The Unit of Selection

There has been much debate in the scientific community as to the specific unit of selection. Basically, the contenders are: the group or species, the organism, or the gene. Advocates of group selection argue that natural selection discriminates among better- or worse-adapted groups of organisms, typically interbreeding populations or entire species. Some of these supporters state that adaptations are for the good of the population or species, and not for the individual organism's benefit. Under this paradigm, some organisms in a group may evolve altruistic or self-sacrificial behavior in order to preserve the group's good.

Another type of species selection is advocated in conjunction with Stephen Jay Gould and Niles Eldregde's theory of punctuated equilibrium. This "species sorting" is a type of natural selection taking place at the species level and causing differential survival not of individual organisms but of entire species. Eldredge argues that species sorting determines which adaptations will ultimately survive and prosper in nature.

Another group of scientists in this debate are the advocates of organism-level selection. These scientists argue that adaptations are for the good of the organism and that natural selection takes place at the organism's level. This was for a long time the most popular argument in science and is still commonly seen and presented as correct.

Finally, the last group argues for selection at the genetic level. Richard Dawkins in his book The Selfish Gene became the most famous advocate of this theory. Based on the concept of replicators, or entities that can be copied, Dawkins recognized that genes are types of replicators and argued that selection takes place at the genetic level.

The group-selectionist theory is now generally in disfavor - it has been demonstrated that "altruistic" behavior and other predictions made by the model are better explained by the replicator theory. Gould and Eldredge's theory of punctuated equilibrium is, in reality, not logically distinct from gradualism, though species sorting does sometimes occur. The organism-level position is not totally wrong, but is an incomplete explanation that fails to take into account the importance of genes. The genetic-level selection position is the most complete and most consistent with available evidence.

Genes as Replicators

In his groundbreaking work The Selfish Gene, Richard Dawkins defines a gene as belonging to a class of entities called replicators, or entities that make copies of themselves. Genes are replicators because they commonly duplicate themselves during mitosis (cell division). In the sequel to The Selfish Gene, titled The Extended Phenotype, Dawkins defines two cross-sectional "options" possessed by all replicators: they may be passive or active, and dead-end or germ-line. A passive replicator is one whose particular content has no effect on its likelihood of being copied. Conversely, an active replicator is one whose content or message affects its likelihood of being copied (and those copies' likelihoods of being copied). A dead-end replicator is one which may be copied a finite number of times, but has no expectation of a (potentially) infinite line of descendants. Conversely, a germ-line replicator is one which has the potential to give rise to an infinitely long line of descent.

Most genes, in the sense of physical DNA sequences, are active dead-end replicators. Since most physical genes are in somatic (body) cells, they may be copied a finite number of times through mitotic division, but will not be passed on to future generations. Genes in germ-line cells, the ancestors of gametes (eggs and sperm), are usually active (but may be passive) germ-line replicators that have a potential for infinite replication.

At this point, a clarification should be made. A germ-line replicator has the potential for immortality. This does not mean it will succeed in this goal and be replicated forever. In fact, it is a requirement of natural selection that some replicators with the potential for infinite copying die out as a result of maladaptation. For example, even a replicator causing great damage to an organism has the potential for infinite copying - it simply has virtually zero likelihood of realizing this potential because it is an active replicator, and its detrimental quality negatively affects its likelihood of being copied again.

The "Optimon" as the Beneficiary of Selection

Dawkins proposes that the active germ-line replicator - such as the genes found in human sex cells - is the optimon, or the unit benefiting from adaptations. Under this view, adaptations exist for the propagation not of the species nor of the individual, but of the optimon. An optimon benefits from any adaptation that is likely to increase its chances of being copied or the number of its descendants. Thus, an optimon in human sex cells clearly benefits from the survival and reproduction of its vehicle, or the body in which it finds itself. Thus, it is to the benefit of an optimon to produce successful adaptations in its vehicle.

Natural Selection of Replicators

Obviously, natural selection cannot choose replicators directly, because it does not operate directly on DNA, but rather on the phenotypic effects expressed by genes. Therefore, natural selection does in a sense select at the organism level, because it makes proxy selections based on the adaptation of the phenotypes of various competing organisms. Natural selection operates directly on phenotypic adaptations and thus indirectly on the genes responsible for those phenotypes.

Optimons obviously benefit from the survival of their vehicles, so adaptations appear to favor the survival of the individual organism. However, some optimons do not always favor the individual organism - some may confer a measure of "altruistic" behavior toward kin because they are likely to share genes with a given organism. By coding for altruistic behavior, such genes encourage their own survival in the body of the relative benefiting from the behavior. Conversely, an optimon may "prefer" to favor, say, more offspring for a given organism over that individual organism's survival. Thus the optimon ensures its own replication even at the expense of its vehicle when necessary.

Why Organisms Are Not Replicators

There are two basic reasons why organisms are not replicators, but only vehicles for groups of replicators (optimons). First, anything larger than a gene is too large to be a replicator in the vast majority of cases. Chromosomes are crossed over and recombined during meiosis in sexual organisms, so they do not replicate themselves faithfully. Genomes are unique to each individual in sexually reproducing species, and are recombined at each generation, so they do not qualify as replicators either. Therefore, if genomes are too large and temporary to qualify as replicators, so are organisms.

However, this reason leaves out a very important point: it would seem that asexually reproducing organisms are true replicators, since their genomes are copied into each new generation. But even asexual organisms are not true replicators for a more basic reason: acquired characteristics of an organism cannot be inherited.

Suppose that a female, asexually reproducing insect, such as an aphid, produces a daughter, who in her turn produces a granddaughter, etc. Suppose that the original aphid suffers an injury at some point before reproducing. This injury may persist throughout her lifetime, but will not be passed down to the next generation. Acquired characteristics are not inherited and have no effect on genes, which are passed down. The organism is not a replicator, because a copy of the organism was not made. On the other hand, an error in the insects' genomes will be passed down to all subsequent generations - it is a replicator. Because genes cause phenotypic effects, but phenotypes have no effect on genes, only genes and not phenotypes are replicators. (Note: This analogy is adapted from this edition of Richard Dawkins' The Extended Phenotype, page 97.)

Looking Further: Links and References

The following links and references will be helpful in the study of genes as replicators and natural selection of replicators.

• Books
  • The Blind Watchmaker: Why the Evidence of Evolution Reveals a Universe Without Design by Richard Dawkins
  • The Extended Phenotype by Richard Dawkins
  • The Selfish Gene by Richard Dawkins

• Websites
  • The World of Dawkins (unofficial)
  • Levels of Selection by Barry Sinervo

Biological Evolution | Cultural Evolution | Languages Philosophy | History of Evolution | Interactivities FAQ | Forum | Guestbook | References | About Us Home | Search | Index | Features | Newsflash! Site News | Our Awards | New!  Featured Links