Much of life history theory follows from the idea that natural selection acts on the allocation of resources to competing and independent demographic functions. This paradigm has stimulated much research on the life histories of annual plants. Models of whole-plant resource budgets that use optimal control theory predict periods of 100% vegetative and 100% reproductive growth, sometimes with periods of mixed growth. I show here that this prediction follows from the assumption of independence of the competing 'vegetative' and 'reproductive' compartments. The prediction is qualitatively unchanged even after relaxing important simplifying assumptions used in most models. Although it follows naturally from the assumptions of the models, this kind of allocation pattern is unlikely to occur in many plants, because it requires that (1) leaf and flower buds can never simultaneously be carbon sinks; and (2) organs that accompany flowers, such as internodes and bracts, can never be net sources of photosynthate. Thus while resources are doubtless important for annual plants, an exclusively resource-based perspective may be inadequate to understand the evolution of their life histories. Progress in research may require models that incorporate, or are at least phenomenologically consistent with, the basic developmental rules of angiosperms.
