524 CHAPTER 20different cis-regulatory regions of a single gene, shavenbaby, that regulates expression
of three downstream batteries of genes that are necessary for trichome development.
The shavenbaby gene itself is regulated by an array of upstream genes that determine
where it is expressed (FIGURE 20.5B). Mutation of any single downstream gene would
not suffice to alter trichome development, and mutation of the upstream genes, all of
which have multiple functions, might alter other organs as well.The Evolution of Novelty
How do major changes in characters evolve, and how do new features originate?
These questions have two distinct meanings. First, we can ask what role natu-
ral selection plays in the evolution of such changes. For instance, we may well
ask whether each step, from the slightest initial alteration of a feature to the full
complexity of form displayed by later descendants, could have been guided by
selection. Second, we can ask what the genetic and developmental bases of such
changes are (see Chapter 15). Both questions bear on the problem of how complex
characters could have evolved if their proper function depends on the mutually
adjusted form of their many components.Incipient and novel features:
Permissive conditions and natural selection
Many features are modifications of ancestral structures that have been shaped by nat-
ural selection for new functions. This principle, already recognized by Darwin, is one
of the most important in macroevolution [70], and every group of organisms presents
numerous examples. A bee’s sting is a modified ovipositor, or egg-laying organ. The
wings of auks and several other aquatic birds are used in the same way in both air
and water; in penguins, the wings have become entirely modified for underwater
flight (see Figure 3.13). Many proteins have been co-opted or modified for new func-
tions, such as a heat-shock protein and other proteins that, with little or no modifi-
cation, form the crystallin lens in vertebrate eyes (see C hapter 14). In some cases, a
feature may be an initially nonadaptive by-product of other adaptive features and has
been recruited or modified to serve an adaptive function. For instance, by excreting
nitrogenous wastes as crystalline uric acid, insects lose less water than if they excreted
ammonia or urea. Excreting uric acid is surely an adaptation, but the white color of
uric acid is not. However, pierine butterflies such as the cabbage white butterfly (Pieris
rapae) sequester uric acid in their wing scales, imparting to the wings a white color
that plays a role in thermoregulation and probably in sexual interactions.
By their behavior, animal species often affect or even determine the sources of
natural selection on morphological and physiological traits [29, 70, 83]. Aquatic
mammals would not have started to evolve adaptations for swimming unless their
ancestors had selected wet habitats to live in; insects are selected to adapt to a
plant’s toxic chemicals only if some fraction of the population chooses to eat that
species of plant [26, 89]. Behaviorally flexible species are frequently seen doing
things they are not specifically adapted for; some species of gulls, for example, will
feed on swarms of flying ants or termites, even though they normally eat aquatic
animals. Changes in behavior may often be the first step in the evolution of a new
ecological niche, to which other features become adapted [70].
Some aspects of organisms’ form and function permit or facilitate the evolution
of new characteristics. For example, decoupling the multiple functions of an ances-
tral feature relieves functional constraints, so the feature may be free to evolve in
new ways. The loss of lungs in the largest family of salamanders (Plethodontidae)
may have relieved a functional constraint on the evolution of the tongue [113]. In
other salamanders, the bones that support the tongue are also used for moving air20_EVOL4E_CH20.indd 524 3/22/17 1:44 PM