Figures
Abstract
Knowledge of spatiotemporal distribution of biodiversity is still very incomplete in the tropics. This is one of the major problems preventing the assessment and effectiveness of conservation actions. Mega-diverse tropical regions are being exposed to fast and profound environmental changes, and the amount of resources available to describe the distribution of species is generally limited. Thus, the tropics is losing species at unprecedented rates, without a proper assessment of its biodiversity. Species distribution models (SDMs) can be used to fill such biogeographic gaps within a species’ range and, when allied with systematic conservation planning (e.g. analyses of representativeness, gap analysis), help transcend such data shortage and support practical conservation actions. Within the Neotropics, eastern Amazon and northern Cerrado present a high variety of environments and are some of the most interesting ecotonal areas within South America, but are also among the most threatened biogeographic provinces in the world. Here, we test the effectiveness of the current system of Protected Areas (PAs), in protecting 24 threatened and endemic bird species using SDMs. We found that taxa with wider distributions are potentially as protected as taxa with smaller ranges, and larger PAs were more efficient than smaller PAs, while protecting these bird species. Nonetheless, Cerrado PAs are mostly misallocated. We suggest six priority areas for conservation of Neotropical birds. Finally, we highlight the importance of indigenous lands in the conservation of Neotropical biodiversity, and recommend the development of community management plans to conserve the biological resources of the region.
Citation: de Carvalho DL, Sousa-Neves T, Cerqueira PV, Gonsioroski G, Silva SM, Silva DP, et al. (2017) Delimiting priority areas for the conservation of endemic and threatened Neotropical birds using a niche-based gap analysis. PLoS ONE 12(2): e0171838. https://doi.org/10.1371/journal.pone.0171838
Editor: Govindhaswamy Umapathy, Centre for Cellular and Molecular Biology, INDIA
Received: August 28, 2016; Accepted: January 26, 2017; Published: February 10, 2017
Copyright: © 2017 de Carvalho et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) doctoral fellowships: DLC and PVC. Fundação Amazônia de Amparo a Estudos e Pesquisas (FAPESPA) doctoral fellowship: TSN. Programa de Pós-Graduação em Zoologia Museu Paraense Emílio Goeldi / Universidade Federal do Pará (PPGZOO MPEG/UFPA) Programa Nacional de Pós-Doutorado (PNPD) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) fellowship: SMS. Eudocimus Consultoria Ambiental provided support in the form of salaries for GG, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Thus, I assure that GG´s commercial affiliation did not play a role in our study. The specific roles of all the authors are shown in the ‘author contributions’ section.
Competing interests: We have the following interests: Gustavo Gonsioroski is employed by Eudocimus Consultoria Ambiental. There are no patents, products in development or marketed products to declare. This does not alter our adherence to all the PLOS ONE policies on sharing data and materials, as detailed online in the guide for authors.
Introduction
The world is undergoing rapid and intense environmental changes that are, directly or indirectly, caused by human activities. Habitat loss and fragmentation, deposition of anthropogenic fixed nitrogenous substances, and the rise of atmospheric carbon dioxide concentration related to climatic changes are or will be the main drivers of such alterations [1,2]. Under this scenario, high-quality species distributional data are essential to set efficient conservation actions [3–5]. However, those biogeographic information are often lacking, being one of the major setbacks preventing the assessment of need and effectiveness of these actions (the Wallacean shortfall) [4,6]. Such a scenario is even more concerning in tropical regions [7–9], because these are mega-diverse areas, that have been suffering fast environmental changes [10,11], and in general, the amount of resources to describe the distribution of species is limited [12–14]. Consequently, the tropics are losing species at unprecedented rates [15–17], often without properly identifying and describing their biodiversity (the Linnean shortfall) [4,18].
One way to fight back against the Wallacean shortfall is to use species distribution models (SDMs) [19–21]. These models correlate known occurrences of target species with climatic, land-use, and topographic data to delimit the multidimensional bioclimatic requirements for the modeled taxa, reflecting their environmental preferences [22]. SDMs can be overlaid upon the geographic range space to fill biogeographic gaps within species’ ranges [23], even for elusive and seldom recorded species [24–27]. In conservation, SDMs have been widely and successfully used to 1) predict the distribution of rare, endemic and threatened species [25,28–31], 2) perform niche-based gap analyses and discover species that are not protected (i.e. do not occur in any protected area; PA hereon) [32–34], 3) predict suitable areas for the invasion of exotic species [35–38], 4) evaluate the potential effects of future climate changes [39,40], 5) determine suitable areas for the reintroduction of rescued fauna [41,42], and 6) establish and evaluate priority areas for conservation [43–45], amongst other examples.
Therefore, one of the first steps for setting a conservation plan may be to ally SDMs with systematic conservation planning [46], particularly in the analysis of representativeness, also known as gap analysis [20,34,47–49]. Gap analysis consists of the identification, classification, and examination of the existing system of PAs based on the assessment of the representation of species, vegetation types or biomes within those PAs network and identification of gaps of distribution in its coverage [23,50]. Representativeness is one of the four main principles of systematic conservation planning, the others: comprehensiveness, adequacy and efficiency [46,51–55].
The region encompassing eastern Amazon and northern Cerrado (Fig 1) is one of the most heterogeneous regions throughout the Amazon basin. This area presents a high variety of environments, as tropical rainforests (terra firme and várzea), floodplains, campinas, extensive mangroves in the coastal zone, being a large ecotonal area with the Cerrado [56–58]. All this diversity has been affected by intense anthropogenic pressure, mainly due to high deforestation rates and a strong expansion of agribusiness. Noteworthy, 61% of the endangered birds in the Brazilian Amazon occur primarily or exclusively in this portion of the Amazon basin, which constitutes the Belém area of endemism (BAE) [59], while Cerrado is the second most threatened biodiversity hotspot of Brazil [3,60–63].
Location of Protected Areas (PAs): conservation units (dark brown) and indigenous lands (light brown), within the Amazon and Cerrado biomes. See Table 1 for PAs identification (ID).
Within Eastern Amazon, the state of Maranhão (Fig 1) has already lost about 30% (24,412 km2) of its historical Amazon forest, and 21% (436 km2) of Cerrado's native vegetation [64]. Nonetheless, this is the less affected area (within BAE and Cerrado), and still preserves one of the richest avifauna of Brazil, mainly due to the aforementioned diversity of ecosystems [65,66]. The birds’ list of Maranhão has over 640 species [67], corresponding to 34% of the total number of species in Brazil [68]. Of those, 49 (21%) are included on the Brazilian official list of threatened species [59], and several taxa from the BAE have the highest level of local vulnerability [67]. The state further includes 10 endemic bird species within Cerrado (DL Carvalho et al. in prep). These features make Maranhão an interesting region to test the effectiveness of the state system of PAs (both in Amazonian forest and the Cerrado biomes), in protecting threatened and endemic bird species using SDMs. Specifically, here we used SDMs to perform a gap analysis, and seek to know if 1) taxa with relatively wider distributions are more protected (i.e. have higher percentage of area within PAs) than taxa with smaller distributions and 2) if relatively larger PAs are more efficient (i.e. have higher species richness) than smaller PAs. Finally, based on our results, we also suggested new conservation areas for the target taxa and discuss the effective implementation of new conservation practices in the Neotropical region, in order to allow a more significant conservation of its bird species.
Materials and methods
Study area
The study area has a total area of 331,983.29 km2, corresponding to the Brazilian state of Maranhão, the eighth largest Brazilian state. Political borders are biologically meaningless, but conservation actions mostly depend on political actions, so we chose to use this political delimitation to study a biologically relevant ecotonal area, located between the Amazon forest (west), Cerrado (south and southwest), and small patches of Caatinga biome (east) (Fig 1) [69]. The state’s economy is structured in two main areas of development and integration: extensive cattle ranching and logging in the Amazonian portion; and mineral and metallurgical complex, agriculture and production of energy in the Cerrado [70]. As aforementioned, despite the intense anthropogenic actions, Maranhão holds one of the largest patches of forest within BAE, and native Cerrado vegetation [64].
In this study, we considered 39 protected areas (PAs) distributed along the biomes of Amazon and Cerrado (Fig 1). Of these, 24 are conservation units: 13 are Federal PAs (eight of sustainable use and four of full protection), while 10 are state PAs (six of sustainable use and two of full protection). The other 18 are indigenous lands delimited and homologated by the Federal government (Table 1).
Instance, kind of usage, biome, extent area according to World Wildlife (www.worldwildlife.org), and priority according to MMA [71].
Target species and occurrence dataset
Our total dataset includes 24 terrestrial, non-migratory bird taxa, with enough occurrence records available (Table 2). Fourteen are classified as threatened in the Belém area of endemism (BAE) in Amazonia [59,72], and 10 are endemics to the Cerrado biome [73–76], including two species considered Vulnerable to extinction [59,72].
Points: number of records (4.5 × 4.5 km cells), AUC and TSS: mean values, and respective standard deviation values, ER/study area: estimated range in number of cells in the study area, ER/biome: estimated range in number of cells in the study area by biome (Amazon/Cerrado), %PA/biome: percentage of occurrence in protected areas, Status: conservation status according to IBAMA[59] and IUCN[72], and biome of occurrence.
We gathered distribution data for each taxon from literature records, online databases [VertNet (http://vertnet.org/), Species Link (http://splink.cria.org.br), Global Biodiversity Information Facility (http://www.gbif.org), Wikiaves (http://www.wikiaves.com.br), xenocanto (http://www.xeno-canto.org)], museum collections (Louisiana Museum of Natural History, Museu Paraense Emílio Goeldi and Museu Nacional do Rio de Janeiro) and personal observation (DLC, GG, and PVC; see S1 Fig for a complete list of records). We checked all occurrences and excluded dubious records based on the known distribution of the species [72]. Geographical coordinates were obtained directly from the original sources or from Ornithological Gazetteer of Brazil [77]. Bird nomenclature follows the Brazilian Ornithological Records Committee [68].
Environmental layers, modeling procedures, thresholds and evaluation
Occurrence records were overlaid on grid of cells of 2.5 arc-min (~4.5 x 4.5 km). A buffer of 200 km was set around all records to define total extent area (S2 Fig). Using this same grid and considering all 19 bioclimatic variables from WorldClim (http://www.worldclim.org/), we performed a pair-wise Pearson correlation test of all variables to remove those highly correlated and reduce their collinearity [78]. In the case of high correlation (r> 0.8 or r< -0.8), we used only one of the variables in the distribution modeling. We selected nine predictor variables as our environmental variables (Annual Mean Temperature, Mean Diurnal Range, Isothermality, Max Temperature of Warmest Month, Annual Precipitation, Precipitation of Driest Month, Precipitation Seasonality, Precipitation of Warmest Quarter and Precipitation of Coldest Quarter).
All models were trained with MaxEnt 3.3.3 [79,80]. This method computes the suitable distribution of maximum entropy for the set of climatic variables associated to the occurrence records of the target species, however this procedure can be constrained by the incomplete knowledge about the distribution of the species [79,81]. MaxEnt is a presence/background method that requires only presence data as input, and consistently performed well in comparison to other methods [48], especially at low samples sizes [28,82,83].
Due to limited availability of suitable occurrence data for five modeled taxa (<10 records), we used the Jackknife approach, also known as leave-one-out method [28], to predict their potential distributions. Then, we evaluated the resulting distributions with the same subsets. For the remaining 19 taxa, we used 10 subsets dividing the occurrences into 70% training and 30% testing records. We used the threshold that balances both omission and commission errors while modeling the species distributions to cut the suitability matrices of the modeled species in modeling algorithm into presence-absence maps [84,85]. The statistically significant probability (p<0.05) indicate that the model predictions are reliable, despite some eventual omission and/or commission. All probabilities were evaluated in R 3.3.1 (www.r-project.org). We further used both Area Under the receiver–operator Curve (AUC) [86] and the True Skilled Statistics (TSS) [87] to assess models’ performance. AUC and TSS account for the sensitivity (quantifies omission errors), and specificity (quantifies commission errors) of the models [87]. AUC values vary between 0 and 1, with values ≤0.5 representing models no better than random and values around 1 representing a perfect fitting between the observed and the predicted species distribution. Thus, we considered acceptable distribution models to be those with AUC≥0.7 [86,88]. TSS varies from −1 to +1, where negative and around zero values indicate that distributions are no better than random, while values near +1 represent perfect agreement between the observed and the modeled distributions. Acceptable and excellent models were those with TSS values of at least 0.5 and ≥0.7, respectively [87]. We used 10.000 random pseudo-absences in all model evaluation procedures. A mean consensual distribution map for each taxon was made with those models which achieved TSS >0.4. This method was considered to be the best to delimit the final distribution of a given taxon for several different modeling algorithms [89].
Estimated protected range, species richness and identification of priority areas for conservation
As our group of target taxa occurs in two distinct biomes, one predominantly forested (Amazon) and another dominated by savanna (Cerrado), it is expected that predominantly forest-dependent taxa from the Amazon biome are not present in protected areas dominated by savanna, while endemic species from Cerrado are not expected to occur in PAs dominated by forest. Accordingly, we considered two different approaches to estimate the proportion of protected area for each taxon, and species richness, and so to perform the statistical tests related to the gap analysis. In the first approach, we considered all 24 taxa together, and in the second, taxa from each biome were considered separately. Neomorphus geoffroyi and Celeus obrieni were considered in both Amazon and Cerrado analyses, since each are known to occur in both biomes. We obtained the modeled species richness by summing the final distribution of each taxon.
We used the shapefile of the world ecoregions available at the World Wildlife website (http://www.worldwildlife.org/publications/terrestrial-ecoregions-of-the-world) to depict PAs (conservation units and indigenous lands) within the study area. PAs were converted to raster files with grid cells of approximately the same resolution used in the modelling procedures (0.041° or ~4 km near the Ecuador). We identified the grid cells in which each bird taxon was predicted to occur in protected area.
We used linear regressions and power functions to evaluate the effectiveness of PAs. To evaluate the relationship between the size of PAs individually and the estimated species richness covered in each PAs, we used the same two approaches, and so considered three scenarios: 1) the maximum value of Amazon species richness present in the Amazon biome; 2) the maximum value of endemic Cerrado's species richness present in Cerrado; 3) the maximum value of all target taxa richness throughout the study area. A 95% confidence interval for the slope for all regression analyses was selected.
Finally, to identify priority areas for conservation, we overlaid the predicted species richness for each biome separately with maps of deforestation in the Amazon biome (PRODES data from [64]), and remnants of native vegetation in Cerrado [90]. Protected Areas fully covered by native vegetation, and in which more than half of the target taxa potentially occurs, were considered priority areas for conservation.
Results
Species distribution models
We collected a total of 1,534 occurrence records, from which 929 were used (sample size varied between 5 and 110; S1 Fig) to generate the potential distribution maps of the 24 target bird taxa (S2 Fig; Fig 2). Six out of the 24 SDMs presented errors of omission and/or commission (Dendrocincla m. badia, Alipiopsitta xanthops, Herpsilochmus longirostris, Antilophia galeata, Suiriri affinis and N. geoffroyi). SDMs for A. xanthops, and H. longirostris had considerably reduced potential area of occurrence within the study area (Fig 2; S2 Fig). Yet, all SDMs showed fair to excellent predictive capability (Table 2). AUC values varied between 0.78 and 0.98. TSS values were always higher than 0.5, except for the endemic species from Cerrado, Saltatricula atricollis (TSS = 0.4). Models for taxa with less than 10 records (Crax f. pinima, Psophia obscura, Neomorphus geoffroyi, Piculus paraensis, Dendrexetastes r. paraensis) predicted taxa distributions better than random (p<0.05), according to the leave-one-out method (Table 2). Thus, all taxa were considered in the following analyses. In the study area, predominantly forest-dependent taxa had higher probabilities of occurrence in the Amazon biome, and Cerrado endemics were mostly assigned to occur in this biome (Fig 2). Also, as expected, C. obrieni was predicted to be present in both biomes. However, SDM for N. geoffroyi estimated the distribution of this species to be mostly restricted to the Amazon (Fig 2). Therefore, further results were mostly focused in the approach separating taxa by biome (with N. geoffroyi excluded from analyses considering Cerrado taxa), and results considering total target taxa and the entire study area are only presented for comparison.
Estimated range of taxa from Amazon 1) and Cerrado 2) biomes.
Estimated protected range and species richness
Percentage of protected area for each taxon varied between 41% (Pyrrhura coerulescens) and 94% (C. obrieni) considering the Amazon biome only, and 0% (A. xanthops) and 11% (S. affinis) for Cerrado taxa (Table 2).
Our linear regression analyses indicated that both conservation units and indigenous lands are protecting the target taxa better than random (Fig 3). For the species within the Amazon, for every 202.5 km2 of distribution range (10 grid cells), there was a gain of protection of about 81 km2 (four cells; Fig 3A). On average, 59%±13% of estimated ranges for the Amazonian species is protected in this biome. In Cerrado portion of the study area, for every ≈69.000 km2 (3400 cells of distribution), there was a gain of protection of only 20.25 km2 (one cell; Fig 3B). The protected range of species from Cerrado averaged only 12%±11% of their distribution. Using the entire study area and the total dataset of target taxa, we obtained a gain of protection of 40.5 km2 (two cells), for every 202.5 km2 (10 cells of distribution) (Fig 3C). For all 24 species in the whole study area, the average of protected range was 38%±26%.
We observed a positive relationship between the total range and the protected range size of threatened taxa from Amazon (A), Cerrado (B) and all target species in the whole study area (C).
Considering Amazonian threatened taxa (n = 14), within this biome, we observed that areas with higher estimated species richness (n≥7, i.e. ≥50%) are located in the north-western region of the study area (Fig 4A). The relationship between the sizes of each PA and the estimated species richness in this biome was positive (Fig 5A), i.e. the largest protected areas in this region have a wider number of species. For each 2.025 Km2 (100 grid cells), a gain of protection of one species was obtained (Fig 5A). Regarding only the Cerrado’s potential species richness, areas with higher values (n≥6, i.e. ≥50%) are inserted in patches in the southeast and south of the study area (Fig 4B). There was no relationship between the size of PAs of this biome and the estimated species richness, since the random model was sufficient to explain the observed variation (R2 = 0.201, p >0.05, y = 0.8253*x^0.2276).
A) Amazon biome and B) the whole study area.
Considering the 24 target taxa together in the entire study area, the highest values of species richness (n≥12, i.e. ≥50%) were in the Amazon biome, and only some patches were highlighted in Cerrado, mainly in the southern part of the study area (Fig 4C). Within Cerrado, the estimated species richness with all target taxa reached a maximum of only 16% (Fig 4C). The relationship between estimated species richness and the size of each PA in the entire study area was positive (Fig 5B). For each 1.012 Km2 (50 grid cells) a gain of protection of one taxon was obtained (Fig 5B).
Identification priority areas for conservation
Again, using the approach of analyzing Amazon and Cerrado, and their taxa, separately, we highlight two priority areas for conservation in the Amazon biome, and four areas in Cerrado (Fig 6). As aforementioned, these priority areas for conservation have a species richness ≥50% and still maintain native vegetation.
Study area showing the currently existing protected areas and indicating new priority areas for conservation according to the estimated species richness in the Amazon and Cerrado biomes, separately. 1—Regions connecting APA Baixada Maranhense with indigenous land Alto Turiaçu; 2 –Extension of indigenous land Alto Turiaçu; 3—Polígono das Águas in southern Maranhão; 4—Southwestern Plateau; 5—Mirador/ Uruçuí; 6—Extension of the Maranhão semideciduous forest area, in central Maranhão.
Discussion
Species distribution models
Overall, our species distribution models (SDMs) agree with known distributions and ecological requirements of target taxa. Most target Amazon taxa have a high specificity for forested habitats [67], and in fact had predicted distributions mostly restricted to those regions. Only for Crax f. pinima, Guaruba guarouba, Neomorphus geoffroyi, Hylopezus paraensis and Dendrocolaptes medius, SDMs predicted their occurrence also in Cerrado, although these are forest-dependent taxa. In fact, these were predicted to occur in mangrove areas in coastal zones, because of the similar climatic features of mangroves and the neighboring rainforests. Cerrado and open areas taxa had SDMs consistent with their known habitat affinity [91–93]. For conservation purposes, models that estimate the full niche requirements are preferred [34]. Thus, our SDMs express the full niche of our target taxa, and so are adequate to perform gap analysis [86,87], even for taxa with low numbers of records (this study; [28,94]). As detailed in the next sections, we add support to the use of SDMs in the systematic conservation planning of Neotropical organisms, as previously shown not only for birds, but also other different taxonomic groups such as odonata [34], anura [6], turtles [49], and mammals [95].
Nonetheless, we know that the potential distribution maps are an abstraction that might not reflect species occurrence at fine geographic scales [96,97]. Furthermore, in ecotonal areas such as ours, which might represent a limit of distribution for several distinct taxa, species might have different ecological requirements than at the core of their distributions, and do not fully express their niches (incomplete niche expression), and finer-scale studies may be necessary to addresses some species-specific questions [98,99]. This is particularly important in the current fast climatic changing scenario all natural species are facing. As our SDMs show, Amazonian and Cerrado birds seem to have totally distinct climatic requirements, and so may respond differently to climate change. Thus, to decrease the Wallacean shortfall, we propose that more studies are necessary in such transition regions in the Neotropics, considering the lack of information and that the ecotonal condition may promote a high species richness [100].
Species estimated protected distribution
Taxa with broad distributions are potentially as protected as taxa with smaller distributions within the study area. Nonetheless, despite the positive relationship between the amount of potential range and the amount of potential range that is protected, this latter proportion varied enormously between biomes.
Taxa with a potential area of distribution wider than 250,000 km2 must have at least 10% of its distribution protected, and taxa with smaller distributions (around 103 km2), should be fully protected, i.e. 100% of its potential distribution must be included in PAs [101]. Within the study area, most target taxa had potential distributions greater than 104 km2. Thus, no less than about 60% of their potential ranges should be protected [102]. For the target Amazonian taxa, this is close to the mean percentage of potential protected area estimated, suggesting these species are well protected. However, the Critically Endangered C. f. pinima and P. obscurus, and the Threatened P. paraensis, N. geoffroyi and D. r. paraensis occur at low population densities, even in well-preserved areas, and most of them have already been indicated as likely extinct at a regional level, even in still forested areas, due to degradation and hunting, particularly in western BAE [67,103–108]. This implies that, despite our results on linear regression analyses, these taxa might need further conservation actions. More importantly, they exemplify the need to gather the most up to date information available on Neotropical species; otherwise more recent impacts of habitat loss and degradation might be overlooked.
Within the Cerrado portion of the study area, the mean potential area of distribution estimated to be currently protected was only of 12%; a percentage close to the 10% recommended by Rodrigues’ et al.[101] for widespread species. Yet, it is alarming that, even considering the whole biome, hardly any Cerrado species will overcome this threshold, because current PAs system within Cerrado is highly inefficient in conserving bird species [39]. Only 12 (32%) of the target species analyzed by Nóbrega & De Marco [33] had 5% of their distribution protected, even when authors considered all Brazil, and none of them even reached the 10% threshold if considering only large reserves [33]. Furthermore, protected potential distribution of Cerrado species might be insufficient to maintain viable populations, due to the high level of fragmentation, especially within the southern part of the biome [109]. In fact, endemic Cerrado birds are already presenting signs of a decreasing gene flow due to anthropogenic habitat fragmentation and degradation [110]. Moreover, Cerrado endemics have highly specific habitat requirements, such as the Vulnerable C. ferdinandi [59,72,111]. Its potential protected distribution was estimated to be only 3% in the study area, despite occurring in more than about 22,000 km2. The demand for a specific conservation plan for this species was already stressed elsewhere [39], and our results further support this recommendation.
Our study reinforces the need to overcome the huge Wallacean shortfalls that prevents proper conservation planning of Neotropical species. Assessments based on species-specific information, not only occurrence data, but also biological and ecological data, should be added to general conservation plans [46], and must be thorough and updated frequently, due to fast land-use changes. Not only the Amazon and Cerrado are losing native vegetation at a fast rate, but many other Neotropical regions are equally or more threatened. For instance, the Brazilian Atlantic forest, the tropical Andes, and the Chilean Winter Rainfall-Valdivian Forests were, almost two decades ago, highlighted as hotspots of biodiversity [3], and their degree of threat has still not changed [15].
Estimated endemic and threatened species richness
According to our data, larger PAs are more efficient, i.e. have higher species richness, than smaller PAs. Relationship between the size of Amazonian PAs and potential Amazon species richness was positive, but with a low explaining power. The low coefficient of determination obtained (R2 = 0.43) was influenced by the indigenous lands Alto Turiaçu, Alto Rio Guama, Awa, and Caru, and the conservation unit REBIO Gurupi, which altogether assemble the forest block of Gurupi, totaling 13,900 km2. These distinct PAs, have distinct kinds of usage, and so were analyzed separately, but biologically they seem to be in fact a unit, having a similar species richness as bigger PAs (APA Baixada Maranhense, 17,285 km2 and APA Reentrâncias Maranhenses, 26,285 km2). Considering the forest block of Gurupi as a unique PA would increase the coefficient of determination, and so the positive relationship (data not shown). This relationship is in agreement with previous findings [112,113]. Peres [112] further states that only a well-connected network of mega-reserves, exceeding an area of 10,000 Km2, would cover a major portion of regional biodiversity, preserving populations of rare predators, but also species with seasonal movements (e.g. G. guarouba and A. xanthops), and animals impacted by hunting (e.g. among our target taxa, P. obscurus and C. f. pinima). In fact, within eastern Amazon, roads seem to impact on avian species richness and composition due to habitat fragmentation but also by facilitating logging, fire, hunting, and other traffic disturbances [114]. Additionally, considering the potential effects of the predicted climate changes upon overall biodiversity, patch connectivity may become even more important to guarantee species dispersal in the future. Mega-reserves are considered to enable species to better overcome climatic changes than smaller PAs [115], since larger areas potentially enable species to maintain larger population sizes, with greater genetic diversity, allowing them to adapt their niches and distributions in changing environments [113].
Within Cerrado, we did not find a relationship between PA size and species richness, but Cerrado PAs are mostly misallocated, covering areas of low species richness, and not suitable for cultivation. As abovementioned, Cerrado PAs are failing in protecting the biome´s biodiversity (this study;[39]).
Our results agree with Marini et al. [39], Bini et al. [6], and Peres [112] in that more (and larger) PAs are needed to maintain eastern Amazon and Cerrado biodiversity. Noteworthy, both Amazon and Cerrado PAs face the same anthropogenic pressure as other Neotropical regions [116–118], so similar studies are still needed throughout the Neotropics to review the systems of PAs (but see [119–121]).
Identification of priority areas for conservation
Considering the entire BAE, less than 17% of its area is currently protected (1.4% conservation units of full protection, 9.77% conservation units of sustainable use, and 6.49% indigenous lands [65]). Currently, the greatest rates of deforestation, within the Amazon, occur precisely in the east, due to a stronger pressure from economic groups that occupy public and private lands for the development of agricultural production, logging and cattle-raising [122]. One of the most important areas currently protected, REBIO Gurupi, has also lost 20% of its area due to illegal occupation for agricultural exploitation, extraction of wood, burning and deforestation made by squatters and other landowners [67,123–125]. REBIO Gurupi is part of the forest block of Gurupi, that together with APA Baixada Maranhense, and APA Reentrâncias Maranhenses were identified as Important Bird Areas (IBAs), with the occurrence of Endangered and Near Threatened species’ populations and “trigger species”, and also considered of "extreme importance" (this study;[71,103,124]). Thus, and according to Peres’ [112] recommendations to extend PAs networks into mega-reserves, we highlight the regions connecting APA Baixada Maranhense with indigenous land Alto Turiaçu, and an extension of this last PA as priority for conservation actions.
Cerrado holds 5% of the planet's biodiversity and is considered the richest savanna in the world, but one of the most threatened regions in Brazil, which has lost about 48% of its native vegetation until 2008, and has only 2.2% of protected area [60,62,69]. Estimates indicate that at least 20% of endemic and threatened species within the whole biome remain outside parks and reserves [126]. As extensively debated above, Cerrado system of PAs needs to be revised (this study, [6,39]), but the better location of new PAs has been contentious. According to Bini et al. [6], weighting for the knowledge on species distribution, new areas in the north of the biome should be priority, but for Marini et al. [39] and Diniz-Filho et al. [127,128], these new areas should be in the southern part of the biome, since species richness was higher there. The priority areas for conservation we suggest, not only had higher species richness and still hold extensive native vegetation, but also were already recommended to acquire conservation unit status by MMA [71] (Polígono das Águas in southern Maranhão, Southwestern Plateau, and Mirador/ Uruçuí). The last area was also indicated as an IBA for the presence of endemic species as A. xanthops, M. torquata, C. cristatellus, P. caerulescens, C. eucosma, S. atricollis and the threatened C. obrieni [124]. In central Maranhão, we recommend the extension of the Maranhão semideciduous forest area (also already highlighted by [71]).
Finally, our results highlight the importance of indigenous lands in the conservation of Neotropical biodiversity. Among the areas with higher species richness (≥50% of taxa), more than a half were indigenous lands. Brazil's forestry code (Law 12651, Article 3, 25 May 2012) classifies indigenous lands as areas of full protection. However, Rylands [129] and Instituto Socioambiental et al. [130] categorize indigenous lands as areas that allow human occupation and/or sustainable management activities, having a conflicting view about the land uses that should be allowed in these areas [131]. Given the value of indigenous lands for conservation, the development of community management plans is essential to conserve the biological resources of the region, and is beneficial for all society [67].
Conclusions
Protected Areas in Eastern Amazon are large and, at least in part, well connected, holding high biodiversity. Nonetheless, the lack of overall biological knowledge, and the high rate of deforestation, habitat degradation, and mostly economic pressures make studies such as ours only useful if accompanied by an increase of public awareness, adequate governmental policy, and proper conservation planning. Noteworthy, this is most striking in Cerrado, where scientific debate on conservation actions has been quite intense and controversial, but habitat degradation has increased. Nonetheless, our results further validate governmental reports on the implementation of new PAs, and encourage putting these findings into practice.
Supporting information
S1 Fig. Occurrences for the taxa analyzed.
A-X) Twenty-four maps depicting 929 records obtained from the literature (triangles), museum collections (diamonds), online databases (crosses), and field expeditions (circles).
https://doi.org/10.1371/journal.pone.0171838.s001
(PDF)
S2 Fig. Species distribution models.
First map depicts the extent area in yellow, and the study area in red, Amazon 1) and Cerrado 2) biomes, and all the other maps represent the 24 SDMs estimated. SDMs for Herpsilochmus longirostris and Alipiopsitta xanthops include potential distributions overlaid by the border of the study area.
https://doi.org/10.1371/journal.pone.0171838.s002
(TIF)
Acknowledgments
We thank L. Carneiro and the Project Biodiversidade Ameaçada / PróVida Brasil 002/2014/SEMA for logistical support in the field, F. Freitas Filho for permission to work in his area and for logistical support, S. Dantas, A. Lees, N. Moura, J. V. Remsen (LSUMNH), M. Raposo (MNRJ), A. Aleixo (MPEG), F. Lima (MPEG), and all the birdwatchers in WikiAves for access to and information about occurrence records. Finally, we acknowledge the suggestions of A. Lees and two anonymous reviewers, which improved this manuscript.
Author Contributions
- Conceptualization: DLC DPS MPDS.
- Formal analysis: DLC TSN DPS.
- Investigation: DLC PVC GG.
- Methodology: DLC DPS MPDS.
- Software: DLC TSN DPS.
- Supervision: DPS MPDS.
- Validation: DLC TSN DPS.
- Visualization: DLC TSN SMS DPS MPDS.
- Writing – original draft: DLC TSN SMS DPS MPDS.
- Writing – review & editing: DLC TSN SMS DPS MPDS.
References
- 1. Tylianakis JM, Didham RK, Bascompte J, Wardle DA. Global change and species interactions in terrestrial ecosystems. Ecol Lett. 2008;11: 1351–1363. pmid:19062363
- 2. Dobrovolski R, Diniz-Filho JAF, Loyola RD, Júnior PDM. Agricultural expansion and the fate of global conservation priorities. Biodivers Conserv. 2011;20: 2445–2459.
- 3. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GA, Kent J. Biodiversity hotspots for conservation priorities. Nature. 2000;403: 853–858. pmid:10706275
- 4. Whittaker RJ, Araújo MB, Jepson P, Ladle RJ, Watson JEM, Willis KJ, et al. Conservation Biogeography: assessment and prospect. Divers Distrib. 2005;11: 3–23.
- 5. Brooks TM, Mittermeier RA, da Fonseca GAB, Gerlach J, Hoffmann M, Lamoreux JF, et al. Global biodiversity conservation priorities. Science. 2006;313: 58–61. pmid:16825561
- 6. Bini LM, Diniz-Filho JAF, Rangel TF, Bastos RP, Pinto MP. Challenging Wallacean and Linnean shortfalls: knowledge gradients and conservation planning in a biodiversity hotspot. Divers Distrib. 2006;12: 475–482.
- 7. Graham CH, Ferrier S, Huettman F, Moritz C, Peterson AT. New developments in museum-based informatics and applications in biodiversity analysis. Trends Ecol Evol. 2004;19: 497–503. pmid:16701313
- 8. Newbold T. Applications and limitations of museum data for conservation and ecology, with particular attention to species distribution models. Prog Phys Geogr. 2010;34: 3–22.
- 9. Kamino LHY, Stehmann JR, Amaral S, De Marco P, Rangel TF, de Siqueira MF, et al. Challenges and perspectives for species distribution modelling in the neotropics. Biol Lett. 2012;8: 324–326. pmid:22031720
- 10. Bawa KS, Kress WJ, Nadkarni NM, Lele S. Beyond paradise–meeting the challenges in tropical biology in the 21st century. Biotropica. 2004;36: 437–446.
- 11. Hong SK, Lee JA. Global environmental changes in terrestrial ecosystems. International issues and strategic solutions: introduction. Ecol Res. 2006;21: 783–787.
- 12. Dutra RCD, Oliveira AB, Prado AC de A. Execução orçamentária do Ministério do Meio Ambiente entre 2000 e 2005. Política Ambient. 2006;1: 3–14.
- 13. Borges SH, Iwanaga S, Moreira M, Durigan CC. Uma análise geopolítica do atual sistema de unidades de conservação na Amazônia Brasileira. Política Ambiental. 2007;4: 1–42.
- 14.
Nitta R. Budget and Staff analysis for Management of Forest Protected Areas in Brazil. University of East Anglia. 2009.
- 15. Pimm SL, Jenkins CN, Abell R, Brooks TM, Gittleman JL, Joppa LN, et al. The biodiversity of species and their rates of extinction, distribution, and protection. Science. 2014;344: 1246752. pmid:24876501
- 16. Ceballos G, Ehrlich PR, Barnosky AD, García A, Pringle RM, Palmer TM. Accelerated modern human–induced species losses: Entering the sixth mass extinction. Sci Adv. 2015;1: e1400253. pmid:26601195
- 17. Lees AC, Pimm SL. Species, extinct before we know them? Curr Biol. 2015;25: R177–R180. pmid:25734261
- 18. Costello MJ, May RM, Stork NE. Can we name Earth’s species before they go extinct? Science. American Association for the Advancement of Science; 2013;339: 413–416. pmid:23349283
- 19. Guisan A, Zimmermann NE. Predictive habitat distribution models in ecology. Ecol Modell. 2000;135: 147–186.
- 20. Elith J, Leathwick JR. Species distribution models: ecological explanation and prediction across space and time. Annu Rev Ecol Evol Syst. Annual Reviews; 2009;40: 677–697.
- 21. Rangel TF, Loyola RD. Labeling ecological niche models. Nat Conserv. 2012;10: 119–126.
- 22. Anderson RP, Lew D, Peterson AT. Evaluating predictive models of species’ distributions: criteria for selecting optimal models. Ecol Modell. 2003;162: 211–232.
- 23. Scott JM, Davis F, Csuti B, Noss R, Butterfield B, Groves C, et al. Gap analysis: a geographic approach to protection of biological diversity. Wildl Monogr. 1993; 3–41.
- 24. Elith J, Burgman MA. Predictions and their validation: rare plants in the Central Highlands, Victoria, Australia. Predict species Occur issues accuracy scale. 2002; 303–314.
- 25. Guisan A, Broennimann O, Engler R, Vust M, Yoccoz NG, Lehmann A, et al. Using niche-based models to improve the sampling of rare species. Conserv Biol. 2006;20: 501–511. pmid:16903111
- 26. Williams JN, Seo C, Thorne J, Nelson JK, Erwin S, O’Brien JM, et al. Using species distribution models to predict new occurrences for rare plants. Divers Distrib. 2009;15: 565–576.
- 27. Zhao L, Gong J, Zhang X, Liu Y, Ma X, Ren Y. Floral organogenesis in Urophysa rockii, a rediscovered endangered and rare species of Ranunculaceae. Botany. 2016;94: 215–224.
- 28. Pearson RG, Raxworthy CJ, Nakamura M, Townsend Peterson A. Predicting species distributions from small numbers of occurrence records: a test case using cryptic geckos in Madagascar. J Biogeogr. 2007;34: 102–117.
- 29. Marini MÂ, Barbet-Massin M, Lopes LE, Jiguet F. Predicting the occurrence of rare Brazilian birds with species distribution models. J Ornithol. 2010;151: 857–866.
- 30. De Siqueira MF, Durigan G, de Marco Júnior P, Peterson AT. Something from nothing: using landscape similarity and ecological niche modeling to find rare plant species. J Nat Conserv. 2009;17: 25–32.
- 31. Silva DP, Aguiar AJC, Melo GAR, Anjos-Silva EJ, De Marco P Jr. Amazonian species within the Cerrado savanna: new records and potential distribution for Aglae caerulea (Apidae: Euglossini). Apidologie. 2013;44: 673–683.
- 32. Jimenez AL, Nekaris A, Lee J, Thompson S. Modelling distributions for Colombian spider monkeys (Ateles sp.) to find priority areas for conservation. American Journal of Primatology. 2005. p. 131.
- 33. Martínez I, Carreño F, Escudero A, Rubio A. Are threatened lichen species well-protected in Spain? Effectiveness of a protected areas network. Biol Conserv. 2006;133: 500–511.
- 34. Nóbrega CC, De Marco P. Unprotecting the rare species: a niche-based gap analysis for odonates in a core Cerrado area. Divers Distrib. 2011;17: 491–505.
- 35. Peterson AT, Robins CR. Using Ecological-Niche Modeling to Predict Barred Owl Invasions with Implications for Spotted Owl Conservation. Conserv Biol. 2003;17: 1161–1165.
- 36. Thuiller W, Richardson DM, Pysek P, Midgley GF, Hughes GO, Rouget M. Niche-based modelling as a tool for predicting the risk of alien plant invasions at a global scale. Glob Chang Biol. 2005;11: 2234–2250.
- 37. Peterson AT, Papeș M, Reynolds MG, Perry ND, Hanson B, Regnery RL, et al. Native-range ecology and invasive potential of Cricetomys in North America. J Mammal. 2006;87: 427–432.
- 38. Silva DP, Gonzalez VH, Melo GAR, Lucia M, Alvarez LJ, De Marco P. Seeking the flowers for the bees: integrating biotic interactions into niche models to assess the distribution of the exotic bee species Lithurgus huberi in South America. Ecol Modell. Elsevier; 2014;273: 200–209.
- 39. Marini MÂ, Barbet-Massin M, Lopes LE, Jiguet F. Predicted Climate-Driven Bird Distribution Changes and Forecasted Conservation Conflicts in a Neotropical Savanna. Conserv Biol. 2009;23: 1558–1567. pmid:19500118
- 40. Silva DP, Macêdo ACBA, Ascher JS, De Marco P Jr. Range increase of a Neotropical orchid bee under future scenarios of climate change. J Insect Conserv. 2015;19: 901–910.
- 41. Hirzel AH, Hausser J, Chessel D, Perrin N. Ecological-niche factor analysis: how to compute habitat-suitability maps without absence data? Ecology. 2002;83: 2027–2036.
- 42. Martinez-Meyer E, Peterson AT, Servín JI, Kiff LF. Ecological niche modelling and prioritizing areas for species reintroductions. Oryx. 2006;40: 411–418.
- 43. Loiselle BA, Howell CA, Graham CH, Goerck JM, Brooks T, Smith KG, et al. Avoiding pitfalls of using species distribution models in conservation planning. Conserv Biol. 2003;17: 1591–1600.
- 44. Wilson KA, Westphal MI, Possingham HP, Elith J. Sensitivity of conservation planning to different approaches to using predicted species distribution data. Biol Conserv. 2005;122: 99–112.
- 45. Ko C-Y, Lin R-S, Ding T-S, Hsieh C-H, Lee P-F. Identifying biodiversity hotspots by predictive models: a case study using Taiwan’s endemic bird species. Zool Stud. 2009;48: 418–431.
- 46. Margules CR, Pressey RL. Systematic conservation planning. Nature. 2000;405: 243–253. pmid:10821285
- 47. Austin M. Species distribution models and ecological theory: a critical assessment and some possible new approaches. Ecol Modell. 2007;200: 1–19.
- 48. Elith J, Graham CH, Anderson RP, Dudík M, Ferrier S, Guisan A, et al. Novel methods improve prediction of species’ distributions from occurrence data. Ecography. 2006;29: 129–151.
- 49. Fagundes CK, Vogt RC, De Marco P Júnior. Testing the efficiency of protected areas in the Amazon for conserving freshwater turtles. Divers Distrib. 2016;22: 123–135.
- 50. Jennings MD. Gap analysis: concepts, methods, and recent results. Landsc Ecol. 2000;15: 5–20.
- 51. Wilson JA, Acheson JM, Metcalfe M, Kleban P. Chaos, complexity and community management of fisheries. Mar Policy. 1994;18: 291–305.
- 52. Camm JD, Polasky S, Solow A, Csuti B. A note on optimization models for reserve site selection. Biol Conserv. 1994;
- 53. Nantel P, Bouchard A, Brouillet L, Hay S. Selection of areas for protecting rare plants with integration of land use conflicts: a case study for the west coast of Newfoundland, Canada. Biol Conserv. 1998;84: 223–234.
- 54. Araujo MB. Distribution patterns of biodiversity and the design of a representative reserve network in Portugal. Divers Distrib. 1999;5: 151–163.
- 55. Pressey RL, Watts ME, Barrett TW, Ridges MJ. The C-Plan conservation planning system: origins, applications, and possible futures. Spat Conserv prioritization Quant methods Comput tools. 2009; 211–234.
- 56.
Ab’Saber AN. Espaços ocupados pela expansão dos climas secos na América do Sul, por ocasião dos períodos glaciais quaternários. Universidade de São Paulo/Instituto de Geografia; 1977.
- 57.
Mello CF, Mochel FR, Silveira OFM, Santos VF, Prost MT, Mendes A, et al. IBAMA: Diagnóstico para avaliação e ações prioritárias para conservação da biodiversidade da zona costeira-estuarina dos Estados do Piauí, Maranhão, Pará e Amapá. 2000.
- 58.
Instituto Brasileiro de Geografia e Estatística. Mapas temáticos do Brasil. 2013. mapas.ibge.gov.br
- 59.
Instituto Brasileiro do Meio Ambiente, Ministério do Meio Ambiente. Lista das espécies da fauna brasileira ameaçada de extinção. Brasilia; 2014. www.ibama.gov.br
- 60. Ratter JA, Ribeiro JF, Bridgewater S. The Brazilian cerrado vegetation and threats to its biodiversity. Ann Bot. 1997;80: 223–230.
- 61.
Klink CA, Moreira AG. Past and current human occupation, and land use. In: Oliveira PS, Marquis RJ, editors. The cerrados of Brazil: ecology and natural history of a neotropical savanna. New York: Columbia University Press; 2002. pp. 69–88.
- 62. Klink CA, Machado RB. Conservation of the Brazilian cerrado. Conserv Biol. 2005;19: 707–713.
- 63. Carvalho FM V, De Marco P, Ferreira LG. The Cerrado into-pieces: Habitat fragmentation as a function of landscape use in the savannas of central Brazil. Biol Conserv. 2009;142: 1392–1403.
- 64.
Instituto Nacional de Pesquisa Espacial. Taxa estimada do desmatamento da Amazônia Legal para período ago/2014 –jul/2015. Brasília; 2015. www.obt.inpe.br
- 65. Da Silva J, Cardoso M, Rylands AB, Fonseca D, Gustavo AB. The fate of the Amazonian areas of endemism. Conserv Biol. 2005;19: 689–694.
- 66.
Oren DC. Aves do estado do Maranhão, Brasil. Belém: Museu Paraense Emílio Goeldi; 1991.
- 67.
Oren DC, Roma JC. Composição e vulnerabilidade da avifauna da Amazônia Maranhense, Brasil. In: Martins MB, Oliveira TG, editors. Amazônia Maranhense—diversidade e conservação. Belém: Museu Paraense Emílio Goeldi; 2011. pp. 221–248.
- 68.
Comitê Brasileiro de Ornitologia. Listas das aves do Brasil. 2014. www.cbro.org.br
- 69.
Ministério do Meio Ambiente. Plano de ação para a prevenção e controle do desmatamento no estado no Maranhão. 2011.
- 70.
FIEMA. Plano Estratégico de Desenvolvimento Industrial do Maranhão. 2009. www.fiema.org.br
- 71.
Ministério do Meio Ambiente. Áreas prioritárias para a conservação, utilização sustentável e repartição de benefícios da biodiversidade Brasileira. Brasília; 2003.
- 72.
International Union for Conservation of Nature. The IUCN Red List of Threatened Species. 2012. www.iucnredlist.org
- 73. Da Silva JMC. Birds of the cerrado region, South America. Steenstrupia. 1995;21: 69–92.
- 74. Da Silva JMC. Endemic bird species and conservation in the Cerrado region, South America. Biodivers Conserv. 1997;6: 435–450.
- 75. Silva JMC, Bates JM. Biogeographic Patterns and Conservation in the South American Cerrado: A Tropical Savanna Hotspot. Bioscience. 2002;52: 225–233.
- 76.
SILVA JMC da, Santos MPD, Scariot A. A importância relativa dos processos biogeográficos na formação da avifauna do Cerrado e de outros biomas brasileiros. Cerrado Ecol biodiversidade e Conserv MMA Brasília, DF. 2005;
- 77.
Paynter RA, Traylor MA. Ornithological gazetteer of Brazil. Bird Department, Museum of Comparative Zoology, Harvard University; 1991.
- 78. Guisan A, Thuiller W. Predicting species distribution: offering more than simple habitat models. Ecol Lett. 2005;8: 993–1009.
- 79. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol Modell. 2006;190: 231–259.
- 80. Phillips SJ, Dudík M. Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography. 2008;31: 161–175.
- 81. Graham CH, Elith J, Hijmans RJ, Guisan A, Townsend Peterson A, Loiselle BA. The influence of spatial errors in species occurrence data used in distribution models. J Appl Ecol. 2008;45: 239–247.
- 82. Hernandez PA, Graham CH, Master LL, Albert DL. The effect of sample size and species characteristics on performance of different species distribution modeling methods. Ecography. 2006;29: 773–785.
- 83. Wisz MS, Hijmans RJ, Li J, Peterson AT, Graham CH, Guisan A. Effects of sample size on the performance of species distribution models. Divers Distrib. 2008;14: 763–773.
- 84. Liu C, Berry PM, Dawson TP, Pearson RG. Selecting thresholds of occurrence in the prediction of species distributions. Ecography. 2005;28: 385–393.
- 85. Liu C, White M, Newell G. Measuring and comparing the accuracy of species distribution models with presence–absence data. Ecography. 2011;34: 232–243.
- 86. Swets JA. Measuring the accuracy of diagnostic systems. Science. 1988;240: 1285–1293. pmid:3287615
- 87. Allouche O, Tsoar A, Kadmon R. Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS). J Appl Ecol. 2006;43: 1223–1232.
- 88. Girardello M, Griggio M, Whittingham MJ, Rushton SP. Identifying important areas for butterfly conservation in Italy. Anim Conserv. 2009;12: 20–28.
- 89. Marmion M, Parviainen M, Luoto M, Heikkinen RK, Thuiller W. Evaluation of consensus methods in predictive species distribution modelling. Divers Dstributions. 2009;15: 59–69.
- 90.
Ministério do Meio Ambiente. Monitoramento do Desmatamento dos Biomas Brasileiros. 2008. mapas.mma.gov.br
- 91.
Leite LO. Análise de endemismo, variação geográfica e distribuição potencial das espécies de aves endêmicas do cerrado. 2006. Universidade de Brasília, Brasília, pp 187.
- 92. Pinheiro RT, Dornas T. Distribuição e conservação das aves na região do Cantão, Tocantins: ecótono Amazônia/Cerrado. Biota Neotrop. 2009;9: 187.
- 93. Rego MA, Silveira LF, Piacentini V de Q, Schunck F, Machado É, Pinheiro RT, et al. The birds of Estação Ecológica Serra Geral do Tocantins. Biota Neotrop. 2011;11: 283–297.
- 94. Lima-Ribeiro MS, Diniz-Filho JAF. Modelando a distribuição geográfica das espécies no passado: Uma abordagem promissora em paleoecologia. Rev Bras Paleontol. 2012;15: 371–385.
- 95. Fuller T, Munguía M, Mayfield M, Sánchez-Cordero V, Sarkar S. Incorporating connectivity into conservation planning: a multi-criteria case study from central Mexico. Biol Conserv. 2006;133: 131–142.
- 96. Hurlbert AH, White EP. Disparity between range map and survey-based analyses of species richness: patterns, processes and implications. Ecol Lett. 2005;8: 319–327.
- 97. Heikkinen RK, Luoto M, Araújo MB, Virkkala R, Thuiller W, Sykes MT. Methods and uncertainties in bioclimatic envelope modelling under climate change. Prog Phys Geogr. 2006;30: 751–777.
- 98. Tarroso P, Pereira RJ, Martínez-Freiría F, Godinho R, Brito JC. Hybridization at an ecotone: ecological and genetic barriers between three Iberian vipers. Mol Ecol. 2014;23: 1108–1123. pmid:24447270
- 99. Vale CG, Tarroso P, Brito JC. Predicting species distribution at range margins: testing the effects of study area extent, resolution and threshold selection in the Sahara–Sahel transition zone. Divers Distrib. 2014;20: 20–33.
- 100. Dornas Túlio, Ramos Leandro, Pinheiro RT, Barbosa M de O. importantes e inéditos registros de aves para o ecótono amazônia/cerrado no centro norte do estado do tocantins: implicações biogeográficas e extensão de distribuição geográfica de aves amazônicas. Rev Bras Ornitol. 2012;20: 119–127.
- 101. Rodrigues ASL, Akcakaya HR, Andelman SJ, Bakarr MI, Boitani L, Brooks TM, et al. Global gap analysis: priority regions for expanding the global protected-area network. Bioscience. 2004;54: 1092–1100.
- 102. Rodrigues ASL, Andelman SJ, Bakarr MI, Boitani L, Brooks TM, Cowling RM, et al. Effectiveness of the global protected area network in representing species diversity. Nature. 2004;428: 640–643. pmid:15071592
- 103. Lees AC, Moura NG, Santana A, Aleixo A, Barlow J, Berenguer E, et al. Paragominas: a quantitative baseline inventory of an eastern Amazonian avifauna. Rev Bras Ornitol. 2012;20: 93–118.
- 104.
Payne RB. Family Cuculidae (cuckoos). In: Hoyo J del, Elliott A, Christie D, editors. Handbook of the Birds of the World, vol 4: Sandgrouse to cuckoos. Lynx Ediciones. Barcelona; 1997.
- 105.
Del Hoyo J. Family Cracidae. In: Hoyo J del, Elliot A, Sargatal J, editors. Handbook of the Birds of the World, vol 2: New World vultures to guineafowl. Lynx Ediciones. Barcelona; 1994.
- 106. Portes CEB, Carneiro LS, Schunk F, e Silva MS, Zimmer KJ, Whittaker A, et al. Annotated checklist of birds recorded between 1998 and 2009 at nine areas in the Belém area of endemism, with notes on some range extensions and the conservation status of endangered species. Brazilian J Ornithol. 2013;19: 19.
- 107.
Silveira LF. Crax fasciolata pinima. In: Machado ABM, Drummond GM, Paglia AP, editors. Livro vermelho da fauna brasileira ameaçada de extinção. Brasília: Ministério do Meio Ambiente; 2008.
- 108. Moura NG, Lees AC, Aleixo A, Barlow J, Dantas SM, Ferreira J, et al. Two hundred years of local avian extinctions in Eastern Amazonia. Conserv Biol. 2014;28: 1271–1281. pmid:24779443
- 109. Marini MÂ, Barbet-Massin M, Lopes LE, Jiguet F. Major current and future gaps of Brazilian reserves to protect Neotropical savanna birds. Biol Conserv. 2009;142: 3039–3050.
- 110. Miller MP, Bianchi CA, Mullins TD, Haig SM. Associations between forest fragmentation patterns and genetic structure in Pfrimer’s Parakeet (Pyrrhura pfrimeri), an endangered endemic to central Brazil’s dry forests. Conserv Genet. 2013;14: 333–343.
- 111.
Del Hoyo J. AE& DAC. Handbook of the birds of the world. Broadbills to Tapaculos. vol. 8. Lynx Edici. Del Hoyo J. AE & DAC, editor. Barcelona; 2003.
- 112. Peres CA. Why we need megareserves in Amazonia. Conserv Biol. 2005;19: 728–733.
- 113. Laurance WF. When bigger is better: the need for Amazonian mega-reserves. Trends Ecol Evol. 2005;20: 645–648. pmid:16701449
- 114. Ahmed SE, Lees AC, Moura NG, Gardner TA, Barlow J, Ferreira J, et al. Road networks predict human influence on Amazonian bird communities. Proc R Soc B. 2014. p. 20141742. pmid:25274363
- 115. Noss RF. Beyond Kyoto: forest management in a time of rapid climate change. Conserv Biol. 2001;15: 578–590.
- 116. Fearnside PM. Soybean cultivation as a threat to the environment in Brazil. Environ Conserv. 2001;28: 23–38.
- 117. Laurance WF, Cochrane MA, Bergen S, Fearnside PM, Delamônica P, Barber C, et al. The future of the Brazilian Amazon. Science. American Association for the Advancement of Science; 2001;291: 438–439. pmid:11228139
- 118. Redford KH, Stearman AM. Forest-Dwelling Native Amazonians and the Conservation of Biodiversity: Interests in Common or in Collision? Conserv Biol. 1993;7: 248–255.
- 119.
Leal IR, da Silva JMC. Ecologia e conservação da Caatinga. Editora Universitária UFPE; 2003.
- 120.
Ramos-Fernandez G, Wallace RB. 13 Spider monkey conservation in the twenty-first century: recognizing risks and opportunities. Spider Monkeys Biol Behav Ecol Genus Ateles. Cambridge University Press; 2008;55: 351.
- 121. Galetti M, Giacomini HC, Bueno RS, Bernardo CSS, Marques RM, Bovendorp RS, et al. Priority areas for the conservation of Atlantic forest large mammals. Biol Conserv. 2009;142: 1229–1241.
- 122. Vieira ICG, Toledo PM de, Silva JMC da, Higuchi H. Deforestation and threats to the biodiversity of Amazonia. Brazilian J Biol. 2008;68: 949–956.
- 123.
Rylands AB, Pinto LP de S. Conservação da biodiversidade na Amazônia brasileira: uma análise do sistema de unidades de conservação. Fundação Brasileira para o Desenvolvimento Sustentável; 1998.
- 124.
De Luca AC, Develey PF, Bencke GA, Goerck JM. Áreas Importantes para a Conservação das Aves no Brasil. Parte II-Amazônia, Cerrado e Pantanal. São Paulo: SAVE Brasil; 2009.
- 125.
Couto RG. Atlas de conservação da natureza brasileira–Unidades Federais. São Paulo: Metalivros; 2004.
- 126.
Machado RB, Ramos Neto MB, Pereira PGP, Caldas EF, Gonçalves DA, Santos NS, et al. Estimativas de perda da área do Cerrado brasileiro. Conservation International do Brasil. Brasília; 2004.
- 127. Diniz-Filho JAF, Bini LM, Vieira CM, De Souza MC, Bastos RP, Brandão D, et al. Spatial patterns in species richness and priority areas for conservation of anurans in the Cerrado region, Central Brazil. Amphibia-Reptilia. 2004;25: 63–75.
- 128. Diniz-Filho JAF, Bini LM, Pinto MíP, Terribile LC, de Oliveira G, Vieira CM, et al. Conservation planning: a macroecological approach using the endemic terrestrial vertebrates of the Brazilian Cerrado. Oryx. 2008;42: 567–577.
- 129.
Rylands AB. The status of conservation areas in the Brazilian Amazon. WWF; 1991.
- 130.
Instituto Socioambiental, Instituto do Homem e Meio Ambiente da Amazônia, Instituto de Pesquisa Ambiental da Amazônia, Instituto Sociedade População e Natureza, Grupo de Trabalho Amazônico, Conservation International. Seminário Consulta de Macapá 99: avaliação e identificação de ações prioritárias para a conservação, utilização sustentável e repartição dos benefícios da biodiversidade na Amazônia. São Paulo; 1999.
- 131. Veríssimo A, Cochrane MA, Souza C Jr, Salomão R. Priority areas for establishing national forests in the Brazilian Amazon. Conserv Ecol. 2002;6: 4.