Input - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 23

other allele may cause major phenotypic differences (e.g., see Altshuler et
al. 2005). One can examine how variation in different factors (e.g., selfi ng
rates, distribution or number of founders, dispersal) affects changes in
SNP allele frequencies over generations for trials with different initial
frequencies of each SNP allele. Alternatively, some loci are known to carry a
large amount of allelic variation in large source populations (e.g., the major
histocompatibility complex system, self-incompatibility genes, some control
genes, repetitive sequences). In some cases, rare alleles are maintained at
random or because they are selected for (Hamilton 1982). Examples of
possible source population loci assemblages, each with a different pattern of
allelic diversity across varying numbers of loci, are provided in appendices
and with the program.
Diversity at loci does not just take the form of numerous alleles; loci
may vary in the distribution of frequencies of their alleles. Perhaps a user
is interested in what will happen to alleles of lower or rare frequency
under different founder introduction patterns, or different age-specifi c
reproduction schedules. Another may be interested in maintenance of
diversity when comparing a species with only two alleles per locus, these
alleles occurring at varying frequencies in the source population.
In many of the representative trials discussed below, our central
question is, “What is the effect of varying a particular input condition on the
preservation of genetic diversity?” As will be seen, one of the more sensitive
ways to examine this question is to use only loci with high diversity. Thus,
in many of the examples below, we use loci with 100 unique alleles of
equal frequency (frequency = 0.01) in the source population. Obviously,
this will result in different allele frequencies in the founding generation,
especially when different numbers of founders are involved. However, since
the number of replicate runs for a set of trial conditions is set by the user,
one can statistically examine averages and the degree of variation that are
produced by changing loci and founder input conditions.
For some species that are subjects of restoration projects, measures
of genetic diversity may already be in hand, and the user may want to
construct loci panels refl ecting that knowledge. For many other species,
little if any knowledge will be available, and even when some information
has been obtained, to maximize the preservation of genetic diversity it may
be best to assume that genetic variation is high, at least at a large number
of loci. Restoration models designed according to that assumption are
most conservative: if it turns out that genetic diversity in the species is less
than modeled, genetic diversity loss should be less than predicted by the
modeling. Although the majority of analyses conducted in the remainder
of this topic thus involve loci with numerous alleles, each of low frequency,
in Chapter 18 (Conclusions) we demonstrate that the population genetic