Land size limits

Why does the largest animal on land fall so far short of the largest in the sea?

A blue whale runs to something like 150 tonnes. The largest bull elephants reach around six, with exceptional individuals heavier. Even the largest land animals that ever lived, the great sauropods, are estimated only in the tens of tonnes — and the estimates carry error bars wide enough that the ranking among the contenders keeps shifting. So the sea holds animals an order of magnitude beyond anything the land has managed, twice over. Why should the medium an animal lives in set a ceiling on how big it can be?

Take the simplest possible model first and see where it fails. Imagine an animal, and imagine doubling every one of its linear dimensions — twice as long, twice as tall, twice as wide, same shape, same materials. What happens to its weight? Weight follows volume, and volume goes up as the cube of length: eight times heavier. What happens to the leg bones that must carry that weight? A bone resists compression in proportion to its cross-sectional area, and area goes up only as the square: four times stronger.

So the load doubles relative to the support. Double the size again and it doubles again. This is Galileo's observation, published in 1638, and it is the whole of the structural argument: in a body scaled up without changing its design, the stress in the skeleton rises in direct proportion to linear size. Nothing has been assumed about biology at all. It follows from geometry.

An animal facing that can do two things. It can build disproportionately thick bones — and large mammals do have relatively stouter limb bones than small ones. Or it can change its posture, which turns out to matter more: a shrew runs with crouched, bent limbs, while an elephant stands on nearly straight columnar ones, holding the bones close to the line of the load so the muscles need far less force to resist collapse. Both fixes cost something, though. Straight columns cannot jump, dodge or gallop properly, and past a certain size a fall is fatal on its own — as Haldane put it in 1928, drop a mouse down a mine shaft and it walks away, drop a horse and it splashes.

Now ask what changes in water. The animal's weight is opposed by buoyancy — the upward push of the water it displaces — and for a body of roughly the same density as seawater those two nearly cancel. The skeleton is no longer holding the animal up. The square-cube problem does not go away, it simply stops being the binding constraint, which is why a stranded whale is crushed by its own bulk within hours of losing the support it evolved to rely on.

But it would be too tidy to say bones alone explain the land ceiling. Other limits appear well before the skeleton snaps. Heat is one: an animal produces heat roughly in proportion to its mass and sheds it across its surface, so bigger animals cook more easily — which is part of why elephants have those ears. Food is another: a body eight times heavier needs vastly more fuel, gathered by a mouth whose area has only quadrupled, so the hours in the day become a constraint. And a mammal carrying a proportionate foetus internally, for a proportionately long gestation, hits limits of its own.

The most common misreading is that the square-cube law predicts a specific maximum size, and that sauropods somehow violated it. They did not violate anything; they evaded the corollaries, and how they did it is why the subject is still interesting rather than closed.

Their skeletons were pneumatic — the vertebrae were invaded by air sacs of the kind birds still have, hollowing out much of the backbone and cutting the mass a leg had to carry, while quite possibly helping dump heat as well. Their heads were small and they did not chew: food was cropped and swallowed, so intake rate was not limited by a heavy grinding apparatus on the end of a very long neck. That neck let a stationary animal harvest an enormous volume of vegetation without paying to move its body. And they laid eggs, many at a time, which removes the reproductive ceiling that a gestating mammal runs into.

Two honest caveats. The mass estimates for the largest extinct giants are reconstructions from incomplete skeletons and vary substantially between methods, so precise comparisons should be treated gently. And the question "why is nothing on land that big now" is not purely a physics question — it involves the history of what went extinct, what habitats and plant communities exist, and how long a lineage has had. Physics sets the terms of the problem. It does not by itself pick the winner.

• Why metabolic rate does not scale in simple proportion to mass, and what that implies for very large animals.
• What actually sets the upper limit on whales, once the support problem is removed.

Every term the collection defines is gathered in the glossary.