Using Multiple Sources of Information to Estimate Tiger Densities

In landscapes where wildlife occurs in low densities, gathering information from a single data source often does not permit accurate estimation of population densities and abundance. In such cases, using multiple data sources may allow us to overcome ecological and logistical constraints to estimate densities of elusive carnivores such as tigers. In particular, innovative spatially explicit capture-recapture modeling approaches integrate information from photographic capture-recapture and genetic data to derive more robust estimates of tiger densities in India.

Authors A. M. Gopalaswamy, K. U. Karanth, J. A. Royle, M. Delampady, J. D. Nichols and D. W. Macdonald, demonstrate approaches to reliably estimate tiger abundance in Bandipur Tiger Reserve, Karnataka. These are the highlights of their study published in Ecology in 2012.

Study site

The study was conducted to estimate abundance and density of tigers (Panthera tigris) in the 871 sq km Bandipur National Park and Tiger Reserve located in Karnataka, South India. This dry deciduous reserve receiving an annual rainfall of 625-1250 mm with varying temperatures from 18-29°C supports high ungulate densities (35.2 animals/ sq km) including sambar, chital, four-horned antelope, wild pig, gaur and elephant.

Methods

Data were collected using two intensive field survey methods. The photographic survey method using camera traps collected data from November 2005 to January 2006 and the fecal samples of tigers containing DNA were also collected from the same site. The photographic survey involved setting up 118 camera trap locations across key trails. Covering a walk effort of 237 km the fecal DNA survey contained 18 routes distributed along camera trap trails. The scats were collected over six successive days and repeated for six weeks. Models were constructed by combining data to identify ecological and sampling processes that produce the most reliable estimate of tiger densities.

The Results

1. The photographic survey resulted in 35 captures of 29 individual tigers.
2. The DNA survey resulted in 63 samples from 24 individually identifiable tigers.
3. Three models were applied to analyze data from the two surveys. In Model 1, the population size was assumed to be constant in contrast to Model 2 where the detection parameter was assumed to be survey specific and for Model 3 all parameters were assumed to be survey-specific.
4. The combined information from two surveys Models 1 and 2 produced a different and more precise estimate of tiger densities (8.7 and 8.5 tigers /100 km2) compared to Model 3 (12.8 and 6.65 tigers/100 km2).
5. Direct and step-wise approaches produced similar estimates of tiger densities: 8.7 and 8.25 tigers/100km2 from Model 1 and 8.5 and 8.69 tigers/100 km2 from Model 2.

Conservation Implications

Data on tigers from the photographic capture-recapture and the fecal DNA methods were compared through several proposed models resulting in multiple estimates of tiger densities for Bandipura National Park. This study suggests that using multiple sources of data is important & provides better results than a single source when studying rare or elusive species that occur at low densities such as tigers. The authors recommend that future research efforts focus on developing spatially explicit models that can combine information from multiple data sources to improve existing population density estimates for many wildlife species.