Groups of organisms, termed populations and species, are formed by the division of ancestral populations or species, and the descendant groups then change independently. Hence, from a long-term perspective, evolution is the descent, with modification, of different lineages from common ancestors. Thus, the history of evolution has two major components: the branching of lineages, and changes within lineages (including extinction). Initially similar species become ever more different, so that over the course of sufficient time, they may come to differ profoundly.

All forms of life, from viruses to redwoods to humans, are related by unbroken chains of descent. The hierarchically organized patterns of commonality among species--such as the common features of all primates, all mammals, all vertebrates, all eukaryotes, and all living things--reflect a history in which all living species can be traced back through time to fewer and fewer common ancestors. This history can be described by the metaphor of the phylogenetic tree (see sidebar 1). Some of this history is recorded in the fossil record, which documents simple, bacteria-like life as far back as 3.5 billion years ago, followed by a long history of diversification, modification, and extinction. The evidence for descent from common ancestors lies also in the common characteristics of living organisms, including their anatomy, embryological development, and DNA. On such grounds, for example, we can conclude that humans and apes had a relatively recent common ancestor; that a more remote common ancestor gave rise to all primates; and that successively more remote ancestors gave rise to all mammals, to all four-legged vertebrates, and to all vertebrates, including fishes.

Evolutionary theory is a body of statements about the processes of evolution that are believed to have caused the history of evolutionary events. Biological (or organic) evolution occurs as the consequence of several fundamental processes. These processes are both random and nonrandom.

Variation in the characteristics of organisms in a population originates through random mutation of DNA sequences (genes) that affect the characteristics. "Random" here means that the mutations occur irrespective of their possible consequences for survival or reproduction. Variant forms of a gene that arise by mutation are often called alleles. Genetic variation is augmented by recombination during sexual reproduction, which results in new combinations of genes. Variation is also augmented by gene flow, the input of new genes from other populations.

Evolutionary change within a population consists of a change in the proportions (frequencies) of alleles in the population. For example, the proportion of a rare allele may increase so that it completely replaces the formerly common allele. Changes in the proportions of alleles can be due to either of two processes whereby some individuals leave more descendants than others, and therefore bequeath more genes to subsequent generations. One such process, genetic drift, results from random variation in the survival and reproduction of different genotypes. In genetic drift, the frequencies of alleles fluctuate by pure chance. Eventually, one allele will replace the others (i.e., it will be fixed in the population). Genetic drift is most important when the alleles of a gene are neutral--that is, when they do not substantially differ in their effects on survival or reproduction--and it proceeds faster, the smaller the population is. Genetic drift results in evolutionary change, but not in adaptation.

The other major cause of change in the frequencies of alleles is natural selection, which is a name for any consistent (nonrandom) difference among organisms bearing different alleles or genotypes in their rate of survival or reproduction (i.e., their fitness) due to differences in one or more characteristics. In most cases, environmental circumstances affect which variant has the higher fitness. The relevant environmental circumstances depend greatly on an organism's way of life, and they include not only physical factors such as temperature, but also other species, as well as other members of its own species with which the organism competes, mates, or has other social interactions.

A common consequence of natural selection is adaptation, an improvement in the average ability of the population's members to survive and reproduce in their environment. (The word "adaptation" is also used for a feature that has evolved as a consequence of natural selection.) Natural selection tends to eliminate alleles and characteristics that reduce fitness (such as mutations that cause severe birth defects in humans and other species), and it also acts as a "sieve" that preserves and increases the abundance of combinations of genes and characteristics that increase fitness, but which would occur only rarely by chance alone. Thus, selection plays a "creative" role by making the improbable much more probable. Often the effect of selection will be the complete replacement of formerly common genes and characteristics with new ones (a process called directional selection), but under some circumstances, "balancing selection" can maintain several genetic variants indefinitely in a population (a state called genetic polymorphism, as in the case of the sickle-cell and "normal" hemoglobins found in some human populations in Africa).

Natural selection is the ultimate cause of adaptations such as eyes, hormonal controls on development, and courtship behaviors that attract mates, but it cannot produce such adaptations unless mutation and recombination generate genetic variation on which it can act. Over a long enough time, new mutations and recombinations, sorted by genetic drift or natural selection, can alter many characteristics, and can alter each characteristic both quantitatively and qualitatively. The result can be indefinitely great change, so great that a descendant species differs strikingly from its remote ancestor.

The movement of individuals among populations followed by interbreeding (i.e., gene flow) allows new genes and characteristics to spread from their population of origin throughout the species as a whole. If gene flow among different geographically separated populations is slight, different genetic changes can transpire in those populations. Because the populations experience different histories of mutation, genetic drift, and natural selection (the latter being especially likely if their environments differ), they follow different paths of change, diverging in their genetic constitutions and in the individual organisms' characteristics (geographic variation). The differences that accumulate eventually cause the different populations to be reproductively isolated: that is, if their members should encounter each other, they will not exchange genes because they will not mate with each other, or if they do, the "hybrid" offspring will be inviable or infertile. The different populations are now different species. The significance of this process of speciation is that the new species are likely to evolve independently from then on. Some may give rise to yet other species, which ultimately may become exceedingly different from one another. Successive speciation events, coupled with divergence, give rise to clusters of branches on the phylogenetic tree of living things.

Although each of the separate processes involved in evolution seems relatively simple, evolution is not as straightforward as this summary might make it appear. The various processes of evolution interact in complex ways, and each of them itself has many nuances and complexities. One gene may affect several characters, several genes may affect one character, natural selection may change in rate or even direction from year to year, or conflicting selection pressures may affect a character. When such complexities are taken into account, it can be quite difficult to predict when and how a character will evolve. Mathematical theory and computer modeling are invaluable tools for understanding how the evolution of a character is likely to proceed. A great deal of evolutionary research consists of formulating precise, often quantitative models, then testing them by experiment or observation.

It is important to distinguish between the history of evolution and the processes held to explain this history. Most biologists regard the history of evolution--the proposition that all species have descended, with modification, from common ancestors--as a fact--that is, a claim supported by such overwhelming evidence that it is accepted as true. The body of principles that describe the causal processes of evolution, such as mutation, genetic drift, and natural selection, constitutes the theory of evolution. "Theory" is used here as it is used throughout science, as in "quantum theory" or "atomic theory," to mean not mere speculation, but a well-established system or body of statements that explain a group of phenomena. Although most of the details of the history of evolution remain to be described (as is true also of human history), the statement that there has been a history of common ancestry and modification is as fully confirmed a fact as any in biology. In contrast, the theory of evolution, like all scientific theories, continues to develop as new information and ideas deepen our understanding. Evolutionary biologists have great confidence that the major causes of evolution have been identified. However, views on the relative importance of the various processes continue to change as new information adds detail and modifies our understanding. Yet, to cite evolution as a fact can invite controversy, for probably no claim in all of science evokes as much emotional opposition. Thus we include Appendix I, entitled "Evolution: Fact, Theory, Controversy."