Feeding behavior of Geoffroy's marmoset (Callithrix geoffroyi) in an Atlantic forest fragment of South-eastern Brazil

We report on the diet and feeding behaviour of a group of Geoffroy's marmosets (Callithrix geoffroyi) in an Atlantic forest fragment in south-eastern Brazil, studied during the period February 1993 to Januaray 1994. Major food categories were gums (68.6%) fruits (15%), and small animal prey (invertebrates 14.6% and vertebrates 0.8%). Dietary changes were observed between the wet and dry seasons, although they were not statistically significant. Insects and gums were consumed throughout the year but fruits were more prevalent in the diet in the wet season. Plant species exploited for their gums includedInga stipularis, followed byAcacia paniculata, Paulinia carpopodia, andBauhinia angulosa. Chemical analysis of the gum of the four species most used all presented high values for carbohydrates and crude proteins. Fruits of an unidentified species of Myrtaceae (sp. 2) were consumed the most. Animal prey consisted mainly of insects, arachnids, snails, and in three cases, lizards and frogs, with orthopterans being the most prevalent insect prey. This study demonstrates thatC. geoffroyi efficiently exploits resources typically available in secondary and disturbed forest habitat. The main threat to the species is forest destruction, degradation, and fragmentation, but the management of small forest fragments may be an effective corservation strategy.     

Craniometrical variations among eastern Brazilian marmosets and their systematic relationships

The craniometrical variations among marmosets on the eastern coast of Brazil or theCallithrix jacchus group (C. jacchus, C. penicillata, C. kuhli, C. geoffroyi, C. flaviceps, andC. aurita) were analyzed to test various hypotheses for the arrangement of the species and phylogenetic relationships ofC. kuhli andC. flaviceps. Q-mode correlation coefficients and principal component analysis were employed to extract important shape factors from the craniometrical data. On this basis, the shape distances among the eastern Brazilian marmosets were found to be larger than those between the marmosets which are recognized as distinct species. This evidence supports the hypothesis that the members of theC. jacchus group are distinct species.C. kuhli shows the strongest resemblance toC. geoffroyi within theC. jacchus group.C. flaviceps is the most similar toC. aurita in its cranial shape. These findings indicate thatC. kuhli is most closely related toC. geoffroyi, whileC. flaviceps is toC. aurita.     

Diet and feeding ecology of gray woolly monkeys (lagothrix lagotricha cana) in Central Amazonia: Comparisons with other Atelines

I studied gray woolly monkeys (Lagothrix lagotricha cana)in an undisturbed central Amazonian terra firme forest, near the headwaters of the Urucu river, Tefé, Amazonas, Brazil (5°50’S, 65°16’W). I report the diet and feeding ecology of a group of 39–41 individuals, based on systematic feeding observations obtained during 11 months. Woolly monkeys are primarily frugivorous; mature fruits and young seeds account for 83 and 7% of 3298 feeding records, respectively. On a seasonal basis, however, they relied heavily on young foliage (16%), seed-pod exudates (6%), and flowers (3%), particularly during the greatest annual period of ripe fruit scarcity, as determined by a phenological survey. Animals represent only 0.1% of their year-round diet, and they spent little time capturing arthropods and other prey items. Although at least 225 plant species, belonging to 116 genera and 48 families, are in their diet, the three top-ranking families (Moraceae, Sapotaceae, and Leguminosae) account for 43% of their food species and 63% of the time they spent feeding on a year-round basis. I compare the feeding ecology and diet of L. 1. canain the Urucu and other taxa of Lagothrixin upper Amazonia — the last large-bodied Neotropical primates to be studied — to those of other ateline genera: Atelesand Brachyteles. An erratum to this article is available at .     

Feeding ecology of a group of buffy‐headed marmosets (Callithrix flaviceps): fungi as a preferred resource

Mycophagy is a relatively rare behavior in primates and has only been recorded in five callitrichid species. Here, we present data on the feeding ecology of a free‐ranging group of Callithrix flaviceps, which was studied in the Augusto Ruschi Biological Reserve, Southeastern Brazil, in 2008. In contrast with other marmosets, which are typically gummivorous, the study group was predominantly mycophagous–insectivorous, with fungi corresponding to 64.8% of total feeding records, and gum (6.1%) and fruit (3.3%) together providing only a minor part of the diet. Prey corresponded to 25.8% of the group's diet. The fungi (Mycocitrus spp.) consumed by the marmosets were found attached to the stems of Merostachys bamboo. As the animal component of the group's diet was similar to that recorded in studies of other marmosets, we propose that fungi were exploited primarily as a substitute for plant material, in particular exudates. This highly mycophagous diet may be determined by two principal factors: (1) the abundance of fungi within the study area, and (2) the avoidance of bark gouging, for which C. flaviceps may be less specialized than most other marmosets. These conclusions are supported by comparisons with other marmoset groups, which indicate an ecological specialization for mycophagy in C. flaviceps, and that the species will resort to gummivory in habitats where fungi are scarce. Am. J. Primatol. 72:515–521, 2010. © 2010 Wiley‐Liss, Inc.     

Density and Spatial Distribution of Buffy-tufted-ear Marmosets (<Emphasis Type="Italic">Callithrix aurita</Emphasis>) in a Continuous Atlantic Forest

The continued degradation of forest habitats and isolation of fragmented populations means that the conservation of endemic marmosets in the Brazilian Atlantic forest depends on human interventions including legal protection. Population monitoring is required to ensure effective management and appropriate allocation of conservation resources; however, deriving estimates of population metrics such as density within heterogeneous environments is challenging. We aimed to quantify the population density and spatial distribution of buffy-tufted-ear marmosets (Callithrix aurita) in the northern region of Serra-do-Mar State Park. We incorporated habitat suitability as quantified by a niche modeling algorithm (MAXENT) to refine density estimates obtained via distance methods. We used 6 environmental predictors to model the distribution of Callithrix aurita and used the resulting MAXENT niche model to identify environmental conditions that represent suitable habitat for this species. We used 877.7 km of line transect surveys and distance methods to derive estimates of 2.19 groups or 7.55 individuals/km² from direct observations (n = 40), providing an overall population estimate of 1892 (95% CI = 1155–3068) individuals in 250.7 km² of Atlantic forest. Our refined density estimate, obtained by combining distance methods and a niche model, yielded a result of 1386 individuals. Suitable habitat was not uniformly distributed across the study area and was most strongly associated with altitude and the type of vegetation cover. We provide a review of previous surveys and find this is the largest known population of Callithrix aurita. Our refinement of density estimates provides a simple and informative addition to the primatologist’s toolbox.     

Feeding ecology of Saguinus bicolor bicolor (Callitrichidae: Primates) in a relict forest in Manaus, Brazilian Amazonia.

This study is part of a long-term ecological study of habitat and dietary requirements of the pied bare-face tamarin (Saguinus bicolor bicolor). One group was studied for 11 months in an area of secondary forest in a suburb of Manaus, Amazonas, Brazil. Three main vegetation types occurred inside the group's home range (12 ha): capoeira, older secondary forest and campinarana (white sand forest). The tamarins ate fruits (21 species), flowers (1 species), exudates (4 species), and arthropods (insects and spiders). They spent 14.3% of total activity time seeking and eating animal prey, and 9.9% feeding on plant material, mostly fruits. In general, fruits consumed were ripe, small and succulent. Trees used for feeding were low and had small crown diameters. Three plant species (Protium aracouchinni, Myrcia cf. fallax, and Couma utilis) were used intensively during the three seasons covered by the study period. The concentrated use of 3 fruit species, each for an extended period (one fruiting species per season), provided the tamarins with a regular food supply. Tamarins consumed exudates from holes in the bark of trees of the families Anacardiaceae and Vochysiaceae, as well as gum exuded from seed pods of Mimosaceae. Exudates were exploited during the dry season and at the beginning of the wet season. Group travel was primarily based on routes connecting the fruiting trees exploited, with foraging for animal prey occurring during travel. Tamarins searched for arthropods on trunks, branches and leaves and in trunk holes. The foraging and feeding tactics displayed by S. b. bicolor are closely linked to morphological characteristics (small size and weight, claw-like nails) that allowed access to energy-rich resources (arthropods and plant exudates) in different strata of the vegetation.     

Foraging for animal prey by outdoor groups of Geoffroy’s marmosets (Callithrix geoffroyi)

I studied animal prey foraging by three outdoor groups of Geoffroy’s marmosets, Callithrix geoffroyi, over the course of two summers (June–August 1994; August–September, 1995. Marmosets are highly motivated to forage for animal prey, as demonstrated by the amount of time spent foraging for insects and small vertebrates even in the presence of provisioned, high-quality food. Like their wild congeners, the marmosets engaged in prolonged visual searches of selected areas while slowly locomoting. They seldom used their hands to manipulate substrates. The marmosets were startled proportionately more often while foraging for animal prey than would be predicted by their overall time budgets, and they regularly, if briefly, interrupted their search/locomote pattern with sweeping scans of the surroundings that may contribute to a vigilance function. While carrying infants, they almost never foraged. These results are consistent with the notion that marmosets may be particularly vulnerable to predation while searching for animal prey. Most of the data on marmoset and tamarin feeding ecology have focused on fruits and gums. My data underscore the need to understand better the role of animal prey in the behavioral ecology of callitrichids.     

Primate Group Size and Abundance in the Caatinga Dry Forest, Northeastern Brazil

The Caatinga dry forest poses a series of ecological challenges for mammals in general and primates in particular. The erratic rainfall pattern impacts on plant diversity and phenological patterns; from year to year there is marked variability in fruit production and failure to fruit is common. The harshness apparently accounts for the impoverished mammalian fauna. However, data on primate abundance, distribution, and possible environmental effects on primate density are lacking in this type of dry forest. I censused the primate community in 3 habitats of the Serra da Capivara National Park, Piaui, NE Brazil, over a total distance of 318 km. Overall, the abundance of primates in the Caatinga dry forest is very low as a consequence of low abundance of food resources both in space and time. Alouatta caraya (predominantly folivorous) occurs at extremely low density, and during the dry season are apparently confined to canyon areas, where trees retain their leaves. Callithrix jacchus has morphological feeding specializations for gum-eating, and gum is an important resource during food bottleneck periods. Nonetheless, Callithrix jacchus occurs at comparatively low densities. Group sizes for howlers and marmosets in the Caatinga are significantly smaller than in other forest types. Contrarily, Cebus apella libidinosus had an average group size within the range reported for Amazonian and Atlantic forests. Researchers consider the generalized diet of capuchins as the explanation for their similar abundance in different habitats, indicating relative independence from ecological constraints. However, I suggest that capuchin foraging style and cognitive abilities are important factors accounting for their unreduced group size and density even under extreme conditions.     

Gum feeder as environmental enrichment for zoo marmosets and tamarins

Tamarins and marmosets are small-bodied social callitrichines. Wild callitrichines feed on exudates, such as sap and gum; particularly, marmosets are mainly gummivores, while tamarins consume gums only occasionally and opportunistically. Zoo marmosets and tamarins are usually provided with gum arabic as an alternative to the exudates normally found in the wild. The aim of this study was to evaluate the effects of a gum feeder on the behavior and well-being of four zoo-managed callitrichines. We studied four cotton-top tamarins (Saguinus oedipus), four red-handed tamarins (S. midas), two pygmy marmosets (Cebuella pygmaea), and three Geoffroy's marmosets (Callithrix geoffroyi) housed at Parco Natura Viva (Italy). We conducted the study over two different periods, a baseline (control, without the gum feeder) and then a gum feeder (when the gum feeder was provided) period. We used continuous focal animal sampling to collect behavioral data, including durations of social and individual behaviors. We collected 240 min of observations per period per study subject, with a total of 3,120 min for all the subjects in the same period and of 6,240 min in both periods. We analyzed data by using nonparametric statistical tests. First, we found that the gum feeder promoted species-specific behaviors, such as exploration, and diminished self-directed behaviors, suggesting an enriching effect on tamarin and marmoset behavior. Moreover, in red-handed tamarins, the provision of the gum feeder reduced the performance of self-directed and abnormal behavior, specifically coprophagy. These results confirm that gum feeders are effective foraging enrichment tools for zoo marmosets and tamarins.     

Ecology of <Emphasis Type="Italic">Callithrix kuhlii</Emphasis> and a Review of Eastern Brazilian Marmosets

We provide data on wild Callithrix kuhlii, a little studied species. We describe their ecology, focusing on demography, diet, and ranging patterns in 8 groups studied from 1995 to 1999. The groups averaged 4.3 individuals, with 1 breeding female and 1–2 adult males, and reproduction was seasonal. Evidence from nonstudy groups indicated the possibility of polygynous group structure. We observed group formation and the transfer of adults between groups. The marmosets ate 20 species of fruit and nectar and 5 exudate species. Home ranges were 34–39 ha, and daily path lengths averaged 1498 m. We include our data in a comprehensive review of eastern Brazilian marmosets to characterize their ecology. Callithrix kuhlii ate more fruit species and less gum species, and had larger home ranges and daily path lengths versus those of most congeners. There was also a tradeoff between travel time and other behaviors. Species with smaller home ranges spent more time foraging and resting and less time traveling than the ones with larger ranges. Attributes consistent with all species included their reliance on gum during fruit scarcity and preference for disturbed habitat. We explored factors relating to the variation in diet and ranging patterns within the eastern Brazilian marmosets, in particular by considering the species in relation to elevation and the extent and proximity of their distribution to the sea. However, large intraspecific variation resulting in part from variable levels of anthropogenic disturbance across sites confounded our ability to verify trends corresponding to the geographic characteristics.     

Ranging behaviour and habitat preference of a wild marmoset group, Callithrix humeralifer (Callitrichidae, Primates)

The ranging behaviour of a group of marmosets ( Callithrix humeralifer ) in seasonal Amazonian rain forest was studied during one year. Range sizes (monthly and daily), day range lengths and patterns of range use are examined for correlations with feeding behaviour and the distribution of three forest types within the marmoset's range. Seasonal differences in ranging are associated with changes in the abundance and distribution of plant food sources. The marmosets ranged more widely and used more sources of a greater diversity of plant food species, which were distributed over a wider area, in the wet season than in the dry season. In the dry season, they ranged over a smaller area and, although they used fewer sources of a reduced diversity of plant species overall, they exploited a larger number of sources of the five highest ranked plant species in the diet. Throughout the year, they showed a preference for disturbed primary forest, characterized by dense understoreys and abundant second growth patches. Reasons for this preference are discussed, taking into account their use of fruits of typical pioneer species (particularly in the dry season), insect prey abundance, sleeping site availability and defence against predators.     

Differences among captive callitrichids in the digestive responses to dietary gum

Although many members of the Callitrichidae, a monophyletic family of small, New World monkeys, have been observed to feed on plant exudates, available field data support the generalization that pygmy and common marmosets (Cebuella pygmaea and Callithrix jacchus) feed on gums to a greater extent than most other callitrichids. Because microbial fermentation is required for vertebrates to digest gums, gum-feeding primates may react differently to dietary gum from their relatives that eat little gum. To test this hypothesis, digestion trials were conducted on animals from the two marmoset species, two tamarin species (Saguinus fuscicollis and S. oedipus), and a species of lion tamarin (Leontopithecus rosalia). These species span the range of body sizes within the Callitrichidae. All animals were fed two variations of a homogeneous diet, which differed only in that gum arabic was added to one. Transit time of digesta (TFA) and digestive efficiency (as measured by the coefficients of apparent digestibility of dry matter and energy [ADDM and ADE, respectively]) were compared between diets for each individual. As predicted, the digestive responses of marmosets differed from the responses of the other study species. In marmosets, TFA tended to be longer when gum was added to the diet, while TFA did not change in the other three species. Digestive efficiency decreased in tamarins and lion tamarins with the addition of gum to the diet; marmoset digestive efficiency was unaffected by diet. The results of this research are consistent with the hypothesis that marmosets have digestive adaptations that aid in the digestion of gum that other callitrichids lack. © 1996 Wiley-Liss, Inc.     

Predator avoidance behaviour in the buffy-headed marmoset,Callithrix flaviceps

The predator avoidance behaviour of a free-ranging group of buffy-headed marmosets,Callithrix flaviceps, was recorded in detail during the course of a long-term study of behavioural ecology at the Fazenda Montes Claros, southeastern Brazil. Four distinct patterns of predator avoidance behaviour, each with specific vocalisations, were recognised and are described here. The selection and use of sleeping sites by the study group are also described. An analysis of the records indicates that these small monkeys are generally most vulnerable to predation by aerial raptors. Variations in the frequency of alarm calls also indicate that the marmosets tend to be more vigilant at higher levels in the forest and when the leaf cover is less extensive. The implications of group size and social structure for both the evolution and the efficacy of the anti-predator behaviour of marmosets are also discussed.     

Habitat and the evolution of social and reproductive behavior in callitrichidae

The numerous remarkable traits characterizing the ecology and behavior of callitrichids have inspired considerable research and discussion of the flexible mating system (cases of monogamy, polygyny, and polyandry), cooperative breeding, reproductive inhibition by dominant females, rapid reproductive rate, significance of dietary differences (gum feeding, frugivory, and insectivory), and demographics and social characteristics of the four genera, Cebuella, Callithrix, Saguinus, and Leontopithecus. The majority of functional explanations evaluate costs, benefits, and alternative strategies but do not address the critical selective forces that led to the behaviors in the first place. In this paper, it is argued that Callithrix and Saguinus evolved to occupy a small insectivore/frugivore niche in secondary growth forest patches (gaps arising from tree falls), and other successional forests and edge habitats; that Cebuella evolved to occupy a gum-feeding/insectivore niche in inundated forest; and that Leontopithecus is above all a small animal predator/frugivore and a mature forest genus. The keys to explanations concerning the evolution of the social and reproductive systems of these animals lie in an understanding of the resource base in these different habitats. Finally, it should be remembered that studies of callitrichids have in the main part been carried out in habitats highly altered by human activities (especially the Brazilian Atlantic forest, home to marmosets and lion tamarins) and are not the environments in which their social and breeding behavior evolved. © 1996 Wiley-Liss, Inc.     

Home range use and the exploitation of gum in the marmosetCallithrix jacchus jacchus

Three wild groups of common marmoset, Callithrix jacchus jacchus,in north-east Brazil, of approximately similar size, had home ranges between 2.5 and 6.5 ha. But their core areas were similar in size between 1.0 and 1.5 ha, with a monthly area of heavy use between 1.1 and 1.6 ha. The groups were selective in the use of their home ranges, even though they were small: they used some areas heavily and others lightly. The core areas had higher densities of trees that produced gum exudates than did other parts of the home ranges. Our data suggest that a group of marmosets in this habitat may require a minimum of about 50 gum trees in its home range at a minimum density of about 50 trees/ha. In addition, the animals require suitable trees in which to sleep. We suggest that patches of forest with these desirable properties remain relatively fixed in size and location over the years and that individual animals are constantly in flux between them.     

Fragmentation causes rarity in common marmosets in the Atlantic forest of northeastern Brazil

The purpose of this study was to investigate the population status of the common marmosets, Callithrix jacchus, in one of their main habitats, the northeastern Atlantic forest of Brazil, where only 2% of its original area remains as small and isolated forest fragments, to determine long-term viable populations. The study was carried out in 15 forest fragments, measuring from 3,478.3 ha (the largest) to 6 ha, which were invariably subjected to high human impact. Line transect surveys were carried out between January 2002 and December 2004, along transects measuring from 350 to 4,000 m, between 0500 h and 0900 h. Common marmosets were registered in 73% (n = 11) of the forest fragments, were not recorded in the largest one, the Coimbra Forest, and group sizes varied from 1 to 4 individuals. A negative significant relationship was detected between the size of the fragment and the number of sightings of common marmosets. It is shown that the number of groups of common marmosets dwelling in this highly impacted landscape is nowadays lower than necessary for long-term survival. Additionally, smaller fragments having more groups suggests that they live in total association with humans, which assure their subsistence through the supply of introduced and exotic foods.     

Diet and Feeding Ecology of Woolly Monkeys in a Western Amazonian Rain Forest

I investigated the diet and feeding ecology of two social groups of woolly monkeys (Lagothrix lagotricha poeppigii) in Yasuní National Park, Ecuador between April 1995 and March 1996. Woolly monkeys in Yasuní were predominantly frugivorous, with fruits comprising ca. 77% of the yearly diet; the next most common food type in the diet was insect and other animal prey. The fruit diet of woolly monkeys in Yasuní is the most diverse yet recorded for any ateline primate, including spider monkeys (Ateles), which are often regarded as ripe fruit specialists: 208 distinct morphospecies of fruits were consumed by woolly monkeys either during the study or during several preceding months of pilot work. Nonetheless, close to one-third of the yearly diet came from just 3 plant genera—Inga, Ficus, and Spondias—and only 20 genera each contributed to 1% of the diet. For one study group, the proportion of ripe fruit in the diet each month was correlated with the habitat-wide availability of this resource, a pattern evidenced by several other ateline species. However, the relationship was not apparent in the second study group. The modal party size for feeding bouts on all food types was a single monkey, and, contrary to reports for other atelines, neither feeding party size nor the total number of feeding minutes that groups spent in food patches was well predicted by patch size. Both results highlight the independent nature of woolly monkey foraging. Given that woolly monkeys and closely-related spider monkeys focus so heavily on ripe fruits, their very different patterns of social organization are intriguing and raise the question of just how their ecological strategies differ. Two important differences appear to be in the use of animal prey and in the phytochemical composition of the ripe fruits that they consume: spider monkeys rarely forage for animal prey, and woolly monkeys seldom consume the lipid-rich fruits that are an important part of spider monkey diets.     

Hunting strategies in wild common marmosets are prey and age dependent

We investigated the hunting strategies of wild common marmosets (Callithrix jacchus) to determine whether the strategies differed among animals of different age classes and/or prey type. The study was conducted in a fragment of Atlantic Rain Forest, situated 40 km from Recife (PE/Brazil). Twenty‐seven individuals from four social groups were observed. Captured prey items were divided into three categories. The hunting strategies of the common marmosets were ranked into four categories. The acquisition of larger prey (items more than 2.0 cm) involved the appropriate body movements and postures that concealed the approaching marmosets, whereas the acquisition of smaller prey (items under 2.0 cm) involved less concealing behaviors. Furthermore, adults and juveniles (age ≥5 months) were more capable of capturing larger prey than were younger (1–2 months) or older infants (3–4 months). Although older infants were successful in capturing certain prey, they often failed when they attempted to capture larger prey that jumped and/or used flight to escape. The results suggest that both the experience of the monkeys and escape behavior of the prey affect predation efficiency in wild common marmosets. Am. J. Primatol. 72:1039–1046, 2010. © 2010 Wiley‐Liss, Inc.     

Feeding and ranging patterns of forest hanuman langurs (Presbytis entellus)

The feeding and ranging patterns of a troop of hanuman langurs (Presbytis entellus, Colobinae) were studied in Kanha Tiger Reserve, central Indian Highlands for 1850 hr (1981–1982), in a mosaic of moist deciduous forest and anthropogenic meadow. The location, size, and species of each tree within the 74.5-ha troop annual range was known and the phenology of all tree species was sampled. According to scan sampling, the troop spent 25.7% of the daytime feeding, with range use concentrated on an island of dry deciduous forest. Whereas adjacent troops occupied only the periphery of the focal troop's range, all-male bands occupied its center, especially during takeover and infanticidal attacks. The troop consumed items from 60 of the 67 species of trees and woody climbers available; mature leaves (34.9% of feeding time), fruits (24.4%), leaf buds (10.6%), flowers and flower buds (9.5%), young leaves (3.6%), insects (3.0%), and gum (1%). The monthly utilization of fruit, open leaf buds, and flower buds is correlated significantly with their abundance, and the troop spent significantly more time feeding and less time moving when consuming mature leaves. Comparison of tree dispersion and langur ranging patterns suggests that the distribution of the most important food trees is a major influence on their range use.     

Diet and Feeding Behavior of <Emphasis Type="Italic">Leontopithecus chrysomelas</Emphasis> (Callitrichidae) in Degraded Areas of the Atlantic Forest of South-Bahia,Brazil

Habitat disturbance alters plant diversity and food resource availability, affecting the ecology, and ultimately the survival and reproduction, of species depending on those plants. Studies in degraded areas serve to improve our understanding of the consequences of habitat modification for endangered species and to guide conservation actions. We studied diet composition, monthly variation in feeding behavior and fruit feeding time, and dietary diversity in two golden-headed lion tamarin (Leontopithecus chrysomelas) groups ranging in a degraded area of the Atlantic Forest in South-Bahia, Brazil, over a 12-mo period. We recorded feeding behavior and food items consumed through continuous observations and focal animal sampling, and performed dietary analyses on samples from consumed fruits. Substrate manipulation to search for animal prey was the feeding activity most frequently observed, followed by fruit consumption. We observed nectar drinking and exudate feeding at low frequencies from April through July. Bromeliads were the principal foraging substrate. We observed the use of 114 plant species, confirming the large dietary diversity reported for the species. Individual trees from the family Melastomataceae, common in degraded forests, accounted for the highest visiting frequencies (40%). Fruit availability was the main factor explaining variation in monthly fruit feeding time, despite the absence of climatic seasonality. Nutritional or energetic characteristics did not affect fruit choice. Differences in floristic composition appear to be a major determinant of the species’ diet in different study areas. Regional forest restoration programs should consider including advanced forest species, to improve both forest quality and animal mobility between fragments.