Abstract The metabolic theory of ecology (MTE) has been an important strand in ecology for almost a quarter of a century, renewing interest in the importance of body size and the role of energy. The core of the MTE is a hydrodynamic model of the vertebrate cardiovascular system that predicts allometric scaling of metabolic rate with exponents in the range 0.75 at infinite size to ~0.80 at more realistic sizes, though most studies using the model have assumed an exponent of 0.75. The model is broadly supported by data for resting and routine metabolic rate in ectothermic vertebrates and also a wide range of invertebrates with a circulatory system. Scaling in endotherms is influenced by additional factors, possibly associated with heat flow, and is essentially isometric in prokaryotes, unicellular eukaryotes, and diploblastic invertebrates. This suggests that the presence of any form of circulatory system, even one much simpler than the closed high‐pressure system that is the basis of the model, results in allometric scaling of metabolic rate, though the value of the scaling exponent varies across taxa. The temperature sensitivity of metabolism is captured by a simple Boltzmann factor, with an assumed apparent activation energy of 0.65 eV ( Q 10 ~ 2.4). Empirical data are frequently lower than this, typically in the range 0.52–0.57 eV ( Q 10 ~ 2.0–2.2). Attempts to broaden the scope of the MTE into areas such as growth, speciation, and life‐history have met with mixed success. The major use of the MTE has been to explore the consequences of the central scaling tendency for topics as diverse as migration, acoustic communication, trophic interactions, ecosystem structure, and the energetics of deep‐sea or extinct taxa. Although it cannot predict absolute metabolic rates, the MTE has been an important tool for exploring how energy flow influences ecology. Its greatest potential for future use is likely to come from building energetics into ecosystem models and in exploring potential consequences of climate change. In both cases, however, it will be important to encompass the range of empirical data for both scaling and temperature sensitivity rather than the widely assumed canonical values.
Andrew Clarke (Fri,) studied this question.