Cornufer (Batrachylodes) exedrus, Travers & Richards & Broadhead & Brown, 2018

Travers, Scott L., Richards, Stephen J., Broadhead, Taylor S. & Brown, Rafe M., 2018, A new miniature Melanesian Forest Frog (Ceratobatrachidae: CORNufeR) from New Britain Island, constituting the first record of the subgenus BATRAchylOdes from outside of the Solomon Archipelago, Zootaxa 4370 (1), pp. 23-44 : 26-35

Cornufer (Batrachylodes) exedrus sp. nov.

Figs. 2a–c , 4 , 5

Holotype. South Australia Museum ( SAMA) R64760 (SJR Field No. 10784), adult male, collected by S. J. Richards at Tompoi Camp, 1590 m above sea level (05°20.623'S, 151°18.873'E; WGS-84), Nakanai Mountains, East New Britain Province, Papua New Guinea, 19 April 2009 .

Paratypes. SAMA R64762 and R64764 (SJR Field Nos. 10815 and 10820 respectively), two adult males, and SAMA R64763 (SJR 10819), adult female, all collected at the type locality, 21 April 2009; SAMA R64761 and R71008 (SJR 10798 and 10821 respectively), two adult males, same locality, 20 and 21 April 2009, respectively; SAMA R64765 (SJR 10843), juvenile, and SAMA R71007 (SJR 10846), adult male, same locality, 23 April 2009.

Etymology. The specific epithet is derived from the masculine formation of the Greek adjective exedros, meaning away from home, strange, or extraordinary. The surprising discovery of Cornufer (Batrachylodes) exedrus from a high elevation site in New Britian is a taxonomic and biogeographic enigma, given that all other species in the subgenus are restricted to the Solomon Archipelago.

Diagnosis. The new species is a member of the genus Cornufer based on molecular phylogenetic evidence. There are currently no morphological synapomorphies known for the genus Cornufer , however, based on the updated taxonomic arrangement of this clade ( Brown et al. 2015) into several morpholgically diagnosible subgenera the new species can be assigned to the subgenus Batrachylodes , as defined by Brown et al. (2015), on the basis of its small, triangular shaped body, pointed snout, absence of interdigital webbing, absence of vomerine teeth, presence of a darkened loreal stripe extending diagonally across the flanks, and molecular phylogenetic placement ( Brown et al. 2015:fig. 2). Cornufer exedrus is distinguished from congeners by a combination of (1) adult body size small (SVL 17.6–18.8 mm for six adult males, mean = 18.3 mm; SVL 18.3 mm for one adult female), (2) body shape triangular with a small, pointed head, (3) terminal disks of the fingers and toes narrowly expanded (among adults: Fin3DW 0.3–0.6 mm, mean = 0.5 mm; Toe4DW 0.4–0.7 mm, mean = 0.6 mm) ( Figs. 2 , 4 , 5 ), (4) vomerine teeth absent, (5) digits lacking all vestiges of lateral flange and interdigital webbing, (6) dorsal skin smooth to finely granular, lacking prominent dermal tubercles or ridges, (7) subarticular tubercles of fingers, toes, hands and feet uniformly low and flattened, (8) dark band extending along the canthus rostralis and onto the flank with a smooth dorsal margin demarcating lighter dorsal and darker lateral colors, and (9) unique advertisement call consisting of a series of brief, high-frequency atonal notes.

Comparisons with other species. Among members of the subgenus Batrachylodes , Cornufer exedrus can be easily distinguished from C. elegans , C. gigas , C. montanus , C. vertebralis , and C. wolfi by the narrowly expanded (versus widely expanded) terminal finger disks except for the first finger (Fin3DW/SVL = 0.02–0.03, Fin3DW/ Fin3L = 0.1–0.2, and Fin3DW/TD = 0.2–0.3 in C. exedrus versus Fin3DW/SVL> 0.03, Fin3DW/Fin3L> 0.2, and Fin3DW/TD> 0.3 in the aforementioned species) and smaller adult body size (SVL 17.6–18.8 mm in C. exedrus versus SVL> 20.0 mm in the aforementioned species) (Table 2). Of the Batrachylodes that lack prominently dilated finger disks, Cornufer exedrus differs from C. mediodiscus in having a smaller adult body size (SVL 17.6– 18.8 mm in C. exedrus versus 20.2–26.0 mm in C. mediodiscus ), slightly smaller third finger disk width (reflected in the Fin3DW/SVL, Fin3DW/Fin3L, and Fin3DW/TD ratios, Table 2), and lacking raised dorsal skin ridges ( Brown & Parker 1970). Cornufer exedrus differs from C. minutus by its slightly larger body size (SVL 17.6–18.8 mm in C. exedrus versus 15.2–17.3 mm in C. minutus ) and by differences in the proportions related to hind limb length (reflected in the HW/TBL, TBL/SVL, and PL/SVL ratios, Table 2). Cornufer minutus also differs in color pattern, with the dark band along the canthus rostralis more poorly defined, often broken or with an irregular dorsal margin, than in Cornufer exedrus . The new species differs from C. trossulus by its slightly smaller body size (SVL 17.6–18.8 mm in C. exedrus versus 18.9–24.3 mm in C. trossulus ) and by differences in the proportions related to hind limb length (reflected in the HW/TBL, TBL/SVL, and PL/SVL ratios, Table 2).

On New Britain Island, C. exedrus can be differentiated from most congeners by its smaller adult body size (<25 mm), narrowly expanded finger disks, and/or absence of pedal webbing. Of the small-bodied Cornufer species on New Britain (<25 mm), C. exedrus can be further distinguished from C. caesiops (and the morphologically similar C. browni , on nearby New Ireland Island; Allison & Kraus 2001; Kraus & Allison 2009) in having narrowly expanded finger disks (versus widely expanded in C. caesiops and C. browni ); from C. akarithymus , C. bufonulus , and C. sulcatus in having a smooth dorsal surface (versus tuberculate or with dorsal ridges in C. akarithymus , C. bufonulus , and C. sulcatus ); and from C. boulengeri (males and juveniles previously assigned to C. rhipiphalcus ; Kraus 2008), by the lack of a fanlike array of dorsal ridges in the scapular region and a less prominent dark face mask, which does not extend onto the flanks in C. boulengeri .

Description of holotype. Adult male, in good condition, preserved with mouth open, incision in left thigh where SAMA tag was threaded; habitus moderately robust; head not distinct, slightly narrower in dorsal aspect than body, length 41% of SVL; head length 120% of head width; snout short, tip acuminate in lateral aspect, rounded dorsal aspect, protruding prominently beyond lower jaw, unpigmented at tip; lower jaw notched at tip; eyes protrude only slightly beyond silhouette of head in both dorsal and lateral aspects; labial region rounded and smooth, lips not swollen, not extending past eyes when viewed in dorsal aspect; interorbital region slightly convex; eye diameter 90% of interorbital distance; pupil horizontally ovoid; canthus rostralis medially bowed; loreal region slightly concave; eye diameter 55% of snout length; nostrils slightly laterally protuberant; eye-narial distance 1.25 times the distance from nostril to snout tip; internarial region flat; tympanum distinct, its diameter 89% of eye diameter; slight supratympanic fold present, extending from posterior edge of eye, over dorsal edge of tympanum, and terminating at supra-axillary (post-rictal) region; tongue ovoid, with a slightly narrower anterior attachment; choanae round, minute, at anterolateral edge of palate, separated by a distance five to seven times greater than their diameter, obscured in ventral aspect by palatal shelf; vomerine teeth absent; openings to vocal sac minute slits, at level of the angle of jaw.

Dorsal skin of body, head, and limbs generally smooth, lacking dermal folds, tubercles, or ridges; ventral surfaces of trunk, head, throat, and limbs smooth, with the exception of fanlike array of 12–14 small white tubercles on anterior margin of chin.

Cornufer Batraychylodes ......continued on the next page 9 5: 6;!#") C. montanus C. gigas & Manus length 50% of foot length; digits of manus narrow ( Figs. 4–5 ), ovoid in cross-section; dermal flanges absent on lateral edges of digits; terminal finger disks barely expanded (<2 times width of penultimate phalanges in Fingers I–IV), with short circum-marginal folds on distal tips of digits; minute supra-articular flaps present above penultimate-ultimate phalangeal articulation; fingers unwebbed; decreasing finger length III, IV, I, II; subarticular, supernumerary, and metacarpal tubercles poorly developed and unpointed; subarticular tubercles not prominent, low and rounded on ventral surfaces; one subarticular tubercle on Fingers I–II, two tubercles under Fingers III–IV; supernumerary tubercles indistinct, most distinguishable at base of Finger III; palmar surfaces basal to supernumary tubercles smooth, with slightly convex round tubercule between the supernumerary tubercle at base of Finger III and the medial palmar tubercle; thenar (inner metacarpal), medial palmar and outer metacarpal tubercles indistinct, flat, with poorly defined edges and distinguishable only by the lack of pigment; thenal tubercle elongate, along the medial edge of Finger I; medial palmar tubercle ovoid and larger than elongate, outer metacarpal; medial and outer metacarpal tubercles not separated at base; nuptial pads absent, forearm musculature not hypertrophied.

Hindlimbs short; tibia length 41% of snout-vent length, pes length 99% of tibia length; tarsus smooth, lacking folds, flaps, or tubercles; terminal disks of toes slightly expanded and ovoid (<2 times width of penultimate phalanges in Toes I–V); disk width of Toe IV 130% of disk width of Finger III; lateral dermal flanges absent along distal phalanges; circum-marginal grooves on distal ends of digits; supra-articular cutaneous folds slight; plantar surfaces of pes smooth ( Figs. 4–5 ), lacking supernumerary tubercles or texture on plantar surfaces, but with moderately developed, rounded (not pointed) subarticular tubercles, numbering one under Toes I–II and two under Toes III–V; decreasing toe length IV, III, V, II, I; outer metatarsal tubercle small, low, and rounded; inner metatarsal tubercle moderate, low, flat, and oblong; toes without interdigital webbing. Cloacal region lacking tubercles, but with small supracloacal dermal flap.

Measurements of holotype. SVL 18.1; ED 1.8; TD 1.6; HL 7.4; SNL 3.3; IOD 2.0; HW 6.2; FL 8.1; TBL 7.5; TSL 4.5; PL 7.4; ML 3.7; FA 3.7; Toe4L 4.1; Fin1L 1.1; Fin3L 2.2; Fin1DW 0.3; Fin3DW 0.3; PpFin3 0.2; Toe4DW 0.4; PpToe4 0.3.

Color of holotype in preservative. Dorsal surfaces of body and head exhibit nearly homogeneous light brown coloration; under magnification, background color of dorsal surfaces is pale cream and unpigmented with dense and nearly uniform speckling of brown chromatophores; less brown pigmentation present in interorbital region, extending posteriorly in a U-shaped pattern to a midpoint between the scapulae, giving slightly paler appearance to region along top of head than rest of body; two small, dark brown interorbital spots between anterior margin of eyes; tip of snout unpigmented; dorsal surface of limbs similar in background coloration to rest of body, but with darker brown transverse bars (three on forearm, three on femur, three on tibia, one on tarsus, two on metatarsus) with irregular margins and poorly defined; dorsal surfaces of hands, feet, and digits similarly colored as rest of the body and limbs with irregular dark spots and bars; the lateral surfaces of the head and body have a distinct dark brown band (darker than dorsal coloration) covering entire loreal and supralabial region, extending posteriorly along flanks diagonally (anteriodorsally to posteroventrally) towards inguinal region; this lateral band is darkest and near uniformly shaded along head (although tympanum is slightly lighter) becoming lighter and more diffusely speckled posterior to the humerus, into light tan irregular blotching in the inguinal region; the dorsal margin of this lateral band is nearly smooth and darkest brown in coloration (with respect to the rest of band), forming a clear dark line demarcating lighter dorsal and darker lateral colors extending from snout tip (although absent from unpigmented snout tip) along the canthus rostralis (above nostril) and upper eyelid, and posterior onto flanks slightly above tympanum; dark dorsolateral line fades in the inguinal region becoming prominent again on hindlimbs (although with more irregular margin) particularly on anterior edge of thigh and knee, fading into diffuse irregular tan blotching on tibia; underside of throat dark brown with lighter unpigmented mottling throughout; brown ventral coloration of throat fades posterior to the pectoral region becoming diffuse irregular tan blotching on pale cream midventer; ventral surfaces of forearms tan, darker brown towards outer edge of forearms; inner forearm with unpigmented band extending to underside of humerous; underside of hindlimbs similar in coloration to midventer, pale cream with brown irregular mottling; when viewed in posterior aspect cloacal region nearly uniform dark brown with smooth, dark dorsal margin above cloaca extending along the posterior of thighs clearly demarcating it from lighter dorsum; when in a seated position, this dark cloacal-femoral band grades into similarly colored dark band along outer edge of tarsus; small dark brown blotch just above cloaca; ventral surfaces of hands and feet dark brown with pale subarticular banding.

Variation. Compared to the holotype (SAMA R64760), dorsal coloration of the head and body of the paratypes ranges from a darker brown (SAMA R64763, SAMA R64764, SAMA R71008, SJR 10843) to a lighter cream (SAMA R64761, SAMA R64762, SAMA R71007) background coloration, and whereas the holotype has a nearly uniform dorsal coloration, all of the paratypes exhibit some degree of dorsal markings. SAMA R71008, SAMA R64763, and SJR 10843 possess a faint, thin and lighter vertebral stripe. Four of the paratypes (SAMA R71007, SAMA R71008, SAMA R64762, SAMA R64761) possess a pair of discrete darker brown to black dorsal spots that are posteroventrally located just behind the projection of the sacrum. Several paratypes also have irregular darker paravertebral spots between the scapulae, which are similar in size to the posterior dorsal spots in some (SAMA R64761, SAMA R64762) or a faint speckling in others (SJR 10843, SAMA R64763). SAMA R64761 exhibits more extensive irregular dark flecking and blotching along the dorsum. Three paratypes (SAMA R64764, SAMA R64761, SAMA R71007) possess a faint, thin and darker interorbital bar. Several paratypes also exhibit faint darker banding extending from behind eyes and/or the sacral region intersecting at the midscapular region, leaving a faint x or v-shaped pattern along the dorsum. All of the paratypes possess the transverse bars on the limbs as described in the holotype.

There is variation in breadth and extent of the dark lateral band that extends from the loreal region onto the flanks. In some individuals, the lateral band is darker and more filled in throughout the snout and flank region, giving appearance of dark face mask which remains solid (not fading) along the flanks. However, in others the lateral band is limited to more of a dark stripe from the tip of the snout, along the canthus rostralis, and posterior towards the inguinal region, which quickly fades ventrally into a light-brown coloration similar to the color of the dorsum. In two individuals (SAMA R64761, SAMA R71007), the lateral band is almost entirely limited to the snout along the canthus rostralis and quickly broken and fading posteriorly behind the eye, and in these individuals there is almost no darker shading below the band in the loreal and supralabial region of the snout or along the flanks. Ventral coloration of several of the paratypes is similar to that of the holotype, with a darker brown chin and throat region fading into a lighter cream midventral region and the underside of the hindlimbs with diffuse brown speckling. However, in one individual (SAMA R64762) the entire vertrum is similar in coloration to the brown throat, and in three individuals (SAMA R64761, SAMA R71007, SJR 10843) the brown pigmentation has faded, leaving the entire ventrum pale cream in coloration.

In most of the paratypes the skin texture is smooth, similar to that of the holotype, however several specimens exhibit finely granular skin along the dorsum, and one indiviudal (SAMA R64761) possesses small raised dorsal ridges and minute tubercles associated with the darker brown spots and flecks along the dorsal surface of the body. The small unpigmented tubercles on the chin and throat of the holotype are also present on most of the paratypes (although varying in number and extend along the throat); however, they are absent on the juvenile (SJR 10843) and female (SAMA R64763) specimens, suggesting that it may be a sexually dimorphic character. Additionally, the unpigmented projection of the snout on the holotype is prominent on most of the paratypes except the aforementioned juvenile and female paratypes, suggesting it may also be a sexually dimorphic character. Mensural variation is presented in Table 2. All mensural ratios for the single female paratype are within the ranges found in males, possibly indicating little sexual dimorphism in morphometric characters.

Color in life. Based on color images of the holotype (SAMA R64760, Figs. 2–3 ) and two paratypes (SAMA R64763 and SAMA R64764, Fig. 2 ) in life taken during the daytime, prior to preservation: Dorsal ground coloration of body and limbs of the holotype (SAMA R64760) a pale pinkish-salmon that is overlain with small darker pinkish-orange spots or streaking (streaks most prominent on the tibia) with some small irregular whitish spotting throughout; thicker (but faint) darker brown transverse bars extending across the hindlimbs and three thin (broken) darker brown transverse bars on the forearm; dorsal coloration of the head orangish, giving the appearance of a head cap (from just behind the unpigmented snout projection extending in a U-shaped pattern posterior to the eyes to the mid-scapular region); three small whitish and two small dark brown interorbital spots between the anterior margins of the eyes; margin of the upper eyelid with a thin light whitish band speckled with diffuse brown flecking. Dorsal coloration of the two paratypes different than that of the holotype. One paratype (SAMA R64763) has a more solid and uniformly darker brown coloration across the head, body, and limbs flecked with minute irregular dark brown, whitish, and reddish spots; limbs also with some suffused lighter orangishbrown highlights; a thin and slightly broken vertebral stripe is present, and a cluster of small dark-brown paravertebral spots in the midscapular region; faint darker brown transverse bars present on the limbs. The other paratype (SAMA R64764) possesses lighter tan dorsal coloration on the head and body, fading to darker brownishgray hindlimbs (forearms of similar color as hindlimbs), leaving the transverse bars on the hindlimbs poorly diagnosable; faint minute irregular dark-brown, whitish, and reddish spots or blotching on the dorsal surfaces of the body and limbs; light yellow-gold pigment faintly dusting the forelimbs and hands; two darker brown dorsal bands also apparent, one interorbital band between the midpoints of the orbits and upper eyelids, and one extending as a forward-facing V-shape connecting at the midscapular region and extending posteroventrally to the mid-dorsum.

The shading of the dark-brown lateral band (from the snout tip to the flanks) also varies slightly. In all three specimens the dorsal margin of the lateral band is the darkest brown, particularly in the loreal region along the canthus rostralis; in SAMA R64763 the dorsal margin of this lateral band is edged by a thin (slightly broken) white line that extends from the inguinal region diagonally (posteroventrally to anteriodorsally) across the distal margin of the upper eyelid, along the canthus rostralis, over the nostril, and around the tip of the snout (which lacks the unpigmented protrusion), and there is also some reddish-orange coloration associated with this white line along the flanks; this thin white line is absent in SAMA R64760 and SAMA R64764, which only have a lighter whitish distal margin along the upper eyelid; in SAMA R64763 and SAMA R64764 the shading of the dark brown lateral band is nearly homogeneous throughout the loreal and labial region and posteriorly onto the flanks (only becoming slightly paler on the tympanum and flanks, with scattered white flecks on the flanks), whereas in the holotype (SAMA R64760) the dark brown rapidly fades to a lighter brown ventrally towards the labial region and on the flanks posterior to the tympanum (where the background coloration is more pinkish-salmon, similar to the dorsum); in all three specimens (SAMA R64760, SAMA R64763, and SAMA R64764) this dark lateral band transitions onto the inner thigh out to the knee, although its dorsal margin is irregular.

Background ventral coloration of the holotype a light yellowish-straw overlaid with brown flecks that are suffused throughout the chin, throat, and pectoral region and fading posteriorly to faint brown blotches, giving the venter and overall dark (anterior) to light (posterior) appearance. Orange pigmentation in the form of blotching is also diffusely scattered throughout the ventral surfaces and most concentrated on the throat. The ventral surfaces of the hands and feet are mostly brown, which extends as an irregular dark band on the underside of the forearm and tarsus. Iris reddish with black streaking above and below the pupil.

Advertisement call. We quantified both temporal properties (calling rate, call duration, intercall interval, note repetition rate, internote interval), structure (number of notes), spectral characters (dominant frequency, other frequency components, harmonics), and variation in peak amplitude across distinct frequency components of the call. Our description of the advertisement call of male C. exedrus is based on recording segments (10 complete calls; 2 per subject, separated by a single intercall interval per male) from five individual males, one of which was captured and preserved (SAMA R64762). The other four calls corresponded to males that eluded capture. Because the majority of recorded males escaped, we report only ambient temperatures, which ranged only from 18.2– 19.2°C. Typical calling behavior in focal males was initiated by stimulation by other, nearby calling individuals; the majority of our recordings (4/5) contain calls of other individual males, precisely alternating notes, in the background of the focal subject ( Fig. 6 ).

The call of C. exedrus is a series of brief, high-frequency atonal notes, delivered in a slow train of variable duration, numbers of notes, and note repetition rates. We define the call as distinct group of atonal notes clustered temporally, and corresponding to times when calling males’ posture was upright, and when vocal sac inflation cycles and lateral body wall (flank) compressions were observed during sound production. Discrete calling efforts were unambiguously offset from periods of silence before and after sound production, corresponding to periods when the individual’s posture was less erect, and no vocal sac inflation, nor flank compression was observed. The typical call of C. exedrus begins with a series of slow, low-amplitude notes in the form of a quiet “tink….tink…tink,” which gradually increases in note repetition rate and amplitude, until a near constant rate and amplitude are achieved ( Fig. 6 ). Other than this initial “warm-up” period, rates and amplitudes of notes remain remarkably constant across the call of the new species.

Across all recording segments (n = 5 males), calling rate varied from 0.007 to 0.014 (mean = 0.011 ± 0.03 SD; n = 4) calls/s, and average call duration ranged from 16.0 to 93.9 (mean = 54.3 ± 30.1 SD; n = 5) s, with 17–75 (mean = 47.6 ± 24.4) notes/call. Intercall interval per male varied from 24.5 to 66 (mean = 39.5 ± 19 SD; n = 4) s. Within calls, internote intervals ranged from 0.80 to 0.95 (mean = 0.90 ± 0.06 SD) s, and note repetition rate varied from 0.79 to 1.1 (mean = 0.90 ± 0.11; n = 5) notes/s.

The spectral structure of calls of the new species are remarkably stereotyped ( Fig. 6 ). Individual notes contain discrete frequency components that are invariant across the call and apparently fixed within individuals. No frequency modulation or shifts in emphasis of individual frequency components were observed in any of the calls studied here (n = 5). Calls consist of internally identical notes (no switching of dominant and fundamental frequency observed), each containing a fundamental (lowest; mean = 2.2 ± 0.1 SD; range 2.0–2.3 kHz; n=5) frequency, differing discretely from the dominant (= emphasized; mean = 4.3 ± 0.2 SD; range = 4.1–4.7 kHz; n = 5) frequency. Calls possess 2–5 additional frequency components at means of 6.7 (± 0.3 SD; range = 6.3–7.2; n = 5), 8.6 (± 0.4 SD; range = 8.3–9.2; n = 5), 11.4 (± 0.4 SD; range = 10.9–11.7; n= 3), and approximately 13 and 15 (n = 2 each) kHz, respectively. Relative amplitude of the fundamental frequency ranged from 0.88–0.96 (mean = 0.91 ± 0.04; n = 5).

The call of the new species is similar to C. vertebralis (the only other Cornufer [subgenus Batrachylodes ] with recorded calls available for comparison) in overall structure: a series of slow, high-frequency notes. However the call of C. (B.) vertebralis is more rapid (calling rate mean = 0.019 ± 0.05 SD; range = 0.014–0.023; n = 3) calls/s, generally consists of fewer (11–39; mean = 21.3 ± 15.4; n= 3) notes/call, which are delivered at a more rapid rate, and pulse repetition rate varied from 0.96 to 1.6 (mean = 1.3 ± 0.33; n = 3). Cornufer (Batrachylodes) vertebralis calls also have distinct intranote structure, whereas C. exedrus calls contain only simple, unstructured atonal notes, entirely lacking within-note structure or variable amplitude modulation.

Ecology, Distribution, and Natural History. Cornufer exedrus is known only from elevations of 1500–1700 m above sea level, based on our surveys in the vicinity of the type locality, Tompoi Camp, in the Nakanai Mountains of East New Britain Province in the Bismarck Archipelago. ( Fig. 1 ). The camp was located in a patch of mossy montane forest surrounded by dense, near-impenetrable thickets of thin, scrambling Nastus bamboo. During the survey period (19–25 April, 2009), this site was extremely wet from the frequent rain and dense fog that shrouded the forest on most days. Males called from hidden positions in wet litter on the forest floor, and densities were highest where the forest had not been invaded by dense bamboo thickets. The species appears to be crepuscular (calling activity peaked at dawn and dusk) but many non-calling individuals were active in wet litter during the day, particularly after rain. In contrast, calling activity was greatly reduced at night, and few animals were observed to be active on the forest floor after dusk. Sympatric species included Cornufer adiastolus , C. citrinospilus , C. mamusiorum , C. nakanaiorum , C. cf. sulcatus , and Oreophryne brachypus .

Figures

FIGURE 1. Map of the Bismarck and Solomon Archipelagos showing the distributions of all species of Cornufer within the subgenus Batrachylodes, including the type locality of Cornufer exedrus (black circle) at Tompoi Camp (~1,600 m above sea level) in the Nakanai Mountainsof eastern New Britain.

FIGURE 2. Photographs in life of the holotype(A:SAMA R64760) and paratypes (B: SAMA R64764; C: SAMA R64763) of Cornufer (Batrachylodes) exedrus, as well as other representatives of the subgenus Batrachylodes from the Solomon Archipelago (D: KU 341539, Batrachylodes trossulus, Choiseul Is., Solomon Islands; E: KU 341494–5, Batrachylodes minutus, Choiseul Is., Solomon Islands; F: KU 341497, Batrachylodes wolfi, Shortland Is., Solomon Islands).

FIGURE 4. Close-up images of the holotype of C. exedrus (SAMA R64760) in life showing ventral (A), lateral (B), palmar (C), and plantar and cloacal(D) regions.

FIGURE 5. Close-up images of the ventral surfaces of the right hand (A) and left foot (B) of the holotype of C. exedrus (SAMA R64760)in preservative (scale bar = 1 mm).

FIGURE 6. The advertisement callof a male C.exedrus paratype (SAMA R64762)recorded at ambient temperature of 18.4°C on 21 April 2009 (1830 hr, by SJR): (A) oscillogram (relative amplitude vs. time in s) of three typical notes and (B) a 2 s audiospectrogram (frequency in kHzvs.time ins) from themiddle of a60 s, 44-pulsed call (D, relativeamplitude vs. time ins). Note (C) the corresponding power spectrum (Fast Fourier Transformation; relative amplitude vs. frequency in kHz) of a single note, depicting a fundamental (lowest) frequency of 2.3 kHz and the dominant (emphasized) frequency of 4.7 kHz.