Why the ‘Pugmark Census’ Used to Monitor Tiger Populations Failed

In 2003, some highly respected conservationists got together to write a paper for an international journal, in which they laid bare the loopholes in the pugmark method of counting tigers in India. Shortly thereafter, in an extraordinary sequence of events, newer and more advanced methods were used to assess the tiger population, which resulted in the shocking denouement that there were only about 1411 tigers in India. Read on to discover how the scientists rated the old ‘pugmark census’.

The ‘pugmark census’ was invented in 1966 by Indian forester SR Choudhury. In this method, during a 1–2-week period, thousands of personnel would simultaneously fan out across India to search for tiger tracks. They were expected to locate tiger tracks and obtain plaster casts or tracings of the left hind pugmark. The pugmarks collected would be later compared to identify individual tigers relying on perceived differences in shape and other measurements. These ‘individual tiger identifications’ would be then refined through cross-comparisons among census blocks, reserves and larger regions to obtain ‘reliable estimates’ of wild tiger numbers in India.

The census-takers were expected to locate the tracks and obtain plaster casts or tracings of the left hind paws of nearly all the tigers in the entire country. The pugmark method was questioned by scientists, who pointed out that the following assumptions must hold true for the pugmark method to be statistically valid:

1. The entire potential tiger habitat in India had to be effectively covered during the pugmark census.
2. All the four paw prints of every individual tiger in the surveyed area had to be detected during the censuses.
3. The same hind pugmark of each one of these individual tigers must be lifted from suitable and comparable substrates or from standardized soil track-plots.
4. The shape of each pugmark lifted had to be recorded without distortion by the thousands of census personnel involved in the operation.
5. Supervisory officials were expected to be subsequently able to segregate the pugmarks of each individual tiger correctly, based on footprint shape, track measurements, and prior local knowledge.

Failure of assumptions 1 and 2 would lead to under-counts; failure of assumptions 3 and 4 would lead to overcounts; and failure of assumption 5 could lead to either undercounts or overcounts.

Unsurprisingly, the ‘pugmark census’ could not live up to this scrutiny:

1. Surveying the entire area: In reality, only an unknown fraction of the 3,00,000 km2 that is considered habitable by tigers, is searched intensively during censuses. For example, logistical constraints (e.g. in Namdapha, Sundarban), security concerns (e.g. in Nagarjuna Sagar, Indravati and Manas reserves) or staff shortages (almost everywhere) restrict the proportion of the area covered by field teams.
2. Locating the tracks of every individual tiger: The probabilities of finding the tracks of each individual tiger in the surveyed area are low, except in a few reserves with high road density and suitable soil type.
3. Selecting the appropriate pugmark: Lifting pugmarks from firm soil overlaid with dust or sand is an essential precondition to obtaining accurate pugmarks. Unless clear impressions of all the four paws on the right soil are detected for each individual tiger, it is impossible to pick the same hind pugmark of each individual for comparisons, as prescribed by the pugmark method. In reality, census personnel often do not find clear prints of all four paws, and consequently lift prints of the different paws of the same animal from different localities.
4. Recording and recognizing tiger tracks: The shape of the same tiger footprint traced by different persons may vary considerably. This variation can be reduced by using well-trained persons in controlled trials, but not in field censuses involving thousands of personnel with varying levels of skills.

Three decades of tiger monitoring following this unrealistic method essentially failed in India—despite being backed by massive investments and the best of intentions. This failure has, inevitably, led to poor conservation practices. For example, field managers initially reported an increase in tiger population in 1994, despite mounting evidence of deteriorating reserve protection and increasing poaching pressure.

Possible alternative approaches to monitoring tigers at different levels of resolution

In areas that are fragmented or degraded: It is critical to document annually the presence or absence of tigers in such areas. The present large-scale, labour-intensive pugmark censuses can be modified into surveys of tiger-sign — this involves merely recording the presence of tiger tracks and other signs under a rigorous sampling design. Such surveys will not involve the impossible task of individually trying to identify all wild tigers from track prints under field conditions. Positive results from such surveys could enable restoration and reconnection of fragmented habitats, establishing new or larger populations.

In individual reserves: Managers need to keep track of tiger populations to evaluate the effectiveness of conservation schemes. Even under resource constraints, densities of tigers can be derived annually using encounter rates of tiger-sign (e.g. number of tiger track sets or scats encountered per 10 km walked, the proportion of 1 km trail segments in which tiger tracks were detected, number of fresh tiger tracks crossing the path of a boat or vehicle). This will, however, not translate into accurate predictions of tiger populations.

At priority sites: In some areas, absolute densities of prey and tigers may need to be estimated. In such cases, there is no escape from employing advanced equipment and techniques, and skilled personnel. Such techniques will include line transect surveys of prey densities and camera-trap capture–recapture surveys of tigers.

The original article first published as “A Review of Tiger Population Monitoring Issues” by K. Ullas Karanth, James D. Nichols, John Seidensticker, Eric Dinerstein, James L. David Smith, Charles McDougal, A. J. T. Johnsingh, Raghunandan S. Chundawat and Valmik Thapar. Animal Conservation (2003) 6, 14