Pollen presentation theory (PPT) allows for a re-examination of some classic themes in pollination biology. Here, we outline its implications in the context of bee-and birdadapted species of Penstemon and Keckiella (Scrophulariaceae). PPT models the optimal schedule of pollen presentation, based on the frequency of visits by pollinators, and the capacities of those pollinators to remove and deposit pollen. High visitation rates, high removal and low deposition all favor plants that present pollen in many small doses. Dosing is achieved through gradual opening of anthers and through anthers opening only narrowly. We hypothesize that bees have higher rates of removal and lower rates of deposition than birds; therefore, bee-pollinated species should have anthers that open more gradually and less completely than bird-pollinated species. Before presenting preliminary results that affirm this prediction, we critically discuss the characterization of species by pollination syndrome. PPT sheds new light on why plants may specialize on particular pollinators. Stebbins' most effective pollinator can be recast as the pollinator that deposits more of the pollen that it removes, thereby making other visitors into conditional parasites. Pollinator shifts might occur when a pollinator with low removal and high deposition becomes abundant; the plants would then be selected to discourage their previous pollinators who are now parasites. Bird-pollination may favor anthers that open quickly and widely, thereby making bees wasteful parasites. Bee-pollination may favor anthers that open slowly and narrowly, thereby making birds ineffective pollinators. In paired comparisons of closely related species, the hummingbird-visited species were redder, had narrower or longer floral tubes, more exserted anthers and stigmas, less pronounced landing platforms, more inclined orientation, produced more nectar of a lower concentration, and had anthers that dehisce faster and more extensively. Extending pollen presentation theory to syndromesThis paper attempts to link classical themes in floral evolution to a new body of theory, using examples from the large genus Penstemon and its segregate genus Keckiella. Adaptation of flowers to pollinators traditionally has been considered in terms of characters such as corolla shape, color, and the placement of sexual organs, but 'pollen presentation theory' (PPT) (Percival 1955) suggests that the scheduling and control of pollen release to visitors should also reflect selection for successful pollination. We first outline the most relevant insights from PPT then review the classical notion of pollination syndromes, to which we add new characters derived from PPT. Turning to organisms, we consider seven pairs of plant species that probably represent independent evolutionary transitions between bee pollination and hummingbird pollination. We gauge the conformity of each P O L L E N P R E S E N TAT I O N A N