One of the major problems facing tiger conservation
today is our inability to accurately estimate the
population in a given tract. Population estimation
based on pugmarks, the traditional method (Choudhury,
1970; Panwar 1979), has been questioned (Karanth
1995) as most of the parameters used for population
estimation have not been statistically validated.
Gore et al. (1993) concluded that sex can be
identified, but the technique for individual
identification needs to be refined.
It is believed that individual tigers can be
identified based on facial markings and stripe
patterns (Champion 1927, Schaller 1967, McDougal
1977). Based on this assumption, Karanth (1995) had
applied capture-recapture technique for tiger
population estimation. He used self-activated
cameras to identify individual tigers based on
stripe pattern. Data on individuals identified
during a time period were then used to estimate the
population.
Since December 1994, in Dholkhand, the mini-core
area in Rajaji National Park in north-western India,
we have been using self-activated camera units to
photograph tigers. Till November 1995, tigers were
photographed six times over 85 camera trap nights.
Three new facets about the use of camera trap
technique arose when we asked our colleagues at the
Institute to identify individual tigers based on
these photographs. In one case, the face of a tiger
was photographed twice with a time difference of
nine seconds. In the first picture, one of the face
stripes is connected with the eye and in the second,
due to a slight change in posture, the stripe looks
as if separated from the eye. This had made 100% of
our colleagues (N=20) identify the photographs as
belonging to two different tigers.
Another time, two photographs of the lateral side of
a tiger were taken one after the other. All (N=l3)
identified them as two different individuals. The
track data, however, had shown that only one tiger
had walked in front of the camera. When we examined
the face pictures of these two tiger photographs, 10
major stripes were seen in both the pictures. But in
the second picture a loop had been formed due to a
change in posture, Seventy-eight per cent of our
colleagues (N=9) identified these face pictures as
belonging to two animals.
When we critically looked at the reason for this
confusion we discovered that 24 stripes above the
belly and shoulder were identical in both the
photographs. There were considerable variations,
however, between the two photographs when the stripe
pattern on the flank above the elbow joint and hind
quarters of the body were compared. We observed
similar variations in the stripe pattern of a tiger
photographed in Delhi Zoo. The left side of the
tiger was photographed three times and the right
side four times. When photographs of a particular
side were compared to one another, all showed
individual variations. These variations are
attributed to the loose nature of the skin on the
upper parts of the body. As a result, slight changes
in the body posture bring about variations in the
way stripes appear from one photograph to the
other.
1.Movements are likely to cause least variations in
the stripes on the face as the skin is tight. But we
have to ascertain, based on photographs of captive
animals, whether there are enough variations in face
stripes to identify individuals.
2.Shift in posture causes considerable variations in
the stripe patterns on the hind quarters and flank
above the elbow joint and this is attributed to the
loose nature of the skin on these parts of the
body.
3.Stripes above the belly and shoulder show least
variations due to movements. This, however, needs to
be ascertained with a large sample from captive
tigers.
We conclude that identifying individual tigers based
on stripe patterns may not be as easy as it appears
to be. Variations could appear in photographs of the
same animal due to slight changes in posture. These
findings need to be taken into account in any future
programme to estimate the population of tigers based
on camera traps.