Structure of understory bird communities in Atlantic Forest Remnants in Northeastern Brazil

. The Atlantic Forest in the State of Paraíba integrates the Pernambuco Center of Endemism. The present work investigated understory bird communities of three of its remnants using ten mist nets installed at eight different points of the forest understory of each area. We captured 333 birds of 44 species and 19 families. Bird richness and abundance did not differ between the three remnants (KW = 6.08; p = 0.04 and KW = 10.54; p = 0.004). The composition was little similar among them, and each area presented high species diversity and evenness. The most captured species were Arremon taciturnus (n = 40), Chiroxiphia pareola (n = 32) and Conopophaga melanops (n = 28). C. melanops has a local threat status of “vulnerable” and is considered endemic. Endangered and endemic species each represented 15.91% of the captures. Both Reserva Biológica Guaribas areas had the highest number of threatened species, while Reserva Particular do Patrimônio Natural Engenho Gargaú showed the highest number of endemic and understory insectivorous species, which indicates a better quality status in comparison to the other two. These results show the importance of the remnants to bird conservation in Northeastern Brazil


Introduction
The tropical region presents high levels of biodiversity and endemism (Sodhi, Sekercioglu, Barlow, & Robinson, 2011).The Brazilian Atlantic Forest, for example, is considered one of five hotspots with the highest number of endemic species of plants and vertebrates (Myers, Mittermeier, Mittermeier, Fonseca, & Kent, 2000).Extending for roughly 150 million hectares (Ribeiro, Metzger, Martensen, & Hirota, 2009), the biome has been reduced to small forest remnants and holds a large proportion of threatened species (Goerck, 1997).
Acta Scientiarum.Biological Sciences, v. 40, e37145, 2018 Birds participate in a large range of ecological functions, such as predation, pollination, seed dispersion, nutrient deposition through feces, and even decomposition, according to the species present in the ecosystem (Lundberg & Moberg, 2003).However, ecological information on the avifauna and the relationships between habitat and species is still limited (Sodhi, Brook, & Bradshaw, 2007), even though this knowledge can help understand the consequences of habitat change (MacNally & Bennett, 1997).
One of the areas where birds need studies is the northern portion of the Brazilian Atlantic Forest, which is considered the most vulnerable region of the biome (Silva & Tabarelli, 2000).The Pernambuco Center of Endemism, situated in this area, holds two thirds of the Brazilian Atlantic Forest birds (Roda, Pereira, & Albano, 2011) and, together with the Central Atlantic Forest Ecological Corridor, form the areas with the highest number of endemic species in the country (Olmos, 2005).The Center has 40 taxa threatened to some extent and one extinction, the species Pauxi mitu (Linnaeus, 1766) (Roda et al., 2011).Nonetheless, some states in the Center, such as Paraíba and Rio Grande do Norte, have been little studied.
Knowledge of the avifauna in the state of Paraíba is still incipient.There are 337 bird species registered in the state (Schulz-Neto, 1995), with few of the studies developed exclusively in the Atlantic Forest (Pinto & Camargo, 1961;Almeida & Teixeira, 2010;Pereira et al., 2014).Therefore, this research aimed to measure species richness, composition, diversity and similarity of three Atlantic Forest remnants in Northeastern Brazil.

Study area
This research was carried out in three Atlantic Forest remnants of the coastal portion of the State of Paraíba, Northeastern Brazil.The weather is predominantly hot and humid with the mean annual temperature ranging from 24 to 26°C, with maximum annual values of 36°C; the mean pluviosity is from 1,750 to 2,000 mm year -1 and the mean relative humidity reaches a maximum of 90% (Ministério do Meio Ambiente / Instituto Brasileiro do Meio Ambiente [MMA/IBAMA], 2003).
The areas present vegetation dominated by secondary semideciduous forest and alluvial ombrophilous dense forest, in the highest portions and along watercourses, with canopies 30 m in height; and savannic formations and transition regions between both, in Tabuleiro areas (MMA/IBAMA, 2003).The remnants are undergoing a natural restoration process after wood extraction 31 to 74 years ago, with better quality portions in RBG2 (MMA/IBAMA, 2003) and RPEG, as can be observed by the tree heights and the presence of more lianas and epiphytes.The forest of RBG1 presents more signs of degradation (MMA/IBAMA, 2003).The areas are inserted in an anthropic matrix with fruit and sugarcane farming, as well as pastures and small farms.The remnant RBG1 is surrounded mainly by fruit plantations, while RBG2 has a small city and mostly sugarcane around its borders (MMA/IBAMA, 2003) and RPEG has its edge facing mainly sugarcane and pasture.

Methods
Eight expeditions were carried out on each remnant from October 2011 to September 2012, always covering different portions.We installed ten mist nets (12 × 3 m, five shelves, 19 mm mesh) to capture birds in the forest understory from 05:00 to 11:00 for two consecutive days on each visit (Keyes & Grue, 1982).The mist nets were installed as line clusters of five in the first 60 m from the forest edge and after 180 m, in the forest interior.The total mist-net effort was 2,880h.
After capture, birds were identified using field guides (Erize, Mata, & Rumboll, 2006;Sigrist, 2007) and marked with CEMAVE/IBAMA metal rings on permit number 32568-1 and SISBIO authentication code 81552292.Systematics and taxonomy followed the list of the Brazilian Ornithological Records Committee of 2015 (Piacentini et al., 2015) and trophic information followed Sick (1997).The threat level of the species followed Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio, 2016) and the endemism status for Atlantic Forest and Pernambuco Center of Endemism followed Parker III, Stotz, and Fitzpatrick (1996) and Roda et al. (2011) respectively.

Data analysis
We compared remnant richness and abundance from the capture data using a Kruskal-Wallis test followed by Dunn's post-hoc test (p < 0.05).The Shannon-Wiener diversity index was used to access bird diversity of the areas, while the Pielou index was used to verify their evenness.The similarity between remnants was measured through the Jaccard similarity index.Specimen recaptures were removed from all data analysis to avoid overestimation.
On the other hand, some threatened species and those of higher sensitivity levels, such as P. mystaceus, P. leuconota and C. melanops (Parker III et al., 1996;ICMBio, 2016), registered at a high frequency here, were represented by low numbers in other alluvial ombrophilous dense forest remnants studies in Pernambuco (Lyra-Neves et al., 2004;Farias, 2009).
These species feed mainly on insects in the understory and sometimes follow army ants, aiming for exposed arthropods flushed by the swarms (Remsen et al., 1983), which was seen during this study (mainly in RPEG) and may have influenced their presence.This disparity between research and areas within this study may also occur because of differences in the vegetation, size and quality of the areas, as well as the different anthropic use and occupation history (Martin, Gaston, & Hitier, 1995;Farias et al., 2007).
The highest richness and diversity found in RBG1 may be mainly due to its larger size in comparison to the other remnants (Gimenes & Anjos, 2003;Telles & Dias, 2010), since several factors, such as size, shape, isolation degree and characteristics of remnants may influence species presence, richness and diversity (Farias et al., 2007;Martin et al., 1995).Also, the habitat chosen by some bird species may be highly specific and influenced by the vegetation of the forest remnant and surrounding matrix structures (Gimenes & Anjos, 2003).The dissimilarity found between the bird communities of each area may have been influenced by the different surrounding matrices and the fact that RBG1 and RBG2 are administrated by government and legal protection, while RPEG is a private area (MMA/IBAMA, 2003).The surrounding matrices of forest remnants directly influence habitat quality, species composition and their dispersion (Bierregaard & Stouffer, 1997;Sisk, Haddad, & Ehrlich, 1997), as the distance between remnants may make dispersion of individuals between areas difficult and select distinct species (Sisk et al., 1997;Telles & Dias, 2010;Ferraz et al., 2007).
The presence of some threatened and endemic species in this portion of the Atlantic Forest was one of the main aspects found in this study (Roda et al., 2011;Telino-Júnior et al., (2005), Farias et al., 2007;Farias, 2009).This may suggest that the investigated areas are still capable of presenting good habitat quality (Gouvêa-Maia et al., 2005).On the other hand, even though our samples were limited by the use of mist nets, the presence of species with special habitat requirements, such as understory insectivores (e.g.P. leuconota and C. laeta in RPEG), and their absence in both Reserva Biológica Guaribas areas may indicate that their local conditions may not be as good as those found in RPEG.

Conclusion
The data suggests the need to understand the factors capable of maintaining a group of species in fragmented areas; this particularly concerns species Acta Scientiarum.Biological Sciences, v. 40, e37145, 2018 that are highly sensitive to disturbance.More detailed studies may help define their actual conservation status.
Even though the remnants present a similar history of anthropic pressure and a great portion of the areas were intensely disturbed in recent decades (MMA/IBAMA, 2003), the high diversity found, as well as the presence of endangered and endemic species shows the importance for the bird conservation in Northeastern Brazil and the importance of studying forest and their biodiversity.