EVoluTioN iN SPACE 199
from comparisons between the tolerance of plants grown from seeds and of adult
plants (see Figure 8.3). On the mine, the adults are more tolerant of copper than
are plants grown from seed. Far from the mine, the pattern is reversed: plants
grown from seed are more tolerant than the adults. These differences result from
powerful selection. Among the plants that germinate on the mine, only the few
that can tolerate very high levels of copper in the soil survive in each generation.
Surviving adults therefore have higher tolerance than the seeds, which have not
yet experienced selection in the current generation. Selection works in the opposite
direction off the mine. Plants that are tolerant grow more slowly in soil that has no
copper, likely because the alleles that make plants tolerant have deleterious pleio-
tropic effects (see Chapter 4). Gene flow prevents the population that is growing
on the mine from becoming fixed for copper-tolerance alleles: each generation,
pollen and seeds from the pasture are blown onto the mine. Likewise, gene flow
from the mine prevents alleles for tolerance from being completely eliminated in
the pasture nearby.
That hypothesis is confirmed by comparing the clines from the two transects on
the left and the right in Figure 8.3. The cline on the left is very steep: copper toler-
ance declines quickly just a few meters from the mine. But the cline on the right is
much more gradual. This transect goes downwind from the mine. Pollen from the
plants on the mine is blown far by the prevailing winds. That pollen continually
introduces alleles for high tolerance into the populations far downwind.
How the compromise between gene flow and local adaptation is struck depends
on the relative strengths of selection and migration. The simplest situation is
when an island (or other small region) receives migrants from a nearby continent
(or other large region). Imagine that the continent is fixed for allele A 1. Different
ecological conditions on the island give allele A 2 higher fitness there, so that the
relative fitnesses of the A 1 A 1 , A 1 A 2 , and A 2 A 2 genotypes are 1, 1 + s, and 1 + 2s.
Migrants from the continent arrive on the island at a rate m per generation.
How do allele frequencies on the island evolve? The answer depends on the
relative sizes of the selection coefficient s and the migration rate m. If selection
is much stronger than migration (m << s), then gene flow will be largely over-
whelmed and the locally adapted A 2 allele will become nearly fixed. As m grows
relative to s, the locally adapted allele will decline in frequency. In the simple case
of no dominance, the equilibrium frequency of A 2 on the island is simply(8.3)(This is an approximation that is accurate so long as m is much smaller than s.)
We can exploit theoretical results like Equation 8.3 to study selection in nat-
ural populations. If we have independent measures of allele frequencies in two
populations and the migration rate between them, we can estimate the selection
coefficient. The rock pocket mouse (Chaetodipus intermedius) lives in the desert
southwest of the United States where the landscape is a patchwork of dark fields of
lava (much like islands) surrounded by light-colored granite and sand. While most
populations of this mouse are light colored, a dark form is common on the lava,
where it is camouflaged from the owls that prey on it (FIGURE 8.10). The dark col-
oration is caused by a melanic allele at the Mc1r locus (see Figure 6.29). Researchers
estimated the migration rate, m, of mice from the granite habitats onto lava with
the indirect methods that we will discuss later in this chapter [12]. Using that value
and the observed allele frequencies at Mc1r, the selection coefficient s favoring the
melanic allele on the lava is estimated to be as large as s = 0.4. This is extremely
strong selection.p
m
s
2 ≈1–08_EVOL4E_CH08.indd 199 3/23/17 9:12 AM