A new species of Pristimantis (Anura: Strabomantidae) from white-sand forests of central Amazonia, Brazil

The white-sand ecosystems in the Solimões-Negro Interfluve are among the less studied in Amazonia. Recent herpetological surveys conducted west of Manaus, Brazil (central Amazonia) indicate that white-sand forests host a unique anuran fauna comprising habitat specialized and endemic species. In the present study we describe a new species of rain frog belonging to the Pristimantis unistrigatus species group from the white-sand forest locally called “campinarana” (thin-trunked forests with canopy height below 20 m). The new species is phylogenetically close to rain frogs from western Amazonian lowlands (P. delius, P. librarius, P. matidiktyo and P. ockendeni). It differs from its closest relatives mainly by its size (male SVL of 17.3–20.1 mm, n = 16; female SVL of 23.2–26.5 mm, n = 6), presence of tympanum, tarsal tubercles and dentigerous processes of vomers, its translucent groin without bright colored blotches or marks, and by its advertisement call (composed of 5–10 notes, call duration of 550–1,061 ms, dominant frequency of 3,295–3,919 Hz). Like other anuran species recently discovered in the white-sand forests west of Manaus, the new species seems to be restricted to this peculiar ecosystem.


INTRODUCTION
The genus Pristimantis Jiménez de la Espada, 1870 is the most species-rich among vertebrates. Nevertheless, its total species diversity remains highly underestimated as suggested by the numerous new species described each year (e.g., De Oliveira et al., 2020;Ron et al., 2020;Gete, Buckely & Ron, 2021;Fouquet et al., 2022a;Ortega, Brito & Ron, 2022;Roberto et al., 2022). Since 2020 around 60 species of Pristimantis were described (Frost, 2023). The Andes is the most species-rich region for Pristimantis (Pinto-Sanchez et al., 2012), hosting more than 60% of the nominal species of the genus (Meza-Joya & Torres, 2015). Many other species are distributed throughout Amazonia, and the rest occurs in the Cerrado, Atlantic Forest, Pantepui and in Trans-Andean forest areas (Frost, 2023).
This species-richness disparity is at least partially explained by the evolutionary history of the genus that started to diversify in the Andes and dispersed many times to Amazonian lowlands (Mendoza et al., 2015) with subsequent dispersals back to the Andes (Fouquet et al., 2022b). However, such disparity is also a consequence of the large amount of poorly sampled areas in Amazonia, where many undescribed species of Pristimantis occur (e.g., Vacher et al., 2020;Fouquet et al., 2022a). Finally, the low number of taxonomists working in lowland Amazonia also contributes to this knowledge gap (Melo-Sampaio, Ferrão & Moraes, 2021).
Anuran sampling in Amazonia is concentrated near urban areas, navigable rivers, roads and highways (Jenkins et al., 2015;see map in Vacher et al., 2020). This is especially true for the Negro-Solimões Interfluve (NSI), a region with the largest amount of WSE (see Adeney et al., 2016) and only a few documented anuran assemblages (Neckel-Oliveira & Gordo, 2004;Menin et al., 2017;Moraes et al., 2022a;Moraes et al., 2022b). Even regions geographically close to the largest city in Brazilian Amazonia (Manaus) and regions easy to access remain poorly sampled and studied. However, this picture is changing. The Reserva do Desenvolvimento Sustentável do Rio Negro (RDS Rio Negro)-a NSI reserve covered mainly by a mosaic of dense forests and WSE patches lying ca. 100 km west of Manaus-has recently become a research center for biodiversity studies focussing on this ecosystem. The first attempt to document the anuran communities of the RDS Rio Negro has notably resulted in the rediscovery of an overlooked spiny-backed treefrog (Ferrão et al., 2019) and the description of a new snouted treefrog (Ferrão et al., 2022).
Herpetological surveys conducted in the RDS Rio Negro and nearby WSE patches in 2018 and 2020 resulted in the discovery of an unknown Pristimantis species associated with campinarana. The external morphology and the advertisement call of this species indicated that it represents an unnamed species and preliminary molecular comparisons confirmed this assumption. Herein, we use an integrative taxonomy approach and describe this new species of Pristimantis as well as its phylogenetic position, geographic distribution and natural history.
According to BLAST, sequences of the new species were highly similar to species currently assigned to the Pristimantis unistrigatus species group. To infer the phylogenetic relationships among the new species and its close relatives, newly generated sequences were inserted into a data set retrieved from GenBank containing selected homologous sequences (Appendix 1). Sequence selection in GenBank was focused on specimens of the P. unistrigatus species group from the Andes, Pantepui and Amazonian lowlands. Additionally, sequences of two species of the genus Oreobates were retrieved to root the tree. In total, 265 sequences (16S = 136; COI = 70; RAG1 = 59) corresponding to 137 specimens were selected. We aligned sequences of each gene using MAFFT online server (https://mafft.cbrc.jp/alignment/server/) with default parameters, except by the use of the E-INS-i strategy for the 16S and G-INS-i strategy for protein-coding genes (Katoh & Standley, 2013). The final matrix was concatenated in Mesquite (Maddison & Maddison, 2021) and composed of 137 terminals with 1,827 bp (16S = 561 pb; COI = 636 pb; RAG1 = 630 pb).
Best-fit evolutionary models and partition schemes were determined through ModelFinder (Kalyaanamoorthy et al., 2017) using seven initial partitions: one for the 16S and one for each codon of protein-coding genes. The best evolutionary models for partitions in the concatenated matrix were: TIM2 + F + R5 for 16S, GTR + F + I +G4 for COI 1st and RAG 3rd codons, HKY + F +G4 for COI 2nd, RAG 1st and 2nd codons, and GTR + F +ASC + G4 for COI 3rd position. Phylogenetic relationships were reconstructed using Maximum Likelihood inference (ML). The ML tree was inferred with IQTREE (Nguyen et al., 2015) as implemented in the webserver (http://iqtree.cibiv.univie.ac.at; Trifinopoulos et al., 2016). Clade support was estimated with 10,000 ultrafast bootstrap replications (Hoang et al., 2018), 1,000 maximum iterations, and a minimum correlation coefficient of 0.99. We calculated pairwise genetic distances (p-distance and Kimura-two-parameter distance; Kimura, 1980) among the populations of new species and close relatives using MEGA 11 (Tamura, Stecher & Kumar, 2021). Genetic distances were calculated using pairwise deletion.

Morphology
Twenty five morphometric measurements were taken from 16 adult males and six adult females of the new species following Duellman & Lehr (2009) (eye diameter-ED, eyenostril distance-EN, foot length-FL, interorbital distance-IOD, internarial distance-IND, head length-HL, head width-HW, snout-vent length-SVL, tibia length-TL, and tympanum diameter-TD), Caldwell, Lima & Keller (2002) (forearm length-FAL, hand length-HAND, snout length-SL, disc width of Finger III-WFD), Heyer et al. (1990) (tarsus length-TAL, thigh length-THL, upper arm length-UAL), Lima, Sanchez & Souza (2007) (hand length from proximal edge of palmar tubercle to tip of Finger I-HANDI, Finger II-HANDII, and Finger IV-HANDIV) and Mônico et al. (2022) (foot length from proximal edge of outer metatarsal tubercle to tip of Toe I-FLI, Toe II-FLII, Toe III-FLIII and Toe V-FLV, and disc width of Toe IV-WTD). Measurements were taken to the nearest 0.01 mm using a Leica stereomicroscope (model S8APO) coupled to a Leica DFC295 camera, except for SVL, measured to the nearest 0.01 mm with a digital caliper. Raw data are provided in Table S1.
Format of the description and terminology of morphological characters follow Kok & Kalamandeen (2008), Duellman &Lehr (2009), andKok, Means &Bossuyt (2011). Color in life was described based on photographs taken in the field, following the color catalog provided by Köhler (2012).

Bioacoustics
Bioacoustic variables were analyzed with Raven Pro 1.6 (Bioacoustics Research Program, 2014) with the following configuration: window = Blackman, Discrete Fourier Transform = 2,048 samples and 3dB filter bandwidth = 80.0 Hz. The following temporal and spectral traits were measured: call duration-CD, number of notes per call-NN, note duration-ND, inter-note interval-SBN, and minimum-LF, maximum-HF and dominant frequency-DF. Inter-call interval was not measured because it is usually affected by microclimatic conditions at the time of recording (i.e., on rainy days, males call more often in a short period of time than on days without rain) and poorly informative. Dominant frequency was measured using the Peak frequency function; maximum and minimum frequencies were measured 20dB below the peak frequency to avoid background noise interference. Call description follows the call centered approach of Köhler et al. (2017). Spectrogram and oscillogram were generated in R environment (R Core Team, 2019) through the 'seewave' package 2.0.5 (Sueur, Aubin & Simonis, 2008) using a Hanning window, 256 points of resolution (Fast Fourier Transform) and an overlap of 85%. Bioacoustic raw data are provided in Tables S2.

Nomenclatural acts
The electronic version of this article in Portable Document Format (PDF) will represent a published work according to the International Commission on Zoological Nomenclature (ICZN), hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is: urn:lsid:zoobank.org:pub:F8ED54F6-9E18-49BD-8178-659BDFB79C65. The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central and CLOCKSS.

Description of holotype
INPA-H 44426 (field number APL 23164). Morphometric measurements are presented in Table 1. An adult male (Fig. 3), SVL 19.1 mm; head slightly longer than wide (HL 103% of HW); head width 35.9% of SVL; head length 36.9% of SVL; cranial crest absent. Snout moderately long(SL 131% and 106% of EN and ED, respectively), subacuminate in dorsal view (Fig. 3A) and moderately truncated in lateral view (Fig. 3C); nostril ovoid, slightly protuberant, directed dorsolaterally; IND 87.3% of IOD; internarial region almost straight; canthus rostralis almost straight in dorsal view, slightly angular in profile; loreal region concave; lips not flared; one small tubercle on upper eyelid; interorbital region straight, IOD 32.1% of HW; eyes large (ED/TD = 2.7), pupil horizontally elliptical; supratympanic fold slightly distinct, extending from posterior margin of eyelid angling posteroventrally behind tympanic annulus; tympanum visible and rounded, TD 36.5% of ED; tympanic membrane poorly prominent, directed laterally; tympanic annulus poorly distinct, obscured dorsally by the supratympanic fold; one small postrictal tubercle, poorly visible; choanae of moderate sized, rounded to ovoid, not concealed by palatal sheath of maxilla; dentigerous processes of vomers present, with two ill-defined teeth, small, oblique and positioned posterior to level of choanae; tongue ovoid, longer than wide; short vocal slits, located in posterior half of mouth floor between tongue and margin of jaw; vocal sac small, simple and subgular.
Forearm shorter than hand (FAL 81.2% of HAND), notched posteriorly, nearly 78% free; three small ulnar tubercles ill-defined and aligned, almost indistinct after preservation; relative length of fingers I <II <IV <III (Fig. 3D); discs small and rounded on fingers I and II, expanded on fingers III and IV, with circumferential grooves; thenar tubercle poorly distinct, ovoid; palmar tubercle distinct, bifid; subarticular tubercles ill-defined, most prominent on fingers III and IV, rounded in dorsal and lateral view; small supernumerary tubercles present, but poorly visible; ventral pads well-defined on fingers III and IV.
Hindlimbs slender; tibia length 49% of SVL; heel without tubercles; tarsus with a row of small, poorly defined tubercles; tarsal fold absent; foot length 41% of SVL; relative length of toes I <II <III <V <IV (Toe III reaches the second subarticular tubercle of the Toe IV; Fig. 3E); toes with lateral fringes, more developed on toes III-V, webbing basal between toes IV-V; discs small and rounded on toes I and II, expanded on toes III-V; inner metatarsal tubercle large and ovoid, more than two times the size of ovoid outer metatarsal tubercle; subarticular tubercles large, protuberant, single, round on toes I-III and elliptical on toes IV and V; small supernumerary tubercles more visible on toes II-IV; ventral pads well-defined on toes III-V. Dorsal skin shagreen (Fig. 3A), with longitudinal stripes and an irregular dorsolateral fold composed of spaced tubercles; small tubercles on scapular region; upper eyelid shagreen, with small tubercles; skin on flanks and chest smooth; skin on belly slightly areolate; upper and posterior surfaces of hindlimbs smooth, with small flat tubercles on thigh; dorsolateral folds absent; pectoral and discoidal folds not visible; cloaca protuberant, cloacal region without tubercles.
In life, dorsum yellow ocher (color 14 by Köhler, 2012) with dark brown mid-dorsal stripe running from snout to cloaca. Canthal stripe present, black, running from the tip of snout to anterior margin of eyelid. Dorsolateral stripe present, irregular, formed by a series of small, dark brown dots and blotches, running from posterior eye to cloacal region (Fig. 4A). Upper lip with dark brown subocular and dark brown supratympanic bar. Dark brown transversal bars on forearms; light brown transversal bars on thigh and tibia. Posterior surfaces of thigs uniformly brown. Groin white, translucent, with sparse brown melanophores. Throat, chest, belly and ventral surfaces of legs white, translucent, densely covered by tiny brown melanophores (Fig. 4B). Iris pale bronze with dark brown In alcohol, color pattern faded (Fig. 3), upper lip with five dark bars (Fig. 3C), bars on dorsal surfaces of forearms, thighs and tibias dark brown; venter whitish with dense dark brown melanophores.

Variation in the type series
SVL ranges from 17.3 to 20.1 mm in males (n = 16) and from 23.2 to 26.5 mm in females (n = 6) ( Table 1). Canthus rostralis almost straight in dorsal view (e.g., Fig. 3B, holotype) to slightly curved in some individuals (e.g., Figs. 5E and 5F). Three to four ulnar tubercles are present in males (barely visible in 63% of them), ulnar tubercles absent in females. Dorsal skin texture varies from shagreen to granular, small tubercles are present or absent (character of skin texture probably depends on individual activity at the moment of sampling; see Guayasamin et al., 2015;Kok et al., 2018).
In preservative, three types of basic dorsal color patterns can be detected in the type series of Pristimantis campinarana sp. nov.: dorsum with irregular dark brown markings (68%; Figs. 5A, 5D and 5E), dorsal coloration sharply outlined against the flanks (9%; Fig. 5B), dorsum with dark brown stripes as in the holotype (23%; Figs. 5C and 5F). In addition, a dark brown interorbital bar is present in 73% of the specimens (Figs. 5A, 5D and 5E); a dark brown W-shaped mark is presented on the scapular region of 18% of the individuals, in some of them more conspicuous (Figs. 5A and 5D) than in others (Figs. 5D and 5E). Dark brown bars and blotches are present on the upper lip of all type specimens, dark and distinct in 73% of the specimens, faded or less conspicuous in the others. Obvious transversal dark brown bars are present on the arms and hands of 63% of the individuals, less conspicuous or absent in 23% and 14% of the specimens, respectively. Distinct transversal dark bars are present on the thigh and tibia of 32% of the specimens (Figs. 5C and 5F), poorly conspicuous in 54% (Fig. 5A) and absent in 14% of the individuals (Fig. 5B). Dark supratympanic stripe present in all specimens. Ventral surface is whitish cream to yellowish white, with a moderate amount of melanophores in

Etymology
The specific epithet 'campinarana' is used as a noun in apposition and refers to the word in Portuguese that defines the type of forest that the new species occupies: the white-sand forest campinarana.

Distribution, natural history and conservation
Currently, Pristimantis campinarana sp. nov. is known only from primary and slightly anthropized forests at two localities in the municipality of Iranduba, state of Amazonas, Brazil (Fig. 9). All individuals were recorded in WSE characterized as campinarana (a forest with canopy height below 20 m; Fig. 10A), where the species is locally abundant. However, none was found in campina, a white-sand forest with canopy height below 10 m and large patches of exposed white sandy soils. Pristimantis campinarana sp. nov. is a crepuscular and nocturnal species, with peak activity at crepuscule. Its breeding activity takes place in the rainy season (November to February). In the dry season, we found males sheltering among the leaves of terrestrial bromeliads of the genus Guzmania (Fig. 10B). In the rainy season, males start calling at dusk (∼18:00 h) and are very active until ∼20:00 h. Then their activity decrease and end usually around 22:00 h. In rainy days, the vocalization is sustained throughout the night. Males were observed calling perched on the vegetation (Fig. 10C) usually 1 m above the ground, rarely above 3 m. Calling males aggregate in groups of up to ten individuals, separated from each other by ∼4-5 m, but it is not uncommon to observe smaller groups of three to four calling individuals that are spatially more spaced. The amplexus (n = 3) is axillary (Fig. 10D). Clutches were not found in situ, but females produce about 14-17 large oocytes (n = 4). The new species occurs in sympatry with a candidate species closely related to Pristimantis orcus (AT Mônico, 2022, unpublished data). We do not have sufficient data to categorize the new species following the criteria of the International Union for Conservation of Nature (IUCN), it should be thus considered Data Deficient (DD).

Notes.
*It includes the holotype's measurements.
herpetological surveys were conducted in dense ombrophilous forests east of Negro River and in the Purus-Madeira Interfluve, the new species was never found there. Therefore, we assume that the new species is endemic of the Negro-Solimões interfluve and only found in WSE. This distributional pattern is possibly similar to the ones of other anuran species (Trachycephalus venezolanus [Mertens, 1950], Osteocephalus vilarsi [Melin, 1941] and Scinax albertinae Ferrão et al., 2022) recently discovered from the Negro-Solimões interfluve (Ferrão et al., 2019;Ferrão et al., 2022). Moreover, additional undescribed species (e.g., species of Adenomera; M Ferrão, 2022, unpublished data; Phyllomedusa AP Lima, 2022, unpublished data; Pristimantis aff. orcus, AT Mônico, 2022, unpublished data) are also found associated to WSE, thus totaling at least seven frog species sharing this habitat specialization and distribution pattern endemic to the Negro-Solimões interfluve. In fact, the Jaú region was recently defined as an additional independent area of endemism in Amazonia due to the coocurrence of six bird species (Borges & Da Silva, 2012). Our finding, thus, strengthen the idea that white-sand ecosystems of the Negro-Solimões interfluve harbors a unique biodiversity that deserves effective protection. Pristimantis campinarana is nested within a clade formed by species otherwise restricted to western Amazonia lowlands: P. matidiktyo, P. ockendeni, P. delius and P. librarius. This nested position of P. campinarana suggests that the speciation occurred after a dispersal from the west. A possible scenario involves historical changes in the Amazon basin drainage (Hoorn et al., 2010;Albert, Val & Hoorn, 2018) such as the disappearance of a riverine barrier that connected the Japurá River to Negro River (Ruokolainen et al., 2019), in the Jaú region, that could have favored the dispersal P. campinarana ancestors. Relatively recent eastward dispersals have been reported for other anuran groups, from small leaflitter toads (i.e., Allobates caeruleodactylus and A. trilineatus clades, Réjaud et al., 2020;Ameerega, Roberts et al., 2006) to arboreal treefrogs(i.e., Osteocephalus taurinus and O. buckleyi groups, Ortiz et al., 2022). Alternative scenarios involve dispersal across rivers (Smith et al., 2014;Moraes et al., 2016;Pirani et al., 2019). Both hypotheses, however, involve subsequent isolation by river and eventually speciation (Wallace, 1852;Haffer, 1974;e.g., Ribas et al., 2012;Rojas et al., 2018).
Despite large genetic divergence and consistent diagnosis characters, the species forming this clade remain very similar in morphology (e.g., SVL of males, ventral skin texture and iris in life) illustrating the trend in the genus of highly conserved morphology and the challenge that describing its diversity Waddell et al., 2018;De Oliveira et al., 2020). This phenotypic conservatism have been discussed for other Amazonian frogs (Gonzalez-Voyer et al., 2011;Kaefer et al., 2013;Guayasamin et al., 2015;De Oliveira et al., 2020), as well as its implications for taxonomy (Funk et al., 2007;Kaefer et al., 2013;Waddell et al., 2018).
New species are being described recurrently in Amazonia, even from areas close to large urban and research centers. For example, in the last years, several new species were described from the Reserva Florestal Adolpho Ducke (Manaus, Brazil), an intensively studied area (e.g., Amazophrynella manaos Rojas-Zamora et al., 2014;Atelopus manauensis Jorge, Ferrão & Lima, 2020;Synapturanus ajuricaba Fouquet et al., 2021; and Pristimantis guianensis Mônico et al., 2022). The lowlands in the Negro-Solimões interfluve are also close to Manaus, and also harbor newly described species notably the discovery of a new lizard genus (i.e., Marinussaurus Peloso et al., 2011). This illustrates how far we are from understanding the species diversity in Amazonia particularly for small and cryptic species like the P. unistrigatus group (Fouquet et al., 2022a;Mônico et al., 2022).

CONCLUSIONS
Using morphology, bioacoustics and molecular data from three markers (16S, COI and RAG-1), we described a novel species of rain frog (genus Pristimantis) from an unexplored environment in the Amazonia: the white sand ecosystems. Description of Pristimantis campinarana sp. nov. reaffirms that the species richness of the west Amazonian anurans remains considerably underestimated even near to the largest and dynamically developing Amazonian metropole-Manaus. The congruence of seven frog species sharing a WSE specialization and distribution pattern likely endemic to the Negro-Solimões interfluve seem to reinforce the area of endemism proposed by Borges & Da Silva (2012) in the Jaú region. Also, the vertebrate fauna of white sand forests is unique and recent studies have revealed notable discoveries, reinforcing the need for effective protection of these environments.

APPENDIX 1
Appendix 1 Specimens of Pristimantis and Oreobates used in phylogenetic analyses, with respective voucher, GenBank acession number and references.

Species
Voucher • Miquéias Ferrão conceived and designed the experiments, performed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft.
• Jiří Moravec conceived and designed the experiments, performed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft.
• Antoine Fouquet conceived and designed the experiments, performed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft.
• Albertina P. Lima conceived and designed the experiments, performed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft.

Animal Ethics
The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers): Protocols of collection and animal care follow the Brazilian Federal Council for Biology (Resolution number 148/2012) and the Ethics Committee on the Use of Animals of the Instituto Nacional de Pesquisas da Amazônia-CEUA-INPA (Process n • 35/2020, SEI 01280.001134/2020-63).

Data Availability
The following information was supplied regarding data availability: The raw morphometric and bioacoustic data of adults of the new species are available in the Supplemental Files.

Supplemental Information
Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.15399#supplemental-information.