AbstractAutotrophic organisms reveal an astounding flexibility in their elemental stoichiometry, with potentially major implications on biogeochemical cycles and ecological functioning. Notwithstanding, stoichiometric regulation, and co-limitation by multiple resources in autotrophs were in the past often described by heuristic formulations. In this study, we present a mechanistic model of autotroph growth, which features two major improvements over the existing schemes. First, we introduce the concept of metabolic network independence that defines the degree of phase-locking between accessory machines. Network independence is in particular suggested to be proportional to protein synthesis capability as quantified by variable intracellular N:C. Consequently, the degree of co-limitation becomes variable, contrasting with the dichotomous debate on the use of Liebig's law or the product rule, standing for constantly low and high co-limitation, respectively. Second, we resolve dynamic protein partitioning to light harvesting, carboxylation processes, and to an arbitrary number of nutrient acquisition machineries, as well as instantaneous activity regulation of nutrient uptake. For all regulatory processes we assume growth rate optimality, here extended by an explicit consideration of indirect feed-back effects. The combination of network independence and optimal regulation displays unprecedented skill in reproducing rich stoichiometric patterns collected from a large number of published chemostat experiments. This high skill indicates (1) that the current paradigm of fixed co-limitation is a critical short-coming of conventional models, and (2) that stoichiometric flexibility in autotrophs possibly reflects an optimality strategy. Numerical experiments furthermore show that regulatory mechanisms homogenize the effect of multiple stressors. Extended optimality alleviates the effect of the most limiting resource(s) while down-regulating machineries for the less limiting ones, which induces an ubiquitous response surface of growth rate over ambient resource levels. Our approach constitutes a basis for improved mechanistic understanding and modeling of acclimative processes in autotrophic organisms. It hence may serve future experimental and theoretical investigations on the role of those processes in aquatic and terrestrial ecosystems.