Cophixalus verrucosus (Boulenger, 1898)

Kraus, Fred, 2012, Papuan frogs of the genus Cophixalus (Anura: Microhylidae): new synonyms, new species, and a dichotomous key, Zootaxa 3559, pp. 1-36 : 6-7

Cophixalus verrucosus (Boulenger, 1898)

Boulenger (1898) also described Cophixalus verrucosus on the basis of two specimens from Moroka and six specimens from Mt Victoria. Both localities are within approximately 70 km of each other and outside of Port Moresby in the central Owen Stanley Range, Central Province, Papua New Guinea. The diagnostic features mentioned by Boulenger are a sharp canthus, vertical lores, projecting snout, head as long as broad, large finger discs, tuberculate skin, and hidden surfaces of thighs and groin with a pattern of large white spots on a black ground. Zweifel (1956a) emphasized the color pattern on the thighs as uniquely diagnostic within Cophixalus , and Zweifel (1979) and Menzies (2006) added the detail that these colors are yellow on a dark-brown ground in life. Menzies (2006) further noted that the venter is gray anteriorly and bright lemon yellow posteriorly and on the hind limbs. In my experience, this too appears unique among the known species of the genus. This species ranges from sea level to 1600 m elevation throughout most of the Papuan Peninsula and on several of the offshore islands of Milne Bay Province (unpubl. data).

Hiaso (2002) described Cophixalus aimbensis on the basis of 11 specimens (one cleared and stained) collected from along the Aimba River, Sudest Island, Milne Bay Province, Papua New Guinea. He claimed that all specimens were male, but in fact three of them (UPNG 8464, 8466, 8469) are female. Hiaso diagnosed his new species from its congeners with the characters: (1) head as long as broad, (2) snout rounded dorsally and projecting in profile, (3) fingers and toes with small terminal discs with marginal grooves, and (4) width of finger discs slightly larger than width of toe discs. The first two features were both mentioned by Boulenger (1898) in his description of C. verrucosus . Furthermore, the first does not serve to distinguish C. aimbensis from most species of Cophixalus . My remeasurements of the nine remaining type specimens of C. aimbensis (UPNG 8468 and 8470 are missing and presumably destroyed, cf. Shea and Kraus, 2007) show that HL/HW varies in that series from 0.88–0.99, following the mode shown by most members of the genus, which tend to have the head slightly shorter than wide. These figures differ from those provided by Hiaso (HL/HW = 0.99–1.14), but he didn’t define the landmarks used for his measurements, so the source of this difference is uncertain. The second feature is distinctive within Papuan Cophixalus , but it is also shared with C. cheesmanae Parker, 1934 and C. verrucosus . Hiaso’s third diagnostic feature characterizes almost all members of Cophixalus and is, hence, uninformative. His fourth feature characterizes almost half the Papuan species of Cophixalus . Indeed, Menzies and Tyler (1977) thought the greater breadth of the finger discs relative to the toe discs was a diagnostic feature of Cophixalus vis-à-vis Copiula Méhely, 1901 , but there is more variation in that feature within Cophixalus (as currently constituted) than they thought, and a large number of species do not fit this description ( Kraus and Allison, 2009).

Although not used in his diagnosis, Hiaso (2002) also mentioned that his specimens were bright yellow on the posterior of the abdomen and on the hidden surfaces of the limbs, a feature noted above to be unique to C. verrucosus . In his description, Hiaso (2006) did not explicitly compare his new species to any other member of the genus Cophixalus , but, instead, concentrated on distinguishing it from members of the genus Copiula , which it resembles only in having a projecting, though differently shaped, snout. From the original description, then, C. aimbensis cannot be distinguished from C. verrucosus .

In 2004 I obtained a large series of Cophixalus verrucosus on Sudest Island, the type locality of C. aimbensis . There is nothing to distinguish these specimens from the type series of C. aimbensis : they show complete overlap in bivariate morphometric ratios with the type series of C. aimbensis and they match those specimens in all details of morphology and color pattern. They also closely match specimens of C. verrucosus I have collected from several localities throughout the Papuan Peninsula in everything but their larger body size (Sudest male SV = 24.5–31.0 mm vs. 19.5–27.5 mm elsewhere; Sudest female SV = 29.6–37.0 mm vs. 23.0–31.7 elsewhere, Table 2). In particular, they share the projecting snout, vertical lores, hidden surfaces of thighs and groin dark brown with large yellow spots, grey chin and throat, and lemon-yellow abdomen and undersurfaces of legs that are characteristic of C. verrucosus . That all of these specimens represent C. verrucosus is further indicated by principal components analysis of the Sudest specimens compared to C. verrucosus from several localities elsewhere in their range. In both bivariate ( Fig. 3 ) and 3-dimensional (not shown) plots, the Sudest specimens cannot be distinguished from C. verrucosus , although they lie to the extreme of that species along PC1. However, this axis largely reflects specimen size, judging by the uniformly positive loadings of the constituent variables ( Table 3 ), and this merely confirms that Sudest animals are on average larger than those from other populations of C. verrucosus . As can be seen from the loadings and from examination of scatter plots, there is no information expressed in PC2 or PC3 that serves to distinguish among populations ( Fig. 3 B). Smaller-sized specimens from Sudest overlap the morphological distribution of specimens from Moroka and approach those from Mt. Victoria, the type localities for C. verrucosus . More importantly, there is no obvious break in morphotype between mainland and Sudest populations. This holds as well if the mainland localities are each distinguished separately on the plot (graph not shown), in which case the Sudest specimens adjoin and overlap those from the Cloudy Mts., the geographically most proximate population on the mainland. That population, in turn, overlaps populations from elsewhere in Milne Bay and Central provinces.

SV (mm) Males Females

mean (n) range mean (n) range Mt Victoria 24.1 (2) 23.2–25.0 28.6 (4) 26.8–31.7 Moroka 22.8 (2) 22.3–23.3 26.2 (2) 26.0–26.4 Mt Simpson 20.6 (4) 19.5–22.5 24.7 (3) 24.1–25.1 Oya Tabu 25.1 (7) 23.3–27.5 28.1 (3) 23.9–31.5 Cloudy Mts ― 27.3 (10) 23.0–31.1 Sudest Island 28.6 (5) 24.5–31.0 33.0 (5) 29.6–37.0 It could perhaps be posited that the Sudest population in fact represents a distinct, separate species that is merely closely related to C. verrucosus . If true, one might expect differences to appear in call parameters between the Sudest animals and representatives of C. verrucosus from the mainland. This too is not the case. The call of C. verrucosus is a single pulsed note emitted at time intervals ranging from 1.5–52 s in animals recorded by me. To the human ear the call sounds raspy, similar to a thumb dragged over the teeth of a comb. No call parameter serves to distinguish the Sudest population from mainland populations of C. verrucosus ( Table 4 , Figs. 2 , 4 ); instead, there is considerable overlap with mainland populations in all these parameters. Of particular interest is that the range of variation in the call parameters from the Sudest sample, with the exception of pulse rate, are entirely contained within the range seen for those same parameters from the Cloudy Mts sample ( Table 4 ), its geographically nearest neighbor. Because no morphological, color-pattern, or call differences serve to distinguish C. aimbensis from C. verrucosus , and both taxa share a number of color-pattern features unique within Cophixalus , I synonymize the former with the latter. The taxonomic history for the species becomes:

Figures

FIGURE 2. Map of southeastern New Guinea and adjacent islands showing type localities for Cophixalus ateles (filled star), C. sisyphus and C. nexipus sp. nov. (filled circle), C. verrucosus (stars), C. albolineatus sp. nov. (filled square), C. clapporum and C. amabilis (filled triangle), and an additional locality for C. albolineatus sp. nov. (open square). Five additional numbered localities from which call samples of C. verrucosus were obtained for this study are: 1 = Sudest Island, type locality for C. aimbensis; 2 = Cloudy Mts.; 3 = N slope of Mt. Simpson; 4 = W slope of Mt. Dayman; and 5 = Fane. Elevational contours are at 1000 m (light gray) and 1500 m (dark gray).

FIGURE 3. Bivariate plot of (A) principal component 1 (eigenvalue = 0.168) versus principal component 2 (eigenvalue = 0.012), and (B) principal component 2 versus principal component 3 (eigenvalue = 0.008) for eight type specimens of Cophixalus aimbensis (closed circles), ten other specimens of C. verrucosus from Sudest Island (open circles), ten specimens of C. verrucosus from Mt. Victoria and Moroka, the localities from which the type series was obtained (diamonds), and 27 other specimens of C. verrucosus from Milne Bay Province in southeastern Papua New Guinea (squares). Component loadings presented in Table 2.

FIGURE 4. (A) Waveform, (B) power spectrum, and (C) spectrogram of three calls of Cophixalus verrucosus (BPBM 15261) recorded in the Cloudy Mts., 10 April 2002, air temperature 24.0 C; and (D) waveform, (E) power spectrum, and (F) spectrogram of three calls of C. verrucosus (BPBM 20185) recorded on Sudest Island, 16 April 2004, 26.0 C.