Aim Long-distance dispersal is important for plant population dynamics at larger spatial scales, but our understanding of this phenomenon is mostly based on computer modelling rather than field data. This paper, by combining field data and a simulation model, quantifies the fraction of the seed of the alien species Heracleum mantegazzianum that needs to disperse over a long distance for successful invasion.
Location Central Europe, Czech Republic.
Methods To assess the role of random dispersal in long-term population dynamics of the studied species, we combined longitudinal data covering 50 years of the invasion of this plant from its very start, inferred from a series of aerial photographs of 60-ha plots, with data on population dynamics at a fine scale of 10-m2 plots.
Results A simulation model based on field data indicates that the fraction of seed that is dispersed from source plants not described by the short-distance dispersal kernel ranges from 0.1 to 7.5% of the total seed set. The fraction of long-distance dispersed seed that provides the best prediction of the observed spread was significantly negatively correlated with the percentage of habitats suitable for invasion.
Main conclusions Our results indicate that the fraction of seeds that needed to be dispersed over long distances to account for the observed invasion dynamics decreased with increasing proportion of invasible habitats, indicating that the spatial pattern of propagule pressure differs in landscapes prone to invasion. Long-distance dispersal is an important component of the population dynamics of an invasive species even at relatively small scales.