Facing the estimation of effective population size based on molecular markers: comparison of estimators

Abstract Text: We performed a simulation study of several estimators of the effective population size (N_e) (N_e based on rate of decrease in heterozygosity, N_eH ; N_e based on the temporal method, N_eT, and linkage disequilibrium-based method, N_eLD). We first focused on N_eH, which presented an increase in the variability of values over time. The distance from the mean and the median to the true N_e increased over time too. This was caused by the fixation of alleles through time due to genetic drift and changes in the distribution of allele frequencies. We compared the three estimators of N_e under scenarios of 3 and 20 bi-allelic loci. Increasing the number of loci largely improved the performance of N_eT and N_eLD. We highlight the value of N_eT and N_eLD when large numbers of bi-allelic loci are available, which is nowadays the case for SNPs markers.

Keywords:

Effective population size

Heterozygosity

Genetic drift