Australasian Science: Australia's authority on science since 1938

Were dinosaurs warm-blooded?

T. rex

There are several lines of evidence that the basal archosaurs were endotherms

By Roger Seymour

An analysis of muscular power reveals that cold-blooded crocodiles are poor models for our beliefs about dinosaur physiology.

When thinking about dinosaurs, people typically think of large, powerful reptiles that were cold-blooded killers – just like modern crocodiles. The first dinosaur fossils were unquestionably associated with reptiles and just as unquestionably thought to be cold-blooded.

We call cold-blooded animals such as reptiles, amphibians and fish “ectotherms” because their body temperatures are influenced mainly by the outside environment. In contrast, warm-blooded birds and mammals are called “endotherms” because they heat their bodies internally. The strategies are so different that it makes us wonder what the dinosaurs were like.

We know that dinosaurs, crocodiles and birds form a group called the “archosaurs” (ruling reptiles), which evolved separately from the rest of the reptiles. Two lineages arose from the basal archosaurs – the crocodilian line and the dinosaurian line, the latter giving rise to birds.

The trouble is that the crocodilian line is represented today by cold-blooded crocodiles and alligators while the dinosaur line is represented by warm-blooded birds. This implies that either the dinosaur-bird line evolved endothermy from ectothermic basal archosaurs, or the crocodilian line evolved ectothermy from endothermic basal archosaurs.

I believe that all archosaurs, including all dinosaurs, were actually warm-blooded, while the crocodile line assumed cold-bloodedness to suit their sit-and-wait lifestyle. The alternative scenario proposes that dinosaurs were cold-blooded and that living large crocodiles are a good model for dinosaurs.

Saltwater crocodiles (Crocodylus porosus) are the largest living crocodiles in the world, reaching over 1 tonne in weight. They are typical ectotherms in that their body temperatures are variable and subject to the temperature of their environments. However, crocodiles are able to raise their body temperatures above 30°C by basking in the sun during the day.

Furthermore, the larger they are, the more stable the body temperature is. A crocodile weighing 8 tonnes, like the extinct Sarcosuchus from the Early Cretaceous (146–100 million years ago), would have changed body temperature only about 0.1°C during a 24-hour period.

Temperature stability due to large body size is a physical fact called “inertial homeothermy”. Some palaeontologists have coined the word “gigantothermy” and applied the idea to large dinosaurs, suggesting that their size would have made their temperatures almost constant.

The advantage of high body temperature is that practically everything that keeps an animal alive depends on temperature, so essential activities such as muscle contraction (for locomotion) nervous activity (sensors such as the eye’s retina and speed of nerve conduction) and functions of the internal organs (e.g. digestion, circulation) can operate effectively and efficiently. Stable body temperature is also advantageous to allow an animal to be independent of environmental temperature changes.

In short, animals with warm and stable body temperatures have advantages over those with cool and variable body temperatures.

However, regulating body temperature internally is energetically expensive. Endothermic birds and mammals require about 5–10 times the amount of food as ectothermic reptiles of the same size, even when resting. Hence the idea of giganto­thermy in dinosaurs is compelling to some palaeontologists because the extinct giants could have achieved high and stable temperatures as ectotherms without the need to spend the amount of energy expected for an endotherm.

It is clear that large dinosaurs could have been similar to large crocodiles in being ectotherms with moderately high and stable body temperatures, but how powerful would they have been with an ectothermic, crocodile-like physiology?

The power output of ectothermic lizards is similar to endothermic mammals, but the origin of that power is very different. While the maximum muscular power of lizards is largely derived from the anaerobic biochemical pathway (which does not rely on oxygen and produces lactic acid as waste), mammals primarily use the aerobic pathway (which needs oxygen and produces CO2 as waste).

Because of the build-up of lactic acid and other biochemical changes, anaerobic power generation is not sustainable and the animal fatigues quickly. In contrast, aerobic energy production is sustainable because CO2 is lost through the lungs.

These conclusions are based on measuring the maximum rate of oxygen consumption (to assess the aerobic component) and the maximum rate of lactic acid production (to assess the anaerobic component). Each molecule of oxygen or lactic acid can be converted into an equivalent amount of energy available for use in the muscles.

Measuring oxygen consumption is easy, but measuring lactic acid is difficult because it builds up in the muscles, blood and other tissues to different extents. Research on small reptiles in the 1970s involved killing the animals after strenuous activity and homogenising the entire body to extract the lactic acid.

The conclusions were based on very small animals weighing only a few hundred grams. Would the similarity of power production between reptiles and mammals persist if they weighed a few hundred kilograms? The answer seems to be that it doesn’t.

In order to understand the exercise physiology of large saltwater crocodiles, I analysed blood and muscle lactate measurements of “salties” with John Baldwin of Monash University and Grahame Webb of Wildlife Management Inter­national (Northern Territory). With ethical approval from the University of Adelaide, we captured crocodiles ranging in size from 240 grams to 188 kg from the Adelaide River near Darwin. The crocodiles were approached by boat at night and secured to a thin cord by a small barb through their skin. This caused the animals to thrash violently to the point of complete exhaustion in water, and they failed to right themselves on landing.

The duration of the exercise was recorded, and blood and tail muscle biopsies were obtained at exhaustion. Both samples were required to account for all of the lactate.

Obviously we could not homogenise an entire crocodile, so only small samples were taken. All animals recovered completely and were returned to the river.

The results of that study were published in 1995, but we did not think to compare the metabolic power of crocodiles with mammals because we assumed that they would be similar. That assumption was wrong.

This year I revisited the crocodile data and compared it with published information on the effect of body size on aerobic and anaerobic muscular power in mammals. During maximal, non-sustainable activity, mammals produce about half of their energy with aerobic pathways and half with anaerobic pathways.

Crocodiles, in contrast, are about 90% anaerobic during activity, so they eventually fatigue. A 1 kg crocodile fatigues in about 10 minutes, while a 200 kg one can struggle for 50 minutes.

To be fair, therefore, I calculated power generation by crocodiles during the first 10% of activity periods, because that is when they are most powerful.

The results were unexpected. Not only did small crocodiles produce less energy than mammals, but the proportion decreased in larger animals.

The power of a muscle can be measured by the external work that it does or by the amount of energy that it consumes. Muscular power consumption, like the power consumption of any machine, can be calculated in units of Watts, with 1 W defined as 1 Joule of energy per second.

A resting human consumes about 80 W. A 1 kg crocodile consumes 16 W during maximal activity, which is about 60% of the energy consumption of a mammal of that size. A 200 kg crocodile consumes about 400 W, or only 14% of a similar-sized mammal. Judging from this trend, larger crocodiles would consume even a smaller percentage.

The significance of this comparison is that a crocodile-like dinosaur would not produce as much muscular power as a mammal-like dinosaur. It is tempting to imagine a fight between the two dinosaurs of different physiologies – the mammal-like dinosaur would clearly have the advantage.

These results further show that cold-blooded crocodiles lack not only the absolute power for exercise but also the endurance of warm-blooded mammals, even at similar body temperatures. They make it clear that simply raising the body temperature is not the evolutionary objective of being an endotherm, but having a high capacity for aerobic metabolism is.

The evolution of endothermy is thought to be associated with a high capacity for aerobic metabolic power production, which is advantageous for a number of reasons. For one thing, aerobic metabolism is sustainable because no metabolites like lactate build up. For another, aerobic metabolic pathways produce about ten times the energy as anaerobic ones because the energy in the food substrate is more fully extracted.

Aerobic pathways occur in mitochondria, and consequently the density of mitochondria in birds and mammals is about five times higher than in reptiles. However mitochondria are not perfect converters of energy – they leak about 25% of the energy that flows through them, ending up as heat.

Therefore, if an animal develops a high aerobic capacity, the heat loss from the mitochondria could either be wasted or used to raise the body temperature and maintain endothermy.

There are now several lines of evidence that the basal archosaurs were in fact endotherms, and we can see it today in crocodiles and birds.

First, both groups have four-chambered hearts to separate high pressure blood going to a highly active aerobic body from low pressure blood going to the lungs. (Blood vessels in the lungs are at low pressure blood to prevent fluid from squeezing into the air spaces and fatally preventing gas exchange.)

Second, the lungs of birds are highly complex structures that permit an effective one-way flow of air through them to supply oxygen quickly. The lungs of crocodilians also have a highly complex one-way flow, indicative of their endothermic ancestry, yet have one of the lowest gas exchange capacities of any reptile.

Third, birds and crocodiles share an endothermic bone type. Only neonatal and juvenile crocodiles have fibrolamellar bone – reflecting their endothermic past – but they lose it as adults.

There are several other independent lines of anatomical, physiological and behavioural evidence for endothermy in basal archosaurs, so why did the crocodilian lineage abandon endothermy?

Most of them probably didn’t, and were endotherms until they became extinct. More than 150 genera of large crocodilians coexisted with the dinosaurs through most of the 185 million years of the Mesozoic Era in terrestrial, marine and freshwater environments. Many were highly active, often bipedal, predators on land. Only one crocodilian group survived the extinction of large dinosaurs – the crocodiles and alligators that we see today.

Living crocodilians are archetypal, sit-and-wait predators in water that grab their unsuspecting prey with a short burst of power and immediately crush and eat small prey or drown large ones before eating them. They have no need for endothermy or sustained locomotion. In fact, trying to maintain a high body temperature against the cooling power of water would be so energy costly that it would be selected against.

Furthermore, the low metabolic rate and low oxygen demand of ectothermy would be selectively favoured because it would permit longer breath-holding while waiting for prey and while drowning it.

Thus the sit-and-wait aquatic predator niche of modern crocodilians has selected for a shift from endothermy to ectothermy, and from mainly aerobic to anaerobic power, with a consequent reduction in maximal power output to a fraction of what is expected for aerobic endotherms.

Roger Seymour is Professor of Ecology and Evolutionary Biology at the University of Adelaide.