An accurate estimate of population density is essential for the management and protection of endangered species, such as the cheetah Acinonyx jubatus, which is listed as vulnerable by IUCN (2007). As in most large carnivores, the cheetah population is declining (Marker et al., 2003) and the ability to develop an accurate, repeatable and cost-effective method to assess population trends and density is required. Direct methods such as visual counts and physical mark and recapture are often the method of choice; however, they rely on the visual recognition of individuals and in large carnivores such methods are often expensive, difficult and time consuming (Stander, 1998; Gusset & Burgener, 2005). When direct methods are too expensive or impractical, as is often the case in low-density species such as the cheetah (Mills, 1997; Wilson & Delahay, 2001), the use of indirect measures
that rely on the presence and detection of field signs, as an index of true density, is a more favourable option. For example, it is possible to obtain relative estimates of carnivore populations by calculating the number of scat samples, den sites or spoor seen in the study area (Mills, 1997). Spoor surveys, in particular, have been used extensively as a monitoring tool in several species, including leopard Panthera pardus, lion Panthera leo, brown hyena Hyaena brunnea (Stander, 1998; Funston et al., 2001), caracal Caracal caracal (Melville & Bothma, 2006) and mountain lion Felis concolor (Smallwood & Fitzhugh, 1995). They are
less invasive and more cost-effective than direct methods (Jewell, Alibhail & Law, 2001), while remaining repeatable, objective, valid and accurate (Stander et al., 1997; Gusset & Burgener, 2005). However, they only provide a relative estimate of population size and a quantifying technique must be applied to calculate the population density. One technique is to use spoor measurements to identify and count individuals within a population. This has been applied to mountain lion (Lewinson, Fitzhugh & Galentine, 2001), tiger Panthera tigris (Sharma, Jhala & Sawarkar, 2005) and black rhino Diceros bicornis (Jewell et al., 2001). However, it
requires further study before it may be applicable for use in the field with varying substrates (Lewinson et al., 2001).
Alternatively, by double sampling the population by a direct technique such as the physical capture and marking of individuals and an indirect technique such as spoor tracking, the relationship between them may be quantified and a correction factor to calibrate the indirect method may be calculated (Eberhardt & Simmons, 1987; Wilson & Delahay, 2001). Stander (1998) showed a significant linear relationship between spoor counts and true density determined by the recognition of marked or collared lion, leopard and wild dog Lycaon pictus. Gusset & Burgener (2005) used Stander’s regression equation to estimate the leopard population in the Waterberg region of South Africa and showed the result to be similar to that derived from the identification of individuals from spoor measurements. Funston et al. (2001) also found a similar regression equation in lions and extrapolated the slope of the line to brown hyena, spotted hyena Crocuta crocuta, cheetah and leopard.
Stander (1998) found the linear relationship between spoor counts and true density to be species specific and predicted that the slope of the line would vary with habitat use and species behaviour. It is therefore reasonable to assume that differences in lion, leopard and wild dog home range, daily movements and road usage, compared with cheetah, will cause differences in this relationship. However, this correlation has not yet been quantified. This paper’s aim is to compare cheetah spoor counts with capture and radio collaring information in an open and free-ranging population of cheetah in Southern Botswana. Thus, the objectives of this study were to calibrate the spoor survey technique to calculate a correction factor for use in future cheetah spoor
surveys and to subsequently test the correction factor in a spoor survey in the following wet season.