Non-random seed abscission, long-distance wind dispersal and plant migration rates


  • Merel B. Soons,

    Corresponding author
    1. Landscape Ecology Group, Department of Biology, Utrecht University, Sorbonnelaan 16, 3584 CA Utrecht, The Netherlands; and
    Search for more papers by this author
  • James M. Bullock

    1. Centre for Ecology and Hydrology, Maclean Building, Benson Lane, Crowmarsh Gifford, Wallingford Oxfordshire OX10 8BB, UK
    Search for more papers by this author

*Correspondence author. E-mail:


  • 1In plants, long-distance dispersal (LDD) is a widespread phenomenon of great importance, especially in spatial dynamics such as in metapopulations, invasions and migration. Much has become known about the mechanisms underlying LDD by wind, but selective release mechanisms have received little attention. Recent papers call for particular effort in clarification of the seed release stage of the dispersal process, which is our aim here.
  • 2We studied non-random seed release at a range of spatial scales, from the individual plant to species’ spatial dynamics. We did this by combining wind tunnel experiments, field trap data and simulation and analytical models, using two common wind-dispersed heathland plants (Calluna vulgaris and Erica cinerea) as study species.
  • 3In both species, seed release from the plant increases with increasing wind velocity. There is a minimum release threshold, above which the probability of seed release increases with increasing wind velocity. The drag caused by the wind is the motive force for seed release.
  • 4Simulations of seed dispersal with non-random seed release match measured dispersal kernels better, especially in the tails, than ‘conventional’ simulations with random seed release. Seed release during gusts gives the most realistic kernel. This result matches previous findings that seeds are released primarily at the start of gusts and during turbulent wind flow.
  • 5Dispersal kernels assuming non-random seed release had large impacts on analytically modelled population migration rates. Wavespeeds assuming seed release during gusts were, on average, more than double those assuming random seed release. Increases in wavespeeds under different seed release assumptions were due mostly to increases in the tail of the dispersal kernel.
  • 6Synthesis. This study shows how the small-scale process of a seed being released from the plant could affect the large-scale spatial dynamics of plant species in landscapes. The mechanism of non-random seed release, i.e., seed release during gusts, is an important mechanism for attaining LDD by wind. As such, non-random seed release influences landscape-scale species dynamics such as invasions and migration.