Spatial variation in the direction of selection drives the evolution of adaptive differentiation. However, few experimental studies have examined the relative importance of different environmental factors for variation in selection and evolutionary trajectories in natural populations. Here, we combine 8 y of observational data and field experiments to assess the relative importance of mutualistic and antagonistic interactions for spatial variation in selection and short-term evolution of a genetically based floral display dimorphism in the short-lived perennial herb Primula farinosa. Natural populations of this species include two floral morphs: long-scaped plants that present their flowers well above the ground and short-scaped plants with flowers positioned close to the ground. The direction and magnitude of selection on scape morph varied among populations, and so did the frequency of the short morph (median 19%, range 0-100%; n = 69 populations). A field experiment replicated at four sites demonstrated that variation in the strength of interactions with grazers and pollinators were responsible for among-population differences in relative fitness of the two morphs. Selection exerted by grazers favored the short-scaped morph, whereas pollinator-mediated selection favored the long-scaped morph. Moreover, variation in selection among natural populations was associated with differences in morph frequency change, and the experimental removal of grazers at nine sites significantly reduced the frequency of the shortscaped morph over 8 y. The results demonstrate that spatial variation in intensity of grazing and pollination produces a selection mosaic, and that changes in biotic interactions can trigger rapid genetic changes in natural plant populations.patial variation in the intensity of biotic interactions is an integral part of the geographic mosaic model of coevolution (1, 2), and may result in divergent selection and the maintenance of genetic variation in traits influencing the strength and outcome of interactions (3, 4). However, few studies have presented quantitative estimates of spatiotemporal variation in selection on traits influencing the outcome of biotic interactions across more than a handful of populations. In plants, variation in the composition of the mutualist and antagonist assemblages may result in spatially varying selection on morphology, phenology, and lifehistory traits (e.g., 5-12). Of particular interest are traits such as floral display that may be subject to conflicting selection from mutualists and antagonists, and where the magnitude and direction of net selection should depend on the relative strength of these interactions (13)(14)(15)(16)(17)(18)(19)(20).Experimental manipulation of environmental conditions is a powerful approach to identify agents of selection and to determine the evolutionary consequences of changes in the selection regime (21,22). Experimental manipulation of pollen deposition (6,23,24) and interactions with herbivores (25-28) can be used to assess the roles of p...