We collected 11 specimens during fieldwork in Kranji Marshes, Singapore (1°25.1150’N, 103°43.2641’E; Fig. 1A–C) between January and October 2021 (surveys in the CCNR failed to yield any specimens). The site is dominated by marshy young secondary scrubland (Fig. 1C). Historically, the area was a coconut plantation bounded by mangroves along the Kranji River. With the damming of the river, the surrounding mangroves were converted into freshwater marshes and there is a conservation site approximately 200 m northeast of the site sampled. The site is highly disturbed and is littered with trash and construction rubble (Fig. 1C). Several individuals were found throughout the site calling after rain. We collected and dissected the specimens to extract liver tissue, which was preserved in 100% ethanol. Specimens were fixed in 10% formalin and subsequently preserved in 70% ethanol. All voucher specimens collected in this study are deposited in the Zoological Reference Collection (ZRC) at the Lee Kong Chian Natural History Museum, National University of Singapore. Advertisement calls were recorded using an Olympus LS-100 recorder and a Sennheiser ME-66 microphone.

We took the following measurements of collected specimens using Mitutoyo digital calipers to the nearest 0.01 mm: SVL (snout-vent length; tip of snout to cloaca), HL (head length; tip of snout to posterior margin of mandible), HW (head width; measured across mandibular articulations), HD (head depth; measured at the occiput), UEW (upper eye-lid width; widest distance measured from upper margin of orbit to medial margin of upper eye lid), EL (eye length; measured across anterior and posterior corners of orbit), IND (internarial distance; distance across medial margins of nostrils), IOD (inter-orbital distance; smallest distance between medial margins of upper eye lid), SL (snout length; tip of snout to anterior-most point of orbit), N-EL (nostril-eye length; posterior margin of nostril to anterior-most pint of orbit), S-NL (snout-nostril length, anterior margin of nostril to tip of snout), FAL (forearm length; base of outer palmar tubercle to posterior margin of elbow inflection), HAL (hand length; base of outer palmar tubercle to tip of third finger), THL (thigh length; cloaca to outer margin of knee inflection), AGL (axilla-groin length; distance between arm and leg insertions), TFL (tibiofibula length; measured across the outer margins of knee and ankle inflections), and FL (foot length; outer margin of ankle inflection to tip of fourth toe).

A Levene’s test was performed on each dependant variable (character) to test for equal variance. However, the p-values of the Levene’s tests were >0.05 for each dependant variable. Therefore, t-tests were performed to determine whether the Singapore specimens were statistically different from Micryletta inornata s.s. collected from Sumatra following the framework outlined by Chan and Grismer (2021). A principal component analysis was also performed to find the best low-dimensional representation of variation in the data to determine whether morphological variation could form detectable group structure. Measurements for Micryletta inornata s.s. were obtained from Alhadi et al. (2019). Measurements (except for SVL) were corrected for body-size variation using an allometric growth model implemented in the R package GroupStruct (Chan and Grismer 2022) and all morphological analyses were performed on the size-corrected data. All morphological analyses were performed in R v.3.6.1 (R Core Team, 2014).

Advertisement call recordings were cleaned of background noise using the Noise Reduction tool in Audacity version 2.3.3 (Audacity, GNU General Public License). Noise reduction settings are as follow: Noise Reduction (dB) = 40; Sensitivity=20; Frequency Smoothing (bands) = 3. A total of 12 calls (n=12) from a single individual was analyzed. All calls were recorded at an ambient temperature range of 26–27 ºC. We generated oscillograms and spectrograms using the R package of SEEWAVE (Sueur et al. 2008).

We amplified a ~900 bp fragment of the 16S rRNA mitochondrial gene for four specimens from Singapore (ZRC 1.13323, ZRC 1.13369, ZRC 1.13370, ZRC 1.13389) using the primers 16SD-L (5′-GTRGGCCTAAAAGCAGCCAC-3′), and 16SD- H (5′-CTCCGGTCTGAACTCAGATGACGTAG- 3′) (Evans et al. 2003, Chan et al. 2022). Amplification was done using the following PCR thermal profile: 95 °C for 4 min, followed by 35 cycles of 95 °C for 30 s, 52 °C for 30 s, 72 °C for 70 s, and a final extension phase at 72 °C for 7 min.

Sequences were assembled and aligned (MUSCLE algorithm) using Geneious v5.6.7 (Kearse et al. 2012). New sequences generated for this study are cataloged under the GenBank accession numbers ON026063–66. Additional published sequences of other Micryletta species were obtained from GenBank to supplement the phylogenetic analysis. A maximum likelihood phylogeny was inferred using IQTREE 2 (Minh et al. 2020). The best-fit substitution model was determined using ModelFinder (Kalyaanamoorthy et al. 2017) and branch support was assessed using 1000 ultrafast bootstrap replicates.

We also performed two species delimitation analyses using the programs mPTP v.0.2.4 (Kapli et al. 2017) and ASAP (Puillandre et al. 2021) to determine if the Singapore population could represent a distinct species. For the mPTP analysis, the mitochondrial phylogeny was used as the input tree and the minimum branch length was calculated using the –minbr_auto function. The confidence of delimitation schemes was assessed using two independent MCMC chains at 10,000,000 generations each. The ASAP analysis was performed using the Kimura K80 substitution model (ts/tv=2.0) through the web-server accessible at https://bioinfo.mnhn.fr/abi/public/asap. Uncorrected p-distances were calculated in MEGA-X (Kumar et al. 2018) using the Complete Deletion option for missing data.

Raw measurements and summary statistics of all 11 specimens collected in this study are presented in Table 1. The t-test revealed that male specimens from Singapore differed significantly (p < 0.05) from male specimens of Micryletta inornata s.s. from Sumatra in several head-related characters (head width, head length, snout length, nostril-eye length, snout-nostril length, upper eye-lid width) and also tibiofibula length. Of these differences, head width, head length, snout length, snout-nostril length, and upper eye-lid diameter were found to be highly significant (p < 0.01; Table 2). Female and juvenile specimens were not assessed due to limited sample size. The PCA analysis showed that the Singapore population is distinctly separated from the M. inornata s.s. along PC1, which accounted for 53% of the variation (Fig. 2A).

Comparisons of uncorrected p-distances revealed that M. cf. inornata from Singapore is 3.5–4.2% divergent from M. inornata s.s., which is comparable to divergences between M. steinegeri and M. immaculata (2.9–4.5%) and between M. hekouensis and M. immaculata (2.6–4.5%) but higher than divergences between M. erythropoda and M. lineata (2.3–2.9%) and between M. steinegeri and M. hekouensis (1.6–1.9%; Fig. 2B). The Singapore population was reciprocally monophyletic with Micryletta inornata s.s. from Indonesia with strong support (bootstrap=100; Fig. 3). Similar to Miller et al. (2021), sequences from southern Myanmar were recovered within the Micryletta inornata s.s. clade but with weak support. The relationships among other species of Micryletta received mixed support and were different in several regards from previous phylogenies (Liu et al. 2021, Poyarkov et al. 2021, Yang and Poyarkov 2021).

The mPTP species delimitation analysis inferred the Singapore population as a distinct species with strong support (average support value = 99; Fig. 4). Micryletta inornata s.s., M. sumatrana, M. dissimulans, M. nigromaculata, M. aishani, M. erythropoda, and M. lineata were also corroborated as distinct species with strong support. However, M. hekouensis, M. steinegeri, and M. immaculata were lumped as a single species indicating that the distinction between these species is weak. Similarly, the ASAP analysis also inferred the Singapore population as a distinct species across all partitions. The top-ranked partition further split M. inornata s.s. and M. immaculata, while lumping M. steinegeri and M. hekouensis (Fig. 4).

The advertisement call of Micryletta cf. inornata from Singapore is a pulsatile sweeping metallic call that comprises of an average of 21 pulses (mean=21; range=16–28; n=12; Fig. 5). Calls had a mean dominant frequency of 5333.05 Hz (n=12), with a duration of 0.41 s (mean=0.41, range=0.34–0.47, SD=0.04, n=12).