Inventory of birds in the coastal restinga of a Private Natural Heritage Reserve in northeastern Brazil

The knowledge of the characteristics of bird communities found in each habitat type and the ecological requirements of each species helps to define the environmental conditions of an area. In the present study, we provide an inventory of the bird community of the Caju Private Natural Heritage Reserve (RPPN Caju), in Itaporanga d’Ajuda, Sergipe, Brazil. We conducted a rapid assessment of the bird community between October 21st and 31st, 2014, using MacKinnon’s lists and mist-netting. We identified 89 bird species representing 36 families, of which the most diverse were Thraupidae (S=11), Tyrannidae (S=11), Columbidae (S=5), Falconidae (S=4) and Psittacidae (S=4). Five of the species recorded are endemic to Brazil and one (Herpsilochmus pectoralis) is considered vulnerable to extinction, according to IUCN criteria. We recorded bird species that are predominantly insectivorous (40.45%), omnivorous (25.84%), and carnivorous (11.24%). Other guilds (frugivorous, granivorous, detritivorous, piscivorous, and nectarivorous) represented together less than 25% of the records. The bird richness recorded at the RPPN Caju confirms the potential importance of the area as a conservation unit.


Introduction
Birds constitute a highly diverse vertebrate group, with 1,901 species known to occur in Brazil (CBRO, 2014), most of which are habitat specialists.Knowledge of the ecological requirements of some bird families, genera and species may be sufficient, in some cases, to determine the environmental conditions to which the birds are sensitive (Donatelli et al., 2004).These characteristics (habitat and ecological requirements) contribute to the role of birds as one of the principal bio-indicators of environmental quality in terrestrial ecosystems (Canterbury et al., 2000).
Coastal restingas are highly diverse environments, resulting from the formation of distinct phtyophysiognomies, related to the distinct structural features of coastal plains, including the beach line, dunes, swamps, lagoons, and forests (Silva, 1999).Due to their coastal location, restingas are among the most endangered Brazilian ecosystems, which are threatened primarily by urban growth and development (Rocha et al., 2007).
While bird species adapted to open habitats are common in restingas, these systems are assumed to have low levels of bird diversity and endemism (Alves et al., 2004), which are reflected in the limited interest of ornithological researchers (Gonzaga et al., 2000).Another factor that contributes to the lack of knowledge on the fauna of restingas is the intense colonization of the Brazilian coast, which began in the sixteenth century, and has intensified in recent decades, with extensive impacts on the local environments and their resident fauna (Mittermeier et al., 2005;Rocha et al., 2004;Sick, 1997).
One of the most effective strategies for the conservation of natural ecosystems is the establishment of an effective system of protected areas (Bruner et al., 2001).In Brazil, these areas include zones of permanent protection (e.g., riparian forest and mangroves), legal reserves (i.e., the forested portion of private properties), indigenous lands and conservation units.Conservation units can be established on either public land or private properties.In this context, the establishment of Private Natural Heritage Reserves (RPPNs) is an important complement to the conservation of species and ecological services without requiring public resources.
Faunal inventories are an important component of environmental conservation, especially in areas that suffer intense anthropogenic pressures, such as the coast of Sergipe.In this study, we present the results of a rapid assessment of the bird community of the RPPN Caju, providing important new insights into the complexity of the fauna of Brazilian coastal restinga ecosystems.

Study site
Caju Private Natural Heritage Reserve, RPPN Caju (11°6'10.12"S, 37°11'4.59"W), declared a conservation unit in 2011, is located in Itaporanga d'Ajuda, at km 24 of the SE-100 state highway in Sergipe, northeastern Brazil (Figure 1).The RPPN Caju encompasses an association of Atlantic Forest environments, including shrubby-arboreal restinga, mangroves, sand dunes, and salt marshes in an area of 763.37 hectares (EMBRAPA, 2013).The main plant species present in the study area are Schinus terebinthifolius Raddi 1820, Anacardium occidentale L., Psidium sp., Andira fraxinifolium Benth, Byrsonima sericea DC, Inventory of birds in the coastal restinga of a Private Natural Heritage Reserve in northeastern Brazil and Campomonesia sp.(EMBRAPA, 2013).The local climate is Tropical Megathermal with a dry season in the austral summer, corresponding to the As type in Köppen's classification system (Alvares et al., 2013).Mean annual rainfall is typically above 1,250 mm (EMBRAPA, 2013).
Local topography is typical of the coastal lowlands of the state of Sergipe (Silva et al., 2000), which are predominant flat, with subtle undulations and few areas above 10 m in altitude, corresponding to the sand dunes (EMBRA-PA, 2013).The soils are sandy and have low natural fertility (Araujo Filho et al., 1999;Silva et al., 2000).

Inventory of the avian fauna
We conducted the bird inventory, between October 21 st and 31 st , 2014, using two different methods simultaneously, MacKinnon lists and mist-netting.In both methods, we used pre-established trails located within the different vegetation types of the RPPN Caju, and a field guide (Sigrist, 2013) for the identification of the bird species.The MacKinnon list procedure is based on MacKinnon and Phillips (1993), where the observer walks through a pre-established area, listing the bird species encountered up to a predetermined number of taxa.Each list constitutes a sampling unit.In the present study, we used lists of 10 species, in order to maximize the number of sampling units (Herzogh et al., 2002), and minimize the probability of recording a species more than once in the same unit.We (up to three observers) collected MacKinnon lists during the periods of greatest bird activity, i.e., early morning (6h-9h) and late afternoon (15h-18h), using binoculars (8x40), with a total sampling effort of 60 hours, with 60 MacKinnon lists being collected.In order to avoid over-estimates, we used five different transects, which were alternated every five days.
To complement these lists, we captured birds in mist nets, enabling the register of bird species less easily detected by visual or auditory methods (Roos, 2010).We set eight mist nets (2.5 x 12 m; 36 mm mesh) over a total of 128 hours, divided between morning (5-10 h) and afternoon (15-18h) sessions, checking the nets at 20-minute intervals to minimize the risk of predation (Ruiz-Esparza et al., 2011).We changed the sampling point every two days, in order to sample the different microhabitats (freshwater lagoons, mangroves, and grassland, shrub and shrubby-arboreal restingas) found within the RPPN Caju.We removed the trapped birds carefully, and after biometry, photography and banding, we released the birds at the capture site (CEMAVE, 1994).
Opportunistic records of bird species encountered outside the principal sampling areas were included in the species inventory.The taxonomic ordination adopted follows the proposition of the Comitê Brasileiro de Registros Ornitológicos (CBRO, 2014).We estimated the total species richness of the study area (Jackknife 1) using Estimates 9.0 (Colwell et al., 2012), and we calculated the Shannon-Wiener diversity index (H') and Pielou's equitability index, E' (Pielou, 1966;Tramer, 1969).
To assess the residence status of the species, we calculated the Frequency of Occurrence (FO) of the species recorded in the MacKinnon lists, which we presented as a daily rate of occurrence: FO = n/N*100, where n = the number of days on which the species was recorded, and N = the total number of observation days.Residence status was defined as Occasional (O: FO < 15%), Possibly Resident (P: 60% > FO ≥ 15%), and Resident (R: FO ≥ 60%), following Mendonça- Lima and Fontana (2000).

Results
Considering MacKinnon lists, captures in the mist nets and occasional observations, we recorded 89 bird species in the RPPN Caju (Appendix 1).These species are distributed in 36 bird families, of which the most diverse were Thraupidae (S=11), Tyrannidae (S=11), Columbidae (S=5), Falconidae (S=4), and Psittacidae (S=4).If we consider only MacKinnon lists, 78 species were recorded.According to the species richness estimator Jackknife 1, we should have recorded 89 species (Figure 2).This was 13% more than the number recorded in the lists, but it corresponds to the total recorded in the combined dataset (including mist-netting and occasional sightings).The diversity index (H') was 3.82 and the equitability index (E') was 0.87.

Discussion
The 89 bird species recorded in the RPPN Caju represent 26% of the 340 bird species recorded to date in the state of Sergipe (Sousa, 2009).The bird species richness found in the present study is similar to that of other restinga sites in northeastern Brazil, which vary among 79 species recorded in Piaçabuçu, in the state of Alagoas (Cabral et al., 2006), 96 species in Imbassaí, in the state of Bahia (Mota et al., 2012), and 127-136 species in Aracajú, Sergipe (Almeida et al., 2012;Ruiz-Esparza et al., 2014).Inventories conducted in coastal restingas in southern and southeastern Brazil have generally recorded higher species richness (Table 1).In the arboreal restinga of Itanhaém/ Mongaguá, in the state of São Paulo, Lima ( 2010 (2004) found only 84 bird species.These differences may at least partially be the result of the different methods or sampling effort used in each study (Bibby et al., 2000), in addition to the heterogeneity of the habitats, the size of the study area, and the surrounding matrix (Watson et al., 2005).
While coastal restinga ecosystems in Brazil are protected by law, together with mangroves, it is important to note that only four of the 13 sites in Table 2 are located in protected areas, and eight are close to urban zones characterized by intense anthropogenic pressure, including urban development.Despite the legal protection of habitats, the fauna of these areas may be vulnerable, and the total area of conservation units should be increased, especially to endemic birds or migratory species that use the sites to breed.
While some species that occur in the region, such as frigatebirds, terns and sandpipers (Almeida et al., 2012;Cabral et al., 2006;Mota et al., 2012;Sanabria et al., 2010), were not recorded in the present study, we did record one species -Paroaria dominicana (Cracraft, 1985) -which is endemic to the Caatinga biome.This species is popular in the illegal wildlife trade, especially in the Brazilian Northeast (Pereira and Brito, 2005), and is often released illegally.It thus seems possible that the present record represents an expansion of the original species range as a consequence of human activities (i.e., illegal wildlife trade and habitat changes), probably supported by its ecological flexibility and to the mobility of the species in open habitats (Olmos et al., 2005).
Only three species captured in the mist nets were not recorded in the MacKinnon lists -Chloroceryle americana (Gmelin 1788), Hydropsalis albicollis (Gmelin 1789) and Neopelma pallescens.A total of 50 species were recorded only in the MacKinnon lists, however, which reinforces the utility of this technique in rapid assessment surveys (Ribon, 2010).
The majority of the bird species we recorded is insectivorous (40%).Such guild plays an important ecological role in the maintenance of plant community.Insectivorous birds, for example, are important for the biological control of insect pests, which may be potentially detrimental to agriculture (Andrade, 1988).In the specific case of the present study area, insectivorous birds provide an important ecosystem service, given that EMBRAPA keeps an active germplasm reserve of coconut (Cocos nucifera L.) and mangaba (Hancornia speciosa Gomes) in the RPPN Caju (EMBRAPA, 2013).
The other trophic guilds were more poorly represented, although they may also have their ecological importance in the study area.Frugivorous species play an important role in seed dispersion (Fadini and Marco Jr., 2004), for example, and nectarivorous birds are important pollinators.In tropical forest ecosystems, more than 80% of the plant species are dependent on vertebrates for their dispersal (Howe and Smallwood, 1982), and, in the Neotropical region, hummingbirds are responsi-ble or the pollination of up to 15% of all angiosperms (Feinsinger, 1983).
In addition to its overall diversity, the inventory of the RPPN Caju recorded a number of specific cases that highlight the importance of the study area, including the vulnerable Herpsilochmus pectoralis, species typical of coastal habitats (Mimus gilvus and Conirostrum bicolor), and evidence of the range expansion of a Caatinga endemic Paroaria dominicana.The RPPN Caju clearly has a potentially important role to play in the conservation of the region's fauna, and in the ecosystem services it provides, especially in the context of ongoing urban development and expansion.

Figure 2 .
Figure 2. Species accumulation curve (line) and rarefaction curve (dashed) of the Jackknife 1 estimator for the bird species recorded using MacKinnon lists in the Caju Private Natural Heritage Reserve, Itaporanga d'Ajuda, Sergipe, Brazil.The vertical bars correspond to the standard deviation.
) recorded 262 bird species, Vieira et al. (2014) registered 227 species in the Carijós Ecological Station (state of Santa Catarina), and Sanabria et al. (2010) recorded 111 species in northern Rio Grande do Sul.In the Restinga de Jurubatina National Park, in the state of Rio de Janeiro, however, Alves et al.

Figure 4 .
Figure 4. Trophic composition of the bird community at Caju Private Natural Heritage Reserve in Itaporanga d'Ajuda, Sergipe, northeastern Brazil.