In 1922, Walter Garstang declared that ontogeny (an individual's development) does not recapitulate phylogeny (evolutionary history); rather, it creates phylogeny. Evolution is generated by heritable changes in development. “The first bird,” said Garstang, “was hatched from a reptile's egg.” Thus, when we say that the contemporary one-toed horse evolved from a five-toed ancestor, we are saying that heritable changes occurred in the differentiation of the limb mesoderm into chondrocytes during embryogenesis in the horse lineage. This view of evolution as the result of hereditary changes affecting development was lost during the 1940s, when the Modern Synthesis of population genetics and evolutionary biology formed a new framework for research in evolutionary biology.

The Modern Synthesis has been one of the greatest intellectual achievements of biology. By merging the traditions of Darwin and Mendel, evolution within a species could be explained: Diversity within a population arose from the random production of mutations, and the environment acted to select the most fit phenotypes. Those animals capable of reproducing would transmit the genes that gave them their advantage. These genes included, for example, those encoding enzymes with better rates of synthesis and globins with better oxygen-carrying capacity. It was assumed that the same kinds of changes (gene or chromosomal mutations) that caused evolution within a species also caused the evolution of new species. There would need to be an accumulation of these mutations, and a mechanism of reproductive isolation to enable them to accumulate in new ways, if a new phenotype was to be produced.

Not only could the Modern Synthesis explain evolution within a species remarkably well, it also explained medically relevant questions such as why certain alleles that seem deleterious (the hemoglobin gene variant that can result in sickle cell anemia, for example) might be selected for in certain populations. The population genetic approach to evolution was summed up by one of its foremost practitioners and theorists, Theodosius Dobzhansky, when he declared, “Evolution is a change in the genetic composition of populations. The study of the mechanisms of evolution falls within the province of population genetics” (Dobzhansky (1951).

The developmental approach to evolution was excluded from the Modern Synthesis (Hamburger 1980; Gottlieb 1992; Dietrich 1995; Gilbert et al. 1996). It was thought that population genetics could explain evolution, so morphology and development were seen to play little role in modern evolutionary theory (Adams 1991). In other words, macroevolution (the large morphological changes seen between species, classes, and phyla) could be explained by the mechanisms of microevolution, the “differential adaptive values of genotypes or deviations from random mating or both these factors acting together” (Torrey and Feduccia 1979).

The population genetics model contained some major assumptions that have now been called into question.

1.

Gradualism. The supposition that all evolutionary changes occur gradually was debated by Darwin and his friends. Thomas Huxley, for instance, accepted evolution, but he felt that Darwin had burdened his theory with an unnecessary assumption of gradualism. A century later, Eldredge and Gould (1972), Stanley (1979), and others postulated punctuated equilibrium as an alternative to the gradualism that characterized the Modern Synthesis. According to this theory, species were characterized by their morphological stability. Evolutionary changes tended to be rapid, not gradual. At the same time, molecular studies (King and Wilson 1975) showed that 99% of the DNA of humans and chimpanzees was identical, demonstrating that a small change in DNA could cause large and important morphological changes. New findings in paleontology and molecular biology prompted scientists to consider seriously the view that mutations in regulatory genes can create large changes in morphology in a relatively short time.

2.

Extrapolation of microevolution to macroevolution. The idea that accumulations of small mutations result in changes leading to new species has also been criticized. Richard Goldschmidt (1940) began his book The Material Basis of Evolution by asking the population genetic evolutionary biologists to try to explain the evolution of the following features by accumulation and selection of small mutations: hair in mammals; feathers in birds; segmentation in arthropods and vertebrates; the transformation of the gill arches into structures including aortic arches, muscles, and nerves; teeth; shells of molluscs; compound eyes; and the poison apparatus of snakes. Interestingly, both Goldschmidt and Waddington saw homeotic mutations as the kind of genetic change that could alter one structure into another and possibly create new structures or new combinations of structures. These mutations would not be in the structural genes, but in the regulatory genes. Few scientists paid attention to Goldschmidt or Waddington, however, because they were not working under the population genetics paradigm of the Modern Synthesis and because their scientific programs were suspect. (Goldschmidt did not believe in Morgan's notion of the gene as a particulate entity, and Waddington's work was misinterpreted as supporting the inheritance of acquired traits: see Gilbert 1988; 1991; Dietrich 1995.)

3.

Specificity of phenotype from genotype. Developmental biologists have found that life is more complicated than a 1:1 relationship between genotype and phenotype. Chapter 21 documents numerous cases wherein the genotype can permit any of several phenotypes to form. These cases include polyphenisms induced by predators, diet, day length, or antigenic or visual experience. Moreover, development always mediates between genotype and phenotype. The same gene can produce different phenotypes depending on the other genes that are present (Wolf 1995). A mutant gene that produces limblessness in one generation can produce only a mild thumb abnormality in the next (Freire-Maia 1975). That evolution is the result of heritable changes in development (Goldschmidt 1940) is as true for whether a fly has two or three bristles on its back as for whether an appendage is to become a fin or a limb. One way of visualizing this is to use a mathematical analogy (Gilbert et al. 1996):

Functional biology = anatomy, physiology, cell biology, gene expression

Developmental biology = δ [functional biology]/δt

Evalutionary biology = δ/[developmental biology]/δt

To go from functional biology to evolutionary biology without development is like going from displacement to acceleration without dealing with velocity.

4.

Lack of genetic similarity in disparate organisms. We have come a long way from when Ernst Mayr (1966) could state, concerning macroevolution: “Much that has been learned about gene physiology makes it evident that the search for homologous genes is quite futile except in very close relatives.” Indeed, when one considers the Hox genes, the signal transduction cascades, and the families of paracrine factors, adhesion molecules, and transcription factors, the opposite has been seen to be the case. Adult organisms may have dissimilar structures, but the genes instructing the formation of these structures are extremely similar.

The population genetics model was formulated to explain natural selection. It is based on gene differences in adults competing for reproductive advantage. The developmental genetics model is formulated to account for phylogeny—evolution above the species level. It is based on the similarities in regulatory genes that are active in embryos and larvae. We are still approaching evolution in the two ways that Darwin recognized. One can emphasize the similarities or the differences.

When the Modern Synthesis was formulated, developmental biology (and developmental genetics) were not even sciences. Embryology was left out of the Modern Synthesis, as most evolutionary biologists and geneticists felt it had nothing to contribute. However, we know now that it does. The developmental genetics approach to evolution concerns more the arrival of the fittest than the survival of the fittest.

Even critics of the Modern Synthesis (including Goldschmidt and Gould) agree that macroevolutionary change is predicated upon mutation and recombination. However, these macroevolutionary changes are in developmental regulatory genes, not the usual genes for enzymes and structural proteins; and these changes occur in embryos and larvae, not in adults competing for reproductive success (see Waddington 1953; Gilbert 1998).

Developmental biology brings to evolutionary biology, first, a new understanding about the relationships between genotypes and phenotypes, and second, a new understanding about the close genetic relationships between organisms as diverse as flies and frogs. In doing so, developmental biology complements the population genetics approach to evolutionary biology. It also highlights new questions. For instance, there can now be a population genetic approach to the regulatory genes (see Arthur 1997; Macdonald and Goldstein 1999; Zeng et al. 1999). One can also look at how paracrine factors, signal transduction pathways, and transcription factors have changed during the evolution of various phyla. Evolutionary developmental biology can also provide answers to classic evolutionary genetics questions such as these posed by mimicry and industrial melanism. The genes involved in these processes are being identified so the mechanisms of these phenomena can be explained (Koch et al. 1998; Brakefield 1998). To explain evolution, both the population genetics and the developmental genetics accounts are required.

Leaving developmental biology out of the population genetics model of evolution has left evolutionary biology open to attacks by creationists. According to Behe (1996), population genetics cannot explain the origin of structures such as the eye, so Darwinism is false.^* How could such a complicated structure have emerged by a collection of chance mutations? If a mutation caused a change in the lens, how could it be compensated for by changes in the retina? Mutations would serve only to destroy complex organs, not create them. However, once one adds development to the evolutionary synthesis, one can see how the eye can develop through induction, and that the concepts of modularity and correlated progression can readily explain such a phenomenon (Waddington 1940; Gehring 1998). Moreover, when one sees that the formation of eyes in all known phyla is based on the same signal transduction pathway, using the Pax6 gene, it is not difficult to see descent with modification forming the various types of eyes. This was much more difficult before the similarity of eye instructions had been discovered. Indeed, one study based in population genetics claimed that photoreceptors or eyes arose independently over forty times during the history of the animal kingdom (Salvini-Plawen and Mayr 1977).

In his review of evolution in 1953, J. B. S. Haldane expressed his thoughts about evolution with the following developmental analogy: “The current instar of the evolutionary theory may be defined by such books as those of Huxley, Simpson, Dobzhansky, Mayr, and Stebbins [the founders of the Modern Synthesis]. We are certainly not ready for a new moult, but signs of new organs are perhaps visible.” This recognition of developmental ideas “points forward to a broader synthesis in the future.” We have finally broken through the old pupal integument, and a new, broader, developmentally inclusive evolutionary synthesis is taking wing.