Although we know that plants can respond to selection by pollinators and that for many plants pollinators are critical to successful reproduction, it is not clear a) how high spatio-temporal variation in the pollinators to a given plant in nature can lead to selection of the diversity of floral traits we see today, nor b) how plants will respond to the current "pollination crisis." The purpose of the proposed research is to provide clarification of how plants may be affected in terms of success and evolution by pollinators under a natural situation of pollinator change. This will be achieved by testing the assumed and hypothesized consequences of island colonization for plants from the mainland, using a novel study site, system and combination of research approaches. The pollination biology of plant populations will be compared between mainland and island sits, and it will be determined how patterns in reproductive success depend on the composition and frequency of pollinators and floral and reproductive-assurance traits of study populations. This will be done for each of two hummingbird-pollinated plant species - a native (Epilobium canum) and a widespread invasive (Nicotiana glauca) - that have colonized the California Channel Islands.