This is the first in a series of articles on big cat conservation by Ullas Karanth.
Pollsters say tigers are the most popular animal species on this planet. Unfortunately, they are also among the most threatened. Direct killing and overhunting of their prey by humans, as well as habitat reduction by 93% in just two centuries, have all pushed them to the brink. Global efforts are now underway to reverse their decline.
Meanwhile, wildlife biologists like me struggle to monitor the fate of surviving tiger populations. To really know if recovery efforts are working or not, we must count wild tigers. This is not easy: tigers are scarce, secretive and shun human company. History of of tiger counting is littered with failures.
Way back in the 1930s, zoologists tried counting wild tigers across the vast Russian Far East, a region of over 200,000 square kilometers. Doggedly following fresh tiger tracks in the snow, they tried to distinguish individual cats based on distance between track sets, hoping to get a total count. Although followed to this day, this method has poor statistical basis.
For example, what proportion of tigers are not counted or how many tracks are double-counted, remain unknown. A recent modification, bearing the long-winded name of Formozov–Malyshev–Pereleshin formula, tries to impose statistical discipline. It also, however, involves untestable assumptions. Curiously, around the same time a colonial forester in distant India, J.W. Nicholson, also tried to count tigers by segregating tracks located at waterholes that he thought were sufficiently apart. He also had to make untestable assumptions about how far tigers moved and how often they drank water.
In the early 1970’s when India launched its ambitious Project Tiger, a senior forestry official—blissfully untrained in either ecology or statistics—proposed an entirely new tiger counting method. He assumed—quite wrongly it turned out—that the paw-print of each individual tiger was uniquely identifiable. Thousands of his Forest Guards fanned out across India—pencil, paper and glass panes in hand—to trace tiger ‘pugmarks’. They claimed to generate total counts of all tigers in India.
As in Russia, this homegrown and naïve method thrived for decades in an unquestioning bureaucratic culture. In 2005, however, already weakened by tough scientific criticism, the ‘pugmark census’ finally met an inglorious end. It had, quite embarrassingly, failed to detect the poaching of an entire tiger population in Sariska Reserve right next to the rulers in Delhi! A new tiger counting method was desperately sought.
As far back as in 1920s, Indian forester Fred Champion experimented with a cumbersome camera trap developed by William Nesbitt of New York. Wild animals that walked past tripped it. Illuminated by a magnesium powder flash, Champion’s trap was cumbersome and unreliable: for a lifetime of effort, he got just nine decent tiger photos. Inspired by Champion, fifty years later, American tiger enthusiast Chuck McDougal used a pressure-pad activated camera trap to obtain some fine tiger portraits in Nepal’s grasslands.
Beginning my tiger studies in the 1980s in the lush Nagarahole forests of India, I focused on understanding tiger population biology rather than getting pretty pictures. I recognized that while the fuzzy paw prints of tigers were unreliable, stripe patterns on their bodies were demonstrably distinct in photographs. However, to estimate a tiger population, I had to deploy dozens of camera traps across vast forests. Expensive and complicated equipment was ruled out.
Fortunately, just at that time, cheap camera traps—used by American deer hunters to identify trophy bucks—came onto the market. In 1990, setting these traps along the favored tiger trails in Nagarahole, I ‘photo-captured’ several distinct tigers, including three that I had radio-collars on. I now had a method that could photograph, identify, and count tigers!
However, the next challenge appeared insurmountable. I had to estimate real tiger numbers in order to know how their populations were doing. I needed to answer questions, such as: How many tigers were present each year, how many were born, how many survived, and how many moved in or out.
To understand their population dynamics, in addition to counting tigers I had caught, I had to know how many tigers I had missed. There was no guarantee that every tiger in the population would be caught. Effective conservation demanded real tiger numbers, not just the minimum count from photos. This meant I had to know what proportion of the tiger population I had caught. In other words, I had to estimate what the chance (probability) that any tiger in the sampled population was photo-captured. To crack this mystery, I had to penetrate a different jungle: one covered in dense thickets of probability statistics, models, and computers. That is another story.