Estimating demographic parameters from F-statistics


J. Goudet


Dept. Ecology & Evolution, Biophore, UNIL-Sorge, CH-1015, Lausanne - SUISSE



Inferring demographic parameters from genetic data is a long standing goal of population genetics.  Most students of population genetics would say that genetic data allows estimation of "average" demographic parameters, and often as a product of two variables. Typical examples are the estimation of the parameter, proportional to the number of mutants entering each generation in the population, or the number of migrants Nm from FST. In addition to estimate "average" demographic parameters, these estimations are strongly dependent on the model from which they are derived, and on equilibrium conditions.


Here, I will show that if one divides a life cycle in its different elements, and obtain the transition in gene identities (simple function of gene diversities) between the different life stage, instantaneous estimation of migration and effective population size can be obtain, providing that samples from the different life stages have been genotyped.  I will illustrate this method using data from the common shrew, Crocidura russula, where we have been able to estimate dispersal rates of males and females as well as effective population size. I will conclude by discussing possible developments of the method, which might be of interest to mycologists.