Mammalian Biology 76 (2011) 243–250

Contents lists available at ScienceDirect

Mammalian Biology

journal homepage: www.elsevier.de/mambio

Original Investigation The cacao agroforests of the Brazilian Atlantic forest as habitat for the endangered maned Bradypus torquatus

Camila Righetto Cassano a,b, Maria Cecília Martins Kierulff c, Adriano G. Chiarello d,∗ a Instituto de Estudos Socioambientais do Sul da Bahia – IESB, Rua Araujo Pinho, 72/3, Ilhéus, BA, b Programa de Pós-graduac¸ ão em Zoologia, Universidade Estadual de Santa Cruz, km 16, Rodovia Ilhéus-Itabuna, Ilhéus, CEP:45.662-900 BA, Brazil c Instituto Pri-Matas para a Conservac¸ ão da Biodiversidade, Caixa Postal 3304, Belo Horizonte, CEP: 30.112-970, MG, Brazil d Programa de Pós-graduac¸ ão em Zoologia de Vertebrados, Pontifícia Universidade Católica de Minas Gerais – PUC-MG, Av. Dom José Gaspar 500, Prédio 41, Belo Horizonte, CEP: 30.353-610, MG, Brazil article info abstract

Article history: are arboreal strictly dependent upon forested habitats. The southern part of the state of Received 7 December 2009 Bahia in northeastern Brazil harbors important forest remnants and the highest genetic diversity known Accepted 21 June 2010 for the maned sloth (Bradypus torquatus), an endangered species endemic to the Atlantic forest. Large extents of cacao agroforests (cabrucas) connected to forest patches mitigate the effects of fragmentation Keywords: in this region. We radio-tracked three maned sloths during 40 months in a cabruca at the vicinity of Una Biological Reserve, southern Bahia, and estimated their home range using two commonly employed estimators (minimum convex polygon (MCP) and kernel). Overall cabrucas comprised a significant portion Cocoa plantation Home range of the home range of the three study (MCP: 7–100%) and at least a third of the areas of more Radio telemetry intensive use (kernel: 27–99%). The tagged sloths used cabrucas more than expected according to the Theobroma cacao availability of this habitat in their home range and in the surrounding landscape. In addition to the tagged individuals, maned sloths were observed five times in the study area, twice in cabrucas. Eleven tree species present in cabrucas were used as food sources by maned sloths. Results indicate that biologically rich cacao agroforests immersed in a landscape still largely composed of native forests, as is the case here, can provide habitat for the maned sloth. This finding spells good news for the conservation of this species, as southern Bahia is one of the most important strongholds for the maned sloth. However, further actions are necessary to protect the species from local extinction, including active management of protected areas, forest fragments, cabrucas and pastures in an integrated, landscape-level manner. © 2010 Deutsche Gesellschaft für Säugetierkunde. Published by Elsevier GmbH. All rights reserved.

Introduction is covered by “forested environments”, mainly traditional cacao plantations (26% of the landscape), secondary growth forest (19%) Land use change is one of the major forces leading to a decrease and primary forests (9%) (Landau et al. 2008); the remaining 46% is in wildlife population, reduction of species distribution, and dis- covered mainly by pastures and agriculture. In the 1990s, approxi- ruption of ecosystem function (Gutzwiller 2002). In recent years, mately 70% of the 700,000 ha planted with cacao (Theobroma cacao) a large number of studies have focused on the importance of in this region was cabrucas, traditional plantations established in the mosaic of habitats to preserve biodiversity, and some studies the shade of native canopy trees (Araújo et al. 1998). However, as a have suggested that managed environments such as agroforestry result of the cocoa economic crisis of the late 1980s associated with systems may play an important role in sustaining vertebrate biodi- the infection of a fungus disease (witches’ broom, Moniliophthora versity in human-modified landscapes (Rice and Greenberg, 2000; perniciosa), traditional cabrucas have increasingly been replaced Ricketts, 2001; Schroth and Harvey 2007; Cassano et al. 2009). by pasture or cacao plantations shaded by exotic trees or simply The Brazilian Atlantic forest is a biodiversity hotspot that has unshaded cacao. At the same time, some cabrucas have been aban- been severely fragmented. It is currently reduced to less than 12% doned or received little management. In these cases, tree recovery of its original extent (Ribeiro et al. 2009). Southern Bahia is a rela- has resulted in a dense canopy layer (Sambuichi and Haridasan tively well-preserved part of this biome where over 50% of the land 2007) that, while beneficial for native wildlife, negatively impacts cacao production, making it a non-viable management strategy. Oliver and Santos (1991) proposed that sloths and other threat- ∗ Corresponding author. Tel.: +55 31 33194967; fax: +55 31 33194967. ened mammals from the Atlantic Forest make use of cabrucas to E-mail address: [email protected] (A.G. Chiarello). some extent, and more recent research has shown the importance

1616-5047/$ – see front matter © 2010 Deutsche Gesellschaft für Säugetierkunde. Published by Elsevier GmbH. All rights reserved. doi:10.1016/j.mambio.2010.06.008 244 C.R. Cassano et al. / Mammalian Biology 76 (2011) 243–250

Fig. 1. Land cover in the region of the Una Biological Reserve and Ecoparque de Una (insert, right), northeastern Brazil (left, lines indicate state borders). The study site is located to the east of Una Biological Reserve (circle). of those areas as corridors or as additional habitat for some mam- of Una, in the lowlands of southern Bahia, Brazil. The study site is mal species (Moura 1999; Pardini, 2004; Raboy et al. 2004; Pardini located near the Una Biological Reserve and the Private Reserve et al. 2009). The maned sloth (Bradypus torquatus) is one of the six Ecoparque de Una, and is crossed by the Maruim River (Fig. 1). extant species of sloths and is listed as “endangered” by the IUCN Approximately 67% of the land encompassed by Una Biological (IUCN 2008). The maned sloth is endemic to the coastal rain for- Reserve and its buffer zone (extending 10 km beyond the Reserve’s est of eastern Brazil, occurring from the state of Sergipe to Rio de borders) is covered by primary and secondary forests (Instituto Janeiro (Wetzel and Avila-Pires 1980; Oliver and Santos 1991). The de Estudos Socioambientais do Sul da Bahia, unpublished data). species is separated into three functionally isolated populations, Shaded cacao plantations and rubber tree plantations account for each genetically homogeneous but strongly differentiated from the 60% of the cultivated land and represent the main crops in the Una others, to the extent that the northernmost population in Bahia is Biological Reserve buffer zone (Araújo et al. 1998). considered a distinct evolutionary unit requiring separate manage- Average annual temperature is 24–25 ◦C, and annual precipita- ment from the southern populations (Lara-Ruiz et al. 2008). Habitat tion is 1600–2000 mm (Gouvêa 1969). The vegetation in the study destruction is by far the main threat to the conservation of the area is tropical lowland rainforest (Oliveira-Filho and Fontes 2000). maned sloth (Machado et al. 2005; IUCN 2008). The mature humid forest at Una Biological Reserve is characterized Given the great reduction and fragmentation of mature forest by a canopy 25–30 m tall, high abundance of epiphytes and vines, remnants in southern Bahia, a better understanding of the capac- and clearly defined strata (Amorim et al. 2008). In the cabrucas, ity of maned sloths to use agroforests as permanent or temporary the understory of original forest is replaced by cacao shrubs, tree habitat has critical implications for assessing the conditions under canopy density is lower than in mature forests, epiphytes are usu- which the species can persist. Given that agroforests comprise such ally reduced to canopy layer and vines are minimal (Alves 1990, a large proportion of the matrix, we investigate the use of a cabruca 2005; Sambuichi 2006; Sambuichi and Haridasan 2007; Cassano by three maned sloths in the vicinity of Una Biological Reserve in et al. 2009). Traditional cacao agroforests exhibit great variation southern Bahia. Additionally, we describe the tree species found in in vegetation diversity and structure as a consequence of both cabrucas that can be used as a source of food by maned sloths. tree diversity of the original forest and the agroforest management (Sambuichi 2006). The cabruca where the study was conducted has Material and methods well-developed canopy strata, although it is lower and thinner than that of mature forests. Barreto and Cassano (2007) found 30 trees Study site (11 species) with a DBH > 10 cm in 0.1 ha, most were early succes- sional species. Inga edulis, Tapirira guianensis and Senna multijuga, The maned sloths were monitored in a cabruca located at the for example, together comprised over 50% of the trees. The cacao Cabana da Serra Farm (15◦1120S; 39◦233W) in the municipality plantation where the study was conducted received little manage- C.R. Cassano et al. / Mammalian Biology 76 (2011) 243–250 245 ment: weeding happened only twice during the four years of the of its range (i.e., the utility density distribution, Powell (2000)). The study, tree epiphytes were neither pruned nor removed from cacao bandwidth (h) chosen for fixed kernel was 15 m, which we con- plants, and agrochemicals were not used. sidered as a good approximation of the perception radius (Powell 2000) of this species (see also Montgomery and Sunquist 1975). 8 Data collection Scaling factor used was 10 and raster cell size was 10 due to the small areas under analysis. Three maned sloths inhabiting the cabruca, an adult female and Interpretation of aerial photographs provided by Instituto de two of its infants, were tracked with telemetry equipment (radio Estudos Socioambientais do Sul da Bahia (unpublished data) was collars model TW-3 from Biotrack Ltd., a Telonics TR-4 receiver used as an initial source for the land cover of the region. However, and a three-element Yagi antenna from Wildlife Material) between this method did not discriminate cabrucas from secondary logged October 2004 and February 2008. These radio collars weigh less forest, due to similarities in canopy structure between these two than 80 g (<2% of the adult body weight) and have not caused vegetation types. We therefore modified the original maps using injuries, scars, or any other known problems to tagged individuals field observations in order to delimitate more accurately the extent in the 10+ years we have been studying these animals (Chiarello of cabrucas in the study site. The home ranges (MCP and kernel) 1998a,b; Chiarello et al. 2004; Lara-Ruiz and Chiarello 2005; Lara- were clipped from the land cover map using the intersect func- Ruiz et al. 2008). The two cubs were tagged while they were still tion from ArcMap 9.2. We then calculate the proportion of cabruca carried by the adult female, but data on home range were col- falling within each home range using the resulting shape file. lected for these individuals only after they separated (i.e., became We used contingency tables (Chi-square goodness-of-fit-tests) independent) from the mother. complemented by residual analysis to assess if the radio-collared The sloths were located two to three times a month; consecutive sloths used the habitats according to their availabilities. First, we sightings were separated by at least 1 week. Every tree in which contrasted the number of location fixes on each major habitat type the animals were seen was mapped. To minimize location errors inside the MCPs (swamp/pasture, cabruca, early secondary for- two methods were used for mapping trees. When used trees were est and late secondary forest) with the expected frequencies of between 10 and 50 m to each other the position of these trees were fixes based on the availability of these habitats in the MCPs. We mapped using compass and measuring tapes. We avoided using pooled the swamp with pasture as all fixes falling in pasture were the GPS in these circumstances as distances were generally smaller in fact in the very edge between pasture and swamp vegetation. than the GPS error. Later the geographic coordinates (recorded by a In a second Chi-square we contrasted the observed frequency of Garmim Etrex 12 channel GPS) of at least three trees located at the fixes with the expected value based on habitat availability in a periphery of the area used by each were used as a geographic buffer area surrounding the home ranges of the three sloths. To reference to plot these trees on a map using ArcMap 9.2 Program. calculate the width of this buffer area we measured the largest Trees located more than 50 m from the nearest used tree had their distances between fixes inside each individual home range and coordinates recorded directly by the GPS. averaged them. The resulting width (487 m) encompassed an area Information on sloth diet was collected during observations of of 214.7 ha (6.8 times larger than the combined area of the three feeding behavior of the three individuals inhabiting the cabruca MCPs). Although this wide buffer may include land-use types dif- (Correia et al. 2006 and C.R. Cassano, personal observation) and also fering from the immediate study area, a narrower band would not from previous research developed in a mature forest remnant at be representative of the larger landscape context. These tests were Ecoparque de Una (Cassano 2006 and C.R. Cassano, personal obser- performed using BioEstat version 5.0, a free statistical software vation). We searched botanical inventories carried out in southern available at Instituto de Desenvolvimento Sustentável Mamirauá Bahia and compile information on plant species present in cabrucas (http://www.mamiraua.org.br). of this region that are known to be used as source of food by maned sloths. Results

Data analysis Monitoring time for individual sloths ranged from 12 to 40 months. A total of 199 locations of the three radio-collared maned Home ranges were estimated using HawthsTools, for ArcMap sloths were recorded (Table 1). Both MCP and kernel methods 9.2. For comparative purposes, home ranges were calculated using indicated that the adult female (BT464) had the smallest home two methods: the minimum convex polygon (MCP) using 100% of range, followed by the male juvenile (BT162) and the female locations, and fixed kernel using 95% of locations. Because of its juvenile (BT065) (Fig. 2). The cabrucas comprised the major- ease of calculation, many investigators use MCP to describe home ity of home range of the mother and the male juvenile and a ranges, and we therefore present these analyses to facilitate com- significant proportion of the home range of the female juve- parison with other studies. However, the kernel method provides a nile (Table 1). The latter individual stayed in the cabruca tract more detailed home range map, identifying areas of intensive use where it was born during the first year of study. Subsequently, (Powell 2000) through concentric contour lines (95, 50, 20, 10 and it used a mix of land-use types including secondary forest, the 5%) indicating the probability of finding an animal in a given region border of a swamp vegetation and another tract of cabruca

Table 1 Estimates of home range size (in hectares, ha) of the three radio-collared maned sloths based on minimum convex polygon (MCP) and kernel methods and the corresponding percentage of the home range comprising cabruca (in parenthesis). See Methods for details.

Study animal Sex/age Number of fixes MCP (ha) Kernel (ha)

Year 1 Year 2 Year 3 Total

BT464 Female/adult 116 0.44 (90) 0.49 (86) 0.45 (83) 0.56 (83)a 1.00 (76)a BT065 Female/juvenile 58 0.72 (70) 17.59 (7) 29.33 (7)b 4.04 (27)b BT162 Male/juvenile 25 2.28 (100) 2.28 (100) 1.64 (99)

a Based on the total number of fixes (years 1, 2 and 3, plus nine fixes from the forth year). Annual number of fixes were year 1 = 47; year 2 = 31 and year 3 = 29. b Based on the total number of fixes (years 1 and 2). Annual number of fixes were year 1 = 31 and year 2 = 27. 246 C.R. Cassano et al. / Mammalian Biology 76 (2011) 243–250

Fig. 2. Minimum convex polygon (MCP) home ranges of the tree radio-tracked animals showing the main land cover types as well as opportunist sightings of other maned sloths (dotted circles).

(Fig. 2). The female juvenile moved continuously northwards in cas from southern Bahia by Sambuichi (2002, 2006) and Sambuichi what seems to be a dispersal movement (Fig. 2). Areas of more and Haridasan (2007). Most trees used as food in the mature for- intensive use (kernel contours) corresponded to groves or indi- est are considered shade tolerant, i.e., indicative of late succesional vidual trees used repeatedly by the sloths (Fig. 3). These include phases, while the opposite was observed for tree species consumed Ficus clusiaefolia, Inga edulis, Tapirira guianensis, hololeuca, solely in cabrucas (Table 3). Ocotea pretiosa and one individual from the Olacaceae Family cov- ered by an unusually high number of vines. Leaves of all these Discussion species were consumed by maned sloths, except the Olacaceae tree and the vines. The kernels in cabrucas indicated that sloths This is the first unequivocal confirmation of the use of a cacao are using this habitat both for resting and for feeding. In addi- agroforest by maned sloths based on field observation of wild indi- tion to the two tagged juveniles, the mother gave birth to two viduals. Both kernel and residual analyses suggested that cabruca other infants during the study. One was seen when the adult was actively selected by maned sloth, providing both feeding and female was first tagged, in October 2004, and the second when resting trees. Indeed, the range of the adult female (BT464) was the mother was last captured for removal of the radio collar in relatively small, and was comprised almost entirely of cabruca. Nev- February 2008. Besides the tagged individuals, five opportunistic ertheless she was actively reproducing, indicating the availability observations of other maned sloths were recorded in the study of both sufficient food resources and sexually mature males in her cabruca or in adjacent areas, two of which within areas of cabruca home range. Our limited sample of 199 observations in a single (Fig. 2). cabruca limits our ability to conclude that maned sloths area able The frequency with which the three sloths were located in to exploit cabruca throughout their range. However, we observed each vegetation type was significantly different from expected no evidence that using cabruca was detrimental to the sloths, and (Chi-square goodness-of-fit test; 2 = 111.517; d.f. = 3; p < 0.0001). we believe that these areas potentially represent high-quality sloth The residuals showed that swamp/pastures and early secondary habitat. Sloths have been reported in cabruca region of Uruc¸ uca, forests were used less than expected, cabrucas were used more Santa Maria Eterna (Oliver and Santos 1991) and Una municipali- than expected and the late secondary forests were used accord- ties (C.R. Cassano, personal observation) suggesting that cabrucas ing to their availability within the home range’s area (Table 2). in a wider area of southern Bahia might similarly represent usable Similarly, the frequency of fixes differed significantly from the habitat for maned sloths. expected value based on availability of habitat in the buffer area Because the plant species consumed by maned sloths are found (Chi-square goodness-of-fit test; 2 = 119.001; d.f. = 3; p < 0.0001) in a wider sample of cabrucas in the region (Sambuichi 2002, 2006; and the residuals showed the same pattern found in the previous Sambuichi and Haridasan 2007) sloths should be able to forage in analysis (Table 2). many of southern Bahia’s cabrucas. Although they feed from many Sloths were observed feeding on leaves of six tree species from abundant trees (Inga edulis, Tapirira guianensis and Senna multi- cabrucas and on 11 plant species (10 trees and one vine) from juga) the kernel analysis demonstrates that the availability of many mature forests at the Ecoparque de Una (Table 3). Five of the 10 tree relatively rare species (Ficus clusiaefolia, Cecropia hololeuca, Ocotea species found in mature forests and four of the six species found in pretiosa) may be the key to the survival of maned sloths in these cabrucas at Una have also been previously recorded in other cabru- habitats. That single trees have significant influences on patterns of C.R. Cassano et al. / Mammalian Biology 76 (2011) 243–250 247

Fig. 3. Home range (kernel method) of the three tagged animals showing the land cover types. The concentric contours (kernels) represent from outside in, 95, 50, and 5% of the utility density distribution. See Methods for details.

Table 2 Results of two residual analyses on habitat preferences of the three radio-collared maned sloths. In the first analysis (“MCP”), the number of fixes found inside the home ranges (minimum convex polygons – MCP) (“observed”) was contrasted with the expected number based on the proportion of vegetation types available in the MPCs (“expected”). In the second analysis (“Buffer”), the frequency of fixes was contrasted with the expected number based on habitat availability in the buffer surrounding the MCPs (buffer area). See Methods for details.

Vegetation MCP Residual p Buffer Residual p

Observed Expected Observed Expected

Swamp/pasture 12 62 <0.01 12 58 <0.01 Cabruca 123 30 <0.01 123 26 <0.01 Early secondary forest 7 36 <0.01 7 40 <0.01 Late secondary forest 57 71 nsa 57 75 nsa

Total 199 199 199 199

a ns: not significant (p > 0.05). 248 C.R. Cassano et al. / Mammalian Biology 76 (2011) 243–250

Table 3 Tree species from mature forest or cabruca where maned sloths have been found feeding at Ecoparque Una (“feeding trees”). We also indicate which of these species were recorded in other cabrucas sampled in the study region (“other cabrucas”) and the tree sucessional phases: shade intolerant (SI) and shade tolerant (ST).

Family Species Feeding trees Other cabrucasc Sucessional phased

Mature foresta Cabrucab

Anacardiaceae Tapirira guianensis xx SI Burceraceae Tetragastris catuaba x ST Cecropiaceae Cecropia hololeuca xx SI Fabaceae Dalbergia frutencens x SI Dialium guianense xxST Inga edulis xSI Machaerium sp. (vine) x Macrolobium latifolium xxST Senna multijuga xx SI Lauraceae Ocotea pretiosa xST Malvaceae Eriotheca globosa x ST Moraceae Ficus clusiifolia xx ST Brosimum guianense xxSI Brosimum rubescens xxSI Myristicaceae Virola gardneri xxSI Myrtaceae Myrcia fallax x ST Urticaceae Coussapoa pachyphilla x ST

Total 11 6 9

a Cassano (2006). b This study and Correia et al. (2006). c Sambuichi (2002, 2006) and Sambuichi and Haridasan (2007). d Classified by expert taxonomists with experience in the region (Eduardo Mariano Neto, André Amorim, Jomar Gomes Jardim and André Mauricio de Carvalho). habitat use by sloths (Vaughan et al. 2007) and primates (Williams- sloths (Montgomery and Sunquist 1975; see Chiarello 2008, for a Guillén et al. 2006; Oliveira et al. 2010) in similar agroforestry recent review) which, in turn, derives from its anatomy strongly systems suggest that our results may be generalized beyond our adapted to arboreal life (Goffart 1971). Further, sloths are very vul- study population. nerable to predation by mammalian carnivores when on the ground Most tree species eaten by maned sloths in cabruca are shade (Vaughan et al. 2007, pers. obs.). Being a strictly arboreal species, intolerant species, a probable consequence of the tree species maned sloths need high densities of shade and, perhaps more composition of these systems in general. In cacao plantations importantly, uninterrupted connectivity between trees crowns of established under the shade of native trees, climax trees of the the canopy layer. We have observed that maned sloths might cross canopy layer undergo gradual loss, being replaced by pioneer, early small canopy gaps, but mainly when adjacent tree crowns are suf- secondary growth or exotic species (Rolim and Chiarello 2004; ficiently close to each other. Larger gaps might hamper movements Sambuichi and Haridasan 2007). On the other hand, the major- because sloths never jump as primates do and are very reluctant to ity of trees consumed by maned sloths in forest remnants are use the ground as a substrate to pass from tree to tree (Chiarello shade tolerant species, which are probably more abundant there et al., 2004; Chiarello 1998b). We strongly believe therefore that than in cabrucas (Sambuichi and Haridasan 2007; Pardini et al. low canopy connectivity might be a major hindrance for maned 2009). Although detailed dietary studies comparing maned sloths sloths in cabrucas. The tree densities in cabrucas from southern in cabrucas with native forests are still lacking, this finding sug- Bahia are at least half that found in forest remnants of this region, gest that maned sloths might be able to adapt its diet according although it is generally much lower than that. Sambuichi (2006) to local differences in tree species composition and abundance. recorded, for example, densities varying from 47 to 355 individ- This corroborates previous studies on feeding preferences of maned uals/ha in cabrucas of southern Bahia as compared to 596 to 988 sloths carried out in other region of the Atlantic forest (Chiarello individuals/ha in native forests of the region (Martini et al. 2007). 1998a) and suggests that this endangered sloth might thrive in a Tree densities approaching 300 trees/ha are characteristics of little- broader diversity of habitats than was previously thought. In fact managed cabrucas but much lower tree densities are recommended the maned sloths do not seem to be different from the other sloths to increase cacao productivity in traditional agroforests (25–30 in this respect. Such adaptability may result in part for the appar- tree/ha according to Alvim (1972)). It is therefore critically impor- ent ability of sloths to tolerate a variety of plant feeding deterrents tant to investigate what is the maximum tolerance to tree thinning (Cork and Foley 1991), from trees found in primary and secondary for maned sloths. forests (Chiarello 1998a), a trait other sloth species apparently Very likely, the survival of maned sloth is also dependent upon share (Montgomery and Sunquist 1975; Chiarello 2008). the existence and proximity of forests tracts, as has been observed As more information on habitat and diet preferences accumu- in other forest specialists (Faria et al. 2006, 2007; Moura 2008; lates, the more it appears that sloths in general are flexible species Raboy et al. 2008; Pardini et al. 2009), including the golden-headed- (Vaughan et al. 2007; Chiarello 2008). A recent study in Costa Rica lion-tamarin (Leontopithecus chrysomelas)(Dietz et al. 1996; Raboy demonstrated, for example, that two other sloth species, the three- et al. 2004). In agricultural landscapes, the distribution and rich- toed sloth (Bradypus variegatus) and the two-toed sloth (Choloepus ness of large mammals is affected by the presence of forest tracts, hoffmanni), are also able to use cocoa landscape as habitat (Vaughan for example, through the gradual loss of species with increasing et al. 2007). Vaughan et al. (2007) also observed that sloths (mainly distance from forest (Bali et al. 2007), or through the availability Choloepus hoffmanni) can routinely venture into relatively open of more species in forest fragments (Daily et al., 2003; Harvey et pastures in search of preferred food trees. Although this fact doubt- al. 2006). What if these native forest tracts provide feeding trees less stress the behavioral plasticity of sloths and their high capacity in lean times of the year, or function as source habitats for sur- of adaptation to disturbed places, it should be regarded as excep- plus individuals for sloths? These and other possibilities should be tional in the light of what is known about the natural history of investigated in future studies to better frame our results. C.R. Cassano et al. / Mammalian Biology 76 (2011) 243–250 249

Given the severe reduction in the original extent of the Brazil- Amorim, A.M., Thomas, W.W., Carvalho, A.M., Jardim, J.G., 2008. Floristic of the Una ian Atlantic forest (Ribeiro et al. 2009), understanding how maned Biological Reserve, Bahia, Brazil. In: Thomas, W.W. (Ed.), The Atlantic Costal Forest of Northeastern Brazil. New York Botanical Garden Press, New York, pp. sloth use human-modified landscape is essential to prioritize areas 67–146. and actions for the conservation of this endangered species. Our Araújo, M., Alger, K., Rocha, R., Mesquita, C.A.B., 1998. A Mata Atlântica no sul results, although limited in scope, suggest that existing habitat for da Bahia. Cadernos da Reserva da Biosfera da Mata Atlântica, vol. 8. Conselho Nacional da Reserva da Biosfera da Mata Atlântica, São Paulo. this species in southern Bahia is, in fact, greater than the area cov- Bali, A., Kumar, A., Krishnaswamy, J., 2007. The mammalian communities in coffee ered solely by primary and late secondary forest remnants, which plantations around a protected área in the Western Ghats, India. Biol. Conserv. together account for approximately 28% of southern Bahia (“Litoral 139, 93–102. Sul da Bahia” – Landau et al. 2008). Since cabrucas cover a quarter Barreto, R.M.F., Cassano, C.R., 2007. Uso do habitat por uma preguic¸ a-de-coleira (Bradypus torquatus) em plantac¸ ão de cacau (Theobroma cacao L.) sombreada of the landscape, the area available to maned sloth in this land- por árvores nativas no sul da Bahia. In: I Simpósio sobre Paisagem Cacaueira e scape could exceed 50%, but only if cabrucas are managed with both Biodiversidade no Sudeste da Bahia, Universidade Estadual de Santa Cruz, Ilhéus, productivity and wildlife conservation in mind. However cabrucas 8–9 October 2007. Cassano, C.R., 2006. Ecologia e conservac¸ ão da preguic¸ a-de-coleira (Bradypus torqua- vary widely in the diversity and density of tree cover (Sambuichi tus Illiger, 1811) no sul da Bahia. MSc Dissertation, Universidade Estadual de 2006; Schroth and Harvey 2007). Moreover, the presence of large Santa Cruz. tracts of primary forest, certainly contributes for the existence of Cassano, C.R., Götz, S., Faria, D., Delabie, J.H.C., Bede, L., 2009. Landscape and farm scale management to enhance biodiversity conservation in the cocoa production a relatively healthy wildlife in cabrucas of southern Bahia (Faria region of southern Bahia, Brazil. Biodivers. Conserv. 18, 577–603. et al. 2007; Cassano et al. 2009), making them very different from Chiarello, A.G., 1998a. Diet of the Atlantic Forest maned sloth Bradypus torquatus biologically poorer cabrucas of eastern Brazil, such as those found, (Xenarthra: Bradypodidae). J. Zool. (London) 246, 11–19. Chiarello, A.G., 1998b. Activity budgets and ranging patterns of the Atlantic forest for example, in the state of Espírito Santo (Rolim and Chiarello maned sloth Bradypus torquatus (Xenarthra: Bradypodidae). J. Zool. (London) 2004). 246, 1–10. Southern Bahia likely harbors the largest and genetically most Chiarello, A.G., 2008. Sloth ecology: an overview of field studies. In: Vizcaíno, S.F., Loughry, W.J. (Eds.), The Biology of the Xenarthra. University Press of Florida, diverse population of maned sloths (Chiarello and Lara-Ruiz 2004; Gainesville, pp. 269–280. Lara-Ruiz et al. 2008). This, alongside the findings presented here, Chiarello, A.G., Lara-Ruiz, P., 2004. Species discussions: Bradypus torquatus. Edentata indicate that the cacao region of southern Bahia has great potential 6, 7–8. to support perhaps the most viable populations of maned sloth. Chiarello, A.G., Chivers, D.J., Bassi, C., Maciel, M.A.F., Moreira, L.S., Bazzalo, M., 2004. A translocation experiment for the conservation of maned sloths, Bradypus torqua- Nevertheless, deforestation, selective logging, the thinning and tus (Xenarthra, Bradypodidae). Biol Conserv. 118, 421–430. impoverishment of trees in the cabrucas, as well as their conversion Cork, S.J., Foley, W.J., 1991. Digestive and metabolic strategies of arboreal mam- into non-agroforest systems (Johns 1999; Sambuichi and Haridasan malian folivores in relation to chemical defenses in temperate and tropical forests. In: Palo, R.T., Robbins, C.T. (Eds.), Plant Defenses Against Mammalian 2007), might undermine the expectation of long term survival of Herbivory. CRC Press, pp. 133–166. this species. These practices must be reverted or at least better Correia, T.L., Cassano, C.R., Barreto, R.M.F., 2006. Comportamento de filhote desma- controlled particularly in the buffer zones of forest reserves and mado e fêmea adulta de preguic¸ a-de-coleira (Bradypus torquatus), Bahia, Brasil. In: VII Congresso Internacional sobre Manejo de Fauna Silvestre na Amazônia corridors. Canopy shade management within the cabrucas should, e América Latina, Universidade Estadual de Santa Cruz, Ilhéus, 3–7 September for example, consider the maintenance of canopy connectivity, tree 2006. species of conservation interest as well as key tree species for fauna Daily, G.C., Ceballos, G., Pacheco, J., Suzán, G., Sánchez-Azofeifa, A., 2003. Coun- triside biogeography of Neotropical mammals: conservation opportunities in (see Cassano et al. (2009) and Oliveira et al. (2010) for species rec- agricultural landscapes of Costa Rica. Conserv. Biol. 17, 1814–1826. ommendation). Additionally, forest cover in private lands should Dietz, J.M., De Souza, S.N., Billerbeck, R., 1996. Population dynamics of golden- achieve at least the minimum amount mandated by the Brazilian headed Lion tamarins Leontopithecus chrysomelas in Una Reserve, Brazil. Dodo 32, 115–122. environmental legislation (20% of the property’s area). Otherwise Faria, D., Laps, R.R., Baumgarten, J., Cetra, M., 2006. Bat and bird assemblages from the policy makers and society as a whole must be aware that not forests and shade cacao plantations in two contrasting landscape in the Atlantic all cocoa plantations can be regarded as a “green” economic alter- Forest of southern Bahia, Brazil. Biodivers. Conserv. 15, 587–612. native. Faria, D., Paciencia, M.L.B., Dixo, M., Laps, R.R., Baumgarten, J., 2007. Ferns, frogs, lizards, birds and bats in forest fragments and shade cacao plantations in two contrasting landscape in the Atlantic forest, Brazil. Biodivers. Conserv. 16, 2335–2357. Acknowledgements Goffart, M., 1971. Function and Form in the Sloth. Pergamon Press, Oxford. Gouvêa, J.B.S., 1969. Contribuic¸ ão à geomorfologia do Sul da Bahia: área dos baixos cursos dos rios Pardo e Jequitinhonha. CEPEC/CEPLAC Comunicac¸ ão Técnica 35, We would like to thank the Instituto de Estudos Socioambien- 1–11. tais do Sul da Bahia for the logistic support. Fundac¸ ão O Boticário de Gutzwiller, K.J., 2002. Applying Landscape Ecology in Biological Conservation. Protec¸ ão à Natureza, Conservation International Brazil and Seeds of Springer, New York. Harvey, C.A., Gonzalez, J., Somarriba, E., 2006. Dung and terrestrial Change for financial support. We also thank the Brazilian Program diversity in forests, indigenous agroforestry systems and plantain monocultures CAPES for the MSc scholarship to Camila R. Cassano at the Universi- in Salamanca, Costa Rica. Biodivers. Conserv. 15, 555–585. dade de Santa Cruz, and the Brazilian Environmental Agency for the IUCN, 2008. 2008 IUCN Red List of Threatened Species. www.iucnredlist.org. (11 December 2008). research permits (IBAMA # 02001.007588/2002-18). Rubens Vieira Johns, N.D., 1999. Conservation in Brazil’s chocolate forest: the unlike persistence Lopes, Rebeca M. Fonseca Barreto and Tâmara L Correia helped of the traditional cocoa agroecosystem. Environ. Manag. 23, 31–47. during the field work and Rebeca Mascarenhas Fonseca Barreto col- Landau, E.C., Hirsch, A., Musinsky, J., 2008. Vegetation cover and land use in the lected data on tree species from cabruca. Adriano G. Chiarello has Atlantic forest of southern Bahia, Brazil, based on satellite imagery: a comparison among municipalities. In: Thomas, W.W., Britton, E.G. (Eds.), The Atlantic Coastal a research Grant from the Brazilian Science Council (CNPq grant Forest of Northeastern Brazil. Memmoirs of the New York Botanical Garden, vol. # 303273/2008-9). One anonymous referee made a careful and 100. The New York Botanical Garden Press, New York. thorough revision. Lara-Ruiz, P., Chiarello, A.G., 2005. Life-history and sexual dimorphism of the Atlantic forest maned sloth Bradypus torquatus (Xenarthra: Bradypodidae). J. Zool. (Lon- don) 267, 63–73. Lara-Ruiz, P., Chiarello, A.G., Santos, F.R., 2008. Extreme population divergence and References conservation implications for the rare endangered Atlantic Forest sloth, Brady- pus torquatus (Pilosa: Bradypodidae). Biol. Conserv. 141, 1332–1342. Alves, M.C., 1990. The role of cacao plantations in the conservation of the Atlantic Machado, A.B.M., Martins, C.S., Drummond, G.M., 2005. Lista da fauna brasileira Forest of Southern Bahia, Brazil. Dissertation, University of Florida. ameac¸ ada de extinc¸ ão: incluindo as espécies quase ameac¸ adas e deficientes em Alves, T.F., 2005. Distribuic¸ ão geográfica, forófitos e espécies de bromélias epífitas dados. Fundac¸ ão Biodiversitas, Belo Horizonte. nas matas e plantac¸ ões de cacau da região de Una, Bahia. Dissertation, Univer- Martini, A.M.Z., Fiaschi, P., Amorim, A.M., Paixão, J.L., 2007. A hot-point within a sidade de Campinas. hot-spot: a high diversity site in Brazil’s Atlantic Forest. Biodivers. Conserv. 16, Alvim, P.T. 1972. Cacau: ontem e hoje. CEPLAC, Ilhéus. 3111–3128. 250 C.R. Cassano et al. / Mammalian Biology 76 (2011) 243–250

Montgomery, G.G., Sunquist, M.E., 1975. Impact of sloths on Neotropical forest Raboy, B.E., Canale, G.R., Dietz, J.M., 2008. Ecology of Callithrix kuhlii and a review of energy flow and nutrient cycling. In: Golley, F.B., Medina, E. (Eds.), Tropical eastern Brazilian Marmosets. Int. J. Primatol. 29, 449–467. Ecology Systems: Trends in Terrestrial and Aquatic Research. Springer-Verlag, Ribeiro, M.C., Metzger, J.P., Martensen, A.C., Ponzoni, F.J., Hirota, M.M., 2009. Berlin. The Brazilian Atlantic Forest: how much is left, and how is the remaining Moura, R.T. 1999. Análise comparativa da estrutura de comunidades de pequenos forest distributed? Implications for conservation. Biol. Conserv. 142, 1141– mamíferos em remanescente de Mata Atlântica e em plantio de cacau em sis- 1153. tema de cabruca no sul da Bahia. MSc Dissertation, Universidade Federal de Rice, A.R., Greenberg, R., 2000. Cacao cultivation and the conservation of biological Minas Gerais, Belo Horizonte. diversity. Ambio 29, 167–173. Moura, R.T., 2008. Callistomys pictus. In: Machado, A.B.M., Drummond, G.M., Paglia, Ricketts, T.H., 2001. The matrix matters: effective isolation in fragmented land- A.P. (Eds.), Livro Vermelho da Fauna Brasileira Ameac¸ ada de Extinc¸ ão. Ministério scapes. Am. Nat. 158, 87–99. do Meio Ambiente, Brasília. Rolim, S.G., Chiarello, A.G., 2004. Slow death of the Atlantic forest trees in cocoa Oliveira, L.C., Hankerson, S.J., Dietz, J.M., Raboy, B.E., 2010. Key tree species for the agroforestry in southeastern Brazil. Biodivers. Conserv. 13, 2679–2694. golden-headed lion tamarin and implications for shade-cocoa management in Sambuichi, R.H.R., 2002. Fitossociologia e diversidade de espécies arbóreas em southern Bahia, Brazil. Anim. Conserv. 13, 60–70. cabrucas (Mata Atlântica raleada sobre plantac¸ ão de cacau) na região sul da Oliveira-Filho, A.T., Fontes, M.A.L., 2000. Patterns of floristic differentiation among Bahia, Brasil. Acta Bot. Bras. 16, 89–101. Atlantic forests in southeastern Brazil and the influence of climate. Biotropica Sambuichi, R.H.R., 2006. Estrutura e dinâmica do componente arbóreo em área de 32, 793–810. cabruca na região cacaueira do sul da Bahia, Brasil. Acta Bot. Bras. 20, 943–954. Oliver, W.L.R., Santos, I.B., 1991. Threatened endemic mammals of the Atlantic Forest Sambuichi, R.H.R., Haridasan, M., 2007. Recovery of species richness and conserva- region of South-east Brazil. Wildlife Preservation Trust Special Scientific Report tion of native Atlantic forest trees in the cacao plantations of southern Bahia in 4, pp. 21–31. Brazil. Biodivers. Conserv. 16, 3681–3701. Pardini, R., 2004. Effects of forest fragmentation on small mammals in an Atlantic Schroth, G., Harvey, C.A., 2007. Biodiversity conservation in cocoa production land- Forest landscape. Biodivers. Conserv. 13, 2567–2586. scapes: an overview. Biodivers. Conserv. 16, 2237–2244. Pardini, R., Faria, D., Accacio, G.M., Laps, R.R., Mariano-Neto, E., Paciencia, M.L.B., Dixo, Vaughan, C., Ramíez, O., Herrera, G., Guries, R., 2007. Spatial ecology and conserva- M., Baumgarten, J., 2009. The challenge of maintaining Atlantic forest biodiver- tion of two sloth species in a cacao landscape in Limón, Costa Rica. Biodivers. sity: a multi-taxa conservation assessment of specialist and generalist species Conserv. 16, 2293–2310. in an agro-forestry mosaic in southern Bahia. Biol. Conserv. 142, 1178–1190. Wetzel, R.M., Avila-Pires, F.D., 1980. Identification and distribution of the recent Powell, R.A., 2000. Animal home ranges and territories and home ranges estima- slots of Brazil (Edentata). Rev. Bras. Biol. 40, 831–836. tors. In: Bioitani, L., Fuller, T.K. (Eds.), Research Techniques in Animal Ecology: Williams-Guillén, K., McCann, C., Sanchez, J.C.M., Koontz, F., 2006. Resource avail- Controversies and Consequences. Columbia University Press, New York. ability and habitat use by mantled howling monkeys in Nicaraguan coffee Raboy, B.E., Christman, M.C., Dietz, J.M., 2004. The use of degraded and shaded cocoa plantation: can agroforests serve as core habitat for a forest mammal? Anim. forests by endangered golden-headed lion tamarins Leontopithecus chrysomelas. Conserv. 9, 331–338. Oryx 38, 75–83.