Occidozyga tompotika, Iskandar & Arifin & Rachmansah, 2011

Iskandar, Djoko T., Arifin, Umilaela & Rachmansah, Angga, 2011, A New Frog (Anura, Dicroglossidae), Related To Occidozyga Semipalmata Smith, 1927, From The Eastern Peninsula Of Sulawesi, Indonesia, Raffles Bulletin of Zoology 59 (2), pp. 219-228 : 221-225

Occidozyga tompotika , new species

( Fig 2A, B )

Holotype. – MZB (Field number DTI 2816 ), SVL: 34.1 mm, a gravid female from Bantayan , Mount Tompotika, Balantak Mountains (between S 00°39'39.5"; E 123°06'39.1"; 757 m ASL and S 00°39'50.0"; E 123°06'42.2"; 778 m ASL), Desa (=Village) Bualemo, South Sulawesi Province, Indonesia, by Umilaela and A. Rachmansah coll., 21 May 2009.

Paratypes. – MZB ( DTI 02767, 02812 , 02814 , 02815 , 02830 , 02853 ) 6 ex., collected at several camps around the same surrounding area as for the holotype, collected between 18 May and 2 Jun.2009 . MVZ ( DTI 02768) from Longkoga Stream (between 00°39'16.6"S; 123°06'47.2"E; 526 m ASL and 00°39'29.7"S; 123°06'41.0"E; 574 m ASL), Bualemo, South Sulawesi Province, by Umilaela and A. Rachmansah coll., 18 May 2009 . FMNH ( DTI 02813) from Longkoga Stream (between 00°39'39.5"S; 123°06'39.1"E; 757 m ASL and 00°39'50.0"S; 123°06'42.2"E; 778 m ASL), Bualemo, South Sulawesi Province, by Umilaela and A. Rachmansah coll. ,

21 May 2009. RMBR-ZRC ( DTI 02888) from Lemon Nipis Stream, Bantayan (between 00°45'05.0"S; 123°06'08.2"E; 462 m ASL and 00°44'55.5"S; 123°06'03.0"E; 505 m ASL), Bualemo, South Sulawesi Province, by Umilaela and A. Rachmansah coll. , 2 Jun.2009.

The description is based on 10 specimens: 3 males, 4 females and 3 juveniles.

Diagnosis. – A medium large species with half webbed toes, dorsum uniformly coloured, no light dorsolateral stripe, no dark blotch at midbody around the scapular region, fingers and toes with wide expansions at tips, all toes webbed to base of penultimate phalange except the fourth toe, which is webbed up to the position of second subarticular tubercle; tympanic annulus hidden under skin and completely covered under temporal musculature. Vomerine teeth absent, tongue completely rounded, no median notch behind, and fixed to the floor of mouth for about 70% of its length at the middle so that the margins are still free.

Description. – A medium large species, SVL in adults 34.1 ± 2.12 mm (n=7), body robust and plump. Head slightly longer than wide, eyes about equal to length of snout, pupil ovoid, oriented horizontal; snout rounded, with a straight canthus seen from above and rounded in profile, lores concave; nostrils slightly below canthus rostralis, closer to tip of snout than to eye; distance between eyes narrower than internarial distance, tympanic annulus hidden under skin and covered with muscle, but a rounded soft area could be located after incision of the skin covering the tympanic area; a distinct supratympanic fold, extending from slightly below the middle part of eye, curving straight down to dorsal part of fore limb and not looping above the position of the tympanic annulus as usually seen in most frogs. Vomerine teeth absent and lower jaws with a single protuberance at the tip. Upper arm with small tubercles, longer than lower arm; limbs short, hands with two palmar tubercles; subarticular tubercles distinct. Fingers with narrow fringe of skin and wide, truncated and flattened tips devoid of circum-marginal grooves; a rounded inner and elongated outer metacarpal tubercle, no supernumerary metacarpal tubercles. Fingers in length order 2<1<4<3. Disks of toes about equal to fingers, also with wide flattened tips bearing circum-marginal grooves with extensive webbings to base of flattened disks, only in the fourth toe, the last phalange is free from webbings, a single inner metatarsal tubercle, flattened, its length slightly greater than its distance to the base of the widened toe tip; tarsal fold present, short and running across the tarsus and ended at about midline of the tarsus. Tibias do not meet when the femurs are placed perpendicular to body axis, slightly shorter than femur. Toes in length order 1<2<5<3<4. Dorsal skin rough but no large tubercles are present on the head, snout or eyelids. Lower part of body, gular surface and ventrum are essentially smooth.

Secondary sex characters. – Males are slightly smaller compared to females, have enlarged thenar tubercle at the dorsal surface of first finger and have paired vocal sac with openings at inner side of mouth corner near end of jaws. The gular skin is darker compared to that of females and the upper jaw of males are covered with small whitish tubercles.

Detailed measurements of the female holotype (in mm): SVL, 34.1; HL, 12.7; HW, 12.3; IO, 2.9; ED, 5.1; SL, 4.5; EN, 2.4; NT, 2.1; IN, 3.7; UA, 8.8; LA, 6.2; HA, 9.7; FE, 17.3; TI, 14.5; PL, 19.2; F3W, 1.3; T1W, 1.5; T4W, 1.7; IM, 2.4; EM, 2.1.

Coloration in life. – All specimens are dark brown or blackish, usually without faint markings, pattern or mottling. Lower parts are heavily marbled with dark brown and gular surfaces more heavily marked. Lips uniform as other parts of head and body, not barred although the region below the eyes is distinctly darker, bearing a single white dot exactly below the eyes. A similar darker area is on the lip below the nostril. A pineal spot is faintly present. Upper arm is essentially uniform, with no bands visible. The femur and tibia have three indistinct dark cross bands. Few white spots are scattered at the sides of body and head and lower parts of sides and on the hands and limbs ( Fig. 1 ). All fingers and toes have a white bar straddling the middle of the truncated, flat disk and a brownish tint in the middle of the white bar. The iris is dark brown with indistinct reticulation. A middorsal light line is present in two specimens including the holotype (DTI 2812), but there is no indication of lighter dorsolateral band.

Etymology. – Refers to the type locality, Mount Tompotika at the Balantak Mountains, eastern peninsula of Sulawesi.

Ecology. – We have collected a large number of Occidozyga celebensis in freshly ploughed rice fields at Cikoro, together with tadpoles in 1992, indicating that this species has the same habitat as O. lima and O. sumatrana . Occidozyga tompotika and O. semipalmata were collected in very shallow creeks with continuous running water, in contrast to at least three Occidozyga species usually found submerged in stagnant waters. Different features of the head and laterally situated eyes in O. tompotika and O. semipalmata set them apart from other members of the genus, fit with their habits of living in ditches or shallow water and rarely stay submerged under water. This is also more or less the same habitat as many small sized Limnonectes such as L. rhacodus and also in three Ingerana species ( Sailo et al., 2009; Iskandar et al., 2011a). Judging from the nearly fully webbed toes, O. tompotika may be more aquatic compared to O. semipalmata . The species has been collected together with three undescribed species of Limnonectes , one of them is a very large species up to 200 mm. The other two species represent a medium (50–60 mm) and a small species (25–35 mm). The small species lay eggs on land. Otherwise Rana mocquardii and R. celebensis have been collected in the same habitat.

Up to present, many members of the genus have undergone considerable moving from one genus to another even after the genus Ingerana was erected. Many members of Ingerana and some Occidozyga are uncommon in museum collections, even O. semipalmata for which we have examined a good series originating from at least dozen different expeditions beginning 20 years ago.

Comparisons. – When comparing between species and from a literature search, it is evident that aside from Occidozyga lima and O. laevis , other members have tiny disks on the toes. Judging from the overall physiognomy, O. semipalmata and O. tompotika are very closely related. These two forms have a robust head with laterally oriented eyes and ovoid pupil. In other Occidozyga , the snout is more or less pointed and the eyes are situated at the top of the head directing upwards facilitating viewing the surroundings while still submerged under water. The pupil of O. sumatrana is known as diamond shaped, those of O. martensi is ovoid. The pupil form varies within the genus, but can be due to the strength of light when observed during daylight or when photographed; stronger light will force the pupil to contract and giving the diamond aspect. Occidozyga tompotika is most closely related to O. semipalmata , its congener in Sulawesi by the nature of flattened finger and toe disks. Both species also have a rounded tongue, partially adherent to the floor of the mouth. However O. tompotika has more extensive webbing on the toes compared to O. semipalmata . In O. semipalmata , webbing on the first and second toes extend to the base of tips as in O. tompotika . On the outer side of the third toe, the webbing is also starting from the base of the tip, but deeply excised so that practically the last phalange is free from extensive webbing and so is the fifth toe. The fourth toe has the last two phalanges free from extensive webbing although the web starts from the last penultimate phalange. In addition, the tongue is adherent to the mouth floor for about 50% of its length in O. semipalmata and the dorsum has some dark mottling. Occidozyga floresiana is another species to be considered as closely related to O. semipalmata and O. tompotika . The species was described as having distinctive disks on the fingers and toes and suggest a close relationship with O. semipalmata ( Mertens, 1927) . It is a large species approaching the size of O. laevis . In addition this form is also reported in having the same snout form as O. semipalmata . Occidozyga floresiana differs from O. tompotika by being larger and has all toes fully webbed up to the flattened tips and a tiny notch at the tip of tongue. These three species appears to constitute a single clade and are easily distinguished from other members of the genus in the Sundaland by the flattened toe and finger disks as well as having rounded snout and laterally oriented eyes, as no other members of the genus have all of these characters. Occidozyga tompotika appears to be also closely related to O. diminutiva based on the reduced toe webbing and having flattened toe disks without circum-marginal groove. However fingers of O. diminutiva lacked flattened disks and the tongue has a tiny notch behind ( Inger, 1954). In O. martensi , only the fingers have flattened disks, contrasting to O. diminutiva where disks are evident only on the toes. Occidozyga baluensis is similar to O. tompotika by having reduced webbing to the same extent, but differs from the new species by being slightly smaller, having a light saddle at the scapular region, ventral, gular and underside the femur is heavily blotched and a weak scapular marking or broken dorsolateral fold behind the eyes to scapular region. The eyes are directing upwards and the head is also much pointed and not highly built as in the Sulawesi species mentioned above. It is important to stress that O. baluensis also has toes with tiny flattened disks. Occidozyga tompotika differs from O. sumatrana and O. celebensis by its larger size, laterally oriented eyes, reduced toe webbing, having rounded tongue, and by being black instead of grayish. O. sumatrana is described as having a very small toe disks, however this characteristic is not clearly evident compared to O. laevis as it is more likely a conical tip. It differs further from O. tompotika by having dorsum and lateral sides covered with tubercles and at each sides of cloaca there are a dark brown band and some brown stippling at the ventral surface of the thigh. Occidozyga tompotika differs from O. laevis of the Philippines by being much smaller and having reduced webbing aside for different dorsum coloration, despite both species having a same rounded tongue without median notch behind. Occidozyga tompotika differs from O. lima by being larger and from the smooth skin in contrast to the pearly tubercles and the fully webbed toes. In addition, O. lima has distinct markings at the ventral side of the femur. Occidozyga magnapustulosa is also a species having rounded tongue without median notch similar to the new species ( Matsui, 1979). Occidozyga tompotika differs from O. magnapustulosa by the smooth skin in contrast to the tuberculated area at posterior part of the dorsum and blotched skin of the latter. Occidozyga tompotika occupy a similar habitat as Ingerana borealis , share the toe and finger tips characteristics as well as having slightly reduced webbing; however it has a rounded tip of the tongue without terminal notch and no lighter dorsolateral band. The same reason is applicable for comparison between I. tenasserimensis and O. tompotika , which also have reduced toe webbings, small, flattened finger and toe tips and a light dorsolateral region. Ingerana tenasserimensis has a tympanic annulus, easily distinguishable from O. tompotika . The finding of this small frog species with relatively large finger and toe disks at Mount Tompotika enforced our previous idea that Occidozyga semipalmata is distantly related to the other members of the genus (Iskandar & Colijn, 2000) and supported further by the lateral eye orientation and the general head form. Genetic analysis on several mitochondrial and nuclear genes supports our suggestion that O. semipalmata is very distinctive compared to other members of Occidozyga (Wogan, pers. comm., 2009, data not shown). Occidozyga tompotika , O. semipalmata , O. floresiana and potentially O. diminutiva might be placed among the Ingerana species with at least toe tips with flattened disks devoid of circummarginal grooves or to Limnonectes or even a distinct genus of its own. Rana charlesdarwini from Nicobar also has the same type of finger and toe tips ( Das, 1998), and without strong arguments is placed in Ingerana ( Frost, 2006) . A recent finding of a large Ingerana from Borneo lacking external tympanic annulus has weakened its generic definition and makes the species also closely allied to the genus Occidozyga ( Iskandar et al., 2011a) .

A number of small Limnonectes are about the same size as Occidozyga tompotika , and could easily be confused in the field. Most Limnonectes have smooth skin and none have flattened finger and toe tips, hence can be differentiated easily by those two characteristics. Limnonectes hascheanus and L. limborgi differ in smaller size and having very reduced toe webbings. Limnonectes parvus is also a small species, with smooth skin and having very reduced webbings and a pair of dorsolateral folds, so not easily misidentified. Limnonectes palavanensis is easily distinguished from this species by having smooth skin and a pairs of dorsolateral folds and from the inverted V shaped black scapular markings. Limnonectes nitidus is a much larger species, has also smooth skin with dorsolateral folds. Limnonectes tweediei is another small species with smooth skin and no the dorsolateral folds, but does not have expanded finger and toe disks. Limnonectes rhacodus is the most similar species to any Occidozyga and easily confounded with Occidozyga baluensis , as both have more or less a dorsolateral fold. However, L. rhacodus can be excluded from O. tompotika by having a pairs of dorsolateral folds and dorsum with lateral oriented wrinkles. An undescribed Sumatran species has flattened toe and finger disks with circummarginal groove and a white bar straddling on the flattened disks, but it will be placed in Limnonectes as other characters such as enlarged head in the males, a well developed odontoid processes and smooth skin, suggest that it is inappropriate to place it in Ingerana (unpubl.).

Initially the flattened finger and toe tips were thought to be a parallel character and as the tips are rather small, assignment of O. semipalmata and O. floresiana to another genus was not taken into consideration, despite of its having a relatively higher and rounded snout and laterally oriented eyes compared to other Occidozyga . Based on the finding of a distinctive form from Mount Tompotika closely related to O. semipalmata or O. floresiana , it might be justifiable to assign these species to a different genus. Decision about their generic position will be put forward after the tadpole morphology and vocalization characteristics are available and thorough genetic analysis based on nuclear and mitochondrial genes of all members of the genera Ingerana and Occidozyga . Marmayou et al. (2000) found that Ingerana tenasserimensis is closely related to Occidozyga which also enforced our conclusion that a lot more information is needed to serve as the base in placing members of Ingerana and Occidozyga in a proper generic position.

Biogeography. – Considering examples of other vertebrates such as monkeys, lizards, toad and frogs based on both genetic

and morphological analyses, parts of Occidozyga semipalmata populations are potentially distinct and separated from each other and fit with the ancient micro-continent of Sulawesi ( Evans et al., 2001, 2003a, 2003b; McGuire et al., 2007). Recent studies show that numerous cryptic species are surging as molecular techniques are very useful and widely used ( Evans et al., 2003a; Bickford et al., 2007; Inger et al., 2009). Occidozyga can be used as another supporting taxon where Sulawesi and the Philippines are found to be tightly linked as shown in several mammalian genera (i.e. Musser, 1982; Shekelle & Salim, 2009), as well as reptiles and amphibians ( de Haas, 1950; Iskandar & Tjan, 1996; Iskandar & Colijn, 2001; de Lang & Vogel, 2005; Evans et al., 2003b). The result of this study reinforces the fact that biogeographically, the southern Philippines, especially Mindanao, has a very strong connection with the Wallacean region. Conversely, much more information is needed to associate Sulawesi with the Lesser Sunda Islands. Up to present it is not clear how effective Salayar and Tanah Jampea Island served as landbridges in the past. These islands have three endemic snakes of the genera Cylindrophis, Trimeresurus, Boiga , and a Cyrtodactylus which are not enough to be used to explain the close relationships between Sulawesi and the Lesser Sunda Islands ( Hayden et al., 2009; Iskandar et al., 2011b). The presence of endemic species on this Papuan origin microcontinent within Sulawesi supports the urgent need to protect the Balantak Mountains and promote more conservation efforts. At least three reptiles and one amphibian species are now considered endemic to this peninsula (Inger & Marx, 1965; Brown et al., 2000; Iskandar et al., 2011b).

Figures

Fig. 2. Occidozyga semipalmata from Lore Lindu, proximity of Kebun Kopi showing the flattened finger disks ornamented with a white bar. Photo by T. C. Wanger.

Fig. 1. Map of Sulawesi showing the distribution of Sulawesian Occidozyga. Legend: Circle, Occidozyga semipalmata; small circles denote localities with one or two specimens; Circle, Mount Lompobatang (the type locality); Star, Occidozyga tompotika; Inverted triangle, Occidozyga celebensis. Map modified from Google map.