Laserfiche WebLink
Results: <br />General statistics <br />The exact test for genotypic linkage disequilibrium did not show statistically <br />significant associations (P > 0.05 after correction for multiple tests) between any pair of <br />the nine loci used in this study. Running the dataset through the program Micro-Checker <br />(van Oosterhout et al. 2004) did not reveal any scoring anomalies, but did suggest two <br />instances where null alleles may have been present: the CypG5 locus in the upper Virgin <br />River sample, and the CypG2 locus in the sample from the Dexter broodstock. Null <br />alleles are commonly associated with microsatellite markers and are the result of <br />mutations in the flanking regions that cause a mismatch between the sequence of the <br />primer and the targeted DNA fragment (Pemberton et al. 1995). We adjusted the <br />genotype frequencies for the C jpG5 locus in the upper Virgin River sample and the <br />C'ypG2 locus in the sample from the Dexter broodstock to account for the null allele <br />frequencies, and used the adjusted dataset for all further analyses. <br />Estimates of Wright's F-statistics over all five samples and loci were: Fis = -0.006 <br />0.018 (S.E.), FsT = 0.061 ± 0.020 (S.E.), and FIT = 0.055 ± 0.026 (S.E.). The variance <br />components used to estimate the F-statistics were: among sample variance 4 = 0.374 <br />or 6.11% of the total variance; and within sample variance cz, + O, , = 5.744 or 93.89 % <br />of the total variance (Figure 1 a). When the Moapa River sample was excluded, the <br />variance components were: among sample variance Cr 2 , = 0.077 or 1.26% of the total <br />variance, and within sample variance 4 + o2, = 6.042 or 98.74% of the total variance <br />(Figure 1 b). The parameter Fis is a measure of inbreeding or non-random mating within <br />populations. Values of Fis in natural populations are usually very low, but within small <br />10