The amount of carbon and water needed to support floral development changes dramatically between the early stages of bud development and anthesis. In a previous study, we suggest that these changes correspond with a shift in how flowers are hydrated. Early in floral development, when sugars are being imported into a bud and water loss is low, flowers could be hydrated mostly by the phloem, but once flowers open and demand for water is greater, there is a shift to xylem-dominated hydration. This hypothesis is consistent with evidence that some species like Forsythia do not exhibit xylem maturation in the pedicel of their flowers until immediately prior to anthesis. Therefore, it is possible that some flowers not only require minimum input from the xylem but are hydraulically isolated from the stem during the winter. This would explain why ice does not propagate into buds before budburst. To better understand whether there is a shift in the dominance of the xylem and phloem in hydrating flowers and how this shift might relate to floral freezing tolerance, we conducted an in-depth analysis of changes in the anatomy and physiology of Forsythia flowers during the spring. We measured anatomical changes in the vascular tissue, estimated xylem and phloem input based a carbon balance model and used thermal imaging to examine the propagation ice into flowers at different developmental ages. We found evidence that the phloem likely plays a larger role in floral hydration during bud expansion than after anthesis, as hypothesized. The timing of this change appears to correspond with a shift in floral freezing tolerance and the ability of ice to propagate into developing buds. Taken together, these results suggest that for plants that flower early in the spring, there could be a benefit to relying more heavily on the phloem before anthesis because it would reduce the risk of bud damage to young flowers.