_Life History._--At Vista Hermosa, Oaxaca, males were calling on vegetation above small streams on March 30, 1959, and on June 28, 1962; males were found on vegetation overhanging a stream 6 kilometers south of Vista Hermosa on June 27 and July 3, 1962. The call consists of a series of short notes, three to thirteen notes per series, sounding like "raa-raa-raa." The duration of each note is about .08 of a second and has a rate of 123 to 129 pulses per second. The dominant frequency of notes in short series is about 2100 cycles per second, whereas the dominant frequency of notes in long series is about 3150 cycles per second (Pl. 11E).

On June 28, 1962, two tadpoles of this species were found in a quiet pool in a spring-fed rivulet at Vista Hermosa, Oaxaca. Females are unknown.

_Remarks._--The absence of a tarsal fold and of nuptial spines in breeding males, the nature of the breeding call, and the form of tadpole are characters that place _Ptychohyla ignicolor_ in the _P.

schmidtorum_-group.

_Distribution._--This species is known from only two localities at elevations of 1500 and 1850 meters in the cloud forest on the northern (Atlantic) slopes of the Sierra Madre Oriental in northern Oaxaca.

_Specimens examined._--MEXICO: _Oaxaca: Vista Hermosa_, KU 71334, 71716 (tadpoles), UMMZ 119602; 6 km. S of Vista Hermosa, KU 71335-42, 71343 (skeleton), UMMZ 119603, 123327 (2).

DISTRIBUTION AND ECOLOGY

Geographic Distribution of the Species

The distribution of species of _Ptychohyla_ reflects the distribution of cloud forest in southern Mexico and northern Central America.

The frogs are restricted to mountainous areas, usually at elevations higher than 1000 meters above sea level. _Ptychohyla_ does not range to great heights in the mountains, where west of the Isthmus of Tehuantepec the mountain streams are inhabited by frogs of the _Hyla bistincta_ group, and in Chiapas and Guatemala by species of _Plectrohyla_.

Frogs of the _Ptychohyla euthysanota_ group have a greater combined geographic range than the species comprising the _Ptychohyla schmidtorum_ group (Fig. 7). No two species in the same group are sympatric, but members of different groups are sympatric in at least parts of their ranges. Apparently _P. leonhardschultzei_ ranges around the southern edge of the Mexican Highlands, where the species occurs on both Atlantic and Pacific slopes; as can be seen from the distribution map, there are many gaps in the known range of this species. The range of _P. euthysanota euthysanota_ is along the Pacific slopes of the Sierra Madre in Chiapas, Guatemala, and El Salvador, whereas that of _P. euthysanota macrotympanum_ is along the southern interior slopes of the Central Highlands of Chiapas and the Sierra de Cuchumatanes in Guatemala. _Ptychohyla spinipollex_ occurs on the wet Atlantic slopes of the Guatemalan and Honduranean Highlands; the range of the species in Honduras is poorly known.

[Ill.u.s.tration: FIG. 7. Map showing locality records for the species and subspecies of _Ptychohyla_.]

The frogs of the _Ptychohyla schmidtorum_ group have more restricted geographic ranges than members of the former group. _Ptychohyla schmidtorum schmidtorum_ occurs on the Pacific slopes of the Sierra Madre in Chiapas and Guatemala, where it occurs with _P. euthysanota euthysanota_; _P. schmidtorum chamulae_ is known from only two localities on the Atlantic slopes of the Central Highlands of Chiapas, where it occurs close to, but as now known not with, _P. euthysanota macrotympanum_. On the Atlantic slopes of the Sierra Madre Oriental in northern Oaxaca _P. ignicolor_ occurs with _P. leonhardschultzei_.

In the Sierra de los Tuxtlas in southern Veracruz and in the cloud forests along the eastern slopes of the Sierra Madre Oriental northward to Nuevo Leon, _Hyla miotympanum_ seems to be the ecological replacement of _Ptychohyla_. On the Pacific slopes north of Guerrero, Mexico, humid forests in which there are cascading mountain streams are absent; consequently, no _Ptychohyla_ are known from that region.

In the mountains of El Salvador _Ptychohyla euthysanota euthysanota_ occurs sympatrically with another small stream-breeding hylid, _Hyla salvadorensis_. To the south of Honduras the highlands diminish into the lowlands of Nicaragua, where habitat suitable for _Ptychohyla_ apparently does not exist. In the mountains of Costa Rica and Panama, the habitats occupied by _Ptychohyla_ in northern Central America are filled by a variety of stream-breeding _Hyla_, such as _Hyla legleri_, _H. rivularis_, _H. rufioculis_, _H. alleei_, and _H. uranochroa_.

Although members of the genus _Ptychohyla_ occur in the southern part of the Mexican Highlands to the west of the Isthmus of Tehuantepec, the greater distribution and differentiation in the genus is in the Chiapan-Guatemalan Highlands. In this respect _Ptychohyla_ is a counterpart of _Plectrohyla_.

Habitat Preference

Frogs of the genus _Ptychohyla_ are ecologically a.s.sociated with mountain streams at elevations between 650 and 2200 meters; in the geographic region where these frogs occur the vegetation between those elevations consists of cloud forest or pine-oak forest. In some places the frogs have been found in a mixture of oak and semi-deciduous scrub forest. At Vista Hermosa, Oaxaca, _P. leonhardschultzei_ and _P.

ignicolor_ were found in cloud forest, whereas at Agua del Obispo, Guerrero, the former species was found in pine-oak forest. _Ptychohyla schmidtorum_ is known only from cloud forest; _P. euthysanota euthysanota_ and _P. spinipollex_ generally are found in cloud forest, but in some places they live in pine-oak forest. _Ptychohyla euthysanota macrotympanum_ has been found in pine-oak forest and in a mixture of oak and semi-deciduous scrub forest. With the possible exception of the members of the _Ptychohyla schmidtorum_ group, which has been found only in cloud forest, it seems as though the type of vegetation is not the controlling factor in the ecological distribution of these frogs.

_Ptychohyla_ has been found only where there are clear, cascading streams overhung by vegetation, on which adults and young perch at night, or even by day. The presence of these streams, in which the tadpoles live, seems to be an important factor in the distribution of _Ptychohyla_. As has been shown previously, the tadpoles of _Ptychohyla_ are adapted for existence in torrential streams, where the water is cool, and the amount of oxygen is high. Clearly these tadpoles are unsuited for life in ponds or sluggish streams in the lowlands, where the temperature of the water is high, a layer of silt on the bottom is deep, and the amount of oxygen is low. The tadpoles cling to rocks on the bottom of the streams; there they move slowly across the rocks, apparently feeding on the thin covering of algae.

Tadpoles were not observed on rocks having a thick covering of algae or moss. The tadpoles were observed to swim against the current in torrential streams, in which no fishes were found. Therefore, it seems as though the presence of the stream-habitat for the tadpoles is a significant factor in the ecological distribution of the species of _Ptychohyla_.

Interspecific Compet.i.tion

At localities where two species of _Ptychohyla_ occur sympatrically (_P. ignicolor_ and _P. leonhardschultzei_ at Vista Hermosa, Oaxaca, and _P. euthysanota euthysanota_ and _P. schmidtorum schmidtorum_ at Finca La Paz, Depto. San Marcos, Guatemala) effort was made to determine what, if any, ecological interspecific relationships existed. Although adults of the sympatric species were found on adjacent leaves or branches of bushes overhanging the streams at both localities, segregation at the time of breeding seems to be maintained by the notably different breeding calls in sympatric species (see discussion of breeding calls). Thus, as has been shown by Blair (1956), Fouquette (1960), and others working on a variety of pond-breeding frogs and toads, the breeding call in _Ptychohyla_ acts as an important reproductive isolating mechanism.

At no locality were _Ptychohyla_ and a.s.sociated species of hylids found so abundantly as were species of pond-breeding hylids in the lowlands. Apparently reproductive activity is not concentrated in a short breeding season, and it is highly doubtful if the populations of these frogs are as large as those of the lowland pond-breeders. The continual humid conditions and abundance of insect food throughout the year in the cloud forest are perhaps indicative of little interspecific compet.i.tion among adults of _Ptychohyla_ and other sympatric hylids.

At Finca La Paz, Guatemala, tadpoles of two species of _Ptychohyla_ were ecologically segregated. The tadpoles of _P. euthysanota euthysanota_ were found in riffles in the streams, whereas those of _P. schmidtorum schmidtorum_ were found in slower water, chiefly in small pools in the streams. At Vista Hermosa, Oaxaca, tadpoles of _P.

leonhardschultzei_ were found in riffles, and tadpoles of the sympatric _P. ignicolor_ were found in a small pool in a stream.

Similar ecological relationships were observed for several species of Costa Rican hylids. Throughout the range of _Ptychohyla_ east of the Isthmus of Tehuantepec, members of the genus occur with species of _Plectrohyla_, all of which are larger than _Ptychohyla_, and all of which have tadpoles that live in torrential streams. Tadpoles of _Ptychohyla spinipollex_ have been found in streams inhabited by the tadpoles of _Plectrohyla guatemalensis_ and _P. quecchi_; tadpoles of _Ptychohyla euthysanota euthysanota_ and _P. schmidtorum schmidtorum_ were found in streams inhabited by tadpoles of _Plectrohyla guatemalensis_, _P. matudai_, and _P. sagorum_. In some streams great numbers of tadpoles occur. The habitat is rather restricted, and the food supply is limited. Consequently, interspecific compet.i.tion among the various species of hylids whose tadpoles live in the torrential streams probably is highest during the larval stage. Unfortunately, this aspect of salientian population ecology has received no intensive study.

Reproduction and Development

Since the cloud forests inhabited by _Ptychohyla_ are daily bathed in clouds and have a fairly evenly distributed rainfall throughout the year, the frogs living in these forests are active throughout the year. At least some of the species evidently have a long breeding season, for I found calling males of _P. leonhardschultzei_ in February, March, and August, and found tadpoles in February, March, June, and August. Tadpoles of the various species have been obtained throughout much of the year, as follows: _P. euthysanota euthysanota_, February, March, May, and July; _P. euthysanota macrotympanum_, March, June, and August; _P. spinipollex_, February, March, April, June, July, and August; _P. schmidtorum schmidtorum_, March, May, June, July, and August; _P. schmidtorum chamulae_, June and August; _P.

ignicolor_, June. I suspect that this temporal distribution more accurately reflects the seasonal activities of collectors than of the frogs.

Calling frogs usually are on vegetation adjacent to or overhanging streams; some calling males of _P. spinipollex_ were on rocks in or by streams. Clasping pairs of _P. euthysanota_ and _P. schmidtorum_ were observed on vegetation by streams. Despite intensive search, no eggs were found. It is doubtful if _Ptychohyla_ deposit eggs on vegetation overhanging streams, as do centrolenids and _Phyllomedusa_, for egg-clutches of these frogs are easily found. Possibly the eggs are laid separately on vegetation above the stream, in which case they could be overlooked easily. In streams where _Ptychohyla_ and other hylid tadpoles occur, empty egg capsules have been found on the lee sides of rocks, but there is no way to determine which species laid the eggs.

Numbers of eggs were counted in gravid females; the largest eggs have diameters ranging from 2.5 to 3.0 mm. The smaller species, comprising the _Ptychohyla schmidtorum_ group, have fewer eggs than do the larger species. Numbers of eggs found in females of the various species are: _P. euthysanota euthysanota_, 108; _P. euthysanota macrotympanum_, 136, 160; _P. leonhardschultzei_, 141; _P. spinipollex_, 119, 134, 143; _P. schmidtorum schmidtorum_, 59, 61, 90; _P. schmidtorum chamulae_, 60, 71, 89.

Duration of the larval stage is unknown. Metamorphosing young have been found from May through August. From two to six completely metamorphosed young are available for each of the species, except for _P. ignicolor_ of which none is available. The smallest young frog is a _P. euthysanota_ having a snout-vent length of 14.2 mm.; the largest young frog is a _P. schmidtorum schmidtorum_ having a snout-vent length of 17.0 mm.

PHYLOGENY OF PTYCHOHYLA

The preceding data on morphology, life histories, and behavior form the basis for the following interpretation of the phylogeny of _Ptychohyla_. Additional data are needed to support some of the ideas discussed below; many of the data that are available for _Ptychohyla_ are lacking for other, possibly related, hylids. The family Hylidae is composed of several hundred species, and most of the species are poorly known. Consequently, any attempt to place _Ptychohyla_ in the over-all scheme of hylid phylogeny would be premature at this time.

But, as between the five species of two species-groups here recognized as const.i.tuting the genus _Ptychohyla_, some estimate of relationships can be made. First, it is necessary to determine the validity of the genus itself.

Ptychohyla as a Natural a.s.semblage

As stated in the diagnosis of the genus, the only character that sets this group of species apart from other hylids is the presence of ventrolateral glands in the breeding males. To many systematists the thought of being able to identify to genus only breeding males is sufficiently disturbing to cause them to view with disfavor the recognition of the genus. Nonetheless, the question is raised: Do the five species herein placed in the genus _Ptychohyla_ const.i.tute a natural a.s.semblage? If the genus is considered to be more than a category of convenience, that is to say, a group of related species having a common origin, the primary problem is to determine whether or not the five species form a phylogenetic unit.

The species of _Ptychohyla_ are divided into two groups on the basis of external morphology, breeding calls, and tadpoles. The _Ptychohyla euthysanota_ group seems to be a natural group composed of three species, all of which are more closely related to one another than to any other hylid. Likewise, the species comprising the _Ptychohyla schmidtorum_ group seem to represent a natural unit. If the presence of ventrolateral glands in breeding males is ignored, a student of salientian systematics might derive the _Ptychohyla euthysanota_ group from a hylid stock containing _Hyla miotympanum_ and _Hyla mixomaculata_. Likewise, _Ptychohyla schmidtorum_ could be placed with _Hyla uranochroa_ and related species in Costa Rica. Nonetheless, the fact remains that all of the species a.s.signed to the genus _Ptychohyla_ have ventrolateral glands in the breeding males; furthermore, ventrolateral glands are unknown in other hylids. If the _P. schmidtorum_ group and the _P. euthysanota_ group each evolved from separate hylid stocks, then the ventrolateral glands must have developed independently in both groups. That ventrolateral glands developed independently in five species of frogs in southern Mexico and northern Central America and not in any of the other approximately 500 species of hylids in the world is untenable. It is more logical to a.s.sume that the development of the glands took place only once in a stock of hylids that gave rise to the five species herein recognized as members of the genus _Ptychohyla_.

Generic Relationships

The affinities of _Ptychohyla_ apparently are not with any of the other groups that have been generically separated from _Hyla_. Of the daughter genera in Middle America only _Plectrohyla_ has stream-adapted tadpoles, but these large frogs are not closely related to _Ptychohyla_. Stuart (1954:169) suggested that certain montane species of _Hyla_ in lower Central America and _Hyla salvadorensis_ in El Salvador may be related to _Ptychohyla_ or even congeneric. I have had experience with most of these species in the field and believe that Stuart was correct in his suggestion of relationships. The species concerned are four red-eyed stream-breeding _Hyla_ in Costa Rica--_H. alleei_, _H. legleri_, _H. rufioculis_, and _H. uranochroa_, plus _Hyla salvadorensis_ in the mountains of El Salvador.

Morphologically all of the species are similar; _Hyla uranochroa_, _H.

legleri_, and _H. rufioculis_ have a lateral white stripe that is expanded to form a spot beneath the eye, as in _Ptychohyla schmidtorum_. The tadpoles of _Hyla rufioculis_ and _H. uranochroa_ have large funnel-shaped mouths and long slender tails like those of _Ptychohyla schmidtorum_. Lips of the tadpoles of _H. legleri_ and _H.

salvadorensis_ are folded laterally, in this respect resembling those of the _Ptychohyla euthysanota_ group. I do not know the tadpoles and the breeding call of _Hyla alleei_. The breeding calls of _Hyla rufioculis_ and _H. uranochroa_ consist of high melodious notes; the calls of _H. legleri_ and _H. salvadorensis_ consist of series of short notes that have the general characteristics of the call of _Ptychohyla schmidtorum_. Affinities of the genus _Ptychohyla_ seem to me to be with the red-eyed species forming the _Hyla uranochroa_ group in Costa Rica. All of the species in the _Hyla uranochroa_ group have large frontoparietal fontanelles, rather small ethmoids, and small nasals that are not in contact with one another or with the ethmoid.

Some species have a complete quadratojugal-maxillary arch; others do not. a.s.suming that the parental stock that gave rise both to the _Hyla uranochroa_ group and to _Ptychohyla_ was widespread in Central America at a time of cooler, more humid conditions, it is possible that with subsequent warming temperatures and seasonal rainfall in the lowlands the parental stock was restricted to the Costa Rican highlands, where the _Hyla uranochroa_ group developed, and to the Chiapas-Guatemala highlands, where _Ptychohyla_ evolved.