Herpetofauna recorded by a fauna rescue program in a Caatinga area of João Câmara , Rio Grande do Norte , Brazil

We present an annotated list of herpetofauna for the municipality of João Câmara, Rio Grande do Norte State (RN), Brazil, in Caatinga domain. The data were obtained through fauna rescue program activities performed during the installation of two wind energy parks. We recorded 10 species of lizards, eight snakes, four amphisbaenids and three anurans. The results obtained present extensions to known geographic distribution for four species (Amphisbaena littoralis, Enyalius bibronii, Epictia borapeliotes and Corythomantis greeningi) and add information to the natural history of the first two species. This constitutes the second list of herpetofauna for RN. Our results show that such studies may contribute useful data on herpetofauna, especially in poorly studied areas.

Despite advances in the knowledge of the Caatinga biota, the fragmentation of the biome by human activities continues and many localities remain unsampled (Castelleti et al. 2003, Rodrigues 2003).In RN there are 5 lists for reptiles and amphibians, 1 for amphibians, 1 for squamates reptiles, 2 for lizards and 1 for general herpetofauna (Freire 1996, Delfim and Freire 2007, Freire et al. 2009, Magalhães et al. 2013, Caldas et al. 2016).
According to National Biodiversity Policy knowledge of biological diversity in areas for occupancy and management is the responsibility of both public and private institutions.Thus, the aim herein is to present a list of herpetofauna recorded during a fauna rescue program performed during the installation of 2 wind energy parks at João Câmara, RN, Brazil.This list aims to fill gaps in the knowledge of the herpetofauna in a poorly known region of the Caatinga domain and intends to assist proposals of conservation of local biodiversity.

Methods
Study site.Data collection was part of the fauna rescue program conducted prior to the construction of wind energy parks Model I and Model II (05.427° S, 035.911° W, 194 m elev., datum SIRGAS2000).The project's area is 1729.7 ha, located in the rural area of João Câmara, RN (Fig. 1).Installation of the wind energy parks required removal of about 52 ha of vegetation, or 2.99% of the total area.
The study site is located in the "Depressão Sertaneja Setentrional", 1 of 8 ecoregions of Caatinga, which has a hot and semi-arid climate with annual average precipitation of 500 to 800 mm (Velloso et al. 2002).The veg-   comm.).Specimens were captured by hand, identified, photographed, and transported out of the immediate area.Captured animals whose welfare was in doubt or that could not be identified in the field were euthanized in accordance with resolution no.301 of December 8/2012 (CFBIO 2012).Subsequently, these specimens were fixed in 10% formol, preserved in 70% alcohol and sent to the Federal University of Rio Grande do Norte, Biosciences Center, Department of Botany, Ecology and Zoology.Specimens that were observed but not captured were not included in the study because of the possibility of incorrect identification.Activities were performed in accordance with permits issued by the Instituto de Desenvolvimento Sustentável e Meio Ambiente do Rio Grande do Norte (IDEMA) numbers 2012-056408/ TEC/Sveg-0167 and 2012-056409/TEC/Sveg-0168. The nomenclature for reptiles and amphibians followed Uetz et al. (2016) and Frost (2017), respectively.Data analysis.The abundance and species richness data were plotted in a rarefaction curve in order to determine the sampling efficiency, using Chao2P species estimator (Gotelli and Colwell 2001).The rarefaction curve was generated using the statistical program EstimateS 9.1.0(Colwell 2013).

Results
The rescue activities resulted in 115 recorded individuals belonging to 25 species (Table 1; Figs 2-23), of which 3 are Anura (3 individuals) and 22 are Squamata (112 individuals).The anurans were Rhinella jimi (Stevaux, 2002), Corythomantis greeningi Boulenger, 1896, andPhysalaemus cicada Bokermann, 1966.The Squamata, consisted of 68 lizards in 10 species from 8 families and 17 snakes in 8 species from 4 families.We recorded 27 amphisbaenids in 4 species from 1 family.The list of species is shown in Table 1.The rarefaction curve (Fig. 24) did not reach the asymptote, indicating that other species are expected for the area.Species accounts with natural history, diagnosis and notes on distribution are provided below.

Amphibians
Family Bufonidae Gray, 1825 Rhinella jimi Stevaux, 2002: Figure 2 Bufo jimi Stevaux 2002: 19-Kwet et al. 2006: 96;Freire et al. 2009;Caldas et al. 2016: 1929. Chaunnus jimi-Frost et al., 2006: 364.Before the description of R. jimi, the populations of the Brazilian northeast were identified as Bufo paracnemis (= R. schneideri).Stevaux (2002) described and differentiated R. jimi from other species of the marinus group using the following characteristics: presence of glands on forearm, on the external side of the foot and both sides of cloaca; median snout-vent length (SVL) is 147.48 mm in males and 133.8 mm in females (Stevaux, 2002).No individuals of this common species were collected.

Family Hylidae Rafinesque, 1815
Corythomantis greeningi Boulenger, 1896: Figure 3 Corythomantis greeningi Boulenger 1896: 405-Carvalho, 1941:  The following characters distinguish C. greeningi from its congener C. galeata Pombal Jr, Menezes, Fontes, Rocha & Van Sluys, 2012: head with ridge on the posterior border; skin of dorsum granulate and relatively large discs on fingers and toes, which present keratinized spicules.Snout-vent length varies between 34-82 mm (Carvalho 1941).Corythomantis greeningi was recorded 4 times in the RN and never in sympatry with C. galeata in this state (Jared et al. 2005, Pombal et al. 2012, Magalhães et al. 2013, Caldas et al. 2016).This species can be found in the entrance of tree cavities and among rocks.When found in tree cavities, the cavities contain humid organic material (Jared et al. 1999).Our individual was collect in such a cavity.The voucher of the individual is UFRN 4123.

Reptiles
Family Amphisbaenidae Gray, 1825 Amphisbaena littoralis Roberto, Brito & Ávila, 2014: Figure 5 Amphisbaena littoralis resembles 5 other congeners (A. alba, A. fuliginosa, A. ignatiana, A. lumbricalis and A. pretrei) in the Brazilian northeast.However, A. littoralis is distinguished from A. ignatiana and A. lumbricalis by greater SVL, and from A. fuliginosa by the higher number of body annuli and ventral and dorsal segments.Also, A. littoralis may be distinguished from A. alba by a greater number of body annuli, smaller SVL and possessing an autotomic site, and from A. pretrei by higher number of body and tail annuli.The vouchers of individuals are UFRN 4108-4112.

Discussion
The low diversity and abundance of anurans be associated with 2 factors.First, the activity search was performed during the daylight hours and anurans are more active at night.The second factor may be the sampling method.Studies conducted in the same season and in nearby areas recorded high anuran richness (N = 34 and 19; Magalhães et al. 2013, Caldas et al. 2016, respectively).However, those studies used other sampling methods such active search in temporary and semi-permanent ponds, pitfall trapping and glue traps (Heyer et al. 1994).Furthermore, the study of Magalhães et al. (2013) was conducted in an ecotonal region of Caatinga and Atlantic Forest, where rainfall and humidity levels are higher than in Caatinga stricto sensu.We highlight the record of Corythomantis greeningi, which is the fifth record for RN and extends this species' geographic distribution more than 80 km to the northeast from the record of Jared et al. (2005) in Angicos, RN (Pombal et al. 2012, Magalhães et al. 2013, Caldas et al. 2016).
For reptiles, we recorded diversity similar to that in Caldas et al. (2016), who recorded 22 species (13 lizards, 8 snakes, and 1 chelonian), the same number as this study.However, less than half (N = 12) of their species were found in this study (Table 1).
In our study, lizards were more abundant, possibly because most lizard species recorded were heliophilic, which favors their presence during dry season (Abe 1994) and also because of daylight sampling.The record of Enyalius bibronii is the third record from RN and represents an extension of 130 km northeast from the closest record at Tenente Laurentino Cruz, RN (Freire 1996, Gogliath et al. 2010).Enyalius bibronii exhibits relictual distribution, which may include incursions in certain areas of Caatinga.Our record confirms this possibility (Rodrigues 2003, Gogliath et al. 2010, Silva and Moura 2013).However, E. bibronii may be a species complex, and more detailed taxonomic studies are required (Rodrigues 2003).
For snakes, our records represent 32% of the species For Amphisbaenia, 4 species were found in the study area.At least 7 species are known in the state (Perez and Ribeiro 2008, Freire et al. 2009, Roberto et al. 2014, Caldas et al. 2016).Our record of Amphisbaena littoralis is the second record from RN and represents a range extension and a new habitat.The study area is 65 km southeast from the type locality (Guamaré, RN) and is in Caatinga stricto sensu habitat very different from the marine influenced environments where the species was found previously (Roberto et al. 2014).
We estimate that the species sampled in this study should be representative of the local herpetofauna, since the rescue activities relied on active searching.However, the absence of the stabilization of rarefaction curve (Fig. 24) indicates the possible occurrence of other species in the study area.Alternatively, this could be because of the low numbers of individuals collected of most species (N = 11 species; 44% of total richness).Furthermore, the low incidence of anurans influenced the rarefaction curve.We associate this with the fact that sampling occurred during daylight hours in the dry period and because we used only a single search method.During the study, most anuran species were probably in estivation in burrows in soil, termite mounds, or in lower areas where there is a higher moisture concentration (Navas et al. 2004, Borges-Nojosa and Santos 2005, Loebmann and Haddad 2010).Therefore, it is likely that the herpetofauna, particularly amphibians, concentrates in lower areas during the dry season.In any case, we recommend studies of population dynamics of the herpetofauna in different environments of the Caatinga.
In conclusion, the results indicate that, in order to reduce impacts on biodiversity of the Caatinga, devel- opment projects that require the removal of vegetation should perform studies on the dynamics of species in all seasons, to determine the best times and places for carrying out activities.Our data indicate that dry periods are more suitable for vegetation removal activities.Finally, the results demonstrate that programs related to environmental licensing of projects provide meaningful data on communities and species.Therefore, we recommend that studies required by the licensing, in particular fauna rescue and monitoring programs, be published, because they contribute important records in poorly sampled regions.

Figure 1 .
Figure 1.Map showing the location of the wind energy parks Modelo I and Modelo II in João Câmara, Rio Grande do Norte, Brazil.
known for RN (N = 25;Guedes et al. 2014b).Possibly a greater richness of snakes could be recorded with medium to long-term studies along with a variety of sampling methods.The record of the Epictia borapeliotes is the third from RN and it represents an extension of 90 km northeast from the nearest published locality (Angicos, RN;Guedes et al. 2014a).

Figure 24 .
Figure 24.Species accumulation and rarefaction curves for herpetofauna based on number of individual specimens recorded in João Câmara, Rio Grande do Norte, Brazil.Solid line represents collected individuals and dots are total species estimates based on Chao2P.

Table 1 .
List of the herpetofauna species recorded during the fauna rescue program in João Câmara, Rio Grande do Norte, with their respective vouchers.Abbreviations: N: numbers of individuals per species.