What is the primary difference between natural selection and artificial selection?

Introduction

Artificial selection is the process by which humans choose individual organisms with certain phenotypic trait values for breeding. If there is additive genetic variance for the selected trait, it will respond to the selection, that is, the trait will evolve. All of our domesticated species, including crop plants, livestock, and pets, are the products of artificial selection for desirable traits, such as seeds and fruits that do not disperse readily, increased meat and milk production, and docile behavior. The earliest artificial selection may have been unconscious, but it developed into a sophisticated science of plant and animal breeding; indeed, much of the field of quantitative genetics was developed to improve breeding programs.

The importance of artificial selection to the field of evolutionary biology dates back to Darwin, who was likely the first to use the term artificial selection in the ‘Origin of Species’ (Darwin, 1859). Darwin used the obvious evolutionary results of domesticated species to show the power of selective breeding as an analogy to natural selection. One of the earliest uses of experimental artificial selection to address evolutionary questions was by Holtorp (1944). He selectively bred Brassica plants that produced an extra cotyledon and reported an increase in frequency of plants with three and even four cotyledons in subsequent generations. Similarly, Huether (1968) was able to increase and decrease the number of corolla lobes in Linanthus through five generations of artificial selection. These early studies established that even traits that are conserved at higher taxonomic levels could evolve.

Artificial selection differs from what has been called laboratory natural selection (Rose et al., 1990) or controlled natural selection (Conner, 2003). In artificial selection the experimenter chooses specific phenotypic traits to select upon, while in controlled natural selection an environmental factor is manipulated and evolution of the populations in response to this selective agent is monitored. While artificial selection is certainly a form of experimental evolution, often the meaning of the term ‘experimental evolution’ is confined to controlled natural selection, excluding artificial selection (e.g., Kawecki et al., 2012). Because artificial selection applies a known strength and direction of selection to specific phenotypic traits, it is one of the most powerful methods available for understanding the underlying genetic variation and thus evolvability of those traits; in controlled natural selection the strength and direction of selection cannot be determined by the investigator.

Introduction

Artificial selection is distinct from natural selection in that it describes selection applied by humans in order to produce genetic change. When artificial selection is imposed, the trait or traits being selected are known, whereas with natural selection they have to be inferred. In most circumstances and unless otherwise qualified, directional selection is applied, i.e., only high-scoring individuals are favored for a quantitative trait. Artificial selection is the basic method of genetic improvement programs for crop plants or livestock (see Selective Breeding). It is also used as a tool in the laboratory to investigate the genetic properties of a trait in a species or population, for example, the magnitude of genetic variance or heritability, the possible duration of and limits to selection, and the correlations among traits, including with fitness.

3.3 Domestication

Artificial selection and inbred lines allow exploration of behavioral genetics by testing the responses of behavior to selection or to reduction of genetic variation.

Artificial selection, in scientific laboratories and in animal husbandry, has dramatic effects on behavior. Perhaps the broadest range of artificially selected behavior is seen in domestic dogs, which display a wide variety of behavioral attributes. These behavioral patterns are the result of selection for dogs that assist humans in work (e.g., retrievers, shepherds) or as companion animals. Most domestic livestock (such as chickens, horses, cattle, sheep, goats, and swine) reflect the results of artificial selection for manageability in confinement, ease of training, and docility (Figure 3.3). Strong artificial selection, such as that applied by animal breeders to domestic species (e.g., rabbits, chickens, dogs,12,13 cats, and cattle), can have substantial effects over three to five generations. This suggests that populations of species in new environments (such as invasive species) or species that are experiencing rapidly changing environmental conditions could have the flexibility to exhibit rapid evolutionary responses if sufficient genetic variation is present.

Artificial Selection

Artificial selection and inbred lines allow exploration of behavioral genetics by testing the responses of behavior to selection or to reduction of genetic variation. Recall from Chapter 1 that genetic variation is necessary for either natural or artificial selection to produce shifts in gene frequencies, and the only traits that can be selected are those found within the range of variation genetic variation present in the population. The potential for selection to modify a trait is assessed by measuring its heritability (see below), one estimate of genetic variation.

Key Term

An inbred line is a population in which closely related animals, such as siblings or parents and offspring, have been repeatedly mated so that nearly all genetic variation is lost. This is similar in effect to cloning.

Artificial selection, in scientific laboratories and in animal husbandry, has dramatic effects on behavior. Perhaps the broadest range of artificially selected behavior is seen in domestic dogs, which display a wide variety of behavioral attributes. These behavioral patterns are the result of selection for dogs that assist humans in work (e.g., retrievers, shepherds) or as companion animals. Most domestic livestock (such as chickens, horses, cattle, sheep, goats, and swine) reflect the results of artificial selection for manageability in confinement, ease of training, and docility (see Figure 3.9).

Abstract

Artificial selection is a breeding process in which a population of organisms is screened for some quantitative trait or traits and those individuals rated highest are used as parents for the next generation. With selection on phenotype on a single trait, the response in the population is proportional to heritability and selection differential, and can be predicted from these population parameters. Predictions can be extended to include situations in which selection is practiced on an index incorporating data on relatives and/or on multiple traits, and on genomic information on the candidates and the population. Response over the longer term depends on unknown parameters such as the distribution of the frequencies and the effects on the trait of individual genes, on effective population size, and mutation. Artificial selection experiments have, for example, been widely used to test genetic assumptions and develop extreme lines.

10.6 Artificial selection

Artificial selection is that which is practiced by man. Thereby, man determines, to a great extent, the animals that will be used to produce the next generation of offsprings. Some researchers have divided selection in farm animals into two kinds, one known as automatic and the other as a deliberate selection. Litter size in swine may be used as an illustration of the meaning of these two terms. Here automatic selection would result from differences in litter size even if parents were chosen entirely at random from all individuals available at sexual maturity. Under these conditions, there would be twice as much chance of saving offspring for breeding purposes from a litter of eight than from a litter of four. The automatic selection here differs from natural selection only to the extent that the size of the litter in which an individual is reared influences the natural selective advantage of the individual for other traits. Deliberate selection, in this example, is the term applied to selection in swine for litter size above that which was automatic. In one study by Dickerson and coworkers, involving selection in swine, most of the selection for litter size at birth was automatic and very little deliberate; however, the opportunity for deliberate selection among pigs was utilized more fully for growth rate.

Artificial Selection is a form of selection in which we actively choose the desirable traits that are passed on to the offsprings. Humans have used selective breeding long before Darwin's Postulates and the discovery of genetics. Farmers chose cattle with beneficial traits such as larger size or producing more milk, and made them breed; and although they may have known nothing about genes, they knew that the beneficial traits could be hereditary.

Farmers can breed animals in order to improve productivity, and thus, profits. For example, dairy farmers will look for the cows that can produce the most milk and only breed those cows. These cows then pass their genes that contribute to higher milk production onto their offspring, increasing productivity in each generation for the farmers. Selection based on many traits or multitraits selection in terms of progeny testing for male selection and selection indices for female selection becomes effective. A definite difference between breeds and types of farm animals within a species is proof that artificial selection has been effective in many instances. This is true, not only from the standpoint of color patterns that exist in the various breeds but also from the standpoint of differences in performances that involve certain quantitative traits. For instance, in dairy cattle, there are definite breed differences in the amount of milk produced and in butterfat percentages of the milk.

Selective Breeding

Artificial selection or selective breeding is the one of the oldest and most powerful methods in behavioral genetics. In the late 1940s, high and low preferring lines of rats were bred to drink alcohol solutions in preference to water at the University of Chile. Now there are a variety of rat and mouse lines selectively bred to differ in various responses to ethanol, including drinking preference, tolerance and withdrawal severity.42 One of the features of selective breeding is that selecting for one trait leads to the development of correlated pleiotropic differences in many other traits. For example, lines bred for differences in ethanol preference also differ in tolerance.40 Recently, some attempts have also been made to selectively breed lines differing in traits that are comorbid with addiction disorders. Anxiety, impulsivity, antisocial behavior, and depression can be modeled in rat and mouse behavioral assays, although some of these behavioral assays have a bit more than the usual level of difficulty in convincing nonbelievers they possess face validity. Given the intrinsic power of this method to assess genetic pleiotropy, it might be worthwhile developing additional lines of mice or rats that differ in some of the other traits correlated with drug abuse susceptibility in humans.

Experimental Manipulation of Population Genetics

Artificial selection involves experimental imposition of selection to answer questions about potential rates of evolutionary change. In practice, a researcher selects from a population individuals that express phenotypes of interest, for example, drought tolerance or earlier flowering. The selected individuals are interbred with the expectation that, if the trait is heritable, its phenotypic mean will differ in the offspring of the subsequent generation. From eqn [1] we can see that if the strength of selection (S) is experimentally determined and the response to selection (R) is observed, we can estimate the realized heritability (h2) of the trait of interest. An alternative approach is to expose a population to an environment (e.g., higher temperatures) in which genetic variation in fitness is expressed, such that natural selection proceeds (Bennett et al., 1992). Following several to many generations, fitness and other traits are compared between the selected and control populations to determine the extent of evolutionary change. Experimental evolution by either of these approaches may demonstrate evolutionary change in numerous traits.

Information obtained from studies of experimental evolution has improved our understanding of biotic response to climate change in numerous ways. For example, artificial selection studies have shown that it is possible that shifts in flowering time in herbaceous plant species that have been observed in nature (Parmesan and Yohe, 2003) may be due, in part, to evolutionary change (Burgess et al., 2007). In bacteria, studies where organisms evolved under experimental conditions have resulted in dramatic adaptive evolution in thermal tolerance (Lenski, 2001). It is particularly notable that adaptation proceeded in populations that initially lacked genetic variation; thus the adaptation that occurred depended entirely on newly arising mutations, and proceeded even under conditions of fluctuating selection. Mongold et al. (1999) have further shown that genes that can support adaptation beyond the presumed lethal thermal limit of E. coli fail to increase in frequency in competitive conditions. Consequently, only when populations decline in abundance does adaptation to such extreme thermal conditions proceed, a clear example of the intimate connection between demographic and genetic change. In contrast to the previous examples, some experiments to evaluate potential for adaptive evolution in response to aspects of global change have detected only limited evolutionary responses. Potvin and Tousignant (1996) manipulated both temperature and CO2 concentration and found that responses of wild mustard following seven generations in those conditions were primarily through nonadaptive plasticity. In summary, experimental evolution is a powerful approach that will continue to elucidate the rate of evolutionary change, the presence of genetic variation for traits that are under selection, and the role of genetic correlations in evolutionary change.

3.5 Genetic Basis for Selected Behaviors

Artificial selection in the laboratory has been the conventional method for detecting the presence of naturally occurring genetic variation in behavior (Ehrman and Parsons 1981). While such variation has been clearly demonstrable in virtually every behavioral selection attempted, from maze learning in rats to gravity response in Drosophila, there has been no way to identify any of the relevant genes. The behavioral differences are generally due to the contribution of many genes, usually of rather small effect, and thus refractory to conventional mapping techniques. Only their approximate number and putative interactions could be estimated from statistical analyses of selected lines and of F1, F2, and backcross progeny. The advent of QTL mapping (quantitative trait loci) using a widely distributed set of DNA markers has improved the prospects for mapping of such polygenic traits, but significant difficulties still attend the identification of the contributing gene itself.

The identification of natural variants in the foraging locus created the possibility of testing directly its response to selection under conditions likely to be relevant to its role in the wild: differences in population density. In two different experiments, strains of flies mixed for rovers and sitters were cultured at high or low larval density for over 250 generations. In both cases, the resulting high-density strains had become predominantly rover in behavior and the low-density strains predominantly sitter (Sokowloski 1998). That is, selection had eliminated the vast majority of one foraging allele from each population.

These findings contrast with those on intraspecies variants of the period gene, where the possibility of selection between alleles was inferred from their geographical distribution and made more plausible by the subtle functional differences between alleles. The conditions necessary to impose selection on the foraging locus are easier to imagine and implement, and the outcome leads to the conclusion that being genetically rover is advantageous under crowded conditions, whereas being genetically sitter is advantageous under conditions of low density.

The most common sort of selection experiments have also been the most refractory to analysis. The ability to analyze the genetic basis of the strains whose behavior results from selection for variation in many genes is likely to improve with the advent of genome-wide capabilities to measure differences in gene activity (White et al. 1999). This will likely make it feasible to identify genetic differences between selected strains or natural variants based on detecting asymmetries in the activity patterns of specific genes.

Theoretical Models of Recombination Evolution

Artificial selection experiments suggest that almost every population has enough stored genetic variability to ensure response to direct selection for changed recombination frequency (see Selection for Changed Recombination, Genetic Modifiers of Recombination, and Intraspecific Variation for the Rate of Recombination). The observed polymorphism at loci affecting recombination could be either balanced (selected) or transient. Theoretical analysis shows that under a stable environment, a panmictic population should evolve toward a minimum possible level of recombination. This can be formulated in terms of the fate of a selectively neutral modifier locus affecting recombination.

Understanding the forces maintaining sex and recombination is considered one of the most challenging problems in evolutionary theory (Michod and Levin, 1988). Shared genetic control and molecular mechanisms of DNA recombination and DNA repair across life (see Meiotic Mutants as an Analytical Tool in Recombination Studies: Overlapping of DNA Repair, Recombination, and Segregation Systems and Molecular Mechanisms of Recombination) indicate their common evolutionary origin and functional overlap in extant organisms. It could probably be supposed that repair functions played the leading role at early evolutionary stages, having provided opportunities for a large increase in the genome size and the transition from haploidy to diploidy. The latter offered the possibility of recombination repair of two-strand DNA lesions that is impossible in haploid systems. Some authors hypothesize that the subsequent stages in the evolution of recombination and sex were associated with repair alone. These explanations of repair functions were called “physiological” (Maynard Smith, 1978). Another physiological explanation of recombination in sexuals is its association with chromosome segregation (see Meiotic Mutants as an Analytical Tool in Recombination Studies: Overlapping of DNA Repair, Recombination, and Segregation Systems), although numerous examples are known, such as male Drosophila and female silkworm, in which normal segregation is associated with achiasmatic (without crossing-over) meiosis.

Combinative, or generative, hypotheses consider the main function of sex and recombination in shuffling genes. This removes negative correlation between favorable alleles at different loci, thereby increasing the efficiency of natural selection. Generative models can be classified according to the source of linkage disequilibria between selected loci: stochastic or deterministic, caused by new mutations or variation of external conditions (Kondrashov, 1993). In the 1930s, Fisher and Muller proposed that sex may be advantageous by combining beneficial mutations randomly occurring in different individuals (Hartfield et al., 2010). A complementary version of the stochastic–mutation explanation considers the role of recombination in selection against deleterious mutations (Muller, 1964). According to Muller, deleterious mutations tend to be fixed in a finite asexual population due to random drift despite purifying selection (Muller's ratchet), whereas recombination helps to stop this process. In the 1980s, a few deterministic models of selection against deleterious alleles were proposed, with the evolutionary advantage of recombination being dependent on linkage disequilibria resulted from synergistic interaction between harmful mutations produced at a high rate (Kondrashov, 1988). Other models of deterministic sources of linkage disequilibria include (1) adaptation to temporarily or spatially varying environment, (2) antagonistic species interaction (the Red Queen hypothesis), and (3) instraspecific competition (sib competition or tangled bank hypothesis) (Hamilton et al., 1990; Kondrashov, 1993; Korol et al., 1994; Feldman et al., 1997; Barton and Charlesworth, 1998; Otto and Michalakis, 1998). In most of these models, the conditions favoring increased recombination are associated with negative linkage disequilibria among selected loci owing to stringent conditions for epistasis. Some mechanisms (e.g., selection of beneficial mutations) may favor recombination in the absence of epistasis, with linkage disequilibrium caused by finite population size. The diversity of models proposed to explain the evolution of sex and recombination and their rather complicated “relationships” with the evidence from studies of different organisms, lead to the idea of pluralistic explanation (Michod and Levin, 1988; Korol et al., 1994; West et al., 1999). It suggests that different mechanisms could have played different roles in the emergence, evolution, and, especially, maintenance of recombination and sex in diverse phyletic lines of eukaryotes.