Philautus pardus, Meegaskumbura, Madhava, Manamendra-Arachchi, Kelum, Schneider, Christopher J. & Pethiyagoda, And Rohan, 2007

Meegaskumbura, Madhava, Manamendra-Arachchi, Kelum, Schneider, Christopher J. & Pethiyagoda, And Rohan, 2007, New species amongst Sri Lanka’s extinct shrub frogs (Amphibia: Rhacophoridae: Philautus), Zootaxa 1397, pp. 1-15 : 3-6

Philautus pardus new species

Figs 1 , 2

Etymology

The species epithet is the Latinised form of ‘pardos’, Greek for leopard, applied here as a noun in apposition; it alludes to the leopard­like coloration of this frog.

Diagnosis

Philautus pardus is distinguished from all other Sri Lankan species of Philautus by the following combination of characters: (1) distinctive colour pattern, comprising of dark spots on the head and dorsum; (2) angle of snout ~110º in dorsal aspect; (3) absence of a vomerine ridge, calcar, lingual papilla and lateral dermal fringe on fingers.

Description of the holotype

(See Figs. 1 , 2 .) Mature female, 32.1 mm SVL. Body stout. Head dorsally convex; snout rounded in lateral aspect, its angle about 110º in dorsal aspect; canthal edges indistinct; loreal region flat; interorbital space flat; internasal space concave; tympanum oval, vertical; pineal ocellus, vomerine ridge, cephalic ridges, calcar, lingual papilla and co­ossified skin on skull absent. Supratympanic fold indistinct. Lateral dermal fringe on fingers absent. Webbing present on toes (see Fig. 2 E). Tarsal folds absent. Snout, interorbital space, side of head and dorsum smooth; both upper and lower flanks granular; dorsal part of forelimb, thigh, shank and foot smooth; granules on throat smooth, those on chest, belly and underside of thigh rough. Inner vocal slits and nuptial pads absent.

Colour in preservative: Dorsum yellowish brown, scattered with distinct, dark­brown spots of diameter about ¼–¾ that of eye, some of them coalesced to form larger blotches (see Fig. 1 ). Spots extend on to the head, femur, inguinal zone, tibia, tarsus (absent on hands and feet) and sides to the mid­lateral fold. Upper lip yellow. Posterior part of thigh yellow to ashy brown with dark­brown spots. Venter a uniform dull yellow, slightly lighter than background colour of dorsum.

Measurements of holotype (BMNH 1947.2.7.96, in mm): DBE, 12.1; DFE, 6.5; DL, 1.4; DW, 1.7, ED, 4.7; EN, 3.0; ES, 5.1; FEL, 16.1; FL I, 2.5; FL II, 3.2; FL III, 5.3; FL IV, 4.0; FOL, 21.1; HL, 12.9; HW, 14.1; IML, 1.4; IN, 3.2; IO, 3.8; LAL, 6.6; MBE, 4.7; MFE, 8.7; MN, 11.3; NS, 2.2; PAL, 9.0; SVL, 32.1; TBL, 15.8; TL I, 2.3; TL II, 2.8; TL III, 4.6; TL IV, 6.5, TL V, 4.6; TYD, 1.2; TYE, 1.8; UAW, 5.6; UEW, 2.4.

Remarks

The key of Manamendra­Arachchi and Pethiyagoda (2005) refers P. pardus to P. viridis , while the diagnoses of Meegaskumbura and Manamendra­Arachchi (2005) affine it to P. stuarti . It can be distinguished morphologically from these two species, however, by the dorsal colour pattern in preserved specimens of distinct dark spots (vs uniform brown); having the loreal region flat (vs concave); possessing an indistinct supratympanic fold (vs distinct); and the absence of a lateral dermal fringe on fingers (vs present). The holotype of P. pardus is a paralectotype of P. variabilis (type locality: Sri Lanka). It is distinguished from P. v a r i ­ abilis, however, by having the snout rounded in lateral view (vs truncate in P. variabilis ), head dorsally convex (vs flat), loreal region flat (vs concave), supratympanic fold indistinct (vs distinct), and a lateral dermal fringe present on fingers (vs absent).

Philautus pardus also separates well from P. variabilis , P. viridis and P. stuarti in morphological space. Principal components analysis ( Fig. 3A ) shows that the four species are distinguished by a combination of body size, internarial distance, fourth­toe length, interorbital width, palm length and third­finger length. The PC(1) axis, which explains 87 % of the variance, is a size axis (snout–vent length loads most heavily and fourth­toe length least heavily, but all variables have high, positive loadings on this axis; component loadings range from 0.858–0.980, suggesting that the variation relates mostly to size). The PC(2) axis represents 4 % of the variance, with internarial distance, fourth­toe length, interorbital distance, palm length, and third­finger length contributing most heavily. While interorbital width and internarial distance load positively, the other variables load negatively (component loadings 0.302–0.363). All four species separate well on the PC(1) axis, P. variabilis being the largest and P. stuarti the smallest. While the new species also separates from the other three on the PC(2) axis, P. variabilis , P. stuarti and P. v i r i d i s almost completely overlap. Philautus pardus has a relatively greater internarial distance and interorbital diameter, but relatively shorter palm, fingers and toes compared to P. variabilis , P. v i r i d i s and P. s t u a r t i.

The size­adjusted DFA ( Fig. 3 B) shows P. pardus to be clearly distinct from P. viridis , P. stuarti and P. variabilis . The analysis correctly classified 100 % of each of the four species (Wilks’ lambda 0.001, p = 0.5). The first canonical variable best discriminates between the groups and accounts for 69 % of the total dispersion (eigenvalue 51.64), while the second variable accounts for 29 % (eigenvalue 22.32). In the canonical variables plot, the centroids are (2.307, ­3.655) for P. viridis , (­3.039, 2.944) for P. stuarti , (15.183, 1.728) for P. variabilis , and (14.230, 9.134) for P. pardus . The first canonical variable represents mostly distance between back of eyes (standardized canonical discriminant function, SCDF, 8.070), distance between front of eyes (SCDF ­7.352), head length (SCDF 3.376), head width (SCDF 2. 517), internarial distance (SCDF ­ 2.333) and eye­to­snout distance (SCDF ­2.231). The second represents mostly distance between front of eyes (SCDF ­4.528), head width (SCDF 4.230), eye­to­snout distance (SCDF 2.901), eye diameter (SCDF ­2.878), and foot length (SCDF 2. 530).

Philautus variabilis has long been a catchall taxon for difficult­to­identify Philautus in Sri Lanka and India, and a number of distinct species have erroneously been referred to it (e.g., Kirtisinghe 1957; Dutta and Manamendra­Arachchi 1996; Kanamadi et al. 1996; Kadadevaru and Kanamadi 2001; Vasudevan et al. 2001; Daniels 2003). Designation of a lectotype by Bossuyt and Dubois (2001) helped to define this species, which we failed, however, to record in our decade­long (1993–) amphibian survey of Sri Lanka. Accordingly, the GAA ( Stuart et al. 2004; see also Manamendra­Arachchi and Pethiyagoda 2005) concluded that P. variabilis sensu stricto is Extinct: it is now known only from its lectotype.

Conservation status

Not having been recollected since the holotype was collected prior to 1859, and not recorded in the WHT survey (see Manamendra­Arachchi and Pethiyagoda 2005), P. pardus is considered Extinct in terms of the IUCN’s Red List criteria.

Figures

FIGURE 1. Holotype female of Philautus pardus, 32.1 mm SVL, BMNH 1947.2. 7.96. A. Dorsal view. B. Ventral view.

FIGURE 2. Holotype female of Philautus pardus, 32.1 mm SVL, BMNH 1947.2. 7.96. A. Lateral view of head. B. Dorsal view of head. C. Left hand. D. Left foot. E. Diagrammatic representation of left­foot webbing. Scale bars = 1 mm.

FIGURE 3. A. Factor scores (no rotation) of principal components analysis of mensural data of Philautus pardus (square), P. viridis (rhombus), P. stuarti (star), and P. variabilis (circle) (females represented by open symbols, males by solid symbols). B. Canonical variables plot of discriminant function analysis of size­adjusted residuals (the 95 % confidence ellipses are centered on the centroid of each species).