Conclusions and Discussion - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 330

Suppose Locus 2 has only two additional alternate alleles (call them alleles
2a and 2b) that are of approximately equal frequency to one another.
Table 18.2 shows the founding and derived frequencies of these alleles as
determined by values in Table 18.1, and the resultant heterozygosities, in
the newly founded and derived populations X versus Y. Shifts in allele
frequencies of the indicated magnitudes result in approximate founding-
population-to-derived-population differences in heterozygosity of an
increase of approximately 10% for population X, and a decrease of 8% for
population Y.
Table 18.2 Heterozygosities at locus 2 (from Table 18.1) of the founding and derived populations,
assuming that the only two alternate alleles are at more or less equal frequency.

Frequencies of

alleles at locus 2: Locus expected

Population 2g 2a 2b heterozygosity

Founding populations X and Y 0.78 0.11 0.11 0.367

Derived population X 0.75 0.12 0.13 0.406

Derived population Y 0.80 0.10 0.10 0.340

When alleles with higher frequencies are used, do such large changes
in derived population allele frequencies, and consequent shifts in expected
heterozygosity values, occur in NEWGARDEN trial populations of the types
explored earlier in this topic in which only low-frequency alleles were used?
To examine this question, we changed and re-ran the o trial used to produce
the data depicted in Figs. 9.9 and 9.10 as follows. In Chapter 9, the o trial had
30 loci, each with 100 alleles, each allele at a frequency of 0.01. Under the
given conditions, that trial lost well over 25% of its original low-frequency
unique alleles. To investigate whether such losses of low-frequency unique
alleles translate to changes in expected heterozygosity comparable to those
shown in Table 8.2 when equal numbers of copies of more frequent alleles
are lost as o type populations grow, we replaced the 30 loci with 100 alleles
each used in Chapter 9 with a single locus with three alleles of frequencies
0.78, 0.11, and 0.11 in the source populations, equivalent to the founding
frequencies for populations X and Y in Table 18.2.
The question now becomes, can losses of copies of higher-frequency
distinct alleles affect expected heterozygosity values in derived populations
to the degree shown in Table 18.2? If expected heterozygosity values do
change to that degree in our new type o trials using alleles of higher
frequency, then losses of copies of higher-frequency unique alleles are, at
least in part, driving changes in allele frequencies similar to those shown
in Table 18.2. To answer this question, we ran 25 separate, single-run trials
of trial o with the higher-frequency alleles to examine the range of percent
differences between the founding expected heterozygosity and derived
heterozygosity values produced after just one round of mating. We chose