SEARCH

SEARCH BY CITATION

The study of dispersal has undergone a revival over the last decade. This increase in research activity has been marked by edited books (e.g. Bullock et al. 2002; Levey et al. 2002) and collections of papers devoted to the subject in ecological journals (e.g. Cain et al. 2003; Nathan 2005). This Journal of Ecology Special Feature, consisting of 12 papers, aims to examine the current state of knowledge about plant dispersal and to illustrate the major advances that have been made in this rapidly growing field over the last few years.

The key features of dispersal we asked authors to address in writing for this Special Feature are summarized in its title. Plants are generally sessile, and dispersal of seeds or other diaspores links the life cycle of an immobile individual to processes at local, landscape and biogeographic scales. One of the key themes driving the renaissance in plant dispersal research is the growing recognition that easily-observed dispersal at local scales (typically a few tens of metres) cannot be simply extrapolated to yield understanding and prediction of processes at larger scales. Studying plant dispersal across multiple scales requires the development of new statistical, simulation and mathematical models to predict dispersal more precisely and to use the better data to understand ecological processes at these scales. Authors of the papers in this Special Feature were encouraged to present their own interpretation of this general premise. Thus, the papers address a wide range of questions about dispersal across multiple scales. Despite this range, three common issues can be identified.

(1) mechanistic models or inverse modelling?

  1. Top of page
  2. (1) mechanistic models or inverse modelling?
  3. (2) simplicity or complexity?
  4. (3) dispersal or demography?
  5. References

Models of various types have contributed to progress in dispersal research. The Special Feature reports on ways in which mechanistic wind dispersal models have been extended to include more realistic representation of processes such as seed abscission and turbulence (Bohrer et al. 2008; Soons & Bullock 2008). Mechanistic modelling of dispersal by other vectors is now waking from a period of stasis, and papers in the Special Feature report new approaches to data gathering and modelling for seed dispersal by animals (Carlos & Morales 2008; Levey et al. 2008; Soons et al. 2008). While mechanistic models have became a holy grail for many dispersal researchers, inverse modelling is of more immediate use in generating dispersal kernels from seed trap (Muller-Landau et al. 2008) and genetic (Jones & Muller-Landau 2008) data. However, as these papers show, there are statistical problems with this approach. A potential bridge between the two methods is provided by an inverse modelling technique that incorporates spatial heterogeneity explicitly (Schurr et al. 2008).

(2) simplicity or complexity?

  1. Top of page
  2. (1) mechanistic models or inverse modelling?
  3. (2) simplicity or complexity?
  4. (3) dispersal or demography?
  5. References

Papers in this Special Feature conclusively demonstrate that studying plant dispersal across multiple scales requires the examination of complexities beyond the traditional simplifications. We ask, for example, whether small-scale processes affect long-distance dispersal, and conversely, do large-scale processes affect short-distance dispersal? Several papers show that conditions at the seed source, such as wind speed and plant density, affect the initiation of dispersal and so the probability of long-distance dispersal (Bohrer et al. 2008; Carlos & Morales 2008; Soons & Bullock 2008). The crucial role of environmental heterogeneity in affecting dispersal at all scales is demonstrated, both for species dispersed by animals (Levey et al. 2008) and by wind (Bohrer et al. 2008; Pounden et al. 2008; Schurr et al. 2008). The finding that simple life-history attributes cannot explain variation in key features of dispersal (Muller-Landau et al. 2008) also shows the need to add complexity in dispersal studies. Only a few simplifications examined are still supported, such as the assumptions of no inertia (Bohrer et al. 2008) and no collisions for some types of wind-dispersed seeds (Pounden et al. 2008).

(3) dispersal or demography?

  1. Top of page
  2. (1) mechanistic models or inverse modelling?
  3. (2) simplicity or complexity?
  4. (3) dispersal or demography?
  5. References

Dispersal, however important, is just one part of the life cycle. Although demography is at the core of plant population ecology, dispersal has generally been poorly integrated into this research field. The Special Feature shows how models can be used to integrate dispersal with demography to gain a better understanding of spatial population dynamics. This allows analysis of the factors governing both local persistence and regional spread, and the papers show that each is affected by demographic and dispersal processes (Hampe et al. 2008; Higgins et al. 2008; Jongejans et al. 2008; Soons & Bullock 2008; Schurr et al. 2008). At larger scales, these models demonstrate the critical importance of long-distance dispersal; even if only a tiny proportion of seeds travel long-distances, these have a disproportionate effect on large-scale processes (Jongejans et al. 2008; Soons & Bullock 2008). Equally, local dynamics are strongly affected by dispersal at smaller scales, which determines the pattern and magnitude of plant establishment (Hampe et al. 2008; Higgins et al. 2008).

These themes show that the keys to understanding, modelling and predicting plant dynamics at any scale are to check our simplifications and address the challenge of adding complexity, to use a variety of modelling methods to simulate the focal process of dispersal and the resulting patterns, and to combine these tools and insights to examine dispersal consequences more rigorously. This is an area of research of relevance to the interests of many plant ecologists and this Special Feature provides an essential overview of recent progress and current understanding.

References

  1. Top of page
  2. (1) mechanistic models or inverse modelling?
  3. (2) simplicity or complexity?
  4. (3) dispersal or demography?
  5. References