TABLE 1

Results from simulation tests in haploid numbers

F-statistic
estimate
Maximum-
likelihood
estimate
No. locid/sμFσFdFμMLσMLdML
Samples at generations 0 and 4
574/1000847471766446
1015/100075611067495
154/10006843462352
250/2006124058210
500/2005715055110
1000/200561005480
2000/20054605450
Samples at generations 0 and 8
544/1000867243675324
103/10007246360331
150/10006732057230
250/2006421056180
500/200601205490
1000/20058905370
2000/20056605250
Samples at generations 0, 4 and 8
515/100076571460427
102/10006637255210
150/10006324054160
250/1006320054120
500/100591005370
1000/10059805360
2000/10057505240
• Simulated populations had a true haploid size of N = 50. The first column indicates the number of diallelic markers in each sample. Initial allele frequencies for these loci were generated randomly from a uniform distribution. Samples of haploid size 100 were drawn with replacement. Runs were discarded if either estimate of population size was > 500. The number of runs discarded is d and the total number of runs for each simulation test is s. Note that the value of s varies between tests. For each estimator, we list the mean, μ, and the standard deviation, σ, of the estimates of population size from runs in which both estimators were < 500. dF and dML are, respectively, the number of runs in which the F-statistic estimate or the maximum-likelihood estimate of N was > 500.