Spatial Arrangement of Founders - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 127

line, grew the fastest (an 8.7% increase in numbers at generation 16 over
trial g) because these founders generated the least potential interference
competition for the establishment of offspring. In Fig. 9.5B, the v, f and g
populations lose observed heterozygosity at more or less similar rates,
with population e losing observed heterozygosity at a faster rate, at least
initially. While greater loss of unique alleles could be behind this greater
loss of observed heterozygosity for population e (it declines by only 1.3%
compared to trial g), Fig. 9.6A provides evidence that localized inbreeding
(F values; Fig. 9.6A), due to some founders being placed more distantly from
one another, is greater by 25% and thus likely driving the trend in increased
loss of observed heterozygosity. Indeed, population e loses unique alleles
at the lowest rate (Fig. 9.6B). Under the given conditions (e.g., dispersal
distance of 5 units for offspring and pollen), distribution of founders in
a long line (population e) creates less spatial competition for offspring,
resulting in a greater population growth rate, but also more localization
of alleles with less mixing, leading to more inbreeding and retention
of unique alleles (although population e retains only 2.5% more alleles
compared to trial g). These results demonstrate that variations in the spatial
arrangement of founders can result in differences in population growth,
heterozygosity, inbreeding, and unique alleles. Perhaps, for conservationists
and evolutionary biologists, the most important practical fi nding from
these examples is that just the spatial patterning of the founders can alter
population growth rates by more than 8%.
Other geometric patterns of introduction of 20 founders can generate
even greater disparities in some output parameters under otherwise
identical conditions. Figures 9.7 and 9.8 show the effects of altering the
number of grid points separating founders. These trial populations were
initiated with identical input conditions except for the geometric patterning
of the founders, and trial v included one extra generation. Population
growth rate is highest when the 20 founders are placed in two lines of 10
each with 3 grid points in between adjacent individuals (population i; Fig.
9.7A). At generation 16, that population is approximately 33% larger than
when founders are placed in two lines with no spaces between adjacent
founders (trial v). However, observed heterozygosity drops more rapidly for
population i (Fig. 9.7B), most likely because of a higher degree of inbreeding
(Fig. 9.8A) attributable to the greater spacing between individuals generating
more local breeding and less mixing given the input level of offspring and
pollen dispersal distances (5 grid units potentially involving the closest
121 grid points). Population i is more clearly exhibiting a moderate level
of population subdivision according to the guidelines given in Hartl (1988:
90). The more extreme loss in observed heterozygosity and increase in
F values for population i cannot be attributed to loss of alleles increasing
the frequency of other alleles: population i retains the greatest number of