Between one estimate and the next, roughly 369 wolves disappeared from Montana. Not from the landscape, from the spreadsheet. What changed was the math.

In its 2025 Annual Wolf Report, Montana Fish, Wildlife & Parks (FWP) published two different numbers for the same wolves at the same moment in time. The original model, iPOM 1.0, put the statewide population at 1,094 animals for 2025, somewhere between 991 and 1,211. A revised model, iPOM 2.0, put the same population at a preliminary 725 wolves, somewhere between 618 and 836. The gap between them is about 34 percent of the population, gone on paper.

That is a strange thing to reckon with. So we went looking for what actually happened, and why hundreds of paper wolves turned into a very real fight over hunting quotas.

Why Montana models its wolves instead of counting them

Nobody at FWP is walking the state counting wolves one by one. The agency stopped publishing head counts back in 2018, having decided that trying to physically tally every animal across tens of thousands of square miles of rough country was too expensive and too unreliable to continue.

What replaced the head count is a statistical estimate called iPOM. Instead of a tally, it combines several inputs: reports from hunters who spot wolves, known wolf locations, habitat data, and assumptions about how big packs tend to be and how much land they hold.

So every wolf number out of Montana for the last several years is an estimate with a margin of error, not an actual count. That matters for everything that follows. Change the assumptions the model runs on and the number moves, even if not a single animal on the ground has changed.

What iPOM 2.0 changed under the hood

The revision adjusted several assumptions at once. It changed how the model handles habitat, how much land it assumes each pack occupies, and how big it assumes an average pack is. The typical territory in the busier wolf areas grew from roughly 450 square kilometers under the old model to about 510 under the new one. Bigger territories mean fewer packs fit into the same landscape.

The pack math moved too. iPOM 1.0 estimated 181 packs, against 155 under iPOM 2.0. The newer model also draws on a wider set of pack-count data going back to 2007, which lowers its estimate of average pack size. 

FWP’s chief of conservation policy, Quentin Kujala, called it “a refinement of our modeling process”. That is the agency’s own framing.

Warnings sit right next to that 725 figure. The new model has been laid out in a June 2026 technical summary published by the U.S. Geological Survey, but the authors themselves call the results preliminary and say they remain subject to final model refinement and peer review. FWP likewise says iPOM 2.0 has not yet been through the same peer-review process as the original model, and a manuscript describing all of the model updates is still in progress and will be submitted for peer review and publication.

Why hundreds of paper wolves became a real fight

A lower estimate leaves less room for hunting. In response to the new figure, FWP proposed cutting the statewide hunting quota to 250, down from 452 the previous year.

That proposal walked into a five-hour argument. After a heated meeting, the Fish and Wildlife Commission settled on a compromise quota of 300. For context, hunters and trappers took 247 wolves in the 2025-2026 season, the lowest total since 2016.

The deeper stakes are the state’s own floor. Montana’s wolf plan, following direction from the legislature, sets a minimum of 450 wolves and 15 breeding pairs. iPOM 2.0 projects that with an average hunting take, the population could fall below that 450 line within three years. That is the number that turns a modeling change into a possible federal question about protecting wolves again.

Everyone read the same report and reached different conclusions. Lizzy Pennock, a carnivore coexistence attorney with WildEarth Guardians, argued the quota was still too high, noting that “Two-hundred fifty wolves out of 725 is still a third or more of the population”.

From the other side came suspicion that the revision was too convenient. State representative Jedidiah Hinkle, who chairs the House Fish and Game Committee, said of the change, “By just changing the pack size and the distribution of the territory, we’re able to knock off 400 wolves”. On the commission, Region 1’s Ian Wargo voiced his own doubt about the state’s science, saying “It doesn’t make a whole lot of sense to me”. Both are opinions from people who dispute the model, not proof that the model is wrong.

How to read a modeled number without treating it as gospel

The revision did not come out of nowhere. Back in 2023, Robert Crabtree, chief scientist at the Yellowstone Ecological Research Center, released an early draft paper arguing that iPOM’s estimates ran high, a claim FWP disputes. Crabtree’s framing was blunt: “We demonstrated iPOM has an inherent severe overestimation bias”.

The draft had not been peer-reviewed, and FWP along with iPOM’s developer Sarah Sells pushed back hard, calling it “preliminary and weakly supported at best”. It stands as one contested, unreviewed argument that the model ran high, not a settled verdict. The 2026 revision came years after that dispute.

What we take from the whole episode is less about wolves than about numbers. Every one of these figures, the 1,094 and the 725 alike, comes with a margin of error, and those margins are wide. The new estimate could plausibly be anywhere from 618 to 836. Treating any single point in that range as the true, exact wolf count was never really warranted. A model is a best guess with its assumptions written down, and when the assumptions get revised, honest models move.

The one question that would actually settle this hasn’t been answered yet. iPOM 2.0’s full method still has to clear outside peer review and be published. Until that happens, both the people who think 725 is too low and the people who think it is suspiciously convenient are arguing over a number the wider scientific community hasn’t finished checking.