by M. TROLLE

Zoological Museum University of Copenhagen, Mammal Department,
Universitetsparken 15, 2100 Copenhagen, Denmark
e-mail: mtrolle@zmuc.ku.dk

Summary. – The Brazilian part of the Rio Negro Basin, a major region of the Amazon, is one of the least studied regions of the Amazon rainforest. An intensive, four months inventory of the medium to large, non-volant mammal species was conducted in the area encompassing the lower Rio Xixuaú, a minor tributary to the Rio Jauaperí river, to the north of Rio Negro. The main habitats of the study area were seasonally flooded igapó forest, riparian vegetation, higher-lying terra firme forest, secondary forest, minor water courses and lakes. Using walked transects, camera trapping, observations from canoe, nightspotting, interviews, and identification of skulls, it was possible to list 43 species for the study site, including 8 xenarthrans, 7 primates, 12 carnivores and 5 ungulates. The habitat use of the primate species is analyzed.

KEY WORDS : Mammal inventory, camera trapping, primate habitat use, Rio Negro region, the Amazon, Brazil.

INTRODUCTION

The majority of the mammal surveys of the Brazilian Amazon has been carried out to the south of Rio Amazonas and on tributaries to Rio Solimões (Voss and Emmons 1996; Patton et al. 2000) – mainly on white water rivers. The black water Rio Negro Basin is one of the least studied regions of the Amazon rainforest. The first collections on the Rio Negro were conducted by the expedition of A. Rodrigues Ferreira in the end of the eighteenth century, and later, in the nineteenth century, also A.R. Wallace visited the region. Avila Pires (1964) published a paper on mammals collected on the Rio Negro close to Manaus. The only recent intensive mammal inventory of a delimited study site in the Brazilian part of the Rio Negro region was carried out in the Parque Nacional do Jaú, situated at Rio Jaú, to the south of Rio Negro (M.N.F. da Silva, personal communication) - a study that focused on small mammal trapping (Patton and da Silva, unpublished species list). Rio Jauaperí, where this study took place, runs to the north of Rio Negro, which is a major distribution barrier for many mammal species (Emmons and Feer 1997; Eisenberg and Redford 1999).
An intensive, four months mammal inventory was conducted in the Xixuaú Nature Reserve, situated on the western side of the middle Rio Jauaperí, Roraima state, Brazil (headquarters at S 0º48.023’, W 61º33.476’). Xixuaú Nature Reserve is an area of approximately 150,000 ha of mostly pristine rainforest. The study area consists of the typical forest habitat types of the region; higher-lying unflooded terra firme forest, lower-lying seasonally flooded igapó forest and riparian vegetation, as well as secondary forest at various successional stages, a few minor plantations, minor water courses and lakes. During the floodseason the area receives ‘black’ water from Rio Jauaperí, and the igapó is flooded by black water, however, a couple of months a year, during the low water season, the water courses inside the Nature Reserve are exceptionally clear and the water less acidic. The study focused on the area encompassing the lower Rio Xixuaú, a narrow river in the southern part of the Nature Reserve.
The study was conducted starting at the peak of the low water season of the beginning of 2001 when all of the igapó forest was dry. During the study period, the rainy season began and waters rose considerably, and at the end of the study period the igapó was starting to flood. Observations from an earlier one-month visit to Xixuaú in August 1996, at the peak of the flood season, are included.
The area was traditionally inhabitated by indigenous people, who must have carried out subsistance hunting. For more than three decades, the area has been inhabited by only a few caboclo families and, besides poaching of cats and otters, hunting pressure was probably light. Hunting ceased altogether with the establishment of the Xixuaú Nature Reserve about 10 years ago.
The main aim of this study was to inventory the medium to large, non-volant mammal species of the area. In addition, the habitat use of primates in the area was studied in more detail.

MATERIALS AND METHODS

The following methods were used: camera trapping, diurnal and nocturnal walked transects, diurnal and nocturnal observations from canoe, identification of tracks, interviews of knowledgeable reserve staff, and, in the case of one species, identification of skull and skeletal parts.
Six TrailMaster camera traps with infrared trigger mechanism were used (TM 550 Passive Infrared Trail Monitor + TM35-1 Camera Kit) (Goodson and Associates, USA). Trap stations were often baited (Table 2).
A system of nine transects was established and used for the daily surveys. The trails were situated along an 8 km stretch of the lower Rio Xixuaú and its tributaries. The trail system covered all of the above mentioned major terrestrial habitat types. On the walked transects, observations of animals heard close to the trail were included (typically flushed animals and monkeys). Animal signs that could be identified to species level, such as armadillo burrows and tapir droppings, were also recorded.
For the interviews I relied mostly on two very knowledgeable locals: Carlos Roberto Nascimento who has lived for 25 years in Xixuaú, and worked for many years as a poacher of wild cats and Giant Otters, and João Soares Gomes da Silva, who has spent his entire life working inside the forest, and has lived about 10 years in Xixuaú. Both men have worked extensively as guides in the reserve.

RESULTS

More than 750 records of mammals were obtained by sightings, acoustic records and camera trapping. The inventory includes 43 species (Table 1), out of which 23 were sighted, 13 were recorded by camera trapping, and an additional 13 ‘reliable’ species were reported by the locals of Xixuaú. The study design did not allow for precise estimates of abundance of the species, however, several of the species were recorded frequently enough to conclude that they were common. This information is included in Table 1.
Table 2 lists the results of the camera trapping, giving information on the sites and bait types used successfully to record the various species.

Primate observations

The only monkeys observed frequently in the igapó forest during the survey were Howlers and Capuchins. During the low water season the diverse and abundant primate fauna is found in the terra firme.
In order to analyze the habitat use of the primates in the terra firme forest, four of the most walked transects were chosen, representing a grade from very tall, open, undisturbed primary forest to forest with many tree falls and much natural secondary vegetation. The average time it took to walk a transect was about the same on these four trails, so the results are comparable. Two measures were calculated: a sighting index (i.e. the frequency of recording a species per walked transect), and the number of records of a species on a trail divided by the total number of primate records on that trail (i.e. how frequently a species was observed on a trail in comparison to the other species) (Table 3).
All seven species of primates were observed carrying young throughout the study period.
Squirrel Monkeys and Capuchins are known to commonly associate (Emmons and Feer 1997). During our study, we saw 22 examples of such mixed troops. The only other species often seen associating was the Bearded Saki. We had several observations of a single Bearded Saki moving together with either Spider Monkeys (n=3) or mixed Capuchin/Squirrel Monkey troops (n=3).

When opossums (Didelphidae), spiny rats (Echimyidae), and rats and mice (Sigmodontinae) are excluded, the long-term study at the MSCE Reserves north of Manaus recorded 32 species of non-volant mammals (Voss and Emmons 1996). In our study, we found 41 species within this category. The high number compared to the MSCE reserves is mainly explained by the fact that this area consists exclusively of terra firme growth, with no major rivers and associated riparian vegetation.
An additional armadillo species distributed throughout the Amazon, Cabassous unicinctus, could be expected in the Rio Jauaperí region. The species is extremely difficult to record in rainforest; neither L.H. Emmons or J.R. Malcolm, some of the most experienced field workers in the region, found the species during their extensive studies at the MSCE Reserves (Voss and Emmons 1996), and only later has it been recorded in the region by camera trapping (Rittl 1998; Yabe et al. 1998). A second species of tree porcupine, Coendou melanurus, could also occur (Emmons and Feer 1997; Eisenberg and Redford 1999), however, the species was not recorded in the MSCE Reserves, and was not known by locals in Xixuaú.

The data in Table 3 should obviously be viewed with caution, but nevertheless give a good indication of various trends. Using these results along with the rest of our observations, in Table 4 I have given a summary of the overall habitat use of the primate community. The following forest strata classifications for the terra firme were used: A stratum = tall emergents; B stratum = the continuous canopy; C stratum = smaller trees (and palms) below the canopy.

During the dry season when fruit is scarce inside the seasonally flooded igapó the only common monkeys found in this part of the forest are Howlers (specialized folivores) and Capuchins (omnivores) (Emmons and Feer 1997), both able to get by without fruit. During my flood season visit in 1996 I recorded also White-faced Saki, Brown Bearded Saki and Squirrel Monkey in the igapó. My own observations combined with information from locals indicate a strong migration of many of the primate species from the terra firme to the igapó during the flood season, when the amount of fruit increases substantially. However, I never saw Spider Monkeys in igapó, and according to the locals, this species stays in the terra firme. This may also, at least to some extent, be the case with the Tamarin.

Camera trapping has been used successfully for mammal studies in tropical forests of Asia and Africa (Griffiths and van Schaik 1993; Karanth and Nichols 1998; Franklin et al. 1999). This survey, along with the studies of Rittl (1998), Yabe and Higuchi (1998), and Yabe et al. (1998) has shown that also in the Amazon camera trapping is a highly efficient way to relatively quickly record a number of shy and secretive terrestrial mammal species, and survey a fauna that may otherwise take years to inventory. The above mentioned camera trapping studies north of Manaus recorded also Jaguarundi, Ocelot, Margay, Grison, Coati, and a range of smaller species of opossums and rodents.
Ants and termites are known to be a major problem for field workers camera trapping in rainforest in Asia, since they use camera traps as nesting sites. To solve this problem I recommend a simple solution: wrap adhesive tape around the base of the pole on which the camera trap is placed with the sticky side facing outwards. Above this, the bark can be peeled upwards with a machete so that an ‘umbrella’ is constructed; this umbrella protects the tape against rain and probably works against rodents (like the system used on ropes on ships, to prevent rats from entering).
After this study, a similar mammal survey was carried out in Pantanal of SW Brazil. In the coming publication of that study (Trolle in prep.) more detailed recommendations for camera trapping in the Amazon and Pantanal will be given.

ACKNOWLEDGEMENTS

This project was a collaboration between the Zoological Museum University of Copenhagen, National Institute for Amazonian Research, Manaus (INPA) and Museu Nacional, UFRJ, Rio de Janeiro. The project was made economically possible by the generous support of WWF-Denmark/Novo Nordisk, Explorers Club-US, Copenhagen Zoo, University of Copenhagen, Zoological Museum of Copenhagen, Frimodt-Heineke Fonden, Torben og Alice Frimodts Fond, H.R. Frederiksen og Grete Siim Frederiksens Fond, Direktør Peder Mortensen og Hustru Marry Mortensens Fond, Kjebi Fonden, NetTravel, Duracell, Photographica, and Fuji. Thanks are also due to Associação Amazônia for providing the opportunity to carry out the study in Xixuaú Nature Reserve. For collaboration, advice and support thanks to Dr. João Alves de Oliveira, Dra. Maria Nazareth Ferreira da Silva, Dr. Hans J. Baagøe, Prof. Jon Fjeldså, Mogens Andersen, Dr. Leila M. Pessôa and Dr. Louise Emmons. Tim Dempster and João Alves de Oliveira kindly revised the manuscript and Signe Zacchi translated the summary into French. And finally, many thanks are extended to my field assistants in Xixuaú for invaluable help in the field.

BIBLIOGRAPHY