LYNCH, J. D.: A NEW SPECIES OF FROG FROM NORTHEASTERN COLOMBIA (GENUS ELEUTHERODACTYLUS. . . 287 Z OOLOGÍA A NEW SPECIES OF FROG FROM NORTHEASTERN COLOMBIA (GENUS ELEUTHERODACTYLUS: LEPTODACTYLIDAE) por John D. Lynch1 Resumen Lynch, J. D.: A new species of frog from northeastern Colombia (genus Eleutherodactylus: Leptodactylidae). Rev. Acad. Colomb. Cienc. 27 (103): 287-289, 2003. ISSN 0370-3908. Se describe una especie nueva de rana del grupo Eleutherodactylus nigrovittatus de los bosques de niebla de Santander. La especie nueva se parece a E. nigrovittatus de la cuenca Amazónica, pero igualmente esta relacionada con las especies andinas E. latens y E. mantipus de las cordilleras Central y Occidental. Con base en un análisis filogenético, se señala que el hábitat primitivo para las ranas de este grupo es el páramo, desde donde se han desplazado progresivamente hacia menores altitudes. Palabras clave: Cladística, Eleutherodactylus, taxonomía. Abstract A new species of frog of the Eleutherodactylus nigrovittatus group is described from cloudforests in northern Santander. The new species is most similar to E. nigrovittatus of the Amazon basin but is equally closely related to the Andean E. latens and E. mantipus found on the Central and Western cordilleras. Based on a phylogenetic analysis, the primitive habitat for frogs of this species group is paramo with subsequent downward movement. Key words: Cladistics, Eleutherodactylus, taxonomy. Introduction is of some biogeographic interest because while three species are Andean (in Colombia and Ecuador), the fourth The Eleutherodactylus nigrovittatus group was treated also occurs in the lowlands of the Amazon Basin in Bra- by Lynch (1989) who recognized four species. The group sil, Colombia, Ecuador, and northern Peru. Subsequently, Duellman & Pramuk (1999) added an additional Andean species from northern Peru. I recently had occasion to 1 Instituto de Ciencias Naturales, Universidad Nacional de Co- lombia, Bogotá. [email protected] examine a collection of frogs made in Santander and was 288 REV. ACAD. COLOMB. CIENC.: VOLUMEN XXVII, NÚMERO 103-JUNIO DE 2003 surprised to discover an additional species of the E. posteriorly, posterior 1/5 not adherent to floor of mouth; nigrovittatus group, also Andean. Although the species is short vocal slits posterolateral to tongue. represented by a single male specimen, given the group’s Skin of head and dorsum smooth, that on flanks importance in understanding the role of the Andes in the granular, of venter smooth; occipital folds present but no evolution of the group (Lynch, 1999) and my recognition dorsolateral folds; no anal sheath; discoidal fold just an- that I could place it in my earlier cladogram, I decided to terior to groin; no ulnar folds or tubercles; thenar tubercle describe it. oval, smaller than bifid palmar tubercle; supernumerary Eleutherodactylus adercus sp. nov. tubercles few, flattened; subarticular tubercles of fingers round, non-pungent; fingers lacking lateral fringes; Holotype: ICN 47772 (original field number JMD fingers lacking discs, finger III pointed, others rounded at 799), a male obtained by Juan Manuel Daza and Eliana tips, digits lacking circumferential grooves, first finger M. Muñoz on 6 of October 2001. slightly longer than second. Type-locality: COLOMBIA, Departamento de Santan- Heel and tarsus lacking tubercles or folds; inner der, municipio de Betulia, vereda El Centro, 2280 m. metatarsal tubercle oval, length 2 ½ times width; outer metatarsal tubercle 1/3 to ½ size of inner metatarsal Diagnosis: (1) Skin of dorsum and venter smooth, no tubercle; sole lacking plantar supernumary tubercles; dorsolateral folds; discoidal folds present; (2) tympanum subarticular tubercles of toes round, non-conical; toes round, 37.5% length of eye; (3) snout short, subacuminate lacking lateral fringes or keels; toes bearing prominent in dorsal view, sloping toward lips; (4) upper eyelid width discs, those on toes II—IV 1 ½ times width of digit below less than interorbital distance; no cranial crests; (5) vomerine odontophores prominent, triangular in outline; disc, those of Toes I and V scarcely wider than digit; pads (6) vocal slits present; no nuptial excresences; (7) first defined by circumferential grooves; discs pointed, except finger longer than second; fingers lack discs and ventral on Toe V (rounded); Toe V almost reaches penultimate pads, (8) fingers lack lateral fringes or keels; (9) ulnar subarticular tubercle of Toe IV, Toe III reaches ½ way tubercles absent; (10) heel and tarsus lacking tubercles or between penultimate and distal subarticular tubercles of folds; (11) inner metatarsal tubercle oval, 2-3 times size Toe IV; when flexed hindlimbs held perpendicular to of outer metatarsal tubercle; no supernumary plantar sagittal plane, heels barely overlapping. tubercles; (12) toes lacking lateral fringes and webbing; Coloration in alcohol: Dorsum brown with pale brown Toe V shorter than Toe III; toes bearing expanded disks, snout and pale spot on lower back; fleshy protuberance pointed at tip; (13) dorsum brown, flanks darker, throat on snout cream; flanks darker brown; throat brown with brown, venter cream with brown reticulum; posterior cream flecks; venter cream with brown reticulation; dark surfaces of thighs brown with cream flecks and large cream brown anal triangle; posterior surface of thighs brown with spot behind knee; (14) one adult male 18.8 mm SVL. cream flecks and large cream spot behind knee. Eleutherodactylus adercus is most similar to E. nigrovi- Coloration in life: Dorsum dark brown; throat and ttatus but differs in lacking tubercles along the outer edge abdomen dark gray; undersides of feet dirty yellow; iris of the tarsus, in having vocal slits, and in having smooth, dark with some metallic golden flecks; pale brown line not shagreened skin on the dorsum. along canthus rostralis to tip of snout. Description: Head broader than body; snout subacumi- Measurements of holotype in mm: SVL 18.8; shank nate, sloping in lateral view; nostrils weakly protruding, 8.3; head width 7.0; head length 7.6; chord of head length directed dorsolaterally; canthus rostralis straight; loreal 7.8; upper eyelid width 1.4; interorbital diameter 1.9; region concave, sloping abruptly to lips; lips not flared; tympanum length 0.9; eye length 2.4; eye to nostril upper lip bearing fleshy fold; upper eyelid bearing a few distance 1.5. low tubercles; interorbital space flat, broader than upper eyelid; supratympanic fold obvious, obscuring upper edge Etymology: Greek, aderkes, meaning unseen or un- of tympanum; tympanum round, separated from eye by its expected. own diameter; two non-conical postrictal tubercles; choanae small, not concealed by palatal shelf of maxillary Natural history: The holotype is an adult male and arch; vomerine odontophores median and posterior to like most other species has a pale fleshy fold on the snout choanae, triangular in outline, bearing a row of 6—7 teeth, (except E. elassodiscus; males are unknown for E. araio- slanting posteriorly; tongue longer than wide, not notched dactylus but, given its placement in the cladogram, I LYNCH, J. D.: A NEW SPECIES OF FROG FROM NORTHEASTERN COLOMBIA (GENUS ELEUTHERODACTYLUS. . . 289 predict that they will lack the fleshy fold on the snout) . In other species, this structure is seen only in males and is probably involved either in courtship or in excavating a burrow for eggs. The holotype was found in leaf litter in mature forest during the day. Remarks: Lynch (1989) proposed a cladogram, based on three osteological and three non-osteological characters, for the four species then-known: E. elassodiscus, E. latens, E. mantipus, and E. nigrovittatus wherein the species were related as follows—(elassodiscus (nigrovittatus (latens mantipus))). Subsequently Duellman & Pramuk (1999) named E. araiodactylus from the An- des of northern Peru. My interest in this group stemmed from my observation of what I then-thought, and still think, to be a peculiar distribution. Of the five species, four are Andean, whereas E. nigrovittatus is distributed primarily in the western Amazonian lowlands but also on the Amazonian slopes of the Ecuadorian Andes (to 1935 m). With the addition of E. adercus, the Andean dominance is further evidenced. No osteological data are available for E. araiodactylus and the only osteological characters that I can score for the unique holotype of E. adercus are character 1 (large sphenethmoid extending anteriorly— Figure 1. Cladogram of the six species of the Eleutherodactylus the derived condition) and character 6 (quadratojugal not nigrovittatus group. Characters are those identified by Lynch enlarged—primitive condition and uninformative). (1989), some of which are mentioned in the text. The habitat (following Lynch, 1999) is indicated above the species name In spite of not being able to score four characters for E. in upper case letters (see text). araiodactylus and one for E. adercus, the least refuted hypothesis of relationships (Fig. 1) for the six species is largely resolved (a trichotomy at the base). In my evolving Acknowledgments effort to understand what is happening in the evolution of the genus Eleutherodactylus, I (Lynch, 1999) mapped I am grateful to Vivian Páez, Universidad de Antioquia, habitat onto the topologies of several cladograms in order and Martha Patricia Ramírez, Universidad Industrial de San- to discover which was the ancestral habitat and to specify tander, for making available laboratory space and for their all habitat changes, where habitat was assumed to be encouragement to describe this frog. Juan Manuel Daza shared congruent with the altitudinal ranges of the species his fieldnotes. Vivian Páez donated the unique holotype to involved. Lynch’s (1999) simple classification of habitat the Amphibian collection of the Instituto de Ciencias Naturales. was thus: A (0-900 m), B (901-1800 m), C (1801-2700 m), D (2701-3200 m), and E (> 3201 m). Thus, E. araiodactylus Bibliography and E. elassodiscus are scored E, E. latens is scored D, E. adercus is scored C, E. mantipus is scored B, and E. Duellman, W. E. & J. Pramuk. 1999. Frogs of the genus Eleuthero- dactylus (Anura: Leptodactylidae) in the Andes of northern nigrovittatus is scored AB. Using this classification, and Peru. Scientific Papers Natural History Museum University of assuming that E. araiodactylus does not have the derived Kansas (13): 1-78. condition for character 1, the primitive habitat for the Lynch, J. D. 1989. Intrageneric relationships of mainland Eleu- group is páramo (E) and the remaining four species therodactylus (Leptodactylidae). I. A review of the frogs emerged from an ancestor who descended the Andean assigned to the Eleutherodactylus discoidalis species group. slopes. Further resolution of primitive states is not Milwaukee Public Museum Contributions in Biology and Geology (79): 1-25. possible but the conclusion that there were subsequent movements up and downslope is inescapable. This species . 1999. Ranas pequeñas, la geometría de evolución, y la especiación en los Andes colombianos. Revista de la Academia group thus presents the most complex scenario of the ten Colombiana de Ciencias Exactas, Físicas y Naturales 23 (86): cases analyzed. 143-159.