INTRODUCTION

The Brazilian Atlantic Forest is one of the main biodiversity hotspots on earth (Myers et al. 2000). After more than 500 years of exploitation, only about 11.4 to 16% of the original area is left, and about 97% of the remnants are smaller than 250 ha and are exposed to intense anthropogenic pressures (Ribeiro et al. 2009). In this megadiverse biome, the Pernambuco Endemism Center (hereafter PEC) is of special concern, considered to be a hotspot within a hotspot (Tabarelli et al. 2006, Mendes Pontes et al. 2016). The PEC is the portion of the Atlantic Forest isolated in the south by the São Francisco River and in the north and west by the Caatinga biome. This particular forest is distributed mainly along the Atlantic coast in the states of Paraíba, Pernambuco, and Alagoas in northeastern Brazil. Today, the PEC is the most degraded of all Atlantic Forest regions, with only about 2% of the original 56,000 km² of forests remaining (Pereira et al. 2016, Garbino et al. 2018), and especially during the 1970s and 1980s the forest remnants suffered from intense logging to supply the local sugarcane and ethanol industries with timber. Three bird species from the PEC were recently recognized as globally extinct (the Pernambuco Pygmy-Owl, Glaucidium mooreorum, the Cryptic Treehunter, Cichlocolaptes mazarbarnetti, and the Alagoas Foliage-gleaner, Philydor novaesi; Pereira et al. 2014, ICMBio 2018); some were likely extinct before being described by science (Silveira et al. 2003, Lees and Pimm 2015), and many others are on the verge of extinction (Pereira et al. 2014, 2016, ICMBio 2018, Francisco et al. 2020). Furthermore, new endemic species are still being described and immediately listed as threatened (Dickens et al. 2021).

In this scenario of intense degradation, studies on bird population demography are important because they can be the first step in estimating local and global population sizes, which in turn can parameterize population viability analyses, and are elementary for assessing the actual conservation status of the taxa (Sinclair et al. 2006, Cornils et al. 2015, Tonetti and Pizo 2016, Machado et al. 2020). Furthermore, studies on microhabitat selection can contribute to the understanding of local extinctions and can indicate habitat characteristics that must be preserved or restored for the conservation of target species (Sinclair et al. 2006, Vié et al. 2009, Cornils et al. 2015, Tonetti and Pizo 2016).

In the Neotropics, the guild of understory insectivores is of interest to conservation practitioners (Stratford and Stouffer 2013, Powell et al. 2015). These birds are often unwilling to engage in long flights and are reluctant to cross open areas, favoring the isolation of small populations in fragments of habitats and the consequent exposure to inbreeding and to the demographic impacts of isolation (Moore et al. 2008, Oliveira, Jr., et al. 2011, de Camargo et al. 2015, Powell et al. 2015). Although they are not game birds and are not targets of illegal trapping, two of the three bird species from the PEC that were recently uplisted as globally extinct, the Cryptic Treehunter and the Alagoas Foliage-gleaner, were forest insectivore passerines, and of the 14 other endangered taxa, six are forest understory insectivores (Pereira et al. 2014, ICMBio 2018). Some of these taxa, however, still occur in forest fragments in which logging activities may have caused potential microhabitat changes (Pereira et al. 2014, 2016), making the understanding of demographic aspects and habitat requirements of the endangered bird taxa from the PEC of broad theoretical and practical interest.

Here we provide distance sampling–based population density estimates and microhabitat selection information for two endangered forest understory insectivorous suboscine passerines from the PEC, the Pernambuco Fire-eye, Pyriglena pernambucensis, and the Black-cheeked Gnateater, Conopophaga melanops nigrifrons, from an Atlantic Forest remnant from Alagoas state, northeastern Brazil. Further, we extrapolated population density values to areas with reliable records of these species to obtain minimum global population size estimates. Although extrapolating data from only one fragment to other areas can provide only rough estimates, to our knowledge this is the first empirical study to address demographic aspects and microhabitat requirements of endangered bird species from this important global hotspot.

METHODS

Study area

Field work was carried out at the Private Natural Heritage Reserve (PNHR) Mata do Cedro (9°31’23.82” S, 35°55’6.53” W), a 978-ha Atlantic Forest conservation unit surrounded by sugar cane plantations, located in the municipality of Rio Largo, state of Alagoas, northeastern Brazil (Fig. 1). Vegetation is characterized as open ombrophilous forest (Roda and Santos 2005), and like most of the PEC fragments it has been through selective logging during the 1970s and 1980s, but today the area is represented mainly by forest tracts in middle to late regeneration stages, with emergent trees already noticeable (Roda and Santos 2005, Pereira et al. 2014, 2016). According to the Köppen climate classification system, the climate is tropical with a well-defined dry season (AS’). Rains are concentrated from February through September, and the dry season occurs from October through January. Average annual precipitation varies from 1600 to 1700 mm, and average minimum and maximum temperatures vary from 21–22 °C and 30–31 °C, respectively (Roda and Santos 2005, Barros et al. 2012).

Selected bird species

Previously classified as a subspecies of P. leuconota, the Pernambuco Fire-eye (Thamnophilidae), is recognized as a full species (Pacheco et al. 2021), endemic to the PEC and occurring in the states of Alagoas, Pernambuco, and Paraíba. It is forest dependent, inhabiting the low forest strata, including borders and areas in early regeneration stages, but it is assumed to be reluctant to disperse between fragments (ICMBio 2018). It forages on arthropods, often in the leaf litter, and it is among the facultative ant-follower birds (Zimmer and Isler 2020, Martínez et al. 2021).

The Black-cheeked Gnateater, C. m. nigrifrons (Conopophagidae), is a subspecies of the widely distributed C. melanops, and it is restricted to the PEC, in the states of Alagoas, Pernambuco, and Paraíba (Whitney and de Juana 2020). It is sedentary and dependent on well-preserved forests, although it can be seen near borders and regenerating areas (ICMBio 2018). Like the Pernambuco Fire-eye, the Black-cheeked Gnateater also forages on the forest floor and it can follow army ants to capture arthropods (ICMBio 2018, Whitney and de Juana 2020). Both the Pernambuco Fire-eye and the Black-cheeked Gnateater are listed as Vulnerable (VU) in the Brazilian Red List of Endangered Fauna (ICMBio 2018).

Population density and census estimates

Population densities of the Pernambuco Fire-eye and of the Black-cheeked Gnateater were estimated using a line-transects distance sampling approach. In this method, the perpendicular distances between detected individuals and the transect line are estimated, and the set of measures obtained along a transect is used for the delimitation of the sampling area and for estimating the number of individuals within it, following the key assumption that animal detectability decreases with increasing distance from the transect line (Buckland et al. 1993, 2001). Seven line-transects with lengths varying from 0.5 to 2.17 km were distributed in the interior of the forest fragment, totaling 6.36 km (Fig. 1).

Data were collected from October 2019 to January 2020 to increase detectability. These months correspond to the dry season in the region, in which the days are longer and the birds vocalize more intensely. Each transect was walked 10 times (at intervals of at least 10 days) at an average speed of 1.5–2 km/h, always during the first hours after sunrise (4:30–9:00 am). While walking transects, we detected the birds either visually or audibly and measured the perpendicular distance between the initial detection point of each individual and the transect line with a laser measuring tool (Stanley tlm100) to improve data accuracy. Data of all individuals, independently of sex and age, were collected without truncation distance by a single observer (LCP) trained in Pernambuco Fire-eye and Black-cheeked Gnateater identification, and both sides of each transect were considered.

Minimum global population size estimates

We performed a literature review of bird surveys carried out in the PEC fragments and we multiplied our population density estimates by the size of the areas with records of the Pernambuco Fire-eye and of the Black-cheeked Gnateater to obtain the first rough global minimum population size estimates for these taxa. We considered them to be minimum estimates because surveys were not conducted in many other fragments in which these taxa could occur. The literature review was performed by searches in Google Scholar and Web of Science index databases, using the following key words: birds, survey, communities, diversity, Atlantic Forest, and Pernambuco Endemism Center. Only articles published in scientific journals, technical governmental reports, theses, and dissertations were considered.

Microhabitat selection

To address microhabitat selection in the two studied taxa, we compared a set of environmental parameters between sites in which birds were observed and sites chosen at random in the study area. Specifically, we aimed to address microhabitat selection during the period of potential resource scarcity (the dry season). This is because the availability of certain microhabitats may be more important for birds’ survival during this period than during the rainy season, in which arthropods may be more abundant (Devries and Walla 2001, Coutinho-Silva et al. 2017). Therefore, microhabitat data collection also occurred from October 2020 to January 2021, which corresponds to the dry season in the region.

The borders of PNHR Mata do Cedro were delimited using Google base maps in QGIS v. 3.8.1 (QGIS 2020), and random sites were generated throughout the whole forest fragment using “spsample” function from the R-package sp (Pebesma and Bivand 2005). These random points were accessed by following the shortest routes created by a GPS (Garmin eTrex® 20x), departing from the forest border or from one of the trails. Sites used by birds were located when observers were seeking the random points, to ensure that they were also searched at random across the whole area. Random points and locations where individuals were first observed were adopted as the centroids of 5 m radius plots. Flying individuals were not considered, meaning that most birds were foraging or defending territories, and when birds were in pairs or small groups only the location of the first observed individual was considered.

At “use” and “random” plots we measured the following biotic microhabitat parameters: canopy openness, ground leaf litter depth, number of trees or saplings with diameter at breast height (DBH) < 10 cm, number of trees with DBH > 10 cm, presence/absence of lianas, presence/absence of fallen trees or branches, vegetation density 0.5 m above ground, vegetation density 1.5 m above ground, canopy height, and canopy biomass. Further, for abiotic parameters we estimated elevation and declivity. The use of these variables is justifiable because (1) canopy openness is an index of light penetration and many tropical forest birds can be dependent on low-light conditions and are more likely to disappear when forest fragmentation and degradation cause increased light intensity (Patten and Smith-Patten 2012, Pollock et al. 2015); (2) leaf litter depth is positively associated to the presence of certain insectivorous birds that forage on the ground in the Amazon forest (Cintra and Cancelli 2008), but it is negatively correlated to the presence of other taxa in regenerating areas (Stratford and Stouffer 2013); (3) increased densities of small trees and non-wood vegetation, typical of regenerating areas, reduce the occurrence of certain Amazonian ground insectivorous birds, likely because the density compromises their movements (Stratford and Stouffer 2013), whereas other species benefit from forest gaps where vegetation density is higher and increased light incidence favors primary productivity and the abundance of leaf-consumer insects (Banks-Leite and Cintra 2008). It is worth noting that the Pernambuco Fire-eye and the Black-cheeked Gnateater forage primarily on the ground, but they also capture insects in forest understory foliage (ICMBio 2018, Whitney and de Juana 2020) in such a way that vegetation density also could affect their movements or foraging efficiency; (4) canopy height, canopy biomass, and the presence of fallen trees or branches are indexes of forest succession stage (Berveglieri et al. 2021) and can be additional indications of whether the studied taxa avoid or select regenerating areas; and (5) elevation and declivity can be associated with soil parameters, affecting foraging conditions and the abundance of food for species foraging on the ground, with elevation being demonstrated to affect the occurrence of at least one Amazon forest insectivorous bird (Cintra and Cancelli 2008).

Canopy cover was estimated by only one of the researchers (LCP) at the center of the plots using a spherical densiometer (Convex Model-A, Forest Suppliers, Inc., Jackson, MS, USA). The equipment was positioned in the four cardinal directions and the number of illuminated grids was averaged to estimate canopy openness following the manufacturer instructions (Lemmon 1957). Canopy height was estimated by a trained observer (LCP) with the laser measuring tool (Stanley tlm 100 - 30 mt) to improve measurement accuracy. The leaf litter depth was estimated in centimeters (measured with a ruler) in eight different points: in the four cardinal directions at the center of the plots and at the four cardinal directions 1 m from the center of the plots (see also Alves et al. 2017). Values were averaged to obtain the estimation of leaf litter depth in the plot. The number of trees or saplings with DBH > 10 cm and the number with DBH < 10 cm at 5 m radius were estimated by direct counting, and lianas and fallen branches/trees were considered as dummy variables (presence/absence). Vegetation density at 0.5 m and 1.5 m above ground were measured with a 5 m metal rod subdivided into 50 intervals of 10 cm. Then, vegetation density was estimated as the total number of intervals touched by vegetation in the four cardinal directions. Forest canopy biomass at “use” and “random” plots was assessed using remote sensing through an NDVI (Normalized Difference Vegetation Index) layer derived from a Sentinel-2 Level-1C (L1C) MS satellite scene, with 10 m spatial resolution (dated from 19 March 2019). The NDVI estimate for each “use” or “random” point was the value of the 10 x 10 m pixel in which the point fell. NDVI provides an index of forest biomass and is widely used to describe habitat productivity and biomass availability (Riedel et al. 2005, Hansen et al. 2009). The NDVI layer was generated using the band equation from Lange et al. (2017):

(1)

The values were estimated and extracted for each point using the “extract” function of the R-package raster (Hijmans et al. 2021) in R (R Core Team 2020). Elevation and declivity layers were obtained from the Brazilian TOPODATA project (Instituto Nacional de Pesquisas Espaciais - INPE; http://www.webmapit.com.br/inpe/topodata/), and were also extracted for each point using the “extract” function from raster.

Statistical analyses

Population density estimates were obtained by the model selection procedure implemented in the software Distance 7.0 (Thomas et al. 2010). In this approach, a set of pre-defined key detection functions (uniform, hazard-rate, half-normal, and negative exponential key functions) is used to model how the probability of detection decreases with increasing perpendicular distances, which is then used to obtain corrected population density estimates (Buckland et al. 1993, 2001, Thomas et al. 2010). Each detection function was run with cosine, hermite polynomial, and simple polynomial adjustments, resulting in 12 different models. The efficiency of each model to fit the data was assessed by (1) Akaike’s Information Criteria (AIC), (2) the goodness-of-fit test of Kolmogorov-Smirnov, and (3) visual inspection of quantile-quantile plots (Q-q plots). Because of the small sample sizes expected for endangered species restricted to small areas, the different transects were considered as a unique longer transect for the statistical analyses, and all the records obtained in transect replicates were pooled together, which permitted a more accurate modeling (see also Bernardo et al. 2011). The final population density estimate was then divided by 10 (the number of replicates) to report the actual number of individuals per hectare. The coefficients of variation (CV) and lower and upper 95% confidence intervals associated with population density estimates were obtained by the analytical approach, using default parameters settings.

To test microhabitat preferences, “use” and “random” plots were compared by using generalized linear models (GLM) with binomial distribution (use = 1 and random = 0) and logit-link function. All of the non-categorical explanatory variables were standardized using z-score and tested for correlations (Pearson’s r ≥ 0.6). Model parameter estimates, their standard errors (SE), and values of z-tests used to verify the levels of significance of each variable within the models were presented. These statistical procedures were conducted using the software R (R Core Team 2020).

RESULTS

Population density estimates

For both taxa, the best-fitted models had cosine adjustments, and because hermite polynomial and simple polynomial adjustments resulted in only slight variations in the estimates derived from each key function, in Table 1 we report only the results associated with the cosine adjustments. For the Pernambuco Fire-eye we recorded 83 perpendicular distances, varying from 0.5 to 65 m (25.20 ± 17.49 m). The detection function uniform cosine presented the lower ∆AIC, a low CV, and the less significant goodness-of-fit test (Table 1). Further, the graphic of the detection probability versus perpendicular distances (Fig. A1.1) provided visual evidence for a good model fit and the Q-q plot showed no remarkable deviations (Fig. A1.2). Using this selected model, the corrected population density (divided by 10) in the study area was 0.15 individuals/ha, and the estimated population size for the study area was 147 individuals.

For the Black-cheeked Gnateater we recorded 167 perpendicular distances varying from 0.5 to 80 m (22.84 ± 15.61 m). For this taxon, the half-normal detection function presented ∆AIC = 0, but the goodness-of-fit test was marginally significant for this model. Then, we opted for choosing hazard-rate cosine to estimate population density, because it had only a slightly larger ∆AIC value (0.32) and it presented the less significant goodness-of-fit probability (Table 1). Despite the lower concentration of detections in the first band (Fig. A1.3), we did not detect noticeable deviances in the Q-q plot (Fig. A1.4), upper and lower 95% confidence limits were narrow, and the CV was low (Table 1). By using the selected model, the corrected population density was 0.35 individuals/ha, with an estimated population size of 342 individuals for the study area.

Minimum global population size estimates

In our literature review, each taxon was recorded in 39 areas. Despite the equal numbers of areas with records of these taxa, the fragments in which each taxon occurred were not necessarily the same. For both species, the size of fragments varied from 25 to 10,074 ha. For the Pernambuco Fire-eye the areas totaled 32,904 ha, and for the Black-cheeked Gnateater they totaled 36,226 ha. The extrapolation of our population density estimates to these areas resulted in population sizes of 4936 individuals for the Pernambuco Fire-eye and 12,679 individuals for the Black-cheeked Gnateater.

Microhabitat selection

In total, we collected microhabitat parameters from 28 “use” plots for the Pernambuco Fire-eye, 34 plots for the Black-cheeked Gnateater, and from 54 random plots (Fig. 1). Because our plots covered a wide range of the area, multiple sampling of the same individual was unlikely to occur. Values or frequencies of the parameters are shown in Table 2. For both species, the z-score transformed numerical variables elevation and declivity were correlated (Tables A2.1 and A2.2). GLM analyses were then partitioned to address these parameters in separate models (see Table 3). For the Pernambuco Fire-eye, in the model including elevation, vegetation density 1.5 m above ground was positively correlated to the presence of the species, and in the model containing declivity, vegetation densities at both 1.5 m and 50 cm above ground presented significant positive correlations (Table 3). For the Black-cheeked Gnateater, in both model partitions the presence of lianas and vegetation density 1.5 m above ground was positively correlated to the probability of occurrence (Table 3).

DISCUSSION

Population densities and minimum global population size estimates

Population density can be among the main parameters affecting the level of threat to a taxon, especially when it is associated with limited geographic distribution (Goerck 1997, Monroy-Ojeda et al. 2018, Birskis-Barros et al. 2019). However, population density estimates based on distance sampling methods are scarce for both endangered and non-endangered Neotropical forest passerines. For the near-threatened Southern Bristle-Tyrant, Phylloscartes eximius (Tyrannidae), from the Atlantic Forest of southeastern Brazil, distance-sampling estimated population density was 0.13 individuals/ha, a value that was considered very low when compared to unpublished data on another congener (Tonetti and Pizo 2016). For the endangered Black-cheeked Ant-Tanager, Habia atrimaxillaris (Cardinalidae), from tropical forests of Costa Rica, population densities varied from 0.24 to 0.27 individuals/ha (Cornils et al. 2015). The population density of the Pernambuco Fire-eye (0.15 individuals/ha) was relatively low, and the value obtained for the Black-cheeked Gnateater (0.35 individuals/ha) relatively high, when compared to those other endangered tropical forest passerines. Despite the limited amount of data available for comparisons, the population density of the Black-cheeked Gnateater being more than twice that found for the Pernambuco Fire-eye reveals that bird species from the PEC, included in the Brazilian Red List solely on the basis of habitat availability information (ICMBio 2018), can have highly divergent population densities and likely different risks of extinction, even when classified in the same threat category.

For the Black-cheeked Gnateater, the graphic of detection probability versus perpendicular distances presents a remarkably lower concentration of detections in the first detection band and an increased detection probability in the second band. It reveals higher impacts of the observer or trails on this species than on the Pernambuco Fire-eye. This type of residual is not rare in transect-based distance sampling approaches (see also Cornils et al. 2015), but it is often absent in point count-based techniques because the latter permit the observer to cause less disturbance during data collection (see Tonetti and Pizo 2016). In the present case, however, the use of line transects was justifiable because in preliminary samplings (data not shown), point-counts distance-sampling resulted in too few records to reach the minimum number of 50 detections needed to obtain good model fit (Buckland et al. 1993, 2001, Thomas et al. 2010). It is worth noting, however, that the other parameters used to assess model fit suggested that this problem has not caused a relevant impact in the final population density estimate.

Extrapolating data from only one fragment to other areas of occurrence of the taxa was problematic for two main reasons. First, many other PEC fragments in which the Black-cheeked Gnateater and the Pernambuco Fire-eye could occur were never surveyed and their areas were not included in our global population size extrapolations, meaning that we may have generated only minimum population size estimates. Second, population densities can certainly vary among the fragments in which these taxa were recorded because of microhabitat specificities (see below), which may have been another important source of biases. Therefore, the global population sizes we generated are only rough approximations based on primary field data, but at the moment these are the only demographic information available for these taxa. Based on these data, none of the two taxa we addressed were eligible to be uplisted to endangered category (EN) according to the IUCN criterion C. According to this criterion, EN taxa must present a global population smaller than 2500 mature individuals, which is well below our estimates. We also applied our data to the C2a(i) criterion, which refers to the numbers of individuals present within subpopulations, a criterion that is highly applicable to taxa inhabiting fragmented habitats like the PEC. According to this criterion, VU taxa are distributed in populations presenting 250 to 1000 individuals, whereas endangered taxa should have no populations with more than 250 individuals. Based on our population density estimate, an area of approximately 1600 ha would be needed to retain a population of 250 individuals of the Pernambuco Fire-eye, the species with the lowest population density. In our literature review we found at least three areas above this size with reliable records of this species, including one area of about 10,000 ha, suggesting that the EN category is also not applicable on the basis of this criterion.

Microhabitat selection

The presence of these two insectivore forest understory birds is associated with increased forest understory vegetation density, rather than with characteristics of forest canopy, leaf litter, declivity, elevation, or local biomass concentration. In tropical forests, areas with denser understory vegetation can be associated with higher abundance of herbivorous insects (Richards and Windsor 2007, Banks-Leite and Cintra 2008). Because the Pernambuco Fire-eye and the Black-cheeked Gnateater forage both on the ground and in forest understory foliage, the positive relationships between the presence of these taxa and forest understory density may be explained by increased food availability (see also Richards and Windsor 2007, Banks-Leite and Cintra 2008). It is worth noting that our study was conducted during the dry season, when the abundance of insects is often reduced (Devries and Walla 2001, Coutinho-Silva et al. 2017), suggesting that the selected microhabitats may be important to increase foraging efficiency during this period of potential food scarcity.

Our data did not permit us to infer whether these birds were selecting recently formed gaps to forage because we have not addressed the effects of gaps directly, and because canopy openness did not diverge between “use” and “random” points, likely because of reduced sample sizes. However, the associations of these birds to increased forest understory density suggests that they have selected sites where the vegetation was at some stage of regeneration. These findings were consistent with the scattered observations that the Black-cheeked Gnateater and the Pernambuco Fire-eye can occur near forest gaps or borders (ICMBio 2018, Whitney and de Juana 2020). Although they are forest dependent, they may tolerate or even benefit from certain levels of forest disturbance, but the latter is still to be confirmed.

Microhabitat preference divergences were observed between the two addressed taxa. The selection of sites with increased vegetation density at 1.5 m above ground and with lianas by the Black-cheeked Gnateater suggested that this bird selected sites in more advanced levels of vegetation regeneration than the Pernambuco Fire-eye did. The latter occurred in locations with denser understory vegetation even near the ground, where the presence of herbaceous plants indicated a shorter time had elapsed since the occurrence of natural or anthropogenic forest disturbances. These data also contributed to show how Neotropical forest insectivores can respond differently to microhabitat parameters. In the Amazon forest, for instance, increased forest understory density was among the variables accounting for the absence of nine ground insectivorous birds in second-growth areas (Stratford and Stouffer 2013), a result that diverged from our findings.

CONCLUSION

Our data reinforce the previous evidence that microhabitats in Neotropical forests can be heterogeneous and that different insectivore bird species can show preferences for specific habitat parameters (Cintra and Cancelli 2008). We provide evidence that the logging activities that occurred in the past may have contributed to an increase in availability of habitats selected by the Pernambuco Fire-eye and the Black-cheeked Gnateater. As a consequence, their population densities could be temporarily inflated in some forest fragments and could potentially decrease as forest fragments reach more advanced regeneration stages, and their populations should be monitored in the future. It does not reduce, however, the importance of other factors that put these taxa in risk of extinction, such as their limited geographic distributions and population isolation.

Tropical forest understory insectivorous birds are of special concern to conservation practitioners, and in regions like the PEC (already highly fragmented) the endangered taxa will likely persist only with intensive management actions. Knowledge of habitat requirements and census estimates of the most vulnerable taxa are urgently needed. Our study is among the very few providing these types of information for birds of this guild in the Neotropics. Although our estimates are only rough, they will serve as comparative data for other studies and may contribute to practical conservation policies.

LITERATURE CITED

Alves, F., G. M. Lópes-Iborra, D. Stojanovic, and L. F. Silveira. 2017. Habitat selection by the endangered Red-billed Curassow (Crax blumenbachii) in an Atlantic forest remnant. Emu 117(4):316-324. https://doi.org/10.1080/01584197.2017.1326010

Banks-Leite, C., and R. Cintra. 2008. The heterogeneity of Amazonian treefall gaps and bird community composition. Ecotropica 14:1-13.

Barros, A. H. C., J. C. de Araújo Filho, A. B. da Silva, and G. A. C. F. Santiago. 2012. Climatologia do estado de Alagoas. Embrapa Solos, Recife, Brazil.

Bernardo, C. S. S., P. Rubim, R. S. Bueno, R. A. Begotti, F. Meirelles, C. Donatti, C. Denzin, C. E. Steffler, R. M. Marques, R. S. Bovendorp, S. K. Gobbo, and M. Galetti. 2011. Density estimates of the Black-fronted Piping Guan in the Brazilian Atlantic rainforest. Wilson Journal of Ornithology 123(4):690-698. https://doi.org/10.1676/10-140.1

Berveglieri, A., N. N. Imai, L. E. Christovam, M. L. B. T. Galo, A. M. G. Tommaselli, and E. Honkavaara. 2021. Analysis of trends and changes in the successional trajectories of tropical forest using the Landsat NDVI time series. Remote Sensing Applications: Society and Environment 24:10062. https://doi.org/10.1016/j.rsase.2021.100622

Birskis-Barros, I., L. R. V. Alencar, P. I. Prado, M. Böhm, and M. Martins. 2019. Ecological and conservation correlates of rarity in New World Pitvipers. Diversity 11(9):147. https://doi.org/10.3390/d11090147

Buckland, S. T., D. R. Anderson, K. P. Burnham, and J. L. Laake. 1993. Distance sampling: estimating abundance of biological populations. Chapman and Hall, London, UK.

Buckland, S. T., D. R. Anderson, K. P. Burnham, J. L. Laake, D. L. Bochers, and L. Thomas. 2001. Introduction to distance sampling: estimating abundance of biological populations. Oxford University Press, New York, New York, USA.

Cintra, R., and J. Cancelli. 2008. Effects of forest heterogeneity on occurrence and abundance of the Scale-backed Antbird, Hylophylax poecilinotus (Aves: Thamnophilidae), in the Amazon forest. Revista Brasileira de Zoologia 25(4):630-639. https://doi.org/10.1590/S0101-81752008000400008

Cornils, J. S., I. Riedl, J. Fricke, M. Katz, and C. H. Schulze. 2015. Population density and habitat preferences of the Black-cheeked Ant-tanager Habia atrimaxillaris. Bird Conservation International 25(3):306-321. https://doi.org/10.1017/S0959270914000215

Coutinho-Silva, R. D., M. A. Montes, G. F. Oliveira, F. G. de Carvalho-Neto, C. Rohde, and A. C. L. Garcia. 2017. Effects of seasonality on drosophilids (Insecta, Diptera) in the northern part of the Atlantic Forest, Brazil. Bulletin of Entomological Research 107(5):634-644. https://doi.org/10.1017/S0007485317000190

de Camargo, C., H. L. Gibbs, M. C. Costa, G. Del-Rio, L. F. Silveira, A. P. Wasko, and M. R. Francisco. 2015. Marshes as “mountain tops”: genetic analyses of the critically endangered São Paulo Marsh Antwren (Aves: Thamnophilidae). PLoS ONE 10(10):e0140145. https://doi.org/10.1371/journal.pone.0140145

de Oliveira, Jr., P. R. R., C. C. Alberts, and M. R. Francisco. 2011. Impact of road clearings on the movements of three understory insectivorous bird species in the Brazilian Atlantic Forest. Biotropica 43(5):628-632. https://doi.org/10.1111/j.1744-7429.2010.00744.x

Devries, P. J., and T. R. Walla. 2001. Species diversity and community structure in neotropical fruit-feeding butterflies. Biological Journal of the Linnean Society 74(1):1-15. https://doi.org/10.1111/j.1095-8312.2001.tb01372.x

Dickens, J. K., P. P. Bitton, G. A. Bravo, and L. F. Silveira. 2021. Species limits, patterns of secondary contact and a new species in the Trogon rufus complex (Aves: Trogonidae). Zoological Journal of the Linnean Society 193(2):499-540. https://doi.org/10.1093/zoolinnean/zlaa169

Francisco, M. R., M. C. Costa, R. M. A. Azeredo, J. G. P. Simpson, T. C. Dias, A. Fonseca, F. J. M. Pinto, and L. F. Silveira. 2020. Recovered after an extreme bottleneck and saved by ex situ management: lessons from the Alagoas curassow (Pauxi mitu [Linnaeus, 1766]; Aves, Galliformes, Cracidae). Zoo Biology 40:76-78. https://doi.org/10.1002/zoo.21577

Garbino, G. S. T., G. C. Rezende, H. Fernandes-Ferreira, and A. Feijó. 2018. Reconsidering mammal extinctions in the Pernambuco Endemism Center of the Brazilian Atlantic Forest. Animal Biodiversity and Conservation 41:175-184. https://doi.org/10.32800/abc.2018.41.0175

Goerck, J. M. 1997. Patterns of rarity in the birds of the Atlantic Forest of Brazil. Conservation Biology 11:112-118. https://doi.org/10.1046/j.1523-1739.1997.95314.x

Hansen, B. B., I. Herfindal, R. Aanes, B.-E. Saether, and S. Henriksen. 2009. Functional response in habitat selection and the tradeoffs between foraging niche components in a large herbivore. Oikos 118:859-872. https://doi.org/10.1111/j.1600-0706.2009.17098.x

Hijmans, R. J., J. van Etten, M. Mattiuzzi, M. Sumner, J. A. Greenberg, O. P. Lamigueiro, and A. Shortridge. 2021. Raster package in R. https://rspatial.org/raster/pkg/RasterPackage.pdf

Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio). 2018. Livro vermelho da fauna brasileira ameaçada de extinção: Volume III - Aves. ICMBio/MMA, Brasília, DF, Brazil. https://www.gov.br/icmbio/pt-br/centrais-de-conteudo/publicacoes/publicacoes-diversas/livro_vermelho_2018_vol3.pdf

International Union for Conservation of Nature and Natural Resources (IUCN). 2022. IUCN standards and petitions committee. Guidelines for using the IUCN red list categories and criteria. Version 15.1. https://nc.iucnredlist.org/redlist/content/attachment_files/RedListGuidelines.pdf

Lange, M., B. Dechant, C. Rebmann, M. Vohland, M. Cuntz, and D. Doktor. 2017. Validating MODIS and sentinel-2 NDVI products at a temperate deciduous forest site using two independent ground-based sensors. Sensors 17(8):1855. https://doi.org/10.3390/s17081855

Lees, A. C., and S. L. Pimm. 2015. Species, extinct before we know them? Current Biology 25(5):R177-R180. https://doi.org/10.1016/j.cub.2014.12.017

Lemmon, P. E. 1957. A new instrument for measuring forest overstorey density. Journal of Forest Research 55:667-669.

Machado, R. B., L. F. Silveira, M. I. S. G. da Silva, F. K. Ubaid, C. A. Medolago, M. R. Francisco, and J. C. Dianese. 2020. Reintroduction of songbirds from captivity: the case of the Great-billed Seed-finch (Sporophila maximiliani) in Brazil. Biodiversity and Conservation 29:1613-1636. https://doi.org/10.1007/s10531-019-01830-8

Martínez, A. E., H. S. Pollock, P. F. Rodrigues, and J. M. Touchton. 2021. Army-ant following in Neotropical birds: a review and prospectus. Ornithology 138(1):1-16. https://doi.org/10.1093/ornithology/ukaa078

Mendes Pontes, A. R., A. C. M. Beltrão, I. C. Normande, A. de J. R. Malta, A. P. da Silva Júnior, and A. M. M. Santos. 2016. Mass extinction and the disappearance of unknown mammal species: scenario and perspectives of a biodiversity hotspot’s hotspot. PLoS ONE 11(5):e0150887. https://doi.org/10.1371/journal.pone.0150887

Monroy-Ojeda, A., M. Grosselet, A. Panjabi, D. Pool, C. White, and G. Ruiz-Michael. 2018. Density and population estimates of Rose-bellied Bunting (Passerina rositae), Cinnamon-tailed Sparrow (Peucaea sumichrasti), and Orange-breasted Bunting (Passerina lechlancherii) in the Sierra Tolistoque, Oaxaca, Mexico. Wilson Journal of Ornithology 130(2):353-361. https://doi.org/10.1676/16-217.1

Moore, R. P., W. D. Robinson, I. J. Lovette, and T. R. Robinson. 2008. Experimental evidence for extreme dispersal limitation in tropical forest birds. Ecology Letters 11(9):960-968. https://doi.org/10.1111/j.1461-0248.2008.01196.x

Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, and J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403:853-858. https://doi.org/10.1038/35002501

Pacheco, J. F., L. F. Silveira, A. Aleixo, C. E. Agne, G. A. Bencke, et al. 2021. Annotated checklist of the birds of Brazil by the Brazilian Ornithological Records Committee. Second edition. Ornithology Research 29:94-105. https://doi.org/10.1007/s43388-021-00058-x

Patten, M. A., and B. D. Smith-Patten. 2012. Testing the microclimate hypothesis: light environment and population trends of Neotropical birds. Biological Conservation 155:85-93. http://dx.doi.org/10.1016/j.biocon.2012.06.004

Pebesma, E., and R. S. Bivand. 2005. S classes and methods for spatial data: the sp package. R news 5:9-13.

Pereira, G. A., H. F. P. Araújo, and S. M. Azevedo-Júnior. 2016. Distribution and conservation of three important bird groups of the Atlantic Forest in north-east Brazil. Brazilian Journal of Biology 76(4):1004-1020. https://doi.org/10.1590/1519-6984.06815

Pereira, G. A., S. de M. Dantas, L. F. Silveira, S. A. Roda, C. Albano, F. A. Sonntag, S. Leal, M. C. Periquito, G. B. Malacco, and A. C. Lees. 2014. Status of the globally threatened forest birds of northeast Brazil. Papéis Avulsos de Zoologia 54(14):177-194. https://doi.org/10.1590/0031-1049.2014.54.14

Pollock, H. S., Z. A. Cheviron, T. J. Agin, and J. D. Brawn. 2015. Absence of microclimate selectivity in insectivorous birds of the Neotropical forest understory. Biological Conservation 188:116-125. https://doi.org/10.1016/j.biocon.2014.11.013

Powell, L. L., N. J. Cordeiro, and J. A. Stratford. 2015. Ecology and conservation of avian insectivores of the rainforest understory: a pantropical perspective. Biological Conservation 188:1-10. https://doi.org/10.1016/j.biocon.2015.03.025

QGIS Development Team. 2020. QGIS: a free and open source geographic information system. Open Source Geospatial Foundation. http://qgis.org

R Core Team. 2020. R: A language and environment for statistical computing. R version 4.0.2. https://www.r-project.org/

Ribeiro, M. C., J. P. Metzger, A. C. Martensen, F. J. Ponzoni, and M. M. Hirota. 2009. The Brazilian Atlantic Forest: how much is left, and how is the remaining forest distributed? Implications for conservation. Biological Conservation 142(6):1141-1153. https://doi.org/10.1016/j.biocon.2009.02.021

Richards, L. A., and D. M. Windsor. 2007. Seasonal variation of arthropod abundance in gaps and the understorey of a lowland moist forest in Panama. Journal of Tropical Ecology 23(2):169-176. https://doi.org/10.1017/S0266467406003907

Riedel, S. M., H. E. Epstein, D. A. Walker, D. L. Richardson, M. P. Calef, E. Edwards, and A. Moody. 2005. Spatial and temporal heterogeneity of vegetation properties among four tundra plant communities at Ivotuk, Alaska, USA. Arctic, Antarctic, and Alpine Research 37(1):25-33. https://doi.org/10.1657/1523-0430(2005)037[0025:SATHOV]2.0.CO;2

Roda, S. A., and A. M. M. Santos. 2005. Avaliação de fragmentos florestais para uma possível reintrodução do mutum-de-alagoas em seu ambiente natural. Centro de Pesquisas Ambientais do Nordeste - CEPAN, Recife, Pernambuco, Brazil.

Silveira, L. F., F. Olmos, and A. J. Long. 2003. Birds in Atlantic Forest fragments in north-east Brazil. Cotinga 20:32-46. https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.498.2742&rep=rep1&type=pdf

Sinclair, A. R. E., J. M. Fryxell, and G. Caughley. 2006. Wildlife ecology, conservation, and management. Second edition. Blackwell, Oxford, UK.

Stratford, J. A., and P. C. Stouffer. 2013. Microhabitat associations of terrestrial insectivorous birds in Amazonian rainforest and second-growth forests. Journal of Field Ornithology 84(1):1-12. https://doi.org/10.1111/jofo.12000

Tabarelli, M., J. A. de Siqueira Filho, and A. M. M. Santos. 2006. Conservação da Floresta Atlântica ao Norte do Rio São Francisco. Pages 41-48 in K. C. Pôrto, J. S. Almeida-Cortez, and M. Tabarelli, editors. Diversidade Biológica e Conservação da Floresta Atlântica ao Norte do Rio São Francisco. Ministério do Meio Ambiente, Brasília, Brazil.

Thomas, L., S. T. Buckland, E. A. Rexstad, J. L. Laake, S. Strindberg, et al. 2010. Distance software: design and analysis of distance sampling surveys for estimating population size. Journal of Applied Ecology 47:5-14. https://doi.org/10.1111/j.1365-2664.2009.01737.x

Tonetti, V. R., and M. A. Pizo. 2016. Density and microhabitat preference of the Southern Bristle-Tyrant (Phylloscartes eximius): Conservation policy implications. Condor 118(4):791-803. https://doi.org/10.1650/CONDOR-16-89.1

Vié, J.-C., C. Hilton-Taylor, and S. N. Stuart, editors. 2009. Wildlife in a changing world: an analysis of the 2008 IUCN Red List of threatened species. International Union for Conservation of Nature, Gland, Switzerland. https://doi.org/10.2305/IUCN.CH.2009.17.en

Whitney, B., and E. de Juana. 2020. Black-cheeked Gnateater (Conopophaga melanops), version 1.0. In J. del Hoyo, A. Elliott, J. Sargatal, D. A. Christie, and E. de Juana, editors. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.blcgna1.01

Zimmer, K., and M. L. Isler. 2020. White-backed Fire-eye (Pyriglena leuconota), version 1.0. In J. del Hoyo, A. Elliott, J. Sargatal, D. A. Christie, and E. de Juana, editors. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.wbfeye1.01