386 CHAPTER 15Overview: The genetics and development of
phenotypic evolution
These examples show how evolutionary changes in the presence versus absence
of phenotypic traits and in their size, form, and location on the organism can be
understood as consequences of evolutionary alterations in gene regulation. The
differences in gene expression among species, and consequently in the proteins
that determine the form and function of groups of cells and tissues, are the proxi-
mate cause of the differences in morphology: they are the mechanisms by which
the genetic differences among species are expressed as different phenotypes.
Moreover, the genotype-phenotype relationship, or “map,” may show surpris-
ing differences among species; we saw that changes in different genetic pathways
account for pigmented wing spots in different lineages of Drosophila. (In other
cases, similar changes in the same developmental mechanism underlie conver-
gent evolution; for example, mutations in the melanocortin-1 receptor gene account
for differences in pigmentation in diverse vertebrates [see Figure 6.29].) As we
noted earlier, this perspective (sometimes called a structuralist viewpoint) does
not conflict in any way with the theory of allele frequency changes under natural
selection. Instead, it complements the perspective of natural selection as the major
cause—the ultimate cause—of phenotypic evolution (see Chapter 1, p. 7). Genetic
mutations that produce a phenotypic change that enhances fitness increase and
become fixed in populations. What evolutionary developmental biology tries to do
is tell us how those mutations act—how they produce the phenotypic alterations
that may have increased fitness. These mechanisms will help us understand phe-
nomena that we have so far described in rather abstract genetic terms. For exam-
ple, genetic correlations between traits can influence their evolution (see Chapter
6). At this time, we cannot predict very well which traits are likely to be genetically
correlated within species. A sufficient understanding of the genetic regulatory
networks that underlie such correlations will make prediction more feasible, and
so can enhance our understanding of how characters respond to natural selec-
tion. Thus, developmental biology and population genetics can meet to enhance
understanding. We now describe some of the steps evolutionary developmental
biologists are taking toward that goal.Evolvability and developmental Pathways
Understanding the genetic networks that control development may shed light on
evolvability, the ability of a characteristic to evolve, especially under directional
selection [25, 83]. For example, characters that differ in additive genetic vari-
ance differ correspondingly in evolvability, all else being equal (see Chapter 6).
But all else may not be equal, because a character may be genetically correlated
with other characters. Genetic correlations may constrain a character’s ability to
respond to directional selection, or they may enhance selection response if the
correlation points toward a new and better combination of character states. In par-
ticular, functionally interacting features may be more evolvable if they are geneti-
cally correlated. For example, the bill of a bird species that picks small insects
off leaves may more easily evolve to be shorter or longer if variation of the upper
(maxilla) and lower (mandible) parts is correlated (FIGURE 15.20). In many ani-
mal-pollinated flowers, successful pollination depends on a close match among
various dimensions of the flower, such as the length of stamens and styles. These
features’ evolvability—the response to natural selection on their length—would
be enhanced if the structures co-varied.
Are functionally interacting structures especially strongly correlated within
populations? The earliest evidence for this idea was presented by the paleontolo-
gists Everett Olson and Robert Miller, who called such a pattern “morphological15_EVOL4E_CH15.indd 386 3/22/17 1:30 PM