ROM Research: Weighing Giants

Posted: August 24, 2012 - 08:25 , by royal
What's New, Dinosaurs, Palaeontology, Research | Comments () | Comment

by Nicolás Campione, PhD Candidate, University of Toronto

Ever wonder how dinosaurs are weighed? Determining the mass of an animal that has been dead for more than 65 million years can be difficult, but knowing the weight of an animal can provide useful information about how it lived. This is because the size of an animal is strongly related to its physiology (how its body functions) and ecology (how it interacts with the environment). A mouse, for instance, has a much higher metabolism relative to its body mass than does an elephant; but a population of elephants will consume a much larger volume of food and therefore requires a much larger area to live in than that of mice. Given relationships such as these, getting accurate size estimates of extinct animals, like dinosaurs, are important in order to reconstruct the biology of these animals, and perhaps investigate changes in ecological interactions throughout the history of life. So, given the importance of body size (or body mass), what is the best way to estimate the mass of a dinosaur?

Inside the ROM's Dino storage room the shelves are filled with bones.

Figure 1. Standing next to a complete skeleton of an African elephant, Loxodonta africanus, the largest living land animal.

Body mass of extinct animals can be estimated in two ways. The first method is based on creating a scaled, fleshed-out body reconstruction of the dinosaur using its skeleton. This can be done by either creating a physical sculpture or, the more recent technique, by scanning mounted museum skeletons and reconstructing the soft tissue using computers. Once the reconstruction is complete, its volume is calculated. Traditionally, this is done by displacement (e.g. measuring the amount of water the sculpture displaces and scaling up), or through the use of 3D computer software. The problem with using body reconstructions to estimate body mass is that it is somewhat subjective. In general, dinosaurs do not preserve soft tissues and therefore the amount of soft tissue—muscle sizes for instance—added to a skeleton is up to the researcher or artist. The second problem is that in order to convert a volume (which is measured in m3) into a mass (measured in kg), the volume needs to be multiplied by body density (measured in kg/m3). Most mammals living today have densities close to that of water (1000 kg/m3), however, birds can have much lower densities due to their highly pneumatic (or air-filled) bodies. Given that birds are a feathered, flying dinosaurian lineage, and the fact that most species of dinosaurs also have highly pneumatic skeletons, what should we assume the density of a dinosaur to be? Was body density the same in different dinosaurs? These questions are virtually impossible to answer, but can have a big effect on body mass estimation.

Line chart

Figure 2. The relationship between the total humeral and femoral circumfered and body mass in mammals, reptiles, and some amphibians.

The second method relates to the measured relationship of live body mass to the size of the bones used to support the mass in living animals. This method is more objective as it does not require assumptions about the amount of soft tissue and body density. This method begins by determining the relationship between a skeletal measurement (for example the length or circumference of the humerus and femur) and body mass in living animals. Once the relationship is determined, the skeletal measurement can be taken from the dinosaur and placed into the scaling equation to provide an estimate of body mass. The main criticism with this method has been that it assumes that the scaling relationship is similar among different types of animals (for example, mammals and dinosaurs). This is especially important because the posture and gait of different types of animals can be very different. For example, the femur is held vertically and below the body in mammals (and most dinosaurs), but is held horizontally and at a right angle to the body in reptiles—think of the difference between a horse and a crocodile. Such differences have been assumed to diminish the utility of using a scaling relationship derived from living mammals for estimating body mass in dinosaurs. However, these assumption can be tested in living animals, with important implications for estimating body mass in a range of extinct animals, not only dinosaurs.

This was the main goal of a study recently published in the journal BMC Biology by myself and my PhD supervisor Dr. David Evans (Curator of Dinosaurs, Royal Ontario Museum). We set out to test the main criticisms of limb scaling as applied to dinosaur body masses by looking for differences in the scaling of the main front and back leg bones (humerus and femur length/circumference) and body mass using a new, and unprecedented, sample of live-weighed skeletons of living mammals and reptiles (Fig. 1). As was expected, our results revealed significant differences in limb scaling between mammals and reptiles but only when we related the length of the humerus or femur with body mass. To our major surprise, the relationship between the circumference (the size of the bone shaft) of these bones and body mass was remarkably similar between mammals and reptiles—in fact, it was statistically identical (Fig. 2). This simple test allowed us to conclude that the relationship between the circumference of the main weight-bearing limb bones and body mass transcends major differences in limb posture and thus provides a consistent and reliable scaling equation for estimating body mass in land-living four-legged animals, including many dinosaurs.

A large warehouse holding dinosaurs articulated and fully mounted.

Figure 3. Measuring the largest duck-billed dinosaur, Shantungosaurus giganteus.

When we apply our findings to a set of familiar dinosaurs (Fig. 3) our results generally indicate that previous model-based estimates of dinosaurs are reasonable. However, several volumetric body models either under- or overestimate body mass. For example, our equation suggests a body mass of 8 tonnes for the ornithopod Iguanodon, which is approximately double that typically suggested for the species. At the other end of the spectrum, our estimate of the giant sauropod Brachiosaurus at 35 tonnes is significantly below the 70 tonnes suggested by some models, but consistent with others-although we get a heavier mass than the most recent volumetric attempt (see here for a story: skinny brachio). Given that life reconstructions of dinosaurs are important for addressing many biological questions, such as how dinosaurs moved, our results will help to constrain these reconstructions and test whether their assumptions conform to the patterns seen in living animals. Finally, our results only apply to quadrupedal animals, but we are working on new analyses addressing body mass estimation in bipeds, which include all of the meat-eating dinosaurs, so stay tuned to find out what we get for the mass of T. rex and its relatives.

David and I have made this research accessible to everyone via Open Access publishing.  Read the original research paper.


Campione, N. E., and D. C. Evans. 2012. A universal scaling relationship between body mass and proximal limb bone dimensions in quadrupedal terrestrial tetrapods. BMC Biology 10:60 doi:10.1186/1741-7007-10-60 (free online access)