Abstract
Body size often strongly covaries with demography across species. Metabolism has long been invoked as the driver of these patterns, but tests of causal links between size, metabolism and demography within a species are exceedingly rare. We used 400 generations of artificial selection to evolve a 2427% size difference in the microalga Dunaliella tertiolecta . We repeatedly measured size, energy fluxes and demography across the evolved lineages. Then, we used standard metabolic theory to generate predictions of how size and demography should covary based on the scaling of energy fluxes that we measured. The size dependency of energy remained relatively consistent in time, but metabolic theory failed to predict demographic rates, which varied unpredictably in strength and even sign across generations. Classic theory holds that size affects demography via metabolism – our results suggest that both metabolism and size act separately to drive demography and that among‐species patterns may not predict within‐species processes.