Conclusions and Discussion - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 327

species, suggesting that localized formation of genetically differentiated
populations carrying alleles of low frequency arising from founder/small
population effects may be rather common.
Effects on Evolution
Genetic diversity is the substrate on which evolution acts: no diversity, no
evolution (genostasis). Loss of unique alleles, even those of low frequency,
can thus affect the potential for evolutionary change. Rare genes that are at
low frequency because they are not selected for in current populations, or
that have been previously selected against, may take on greater functional
relevance and be selected for as environments change. Mutations of small
phenotypic effect will obviously be initially rare. But mutations of this type
are increasingly being found to be present in populations, where they can
contribute to a portion of the additive genetic variation present in traits that
affect fi tness (e.g., see Tomkins et al. 2010, and references).
Thus far, our main focus has been on the loss of population genetic
diversity through the complete loss of unique alleles that are at low
frequencies. However, as unique alleles drift towards extinction, although
still present in a population, diversity also decreases. But, even when
unique alleles achieve frequencies lower than those used in the bulk of the
NEWGARDEN analyses in this topic, expression of those alleles on which
selection can act may still exist. Consider the following example (Mettler et
al. 1988: 127). Imagine we have 10 functionally important loci, each having
one allele with frequency 0.99. Each locus also has 3 other distinct alleles,
each with frequency 0.0033 (below the 0.01 we have been modeling). By the
conventions used in the earlier analyses in this topic, this panel of 10 loci
would be considered to have very low genetic diversity. However, given
the loci just described, in a large population, 81% of the individuals in the
next generation would be homozygous for the predominant allele at all 10
loci. Thus, approximately 20% would have a heterozygote at at least one
locus (and some at more loci). If these alleles lack complete dominance
and are to some degree additive, then 20% of the next generation would
have a functionally different genotype based on low-frequency alleles just
considering these 10 loci. Across thousands of loci, the implications are even
more complex. Once a unique allele is completely lost from a population,
however, it can no longer make any contribution to levels of heterozygosity
or diversity expression, and thus variation involving that allele is lost forever
(Allendorf and Luikart 2007: 129). Loss of such unique alleles has been a
central focus of this topic.
But are the losses of function of such alleles at single loci the only, or
even the main, evolutionary/conservation consequence arising from the
genetic processes reviewed in this topic? As noted by Templeton (2006: 94),