Hyalinobatrachium taylori (Goin)

Myers, C. W. & Donnelly, M. A., 2008, The Summit Herpetofauna Of Auyantepui, Venezuela: Report From The Robert G. Goelet American Museum-Terramar Expedition, Bulletin of the American Museum of Natural History 2008 (308), pp. 1-147 : 30-41

Hyalinobatrachium taylori (Goin)

Centrolenella taylori Goin, 1968: 115 , fig. 1 (drawings of holotype in dorsal view, side of head, hand and foot). Holotype BMNH 1939.1 .1.65, from along New River , 750 ft elev., Guyana, collected by C. A. Hudson.

C [entrolenella] taylori: Señaris and Ayarzagüena , ‘‘1993’’ [1994]: 124, sector centro-oeste del Auyán-tepui (5 ° 549N, 62 ° 389W), 1850 m, June 1992 [one specimen fide Señaris and Ayarzagüena, 2005: 305].

Hyalinobatrachium taylori: Ruiz-Carranza and Lynch, 1991: 25 (new combination).

Hyalinobatrachium taylori is known from Auyantepui by a single specimen collected at 1850 m, in June 1992. It was found by Ayarzagüena in sympatry with H. crurifasciatum and Centrolene gorzulai . See Señaris and Ayarzagüena (2005: 221–232, 274, 305) for a full species account and illustrations.

SUPERFAMILY DENDROBATOIDEA (FORMERLY FAMILY DENDROBATIDAE ) FAMILY AROMOBATIDAE Anomaloglossus tepuyensis (La Marca) Figures 13–19, 21

Colostethus , undescribed species from Auyantepui, Myers, 1997: 3; Myers and Donnelly, 1997: 24; Grant et al., 1997: 8, 35, fig. 3B (photo of median lingual process).

Colostethus tepuyensis La Marca , ‘‘1996’’ [1998]: 49, fig. 11 (hand and foot). Holotype ULABG 2557 from a small stream ‘‘ entre Danto y Piñón, a casi una hora caminando desde Danto , en el trayecto desde Kamarata hasta las laderas del Auyantepui’ ’, Estado Bolívar, Venezuela, 1650 m, collected May 28, 1988, by D. Lussner and H.-P. Reinthaler. Myers and Donnelly, 2001: 34 (allocation of previous references to this name).

Anomaloglossus tepuyensis (La Marca) . Grant et al., 2006: 158 (new genus for species of Colostethus sensu lato with median lingual process).

This species was named in 1998 by La Marca on the basis of one specimen, an adult female from an elevation of 1650 m on the southeastern slope of Auyantepui. La Marca et al. (2002) subsequently mentioned the existence of several other specimens reported as a range extension of about 40 km to the NW (not ‘‘NE’’ as stated) of the type locality, at lower elevations of 390 and 445 m. in the Río Cucurital drainage near the northwestern base of Auyantepui (map 1).

Based on a sample of 33 specimens and several lots of tadpoles from the summit of Auyantepui, a fuller characterization of adult morphology and descriptions of the advertisement call and larva can now be given.

NEW DESCRIPTION

MATERIAL: 1.1 km NNE Camp 1, 1650 m: AMNH A-164834–164839 (larvae), EBRG 2649 (7 larvae). Camp 1, 1700 m: AMNH A-164817–164822, EBRG 2693– 2699. Camp 3, 1850 m: AMNH A-164823, 164840–164843 (larvae), EBRG 2650 (8 larvae), 2700. Camp 4, 1600 m: AMNH A- 164824–164833, 164844–164848 (larvae), EBRG 2651 (15 larvae), 2701–2709. All from the 1994 AMNH –TERRAMAR Expedition to Auyantepui.

DIAGNOSIS: A small frog (8 about 21– 24 mm, ♀ about 24–28 mm SVL) with an elongated median lingual process; no cloacal tubercles; fingers with keel-like lateral folds, broadest and best developed on fingers II and III; feet moderately webbed, toes with folded flaplike fringing; dorsal blotched pattern; an

incomplete oblique pale lateral line either present or absent. See Discussion and Other Comparisons.

MORPHOLOGY: Adult male Anomaloglossus tepuyensis about 21–24 mm SVL, adult females larger at about 24–28 mm SVL (table 1). Sexual maturity was judged by the presence of open vocal slits in males (open only on right in two specimens of 22 mm SVL), and by the presence of enlarged pigmented ova and convoluted oviducts in females.

Dorsal skin granular in life, becoming somewhat tuberculate on hind limbs (fig. 13), but skin much smoother in preservative; ventral skin virtually smooth in preservative, weakly granular on throats of some males. Dorsal skin forming a welldefined rounded, posteriorly projecting flap well above vent, which opens at upper level of thighs. Proximal thighs coarsely granular below vent, but no pronounced, symmetrically arranged tubercles in cloacal region.

Head slightly wider than long; greatest head width (between angles of jaws) 33 % – 39 % of SVL. Snout sloping, rounded to bluntly pointed in profile, truncate to broadly rounded or bluntly pointed in dorsal and ventral view. Nares situated near tip of snout and directed slightly posterolaterally; nares visible from front, barely visible from above or below; posterior rim of naris bordered behind by a crescent-shaped, finely incised groove; posterior rim not raised but usually bearing a small tuberclelike prominence posterodorsally to naris. Canthus rostralis slightly rounded; loreal region weakly concave, sloping slightly outward to lip. Interorbital region much wider than upper eyelid. Snout subequal to eye length; distance from center of naris to edge of eye/eye length 5 0.55–0.74. Usually two low but inconspicuous postrictal tubercles. Tympanum usually inconspicuous, concealed posterodorsally by a diffuse supratympanic bulge; tympanum less than half of eye length.

Hand moderate, its length 24 % –29 % of SVL, 62 % –82 % of greatest head width. Relative lengths of appressed fingers III. IV. II. I; disc of finger I either failing to reach or else overlapping disc of finger II. Discs of all fingers moderately expanded; third finger disc in adults 1.4–2.0 times wider

TABLE 1

Size (in mm) and Proportions of Anomaloglossus tepuyensis La Marca , from the Summit of Auyantepui a

than distal end of adjacent phalanx, average width greater in females. Third finger broadly keeled or fringed but not swollen. Keel-like lateral folds on fingers usually low and inconspicuous except on finger III and, to a lesser extent, finger II; these keel-like structures appear to be tightly folded ventromedially around the fingers—with the edge of the fold recurved in a groove along the ventral side of the finger, usually best developed on medial side of finger III (figs. 14, 21); no obvious sexual dimorphism in finger keeling, which is also detectable but inconspicuous in small juveniles.

Base of palm with large median metacarpal tubercle, generally distally flattened, but in some specimens slightly cordiform or rounded; smaller elliptical inner metacarpal tubercle on base of first finger; one or two low, rounded subarticular tubercles (one each on fingers I and II, two each on fingers III and IV, with distal tubercle tending to be inconspicuous on IV). A faint outer metacarpal fold extending from lateral side of fourth finger to the large palmar tubercle, sometimes bearing a small tubercle. No ulnar tubercles or fold. Often a small tubercle situated below palmar tubercle toward lateral side of wrist (fig. 14). No fleshy supracarpal pad atop wrist.

Hind legs moderately long, with heel of adpressed limb reaching from between eye and snout to beyond snout; tibia 49 % –55 % of adult SVL, relatively shorter in females (table 1). Relative lengths of appressed toes IV. III $ V. II. I; first toe short, reaching or failing to reach base of subarticular tubercle of second toe. Toe discs moderately expanded (sometimes only weak- ly expanded on first toe), subequal or larger than finger discs. Feet moderately webbed, the webbing distally continuous with distinct lateral fringing on all toes; the lateral fringing is folded downward along sides of toes (‘‘folded flaplike fringes’’, see Discussion and fig. 14). Modal webbing formula of adults: I 1 K –2 K II 1 K –3 III 2 K –3 K IV 4–2 K V. One specimen has a short first toe with only the first phalanx (disc) free of web; otherwise there is slight variation only between the third to fifth toes: III (2–3)– (3 K –4) IV (2–4)–(2–2 K) V.

A small, usually elliptical inner metatarsal tubercle, and a small, usually round outer metatarsal tubercle subequal in size (either may be equal to or larger than the other). One to three nonprotuberant subarticular tubercles (one each on toes I, II, two each on III and V, three on IV). The fringe along outer edge of the fifth toe is continuous with a weak outer metatarsal fold that terminates at the outer metatarsal tubercle. Distinct fringe along free edge of first toe is continuous with an oblique tarsal fold extending proximolaterad on the distal half of the tarsus (fig. 14); the nearly straight tarsal fold terminates abruptly in a curved tuberclelike structure (similar to condition shown in fig. 20A [ A. tamacuarensis ]).

Teeth present on maxillary arch. Tongue longer than wide; free posteriorly, with rounded margin; a usually prominent median lingual process is either pointed or distally rounded and is present in all specimens including juveniles. Vocal slits large, each extending from edge of tongue nearly to angle of jaw.

COLORATION: In life (fig. 13), the dorsal ground color of Anomaloglossus tepuyensis ranged from grayish brown to orangish brown and bronzy brown in adults, mostly orange-brown in subadults, and brown or gray-brown with conspicuous white flecking in small juveniles—all with irregularly pigmented dark brown or blackish brown markings (obscure or indefinite only in a few specimens) usually arranged in the following pattern (fig. 15): (1) a dark interorbital bar, often with a slightly concave (sometimes straight) anterior margin, with or without a posteriorly produced apex; (2) a Vshaped to cordiform (transversely straight in a few) median dark blotch between arms, sometimes connected with the following; (3) a transverse pair of smaller paravertebral blotches at midbody—these markings sometimes fused with the anterior blotch to form a vaguely X-shaped dorsal mark; (4) usually a single small median posterior blotch near end of body.

The dorsal dark markings listed above are formed from irregular distributions of melanophores and usually contain pale areas of brown ground color; the paired paravertebral blotches and posterior median blotch often are circular with pale centers (ocellus-like). The snout anterior to the interorbial marking is brown like the rest of the dorsum, often with one or a few small dark marks or blotches, but the top of the snout is unmarked in about 27 % of the sample and is never darkened overall. The eyelids occasionally have a few unpigmented dots but are usually unicolor except where crossed by the interorbital bar.

An incomplete pale oblique lateral line (in life, pale brown to blue) present or absent; extending from groin and terminating dorso- laterally behind arms when distinct, but often poorly defined and broken, or completely absent (frequently present on one side but absent on the other). A blackish stripe or face mask along side of snout passing through eye and widening above shoulder, tending to continue posteriad just below dorsum but changing along its lower edge on flanks to indefinite brown mottling or reticulum with irregular or ill-defined pale spots of bluish white, pale blue, or pale bronze, this area interrupted or not by a pale oblique lateral line of similar color. Pale spotted pattern of flanks extending to ventrolateral edge of venter (no hint of a pale ventrolateral stripe).

Black face stripe sometimes continuous around tip of snout, which otherwise varies from uniform brown to pale spotted. Upper lip pale to dark brownish, variously and indefinitely marked with small pale areas or spots (sometimes bluish white) or irregular dark blotching. Often a bronze (white to pale brownish in preservative) stripe extending obliquely from eye to forearm, just below the postocular widening of the black face mask; this pale postocular stripe distinct in some specimens, virtually absent in others in which it may lack contrast with a pale lip or else be interrupted and obscured by infusion of dark pigment.

Forearms and hind legs brown with distinct to vague dark bars or irregular spots, and with a tendency toward orange suffusion on the thighs. Often a dark brown stripe along anterior edge of upper arm, sometimes extending forward to corner of mouth, with a similar horizontal brown stripe varying from distinct to absent along anterior surface of thigh. Posterior thigh brown with indefinite pattern of pale flecks or small spots, only rarely with an ill-defined horizontal brown line or stripe on one or both sides (EBRG 2695, 2696, 2701). Some specimens have conspicuous pale spots or crossbands on the joints of the digits of hands and feet, but this pattern usually vague or digits uniformly pigmented. Digit tips noted as being ‘‘white marked’’ in life, but in preservative all digital scutes tending to be completely dark above, only occasionally with a pair of small inconspicuous pale spots.

Ventral color varies ontogenetically in life: (1) small juveniles are green or greenish gray, with silvery white flecking; (2) larger juveniles had all ventral surfaces orange except greenish on underside of thigh and shank and gray on palms and soles; (3) a few adult males at camp 1 were greenish gray ventrally becoming yellowish green under the limbs, whereas an adult male at camp 3 had the throat and chest gray with conspicuous white flecking that continued along the sides of the belly, the belly and undersides of limbs otherwise bright greenish yellow; ventral color at camp 4 (including adult males and females) was mostly gray with white flecking on throat and/or chest, turning bright greenish yellow or even (in a few) golden yellow posteriorly and under limbs, but a few adults had even the throat and chest bright colored like the belly and undersides of limbs. (Some degree of sexual dimorphism in ventral coloring is probable, but specimens could not be reliably sexed in the field.)

In preservative, ventral surfaces (fig. 16) vary from pale tan with a weak suffusion of melanophores, most dense on throat, to overall darker brown owing to dense suffusion of melanophores, which sometimes form a reticulum around pale spots. The axillary region tends to be weakly pigmented or unpigmented (but not forming in life a flash mark differently colored from ventral surfaces), this area being continuous ventrally with a small unpigmented area proximal to the arm insertion, which stands out as a small distinct spot (except in specimens with the palest venters). Undersides of hands and feet dark brown.

Iris pale bronze or orange-bronze, darkened with black suffusion, sometimes paler below pupil than above; pale bronze or orange-bronze pupillary ring usually interrupted by a black tick mark below pupil. Nictitating membrane translucent, with brown pigment along free edge.

TADPOLES

The following description is based on Anomaloglossus tepuyensis larvae in stage 36 (fig. 17), followed by notes on ontogenetic change. See table 2 for additional measure- ments. See Natural History and Vocalization for larval habitats.

HABITUS AND PROPORTIONS: Head and body elliptical and rounded at both ends in dorsal view; midbody width 70 % –75 % of head-body length. Head-body depressed (midbody depth 76 % –80 % of midbody width), somewhat flattened dorsally and ventrally. Snout rounded in dorsal and lateral view. Eyes dorsal, directed laterally; interorbital distance (between centers of pupils) 3.9 mm. Nares dorsal, directed laterally; internarial distance (between centers of nares) 2.6 mm. Nares approximately equidistant to tip of snout and anterior edge of eye. Spiracle sinistral, directed posterodorsally, positioned low on body, a tube (1.5–2.1 mm long) with the medial wall attached to the body; opening 6.9–7.6 mm from snout, at 55 % –58 % of head-body length. Vent tube dextral to caudal fin.

Lateral line system with a supraorbital and infraorbital branch originating from near upper labium and connecting behind eye. Angular branch originating ventral to the infraorbital branch near the upper lip and continuing posteriorly as the middle body branch; a branch connecting the angular

164837). Scale lines 5 10 mm above and 1 mm below.

TABLE 2 Measurements (in mm) of Anomaloglossus tepuyensis Tadpoles (Values are ranges, followed below by means ± 1 SD when N. 2)

branch to the infraorbital branch posterior to the eye, this branch continuing ventrally onto the ventral surface of the body anterior to the spiracle. Superior trunk branch continuing onto tail in the dorsal fin just above the musculature, for about 25 % of tail length.

Tail 67 % –68 % of total length, with low fins, maximal height approximately equal to body depth; dorsal fin not extending onto body; dorsal fin higher than ventral fin; tail tip rounded, extending slightly past tip of musculature.

PIGMENTATION: In preservative, headbody brown dorsally and laterally; ventral body clear, flecked with melanophores and some light brown blotches; tail musculature tan with irregularly shaped dark brown blotches and flecks; ventral fin lacking pigment near body, with dark blotches and flecks posteriorly; dorsal fin dark near body, with flecks and blotches posteriorly.

MOUTH PARTS: Mouth ventral. Oral disc laterally emarginate. Labial teeth in 2/3 rows; anterior rows nearly equal in length, A2 row with median gap; posterior rows 2 and 3 equal in length, slightly longer than P1; P3 not as heavily keratinized as P1 and P2. Upper jaw sheath a broad arch with slender lateral processes and blunt serrations; lower jaw sheath V-shaped with blunt serrations like those on upper jaw. Marginal papillae from lateral edges of upper labium in single row above emargination; in weak double row below emargination and then single across middle of lower labium; papillae on lower labium more pointed than those on upper labium.

ONTOGENETIC CHANGES: Changes in body size are summarized in table 2. The ratio of head-body length to total length decreases from 38 % in stage-25 tadpoles to 31 % in stage- 36 specimens. Body width increases during development relative to oral disc width; the ratio of oral disc width to body width changes from 55 % in stage-25 tadpoles to 42 % in the single stage 40-specimen. The lateral line system is not visible in several stage-25 tadpoles, but is distinct by stage 29.

In several stage-25 tadpoles, labial tooth row P3 is lightly keratinized (this row commonly is the last to be keratinized in dendrobatoid larvae [e.g., Myers et al., 1978; Donnelly et al., 1990; Myers and Donnelly, 1997]).

There is some variation in the marginal papillae on the lower labium (a single row or double rows), but we did not detect ontogenetic change in the character. (Ontogenetic change from one to two rows seems to be common among dendrobatoids.)

NATURAL HISTORY AND VOCALIZATION

As would be anticipated from the fringed toes and moderately webbed feet, Anomaloglossus tepuyensis is a riparian frog. At Camp 1, a few individuals were found in vegetation near the river’s edge below camp, but they were most common in and along small shallow, sand-bottomed streams within deep sandstone crevices near camp (figs. 4, 18). They would dive into shallow pools and attempt to hide on the bottom; they were calling infrequently but were difficult to approach and were not recorded at this locality. They were somewhat easier to find along a forest stream at Camp 4, either active or under rocks, and a recording was obtained.

Nurse frogs transporting tadpoles were not seen. Three lots of free-living larvae were dipnetted from pools near three camps. Tadpoles were not collected where frogs were found at Camp 1, but a series of larvae in stages 25–40 were taken from a pool in the river 1.1 km NNE of Camp 1. At Camp 3, larvae in stages 25–33 were taken from an isolated rocky sandstone pool, about 50 cm deep, next to a stream. At Camp 4, larvae in stages 25–32 were netted in the backwater of a river. Tadpoles of Hypsiboas jimenezi were in the same pools as Anomaloglossus near Camp 1 and at Camp 4.

A tape recording was made of two or more males calling at Camp 4. The advertisement call (fig. 19) is a short train of 14–22 notes (x¯ 5 16.0 notes, N 5 8), which are heard as a series of rapid ‘‘peeps’’ given at a variable intracall rate of about 4–7 notes per sec. Calls begin slowly and then speed up. In the illustrated call, the interval is 0.36 sec between the initial two notes and 0.07–0.13 sec between the last several notes. Note length varies from about 0.06 to 0.10 sec within the call.

Narrowband analysis shows a very slight rise in frequency at note initiation, after which the frequency is narrowly tuned and remains nearly flat, varying from about 3270 Hz to 3580 Hz. A few simultaneously calling males show nonoverlapping differences of about 300 Hz. Also, there is slight internote variation in frequency within a call. Waveforms and wideband spectrographic analyses show individual notes within single calls to vary from nonpulsatile to weakly pulsatile, with 2–4 distinct pulses per note.

COMPARISON WITH THE DESCRIPTION OF THE HOLOTYPE OF ANOMALOGLOSSUS TEPUYENSIS (LA MARCA)

We recognized in 1994 when collecting apparently the first specimens of a dendrobatoid from the summit of Auyantepui that we were sampling an unnamed species (Myers, 1997). La Marca (‘‘1996’’ [1998]) has independently provided the specific name tepuyensis , based on a specimen collected in 1988 at 1650 m elevation on the southern slopes of Auyantepui. It seems probable that only a single species is involved, although one aspect of the described color pattern of the holotype of tepuyensis does not agree with the observed variation in summit material. Some relevant points for comparison, from La Marca’s (‘‘1996’’ [1998]: 52–53) description of the holotype (specimen not examined by us), include the following (free translation):

Dorsum dark brown with irregularly distribut- ed blackish marks. These are very small and inconspicuous, so the dorsum at a glance shows no evidence of color pattern. Abundant blackish and some dirty whitish marks on the upper part of the head, making the head seem darker than the rest of the dorsum. This pattern closes behind in a V-shape, from the posterior part of the eyelids with the apex on the occipital region. The eyelids show nearly circular marks that perhaps correspond in life with flat tubercles …. Arms and forearms lighter than the dorsum. The first with a dark brown band that extends to the posterior part of the lips, the second with dark brown marks not forming bracelets …. There is a dark longitudinal stripe on the posterior part of the thighs. Upper part of thighs and tibias the same as the dorsum; there is a dark spot on the middle of the tibia that does not form a [cross]band …. [In the recently preserved holotype]: Dorsal surfaces dark brown, the ventral surfaces yellowish, with intense yellow color on the posterior part of the venter and undersides of thighs, tibias and forearms. There is a yellowish cream band from the loreal and infratympanic regions to near the insertion of the front leg.

The posterior V-shape of the dark head of the holotype obviously corresponds to the posteriorly projecting rear edge of the interorbital bar in many of our specimens, but none of the 33 specimens from the summit collections has the overall head darker than the body; the interorbital bar, when present, is always distinct from any anterior blotching atop the snout (which is unmarked in over 25 % of the sample). Except where crossed by the interorbital bar, the eyelids are unicolor brown except in occasional specimens having a few or several pale dots on each eyelid, possibly but not clearly equivalent to the manchitas casi circulares of the holotype. The dorsal markings in summit specimens are somewhat variable but nonetheless are arranged in a distinct pattern as described above and as shown in figure 15. However, a few summit specimens are virtually patternless (fig. 15, lower right) and in that regard may approach the condition of the holotype.

The holotype has a few dark marks atop the lower arms and on the tibia that do not form crossbands, and the thighs are pale above; the summit specimens tend to have the forearms and hind legs crossbanded, but it is a tendency only and the markings may be vague. Only three summit specimens were noted as having an ill-defined horizontal brown stripe on the rear of the thigh, and two of these had it on one limb only.

The recently preserved holotype had yellow ventral surfaces becoming intensely yellow posteriorly and on the undersides of the limbs; such coloration occurred in a few summit specimens. The yellowish band from the loreal region to the forearms likewise occurs within the variation of the summit specimens, some of which have the labial and infratympanic region pale to the forearm, whereas others do not (fig. 13).

La Marca stressed the absence of a pale oblique lateral line as a diagnostic character of tepuyensis (he also mentioned its presence in the second paragraph of the diagnosis, but reaffirmed the absence in the detailed description following). An oblique lateral line is present in some summit specimens (fig. 13), but it is rarely sharply defined, being often vague or broken and it is frequently absent.

La Marca (‘‘1996’’ [1998]: 49, 50) described the tip of the snout as being broadly rounded (ampliamente redondeada) in dorsal view; however, he inconsistently described the snout tip as being laterally rounded (redondeada en vista lateral) in the diagnosis, but laterally protuberant (protuberante en vista lateral) in the description. In the summit sample, the sloping snout in profile varies from rounded to bluntly pointed, and in dorsal view from from truncate to broadly rounded or bluntly pointed.8

In summary, although the coloration and color pattern of the holotype of Anomaloglossus tepuyensis is not typical of summit specimens, most aspects of color and pattern seem to fit within observed variation in the 33 specimens from the summit of Auyantepui. Only the overall darker head of the holotype seems inconsistent in this regard. The female holotype was measured at 28.3 mm SVL, which at first glance seems well outside the range for six summit females (24.3–26.5 mm). However, leaving aside the confounding

8 Such variation in snout shape seems more or less normal at least in some sections of ‘‘ Colostethus ’’ sensu lato. La Marca used the conditions of ‘‘truncado o casi truncado’’ vs. ‘‘no truncado’’ in the second couplet of a key to eastern Venezuelan Colostethus , but the character is unreliable in these frogs. In addition to normal variation, snout shape is sometimes altered in preservation, as when old, poorly fixed specimens have been crowded in a container. Snout shape also can be subjectively variable depending on angle of view—whether viewed dorsally at a right angle to the long axis of the body (as in fig. 15), or dorsally at a right angle to the upwardly inclined head.

effects of slight differences in measuring technique between investigators (as well as normal measuring error), adding the holotype gives a total range of 4 mm for adult females, which does not seem excessive for frogs of this size.

Therefore, so far as we can judge at this time, our population samples of Anomaloglossus from the summit of Auyantepui are conspecific with the holotype of A. tepuyensis . If the additional specimens recently mentioned by La Marca et al. (2002) are indeed of this species, Anomaloglossus tepuyensis has a noteworthy elevational range of 390–1850 m in the Auyantepui area and is not strictly a ‘‘tepui species’’.