Corridors - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 299

In Fig. 17.1, the effect of removing the corridor was explored in trial
U, where the 20% long (1,028-unit-long) corridor was simply removed.
Those trials were conducted with an annual with short-distance dispersal
conditions (maximum pollen and offspring dispersal = 701 grid units), and
removing the corridor was seen to have a measurable but not pronounced
effect on population growth. But, using a dioecious perennial, what if long-
distance dispersal of pollen and offspring (maximum dispersal distance =
2,804) was involved, and the two preserves were separated only by the width
of a wide river, say, 50 units? Further, in all of the previous examples, the
founders were placed in squares, either one large one in one preserve, or
two smaller squares, one placed in each preserve. Obviously, the universe
of founder positional confi gurations has hardly been explored.
To demonstrate that changes in such configurations can have a
pronounced effect on perennial populations, the next trials compared
examine the following conditions. First, trial 5 is exactly as described earlier
(e.g., dioecious, perennial, 172 founders in one square in one preserve, 40%
high corridor, long-distance dispersal of pollen and offspring (up to 2,804
units)). Trial 9 is exactly like trial 5, except that rather than being connected
by a corridor of length 1,024 units, the two preserves are separated by a
50-unit-gap. Unlike the corridors, any offspring dispersed into that gap
will immediately be removed from further analyses. Trial a is just like trial
9, except that in a the founders are placed in two groups of 86, one in each
preserve. Further, rather than being placed in squares, the founders in each
preserve are placed in three horizontal lines (4 units between founders as
in all previous corridor trials) of 28, 30, and 28 individuals respectively,
the lines being parallel to the border facing the river, and with 100 units
separating the river from the nearest founders. This latter founder placement
is perhaps more typical of a riparian species (e.g., spicebush favors moist,
lowlands).
The results from these trials are shown in Fig. 17.15. While trial 5 has
a corridor of 1,024 units, trial 9 has a river of 50 units, which reduces the
growth of the latter population by 7.8%. Even though the two preserves are
closer in trial 9, the presence of the narrow river (lack of a longer corridor)
reduces population growth. However, when the river is present (corridor
lacking) and the founders are lined up along the “banks” in two groups,
one on either side (trial a), population growth increases by 27% compared
to the case where founders are placed in one preserve in one square 100
units from the river (trial 9). Even though there is a river with no corridor
in trial a, with the founders lining the banks, population growth is also
greater than trial 5 with a corridor but all founders placed in a square in
one preserve. Among these three trials, there are no appreciable differences
in rate of unique alleles lost (approximately one allele is lost from 2,906),
nor do major differences occur in levels of observed heterozygosity, all