Perennials - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 240

occurs when four small groups of 43 founders are used (trial m), this
difference resulting in the largest F values ( Fig. 15.10A ). Note that,
with increased inbreeding in trial m relative to trial M, heterozygosity
will decrease more rapidly in trial m, but allele frequencies will not
change. Under that scenario, since expected heterozygosity for trials
M and m are initially the same, they should remain similar, which they
do. Taken together, the above arguments suggest that the main factor
contributing to increased F values when founders are split into smaller
groups is neither increased inbreeding due to allele loss at borders nor
Wahlund effect due to subdivision; it is primarily increased inbreeding
in smaller isolated groups.
One additional issue is of interest with regard to these four trials. As
shown in Fig. 15.9B, observed heterozygosity declines most rapidly when
founders are placed in small groups of 43, regardless of the total number
or position of the founders, because of increased localized inbreeding.
However, why are expected heterozygosity values for trials M and m, when
the total number of founders is 172, very similar to one another, but very
different to the co-similar values of trials v and W, both of which have 43
founders? In other words, it appears that lowering the number of founders
results in a more rapid decline in expected heterozygosity as the population
passes through bouts of reproduction under the given conditions. Why?
As shown in Fig. 15.10B, trials appear to have rather similar rates of
unique allele loss in all populations independent of the original number of
unique alleles in the founders. However, the percentage of unique alleles
retained (calculated as: (generation alleles retained/original number of
alleles in the founders) * 100) is compared for these trials in Fig. 15.11. Note
that the percentage decline of unique alleles when populations are small (43
founders in trials v and W; lower curve) is at fi rst more rapid than when
populations are initiated with 172 founders (trials M and m; upper curve).
Trials with the same number of founders are very similar to one another in
rates of percentage loss of unique alleles, but very different from the trials
with different numbers of founders in a highly consistent manner. Rate of
percentage total unique founder allele loss is greater in small populations
because it is less likely that the small populations will have several (or a
few, or at least one) duplicates of all alleles. Failure to transmit rare alleles
to future generations is thus greater as founder populations become
smaller. Increased loss of unique alleles drives increased decline in expected
heterozygosity in smaller populations compared to larger populations, no
matter how the individuals are placed.
From the above considerations, we can conclude the following, under
the given conditions: