Body size as emergent property of metabolism

Body size is not an independent variable, but an emergent property: the result of a number of inter-linked eco-physiological processes, like most other quantities that we can measure on organisms. Organisms do not have a particular body size. They are born small and grow to larger sizes during their life trajectory, while changing properties during the growth process in interaction with the environment. The use of maximum body mass of a species as an independent variable when analyzing some other trait, bypasses the important question: What factors control maximum body mass and how do these factors affect the trait of interest? We argue that the old, famous question of why weight-specific respiration decreases for increasing maximum body mass of species was difficult to explain because ecological literature typically treats body size as an independent variable. We demonstrate that Dynamic Energy Budget (DEB) theory could explain this phenomenon by treating body size as an emergent property. The question of why specific respiration decreases with increasing body size then translates to the question of why specific assimilation and/or specific maintenance would vary among species. We discuss the four parameters that control maximum body weight in the DEB theory and study how they co-vary. One of these parameters, the allocation fraction to soma, turned out to follow a beta distribution in the Add-my-Pet collection, with perplexing accuracy. We found the explanation after discovering that the supply stress, i.e. maturity maintenance times squared somatic maintenance divided by cubed assimilation, also followed a (scaled) beta distribution. The allocation fraction can be written as the ratio of somatic maintenance and assimilation for fully-grown individuals and we found that these rates turn out to follow Weibull distributions. Beta-distributions are known to result from appropriate ratios of gamma-distributed variables and we demonstrate that this also applies, to a very good approximation, for ratios of Weibull-distributed variables. We noticed similarities between Weibull distributions and allometric functions and suggest that they fit data well because many factors contribute to the underlying processes. This explains why the allocation fraction, the supply stress and some other ratios of fluxes follow beta distributions. We support our findings with empirical data.