Bats in a restinga area in Sergipe, Northeastern Brazil

There is little known about the bats of the Brazilian restinga as most studies have concentrated on the country’s south and southeast regions. In Sergipe, Northeastern Brazil, the only study previously carried out registered 17 species in different restinga habitats. Thus, this study aimed to characterize the bat community in a restinga area in Sergipe and update the list of species that occur in the area. The study was carried out in the Caju Private Natural Heritage Reserve, on the south coast of the state of Sergipe. Monthly campaigns were carried out from October 2016 to September 2017 over two consecutive nights and alternating between two sites to capture the bats. We captured Bats using 10 mist nets that remained open between 6:00 p.m. and 12:00 p.m. We determined the abundance and trophic guilds of the captured species. In addition, we obtained the occurrence frequency degree through the Constancy Index. We captured 457 individuals distributed over 13 species and two families, where three species represented a new record for the locality. The family Phyllostomidae was the richest and most abundant. Most species were frugivorous (61.5%). According to the Constancy Index, only four species were considered common. Using Jackknife 1 estimator, we estimated 14.83 species for the area, indicating that the richness obtained in this study corresponds to 87.6% of this estimate. This study resulted in an 17.6% increase in bat richness known for the area. The high representativeness of the Phyllostomidae family may be related to the capture method used. The predominance of frugivores bats in this study may be associated with the presence of many fruit trees in the area. The low occurrence of species considered common is often reported and can be explained by the species’ trophic specializations and by the sampling methods. Considering the scarcity of studies in restinga areas in Northeastern Brazil, this work becomes important for the knowledge of the bats in this environment, especially for Sergipe.


INTRODUCTION
Restinga environments are characterized by long strips of sandy marine deposits, with poor soils, a high degree of salinity, high temperature, and intense luminosity (Hay et al., 1981). This environment presents vegetation that varies from herbaceous formations, shrubs, and denser forests, as it moves away from the sea line, whose canopy does not exceed 20 m in height (Silva, 1999;IBGE, 2012). It is characterized as a vegetation subset of the Atlantic Forest (Cerqueira et al., 1990) and, because it is generally close to forest areas, it contains important breeding, feeding, and shelter sites for fauna (Bôlla et al., 2017). In general, the biological composition in restingas is considered a subset of adjacent areas (Rizzini, 1997), and it is one of the least known environments in the Atlantic Forest biome (Rocha et al., 2005;Oprea et al., 2009).
For Chiroptera, studies have revealed that there is not a restricted fauna to restinga environ-ments in Brazil, with the occurrence of generalist and common species, which form a subset of species of the adjacent forest areas (Cerqueira, 2010). Despite the increase in the number of studies with bats in this environment in recent years (Bôlla et al., 2017), most of these are concentrated in the south and southeast regions of the country in the states of Rio de Janeiro (Cerqueira et al., 1990;Pessôa et al., 2010;Luz et al., 2011;Gomes et al., 2016), Espírito Santo (Luz et al., 2009;Oprea et al., 2009), Santa Catarina (Carvalho et al., 2009Bôlla et al., 2017), and Paraná (Fogaça & Reis, 2008). In the northeast of the country, studies with this group in restingas have only been carried out in Paraíba (Campos et al., 2018), Rio Grande do Norte (Soares et al., 2018), and Sergipe (Rocha et al., 2017).
Some of these studies corresponded to a rapid survey with a small sampling effort in different habitats (Luz et al., 2009;Gomes et al., 2016;Rocha et al., 2017;Soares et al., 2018), and most of them revealed a low efficiency of bat capture in the rest-inga areas (e.g., Luz et al., 2009Luz et al., , 2011Oprea et al., 2009;Nogueira et al., 2010;Gomes et al., 2016). Although the majority of species registered in this environment are generalist, the limitation of resources such as food and shelter in the restinga influences the lowest rates recorded (Nogueira et al., 2010), reflecting the lower densities for the group (Luz et al., 2011).
In Sergipe, Northeastern Brazil, Rocha et al. (2017Rocha et al. ( , 2018a recorded 17 species in a rapid survey conducted in different habitats of the restinga area in the south of the state. This study presents an update of this inventory and aimed to characterize the richness and composition of bat species in this area during a year.

Study area
This work was conducted at the Caju Private Natural Heritage Reserve (RPPN do Caju; 11°07′S; 37°11′W) located on the south coast of the state of Sergipe, in the municipality of Itaporanga d'Ajuda (Braghini & Vilar, 2013a;Fig. 1). This area, with approximately 763 ha, corresponds to a remnant of the Atlantic Forest associated with restinga, mangrove, and apicum environments and property of EMBRAPA Tabuleiros Costeiros (Braghini & Vilar, 2013b;EMBRAPA, 2013).
The surrounding areas are characterized by shrimp farming, plant extraction, real estate expansion, and tourism, which negatively influences local conservation (Braghini & Vilar, 2013b;EMBRAPA, 2013). The area has a megathermic tropical climate (Alvares et al., 2013) with an accumulated rainfall during the study period of 1,596.5 mm and greater rainfall between April and September (EMBRAPA, unpublished data).
Two sites in restinga environments 600 m apart ( Fig. 1) were studied. Site 1 is characterized by the presence of dense vegetation with a closed understory and the presence of a temporary pond. Site 2 has a more open understory, bordered by a plantation of Hancornia speciosa Gomes (mangabeira; Fig. 2; Bezerra & Bocchiglieri, 2018).

Data collection and analysis
We carried out the campaigns monthly, for two consecutive nights, from October 2016 to September 2017, alternating between the two sites. We used ten mist nets (9 × 3 m, mesh 20 mm). They remained open between 6:00 p.m. and 12:00 p.m. and we inspected them at 30-minute intervals. We identified each captured bat based on Díaz et al. (2016) and Peracchi & Lima (2017), using specialized keys when necessary. We marked and released the Bats at the same capture site according to the SISBIO Research and Collection License № 54957-1.
The richness was estimated using the EstimateSWin 8.2 program (Colwell, 2011), the non-parametric estimator that presented the less accumulated standard deviation (Jackknife 1), and through the construction of 10,000 accumulation curves of species with an increase in the sampling effort.
Based on Jackknife 1, we estimated 14.83 ± 1.24 species from the sampling effort spent (Fig. 3), indicating that the richness recorded in this study (S = 13) corresponded to 87.6% of the estimate for the area.

DISCUSSION
In the RPPN do Caju, 17 species of bats were known (Rocha et al., 2017(Rocha et al., , 2018a, of which 10 (62.5%) we captured in this study. The previous record of M. nigricans (Schinz 1821) presented by these authors corresponded to M. lavali, also captured in this study. There was an addition of three species to the locality, which resulted in an 17.6% increase in bat richness recorded in the area, for a total of 20 species.
The richness observed in restinga environments varied from 2 (Cerqueira et al., 1990) to 17 species (Luz et al., 2009). This difference was due to different sampling efforts, that varied between 1,440 m².h (3 species; Nogueira et al., 2010) and 63,926 m².h (13 species; Bôlla et al., 2017), and capture methods at each location. The lower richness found in this study, when compared to that previously carried out in the RPPN do Caju, may be   related to the change in the location of the mist nets during sampling, resulting in the record of more species by Rocha et al. (2017). Likewise, Luz et al. (2009Luz et al. ( , 2011 and Oprea et al. (2009), despite the different sampling efforts and low capture rate, recorded between 14 and 17 species in the restinga areas in Southeastern Brazil as a reflection of the greater variety of sampled environments. The predominance of frugivores bats in this study may be associated with the presence of many fruit trees in the area (EMBRAPA, 2013; personal observation) and the fact that this guild is well represented in neotropical environments (Nogueira et al., 2010). In addition, frugivore bats of the Phyllostomidae family are predominant in studies with mist nets in the Neotropical region (Bergallo et al., 2003;Nogueira et al., 2010). The same pattern was also observed in other restinga areas (Carvalho et al., 2009;Oprea et al., 2009;Luz et al., 2011;Soares et al., 2018) as a reflection of the sampling method and the availability of fruits in the areas.
Carollia perspicillata, A. lituratus, A. planirostris, and D. cinerea were the most captured species in this study, being abundant in restinga areas (Oprea et al., 2009;Nogueira et al., 2010;Luz et al., 2011;Gomes et al., 2016;Soares et al., 2018). These species occur in a variety of habitats, from more structured forests to degraded areas and plantations (Bernard, 2002;Passos et al., 2003), and have high ecological flexibility because they are able to explore a wide variety of resources (Mikich, 2002;Passos et al., 2003).
In the RPPN do Caju, these four species (30.7% of the total species captured) were also considered frequent and the most abundant. The capture of bats in Neotropical environments using mist nets is generally composed of a few common species and many rare species (Kalko & Handley, 2001). This low frequency of some species may be related to the trophic specialization of the species or sampling methods (Kalko & Handley, 2001). Thus, it is important to consider the capture method used that may have influenced the abundance of the species (Simmons & Voss, 1998;Larsen et al., 2007).
Sampling in different environments and with complementary methodologies, such as active search and ultrasound detectors, can contribute to the increase in richness (Esbérard & Bergallo, 2008;Carvalho et al., 2009;Gomes et al., 2016) from a high number of captures (Bergallo et al., 2003). In the RPPN do Caju, there is an expectation of greater richness since Rocha et al. (2017) registered representatives of the Molossidae and Emballonuridae families through the use of mist nets close to shelters. Thus, a greater sampling effort with the use of additional collection methods in different environments may reflect an increase in the local number of species.
Considering the scarcity of studies in restinga areas in Northeastern Brazil, this work becomes important for the knowledge about the chiropterofauna in this environment, especially for the state of Sergipe where studies related to the characterization of this group's community are still considered incipient in the region (Mikalauskas, 2005;Rocha et al., 2010Rocha et al., , 2015Rocha et al., , 2018bBrito & Bocchiglieri, 2012).

ACKNOWLEDGMENTS
This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior -Brasil (CAPES) -Finance Code 001. We are grateful to EMBRAPA Tabuleiros Costeiros and Universidade Federal de Sergipe (UFS) for logistical support, Fundação de Apoio à Pesquisa e Inovação Tecnológica do Estado de Sergipe (FAPITEC) (# 019.203.01181/2011-2); and CAPES for financial support (2417/2013; 1941/2017), and the students of the Laboratório de Mastozoologia/UFS for helping us in the field campaigns.