How Big Did Dinosaurs Get? The Sauropods and the Limits of Size

The largest dinosaurs known from good skeletons are the titanosaurs of South America: Argentinosaurus, described in 1993 from a few vertebrae and a leg bone, and Patagotitan, excavated from 2012 and known from several individuals, both estimated at around 37 metres long and 60 to 70 tonnes. Rivals rest on scraps: Maraapunisaurus, known from a single vertebra drawn in 1878 and since lost, might have been bigger, or might have been a misreading. Among the earlier Jurassic sauropods, Brachiosaurus stood tallest, at about 13 metres, and Diplodocus was the longest for a century at 27 metres.

Mass estimates are uncertain by a factor of two, because the soft tissue has to be guessed. The methods are to scale up from the circumference of the leg bones, which must carry the weight, or to build a three-dimensional model of the body and calculate its volume. The two rarely agree, and every headline claim of a new record should be read with that in mind.

A sauropod scaled up from an elephant would have collapsed under its own weight, overheated and suffocated. The sauropods avoided all three with features inherited from small ancestors and refined over a hundred million years:

• •Hollow bones: the vertebrae and ribs were riddled with air sacs, as in birds, cutting the weight of the skeleton by a third or more without losing strength
• •A bird-like lung: air sacs throughout the body pushed air through the lungs in one direction, far more efficient than a mammal's in-and-out breathing and able to ventilate a nine-metre neck
• •A small head: sauropods did not chew; they cropped leaves with peg-like teeth and swallowed, so the head at the end of the long neck weighed almost nothing
• •Egg laying: a sauropod hatched at around ten kilograms and grew to tens of tonnes, so no mother ever had to carry a giant, unlike a whale or elephant
• •Columnar legs: straight, pillar-like limbs with the bones stacked vertically, so that muscle did not have to hold the joints against the weight

The neck is the sauropod's defining feature and its most argued-over one. Mamenchisaurus, from China, had a neck nearly half its total length, and the necks of the largest titanosaurs reached ten to twelve metres. Whether they were held high, like a giraffe's, or swept horizontally like a boom is unresolved: a high neck would need blood pressure roughly double a giraffe's to reach the brain, and a heart to match, while a low neck would let the animal sweep a wide arc of vegetation without moving its body, which for a sixty-tonne animal is a considerable saving. The likeliest answer is that different sauropods did different things, and that the neck's main purpose was to reach food others could not while standing still.

Mammals have never come close. The largest land mammal, the rhinoceros relative Paraceratherium of thirty million years ago, weighed perhaps twenty tonnes, a third of the biggest sauropod. The reasons trace back to the list above: mammals have solid bones, tidal lungs, heavy chewing heads and live young that must be carried and born, and each of those imposes a ceiling. A whale escapes the limits by living in water, where the sea carries its weight, and the blue whale at 150 tonnes is heavier than any sauropod. On land, the sauropod body plan appears to be the only one that has ever worked at that scale, and it needed the specific inheritance of air-sac breathing from the early dinosaurs to work at all.

It also needed a world to feed it. Sauropods lived in warm, high-carbon-dioxide climates with lush conifer and fern growth, and they ate constantly, perhaps a few hundred kilograms of vegetation a day, digested slowly in a vast gut. When that world ended, so did they.

For much of the twentieth century, books and museum paintings showed sauropods wallowing in lakes and swamps. The idea was that such huge animals could not support their own weight on land and needed water to buoy them up. The evidence turned against it. Fossil trackways show sauropods walking on dry ground, their pillar-like legs were built for land, and their teeth suit browsing on trees. Calculations also suggested that an animal standing deep in water would struggle to breathe, because water pressure on the chest would make it very hard to expand the lungs.

Palaeontologists can also study how quickly these giants grew. Thin slices of fossil bone, examined under a microscope, reveal the bone's internal structure. Many animals lay down lines of arrested growth, rather like tree rings, during seasons when growth slows, and counting them gives an estimate of age. Sauropod bones show tissue typical of very rapid growth, similar to that of fast-growing birds and mammals rather than slow-growing reptiles. The evidence suggests they reached their enormous adult size within a few decades, rather than the century or more once assumed for cold-blooded giants.

The predators never came close to the sauropods. The largest meat-eating dinosaurs, such as Tyrannosaurus, Giganotosaurus and Spinosaurus, reached roughly 12 to 15 metres in length, and the heaviest well-known Tyrannosaurus skeletons are estimated at around eight or nine tonnes. That is still a heavyweight, but a fraction of a big titanosaur. Ecology helps explain the gap. Energy is lost at every step of a food chain, so a landscape can feed far fewer tonnes of hunters than of plant eaters, and top predators are typically rarer and smaller than the largest browsers around them.

At the other end of the scale, many dinosaurs were small. Microraptor, a four-winged feathered dinosaur from China, was about the size of a crow, and some of its relatives were smaller still. If you count birds as dinosaurs, as palaeontologists do, the smallest dinosaur alive today is the bee hummingbird of Cuba, which weighs about two grams. That makes dinosaurs one of the most size-diverse groups of land animals that has ever existed, spanning a range of more than ten million times from the lightest to the heaviest.

The square-cube law explains why great size is so difficult. Imagine scaling an elephant up so that every length doubles. Its volume, and so its weight, increases by two cubed, or eight times, but the cross-sectional area of its leg bones, which sets how much load they can bear, increases by only two squared, or four times. Each square centimetre of bone now carries twice the stress. Scale it up three times and the weight rises 27 times while bone area rises only 9 times, tripling the stress. This is why giant animals cannot simply be enlarged copies of small ones.

The same law makes mass estimates slippery. When the titanosaur Dreadnoughtus was described in 2014, its weight was put at about 59 tonnes, using an equation that predicts body mass from the thickness of limb bones in living animals. The following year another team built a three-dimensional digital model of the skeleton, wrapped it in a plausible amount of flesh and calculated its volume, arriving at somewhere between about 22 and 38 tonnes. Both methods were reasonable, yet the answers differed by roughly a factor of two. Because weight grows with the cube of size, small differences in assumed body shape add up quickly.

The largest dinosaurs were the titanosaur sauropods of Cretaceous South America, around 37 metres long and 60 to 70 tonnes, ten times an elephant. They reached that size with hollow, air-filled bones, bird-style one-way lungs, small unchewing heads, egg-laying and pillar legs, a combination no mammal possesses, which is why nothing on land has approached them since.