Actual Data versus NEWGARDEN: American Chestnut - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 262

Another way to consider these NEWGARDEN establishing populations
is to examine the data output for each new cohort. For example, if only
the latest cohort of offspring were to survive as seed (e.g., all other older
individuals, including the founders, being destroyed by fi re), those seeds
forming the future population, what would be the population characteristics
of that latest cohort of seeds? The increase in the number of individuals
produced in each new cohort for NEWGARDEN trials a, b, e, and f, which
was discussed above with regard to the entire trial populations, shows
that the pattern of increase in sequential cohorts approximates the patterns
for growth of the entire populations ( Fig. 16.1A = entire population;
B = individuals in each cohort only). Note that, compared to trial a,
sequential cohort sizes for population b increase at a slower rate after
generation 80 because of the increasing lack of sites for establishment of
new offspring, which are on average dispersed to shorter distances (but
with a longer “tail” of offspring distribution) than in trial a. This difference
in offspring establishment translates to a slower growing overall population
as seen in Fig. 16.1A.
Individual cohort F values are depicted in Fig. 16.2B. As noted above,
all four trials had a total population F value of approximately -0.07 for the
fi rst few generations after founding. F values for early individual cohorts
of all trials were well below this number (trials a, b, and f had F values of
approximately -0.13 for the fi rst few generations), indicating an excess of
heterozygosity due to small numbers of offspring in each cohort and the
high diversity of alleles in the founders and subsequent generations. Early
trial e cohorts exhibited the lowest decline in F values compared to the
founders (dropping to approximately -0.5), refl ecting the very low number
of individuals in early cohorts (Fig. 16.1) and the consequent vagaries in
allele representation. As each of these trial populations grows in number,
F values increase to approximately 0.02 to 0.03 for trials a, b, and f, and
0.0 for trial e, all at the cohort produced after 100 rounds of mating. As
stated above, these later cohort F values are in the range of values for the
entire population in later generations for these trials, and thus, after 100
generations, new cohorts of all NEWGARDEN trial populations developing
under the different respective initial input conditions appear to be exhibiting
little inbreeding or subdivision similar to the entire population.
As for unique allele retention, fi rst consider the following. In each trial,
for the nine founders, 18 alleles were drawn from the source population
for each locus, each source population locus having 100 unique alleles of
equal frequency. Given that for each locus there are 18 draws of alleles from
a pool of 100 unique alleles, the nine founders will have mostly singleton
draws of each unique allele yielding a frequency of 1 allele divided by 18
allelic “slots” (frequency = 0.056) among the founders per locus for most
unique founder alleles. Thus, allele frequencies are much lower on average