archipelagos, and interspecies competition. The evolution of reproductive isolation is considered, but “the genetics of speciation are the genetics of other organisms, mainly Drosophila.”

Peter and Rosemary Grant note differences between speciation in birds and speciation in Drosophila. It is significant that in birds the behavioral barriers that prevent mating evolve first, whereas post-mating isolation typically evolves much later, perhaps after gene exchange has all but ceased. Premating isolation in birds may arise from nongenetic causes, often from factors such as song, which in many groups of birds is culturally inherited through an imprinting-like process. Of the factors involved in pre-mating isolation, such as plumage, morphology, and behavior, some are under single-gene control but most are polygenetically determined.

Patterns of Evolution

The universal tree of life consists of three domains, or “empires,” bacteria, archaea, and eukarya. The three multicellular kingdoms, animals, plants, and fungi, are just 3 of the 10–12 extant major branches of the eukaryote domain. Molecular evolutionary investigations in the last decade have elucidated the large genetic diversity encompassed by the set of all eukaryotes and, hence, the reduced proportion represented by the multicellular kingdoms. The existence and great genetic heterogeneity of the archaea have been discovered by molecular evolutionists also in the last few years, and so have been most of the species and higher taxonomic groups. The reconstruction of the universal tree and the assessment of the genetic diversity of each branch are buttressed by the hypothesis of the molecular clock of evolution, which has multifarious other applications in other evolutionary studies. How good is the molecular clock?

It has been known for some time that the time variance of molecular evolutionary events is larger than would be expected if the molecular clock were a stochastic clock, like the radioactive decay of isotopes. Francisco Ayala in “Vagaries of the Molecular Clock” (47) reviews two clocks, the genes Gpdh and Sod, investigated in his laboratory. Gpdh evolves in Drosophila very slowly, at a rate of 1.1×10-10 amino acid replacements per site per year. But the rate is much faster, ˜4.5×10-10 in mammals, between Dipteran families, between animal phyla, or between plants, animals, and fungi. On the other hand, Sod evolves very fast in Drosophila, ˜16×10-10, which is also the rate in mammals and between Dipteran families; but the rate becomes much slower, 5.3×10-10, between animal phyla, and still, slower, 3.3×10-10, between plants, animals, and fungi. If one were to assume that Gpdh and Sod are good clocks and project the Drosophila rate to estimate the time of divergence of the three multicellular kingdoms, Gpdh would yield an estimate of 3,990 million years, Sod an estimate of 224 million years, both very much off the commonly accepted divergence time of ˜1,100 million years. It is unlikely that many molecular clocks are as erratic as Gpdh or Sod, but molecular clocks should be applied with caution, particularly when remote extrapolations are made.

The hypothesis of the molecular clock was originally predicated on the assumption that the evolutionary replacement of one amino acid for another, or one nucleotide for another is most often of no adaptive consequence. If such assumption would obtain, the process of molecular evolution would be governed by a time-dependent stochastic process. The assumption of adaptive inconsequence seems safest in the case of synonymous nucleotide substitutions, which do not change the amino acids encoded by a gene. Jeffrey Powell and Etsuko Moriyama (48) explore a vexing problem, namely that organisms do not use alternative synonymous codons with the frequencies expected if synonymous substitutions were inconsequential. The deviations from random expectations are large in Drosophila genes and they often persist through long periods of evolution.

Powell and Moriyama (48) exclude differential mutation rates as the cause of the codon bias. Rather, they conclude that natural selection is the cause. The determining factor is the relative abundance of the tRNAs that execute the translation of genes into proteins: genes that are expressed at high rates favor codons that match those tRNAs that are more abundant.

The genes in the nucleus of plants often occur as “families” —i.e., a gene encoding a particular polypeptide may exist in several copies of more or less remote evolutionary origin. Michael Clegg, Michael P.Cummings, and Mary L.Durbin investigate “The Evolution of Plant Nuclear Genes” (49) by focusing on three gene families, rbcS, Chs, and Adh. Additional copies are recruited at different rates in these families: new Chs and rbcS genes are recruited 20 times faster than Adh genes. The multiplication of gene copies and their divergence is particularly notable for Chs genes in the evolution of flowering plants.

The evolution of Adh in monocot plants is not consistent with the molecular clock hypothesis even for synonymous nucleotide substitutions. Clegg and colleagues conclude that natural selection plays a significant role in driving the evolutionary divergence of duplicated genes. They add that new alleles often arise by intragene recombination (49).

Multigene families occur in animals as well as in plants. Notable in humans and other mammals are genes associated with the immune system, such as the MHC genes and immunoglobulin (Ig) genes. Some multigene families, in animals as in plants, arise by concerted evolution, a process that generates new genes by interlocus recombination or gene conversion. Masatoshi Nei, Xun Gu and Tatyana Sitnikova (50) raise the question whether concerted evolution may account for the MHC and Ig families, as some authors have suggested. They note that member genes of these families are often more similar to homologous genes from different species than they are to other member genes within the same species. This would not be expected if concerted evolution were the main originating process of gene multiplication within a family. Phylogenetic analyses of several MHC and Ig multigene families display patterns inconsistent with the concerted evolution hypothesis. The evidence favors the conclusion that the creation of new genes by gene duplication has repeatedly occurred in the evolutionary history of organisms. Some duplicated genes persist in the diversified descendant species for a long time; others effectively disappear, either because they are deleted or have become nonfunctional by deleterious mutations.

We are grateful to the National Academy of Sciences for the generous grant that financed the colloquium and to Kenneth Fulton and Edward Patte, and the staff of the Arnold and Mabel Beckman Center for their skill and generous assistance during the colloquium and its preparation. Special gratitude is owed to Denise Chilcote, who was responsible for the colloquium’s logistics at all stages, and for her gracious and dedicated performance. Most of all we are grateful to the speakers and their co-authors for their wonderful contribution to the colloquium and in the papers that follow. We have borrowed extensively from ref. 16 in the preparation of Dobzhansky’s biographical statement.

1. Dobzhansky, Th. (1937) Genetics and the Origin of Species (Columbia Univ. Press, New York); 2nd Ed., 1941; 3rd Ed., 1951.

2. Darwin, C. (1859) On the Origin of Species by Means of Natural Selection (Murray, London).

3. Mendel, G. (1866) Verh. Naturforsch. Vereines Abhandlungen Brünn 4, 3–47.

4. de Vries, H. (1900) Rev. Gen. Bot. 12, 257–271.

5. Provine, W.G. (1971) The Origins of Theoretical Population Genetics (Univ. of Chicago Press, Chicago).

6. Fisher, R.A. (1930) The Genetical Theory of Natural Selection (Clarendon, Oxford).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement