Evolution, 4th Edition

(Amelia) #1
388 CHAPTER 15

alter another gene’s expression in one developing characteristic but not another.
Such genes have been referred to as relationship quantitative trait loci (rQTL) [54].
Many genes that are known to interact with other genes may have such effects.
For example, a mutation at the S locus in the butterfly Bicyclus anynana (see Figure
6.22) reduces colored spots (“eyespots”) only on the hindwing if the insect has a
certain allele at the R locus, but has this effect on both the hindwing and the fore-
wing in the presence of a different R allele [49].
Suppose, then, that a butterfly population, with alleles S 1 and R 1 , originally has
large spots on both wings, but that selection by a new predator favors small spots
on the hindwing only. A mutation S 2 is pleiotropic: it reduces spots on both wings
(FIGURE 15.22). This mutation increases fitness, but fitness would be still higher if
the forewing were to retain large spots. A mutation R 2 that alters expression of S 2
in the forewing so that the spots are large will be advantageous. The R 2 allele thus
differentiates the two wing spots, reducing the pleiotropic effect of the S gene, and
compensating for its deleterious effect. This scenario illustrates a model in which
body parts that shared pleiotropic genes become differentiated so that they become
less genetically correlated: they become distinct, individualized characters, or mod-
ules [52]. Conversely, selection for phenotypic integration of functionally related
traits could increase pleiotropic correlations between characters.
Many studies of phenotypic variation support the ideas of modularity and phe-
notypic integration [3]. We would expect that when new functional relationships
among characteristics evolve, selection would shape new correlation patterns. For
example, serially homologous organs, such as the fore- and hindlimbs of verte-
brates, are based on similar developmental genetic pathways (e.g., the expression
of Hox genes; see Figure 15.12B). Nathan Young and colleagues found that in qua-
drupedal monkeys, the fore- and hindlimbs are rather similar in length, and the
correlations between the lengths of corresponding parts (humerus and femur;
radius and tibia; metacarpals and metatarsals) are high (FIGURE 15.23) [87]. But
apes use their fore- and hindlimbs quite differently; gibbons, for example, have
very long arms, used for swinging between branches of trees. The correlations
between corresponding bones in the fore- and hindlimbs are lower in apes than in
monkeys. Young and colleagues suggest that the reduced pleiotropic integration of
fore- and hindlimbs enabled them to evolve more independently (i.e., to become
more evolvable). Moreover, this independence may have facilitated the evolution of
the unique limbs of humans, the only species of primate in which legs are much
longer than arms.
The expression of some developmental genes corresponds to morphologi-
cal modules that are recognized by patterns of correlation. For example, digit 1
(thumb) in the hand of primates shows more independent variation, both within
and among species, than digits 2–5, which are consistently more similar to each
other [59]. Corresponding to these two apparent modules, studies of mice show

Futuyma Kirkpatrick Evolution, 4e
Sinauer Associates
Troutt Visual Services
Evolution4e_15.22.ai Date 03-13-17

(B)

S 1 R 1 S 2 R 1 S 1 R 2 S 2 R 2

FIGURE 15.22 Modularity of characters can (A)
evolve by means of changes in pleiotropic ef-
fects. Locus S affects the size of wing spots of
a hypothetical butterfly. (A) In the presence of
allele R 1 at a “relationship gene,” the alleles S 1
and S 2 have pleiotropic effects on the forewing
and hindwing; there appears to be a single
character, “spot size.” (B) The allele R 2 suppresses
the effect of S alleles on the forewing, so they
affect only the hindwing. Thus, the spot sizes are
decoupled and appear to be distinct characters.
Solid blue lines indicate pathways between loci
and traits they affect; broken lines are pathways
that are blocked or inactive. This diagram is
based on a genetic study of wing spots in the
butterfly Bicyclus anynana.

15_EVOL4E_CH15.indd 388 3/22/17 1:30 PM

Free download pdf