Biology Reference
In-Depth Information
to analyze the range after one round of mating because (1) that is the round
in which the greatest number of low-frequency unique alleles are lost
(approximately 15%), and thus loss of copies of alleles at higher frequencies
should also be lost, (2) the number of individuals involved is lower than for
later generations, and so allele losses resulting in random changes in allele
frequencies should be most pronounced, and (3) the effects of inbreeding
and differential maintenance of allele frequencies due to localized spatial
patterning will be of the least importance. Across these 25 replicate, single-
run trials, the percent change in expected heterozygosity from the founding
generation to the fi rst sexually produced generation ranged from a negative
28% to a positive 33%. These results imply that fl uctuating frequencies of
higher-frequency alleles due in part to variable losses of copies of alleles
for each distinct allele (e.g., 2g, 2a, and 2b here) in early generations across
different runs can contribute to changes in heterozygosity from run to run.
These changes in heterozygosity obviously exceed the range of change
demonstrated in Table 18.2.
From these results, we conclude that varying losses of copies of higher-
frequency alleles under trial conditions similar to trial o as shown in the
previous example can lead to changes in the probabilities of particular
multilocus genotypes such as those discussed for the populations described
in Table 18.1. But in Fig. 9.10 it can be seen that trial r loses low-frequency
unique alleles at a signifi cantly higher rate than trial o. Does this higher
rate of loss of such alleles in trial r translate to a correspondingly higher
rate of losses of copies of high-frequency alleles such that there will be
greater variance in their frequencies than in trial o? If the answer is yes, we
should see an increase in the standard deviation of the mean of expected
heterozygosity at loci when using higher-frequency alleles under trial
r compared to trial o conditions. It is diffi cult to demonstrate whether
this is generally the case because (1) loci that differ in allele number and
frequencies may exhibit different patterns of change of values of expected
heterozygosity as copies of alleles are randomly lost, and (2) no matter how
many replicate runs of one set of trial conditions are used, there will always
be variation in comparative trials as to the mean expected heterozygosity
and its standard deviation for the founders. With regard to issue 1, in the
o versus r type of comparative trials discussed below, we examined the
standard deviations of expected heterozygosity for a number of trials that
differed in their solitary locus allele numbers and frequencies (see allele
panels, Table 18.3) . To address issue 2, we examined the percent change
in the standard deviation of the mean expected heterozygosity from the
founding generation to generation 1 (after one round of reproduction) or
generation 3 (Table 18.3).
Search WWH ::




Custom Search