Animals vary dramatically in size and biologists have long been fascinated with how other traits change along with size. Metabolism is one of those traits. Metabolic rates increase with size but not proportionally, and the struggle to describe and understand this non-proportional or allometric scaling relationship has a long history.

In their review of the subject, Craig White and Dustin Marshall initially return to the 1830s where physical properties that governed heat exchange were thought to dictate how metabolic rates scale with size. By the 1900s scientists had moved to disputing the exact nature of the ‘scaling’ relationship between size and metabolism.

After more than a century of study, the most reliable finding is that metabolism almost always scales hypo-allometrically with body mass, that is: metabolism increases with size but at a less than 1:1 ratio. A scaling exponent less than 1. So yes, elephants have faster metabolic rates than mice but not as much as you might expect — as body size increases, relative energy use decreases.

More recently, scientists have focused not only on how — but also why — does metabolism scale with body size? Debating this question has resulted in two polarised schools of thought.

The first, which includes Sarrus and Rameaux’s 1830s prediction around heat exchange, is that physical properties constrain the metabolic rate. So, the surface area of an organism can constrain metabolism through regulating heat dissipation, or the network of vessels delivering (usually) oxygen (e.g. blood vessels, trachea or gills), dictates the scaling of metabolic rate through physical constraints to the rate at which resources can be delivered.

The second school of thought takes an evolutionary approach where life history optimisation considers the combination of traits that maximises fitness. Craig and Dustin align with this philosophy. Their recent publication in Science proposed a life history optimisation model which predicts that animals most successful at reproducing are those that exhibit precisely the kind of disproportionate scaling of metabolism with size that we see in real life. And for Craig and Dustin, their work highlights a vital issue — the strong connection between metabolic traits and fitness components: changing one changes them all.

To demonstrate how polarised these two approaches have become, Craig and Dustin took a ‘research weaving’ approach. They found that papers based on the latest theories around the importance of physical constraints in metabolic scaling with body mass, (Dynamic Equilibrium Theory and Metabolic Theory of Ecology) did not cite or consider the other main theory base (Life History Optimisation and Pace of Life theory) and vice versa.

So where to next? Craig and Dustin think that constraint driven models would benefit from including optimisation and that examining how optimisation changes in the face of absolute constraints (e.g. organisms cannot be infinitely large or small or quick, and metabolic rates cannot be zero or infinitely high) would add some necessary limits to their own work.

Importantly they also provide some examples of how we can test these theories. They focus on the long-standing traditions of quantitative genetics to simplify conceptual arguments about how metabolic rate (co)varies with other traits and fitness and predict its evolution.

For Craig and Dustin, the crux of changing to a more pluralistic approach based on life history optimisation is that we may better understand how metabolic rate evolves in response to global change or anthropogenic pressures.

This research is published in the journal Physiology.