Wildlife & Biodiversity

How robust is India’s tiger census?

Government officials and independent experts help decode the riddle of tiger estimation

 
By Rajat Ghai
Published: Friday 25 August 2023
Photo: iStock__

During the 50th anniversary celebrations of Project Tiger on April 9 this year, Prime Minister Narendra Modi announced that the animal’s population in India had increased to 3,167. Less than four months later, on July 29, a new set of numbers were released that pegged the population at 3,682 tigers; over 500 more than the April estimate.

This has sparked debates over the accuracy of tiger census. Rajat Ghai speaks to government officials and independent experts to decode the riddle of tiger estimation. Edited excerpts:

“We are not here to inflate tiger numbers”

Qamar Qureshi - 
co-author of 
“Status of Tigers Co-predators and prey in India 2022”
Qamar Qureshi, co-author of Status of tigers co-predators and prey in India 2022

Estimating tiger numbers is a huge exercise, with more than 40,000 people involved in data collection.

The first phase of data collection happens at the beat level. The whole of India is divided into 15 square kilometre cells. From each beat, forest guards, volunteers and daily wagers collect data about tigers, leopards, other carnivores, ungulates, large herbivores, vegetation and human disturbance. They use an application to take pictures, collect data and then upload this data in local languages. This then gets translated into English.

Secondary public data and satellite remote sensing data is used for modelling. In phase three, forest department staff and researchers collect camera trapping data including tiger pictures with the help of traps installed across India. We then use artificial intelligence (AI) to sort out those pictures that only feature tigers.

These photos then go through another AI-based programme that identifies individual tigers. Finally, a round of cross checking makes sure everything is in order. We then run models to analyse the data. We thus have data for 83 per cent of the tiger population. The rest is extrapolated.

There is always a margin of error. That is why we use human intelligence as well.

As for murmurs of discontent, people have a lot of misconceptions about the tiger estimation process. We are not here to increase or decrease tiger numbers. Our job is to be as authentic as possible.

Of course, we are human. So, there will always be mistakes. A common mistake people do is adding tigers. For example, you have two locations which have tigers.

But tigers are also shared between them. Karnataka will share tigers with Kerala and Tamil Nadu; Uttarakhand with Uttar Pradesh and so on. When we do population estimation, we use specially explicit models to adjust the tiger numbers between different states. Thus duplication is removed.

“Tiger cubs are not counted due to high mortality”

Sushil Kumar Popli -
Principal Chief Conservator of Forests, Wildlife, Odisha 
Sushil Kumar Popli, principal chief conservator of forests, wildlife, Odisha

The tiger is a fecund animal. Its gestation period is hardly three to three-and-a-half months. Generally, the animal can have a litter once every two or three years, the average litter size being three to four. Animals that are less than a year old are not included in estimation surveys. On the other hand, tiger estimation exercises are usually carried out by the Centre once every four years.

For instance, the data for this latest report — Status of Tigers Co-predators and Prey in India 2022 — was collected in 2021-22. Only those tigers aged one year or older and considered as “sub adults”, who are more likely to survive, are counted. This is justified because human babies less than five years of age are also not considered in population surveys as mortality is more common below five years.

My only point has been that in 2021-22, Odisha did a survey through camera traps. These cameras captured young tigers less than a year old. They must now be over one year of age.

I have evidence of almost 24 or 25 individual animals identified through stripes. But since the data is from 2021-22, we cannot confirm right now whether they are the cubs of 2021. We should have the figures by January 2024, which are definitely going to be more than what has been reported in Status of Tigers Co-predators and Prey in India 2022.

“Tiger numbers could be an overestimation”

Ghazala Shahabuddin -
Senior Adjunct Fellow at Ashoka Trust for Research 
in Ecology and the Environment, Bengaluru
Ghazala Shahabuddin, senior adjunct fellow, Ashoka Trust for Research in Ecology and the Environment, Bengaluru

The effort of the Union government to monitor tigers is commendable, and even the methodology is scientifically valid.

However, the results of the tiger estimate exercise may come into question, as it is not clear from the 2022 report to what extent the assumptions of this method have been met. Another shortcoming in the tiger census is related to the extrapolation of tiger densities from the intensively sampled sites (within each landscape) to the entire landscape. It is mentioned in the report that the intensively sampled site is usually located in areas of high tiger density or conducive habitat within the larger landscape. Due to this reason, extrapolations to the entire landscape are likely to lead to an overestimation of tiger numbers.

Further, to establish population trends in any given location, it is necessary to strictly follow the same protocols and sampling efforts from year to year. Based on this rule of thumb, the trends over time in tiger populations as reported in the 2022 report may not be completely reliable. Although the sampling grids of 100 sq km in the monitoring exercise have remained constant since 2006, the sampling efforts deployed — camera trapping and transect walking — have increased. For instance, the number of camera traps deployed at an all-India level has increased from 26,838 in 2018 to 32,588 in 2022 (a 21.4 per cent increase).

Further, the number of kilometres walked has increased from 522,996 km in 2018 to 641,996 km in 2022, a 22.6 per cent increase. This is possibly reflected in the quantum of increase in the number of unique tiger individuals recorded at an all-India level — from 2,461 in 2018 to 3,080 in 2022 (a 25.1 per cent jump). The attempt to develop indices of tiger density based on visual indirect signs, which is mentioned in the 2022 report, is a good idea for future population monitoring.

“What are we gaining from faulty census”

Gopalaswamy - 
founder and lead scientist of Carnassials Global, Bengaluru
Arjun M Gopalaswamy, founder and lead scientist, Carnassials Global, Bengaluru

Creating an effective wildlife population monitoring programme is a complex task that requires strict adherence to the scientific trinity: formulating pertinent questions, using appropriate methods, and gathering reliable data. The ultimate purpose of this process is to foster a body of credible knowledge that can aid conservation efforts.

India’s official approach to monitoring its tiger population consists of an all-encompassing nationwide survey carried out every four years. While aiming to merely provide up-to-date information on the number of tigers, these surveys fall short of having a clear scientific goal, consequently restricting our comprehension of tiger population and occupancy dynamics.

The survey includes estimation of tiger numbers within reserves using a rigorous method, and data collection on factors influencing tiger abundance (such as prey availability), along with general indices (like tiger signs), across the landscape. These data are then used to build a model correlating tiger numbers with these factors and indices. The model is then used to predict tiger numbers at the landscape scale. The accuracy of these “model-based” estimates relies heavily on the rationale behind the model’s construction and its subsequent stability. Unfortunately, these models have shown considerable inconsistency, a fact well documented in the scientific literature and previous tiger reports. Thus, it is challenging to definitively say whether India’s national tiger numbers have increased, decreased, or remained stable over the past two decades.

At the reserve level, spatial capture-recapture (SCR) methods have been applied to a vast quantity of camera trap data. However, the versions of SCR models used seem poorly suited or potentially misapplied, leading to biases in both the 2018 and the 2022 tiger reports. A cursory reanalysis of the 2022 tiger report suggests that the report may have overestimated the number of tigers by 350-900 within the reserves.

The immediate course of action should be a thorough reanalysis of these data, beginning at the reserve level, ideally utilising custom models, including some that have been developed in India. Alarmingly, the more we invest, the less clarity we appear to gain about tiger population dynamics. Considering the far-reaching implications of these surveys’ findings, it is crucial that we get them as right as we can.

Subscribe to Daily Newsletter :

Comments are moderated and will be published only after the site moderator’s approval. Please use a genuine email ID and provide your name. Selected comments may also be used in the ‘Letters’ section of the Down To Earth print edition.