New and noteworthy distributional records of treefrogs ( Anura ) from southwestern Amazonia

We reviewed the herpetological collection of Universidade Federal do Acre and provide here a detailed list of information about new distributional records. We report for the first time in Brazil two anuran species (Dendropsophus joannae and Scinax ictericus) and we extend the known distribution of Dendropsophus marmoratus, Scinax pedromedinae and Teratohyla midas to Acre state. Besides, we provide the second and southernmost record on Scinax iquitorum and new distributional data of Dendropsophus salli from both Acre and Rondônia states.

Some Amazonian species of anurans recently described in adjacent countries (Peru and Bolivia) are expected to occur in Brazil (Köhler and Lötters 2001;Lehr et al. 2007;Moravec et al. 2006Moravec et al. , 2008Moravec et al. , 2009, something that has been confirmed in recent publications Souza 2009, 2010;Brown et al. 2011, López-Rojas et al. 2013). However, scarce information on these species is available and materials housed at museums constitute an interesting source of information. Besides, many regions throughout Amazon where species are described in recent years remain poorly surveyed (De la Riva et al. 2000;Moravec et al. 2009;Jungfer et al. 2010). This is particularly true for small hylids and centrolenids widely distributed, but with a scarce number of records (Moravec et al. 2011;Melo-Sampaio and Oliveira 2013;Twomey et al. 2014).
Increase of our knowledge about amphibian faunas of the western portion of the Amazon has been improved during the last years, but considerable inaccuracies have existed regarding the appropriate names for small species of "yellow" treefrogs stored in collections from the upper Amazon Basin. For example, Duellman (2005: 215) stated that frogs referred to as Dendropsophus leali from the Iquitos region of Peru by Rodríguez and Duellman (1994) are in fact D. riveroi, and those referred by them as D. rossalleni correspond to D. leali (De la Riva and Duellman 1997). In order to help the understanding of the composition of western Amazonian anuran fauna, we investigated museum specimens, performed comparisons, and now provide important new records associated with detailed literature.
We examined the material housed at the Herpetological Collection of Universidade Federal do Acre at the city of Rio Branco (acronym hereafter UFAC-RB). We were able to identify new records for Acre state, and also new records for Brazil based on original descriptions and diagnosis, or relevant literature published for all species mentioned here following Bokermann (1964), Duellman and Wiens (1993), Rodríguez and Duellman (1994), Köhler and Lötters (2001), Duellman (2005), ), Jungfer et al. (2010 and Twomey et al. (2014). For each species, we provided localities (municipality and country names) with geographic coordinates, and number of voucher specimens.
We provide below a detailed list of 19 new records from Brazil (18 in Acre and one in Rondônia state) and one new record from Peru. We provide individual collection numbers and a distribution map emphasizing the closest records and type localities of new species recorded from Brazil ( Figure 1). (Köhler & Lötters, 2001) Figure 2 Rio Branco: Parque Zoobotânico da Universidade Federal do Acre: 12 individuals (UFAC-RB 0275, 0500-0505, 0565, 1994-1996, 1998

A B
Melo-Sampaio and Souza | New records of anurans  Venâncio et al. (2014) in an extensive work on anuran-pond community misidentified this species as Dendropsophus leali at PACM. According Köhler and Lötters (2001), Dendropsophus joannae is most similar to D. leali from which it is distinguished mainly by smaller size, shorter snout, more protuberant eyes, more tuberculate dorsal skin, and a red inner iris in life. Barrio-Amorós and Neira (2004) reported this species to Tambopata, Peru. Based in our records, near the Peruvian border at rio Juruá, we are confident that this species could reach into river Ucayali drainage. The known distribution of Dendropsophus joannae is extended to Acre in ca. 680 km northwest in a straight line from Cobija (Köhler and Lötters 2001). Remarks: This species is easily confused with Dendropsophus acreanus, a slightly smaller species having fine black reticulations on the venter and limbs with yellow concealed surfaces, whereas D. marmoratus has bold black mottling on the venter and limbs with orange concealed surfaces (Duellman 2005). It was illustrated in Souza (2009), plate 19 as Dendropsophus acreanus. Dendropsophus marmoratus differs from D. acreanus in the less projecting eyes, the expanded web between fingers and toes, and the ventral coloration of abdomen (large blacks markings versus no markings with interstices between granules black) and ventral surface hind limbs (usually black versus pale, with a few black blotches). Although Bokermann (1964) stated that there are no localities where both Dendropsophus  Remarks: this species is known to its type locality in Bolivia (Nacebe) and listed at appendix 1 from Acre, municipality of Tarauacá (MZUSP 116707-19) by Orrico et al. (2014). Thus, our records extend the species' known distribution to Espigão do Oeste, Rondônia. According to its original description (Jungfer et al. 2010), this species was previously confused with Dendropsophus bifurcus. Dendropsophus salli is now recorded to Rondônia, 780 km northeast of its type locality in Provincia Ballivian, Departamento Beni, Bolivia. (Duellman & Wiens, 1993) Figure 5 Mâncio Lima: Parque Nacional da Serra do Divisor: Gibraltar: 12 individuals (UFAC-RB 1517-1528, 1811 (3146)(3147)(3149)(3150).

Scinax ictericus
Remarks: According to its original description (Duellman and Wiens 1993), Scinax ictericus is known from its type and additional localities in Peru. This species was assigned as Scinax cruentommus, S. funereus, or S. ruber in UFAC-RB. Scinax ictericus is now known from Gibraltar, Brazil, ca. 395 northwest in a straight line from Balta, Peru, the closest site known between several localities at elevations of less than 300 m in the drainages of the Rio Purus and Rio Madre de Dios in southern Peru where it occurs (Duellman and Wiens 1993). Moravec, Arista, Pérez & Lehr, 2009  (2) snout rounded, not acuminate in dorsal and lateral views; (3) ulnar and tarsal tubercles indistinct (inconspicuous traces of small tubercles can be detected in proximal outer edge of tarsus); (4) tubercles absent on heel; (5) tubercles absent on lower jaw; (6) skin on dorsum slightly to coarsely shagreen; (7) in life, dorsum light olive-green to brown; (8) in life, flanks bright yellow with distinct round black spots becoming gradually larger from axillae to groin; (9) in life, concealed surfaces of legs black; (10) in life, iris gold to bronze with irregular dark reticulation. Our record extends the known distribution 670 km southeast from Santa Luzia, municipality of Cruzeiro do Sul, Brazil (Machado et al. 2015).   one individual (UFAC-RB 6153). Remarks: we fill a gap in the distribution record of Scinax pedromedinae: REAJ, PNSD and ARIE specimens are between Pucallpa area in Peru (Duellman and Wiens 1993) and Porto Alegria, Loreto, Peru (Hoogmoed and Ávila-Pires 2011), extending the known distribution of the species ca. 145 km N and 495 km southwest, respectively. PEC, REPD and RFH specimens are located between Pakitza (Departamento Madre de Dios) and Nacebe (Departamento Pando) the closest localities in Peru, and extends ca. 140 km north of its distribution (Moravec and Aparicio 2004;Hoogmoed and Ávila-Pires 2011).
Remarks: This species was misidentified as Centrolene sp. and illustrated in the plate 8 from Souza (2009). Recently, this species was found near Andes at San Martin and Pongo de Rentema sites in Peru (Twomey et al. 2014) suggesting that it is not restricted to lowland habitats. This record of Teratohyla midas decrease the gap in the distribution 590 km west from Boca do Acre region, Amazonas (França and Venâncio 2010), 520 km northwest from Pakitza, Madre de Dios, Peru (Duellman 2005) and 540 km south from Letícia, Colômbia (Lynch 2005).
The new distributional records reveal the importance of surveys in fragmented areas, since that many species were collected in this habitat, including the open forests with bamboo, a phytophysiognomy poorly studied concerning herpetofauna. Souza (2009) related more than 100 species throughout western portion of Acre state into two large conservation units (PNSD and REAJ) with different degrees of human occupation (integral protection in the former, and sustainable use in the last one). Major taxonomic changes were followed Many species mentioned here are restricted to western portion of Amazon and Andes slopes are expected to occur in Brazil. Further studies on herpetofauna of open forests with bamboo are needed. Currently, either some candidate species remains unnamed, whereas named species wait synonymization or revalidation in light of integrative taxonomy (Ortega-Andrade and Venegas 2014). The genera Scinax and Dendropsophus still are challenging and we hope that herpetologists pay attention with small treefrogs, and detail every species in life including acoustic and behavioral characters (natural history notes) before assigned a name to checklist, because they are abundant and very common, but its taxonomy is very hard, mainly to fix after preservation leading to misidentification and unwarranted descriptions.