Long-distance dispersal (LDD) is central to species expansion following climate change, re-colonization of disturbed areas and control of pests. The current paradigm is that the frequency and spatial extent of LDD events are extremely difficult to predict. Here we show that mechanistic models coupling seed release and aerodynamics with turbulent transport processes provide accurate probabilistic descriptions of LDD of seeds by wind. The proposed model reliably predicts the vertical distribution of dispersed seeds of five tree species observed along a 45-m high tower in an eastern US deciduous forest. Simulations show that uplifting above the forest canopy is necessary and sufficient for LDD, hence, they provide the means to define LDD quantitatively rather than arbitrarily. Seed uplifting probability thus sets an upper bound on the probability of long-distance colonization. Uplifted yellow poplar seeds are on average lighter than seeds at the forest floor, but also include the heaviest seeds. Because uplifting probabilities are appreciable (as much as 1-5%), and tree seed crops are commonly massive, some LDD events will establish individuals that can critically affect plant dynamics on large scales.
ASJC Scopus subject areas