The movement of genes via pollen among plants within populations and between plant populations is poorly understood. In addition, little is known of the way ecological factors affect pollinator and pollen movement. Quantification of pollen movement is necessary for understanding the evolutionary dynamics of natural populations. In a more practical context, an understanding of pollen movement is critical if we are to assess how to best preserve genetic diversity in populations of endangered species. Understanding the movement of genes via pollen will also help us assess the inherent dangers (and benefits) of "introduced" genes in agricultural plants and trees. Previous measures of pollen flow relied primarily on pollinator observations and on the dispersal of unique markers from selected plants in an experimental setting. These methods have serious drawbacks, and may not accurately assess pollen movement. A statistical technique--paternity analysis-- developed for cases of questionable human paternity and only recently adapted to plant populations--allows refinement of the measurement of effective pollen movement. Paternity analysis will be used to accurately assess pollen movement in a natural and an experimental population of the legume Lathyrus latifolius. This plant was chosen for its level of genetic variation and ease of experimental manipulation. The effects of plant density and distance on pollen movement can thus be more easily measured. This study is one of the first attempts to analyze in detail factors affecting gene movement via pollen under controlled conditions, and to integrate the experimental data with descriptions of pollen movement patterns in a natural population.
