Habitat Suitability for Primate Conservation in North-East Brazil

Habitat suitability for primate conservation in north-east Brazil

B ÁRBARA M ORAES,ORLY R AZGOUR,JOÃO P EDRO S OUZA-ALVES J EAN P. BOUBLI and B RUNA B EZERRA

Abstract Brazil has a high diversity of primates, but in- Introduction creasing anthropogenic pressures and climate change  could influence forest cover in the country and cause fu- razil is home to species of non-human primates   ture changes in the distribution of primate populations. B(Estrada et al., ; Costa-Araújo et al., ; IUCN,   Here we aim to assess the long-term suitability of habitats ), of which occur in the Atlantic Forest, Cerrado for the conservation of three threatened Brazilian primates and Caatinga biomes in the north-east. These three biomes (Alouatta belzebul, Sapajus flavius and Sapajus libidinosus) have been extensively modified by centuries of anthropo- through () estimating their current and future distributions genic forest destruction for the development of agriculture, using species distribution models, () evaluating how much infrastructure and urban areas. Primate populations are in of the areas projected to be suitable is represented within sharp decline, including the most charismatic, elusive and protected areas and priority areas for biodiversity conser- rare species. Half of the primate species in north-east Brazil vation, and () assessing the extent of remaining forest cov- are under threat, including five species categorized as er in areas predicted to be suitable for these species. We Endangered and five categorized as Critically Endangered  found that % of the suitable areas are outside protected on the IUCN Red List of Threatened Species (IUCN, ). areas and only % are located in areas with forest cover. Although much attention has been, rightfully, devoted Although not within protected areas, % of the climatically to the plight of the Brazilian Atlantic Forest, there has been suitable areas are considered priority areas for conserva- little focus on the Cerrado and Caatinga. The Caatinga in par- tion. Future projections, considering a severe climate change ticular has been neglected in terms of conservation action, scenario, indicate that A. belzebul, S. flavius and S. libidi- although nearly half has already been lost (Beuchle et al.,  nosus may lose up to ,  and % of their suitable ). It is predicted that the Atlantic Forest, Cerrado and range, respectively. The establishment of primate popula- Caatinga biomes will be severely affected in the future by con- tions and their long-term survival in these areas are at tinuing anthropogenic impacts in these rapidly developing  risk. Mitigation actions such as the implementation of new areas, and by climate change (Marengo et al., ). Thus it protected areas, forest restoration and reduction of green- is important to determine which areas will be more severely house gas emissions will be essential for the conservation affected and where new protected areas are required to create of Brazilian primates. an effective network of protected areas that supports the future survival of primates in these biomes (Estrada et al., ). Keywords Brazil, climate change, future range shift, gap In  the Instituto Chico Mendes de Conservação da analysis, primates, priority areas, protected areas, species Biodiversidade, the national institution for biodiversity con- distribution models servation in Brazil, developed a conservation action plan for  Supplementary material for this article is available at the primates of north-east Brazil (ICMBio-CPB, ). Six  doi.org/./S of the primate species in the area were included in this conservation action plan: the red-handed howler monkey Alouatta belzebul, Caatinga howler monkey Alouatta ululata, blonde titi monkey Callicebus barbarabrownae, ’ BÁRBARA MORAES (Corresponding author, orcid.org/0000-0002-1804-9828), Coimbra-Filho s titi monkey Callicebus coimbrai, blonde JOÃO PEDRO SOUZA-ALVES ( orcid.org/0000-0002-8517-1276) and BRUNA capuchin Sapajus flavius and yellow-breasted capuchin BEZERRA ( orcid.org/0000-0003-3039-121X) Programa de Pós-Graduação Sapajus xanthosternos. These species occur in one or more em Biologia Animal, Departamento de Zoologia, Centro de Biociências, Universidade Federal de Pernambuco, Av. Prof. Moraes Rego, 1235 Cidade of the three biomes occurring in north-east Brazil, the Universitária, Recife PE, 50670-901, Brazil Atlantic Forest, Cerrado and Caatinga. The main threats to E-mail [email protected] these primates are habitat destruction and fragmentation, ORLY RAZGOUR*( orcid.org/0000-0003-3186-0313) Biosciences, University of hunting and the pet trade (ICMBio, ; ICMBio-CPB, Exeter, Exeter, UK ). One of the actions proposed in the conservation JEAN BOUBLI ( orcid.org/0000-0002-5833-9264) School of Environment and Life Sciences, University of Salford, Salford, UK action plan is to determine more accurately the current distribution of all six species and to evaluate how these dis- *Also at: Biological Sciences, University of Southampton, Southampton, UK     tributions will be affected by future human activities and Received May . Revision requested July .  Accepted  November . First published online  September . climate change (ICMBio-CPB, ). Such data are essential

Oryx, 2020, 54(6), 803–813 © The Author(s), 2020. Published by Cambridge University Press on behalf of Fauna & Flora International doi:10.1017/S0030605319001388 Downloaded from https://www.cambridge.org/core. IP address: 170.106.202.58, on 26 Sep 2021 at 02:14:37, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605319001388 804 B. Moraes et al.

for conservation planning, to ensure there is sufficient north-east of Brazil, but also areas in the north and in the suitable habitat for the survival of these species through centre-west (Fig. ). The Atlantic Forest in north-east creating new protected areas, connecting existing ones and Brazil is at low altitudes (– m; Tabarelli et al., restoring key habitats. ) with annual rainfall of ,–, mm (Rêgo & Spatial analyses such as species distribution models and Hoeflich, ). The Caatinga and the Cerrado biomes are gap analyses are useful for assessing the impact of habitat semiarid environments, with annual precipitation of – loss and fragmentation on species (Beuchle et al., ; , mm (Ratter et al., ;Prado,)and–, mm Titeux et al., ; Zwiener et al., ). Distribution models (Hunke et al., ), respectively. Annual precipitation in project potentially suitable areas for a particular species the Amazon rainforest biome is ,–, mm (Villar based on presence location records and abiotic environ- et al., ). mental data (Elith & Leathwick, ). Gap analysis assesses Alouatta belzebul has a disjunct distribution, occur- whether species or ecosystems are represented within exist- ring in the north-eastern Atlantic Forest and lower east- ing protected areas; i.e. it identifies potential conservation ern Amazon in the Brazilian states of Amapá, Pará and gaps (Rodrigues et al., ). Both techniques provide infor- Maranhão (Veiga et al., ). It is folivorous–frugivorous mation to guide efficient management actions for the con- (Pinto et al., ) and categorized as Vulnerable on the servation of a greater number of species (Rodrigues et al., IUCN Red List because its population has declined by ), and, if integrated, improve the interpretation of % over  years (Veiga et al., ). It is estimated that impacts of global change scenarios on biodiversity (Titeux the population restricted to the Atlantic Forest has only et al., ). The combination of such approaches is thus  individuals (Veiga et al., ). important for conservation planning, particularly because Sapajus flavius occurs in the Atlantic Forest and Caatin- research and conservation efforts have traditionally focused ga of north-east Brazil (Martins et al., ; Valença on charismatic species and protected areas, leaving some Montenegro et al., ). This species has a generalist diet species and their habitats at high risk of extinction (de Souza & Ferreira, ; Medeiros et al., ) and is (Bezanson & McNamara, ). categorized as Endangered on the IUCN Red List because Here we combine species distribution models and gap of habitat loss and fragmentation resulting from coastal analysis to assess the long-term suitability of habitats for development and sugar cane plantations (Valença Monte- the conservation of two of the six target species included negro et al., ). It is recognized as one of the most threa- in the conservation action plan: A. belzebul and S. flavius. tened primates globally (Mittermeier et al., ), although it We also include a third species, the bearded capuchin is no longer included in the list of the top  most endangered Sapajus libidinosus. This species is not included in the na- primate species (Schwitzer et al., ). tional action plan but is strongly affected by habitat loss Sapajus libidinosus inhabits dry forests in semiarid (Beuchle et al., ; Rylands & Kierulff, ), the illegal areas, including the Caatinga and Cerrado biomes (Rylands pet trade (Nascimento et al., ) and hunting for use & Kierulff, ). Although categorized as Least Concern in traditional medicine and in retaliation for crop use on the IUCN Red List (Rylands & Kierulff, ), the (Torres Junior et al., ; Freire-Filho et al., ; Souto Brazilian government considers it to be Near Threatened et al., ). We chose these three species because their ha- (ICMBio, ) and it is likely to become more threatened bitats represent the main biomes of north-east Brazil and because of habitat loss (Beuchle et al., ; Rylands & they could thus serve as flagship species, with their pro- Kierulff, ) and illegal pet trade (Nascimento et al., ). tection providing wider benefits for the conservation of thesehabitatsandotherwildlifewithinthem.We()estimate the current potential range of each species and project the Methods effects of future climate change on their ranges, () evaluate the extent of suitable areas that overlap with existing pro- Species distribution modelling tected areas and proposed priority areas for biodiversity conservation, and () assess how much forest cover still re- We obtained occurrence data for the three target species mains in the areas predicted as suitable for the occurrence from the Global Biodiversity Information Facility ()and of these species. speciesLink (). We also retrieved location records from the literature using the search terms Sapajus, Sapajus libidinosus, capuchin monkeys, Cebus, Cebus libidinosus, flavius, Study area and species Cebus flavius, Alouatta, Alouatta belzebul, guariba, bugio, bugio-de-mãos-ruivas, macaco-prego, macaco-prego-galego, The study area comprises the known distribution area of macaco-prego-da-caatinga, red-handed howler monkey, the three target species, the Caatinga, Cerrado, Amazon howler monkey, blonde capuchin monkey, and bearded and Atlantic Forest biomes, encompassing mainly the capuchin monkey in ScienceDirect (Elsevier, Amsterdam,

same buffer but in pixels with different environmental characteristics. This procedure was performed to avoid a sampling bias, whereby clusters tend to give greater weight to environmental variables (Renner et al., ). Models were projected into the future () based on  general circulation models used in the th Inter- governmental Panel on Climate Change (IPCC) report (Flato et al., ): ACCESS-,HadGem-ES, Miroc-ESM, BCC-CSM-,CCSM,CNRM-CM,GFDL-CM,GISS- E-R, INMCM,IPSL-CMA-LR, MPI-ESM-LR, MRI- CGCM and NorESM-M. We selected the  scenarios, which predict climate towards the end of the century, because predictions based on shorter time frames would not provide an adequate parameter representation of the impact FIG. 1 Occurrence records for Alouatta belzebul, Sapajus flavius of climate change on the species. We considered two repre- and Sapajus libidinosus in the biomes and the Brazilian states sentative concentration pathways, defined by the trajectory included in our study area. of greenhouse gas emissions and subsequent radiative for-  cing (Wayne, ): . W/m (moderate climate change  The Netherlands), Web of Science (Clarivate Analytics, scenario) and . W/m (severe climate change scenario). Philadelphia, USA), Periódicos CAPES (Coordenação de We converted the continuous model output into binary Aperfeiçoamento de Pessoal de Nível Superior, Brasilia, maps using the thresholding method, which maximizes Brazil) and Google Scholar (Google, Mountain View, USA). the sum of sensitivity and specificity (Liu et al., ). To We also collected new location data for S. libidinosus during incorporate model variability while avoiding biases result-  expeditions to nine localities in the state of Pernambuco, ing from outlier model outputs we generated the final fu- Brazil, during May –March . All records were vali- ture maps for each scenario based on agreement between dated using Google Earth satellite maps (Google, Mountain . % of the Maxent general circulation models outputs View, USA) to exclude records outside forested areas, which (upper quantile). This was done using ArcGIS, by adding were probably the results of inaccurate coordinates. the binary model outputs generated from the  general cir- We generated species distribution models with MaxEnt culation models and reclassifying the resulting map, assign- .. (Phillips et al., ). This tool uses a maximum entropy ing a value of  (unsuitable) to cells that were identified as algorithm to select environmental variables that explain spe- suitable by no more than three models, and  (suitable) to cies distribution using presence-only data (Phillips et al., cells identified as suitable by more than three models. ). The background was delimited, for each species, as a We ran models with , interactions using the cloglog buffer of  km generated around the minimum convex model output. We used the ENMeval package in R .. (R polygon of all known occurrence records (Supplementary Core Team, ) to evaluate and select the best model par- Fig. ). We obtained climatic variables from WorldClim . ameterization (regularization multiplier value and number (Hijmans et al., ), Brazil ecoregions (MMA, ) and of parameters) based on the Akaike information criterion geomorphology databases (INPE, ), and generated the corrected for small sample sizes (AICc; Muscarella et al., slope layer from the altitude layer (Diva-GIS, ). To avoid ). The best fit model included three features (linear, collinearity, we only included variables that were not highly quadratic and hinge) and a regularization multiplier of . correlated (r , .; Supplementary Table ). Through prin- We evaluated the performance of the models using -fold cipal component analysis (Supplementary Table ) we se- cross-validations and the area under the receiver operator lected the most important variables to include in each curve (AUC), a measure of the ability of the model to distin- species’ model. These variables explained % of the distri- guish between presence locations and background/pseudo- bution models. All variables were downloaded at (or con- absences. We compared model AUC scores with  null  verted to) a resolution of  arc-seconds (c.  km ), which models, generated through resampling the isothermality was the cell size for the analyses, using ArcGIS . (Esri, layer in ENMTools (Warren et al., ), to determine Redlands, USA). whether our models performed significantly better than We reduced spatial autocorrelation among location re- random (Raes & ter Steege, ). cords through an environmental heterogeneity rarefaction  analysis using SDMTools box (Brown, )inArcGIS.We Gap analysis created a buffer of  km around each occurrence record and randomly removed duplicate points within the zones We carried out gap and range change analyses using the re- of the buffers. We retained records that were within the classified binary maps. We calculated the representation of

TABLE 1 Results of the species distribution models for the three studied primates, number of location records included in the models (N), results of the statistical tests used to evaluate model discrimination ability (area under the receiver operator curve; AUC) for training and test datasets and the per cent contribution of the different environmental variables.

Environmental variables contribution (%)1 Species (N) AUC train AUC test Bio2 Bio3 Bio8 Bio11 Bio12 Bio15 Bio18 Slope Eco Geom Alouatta belzebul (66) 0.91 0.87 21.30 6.41 3.97 10.76 7.97 0.43 3.68 45.42 Sapajus flavius (33) 0.98 0.97 0.69 7.78 0.15 1.65 37.46 13.75 3.08 35.41 Sapajus libidinosus (77) 0.88 0.82 4.48 0.06 0.36 3.30 9.10 6.95 60.52 15.19

 Bio, mean diurnal range (mean of the difference of the monthly maximum and minimum temperatures over  year); Bio, isothermality; Bio, mean temperature of wettest quarter; Bio, mean temperature of coldest quarter; Bio, annual precipitation; Bio, precipitation seasonality; Bio, precipitation of warmest quarter; Eco, ecoregion; Geom, geomorphology.

predicted current and future suitable areas for each species associated land, areas occupied by industrial and commer- by overlaying in ArcGIS the outputs of our models with cial complexes, buildings (which may in some cases be maps of Brazilian protected areas (ICMBio, ; MMA, located in periurban areas), Indigenous villages and mining b), priority areas for biodiversity conservation (MMA, areas. a) and forest cover (IBGE, ). To identify protected areas and priority areas that will retain climatic suitability in the future, we overlapped areas that were predicted to be Results suitable under both present and future conditions with existing protected areas and priority areas. We overlapped pri- We found  occurrence records for the three target spe- ority areas and protected areas with our modelled suitable cies,  of which were retained after validation. These com- areas to identify relevant locations for the expansion or cre- prised  records of A. belzebul,  of S. flavius and  of ation of protected areas. We also identified protected areas S. libidinosus (Supplementary Table ). that are likely to be under threat because of hunting and All species distribution models were able to discriminate other anthropogenic impacts by overlapping model outputs between true presence and pseudo-absences (AUCtest range: with a human settlement map (IBGE, ). .–.; Table ) and performed better than null models To categorize the degree of protection of the areas pre- because they fell outside the range of AUC values gener-   –  dicted by our models to be suitable for the target species, ated from null models (AUCtrain range: . . ). we considered the following categories: high protection Geomorphology, annual precipitation and ecoregion were status (protected areas of integral protection, where human the main environmental variables affecting habitat suitabil- settlement is not permitted, but certain activities such as sci- ity for the target species (Table ). The model considering entific research and ecotourism are); medium protection current conditions predicted suitable areas of , ,  status (areas under permanent protection in which sustain- and ,, km for A. belzebul,S.flavius and S. libidi- able use of natural resources and human occupation are per- nosus, respectively (Table , Fig. ). Our future models con- mitted); low protection status (protected areas in which sidering climate change predicted a reduction in the areas human settlement and development are permitted as well suitable for all species (Table , Fig. ). sustainable use of natural resources); and unprotected (areas Gap analysis showed that only ,  and % of the areas not formally protected). predicted to be suitable under current climatic conditions Areas defined by the Brazilian government as priority for A. belzebul, S. flavius and S. libidinosus, respectively, fall areas for biodiversity conservation are areas where conser- within existing protected areas (Table , Fig. ), and %of vation efforts should be directed for the planning and imple- these areas are of low protection status. Approximately % mentation of actions such as the creation of protected areas, of the areas predicted to be suitable are unprotected. licensing, inspection, and promotion of sustainable use. We In our models, we overlapped areas predicted to be suit- considered areas occupied by forests as those with trees able for the occurrence of the three target species with gov- .  m tall, including areas of dense, open, seasonal and ernment priority areas and forest cover layers. We found mixed ombrophilous forest, as well as forested savannah, that % of the suitable areas for all three target species to- forested campinarana and mangroves (IBGE, ). We de- gether fall within government priority areas for conserva- fined human settlements as areas characterized by urban tion (% in the Amazon forest, % in the Cerrado, % use, structured by buildings and road systems, where non- in the Caatinga and % in the Atlantic Forest). Only % agricultural artificial surfaces predominate (IBGE, ). of the suitable areas are currently forested (%inthe This category includes cities, towns, roads, services and Amazon forest, % in the Caatinga, .% in the Cerrado transport, power grids, communication infrastructure and and .% in the Atlantic Forest; Table ). Binary maps of

 TABLE 2 Area predicted to be suitable for the target species (km ), considering current conditions, and future () conditions moderate (RCP .) and severe (RCP .) climate change emission scenarios, and including their geographical range and future range loss.

Species Range Distribution models Area predicted to be suitable (km2) Range loss (%) A. belzebul Atlantic Forest, Amazon forest Current 671,135 Future moderate 131,792 80.36 Future severe 40,607 93.94 S. flavius Atlantic Forest Current 47,184 Future moderate 10,331 78.00 Future severe 1,102 97.66 S. libidinosus Caatinga, Cerrado Current 1,059,360 Future moderate 726,892 31.38 Future severe 486,616 54.00

FIG. 2 Predicted current distribution of suitable areas for the occurrence of (a) A. belzebul, (b) S. flavius and (c) S. libidinosus. Suitability ranges from low ()tohigh().

current predictive distribution with their priority areas and of government priority areas and conservation actions for forest cover are detailed in Supplementary Figs  and , each area). respectively. To identify relevant areas for the expansion or creation of Discussion protected areas, and potential connectivity between suitable areas, we overlapped government priority areas and pro- Our findings indicate concern for the future of A. belzebul, tected areas with our modelled suitable areas. We found S. flavius and S. libidinosus,as% of the areas predicted to that ,  and % of the government priority areas with- be suitable for these species are unprotected. The remaining in A. belzebul, S. flavius and S. libidinosus suitability areas, % of the suitable areas fall within protected areas, but % respectively, are outside protected areas (Supplementary of these have low protection status. Only % of the overall Table ). areas predicted to be suitable for the three species overlap Our analysis showed that %(), %() and % with priority areas for conservation, and only % are cur- () of protected areas will maintain climatic suitability rently forested. Future models predict a near total loss of under the moderate future climate change scenario for A. climatic suitability for the three species in tropical forests belzebul, S. flavius and S. libidinosus, respectively. These num- (Amazon and/or Atlantic Forest) and loss of a quarter of bers decrease to %(), %()and%()whenwecon- suitable areas in the semiarid regions (Caatinga and Cer- sider the more severe future scenario (see Supplementary rado). Table  for a list of protected areas). Of all the protected Our models were able to distinguish between the habitats areas identified as climatically suitable for the three primate of the three species. We found that habitat suitability is species under present and future conditions, %()overlap affected by geomorphology, annual precipitation and eco- with human settlements. For S. flavius, in particular, %of region, which together influence vegetation structure and climatically suitable protected areas overlap with human composition (Shi-kui et al., ). Characteristics such as settlements (Supplementary Table ). We also identified temperature variation and deciduous and semideciduous that %(), %() and %() of government priority vegetation differentiate the species’ habitat needs from areas will remain climatically suitable under the moderate those of other primates (Ratter et al., ; Prado, ). future climate change scenario for A. belzebul, S. flavius Water availability and changes in soil composition also and S. libidinosus, respectively. These numbers decrease to play an important role by controlling the type of vegetation %(), .%() and %() when we consider the more that can grow in different landscapes (El-Keblawy et al., severe future scenario (see Supplementary Table  for a list ; Cowles et al., ).

FIG. 3 Binary maps of present and future predictive distribution for (a) A. belzebul, (b) S. flavius and (c) S. libidinosus, under () current conditions, and future conditions under scenarios of () moderate (, RCP .) and () severe (, RCP .) climate change. Future predictions are based on  general circulation models for each scenario. Shaded areas indicate predicted suitable habitats above the maximum training sensitivity plus specificity threshold.

Our findings reflect the current situation of the protected Only two government priority areas were considered in areas system in Brazil, with a large proportion categorized north-east Brazil and these are located in the state of as low protection status. Currently, only % of the Atlantic Bahia, in areas where the target species do not occur. Al- Forest is under protection and only .% falls within the though government priority areas cover c. % of biomes high protection status (MMA, b). There is even less such as the Amazon, Caatinga and Cerrado, only –% protection for the Caatinga and Cerrado biomes, with % of these areas were predicted as suitable for the three of each protected and only . and %, respectively, within primate species in our study. Nevertheless, because a large the integral protection areas (MMA, b). In contrast, proportion of government priority areas within the mod- % of the Amazon falls within protected areas and % elled suitable areas is not within protected areas, it will be within integral protection areas (MMA, b). In areas necessary to identify potential connectivity areas, and areas with low protection status, extractive activities are permit- for the expansion or creation of new protected areas, to con- ted in accordance with applicable law, resulting in inade- serve these primate species. The government priority areas quate protection for species at imminent risk of extinction within the areas suitable for the target species that will remain (Schulze et al., ). In addition, some areas officially desig- suitable under future climate change (Supplementary Table ) nated as protected lack essential infrastructure or resources, are potential sites for reintroducing confiscated individuals resulting in little if any actual protection (Saout et al., ; from the illegal wildlife trade, provided these areas are under Oliveira & Bernard, ). Despite problems related to the some level of legal protection. poor management of these areas, protected areas are rele- The low forest cover in areas predicted to be suitable for vant because they prevent the conversion of natural ecosys- the three target species is mainly a result of the high an- tems (Geldmann et al., ) and support a greater diversity thropogenic impact on biomes such as the Atlantic Forest, and abundance of species than unprotected areas (Gray Caatinga and Cerrado. The north-eastern Atlantic Forest et al., ). is highly fragmented (Ribeiro et al., ), and in the Per- We observed the least overlap between habitat suitabil- nambuco Endemism Center, which includes the distribu- ity for the primate species and government priority areas tion of A. belzebul and S. flavius, % of the remnant for- for biodiversity conservation in the north-eastern Atlantic est fragments are smaller than  ha (Silva & Fialho, ). Forest. This is because most of the Atlantic Forest priority Similarly, only % of the original vegetation remains in the areas are concentrated in southern Brazil (MMA, a). Caatinga and Cerrado (MMA, ; Strassburg et al., ).

TABLE 3 Area predicted to be suitable for occurrence of the three primates and their representation in areas with different protections status (high, medium, low and unprotected), in priority areas for biodiversity conservation and in areas with forest cover.

Priority area for biodiversity Protection status (%) conservation (%) Forest cover area (%) 1 1 Predicted suitable Biome Biome Species area (km2) High Medium Low Unprotected AF AM CA CE AF AM CA CE A. belzebul 67,1135 4.56 8.03 11.80 75.61 0.27 25.39 0.98 0.44 0.28 41.93 0.49 0.76 S. flavius 47,184 0.44 0.48 7.13 91.95 1.76 8.07 6.61 1.70 S. libidinosus 1,059,360 2.32 0.35 6.75 90.58 1.09 0.80 10.12 15.22 0.57 1.29 6.17 3.88

 Biomes: AF, Atlantic Forest; AM, Amazon forest; CA, Caatinga; CE, Cerrado.

FIG. 4 Predicted suitable areas for the occurrence of (a) A. belzebul, (b) S. flavius and (c) S. libidinosus, and their overlap with protected areas of high, medium and low protection status.

In addition, habitat remnants within the Atlantic Forest and them into closer contact with people, potentially increasing semiarid zones are surrounded by an inhospitable agricul- negative human–wildlife interactions. Primates are threat- tural matrix that presents a barrier to movement and dis- ened by dogs, contact with electrical wires, the illegal pet persal (Portillo-Quintero & Sánchez-Azofeifa, ;Ribeiro trade and retaliation from farmers because of crop use et al., ). Areas of occurrence of S. flavius are also affected (Fuentes, ). by mining (e.g. Bezerra et al., ). Although some primate Our data corroborate other studies suggesting that both species are able to cross and benefit from non-forested ma- the Amazon and the Atlantic Forest will suffer significant trices, such as agricultural areas (Mandujano et al., ; biodiversity losses as a result of the combined effect of de- Canale et al., ; Souza-Alves et al., ), this brings forestation and climate change (Pires & Costa, ; Bellard

et al., ). In the Amazon, deforestation can alter the conditions we have identified areas where conservation balance of the forest and transform it into a savannah efforts should focus, to reduce habitat destruction and environment (Costa & Pires, ). Future climate change fragmentation. Creating and maintaining protected areas projections indicate a % increase in aridity, both for could help to preserve forested areas and contribute to the the Amazon rainforest and for the north-east of Brazil survival of these species. (Franchito et al., ), and an increase in temperature Our models show that future climate change could lead (Marengo et al., ). Biodiversity hotspots, such as the to substantial range losses for the three primate species. This Cerrado and the Atlantic Forest, are predicted to be adverse- is of concern as these changes can affect the establishment ly affected by future climate change and to lose c. %of of populations and their ability to survive in these areas in their endemic species (Bellard et al., ) because they the long term. Our findings will facilitate assessment of the are projected to become arid lands (Costa & Pires, ; conservation status of each species and the establishment of Franchito et al., ). It is believed that there will be goals in action plans for the conservation of other pri- changes in land use because of the projected decrease in mate species inhabiting the same regions. We will share herbaceous vegetation in the Cerrado and an increase in our findings and recommendations with key stakeholders the extent of fragmentation and conversion to pasture in such as the Centro Nacional de Pesquisa e Conservação de the Atlantic Forest (Bellard et al., ). Changes in spatial Primatas Brasileiros from the Chico Mendes Institute for configuration and habitat quality may affect the distribution Biodiversity Conservation, an administrative arm respon- and density of primate species (Estrada et al., ), as well sible for primate conservation action plans in Brazil. The as the quantity and quality of food resources available to sharing of our impartial scientific study may help such them (Dunn et al., ; Morellato et al., ). institutions plan future conservation actions, considering To suggest important areas for conservation action, we that the current Brazilian government is weakening envir- identified protected areas that will maintain climate suitabil- onmental laws, leading to further deforestation, fragmen- ity in the future and may serve as a refuge for the target spe- tation and destruction of wildlife habitats in the country cies. These areas were well represented in the areas suitable (Ferrante & Fearnside, ). for S. libidinosus, but much less so in the areas suitable for A. belzebul and S. flavius. This indicates the importance of Acknowledgements We thank Felipe França, José da Silva, Silvanete da Silva, Antônio Alencar Sampaio, Guga do Pinheiro, João da Silva expanding or creating new protected areas for the latter and Yuri Marinho for assistance with fieldwork, Agência Estadual de two species, especially in the north-east Atlantic Forest. Al- Meio Ambiente in the Pimenteira State Park, the Parnamirim Irrigated though the overall percentage of protected areas affected by Agriculture Station—UFRPE and the Programa de Pós-graduação em urban settlements was relatively low in our study (c. %), Biologia Animal (Graduate Programme in Animal Biology) for logistical a larger proportion of the areas suitable for S. flavius was support. We thank Enrico Bernard, Maria Adelia Oliveira and João Lucas Feitosa for fruitful discussions on earlier stages of this study. We affected by human settlements (c. %). Human settlements acknowledge funding to BLCM from Fundação de Amparo a Ciência close to areas suitable for these primates may result in hunt- e Tecnologia do Estado de Pernambuco (FACEPE; grant number: ing, but no data are available on current hunting pressure IBPG-1013-2.04/14) and Coordination for the Improvement of Higher on S. flavius. Overexploitation has already eradicated sev- Education Personal (CAPES; grant number: PDSE 88881.134891/ eral primate populations in Brazil, including populations of 2016-01); to JPS-A from CAPES (grant number: 527091) and FACEPE capuchin monkeys (Estrada et al., ). Strengthening en- (grant numbers: BCT-0025-2.05/17; BFP0149-2.05/10); to OR through a Natural Environment Research Council Independent Research vironmental policies and enforcing laws is key to preventing Fellowship (NE/M018660/1). This study is a new contribution from hunting and further deforestation (Brancalion et al., ; the blonde capuchin research and conservation project supported by Estrada et al., ). the Mohamed bin Zayed Species Conservation Fund, Margot Marsh Our study highlights concern regarding the conservation Biodiversity Foundation, Rufford Small Grant Foundation, FACEPE of Neotropical primates in general. We provide important (grant numbers: APQ-1534-2.04/10; APQ-0143-2.04/14; BFT-0160- 2.04/17; BFT-0014-2.05/20) and Brazilian National Council for Scien- information for the conservation of three primate species, tific and Technological Development (CNPq; grant numbers: Universal two of which are part of the Brazilian Action Plan for the 445071/2014-1; Pq2 309256/2019-4). Conservation of Primates in Northeast Brazil. Although the three target species inhabit areas considered to be of Author contributions Study design: BM, BB; fieldwork: BM; data global conservation importance (Brooks et al., ), the analysis and writing: all authors. areas predicted to be suitable for these species are mostly outside protected areas and have low forest cover, espe- Conflicts of interest None. cially in the Atlantic Forest. The Brazilian government has designated priority areas for conservation, but these are Ethical standards This research abides by the Oryx guidelines on ethical standards and was carried out in compliance with Brazilian not sufficient to maintain the primate populations we law (fieldwork licence: SISBIO/ICMBio—52404-1; license for inter- have studied. By determining suitable areas for the occur- views: Plataforma Brasil—CAAE-49198215.3.0000.5208/Approval- rence of the target species under present and future 1.266.360).

References extinction crisis of the world’s primates: why primates matter. Science Advances, ,e. BELLARD, C., LECLERC, C., LEROY, B., BAKKENES, M., VELOZ, S., ESTRADA, R.A., GARBER, P.A., MITTERMEIER, R.A., WICH, S.,  THUILLER,W.&COURCHAMP,F.() Vulnerability of GOUVEIA, S., DOBROVOLSKI, R. et al. ( ) Primates in peril: the biodiversity hotspots to global change. Global Ecology and significance of Brazil, Madagascar, Indonesia and the Democratic Biogeography, , –. Republic of the Congo for global primate conservation. PeerJ,   BEUCHLE, R., GRECCHI, R.C., SHIMABUKURO, Y.E., SELIGER, R., EVA, ,e .  ’ H.D., SANO,E.&ACHARD,F.() Land cover changes in the FERRANTE,L.&FEARNSIDE, P.M. ( ) Brazil s new president and Brazilian Cerrado and Caatinga biomes from  to  based on ‘ruralists’ threaten Amazonia’s environment, traditional peoples a systematic remote sensing sampling approach. Applied Geography, and the global climate. Environmental Conservation, , –. , –. FLATO, G., MAROTZKE, J., ABIODUN, B., BRACONNOT, P., CHOU, S.C.,  BEZANSON,M.&MCNAMARA,A.() The what and where of COLLINS,W.etal.( ) Evaluation of climate models. In Climate primate field research may be failing primate conservation. Change : The Physical Science Basis. Contribution of Working Evolutionary Anthropology: Issues, News, and Reviews, Group I to the Fifth Assessment Report of the Intergovernmental , –. Panel on Climate Change (eds T.F. Stocker, D. Qin, G.-K. Plattner, BEZERRA, B.M., BASTOS, M., SOUTO,A.&JONES,G.() Camera M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex & trap observations of nonhabituated Critically Endangered wild P.M. Midgley), pp. –. Cambridge University Press, blonde capuchins, Sapajus flavius (formerly Cebus flavius). Cambridge, UK and New York, USA. International Journal of Primatology, , –. FRANCHITO, S.H., FERNANDEZ, J.P.R. & PAREJA,D.() Surrogate BROOKS, T.M., MITTERMEIER, R.A., DA FONSECA, G.A.B., GERLACH, climate change scenario and projections with a regional climate J., HOFFMANN, M., LAMOREUX, J.F. et al. () Global biodiversity model: impact on the aridity in South America. American Journal conservation priorities. Science, , –. of Climate Change, , –.  BROWN, J.L. () SDMtoolbox: a python-based GIS toolkit for FREIRE-FILHO, R., PINTO,T.&BEZERRA,B.( ) Using local landscape genetic, biogeographic and species distribution model ecological knowledge to access the distribution of Endangered analyses. Methods in Ecology and Evolution, , –. Caatinga howler monkey (Alouatta ululata). Ethnobiology and   BRANCALION, L.C., GARCIA, R., LOYOLA, R.D., RODRIGUES, R.R., Conservation, , .  PILLAR, V.D. & LEWINSOHN, T.M. () A critical analysis of the FUENTES,A.( ) Human-Nonhuman primate interconnections Native Vegetation Protection Law of Brazil (): updates and and their relevance to anthropology. Ecological and Environmental ongoing initiatives. Natureza & Conservação, , –. Anthropology, , –.  CANALE, G.R., KIERULFF, M.C.M. & CHIVERS, D.J. () A Critically GLOBAL BIODIVERSITY INFORMATION FACILITY ( ) Free and Open Endangered capuchin monkey (Sapajus xanthosternos) living in a Access to Biodiversity Data. GBIF Secretariat, Copenhagen, highly fragmented hotspot. In Primates in Fragments. Developments Denmark. gbif.org [accessed April ]. in Primatology: Progress and Prospects (eds L. Marsh & GELDMANN, J., BARNES, M., COAD, L., CRAIGIE, I.D., HOCKINGS,M. C. Chapman), pp. –. Springer, New York, USA. &BURGESS, N.D. () Effectiveness of terrestrial protected areas COSTA, M.H. & PIRES, G.F. () Effects of Amazon and central in reducing habitat loss and population declines. Biological Brazil deforestation scenarios on the duration of the dry season Conservation, , –. in the arc of deforestation. International Journal of Climatology, GRAY, C.L., HILL, S.L.L., NEWBOLD, T., HUDSON, L.N., BÖRGER, L., , –. CONTU, S. et al. () Local biodiversity is higher inside than COSTA-ARAÚJO, R., MELO, F.R., CANALE, G.R., HERNÁNDEZ- outside terrestrial protected areas worldwide. Nature   RANGEL, S.M., MESSIAS, M.R., ROSSI, R.V. et al. () The Communications, , . Munduruku marmoset: a new monkey species from southern HIJMANS, R.J., CAMERON, S.E., PARRA, J.L., JONES, P.G. & JARVIS,A. Amazonia. PeerJ, ,e. () Very high resolution interpolated climate surfaces for global  – COWLES, J., BOLDGIV, B., LIANCOURT, P., PETRAITIS, P.S. & CASPER, land areas. International Journal Climatology, , . B.B. () Effects of increased temperature on plant communities HUNKE, P., MUELLER, E.N., SCHRÖDER,B.&ZEILHOFER,P.() depend on landscape location and precipitation. Ecology and The Brazilian Cerrado: assessment of water and soil degradation in Evolution, , –. catchments under intensive agricultural use. Ecohydrology: DE SOUZA LINS, P.G.A. & FERREIRA, R.G. () Competition during Ecosystems, Land and Water Process Interactions, sugarcane crop raiding by blond capuchin monkeys (Sapajus Ecohydrogeomorphology, , –. flavius). Primates, , –. IBGE (INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICAS)  – – DIVA-GIS () Free Spatial Data – Country Level – Elevation. diva- ( ) Mudanças na Cobertura e Uso da Terra do Brasil gis.org/data [accessed March ]. –. Ministério de Planejamento, Orçamento e Gestão, DUNN, J.C., CRISTÓBAL-AZKARATE,J.&VEA,J.() Differences in Rio de Janeiro, Brazil. diet and activity pattern between two groups of Alouatta palliata ICMBIO (INSTITUTO CHICO MENDES DE CONSERVAÇÃO DA associated with the availability of big trees and fruit of top food taxa. BIODIVERSIDADE)() Listas das Espécies da Fauna Brasileira American Journal Primatology, , –. Ameaçadas de Extinção Vigentes. icmbio.gov.br/portal/   ELITH,J.&LEATHWICK, J.R. () Species distribution models: faunabrasileira/lista-de-especies [accessed August ]. ecological explanation and prediction across space and time. Annual ICMBIO (INSTITUTO CHICO MENDES DE CONSERVAÇÃO DA Review of Ecology, Evolution and Systematics, , –. BIODIVERSIDADE)() Unidades de Conservação Federais. icmbio.    EL-KEBLAWY, A., ABDELFATTAH, M.A. & KHEDR,A.() gov.br/portal/geoprocessamento [accessed December ]. Relationships between landforms, soil characteristics and dominant ICMBIO (INSTITUTO CHICO MENDES DE CONSERVAÇÃO DA xerophytes in the hyper-arid northern United Arab Emirates. BIODIVERSIDADE)—CPB (CENTRO NACIONAL DE PESQUISA E Journal of Arid Environments , –. CONSERVAÇÃO DE PRIMATAS)() Plano de Ação Nacional para   ESTRADA, A., GARBER, P.A., RYLANDS, A.B., ROOS, C., Conservação de Primatas do Nordeste. Portaria no. ,de de  FERNANDEZ-DUQUE,E.&DI FIORE,A.() Impending março de . icmbio.gov.br/portal/faunabrasileira/

plano-de-acao-nacional-lista/-pan-primatas-do-nordeste areas of logged and unlogged forest in eastern Amazonia. Animal [accessed  February ]. Biodiversity and Conservation, , –. INPE (INSTITUTO NACIONAL DE PESQUISAS ESPACIAIS)() PIRES, G.F. & COSTA, M.H. () Deforestation causes different Variáveis Ambientais para Modelagem de Distribuição de Espécies. subregional effects on the Amazon bioclimatic equilibrium. dpi.inpe.br/ambdata/mapa_solos.php [accessed November ]. Geophysical Research Letters, , –. IUCN () The Red List of Threatened Species. Version -. PORTILLO-QUINTERO, C.A. & SÁNCHEZ-AZOFEIFA, G.A. () iucnredlist.org [accessed  August ]. Extent and conservation of tropical dry forests in the Americas. LIU, C., WHITE,M.&NEWELL,G.() Selecting thresholds for the Biological Conservation, , –. prediction of species occurrence with presence-only data. Journal PRADO, D.E. () As caatingas da América do Sul. In Ecologia e of Biogeography, , –. Conservação da Caatinga (eds I.R. Leal, M. Tabarelli & J.M.C. Silva), MANDUJANO, S., ESCOBEDO-MORALES, L.A. & PALACIOS-SILVA,R. pp. –. Universitária da UFPE, Recife, Brazil. () Movements of Alouatta palliata among forest fragments in RAES,N.&TER STEEGE,H.() A null-model for significance Los Tuxtla, México. Neotropical Primates, , –. testing of presence-only species distribution models. Ecography, MARENGO, J.A., TORRES, R.R. & ALVES, L.M. () Drought in , –. Northeast Brazil—past, present, and future. Theoretical and Applied RATTER, J.A., RIBEIRO, J.F. & BRIDGEWATER,S.() The Brazilian Climatology, , –. Cerrado vegetation and threats to its biodiversity. Annals of Botany, MARTINS, A.B., VALENÇA-MONTENEGRO, M.M., FIALHO, M.S., , –. LAROQUE, P.O. & DI FIORE,A.() Range expansion and RCORE TEAM () A Language and Environment for Statistical observation of tool use by blonde capuchins, Sapajus flavius, in the Computing. R Foundation for Statistical Computin, Vienna, Austria. Caatinga biome of Brazil. Presented at Meeting of the International R-project.org [accessed  March ]. Primatological Society and the American Society of Primatologists, RÊGO, G.M. & HOEFLICH, V.A. () Contribuição da Pesquisa – August, , Chicago, USA. Florestal para um Ecossistema em Extinção: Floresta Atlântica do MEDEIROS, K., BASTOS, M., JONES,G.&BEZERRA,B.() Nordeste do Brasil. Embrapa Tabuleiros Costeiros, Aracaju, Brazil. Behavior, diet, and habitat use by blonde capuchin monkeys RENNER, I.W., ELITH, J., BADDELEY, A., FITHIAN, W., HASTIE,T., (Sapajus flavius) in a coastal area prone to flooding: direct PHILLIPS, S.J. et al. () Point process models for presence-only observations and camera trapping. International Journal of analysis. Methods in Ecology and Evolution, , –. Primatology, , –. RIBEIRO, M.C., MARTENSEN, A.C., METZGER, J.P., TABARELLI, M., MITTERMEIER, R.A., RYLANDS, A.B., SCHWITZER, C., TAYLOR, L.A., SCARANO,F.&FORTIN, M.-J. () The Brazilian Atlantic Forest: CHIOZZA,F.&WILLIAMSON, E.A. (eds) () Primates in Peril: a shrinking biodiversity hotspot. In Biodiversity Hotspots (eds The World’s  Most Endangered Primates –. IUCN/SSC F.E. Zachos & J.C. Habel), pp. –. Springer, Berlin and Primate Specialist Group, International Primatological Society Heidelberg, Germany. and Conservation International, Arlington, USA. RODRIGUES, A.S.L., ANDELMAN, S.J., BAKARR, M.I., BOITANI, L., MMA (MINISTÉRIO DO MEIO AMBIENTE)() ° Relatório BROOKS, T.M., COWLING, R.M. et al. () Effectiveness of the Nacional para a Convenção sobre Diversidade Biológica. Série global protected area network in representing species diversity. Biodiversidade, v. . Brasília, Brasil: Ministério do Meio Ambiente, Nature, , –. Brasilia, Brazil. RYLANDS, A.B. & KIERULFF, M.C.M. () Sapajus libidinosus. In The MMA (MINISTÉRIO DO MEIO AMBIENTE)() World Wildlife IUCN Red List of Threatened Species : e.TA. Fund Terrestrial Ecoregions. mapas.mma.gov.br/igeo/ dx.doi.org/./IUCN.UK.-.RLTS.TA.en datadownload.htm [accessed March ]. [accessed  February ]. MMA (MINISTÉRIO DO MEIO AMBIENTE)(a) Áreas Prioritárias SAOUT, S.L., HOFFMANN, M., SHI, Y., HUGHES, A., BERNARD, C., para a Conservação da Biodiversidade. areasprioritarias.mma.gov. BROOKS, T.M. et al. () Protected areas and effective biodiversity br/oque-e [accessed  June ]. conservation. Science, , –. MMA (MINISTÉRIO DO MEIO AMBIENTE)(b) Cadastro SCHULZE, K., KNIGHS, K., COAD, L., GELDMANN, J., LEVERINGTON,F., Nacional de Unidades de Conservação. mma.gov.br/ EASSOM, A. et al. () An assessment of threats to terrestrial areas-protegidas/cadastro-nacional-de-ucs [accessed  June ]. protected areas. Conservation Letters, ,e. MORELLATO, L.P.C., ALBERTON, B., ALVARADO, S.T., BORGES, B., SHI-KUI, D., WEIA, S., XU-KUNB, S., YONGC, Z., SHUAIA, L., BUISSON, E., GABRIELA, M. et al. () Linking plant phenology XIAOXAIA, G. et al. () The impacts of geographic, soil and to conservation biology. Biology Conservation, , –. climatic factors on plant diversity, biomass and their MUSCARELLA, R., GALANTE, P.J., SOLEY-GUARDIA, M., BORIA, R.A., relationships of the alpine dry ecosystems: cases from the Aerjin KASS, J.M., URIARTE,M.&ANDERSON, R.P. () ENMeval:anR Mountain Nature Reserve, China. Ecological Engineering, package for conducting spatially independent evaluations and , –. estimating optimal model complexity for Maxent ecological niche SILVA, E.A. & FIALHO, M.S. () Geoprocessamento Aplicado ao models. Methods in Ecology and Evolution, , –. Plano de Ação Nacional para a Conservação dos Primatas do NASCIMENTO, R.A., SCHIAVETTI,A.&MONTAÑO, R.A.M. ()An Nordeste—PAN: Identificação, Monitoramento e Manejo de Áreas assessment of illegal capuchin monkey trade in Bahia State, Brazil. Estratégicas à Conectividade Funcional das Espécies-Alvo. Centro Neotropical Biology and Conservation, , –. Nacional de Pesquisa e Conservação de Primatas Brasileiros—CPB/ OLIVEIRA, A.P.C. & BERNARD,E.() The financial needs vs. the ICMBio, João Pessoa, Brazil. realities of in situ conservation: an analysis of federal funding for SPECIESLINK () speciesLink: a Distributed Information System protected areas in Brazil’s Caatinga. Biotropica, , –. that Integrates Primary Data from Biological Collections. PHILLIPS, S.J., ANDERSON, R.P. & SCHAPIRE, R.E. () Maximum splink.cria.org.br [accessed April ]. entropy modelling of species geographic distributions. Ecological SCHWITZER, C., MITTERMEIER, R.A., RYLANDS, A.B., CHIOZZA,F., Modelling, , –. WILLIAMSON, E.A., BYLER, D. et al. (eds) () Primates in Peril: PINTO, A.C.B., AZEVEDO-RAMOS,C.&DE CARVALHO,JR,O.() The World’s  Most Endangered Primates –. IUCN SSC Activity patterns and diet of the howler monkey Alouatta belzebul in Primate Specialist Group, International Primatological Society,

Bristol Zoological Society, and Global Wildlife Conservation, in a rural community in Paraíba, Brazil. Folia Primatologica, Washington, DC, USA. , –. SOUTO, W.M.S., BARBOZA, R.R.D., F ERNANDES-FERREIRA,H., VALENÇA MONTENEGRO, M., BEZERRA, B., FIALHO, M., JÚNIOR,A.J.C.M.,MONTEIRO,J.M.,ABI-CHACRA,E.A.&ALVES, JERUSALINSKY, L., LYNCH ALFARO,J.&MARTINS, A.B. () R.R.N. () Zootherapeutic uses of wildmeat and associated Sapajus flavius.InThe IUCN Red List of Threatened Species : products in the semiarid region of Brazil: general aspects and e.TA. dx.doi.org/./IUCN.UK.-.RLTS. challenges for conservation. Journal of Ethnobiology and TA.en [accessed  August ]. Ethnomedicine , . VEIGA, L.M., KIERULFF,C.&DE OLIVEIRA, M.M. () Alouatta SOUZA-ALVES, J.P., MOURTHE, I., HILÁRIO, R.R., BICCA-MARQUES, belzebul. In The IUCN Red List of Threatened Species : J.C., REHG, J., GESTICH, C.C. et al. () Terrestrial behavior in titi e.TA. dx.doi.org/./IUCN.UK..RLTS. monkeys (Callicebus, Cheracebus, and Plecturocebus): potential TA.en [accessed  February ]. correlates, patterns, and differences between genera. International VILLAR, J.C.E., RNCHAIL, J., GUYOT, J.L., COCHONNEAU, G., Journal of Primatology, , –. NAZIANO, F., LAVADO, W., DE OLIVEIRA, E. et al. () STRASSBURG, B.B.N., BROOKS, T., FELTRAN-BARBIERI, R., Spatio-temporal rainfall variability in the Amazon basin countries IRIBARREM, A., CROUZEILLES, R., LOYOLA,R.etal.()Moment (Brazil, Peru, Bolivia, Colombia, and Ecuador). International of truth for the Cerrado hotspot. Nature Ecology & Evolution, , . Journal Climatology, , –. TABARELLI, M., AGUIAR, A.V., RIBEIRO, M.C., METZGER, J.P. & WARREN, D.L., GLOR, R.E. & TURELLI,M.() ENMTools: a toolbox PERES, C.A. () Prospects for biodiversity conservation in the for comparative studies of environmental niche models. Ecography, Atlantic Forest: lessons from aging human-modified landscapes. , –. Biological Conservation, , –. WAYNE, G.P. () The Beginner’s Guide to Representative TITEUX, N., HENLE, K., MIHOUB, J.B., REGOS, A., GEIJZENDORFFER, Concentration Pathways. Skeptical Science. skepticalscience.com/ I.R. & CRAMER,W.etal.() Global scenarios for biodiversity rcp.php [accessed  March ]. need to better integrate climate and land use change. Diversity and ZWIENER, V.P., LIRA-NORIEGA, A., GRADY, C.J., PADIAL, A.A. & Distributions, , –. VITULE, J.R.S. () Climate change as a driver of biotic TORRES JUNIOR, E.U., VALENÇA-MONTENEGRO, M.M. & DE CASTRO, homogenization of woody plants in the Atlantic Forest. Global C.S.S. () Local ecological knowledge about endangered primates Ecology and Biogeography, , –.