Bird diversity in ‘paricá’ (Schizolobium amazonicum Huber ex Ducke) plantations and forest fragments in Eastern Amazon: taxonomic diversity, ecological guilds, and functional trait composition

ABSTRACT Monoculture may drastically reduce biodiversity and change communities’ structure due to habitat simplification. Plantations of Schizolobium amazonicum Huber ex Ducke (‘paricá’) has been expanding in the Amazon region due to the increased demand for veneer and plywood production. Our objective was to assess whether bird communities in paricá plantations and forested areas differ in terms of composition, richness, abundance, trophic structure, and functional trait composition. This study was conducted in young paricá plantations (1 to 5 years) and forest fragments in eastern Amazon, state of Pará, Brazil. We sampled birds in five plantation sites (each paired with a nearby forested area) using both the point-count method and mist nets. We recorded 208 bird species, 132 in forest fragments and 94 in plantations. Comparatively, habitats showed distinct species composition and trait composition, with few shared species (n = 18). Paricá plantations did not support forest species, while forest fragments harbored specialized bird communities that are more sensitive to environmental change as well as endangered species. In this way, the protection of these forested areas surrounding paricá plantations is a priority for forest biodiversity conservation, given the rapid deforestation in the Brazilian Amazon.


Introduction
Currently, one of the major threats to tropical biodiversity is the conversion of natural forests into crops (Fitzherbert et al. 2008;Malhi et al. 2014;Neate-Clegg & Sekercioglu 2020). In 2019, planted forests corresponded to 9 million hectares in Brazil, represented mainly by the genera Eucalyptus (77%) and Pinus (18%) in the Southeast and South regions of the country, where the large industries are located (IBÁ 2020). In the Brazilian Amazon, most agricultural expansion has occurred in the southwestern region known as the 'arc of deforestation', a region that contributes with about 80% of the deforestation recorded in the Legal Amazon (Vieira et al. 2008;INPE 2020;Neate-Clegg & Sekercioglu 2020). In this region, the highest yielding crop is oil palm (Elaeis guineensis Jacq.), driven by the demand for low-cost vegetable oils for food-processing and biodiesel (Fitzherbert et al. 2008;Malhi et al. 2014;Neate-Clegg & Sekercioglu 2020). Also in the Amazon region, a native species commonly known as 'paricá' (Schizolobium amazonicum Huber ex. Ducke) is used by the tree planting industry mainly in the wood panel and laminate flooring segments (Marques et al. 2006). In 2018, the area occupied by paricá plantations in Brazil totaled 90.811 ha of the planted forests and concentrated mainly in the state of Pará (IBÁ 2019).
In forest environments, where vegetation structure is complex and vertical stratification of resources is high, species occupy a wide variety of niches, which consequently increases local diversity since these conditions allow species with different ecological features to coexist (MacArthur 1972;Terborgh et al. 1990;Tews et al. 2004). However, the expansion of agriculture, livestock, illegal logging, and mining activities modify the natural landscape in the Amazon region. These activities cause the suppression and fragmentation of primary forests, generating impacts such as massive losses of biodiversity due to forest conversion (e.g. for cattle pasture or monocultures), while the forest remnants retain increasingly impoverished communities CONTACT Fernanda de Carvalho Barros nandabarros18@gmail.com (Neate-Clegg & Sekercioglu 2020;Stouffer et al. 2021). In general, bird diversity is significantly reduced in monoculture areas due to habitat simplification, reducing the number of resources and leading to a systematic loss or even local extinction of many forest species (Barlow et al. 2007; Lees et al. 2015;Jacoboski et al. 2016;Almeida et al. 2016;Goded et al. 2019). However, other authors argue that plantations can accelerate natural forest regeneration on degraded lands and abandoned pastures, facilitating recolonization by native forest species (Carnus et al. 2006).
In addition to changes in species richness and composition, crops may alter the trophic structure of bird communities through the loss of diet-specialist species (e.g. understory insectivores and frugivores). On the other hand, more opportunistic species (e.g. omnivores and edge insectivores) increase in richness and abundance (Barlow et al. 2007;Almeida et al. 2016). The presence or absence of specific guilds may provide valuable information for the conservation and management of a given area (Sekercioglu 2012). Thus, it is also important to understand how the different facets of bird diversity are affected by monoculture. Diversity metrics that consider variation in ecological traits of communities across environmental gradients may be a complementary tool to the taxonomic approach. Morphological characteristics of bird species (e.g. body mass, bill shapes) can be related to the number and type of resources they exploit as well as their foraging strategies (Sekercioglu 2006;Luck et al. 2012).
In the Eastern Amazon, the landscape is characterized by a mosaic of primary and secondary forest fragments, logging areas, abandoned and active pastures, and crops (Villela et al. 2014;INPE 2020). Additionally, plantations of Schizolobium amazonicum Huber ex Ducke (Fabaceae) has also been expanding in this region, especially in the states of Pará and Maranhão. In a preliminary study conducted in eastern Amazon, Henriques (2003) evaluated the importance of paricá reforestation to bird communities and concluded that this crop holds little value for maintaining regional biological diversity since it has a low attractiveness to forest birds (Henriques 2003). However, this study did not compare the diversity of birds in the plantations with adjacent forested areas because 90% of the surrounding area consisted of active pasture. Here, we evaluated species richness, composition, and abundance, as well as trophic groups and functional trait composition of bird communities in paricá plantations and adjacent forest fragments in a degraded landscape located in the municipalities of Ipixuna do Pará and Paragominas, state of Pará, Brazil. We hypothesize that bird communities from the two environments differ in composition, species richness, trophic structure, and functional traits. Therefore, we predict: 1) higher species richness in forest areas as a result of both higher structural complexity of vegetation and resource availability; 2) different dominant species in forest fragments and plantations, with species composition differing between habitats; 3) given the degree of habitat specificity of many Amazonian forest birds, that forest fragments will harbor more diet-specialist birds, while plantations will harbor species that predominantly show a generalist diet since they are adapted to disturbed environments. Lastly, 4) considering that habitat structure and niche variability could differ between plantations and forested areas, that paricá plantations and forest fragments will differ in the functional composition of bird communities.

Study area
The study was conducted from 6 to 16 August 2014, in 10 transects distributed across five farms located in the municipalities of Paragominas (Flamboyant, Ilha Verde e Taquarussu Farms), and Ipixuna do Pará (Soberana and Brejeira Farms), state of Pará (Figure 1), Eastern Amazon. The area consists of a mixed landscape formed by a farmland-forest mosaic. In the past, these farms were used for livestock. The description of the farm areas can be found in the supplementary material (Table S1).
At each site, we sampled one forest fragment and one paricá plantation (Schizolobium amazonicum, hereafter referred to as plantation), totaling five forest sites and five plantation sites. The climate of the region is characterized by high temperature, with mean annual temperatures ranging from 22 to 33°C and mean relative air humidity of 81%. There is no pronounced dry season, although rain is less frequent between June and November, with mean annual precipitation below 2,000 mm (Bastos et al. 2006). The sampled forested sites are fragments of non-flooded 'terra firme' forests which can be characterized as lowland dense ombrophilous forest with a canopy of about 40 m in height. However, the company carries out low impact logging activities in the forest fragments, and we observed many clearings and a high abundance of lianas in these areas. In addition, the company recorded evidence of hunting and fishing in the fragments, indicating that they have been undergoing some degradation. The vegetation in paricá plantations consisted of individuals of S. amazonicum, presenting an open canopy, with plants ranging from 20 to 30 m in height. The understory of these plantations was predominantly occupied by herbaceous plants, mainly grasses (SMA pers. obs.).

Bird survey
We sampled birds across ten transects, five within paricá plantations and five in forest fragments ( Figure 1). We obtained quantitative data on bird abundance and species diversity through point counts. Along each transect, a single experienced observer conducted from 7 to 12 tenminute point counts between 06:00 AM and 10:00 AM. Census stations were located at intervals of 200 m across the transects. The transects were sampled on different days, each being sampled once. In Sites 1, 2, and 5, we conducted 8-point counts (forest and plantation) along 1400-m-long transects, one transect in each habitat type. Site 2 comprised two 1200-m-long transects with 7-point counts, one in forest and one in plantation areas, while in Site 4 we conducted 12 point counts along 2200-m-long transects, one in each habitat type. To avoid edge effects, point counts started at a distance of at least 500 m from the edge of the forest or plantation. Bird surveys were carried out only in the absence of rain or heavy mist. We recorded all individual birds seen or heard within a fixed 50-m radius. We performed 86-point counts, 43 in each habitat type. The same observer conducted all pointcount samplings. Simultaneously to the quantitative survey, we captured birds within each transect using 10-m-long mist nets (2.5 m high) because some Amazonian bird species are cryptic and difficult to detect. In each transect, two mist-net sequences were allocated 500 m apart from each other and arranged perpendicularly to the transect. Each sequence consisted of 10 mist nets (36-mm mesh) that were linearly set, covering an area of 250 m 2 in each plot. The mist nets remained open from 06:00 AM to 10:00 AM at each plot. The sampling effort in each transect was 2,000 m 2 .h −1 (sensu Straube & Bianconi 2002).

Bird functional traits
To assess the functional composition of bird communities, we used three functional traits (diet, foraging stratum, and body mass). Diet was expressed as the percentage use of each of the 8 considered categories of food items: 1-invertebrates; 2-mammals, birds; 3 -reptiles, amphibians; 4vertebrates in general or unknown (for species where it was not clear what kind of vertebrates they were eating), 5fruit, drupes; 6 -nectar, pollen, plant exudates, gums; 7seed, maize, nuts, spores, grains; 8 -other plant material. Foraging stratum was expressed as the estimated percentage use of each of the five considered strata: ground, understory, mid to high levels, canopy, aerial. Diet and foraging stratum were based on literature (Wilman et al. 2014). Additionally, we considered four morphological traits, viz. body mass, bill length, bill width, and tarsus length. Body mass indicates the amount and size of food required by a given individual (Luck et al. 2012). Bill shape is a morphological adaptation related to food access and type of consumed food item, influencing pollination effectiveness, and handling of fruit and seeds (Luck et al. 2012). The tarsus is related to grasping and perching ability and can influence foraging behavior (Luck et al. 2012). These morphological traits were obtained from the ornithological collection of the Museu Paraense Emilio Goeldi, in Belém, state of Pará, Brazil. Three male specimens of each species were measured using a digital 0.1-mm-precision caliper to obtain a mean value that we used for the analysis. A single observer took all the measurements. We removed the influence of differences in species' body mass on morphological traits (bill length, bill width, and tarsus length), which cause allometry in these traits. To accomplish this, the traits were transformed by dividing the trait value by the cube root of the respective species' body mass (West et al. 1997).

Data analysis
To estimate the abundance of species in each environment, we calculated the Punctual Index of Abundance (PIA), which represents the relative abundance of the species and is obtained from the ratio between the number of contacts with a given species and the total number of point counts (Donatelli et al. 2007;Vielliard et al. 2010). We ran paired sample t-tests to test for differences in species richness and the number of individuals between forest and paricá plantations considering the transect as the sample unit. For this purpose, individuals counted at the point counts (subsamples) of each transect were summed up. Levene's tests confirmed the homogeneity of variance across sampled areas for both metrics. We used iNEXT (Hsieh et al. 2016) to calculate interpolated and extrapolated species richness within each habitat type (using the richness estimators derived by Chao et al. (2014) and plotted these as species accumulation curves. Differences in species composition between natural areas and paricá plantation were evaluated using a Permutational Multivariate Analysis of Variance (PERMANOVA) with 999 permutations (Anderson 2001) through the function 'adonis' using Bray-Curtis dissimilarity index. To visualize patterns of differences in species composition between plantation and forested areas, we run a principal coordinate analysis (PCoA) with Bray-Curtis distance. Afterward, we investigated the relative contribution of individual species to the overall difference in composition using similarity percentages routine (SIMPER). This analysis calculates the contribution of each species (%) to the dissimilarity between the two groups (forest and plantation, in our study) using Bray-Curtis dissimilarities. We ran this analysis using the vegan package (Oksanen et al. 2020) in R environment.
To assess which functional traits are different between forest and plantation sites and understand the influence of community-level mean trait responses to habitat modification, we calculated community-weighted mean (CWM) values for each trait. A similar approach was used by Jacoboski et al. (2016) and Betancurt-Grisales et al. (2021). CWM trait values represent the mean of each trait weighted by the relative abundance of the species. The CWM can define the functional composition and informs us about dominant traits in the community (Duarte et al. 2018). To calculate CWM values, we used the 'matrix.t' function available in the 'SYNCSA' package (Debastiani & Pillar 2012). We plotted CWM for all traits using a principal component analysis (PCA) to interpret the dominant traits in plantation and forest sites. Then, to test for differences in trait composition (CWM values) between forest and plantation sites we ran a global test using Permutational Multivariate Analysis of Variance (PERMANOVA) in the vegan package, employing 999 permutations (Anderson et al. 2006).
We conducted all these analyses with R version 3.5.2 (R Core Team 2018) using data from the quantitative method (point counts).

Overview
We recorded 208 bird species belonging to 45 families and 17 orders, considering both point count and mist nets methods (Table S2). Of this total, we recorded 132 bird species in forest fragments, of which 114 were observed only in this habitat. In plantations, we found 94 species, of which 76 were recorded exclusively in this environment. Eighteen species occurred in both environments. Passerine birds (order Passeriformes) represented 129 species, while 79 species were from other orders. The most representative families were Thamnophilidae (19 species), Tyrannidae (18), Thraupidae (18), Psittacidae (12), and Trochilidae (11). Among the recorded species, the following eleven fall within some extinction threat category according to the Red Book of Threatened Brazilian Fauna (ICMBio 2018), the list of threatened species of the state of Pará (SEMAS 2007) and the IUCN's red list (IUCN 2020) (see Appendix S1 for more details): Psophia obscura, Guaruba guarouba, Pteroglossus bitorquatus bitorquatus, Pyrrhura lepida lepida, Phlegopsis nigromaculata paraensis, Thamnophilus aethiops incertus, Dendrocolaptes medius, Dendrocincla merula badia, Synallaxis rutilans omissa, Piprites chloris griseicens, Tangara velia signata. All of them were recorded in the forest fragments, and none was found in the paricá plantations.
The plantation areas had a significantly lower species richness (t = 2.78, df = 4, p = 0.04; Figure 2a). Species accumulation curves in both forest and plantation were starting to asymptote, indicating that we had sampled a high proportion of all species within each habitat (an estimated 87.10% of species in the forested areas, and 90.31% of species in the plantations) ( Figure 3). We observed different species composition when comparing paricá plantations with forest fragments (Pseudo-F = 11,03; p (perm) = 0.008, Figure 4). Only 13 species were recorded in both habitats using the point counts method.

Bird captures
We captured 314 individuals of 64 species using mist nets, of which 14 species were recorded exclusively through this survey method. In forested areas, we captured 138 individuals belonging to 35 species, while, in plantation areas, we captured 176 individuals belonging to 32 species. Only three species were recorded in both environments by this method: Campylopterus largipennis, Glaucis hirsutus, and Phaethornis superciliosus. The species with the most significant number of individuals captured in the paricá plantations were Volatinia jacarina (50), Ramphocelus carbo (14), and Sporophila angolensis (12). In the sampled forest fragments, the most abundant species were Pyriglena leuconota (17), Glyphorynchus spirurus (16), and Willisornis vidua (14).

Functional trait composition
Bird functional-trait composition (dominant traits) differed between plantations and forest fragments (diet: PERMANOVA pseudo-F = 7.003, p (perm) = 0.02; morphological traits and foraging stratum: PERMANOVA pseudo-F = 5.95, p (perm) = 0.009). Concerning diet traits, the first two PCA axes accounted for 72.81% of the bird trait composition (Axis 1 = 47.26%, Axis      Figure 7a). Bird diet traits that dominated the paricá plantations were seed, plant material, and endothermic vertebrates, while forested areas were dominated by nectar, fruits, and vertebrates in general (Figure 7b). Regarding morphological traits and foraging stratum, we also observed an association of species traits with habitat. The first two PCA axes explained 81.39% of the variation (Axis 1 = 67.56%, Axis 2 = 13.83%, Figure 7b). In forested areas, we found a dominance of species with longer bills and that forage in the higher strata of vegetation (i.e. canopy, medium, and high strata). In plantations, the dominant traits were foraging in the lower strata of vegetation (ground and understory) and foraging in the air, and species had, on average, a longer tarsus (Figure 2b).

Discussion
This study aimed to examine how bird communities respond to paricá (Schizolobium amazonicum) plantations in eastern Amazon. Our results indicate that species richness is reduced in plantations, and species composition was markedly different between the two habitats. Moreover, forested areas and plantations shared a small number of species. However, we found no significant difference in the number of individuals between forest and plantation areas. Threatened species were not recorded in paricá plantations. In these areas, bird communities consisted mainly of species that are more adapted to open environments, that is, that show positive responses to monoculture, such as edge granivores, edge insectivores, and edge omnivores. We also observed different trait compositions between birds from paricá plantations and adjacent forest fragments.
The reduction in bird diversity and changes in community structure found in the present study agree with similar studies that evaluated bird diversity in other monocultures in the Brazilian Amazon, such as oil palm (Almeida et al. 2016) and eucalyptus plantations (Barlow et al. 2007). Monoculture environments are less favorable as habitat to birds than natural environments due to the lower resource availability and lower niche diversity (Tews et al. 2004;Carnus et al. 2006;Barlow et al. 2007) and can lead to biotic homogenization of bird communities (Goded et al. 2019). Henriques (2003) assessed the importance of paricá plantations to reforestation and concluded that this type of monoculture was unattractive to forest bird species and had little importance in the maintenance of biological diversity in eastern Amazon. However, unlike our study, Henriques (2003) studied paricá plantations inserted in a more degraded landscape that consisted of 90% of active pasture, secondary forests in various stages of regeneration and only a single forest fragment with a logging and fire history. This author recorded only 36 bird species in the paricá plantation, and this low diversity was attributed to factors resulting from plantation management, such as cutting the understory during the first four years of planting and fighting caterpillar pests with biocides. In addition, there was a larger area of contact with pasture, which made this habitat the source of species or even individuals for paricá plantations (which acted as 'sink' habitats). On the other hand, in our study, the plantation sites were surrounded by forest fragments and the understory was dense, consisting mainly of grasses and making up a less harsh environment for birds compared to the one studied by Henriques (2003).
The highest contributions to community dissimilarity between habitats are birds that are typical of open areas/ forest edge and forest specialists, respectively, for plantations and forests. Crotophaga ani, Formicivora grisea, and Volatinia jacarina, frequently found in plantation areas and open environments, are species that can benefit from changes in natural habitats in the Amazon (Barlow et al. 2007;Lees et al. 2015;Almeida et al. 2016) and are widely distributed in Brazil (Del Hoyo et al. 2019). In forested areas, the most common species were Cercomacra cinerascens, Lipaugus vociferans, and Pyriglena leuconota, which are common in the forest understory and frequently recorded in primary forests in the Brazilian Amazon (e.g. Barlow et al. 2007;Lees et al. 2015). Pyriglena leuconota, like other four understory birds recorded in the present study (i.e. Phlegopsis nigromaculata, Dendrocincla merula, D. fuliginosa, and Dendrocolaptes medius), are antfollowing insectivores that are typical of primary forests. Forest birds that prey on arthropods flushed by army ants are disturbance-sensitive because they fail to accompany swarms that stray into more open habitats (Oniki 1972;Johns 1991;Roberts et al. 2000) and are unable to have vast territories to monitor several different ant colonies nearly simultaneously (Pearman 2002).
In forest environments, species can exploit a variety of trophic resources in different vegetation strata. Variation in forest cover between types of land use can affect bird guilds, especially forest specialists such as understory insectivores, ant-following species, and mixed-flock participants (Pearman 2002). In this study, mixed-flock insectivores (e.g. antbirds, Thamnophilidae; woodcreepers, Dendrocolaptidae), understory frugivores (e.g. manakins, Pipridae) and canopy omnivores (e.g. toucans, Ramphastidae) were not recorded in plantation areas. Also, canopy frugivores, mixed-flock insectivores, and understory insectivores were the most representative guilds in forest fragments. These guilds are sensitive to environmental changes, and their presence in forested areas indicates good-quality habitat (Gimenes & Anjos 2003;Laps et al. 2003). Monocultures, therefore, are unlikely to conserve them (Barlow et al. 2007). Frugivorous birds such as Psittacidae, Cotingidae, Trogonidae, and Pipridae play a crucial role in maintaining forested areas since they act as seed-dispersing agents and depend on different plant species that produce fruit in different seasons of the year (Gimenes & Anjos 2003).
In paricá plantations, the more representative guilds were insectivores, omnivores and granivores of forest edge, and nectar-insect eaters (Trochilidae and Coereba flaveola). These guilds were positively affected by paricá monocultures because, in general, they depend on more open habitats and are usually associated with changes in land use in the Amazon (Barlow 2007;Almeida et al. 2016). Edge omnivorous adapt well to anthropogenic habitats due to their ability to exploit different resources in the environment, and insects are not a limiting resource to insectivores as well (Fagan 1997;Henriques 2003). Granivorous birds (e.g. Sporophila spp. and Volatinia jacarina), in turn, were benefited in plantation areas due to grass proliferation in this environment. This result is similar to the ones found for bird communities in coffee crops. In these areas, the herbaceous stratum is more developed than in undisturbed forest systems since canopy cover is more open, and more sunlight is available (Tejeda-Cruz & Sutherland 2004).
We observed that nectarivores were more representative in plantation areas, a result that is consistent with other studies in the Brazilian Amazon. This demonstrates that this guild has little sensitivity to human disturbance and variation in forest cover (Stouffer & Bierregaard 1995;Pearman 2002). Therefore, nectarivorous birds should be more influenced by other local conditions such as microclimate conditions and food resources (Pearman 2002). In Eucalyptus plantations, Barlow et al. (2007) also found a high abundance of nectarivores such as hummingbirds (Amazilia spp.) and Coereba flaveola. In contrast, Henriques (2003) recorded only one nectarivorous species in paricá plantation; a result attributed to the study period (November) not matching the flowering period of this tree species (Henriques 2003), which generally happens from May to July (Rosa 2006). Although S. amazonicum flowers have a large amount of nectar and may be attractive to nectarivores (Venturieri 1999), its fruit is dry and hard, so it is not highly attractive to frugivores (Henriques 2003).
We found an association of species traits with the habitat, with only species that have traits that easily adapt to open areas persisting in plantations (e.g. ground-foraging granivores). Changes in specific ecological traits reflect the species turnover between the studied habitats and an increase in ecological functions provided by the community in forested areas (Jacoboski et al. 2016;De Arruda-almeida et al. 2019). When environmental differences between sites are not pronounced, differences in species richness and composition can occur without generating changes in functional composition (Oliveira et al. 2019). Bird communities in forest fragments presented species with longer bills (e.g. toucan, woodpeckers, woodcreepers) while in paricá plantations the birds had small-sized beaks which reflects their diet (mainly grains and insects). This found agrees with Bregman et al. (2016), who found that as the intensity of land use increases, birds tend to have shorter and wider beaks. Tarsus length was associated to the plantations, reflecting the bird community that is dominated by species that forage on the ground. On the other hand, when the community is dominated by species that capture prey in flight or from canopy foliage the average tarsus size is smaller (Bregman et al. 2016). Still considering the CWM values, in forested areas, we observed a high representation of species that feed on fruits and nectar. This result apparently contrasts with what we found for the classification in trophic guilds, where the guild of nectar-insect eaters presented a greater number of species in the plantation compared to the forest. The term guild is used to designate groups of animals exploiting similar resources (Blondel 2003), and the classification of Wilman et al. (2014) is based on the proportion of each food item consumed by a species. Therefore, the same species can consume different types of resources, from which the CWM values were calculated. For example, some species not classified in the nectar-insect eater guild may consume nectar according to the Wilman's classification (e.g. Cyanerpes caeruleus, Cacicus haemorrhous, Saltator spp., Tangara palmarum, Ara macao in this study). However, this is not their main diet. These results also demonstrate that trophic guild and functional structure measures can reveal different aspects of bird communities, requiring a joint approach to better understand birds' responses to environmental changes.
The complete absence of more specialized species such as mixed-flock insectivores and understory frugivores and the reduced abundance of canopy frugivores and understory insectivores reflect the reduced resource diversity and availability for birds in paricá plantations. Additionally, changes in bird functional traits indicate that the replacement of forests with paricá plantations can lead to shifts toward less specialized bird communities with altered proportions of functional groups (Sekercioglu 2012). In this way, paricá plantations have no potential to support forest bird species. Replacing native forests with this monoculture would lead to a significant loss of biodiversity, with the most substantial effect on forest specialists, and affecting the ecosystem services provided by birds in agricultural landscapes. It is important to note that the forested areas considered in the present study have already undergone some change in their structure (e.g. fragmentation, logging, hunting). Thus, the biodiversity loss in paricá plantations must be even greater when compared to primary continuous forests. However, the forest fragments surrounding these plantations have significant bird biodiversity, including species of national and global conservation concern such as Guaruba guarouba. Thus, the protection of these privately owned forest fragments is of vital importance for the conservation of forest biodiversity, given the accelerated deforestation in the Brazilian Amazon.