Part 4

The type of _Hyla depressa_ is a male having a snout-vent length of 69.8 mm. The diameter of the tympanum is 5.2 mm, 77.6 percent of the diameter of the eye. The dorsal roofing bones of the skull are moderately exostosed, and the lateral edges of the frontoparietals are elevated. The skin on the dorsum is tuberculate. The dorsum is dull brown with a broad darker brown longitudinal mark having indistinct lateral edges from the snout to the post-sacral area. A narrow cream middorsal line extends from the snout to the vent. The side of the head is dark brown, palest posteroventral to the orbit. The posterior surfaces of the thighs are dull brown; the flanks are pale brown, and the ventral surfaces are pale creamy tan. Dark brown transverse bars are present on the limbs. When we examined the type on 3 January 1969, it was in excellent condition. Andersson (1945:75) contrasted the type of _Hyla depressa_ with _leprieurii_ and _buckleyi_, but he did not compare his specimen with _taurinus_, from which it exhibits no distinguishing features.

_Osteocephalus taurinus_ is a widespread and variable species, and it has received several specific names. We are convinced that _Osteocephalus taurinus_ Steindachner, 1862, is the oldest available name for this large Amazonian species. The following names are junior synonyms: _Osteocephalus flavolineatus_ Steindachner, 1862; _Osteocephalus planiceps_ Cope, 1874; _Hyla (Trachycephalus) vilarsi_ Melin, 1941; _Hyla depressa_ Andersson, 1945.

_Diagnosis._--1) Size large; sexual dimorphism evident; maximum observed snout-vent length in males 84.6 mm, in females 104 mm; 2) skin on dorsum in males bearing many moderately large, spinous tubercles; 3) skin on flanks smooth; 4) web extending to middle of antepenultimate phalange on inner edge of third finger; 5) dorsum brown usually with a large medial dark brown blotch or, less frequently, several dark spots; narrow middorsal yellow line present in some; 6) venter cream or tan with or without small, irregular brown flecks; 7) lips brown with vertical cream bar below eye in some, expanded into pale labial stripe posteriorly in some females; 8) flanks tan or cream with or without small, irregular brown spots; 9) dermal roofing bones of skull exostosed, casqued, and co-ossified (in large adults); 10) dermal sphenethmoid present; 11) nasals juxtaposed medially; 12) anteromedial margin of frontoparietals at mid-level of orbit; 13) frontoparietal fontanelle covered; 14) palatine serrate; 15) parasphenoid bearing odontoids; 16) zygomatic ramus of squamosal usually articulating with maxillary arch; 17) transverse processes of third presacral vertebra approximately equal in width to sacral diapophyses; transverse processes of presacral vertebrae 3-8 subequal in width; 18) intermandibularis and submentalis muscles connected; 19) supramandibular portion of interhyoideus extensively developed; associated skin forming everted pouch.

The moderately rugose dorsum (in males), large size, extensive webbing on the hand, and frontoparietal flanges in adults serve to distinguish _taurinus_ from other members of the genus.

_Distribution._--The Amazon Basin, the upper Orinoco Basin, and the Guianas. Most localities are below 500 m, but the species ascends the lower Amazonian slopes of the Andes to elevations of about 1000 m (Fig. 11). A record from Caracas, Venezuela, and those from Provincia Carchi and Provincia Esmeraldas, Ecuador, are considered to be erroneous. The latter specimens were included in a collection sold to the University of Illinois; contained in the collection are many common Amazonian species unknown from the Pacific lowlands. 516 specimens from 151 localities.

_Remarks._--This widespread species is highly variable in size and coloration. Striking differences in snout-vent length are evident in series from various parts of the range. The smallest calling males (CAS-SU 12351-6 from Rio Tapirapé, Brasil) have snout-vent lengths of 46.5-60.3 (mean 53.3) mm, whereas the largest (FMNH 140254, KU 92243-6, WCAB 9997, 10001, 10003-4 from Igarapé Marmelo, Brasil) have snout-vent lengths of 71.5-84.6 (mean 77.6) mm. Mean values of snout-vent lengths of males from other localities are: Río Pastaza drainage, Ecuador 73.8 mm, Surinam 67.7 mm, Río Ucayali drainage, Perú 57.6 mm, and Guyana 55.5 mm. Although the difference between the smallest and largest adults is highly significant, populations bridging the gap do exist. Furthermore, the geographic arrangement of small versus large frogs is a confusing mosaic. We have entertained the thought that we have included more than one species in _taurinus_, but on the basis of preserved specimens we are unable to detect consistent differences distinguishing two or more taxa.

The coloration and pattern of _taurinus_ are so variable that no one series of statements can describe samples drawn from the entire range of species. We have been unable to determine geographic trends in color pattern; instead the variation within a given sample can encompass the variety known in most other samples. Two minor exceptions do exist. A narrow middorsal light stripe is present in some individuals from throughout the range, but striped specimens are most common in the upper Amazon Basin. The absence of dorsal markings is uncommon in the entire species, but it is most frequent in individuals from the Guianas. A few individuals, such as KU 105230, have scattered white spots on the dorsum.

The coloration of four males in life from Lago Agrio, Ecuador (KU 126652-5) was: "Dorsal ground color tan to dark brown with darker brown markings. Flanks creamy tan to yellow with brown or black flecks or mottling. Venter uniform creamy yellow or yellow with brown spots or reticulations. Iris greenish yellow with radiating black streaks and a median, horizontal reddish brown streak." (W. E. Duellman, field notes, 12 May 1969.) A female from Santa Cecilia, Ecuador (KU 123173), was: "Dorsum mottled olive-green and tan. Flanks tan with brown spots. Belly and throat creamy white, becoming tan posteriorly. Edge of upper jaw olive-green." (W. E. Duellman, field notes, 16 June 1968.) Another female from Santa Cecilia (KU 123175), was: "Brown dorsally with cream-colored mottling. Transverse bars on legs darker brown with cream-colored edges. Margin of upper lip creamy yellow. Anterior and posterior surfaces of thighs tan. Flanks white with brown spots. Venter creamy white. Iris greenish bronze with heavy radiating reticulations of black." (W. E. Duellman, field notes, 22 July 1968.)

The tendency for females to have a labial stripe posteriorly and the absence of dorsal tubercles in females has resulted in the identification of many such specimens as _O. leprieurii_.

Ontogenetic change in coloration is slight in _taurinus_. Most juveniles (less than 40 mm in snout-vent length) can be identified readily. There is a tendency for the dorsal markings of juveniles to consist of several small spots. Apparently with growth the spots usually coalesce, forming a large median blotch, but some adults retain the juvenile pattern. Cochran and Goin (1970:251) erroneously identified several juveniles from Colombia as _Hyla palpebrogranulata_ Andersson.

=Osteocephalus verrucigerus= (Werner)

_Hyla verrucigera_ Werner, 1901:601 [Holotype.--ZMB 16589 from "Ecuador"; Richard Haensch collector].

_Hyla riopastazae_ Andersson, 1945:72 [Holotype.--NHRM 1960 from Baños, Río Pastaza, Provincia Tungurahua, Ecuador; William Clarke-MacIntyre collector].

_Hyla orcesi_ Funkhouser, 1956:78 [Holotype.--CAS-SU 13150 from Río Pacayacu, tributary of Río Cotapino, Provincia Napo, Ecuador; collector unknown].

_Osteocephalus orcesi_--Cochran and Goin, 1970:317.

_Osteocephalus verrucigerus_--Trueb and Duellman, 1970:601 [Synonymized _Hyla riopastazae_ Andersson, 1945, and _Hyla orcesi_ Funkhouser, 1956, with _Hyla verrucigera_ Werner, 1901].

_Justification of Synonymy._--Trueb and Duellman (1970:605) discussed the assignment of the names in the synonymy of _O. verrucigerus_; only a brief resumé is given here.

The extant type of _Hyla verrucigera_ is a juvenile male having a snout-vent length of 32.0 mm. The dorsum is smooth except for tubercles on the eyelids; the skin is loose, and the body is soft. The specimen is faded to a pale brown; indistinct dark spots are present on the back, and transverse bars are evident on the limbs.

The holotype of _Hyla riopastazae_ is a gravid female having a snout-vent length of 64.7 mm. The dorsum is smooth. The dorsal ground color is pale brown with indistinct brown transverse bars on the limbs. The throat, chest, and belly are cream with brown spots and mottling.

The holotype of _Hyla orcesi_ is an adult male having a snout-vent length of 52.6 mm. The dorsum is heavily tuberculate. The dorsum is dark brown with faint transverse bars on the forearms and feet; the ventral surfaces are creamy brown.

Trueb and Duellman (1970) provided conclusive evidence that the types of _H. verrucigera_, _riopastazae_, and _orcesi_ are a juvenile, adult female, and adult male, respectively, of one species, the earliest available name for which is _Hyla verrucigera_ Werner, 1901.

_Diagnosis._--1) Size moderate, sexual dimorphism evident; maximum observed snout-vent length in males 54.3 mm, in females 65.8 mm; 2) skin on dorsum in males bearing large, keratinized tubercles; 3) skin on flanks smooth; 4) web extending to base of antepenultimate phalange on inner edge of third finger; 5) dorsum uniformly dark brown or black, with tan snout in females; 6) venter creamy white, heavily mottled with black or dark brown, especially in females; 7) lips marked with pale tan labial stripe and suborbital bar; 8) flanks dull reddish brown; 9) dermal roofing bones of skull lacking exostosis; 10) dermal sphenethmoid absent; 11) nasals widely separated medially; 12) anteromedial margin of frontoparietals at anterior border of orbit; 13) frontoparietal fontanelle covered; 14) palatine serrate; 15) parasphenoid bearing odontoids; 16) zygomatic ramus of squamosal extending approximately one-half of distance to maxillary arch; 17) transverse processes of third presacral vertebra approximately equal in width to sacral diapophyses; transverse processes of presacral vertebrae 3-8 subequal in width; 18) intermandibularis and submentalis muscles connected; 19) supramandibular portion of interhyoideus forming simple, tubular, posterolateral extension; associated skin unmodified.

_Osteocephalus verrucigerus_ can be distinguished from other members of the genus by its uniformly dark dorsum, heavily mottled venter, and large, spinous tubercles on the dorsum in males.

_Distribution._--Lower Amazonian slopes (500-1840 m) of the Andes and on the western fringe of the Amazon Basin in Ecuador and Perú; one locality (Acevedo) in upper Río Magdalena drainage in Colombia (Fig. 9). 40 specimens from 13 localities.

_Remarks._--In life the dorsum in males is dull olive-green; the groin, anterior and posterior surfaces of the thighs, inner surfaces of limbs, and upper arms are dark brown. The ventral surfaces of the limbs are pinkish tan; the other ventral surfaces are pale creamy tan with reddish brown spots. The suborbital spot is pale greenish tan, and the iris is deep reddish brown. In females the dorsum is dull olive-brown; the anterior part of the head is tan, and the suborbital spot is yellowish tan. The groin and hidden surfaces of the limbs are dark reddish brown. The ventral surfaces of the limbs are brown; the throat and chest are creamy white, and the belly is reddish tan, both with dark brown mottling.

Considerable ontogenetic change occurs in coloration. Juveniles are pale above with a dark median dorsal blotch and dark transverse bars on the limbs. The venter is white. The change consists principally of an increase in dark pigment and subsequent obliteration of the juvenile pattern.

Tadpoles of this species have moderately long tails with low fins, robust bodies, two rows of labial papillae with median part of the upper lip bare, and two upper and five lower rows of teeth. Trueb and Duellman (1970) described the eggs, tadpoles, mating call, and variation in the adults.

GENERIC RELATIONSHIPS

Among the 33 genera currently recognized in the family Hylidae, there are two basic types of vocal sac structure (Duellman, 1970b), namely the subgular type and the lateral type. Only four hylid genera, all Neotropical lowland groups, are known to possess paired lateral vocal sacs; these are _Osteocephalus_, _Argenteohyla_, _Phrynohyas_, and _Trachycephalus_. The geographical distributions and morphological characteristics of these four genera suggest that they are more closely related to one another than with any other hylid genera.

Of the four genera, _Osteocephalus_ is the most generalized in morphology, and, like _Phrynohyas_, has no specialized habits. _Osteocephalus_ and _Argenteohyla_ are similarly distinguished from _Phrynohyas_ and _Trachycephalus_ on the basis of vocal sac structure. The vocal sacs of _Osteocephalus_ and _Argenteohyla_ are posterior and protrude posterolateral to the angles of the jaws when they are inflated, whereas those of _Phrynohyas_ and _Trachycephalus_ are more lateral and protrude posterior to the angles of the jaws when inflated.

Although _Osteocephalus_ and _Argenteohyla_ have similar vocal sac structure, they are obviously distinct. The monotypic _Argenteohyla_ is a rather specialized, semifossorial frog (Trueb, 1970b), characterized by smooth skin, moderate-sized digital discs, and a large inner metatarsal tubercle. The general architecture of the skull is not unlike that of _Osteocephalus_; the skulls of both are well roofed, broader than long, and characterized by posterolaterally oriented parasphenoid alae. _Argenteohyla_ bears small, slightly curved prevomerine dentigerous processes in contrast to the large, angular processes of _Osteocephalus_. The skull of _Argenteohyla_ shows specializations, apparently adaptations to its semifossorial mode of existence, which further distinguish the genus from _Osteocephalus_. In comparison with _Osteocephalus_, the cranium of _Argenteohyla_ is slightly depressed anteriorly, the roofing bones extensively casqued, and the palatines robust.

Osteologically, _Osteocephalus_ more closely resembles _Phrynohyas_ than either of the other two genera, but _Osteocephalus_ and _Phrynohyas_ are clearly distinct on the basis of their respective vocal sac structure. Like _Osteocephalus_, skulls of the members of the genus _Phrynohyas_ are broader than long, have extensive dermal roofing bones, and have posterolaterally oriented parasphenoid alae. In contrast to _Osteocephalus_, the dentigerous processes of the prevomers are curved, rather than angular in _Phrynohyas_. Furthermore, the latter genus is singularly distinguished from _Osteocephalus_, _Argenteohyla_, and _Trachycephalus_ by having extensively developed parotoid glands that produce a viscous, milky volatile secretion.

_Trachycephalus_ is the most readily identifiable of the four genera under discussion. Members of this genus are large frogs with heavily casqued and co-ossified skulls (Trueb, 1970a). The dermal roofing bones bear ornate and characteristic patterns of sculpturing. The medial ramus of the pterygoid does not articulate with the otic capsule, and the parasphenoid alae are laterally, rather than posterolaterally, oriented. A dermal sphenethmoid is present, and the parasphenoid bears odontoids. The basic structure of the skull has many characters in common with both _Osteocephalus_ and _Phrynohyas_. The obvious modifications of dermal roofing bones and of palatal and suspensory elements seem to be specializations adapting members of the genus _Trachycephalus_ to their peculiar phragmotic habits. The vocal sac structure of _Trachycephalus_ is like that of _Phrynohyas_ and therefore further distinguishes it from _Osteocephalus_.

Morphologically, _Osteocephalus_ seems to be sufficiently diverse and generalized so as to represent a modern derivative of an ancestral type which might have given rise to _Phrynohyas_, _Trachycephalus_, and _Argenteohyla_. The specialized vocal sac structure in _Phrynohyas_ and _Trachycephalus_ suggests that these two genera may be rather closely allied and represent a single phyletic line from an ancestral stock similar to _Osteocephalus_. _Argenteohyla_ is quite distinct from _Phrynohyas_ and _Trachycephalus_ and apparently represents a distinct phyletic line from the ancestral stock.

OCCURRENCE OF _OSTEOCEPHALUS_ IN AMAZONIAN ECUADOR

All of our observations on members of this genus have been made at four localities: 1) Santa Cecilia at an elevation of 340 meters on the Río Aguarico, a tributary of the Río Napo, 2) Lago Agrio, 330 meters, about 14 kilometers east of Santa Cecilia, 3) Puerto Libre, 570 meters, on the Río Aguarico just east of its formation by the confluence of the Río Cofanes and Río Chingua, and 4) south slope of the Cordillera del Dué, above the Río Coca, 1150 meters. _Osteocephalus leprieurii_ was found at all four localities, and _buckleyi_ was found at all but the last; _taurinus_ was found at Santa Cecilia and Lago Agrio, and _verrucigerus_ was found only in the Cordillera del Dué. Our data are based on collections of 113 frogs and three lots of tadpoles, as well as observations on calling sites and young. The observations are summarized by species, as follows:

_Osteocephalus buckleyi._--No breeding activity was observed. Males were found only at night in March, June, and July. One was perched on a _Heliconia_ leaf in a swamp at Puerto Libre, and two were on bushes in the forest at Santa Cecilia. A gravid female was found on a recently felled tree at Lago Agrio on the night of 12 May 1969.

_Osteocephalus leprieurii._--Males were heard calling sporadically at Puerto Libre in July 1968, and at Santa Cecilia in May 1969. A small chorus was found on the night of 12 May 1969 at Lago Agrio, where the frogs were perched on branches of fallen trees over a temporary pool. The call is a soft rattling chuckle. In late April and May many gravid females and males with well-developed nuptial excrescences were obtained from trees as they were felled at Lago Agrio. The reproductive condition of the frogs indicates that they probably breed in May. One individual called nearly every night from a large tree at Puerto Libre between 4-17 July 1968. The tree was felled on the latter date, but no frog was found. Two nights later apparently the same individual called from a bromeliad at a height of about 10 m on a large bamboo adjacent to the felled tree; the frog was collected when the bamboo was cut down.

Throughout the rainy months that we have worked in Ecuador (April-August) we have found occasional individuals perched on bushes or low trees at night. Large numbers of adults were observed only during a clearing operation which resulted in the felling of many large trees. Thus, it seems likely that _leprieurii_ is a tree-top inhabitant. A partially digested adult male was removed from the stomach of a _Hemiphractus proboscideus_.

At Santa Cecilia many recently metamorphosed young and juveniles were found in June and July 1968. Most of these were on low bushes or herbs in swamp forest at night; some were found in unfolded _Heliconia_ leaves by day, and one was observed on the forest floor by day. Snout-vent lengths of 18 specimens are 12.3-17.0 (mean 15.1) mm. The smaller frogs were recently metamorphosed as evidenced by the melanophore deposits above the vent. The coloration of the young is strikingly different from that of the adults (see account of _O. leprieurii_), so the association of the young and adults was not made until individuals with intermediate patterns were obtained at Lago Agrio in May 1969. Probably juveniles obtained in June and July are the offspring of an April or May breeding. We have been unable to associate tadpoles with this species.

_Osteocephalus taurinus._--A small chorus occurred at Lago Agrio on 12 May 1969. Males were calling from the ground adjacent to a small pool amidst recently felled trees. The males were very wary and, when approached, jumped onto limbs and ran up branches; this behavior was noted by Bokermann (1964). The call consists of a series of low-pitched, short notes--like a slow trill--four to six notes per call group. Call groups are repeated two, three, or four times followed by a lapse of several minutes. Although no amplectant pairs were found, several gravid females were collected at Lago Agrio in May, so it can be safely assumed that the species breeds in May. From April through July occasional individuals were observed on bushes and trees at night. During clearing operations at Lago Agrio several individuals were obtained from the tops of trees as they were felled.

_Osteocephalus verrucigerus._--Observations were made in a broad, shallow ravine, in which there was a small stream. On 2-4 August 1968, males were observed calling from low bushes and rocks at the edge of a quiet pool in the stream. The call consists of a series of well-pulsed, low-pitched, guttural notes produced at the rate of 5-10 per minute. One amplectant pair was found at the base of a bush adjacent to the pool on 3 August. Another female was found on a branch of a tree 2 m above the ground and 10 m from the stream. Tadpoles of this species were found in the quiet silt-bottomed pool.

SPECIMENS EXAMINED

The localities for each of the specimens examined are given in the following paragraphs. The arrangement of the data is as follows: alphabetically by country, state (department or province), and locality; alphabetically by the first letter in the abbreviations for the museums, and numerically after each museum abbreviation. Specimens lacking precise locality data are listed first in the most restricted political unit possible; localities which have not been found on maps or the positions of which are not known to us are given in quotation marks. Where more than one specimen is included under one museum number, the number of specimens is given in parentheses after the museum number. Unless noted otherwise, all specimens are alcoholics.

_Osteocephalus buckleyi_

BOLIVIA: _El Beni_: Ivón, BMNH 1967.2070-1. _Santa Cruz_: Buenavista, CM 4333, 4339, UMMZ 66563-5.

BRASIL: _Amapá_: No specific locality, WCAB 13284.

COLOMBIA: _Amazonas_: Río Guacaya, USNM 152759. _Huila_: Acevedo, Río Suaza, FMNH 69702. _Nariño_: Rumiyacu, FMNH 54756. _Meta_: Río Guejar, Campamento La Macarena, USNM 152199.

ECUADOR: No specific locality, NHMW 6209, WCAB 35499. _Chimborazo_: Pallatanga, BMNH 1947.2.13.46; Santiago, FMNH 42529. _Morona-Santiago_: "Río Santiago" (= Río Zamora), MIZS 2950. _Napo_: Lago Agrio, KU 126646; Puerto Libre, Río Aguarico, KU 123172; Santa Cecilia, AUM 8138, KU 105208-9, 109506, 123171. _Pastaza_: Alpayacu, BMNH 1912.11.1.64; Canelos, BMNH 1947.2.13.40-1, 1947.2.13.43-5; Colonia Mena, Río Conambo, ZSM 33/1962; Don Tomás, USNM 166014; Guaché, Río Pastaza, AMNH 79986; Río Bobonaza, USNM 166005; Río Capahuari, USNM 166554; Río Conambo at Río Shiona-yacu, USNM 166018; Río Copataza, upper Río Pastaza, USNM 166007-13; Río Pastaza, NHRM 1946; Río Pucyacu, USNM 165997 (skeleton), 165998-6001; Río Rutuno, USNM 166006; Río Villano, USNM 166002-4; Sarayacu, BMNH 1947.2.13.36-9, MCZ 26090, ZMB 10166.