Platypelis ando, Scherz, Mark D., Koehler, Joern, Vences, Miguel & Glaw, Frank, 2019

Scherz, Mark D., Koehler, Joern, Vences, Miguel & Glaw, Frank, 2019, A new yellow-toed Platypelis species (Anura, Microhylidae, Cophylinae) from the Maroantsetra region, northeastern Madagascar, Evolutionary Systematics 3 (1), pp. 75-83 : 75

Platypelis ando sp. nov. Figs 1, 2, 3, 4, Table 1

Holotype.

ZSM 293/2010 (FGZC 4285), adult male, collected on 3 April 2010 in Ambodivoangy (15.2899S, 49.6203E, ca. 100 m a.s.l.), Analanjirofo Region, northeastern Madagascar, by P.-S. Gehring, F. Glaw, J. Köhler, M. Pabijan, and F. M. Ratsoavina.

Paratypes.

ZSM 291/2010 (FGZC 4200), adult male, and ZSM 292/2010 (FGZC 4226), probably a male, collected on 31 March 2010 from the same locality as the holotype by the same collectors.

Diagnosis.

The new species is assigned to the genus Platypelis based on molecular phylogenetic relationships (Fig. 1). Platypelis ando sp. nov. is characterised by the following combination of characters: (1) Small size, with adult male SVL 16.9-18.7 mm; (2) manus with second finger shorter than fourth, pes with fifth toe shorter than third; (3) discs of fingers and toes yellowish to orangish in life; (4) presence of a dark dorsal chevron; (5) presence of dorsal tubercles; (6) short supratympanic dark brown marking; (7) males with prepollical tubercle but lacking a finger-like prepollex as typical for Anodonthyla Müller, 1892.

The new species is distinguished from Platypelis cowanii , P. mavomavo , P. grandis , P. tsaratananaensis , P. pollicaris , P. alticola , P. olgae , P. tuberifera , P. barbouri and P. milloti by considerably smaller size (16.9-18.7 vs>25 mm). Among Platypelis species of similar size, it can be distinguished from P. tetra by its smaller dorsal tubercles, absence of large white spots on the dorsum (vs presence), and presence of a brown chevron-shaped marking on the dorsum (vs absence); and from P. karenae by its brown colouration and dorsal patterning (vs yellow colouration and lack of dorsal patterning), short supratympanic dark brown marking (vs extended along the flank), and less pointed snout. Morphologically and genetically, P. ando sp. nov. most closely resembles P. ravus . It differs from that species in the lack of yellowish colour on its venter (vs present), yellowish to orangish dorsal finger and toe tip colouration (vs brownish), and by a chevron-shaped brown marking on dorsum (vs W-shaped).

From all members of the externally similar Cophyla , except C. occultans and C. sp. ‘fortuna’ ( Rakotoarison et al. in press), the new species differs in having a smaller body size (16.9-18.7 mm vs 21.6-33.6 mm). This includes C. maharipeo , which is similar in having yellow or orange finger and toe tips, but is larger in size. From C. occultans , P. ando sp. nov. differs in several call parameters (see below), but is very difficult to distinguish in external morphology, despite clear genetic evidence that these two species are not closely related. From C. sp. ‘fortuna’, it differs in having the fifth toe distinctly shorter than the third (vs slightly longer than the third), and presence of a brown chevron on the dorsum (vs absence).

From all members of the genus Anodonthyla , the species can be distinguished by the absence of a distinct finger-like prepollex in males.

The new species differs bioacoustically from other Platypelis species with known advertisement calls as follows: from P. barbouri , P. karenae , P. milloti , P. pollicaris , P. tsaratananaensis , and P. tuberifera by significantly longer call duration (= note duration; single note calls); and in addition from P. barbouri , P. milloti , P. pollicaris , P. ravus , P. tsaratananaensis , and P. tuberifera by significantly higher dominant frequency (see bioacoustics section below).

Description of the holotype.

Adult male in a good state of preservation, tongue taken as tissue sample. Snout-vent length 18.7 mm; for other measurements, see Table 1. Body long and rather round in preservative (more slender in life; see Fig. 2 a–b); head slightly wider than long (HW/HL 1.04), snout rounded in dorsal and lateral view; nostrils not protuberant, nearer to tip of snout than to eye; canthus rostralis distinct, concave; loreal region slightly concave; tympanum hidden; supratympanic fold indistinct, starting at the posterior corner of the eye and ending anterior to the insertion of the forelimb, dark in colour; tongue attached anteriorly and was posteriorly free; maxillary teeth present; vomerine teeth not visible or palpable (presence of rudimentary vomerine teeth cannot be excluded and would require osteological examination); choanae diminutive, round. Forelimbs slender; subarticular tubercles small, single; outer metacarpal tubercle not visible, inner metacarpal tubercle distinct, forming a large and distinct protuberance at the base of the first finger; hand without webbing; finger discs distinctly broadly rounded, somewhat truncate, with small lateral fringes; relative length of fingers 1<2<4<3; nuptial pads absent. Hindlimbs slender, tibiotarsal articulation reaching tympanum when hindlimb adpressed along body; tibia length 37.9% of SVL; inner metatarsal tubercle small, oblong; outer metatarsal tubercle absent; webbing between third, fourth, and fifth toes rather well developed, webbing formula 1(1) 2i(2) 2e(2) 3i(3) 3e(2) 4i(2.5) 4e(2.5) 5(1); subarticular tubercles on toes indistinct; toes flattened and their discs relatively broad and truncate; relative length of toes 1<2<5<3<4; third toe distinctly shorter than fifth. Dorsal skin smooth, without dorsolateral folds. A very weak mid-dorsal ridge was present in life, but is not evident in preservative. Ventral skin smooth on throat, weakly granular on abdomen and ventral legs.

In life, the holotype was olive brown in dorsal colouration with a slightly green-tinged cream saddle marking on its middle, demarcated posteriorly with a dark brown broken border, and anteriorly bordering a dark brown chevron over the suprascapular region that extended to the middle of the eyes, where it stopped abruptly behind an olive-green bar between the eyes (Fig. 2a). The lateral head surface was as the surface of the snout, mottled olive and brown. The same colour was present on the dorsum behind the saddle marking. The dorsal surface of the hindlimbs was a more muted version of this colour, with several dark grey crossbands on each limb segment. The forelimbs were yellowish over the brachium, becoming more orange-tinged distally, with a single, nearly black crossband on the antebrachium and a spot of the same colour on the outer manus. A whitish annulus was present at the base of each terminal phalange. Finger and toe tips were yellowish in colour. Ventrally, it was pale mauve in colour, and the skin was quite transparent, flecked with diminutive cream spots. The digit tips were ventrally also clearly yellow in colouration (Fig. 2b). The iris was gold with black reticulations.

After almost nine years in preservative, the specimen has faded considerably, resulting in the loss of distinction in its pattern (Fig. 3). The body and legs are overall beige, with dark oval markings in the suprascapular region, curving over the tympanum, and scattered irregularly over the rest of the dorsum, including some markings in the inguinal region. Faint crossbands on limbs, one of which is distinct on the forearm. A broad, poorly-defined, pale chevron is indistinctly visible on the dorsum.

Variation.

For variation in measurements, see Table 1. In general, the paratypes agree well with the holotype, but with the following noteworthy differences: The holotype is the plumpest specimen in the type series, with ZSM 291/2010 and ZSM 292/2010 (Fig. 2) being rather slim. In colouration, the holotype is the lightest specimen of the type series in preservative. Its pattern resembles strongly those of ZSM 291/2010 and 292/2010 (Fig. 2). ZSM 292/2010 in life (Fig. 2 c–d) apparently possessed a distinctly greenish marking in its inguinal region, of which there are unfortunately no clear photographs. This specimen also had a whiter venter than the holotype.

Bioacoustics.

The advertisement call recorded on the night of the 3rd of April 2010 in Ambodivoangy (estimated air temperature ca. 25 °C) from the holotype, ZSM 293/2010, consists of a single moderately long, high-pitched tonal whistle, repeated at regular intervals (Fig. 4). Numerical call parameters of 14 calls are as follows: call duration (= note duration) 433 ± 5.8 ms (424-441 ms); inter-call intervals 2655 ± 365 ms (2200-3567 ms); call repetition rate within call series approximately 20 calls/minute; dominant frequency 5402 ± 22 Hz (5380-5432 Hz); prevalent bandwidth 5100-5550 Hz; second frequency band at app. 7800-8200 Hz, and third at 10200-11000 Hz, the latter with the lowest energy of all three recognisable bands. Each note is characterized by a distinct upward modulation of the dominant frequency, starting at around 5250 Hz and increasing up to 5440 Hz before ending with a slight final drop in dominant frequency at around 5300 Hz.

Call comparison: The advertisement call of the sister species P. ravus (see Glaw et al. 2012) is rather similar and temporal call parameters overlap with those of P. ando sp. nov.: call duration 384-443 ms (vs. 424-441 ms); inter-call interval 2504-3200 ms (vs. 2200-3567 ms). However, P. ravus has a distinctly lower dominant frequency, with a mean value of 4010 Hz versus 5402 Hz in the new species. Despite the great similarity in structure of the calls of both species, such differences in dominant frequency are barely explainable with the slight differences in body size of calling males (SVL 19.1 versus 18.7 mm) and thus argue for species-specific differences (see Köhler et al. 2017).

Compared to the call of P. ando sp. nov., the advertisement calls of other Platypelis species differ significantly. The call of P. tuberifera is shorter (280 ms) and has a lower dominant frequency of 2300-3000 Hz ( Glaw and Vences 1994). The call of P. barbouri has significantly shorter duration (160 ms) and is repeated at much longer intervals (3200 ms), with a lower dominant frequency of 3850 Hz ( Glaw and Vences 1994). Calls of P. milloti are very short (55-65 ms) and exhibit a dominant frequency of approximately 3000 Hz ( Glaw and Vences 1994). Calls of P. pollicaris from Andasibe have shorter call duration (160-180 ms) and a dominant frequency of about 3000 Hz ( Glaw and Vences 1994). The call of P. tsaratananaensis is very short (79-145 ms duration) at a dominant frequency of 3057-3186 Hz ( Rakotoarison et al. 2012), that of P. karenae has a duration of 131-145 ms and a dominant frequency 4600-5200 Hz ( Rosa et al. 2014). The call of the morphologically similar Cophyla occultans from Nosy Be differs by a slightly longer note duration of 500-550 ms and shorter inter-call intervals (1210-1360 ms) at a dominant frequency of approximately 4000 Hz ( Glaw and Vences 1994).

Natural history.

As is typical for Platypelis species, calling activity was only heard after dusk. ZSM 291/2010 was found calling 1.8 m above the ground. Nothing further is known about the habits of this species, but based on the reproductive ecology of congeners, it is likely to reproduce in phytotelms and have endotrophic nidicolous tadpoles.

Available names.

Only two available synonyms of any Cophyla or Platypelis refer to small-sized species that could possibly refer to our new species. Cophyla tuberculata Ahl, 1929 ‘1928’ is currently a synonym of P. grandis . The two syntypes are juveniles according to Blommers-Schlösser and Blanc (1991), but have an SVL of 26 mm, and are therefore larger than the new species. Paracophyla tuberculata Millot & Guibé, 1951 is currently considered a synonym of P. barbouri . The holotype of that species, MNHN-RA-1957.715, differs from our new species in having a more rugose dorsum, broader finger discs, and a darker venter. Additionally, it is from Périnet (=Analamazaotra) in the Central East of Madagascar, more than 400 km south of Ambodivoangy. Blommers-Schlösser and Blanc (1991) concluded that it is conspecific with P. barbouri , and we agree that it is a member of that species complex, which is in need of revision.

Etymology.

We dedicate this species to our friend and colleague, Dr. Andolalao Rakotoarison, in recognition of her valuable contributions to the systematics and taxonomy of the Malagasy microhylid fauna. The name is to be treated as an invariable noun in the nominative singular.

Distribution.

The new species is reliably known only from the type locality Ambodivoangy, but the species is likely to be more widespread in low altitude forest of the adjacent Makira Natural Park. Glaw and Vences (1992) found a small Platypelis species (assigned to and figured as P. occultans ) near Voloina (15.5775S, 49.6042S; voucher specimens ZFMK 52777-52779), ca. 30 km south of the type locality with similar calls and morphology, which is possibly conspecific with Platypelis ando , but further studies are necessary to confirm its identity.