Laserfiche WebLink
from randomly generated data indicates significant haplotype frequency differences among <br />populations. <br />8, an analog of FST, was calculated using the program HAPLOID (Weir, 1990), <br />providing a measure of the proportion of variation distributed among populations. Values <br />range from 0 to 1, with higher values indicative of greater isolation. The hypothesis 8~ (i.e., <br />no population subdivision) was tested by jackknifing across samples. In jackknifing, one <br />population is removed and A calculated from the remaining samples. Each population is <br />removed in turn until 8 is calculated for all possible combinations. The mean and standard <br />deviation of this set of values provides the statistical test; populations exhibit significant <br />difference in allele frequenciPS if the confidence interval for 8 does not include 0. <br />Gene diversity analysis (Nei, 1987) was performed using the program REAP (McElroy <br />et al., 1992), partitioning the total average estimate of sequence divergence among all <br />individuals (total nucleotide diversity) into within- and among-population components <br />(population nucleotide diversity and nucleotide divergence, respectively). When <br />subpopulations aze connected by considerable gene exchange, most of the variation is found <br />within populations (i.e., diversity is large and divergence small). When subpopulations are <br />isolated, however, moss of the variation is partitioned among populations (i.e., diversity is <br />small and divergence large). <br />The extent of gene flow among populations was also inferred from haplotype networks <br />(Slatkin and Maddison, 1989) generated in two ways: 1) estimates of sequence divergence <br />were obtained among all possible pairwise combinations of restriction site haplotypes (Nei, <br />1987), and the resulting distance matrix was used to cluster haplotypes using the UPGMA <br />option in NTSYS (Sokol, 1992); and 2) haplotypes were grouped using variable restriction <br />sites (presence/absence) in a parsimony analysis providee~ by PAUP (Swofford, 190). <br />6 <br />