Evolution, 4th Edition

EvoluTion And dEvEloPmEnT 391

the most sensitive to the colchicine treatment. Furthermore, the results strongly

reflected evolutionary trends: salamanders have often lost postaxial digits during

evolution (FIGURE 15.25C), and frogs have repeatedly experienced preaxial digit

reduction. We do not know if postaxial reduction would ever have been advanta-

geous to any frogs, but any such evolution would have had to overcome a develop-

mental barrier.

Developmental studies are beginning to shed light on some constraints. Among

birds, swans have longer necks with more vertebrae than do ducks and most other

birds. But almost all mammals—from whales to giraffes—have seven cervical

(neck) vertebrae (see Figure 3.21). One might well suppose that a giraffe would

profit from more. However, a study of abnormal vertebrae in deceased human

fetuses and infants showed that the vertebrae have been homeotically trans-

formed: a cervical vertebra, for example, often has the shape and ribs of a thoracic

vertebra. These transformations are typically associated with malformations of the

skull and face and of the heart, lungs, and other organs, evidently due to pleiotropy

[77]. These harmful pleiotropic effects may have limited cervical vertebra evolution

in mammals.

Phenotypic Plasticity and Canalization

The correspondence between genotypic differences and phenotypic differences

depends not only on the effects of genotype, but also on environmental conditions

that may affect the developmental expression of the genotype.

The reaction norm of a genotype is the set of phenotypes that genotype is capa-

ble of expressing under different environmental conditions, and can be visualized

by plotting the genotype’s phenotypic value in two or more environments (FIGURE

15.26A–C; see also Figure 6.25). In some cases, a single genotype may produce

different phenotypes in response to environmental stimuli, a phenomenon called

Futuyma Kirkpatrick Evolution, 4e

Sinauer Associates

Troutt Visual Services

Evolution4e_15.26.ai Date 02-02-2017

E 1 E 2

z^1

Phenotype

Phenotype

2 z

Phenotype

2 z

Phenotype

2 z

(A)

(D)

E 1 E 2

z^1

Phenotype

(B)

E 1 E 2

z^1

Phenotype

(C)

Environment

Reaction norm

of a genotype

25

30

35

14 ° 21 ° 26 °

Temperature

Mean number of bristles

FIGURE 15.26 (A–C) Genotype × environment

interaction and the evolution of reaction norms.

Each line represents the reaction norm of a

genotype—its expression of a phenotypic char-

acter (the states expressed are labeled z 1 and z 2 )

in environments E 1 and E 2. The arrows indicate

the optimal phenotype in each environment.

(A) The genotypes do not differ in the effect of

environment on phenotype; there is no G×E

interaction. The optimal norm of reaction can-

not evolve in this case because no genotype

matches the arrows. (B) The effect of environ-

ment on phenotype differs among genotypes;

G×E interaction exists. The genotype with the

norm of reaction closest to the optima in E 1 and

E 2 (red line) will be fixed. New mutations that

bring the phenotype closer to the optimum

for each environment may be fixed thereafter.

(C) Selection may favor a constant pheno-

type, irrespective of environment, resulting in

canalization. (D) The number of bristles on the

abdomen of male Drosophila pseudoobscura

of ten genotypes, each reared at three differ-

ent temperatures. The variation indicates a G×E

interaction. (D after [24].)

15_EVOL4E_CH15.indd 391 3/22/17 1:30 PM