Biology Reference
In-Depth Information
number of founders increases. In the fi gure, each replicate trial is indicated
by the letter A: for example, there are fi ve different means (A) for separate
replicate trials (each with 25 runs) in which only one founder was selected.
Only four A values are shown when the number of founders drawn is two
because two of the means (A's) overlap. All fi ve means are identical when
founder number is 10 or greater (all A's overlap). In other words, as expected,
when just one individual is selected as the only founder, there is extensive
variation as to the average number of unique alleles retained in the founding
population of one individual across 25 runs in each of fi ve separate trials.
In these trials, only one of the unique alleles or both will be found in the
solitary founder, depending on whether it is homozygous or heterozygous,
respectively. Variation in unique alleles retained decreases when two founders
are selected per founding event, and even more so when three individuals
are selected. Note that once 10 founders are drawn, there is no detectable
variation as to whether both alleles will appear in the founding population.
Recall that since individuals are diploid, 10 individuals represent 20 allelic
draws, and it is apparent that in all 125 cases of drawing 10 individuals from a
source population with two alleles of equal frequency both alleles are always
passed into the founding population.
This demonstration of the obvious fact that increasing the number of
founders will increase the chances that the founding population will include
both source population alleles is supported by standard deviation values
for each trial as shown in Fig. 7.2. With 10 or more founders, standard
deviation in the number of alleles retained drops to 0, refl ecting that both
0.54
A
A
A
0.45
A
0.36
A
0.27
A
A
A
A
A
0.18
0.09
0.0
A
A
A
A
A
A
1
10 100
FOUNDERS
Fig. 7.2
Standard deviation for the trials described in
Fig. 7.1.
As in Fig. 7.1, here there are
fi ve trials (each with 25 replicate runs) for each level of number of founders, and fewer than
fi ve symbols indicates that some or all of them overlap. Standard deviation is essentially 0
for trials with 10 or more founders.
Search WWH ::
Custom Search