THE DEVELOPMENT AND APPLICATION OF A REFINED METHOD FOR ESTIMATING GENE FLOW FROM ANGIOSPERM PATERNITY ANALYSIS

While gene flow is an important factor determining the genetic structure of populations, there are few studies that quantify it empirically. Paternity-exclusion analysis has recently been employed to assess the number of seeds fathered by individuals from outside of a study population (gene flow by pollen), but appropriate estimation procedures have been limited to special cases. In this report, we illustrate a general Monte Carlo method that provides an approximate maximum-likelihood estimate of gene flow by pollen from paternity-exclusion analysis. We also show that the method can be used to estimate the number of foreign gametes received by individuals in the study population. Using these methods, we estimated that 7% and 6.3% of the seeds assayed from two wild radish populations were fathered by foreign pollen (95% confidence intervals = 5.0-9.0% and 5.4-7.2%, respectively). Each population was isolated from other radish populations by at least 150 m. Estimates of the number of foreign gametes received by individuals in a population were not correlated with selected reproductive or genetic characters, which included total flower production, total fruit production, seeds per fruit, flower color, floral phenology, and number of heterozygous marker loci.