Carnivores: felids, canids, ursids

About ten wolves on a snowy bank, seen between the trees. — The Wapiti Lake wolf pack near Otter Creek, in Yellowstone Park, photographed in 2019 by A. Falgoust for the US National Park Service. YellowstoneNPS, Public domain, source

The alpha wolf does not exist.

This claim is neither an opinion nor a recent one. The man who spread the idea spent the last twenty years of his working life trying to take it back.

He did not succeed. That is the story worth telling.

The origin of the idea

In the mid-twentieth century, nobody knew how to follow wild wolves. Winter, the forest, the distances: it was impossible.

So they were studied in captivity.

Adult wolves, from different places and unrelated to one another, were taken and put together in an enclosure — the one at Basel Zoo, where observations began as early as 1934.¹

This set-up deserves close examination, because everything follows from it. Animals that, in the wild, range in families over territories of several hundred square kilometres, shut up together in an enclosure, among strangers.

What happened? What was bound to happen. They fought over food, space and right of way. The biggest and most aggressive gained the upper hand. The ranks were given names: alpha, beta, omega. It was a Swiss zoologist, Rudolf Schenkel, who used the word “alpha” in 1947, after observing wolves at Basel Zoo.²

The conclusion drawn was that this is how wolves lived.

A comparison to test the reasoning

In 2010, thirty-three miners were trapped underground in Chile for sixty-nine days.

One could have studied those thirty-three men, in that hole, during those sixty-nine days, and drawn conclusions from them about human behaviour in general.

The approach would seem absurd. Yet that is exactly what was done with wolves.

The observation set-up produced what it observed.

This mechanism appears elsewhere on this site, with the canals people thought they saw on Mars. Nobody lied, and nobody observed carelessly. It was the way of observing that manufactured the result.

What he did next

The biologist who had written the book was L. David Mech. His book came out in 1970 and became a lasting success. In it he took up Schenkel’s word “alpha”, because that was all that science knew at the time.³

Then, from 1986 onwards, he went to Ellesmere Island, in the Canadian Arctic, where the wolves tolerate the presence of humans. He spent whole summers there observing free-living packs.⁴

What he observed bore no resemblance to a power struggle.

A wolf pack is a family. A pair that has young, and its offspring from previous years. The “leaders” are not the strongest: they are the parents.⁵

In 1999, he published the correction in a scientific journal.⁶

The real subject of this page

Mech could have stopped there. He did not.

He publicly asked people to stop using the word, and he repeated this for years.⁷

His book, on the other hand, went on selling. As late as 2020, Mech acknowledged that he was partly responsible for spreading the idea, through this book that was still on sale.⁸

The book remained on sale for fifty-two years, until it was withdrawn in 2022.⁹

This fact is worth pausing on. A scientist understood that he had been wrong. He wrote it, he published it, he repeated it for twenty years. The book went on selling in bookshops, with the error it contained.

The word’s career outside biology

The word “alpha”, for its part, did not stay with wolves.

It was used to justify dog-training methods based on dominance — since the dog descends from the wolf, and the wolf supposedly obeys the strongest.

It was used to give a veneer of nature to a certain idea of leadership among humans: supposedly there are alphas and the rest, and this comes from nature.

But it does not come from nature. It comes from an enclosure.

Three families, three ways of being in the world

The order Carnivora comprises about 300 species of placental mammals (see Entry 5.7, Rodents and small mammals). Within this order, ways of hunting and of living together vary widely; felids, canids and ursids offer three examples.¹⁰

Canids (wolves, foxes, African wild dogs): specialists in endurance and the group. Their strategy rests on cooperation, tracking over long distances and a family-based social structure organised around clear rules.
Felids (lions, tigers, lynxes): hypercarnivory and the perfection of ambush hunting. Retractable claws, night vision, lightning acceleration. With the notable exception of the lion — which lives in prides to defend a territory against other predators — felids are strictly solitary. Among them, interaction with others of their kind is limited to reproduction and territory.
Ursids (brown bears, grizzlies, pandas): power and opportunism. Although classified among the carnivores, most are generalist omnivores. Solitary, they adapt their behaviour through the seasons, building up massive reserves before entering winter torpor. They have no pack and no permanent dominance: their contacts are dictated solely by the availability of resources (as during salmon runs).

Dominance is not what drives these species. For the solitary felid, it has no meaning. For the bear, it comes down to mutual avoidance or temporary precedence over food. For the canid, it gives way to kinship ties.

The most widely told wolf story in the world

You may know it: biology teachers all over the world use it.

Wolves had been eradicated from Yellowstone Park in 1926. They were reintroduced there in 1995. From then on, according to the story, the forest came back, the rivers changed course, the whole ecosystem regained its balance.¹¹

This story is being unravelled, and has been for twenty years.

The timeline

2004 — in the upper Gallatin (Montana), on the edge of the park, photographs document a recovery of willows.
2008 — other researchers publish photographs that dispute it.
2010 — the aspens had not grown back, even though the elk population of the northern winter range had fallen by about 60%. (Caution about the scope: depending on whether one counts the whole park or a specific study area, the reported decline ranges from 20% to over 60%. One must always ask what territory a percentage refers to.)
2011 — across 315 stands monitored, densities of tall stems fell by 12% at the landscape scale.
2024 — a twenty-year field experiment concludes that the return of wolves was not enough to restore the habitat.¹²

The latest episode, in 2025

In early 2025, one team reported the opposite: a strong cascade. Its figure went round the world: a 1,500% increase in willow crown volume, measured at twenty-five riverside sites between 2001 and 2020. The authors claim the effect exceeds most trophic cascades documented worldwide.¹³

In October 2025, another team published in the same journal a comment demonstrating that this conclusion is invalid.

The main criticism deserves to be understood, because it is crystal clear. The authors had not measured crown volume: they had calculated it from plant height, using a formula that defines volume precisely as a function of that same height.¹⁴

In other words: you supply a height, the formula returns a volume, and you then announce that the volume has increased. But the volume was never anything other than the height, dressed up differently. An argument that assumes what it claims to prove is called circular.

Other criticisms follow, and the list is in itself a lesson in method: violated calculation assumptions, plots compared without being matched, equilibrium indicators applied to a system that is not at equilibrium, selectively framed photographs, and the omission of human hunting among the possible causes.¹⁵

The debate continues: in 2026, in another journal, the challenged authors replied to a new critique by Daniel MacNulty and his colleagues, this time targeting their study on aspens.¹⁶

The two teams worked on the same height measurements.

The disagreement is not about the facts. It is about how to treat them.

What the researchers say

The ecologist Tom Hobbs, of Colorado State University, who studies how willows have responded to the reintroduction:

“The story of wolves in Yellowstone has been made true by repeated telling, not by good science.”¹⁷

He adds that the cascade story is presented as an indisputable fact when it is not: a fine story, simple and clear, but in reality more nuanced, more complex, and perhaps false.

The supposed mechanism — that of fear — does not stand up well to measurement. A 2009 study based on 727 elk faecal samples showed that their willow consumption depended more on snow conditions than on the presence of wolves: if there is a cascade, it would work through the fall in numbers, not through fear.¹⁸

The story also glosses over a crucial point: if the vegetation changed when elk numbers fell, human hunting may have contributed as much as the wolves. This omission is one of the criticisms made in 2025.

A conclusion to avoid

It would be wrong to conclude: “scientists contradict each other, so we know nothing”.

It is the reverse.

We know precisely where the disagreement lies, on what data, and for what reasons of method. Both articles can be consulted, and one can check who is right on each point.

A well-documented disagreement is worth more than a vague agreement.

Honesty cuts both ways: yes, stands of aspen and willow are thriving again — in places. But large areas remain damaged and may never recover.

Wolves are not a magic wand.

Large predators and us: the question of safeguards

Wherever the wolf, the bear or the lynx returns, the question ceases to be purely biological: it becomes social and economic.

The debate revolves around two realities that clash on the ground:

The protection of biodiversity: large carnivores play an important role in maintaining natural balances and removing sick animals. Their presence is protected by international regulations (Bern Convention, Habitats Directive).
The vulnerability of livestock farming: the return of predators imposes heavy constraints on farmers (livestock guardian dogs such as Pyrenean mountain dogs, electric fences, human presence). Despite these measures, predation causes stress, financial losses and a significant mental workload.

There is no simple solution. Management tools range from strict protection, through compensation for losses, to regulated defensive or cull shooting. The task is not to pass moral judgement, but to measure the real costs — financial and human — of sharing territory.

One last clarification, to avoid the opposite error

Saying that the alpha wolf does not exist does not mean that wolves live in disorder.

They have an organisation, and it is highly structured.

It is that of a family.

Note 1

Die Beobachtungen, welche dieser Arbeit zugrundeliegen, wurden 1934 begonnen. Sie galten zuerst fast ausschliesslich […] den Wolfsrudeln, die der Basler zoologische Garten […] beherbergte

The observations on which this work is based were begun in 1934. At first they concerned almost exclusively […] the wolf packs that the Basel zoological garden […] housed

Schenkel, Rudolf, “Ausdrucks-Studien an Wölfen: Gefangenschafts-Beobachtungen”, Behaviour, vol. 1, no. 2, 1947, p. 81-129