9718 - Laserfiche WebLink

Home Browse Search

4 DRAFT same chromosome. If this were the case in a series of closely related populations, one would expect the same loci to be associated in each population. Alternatively, linkage disequi- librium could be caused by selection on particular combinations of alleles at different loci. Disequilibrium could also be caused by colonization of new habitat. If relatively few individ- uals colonize a new site, the arrangement of alleles that they bring with them may persist for a short time. Continuous dispersal into a population from a differentiated source could also cause disequilibrium if historic residents of a site are compared to recent immigrants In the captive population, loci 300 and 267 appear to be statistically associated (table 2). Because it is not repeated across populations and the evidence for disequilibrium is weak, it is likely this pattern results from the sampling effort associated with founding the Willow Beach population. 3.6 Genetic variability Neis.Diversity Sample.Size Boulders 0.53 4.00 Coyote Camp 0.63 15.00 ?$ Salt Camp 0.65 17.007 Willow Beach 0.68 63.00 Table 3: Estimates of average Nei's Gene Diversity across loci for three wild-sampled pop- ulations and one captive population (Willow Beach). Sample sizes refer to the number of individuals sampled for at least one of six loci used to calculate averages Table 3 provides estimates of Nei's gene diversity, 1- Ep' (Nei; 1973), for each population considered in this study. The average diversities indicate little overall differences among wild and captive populations. Furthermore, there is little evidence for differences in diversity among wild-sampled populations. Diversity estimates are sample-size dependent. As a result/, the marginally lower values for the Boulders site can be explained by the relatively few fish sampled. Likewise, the slightly higher value for the captive population is also likely to be due to the significantly larger sample size for this group. 3.7 Genotypic differentiation Table 4 presents contingency table-based tests of differences in genotypic frequencies among populations. Because so many genotypes are present in low frequencies in microsatellite data in general and this dataset in particular, the significance of these comparisons was assessed by simulating the null-distribution using Monte Carlo simulation methods (Rousset and Raymond; 1997). Basically, this approach simulates the expected distribution of genotypes among populations assuming no difference among the population pairs. The actual data are then compared to this null-distribution and if they appear more extreme than the simulated 6