Evolution
Introduction
�Creation and Evolution, between them, exhaust the possible explanations for the origin of living things. Organisms either appeared on the earth fully developed or they did not. If they did not, they must have developed from preexisting species by some process of modification. If they did appear in a fully developed state, they must have been created by some omnipotent intelligence.� - D. J. Futuyma
There is plenty of diversity in nature, wherever one may choose to look. The question plaguing us, then, is what causes all this diversity? Is there a mechanistic cause for this spread? Even as we ask this question, we can see that there is much in common among the differences. Look at bone structure � the presence of similar forearm structure, tails (even vestigial) � but also the larger fact that all complex creatures share the same type of cell, the eukaryotic cell, that changes little among species. In an even broader sense, there is unity, as all creatures share the same basic amino acids to make up the proteins needed for survival, and we all share the same genetic code. Also, we all have adaptations which have complete functionality, and also �mistakes�, which are signs of lousy design. A good example of this is the human appendix, which serves no purpose but to rupture at inopportune moments. Adaptation can only be explained by natural selection and molecular genetics; its abstraction worried Charles Darwin as he formulated his theories of natural selection.
Evolutionary theorists ask themselves many questions, some of which have been answered. Why does sex exist? Why are there different numbers of sexes among different species? Why do the rules of Mendelian genetics exist? Why do we grow old and die? How did the diversity of enzymes and biochemical nature of cells develop or evolve from their simple precursors? Why are there different species? If we had one single ancestor, why did it become more complex? Why is there mutation? How does it happen, when random changes are usually harmful to a population? What sets the rate of mutation?
Ernst Mayr distinguishes between ultimate causes and proximate causes. Proximate causes relate to the how � how does a process work, what is its mechanism � but ultimate causes are interested in the general �why?� Why is a certain species here, and what put them here? Why does evolution exist in the first place? This can become increasingly complicated and philosophical as we put aside religion and the idea of a Divine Creator. In any case, evolution itself combines the theories of philosophy, mathematics, molecular genetics, computer science, geology, ecology and many other fields. Theodore Dobzhansky once said that �nothing makes sense in biology except through evolution.� It is more though, as it is a unifying theory of change in the world.
Evolution does not so much reveal hidden potentialities as reveal change: change in systems, in organisms, and in cultures, where it is termed cultural evolution. Cultural evolution follows Lamarckian inheritance and is best represented by the development of language. Like biological evolution, language has populations that �show� characters. Language evolves; before dictionaries, it changed faster in terms of meanings, pronunciations and spellings. Fashions and styles evolve, though it�s not quite as obvious. Foods also change, look at coffee! Even computer programs change, through direct programmer input or the input of others, like in open source Linux. To think about evolution, we must think of change and use this particular approach in resolving the evolutionary problem. Darwin utilizes this thinking all through his Origin of Species; even if he was wrong on several counts, it has stood the test of time despite rigorous scientific, moral, philosophical and religious objections. Even the popular media has reported extensively on change, as the New York Times and other magazines are filled with articles on evolutionary topics. Recently, it was reported that a study by Wallace had suggested that mitochondrial differences in persons of African descent and northern European descent were due to place � African mitochondria were more efficient and generated less heat than northern European mitochondria, to match the temperature differences in the two areas.
When we think about things that change, we must all think of what is common. First of all, there are always populations of species (which are large in number), and variability in traits among populations tended to be transmitted through time (heredity) in certain proportions. This is called descent with modification � propagation with adaptation. Evolution studies populations, not individuals as mot people think. There are key differences: populations evolve rather than develop. Studies of evolution in populations are known as phylogeny and in individuals, ontogeny. So, we have evolution, which is then change between generations � and not within a lifetime, as developmental changes must be excluded and Lamarckian adaptation is wrong.
Notice this: no two things in nature are identical, yet we can identify them as something, like a child viewing a puppy and a big Husky, and identifying both as dog. We abstract from nature to an ideal, and so we ignore variability to some degree to be able to function. Another example is the red apple � it is the ideal, though a lot of apples are not red. Until some limit, we identify everything to that limit as a certain species, so therefore, the dilemma in evolution is how to categorize change and when we can point out that something has definitely changed as to become another species. The Greek thinkers Plato and Aristotle had a strong belief in idealism and the true nature, or the perfect ideal, from which everything deviates from. This strong belief has passed on to modern times, where we still claim the �ideal.�
Look at a system. Usually, we see the product as a result of the mechanism, and therefore the system is goal-directed. This is a teleological argument (like the chicken and the egg). This the Greeks used often. They also believed in fixed species and the idea of spontaneous generation, which was supported ludicrously by experiments involving used underwear and wheat in a jar spontaneously making adult mice of both sexes. The first proof against this was an experiment with covered meat that would not get maggoty. The second proof came from Leuwenhoek�s observations through his microscope of dividing �animolecules.� Spalanzoni went further and showed that boiled broth that was covered produced no mold or fungus; this was sterilization. Pasteur went further on this theme by boiling the broth in a swan-necked flask that was exposed to the air. The spores settled in the swan neck and didn�t get into the broth to spoil it.
Aristotle�s scale of nature involved a laddering of existence with God at the top, that usually led from inanimate matter to plants to fish to birds to mammals to humans to angels to God. This was much appreciated by the Christian perspective, which was taken up by Goethe and others in the school of natural philosophy. They believed both in the scale of nature and in archetypes, a common body plan for animals. This was partially true, as common body structures do exist among different species. Other supporters of the scale of nature included Jennings, who said that the universe resembles a well-regulated family, with just subordination among the various rankings of family members. Voltaire was against this, as it resembled the hierarchy of the church � basically, wee see things as we want to see them. In this case, it is how we order ourselves that we order nature. He also suggested the idea of gaps, where species have disappeared by going extinct. Similarly, Leibniz believed in extinction, but he also believed in a perfecting process that allowed animals to better themselves and change their placings on the ladder, nullifying the idea of fixity.
In the Middle Ages, classifications were based on esoteric qualities that usually involved the culinary and medical qualities of plants, or the physical similarities of animals. Often enough, they were ridiculous and had infinite variation, thus useless. Linnaeus presented the first great classification system that is still used today; his hierarchies (genus, species, kingdom, etc.) and binomial nomenclature are still basically correct. He derived his concepts of ancestral affinity and the inability of species to be borne of other species from John Ray (17th c.).
Evolutionary thinking continued with the Comte de Buffon, who attacked the idea of immutable species. �Species are the only true beings of nature.� He introduced the idea of sterility by saying that if species don�t reproduce together, they must be different species. He admitted variation among species but not change from one species to another (transmutation) with three objections: new species have not appeared in recorded history, the missing links have not been found, and intersterility defines species, so how can change arise within species?
Geology
What is the role of geology in evolutionary thought? The first issue to address here is the actual age of the Earth, to determine how long evolution will have taken place. Herodotus believed in an immensity of time, and he had made note of the ancient fossils of sometimes non-existent animals that were found in odd places like mountaintops, which didn�t make much sense. These thoughts and notations were pushed aside by Christian theology, especially when Archbishop Usher was able to identify Earth�s age as 6000 years old. This, however, didn�t explain the presence of fossils. Christians then argued that fossils were a means of testing our faith. This is wrong in the true Christian sense, because it can be argued with some biblical evidence that God would never test us like that. Other reasons came up after the first was refuted: the catastrophist theory brought in the Flood as a means of extinction. Still others said that there was no such thing as extinction, that the fossils were just representative of creatures that we haven�t found as of yet � Thomas Jefferson was an avid supporter of this theory. There were some that just attributed fossils to God�s mystery; many fundamentalist Christians in the clergy were great fossil collectors. The age of the earth as calculated today is 4.6 billion years, from radioisotope data. This differs from Darwin�s number of 300,000,000 � though much smaller, it was an outrageous estimate, as it was morally offensive to society in his day.
In the 19th century, two view of the geological record were at odds: catastrophism, or punctuated equilibrium, as supported by respected scientists Cuvier and Agassiz, and the gradualism of Lamarck, Hutton, Lyell and Darwin. Catastrophism was able to explain the existence of marine fossils in the mountains and the several layers of fossils in the earth. Gradualism or uniformitarianism, however, believed in the ancient age of the earth (Hutton remarked �no vestige of a beginning and no prospect of an end.�) and a gradual change in the species, whether for adaptation or other reasons. Hutton prepared a rebuttal of catastrophism that said that sharp discontinuities in strata are usually just local phenomena, and that the laws and effects of nature are most parsimoniously viewed as acting in the same fashion throughout time. The latter, in its weaker form, could make room for the odd catastrophe or two. The arguments went nowhere, as both sides had identified laws that they assumed were correct, and for good reason too. The gradualists especially believed in the fixed nature of certain laws, and non-reliance on a huge accident to fix things up.
Darwinism and Detractors
Charles Darwin (1809-1882) was the son of a middle class family of scientists and doctors. He first went to medical school, but soon dropped out to study theology before becoming a naturalist. His chance to go on the H.M.S. Beagle started his career in naturalism, which started after his analysis of the materials he had gathered on the trip. He then attacked the question of species using influences from Lyell�s �Principle of Geology� to Thomas Malthus� essay on the �Growth of Populations� to the proceedings of the Royal Society. He worked for 20 years on his ideas, and gathered evidence through experimentation during his semi-retirement at his home in Down, in County Kent by London. It wasn�t until he received correspondence from Alfred Russell Wallace that he was pushed to publish. Wallace was a hard on his luck animal collector who has also thought of natural selection and adaptation, and Darwin felt threatened by this newcomer � though Wallace had arrived twenty years later, ample time for Darwin to refine his theory. Eventually, they worked out a compromise wherein Wallace and Darwin would both present in front of the Royal Society in London, but evidently it wasn�t momentous, as the secretary noted: �nothing of much note happened [this year].� Afterwards, Darwin quickly wrote a lengthy abstract titled the �Origin of Species� that, upon release, sold out in two days and caused an uproar. Six editions later, it convinced the scientific world of the reality of evolution.
However, during his lifetime, he had trouble persuading anyone of the mechanisms of diversity as he saw it. The theory of Divine Design had been popular � if there is the appearance of design, there was a designer, who naturally would be God. This was perfectly reasonable to the intelligent person who wasn�t familiar with any of the tenets of evolutionary biology. William Paley, in Natural Theology (1802) writes thus: ��that the watch must have had a maker: that there must have existed, at some time, and at some place or other, an artificer or artificers, who formed it for the purpose which we find it actually to answer; who comprehended its construction and designed its use.� God�s creation covers not only the physical creation, but its adaptations as well. Even today, natural theology and plain old creationism are still an issue. In McLean vs. Arkansas Board of Education, the issue was an act signed into law that had promised equal treatment to both creation-science and evolution-science. The judge ruled in favor of the plaintiff, who said that this act is in establishment of religion. In the opinion, it was made clear that most of the creation-science arguments were false or academically unsound. Evolutionists defend themselves in many ways: by emphasizing that evolution is a theory; by showing that adaptation and �survival of the fittest� is not circular reasoning; it can be studied (micro or macro) in the laboratory or in nature through direct observation or the fossil record; it does not attempt to establish an origin of life (Big Bang is not evolution!); it has debatable points and is backed up by most of the scientific community; and most of what the public believes is actually wrong. When these points are brought up, many then turn to an intelligent-design theory, which is another name for natural theology. The prevalence of creationist-based reasoning is astounding, but not unexpected.
Darwin�s main opponent was Lamarck, whose individual-level theory of acquired characteristics said that : 1) through their own efforts, individuals can change in ways favored by the environment, and 2) these changes can be passed on to the offspring. Essentially, the causes and consequences are one and the same. Evolution can then happen in a population of one. Lamarck attributed this to some sort of �internal force� that can develop and then pass on these �acquired characters.� Lamarckian inheritance then became the most popular theory of directed variation, where offspring consistently tend to differ from their parents in a certain direction. August Weissman, in his famous refutation of this theory, grew mice in 22 generations and snipped off their tails before mating to see if the lost tail or any loss to the tail was inherited. He found no such thing � but it is wrong to say that this should be inherited, because the lost tail is not a self-induced process to adapt to better fit the environment. It must be useful, and the body must recognize it as such. This was an increasingly progressive theory, as it aspired to greater and greater complexity in animals, and was receptive to the ideal of the self-made man, for this was change brought about by the individual. The theory, however, was shadowed by his belief in continuous evolution � which was both incorrect and correct, for it did rightly theorize the existence of intermediate forms that tied all organisms together. His belief in extinction was much less controversial. Even today, Lamarckian ideas still pop up, and as late as the 1980�s, scientists were arguing that there was evidence of Lamarckian evolution, but it was never widely accepted.
Darwin, in his �Origin of Species�, argued for the reality of evolution (namely, a common descent with modification) and natural selection as the main mechanism of evolution. He argued that adaptation is best explained by the competition to survive or the appearance of design. In this population-level theory, he posited that: 1) organisms are variable and variation tends to be inherited, and 2) some of the variation affects survival and reproduction, and 3) on average, later generations are better adapted. He saw in the reality of life producing more offspring than survive as evidence for a struggle for existence. This theory separates causes from consequences, as the environment does not change what kid comes out, only how the kid lives after that moment. This sorting and disposing was attacked as �executioners of the unfit�, and it was argued that natural selection had not been observed or tested directly. Others argued that a population level theory should not concern itself with type characteristics. Herschel, a famous scientist of the time, called it the law of �higgledy piggledy� and didn�t like the random change inherent in natural selection. Actually, sexual reproduction contained the random changes, and natural selection was non-random, and directional. The argument for natural selection was bolstered by the algorithm argument: fulfill the conditions, and the process will happen. This was unbelievable to some scientists, and so they were unable to accept the theory.
Darwin�s example of island life was probably the strongest evidence for his theory. His oceanic islands showed dispersing species and an absence of amphibians and mammals; fewer species than a comparably sized continental area; strange fauna and flora, with large numbers of similar species and an under-representation of broader taxonomic groups; and most puzzling, the islands contained fauna and flora that were most closely similar in morphology to the forms present on the nearest continent, even as habits differ in both places. He found that the relationship of extinct forms to new ones through linking forms were also evidence as were the presence of vestigial organs and similarities in bone structure, and the stratigraphy filled with fossils of increasing complexity in the �correct� rock layers.
�He who believes that some ancient form was transformed suddenly through an internal force or tendency into, for instance, one furnished with wings, will be almost compelled to assume, in opposition to all analogy, that many individuals varied simultaneously. He will further be compelled to believe that many structures beautifully adapted to all the other parts of the same creature and to the surrounding conditions, have been suddenly produced; and of such complex and wonderful co-adaptations, he will not be able to assign a shadow of an explanation. To admit all this is, as it seems to me, to enter into the realms of miracle, and to leave those of science.� - Charles Darwin
To explain the nature of adaptation, Darwin wrote out his incorrect theory of pangenesis in the �Descent of Man�; in it, the body is filled with gemmules that respond to the environment by going to the gonads to be passed on, or not. Darwin�s nephew Francis Galton tested this theory by fully replacing the blood of rabbits and looking at their progeny. The progeny looked like its original parents and not the transfused rabbit. When it was published, Darwin was forced to reply: �I didn�t say it was in the blood.� Darwin scrambled mightily to explain the imperfections in his theory, but in the end, nobody preferred natural selection.
One more argument against evolution was the fact that heredity as understood then did not back up evolution and natural selection. Mendel, who worked during the 1860�s, worked with pea crosses and discovered the laws of heredity. He later posited that heredity was particulate. Actually, Darwin had a copy of the paper as sent by Mendel, but he had never read it � the pages weren�t cut. In general, Mendel�s ideas fit with Darwin�s theory, but Darwin refused to believe. Darwin believed in a blending theory of heredity whereas Mendel believed in the radical invention of new forms. Even in his day, most people believed in blending, wherein a pure black and a pure white parent would have children that were an intermediate blended form, or tan in this case. It was also assumed that the parental lineage could never be recovered; Jenkins, a Scottish engineer, said that blending reduced genetic variability by half each generation, and thus it homogenized the population. Questions invariably rose when it was realized that characteristics can reappear � which can be explained off by infidelity and the like, but not effectively � but no one understood enough to give the question a try. In any case, the blending theory was weak: if blending did occur, mutations would be rapidly flushed out of a system before it could achieve dominance. Also, his theory of use and disuse was found to be wrong. Darwin�s followers, the biometricians, fought famously with the Mendelians, but got nowhere, as there was no theoretical framework to start from. Eventually, Mendelism was accepted in the late 1900�s, and both theories reconciled by the Modern Synthesis.
Evolution today differs from the traditional sciences by its emphasis on a historical timeline � any evolutionary thought must take into account the history of species development, or whatever subject it may be treating. However, this history, like traditional history, emphasizes the �why?� more than the �when?� The adaptationist program, as described by Mayr, seeks to find out why certain adaptations occurred. As Lamarckianism, directive forces and saltational evolution have been discredited, these explanations must rely on chance or selective forces. Gould and Lewontin, in their critically acclaimed paper, attack evolutionary explanations for adaptations by claiming that an analysis of each and every immediate adaptation is wrong � that we may be fitting explanations to the present state of the trait, and not emphasizing the evolution of the whole, possibly by some designated Bauplan. This is termed the Panglossian paradigm. Evolutionary thought was held to be erroneous for distinguishing current utility from its origins (different answers to �why?� throughout time), for using plausibility as an important criterion, for not including random fixations, allometry and other forces, and other such reasons for inadequate support. Mayr counters that evolution is like other sciences in that it must test hypotheses until it finds one that is correct. In looser terms, until it finds one that fits. Unsupported answers to the �why?� are of course discouraged as irresponsible, and current evolutionary thinking is gradually including ideas of speciation, selection vs. adaptation, allometry, multiple peaks, etc.
A Brief Review of Molecular Genetics
Double-stranded DNA is made up of a phosphate, a sugar and a base; these nucleotides make up a strand that is arranged in nucleotide triplets. These code for RNA, and through RNA, protein � with the millions of genes found in our body, many proteins can be made in which to regulate and maintain bodily functions. Lots of non-coding DNA exist, which in theory do nothing. They usually exist due to mutations that change one nucleotide, or shift frames, or change whole stretches by slippage and transposition, or change an entire chromosome. To study mutation as it relates to evolution, mutation rates can be measured. Genes that are mutated exist in copies known as alleles; they are rearranged during reproduction, leaving Mendelian ratios of 1:2:1 in a heterozygote cross at one loci.
Evolution and Natural Selection
Though evolution is virtually accepted today, there still exists two other theories of the history of life: creationism and transformism. In creationism, species have separate origins and don�t change, and transformist accounts are similar to the Lamarckian theory. However, evolution provides stronger evidence. Evolution can be observed on a small scale, both in nature and in the laboratory. As for species � well, we must define a species. Most living creatures are classified using the Linnaean hierarchy. Those that are in each species category are defined by the ability to interbreed and by similarity of phenotypic appearance. However, these definitions are not solid, as there are many exceptions. Experiments have shown that variation within a species sometimes creates new species, due to isolation and other mechanisms, especially in ring species. Though they can interbreed and are similar, there still exist many differences. Even in the lab, new species have been produced or hybridized, that interbreed only within themselves and not their parent forms. Other evidence for evolution includes the fossil record, that suggests evolutionary relationships by geological level and fossil form, as well as homologies like similar limbs with different functions, vestigial organs and the genetic code. Crick called the genetic code a frozen accident, as a mutation made the genetic code what it was, but it was �frozen� in that it doesn�t allow new mutations. The evidence for evolution explains adaptation whereas creationism does not. However, it must be understood that religion is still compatible with science � there could have been a primary creator.
In nature, there is a continuous struggle for existence that makes it necessary for some species to have excess fecundity; that way, it is ensured that some will survive to continue the lineage. This is so as there are limited resources, and competition for those limited resources. Thus, some will die and some will win � and live. Those are the preconditions for natural selection. The argument requires four conditions: reproduction, heredity, variation in individual characters among the population, and variation in the fitness of organisms according to the presence of that heritable character. Natural selection will drive evolutionary change and generate adaptation; if the conditions are right, it can keep a population in stasis too. There are three kinds of selection that act on traits with continuous distributions: directional, toward a more favored form; stabilizing, toward an intermediate form; and disruptive, toward both extremes with no intermediate, possibly leading to two different populations. Or, no selection. Variation exists at all levels, but it doesn�t stay the same constantly. New variation is generated by mutation and recombination in sexual populations.
Hardy-Weinberg Equilibrium and Violations
In 1908, Hardy, a British mathematician, and Weinberg, a German pediatrician, published separate papers that provided support for a mixture of Darwinism and Mendelism. They posited that: 1) there was no loss of variability under Mendelism, basically rejecting blending, and 2) the dominant allele doesn�t spread in a population because of its dominance. This proved the start of modern population genetics, and what was known as the Modern Synthesis. Their basic law of population genetics assumed a single locus with two alleles (A, a) and five conditions: no mutation, no gene flow, no natural selection, infinite population size and random mating. The frequency of allele A is p, and of the allele a, q. Thus the total frequency can be illustrated thus: p + q = 1. If we calculate the probability of recombination, A-A is p2, a-a is q2 and A-a/a-A is 2pq. The total probability can be represented thus: p2 + 2pq + q2 = 1, or (p + q)2 = 1. We can introduce violations to the basic assumptions to determine the new models for evolutionary forces.
1. There is natural selection. This model assumes generation N, in which genotype frequency in adults is projected to genotype frequency at birth using Hardy-Weinberg. In the next generation, the new non-Hardy-Weinberg genotype frequency in adults can be calculated after the effects of natural selection.
Disfavored recessive allele
Genotypes AA Aa aa
Frequency p2 2pq q2
Relative fitness 1 1 1 � s (s = selection coefficient)
Average fitness 1 - sq2
Allele frequency p� = p / (1- sq2)
Change in p ^ p = spq2 / (1-sq2)
Change in q ^ q = - spq2 / (1-sq2)
Disfavored dominant allele
Genotypes AA Aa aa
Frequency p2 2pq q2
Relative fitness 1 � s 1 � s 1
Average fitness 1 � ps � pqs
Allele frequency p� = p / (1 � ps � pqs)
Change in p ^ p = p2s (1+q) / 1 � ps (1 + q)
Change in q ^ q = - p2s (1+q) / 1 � ps (1 + q)
Does this mathematical model really matter then, if it has unrealistic assumptions? A realistic model of evolution should take into account variability in loci, but this is impossible, as even 3-loci models are a headache to do. Quantitative genetics, however, can handle it in a more abstract form. This mathematical model is useful to study complex phenotypes with many evolutionary changes because we can study one loci and see how fast that it can take to bring a favored allele to popularity � a good example is trying to get a phrase from a series of random letters. If there was no selection, it would take quadrillions of tries to make it, but with selection, only 300 rounds achieves the desired effect. This, unfortunately, has a bad relevance to eugenic selection within human populations. It would seem that it wouldn�t be sensible if one studied the mathematics: it would take 9000 generations to lower the frequency of the recessive allele, even if all the homozygous recessive were killed off and the original recessive frequency was 0.001. However, we take into account modern advances, which allow us to identify heterozygotes � this makes eugenics a real problem should it be implemented.
2. There is mutation. Mutation doesn�t cause purposeful adaptation, only accidentally. It is a weak evolutionary force that can be measured by rate and net amount. Within a 2-allele system (A1 and A2), where m and n signify the forward and back mutation rates between the two different forms, m usually equals 10-6 if A1 is functional. For A2 to be converted back to A1 specifically, and if A1 is 1000 base pairs long, then n will equal 10-6 / 1000 = 10-9. Forward mutations were thought to proceed faster than back mutations.
Functional allele of one gene loci
Frequency F0 A1: p A2 : q
Frequency F1 p� = p (1-m) + q(n)
Equilibrium p* = n / (m + n) q* = m/ (m + n)
Using m = 10-6 and n = 10-9, p* turns out to be 0.001 � but this seems counter-logical, as A is supposed to be functional. However, there is no selection in this instance. Mutation is an effective randomizing force that spreads non-functional forms, but it only shows reality if selection is inserted as well. Mutation is a weak force in changing equilibrium frequency, as it takes several thousands of generations to change the frequency with any significance. Coincidentally, caffeine is a good mutagen and some time teratogen.
3. There is mutation and selection. Each person has at least one deleterious allele, in heterozygous form. Both mutation and selection select against the unfortunates who are homozygotic for these deleterious alleles. In this model, A represent a dominant, deleterious allele with frequency q. The wild type allele is represented by a, with frequency p. Again, mutation rates between the two forms are m and n. In our calculations, we usually don�t include n, as m is 1000 times bigger � and m is already small as it is (m = 10-6). Also, carriers of the dominant allele usually don�t pass it along anyway.
Deleterious dominant allele
Change in p and q ^ p = -spq2 / [1-s (1- q2)]; ^ q = spq2 / [1-s (1- q2)]
Equilibrium p* = m / s
In one example, with s = 0.804 and the heterozygote frequency at 10 of 94075 births, m was found to be 4 * 10-5. This seems right, but in nature, it isn�t usually that simple. Many allelic forms usually exist, in high quantities. Selection usually works against the mutated alleles, because most of the time mutations have a bad effect. The mutation-selection balance frequency for such a rare allele is typically less than 0.001, a very smaller number. Oppositely, the most common allele is 99% or lower because of the preservation of polymorphisms. These polymorphisms are in place because sometimes allelic differences are neutral, causing no difference in fitness. Natural selection can act to conserve variation, using heterozygote advantage, variable selection in space and time, and negative frequency-dependent selection, where fitness can change as frequencies change (a good example is the 1:1 sex ratio).
4. There is heterozygote advantage. In this case, the heterozygote is better adapted than homozygotes, and more than one gene can be maintained. This is called a polymorphism. Here, the A and a allele have chance of death s and t, respectively.
Heterozygote advantage
Genotypes AA Aa aa
Relative fitness 1 � s 1 1 � t
Equilibrium p* = t/ (s + t), q* = s/ ( s + t)
5. In populations where fitness depends on frequency, there is frequency-dependent selection, with either negative or positive frequency dependences. Negative frequency dependence arises in host-parasite interactions; when the host is in high frequency, the parasite is selected for, which then leads to selection against the host. Positive frequency dependence has increasing fitness as frequency goes up; however, it tends to eliminate polymorphisms.
6. There is gene flow/migration. Species can be made of a number of separate populations, which each have their own selection forces working within each subpopulation. The Wahlund effect is usually in play: when a subdivided set of populations have a higher proportion of homozygotes than an equivalent fused population. Migration allows subpopulation genes to be transmitted between populations � this is gene flow. We compare two populations with one loci, where m is the migration rate rather than the mutation rate, and t is the number of generations.
Alleles at one gene loci
Frequency of a in population 1, generation (t + 1) q1(t+1) = (1 - m) q1(t) + mqm
Frequency of a in population 1, generation t q1(t) = qm + (q1(0) - qm) (1 - m)t
7. There is migration and selection. With one loci, a balance of the two forces can maintain genetic differences between subpopulations. There are three possibilities, it turns out: The A allele will proceed to domination, or to elimination, or there�s an exact balance.
Alleles at one gene loci
Genotypes AA Aa aa
Frequency p2 2pq q2
Relative fitness 1 � s 1 � s 1
Equilibrium p* = pm (m / s + m)
8. There is genetic drift, whose actions are governed by four principles. a) in a finite population with no mutation, selection or migration, allele frequencies wander at random between generations and ultimately lowers variability by fixing one allele or other. b) The amount of time until fixation by drift is proportional to population size. c) The probability of fixation of an allele by drift is equal to its starting frequency. d) Drift can overwhelm selection in sufficiently small populations; drift prevails if s < (1/N). For example, mammals exist in smaller populations by comparison to bacteria and offer no codon usage bias. However, in large populations of bacteria, more transcribed genes use one codon set, and less transcribed genes use another, usually less common codon (possible because of the redundancy of the code).
To summarize one-loci populations genetics, we understand there is no loss of variation when using the Hardy-Weinberg rules with Mendelian principles. Evolution occurs by the favoring of one allele over another through differing models of selection. Mutation cannot work alone, as it would fill the population with deleterious alleles � there is a balance with selection. This is shown when we see the prevalence of a wild-type and the few mutants that exist around. Other forms of selections help maintain variation, like heterozygote advantage. These models, as well as drift models, can also account for the polymorphisms. These usually come up as intermediates on the way to fixation. Last of all, we can say the genetic drift leads to loss of variation.
But that is only one loci genetics. To understand the full picture, scientists study 2,3-loci genetics and above. It isn�t feasible any higher than 3 actually, so those populations are studied under quantitative genetics instead. In multiple loci population genetics, we can assume 2 loci with 2 alleles each, giving A1, A2, B1 and B2, with possibilities p1, p2, q1 and q2 respectively. The frequency of the haplotype with randomness is the product of the two allele frequencies in the gamete; thus, it comes out as p1q1 and etc.. The observed frequencies are labeled P11, P12, P21 and P22. If the haplotypes are truly random, the following should be true: P11 * P22 - P12 * P21 = 0. When this is not 0, it is known as D, or the index of linkage disequilibrium. D can change during recombination. Its value can be calculated with Dt = D0 (1-r)t.
Formula Verbal Description
p2 + 2pq + q2 = 1 Hardy-Weinberg equation
^q = -spq2 /(1 - sq2) change in disfavored recessive frequency
^q = spq2 /[1 - s (1 - q2)] change in favored recessive frequency
p' = p(1-m) + q(n) new allele freq. after 1 generation of forward and back-mutation
p* = n/ (m + n) equilibrium allele frequency under forward and back-mutation
p* = m/s mutation-selection balance freq. for deleterious dominant
q* = (m/s)1/2 mutation-selection balance freq. for deleterious recessive
p* = t/ ( s + t) equilibrium frequencies for two alleles with heterozygote advantage
q* = s/ ( s + t)
q1(t+1) = (1 - m) q1(t) + mqm allele frequency after generation of migration from population with frequency qm
q1(t) = qm + (q1(0) - qm) (1 - m)t allele frequency after t generations of migration in above model
qt = q0 / (1 + tq0) freq. of a lethal recessive after t generations
D = P11P22 - P12P21 coefficient of linkage disequilibrium, where P11 represents frequency of genotype A1B1, P12 represents frequency of A1B2, and so forth
Dt = D0 (1-r)t decline in coefficient of linkage disequilibrium in t generations with recombinant fraction r
Quantitative Genetics
Quantitative genetics, unlike population genetics, seeks to outline the genetic architecture of quantitative traits; that is, traits that are distributed continuously and don�t exist just as 2 different alleles/ forms. Its other goals include predicting the responses to evolution, whether natural or artificial; predicting how selection on a trait influences phenotypic distribution of other traits (these are genetic correlations); and determine the average effect of genes or the genes of major effect, and determining the quantitative trait loci attached to that particular gene or set of genes. The genetic architecture is known as a genotype-phenotype map, and is affected by the number of gene loci and environmental variables. Unlike true evolutionary genetics, quantitative genetics rose up mainly to support agriculture and food production. Models were generated to design the most efficient animal or plant: the hardiest wheat and corn, the bigger-breasted turkey, the super-milk-producing cow.
In the study of quantitative traits, distribution of loci generally fall under a normal distribution. As population genetics do not apply, there is much variation in phenotype, which can be affected by several genes of major effect or possibly hundreds of genes with small cumulative effects. In any case, phenotypic variance (VP) is made up of genetic (VG) and environmental variance (VE). Genetic variance, it turns out, is equal to the total of epistasis (VI), the additive genetic variance component (VA), and the dominance component (VD). Epistasis is the interaction of genes that don�t have additive effects. Conversely, the additive component is the proportion of genetic variance that can be passed on within a family. The dominance component is the different spread of phenotypes available when dominance is put in the picture (think 1:2:1 distribution compared to 3:1 dominance distribution).
From here we define heritability (h2), a ratio that describes the amount of phenotypic variation attributable to the alleles contributing to the phenotype. It varies from 0 to 1, with total non-likeness at one end and 100% additive components at the other. Broad-sense heritability is the genetic effect ratio: h2 = VG /VP. Narrow-sense heritability is more particularly the additive component ratio: h2 = VA /VP. It can be proven in two different ways. The first way looks at the covariance of the offspring versus parent numbers. This gives a summation of � A (A + D), from which we can ignore the term AD, as it balances each other out, no matter which direction it proceeds. At the end, the value for the covariance is � VA. If we divide that by the variation of that same distribution, we get the formula for narrow-sense heritability. Usually, one parent one offspring families are represented by � VA, average of the parents with one offspring by � VA, half siblings by � VA, and full siblings by � VA + � VD. Usually, the full sibling value is not used, as it involves the dominance factor. If the offspring-parent regression line is drawn, the slope of that lines also gives the correct value of the narrow-sense heritability. The other way to calculate this is to use the equation R = h2S, where S is the selection differential and R is the response to selection. In layman terms, S is the difference between the mean for parents of the next generation and the mean of the whole population. R is the change in mean phenotype after selection.
One common application of heritability analysis is the Illinois Corn Oil content experiment. Constant removal of subsets select for highest and lowest oil content. Heritability for either went from 0.32 and 0.5 respectively to 0.12 and 0.15. It slowly precedes to fixation, but never quite reaches there, as mutational selection, genetic correlations and selection in different directions will always prevent heritability from reaching zero. A good example of this is the fecundity/longevity comparison. As one gets older, there is less and less chance and energy available to have new children; thus heritability will never reach zero.
The very nature of heritability assumes the presence of much genetic variability; but how do we account for so much of it? Most populations are affected by stabilizing selection, which selects for an optimum value; deviations are selected against. However, these populations have genetic variance. Why? For one, there is a mutation-selection balance. That is, as selection reduces variability (which seems counter-intuitive), mutation will return it. Since clearing out mutations is neither immediate or 100% efficient, at least some variation will always exist. Also, mutation rates are slow, but the more loci, the higher chance of mutation � thus, higher variation will remain present. How much variation does this mean, as mutations can be deleterious? Lande proposes that stabilizing selection is controlled by many loci, and the mutation at one loci affects the phenotype. There is a 50% possibility that the mutation is beneficial; it may be either above or below the optimum value. Turelli proposed that mutations were on the whole deleterious, and occurred at many loci. Variation can also be caused by variable selection pressures on the genotype by environmental interactions. Using a reaction norm experiment, we can look at different phenotypes with the same genotypes under the influence of different environments. Graphs illustrating the effects of genetics and environment generally show no difference in phenotype if there is no influence and a sloped line if there is influence, or combinations. We currently don�t understand levels of genetic variation very well.
In one model of phenotypic variance, additive genetic variance affects morphology, which is also affected by the environment. From there, morphology affects life history, which is also affected by the environment. One can see that life history has double the effect of the environment that morphology has. In a similar model, everything is the same, except additive genetic variance affects life history directly as well � thus, life history is affected by double the genetic effect and double the environmental effect. Though these models are only theory, what is known for sure is that environment definitely has an effect on both morphology and life history. Take diet. With a crappy diet, animals have a lower life span, lower body size and lower fecundity; oppositely, a healthy diet extends all those factors. In analyzing heritabilities, one can see that life history traits generally have lower heritabilities. This is because the additive genetic variance is decreased just as the environmental variance is increased. As the heritability equals Va / Vp, the addition and subtraction definitely lower the heritability value. However, that is not all � in the total genetic variance, if the additive variance is decreased, dominance and epistasis have a larger effect. Epistasis is usually ignored, so we assume that dominance is the main factor increased. We estimate the proportion of genetic variance due to dominance with DG = VD / (VD + VA), and the proportion of total variance due to dominance with DP = VD / VP.
Certain organisms are more susceptible to environmental effects; as in, they respond differently to lab and field conditions and generate different heritabilities. This is because there is more stress in the field. There are subsequently more different phenotypes, as there is less control over outside conditions. However, we can accept lab data because comparing lab and field heritabilities show a fairly linear regression. The response to artificial selection is then varied. A linear relation means a 1:1 ratio of genotype-phenotype comparisons; the canalized relation means a threshold is present which is harder to pass, as there is little correspondence between genotype and phenotype; and the threshold relation has a threshold value which has two different phenotypes that are separated by the average of the genotypic distribution.
Formula Verbal Description
VP = VG + VE phenotypic variance
VG = VA + VD + VI genetic variance
h2 = VG /VP broad-sense heritability
h2 = VA /VP narrow-sense heritability
R = h2 S change in mean phenotype after selection
DG = VD / (VD + VA) proportion of genetic variance due to dominance
DP = VD / VP proportion of total phenotypic variance due to dominance
Back to the Beginning: Adaptation and Natural Selection
Darwinism involves undirected variation and selection provides the adaptive direction in evolution; so far, it holds philosophically, factually and theoretically. However much scientists are believers in adaptation, we are advised to adopt a pluralistic course in studying evolution, as we also include other factors like drift in evolutionary thought. But can Darwinism actually explain all adaptation? Take complex organs, like the eye. It is said that evolution cannot make the leap to such complex adaptation � but possible, they evolve in many small steps, which a simulation by Nilsson and Pelger showed, in about 400,000 generations.. Or, some critics mention, how about rudimentary stages that seem to be functionless or disadvantageous? Darwinist can argue that gradual change produces novelties, and that functions of adaptations can change with very little change in form. Also, new adaptations may evolve by combining unrelated parts, by symbiosis. Thus, there is a continuity in adaptive evolution.
When explaining the genetics of adaptation, there is a Goldschmidtian view and a Fisherian view. Goldschmidt argued for macromutation � this Fisher had doubts about. He noted that living things are suitably adapted to their environment, or they would be dead. The direction of mutation is such that mutation has a 50% chance of increasing or decreasing the character state, which affects the fitness. There is an optimum value for a trait, so small mutations are more likely to improve fitness, especially when closer to the adaptive peak. Another alternative is Wright�s shifting balance theory, which has multiple peaks. Fisher assumes that living things are close to their peak, but if they are not, then an expanded adaptive theory is required. Kimura suggest that sometimes large mutations, when more advantageous, will have a higher selective advantage. Thus, the change that a mutation is substituted is the chance that the mutations arise * chance that the mutations are advantageous * 2 * the selective advantage of the mutations. More recent experiments suggest that population size matters, because large mutations may or may not occur, and in smaller populations, mutation is likely to go by smaller steps.
Adaptation is studied in three conceptual stages: 1) identify what kinds of genetic variants could exist; 2) develop a hypothesis for its function; and 3) test the hypothesis�s predictions. Additionally, there are three practical methods. The first is to see if the actual form matches the hypothetical predictions. Another is to do experiments, if the organ or behavior can be altered experimentally. Finally, the comparative method can be used. This can only be used if hypotheses predict that different kinds of species will have different adaptations.
Adaptations in nature are not perfect, because of time lags � they coevolve with animals, or appear adaptively out of date, or are adapted for a recently extinct megaclassification � or genetic constraints (heterozygote advantage), or developmental constraints (pleiotropy, canalizing selection, developmental asymmetry), or historic constraints (stranded at local optimums, non-adaptive differences), or tradeoffs between different adaptive needs. Selection and constraint can be separated using adaptive prediction, measures of selection, measures of character heritabilities, and by comparative evidence. Most comparative graphs show allometric relations; that is, there is a correlation of growth, and this is considered non-adaptive. Methods of studying adaptations can only work is we are studying an adaptation � that is why it is important to understand what is or what is not adaptation. An organ�s function must be distinguished from its effects � for example, the aerial glide of a flying fish is not adaptation to get back in the water, because the fish obeys gravity too. Adaptations can be defined by engineering design or reproductive fitness.
Mayr emphasizes that adaptation has some constraints. Species have multiple pathways to the same genetic adaptation, and at the same time, they also have a capacity for nongenetic modification. Also, selection acts on the whole individual, and cannot be thought of as deterministic. Hindsight is 20/20.
What are the units of selection?
Life can be divided into a series of levels of organization, and it must be understood that an adaptation may benefit one level of organization but not another. Lewontin argues that for natural selection to affect a population at any level, there must be phenotypic variation, differential fitness and heritable fitness. If they hold, natural selection can produce beneficial adaptations for that selected level of organism. The segregation distorter gene, for example, breaks Mendel�s laws by allowing meiotic drive, in which one allele is consistently overrepresented. There is an advantage to itself that is a cost to the rest of the body � an intragenomic conflict. Also, there is favored selection in cell lines as well. Weismannist life cycles distinguish between germ and somatic cell lines, with the somatic cell lines inevitably dying and not contributing its adaptations. Buss points out that many life forms do not have this type of development, and can form new offspring from different cell lineages, introducing competition. Organismal selection obviously occurs. Natural selection can also work on groups of close genetic relatives; this is called kin selection. Kin selection can explain altruistic behavior under a condition called Hamilton�s Rule, that states that rb > c, where r is the chance that the altruistic gene is in the recipient, b is the benefit and c is the cost. B can be estimated by the difference between the survival rate of young in nests with or without helpers and c can be estimated as the reproductive success that a helper would have had had it not helped, as outlined by Mumme. Most biologists, however, feel kin selection is a weak force � as it can conflict with individual selection and it can be undermined by migration. Kin recognition � so kin selection can be put into effect � is done by direct or indirect mechanisms. Chemical or visual cues can allow direct identification, and external time and place cues can give indirect clues to identification. Strangely enough, there are some cannibalistic species. These rely, of course, on evolutionary forces. They only become cannibals if costs of kin selection outweigh the benefits.
Generally, the level of organization that will evolve adaptations is the one in which heritability shows. Natural selection can occur if a trait can be selected for. It can also be seen this way. What can natural selection directly adjust the frequency of? The gene. It selects for various favored genetic loci � the organism might cease to exist, even with the same genome, so the organism cannot be selected for in this case. One could argue that the DNA can also cease to exist at points. However, what is important is the information of the gene, not is physical continuity. The longevity of the genetic unit matters relative to the time evolutionary change takes, and is thus important. Also, the �gene� is used in a general sense, as certain types of DNA or more or less permanent and that is not necessarily taken into account. The change of gene frequency over time is a very active part of natural selection. The organism or population with the shorter life cycle will generally evolve faster, and thus be selected for. So, we have covered two senses of �unit of selection�: the entity with phenotypic adaptation, and the entity whose frequency is adjusted by natural selection, referred to as interactors or vehicles, and replicators.
Adaptations in Sexual Reproduction
Sex seems to be a problem because it seems less efficient than cloning; that is, there is a 50% cost in genes from each parent. In some lifeforms, sex has no cost, but the gradual association of sex with reproduction increased the cost. In compensation, the offspring should be more fit than the parent to outweigh the cost � Williams refers to this as the outstanding puzzle in evolutionary biology. It is curious that mutation can remove sex rather easily by eliminating meiosis. Asexual reproduction seems to be the standard in many forms of life, so there has to be a reason why sex exists. Sex accelerates evolutionary change only if the rate of favorable mutation is high. This is a group selection theory, and it comes from the taxonomic distribution of asexual reproduction. It has a spindly phylogenetic distribution that has a high extinction rate � sex prevails, because it has a much lower extinction rate. This argument is not watertight, because there could be different circumstances present, requiring either sexual or asexual reproduction, with some sort of balanced advantages preserving both mechanisms in the species� life cycle. This is known as the balance argument. Also, biologists are wary of accepting group selection theories.
The mutational theory of sex proposes that sex enhances the power of selection against deleterious mutations. For this theory to work, the deleterious mutation rate (U) must be high enough to have any major advantage for sex. Sex becomes advantageous if U is more than 1 � the most controversial prediction of the theory. U can be measured using mutation accumulation experiments or rates of DNA sequence evolution. This theory leaves a paradox if it turns out to be wrong. How can humans exist, given a high deleterious mutation rate? The other prediction says that a slope of log fitness by number of mutations should go down, meaning the more mutations, the worse off the species. The parasitic theory says that sex may be advantageous in changeable environments. Parasite-host coevolution has maintained constant variability in fitnesses of different genotypes in the parasite-host relation, as each adapts to the methods of the other. This ensures the ability to recreate genotypes that have been disadvantageous in the past, but are favored again. However, it is not conclusive if either theory is correct.
Sexual selection is used to explain many differences between the sexes. The peacock tail, for example, reduces survival by limiting movement, flight and being more conspicuous, but they exist because they increase reproduction. Sexual selection maintains that sexual characters are maintained by male competition and female choice. With the tail, it obviously cannot function in warfare, but Darwin suggested that females prefer pretty tails. Sexual selection is dimorphic in nature, and also functions in a runaway sense. Fisher proposed that the basis of some sexual selection was the reinforcement of evolutionary fact with female preference. Peacocks can survive better without the tail, but their offspring won�t get their genes passed on � and peacocks with the tail will. Selection will not only emphasize the peacock with the tail, but the one with the biggest tail, thus allowing the optimum tail size to increase. So, survival advantage progresses to mating advantage progresses to pure female choice. The handicap theory states that a male surviving some handicap has better genes than ones not able to survive. If females mate with the male with the better genes, the net quality of the offspring will be higher, as will their selective advantage. The costliness of this handicap, or signal, means that it is a reliable test. These theories predict that female choice is open ended rather than absolute, as in females like larger tails rather than specifically 12 inch tails. Generally, Fisher�s theory requires heritable variation in the male character as it requires changes in the inherited character, and the handicap theory requires heritable variations in fitness as it only requires changes in genetic quality. The common difficulty is that selection removes heritability. This is a weakness of the handicap theory. Another weakness is the vagueness of variation in genetic quality � but for all its weaknesses, experiments have shown female choice is influenced by genetic quality. It must be kept in mind that natural selection can apply different selective pressures on males and females, that can possibly oppose each other.
Selection favors a 50:50 sex ratio. Group selection favors the female, but over time, it will become more advantageous to the opposite sex. Thus, individual selection will favor the 50:50 sex ratio. This is unaffected by sex differences in mortality. Sex ratios will become biased at some points, however, especially is sons or daughters act as �helpers at the nest�. The helper sex will be favored if advantageous to the parents. If not, then it is the other way around. How this mechanism is established is unknown.