Conclusions and Discussion - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 326

Obviously, numerous low-frequency alleles do not always occur together
at single loci similar to the loci arrays used in most of the examples in this
book. Many (most?) loci will not have such a pattern of diversity, but rather
will have a lower number of distinct alleles with a more uneven distribution
of frequencies to varying degrees. Suppose that there are 30,000 structural
loci in plants on average. Each of these must have control regions often of
several kinds (shadow genes, for example; Hong et al. 2008), and, beyond
structural genes, there are many other kinds of sequences of functional
importance (e.g., sequences involved in the structuring and positioning of
chromosomes, telomeres, sequences for small RNAs). Further, minor gene
changes can sometimes result in major phenotypic effects (Prabhakar et
al. 2008). So, let us consider a plant with 70,000 distinct sequences, each
with functional signifi cance (likely a conservative estimate). Suppose that
1/1000th, or 70, of these have some of their alleles at higher frequencies,
but also, some of their alleles with lower frequencies, for example: four
alleles at frequency = 0.04; fi ve alleles with frequency = 0.03; and fi ve alleles
with frequency = 0.01 at each of the 70 loci. The 14 unique alleles of low
frequency at each locus account for 980 alleles across the 70 loci that are at
varying degrees of risk of loss, under circumstances approximating those
in the previous NEWGARDEN trials in this topic, stemming merely from
differences in the positioning of founders similar to what was seen in Figs.
18.1 and 18.2. Surely, out of those 70,000 loci, there are also an appreciable
number that have one or a few common distinct alleles with only one or
two alleles at relatively lower frequencies. As was demonstrated earlier,
low-frequency alleles that occur as singletons or with only a few other rare
alleles at a locus, but that are distributed across numerous loci, are randomly
lost at slightly reduced rates compared to when several low-frequency
unique alleles co-occur at one locus. In any case, such large numbers of
low-frequency distinct alleles are subject to variable loss at appreciable rates
just due to the idiosyncrasies of the spationumerics of founding.
Little is known about rare alleles in most organisms and populations
because they are rare. In one of the most intense surveys yet conducted,
3,000 European individuals were genotyped at more than 500,000 SNP loci
(Novembre et al. 2008). Rare alleles were common enough across loci to
contribute to the recognizable fi ne-scale differentiation of local European
populations. In another even larger study (The Wellcome Trust Case
Control Consortium 2010), approximately 19,000 humans were typed at
3,432 polymorphic copy number variants loci, and 44% of the loci included
minor alleles with frequencies of less than 0.05. Comparative genome scans
are showing that numerous species harbor enormous amounts of genetic
diversity (e.g., Clark et al. 2007), and surely some of this diversity is resident
as low-frequency alleles in isolated, newly establishing populations. It
is likely that such patterns of high genetic diversity are common among