Bolitoglossa robinsoni, Bolaños, Federico & Wake, David B., 2009

Bolaños, Federico & Wake, David B., 2009, Two new species of montane web-footed salamanders (Plethodontidae: Bolitoglossa) from the Costa Rica-Panamá border region, Zootaxa 1981, pp. 57-68 : 61-66

Bolitoglossa robinsoni sp. nov.

Robinson’s Web-footed Salamander Figure 4

Holotype. UCR 11216, an adult male from Cerro Echandi (top of Cerro 3167 m), Prov. Puntarenas, Costa Rica, latitude 9°01’50” N (9.0306), longitude 82°49’20” W (-82.8222) (3.4 km max. error, WGS84 datum), elevation 3162 m, collected on 6 January 1992 by Eugenia Arguedas.

Paratypes. Three individuals: UCR 11217-11218, same data as holotype; UCR 10921, Cerro Burú, below Cerro Echandi, Prov. Puntarenas, Costa Rica, latitude 9°00’ 45” N (9.0125, longitude 82° 49’ 50” W (- 82.8306), elevation 2300 m, collected 16 October 1991 by M. Zumbado and E. Arguedas.

Referred material. We tentatively assign morphologically divergent populations from nearby peaks to this taxon. See section on variation for detailed comments.

Diagnosis. This is a relatively large, stout-bodied species assigned to Bolitoglossa because it lacks a sublingual fold. It resembles B. cerroensis in general form, but differs by having more slender habitus, more maxillary teeth and broader hands and feet, which however are less broad than those of the even larger B. marmorea . Differs from B. sombra by having more maxillary teeth, broader feet, shorter limbs, and lacking a uniform black coloration. Differs from B. compacta in having many more maxillary teeth and lacking extensive red or red-orange blotching or striping. Differs from all other salamanders in the region by its large size, relatively long and moderately webbed digits and extensive yellowish spotting on the body, limbs and tail.

Description. This is a relatively large, stout species. Standard length (SL) in two adult males is 45.9 and 63.5. Tails are moderate, shorter than SL; SL divided by tail length in two males is 1.34 and 1.06. The head is broad; SL divided by head width is 5.7 and 6.2. The head is well demarcated from the neck. The snout is broadly rounded and not prominent. Nostrils are small and nasolabial protuberances are poorly developed.

Eyes are moderately prominent; they protrude slightly beyond the lateral margins of the head and are relatively frontal in orientation. Teeth are numerous. There are 2 and 6 premaxillary teeth in males. Maxillary teeth are 50 and 65 in the two adult males, and vomerine teeth are 22 in each. Limbs are stout but moderate in length; limb interval is 1.5 and 2 in males. Hands and feet are broad (6.0 and 7.4) with little webbing; fewer than two distal-most phalanges of the longest digits are free of webbing ( Fig. 2 , B). Digital tips are truncate and bear well-developed subterminal pads. Fingers, in order of decreasing length, are 3-4-2-1; toes are 3-4-2- 5-1. Postiliac glands are not evident.

Measurements (in mm), limb interval and tooth counts of the male holotype. Head width 10.2; snout to gular fold (head length) 13.9; head depth at posterior angle of jaw 6.6; eyelid width 2.0; eyelid length 4.4; anterior rim of orbit to tip of snout 3.4; horizontal orbit diameter 2.8; interorbital distance between angle of eyes 4.9; interorbital distance between eyelids 4.9; length of groove extending posteriorly from eye 5.8; distance between nuchal groove and gular fold 6.3; snout to forelimb 20.0; distance separating external nares 2.8; snout projection beyond mandible 0.8; snout to posterior angle of vent (standard length) 63.5; snout to anterior angle of vent 58.6; axilla to groin 34.0; limb interval 2.0; tail length 59.7; tail width at base 5.4; tail depth at base 5.6; forelimb length (to tip of longest finger) 16.1; hind limb length 16.6; hand width 6.0; foot width 7.4; length of fifth toe 1.2; length of third toe 2.7. Numbers of teeth: premaxillary 6; maxillary 31-34; vomerine 10-12, arranged in a single row.

Coloration of the holotype in alcohol ( Fig. 4 ). Dorsal ground color dark brownish-black but with so much pale spotting that the animal appears light. Patches of light yellow have joined together along the trunk to produce a very complex pattern of light on dark. Costal grooves are mainly dark, and the intercostal folds are light. Limbs and tail are mottled dark brown and yellowish-gray. Venter mottled with dark and light but mainly pale because of spaces in the punctate melanophore background. Throat pale but with punctate melanophores. Tail venter dark with ragged light spots.

Variation. We tentatively assign specimens from four localities on or near the continental divide in Panamá and Costa Rica to this species ( Fig. 3 ). However, these specimens differ in several respects from the type specimens and they may represent as many as four additional undescribed species. The four localities and the specimens associated with them are Valle del Silencio (Silencio in Fig. 3 and Table 1 , elevation 2500 m, 9°06’30” [9.1083], 82°58’00” [-82.9667]), MVZ 193550–193559 (193550, 193553, 193556, 193558, 153550 used in morphometrical analysis, see below); Cerro Biné (3200 m, 9°05’12” [9.0867], 82°54’34” [-82.8928]), MVZ 222487–89 and UCR 11787 (MVZ specimens used in morphometrical analysis); Cerro Itamut (3000 m, 9°05’40” [9.0944], 82°53’00” [-82.8833]), MVZ 222485–86 and UCR 11783–11786 (all used in morphometrical analysis); Cerro Fábrega (3335 m, 9°07’00” [9.1167], 82°52’40” [-82.8778]); MVZ 222477–222483, 222484 and UCR 11775–11782 (MVZ 222478 and 222480, and UCR 11775–11778 used in morphometrical analysis). Specimens as indicated plus the type series of B. robinsoni were included in a discriminant function analysis. We conducted several slightly different analyses. In the first we used SL, tail length, head width, head length, forelimb and hind limb length, axilla-groin, shoulder width, foot width, and total numbers of maxillary, premaxillary and vomerine teeth as separate characters ( Table 1 , Fig. 5 ). A second analysis excluded tails, because we could not always be certain whether a tail had regenerated completely or in part from external morphology. A third analysis included the tail but excluded the teeth (on the grounds that they are counts rather than measurements). A final analysis included only teeth.

All specimens in the discriminant function analysis were classified in the appropriate group using all measurements (Wilks’ Lambda=0.0025, F[60, 69]=2.9923 p=<0.000). Each of the populations is surprisingly well differentiated from the others although the samples are small. In the classification matrix (not shown) all specimens were correctly classified except a single specimen from Fábrega misclassified as from Silencio. In the analysis without the tail one specimen from Fábrega was again misclassified as from Silencio. In the analysis without teeth four specimens were misclassified, one of them a specimen from Echandi placed with Burú. A Biné specimen and two Fábrega specimens were classified as Silencio. In the analysis of teeth only, discrimination was far poorer than in the other analyses (Wilks’ Lambda=0.4095, F[15, 88]=2.2400 p=0.0103) and only 55% of the specimens were classified correctly. One specimen from Echandi was classified as Silencio and single specimens from Fábrega, Itamut and Silencio were classified as Burú.

*Variable not included in discriminant function analysis

Despite the results of the discriminant function analysis there is little to distinguish these populations in particular characters. The number of costal folds between adpressed limbs gives an integrated measure (although inexact) that is useful for quick identification of many species of salamanders. In the small Echandi population this value ranges from 1.5-3.5, which is a value larger than for any of the other populations (total range of values from -1.5 to +1.5).

Habitat and range. Bolitoglossa robinsoni is known from a section of the Cordillera de Talamanca on the continental divide that forms the boundary between Costa Rica and Panamá, and in the high Fábrega Massif, which extends northward into Panamá ( Fig. 3 ). Some specimens were obtained from areas of paramo, but others are from nearby montane forests.

Etymology. The species is named in honor of the late Douglas C. Robinson, who did a significant survey of amphibians and reptiles in Costa Rica, consolidated the UCR collection and was the mentor of most herpetologists still present in the country.

Figures

FIGURE 2. Dorsal views of the left foot (width 2.9 mm) of the holotype of Bolitoglossa pygmaea (UCR 11788) (A), and of the right foot (width 7.4 mm) of the holotype of Bolitoglossa robinsoni (UCR 11216) (B).

FIGURE 3. The border region between Costa Rica and Panamá showing the six localities from which species described in this paper have been collected. The two species are sympatric on the Fábrega Massif. The 2400 m contour line is indicated. Elevations within this contour are as high as 3335 m.

FIGURE 4. Dorsal (A) and ventral (B) views of the adult male holotype (UCR 11216) of Bolitoglossa robinsoni, 63.5 mm SL.

FIGURE 5. Results of a discriminant function analysis of morphometrical data for six samples of specimens assigned to Bolitoglossa robinsoni. A. Plot of individuals using the scores of the first two canonical axes. B. Plot of individuals with scores of canonical axis three plotted against canonical axis one.

Tables

TABLE 1. Morphometrical data (average ± standard deviation, range in parenthesis) for six samples assigned to Bolitoglossa robinsoni. The type locality is Echandi and one paratype is from Burú (cf. Figs. 3 and 5).

Locality	Echandi	Burú	Fábrega	Itamut	Biné	Silencio
n	3	1	9	6	3	9
Standard Length (SL, mm)	46.5 ± 16.7 (30.2–63.5)	30.2	55.4 ± 11.2 (34.4–72.5)	48.0 ± 6.5 (37.2–55.7)	53.3 ± 12.4 (39.0–60.6)	36.2 ± 7.4 (26.2–51.8)
Tail Length (mm)	37.3 ± 21.1 (17.8–59.7)	17.8	53.3 ± 13.8 (25.4–70.7)	45.8 ± 7.2 (33.5–52.8)	50.5 ± 17.1 (30.8–62.3)	29.9 ± 11.8 (16.3–55.1)
Head Width (mm)	7.7 ± 2.7 (4.8–10.2)	4.8	8.2 ± 1.3 (5.4–9.9)	8.1 ± 1.0 (6.6–9.3)	8.4 ± 2.2 (5.9–9.9)	6.0 ± 1.0 (4.6–8.2)
Head Length (mm)	10.6 ± 3.3 (7.3–13.9)	7.3	12.7 ± 2.0 (9.3–15.5)	11.8 ± 1.2 (9.7–13.0)	11.9 ± 3.0 (8.5–14.1)	9.0 ± 1.7 (6.7–12.6)
Hindlimb Length (mm)	12.4 ± 4.5 (7.7–16.6)	7.7	15.0 ± 3.3 (8.0–18.8)	13.2 ± 2.0 (10.3–15.2)	13.7 ± 3.1 (10.1–15.6)	9.4 ± 2.5 (6.0–14.3)
Forelimb Length (mm)	11.9 ± 4.7 (6.8–16.1)	6.8	13.4 ± 2.9 (6.6–16.7)	13.0 ± 1.9 (10.1–15.0)	13.1 ± 3.1 (9.5–15.3)	8.9 ± 2.5 (5.9–14.1)
Axilla to Groin Length (mm)	23.9 ± 9.5 (15.2–34.0)	15.2	29.8 ± 6.5 (16.8–37.4)	25.6 ± 3.7 (19.3–30.6)	29.0 ± 7.1 (20.8–33.9)	18.9 ± 4.2 (14.5–28.3)
Shoulder Width (mm)	5.8 ± 1.9 (4.1–7.9)	3.5	5.5 ± 1.2 (3.3–7.3)	5.8 ± 0.7 (4.8–6.6)	5.6 ± 1.8 (3.6–7.1)	4.0 ± 0.7 (3.1–5.1)
Foot Width (mm)	5.4 ± 2.3 (2.9–7.4)	2.9	5.7 ± 1.3 (3.3–8.0)	6.3 ± 1.1 (4.5–7.3)	5.7 ± 1.5 (4.1–6.9)	3.8 ± 1.2 (2.2–6.3)
Number of Maxillary Teeth	53 ± 11 (43–65)	43	49 ± 10 (30–64)	46 ± 9 (32–54)	41 ± 27 (9–58)	29 ± 14 (14–56)
Number of Premaxillary Teeth	4 ± 2 (2–6)	3	5 ± 1 (3–6)	4 ± 1 (3–5)	5 ± 3 (2–7)	2 ± 1 (0–4)
Number of Vomerine Teeth	18 ± 6 (11–22)	23	21 ± 5 (16–32)	20 ± 3 (17–25)	25 ± 10 (14–33)	16 ± 5 (11–25)
Number of Grooves when Limbs Adpressed *	(1.5–3.5)	(3.5)	(0.0–2.0)	(-1.5–1.0)	(1.0–1.5)	(0.5–2.0)