CSIRO PUBLISHING

Wildlife Research
http://dx.doi.org/10.1071/WR11176

First telemetry study of bush dogs: home range, activity

and habitat selection

Edson de Souza Lima A,B,G,I, Karen E. DeMatteo C,D, Rodrigo S. P. Jorge B,E,G, Maria Luisa S. P.
JorgeF,G, Julio Cesar DalponteG, Herson Souza Lima A and Stuart A. KlorfineH
A
UNEMAT, Departamento de Ciência Biológicas, Nova Xavantina, MT, Brazil.
B
Centro Nacional de Pesquisa e Conservação de Mamíferos Carnívoros, Instituto Chico Mendes de Conservação da
Biodiversidade (CENAP/ICMBio), Atibaia, SP, Brazil.
C
University of Missouri-St Louis, Department of Biology, St Louis, Missouri, USA.
D
University of Missouri-Columbia, Division of Biological Sciences, Columbia, Missouri, USA.
E
Instituto Brasileiro para a Medicina da Conservação – TRÍADE, Recife, PE, Brazil.
F
Instituto de Biociências, Departamento de Ecologia, UNESP – Campus Rio Claro, Rio Claro, SP, Brazil.
G
Instituto para a Conservação dos Carnivoros Neotropicais/Pro-Carnívoros, Atibaia, SP, Brazil.
H
University of Washington, Department of Professional and Continuing Education, Seattle, Washington, USA.
I
Corresponding author. Email: edsolima@hotmail.com

Abstract
Context. The bush dog (Speothos venaticus) is difficult to observe, capture, and study. To date, indirect evidence and
opportunistic field observations have been the primary sources of information about the species’ ecology. Field data are
urgently needed to clarify the species’ ecological requirements, behaviour and movement patterns.
Aims. The present study uses 13 months of telemetry data from a group of bush dogs to begin to address questions about
area requirements, habitat preferences and movement patterns of this difficult-to-study species.
Methods. We tracked a group of bush dogs (two adults, one juvenile, four young) in an area of intact and altered Cerrado
(woodland–savanna biome) in eastern Mato Grosso, Brazil (Nova Xavantina District).
Key results. The group had a total home range of 140 km2 (fixed kernel 95%), with smaller seasonal ‘subareas’ (areas used
for 1–2 months before moving to another area, with repetition of some areas over time) and demonstrated a preference for
native habitats.
Conclusions. The bush dog’s home range is greater than that of other canids of the same size, even correcting for group
size. Patterns of seasonal movement are also different from what has been observed in other South American canids.
Implications. From our observations in the Brazilian savanna, bush dogs need large tracks of native habitat for their long-
term persistence. Although the present study is based on a single pack, it is highly relevant for bush dog conservation because
it provides novel information on the species’ spatial requirements and habitat preferences.

Additional keywords: canid, habitat use, social living, Speothos venaticus.

Received 18 October 2011, accepted 17 May 2012, published online 1 August 2012

Introduction and reproduction, comes from captive animals (Kleiman 1972;

The bush dog (Speothos venaticus), a highly social South Porton et al. 1987; Montalli and Kelly 1989; DeMatteo et al.
American canid, is considered Threatened by the Brazilian 2006). Knowledge of its ecology has been primarily gained
Ministry of Environment (MMA 2003) and Near Threatened through indirect evidence (scats and footprints, DeMatteo
by the IUCN (Zuercher et al. 2008). They hunt in groups of 2–10 et al. 2004; Zuercher et al. 2005; Lima et al. 2009; interviews,
individuals (Aquino and Puertas 1997; Barnett et al. 2001; DeMatteo 2008; DeMatteo and Loiselle 2008; detection dogs,
Beisiegel and Zuercher 2005) for a variety of small- and DeMatteo et al. 2009) or opportunistic field observations (Peres
medium-sized prey (Zuercher et al. 2005; DeMatteo 2008; 1991; Beisiegel 2009; Oliveira 2009; Michalski 2010).
Lima et al. 2009). The species appears to be naturally rare Systematic field data are urgently needed to clarify the
(Beisiegel and Zuercher 2005; DeMatteo 2008) and difficult to species’ ecological requirements, behaviour and movement
observe, frequently hiding under low vegetation (Lima et al. patterns. For example, bush dogs are associated with a variety
2009). Most information about bush dogs, including behaviour of intact (fields, savannas, forest and flooded areas, Zuercher et al.
Journal compilation
CSIRO 2012 www.publish.csiro.au/journals/wr
B Wildlife Research E. S. Lima et al.

2005; DeMatteo 2008; Lima et al. 2009; Oliveira 2009) and Monitoring and activity pattern
modified habitats (DeMatteo and Loiselle 2008; Oliveira 2009); For 13 months (from 13 July 2004 to 13 August 2005), radio-
however, it is unknown whether their ecological requirements collared animals were monitored on the ground with a TR4 radio
vary within those. In addition, although estimates of home- receiver and RA-14 H-directional antenna (Telonics, Mesa,
range size exist, they vary tremendously (4.56–150 km2; Arizona, USA). Even though two individuals were radio-
Silveira et al. 1998; Van Humbeck and Pérez 1998; Beisiegel collared, they were considered a single unit for all movement
1999) and are indirectly inferred from body size and diet. The analyses because they typically travelled together. Twenty-four-
present study uses 13 months of telemetry data from a group of hour monitoring sessions took place 3–4 days per week, with an
bush dogs to begin to address questions related to area average of 13.9  2.9 sampling days per month. During each
requirements, habitat preferences and movement patterns. monitoring session, locations were taken every 3 h, so as to ensure
an unbiased sampling of overall space use (Aebischer et al. 1993).
Materials and methods All coordinates were recorded with a Rino 110 GPS (Garmin,
Olathe, Kansas, USA). The location of each individual was
Study area determined either directly (recording the point where the group
The study was conducted between the São Rafael and Índio was determined to be present, such as resting in a den site) or by
Rivers, ~27 km north of the municipal district of Nova Xavantina triangulation (Jacob and Rudran 2003). Estimated location by
(between 14.365S, 52.554W and 14.549S, 52.354W). Nova triangulation was determined using geographic coordinates
Xavantina is located in the state of Mato Grosso in the eastern and respective angular direction of two points of high-quality
basin of the Mortes River that flows into the Araguaia River and signal and visual angle of ~90 to each other relative to the
has an economy based on agriculture and cattle raising. Since animal. Magnetic corrections were made in all directions for
1970, a large portion of its native vegetation has been cleared for each point with the program GPS TrackMaker 12.0 (http://
cultivation; however, some native vegetation still remains, www.gpstm.com/eng/contract.htm, verified 1 December 2010).
especially where the terrain is elevated, because local ranchers Activity pattern of each marked individual was monitored every
lack equipment to clear it. In the study area, elevation varies 15 min, during 24-h monitoring sessions by listening to the
and Cerrado (woodland–savanna biome) is the typical native activity switch on the radio-collar and scoring as either active
vegetation. The soils are predominantly Red–Yellow Dystrophic (moving) or inactive (at rest).
Oxisols, with a moderate horizon, medium texture, with or
without lateritic concretions (MME 1981). The climate
is characterised as type Aw in the Köppen system, with Home range and habitat use
average annual rainfall of 1600 mm (Nimer 1989), and two The home-range size of the monitored animals was determined
distinct seasons, namely, wet (October–March) and dry with fixed kernel 95% (FK 95%; Worton 1987, 1989), using the
(April–September). Animal Movement Extension of ArcView 3.1 (ESRI, Redlands,
California, USA). FK is considered to be a more accurate
estimator of the area an animal uses during its normal
Animal live-trapping activities (Worton 1989). Although home-range estimates can
Initial capture attempts were made with leg-hold (Soft Catch be affected by sample size (Harris et al. 1990; Seaman et al. 1999;
Spring Trap, Forestry Suppliers, Jackson, Mississippi, USA) Börger et al. 2006; Wauters et al. 2007), our FK estimates were
and cage (Tomahawk Live Trap, Tomahawk, Wisconsin, calculated with 30 locations, the recommended minimum,
USA) traps in areas with high potential for presence of bush and most with 50 locations, the preferred minimum (Seaman
dogs, such as locations where groups were observed by local et al. 1999). The area (km2) used by a single animal in the
ranchers. Additionally, we opportunistically surveyed burrows monitored group was calculated by dividing the home-range
looking for fresh signs of recent burrow activity by bush dogs size for the group by the weighted average number of
(e.g. scat, tracks). When an active burrow was located, its entrance individuals in the group during the monitoring period. This
was hand-excavated until the animals were within ~1.5 m and calculation, when compared with data collected in future field
could be captured with a restraining pole lasso. Once restrained, studies, can be used to understand how group size affects the
adults were removed from the burrow and sedated with a species’ spatial requirements.
5 mg kg–1 intramuscular injection of tiletamine hydrochloride To evaluate whether habitat use was random in relation to
and zolazepam hydrochloride (Zoletil 50, Virbac, France). availability, proportions of use were generated from an
Each adult was sedated and fit with a radio-collar (Advanced utilisation distribution (UD) instead of the direct locations,
Telemetry Systems, Isanti, Minnesota, USA) that had a design because the UD quantifies use within the home range with a
specific for the species (DeMatteo and Kochanny 2004), weight probabilistic and continuous metric and eliminates concerns
of 65 g, VHF frequency of 164 MHz and an activity switch. about independence between points (Marzluff et al. 2004;
After a full recovery from sedation, individuals were released Millspaugh et al. 2006). We also calculated the proportion of
into the burrow system where they were captured. Individuals resources from a vegetation grid, which correlates, cell-by-cell,
were noted to remain in the burrow until the following day when with the grid generated for the UD.
their daily movement patterns resumed. All handling of live The vegetation grid was derived from a colour composition,
animals was approved by the Instituto Brasileiro de Meio created with Spring 4.3.3. (Ministério da Ciência e Tecnologia
Ambiente e dos Recursos Naturais Renováveis (Licence (MCT) – Instituto de Pesquisas Espaciais (INPE)), of a 2004
02027.002452/2004-95). CBERS image (INPE), using ArcView 3.2 (Esri). By using an
First telemetry study of bush dogs Wildlife Research C

image from 2004, versus a more recent image, we were able to locations and 142 (14.2%) as triangulations, and estimated
accurately access habitat use relative to availability for the period distance to the animal of ~300 m. In total, 20 211 activity
the group was monitored. The image was classified into the records were collected (adult male = 10 770 and adult
following three vegetation categories: savanna (all native female = 9441). An average of 112  8.5 records (range:
grassland formations, including swampy grassland, grassland, 91–125) and 98  6.5 records (range: 82–105) were collected
shrub grassland, grassland with scattered short trees, Cerrado in each monitoring session for the male and female, respectively.
sensu stricto and dense Cerrado), forest (gallery forest, dry forest, Both animals were primarily active during early morning
valley forest and Cerradão) and cultivated areas (pasture and (0600 hours to 0700 hours) and night (2000 hours to
agricultural crops). 2300 hours), with a higher proportion of activity during the
We then created a resource utilisation function (RUF) with morning than the night period (Fig. 1).
the Focal Patch extension of ArcView 3.2 (Marzluff et al. 2004;
Millspaugh et al. 2006). The RUF expresses the correlation Home range and habitat use
between the UD (group’s relative use of space within its home
The home-range estimate using 95% FK (n = 1002) was
range) and the three vegetation categories. The program
140.86 km2. The average area used per individual in the
extension creates a table, with use and vegetation values for
monitored group was 40 km2 (weighted average number of
each cell of the home-range grid. The use values vary from 1 to 99
individuals in the group = 3.5) and ranged from 20 km2
(1 being the lowest use) and the vegetation categories from 1 to 3
(140 km2 per 7 individuals) to 47 km2 (140 km2 per 3
(each being a single vegetation category). We separated relative
individuals). The 95% minimum convex polygon estimate for
use into two (low v. high – 50% of the cells each) and three
the entire period was 160.97 km2 and not statistically different
categories (low, medium, and high – 33% of the cells each) and
from 95% FK (c2 = 1.272, P = 0.260).
tested for statistically significant differences in proportion of use
Focal patch analysis (ArcView 3.2) generated 6656 cells,
and availability per vegetation type with a G-test.
each of them with a correspondent intensity-of-use value
Movement patterns (going from 1 to 99, based on the FK home-range estimate)
and a resource categorical-value (based on the vegetation grid
To evaluate bush-dog movement patterns, we considered the with 1 = cultivated land, 2 = forest, and 3 = savanna). Cell size was
following three possibilities: (1) nomadic, if the group never 130 m2 and it was determined in the FK analysis of home range,
repeated the use of a single area and constantly moved to a through a least-square cross-validation. The vegetation grid was
different area; (2) semi-nomadic, if the group used an area for a classified as 14% cultivated land (n = 961), 14% forest (n = 930)
short period (1–2 months) before moving to another area, with the and 72% savanna (n = 4765).
group repeating previous sites over time and (3) territorial, if the In both analyses (two and three categories), habitat use was
group repeatedly used the same overall area. All burrows used by not proportional to habitat availability for two habitat types,
the monitored group were geo-referenced to evaluate the use and namely, cultivated lands and savanna (Fig. 2). In cultivated
reuse of individual burrows. lands, there was a higher-than-expected proportion of low-use
cells and a lower-than-expected proportion of medium- and high-
Results
use cells (G2cat = 212.28, d.f. = 1, P < 0.001; G3cat = 232.89,
Animal live-trapping d.f. = 2, P < 0.001). The opposite pattern was found in savanna
On 13 July 2004, a group of three individuals (a pair of adults and a habitat, with a lower-than-expected proportion of low-use cells
juvenile about 6 months of age) was captured inside a giant and a higher-than-expected proportion of medium- and high-use
armadillo (Priodontes maximus) burrow. Radio-collars were cells (G2cat = 51.22, d.f. = 1, P < 0.001; G3cat = 40.76, d.f. = 2,
placed on the two adults. On 20 November 2004, the group P < 0.001). In the forest habitat, results diverged depending on
grew in size with the birth of four pups (two males, one female the number of categories. With two categories (low and high use),
and one unknown). On 20 December 2004, the juvenile results were marginally significant, with low-use cells marginally
(approximately 1-year old) disappeared from the group and more frequent than was expected (O/E = 1.06, or 6% more,
was never found. Approximately 1 week later, the pup of G2cat = 3.01, d.f. = 1, P = 0.08). With three categories (low,
unknown sex disappeared and was assumed deceased. Two of medium and high), the marginal significance disappeared
the three remaining pups (one female and one male) were noted (G3cat = 0.135, d.f. = 2, P = 0.935).
absent from the group in February 2005 and April 2005,
respectively. Movement patterns
In March 2005, the adult female began to lose hair, evolving to During 12 of the 13 months that the group was monitored, it
a severe generalised hair loss. In April 2005, both males (adult and covered five separate sites within the larger home range
7-month juvenile) also started to lose hair, progressing more (‘subareas’ or ‘temporary areas’; Table 1, Fig. 3). The
gradually to a severe generalised hair loss, similar to the female, ‘subareas’ were defined on the basis of the concentration of
with the loss more apparent in the adult male. The female died on locations in a given area for periods of 1–3 months. The
13 June 2005 and both males died ~13 August 2005. overall movement pattern did not involve the group repeatedly
covering the entire area within few weeks (territorial), nor did
Monitoring and activity pattern it involve the group moving from one area to another without
During the 13 months when the group was monitored, a total of ever repeating the same area (nomadic). Instead, the group
1002 point locations was recorded, with 860 (85.8%) as exact stayed for 1–2 months in one area, moved to another, stayed
D Wildlife Research E. S. Lima et al.

100

Male

80 Female
Proportion active (%)

60

40

20

0
00

0
00

00

0
00

0
00

0
00

00

00

0
00

00

0
0

0
0
0

:0
:0
:0
:0

:0

:0

:0
:0

:0
:0

:0

:0
:0
:0
0:

2:

9:
3:

7:
1:

6:

8:
4:

5:

19
16
11
10

12

15

20
17

18
14

21

23
22
13

Time of the day

Fig. 1. Daily activity patterns for both adult bush dogs where records were collected every 15 min during a 24-h monitoring
session. The male (grey) had an average of 112  8.5 records each 24-h monitoring session and the female (black) had an
average of 98  6.5 records each 24-h monitoring session.

for another 1–2 months, and either returned to a previous area, or sympatric South American canids, which differ in degree of
moved to another adjacent area, staying for another 1–2-month sociality (solitary or live in pairs) and diet (omnivorous), bush-
period, and moving again, on a semi-nomadic movement pattern dog home range is dramatically higher, irrespective of body
(Fig. 3). The average 95% FK per ‘subarea’ was 17.9  12.8 km2 size, e.g. maned wolf (Chrysocyon brachyurus) = 67.69 km2
(Table 2). (95% MCP) or 80.18 km2 (FK; Jácomo et al. 2009), crab-
It was very rare for bush dogs to use a den for more than a eating fox (Cerdocyon thous) = 5.32 km2 (restricted polygon
single day or to return to a den that they had previously used. Of (RP); Macdonald and Courtenay 1996) and hoary fox
the 150 geo-referenced lairs that the group used, 147 were used (Pseudalopex vetulus) = 456 ha (95% RP; Courtenay et al.
once, one was used twice and two were used three times. The re- 2006). It is possible that the bush dog’s extensive home range
used dens were associated with reproduction, specifically is associated with its sociality (group living) and restricted diet
mating and birthing. The group used dens of three armadillo (hyper-carnivorous – Zuercher et al. 2005). Although additional
species (nine-banded armadillo/Dasypus novemcinctus, giant data are needed to develop a population-level model for the
armadillo/Priodontes maximus, six-banded armadillo/ species and understand how home range may vary with habitat
Euphractus sexcinctus) and was observed digging and type, degree of fragmentation, prey density and group size, this
cleaning out abandoned armadillo dens, using them as a night value provides an important baseline for initiating species-
shelter, and then leaving the burrow the following morning. specific conservation efforts.
Although the entire group was typically found in a single den, We believe that this semi-nomadic movement pattern is
there were exceptions when the pups were 2–3 months of age, tightly associated with the bush dog’s observed intensive
namely, the adults would leave them in a separate den during predation rates (1 armadillo per 1–2 days); however, additional
the day and retrieve them when starting their night-time activity. studies on prey density are needed to confirm whether their
During the day, the pups were unprotected and free to exit the constant movement within an area and rotational movement
den, which could explain some of the losses of young (e.g. among areas functions to minimise negative effects on prey
predation). density. The group’s constant changing of den sites may be an
energy-saving adaptation to this semi-nomadic pattern, because
Discussion the group uses readily available local dens when they are ready
Home-range estimate (95% FK = 140.86 km2) for the group of to rest instead of traveling back to their previous den, which can
bush dogs (three to seven individuals) monitored in the present be kilometres away. The only period when they change their
study is very close to indirect estimates of bush-dog home pattern of den use is when newborns are present. In those periods,
ranges (150.31 km2 per a group of six animals; Beisiegel the group utilises the same burrow for a longer period, possibly
1999). When the bush-dog home range is compared with other to lower the pup’s exposure to predators.
First telemetry study of bush dogs Wildlife Research E

Intensity of use Table 1. Period that the monitored group of bush dogs visited each
of the five separate sites over a 12-month period
(a) Low High The pattern of staying temporarily in different locations but repeatedly visiting
1.6
one of the areas (Area 2) indicates a semi-nomadic movement pattern

1.2 Date Duration of stay Subarea

(days/months) no.

0.8
13 Jul. 2004 to 14 Aug. 2004 33/1.1 1
15 Aug. 2004 to 28 Oct. 2004 74/2.5 2a
11 Nov. 2004 to 17 Jan. 2005 67/2.2 3 – birth
0.4 18 Jan. 2005 to 31 Mar. 2005 73/2.4 2b
Observed/Expected

7 Apr. 2005 to 21 Apr. 2005 14/0.5 4

4 May 2005 to 24 May 2005 20/0.7 2c
0.0 3 Jun. 2005 to 13 Jul. 2005 40/1.3 5
A
14 Jul. 2005 to 13 Aug. 2005 31/1.0
A
(b) Although the adult male and juvenile remained within the overall home
1.6
Low Medium High range of the group during the last 31 days of monitoring, they shifted to an
aberrant movement pattern. Specifically, they moved among Areas 2, 1,
1.2
and 4, before returning to Area 5, where both died (13 August 2005). This
could be a side-effect of losing the adult breeding female (13 July 2005) in
the group and possibly searching for a female companion, and we decided
0.8 to error on the side of caution and not include this data as an independent
subarea.

0.4
The group used savanna habitat more than predicted by its
0.0
availability, and cultivated lands less than predicted by their
availability. The group’s movement and behaviour support this
Cultivated Forest Savanna
higher-than-expected savanna use, which we believe is
Habitat type associated with widely available shelters (e.g. high grass for
Fig. 2. Habitat-use analyses (a) with two categories of use (low v. high) or hiding, dens for escape), and more abundant prey in native
(b) with three categories of use (low, medium and high) demonstrated that than in cultivated habitats. The latter is supported by the
the monitored bush dogs used cultivated areas less than expected by their group’s lower observed kill rate in cultivated areas (32
availability, forests at the same level as expected by their availability, and armadillo-predation events were in the savanna, eight were in
savannas more than expected by their availability (see text for results of the forest and one was in the cultivated habitat), possibly an
statistical analyses). indication of low armadillo abundance, which appears to be their
preferred prey. In general, when the group was encountered in
The daily activity pattern of the bush dog is similar to three cultivated habitat, they never stayed for extended periods of time,
other South American canids (maned wolf, crab-eating fox, hoary and instead travelled quickly until native habitat was reached. The
fox, Jácomo et al. 2004); however, unlike for these other canids, group’s avoidance of cultivated habitat and preference for native
the bush dog’s morning peak is more intense (shorter in total habitat demonstrated the importance for understanding the
duration) than the night period. These activity peaks are possibly range of effects that fragmentation may have on the ecological
associated with their prey’s observed activity patterns. Nine- requirements of the bush dog. With fragmentation continually
banded armadillos were observed to be the main prey of bush increasing throughout much of the bush-dog distribution
dogs in the present study (97.6%, or 41 of the 42 predation (DeMatteo and Loiselle 2008), it is possible that the bush dog
observations) and in another study in a similar habitat (94.1% will require larger home ranges so as to compensate for lower
occurrence in scats; Lima et al. 2009). Nine-banded armadillos density of basic ecological requirements (e.g. prey, shelter) in
are primarily nocturnal but return to their burrows in early cultivated habitats.
morning (Emmons and Feer 1990). Bush dogs have a body During this 13-month study, there were several changes in
structure well adapted to remove prey from burrows (Lima group size. Although predation is the possible cause of the
et al. 2009) and observations of the monitored group indicate disappearance of the pups, it is not the only explanation for
that once the group located an armadillo in its den, they could very the disappearance of the 1-year-old juvenile. With the latter, it
efficiently dig to remove and kill it. The majority of the observed is possible that the individual dispersed from the group.
predation events (armadillos removed from their burrows) Although this contradicts captive data (extended family
occurred during the morning activity peak when fresh groups) and other opportunistic field observations (e.g. 10–12
armadillo routes might be easier to track. Few predation events individuals per group; Beisiegel and Zuercher 2005), pack size
were recorded during the night, when armadillos were outside may be a plastic trait dependent on many factors (habitat
their burrows. Overall, observations indicated that it seemed quality, time since pack formation, prey density; Sillero-Zubiri
more difficult for the bush dogs to secure an armadillo foraging et al. 2004); however, additional field data are needed to clarify
in open areas than it was to secure one confined in a burrow. this question.
F Wildlife Research E. S. Lima et al.

Area 1 (13 Jul. to 14 Aug. 2004)

Area 2a (15 Aug. to 28 Oct. 2004)

Area 3 - Birth (11 Nov. 2004 to 17 Jan. 2005)

Area 2b (18 Jan. to 31 Mar. 2005)

Area 4 (7 Apr. to 21 Apr. 2005)

Area 2c (4 May to 24 May 2005)

Area 5 (3 Jun. to 13 Jul. 2005)

0 6 12 km
Streams

Fig. 3. Map of the five separate sites, or ‘subareas’, within the larger home range of the monitored bush-dog group. Each black dot
represents a location recorded between 13 July 2004 and 13 August 2005. The duration of stay, number of locations and estimated area
for each ‘subarea’ are presented in Tables 1 and 2. The last period of monitoring (14 July to 13 August 2005) was not defined as a
subarea (see Table 1).

Table 2. Summary of total home-range size and area for each separate
site, or ‘subarea’, with 95% fixed kernel (FK 95%) Acknowledgements
This project was conducted with the help of Centro Nacional de Pesquisa e
Subarea no. FK 95% (km2) No. of points Conservação de Mamíferos Carnívoros of the Instituto Chico Mendes de
1 9.9 122 Conservação da Biodiversidade (CENAP/ICMBio). We thank Instituto
2a 33.9 255 Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis
3 – birth 13.6 181 (IBAMA) for the necessary permits to conduct this study. We are grateful
2b 33.6 152 to the generous financial support provided by the Zoological Society of
4 6.8 40 Chicago, Palm Beach Zoo Conservation Fund, Cleveland Metroparks Zoo
2c 20.2 61 Small Grants Program, Pittsburgh Zoo & PPG Aquarium Conservation Fund,
5 40 74 and Idea Wild. The CBERS satellite image was kindly provided by the
Instituto Nacional de Pesquisas Espaciais (INPE). We thank all of the field
assistants who help with the collection of data, including R. Sollmann and
N. Luz. We appreciate the services provided by both veterinarians in the
Another possible explanation for the loss or disappearance of capture and handling of animals, namely, D. H. de Morais and M. R. Santos.
individuals from the group is the generalised hair loss that was A special thanks goes to all the private land owners that allowed the research
seen in the individuals from the study group. Although it was not to occur in their properties, especially the family of D. Stanke, who also aided
possible to definitely determine the cause of the generalised hair in many other aspects of the research project. We appreciate the time and
loss, it is suspected that it was a type of mange, one of the diseases effort of two anonymous reviewers whose comments helped improve earlier
drafts of the manuscript.
reported to occur in bush dogs (DeMatteo 2008). This is
supported by the diagnosis of Sarcoptes scabiei in two bush
dogs in an area near the area of the monitored group (Jorge et al.
2008). Additional research is also needed to understand the range References
and impact of diseases which bush dogs are susceptible to, Aebischer, N. J., Robertson, P. A., and Kenward, R. E. (1993). Compositional
especially with the increasing loss of habitat and their analyses of habitat use from animal radio-tracking date. Ecology 74,
inevitable interactions with domestic dogs. 1313–1325. doi:10.2307/1940062
In summary, although only one group of bush dogs was Aquino, R., and Puertas, P. (1997). Observations of Speothos venaticus
monitored, our study is the first to shed light on important (Canidae: Carnivora) in its natural habitat in Peruvian Amazonia.
Zeitschrift fur Saugetierkunde 62, 117–118.
questions related to ecological requirements of the species,
Barnett, A., Shapley, R., and Engstrom, M. (2001). Records of the bush dog,
helping develop more sounding conservation and management Speothos venaticus (Lund, 1842) from Guyana. Mammalia 65, 232–237.
plans. Continued effort is greatly needed if we want to better Beisiegel, B. M. (1999). Contribuição ao estudo da história natural do
understand the responses of one the least known species of cachorro do mato, Cerdocyon thous, e do cachorro vinagre, Speothos
Neotropical canid to the increasing fragmentation and venaticus. Ph.D. Thesis, Universidade de São Paulo, São Paulo,
degradation of native habitats. Brazil.
First telemetry study of bush dogs Wildlife Research G

Beisiegel, B. M. (2009). First camera trap record of bush dogs in the state of Kleiman, D. G. (1972). Social behavior of the maned wolf (Chrysocyon
São Paulo, Brazil. Canid News 12.5 [online]. Available at http://www. brachyurus) and bush dog (Speothos venaticus): a study in contrast.
canids.org/canidnews/12/Bush_dogs_in Sao_Paulo.pdf [verified 20 Journal of Mammalogy 53, 791–806. doi:10.2307/1379214
January 2011]. Lima, E. S., Jorge, R. S. P., and Dalponte, J. C. (2009). Habitat use and diet of
Beisiegel, B. M., and Zuercher, G. L. (2005). Speothos venaticus. Mammalian bush dogs, Speothos venaticus, in the northern Pantanal, Mato Grosso,
Species 783, 1–6. doi:10.1644/783.1 Brazil. Mammalia 73, 13–19. doi:10.1515/MAMM.2009.002
Börger, I., Franconi, N., De Michele, G., Gantz, A., Meschi, F., Manica, A., Macdonald, D. W., and Courtenay, O. (1996). Enduring social relationships in
Lovari, S., and Coulson, T. (2006). Effects of sampling regime on the a population of crab-eating zorros, Cerdocyon thous, in Amazonian Brazil
mean and variance of home range size estimates. Journal of Animal (Carnivora, Canidae). Journal of Zoology 239, 329–355. doi:10.1111/
Ecology 75, 1393–1405. doi:10.1111/j.1365-2656.2006.01164.x j.1469-7998.1996.tb05454.x
Courtenay, O., Macdonald, D. W., Gillingham, S., Almeida, G., and Dias, R. Marzluff, J. M., Millspaugh, J. J., Hurvitz, P., and Handcock, M. S. (2004).
(2006). First observations on South America’s largely insectivorous Relating resources to a probabilistic measure of space use: forest
canid: the hoary fox (Pseudalopex vetulus). Journal of Zoology 268, fragments and Steller’s jays. Ecology 85, 1411–1427. doi:10.1890/03-
45–54. doi:10.1111/j.1469-7998.2005.00021.x 0114
DeMatteo, K. E. (2008). Using a survey of carnivore conservationists to gain Michalski, F. (2010). The bush dog Speothos venaticus and short-eared dog
insight into the ecology and conservation status of the bush dog. Canid Atelocynus microtis in a fragmented landscape in southern Amazonia.
News 11.3 [online]. Available at http://www.canids.org/canidnews/11/ Oryx 44, 300–303. doi:10.1017/S0030605309990871
Bush_dog_data_survey.pdf [verified 20 December 2010]. Millspaugh, J. M, Nielson, R. M., McDonald, L., Marzluff, J. M., Gitzen,
DeMatteo, K. E., and Kochanny, C. (2004). Determining an effective and R. A., Rittenhouse, C. D., Hubbard, M. W., and Sheriff, S. L. (2006).
safe radio-tracking collar for bush dogs (Speothos venaticus) Canid Analysis of resources selection using utilization distributions. The Journal
News 7.7 [online]. Available at http://www.canids.org/canidnews/ of Wildlife Management 70, 384–395. doi:10.2193/0022-541X(2006)70
7/Radiotracking_collar_for_bush_dogs.pdf [verified 12 January 2011]. [384:AORSUU]2.0.CO;2
DeMatteo, K. E., and Loiselle, B. A. (2008). New data on the status and MMA (Ministry of the Environment of Brazil) (2003). Instrução normativa,
distribution of the bush dog (Speothos venaticus): evaluating its quality of n. 3. In ‘Diário Oficial de União’, 101. (Ed. Brazilian Government.)
protection and directing research efforts. Biological Conservation 141, pp. 88–97 (Imprensa Nacional: Brasília, Brazil.)
2494–2505. doi:10.1016/j.biocon.2008.07.010 MME (Ministry of Mining and Energy of Brazil) (1981). Goias: geologia,
DeMatteo, K. E., Carrillo, O., Zuercher, G. L., Ramírez, S., Smith, K., and geomorfologia, vegetação, uso potencial da terra. In ‘Projeto
Porton, I. J. (2004). A technique for attracting bush dogs (Speothos RADAMBRASIL. Levantamento de Recursos Naturais’. (Ed. Ministry
venaticus) in the wild. Canid News 7.6 [online]. Available at http:// of Mining and Energy of Brazil.) p. SD.22. (Ministry of Mining and
www.canids.org/canidnews/7/Attracting_bush_dogs.pdf [verified 10 Energy of Brazil: Rio de Janeiro.)
December 2010]. Montalli, R. J., and Kelly, K. (1989). Pathologic survey and review of diseases
DeMatteo, K. E., Porton, I. J., Kleiman, D. G., and Asa, C. S. (2006). The effect of captive maned wolves (Chrysocyon brachyurus) and bush dogs
of the male bush dog (Speothos venaticus) on the female reproductive (Speothos venaticus). Erkrankungen Der Zootiere 61, 67–79.
cycle. Journal of Mammalogy 87, 723–732. doi:10.1644/05-MAMM-A- Nimer, E. (Ed.) (1989). ‘Climatologia do Brasil.’ (Fundação IBGE: Rio de
342R1.1 Janeiro.)
DeMatteo, K. E., Rinas, M. A., Sede, M. M., Davenport, B., Argüelles, C., Oliveira, T. G. (2009). Distribution, habitat utilization and conservation of the
Lovett, K., and Parker, P. G. (2009). Detector dogs: an effective technique vulnerable bush dog Speothos venaticus in northern Brazil. Oryx 43,
for bush dogs (Speothos venaticus) surveys. The Journal of Wildlife 247–253. doi:10.1017/S0030605307002347
Management 73, 1436–1440. doi:10.2193/2008-545 Peres, C. A. (1991). Observations on hunting by small-eared (Atelocynus
Emmons, L. H., and Feer, F. (1990). ‘Neotropical Rainforest Mammals: a microtis) and bush dogs (Speothos venaticus) in central-western
Field Guide.’ (University of Chicago Press: Chicago, IL.) Amazonia. Mammalia 55, 635–639.
Harris, S., Cresswell, W. J., Forde, P. G., Trewhella, W. J., Woollard, T., and Porton, I., Kleiman, D. G., and Rodden, M. (1987). Aseasonality of bush dog
Wray, S. (1990). Home-range analysis using radio-tracking data: a review reproduction and the influence of social factors on the estrous cycle.
of problems and techniques particularly as applied to the study of Journal of Mammalogy 68, 867–871. doi:10.2307/1381569
mammals. Mammal Review 20, 97–123. doi:10.1111/j.1365- Seaman, D. E., Millspaugh, J. J., Kernohan, B. J., Brundige, G. C., Raedeke,
2907.1990.tb00106.x K. J., and Gitzen, R. A. (1999). Effects of sample size on kernal home
Jacob, A. A., and Rudran, R. (2003). Radiotelemetria em estudos range estimates. The Journal of Wildlife Management 63, 739–747.
populacionais. In ‘Métodos de Estudos em Biologia da Conservação e doi:10.2307/3802664
Manejo da Vida Silvestre’. (Eds L. Cullen Jr, R. Rudran and C. Valladares- Sillero-Zubiri, C., Hoffmann, M., and Macdonald, D. W. (2004). ’Canids:
Padua.) pp. 285–342. (Editora UFPR: Curitiba, Brazil.) Foxes, Wolves, Jackals, and Dogs.’ (IUCN Publication Services Unit:
Jácomo, A. T. A., Silveira, L., and Diniz-Filho, J. A. F. (2004). Niche Cambridge, UK.)
separation between the maned wolf (Chysocyon brachyurus), the Silveira, L., Jácomo, A. T. A., Rodrigues, F. H. G., and Diniz-Filho, J. A. F.
creab-eating fox (Dusicyon thous) and the hoary fox (Dusicyon (1998). Bush dogs (Speothos venaticus), in Emas National Park, Central
vetulus) in Central Brazil. Journal of Zoology 262, 99–106. Brazil. Mammalia 62, 446–449.
doi:10.1017/S0952836903004473 Van Humbeck, J., and Pérez, N. (Eds.) (1998). ‘Estudios del Jagua Yvyguy,
Jácomo, A. T. A., Kashivakura, C. K., Ferro, C., Furtado, M. M., Astete, S. P., Speothos venaticus, en el Centro de Investigación de Animales Silvestres
Tôrres, N. M., Sollmann, R., and Silveira, L. (2009). Home range and de Itaipu, CIASI.’ (Superintendência de Médio Ambiente de la Itaipu
spatial organization of maned wolves in the Brazilian grasslands. Journal Binacional: Ciudad del Este, Paraguay.)
of Mammalogy 90, 150–157. doi:10.1644/07-MAMM-A-380.1 Wauters, L. A., Preatoni, D. G., Molinari, A., and Tosi, G. (2007). Radio-
Jorge, R. S. P., Lima, E. S., and Lucarts, L. E. B. (2008). Sarna sarcóptica tracking squirrels: performance of home range density and linkage
ameaçando cachorros-vinagres (Speothos venaticus) de vida livre em estimators with small range and sample size. Ecological Modelling
Nova Xavantina – MT. In ‘Proceedings of the Annual Meeting of the 202, 333–344. doi:10.1016/j.ecolmodel.2006.11.001
Brazilian Society of Zoos (SZB)’. Available at http://www.szb.org.br/ Worton, B. J. (1987). A review of models of home range for animal movement.
[verified 15 December 2010]. Ecological Modelling 38, 277–298. doi:10.1016/0304-3800(87)90101-3
H Wildlife Research E. S. Lima et al.

Worton, B. J. (1989). Kernel methods for estimating the utilization Zuercher, G. L., Swarner, M., Silveira, L., and Carrillo, O. (2008). Speothos
distribution in home-range studies. Ecology 70, 164–168. doi:10.2307/ venaticus. In ‘IUCN 2011. IUCN Red List of Threatened Species. Version
1938423 2011.1’. Available at http://www.iucnredlist.org/apps/redlist/details/
Zuercher, G. L., Gipson, P. S., and Carrillo, O. (2005). Diet and habitat 20468/0 [verified 17 October 2011].
associations of bush dogs Speothos venaticus in the Interior Atlantic
Forest of eastern Paraguay. Oryx 39, 86–89. doi:10.1017/S0030605305
000153

www.publish.csiro.au/journals/wr