density 19–24 teeth/mm, with fewer teeth in rows proximal to jaw sheaths. Teeth in rows P-l through P-4 more robust and about one-third to one-half longer than those of more posterior rows. Several (3–5), short tooth rows scattered lateral to primary tooth rows and on folded area of oral disc. Jaw sheaths pigmented; upper sheath narrow, slightly wider (M-shaped) medially, long and shallowly arched with short serrations; lower sheath moderately wide, V-shaped with small, pointed serrations, that are larger than those on upper jaw sheath.

Body ovoid; snout rounded in dorsal view, sloping in profile. Eyes dorsolateral, not visible in ventral view. Nostrils oval, with low, medial, fleshy projection and darker fleshy rim. Spiracle sinistral with its medial wall fused with the body wall, and the lateral wall shorter than the medial wall; spiracular opening directed slightly dorsad. Vent tube dextral, twice as long as wide. Dorsal fin arises just anterior to body-tail juncture; lower fin arises at base of tail and masked by vent tube; fin tip pointed, transparent. Neuromasts present, whitish, visible beneath and posterior to eyes, on top of head, and dorsolaterally on body. Melanophores rod shaped.

Head and body brown with irregular pigment; a pair of pale, elongate marks visible dorsolaterally. Coloration slightly paler on snout, around oral disc, and on spiracle. Venter brown, gut visible through body wall. Tail and fins brown except for anterior 20% and posterior 15% which are pale, giving a bicolor appearance to the tail; tail tip transparent. Other specimens from the same locality are entirely brown or bicolored and some from camps I and III are completely black (see below). Fins may be transparent, variably mottled, or completely black. Smaller tadpoles are generally paler overall.

Neuromasts are usually visible in tadpoles of Myersiohyla chamaeleo and appear as lines of round, whitish dots on the head, body, and tail. Their topography is complex and most obvious in darkly colored specimens from Camp III (USNM 562721); they are difficult to see in pale or bicolored specimens. In the Camp III tadpoles, the supraorbital line includes 13–15 neuromasts that converge anteriorly on the head between the nares and continue anteroventrally onto the snout where it approaches the anterior extent of the infraorbital series of 15–19 neuromasts; occasionally a single preorbital neuromast is visible. The oral series is indistinct and variable (also noted for many species by Lannoo, 1987). In some tadpoles it consists of a longitudinal row of 4–6 neuromasts and an anterior, ascending line, sometimes V-shaped, of 3–5 neuromasts; in others, only a few scattered neuromasts occur in the oral region. An angular series of 4–5 neuromasts is visible below the eye. A postorbital series consists of two dorsal (supra-) and two ventral (infra-) neuromasts. A row of 8–10 neuromasts located dorsolaterally on the body is continuous with 4–5 neuromasts in the dorsal row at the base of the tail fin; on the body this dorsolateral series forms a loop and joins a lateral series of 10–12 that continues onto the tail as the middle caudal series consisting of 6–8 neuromasts. About halfway along the tail this middle series projects dorsally (3–4 neuromasts) and continues posteriorly (11–13 neuromasts) along the tail-fin juncture to near the tail tip. A ventral body row of 13–14 neuromasts extends from near the vent tube anterodorsally above the spiracle to a point between the dorsal loop and the postorbital series.

Natural History: Tadpoles of Myersiohyla chamaeleo were collected in black water streams at six sites on Cerro de la Neblina. A series of 26 specimens (AMNH A-131171; stage 25; TTL 33–80 mm; LTRF 5/9) collected on February 7, 1984, by Cole from pools in a small stream near Camp I were black with globular bodies. Clear evidence of ontogenetic change from clear fins to black fins can be seen in this series. Tail tips were usually clear. Five other specimens from Camp I (USNM 562717 [4], USNM 562718, tissue voucher; stage 25; TTL 39–65 mm; LTRF 4/7–4/9) were collected by Co croft from pools in another small stream (1–2 m wide, 10–30 cm deep) ESE of camp, January 31, 1985, between 10:00 and 13:00 hr. These tadpoles also were dark brownish black with clear tail tips.

Three tadpoles were collected from a small stream (0.6–1.2 m wide, 30–60 cm deep) near Camp II. One (USNM 562719; stage 25; TTL 40 mm; LTRF 4/9; Feb. 18, 1984) is light brownish with nearly transparent fins with small, scattered patches of melanin, a clear tail tip, and transparent belly. Two other pale specimens (USNM 562720; stage 25; TTL 54–56 mm; LTRF 5/9) were collected by Paolillo on January 30, 1985, from a small hollow beneath a rock; his notes indicate that they were feeding on the white sand bottom of the stream.

Six black specimens (USNM 562721; stage 25; TTL 56–71 mm; LTRF 4/7 [damaged mouth], 4/8–9 [broken series], 4/9, 5/9[3]) were collected in a stream near Camp III by Cole on February 18, 1984. Several of these had broken tail tips, but it is not clear whether this reflected attempted predation or collection damage; all intact tips were clear.

Cocroft and Paolillo collected tadpoles during the days of February 10–11 and on the night of February 12, 1985, at Camp VII. These 29 specimens (USNM 562722 [28], USNM 562723, tissue voucher; stage 25; TTL 23–71 mm; LTRF 3/4–5/11) were taken from quiet pools along the stream near camp; most were in water <1 m deep; several larger tads seen during the day in a large pool >1 m deep eluded capture. At night the tadpoles were sensitive to light and substrate vibrations and quickly took cover under rocks. The water temperature at 11:30 hr was 14° C. Two color phases were seen (Cocroft's field notes, Feb. 11–12, 1985): black, sometimes with a lighter tail, and greenish gray (we assume these greenish tadpoles were larvae of Myersiohyla neblinaria because adults of this species were collected here). Fin tips of the darker ones varied from clear to dark. One tadpole (TTL 35–40 mm estimate, tail broken at 29 mm) had a damaged tail that appears to have resulted from attempted predation. Another (TTL 45 mm) had a regenerated tail tip (distal 7 mm) that was transparent and slightly narrower. Several tadpoles in this series had irregular clumps of melanophores on their bodies and tail, and one in poor condition had what appeared to be a fungal infection of the oral disc near the mouth.

Three other samples of tadpoles were collected by Myers and Ford on December 1–2, 1984, at the same site. Fifteen of 16 tadpoles (AMNH A-123713) were in stage 25 and measured 24–61 mm TTL; the largest from this series was 75 mm TTL at stage 35with an LTRF of 6/10, while the others ranged from 4/7 to 6/10. Two other specimens collected at the same time were metamorphs at stage 41; their tooth rows were starting to break up (AMNH A-123711 at 68 mm TTL had an LTRF of 5/10, and AMNH A-123712 at the same size had an LTRF of 417).

A single brownish tadpole was collected during the day (Feb. 2, 1985) in a stream that formed the headwaters of the Río Baría and flowed SW into the Cañon Grande (Camp IX; 1780–1830 m; 1°00′00″ N, 65°53′00″ W). The tadpole (USNM 562724; stage 25; TTL 29 mm; LTRF 4/7) has an obviously bicolored tail, similar to that illustrated in figure 14; the anterior third is light and the fins are transparent, while the posterior two-thirds is dark with densely distributed melanophores in the fins; the tail tip is transparent.

Several tadpoles collected in the stream below Camp XI are also in stage 25. A single specimen (USNM 562725; TTL 62 mm; LTRF 5/10) was taken in slow-moving water about 5 cm deep on February 25, 1985. Nine tadpoles (USNM 562726, USNM 562727 [described]; TTL 37–61 mm; LTRF 4/8[4], 4/9, 5/8, 5/10[2], 5/11) netted by McDiarmid on February 26, 1985 (19:30–21:30 hr), were uniform dark brown or bicolored; all had clear tail tips. Nine other specimens (USNM 562728; TTL 24–64 mm; LTRF 4/7[1], 4/8[4], 4/9, 5/9[2], 6/9) collected on February 27, 1985, by Paul Spangler during the day were all dark. Two other specimens (USNM 562729; TTL 47–49 mm; LTRF 5/9, 5/8) were both uniform dark. Most of the tadpoles from Camp XI had lateral tooth rows.

Comments: Twelve lots of 109 larvae of Myersiohyla chamaeleo were collected at six sites on Cerro de la Neblina. Even though the size ranged from 23 to 80 mm TTL, all but three specimens were in stage 25. A 75 mm TTL tadpole was in stage 35, and two 68 mm TTL individuals in stage 41; tooth rows in both of the latter were starting to fall out down. LTRF s varied from 4/7 to 5/11 and 6/10, with the lower numbers found in smaller individuals. No variation was observed in the positions or sizes of the too throw gaps. All specimens from camps I and III were more globular (body shape) and densely pigmented, being nearly all black (body, tail, and fins) compared to most tadpoles from the other sites. Tadpoles from camps II, VII, IX, and XI were either uniform brown or bicolored with anterior part of tail pale. The latter specimens were usually found in smaller, often faster-flowing streams bordered by gallery forests that provided more shaded habitats. In contrast, the larger, black specimens generally were from open, exposed streams with larger, pool-type habitats. Intraspecific morphological differences among tadpoles have been noted by previous workers (Jennings and Scott, 1993; R.W.M., personal obs.), and we interpret the differences in coloration and body proportions of M. chamaeleo tadpoles as reflecting differences between habitats. Increased melanic pigmentation might counter the potential negative consequences of increased exposure to predation and solar radiation in more open habitats. Whether ontogenetic increases in melanism and larger body size reflect differences in food availability and growth (Altig and McDiarmid, 1999b) or are associated with distastefulness and aposematic coloration (Brodie and Formanowicz, 1987) remain to be demonstrated. See table 1 for 16S distances among samples from different camps. The few samples available do not allow us to establish any significant association between these morphological differences and the p-distances that we observe.

In an effort to determine any ontogenetic change in LTRFs, we divided the Camp VII sample of 29 specimens (USNM 562722 [28], USNM 562723) into five size categories and recorded the LTRF for each tadpole in each size class (those with damaged oral discs or broken tails were not included). The results are as follows: 1. TTL <30 mm: LTRFs 3/4 (N = 1), 4/6(1); 2. TTL 30–39 mm: LTRFs 4/7(3), 4/8(3), 4/9(1), 5/6(1), 5/8(1), 5/9(1); 3. TTL 40–49 mm: LTRFs 4/8(1), 4/9(3), 5/9(4); 4. TTL 50–59 mm: LTRFs 4/7(1), 4/8 (1), 5/10(2) 5/11(1); TTL >59: LTRFs 4/7(1). Though preliminary, three patterns emerge. First and not unexpectedly, smaller individuals generally have smaller LTRFs, and the smallest value (3/4) was in the smallest tadpole (23 mm). In contrast and in spite of the tendency for larger tadpoles to have more labial tooth rows, the largest LTRF value (5/11) was not in the largest tadpole (TTL 71 mm) but in a smaller one (TTL 59 mm); in this specimen the P-11 row was weakly developed and barely discernible. That all of these specimens were in stage 25 and showed more than a threefold increase⁴ in size (TTL 23–71 mm) without a concomitant change in development suggests a long-lived tadpole. Another sample of 26 specimens from Camp I (AMNH A-131171) showed the same pattern; all were in stage 25 and ranged between 33 and 80 mm TTL. The low water temperatures (14° C) and relatively low nutrient content of most of the Neblina streams likely contributes to a relatively slow growth rate. The size range in a single developmental stage (stage 25) was surprising and reinforces previous observations of a decoupling of size, developmental stage, and age in many tadpoles. One of the largest M. chamaeleo tadpoles (AMNH A-123713) was also collected at Camp VII; it was in stage 35, measured 75 mm TTL, and had an LTRF of 6/10. That it was only 4 mm larger than the largest stage 25 individual collected from this same stream and smaller than the largest specimen (80 mm TTL) at the same stage from Camp I, suggests a larval growth curve in this species different from that reported for most other anurans for which data are available (e.g., see table and plots from Middle American hylids in Duellman, 1970). Why much of the growth is apparently concentrated in the earliest larval stage in this species, as contrasted to a less skewed and more evenly distributed size range with a typical sigmoid-shaped growth curve as reported for most tadpoles, is unknown. We saw no evidence of size-related predator pressure in these environments that might select for a faster growth and larger size early in development. A cursory review of growth data for Middle American tadpoles (Duellman, 1970) showed that 25%–50% of the maximum size is attained in pond tadpoles by stage 25, while stream species may reach more than 65% of their maximum size in the same stage. Together, these data imply a slower developmental rate for tadpoles living in stream compared to pond habitats. The lower productivity and generally cooler temperatures found in stream habitats may account for most of this pattern. Whether this is age dependent remains to be shown.

A second observation relates to the lack of a detectable pattern in either sequence or synchrony in the addition of anterior and posterior tooth rows. In the Camp VII sample, anterior rows increase by only a single row within a size class (e.g., 4–5 in categories 3 and 4) and two rows overall (3–5), while posterior rows increase within size categories by up to four rows (6–9 in category 2) and as many as seven rows (4–11) overall. The smaller size of teeth in the distal rows (fig. 14) indicates that rows are added distally on both labia; this pattern was reported for many stream dwelling hylids by Altig and Johnson (1989) and has been observed in other forms as well (R.W.M., personal obs.).

Finally, we suspect that individual variation (either ecological or genetic) in growth rates of tadpoles of certain species, especially of long-lived, stream inhabiting forms, may override any ontogenetic pattern of toothrow addition, if such exists, and result in an LTRF in the largest specimen that is less than the maximum possible value. Accordingly, some individual larvae may reach a developmental stage (stage 41) approaching metamorphosis without ever having achieved the highest possible LTRF value for the population. Based on examination of all tadpoles of Myersiohyla chamaeleo taken on Cerro de la Neblina, the maximum possible LTRF value is 6(6)/11(1). No single specimen examined had that value. More detailed studies examining the trade-offs and interactions between size, developmental stage, and age, especially with species that have high LTRFs, are needed before we understand what is involved in the ontogeny of stream-breeding hylid tadpoles.

Hylid sp. nov. b - McDiarmid and Paolillo (1988). Species list.

Hylid sp. D (Neblina) - McDiarmid and Donnelly (2005). Species list.

Hyla inparquesi (non Ayarzagüena and Señaris, “1993” [1994]) - Faivovich et al. (2005). Misidentification of specimens from Cerro de la Neblina.

Myersiohyla inparquesi (non Ayarzagüena and Señaris,“1993” [1994]) - Faivovich et al. (2005). First combination with Myersiohyla.

Holotype: USNM 562071, adult ♂, with muscle tissue removed from left thigh. Venezuela: Departamento Amazonas: Cerro de la Neblina: Camp VII, 1730 m, collected by A. Gardner on February 1, 1985.

Paratopotypes: USNM 562072, AMNH A-123715 (♀ ♀).

Paratypes: Venezuela: Departamento Amazonas: Cerro de la Neblina: Camp I, 1820–1880 m: USNM 562070 (♀). Camp II, 2085–2100 m: AMNH A-131172 (♀). Camp XI, 1450 m: USNM 562073-562082 (♂ ♂).

Etymology: The species name neblinaria is an adjective derived from the mountain name Neblina plus the Latin suffix -aria (fem. to agree with -hyla of genus name), denoting a place or a connection to a place.

Diagnosis: Myersiohyla neblinaria is characterized by: (1) SVL of male 47.7–52.3; female SVL 54.0–61.6; (2) a single, thick, nuptial pad; (3) a row of tubercles in the forearm; (4) mental gland present but evident only through dissection; (5) thighs with longitudinal bars; (6) dorsum brown, with variable presence of a dorsal line and blotches; (7) m. depressor mandibulae without an origin on the dorsal fascia at the level of the m. levator scapulae; (8) unpigmented eggs; (9) tadpoles with a dorsoventrally flattened body; (10) fins reaching the base of the tail; (11) labial toothrow formula (LTRF) 16 (16)/21 (1).

Comparison with other species: Myersiohyla neblinaria is most similar to M. inparquesi but differs in having: hand and toe discs without sagittal black lines (present in M. inparquesi); larvae with dorsal and ventral tail fins reaching the base of the tail (extending only a third in M. inparquesi); larvae with scattered submarginal papillae in the upper and lower labium (apparently absent in M. inparquesi); and larval oral disc with a single row of marginal papillae (three rows in M. inparquesi). Furthermore, the advertisement call of M. neblinaria is composed of a single note with 7–9 pulses and a duration of 295–382 ms; in contrast, the advertisement call of M. inparquesi is composed of two notes, one of 6 pulses and a duration of 90 ms, and the other with 7 pulses and a duration of 80 ms (Ayarzagüena and Señaris, “1993” [1994]). Myersiohyla neblinaria differs from M. loveridgei in its larger SVL (M. neblinaria 4 7.7–52.0 mm; M. lover idgei 42–45 mm; Rivero, 1961; “1971” [1972]), and in having a single nuptial pad (two in M. loveridgei). From M. aromatica it differs in having a single nuptial pad, instead of two (Ayarzagüena and Señaris, “1993” [1994]: 130); larger SVL (M. aromatica, 43.6–46.6 mm; M. neblinaria 47.7–52.0 mm); tadpole with dorsoventrally flattened body (globular in M. aromatica) and more rows of labial teeth (LTRF 10–13/14–18 in M. aromatica). M. neblinaria is distinguished from M. kanaima by its larger SVL (males: 37.0–37.8 mm, = 37.4, n = 5; females: 46.3–48.0 mm, = 47.05, n = 4; Goin and Woodley, 1969, and this paper), more robust body (slender in M. kanaima), more developed prepollex (proportionally reduced in M. kanaima), more extensive and thick nuptial pad (reduced to the dorsal margin of the prepollex, and composed of a thin layer of glandular tissue, apparently without colored epidermal projections), unpigmented eggs (completely pigmented in M. kanaima), color pattern, tadpole with dorsoventrally flattened body (depressed in M. kanaima) and more rows of labial teeth (LTRF 2/4 in M. kanaima, but see discussion).

Description of the holotype (figs. 15a, 16–17): Body robust; head slightly longer than wide; as wide as body; head width 33% SVL; head length 36% SVL. Snout in dorsal view rounded, in profile truncate; canthus rostralis rounded; loreal region concave; lips not flared; nares slightly protuberant, directed dorsolaterally, and posterior to anterior margin of lower jaw. Internarial region and top of head slightly depressed. Interorbital distance shorter than eyelid. Eye prominent, its diameter 125% eye-nostril distance. Tympanum rounded, its diameter 39% eye diameter. Supratympanic fold heavy, beginning behind eye and extending to anterior margin of insertion of arm, obscuring dorsal margin of tympanic membrane.

Vomerine teeth in two medially convergent series that contour posterior margin of choanae giving shallow S-shape to each series; 16 teeth on right and 13 on left. Choanae kidney shaped; separated by distance larger than their maximum diameter. Tongue ovoid, attached overall (narrowly free around lateral and posterior margin), posterior margin entire. Vocal slits present, longitudinal, originating at sides of tongue and extending to corners of mouth.

Vocal sac not evident externally. Glandular tissue on mental area visible at 60x, but otherwise not evident externally.

Forearm robust and prominent. Row of low, irregularly spaced tubercles along ventrolateral edge of forearm. Fingers long and slender, with clearly defined circumferential groove, pad smaller (∼75%) than disc; width of disc on third finger similar to tympanum diameter. Relative lengths of fingers 1<2<4<3. Fingers webbed basally, without dermal fringes; webbing formula of outer fingers II III 3⁺-3⁺ IV. Subarticular tubercles large; distal tubercle of finger IV irregularly shaped, others round. Large supernumerary tubercles, round or irregularly shaped. Outer metacarpal tubercle small and bifid, nearly round. Inner metacarpal tubercle large, elliptical. Prepollex enlarged, semicircular. Nuptial pad present, covering internal margin of thumb and dorsal margin of enlarged prepollex; consisting of thick, cream glandular pad, covered by minute dark epidermal projections (fig. 18).

Hind limbs moderately robust; tibia length 55% of SVL; foot length 43% SVL. Calcar absent; tarsal fold absent, but skin subtly thickened. Inner metatarsal tubercle large, elliptical; outer metatarsal tubercle round. Toes short, bearing discs smaller than those on fingers; disc of toe I smaller than others; relative length 1<2<3≈5<4; webbing formula I 2–2 II 1½-2½ III 2–3 IV 2½-1½ V. Subarticular tubercles large, round. Few flat supernumerary tubercles in longitudinal series at base of toes. Cloacal opening directed posteroventrally at midlevel of thighs; several flat tubercles scattered laterally and below cloaca. Dorsal skin smooth, finely granular on abdomen. Pectoral fold absent.

Measurements of holotype (all in mm): SVL 49.8; HL 18.0; HW 16.8; IND 4; IO 5.4; ED 5.9; EN 4.7; TD 2.3; TL 27.7; FL 21.3.

Coloration of holotype in life: No field notes are available for the holotype. Visible details from a color slide of RWM, here reproduced as figure 15A, diifer little from the coloration in preservative.

Coloration in of holotype in preservative: Dorsally brown, turning darker brown toward the flanks, and with many large, irregular tan blotches; these blotches are smaller on the sides. Thick dark brown to black uniform line from tip of the snout to approximately the distal tip of urostyle. Dark, poorly defined, diffuse canthal stripe. Discs dorsally dark gray, contrasting with the brown coloration of the limbs. Hind limbs dorsally and ventrally brown; each thigh with five tan dorsal longitudinal stripes with diffuse margins that extend into the hidden areas. Shank with irregular tan blotches, some of which look dorsally continuous with those on thigh, smaller blotches on the outer surface of the foot. The only pattern present on rear surfaces of thighs are the posterior margins of the tan bars. A few tan blotches surround the cloacal region. Paracloacal tubercles tan. Ventral surfaces of body and limbs brown; subarticular, palmar, and thenar tubercles light gray. Irregular series of small tan blotches across abdomen below the pectoral region.

Variation of paratypes: Morphologically the paratypes are similar to the holotype. The variation in toe-webbing formula can be expressed as: I (2-2½)- (2-2½) II (1⁺-2^-)-(2½-3) III (1½-2)-(2–3) IV (2½-3)-(1⁺-2 ) V. Vomerine teeth (n = 12): 12–17 (right, = 14.1) and 11–16 (left, = 13.4). The structure that has been described in the holotype as a subtle dermal ridge along the tarsus is equally subtle in most paratypes, with the exception of three females (AMNH A-131172, USNM 562070 and 562072), where it is quite distinct, and proximally forms a few flat tubercles.

Most notable variation involves the coloration and color pattern, particularly regarding presence and extent of blotches (figs. 15, 19–20). We describe the variation according to the camps at which specimens were collected. Unless otherwise stated, all descriptions refer to preserved specimens. Camp I. The single specimen collected at Camp I (USNM 562070), an adult female, has a barely defined dorsal pattern, being light tan with a diffuse darker reticulum. Exposed surfaces of shank and foot are light tan as well, with diffuse brown blotches. Thighs dark brown with light tan blotches. Undersides of hind limbs and forelimbs reddish brown. Abdomen and throat creamy; areas surrounding the insertion of the arms dark brown. Diffuse line borders the lower jaw. An irregular brown line extends from the mandibular symphysis up to the beginning of the pectoral region. Camp VII. Coloration of two females differs notably from the holotype. USNM 562072 is dorsally tan, becoming brown on the sides. It lacks the dorsal line and other discernible pattern, with the exception of some irregular, small, tan blotches on the sides. Coloration of this frog during the day was noted by Cocroft (field notes, Dec. 2, 1985) as: iris coppery yellow with black reticulations; dorsal surfaces of fore- and hind limbs tan; dorsum from snout to vent brownish tan, fading to lighter tan laterally in center of body; feet mottled light tan and dark brown dorsally, toe pads dorsally dark brown, almost black in center outlined with light tan, brownish gray ventrally; sides and ventral surfaces lavender brown mottled with light tan; anterior and posterior thighs same as sides; toe webbing dark brown; light tan upper lips; tympanum coppery brown. In AMNH A-123715 the dorsal pattern is reversed with respect to the holotype, with the ground coloration being tan and the blotches, which are smaller and fewer, being brown. Middorsal dorsal dark line present. Seven longitudinal bars on the thighs. In his field notes Myers described this specimen in life as “Brown with vague darker blotching and blackish brown vertebral line. Flanks and thigh mottled and banded tan and purplish brown. Ventral surfaces purplish brown with suffusion of tan across chest. Iris golden yellow with black reticulation (conspicuous but not sharply defined).” Camp XI. Nine of the 10 specimens collected in Camp XI have the dark vertebral line. With the exception of USNM 562079, all are like the holotype in the dark brown flanks with tan blotches of variable size. Most variation involves the dorsal and thigh pattern, and the hues of brown and tan. On specimens USNM 56207374, USNM 562076, and USNM 562079-80, the dorsal pattern is barely distinguishable. USNM 562075 is completely light tan without dorsal blotches. Specimens USNM 562077-78 and USNM 562081-82 have dorsal blotches, but their pattern is reversed from that of the holotype, with their ground coloration being tan or light tan and the blotches brown.

Measurements of the paratypes (all in mm, ± 1 standard error): Males (n = 9): SVL 47.7-52.3 ( = 50.5 ± 0.4); HL 17.9-18.9 ( = 18.4 ± 0.1); HW 16.8.0 - 18.1 ( = 17.4 ± 0.1); IND 3.7-5.0 ( = 4.0 ± 0.1); IO 4.9-5.6 ( = 5.2 ± 0.1); ED 5.0 - 6.5 ( = 5.7 ± 0.1); EN 4.1 - 5.2 ( = 4.6 ± 0.1); TD 2.3 - 2.6 ( = 2.5 ± 0.1); TL 26.0-28.3 ( = 27.6 ± 0.2); FL 19.7 - 21.8 ( = 20.5 ± 0.2). Immature male (USNM 562081): SVL 40.0; HL 15.0; HW 14.0; IND 3.4; IO 4.7; ED 4.8; EN 3.9; TD 1.9; TL 23.1; FL 15.9. Females (n = 4): SVL 54.0 - 61.6 ( = 58.9 ± 1.7); HL 20.4 - 22.3 ( = 21.5 ± 0.4); HW 19.1 - 21.0 ( = 20.3 ± 0.4); IND 4.6 - 5.0 ( = 4.8 ± 0.1); IO 5.9 - 7.4 ( = 6.4 ± 0.4); ED 5.5 - 6.9 ( = 6.3 ± 0.3); EN 5.2 - 5.8 ( = 5.4 ± 0.1); TD 2.6 - 3.1 ( = 2.9 ± 0.1); TL 31.1 - 34.7 ( = 33.3 ± 0.8); FL 22.6 - 25.8 ( = 24.2 ± 0.7).

Remarks: The presence of a mental gland is not externally evident in male paratypes of Myersiohyla neblinaria, but is evident under high power magnification (60x) or through dissections. Superficial dissections in male USNM 562074 reveal a roughly circular patch of glandular tissue in the dermis of the mental region (fig. 21). A dissection of female USNM 562072 did not indicate the presence of a mental gland.

The vocal sac appears to be single and subgular, as indicated by the slightly distended sacs of USNM 562073, USNM 562076-77, and USNM 562080. The forearm is robust and prominent in all males, and proportionally slender in females. The prepollex is equally enlarged in males and females (figs. 17, 22). Two specimens partially dissected when tissue samples were removed (USNM 562072 ♀, and USNM 562077 ♂) show a translucent peritoneum and unpigmented urinary bladder. The female (SVL 60.2 mm) has large, unpigmented ovarian eggs. The dissected right ovary (left one not dissected) contained 119 eggs with largest diameter of 2.7-2.9 mm (= 2.8; s = 0.08; n = 15). The male has elongated, almost cylindrical, unpigmented testes. Females are proportionally larger than males (SVL 54.0-61.6 mm, males 47.7-52.3 mm). Superficial dissections in male USNM 562074 reveal that the m. intermandibularis has a large diamond-shaped aponeurosis; its fibers barely contact the oval and araphic m. submentalis. Supplementary elements of the m. intermandibularis absent. The m. depressor mandibulae originates from the zygomatic ramus of the squamosal, and, contiguously, from the posterior margin of the tympanic annuli; the tympanic fibers are not differentiated as a slip. There are no fibers originating from the dorsal fascia at the level of the m. levator scapulae.

Ecology and natural history: Near Camp I, Buck found an adult female (USNM 562070) asleep during the day in a leaf axil of Bonnetia maguireorum (Theaceae) on a small ridge northeast of Camp I (fig. 23). At Camp VII, a gravid female (USNM 562072) was collected on a vine in the forest and produced copious amounts of a milky white exudate on its back after capture. Specimens from Camp XI were collected by McDiarmid, Cocroft, and Paolillo along a small (1-4 m wide) forest stream with large boulders and cascades. On February 25, 1985, USNM 562073 was recorded as it called beside a cascade in a clump of grass growing out of a sloping, mossy rock face about 1 m from the water; three other males (USNM 562074, USNM 562077-78) were on leaves and in grass clumps near cascades on sloping rocks along the same stream. USNM 56207576, USNM 562079-80, USNM 562082 were all collected from a rocky grotto along the same stream on the following night. An immature male (see below) was collected on a leaf of a small plant about 15 m from the stream that same night.

Two of three females (USNM 562070, SVL 59.8; USNM 562072, SVL 60.2) collected on February 10, 1984, and February 10, 1985, respectively, were gravid. A comparable-sized female (AMNH A-123715, SVL 61.6 mm) collected on November 29, 1984 also was gravid. The fourth female (AMNH A-131172), collected February 9–11, 1984, was smaller (SVL 54.0 mm) and had oocytes in very early stages of development, suggesting that she was approaching maturity Most collected males (SVL 47.7–52.3 mm, n = 10) had well-developed nuptial pads and presumably were capable of breeding. Calling males were heard in late February 1985 at Camp XI. A single smaller male (USNM 562081, SVL 40.0 mm) collected at the same time and place lacked nuptial pads and likely was immature.

Vocalization: The advertisement call of a male paratype (USNM 562073) calling alone was recorded in February along a cascading stream near Camp XI (fig. 24). The frog was calling at the base of a clump of grass, about 1 m from the water. The recording includes 31 notes being continuously given in 80 sec of tape (the recording may contain only the terminal part of a long call).

The call is a long train of well-spaced notes given at a rate of 23.4 notes/min. Note duration is about 0.3–0.4 sec (295–382 ms, = 338 ± 21 ms, n = 20). The individual note is composed of 7–9 pulses ( = 8.05 ± 0.69, n = 20 notes) that are resolvable even by narrowband analysis (fig. 24C). The individual note is a brief musical trill, being narrowly tuned with frequency rising about 113–151 Hz throughout the note from 1522–1560 Hz at initiation to 1673 Hz at the end (n = 9 notes analyzed).

Ayarzagüena and Señaris (“1993” [1994]) described the advertisement call of Myersiohyla inparquesi as a train of 240 ms, composed of two notes, the first with six pulses (90 ms) and the second with seven pulses (70 ms). The call was described as spanning the spectrum between 1200–2000 Hz, with a harmonic at 4400 Hz and the fundamental frequency at 1740 Hz. The call repetition rate was 20 notes/min. While no illustrations or temperature data were provided by Ayarzagüena and Señaris, their description suggests a call quite different from that of M. neblinaria.

Although Myersiohyla neblinaria and M. aromatica are readily diagnosable by several morphological characters (see diagnosis), the calls seem similar. The call of M. aromatica was described by Ayarzagüena and Señaris (“1993” [1994]), on the basis of recordings done at 19° C. The call was composed of single notes with an average duration of 420 ms, with 10–11 wellspaced pulses, a dominant frequency of 2000 Hz, and a bandwidth between 1250 and 2600 Hz. The note repetition rate was 44 notes/min.

Tadpole: The uncorrected pairwise distances of 16S sequences (table 2), between the tadpole (USNM 562732) and those of the holotype, and paratypes (USNM 562072, USNM 562074) are 0%–0.2%. Conversely, they are 12.3%–13.6% different from the sequences of available specimens of M. chamaeleo. Considering the identical or almost identical sequences, we consider this tadpole and all others with the same general morphology (see below) to be larvae of M. neblinaria.

Five larvae of Myersiohyla neblinaria in stage 25 were taken in the stream below Camp IX. They were found among rocks in shallow water (∼5 cm deep) in slow-moving portions of the stream by day and night. A few tadpoles presumed to be of the same species (greenish gray) were seen at Camp VII (Cocroft, field notes) but none were collected. Specimens from Camp XI are quite fragile (possibly because of problems with the formalin [low pH?] at the time of fixation), especially the largest individual (USNM 562370) that was illustrated shortly after its capture and shown in figure 25. Since that time, the specimen has progressively deteriorated, particularly in that its tooth rows are separating from the tooth ridges; otherwise the remainder of the specimen is in good condition. Based on notes on the LTRF taken when the specimen was illustrated and additional observations, we describe it below and compare it with other available specimens.

Description: A tadpole (USNM 562370, fig. 25) in stage 25 from Camp XI has the following measurements (all in mm): total length 76.0; body length 27.0; basal tail muscle height 7.8; basal tail muscle width 7.8; maximum dorsal fin height 2.8, this point being 20 mm away from body; maximum ventral fin height 2.8, this point being 24.4 mm away from body; body width 16.7; body depth 12.1; eye diameter 2.0; pupil diameter 0.8; interorbital distance 5.0; internarial distance 4.9; snout to nostril distance 6.5; snout to eye 2.5; snout to spiracle 21.7; oral disc diameter 14.9.

Oral disc enlarged and ventral; marginal papillae in single row, with contiguous papillae in offset disposition, but sometimes also forming second row. Few submarginal papillae scattered between distal labial tooth rows and labial edge. LTRF 16(16)/21(1). Gaps in rows A-16 and P-1 small, of about equal width. Teeth on proximal (closest to mouth) rows wider than teeth on distal rows; distal teeth 2–3 times more numerous per mm. Tooth density on P-l about 35 teeth/mm. Upper jaw sheath long and slender, with uniform and small serrations. Lower jaw sheath wide, V-shaped, about four times as wide as upper jaw sheath, and with low, rounded serrations.

Body dorsoventrally flattened anteriorly, ovoid posteriorly. Eyes small, dorsolateral. Snout slightly truncated in dorsal view, gently slopping in profile. Nostrils relatively small (about 60% of eye diameter), with fleshy rim and two medial, contiguous, triangular projections, anterior one quite prominent and about twice as large as the posterior one. Spiracle lateral, sinistral, its inner wall fused with body wall along its distal half; opening oriented slightly dorsally. Vent tube dextral, twice as long as wide, fused with lower fin; angular aperture. Dorsal fin originating on terminal eighth of the body, anterior to origin of tail; lower fin originating at base of tail. Neuromasts not visible but presumably present and undetected because of the almost complete loss of pigmentation as an artifact of fixation. Melanophores also not visible but presumably present (see below).

Some variation was observed in comparing the described specimen (USNM 562370) with four other smaller specimens also in stage 25 (USNM 562731: TTL 48 mm, LTRF 14/15; USNM 567732: TTL ? [tail taken for tissue], LTRF 14/18; USNM 562731 [2]: TTL 41 and 22 mm, LTRF 13/15 and 9/10, respectively). We did not observe the second, minor triangular projection from the nostril rim in these specimens. Further studies will determine whether these projections are the same structures noticed by Mijares-Urrutia (1992) and Sanchez (2010) on several species of Hyloscirtus. Coloration of the body of USNM 567732 (a tissue voucher that was frozen in the field in liquid nitrogen) when it was thawed and preserved (Sept., 2005) was overall greenish with extensive concentrations of pigment on the head and body and patches of iridophores above the eyes; scattered patches of greenish color on the snout; neuromasts not visible and perhaps absent; melanophores (not visible in the other field-fixed specimens) visible and rodlike.

The LTRF of 16/21 reported for the largest tadpole (fig. 25) represents the largest labial toothrow formula known; it was erroneously reported as 17/21 and the largest value in hylids by Altig and McDiarmid (1999b: 308). The only tadpoles having large formulae that are close to the one reported here are those of M. inparquesi (a maximum of 14/21), M. aromatica (a maximum of 13/18), and the Hyloscirtus armatus group (a maximum of 14/17, Cadle and Altig, 1991; Lötters et al, 2005). Interestingly, the smaller tadpoles of M. neblinaria also have smaller formulae (ranging from 9/10 to 14/18). This apparent positive association between increased total length and LTRF could be related to a similar observation recently reported by Sanchez (2010) in tadpoles of Hyloscirtus; however, the underlying mechanism of ontogenetic increase in the number of labial tooth rows requires further study.

Hyla loveridgei Rivero, 1961. Type locality: “Pico Culebra, Mt. Duida, 3000 it., Territorio Amazonas.”

Hyla ginesi Rivero, 1963. Replacement name for Hyla loveridgei, incorrectly considered a junior secondary homonym of Nyctimystes loveridgei Neill, 1954 (see La Marca and Smith, 1982).

Hyla loveridgei: La Marca and Smith, 1982.

Myersiohyla loveridgei: Faivovich et al., 2005. First combination with Myersiohyla.

Diagnosis: A species of Myersiohyla characterized by: (1) male SVL 38.2, females unknown; (2) two nuptial pads; (3) a row of tubercles along the forearm; (4) an externally evident mental gland in males; (5) thighs patterned with longitudinal bars; (6) dorsum with irregular brown blotches; (7) condition of the m. depressor mandibulae unknown; (8) eggs unknown; (9), (10), (11) tadpoles unknown.

Comparison with Other Species: Myersiohyla loveridgei is most similar to M. aromatica, from which it apparently differs by the presence of a distinct tarsal ridge (absent in M.aromatica; Ayarzagüena and Señaris, “1993” [1994]). The presence of two nuptial pads differentiates M. loveridgei from M. inparquesi, M. kanaima, and M. neblinaria. Dorsal color pattern and the presence of a row of tubercles along the forearm separate M. loveridgei from M. chamaeleo. The presence of an externally evident mental gland separates M. loveridgei from M. kanaima and M. neblinaria.

Description: The descriptions provided by Rivero (1961, “1971” [1972]) are adequate, but a few details are worth noting. Rivero (1961) described the nuptial pad of the holotype as “male with a brown rugosity on the pollex and inner finger.…” Later Rivero (“1971” [1972]) described the second specimen (UPRM. 2854) as having “a smooth nuptial excrescence that extends up to the second finger.” We had access only to the holotype MCZ 28565, an adult male (fig. 26). Like M. aromatica and M. chamaeleo, it shows two contiguous, discrete nuptial pads covered with minute epidermal projections. One of the pads is on the prepollex, and the other is on the medial-proximal surface of the first digit (fig. 27). The presence of a mental gland that occupies most of the gular region (fig. 28) is another characteristic overlooked by Rivero. Rivero (1961) reported that the holotype has the “vomerine odontoids forming a /--\ figure.” Our study of the same specimen suggests that the expression used by Ayarzagüena and Señaris (“1993” [1994]), “slightly S-shaped” is quite appropriate to describe the disposition of each vomerine tooth series. There are 12 teeth on the right and 13 on the left. Rivero (1961) described a slight tarsal fold in the holotype; we consider it more appropriate to describe it as a distinct tarsal ridge. Foot-webbing formula for the holotype is I 2–2⁺ II 1 ⁺-2½ III 1½–2½ IV 2⁺-1⁺ V. Some approximate measurements of the holotype (the specimen is quite contorted, all in mm) are: SVL 38.2; HL 14.1; HW 13.3; IND 3.1; IO 4.8; EN 3.7; ED 4.2; TD 2.1; TL 22.0; FL 16.4. Note that Rivero (1961) reported the SVL of the same specimen to be 42.0 mm.

Remarks: The former Hyla loveridgei as described by Rivero (1961) was based on a single male specimen from Cerro Duida, and that description was supplemented (Rivero, “1971” [1972]) with an additional male (SVL reported as 45 mm) from the same locality. Since its description, the species has rarely been mentioned in the literature (see Rivero, 1963; La Marca and Smith, 1982). While we still do not know any morphological synapomorphy for Myersiohyla, we find loveridgei to be very close morphologically (if not a senior synonym of M. aromatica, see below) to various species included in Myersiohyla, and for this reason we pose the testable hypothesis that loveridgei is congeneric. As morphological synapomorphies are discovered for Myersiohyla, or tissues of loveridgei become available, its generic status should be reevaluated.

We did not study specimens of Myersiohyla aromatica. Based solely on the thorough description provided by Ayarzagüena and Señaris (“1993” [1994]), the only character states differentiating this species from M. loveridgei seem to be the presence of a distinct tarsal ridge in the latter (Ayarzagüena and Señaris, “1993” [1994]). Since the distinctiveness of this ridge as described earlier in this paper is variable in M. neblinaria, the diagnostic value of the character is tentative and needs further evaluation. Furthermore, Ayarzagüena and Señaris (“1993” [1994]) reported 30–34 total vomerine teeth, whereas the holotype of M. loveridgei has 25. However, the lack of other specimens of M. loveridgei and potential variation in vomerine tooth counts precludes any conclusion regarding the diagnostic value of vomerine tooth number. Given that the type locality of M. loveridgei (“Pico Culebra, Mt. Duida, 3000 ft., Territorio Amazonas”) is approximately 40 km (straight line) from the type locality of M. aromatica (“Cumbre del Tepui Huachamacari, Estado Amazonas, Venezuela”), and that Ayarzagüena and Señaris (“1993” [1994]) did not compare the species, we tentatively consider both species as valid, but suggest that their status deserves further study.

Ecology and Natural History: The holotype was collected during the day “as it sprang from beneath (or near?) a rock in a treeless, rocky area of Mt. Duida” (Rivero, 1961: 109). The other known specimen was collected on the mossy stones along a fast-running stream (Rivero, “1971” [1972]: 185).

Advertisement Call: Unknown.

Tadpole: Unknown.

The second author collected an adult female specimen (USNM 550357) of Myersiohyla on the summit of Cerro Duida (1850 m, 03°25′N, 065°40′W). The point of capture is approximately 25 km SE (straight line) from Cerro Culebra, the type locality of M. loveridgei. The specimen has differences in pattern with the two specimens described by Rivero (1961; “1971” [1972]) and subtle morphological differences; for these reasons we are hesitant to assign it to M. loveridgei until more data on the variation of this species becomes available. The specimen (fig. 29) differs from the holotype of M. loveridgei and the specimen described by Rivero (“1971” [1972]) in having the thighs with irregular tan spots, instead of welldefined vertical bars, and in having the tympanic annuli more inclined medially. Some measurements of this specimen are (all in mm): SVL 47.2; HL 18.0; HW 17.4; IND 3.7; IO 6.0; ED 5.3; EN 4.5; TD 2.5; TL 26.4; FL 20.1. Other morphological characteristics include a row of tubercles along the arm, quite prominent palmar and plantar tubercles (including an enlarged inner metacarpal tubercle; see fig. 30), a well-developed ridge along the tarsus, and foot webbing formula I 2–2 II 1½-2 III 2⁺-2½ IV 2½-1⁺ V. Dorsally the background coloration is dark reddish brown, with multiple, irregular tan mottling. The discs are pale gray. A superficial dissection reveals a group of fibers of the m. depressor mandibulae that originates on the sheath of connective tissue that extends from the dorsal fascia into the space between the head and the body, with a few fibers originating at the level of the suprascapulae. This morphology is different from that seen in M. chamaeleo, as in that species the fibers originating at the level of the suprascapula form a distinct fan over the suprascapulae, while in Myersiohyla sp, most of the fibers originate in the connective tissue between head and body, with only the proximal part of approximately half of the fibers marginally reposing on the levator scapulae. The specimen is a gravid female. The dissected right ovary (left one intact) contained 47 eggs with largest diameters of 2.5–3.0 mm = 2.73; s = 0.15; n = 15), mostly unpigmented, with the exception of a reduced, pale brown animal pole.

Orejas-Miranda and Quesada (1976) referred to an unidentified hylid from the summit of Cerro Jaua and also mentioned tadpoles with numerous labial tooth rows. Further, they published a black-and-white photograph (their fig. 6) of an adult frog that unequivocally looks like a Myersiohyla. An adult male of Myersiohyla and some tadpoles with large LTRFs from Jaua are housed in the USNM (USNM 561349) and will be dealt with in a separate contribution.