Measuring plant dispersal: an introduction to field methods and experimental design

The measurement of plant dispersal is vital for understanding plant distribution and abundance at different scales. However, dispersal is difficult to measure and there is a lack of guidance for researchers new to the subject. In this paper we provide advice on methods for measuring dispersal in the field and approaches to experimental design. First, we encourage clear exposition of the aims of the dispersal study and the ultimate use to which the data will be put (e.g. local dynamics, invasion processes, etc). We outline the types of dispersal exhibited by plants and emphasise that many species are dispersed by multiple processes, which are not necessarily related to putative adaptations. Few studies properly address the full range of processes by which a species is dispersed. We review methods for measuring plant dispersal, summarising the type of dispersal measured and problems with each method. We then outline the major questions about effort to be considered in sampling protocols and present an optimisation algorithm for designing dispersal studies given a suite of options, and biological and resource constraints. We propose and demonstrate a simulation modelling approach to comparing the data quality obtained by alternative experimental designs. Integrating simulation models with pilot studies offers a rapid route to improved estimation methods. We then discuss functions commonly fit to dispersal data and recommend caution as none is a priori the best description of the dispersal process. Finally, we call for a better description and understanding of dispersal kernels by: a more rigorous approach to designing dispersal measurement; better targeting of dispersal studies to particular questions; and achieving a deeper understanding of the mechanisms underlying dispersal, so that we can move from descriptions of pattern to a grasp of process.