Download Habitat selection by feral cats and dingoes in a

Austral Ecology (2002) 27, 26–31
Habitat selection by feral cats and dingoes in a semi-arid
woodland environment in central Australia
G. P. EDWARDS,1* N. DE PREU,2 I. V. CREALY1 AND B. J. SHAKESHAFT1
Parks and Wildlife Commission of the Northern Territory, PO Box 1046, Alice Springs,
Northern Territory 0871, Australia (Email: [email protected]) and 2Department of
Environment and Heritage, Hawker, South Australia, Australia
1
Abstract Habitat use by feral cats and dingoes was examined within a heterogeneous semi-arid woodland site
in central Australia over 2 years. Density estimates of feral cats based on tracks were higher in mulga habitat than
in open habitat. Isodar analysis implied that this pattern of habitat use by feral cats was consistent with the consumer-resource model of density-dependent habitat selection, which is an ideal free solution. The reason why mulga
supported higher densities of feral cats was unclear. Foraging success of feral cats may be higher in the mulga
because the stalk and ambush hunting tactics typically employed by felids are well suited to dense cover. Mulga
may also have offered feral cats more protection from dingo predation. Dingo activity was distributed uniformly
across habitats. The dingo isodar was statistically non-significant, suggesting that habitat selection by dingoes was
independent of density.
Key words: density-dependent habitat selection, dingo, dispersion, feral cat, habitat, isodar analysis.
INTRODUCTION
In this study we examined habitat use by feral cats
(Felis catus) and dingoes (Canis lupus dingo) within a
heterogeneous semi-arid woodland site in central
Australia. Although differential use of habitats has
been shown in both dingoes (Thomson 1992) and
feral cats (Konecny 1987; Alterio et al. 1998), no study
has examined the possible reasons underlying these
phenomena.
Habitat use in animals has been much studied over
the past 75 years (Rosenzweig 1991). From early on
it was recognized that within some species, individuals
actually choose which habitat to occupy and the range
of habitats occupied often increases with population
density (Mayr 1926; Svärdson 1949; Morisita 1952).
This phenomenon is known as density-dependent habitat selection. It appears to be a key factor underpinning
patterns of habitat use in many taxa (Rosenzweig 1991)
and it is potentially a powerful force influencing population dynamics and community organization (Morris
1988, 1992). As a result, the topic has received
considerable attention in recent years (e.g. Morris
1987, 1988, 1989, 1992, 1994; Messier et al. 1990;
Halama & Dueser 1994; Abramsky et al. 1997).
Fretwell and Lucas (1970) proposed the concept of
the ideal free distribution to explain density-dependent
habitat selection in a single species. The ideal free
distribution is based on the principles of optimal
*Corresponding author.
Accepted for publication June 2001.
foraging and intraspecific competition (Rosenzweig
1991) and so incorporates three additional elements
known to influence habitat selection in animals:
food availability, foraging success and competition.
The original concept of the ideal free distribution
assumed continuous resource input and instantaneous
consumption with organisms equilibrating their fitnesses by distributing themselves among the available
habitats in proportion to the resources available in
those habitats (Fretwell & Lucas 1970; Rosenzweig
1991). The ideal free distribution based on continuous
input is the simplest mathematical model of densitydependent habitat selection (Rosenzweig 1991). There
are other models, including ideal free models based
on discontinuous input (Morris 1994). Isodars, which
are regressions of density between two habitats, can
discriminate between the various models of habitat
selection (Morris 1988, 1994). Isodars have been used
to identify possible mechanisms underlying habitat
selection and population regulation in rodents (Morris
1987, 1988, 1989, 1992, 1994; Abramsky et al. 1997).
The present study is the first to apply a similar habitat
analysis to mammalian carnivores.
Another factor that can influence habitat selection
in animals is predation (Lima & Dill 1990). A large
number of studies have shown that in the presence of
a predator, prey species select protected habitats or
areas within habitats where the threat of predation is
diminished (e.g. Kotler 1984; Carey 1985; Gilliam &
Fraser 1987; Hughes et al. 1994). For the simple case
where predation risk is constant within a habitat, predation can be viewed as part of the cost of habitat use
HABITAT SELECTION IN FERAL CATS AND DINGOES
in much the same way as food availability and is already
accommodated in models of density-dependent habitat selection such as the ideal free distribution and its
derivatives (Rosenzweig 1991). Sometimes, however,
the influence of predation on habitat selection is more
complex. A range of foraging animals including
birds (Lima 1985) and mammals (Lima et al. 1985;
Brown 1988; Kotler et al. 1991) demonstrate a clear
predation–energy trade-off where the distance from
refuge or amount of illumination influences the time
spent in risky habitats. In the present study we considered the effects of dingo density on habitat use
by feral cats, as dingoes are a potential predator and
competitor of cats (Corbett 1995).
METHODS
Study site
The study site in the Northern Territory, approximately
110 km north-west of Alice Springs, was approximately
550 km2 in area and has been described in detail in
Edwards et al. (2000). In brief, it is dominated by a flat
plain that is elevated at 650 m a.s.l. The plain comprises
two land systems (= habitats) with different vegetation.
The Hamilton land system has texture contrast soils,
some red earths and red clay soils that support a mosaic
of open grassland dominated by wiregrass (Aristida
contorta) and open mixed woodland with scattered
mulga (Acacia aneura), ironwood (Acacia estrophiolata),
witchetty bush (Acacia kempeana), whitewood (Atalaya
hemiglauca) and bloodwood (Corymbia opaca). The
Bushy Park land system is flanked by the Hamilton land
system (Perry et al. 1962). It has red earths, which
support open to dense stands of mulga woodland with
scattered ironwood, witchetty bush and bloodwood
over grassland with shrubs (mainly Eremophila spp.).
Hereafter we refer to these land systems, respectively,
as ‘open’ habitat and ‘mulga’ habitat. Dingoes and
feral cats occur throughout the study area but red
foxes (Vulpes vulpes) and rabbits (Oryctolagus cuniculus)
are rare or absent (Edwards et al. 2000). Cattle (Bos
taurus) are grazed throughout the study site.
Track monitoring
The tracks of feral cats and dingoes were counted to
assess populations of these species on nine occasions
at approximately 3-month intervals starting in August
1995 and finishing in August 1997. Tracks were
counted along two 25-km stretches of dirt road
(= transects) in the eastern and western sections of
the study site. The roads had a sandy surface substrate
that allowed recording of tracks and other signs of
27
animals and provided broad coverage of the study
site. During each survey period, tracks were counted
along each transect every day for 3 consecutive days.
Twenty-four hours before the first count of a survey
period, each transect was cleared by towing a 1.8-m
long heavy steel bar behind a four-wheel-drive vehicle
to erase all previous tracks and signs. Track counts were
conducted during the following morning by two
observers driving all-terrain vehicles at average speeds
of 5–8 km h-1, each observer counting along one transect. Track counts commenced approximately 0.5 h
after sunrise and were completed within 3.5 h. Each
time that new feral cat or dingo tracks were detected
during a count they were closely inspected and identified to species. The following data were also recorded
for each new set of tracks: (i) the distance of the tracks
from the start point; (ii) the distance that the tracks
remained on the road; (iii) the behaviour of the animal
(whether it walked along the road or simply walked
across the road); (iv) whether the prints were made
by a small or large animal; (v) whether one or more
animals had made the tracks; and (vi) the habitat in
which the tracks were observed. The road was prepared
for the next day’s count by towing a 1.2-m length of
steel as before.
Habitat assessment
The proportion of the study site occupied by each habitat was determined by assessing the relative incidence
or ‘cover’ of each habitat along each of the two transects. These were formally surveyed in July 1997 using
a modification of the line intercept technique (MuellerDombois & Ellenberg 1974). Habitat observations
were made by an observer in the passenger seat of a
four-wheel-drive vehicle driven at 20 km h-1 with the
vehicle odometer being used to record habitat intercept
distances to the nearest 100 m.
Data analysis
A problem with track-based surveys of population
abundance for territorial species such as feral cats
(Corbett 1979) and dingoes (Thomson 1992) is that
the total number of tracks of a particular species
encountered is likely to be a function of activity levels
of individuals in the population as well as population
density (Edwards et al. 2000). To overcome this
problem, for feral cats we estimated density from
the minimum number of animals responsible for
observed tracks. Minimum numbers were estimated
according to the following criteria. Feral cat tracks
separated by 500 m were deemed to have been made
by different animals. Multiple sets of tracks less than
500 m apart were attributed to one individual. The
28
G. P. EDWARDS ET AL.
500 m classification criterion was based on knowledge
of the size of the daily home range of feral cats in the
study area (core area of approximately 20 ha: Edwards
et al. 2001). A distance of 500 m was also used by
Mahon et al. (1998) to classify individual feral cat tracks
in a similar study.
Dingoes are much more mobile than feral cats; individual dingoes have been shown to travel up to 20 km
in just 7.5 h (Corbett 1995). Also, dingoes in arid areas
have much larger long-term home ranges than feral
cats (dingoes 37–160 km2: Thomson 1992; feral cats
1.7–22 km2: Jones & Coman 1982; Edwards et al.
2001). As the spatial scale of the habitat patches (see
Results section) was small in relation to dingo movements, it was neither practical nor appropriate to assess
dingo density in each habitat in the manner described
previously for feral cats. Rather, to gauge the relative
use of habitats by dingoes we viewed our 25-km transects as being 25 contiguous 1-km cells and tallied the
number of cells in each habitat that registered dingo
tracks. In cases where a cell containing both habitats
registered dingo tracks, both habitats were deemed to
have registered tracks. We summed our estimates of
the number of ‘individual’ cats and cells with dingo
tracks observed across days and transects for each
habitat within survey periods. We then divided the
total number of cats and cells with dingo tracks for
each habitat by the relative area covered by the respective habitats within the study site to give an estimate
of the relative ‘density’ of each species for each habitat. For dingoes in particular, this measure is similar
to the ‘activity density’ measure used by Abramsky and
Pinshow (1989) and Abramsky et al. (1991, 1997),
which is also based on tracks.
The dispersion of feral cats and dingoes in relation
to habitat was analysed with 2 goodness of fit. A
heterogeneity 2 analysis was performed initially on
uncorrected data to test whether pooling of data across
sampling periods was justified (Zar 1996). Pooling can
improve statistical power (Zar 1996). Where pooling
was justified, the Yates correction for continuity was
used to overcome the discrete distribution problem
with 2 values based on single degrees of freedom
(Zar 1996). Isodars were constructed to determine
whether the patterns of habitat use by feral cats and
dingoes fitted any of the currently proposed models of
density-dependent habitat selection: continuous input
(ideal free), consumer-resource (ideal free) and preemptive (Morris 1994). Isodars were constructed
using central trend lines based on geometric mean
regression following Krebs (1999), as both the x and
y variables had measurement error. In cases where
isodars were linear on an arithmetic scale and had zero
intercepts, the data were logarithmically transformed
to test whether logarithmic solutions were better than
arithmetic ones (Morris 1994). Pearson correlation
analysis was used to examine whether the relative
density of feral cats in a particular habitat was related
to the density of dingoes in that habitat or with dingo
abundance across the study site (habitats pooled). The
total number of cells with dingo tracks was used as the
latter measure in the correlations. Correlation analyses
were performed using SYSTAT 8.02 (SYSTAT Inc.,
Evanston, Illinois, USA).
RESULTS
Fig. 1. Arithmetic relationship between the population
density of feral cats in open and mulga habitats in central
Australia. Relative density is the total number of cats for each
habitat divided by the relative area covered by each habitat
within the study site. The fitted isodar based on geometric
mean regression is shown. y = 1.033x + 15.981.
The cover data pooled between the transects showed
that the study site had 36% mulga habitat and 64%
open habitat. The two habitats tended not to be interspersed in a fine-scale matrix but rather occurred as
large patches greater than 10 km2 in size. Aerial photographs of the study site supported this interpretation.
The habitat proportions did not change during the
study, thereby satisfying the independence from landscape assumption (Morris 1994). This precluded
having to use weighted means of habitat supply in
the isodar analyses.
Heterogeneity 2 analysis of the cat and dingo track
data indicated that pooling of data between sample
periods was justified (cats: 28 = 8.7, 0.25 < P < 0.50;
dingoes: 28 = 7.8, 0.25 < P < 0.50). The pooled tests
with Yates correction showed that dingoes were
distributed uniformly across habitats (21 = 0.3,
0.50 < P < 0.75) but feral cats were not (21 = 20.7,
P < 0.001). Relative density of feral cats was higher
in mulga than in open habitat.
For feral cats, the arithmetic isodar solution was
HABITAT SELECTION IN FERAL CATS AND DINGOES
linear (Fig. 1) and is described by the geometric mean
regression equation:
(relative density mulga) = 1.033 (relative density
open) + 15.981; r = 0.70; P = 0.035.
The slope was not significantly different from one
(95% confidence interval = 0.375–1.690) but the intercept was significantly greater than zero (95% confidence interval = 0.204–31.759).
The arithmetic dingo isodar regression (Fig. 2) was
statistically non-significant (P = 0.34).
Although the relative density of feral cats in mulga
appeared to be negatively correlated with dingo activity in mulga (r = –0.47), the trend was not significant
(Bartlett 21 = 1.589, P = 0.21). Although there
appeared to be a weak positive correlation between the
relative density of feral cats in open habitat and dingo
activity in the open (r = 0.27), this trend also lacked
significance (Bartlett 21 = 0.50, P = 0.48). Although
the relative density of feral cats in mulga appeared to
be negatively correlated with dingo activity across the
study site (habitats pooled; r = – 0.41), the trend was
not significant (Bartlett 21 = 1.17, P = 0.28). Although
there appeared to be a weak positive correlation
between the relative density of feral cats in open habitat and dingo activity across the study site (r = 0.41),
this trend also lacked significance (Bartlett 21 = 1.17,
P = 0.28).
DISCUSSION
Chi-squared analyses indicated that feral cats were consistently more abundant in mulga than in open habitat at our study site. The isodar solution supported this
result. This is within expectations because preferential
use of cover has been widely reported in the Felidae.
Scottish wildcats (Felis silvestris) were found to favour
dense pine forests and scrub (Corbett 1979), F. catus
on the Galápagos Islands were found more often in
lava/shrub habitat than in open habitats (Konecny
1987), F. catus in New Zealand were attracted to vegetation buffers surrounding open habitat (Alterio et al.
1998), pumas (Puma concolor) were found to use mixed
swamp and hammock forests preferentially, which
provided more cover than other habitats (Belden et al.
1988), and bobcats (F. rufus) were found to use habitats with dense understorey more than open habitats
(Progulske 1982, cited in Wassmer et al. 1988).
The linear signature of the feral cat isodar on the
arithmetic scale implied density-dependent population
regulation (Morris 1988) consistent with the consumerresource model of habitat selection (Morris 1988,
1994). This is an ideal free solution where individuals
are incapable of consuming resources as quickly as they
are renewed (Morris 1994). The non-zero intercept
implied that the two habitats differed quantitatively
29
rather than qualitatively (Morris 1988). Qualitative
differences between habitats may include differences
in structure or in the kinds of resources or kinds of
interacting species that they contain (Morris 1988).
Quantitative differences are differences in the availability or richness of resources such as food and shelter
(Morris 1988).
It is intuitively appealing that habitat selection by
feral cats appears consistent with the consumerresource model. At our study site, resources are
unlikely to be renewed at a constant rate given the
vagaries of arid ecosystems (Noy-Meir 1973; Caughley
1987) nor consumed by feral cats as soon as they are
renewed given that: (i) feral cats occur at relatively low
densities at the study site (Edwards et al. 2000, 2001);
and (ii) the day-to-day movements of feral cats occur
over a relatively large spatial scale (Edwards et al.
2001). The feral cat isodar indicates that there are
initial differences in reproductive rewards between the
mulga and open habitats, the former having the higher
reward through the eyes of the feral cats. At very low
densities, only mulga should be occupied by feral cats
until increased density reduces the fitness rewards in
mulga to those initially available in the open (Morris
1988). Thereafter, as density increases, an equal
number of individuals is added to each of the habitats
(= parallel regulation: Morris 1988). However, the
difference in density between the habitats will fall as
the overall population density increases. With parallel
regulation, density-dependent feedback on reproductive fitness is linear and alike in both habitats
(Morris 1988). Parallel regulation is consistent with the
notion that dispersal from high density (source) to low
Fig. 2. Arithmetic relationship between the activity of
dingoes in open and mulga habitats in central Australia.
Relative activity is the total number of 1-km cells with dingo
tracks for each habitat divided by the relative area covered
by each habitat within the study site.
30
G. P. EDWARDS ET AL.
density (sink) habitats is an important population regulation mechanism (Morris 1988 and others therein).
It is interesting to consider why feral cats at our study
site were more abundant in mulga habitat than open
habitat. In central Australia, small mammals are the
most important prey of feral cats, but birds, reptiles
and invertebrates are also eaten (Paltridge et al. 1997).
Three-monthly monitoring of feral cat prey species at
our study site indicated that small mammal species’
composition and abundance were often similar across
habitats. However, there were marked differences in
the bird and reptile faunas in each habitat and both
birds and reptiles were more abundant in the open
habitat (G. P. Edwards & N. de Preu, unpubl. data,
1994–1997). This implies that the open habitat was in
fact richer in food resources than was mulga. However,
some prey are more susceptible to cat predation than
others (Paltridge et al. 1997) and the apparent irregularity between total prey abundance and habitat use by
feral cats may be due to a lack of correlation between
total prey abundance and available prey. An additional
point to consider in this regard is that the stalk and
ambush hunting tactics typically employed by felids are
well suited to dense cover (Murray et al. 1995), which
is typical of mulga habitat. This may confer greater
foraging success in mulga than in the open, thereby
making mulga the more productive habitat.
Although habitat selection in feral cats was not
related to changes in dingo activity, this does not mean
that predation risk was not important. For feral cats,
the risk of predation by dingoes may have been similar
between habitats or the mere presence of just one
dingo may have been enough to influence feral cats to
choose the habitat richer in protective cover (i.e. mulga
habitat).
The results of this study suggest that in conservation
programs where an objective is to mitigate predation
by feral cats, habitats affording cover are prime locations for the placement of poison baits and/or traps
targeting the species.
The 2 analyses indicated that dingoes showed a
preference for neither mulga nor open habitat. The
statistical non-significance of the dingo isodar regression suggested that habitat selection by dingoes was
density independent (Morris 1988), at least at the scale
considered. At a dissimilar site in north-western
Australia, the use of habitats by radio-collared dingoes
was disproportionate to availability: riverine areas were
used much more intensively than floodplains, hills or
stony habitats (Thomson 1992). The observed pattern
of habitat use in the north-western Australian study is
not necessarily incongruous with the findings of the
present investigation. Isodar analysis was not performed
in the north-western Australian study and distribution
data alone tell us little about mechanisms underlying
habitat use (Morris 1994). If the riverine habitat were
qualitatively and/or quantitatively superior to the other
habitats, dingoes should have favoured the riverine
habitat irrespective of whether density-dependent or
density-independent factors were at play.
ACKNOWLEDGEMENTS
This study was jointly funded by the Parks and
Wildlife Commission of the Northern Territory and
Environment Australia (formerly the Australian Nature
Conservation Agency). The study was approved by the
Alice Springs Animal Ethics Committee. We thank
the many volunteers who assisted with field work,
particularly Jok Markham. Ken Johnson, Dave Lawson,
Bill Freeland, Catherine Meathrel and Doug Morris in
particular provided useful comments on drafts of this
paper. We also thank the owners and managers for
allowing us to work on Hamilton Downs Station.
REFERENCES
Abramsky Z. & Pinshow B. (1989) Changes in foraging effort in
two gerbil species correlate with habitat type and intra- and
interspecific activity. Oikos 56, 43–53.
Abramsky Z., Rosenzweig M. L. & Pinshow B. (1991) The shape
of a gerbil isocline measured using principles of optimal
habitat selection. Ecology 72, 329–40.
Abramsky Z., Rosenzweig M. L. & Subach A. (1997) Gerbils
under threat of owl predation: Isoclines and isodars. Oikos
78, 81–90.
Alterio N., Moller H. & Ratz H. (1998) Movements and habitat
use of feral house cats Felis catus, stoats Mustela erminea and
ferrets Mustelo furo, in grassland surrounding yellow-eyed
penguin Megadyptes antipodes breeding areas in spring. Biol.
Conserv. 83, 187–94.
Belden R. C., Frankenberger W. B. & Schwikert S. T. (1988)
Panther habitat use in Southern Florida. J. Wildl. Manag.
52, 660–3.
Brown J. S. (1988) Patch use as an indicator of habitat preference,
predation risk, and competition. Behav. Ecol. Sociobiol. 22,
37–47.
Carey H. V. (1985) The use of foraging areas by yellow-bellied
marmots. Oikos 44, 273–9.
Caughley G. (1987) Ecological relationships. In: Kangaroos: Their
Ecology and Management in the Sheep Rangelands of Australia
(eds G. Caughley, N. Shepherd & J. Short) pp. 159–88.
Cambridge University Press, Cambridge.
Corbett L. K. (1979) Feeding ecology and social organisation
of wildcats (Felis silvestris) and domestic cats (Felis catus) in
Scotland. PhD Dissertation, University of Aberdeen,
Scotland.
Corbett L. (1995) The Dingo in Australia and Asia. University of
New South Wales Press, Sydney.
Edwards G. P., de Preu N. D., Shakeshaft B. J. & Crealy I. V.
(2000) An evaluation of two methods of assessing feral cat
and dingo abundance in central Australia. Wildl. Res. 27,
143–9.
Edwards G. P., de Preu N., Shakeshaft B. J., Crealy I. V. &
Paltridge R. M. (2001) Home range and movements of male
feral cats in a semi arid woodland environment in central
Australia. Austral Ecol. 26, 93–101.
HABITAT SELECTION IN FERAL CATS AND DINGOES
Fretwell S. D. & Lucas H. L. Jr (1970) On territorial behaviour
and other factors influencing habitat distribution in birds. I.
Theoretical development. Acta Bioth. 19, 16–36.
Gilliam J. F. & Fraser D. F. (1987) Habitat selection under predation hazard: A test of a model with foraging minnows.
Ecology 68, 1856–62.
Halama K. J. & Dueser R. D. (1994) Of mice and habitats:
Tests for density-dependent habitat selection. Oikos 69,
107–14.
Hughes J. J., Ward D. & Perrin M. R. (1994) Predation risk and
competition affect habitat selection and activity of Namib
Desert gerbils. Ecology 75, 1397–405.
Jones E. & Coman B. J. (1982) Ecology of the feral cat, Felis catus
(L.), in south-eastern Australia III. Home ranges and population ecology in semiarid north-west Victoria. Aust. Wildl.
Res. 9, 409–20.
Konecny M. J. (1987) Home range and activity patterns of
feral house cats in the Galápagos Islands. Oikos 50,
17–23.
Kotler B. K. (1984) Risk of predation and the structure of desert
rodent communities. Ecology 65, 689–701.
Kotler B. P., Brown J. S. & Hasson O. (1991) Factors affecting
gerbil foraging behavior and rates of owl predation. Ecology
72, 2249–60.
Krebs C. J. (1999) Ecological Methodology. Addison-Welsey
Educational Publishers, Menlo Park.
Lima S. L. (1985) Maximizing feeding efficiency and minimizing
time exposed to predators: A trade-off in the black-capped
chickadee. Oecologia 66, 60–7.
Lima S. L. & Dill L. M. (1990) Behavioural decisions made under
the risk of predation: Review and prospectus. Can. J. Zool.
68, 619–40.
Lima S. L., Valone T. J. & Caraco T. (1985) Foraging-efficiencypredation-risk trade-off in the grey squirrel. Anim. Behav. 33,
155–65.
Mahon P. S., Banks P. B. & Dickman C. R. (1998) Population
indices for feral carnivores: A critical study in sand-dune
habitat, south-western Queensland. Wildl. Res. 25, 11–22.
Mayr E. (1926) Die Ausbreitung des Girlitz (Serinus canaria
serinus L.). J. fur Ornit. 74, 571–671.
Messier F., Virgl J. A. & Marinelli L. (1990) Density-dependent
habitat selection in muskrats: A test of the ideal free distribution model. Oecologia 84, 380–5.
31
Morisita M. (1952) Habitat preference and evaluation of
environment of an animal: Experimental studies on the
population density of ant lion, Glenuriodes japonicus M’L. I
(in Japanese, with English summary). Physiol. Ecol. 5,
1–16.
Morris D. W. (1987) Spatial scale and the cost of densitydependent habitat selection. Evol. Ecol. 1, 379–88.
Morris D. W. (1988) Habitat-dependent population regulation
and community structure. Evol. Ecol. 2, 253–69.
Morris D. W. (1989) Density-dependent habitat selection:
Testing the theory with fitness data. Evol. Ecol. 3, 80–94.
Morris D. W. (1992) Scales and costs of habitat selection in
heterogeneous landscapes. Evol. Ecol. 6, 412–32.
Morris D. W. (1994) Habitat matching: Alternatives and implications to populations and communities. Evol. Ecol. 8,
387–406.
Mueller-Dombois D. & Ellenberg H. (1974) Aims and Methods
of Vegetation Ecology. Wiley, New York.
Murray D. L., Boutin S., O’Donoghue M. & Nams V. O. (1995)
Hunting behaviour of a sympatric felid and canid in relation
to vegetative cover. Anim. Behav. 50, 1203–10.
Noy-Meir I. (1973) Desert ecosystems: Environment and
producers. Ann. Rev. Ecol. Syst. 4, 25–51.
Paltridge R., Gibson D. & Edwards G. (1997) Diet of the
feral cat (Felis catus) in central Australia. Wildl. Res. 24,
67–76.
Perry R. A., Mabbutt J. A., Litchfield W. H. & Quinlan T. (1962)
Part II. Land systems of the Alice Springs area. In: Lands
of the Alice Springs Area, Northern Territory, 1956–57 (ed.
R. A. Perry) pp. 20–128. CSIRO, Melbourne.
Rosenzweig M. L. (1991) Habitat selection and population
interactions: The search for mechanism. Am. Nat. 137,
S5–28.
Svärdson G. (1949) Competition and habitat selection in birds.
Oikos 1, 157–74.
Thomson P. C. (1992) The behavioural ecology of dingoes in
north-western Australia. IV. Social and spatial organisation,
and movements. Wildl. Res. 19, 543–63.
Wassmer D. A., Guenther D. D. & Layne J. N. (1988) Ecology
of the bobcat in south-central Florida. Bull. Florida State
Mus. Biol. Sci. 33, 159–228.
Zar J. H. (1996) Biostatistical Analysis, 3rd edn. Prentice Hall,
Upper Saddle River.