Acanthosaura cardamomensis

A. crucigera A. cf. crucigera Sample size Mean ± SD Sample Size Mean ± SD Test statistic P-value PS / HL n = 8 0.18 ± 0.10 n = 29 0.33 ± 0.07 t = - 4.2 p <0.001 OS / HL n = 8 0.19 ± 0.13 n = 27 0.33 ± 0.08 t = - 3.57 p <0.002 NS / HL n = 8 0.23 ± 0.82 n = 28 0.41 ± 0.09 t = - 4.69 p <0.001 NCS n = 8 10 ± 1 n = 34 12 ± 2 t = - 4.07 p <0.001 A. crucigera A. cf. crucigera Sample Size Sample Size Test statistic P - value DS n = 8 n = 33 z = - 3.40 p <0.002 NS n = 8 n = 33 z = - 3.18 p <0.002 DS / HL n = 5 n = 31 z = - 4.00 p <0.001 DIA / SVL n = 5 n = 32 z = - 2.49 p <0.010 SUPRA n = 5 n = 34 z = - 3.13 p <0.015 A. armata A. cf. crucigera Sample size Mean ± SD Sample Size Mean ± SD Test statistic P - value PS / HL n = 13 0.43 ± 0.08 n = 29 0.32 ± 0.07 t = - 4.01 p <0.001 OS / HL n = 13 0.23 ± 0.06 n = 27 0.32 ± 0.07 t = 3.75 p <0.002 NS / HL n = 13 0.32 ± 0.07 n = 28 0.40 ± 0.09 t = 2.81 p <0.009 DIASN n = 14 5.21 ± 2.19 n = 33 10.8 ± 2.55 t = 7.10 p <0.001 VENT n = 14 62.2 ± 5.01 n = 33 56.2 ± 3.11 t = - 8.41 p <0.001 NSSOS n = 14 4.5 ± 0.51 n = 32 4.96 ± 0.64 t = 2.39 p <0.022 A. armata A. cf. crucigera Sample Size Sample Size Test statistic P - value SUPRAL n = 14 n = 34 z = - 2.41 p <0.017 INFRAL n = 14 n = 34 z = - 4.42 p <0.001 CS n = 14 n = 34 z = - 2.71 p <0.008 SDLT n = 14 n = 34 z = - 1.97 p <0.050 SDLF n = 14 n = 34 z = - 3.00 p <0.004 Paratypes. LSUHC 7406; adult male from Camp 1, Phnom Aural, Kampong Speu Province, Cambodia (12 ° 01 ’ N, 104 ° 08 ’ E) by Jesse L. Grismer on 5 August 2005 at 549 m a. s. l. LSUHC 7877; adult male Camp 1, Pursat Province, Cambodia (12 ° 01 ’ N, 104 ° 08 ’ E) by L. Lee Grismer on 6 August 2006 at 501 m a. s. l. BMNH 1932.10.4.5; sub adult male from Bokor, Kamchay Mountains, Cambodia, by M. A. Smith 1935 at 915 m a. s. l. BMNH 1915.8.14.11; sub adult female from Koh Chang Island, South East Siam M. A. Smith and C. B. Kloss. BMNH 61.4.12.43; sub adult male from Chartaboum ” (= Chantaburi Province) from the Thai Cardamom Mountains, H. Mouhot. FMNH 180309; adult female from Nakon Ratchasima Prov, Thailand, Khao Yai National Park by W. Ronald Heyer on 10 June 1969. FMNH 180307; sub adult male from Nakon Ratchasima Prov, Thailand, Amphoe Pak Thong Chai, Sakaerat Exp Station collected by W. Ronlad W. Ronald Heyer on 9 May 1969. LSUHC 7446; juvenile from Camp 2, Kampong Speu Province, Phnom Aural, Cambodia (12 ° 01 ’ N, 104 ° 09 ’ E) collected by L. Lee Grismer on 8 August 2005 above 1,400 m a. s. l. FMNH 263260; adult female from hill evergreen forest along Prek Kaoh River, Bokor National Park, Kampot Province, near 10 ° 37 ’ 19 ” N; 104 ° 02 ’ 52 ” E Cambodia, at 800 - 900 m a. s. l. collected by Bryan L. Stuart on 14 may 2000. For referred material see Appendix I.

Acanthosaura cardamomensis differs from A. capra by having more supralabial scales (SUPRAL, 11 – 15 vs. 10), by having an occipital spine verses its absence in A. capra, and by not having uniform dorsal scales as in A. capra. Acanthosaura cardamomensis differs from A. coronata by having a dorsal crest, postorbital spines, nuchal spines, dorsal spines, occipital spines, nuchal diamond, black eye patch, gular pouch and a diastema. Acanthosaura cardamomensis can further be differentiated from A. coronata by having more subdigital lamellae beneath the fourth finger (FI, 15 – 20 vs. 13 – 14), more subdigital lamellae beneath the fourth toe (TO, 20 – 25 vs. 17 – 19), more scales between the fifth canthal scale to the fifth supralabial (NSCSL, 7 – 12 vs. 5 – 6), and having fewer scales between the nasals and the rostrals (NR, 1 – 2 vs. 3 – 4). Acanthosaura cardamomensis differs from A. nataliae by having a longer diastema (DIAS, 2.7 – 8.3 mm vs. 2.5 mm), an occipital spine, black nuchal diamond, and light knee patches. Acanthosaura cardamomensis differs from A. lepidogaster by having a larger postorbital spines to head length ratio (PS / HL, 14 – 45 % vs. 10 – 11 %), by having a larger nuchal scale to the head length ratio (NS / HL, 17 – 66 % vs. 15) having a larger dorsal spine to head length ratio (DS / HL, 14 – 45 % vs. 7 – 12 %) and by having a larger occipital spine to head length ratio (OS / HL, 24 – 56 % vs. 14 – 15 %). Acanthosaura cardamomensis differs from A. titiwangsaensis by having larger dorsal spine to head length ratio (DS / HL, 14 – 45 % vs. 7 – 9 %), having a larger occipital spine divided by the head length ratio (OS / HL, 24 – 56 % vs. 9 – 10 %) and by having a light knee patch present.

Description of the holotype. (Fig. 2). Adult female, SVL 149.4 mm, TL 174.3 mm. Head length (HL 42.4 mm), head somewhat narrow (HW / SVL 17 %), moderately tall (HD / HL 51 %), triangular in dorsal and lateral profile; moderate snout length (SL / HL 44 % mm), rostrum wide (RW / RH 219 %), steeply sloping anteriorly; canthus rostralis prominent, forming a large projecting shelf extending above eye, composed of 13 enlarged scales; shelf terminates with a notch anterior to postorbital spine; rostral large in size, rectangular, bordered laterally by first supralabials and posteriorly by four smaller scales; nasals crater-like, surrounded by one prenasal, four postnasal, and three small subnasal scales; eight scales between nasal scales; supranasals somewhat elongate; enlarged scales above orbit weakly keeled; three rows of moderately keeled scales below orbit extending from anterior margin of eye to middle of eye; eye moderate (ORBIT / HL 17 %), orbit large (ORBIT / HL 25 %); interorbital, frontal, and prefrontal scales keeled and smaller than enlarged row of keeled scales below the orbit; four enlarged, keeled, azygous prefrontal scales arranged in a Y-shaped pattern; scales of occiput flat to weakly keeled; parietal eyespot surrounded by a larger row of scales; large cylindrical epidermal spine above posterior margin of eye, surrounded by four lancolate scales; occipital spine surrounded by four enlarged keeled scales; single row of five enlarged, keeled suborbital scales extending from below posterior margin of eye to anterior margin of tympanum, decreasing in size posteriorly, terminal scale spine-like; elongate cylindrical epidermal spine on lateral margin of nape surrounded by a rosette of four smaller spines; tympanum naked, round, surrounded by small conical scales; 13 rectangular supralabials decreasing in size posteriorly; mental square above becoming triangular below, larger than adjacent infralabials; two postmentals similar in size, four scales contacting mental; chinshields enlarged, extending posteriorly to angle of jaw, separated from infralabials by one scale row anteriorly and three at angle of jaw; 12 rectangular infralabials, scales in center of series largest; gulars sharply keeled and spinose with a larger midventral row; extensible dewlap present; nuchal crest bearing a diastema of 11 scales at base of nape; nuchal crest composed of eight elongate, lancolate scales bordered on each side by two rows of enlarged, flat spinose scales; dorsal body crest present, extending from posterior margin of diastema onto base of tail, vertebral crest composed of enlarged epidermal, laterally compressed, spinose scales, bordered by two rows of slightly smaller paravertebral spinose scales; vertebral crest tapers in height three-fourth the way down vertebral column then fades at base of tail. Body robust, triangular in cross-section; dorsal scales small with medially projecting spines; larger randomly arranged scales among dorsals; scales of pectoral region and abdomen larger than dorsals, keeled, spinose, arranged in semi-transverse rows; enlarged keeled scales anterior to vent; limbs relatively long; dorsal and ventral scales of forelimbs keeled, spinose, and nearly same size; five digits on manus; subdigital scales keeled; scales of hind limbs keeled and spinose; postfemoral scales small, interspersed with larger spinose scales; five digits on pes; subdigital scales keeled; 17 subdigital lamellae under fourth finger; 22 subdigital lamellae under fourth toe; tail 117 % of SVL, covered with keeled spinose scales; keels on subcaudals posteriorly directed, same size as keels of other caudal scales; base of tail in adult males nearly twice as wide as in female. continued next page cardamomensis. nov. armata bintangensis capra coronata crucigera nataliae lepidogaster titiwangsaensis. A sp. A. A A .. A A. A. A. A. n = 34 n = 14 n = 3 n = 3 n = 3 n = 5 n = 1 n = 3 n = 3 FOREL 31.7 – 56.8 33.7 – 48.9 33.9 – 61.5 54.2 – 83.8 30.2 – 35.3 35.2 – 49.8 60 – 85 33.0 – 37.1 38 – 51.7 HINDL 42 – 77.1 39 – 69.6 43.3 – 68.6 78.5 – 107.2 38.4 – 47.8 52.5 – 65 85 – 129.7 49.4 – 50.4 48.5 – 65.6 OS 4.1 – 13.3 4 – 9.4 1.2 – 2.6 absent 0 2.4 – 4.9 absent 3.2 – 3.4 1.8 – 2.3 OS / HL 0.24 – 0.56 0.16 – 0.38 0.1 – 0.11 0 0 0.13 – 0.5 0 0.14 – 0.15 0.09 – 0.1 NSSOS 4 – 6 4 – 6 6 – 7 absent 4 – 5 4 – 6 absent 5 4 – 5 NR 1 – 2 1 – 2 1 1 – 2 3 – 4 1 1 1 1 – 2 NSSLC 10 – 19 10 – 22 9 – 12 9 – 11 6 – 11 9 – 14 16 10 11 – 14 Coloration in life. Dorsal surface of head, body and limbs greenish-yellow, sides and top of the head lighter than body; black eye patch extending from posterior margin of nostrils through orbit and tympanum meeting dark nuchal marking on nape; lips cream colored, mottled in black on anterior labials and decreasing in mottling posteriorly; iris orangish brown; black collar complete dorsally, fading towards nape; gular region nearly uniformly black with a cream medial stripe on dewlap, extending form gular region to anterior edge of dewlap; body marbled with greenish black; marbled base pattern bearing randomly distributed yellow markings extending on to base of tail, where they fade; limbs greenish-yellow with scales outlined in black fading distally; tail faintly banded anteriorly, becoming nearly solid black posteriorly; undersides of limbs and belly creamed colored (Fig. 2). Variation. (Figs. 3 – 5). The paratypes resemble the holotype in all aspects except LSUHC 7877 (male) differs from the holotype by lacking green in the dorsal pattern, the marbling is larger and more uniform and it lacks the dark colored gular region. LSUHC 7406 (male) differs from the holotype in having a cream colored body with a faint marbled pattern bearing small randomly distributed yellow markings and lacking a dark colored gular region. FMNH 180309 (female) differs from the holotype in having a more distinct modeled pattern on the dorsum, more prominent banding on the tail and having a darker gular region. FMNH 263260 (female) has larger randomly distributed markings in the dorsal pattern. FMNH 180307 (male) differs from the holotype by having a more prominent marbled dorsal pattern and lacking dark coloration in the gular region. LSUHC 7446 (juvenile) differs from the holotype by lacking a distinct dorsal color pattern and coloration in the gular region. BMNH 1932.10.4.5 (male) differs from the holotype in having a more prominent dorsal pattern with more marbled markings, lacking the dark coloration in the gular region. BMNH 1915.8.14.11 (female) differs from the holotype in having larger marbled patterns on the dorsum and in having less coloration in the gular region. BMNH 61.4.12.43 (male) differs from the hollow type in having a much lighter dorsal pattern and by lacking coloration in the gular region. The black eye patch and dorsal pattern is extremely faded. Scalation and morphometric variation is presented in Table 6.

Diagnosis. A large species (maximum SVL 149 mm in females, maximum SVL 133 mm in males) with a single long cylindrical spine above posterior margin of eye; large spine on occiput between tympanum and nuchal crest; tympanum naked, round; large rectangular rostral scale; dewlap present in males and females; females with dark coloration in the gular region; scales on flanks intermixed with randomly arranged clusters of larger keeled scales; nuchal crest present; 6 – 15 scales in nuchal diastema; vertebral crest composed of enlarged, pointed scales beginning at shoulder region continuing to ¾ down vertebral column then fading at base of tail; tail 120 – 160 % of SVL; black collar and black eye patch present; posterior 2 / 3 of tail black to dark greenish; yellow to greenish spots on dorsum; dorsum light tan to dark green (Fig. 2; Table 5 compares these characters across all species of Acanthosaura).

Distribution. Acanthosaura cardamomensis occurs in the Cambodian and Thai Cardamom Mountains, with populations in Cambodia at Bokor National Park, Kampot Province; Phnom Samkos, Pursat Province; Phnom Dalai, Pursat Province; Phnom Aural, Kampong Speu Province to Thmai Baing Koh kang province. In Thailand this species is known from and Koh Chang Island, Trat Province; Khao Soi Dao District, Chantaburi Province; Khao Kitchakut National Park, Chantaburi Province; Amphoe Pak Thong Chai, Sakaerat, Nakhon Ratchasima Province; Khao Yai National Park, Nakhon Ratchasima Province, (Fig. 1). Natural history. This species can be found both during the day and at night. Individuals have been found on various substrates such as ferns approximately 0.5 meters above the ground, leaf litter, on sapling trees sleeping at night 2 m above the forest floor, and sleeping on vines 0.75 m up near the streams edge. Juveniles have been seen running across the leaf litter, found under pieces of bark in the leaf litter, and on small logs low to the ground. The majority of the juveniles were found close to the ground. When approached or provoked during the day some lizards will climb up the trees while others may drop to the ground and seek refuge under fallen or within hollow logs. Acanthosaura cardamomensis ranges from 400 - 1400 m a. s. l. The holotype is a gravid female carrying 18 eggs, collected during August 2005. A gravid female (FMNH 180309) collected on 10 July 1969 contained 23 eggs. We hypothesize that this species is breeding year round based on a number of juveniles collected at various times of the year: FMNH 180308, 10 May 1969; FMNH 263335, 30 January 2004; LSUHC 7394 - 95, 6 August 2005.

Etymology. The specific epithet cardamomensis is in reference to the type locality, the Cardamom Mountains. Comparisons to other species. Acanthosaura cardamomensis most closely resembles A. crucigera but can be differentiated from it by having significantly longer spines in the nuchal crest to head length ratio (NS / HL, 1.7 – 66 % vs. 2.0 – 38 %, n = 33; p <0.008), longer postorbital spine to the head length ratio (PS / HL, 14 – 45 % vs. 14 – 33 %, n = 34; p <0.039), longer dorsal spine to head length ratio (DS / HL, 12 – 49 % vs. 9 – 24 %, n = 36; p <0.001) than A. crucigera. Acanthosaura cardamomensis has a significantly smaller diastema to snout vent length ratio than A. crucigera (DIASVL, 3 – 7 % vs. 5 – 8 %, n = 37; p <0.014). Acanthosaura cardamomensis has significantly more infralabials (INFRAL, 10 - 14 vs. 10, n = 39; p <0.001), more canthus-rostralis-supracilliary scales (CS, 11 – 17 vs. 12 – 13, n = 39; p <0.039), more supralabials (SUPRAL, 11 – 15) 10 – 12, n = 39; p <0.01) than A. crucigera. Acanthosaura cardamomensis has significantly fewer scales in the diastema (DIASN, 6 – 15 vs. 12 – 25, n = 38; p <0.001) and subdigital lamellae beneath the fourth toe compared to A. crucigera (TO, 20 – 25 vs. 23 - 26, n = 39; p <0.010). These data are summarized in Table 1 – 5. abbreviations see materials and methods and Table 5. continued next page Acanthosaura cardamomensis differs from A. armata by having significantly larger occipital spine to head length ratio (OS / HL, 14 – 45 % vs. 22 – 56 %, n = 42; p <0.002), larger nuchal scale length to head length ratio (NS / HL, 17 – 66 % vs. 22 – 51 %, n = 41; p <0.009 Acanthosaura cardamomensis has a significantly smaller postorbital spine to head length ratio (PS / HL, 14 – 45 % vs. 22 – 56 %, n = 42; p <0.001) than A. crucigera. Acanthosaura cardamomensis has significantly more scales between the diastema (DIASN, 6 – 15 vs. 1 – 8, n = 47; p <0.001), more supralabials (SUPRAL 11 – 15 vs. 10 – 14, n = 48; p <0.017), more canthus-rostralissupracilliary scales (CS, 11 – 17 vs. 11 – 15, n = 48; p <0.008), more subdigital lamellae beneath the fourth finger (FI, 15 - 20 vs. 14 - 17, n = 48; p <0.003), and more scales bordering the occipital spine (NSSOS, 4 – 6 vs. 4 – 5, n = 46; p <0.022) than A. crucigera. Acanthosaura cardamomensis has significantly fewer infralabial scales (INFRAL, 10 – 14 vs. 12 – 15, n = 48; p <0.001), fewer scales in between the nasals (NS, 6 – 10 vs. 7 – 10, n = 48; p <0.022), fewer ventral scales (VENT, 50 – 65 vs. 51 – 68, n = 47; p <0.001), and fewer subdigital lamellae beneath the fourth toe (TO, 20 – 25 vs. 19 – 26, n = 48; p <0.05) than A. crucigera. Acanthosaura cardamomensis has a black eye patch, which is absent in A. armata. Acanthosaura cardamomensis differs from A. bintangensis by having a larger dorsal spine to the head length ratio (DS / HL 14 – 45 % vs. 8 – 9 %), fewer subdigital lamellae beneath the fourth finger (FI, 15 – 20 vs. 23), a larger occipital spine to head length ratio (OS / HL, 24 – 56 % vs. 10 – 11 %), and by having a light knee patch and lacking an elliptical scale below the orbit.