- as natural selection (formerly also natural selection ) in the reduction of the reproductive success of certain individuals in a population, with the result that other individuals who, in retrospect, are recognizable as “more capable of survival”, reproduce more. The decisive influences are external factors of the environment , so-called selection factors . “Fitness to survive” does not mean “survival of the fittest”. It can also include cooperation and altruism . It is crucial that the genetic makeup of individuals is not passed on with the same probability.
- as sexual selection in the selection of individuals by sexual partners . It is crucial that the genetic make-up of those traits that are preferred by the sexual partners are passed on.
- as an artificial selection in a human controlled selection . It increases the reproductive success of those individuals who have the characteristics promoted by the breeder.
The designation natural selection was coined by Charles Darwin . The selection theory is an aspect of his evolution theory and was adopted as an essential part of the synthetic evolution theory in the modern evolutionary biology . Selection is one of the evolutionary factors .
The basis of natural selection is the respective probability with which individuals pass on their genetic makeup to the next generation. If the individuals of a population vary in one or more characteristics (this is usually the case in natural populations), the selection causes different reproductive successes in that some individuals can survive longer, produce more offspring, escape predators better or withstand more resistance are against diseases, etc. It is said that individuals with greater reproductive success are more fit . The selected traits that cause the higher fitness can be genetic (hereditary), or they can be environmental variants or modifications. Only the selection of hereditary traits is evolutionarily effective. The hereditary traits of the fitter individuals are then represented in the following generation with a larger proportion than in the parental generation , which inevitably means at the same time that individuals with (in their current environment) less favorable traits are represented in the following generation with less frequency. The different individuals have completely different genes only in rare exceptional cases. As a rule, the inherited differences can be traced back to minor variants of the same gene ( alleles ). (Most of these differences even affect only a single base pair: SNPs .) For population genetics, selection therefore means the same thing as directed (i.e. not random) shifting of the allele frequency in the population.
The subject of the selection are all hereditary traits that can lead to a difference in the reproductive rate. In addition to the death of the individual, z. B. those individuals are selectively disadvantaged who have a lower natural rate of reproduction or who are disadvantaged in competition with other species or individuals of other species . The decisive factor here is not the reproduction rate as such, but only the proportion of successful offspring (i.e. those who leave offspring themselves).
Sexual selection is a special case of natural selection. It arises from the competition between the sexes for reproductive partners of the opposite sex within a species. Numerous characteristics of species that cannot be explained by natural selection because they are associated with a survival disadvantage for their carrier can be explained by the fact that they increase the likelihood of their bearer mating successfully and thereby increasing the number of offspring. Sexual selection is important, for example, to explain the sexual dimorphism between the sexes, to explain the relationship between the sexes and to interpret the behavior and social systems of numerous animal species.
Artificial selection is a special case of selection, namely selection by humans. A selected trait or a combination of traits in a population is promoted by humans. Individuals who do not have these characteristics are excluded from reproduction. Since the advances in DNA sequencing in many animal and plant species, it is now also possible to carry out an artificial selection with regard to genetic characteristics. In this way, individuals with undesirable genetic traits ( genotype ) who do not appear on the phenotype because they are recessive can be excluded from reproduction.
Typically, human selection takes place in domestic animals or cultivated plants, but also in laboratories or in fished waters. It differs from natural selection in that the survival and reproduction criteria are selected by humans and are linked to a specific goal, usually for the genetic transformation or reinforcement of wanted or the suppression of unwanted properties.
Natural selection under the influence of man
Due to the overwhelming influence that the human species has on our planet, humans have recently become an important selection factor. If its influence on the population under consideration is not due to conscious selection, but rather to unintended consequences of its intervention, it is not an artificial, but a form of "natural" selection, even if the term here seems unhappy. This applies e.g. B. already for the well-known textbook example of industrial melanism in the birch moth .
A striking example of such a selection factor, the industrial deep-sea fishing . Due to the over-fishing of edible fish populations a strong is selective pressure applied to certain species of fish, where large and capable of reproduction fish are removed from the population. In this way, the survival of smaller and more precocious fish is artificially promoted. These are the only ones who have the chance to escape through the close-knit nets and then reproduce. The fish invest more energy in their reproduction than in growth. So were cod from the North-East Atlantic 60 years ago, when the hunt was on them cm on average still 95 large, today they only reach a height of 65 cm. Similarly, sexual maturity begins three years earlier today, namely at the age of six. With the help of computer models developed by Ulf Dieckmann (ecosystem researcher at the Institute for Applied Systems Analysis IASA, Laxenburg in Austria), it can be shown that fish populations can change significantly within 40 years under the pressure of the fishing fleets, for example by increasing the average size sinks. If fishing were to be stopped today, it would take model calculations up to 250 years before the fish stocks would have regained their original size distribution. This has to do with the fact that nature does not exert as much selection pressure as fisheries. According to Dieckmann, the early maturity of the fish is due to the lack of the larger fish that act as competitors. The fish find more food, grow faster and reach sexual maturity earlier.
The models for the rapid effect of selection and the resulting microevolution were simulated with experiments in various laboratories. For this purpose, fish such as guppies and ear fish, which have a relatively short generation duration, were experimented with. See also various articles by David Conover (State University of New York) and David Reznick (University of California). Both researchers were able to show that with selective fishing of the populations in the aquariums - that is, only the largest fish were removed - the fish were significantly smaller and less fertile after just a few generations. In addition, they put on less meat with the same feed as the control fish. "If you convert the evolution rate of the guppies to the development of commercially used fish, it corresponds to a few decades," says Reznick, summarizing his results. “Large-scale fishing has a genetic selection effect on stocks.” A field experiment by Reznick in Trinidad also produced comparable results (see guppies and “rapid evolution” ).
The level of selection
Within biology there is a dispute about which biological units the selection affects. A distinction must be made here between group selection , individual selection and gene selection.
In order for evolution to work, selection must be based on the hereditary traits of an individual. A selection of non-inherited properties is just as possible, but does not lead to evolution. Since genes cannot act directly, but rather require machinery or a "vehicle" to be active, a survival unit that is at least partially controlled by genes is also required: an individual. Individuals can join together to form groups, but at least as a rule biological groups have no individuality (exceptions are discussed for some special cases, especially insect states or state jellyfish ). This means that the characteristics of the group result from those of the individuals. The level on which the selection factors apply (the level of the “interactors”) is therefore always the level of the individual. Thus the individual is the object of selection. Nevertheless, depending on the question, it can make sense to focus on the gene or the group. This is less about actual differences than about different perspectives on the same thing. Population geneticists are e.g. B. particularly interested in the effect of selection on different alleles or genes. Consider z. B. the effects of a selection factor on the gene frequency, the level of the individual is irrelevant for this question, since it is a property of the population. Here it is less a question of the selection mechanism than of its effects.
The notion of an instinct for conservation , which was once influential and still widespread outside of science, is only of historical interest. It has been overcome by the synthetic theory of evolution (based on populations). Nevertheless, it is of course still useful to make comparisons at the species level if necessary. Here, too, the difference lies in the mechanisms.
The term gene selection (or gene level selection) is sometimes used today in a different sense to describe selection processes between genes within a single genome, e.g. B. in connection with Meiotic Drive .
A mathematically elegant synthesis of the described selection mechanisms is the Price equation , in which individual and group selection are taken into account.
Evolutionary developmental biology deals with the topic of how the control of the individual development of living beings ( ontogenesis ), which are subject to natural selection processes, has developed in the course of evolutionary history .
Group selection, kinship selection and mutualism
While the selection theory can in principle be applied to most morphological and behavioral traits without any problems, it is a problem to explain social behavior patterns by the theory of evolution that do not favor the respective individual himself, but other individuals. The explanation of such altruistic behavior has posed problems for theory since Darwin itself and is still intensely discussed in science today. The problem also exists when the behavior ultimately affects all, i. H. also benefits the helping individual himself. This is due to the fact that a “cheater” who only takes advantage of helping individuals without contributing anything himself would always have to be more fit than a helper. Altruism would therefore not be an evolutionarily stable strategy . As soon as a single “fraudster” emerged in the population as a result of a mutation, it would inevitably have to prevail, even if in the end everyone was worse off.
Science has developed a variety of theories as to how this problem could be solved. It is necessary to explain what the long-term evolutionary advantage of the helping strategy can be (in the short term this is not a question, since in the short term the cheater is always better off); one speaks of the “ultimatums” reasons (“why” questions). On the other hand, the prerequisites and mechanisms under which helpful behavior can arise must also be clarified; this is called the “proximate” reasons (“how” questions). Both levels of consideration must be strictly separated, otherwise confusion will inevitably arise (see also Proximate and ultimate causes of behavior ).
Classic group selection
The evolutionary biologist Vero Wynne-Edwards is the founder of a theory according to which evolution of helping behavior can be explained by selection not between individuals but between groups of them. According to his hypothesis, groups in which the individuals are considerate of each other have a higher overall fitness than groups of individuals in which each only seeks his own interest. Such 'egoist' groups are more likely to die out as a whole, e.g. B. because they do not adapt their rate of reproduction to the carrying capacity of their habitat, overexploit it and then perish. The cooperating groups remain. In the end, the “cooperation” feature prevails overall.
Within the theory of evolution, Wynne-Edwards' theory of group selection is overwhelmingly rejected today. Theorists have shown that the selection mechanism proposed by Wynne-Edwards would work. However, this requires very narrow framework conditions that are hardly conceivable in natural populations. For example, even a very moderate exchange of individuals between groups (and the gene flow that this causes ) destroys the mechanism. Only in a few cases has it been possible to identify natural populations for which the mechanism would be plausible. Usually you can explain them better with one of the competing theories.
The theory of kin selection goes back to the biologists John Maynard Smith and William D. Hamilton . According to this theory, helping behavior is only favored evolutionarily if the individuals who are being helped are also carriers of the “helper gene”, which genetically determines the helping behavior. In this case, the gene frequency of the helper gene in the next generation increases not only through the descendants of the helper himself, but also through the descendants of those who were helped. As a result, this gene can ultimately prevail in the population. When applying the hypothesis, one has to redefine the concept of fitness, since the decisive advantage of selection does not turn out to be when comparing different individuals, but is partly realized in a large number of individuals. The term “ inclusive fitness ” has become established for this . The theory of relative selection therefore suggests that the focus should not be placed on the advantage of the individuals involved, but of their respective genes. The biologist Richard Dawkins carried this out in a particularly concise way with his formulation of the “egoistic gene”, which goes back in large part to the concept of the selection of relatives.
The frequency of the helper gene can increase particularly easily in the population if helpers only help relatives (since they inevitably have a large proportion of their genes in common with the helper), ideally graded according to the degree of relationship (according to Hamilton's rule ) . This is the relative selection in the strict sense. In addition, a relative selection in the broader sense must also be taken into account. The helper can z. B. help all neighboring individuals equally if individuals of the species only rarely and little spread - in this case each neighbor is with a sufficiently high probability a relative without the helper having to be aware of this. In addition, it is theoretically entirely plausible that, according to the same mechanism, carriers of the helper gene simply help one another without having to know their relationship if they can recognize one another. After an early thought experiment, they could e.g. B. all have a green beard. After that, the mechanism entered science as the “green beard effect”.
Especially for the study of human behavior, models based on game theory have been developed , in particular on the so-called prisoner's dilemma . The concept was introduced into biology by Robert Trivers . According to this, a disposition to help in a population could also be established between unrelated individuals simply because the helper only helps those from whom he himself receives help in return (mutual help, according to the technical term reciprocity). For this hypothesis, sanctions and punishments are particularly important, through which a helper punishes those who do not “repay” the help with counter-help. In contrast to many biological approaches, the focus here is not on behaviors themselves (which can be differently determined or motivated), but interactions mediated by them, which are linked to one another in the form of an interaction matrix without the actor's motivation having to be known only plays a special role.
According to empirical studies, however, the principle of reciprocity seems to be very little widespread outside of the type of human being, so that its explanatory power is today mostly rated as low despite theoretical plausibility.
New group selection and multilevel selection
A group of evolutionary biologists brought the concept of group selection back into the debate with a slightly different definition compared to the original version, mainly in order to be able to explain cases in which, from their point of view, the model of relative selection is inadequate. The most prominent proponent of this theory is David Sloan Wilson . (His better-known namesake Edward O. Wilson has joined the theory.) According to the theory, natural selection acts on different levels at the same time: the individual, the social group and the population can thus represent levels of selection. In contrast to the classic group selection (see above), the selection works primarily within the population, not so much by selecting whole populations against each other. According to the theory, one can divide total fitness into one part, which acts between the individuals in a population, and another part, which describes the relative selective advantage of two populations against each other. The selection value for the individual is therefore the sum of the selection values on the different levels.
Various mathematical theorists have shown that the theories of kinship selection and those of multilevel selection are, for the most part, simply different ways of conceptualizing the same facts. This means that the underlying mathematical models can be converted into one another. The multilevel selection is therefore an alternative version of the relative selection (in the broader sense!).
For ecology, there is no fundamental difference between actors belonging to the same species and interactions between actors of different species, which broadens the perspective. If one individual helps another, who belongs to a different species, in the interests of both parties, this corresponds to the definition of mutualism , which is not fundamentally different from cooperation between conspecifics.
At the center of the ecological consideration are the mechanisms that can realize the advantages of a cooperative strategy (justified in a different way). Are considered z. B. species living together in groups compared to their individual relatives. Research has identified conditions under which group formation is ecologically beneficial. Is there e.g. For example, if there is a simple relationship between group size and success in aggressive encounters, it is advantageous to include non-relatives in the group. Under these conditions it is advantageous for the individual to invest in the cohesion of the group. However, (genetic and / or social) mechanisms must then be developed to exclude fraudsters. However, depending on the environmental conditions, it can pay off differently to cooperate or not.
How the selection works
Selection can affect any hereditary trait in an individual that may, directly or indirectly, affect its reproductive rate. Characteristics optimized by the effect of the selection are called adaptations .
As soon as individuals differ, a selection between them is always effective. If no other factor acts as a limiting factor, this is directed towards the rate of reproduction itself. In this case, the individuals who can produce the most offspring are selected for. Normally, however, sooner or later growth will always be limited by competition for resources or by other antagonistic relationships such as B. Predation can be limited. The selection then has the greatest effect on those characteristics which have the greatest effect on growth-limiting. This allows other characteristics. at least temporarily, remain largely excluded from the selection.
For selection to bring about evolution, the selected trait must have genetic variability. If this is missing, even strong selection pressure does not lead to evolution. As a rule, the selection always leads to a reduction in the variability (there are exceptions, especially the disruptive selection, see below). In an evolutionarily short time, strong selection inevitably leads to the fact that gradually all alleles with effects on the selected trait are fixed (i.e. either occur in all or in none of the individuals), which makes further changes to the trait increasingly difficult.
Since individuals in a population inevitably have very similar needs and are subject to selection pressure in the same direction, the decisive factor is the competition between individuals of the same species (called: intraspecific competition ). The conspecifics are, however, not only competitors in sexually reproducing species, but also mating partners. By mixing the genetic make-up during reproduction ( recombination ), adaptations acquired in different individuals within the population can be combined with one another and thus increase fitness more quickly. Initially, however, the mixing (as a gene flow ) has a homogenizing effect and thus counteracts very rapid selective adaptation. The most important effect, however, is that the variability increases, whereby the selection always gets new starting points.
The three known forms of selection - natural, sexual and artificial selection - can each appear in three types: as stabilizing, directed or disruptive selection .
On the right there is a legend for the graphics that illustrate the individual selection types ( selection types ).
Stabilizing selection (or selective stabilization / stabilization or selective retention ) takes place when the individuals in a population live under constant environmental conditions over many generations. Individuals whose characteristics are close to the mean of the population show greater fitness. Extreme phenotypes or phenotypes deviating from the mean value cannot prevail. Thus, stabilizing selection leads to less phenotypic variability.
One example is the wing design of some species of birds. Longer or shorter wings have poorer aerodynamic properties than those with ideal length, which leads to disadvantages in terms of food procurement and escape speed.
Transformative or directional selection
Transformative , dynamic , directional , shifting or directed selection occurs when the carriers of a trait that is at the edge of the population's range of traits are favored. Must z. Adjust as a population to new environmental factors, individuals are preferred whose characteristics have previously happened to fit best on the changed environment ( Präadaption ) and / or individuals whose adaptation are better suited to the new conditions. This leads to a change in the gene pool. A very strong directional selection comes about through targeted breeding.
- A population shows certain variability in the expression of a certain characteristic.
- Extreme characteristics, due to selection pressure due to changed environmental conditions
- One characteristic expression has selection advantages over the other extreme.
- Constant adaptation to changing environmental conditions.
For example, small animals with a higher escape speed have better survival rates, which can result in a selection-related increase in speed.
In disruptive (splitting) selection , the forms that occur most frequently are pushed back, e.g. B. due to parasites , predators or contagious diseases. Individuals who have rare characteristics then have an advantage (for example the particularly small and the particularly large individuals). Due to their specific characteristics, these individuals can occupy so-called ecological niches , which can bring them an evolutionary advantage, for example in obtaining food. You are favored selection. The selection pressure ensures a lower frequency of animals with “average” characteristics; those with the extreme phenotypes are selection-favored. One speaks here of polymorphism .
The technical term for the occupation of ecological niches in the case of new species formation through disruptive selection is adaptive radiation . Such a disruptive selection can lead to a bimodal frequency distribution and thus to a division of the populations into two separate species.
A classic example are the so-called Darwin's finches , the species splitting of which has already been studied by numerous evolutionary researchers. A more recent example are the mosquitoes of the genus Anopheles , which transmit malaria pathogens : After mosquito nets impregnated with insecticides were spread in Africa , the mosquitos' flight time, which was previously mostly nocturnal, was postponed to the late evening and early morning hours.
Selection at the gene level
Another classification is used for the effect of selection at the level of the genes themselves. In other words, selection at the gene level means changing the allele frequency. Various processes can take place here, each of which is important in different situations.
First of all, there are basically two ways in which selection can affect the allele frequency:
- negative selection (also: purifying selection). It consists in removing adverse alleles. The role of negative selection becomes particularly clear in connection with the neutral theory of molecular evolution . According to this theory (now widely accepted), the influence of genetic drift in small and medium-sized populations is so great that the numerous neutral mutations (i.e. those with no effect on the fitness of the phenotype, often even without effects of any kind) are often fixed by chance . So these mutations are not subject to selection. The selection is thus made primarily through the selection of harmful mutations, i. H. negative selection, noticeable. The theory does not deny the importance of positive selection, but predicts that it is much rarer in proportion.
- positive selection . It causes the selection of certain alleles.
Positive selection basically comes in two forms:
- directional selection (engl .: directional selection). In this case, an allele is read out and is consequently more frequently present in a population. This ultimately leads to the fixation of the allele. Directed selection thus reduces the variability, at the gene level one speaks of polymorphism . Indirect methods are established for the detection of previous directed selection at the gene level. So z. B. in point mutations the rate of non-synonymous substitutions compared with that of synonymous (due to the redundancy of the genetic code only certain nucleotide substitutions lead to a change in the amino acid sequence in the protein, others lead to synonymous sequences and are therefore "silent"). In the case of directed selection, this should be higher than expected at random. Another method tries to make use of the effect of directed selection on the polymorphism of alleles. If one compares the polymorphism in the genome at different loci, it should be lower in areas that are subject to more directed selection. This effect is called "genetic sweep" (roughly: genetic wiping). Genetic sweep leads to a remarkably low polymorphism in sections of the DNA of different lengths that contain the selected gene. The segment is usually longer than the gene itself, because the selection cannot start on the gene itself, but only on a DNA segment that is randomly delimited by recombination processes and that contains the gene.
- balancing selection . With this, an allele is read out differently depending on its frequency. When it is rare, it is selectively preferred. If it becomes more frequent, however, it is selectively disadvantaged. Compensating selection thus preserves the polymorphism. There are essentially two processes that are important for compensatory selection: On the one hand, heterozygous individuals can be preferred over homozygous individuals on certain gene loci (English: overdominance). An important example of this is the evolution of the vertebrate immune system. The MHC genes are primarily responsible for the variety of antibodies that enable the recognition of almost every foreign organism . Heterozygous individuals logically have more MHC genes than homozygous individuals. They have a more powerful immune system and are therefore preferred for selection. The advantage is not that individual MHC genes are "better" or "worse", it simply promotes their diversity. As a result, rare gene sequences are promoted by selection. Another example with a similar basis is the evolution of self-incompatibility in plants . The same effect can also occur because an allele in the heterozygous case confers favorable properties, which in homozygous cases are harmful or even lethal. This case has become famous in sickle cell anemia , a human hereditary disease, the gene of which causes resistance to malaria in the heterozygous case and has thus been selectively promoted in malaria areas in a short evolutionary time. On the other hand, a rare allele may occasionally also be directly preferred, e.g. B. when predators in the environment prefer frequent prey over rarer. An example in which rare combinations of characteristics were directly promoted was shown with the guppy . In this fish species, there are numerous, extremely different color morphs, especially of the males. It could be proven that males with rarer color patterns are less likely to be eaten by predators. In addition, the same characteristic is also subject to the effect of sexual selection. In many animal species, the females specifically prefer males with unusual and rare characteristics over familiar forms. B. Males with rare color patterns are significantly preferred as mating partners. This effect can go so far that individuals with artificial markings made by humans (e.g. for behavioral experiments) have greater mating chances, as observed in zebra finches .
Effects of other variations at the gene level
Mutations in the protein-coding sequence of genes and alleles generated thereby are the most well-studied mechanism that provides variations on the existence of which the action of selection in the design of adaptations depends. There are, however, extensive theoretical and empirical references to other mechanisms. Important examples are:
- Mutations in gene regulatory sequences
- Gene duplication (through mutations during recombination or meiosis) with a change in the function of one of the copies
- Gene recruitment: taking on new functions for existing genes
- Exon shuffling
- Insertion of transposons into existing genes.
An overview of these cases is given. Even a completely new generation of genes from functionless DNA appears (albeit very rarely) possible. More recent evolution theories such as evolutionary developmental biology place these processes at the center of their interest, even if the empirical evidence is still controversial. While the cases discussed may be less frequent (or at least less well researched) than mutations in protein coding sequences, their potential evolutionary implications are considerable. Some researchers have deduced from this that the power of selection as a key driver of evolution may have been overestimated. The effect of selection on such rare events is much more difficult to research in the laboratory than that of the fairly well understood effects of "classic" mutations. Despite their possibly great importance for evolution, there are still no closed theories.
In principle, it is possible to measure the effect of selection in a population directly. The following is required for a measurement:
- a trait that varies within the population that is believed to have a selective effect
- the effect of this trait on fitness. As always, the measure of fitness is the rate of reproduction. It is therefore necessary to measure the reproductive success of the individuals.
If one plots the variation of the feature against the reproductive success in a graph, the following relationships can result:
- No selection effective: There is no connection between the measured trait and the reproductive success. The graph results in a straight line parallel to the x-axis (not perfectly realized in real terms due to measurement errors and natural variability).
- Directed selection effective: The reproductive success increases linearly with the trait (or decreases linearly). The graph shows a straight line with a slope that can be used as a measure of the strength of the selection.
If one of the other forms of selection is effective, there is no longer a linear relationship between trait and reproductive success.
- a stabilizing selection results in a curve that rises up to a certain maximum (the optimal feature value) and then falls again.
- In the case of disruptive selection, a curve results that falls to a minimum and then rises again.
Both forms therefore result in similar distributions, only the sign being reversed. Compared to the linear selection, higher-dimensional (quadratic) terms are required for the mathematical description of the curve.
Although the measurement of the selection initially appears relatively simple in principle, there are numerous pitfalls in practice that make the actual measurement a demanding task. First of all, in natural populations it is extremely difficult and (depending on the lifespan of the species examined) also time-consuming to measure the real rate of reproduction over the entire lifespan. It is not permissible to measure over shorter periods of time because the reproductive success of almost every species is known to be dependent on age. Measuring in the laboratory is easier, but it is usually not acceptable to transfer the results to the field, as fitness is dependent on the environment. To circumvent the problem, a more easily measurable criterion is often used, which in the examined case can explain a significant part of fitness, e.g. B. the survival rate. An even bigger problem is that the measured characteristic may not be directly causally responsible for the observed effect, but rather correlates it with the actually decisive characteristic. In practice, it can almost always be assumed that the selection acts on numerous characteristics at the same time and that a certain effect is partly caused by numerous characteristics. In order to extract the proportion of the characteristic under consideration, its partial regression has to be determined.
In the meantime, attempts have been made in numerous natural populations using the method described to measure the strength of the selection involved. In most cases, however, the results obtained are not yet sufficiently informative because the size of the populations measured was still too small. In addition, it was shown that the strength of the selection apparently became smaller and smaller in the larger studies. This indicates a publication prejudice (“drawer effect”), because studies that seem to prove a significant connection between effect and characteristic, perhaps only by chance, are published more frequently and accepted for publication than those without this, which is the result is distorted in perception.
Selection in non-living systems
In his research on self-organization as the origin of life, Manfred Eigen developed the concept of the quasi-species . In doing so, he transferred the concept of selection to chemical systems on the early earth before life came into being. According to Eigen's theory, evolution is inevitable as soon as autocatalytic macromolecules and an energy source are present. For him, selection is an optimal principle that can be derived from physical quantities.
- Benno Müller-Hill : Selection. The science of the biological selection of humans by humans. In: Norbert Frei (Hrsg.): Medicine and health policy in the Nazi era. R. Oldenbourg Verlag, Munich 1991 (= writings of the quarterly books for contemporary history. Special issue), ISBN 3-486-64534-X , pp. 137–155.
- Douglas J. Futuyma : Evolutionary Biology . Springer Verlag, 1990, ISBN 3-7643-2200-4 .
- Charles Darwin : The Origin of Species .
- Richard Dawkins : The Blind Watchmaker . Kindler Verlag, 1992, ISBN 3-463-40078-2 . (orig: (1986) The blind Watchmaker )
- Ernst Mayr : ... and Darwin is right Piper, Munich 1994. (orig. (1991): One Long Argument: Charles Darwin and the Genesis of Modern Evolutionary Thought )
- George C. Williams : Adaptation and natural selection . Princeton University Press, 1996, ISBN 0-691-02615-7 .
- Selection and Fitness. Definitions and theoretical foundations
- Types of biological selection
- Robert Brandon: Natural Selection. In: Edward N. Zalta (Ed.): Stanford Encyclopedia of Philosophy .
- Link to an article by Ulf Dieckmann, see also “Recent examples for artificial selection” / “Unexpected Discontinuities in Life-History Evolution under Size-dependent Mortality”: PDF
- Otto Ammon : The natural selection in humans. Based on the results of the anthropological studies of the conscripts in Baden and other materials. Jena 1893.
- Arnd Krüger : A Horse Breeder's Perspective. Scientific Racism in Germany. 1870-1933. In: Norbert Finzsch , Dietmar Schirmer (Ed.): Identity and Intolerance. Nationalism, Racism, and Xenophobia in Germany and the United States. University Press Cambridge, Cambridge 1998, ISBN 0-521-59158-9 , pp. 371-396.
- Richard Law, Fishing, Selection, and Phenotypic Evolution. In: ICES Journal of Marine Science , Volume 57, 2000, pp. 659-668.
- Esben Moland Olsen, George R. Lilly, Mikko Heino, M. Joanne Morgan, John Brattey, Ulf Dieckmann: Assessing changes in age and size at maturation in collapsing populations of Atlantic cod (Gadus morhua). In: Canadian Journal of Fishery and Aquatic Sciences , Volume 62, 2005, pp. 811-823.
- David O. Conover & Stephan B. Munch: Sustaining fisheries yields over evolutionary time scales. In: Science , Vol. 297, No. 5578, 2002, pp. 94-96.
- David N Reznick & Cameron K Ghalambor: Can commercial fishing cause evolution? Answers from guppies (Poecilia reticulata). In: Canadian Journal of Fisheries and Aquatic Sciences , Vol. 62, No. 4, 2005, pp. 791-801.
- An overview of the debate in E. Lloyd: Units and Levels of Selection. The Stanford Encyclopedia of Philosophy (Winter 2010 Edition), Edward N. Zalta (Ed.) 
- Samir Okasha: Evolution and the Levels of Selection. Oxford University Press, 2006, ISBN 0-19-926797-9 .
- Ernst Mayr : The objects of selection. In: Proceedings of the National Academy of Science USA , Volume 94, 1997, pp. 2091-2094.
- SA West, AS Griffin, A. Gardner: Social semantics: altruism, cooperation, mutualism, strong reciprocity and group selection. In: Journal of Evolutionary Biology , Vol. 20, No. 2, 2007, pp. 415-32.
- R. Bshary & R. Bergmüller: Distinguishing four fundamental Approaches to the evolution of helping. In: Journal of Evolutionary Biology , Volume 21, 2008, pp. 405-420.
- SF Brosnan & R. Bshary: Cooperation and deception: from evolution to mechanisms. In: Philosophical Transactions of the Royal Society Series B , Volume 365, 2010, pp. 2593-2598.
- Vero Wynne-Edwards: Animal Dispersion in Relation to Social Behavior. Oliver and Boyd, (Edinburgh), 1962.
- RL Trivers: The evolution of reciprocal altruism. In: Quarterly Revue of Biology , Volume 46, 1971, pp. 35-57.
- DS Wilson: A theory of group selection. In: Proceedings of the National Acadademy of Science USA. Volume 72, 1975, pp. 143-146.
- DS Wilson & EO Wilson: Rethinking the theoretical foundations of sociobiology. In: Quarterly Revue of Biology. Volume 82, 2007, pp. 327-348.
- A clear derivation (based on the Price equation) z. B. in A. Gardner, SA West, NH Barton: The Relation between Multilocus Population Genetics and Social Evolution Theory. In: American Naturalist. Volume 169, No. 2, 2007, pp. 207-226 + 2 app.
- T. Clutton-Brock: Cooperation between non-kin in animal societies. In: Nature , Volume 462, 2009, pp. 51-57.
- See e.g. E.g. J. Field, A. Cronin, C. Bridge: Future fitness and helping in social queues. In: Nature , Volume 441, 2006, pp. 214-217.
Nicolas Moiroux include: Changes in Anopheles funestus Biting Behavior Following Universal Coverage of Long-Lasting Insecticidal Nets in Benin. In: The Journal of Infectious Diseases . September 2012, doi: 10.1093 / infdis / jis565
Jennifer Stevenson et al: Novel Vectors of Malaria Parasites in the Western Highlands of Kenya. In: Emerging Infectious Diseases . September 2012, doi: 10.3201 / eid1809.120283
- PC Doherty & RM Zinkernagel: Enhanced immunological surveillance in mice heterozygous at the H-2 gene complex. In: Nature , Vol. 256, 1975, pp. 50-52.
- R. Olendorf, FH Rodd, D. Punzalan, AE Houde, C. Hurt, DN Reznick, KA Hughes: Frequency-dependent survival in natural populations guppy. In: Nature , Volume 441, 2006, pp. 633-636.
- KA Hughes, L. Du, FH Rodd, DN Reznick: Familiarity leads to female mate preference for novel males in the guppy, Poecilia reticulata. In: Animal Behavior , Volume 58, 1999, pp. 907-916.
- Nancy Burley: Wild zebra finches have band-color preferences. In: Animal Behavior , Volume 36, No. 4, 1988, pp. 1235-1237.
- Manyuan Long, Esther Betrán, Kevin Thornton, Wen Wang: The origin of new genes: glimpses from the young and old. In: Nature Reviews Genetics , Volume 4, 2007, pp. 865-875.
- Hopi E. Hoekstra & Jerry A. Coyne : The locus of evolution: evo devo and the genetics of adaptation. In: Evolution , Volume 61, No. 5, 2007, pp. 995-1016.
- Overview in: John A. Endler: Natural selection in the wild. Princeton University Press (Monographs in Population Biology), 1986, ISBN 0-691-08387-8 .
- Russell Lande & Stevan J. Arnold: The measurement of selection on correlated characters. In: Evolution , Vol. 37, No. 6, 1983, pp. 1210-1226.
- an overview in: JG Kingsolver, HE Hoekstra, JM Hoekstra, D. Berrigan, SN Vignieri, CE Hill, A. Hoang, P. Gibert, P. Beerli: The strength of phenotypic selection in natural populations. In: American Naturalist , Vol. 157, No. 3, 2001, pp. 245-261.
- EI Hersch & PC Phillips: Power and potential bias in field studies of natural selection. In: Evolution , Volume 58, No. 3, 2004, pp. 479-485.
- Manfred Eigen: Selforganization of matter and the evolution of biological macromolecules. In: The natural sciences . Volume 58, No. 10, 1971, pp. 465-523.