For most of the time anyone has studied this animal, finding one meant standing still and listening.
A radio collar sends a beep, a researcher turns a handheld aerial until the beep is loudest, and a line gets drawn on a map.
Two lines from two ridges give you a rough point.
That is how the numbers in the textbooks were built, in a landscape where getting to a second ridge can take a day.
Between 2013 and 2017, four cats in eastern Nepal carried something that did not need anyone to be listening, and the results were published in 2025.
The map changed more than the animal.
A polygon can only hold the places you went
The standard way to turn locations into an area is to draw the tightest shape that contains them all.
That method inherits every bias in the collecting. If an animal is only ever located from places a person can reach, the polygon stops where the people stopped, and it looks like a boundary when it is a logistics limit.
Radio tracking in this terrain produced a handful of positions a week, in good weather, from the valleys, which makes the resulting shape small by construction.
A satellite collar takes a fix on a timer. It does not care whether anyone is within range, whether the weather closed in, or whether the cat walked over a pass into another country.
So the same animal, measured the second way, occupies an area that can be dozens of times larger without having changed a single habit.
The instrument moved the answer.
A conservation area with a border along its roof
The study ground is a protected area in Nepal’s far northeast, wrapped around the southern flanks of the world’s third highest mountain.
Its northern edge is the frontier with the Tibetan plateau and its eastern edge touches the Indian state of Sikkim, which means the boundary runs along the high crest rather than through a valley.
That matters because a ridgeline is not a barrier to this species. It is the preferred route, the ground where broken rock and steep relief give an ambush predator everything it needs.
The animals were recorded as high as 19,186 feet, and the terrain falls away into gorges more than three miles deep nearby.
A line drawn across that country on a map has no physical presence at all.
The border is a line on paper.
Four collars, 4,707 fixes, twenty days to twenty two months
The collars returned 4,707 locations in total, and the tracking periods ran from 20 days at the short end to 659 days at the long one.
Two adult females and one adult male are reported individually. Measured with a minimum convex polygon, their ranges came out at about 398, 282 and 81 square miles.
Measured with a local hull method on the same data, the same three animals came out at roughly 120, 120 and 39 square miles.
That comparison against older Nepali estimates is where the 6 to 97 times figure comes from, and the top of that range sits on the convex polygon numbers.
The paper carries 22 authors and appeared in a conservation journal last year.
Methods disagree with each other.
Two estimators, one cat, a factor of three
The honest reading starts with that spread, because the male comes out at 398 square miles by one method and 120 by another.
Neither is wrong. A convex polygon includes every acre inside the outline, including glaciers and cliffs the animal never used, while a hull method hugs the actual tracks, in the way that a single count or a counted rescue turns into a headline figure.
Four animals is also four animals. One collar reported for under three weeks, the fieldwork is now eight to thirteen years old, and nothing here establishes a population trend.
The estimates and the methods are set out in the published study with the author list.
The multiplier is method dependent.
The finding that does not depend on the method
Strip the estimator argument away and one result survives untouched.
Three of the four cats crossed an international border five to seven times each, spent between 10 and 34 percent of their tracked time in India or China, and had between 28 and 50 percent of their mapped range lying inside India.
No choice of polygon changes that, because a fix either falls on one side of the line or the other.
Nepal counted 397 of these cats in a national survey last year, with roughly 42 percent of the country’s suitable habitat sitting outside any protected area, which the reporting set alongside the tracking.
The GPS collars settled one thing cleanly.
Half a cat’s ground sits in another country.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
