Natural selection is the term Charles Darwin gave to what he considered the most powerful force of evolutionary change, and virtually all modern evolutionary biologists agree. In fact, the thesis that evolution is primarily driven by natural selection is sometimes called Darwinism. Unfortunately, many people misapply the term to refer to the concept of descent with modification itself, which is erroneous. Natural selection is not the same as evolution. As discussed in chapter 1, there is a conceptual difference between a phenomenon and the mechanisms or processes that bring it about.
When Darwin's friend T. H. Huxley learned of the concept of natural selection, he said, "How extremely stupid not to have thought of that!" (Huxley 1888), so obvious did the principle seem to him—after it was formulated. And indeed, it is a very basic, very powerful idea. The philosopher Daniel Dennett has called natural selection "the single best idea anyone has ever had" (Dennett 1995: 21). Because of its generality, natural selection is widely found not only in nature but also increasingly in engineering, computer programming, the design of new drugs, and other applications.
The principle is simple: generate a variety of possible solutions, and then pick the one that works best for the problem at hand. The first solution is not necessarily the best one—in fact, natural selection rarely results in even a good solution to a problem in one pass. But repeated iterations of randomly generated solutions combined with selection of the characteristics that meet (or come close to meeting) the necessary criteria result in a series of solutions that more closely approximate a good solution. Engineers attempting to design new airplane wings have used natural selection approaches; molecular biologists trying to develop new drugs have also used the approach
(Felton 2000). In living things, the problem at hand, most broadly conceived, is survival and reproduction—passing on genes to the next generation. More narrowly, the problem at hand might be withstanding a parasite, finding a nesting site, being able to attract a mate, or being able to eat bigger seeds than usual when a drought reduces the number of small seeds. What is selected for depends on what, in the organism's particular circumstances, will be conducive to its survival and reproduction. The variety of possible solutions consists of genetically based variations that allow the organism to solve the problem.
variation among members of a species is essential to natural selection, and it is common in sexually reproducing organisms. Some of these variations are obvious to us, such as differences in size, shape, or color. Other variations are invisible, such as genetically based biochemical and molecular differences that may be related to disease or parasite resistance, or the ability to digest certain foods. If the environment of a group of plants or animals presents a challenge—say, heat, aridity, a shortage of hiding places, or a new predator—the individuals that just happen to have the genetic characteristics allowing them to survive longer and reproduce in that environment are the ones most likely to pass on their genes to the next generation. The genes of these individuals increase in proportion to those of other individuals as the population reproduces itself generation after generation. The environment naturally selects those individuals with the characteristics that provide for a higher probability of survival, and thus those characteristics tend to increase in the population over time.
So, the essence of natural selection is genetic variation within a population, an environmental condition that favors some of these variations more than others, and differential reproduction (some have more offspring than others) of the individuals that happen to have the favored variations.
A classic example of natural selection followed the introduction of rabbits into Australia—an island continent where rabbits were not native. In 1859, an English immigrant, Thomas Austin, released twelve pairs of rabbits so that he could go rabbit hunting. Unfortunately, except for the wedge-tailed eagle, a few large hawks, and dingoes (wild dogs)—and human hunters—rabbits have no natural enemies in Australia, and they reproduced like, well, rabbits. Within a few years, the rabbit population had expanded to such a large number that rabbits became a major pest, competing for grass with cattle, other domestic animals, and native Australian wildlife. Regions of the Australian outback that were infested with rabbits became virtual dust bowls as the little herbivores nibbled down anything that was green. How could rabbit numbers be controlled?
Officials in Australia decided to import a virus from Great Britain that was fatal to rabbits but that was not known to be hazardous to native Australian mammals. The virus produced myxomatosis, or rabbit fever, which causes death fairly rapidly. It is spread from rabbit to rabbit by fleas or other blood-sucking insects. The virus first was applied to a test population of rabbits in 1950. Results were extremely gratifying: in some areas the count of rabbits decreased from five thousand to fifty within six weeks. However, not all the rabbits were killed; some survived to reproduce. When the rabbit population rebounded, myxomatosis virus was reintroduced, but the positive effects of the first application were not repeated: many rabbits were killed, of course, but a larger percentage survived this time than had survived the first treatment. Eventually, myxomatosis virus no longer proved effective in reducing the rabbit population. Subsequently, Australians have resorted to putting up thousands of miles of rabbit-proof fencing to try to keep the rabbits out of at least some parts of the country.
How is this an example of natural selection? Consider how the three requirements outlined for natural selection were met:
1. variation: The Australian rabbit population consisted of individuals that varied genetically in their ability to withstand the virus causing myxomatosis.
2. Environmental condition: Myxomatosis virus was introduced into the environment, making some of the variations naturally present in the population of rabbits more valuable than others.
3. Differential reproduction: Rabbits that happened to have variations allowing them to survive this viral disease reproduced more than others, leaving more copies of their genes in future generations. Eventually the population of Australian rabbits consisted of individuals that were more likely to have the beneficial variation. When myxomatosis virus again was introduced into the environment, fewer rabbits were killed.
Natural selection involves adaptation: having characteristics that allow an organism to survive and reproduce in its environment. Which characteristics increase or decrease in the population through time depends on the value of the characteristic, and that depends on the particular environment—adaptation is not one size fits all. Because environments can change, it is difficult to precisely predict which characteristics will increase or decrease, though general predictions can be made. (No evolutionary biologist would predict that natural selection would produce naked mole rats in the Arctic, for example.) As a result, natural selection is sometimes defined as adaptive differential reproduction. It is differential reproduction because some individuals reproduce more or less than others. It is adaptive because the reason for the differential in reproduction has to do with a value that a trait or set of traits has in a particular environment.
Natural Selection and Chance. The myxomatosis example illustrates two important aspects of natural selection: it is dependent on the genetic variation present in the population and on the value of some of the genes in the population. Some individual rabbits just happened to have the genetically based resistance to myxomatosis virus even before the virus was introduced; the ability to tolerate the virus wasn't generated by the need to survive under tough circumstances. It is a matter of chance which particular rabbits were lucky enough to have the set of genes conferring resistance. So, is it correct to say that natural selection is a chance process?
Quite the contrary. Natural selection is the opposite of chance. It is adaptive differential reproduction: the individuals that survive to pass on their genes do so because they have genes that are helpful (or at least not negative) in a particular environment. Indeed, there are chance aspects to the production of genetic variability in a population: Mendel's laws of genetic recombination are, after all, based on probability. However, the chance elements are restricted to affecting the genetic variation on which natural selection works, not natural selection itself. If indeed evolution is driven primarily by natural selection, then evolution is not the result of chance.
Now, during the course of a species' evolution, unusual things may happen that are outside anything genetics or adaptation can affect, such as a mass extinction caused by an asteroid that strikes Earth, but such events—though they may be dramatic—are exceedingly rare. Such contingencies do not make evolution a chance phenomenon any more than your life is governed by chance because there is a 1 in 2.8 million chance that you will be struck by lightning.
Natural Selection and Perfection of Adaptation. The first batch of Australian rabbits to be exposed to myxomatosis virus died in droves, though some survived to reproduce. Why weren't the offspring of these surviving rabbits completely resistant to the disease? A lot of them died, too, though a smaller proportion than that of the parent's generation. This is because natural selection usually does not result in perfectly adapted structures or individuals. There are several reasons for this, and one has to do with the genetic basis of heredity.
Genes are the elements that control the traits of an organism. They are located on chromosomes, in the cells of organisms. Because chromosomes are paired, genes also come in pairs, and for some traits, the two genes are identical. For mammals, genes that contribute to building a four-chambered heart do not vary—or at least if there are any variants, the organisms that have them don't survive. But many genetic features do vary from individual to individual. Variation can be produced when the two genes of a pair differ, as they do for many traits. Some traits (perhaps most) are influenced by more than one gene, and similarly, one gene may have more than one effect. The nature of the genetic material and how it behaves is a major source of variation in each generation.
The rabbits that survived the first application of myxomatosis bred with one another, and because of genetic recombination, some offspring were produced that had myxomatosis resistance, and others were produced that lacked the adaptation. The latter were the ones that died in the second round when exposed to the virus. Back in Darwin's day, a contemporary of his invented a sound bite for natural selection: he called it "survival of the fittest," with fit meaning best adapted—not necessarily the biggest and strongest. Correctly understood, though, natural selection is survival of the fit enough. It is not, in fact, only the individuals who are most perfectly suited to the environment that survive; reproduction, after all, is a matter of degree, with some rabbits (or humans or spiders or oak trees) reproducing at higher than the average rate and some at lower than the average rate. As long as an individual reproduces at all, though, it is fit, even if some are fitter than others.
Furthermore, just as there is selection within the rabbit population for resistance to the virus, so there is selection among the viruses that cause myxomatosis. The only way that viruses can reproduce is in the body of a live rabbit. If the infected rabbit dies too quickly, the virus doesn't have a chance to spread. Viruses that are too virulent tend to be selected against, just as the rabbits that are too susceptible will also be selected against. The result is an evolutionary contest between host and pathogen, which reduces the probability that the rabbit species will ever be fully free of the virus but also reduces the likelihood that the virus will wipe out the rabbit species.
Another reason natural selection doesn't result in perfection of adaptation is that once there has been any evolution at all (and there has been considerable animal evolution since the appearance of the first metazoan), there are constraints on the direction in which evolution can go. As discussed elsewhere, if a vertebrate's forelimb is shaped for running, it would not be expected to become a wing at a later time; that is one kind of constraint. Another constraint is that natural selection has to work with structures and variations that are available, regardless of what sort of architecture could best do the job. If you need a guitar but all you have is a toilet seat, you could make a sort of guitar by running strings across the opening, but it wouldn't be a perfect design. The process of natural selection works more like a tinkerer than an engineer (Monod 1971), and these two specialists work quite differently.
Evolution and Tinkering. Some builders are engineers and some are tinkerers, and the way they go about constructing something differs quite a lot. An engineering approach to building a swing for little Charlie is to measure the distance from the tree branch to a few feet off the ground; to go to the hardware store to buy some chain, hardware, and a piece of wood for the bench; and to assemble the parts, using the appropriate tools: measuring devices, a drill, a screwdriver, screws, a saw, sandpaper, and paint. Charlie ends up with a really nice, sturdy swing that avoids the "down will come baby, cradle, and all" problem and that won't give him slivers in his little backside when he sits on it. A tinkerer, on the other hand, building a swing for little Mary, might look around the garage for a piece of rope, throw it over the branch to see if it is long enough, and tie it around an old tire. Little Mary has a swing, but it isn't quite the same as Charlie's. it gets the job done, but it certainly isn't an optimal design: the rope may suspend little Mary too far off the ground for her to be able to use the swing without someone to help her get into it; the rope may be frayed and break; the swing may be suspended too close to the trunk so Mary careens into it—you get the idea. The tinkering situation, in which a structural problem is solved by taking something extant that can be bent, cut, hammered, twisted, or manipulated into something that more or less works, however crudely, mirrors the process of evolution much more than do the precise procedures of an engineer. Nature is full of structures that work quite well—but it also is full of structures that just barely work, or that, if one were to imagine designing from scratch, one would certainly not have chosen the particular modification that natural selection did.
Several articles by Stephen Jay Gould have discussed the seemingly peculiar ways some organisms get some particular job done. An anglerfish has a clever "lure" resembling a wormlike creature that it waves at smaller fish to attract them close enough to eat. The lure, actually a modified dorsal fin spine, springs from its forehead (Gould 1980a). During embryological development, the panda's wrist bone is converted into a sixth digit, which forms a grasping hand out of the normal five fingers of a bear paw plus a "thumb" that is jury-rigged out of a modifiable bone (Gould 1980b). Like a tinkerer's project, it gets the job done, even if it isn't a great design. After all, natural selection is really about survival of the fit enough.
Natural selection is usually viewed as a mechanism that works on a population or sometimes on a species to produce adaptations. Natural selection can also bring about adaptation on a very large scale through adaptive radiation.
Was this article helpful?