1. ¹Laboratoire Ecologie, Systematique et Evolution, UMR CNRS 8079, Université Paris-Sud, Orsay Cedex, France
2. ²Department of Ecology and Evolutionary Biology, University of California Los Angeles, Los Angeles, CA, USA
3. ³Department of Ecology, Evolution and Natural Resources, Cook College, Rutgers University, New Brunswick, NJ, USA
4. ⁴Donald Bren School of Environmental Science and Management, University of California Santa Barbara, CA, USA
5. ⁵Institute of the Environment, University of California Los Angeles, Los Angeles, CA, USA

Correspondence: Professor VL Sork, Department of Ecology and Evolutionary Biology, University of California Los Angeles, PO Box 951606, Los Angeles, CA 90095-1606, USA. E-mail: vlsork@ucla.edu

⁶These authors contributed equally to this article.

⁷Current address: INRA-Bordeaux, UMR BIOGECO, Domaine de la Grande Ferrade-BP81, F-33883, Villenave d'Ornon, France.

Received 18 April 2006; Revised 2 February 2007; Accepted 19 March 2007; Published online 9 May 2007.

Abstract

The pollen dispersal distribution is an important element of the neighbourhood size of plant populations. Most methods aimed at estimating the dispersal curve assume that pollen dispersal is isotropic, but evidence indicates that this assumption does not hold for many plant species, particularly wind-pollinated species subject to prevailing winds during the pollination season. We propose here a method of detecting anisotropy of pollen dispersal and of gauging its intensity, based on the estimation of the differentiation of maternal pollen clouds (TWOGENER extraction), assuming that pollen dispersal is bivariate and normally distributed. We applied the new method to a case study in Quercus lobata, detecting only a modest level of anisotropy in pollen dispersal in a direction roughly similar to the prevailing wind direction. Finally, we conducted a simulation to explore the conditions under which anisotropy can be detected with this method, and we show that while anisotropy is detectable, in principle, it requires a large volume of data.