The Brazilian Atlantic Forest is one of the most endangered biomes on Earth (Ribeiro et al. 2009), and due to its high biodiversity and rates of endemism it is considered one of the world’s hotspots, and an area of high priority for conservation (Myers et al. 2000). Because of a historical process of exploiting the land for timber, cultivation of coffee, sugar cane, and, more recently, pastures for livestock (Galindo-Leal and Câmara 2005), the original vegetation was reduced to about 11%, and most of the remnant vegetation is currently diffused in small fragments of second-growth forests (Ribeiro et al. 2009). Biodiversity surveys in these remnants are necessary to subsidize conservation management plans for local faunas and floras. Also, the study of local faunas is essential for understanding the regional patterns of biological diversity, and allows better characterizations of the geographic distribution of specific taxa (Soulé and Wilcox 1980).

Bats represent the second most speciose mammalian order, and can make up more than half of the mammal species in some Neotropical communities (Tim 1994). Surveys of bat faunas can be important tools to evaluate the degree of habitat conservation in forest remnants (Medellín et al. 2000, Estrada and Coates-Estrada 2002). According to Esbérard (2003) the richness and diversity of bat species depends on the local availability of food and shelter, so there is a relationship between bat community composition and complexity of habitats available within a given site (Estrada et al. 1993, Sedlock et al. 2008). Thus, different habitats should be sampled to adequately survey local faunas (Kunz and Kurta 1988), including both human-modified and natural environments.

The state of Rio de Janeiro has one of the best studied bat faunas in Brazil (Bergallo et al. 2003, Esbérard and Bergallo 2005, Bernard et al. 2010). However, the mountainous region (i.e., slopes of Serra do Mar and Serra da Mantiqueira) and the northern portion of the state still represent gaps in the knowledge of the bat fauna, and surveys are highly required (Moratelli and Peracchi 2007, Modesto et al. 2008a, Modesto et al. 2008b, Esbérard et al. 2010, Peracchi and Nogueira 2010). Here, we present lists of bat species from three previously unsampled localities in the state of Rio de Janeiro (two in mountainous areas and one in the extreme north of the state), with comments on the importance of sampling in different habitats.

Voucher specimens were deposited at the mammal collection of Museu Nacional, Rio de Janeiro (MN) and at the bat collection of Laboratório de Diversidade de Morcegos (LDM), Universidade Federal Rural do Rio de Janeiro, Rio de Janeiro.

Sumirouro, Rio de Janeiro, Brazil: Anoura caudifer (MN 77690-77695, 77710, 77711-77713, 77715, 77716, 77718-77720); Artibeus lituratus (MN 77696, 77700, 77701); Artibeus obscurus (MN 77698, 77717); Carollia perspicillata (MN 77699, 77702, 77706, 77714, 77732); Chiroderma doriae (LDM 5266, 5267); Chrotopterus auritus (MN 77726, 77729, 77730); Desmodus rotundus (MN 77697, 77708, 77727, 77731, 77733); Diphylla ecaudata (MN 77728); Macrophyllum macrophyllum (MN 77735); Platyrrhinus lineatus (MN 77703); Platyrrhinus recifinus (MN 77705, 77709); Vampyressa pusilla (MN 77704, 77707). Varre-Sai, Rio de Janeiro, Brazil: Artibeus fimbriatus (MN 77725); Carollia perspicillata (MN 77722, 77723); Myotis nigricans (MN 77724); Platyrrhinus lineatus (MN 77734); Sturnira lilium (MN 77721). Cantagalo, Rio de Janeiro, Brazil: Carollia perspicillata (MN 77738, 77746, 77749); Desmodus rotundus (MN 77739-41, 77743, 77745); Diphylla ecaudata (MN 77747, 77748); Glossophaga soricina (MN 77736); Peropteryx macrotis (MN 77737, 77742, 77744, 77750-77752).

We captured a total of 148 bats of 17 species for the three localities together (Sumidouro = 82 individuals of 11 species; Cantagalo = 25 individuals of seven species; Varre-Sai = 41 individuals of nine species). Carollia perspicillata was the most frequent species in the three areas, representing more than 50% of all bats recorded at Varre-Sai and more than 35% of the records from the other two areas (Table 1). Carollia perspicillata seems to be the dominant species in most Atlantic Forest localities in Rio de Janeiro (see Esbérard et al. 2006, Dias et al. 2008, Luz et al. 2011, Delciellos et al. 2012), and one of the most common in the Neotropics.

Individuals of frugivorous bats accounted for most of the captures in the three areas, and in the four habitats sampled as well. The dominance of frugivorous species typical of forest edge, such as C. perspicillata, is common in secondary forest fragments and agricultural areas (Heithaus and Fleming 1978, Faria 2006, Rocha et al. 2010, Novaes et al. 2014), since these species are benefited by the presence of pioneer vegetation and by high fruit production (Herrera et al. 1994, Murcia 1995, Guariguata and Sáenz 2002, Asbjornsen et al. 2004). The use of mist-nets may biased the low sampling of insectivorous bats, becuase it is more efficient for sampling representatives of phytophagous species (such as stenodermatines and carolliines) than animalivorous species, which can detect mist-nets easily (Portfors et al. 2000, Gorresen et al. 2008).

The three areas sampled had similar values for species diversity, equitability and dominance (Table 1). However, there were significant differences between these indices when compared by habitat. The forest interior showed higher species richness and species diversity (H’=2.27) than the other habitats, but there were no dominance (D’=0.12) (Table 2), indicating that the forest interior is subject to less environmental stress than the other habitats. The forest interior also had a high evenness of species composition (J’=0.91), and a greater number of exclusive species (five) when compared with the other habitats.

The higher species richness within the forested areas was expected, since these environments have more heterogeneous habitats, allowing coexistence of more species from different trophic guilds than other habitats, including those species with more specialized feeding habits (Kalko et al. 1996). Nevertheless, five species were recorded neither in the interior nor at the edge of forests, among them: Peropteryx macrotis and Glossophaga soricina were collected only in open areas; Macrophyllum macrophyllum and Platyrrhinus lineatus were collected only near or over water bodies in riparian forest; and Myotis nigricans was collected both in open areas and riparian forest. These observations reinforce the importance of sampling different habitats during short-term species surveys.

The three regions sampled showed low richness and diversity of species compared to other studies carried out in mountainous areas and other close localities (e.g., Esbérard 2007, Moratelli and Peracchi 2007, Dias et al. 2008, Esbérard et al. 2010). This may be due to the low sampling effort employed in the present study, when compared with the aforementioned studies, or even due to the characteristics of the landscape, since most of the previous studies conducted in mountainous regions of Rio de Janeiro were concentrated in areas of continuous forest, which support higher species richness (Medellín et al. 2000). Therefore, it is possible that the high degree of human-induced disturbance and habitat fragmentation in these areas have resulted in the loss of more sensitive species.

Even considering the low species richness, the areas sampled in the present study yielded some interesting records. An individual of Macrophyllum macrophyllum, a species considered rare in the state of Rio de Janeiro, was captured in a mist-net placed over water bodies, in Sumidouro. According to Weinbeer et al. (2005), this species has a strong association with habitats with collections of water, since it forages close to the water, catching aerial insects or “fishing” semi-aquatic insects from the water surface (Meyer et al. 2005).

We known very little about bats inhabiting caves in Southeastern Brazil. Captures with mist-netsset up inside the Novo Tempo cave revealed colonies of Peropteryx macrotis, Desmodus rotundus, Diphylla ecaudata, Carollia perspicillata and Artibeus obscurus.

Considering the three areas combined, the list of species obtained here is still very preliminary, with the species accumulation curve did not reaching an asymptote (Fig. 2). However, the results of this study indicate that sampling different habitats within a given locality increase the efficiency of bat inventories, in particular, during short surveys.