New records of Hemiphractus scutatus

We found three specimens of H. scutatus in two of the 11 sampling sites (Figs. 1–3). It was a rare species in the sampling, recorded at a ratio of one specimen in about each 300 days of sampling, while the most abundant syntopic terrestrial frogs were from genera Adenomera Steindachner, 1867, Pristimantis Jiménez de la Espada, 1870, Allobates Zimmermann and Zimmermann, 1988, and Rhinella Fitzinger, 1826, with 2,700 specimens recorded in this same sampling effort. The three specimens of H. scutatus were only recorded by the active searches, and exclusively in Terra Firme forests (Fig. 4).

On 28 September 2012 one female voucher specimen was collected by D. Pavan close to a large tree and under a palm leaf, on the left bank of Tapajós River, at 19:15 h (76.1 mm SVL; 05°02’S, 56°53’W, 62 m above mean sea level, hereafter referred as asl). On 16 October 2012 a male voucher specimen was collected on the same riverbank by LJCL Moraes hidden inside the leaf-litter at 21:05 h, distant ca. 51 km in straight line from the first record (57.8 mm SVL; 04°39’S, 56°37’W, 60 m asl). On 28 April 2013 a second female voucher specimen was collected also hidden inside the leaf-litter on the same riverbank by J. Cassimiro at 21:30 h (61.7 mm SVL; 04°40’S, 56°37’W, 83 m asl), distant ca. 50 km in straight line from the first record and 430 m from the second record. No evidence of reproductive activity or gaping posture (Trueb 1974) was observed.

These three records represent the easternmost known localities of occurrence of H. scutatus, extending the geographic range of the species and the genus Hemiphractus. They are distant ca. 1,000 –1,500 km from the previously known easternmost points of the species occurrence, in Rondônia (INPA-H 15398, 15399) and Amazonas States, Brazil (GBIF 2017; SpeciesLink 2017) (Fig. 1). Considering only the Amazon Basin at South of Amazon River, these new records even extend to the East the geographic range of the family Hemiphractidae. Furthermore, the elevation level in which these specimens were recorded are among the lowest known elevation for the species (60, 62, and 83 m asl; Fig. 5), and two of them (60 and 62 m asl) also extend downwards the known elevational range of this species, since there are no documented records of individuals below 70 m asl.

Morphologic variation and molecular relationships

The morphologic data confirms the identification of our specimens in accordance to the literature (Trueb 1974) and voucher specimens. Qualitative characters include the triangular head, canthus rostralis rounded in section; tympanum large and vertically ovoid; oblique rows of tubercles on dorsal surfaces of forearm and hind limb (less pronounced in female specimens); small triangular fleshy proboscis, dorsoventrally flattened, on tip of snout; eyelids granular with one (female specimens) or three (male specimen) prominent fleshy conical tubercles; single bony projection at the angle of the jaw; slightly enlarged tubercles at the knee and small tubercles at calcaneum (divergent from the absence of calcar projections reported by Trueb 1974 and Rodríguez and Duellman 1994); fingers and toes with vestigial adhesive discs, well-developed round subarticular tubercles and basal webbing; thenar tubercle elliptical and outer palmar tubercle diffuse, flat and cordiform; no evidence of nuptial pads in male specimen; toes also with well-developed round subarticular tubercles and about onefourth webbed; inner metatarsal tubercle well-developed and elliptical, and outer metatarsal tubercle indistinct; shagreened skin on dorsum and granular on flanks, abdomen and ventral surfaces of thighs.

Dorsal coloration in life varies from reddish brown (INPA-H38116 and 38118) to pale tan background with dark mottling (INPA-H38117), with two dark vertebral spots; dark suborbital marks from the lower margin of the eye expanding posteroventrally but not reaching the lip (more pronounced in INPA-H38117 than in INPA-H38116 and 38118) and scattered dark spots in the tympanic region. Ventrally, gular coloration varies from uniformly brown (INPA-H38116 and 38118) to mottled (INPA-H38117), with a pale mid-ventral stripe reaching the pectoral region; same gular color reaches the pectoral region, and becomes less pigmented posteriorly. A finely dark venate pattern covers the flank areas above the forelimb; forelimbs and hind limbs varies from uniformly brown (INPA-H38116 and 38118) to tan (INPA-H38117), with dark transverse bands, reaching the dorsal surface of hands (more evident in INPA-H38117); iris bronze and darker ventrally, with a longitudinally crossing reddish area and pupil horizontal. Regarding quantitative characters, most of the measurements of the middle Tapajós River specimens agree with the known morphometric range of the species (Table 1), also showing the sexual dimorphism in body size. The only divergence is a small HW compared to SVL in female INPA-H38116.

The 16S mtDNA tree for Hemiphractus species shows, as the results presented by Castroviejo-Fisher et al. (2015), two distinct lineages of H. scutatus. The middle Tapajós River population is more related to the lineage from Peru (Figs. 6, 7), as the sequences have a higher genetic similarity (more than 97%) compared to sequence from Colombia, near the country’s border with Brazil (93%) (Fig. 7).

Discussion

The presence of possible cryptic taxa under the name H. scutatus was suggested based on the results of a phylogeny of molecular and morphologic data (Castroviejo-Fisher et al. 2015), since two genetically distant intraspecific lineages were found. Although we initially worked with the hypothesis that specimens from middle Tapajós River were a new taxon, the morphologic and molecular analysis readily rejected this. Regarding the morphology, despite the possibility of strong variation due to large geographic distance to known distribution area, most of the qualitative and quantitative data of the specimens from middle Tapajós River were inside the known range for the species (Trueb 1974; Rodríguez and Duellman 1994) and other voucher specimens (Table 1). The slightly divergences in colors, shapes, and morphometric characters between this specimens and the known for the species may be part of intraspecific variation. Regarding the molecular data, despite the high geographic distance between the populations from middle Tapajós River and Peru (more than 2,300 km), there is a low genetic distance between the sequences from these regions (between 2% and 3%). As Castroviejo-Fisher et al. (2015) highlighted, the genetic distance between the sequences from Colombia and Peru, and now including the distance of Tapajós sequences, are high and may indicate cryptic speciation (more than 7%). As overall similarity in external morphology and pronounced morphologic variation are common events inside the genus Hemiphractus (Trueb 1974), further broader studies and integrative taxonomic revisions may indicate the extent of morphologic and molecular variability of this species and reveal the taxonomic status of these genetically distant lineages.

Biogeography

After more than 190 years since the original description of H. scutatus (Spix 1824) we recorded this species in the eastern Amazonia, emphasizing the lack of knowledge about the general biogeographic patterns of Amazonian amphibians, which can be mainly generated by sampling difficulties, especially in cases of secretive species. Large forested regions in the Amazonia remain unexplored and have the potential to harbor new species or expanding species distributions (Azevedo-Ramos and Galatti 2002). Therefore, the recognition of broader biogeographic patterns to Amazonian amphibians, as areas of endemism historically recognized in the biome to other vertebrates (e.g., Cracraft 1985; Boubli et al. 2014) depends on the continued expansion of the sampling effort and new analitical techniques that is currently being held in the biome.

Our new records for H. scutatus bring new information to a biogeographic idea historically recognized on the low representation of Hemiphractidae in the eastern Amazonia, probably due to increased seasonality in this region (Sombroek 2001; Duellman 2015). Species of this family that have greater environmental plasticity, as appears to be the case of H. scutatus (the species of the genus with the widest known geographic and elevational range) may reach the preserved forests in this region and establish viable populations, although in less abundance in relation to the more climatically constant (Wang et al. 2017) and humid environments of western Amazonia.

Regarding elevational occurrence, although H. scutatus has already been recorded in high elevations at the Andean mountain range (GBIF 2017), a greater number of individuals is known for the Amazonian lowlands, and 67% of 77 published localities of occurrence are below 600 m asl (Fig. 5). This wide elevational range reinforce the high environmental plasticity, as the life-history strategies of amphibian populations in high and lowlands may drastically differ (Morrison and Hero 2003). The knowledge on the drivers of elevational variation in the distribution of Amazonian amphibians is still incipient (Siqueira and Rocha 2013) and the H. scutatus may be a target taxon for future studies testing this gradient.

Conservation

Hemiphractus scutatus is considered as “Least Concern” by IUCN due to its wide distribution and presumably large and stable populations (Coloma et al. 2004). However, this species is rarely recorded and have poorly known population dynamics to define its conservation status, that can even vary along its wide geographic and elevational range. As the Amazon region has suffered increasing anthropic impact through the advance of cities and highways, forests fragmentation and habitat loss (Fearnside 2015), the H. scutatus may have declining populations in most of its distribution, since they are dependents of undisturbed forests (Rodríguez and Duellman 1994).

The specimens of H. scutatus from middle Tapajós River region may represent a unique population, recorded near and within a federal conservation unit (Parque Nacional da Amazônia), same pattern already described for Peruvian populations (von May et al. 2009), reinforcing the need to maintain large protected forest areas in the Amazonia and adequate land-use on the unprotected (Laurance et al. 2001). In addition to these threats, the biome has been target of dam construction in its larger rivers (Latrubesse et al. 2017), which can negatively affect the biodiversity of the surrounding forests (Moraes et al. 2016). The population of H. scutatus from Tapajós River is in the region affected by the construction of a large hydroelectric plant, part of a complex planned for the basin (Fearnside 2015), thus the implementation of this project may affect the viability of this population.