Specimens Examined.—Panamá: Canal Zone: No specific locality, TNHC 24406; 2.8 km SW Fort Kobbe, KU 101679. Chiriquí: 14.4 km E Concepción, AMNH 69799-801; 6.6 km N David, TNHC 32013-4; 2 km S David, AMNH 68802. Coclé: 1 km NE El Caño, KU 101662-75; El Valle de Antón, AMNH 59601-5, KU 77333-47; 7 km SSW Penonomé, KU 101654-61. Los Santos: Tonosí, KU 101246 (tadpoles), 101697-701. Panamá: 2 km WSW Chepo, KU 101680-8; 6 km WSW Chepo, KU 77324-27; El Cangrejo (Panamá), KU 101676-8; Nueva Gorgona, AMNH 69991, 69798; 1.5 km W Pacora, KU 77328-32; 2 km N Tocumen, KU 101689-95; 8 km NE Tocumen, KU 101696.
Evolutionary History
My assumptions regarding the evolutionary history of the Hyla rubra group in Central America were derived partly from interpretations of the evolutionary history of other animal groups (Simpson, 1943, 1965; Dunn, 1931b; Stuart, 1950; Duellman, 1958, 1960, 1963, 1965; and Duellman and Trueb, 1966). The origin and early evolution of the group probably occurred prior to the Mid-Pliocene in the lowlands of South America, because the greatest diversity of the group is in Brazil. Differentiation into two or more subgroups took place in South America prior to the late Pliocene. At the end of the Pliocene, shortly after the closure of the Colombian Portal, many South American animals migrated into Central America (Simpson, 1943, Maldonado-Koerdell, 1964, and Savage, 1966). It is likely that the Hyla rubra group entered Central America at that time; apparently two stocks (rubra-elaeochroa-staufferi stock and boulengeri-foliamorta stock) migrated into Central America.
Hyla elaeochroa is closely related to rubra and probably differentiated from rubra through spatial isolation. Thus, we have elaeochroa in Central America and rubra in South America; most likely only in relatively recent times has rubra migrated into eastern Panamá from northern South America. The differentiation and dispersal of elaeochroa and staufferi took place in Central America after the Pliocene. Probably the events of the Pleistocene resulted in the isolation of populations. One of these (Hyla staufferi stock) was restricted in the subhumid Pacific lowlands, whereas the Hyla elaeochroa stock occupied the tropical wet forests of the Caribbean lowlands. Hyla elaeochroa apparently more closely resembled the parental stock by being restricted to the tropical rain forests, whereas staufferi adapted to subhumid environments and thereby was able to disperse throughout most of the subhumid regions of Central America.
After geographical separation took place the initial genetic divergence between the two populations was maintained by means of ecological and ethological isolating mechanisms. Under these circumstances it can be supposed that the different ecological preferences of elaeochroa and staufferi depend on the climatic changes that took place during the Pleistocene. On this basis it may be proposed that when the original prototype broke up into the two incipient species, the staufferi stock became physiologically and behaviorally adapted to subhumid conditions and dispersed into dry areas of the lowlands of Middle America. The tropical evergreen forests on the Caribbean side of lower Central America and the uplift of the Talamanca range in the Pliocene were barriers to the dispersal of staufferi. Consequently, this frog dispersed along the Pacific lowlands.
At the present time staufferi occupies the length of the Pacific lowlands in Central America, except in the rainforest of the Golfo Duce region, which apparently is a relict stand and now separates the ranges of two subspecies of Hyla staufferi. This species crossed the central Nicaraguan lowlands and reached the Caribbean lowlands of Nicaragua and nuclear Central America. The species migrated through the subhumid corridor in northern Honduras and eastern Guatemala (Comayagua Valley in Honduras and the Motagua Valley of Guatemala) to the Isthmus of Tehuantepec. Duellman (1960) hypothesized "that during the times of glacial advances (Pleistocene) the lowlands of the Isthmus probably were more extensive and had more semiarid tropical environments than at the present" and that when semiarid environments were continuous from the Pacific slope across the isthmus to the Gulf lowlands staufferi and other amphibians migrated northward to southeastern Tamaulipas, México.
Hyla elaeochroa dispersed along Caribbean lowland routes. This species not only occurs in the wet forests of the Golfo Dulce region but also in Guanacaste. It is possible that elaeochroa entered Guanacaste and moved to the Golfo Dulce region when the intervening area was less xeric than now (Duellman, 1966b). Hyla elaeochroa extended its range to eastern Nicaragua, but even though northeastern Nicaragua has over 2,000 mm. of precipitation annually (Vivo Escoto, 1964), this species has not spread into Honduras and Guatemala.
Hyla boulengeri is widespread in Amazonian and northern South America, whereas foliamorta occurs only in eastern Panamá and in north-central Colombia. The ancestral boulengeri-foliamorta stock probably invaded Central America in the late Pliocene and dispersed through humid forested environments to Nicaragua. Apparently a peripheral population established itself in the dry Pacific lowlands of Panamá. This population differentiated from boulengeri of the humid Caribbean lowlands and evolved into foliamorta, which subsequently expanded its range into Colombia.