Nine additional specimens of Abavorana luctuosa sensu lato from Borneo were sequenced (Table 1). Genomic DNA was isolated from liver or skeletal muscle stored in 95% ethanol using a SPRI magnetic bead extraction protocol (https://github.com/phyletica/lab-protocols/blob/master/extraction-spri.md). A combined fragment (1443 bp) of the mitochondrial genes 16S, ND1 and flanking tRNAs (tRNA-Leu, tRNA-Ile, and tRNA-Gln) was amplified using a double stranded polymerase chain reaction (PCR) under the following conditions: 1.0 µl genomic DNA (10–30 µg), 1.0 µl (concentration 10 µM) light strand primer 16S-frog 5’ TTACCCTGGGGATAACAGCGCAA–3’ (Camargo et al. 2006), 1.0 µl (concentation 10µM) heavy strand primer tmet-frog 5’–TTGGGGTATGGGCCCAAAAGCT–3’ (Camargo et al. 2006), 1.0 µl dinucleotide pairs (1.5 µM), 2.0 µl 5x buffer (1.5 µM), 1.0 µl MgCl 10x buffer (1.5 µM), 0.1 µl Taq polymerase (5units/µl), and 6.4 µl ultra-pure H2O. All PCR reactions were carried out on an Axygen Maxygene II gradient thermocycler under the following thermal profile: initial denaturation at 94°C for 2 min 30s, followed by 30 cycles of a secondary denaturation at 94°C for 1 min, annealing at 54°C for 1 min, elongation at 72°C for 1.5 min. PCR products were visualized via a 1.0% agarose gel electrophoresis. Successful PCR products were sent to GENEWIZ® for PCR purification, cycle sequencing, sequencing purification, and sequencing using the same primers as in the amplification. Sequences were analysed from both the 3’ and the 5’ ends separately to confirm congruence between reads. All forward and reverse sequences were uploaded and edited in Geneious™ version v11 (Kearse et al. 2012). Following sequence editing, 16S, ND1 and flanking tRNAs were aligned using the MAFTT v7.017 (Katoh et al. 2002) plugin under the default settings in Geneious™ (Kearse et al. 2012). Sequences were checked in Mesquite v3.5 (Maddison & Maddison, 2019) to ensure the correct amino acid reading frame for the coding region ND1.

The new sequences were added to the molecular phylogenetic dataset of Quah et al. (2017) along with additional GenBank sequences from Shimada et al. (2011) (Table 1). Ingroup samples consisted of 30 individuals representing three nominal species of Abavorana (Table 1). Outgroups species used to root the tree were Pulchrana sundabarat Chan, Abraham, Grismer & Brown, and Meristogenys stenocephalus Shimada, Matsui, Yambun & Sudin, based in part on the relationships of Quah et al. (2017). New sequence data generated for this study are deposited on GenBank (Table 1).

Maximum likelihood (ML) and Bayesian Inference (BI) were used to generate phylogenetic trees. The data were partitioned into four partitions (ND1 by codon positions [1–3] and 16S+tRNAs) and best-fit models of evolution were determined in the IQ-TREE webserver version (Nguyen et al. 2015; Trifinopoulos et al. 2016) using the Bayesian information criterion (BIC; Schwarz 1978) implemented in ModelFinder (Kalyaanamoorthy et al. 2017). The model K2P+G4 was the best-fit model of evolution for ND1 codon position 1, HKY+I for codon position 2, TN+I for codon position 3 and TIM2e+G4 for 16S+ tRNAs. The ML analysis was performed using the IQ-TREE webserver (Trifinopoulos et al. 2016) with 1000 bootstrap pseudoreplicates using the ultrafast bootstrap (UFBoot) option (Minh et al. 2013; Hoang et al. 2017). The Bayesian analysis was performed in BEAST v2.5.1 (Bouckaert et al. 2014) using bModelTest (Bouckaert & Drummond 2017) to numerically integrate over uncertainty in models of substitution, while estimating the phylogeny using Markov chain Monte Carlo (MCMC). The models selected by bModelTest for ND1 codon position 1 was 123141 (posterior support 35.22%), codon position 2, 121131 (posterior support 6.22%), codon position 3, 121131 (posterior support 29.94%) and 123454 (posterior support 29.01%) for the noncoding partition 16S+tRNAs. We ran the MCMC chain for 100 million generations while sampling every 10,000 generations. Stationarity was checked with Tracer v1.7 (Rambaut et al. 2018) and the effective sample sizes for all parameters were greater than 200. A maximum clade credibility tree was generated by summarizing the posterior distribution of trees after discarding the first 25%, using mean node heights with TreeAnnotator v2.5.1 (Bouckaert et al. 2014). We considered Bayesian posterior probabilities (BPP) of 0.95 and above and ultrafast bootstrap support values (UFBoot) of 95 and above as strong nodal support (Huelsenbeck et al. 2001; Minh et al. 2013). Uncorrected pairwise sequence divergences (p-distance) were calculated in MEGA v6.06 using default settings which used pairwise deletion for the treatment of gaps and missing data in the dataset (Tamura et al. 2013) (Table 2).

Morphological data from 62 adult specimens (34 males; 28 females) of Abavorana luctuosa sensu lato from 21 populations throughout the Thai-Malay Peninsula and Borneo (Appendix 1), and eight specimens of A. nazgul (seven males; one female) from the type locality in Peninsular Malaysia, were taken following the methodology of Quah et al. (2017) (Table 3). Notes on colour pattern were taken from digital images of living and euthanized specimens prior to preservation. Sex was determined by the presence of the humeral glands in adult male Abavorana as they lack vocal sacs and nuptial pads (Inger 1966). The following characters were measured by ESHQ with a Control Company digimatic caliper to the nearest 0.01 mm on the right side of the body for symmetrical characters: snout-vent length (SVL), from the tip of snout to the vent; head length (HL), from the posterior margin of mandible to the tip of snout; head width (HW), measured at the level of the jaw articulation points; snout length (SL), from the anterior margins of eyes to the tip of snout; snout width (SW), distance between the anterior margins of eyes; interorbital diameter (IOD), distance across the top of the head between the medial margins of the upper eyelids at their closest points; internarial distance (IND), distance between the medial margins of the nostrils; eye diameter (ED), distance between the anterior and posterior margins of the eyeball; tympanum diameter (TD), the horizontal width of the tympanum at its widest point; brachium length (BL), distance from the axilla to most distal point of inflection on flexed elbow; forearm length (FAL), distance from outer margin of flexed elbow to the base of the inner metacarpal tubercle; manus length (ML), distance from the proximal edge of the outer metacarpal tubercle to the tip of the third finger; femur length (FL), distance from the vent to the outer margin of the flexed knee; crus length (CL), distance from the outer face of flexed knee to the outer margin of tarsal inflection; tarsal length (TL), the outer margin of flexed tarsus to the base of the inner metatarsal tubercle; pes length (PL), measured from the proximal edge of the inner metatarsal tubercle to the tip of fourth toe; humeral gland length (HG), the horizontal length of the humeral gland. Discrete colour pattern data, along with univariate and multivariate analyses of morphological data (see statistical analyses below) were then used to search for characters and morphospatial patterns bearing statistically significant differences that were consistent with the previous designations of the species-level hypotheses, thus providing independent diagnoses to complement the molecular analyses.

Museum abbreviations are: FMNH (Field Museum of Natural History, Chicago, Illinois); LSUHC (La Sierra University Herpetological Collection, Riverside, California); MNHN (Muséum National d’Histoire Naturelle, Paris); MSNG (Museo Civico di Storia Naturale “Giacomo Doria”, Genova); SP (Sabah Parks Collection, Sabah, Malaysia); USMHC (Universiti Sains Malaysia Herpetological Collection); and ZRC (Zoological Reference Collection, Lee Kong Chian Natural History Museum [LKCNHM], Singapore).

All analyses were performed using the platform R version 3.2.1 (R Core Team 2015). Owing to a lack of genetic material from the type locality of Rana decorata at Mount Kinabalu, Sabah, an a priori decision as to the membership to which species to assign specimens prior to additional downstream analyses was based on the colour pattern of their venters which has been shown to be a distinguishing characteristic for the genus (Fig. 9; Quah et al. 2017). Specimens with bold black and white reticulations along flanks, and banding on the underside of their legs are identified as the decorata pattern morph, while those with generally immaculate venters are identified as the luctuosa pattern morph. To reduce allometric bias as a result of ontogeny, the characters of each species were scaled using the following equation: X_adj = log(X)-β[log(SVL)-log(SVL_mean)], where X_adj = adjusted value; X = measured value; β = unstandardized regression coefficient for each population; SVL = measured snout-vent length; SVL_mean = overall average SVL of all samples (Thorpe, 1983; Lleonart et al. 2000). Each species was scaled separately and the data concatenated so as not to conflate interpopulational variation. The measurements for the sexes were analysed separately to eliminate bias caused by sexual dimorphism. All raw measurements are presented in Table 3.

A MANOVA analysis was performed to test for significantly different (p < 0.05) multivariate structure in the mensural data and subsequently viewed in a principal component analysis (PCA). PCA and discriminant analysis of principal components (DAPC) were performed using the ADEGENET package in R (Jombart et al. 2010) to determine if the morphospatial position of Bornean Abavorana luctuosa populations with boldly marked venters and other populations of A. luctuosa with generally immaculate venters from Borneo and the Thai-Malay Peninsula coincided with the putative species boundaries delimited by the molecular phylogenetic analyses. All mensural data were incorporated in the analyses except HG which is absent in females. Principal components of the DAPC with eigenvalues accounting for 90–99% of the variation in the data were retained for the DAPC according to the criterion of Jombart et al. (2010). For this analysis, the first seven PCs out of 17 (males) and 16 (females) were retained which accounted for 94.9% (males) and 97.6% (females) of the variation.

Independent of the MANOVA, an analysis of variance (ANOVA) was conducted on mensural characters. Since the species and population sample sizes were unequal, a Levene’s test for homogeneity of variances among the mensural characters was conducted prior to an analysis of variance (ANOVA). A Welch’s ANOVA was conducted on characters with unequal variances (p ≥ 0.05) and a standard ANOVA was conducted on characters with equal variances (p ≤ 0.05) to test for the presence of statistically significant mean differences (p ≤ 0.05) between all combinations of species or population pairs in the data set. Characters with equal variances and bearing statistical differences were subjected to a TukeyHSD test to ascertain which population pairs differed significantly from each other for those characters (Table 4). Characters with unequal variances bearing statistical differences were subjected to a Games-Howell test to ascertain which population pairs differed significantly from each other for those characters (Table 4). Summary statistics were generated for the mensural data (Table 5).

The BI and ML analyses recovered trees with identical topologies. Specimen FMNH 231052 with bold black and white mottling on the ventral side of its flanks and banding on the underside of its legs from Danum Valley, Sabah (Fig. 6A & B) was recovered as the strongly supported (BPP=1.0, Ufboot=89) sister lineage of the remaining Abavorana (Fig. 1). The remaining A. luctuosa sensu lato from the Thai-Malay Peninsula and Borneo bearing generally immaculate venters and underside of their legs (Figs 1, 4) formed a strongly supported (BPP=1.0, Ufboot=96) clade, and were recovered as the sister species to A. nazgul (BPP=1.0, Ufboot=99). FMNH 231052 differs by an uncorrected pairwise sequence divergence of 16.0–17.0 % from A. nazgul and other A. luctuosa populations, respectively (Table 2). With the inclusion of additional Bornean specimens in this study, two reciprocally monophyletic lineages (BPP=1.0, Ufboot=96) of A. luctuosa with generally immaculate venters and underside of hind limbs were recovered—one from Peninsular Malaysia and another from Borneo (Fig. 1). Nevertheless, these populations differ from each other by only a 2.0–3.0 % uncorrected pairwise sequence divergence (Table 2).

Figure 1. 

Phylogenetic tree of the genus Abavorana; Maximum likelihood phylogram for species of the genus Abavorana. Nodal support is indicated by Bayesian posterior probability (BPP) / Ultrafast bootstrap (UFboot) respectively.

The MANOVA indicated there were significant (Wilks’ lamda, p=2.2^-16) differences in mensural data between both sexes of the two pattern morphs of Abavorana luctuosa sensu lato and A. nazgul, as visualized in the PCA and DAPC (Fig. 2). For males, the PCA and DAPC showed distinct separation between both the decorata and luctuosa pattern morph populations. In the PCA, the males of both phenotypes were clearly separated along PC1. PC1 accounts for 65.4% of the variation (Fig. 2), and loads most heavily for SVL, HW and CL (Appendix 2). In females, the PCA did not show a separation between the two pattern morphs as there was significant overlap along PC1 and PC2. However, the DAPC showed complete separation in the 66% confidence ellipses between the females of both phenotypes (Fig. 2). FMNH 231052 bearing the decorata pattern morph from Danum Valley, Sabah clustered in morphospace with topotypic material referable to Rana decorata from Mount Kinabalu, Sabah (FMNH 109935 and ZRC 1.709) in the DAPC. In both sexes, the luctuosa pattern morph populations from Borneo and the Thai-Malay Peninsula overlapped in both the PCA and DAPC analyses.

Figure 2. 

PCA and DAPC for the genus Abavorana; PCAs (left) with convex hull polygons and DAPCs (right) showing the morphospatial relationships of male (top) and female (bottom) Abavorana decorata, A. nazgul, and A. luctuosa populations from the Malay Peninsula and Borneo. 1 = FMNH 231052, 2 = ZRC 1.709 and 3 = FMNH 109935.

The analyses also revealed that both venter phenotypes also differ in a number of statistically significant mensural metrics (Table 4 & 5). Across both sexes, the two phenotypes differed significantly in their head width, snout width, interorbital diameter, tympanum diameter, and femur length. The specimens with the reticulated venters and boldly banded undersides of their legs (i.e. the decorata pattern morph) that match the descriptions of Rana decorata presented:

Significantly wider heads in both sexes, with HW of adult males 17.66–19.53 mm, mean 18.60 ± SD 1.32, n = 2 (vs. 12.25–14.41 mm, mean 13.39 ± SD 0.52, n = 22, P < 0.001, in Bornean Abavorana luctuosa, and 13.35–15.58 mm, mean 14.24 ± SD 0.66, n = 10, P < 0.001, in Malay Peninsula A. luctuosa) and adult females 10.62–21.07 mm, mean 17.18 ± SD 4.68, n = 6 (vs. 12.71–15.74 mm, mean 14.69 ± SD 0.83, n = 11, P < 0.001, in Bornean A. luctuosa, and 13.83–16.84 mm, mean 15.77 ± SD 1.01, n = 11, P < 0.01, in Malay Peninsula A. luctuosa); significantly wider snouts in both sexes, with SW of adult males 9.91–10.71 mm, mean 10.31 ± SD 0.57, n = 2 (vs. 7.33–8.22 mm, mean 7.73 ± SD 0.23, n = 22, P < 0.001, in Bornean A. luctuosa, and 7.11–8.62 mm, mean 8.08 ± SD 0.44, n = 10, P < 0.001, in Malay Peninsula A. luctuosa) and adult females 6.25–11.68 mm, mean 9.54 ± SD 2.28, n = 6 (vs. 7.65–9.02 mm, mean 8.33 ± SD 0.51, n = 11, P < 0.001, in Bornean A. luctuosa, and 8.21–9.38 mm, mean 8.84 ± SD 0.40, n = 11, P < 0.05, in Malay Peninsula A. luctuosa); significantly wider interorbital diameter in both sexes, with IOD of adult males 5.59–5.95 mm, mean 5.67 ± SD 0.40, n = 2 (vs. 3.78–4.70 mm, mean 4.24 ± SD 0.30, n = 22, P < 0.001, in Bornean A. luctuosa, and 3.25–5.14 mm, mean 4.35 ± SD 0.52, n = 10, P < 0.001, in Malay Peninsula A. luctuosa) and adult females 3.91–6.78 mm, mean 5.63 ± SD 1.22, n = 6 (vs. 4.15–5.60 mm, mean 4.69 ± SD 0.37, n = 11, P < 0.001, in Bornean A. luctuosa, and 3.80–5.64 mm, mean 4.93 ± SD 0.54, n = 11, P < 0.001, in Malay Peninsula A. luctuosa); significantly wider tympanum diameter in both sexes, with TD of adult males 4.54–4.60 mm, mean 4.57 ± SD 0.04, n = 2 (vs. 2.50–3.83 mm, mean 3.16 ± SD 0.29, n = 22, P < 0.001, in Bornean A. luctuosa, and 2.62–3.66 mm, mean 3.30 ± SD 0.31, n = 10, P < 0.001, in Malay Peninsula A. luctuosa) and adult females 2.27–5.63 mm, mean 4.30 ± SD 1.43, n = 6 (vs. 2.95–3.77 mm, mean 3.44 ± SD 0.25, n = 11, P < 0.001, in Bornean A. luctuosa, and 3.24–4.24 mm, mean 3.87 ± SD 0.27, n = 11, P < 0.05, in Malay Peninsula A. luctuosa); and significantly longer femur length in both sexes, FL of adult males 25.58–26.10 mm, mean 25.84 ± SD 0.37, n = 2 (vs. 17.81–21.57 mm, mean 19.72 ± SD 1.08, n = 22, P < 0.001, in Bornean A. luctuosa, and 18.38–22.04 mm, mean 20.27 ± SD 1.26, n = 10, P < 0.001, in Malay Peninsula A. luctuosa) and adult females 17.19–31.34 mm, mean 25.29 ± SD 6.53, n = 6 (vs. 19.13–23.68 mm, mean 21.73 ± SD 1.32, n = 11, P < 0.001, in Bornean A. luctuosa, and 21.55–26.37 mm, mean 23.83 ± SD 1.51, n = 11, P < 0.05, in Malay Peninsula A. luctuosa). In addition to these, the phenotypes differed significantly from each other and A. nazgul by a suite of other mensural characters (Table 4 & 5).

The analyses demonstrate unequivocally that Abavorana luctuosa on Borneo are composed of two genetically and morphologically distinct forms. Abavorana luctuosa described by Peters (1871) on the basis of a single female specimen from Sarawak (Fig. 3) has only very faint speckling on its venter and underside of the limbs. Therefore, specimens from southern Thailand, Peninsular Malaysia, and Borneo that possess this phenotype shall be referred to as A. luctuosa (Figs 4, 7E–H, 8A–F, 9A–B). The second lineage is readily identifiable by the reticulated pattern on their venters and boldly banded underside of their hind limbs, and is endemic to Borneo. Nine specimens examined from the states of Sabah (Mount Kinabalu, Kota Belud, Maliau Basin, Tambunan and Danum Valley) and Sarawak (Bario) bear this ventral colour patter and match the description of Rana decorata Mocquard, 1890 that was described from Mount Kinabalu, Sabah, East Malaysian Borneo. They resemble the colour pattern of the type series (MNHNP 1889.226–28) (Figs 5, 6, 7A–D, 9C–D). Rana decorata was synonymised with A. luctuosa by Boulenger (1891) without clear justification yet followed in subsequent works (e.g., Boulenger 1920; Van Kampen 1923; Quah et al. 2017 and references therein). Thus, given the preponderance of data, R. decorata Mocquard, 1890, in the combination Abavorana decorata comb. nov. (Mocquard, 1890) is resurrected for the populations bearing this phenotype.

Figure 3. 

Holotype of Abavorana luctuosa; A: Original illustration of the holotype of Abavorana luctuosa from Peters (1872). B: Dorsum of the holotype of A. luctuosa (MSNG 29344). C: Venter of the holotype of A. luctuosa (MSNG 29344). Photographs by Giuliano Doria.

Figure 4. 

Bornean Abavorana luctuosa; Dorsal (A & C) and ventral (B & D) pattern of Abavorana luctuosa from Borneo. A & B: Top row (from left to right); FMNH 234972, FMNH 234973, FMNH 234975 and FMNH 234977 (all specimens from Sipitang District, Sabah). Bottom row (from left to right); FMNH 234980 (Sipitang District, Sabah), FMNH 269142, FMNH 273216 and FMNH 273217 (specimens from Bintulu Division, Sarawak). C & D (from left to right): ZRC 1.11938, ZRC 1.11943 (both from Kubah, Sarawak) and ZRC 1.2270 (unspecified location in Borneo).

Figure 5. 

Syntypes of Abavorana decorata; A & D: Dorsum and venter of MNHN-RA-1889.226. B & E: Dorsum and venter of MNHN-RA-1889.227. C & F: Dorsum and venter of MNHN-RA-1889.228 (Photographs by Christine Reinhardt). G: Original illustration of the syntypes of Abavorana decorata from Mocquard (1890). Illustrations assignable to specimen MNHN-RA-1889.226 (left) and MNHN-RA-1889.228 (right).

Figure 6. 

Examined Abavorana decorata specimens; Dorsal (A & C) and ventral (B & D) pattern of Abavorana decorata from Borneo. A & B (from left to right): FMNH 109935 from Kiau, Mount Kinabalu, Sabah (left), FMNH 109936 from Kiau, Mount Kinabalu, Sabah (middle), FMNH 231052 from Danum Valley Research Center, Lahad Datu District, Sabah (top right), FMNH 128127 from Bario, Sarawak (bottom right). C & D: SP 26873 from Tawau Hills, Sabah (left) and SP 02810 from Maliau Basin, Sabah.

Abavorana decorata comb. nov. (Mocquard, 1890)

Decorated Mahogany Frog Figs 5, 6, 7A–D, 9C–D

Rana decorata: Mocquard 1890:145–146; Guibé 1950:41

Rana luctuosa (in part): Boulenger 1891:341 & 342, 1912:238, 1920:183 & 184; Flower 1896:904, 1899:896 & 897; Hanitsch 1900:73; Butler 1904:199; Smith 1931:16; Van Kampen 1923:196 & 197; Taylor 1962:451–453; Inger 1954:250, 1966:206–208, 1978:312, 1985:iii, 5, 38, 40, 56, 57, 76; Grandison 1972:66 & 67; Berry 1975:74 & 75, Fig. 50 (right); Matsui 1979:333; Inger & Stuebing 1992:42, 2005:19, 27, 33, 149 & 150; Inger & Tan 1996:563; Inger et al. 1996:363; Malkmus 1994:244; Manthey & Grossmann 1997:111 & 112; Khonsue & Thirakhupt 2001:73; Malkmus et al. 2002: 163–165, figs 159, 160; Das 2006a:5.

Rana (Hylorana) luctuosa (in part): Boulenger 1920:126.

Rana (Hylarana) luctuosa (in part): Van Kampen 1923:196 & 197; Bourret 1942:360; Dubois, 1987: 42.

Hylorana luctuosa (in part): Deckert 1938:144

Rana (Pulchrana) luctuosa (in part): Dubois 1992:326

Pulchrana luctuosa (in part): Frost et al. 2006:369; Fei et al. 2010:33.

Hylarana luctuosa (in part): Manthey 1983:22; Che et al. 2007: 1–13 (by implication); Das et al. 2007:159, 160, 167, 170; Imbun 2014:99.

Abavorana luctuosa: Oliver et al. 2015:186 (in part); Quah et al. 2017: 272–288 (in part); Zainudin et al. 2017:876–891 (in part); Haas et al. 2018:92, 93, 100, fig. 4b.

Hylarana (Abavorana) luctuosa (in part): Inger et al. 2017:147 & 148.

Syntypes

MNHNP 1889.226–28 from “Bornéo Kina Balu”, Sabah (Borneo), Malaysia. (Guibé 1950) (Fig. 5).

Diagnosis

In addition to its phylogenetic placement (Fig. 1), Rana decorata is reassigned as a member of the genus Abavorana based on the combination of having a robust, medium-sized body; no vocal sacs in males; length of 1st finger greater than 2nd finger; disc width to finger width ratios of finger 3 and toe 4 is 1–1.5; dorsolateral fold indistinct or absent; the colour of the dorsolateral line being white or yellow; the humeral gland in males is prominent, raised and centrally positioned on the ventral surface of the upper arm; a weak or absent rictal ridge; outer metatarsal tubercle weak or absent; skin of dorsum smooth or finely shagreened; throat, abdomen and flanks smooth; posterior section of venter and back of the thigh rugose; no pale colouration on the margins of the tympanum; flanks dark-brown or black below the dorsal fold grading into a pale venter (Inger 1966; Oliver et al. 2015; Quah et al. 2017). Dorsum reddish-orange to rust-brown, encircled by a white or cream coloured dorsolateral line that encircles the snout, canthus rostralis, outer edge of the upper eyelids, and dorsum along the dorsolateral fold to the vent; lower flanks dark-brown or black below the dorsolateral line grading into a paler venter; dorsal colouration of the limbs light-grey or brown with speckling and prominent dark-brown or black bands. Abavorana decorata can be differentiated from its congeners on the basis of its ventral colour pattern which is reticulated in black and white especially on the lower flanks, underside of the limbs (especially hind-limbs) are boldly banded in black and white, and a prominent white streak under the eye and/or tympanum to the corner of the jaw (Figs 5, 6, 9C–D). Adult males with SVL 46.21–56.0 mm, adult females with SVL 29.66–58.42 mm; adult males with SW 9.91–10.71 mm, adult females with SW 6.25–11.68 mm; adult males with IOD 5.59–5.95 mm, adult females with IOD 3.91–6.78 mm; adult males with TD 4.54–4.60 mm, adult females with TD 2.27–5.63 mm; adult males with FL 25.58–26.10 mm, adult females with FL 17.19–31.34 mm (Table 5).

Distribution (Fig. 10). Endemic to Borneo. The species is known from Sabah state, Malaysian Borneo: Mount Kinabalu, Kampung Kiau Taburi, Danum Valley, Tawau Hills Park, Maliau Basin, Tambunan and Penampang (CalPhotos ID: 0000 0000 0912 1159); Sarawak state, Malaysia Borneo: Kubah National Park (Fig. 7C) and Bario (Fig. 6A–B); and West Kalimantan, Indonesia: Melawi Regency (Fig. 7D). The species is expected to be wider ranging on the island.

Figure 7. 

Living specimens of Abavorana decorata and Bornean A. luctuosa; A: Abavorana decorata from Tawau Hills, Sabah, East Malaysia. B: A. decorata from Maliau Basin, Sabah, East Malaysia. C: A. decorata from Kubah National Park, Sarawak, East Malaysia. D: A. decorata from Melawi Regency, West Kalimantan Province, Indonesia. E & F: A. luctuosa from Kubah National Park, Sarawak, East Malaysia. G: A. luctuosa (FMNH 238720) from Sipitang, Sabah, East Malaysia. H: A. luctuosa (FMNH 242883) from Sipitang, Sabah, East Malaysia. (Photographs by Alexander Haas [A, C & E], Paul Bertner [B], Muhammad Aini [D], Chien C. Lee [F], and Robert F. Inger [G & H], Figs G & H ©Field Museum of Natural History. Created by Field Museum of Natural History, Amphibian and Reptile Collection and licensed under CC-BY-SA 4.0).

Abavorana nazgul Quah, Shahrul Anuar, Grismer, Wood, Siti Azizah & Muin 2017

Gunung Jerai Black Stream-Frog Figs 8G–H, 9E–F

Rana luctuosa Sukumaran, J. 2005: 38

Holotype

An adult male (USMHC 2231) collected on 9 April 2016 from Gunung Jerai, Kedah, Peninsular Malaysia (5°47.719’N; 100°26.222’E, 948 m in elevation) by Evan S.H. Quah and David Chan.

Diagnosis

Abavorana nazgul can be differentiated from its congeners by the following combination of characters: adult males 41.51–44.1 mm SVL and single adult female 52.51 mm; adult males with SW 7.98–8.70 mm, single adult female SW 9.59 mm; adult males with IOD 3.94–4.48 mm, single adult female IOD 4.94 mm; adult males with TD 3.52–4.01 mm, single adult female TD 4.10 mm; adult males with FL 22.20–23.22 mm, single adult female FL 27.80 mm; nuptial pads absent in males; humeral glands in males small (2.4–2.98 mm) (Table 5); dorsolateral stripe continuous, orange to yellow in colour; dorsum between dorsolateral stripes black, with or without faint orange or yellow speckles; flanks black, colouration unstratified; distinct cream-colored spots on flanks, dorsal surfaces of limbs, and upper lip; venter grey-brown with prominent light spots on throat and belly, smaller spots on underside of thigh (Fig. 9E–F; Quah et al. 2017).

Distribution

Abavorana nazgul is thus far confirmed from the upper elevations of Gunung Jerai, Kedah, Peninsular Malaysia (Quah et al. 2017). Specimens resembling this species have been photographed at Thale Ban National Park, Satun Province in southern Thailand at approximately 500–600m in elevation around wild boar wallows but are only tentatively identified as A. cf. nazgul pending positive evidence (Fig. 8H).

Figure 8. 

Peninsular Malaysian, Thai and Sumatran Abavorana luctuosa, and A. nazgul; A: A. luctuosa from Sungai Durian Forest Reserve, Kelantan, Peninsular Malaysia. B: A. luctuosa from Janda Baik, Pahang, Peninsular Malaysia. C: A. luctuosa from Semenyih, Selangor, Peninsular Malaysia. D: A. luctuosa from Fraser’s Hill, Pahang, Peninsular Malaysia. E: A. luctuosa from Hala-Bala Wildlife Sanctuary, Thailand. F: A. luctuosa from Sumatra, Indonesia. G: A. nazgul (USMHC 2572) from Gunung Jerai, Kedah, Peninsular Malaysia. H. A. cf. nazgul from Thale Ban National Park, Satun Province, Thailand. (Photographs by Zaharil Dzulkafly [A], Steven Wong [B], Kurt H.P. Guek [C], Evan Quah [D & G], Ian Dugdale [E], Hellen Kurniati [F] & Bunyarit Dechochai [H]).

Figure 9. 

Colour pattern comparison among species of the genus Abavorana; Dorsum (A,C & E) and venter (B,D & F) in life of A. luctuosa (USMHC 2576) (A & B) from Penang Hill, Penang Island, Malaysia, A. decorata from Maliau Basin, Sabah, East Malaysia, Borneo (C & D) and A. nazgul (USMHC 2233) (E & F) from Gunung Jerai, Kedah, West Malaysia. (Photographs by Evan Quah [A, B, E & F] and Stefan T. Hertwig [C & D]).

Figure 10. 

Distribution map of Abavorana decorata and A. luctuosa in Borneo; Known distribution of Abavorana decorata (red) and A. luctuosa (blue) in Borneo. Star = type locality.