Unravelling the underestimated diversity of Philippine water monitor lizards (Squamata: Varanus salvator complex), with the description of two new species and a new subspecies Koch, André Gaulke, Maren Böhme, Wolfgang Zootaxa 2010 2446 1 54 [151,396,1002,1028] Reptilia Varanidae Varanus Animalia Squamata 27 28 Chordata species rasmusseni sp. nov.    Holotype. ZMUCR42151 (field number E141, Figs. 19–21), an adult male from Tarawakan, north of Batu– Batu, Tawi–Tawi Island, Sulu Archipelago, Philippines; collected by the Noona Dan Expedition, 13.XI.1961.   Paratype. ZFMK89391 (formerly ZMUCR42153, field number E136, Figs. 22–24), a juvenile with the same locality data as the holotype; also collected by the Noona Dan Expedition, 27.X.1961.   Diagnosis.A new species of monitor lizard clearly assignable to the  V. salvatorcomplex, which is characterized by high scale counts in general, (1) particularly around the tail at its base (character Q = 112– 117, mean = 114.50), (2) after one third of its length (character R = 57–60, mean = 58.50), (3) as well as around midbody (character S = 152–157, mean = 154.50), (4) round the neck (character m = 120–129, mean = 124.50), and (5) along the ventral side (character TN = 183–187, mean = 185.00). For remaining scalation features we refer to Table 3. In addition, the colour pattern of the juvenile and adult typespecimens distinguishes the Tawi–Tawi monitor lizards from all other known water monitor populations. In the juvenile paratype, (6) the  salvator-typical dorsal colour pattern is replaced by a finely spotted pattern of small bright dots arranged in about twelve indistinct transverse rows between the fore and the hind limbs. Ventrally, (7) the juvenile paratypespecimen shows about eleven dark crossbands between the fore and hind limbs and (8) distinctive dark spots on the throat. The adult holotype(9) has a melanistic appearance mottled with many single bright or marginally bright scales lacking any distinct dorsal colour pattern, (10) a dark belly with indistinctive bright median bars, and (11) a dark throat with little bright mottling. For details about scalation features and colour pattern of the remaining Philippinemembers of the  V. salvatorcomplex we refer to Tables 3and 5.   Description of the holotype.An adult male with partly everted hemipenes. The habitus is bulky. Total length 1215 mm(SVL = 466 mm, TaL = 749 mm, index 1 = 1.61); head length 82 mm, head width 49 mm(index 10 = 1.67), head height 35 mm(index 11 = 2.34). The nostrils are two and a half times closer to the tip of the snout than to the eye (index 2 = 2.50). The tympanum is oval-shaped. The tail is laterally compressed with a distinct dorsal double keel.  Scalation: The scales of the body are heterogeneous. The scales of the head are irregularly pentagonal or hexagonal; 60 scales from rictus to rictus across the head. The scales around the eye are granulous. 7/7 enlarged supraoculars and 65 (31 right/34 left) supralabials plus one rostral. The scale covering the pineal organ is enlarged, irregularly shaped with a whitish blotch in the centre. The scales of the nape are smooth, oval to roundish, and larger than the dorsals, with 30 scales from a line connecting the hind margin of the ear openings to the opposite of the gular fold; 105 dorsals from opposite of the gular fold to insertion of the hind limbs. The limbs are covered with smooth, oval scales. The caudal scales are rectangular, above slightly keeled, below with prominent keels, twice as large as dorsally; 117 scales around the base of the tail and 57 scales after approximately one third of its length. The two median rows of dorsal tail scales form a double crest starting from the base. The gular scales are smooth, oval, decreasing in size to gular fold, with 92 scales from the tip of the snout to the gular fold; 129 scales around the neck anterior to the gular fold. The ventrals of the belly are rectangular, arranged in 95 regular transverse rows from the gular fold to the insertion of the hind limbs; and 157 scales around the midbody. On the chest, an area of 18 by 6 scales is injured due to a tick infestation. The scales around the cloaca and under the base of tail bear an apical pit posteriorly.  Colour pattern(in preservative): The holotypeshows a dark brown dorsal side lacking a distinctive colour pattern. The dorsum is mottled with many partly or marginally bright scales. Limbs, neck and base of tail are dotted with single bright scales. The digits are covered with dark scales, some of which have a bright spot in the centre. The first scale after each claw exhibits a large whitish dot. After one third, the tail exhibits an indistinctive pattern of bright and dark blocks. The seven bright parts are interspersed with numerous dark scales. The head is dark brown, the regions around the nostrils are somewhat brighter, and distinct bright crossbands on the snout are absent. The tongue is flesh-coloured at its base becoming bluish-grey only dorsally after about one third of its length. The granular scales around the eyes are whitish. A bright temporal streak between eye and tympanum is absent. Ventrally, belly and chest are dark with indistinctive bright median crossbars. The throat is dark with little bright mottling. The chin becomes bright towards the tip of the snout, lacking distinctive crossbands or V-shaped markings. The limbs are covered with dark bordered bright scales. The tail is whitish at its base with an indistinctive pattern of alternating bright and dark crossbands on the second half.  Genital morphology: The left and right hemipenes are party everted in the male holotypespecimen ( Fig. 21). In this state, the copulatory organs are about 40 and 30 mmlong, respectively. The left hemipenis is apically broadened passing into two asymmetrically shaped lobes. As far as visible, both genitalia are fleshcoloured, only the pedicel and the truncus being dark pigmented at its basal smooth paryphasman-free parts. The sulcus spermaticus is basally oblique, mesially running straight to the apex. The outer sulcal lip is more developed, the sulcus being largely open from the pedicel toward the apex. The paryphasman rows (‘frills’) are strongly developed and asymmetrically arranged. Only the proximal asulcal paryphasman rows are continuous and enter the sulcus spermaticus. Eleven paryphasman rows are visible starting from the proximal asulcal side of the truncus. The lower truncus and pedicel are unornamented.  Intraspecific variation.See Table 3for a summary of main morphometrics and scalation features of  V. rasmusseni  sp. nov.While the scale counts of both typespecimens show only little intraspecific variation, clear differences in colour pattern are observable ( Tab. 5). The juvenile paratypeexhibits distinct transverse rows of small bright spots on a dark brown background ( Fig. 22), while the adult specimen lacks any such markings, and rather, has a melanistic appearance interspersed with numerous partly bright scales on dorsum, legs, neck and base of tail ( Fig. 19). This pattern increases towards the tail, where single bright scales form some strongly reduced transverse rows of small dots. Ventrally, the distinct dark crossbands and spots shown on belly and throat of the juvenile are dissolved in the adult specimen, where the throat is entirely dark with little bright mottling ( Figs. 20, 23). A bright temporal streak is visible in the juvenile paratype. Under the chin, three faded dark crossbars are discernable in the juvenile, which are not visible in the adult. The tip of the snout is bright brown with indistinctive dark crossbands.   FIGURE 19.Dorsal view of the adult holotype (ZMUC R42151) of  V. rasmusseni  sp. nov.Note the lack of colour pattern. Photo by André Koch.   FIGURE 20.Ventral view of the adult holotype (ZMUC R42151) of  V. rasmusseni  sp. nov.Photo by André Koch.   FIGURE 21.Partly everted hemipenis of ZMUC R42151, the adult holotype of  V. rasmusseni  sp. nov.Photo by André Koch.  Comparisons with other Philippine  Varanustaxa.In addition to striking differences in colour pattern between  V. rasmusseni  sp. nov.and the remaining water monitor species of the Philippines(with the exception of  V. palawanensis  sp. nov.which also lacks a distinct dorsal colour pattern at least in most adults), the Tawi–Tawi population is distinguished by much smaller and, therefore, more scales on the body. In general, the values for the dorsal scale counts of  V. rasmusseni  sp. nov.(character X = 30–32, mean = 31, character Y = 105–106, mean = 105.5, and character XY= 135–138, mean = 136.5) are intermediate between the low values characteristic of the Philippinespecies (e.g., for  V. marmoratus: X = 24–33, mean = 27.67, Y = 73–97, mean = 83.13, and XY = 101–123, mean = 110.80), and the high values observed in  V. s. macromaculatusfrom Borneo (X = 30–49, mean = 37, Y = 95–123, mean = 114.20, and XY = 125–172, mean = 151.20). Compared with the remaining Philippinewater monitor species (i.e.,  V. marmoratus,  V. nuchalis,  V. cumingi, and  V. palawanensis  sp. nov.),  V. rasmusseni  sp. nov.exhibits the highest scale counts around the base of the tail (character Q = 112–117, mean = 114.50 vs., for instance, 85–107, mean = 94.73 in  V. marmoratus) and after one third of its length (character R = 57–60, mean = 58.50 vs., for instance, 43–55, mean 47.50 in  V. nuchalis). This is also true for the ventral scale counts (e.g., character TN = 183–187, mean = 185.00 vs., for instance, 158–180, mean = 169.00 in  V. marmoratus), and the scales around the neck (character m = 120–129, mean = 124.50 vs., for instance, 72–105, mean = 98.71 in  V. marmoratus).   Distribution.Currently,  V. rasmusseni  sp. nov.is only known with certainty from the typelocality Tawi– Tawi Island, Sulu Archipelago, Philippines( Fig. 34). In all likelihood, the new taxon also inhabits some smaller islands nearby like Siasi, Sanga–Sanga, and Bongao of the Tawi–Tawi island group. Gaulke (1995, 1996) reported that the same kind of water monitor (at that time still allocated to V. s. marmoratus) occurs on all these islands, which formerly formed part of Greater Sulu. Gaulke (1995, 1996), however, did not make any statement about the monitor populations of the Jolo group further northeast towards Basilan and Mindanao, the range of  V. cumingi.   FIGURE 22.Dorsal view of the juvenile paratype (ZFMK 89391) of  V. rasmusseni  sp. nov.The colour pattern of many small bright spots scattered across the dorsum is unique among the members of the V. s a l v a t o rcomplex. Photo by André Koch.   FIGURE 23.Ventral view of the juvenile paratype (ZFMK 89391) of  V. rasmusseni  sp. nov.Photo by André Koch. Whether  V. rasmusseni  sp. nov.lives sympatrically with  V. palawanensis  sp. nov.on Sibutu Island or the Tawi–Tawi group, respectively, cannot be excluded at the moment (see below). According to Gaulke (1992, 1996) only one monitor lizard species was found on each of the Sulu Islands visited. Interestingly, Taylor (1918), who visited more than a dozen of the smaller and larger islands of the Sulu Archipelago, did not record a single water monitor specimen.   Etymology.The specific epithet  rasmusseniis a patronym honouring the late Jens B. Rasmussen ( 1947– 2005), herpetologist at the Zoological Museum of the University of Copenhagen (ZMUC), Denmark, since 1977, for his significant contributions to herpetology in general, and for his support of this discipline at the Copenhagen Museum. Hopefully, his position will be reopened one day.   FIGURE 24.Portrait of the juvenile paratype (ZFMK 89391) of  V. rasmusseni  sp. nov.Photo by André Koch.  Systematic relationships.Due to high scale counts for several specific features (e.g., around midbody or around base of tail; see Table 3) and the lack of enlarged nuchal scales,  V. rasmusseni  sp. nov.shows no similarity to the three established Philippinewater monitor species V. c u m i n g i,  V. nuchalisand  V. marmoratusbut rather resembles  V. s. macromaculatusfrom Borneo and  V. palawanensis  sp. nov.from the Palawan region. The potential close relationship with the latter species is also supported by the similarity in colour pattern. Thus, adult specimens of both species mainly have an unpatterned dark dorsal side. Juveniles of  V. palawanensis  sp. nov.and  V. rasmusseni  sp. nov.show small spots in about eight to twelve transverse rows between the fore and hind limbs, while juveniles of  V. s. macromaculatusfrom Borneo have large bright spots arranged in four to seven transverse rows between the two pairs of limbs.   Biology.No data on the biology of this Philippinemonitor lizard are available except that they inhabit lowland areas such as coconut groves and coastal regions ( Gaulke 1992a, unpubl. observ.).  Conservation status.  V. rasmusseni  sp. nov.is not hunted on the islands of the Sulu Archipelago for food, because the predominantly Islamic population (more than 90%) does not eat the flesh of monitor lizards. No data on the skin trade are available, but it is unlikely that a large-scale trade exists, especially since the complete ban of the skin trade was issued in 1994. Due to the violent problems between the inhabitants of the Sulu Archipelago and the rest of the Philippinesonly few transactions take place even in legal businesses. Nevertheless, this little known, endemic Philippinemonitor lizard taxon deserves further attention by conservationists, and it would be important to identify the monitor lizard species living in the Jolo Island Group.    V. rasmusseni  sp. nov.is restricted to a very small island area. At present it is only known from Tawi– Tawi Island, with an area of 593 km2. Even if it turns out that other islands of the Sulu Archipelago are inhabited by this species (monitor lizards have been recorded from the islands of Bongao, Sanga–Sanga, and Siasi; Gaulke 1995), its range is still extremely small. Even though monitor lizards were rather common on Tawi–Tawi Island, as well as on other visited islands of the Sulu Archipelago in the early 1990s (M. Gaulke pers. observ.), the small distribution area renders them as ‘vulnerable’ according to criterion B (and possibly C) of the International Union for Conservation of Nature (IUCN 2001).