MACROEvOLuTiON: EvOLuTiON AbOvE THE SPECiES LEvEL 521
do show several of the steps leading to modern cetacean morphology, such as reduc-
tion of the pelvis and hindlimbs, the shift in the nostrils to the top of the head, and
the greater uniformity of the teeth. Toothless baleen whales, such as the blue whale,
evolved from the toothed whales (such as today’s dolphins and sperm whale), but
intermediate extinct forms had both baleen and reduced teeth [17].
Some organisms have puzzling features that seem to call for a saltational origin,
because it is hard to see how an intermediate step from their ancestor could have
been advantageous. Turtles are a striking example. Their carapace (upper shell),
largely formed from the vertebrae and modified ribs, encloses the entire pectoral
girdle, including the scapula (shoulder blade). In all other tetrapods, the pectoral
girdle lies outside (above) the rib cage. (Check your own shoulder blades.) A com-
bination of paleontological and developmental studies has begun to show that this
difference could have evolved gradually [57, 60, 80]. In most amniotes, such as birds,
the developing ribs grow laterally and then downward; above them, a muscle plate
does the same (FIGURE 20.4A). The forelimbs (including the pectoral girdle) grow
outward from this dorsal muscle plate, in response to inductive signals from a lateral
ridge (Wolffian ridge). In turtles, however, the developing ribs grow laterally and
then stop, instead of growing downward, partly because of signals from another
external ridge (carapacial ridge), which is a novel feature in modern turtles (FIGURE
20.4B). As a result, the ribs lie above the muscle plate and the developing pecto-
ral girdle. The evolution of this developmental transition may have been easy if a
recently discovered fossil turtle is a reliable guide. The late Triassic Odontochelys sem-
itestacea, one of the oldest members of the turtle lineage yet found, had a lower shell
(plastron), but instead of a carapace, it had only standard-issue ribs. But the anterior
ribs were deflected backward, so that the pectoral girdle and forelimbs lay in front
of the rib cage, instead of above or below it (FIGURE 20.4C). If the alteration of rib
development occurred at this stage in turtle evolution, and the ribs became directed
forward later in evolution, they would lie above the pectoral girdle.
Goldschmidt could point to many mutations that cause large, discontinu-
ous changes that he envisioned might be the basis of saltational evolution. For
example, mutation of the Ultrabithorax (Ubx) gene in Drosophila transforms a fly
with halteres into a fly with two pairs of wings (see Figure 15.8). It may be tempt-
ing to think that a Ubx mutation in the ancestor of the Diptera caused the evolu-
tionary transformation of the second pair of insect wings into halteres. But the
Ubx mutation does not restore “real” hindwings; it transforms the third thoracic
segment into a replicated second segment, including a replicated set of forewings.
The ancestors of flies did not have identical second and third segments, and the
hindwings differ from forewings in all four-winged insects. Mutations that reduce
the function of this master control gene interfere with a complex developmental
pathway, and development is routed into a “default” pathway that produces the
features of the second thoracic segment (including wings). The whole system can
be shut down in a single step by turning a master switch, but that does not mean
the system came into existence by a single step. And—a critical point—this muta-
tion, like many other such “large-effect” mutations, drastically reduces survival
because it so profoundly disturbs normal development.
Certainly, mutations that have fairly large (but not huge) effects can contribute
to evolution. For example, variation within and among species in characters such as
bristle number in Drosophila is often caused by a mixture of quantitative trait loci with
both small and large effects [84]. Alleles with large effects contribute importantly to
Müllerian mimetic phenotypes in butterflies such as Heliconius (see Figures 6.23 and
13.10). Genetic analysis of the color patterns in Heliconius suggests that the evolu-
tion of one mimetic pattern from another was probably initiated by a mutation large
enough to provide substantial resemblance to a different model species, followed by
selection of alleles with smaller effects that “fine-tuned” the phenotype [3].20_EVOL4E_CH20.indd 521 3/22/17 1:44 PM