Estimation of tiger densities in India using photographic captures and recaptures

Ecology, Dec, 1998 by K. Ullas Karanth, James D. Nichols

Densities and sizes of wild tiger populations

The estimates of tiger densities we report (Table 4) are the first such data to be generated using a theoretically well-founded population sampling approach. Our results show that tiger densities did not differ significantly among Kaziranga, Kanha, and Nagarahole areas, but were significantly lower at Pench than in these other areas. The estimated mean tiger densities (Table 4) appear to correspond positively with the mean densities of principal prey (Table 5) at the first three sites, but not at Pench. We suspect that tiger densities at Pench were lower than at limits set by prey abundance, as a result of high incidence of tiger poaching in the region (Wildlife Protection Society of India 1996, unpublished report; K. U. Karanth, personal observations).

In comparing our tiger density estimates to tiger density estimates from other studies, we were constrained by the weak methodologies that underlie most of them (Karanth 1987, 1988). Therefore, for comparisons, we selected three studies that depended on either visual identification or radiotagging of the study animals (Schaller 1967, 1972, Sunquist 1981; E. N. Smirnov and D. G. Miquelle, unpublished manuscript), Because even these studies did not use formal population sampling approaches, they lack comparable estimates of variances for making statistical inferences. However, in view of the scarcity of comparative field data on tigers, in particular, and on large solitary felids in general, we believe it is fruitful to use mean values of the estimates to speculate on the ecological factors underlying the observed patterns of tiger densities across the range of the species.

Our density estimates (Table 4) exclude cubs [less than] 1 yr old. Earlier workers have reported density estimates that include cubs. Therefore, for comparisons, we recalculated tiger density estimates for our sites to include cubs. From direct sightings at Pench, we knew that there were 7 cubs, adding up to a total population size of 12, yielding a mean density of 9.9 tigers/100 [km.sup.2]. However, the rather high percentage of cubs in the Pench population was probably a result of a depressed proportion of juveniles and transients caused by poaching on the park's boundaries (K. U. Karanth, personal observations). Demographic models of typical wild tiger populations (Karanth and Stith, 1998) suggest that cubs may form [approximately]25% of a "normal" tiger population. Because we did not have exact knowledge of the number of cubs at Nagarahole, Kanha, and Kaziranga, we relied on this percentage to estimate total tiger densities from the data in Table 4. The resulting density estimates for Nagarahole, Kanha, and Kaziranga were, respectively, 15.33, 15.60, and 22.40 tigers/100 [km.sup.2]. based on Schaller's (1967) data on visually identified adult tigers and their cubs in a 318-[km.sup.2] area in Kanha during 1964-1965, tiger density was 6.9 tigers/100 [km.sup.2]. Our calculations based on data on home range and social organization from Chitwan, reported by Sunquist (1981) during field work in 1974-1980, yield tiger density estimates of 5.8-8.7 tigers/100 [km.sup.2]. based on snow tracking and radiotelementry (E. N. Smirnov and D. G. Miquelle, unpublished manuscript), tiger densities reported from the temperate forests of the Russian Far East appear to be considerably lower (0.6-1.4 tigers/100 [km.sup.2]). The mean tiger densities estimated by us at all four sites are higher than the densities reported by other researchers.