Mammals of the Brazilian Cerrado

Multi-species monitoring amidst agricultural change
Background

The majority of mortality for carnivores and other wide-ranging species occurs outside of reserves that are intended to provide them safe haven.  Increasing agricultural expansion and land use intensification necessitate that management for large bodied species incorporate the mosaic of land uses outside of nature reserves and national parks.  While the role of these landscapes in species persistence is well-recognized, there is little data about how these areas contribute to populations of interest. 

The Cerrado biome of Brazil comprises the world’s most biological diverse tropical savanna.  It is also one of the world’s most threatened regions.  Given current rates of land conversion for agricultural production, it is expected that all of the natural biome outside of reserves will be entirely converted by the year 2050.  The Cerrado is home to hundreds of endemic species and it comprises the primary habitat of the maned wolf, a wild canid found only in the grasslands of central South America .  The majority of the maned wolves’ range is under conversion for agricultural production, yet scientists have little idea of how this conversion is affecting the population status of the maned wolves. 
The 1200 square kilometer Emas National Park is considered one of a few major strongholds for wolf populations. The Park is quickly becoming an island in a sea of agriculture as conversion has been recent, rapid, and intense.  The impact of the surrounding land conversion on the wolves and other wide-ranging species, however, remains unknown. 
Objectives

This project seeks to provide a population assessment of the wolves, along with stress, nutritional, and reproductive impacts associated with their use of the surrounding landscape.  A secondary objective is to complete a presence-absence survey of other wide-ranging fauna of interest (puma, jaguar, giant armadillo, giant anteater, tapir) that co-occur with the maned wolves.    

Opportunity exists to influence future land management as private landholders are currently required to set aside 30% of their farm as natural habitat.  There is ongoing debate about whether they should be allowed to purchase set aside land off-farm or whether they will continue to be required to protect habitat within this region.  It is thus timely and relevant to understand what type of land mosaic best promotes or reduces species viability. 

The specific objectives of the study are to determine 1) the abundance of maned wolves living in a 5,000 square kilometer region in and around Emas National Park,  2) the degree to which remnant habitat patch configuration affects presence of target species, 3) if crop type and height affect presence, 4) the degree to which distance from park affects stress, nutrition and reproductive hormones of this population of maned wolves, and 5) whether the amount and structure of natural habitat fragmentation affects stress levels, nutritional status, and reproductive function of this population of wolves. 
Methods

We are applying the Scat Detection Dog, fecal DNA, and hormone methodologies pioneered by our Center to monitor impacts of land conversion on movements and physiological health of large mammals of the Brazilian Cerrado.

DNA extracted from scat, combined with our new scat-dog individual-matching technology, is being used to assess the abundance and distribution of maned wolves, puma and jaguar. Physiological health is being assessed using stress and reproductive hormones extracted from these same scat samples. All data are then layered onto a Geographic Information System (GIS), linking observed changes with specific habitat features.
Together, these methods provide means of conducting rapid, long term species monitoring that increases our understanding of barriers to wildlife movement, as well as guides the prioritization of lands for protection.
Results

During a six-week pilot study that occurred August-September 2004, three dog-handler teams collected more than 650 scat samples from puma, jaguar, and maned wolf within the study area.

Figure 1.  Sampling teams located maned wolf, puma samples in all present habitat types.  Jaguar were restricted to grassland, marsh, and Cerrado.

Figure 2.  Number of samples by species for sampling grids inside versus outside of the Park.

We have begun associating the spatial distribution of these scats with land use maps in order to provide some initial insights into how agricultural expansion is affecting wolf distribution.  Initial analyses of maned wolf, for example, suggest that some of the region’s population persist outside of the formally protected area, and that many individuals use both park and disturbed habitat.  Outside the park, the wolves appear to be most heavily concentrated along edge habitat comprised of park and pasture, or in natural habitat fragments found amongst agricultural fields.  Wolves also make extensive use of forest fragments on private lands surrounded by croplands and pasture.  Crop harvest cycle and amount of edge habitat predict presence of maned wolf outside the park.  Indeed, while the wolves are clearly making use of non-park habitat, 90 percent of the samples found outside the park were within 100 meters of natural forest fragment.  Based on known home range sizes, we conservatively estimate that we collected scat from a minimum of 60 individuals in the pilot study.

Figure 3.  Maned wolf samples found in non-natural habitat types (pasture and cropland) were more likely to be found at the edge, whereas those found in natural areas were less likely to be found at the habitat edge.  The sole exception to this was in marsh, which is by definition an edge habitat. 

Figure 4.  This map shows the area covered by our 2006 study and locations of giant anteater scats across the landscape. 

In 335 hours of surveying during our 2006 season, we encountered a total 356 giant anteater scats (256 outside of the Park) and 10 living giant anteaters. While we found little evidence of this species in the hostile matrix (cleared farmlands), individuals are clearly making use of habitat fragments and reproducing pairs are likely to be surviving in remnant habitat on the farms surrounding the park.  In all cases where scats were found on farmlands, they were found at the edge of natural habitat (edge defined as ≤ 100 meters). 
Implications

It is apparent that private lands conservation of even relatively small habitat fragments is enabling persistence of these species beyond Park boundaries.  DNA analysis of the collected scat samples will enable a better understanding of the dynamics of how individuals are using the landscape and the degree to which individuals might occupy isolated habitat patches. 

Giant Armadillo: 

The amazing giant armadillo is listed by IUCN as globally endangered and much of its remaining habitat is being rapidly destroyed.  While several attempts have been made to study giant armadillos, this species has remained elusive and there is very little understanding of their secretive lifestyles.  Giant armadillos are thought to be ant and termite dietary specialists, and spend much of their time underground in burrows. 

Giant armadillos are rarely kept in captivity.  Consequently, we had only two fecal samples (kindly made available by the Jaguar Conservation Fund) for training our detection dogs to find scat from this species.  Despite that limitation, our dogs found up to 29 scats from wild giant armadillos, which are now being confirmed by DNA analyses.

One of 132 giant armadillo burrows located by our survey (see map  below).  The identification of burrows and scats in and around the Park may provide an important snapshot of how these species are utilizing the landscape.

Preliminary assessments of dog accuracy were made by comparing locations of detected scat samples to areas where we found evidence of giant armadillo burrowing – see map).  Giant armadillo scat could be confused with that of several smaller, more abundant armadillo species in the area.  However, if dogs were hitting scat of these more abundant species, they should have acquired many more scat samples, and, these should not have been localized around giant armadillo burrows.  By analogy, our dogs found large numbers of scats from the abundant giant anteaters they were trained to detect in the same study area. 

In summary, preliminary data suggest that the detection dogs were effective and accurate at detecting this cryptic species, providing more information from more individuals than any other study of this globally endangered species to date.  Assuming that DNA analyses confirm these samples to be from giant armadillo, we expect the dogs to perform evev better next year because we now have many more samples to use for dog training.  This would provide an invaluable tool for gaining insights into the life of this cryptic, globally endangered species.

Figure 5. Giant Armadillo Scats and Burrows.  Giant armadillo burrows, which may be identified based on their distinctive size and shape, were recorded when seen by handlers or field assistants.  All white dots indicate a presumed giant armadillo scat (based on size and shape) found by our detector dogs.
Acknowledgements

This work is being supported by the Morris Animal Foundation, Center for Applied Biodiversity Science at Conservational International, the Gordon and Betty Moore Foundation, and the Center for Conservation Biology.  Graduate research fellowships have also been awarded to Carly from the National Science Foundation and the National Security Education Program based on this work.  The project is a partnership with the Jaguar Conservation Fund and the Brazil Program of Conservation International.

Assistance in obtaining scat samples for training purposes has been provided by the Jaguar Conservation Fund, the Smithsonian Conservation and Research Center , the Woodland Park Zoo, and Northwest Trek. Scat detection dogs were trained in collaboration with PackLeader Dog Training.

For Further Information
Carly Vynne, Ph.D. Candidate