Biology2 Test 2

The Random Walk of Evolution

(1) Clarification on Evolution #1: "Evolvedness" is not quantifiable or comparable - Is a mushroom, a dog, or a bacterium more "evolved"? - Most people would say "dog" without hesitation. - But recall that evolution just means "change over time", and does not have a trajectory or direction. - We sometimes treat organisms as if they have "evolved more" if they pass certain organismal feature checkpoints. - Multicellular organisms are treated as more "evolved" than unicellular. - Organisms with dynamic motility are treated as "more evolved", like a mouse over a fungus (in reality their cellular processes are each quite complex). - But all three organisms mutate, and experience drift and selection just the same. - Bacteria can have some amazing adaptations to some really extreme environments where humans are completely incapable of surviving (super-salinated water or hot springs where the water is virtually boiling). - They undergo drift, selection, and mutation in their circumstances and habitat the same as any other organism. - The dog and the bacteria both "evolved" (verb), but neither is "more evolved" (an adjective). - Organisms can be more or less complex (depends on how you measure complexity), but there is no quantity of "evolvedness". - This is applicable to humans because we are not more "evolved" than anything else. - We have some amazing adaptations, but this does not mean we have achieved some "higher level of evolution." (2) Clarification on Evolution #2: Evolution does not proceed toward a long term "goal" - "Teleology" derives from the Greek "telos" meaning "purpose", "goal", or "end". - In the context of evolution, teleological thinking means you interpret evolutionary processes as specifically leading to some fixed end point. - Hopefully, you immediately said "but evolution ultimately comes from random mutations occurring and then selection and drift acting on them, so how could it be driving anything in a particular future direction". - Exactly, evolution cannot predict the future, it is an emergent property of processes operating the present and responding to the past generation. - Organisms cannot predict what traits will be higher fitness in the future, so they cannot set their evolutionary trajectories to work towards any kind of long term goal. - Sometimes biologists use the metaphor of a "random walk of evolution" to keep in mind this stochastic aspect of the evolutionary process. - The pervasive cultural meme of the "March of Progress" has grown up around Figure 6.9.11 and the placard illustration for this tutorial. - These diagrams are intentionally arranged to suggest that modern humans are the endpoint, or apex, of a progressive evolution. - Even the invocation of "road" conjures up imagery of a predefined path to be tread and a sequence of purposeful steps needed to walk that path. - The imagery of the "March of Progress" suggests that chimpanzees or other apes are in some state of "arrested evolution" and are stuck in a more "primitive" state that humans have adapted away from, but this is not so. - Other primates are not waiting around for opportunities to follow humans "down the road" towards bipedalism and language, they are evolving in response to their own evolutionary pressures in their own habitats through their own separate generations. - The March of Progress also misses the fact that we must still be evolving, in ways that are probably not apparent to us at the present moment with the information we have. - Specific examples have been shown of human populations adapting to high altitude, toxic metals in water, and other extreme environmental conditions. - But there is no destination, only random walking. (3) Clarification on Evolution #3: "Living Fossil" and "Missing Links" are imprecise terms - Two other ways in which teleological evolution or progressive evolution invades out lexicon are the terms "living fossils" and "missing link." - Two common examples of species that people will call "living fossils" in Figure 6.9.12: the coelocanth and the ginkgo. - "Living fossils" are extant species that bear a strong phenotypic resemblance to ancestors, while other species may have changed their morphology far more substantially. - This does not mean that the coelacanth has "stopped evolving" or is representative genetically of the ancestral population preserved through time. - Even if strong adaptive selection has not reshaped its morphology, populations experience mutation and drift always. - Genetically, ancient gingkos might have been quite a bit different from moderns ones, but the characteristic fan-shaped leaves have persisted as a phenotype. - Fossils are mineralized impressions of organisms. - Living organisms are alive right now. - Treating any living organism as more "primitive" than any other organism misunderstands the evolutionary process. - "Missing link" is a related concept that means a gap in the fossil record where a new fossil fills some imaged "gap" in a continuum of phenotypes. - But you can see the teleological thinking in the future by the way that the sedentary unsmiling Eustenopteron in the water contrasts with the smiling and dynamic Ichthyostega on land. - What's wrong with the "missing link"? - Aren't some ancestral forms actually morphological steps between the various species? - This is true, but it misses something important. - Just because an organism is "intermediate" in phenotype does not mean it's part of that ancestral chain. - This kind of logic also treats Tiktaalik as if it were embarked on some long evolutionary aspirational quest to walk on land, The problem with the diagram is that it can imply that it's inevitable tetrapods would walk on land if only they could, and semi-aquatic, lobe-finned creatures were just a stepping stone to the grand destiny of terrestrial vertebrates. - See the problem? - We've moved away from the random walk. - Evolution and adaptation does not drive to anything on a predefined trajectory. - Everything that has evolved starts as chance mutations. (4) Clarification #4: Irreducible Complexity One of the oldest criticisms of evolution (going back before Darwin even) were questions about how small changes in populations can produce complex phenotypes and big changes we see among species. - How does a system of parts that works together intricately evolve? - Many analogies have been used, most notably the Watchmaker Analogy and mousetraps. - The premise of the irreducible complexity argument is that if you take any one part of a structure like a mousetrap away it ceases to function, and so how could it evolve in a stepwise manner? - We will look at some biological examples in a moment, but this argument makes two huge mistakes in basic logic. - First, the assumption that systems must evolve one piece at a time. - Why not small alterations to the whole system at once? - As we will see, this is probably in actuality how things happen in biological systems. - We already have covered much about how interconnected genes are in functional systems; really, mutations do not change components in isolation, but instead tend to have more systemic effects. - Second, this is teleological thinking. - The first step toward a complex system was a random occurrence of some new phenotypic aspect that had no adaptive purpose or did not evolve in response to its original purpose. - When something evolves from an existing structure with another function or that was selected for a different effect on phenotype this is called "exaptation." - Recall that mutations happen at random, genotype has to translate into phenotype (with the environment), and then selection and drift can act on the trait. - So all new phenotypes arise at random and then go under drift or selection. - Not every single trait we observe has arisen due to a direct adaptive process on that specific trait (positive selection on a heritable trait). - Traits can arise due to developmental constraints or just by drift if they are not under negative selection. - Therefore, not all steps in a complex adaptation needed to be necessarily major fitness enhancements. - For example, the feathers used for flight in birds did not first emerge for flight at all, but were small downy feathers used for warmth. - Similarly, forelimbs did not originally develop for full flapping flight motions, but evolved their basic bone and muscle structures under other adaptive circumstances. - So the components needed for flight adapted separately and then were exapted into a flight system. - Birds did not go straight from no forelimbs to full feathered flight. - So the failure here is one of time, teleology, and, ultimately, a lack of imagination. - The most commonly invoked biological example of purported irreducible complexity is the camera eye of vertebrates. - If you approach a human eye and say: "you cannot remove any component, so what was the last component added," then this is a flawed argument. - In Figure 6.9.13, you can see how it's not the parts that evolved, but rather the shape. - As with most developmental features, the molecular parts are all present, they just need to be routed to the right place at the right time (think back to Hox genes). - If imagined as Ernst Mayr showed in the figure, you can see a perfectly reasonable progression from eye spots (a) to eyes that use a cavity to check direction of the light (b) to a water lens (c), to a closed eye (d), and finally the development of the iris and cornea. - The system evolves as a whole, not piece by piece. - Another example is the rotor proteins that make bacterial flagellae twist and turn (Figure 6.9.13). - These are amazing proteins whose biomechanical action is coincidentally similar to the common electric motor (this is not a bioinspired design though; electric motors far predate knowledge of the protein structure shown in Fig. 6.9.13). - But how could such a complex biomolecular system evolve? - The answer according to numerous studies of bacterial genetic diversity is that all the major proteins in this structure evolved under selection for some other function in the cell. - Then the final adjustments were made more systematically as refinements later. - Circling back to the mousetrap, it's worth nothing that blocks of wood, staples, springs, metal wire, and cheese were all around before mousetraps. - These components were not invented to catch mice. - Someone just combined what was already around in a novel way to make the system. - As it turns out, biological organisms can do the same with proteins and other structural molecules, and sometimes the novel combination of unrelated proteins can create dramatically new phenotypes.

Genetic Diversity and Equilibrium

(1) Genetic diversity refers to the number of unique genotypes in a population. - Recall that a genotype is the collective combination of all alleles at each locus. - A diploid organism carries two copies of each chromosome, so it's diploid genotype contains two haplotypes (haploid genotypes) of each chromosome. - This means when we measure genetic diversity of alleles in a diploid population, we have to count each individual as two alleles in that population. - A population of 100 Aa heterozygous individuals and a population of 50 AA and 50 aa homozygous individuals both have 50% frequency of the A allele. - The alleles are just arranged in genotypes differently in the two populations. - However, overall genetic diversity is measured not just by alleles at one locus, but by the combinations of alleles that make chromosomes different from one another. - A population with many genotypes and where the genotypes are more genetically distinct (separated by more mutations on average) has higher genetic diversity than a population with few genotypes or where the genotypes are not genetically very distant from each other in sequence. (2) What factors increase genetic diversity? - Mutation should pop into your mind right away. - New alleles originally are "born" by mutation. - The second diversifying force is recombination. - Recombination can take existing variable sites and recombine them into new combinations. - So mutation can make a new allele at the single nucleotide level (or many in the case of larger mutations like insertions/deletions), while recombination can rearrange these variants into new combinations to create a new gene/locus allele of two or more variant sites. (3) What factors decrease genetic diversity? - Genetic drift, given infinite time, will decrease genetic diversity. - Without new mutations, eventually random chance will drive out genetic variation. - Selection will also ultimately reduce genetic diversity. - As selection exerts preference on the phenotypic distribution, high-fitness individuals will be favored and low-fitness individuals will produce less offspring. - This means the next generation will be born from a biased set of parents, and this limitation of the parental pool reduces the diversity in the offspring. (4) Genetic Diversity and Equilibrium - So mutation and recombination add genetic diversity, while drift and selection remove it. - We sometimes call this mutation-drift-selection balance in an equilibrium context. - If mutations and recombinations are creating new alleles faster than drift or selection can remove then, then genetic diversity increases. - If selection or drift act swiftly, then genetic diversity is reduced. - What makes drift act swifty? - Reduction in population size can rapidly change the genetic diversity of a population.

For natural selection to occur, three conditions have to be met:

(1) The population must have variation in a trait (trait variation) (2) The trait must affect the number of offspring produced and/or the reproductive capabilities of the offspring (trait fitness effect) (3) The trait must be able to be inherited by offspring from a parent (trait heritability) - That is natural selection. - Variation in a trait is connected to variation in the number of offspring produced. - If the trait that confers the ability to produce more offspring (higher relative fitness) can be passed to the offspring (i.e., is heritable), then more offspring will be produced with the higher fitness trait in each generation while selection is acting. - Have the trait, use the trait to produce more offspring, then they produce more offspring and the trait increases in frequency.

Humans are phylogenetically related to the rest of life.

- "Modern humans" (Homo sapiens) are the lone living species in our genus. - Our closest living relatives are in the genus Pan: chimpanzees (Pan troglodytes) and bonobos (Pan paniscus). - Rounding out the rest of the family Hominidae are five other extant species: Eastern and Western gorilla (Gorilla beringei and G. gorilla) and Bornean, Sumatran, and Tapanuli orangutan (Pongo pygmaeus, P. abelii, and P. tapanuliensis). - These four genera (hominids, gorillas, chimpanzees, and orangutans) comprise the "Great Apes." - The so-called "Lesser Apes" comprising the four current genera of gibbons are the next-closest relative. - The apes along with the "Old World Monkeys" (vervets, baboons, macaques, mangabeys, colubuses, et al.) comprise the catarrhine primates or "Old World primates" having their distribution across Africa, Eurasia, and Oceania. - This clade joins "New World monkeys" from the Americas (marmosets, capuchins, spider monkeys, howlers, et al.) to form the "monkey" clade (infraorder Simiiformes). - Lemurs, tarsiers, and lorises comprise the remainder of most distant relatives within the order Primates. - Tree shrews are the next closest within mammals, followed by the rodent-lagomorph clade (mice/rats/beavers/squirrels, and rabbits/hares). - After that, we fit into the rest of mammalian diversity, then aminotes (including birds, snakes, lizards, turtles), then tetrapoda (including amphibians), then chordata (including bony, cartilaginous, and jawless fish), animalia/metazoa, eukaryotes, and cellular life.

Heterogametic Chromosomes

- A heterogametic chromosome or sex chromosome is any chromosome pair (two homologous duplexes) that does not occur in every gamete, and so the gametes are heterogeneous for its presence (hence the name). - The X and Y chromosome pair in placental mammals are an example. - All other chromosome pairs are called homogametic or, more commonly, autosomes. - Some human gametes contain an X and some contain a Y, so these are heterogametic. - All human gametes contain the Chromosome 1 pair, so they are autosomes. - This is noteworthy because it means 46,XY human individuals have only one copy of each gene on the X and Y. - This means that the inheritance patterns of genes on heterogametic chromosome pairs is different than the autosomes. - Some offspring get two copies of each gene on the X and zero Y genes, others get one X and one Y. - You can see in this example that XX individuals get two alleles of all X chromosome loci (one from the father (XP) and one from the mother (XM1 or XM2). - XY individuals get one X from their mother and the Y from their father. - What about crossing over recombination? - X and Y chromosomes cannot recombine because they are not homologous! - XX individuals can have recombination on the X chromosome though. - So you might have a recombinant X chromosome from your mother. - If you have two X chromosomes, then one came directly from your paternal grandmother without crossing over in your father. - This also means that alleles on the X chromosome in XY individuals cannot be homozygous or heterozygous. - Why? - There's only one gene copy. - Homozygous and heterozygous refers to the relationship between the two alleles on two chromosomes. - But the X and Y in an XY individual are hemizygous. - So there's only one copy of each gene, and these alleles are hemizygous. - While we are discussing heterogametic chromosome pairs, we should note that other systems occur besides the XY system (a.k.a. male heterogametic; so named because males typically have two chromosomes that are different). - Birds, lepidopterans (butterflies and moths), and many squamates (snakes and lizards) have a ZW system (female heterogametic) where ZZ individuals are male and ZW are female. - By convention, the organism that makes larger gametes is considered "female," and "males" produce smaller gametes, which is a useful universal definition, especially in species where male and females differ little in physiology. - Many plants, turtles, fish, and some invertebrates have no heterogametic chromosome pairs. - If they have separate sexes then usually either an autosome carries a gene or locus with sex-specific alleles or sex determination is environmental. - If you've heard that climate change has disrupted mating patterns of turtles, this is why. - At a key developmental phase, the cascade of factors requires a temperature signal to initiate a male or female pathway. - The warming climate is causing this signal to choose the male pathway too often and so there is an imbalance in sex ratios in these species. - The karyotype configuration is thus different conceptually from sex determination, which refers to the genetic pathways that influence whether an individual produces egg gametes, sperm gametes, or both in the case of hermaphroditic individuals. - This may be directly related to genes on a heterogametic chromosome or not, and still occurs in many species without sex chromosomes. - We should also note that there are a range of genetic factors that can cause humans to be intersexual and display a mix of characteristics that are typically found in only one sex or the other, or characteristics may be absent, or some combination of all of this. - This occurs in about 1.7% of humans, so intersex persons are all around us.

Reading Phylogenetic Trees

- A phylogeny is the most common representation of hereditary relationships among organisms. - Let's take a simple example of a (really over-simplified and incomplete) phylogeny of vertebrates. - We can highlight several features of phylogenies using this example. - Crocodilians (crocodiles, alligators, etc.) and dinosaurs (including birds) form a clade. - A clade is a group of organisms that includes all descendants from a single most recent common ancestor. - At some point in the past, birds and crocodiles are each related to this most recent common ancestor (some extinct lizard-like ancestor). - Then they share another MRCA, a more ancient one, with mammals in a clade called Amniotes that includes all animals that have amniotic eggs or live birth. - The common ancestor of Amniotes is also an amniote and laid eggs with amniotic yolks (as opposed to the gelatinous eggs of amphibians). - Note also that the ancestor of Amniotes is also a common ancestor of birds and crocodiles, but it's not the most recent common ancestor, because it is more ancient than the MRCA. - So, we can read this phylogeny to understand the relationships among organisms and the common ancestors through time. - We can say that humans and birds are more closely related than humans and frogs are, because humans and birds are more related through the more recent Amniote ancestor instead of the earlier Tetrapod ancestor that includes frogs too. - This tree is rooted, meaning that the tree is arranged along a time axis (bottom to top) that tells us the order in which the ancestors appeared in time. - Some rooted trees are placed on a time axis showing the approximate dates of the divergences, or splits, between the lineages shown in the diagram. - But otherwise, they just show the relative order of divergences, without specifying times. - How do we know that this order is correct? - We know from the impressions of organismal forms left in stone (fossils) that shark-like cartilaginous fish ancestors existed earlier in time than frogs or birds, and so forth. -The character changes in this case are phenotypic changes that define various clades (hair for mammals, four limbs for tetrapods, etc.). - Hair in mammals is an example of a shared derived homologous character, which means the character is present in a common ancestor and inherited by all its descendants. - From the time of Carl Linnaeus in the 1700s, phylogenies had to rely on phenotypic characters. - But after DNA sequencing arose in the 1960s, we could analyze DNA changes on a phylogeny as well. - Another example of a phylogeny is shown in Figure 6.8.2, but this time the characters are from DNA sequences. - Molecular phylogenies show the DNA nucleotides or protein amino acids at one or more homologous positions in a gene or genome as characters. - In Figure 6.8.2, the nucleotides at a single homologous position in the genomes of all six species are shown above the icons. - The star represents a substitution from an A to T nucleotide in the primate ancestor, which is then inherited by all three primate descendants of that ancestor. - The interpretations of the tree and relationships are the same as with phenotypic characters; we are just annotating our tree with molecular characters instead of phenotypic characters.

Additive traits

- A special case of epistatic gene interaction occurs when a polygenic trait is affected by several genes whose effects are cumulative on the trait. - This is common in many quantitative traits, which are traits measured on a numerical axis (height, weight, longevity, etc). - Quantitative traits can either be discrete quantitative traits in some kind of discrete units (like fingers and toes), or continuous quantitative traits, like height or mass that adopt all values in a range. (1) Scenario 1: In a population of flowering plants, height is controlled by four genes called A, B, C, and D. - An aabbccdd plant is 5 cm tall, and then each capital letter allele for any of the four genes will add 2cm in height. - A AABBCCDD plant would be 5cm + 8*(2cm) = 21 cm because the eight additional height-associated alleles will each add 2cm. - Since each allele adds to the total effect, they are additive alleles. - This is a special form of epistasis, since rather than two genes simply being redundant for the same effect, they add their contributions together. - So all additive alleles are epistatic, but not all epistatic alleles are additive (since they can be redundant as well. - This chart shows an example of additive alleles with no dominance or recessive relationships at each gene. - Each allele at each gene adds a contribution. - You can see that this roughly forms a normal distribution (a.k.a. bell-shaped curve), although due to the parameters we set there are gaps in the possible values. [We will explore how natural systems close these gaps in the next tutorial.] (2) Scenario 2: Let's keep the same scenario parameters as above, except this time each gene's upper-case alleles are dominant w.r.t. the lower-case ones. - In order to get each of the four genes' height effects, you simply have to have one upper-case allele. - We'll double the height effect to keep the scale consistent too. - Having two will be the same effect size as having one, so AA and Aa will add 4cm while aa will not add anything. - Now the genes are still additive, but there are dominant/recessive considerations. - Below is the result of graphing this type of additive system. - Rather than a normal distribution, now most of the values are at the high end. This is to be expected because now a AaBbCcDd individual and AABBCCDD individual each has the maximum height (as does any genotype with at least one dominant allele for each gene). - So, additive traits do not always form a normal distribution of phenotype values from the genetic contributions. -(3) Scenario 3: Now consider if, instead, the AaBbCcDd plant was merely 7 cm tall. - Furthermore, with any combination of dominant alleles at all the loci were still 7 cm tall. - Only the aabbccdd all-recessive genotype resulted in a 5 cm tall plant. - In this situation, the four genes are only epistatic (and redundant), not epistastic and additive. - So, again, additive is a special case of epistasis where the contributions of the genes are cumulative. - In humans, height is estimated to be affected by hundreds of additive alleles across the genome. - Most other quantitative continuous traits are affected by a similar number of genes. - Because these traits are controlled by so many genes they do not appear to be controlled by a discrete gene or set of genes because the traits appear to be continuous and are not observed in discrete units. - In reality, though, these traits are just controlled by so many discrete genes that the small individual effects of each gene inherited separately is not visible as a separate 3:1 trait ratio.

Humans and their extinct relatives

- As previously said, humans are the lone living species in the genus Homo and in the entire hominid lineage. - But it was not always so, even in the relatively recent past. - Humans and chimpanzees share a common ancestor about 6-8 Mya (million years ago). - The common ancestor of humans and chimpanzees was NOT a "chimpanzee." - Modern humans did not "evolve from chimpanzees" both modern humans and modern chimpanzees (and bonobos) evolved from their shared common ancestor population. - The earliest hominin species after the split from "panins" (the ancestors of modern chimpanzee and bonobo in the genus Pan) are classified in the genera Orrorin, Sahelanthropus, and Ardipithecus. - Ardipithecus (Fig. 6.9.4) is the most recent one and the one most anthropologists place unambiguously in the hominin lineage. - The descendants of these earliest hominins diversified over the next new million years into several species of Australopithecus and Paranthropus (Figure 6.9.5). - The genus Paranthropus is believed to be an extinct lineage that is not ancestral to the Homo lineage. - Instead, Homo and Paranthropus both have common ancestors in the Australopithecus lineages. - Remains of all early hominins, Australopithecus, and Paranthropus remains have only been found in Africa, and Paranthropus is believed to have gone extinct completely by ~1 Mya. - We will next check in on the human lineate at 800 kya. - By this time a new lineage Homo had emerged, and lived for hundreds of thousands of years alongside Paranthropus and Australopithecus in Africa. - However, by this time both Paranthropus and Australopithecus have gone extinct. - The Homo erectus descendants of the Australopithecus ancestors, however, have managed to become the first hominins to leave Africa. - H. erectus spread into continental Europe and across Southern Asia as far as the Indonesian islands of Sumatra and Java. - As H. erectus spread and more time passed, the Asian and European-African populations began to diverge genetically. - Asian populations of H. erectus retain that name in the literature, but African-European populations are called H. heidelbergensis. - Homo sapiens and Homo neanderthalis ("Neandertals;" our closest related extinct species) diverged from their common ancestor (called H. heidelbergensis) about 500kya. - This "human lineage" diverged from H. erectus-H. ergaster about 1.2Mya. - This 1.2Mya is the usual figure quoted as the "start of the ancestral human lineage." - Also note that Neandertals only became extinct about 37-40 thousands years ago. - Think about that! - We were not the only hominid species on Earth as recently the oldest cave paintings. - The map shown in Figure 6.9.8 shows a pivotal point in human history around 90 kya. - You can see that at this point H. neanderthalis inhabits much of Europe (the part not covered by glaciers), H. erectus populations are still in Asia, and H. sapiens inhabits East and North Africa. Around 125 kya - 90 kya marks the second "out of Africa" migration even of hominids. - Recall that H. erectus first migrated out of Africa around 800 kya. - Now H. sapiens begins to migrate out of Africa into the Arabian Peninsula (Figure 6.9.9). - Eventually, they will migrate both west into Europe and east into Asia, displacing first H. erectus (extinct by ~100 kya) and later H. neanderthalis by (extinct by ~37-40 kya). - These displacements were not instantaneous, but occurred slowly after long periods of overlapping habitat ranges. - Most evidence suggests that Neandertals were quite similar in habits, habitat, and morphology to modern humans, and likely had shared cultural practices. - Genomic evidence convincingly has shown that humans with non-African ancestry often have Neandertal-specific alleles from an ancient admixture of human and Neandertal populations. - Finally, within the last 15-20 kya, H. sapiens migrated to the Americas, and to more remote islands in the Pacific. - Now that we have a rough sketch of the migrational and genetic history of humans, we can note a few interesting features of this current understanding. - First, based on a very limited number of fossils, we already can see that there were many species of hominids inhabiting a fairly broad range of Africa for millions of years. - Often, we tend to think of human evolution as a series of species that supplanted each other in a series, but for nearly all of the history of hominids, several species coexisted, inhabited overlapping territory, and likely were not completely reproductively isolated (especially in the case of H. sapiens and H. neanderthalis). - That we have been the lone species of hominid for the last 37 thousand years is a historical exception. - Second, note that many of our extinct relatives had traits (especially behavioral traits) that we tend to think of as uniquely human innovations. - Calling bipedalism, use of fire, stone and wood toolmaking uniquely "human" innovations only works if we include extinct hominids under the moniker "human." - What makes modern humans unique on a biological and (especially) genetic level is still a matter of intense debate among scientists. - Third, H. sapiens was not the first hominid to migrate "out of Africa." - When humans left Africa ~125 kya there were already the descendants of the original H. erectus migration event living all over Eurasia. - The complex overlapping ranges of hominids created several zones of cultural interchange and genetic admixture that are apparent in the genomes of modern humans. - Fourth, H. sapiens have migrated very quickly in relation to generation time. - Humans also experienced a major population bottleneck around 70-90k that drove our population size down to approximately 10,000 individuals (some researchers say we almost went extinct). - While humans have rebounded in that time, this bottleneck means that humans are actually quite low in intraspecific genetic diversity compared to most other species for which we have population genetic sequence data. - Our rapid migration across the globe also created several founder effect bottlenecks. - This means that African populations have the highest genetic diversity, whereas native populations from Europe, Asia, and the Americas have much lower genetic diversity. - It also underscores that all humans are no more than about 0.5% different genetically. - Given that most of this variation is in non-functional parts of the genome, humans are even closer in the functional gene regions. - Finally, it's worth noting that >80% of the variation in humans is intrapopulation variation, while very few differences represent alleles that are fixed differences among human subpopulations.

The Role of Mutation in Evolution

- Before we leave the simulation, we should ask: where do brand new, novel alleles come from in the first place? [ - Mutation! - New alleles come from mutations in the existing alleles. - Let's start with all blue and wait for a mutation. BBBBB BBBBB BBBBB BBBBB BBBBB BBBVB - Aha! - In generation 4, a new allele happens. - Magic? - No. Mutation. - The violet allele emerges by mutation of a blue allele. - But... then on the next generation it doesn't come out of the draw pile. - Maybe that mutant individual doesn't reproduce, or gets hit randomly by an asteroid. - Who knows? - But... the odds were against the new mutation anyway. - This is true of all new mutations. - Since they are originally in only one individual and thus at very low frequency, the general odds are they will be lost immediately under genetic drift. - Let's do another simulation. - This time we will have more mutations of other alleles. - Orange, green, and pink alleles emerge and are lost through drift. - Violet drifts to medium frequency, and finally by change to fixation. BBBBBBBBOBBBBBBV BBBBBBBBBBBBBVVV BBBBBVBBBBBBVVVV BBBBVVBBVVVVVVVV BBBBVVVVVVVVVPVV BBBVVVVVVGVVPPVV - Ok, hopefully now we're solid on what a population is and what random changes in allele frequency mean. - The assumption of all the above simulations, however, is that all the alleles are equally likely to be inherited. - There is no biological advantage in reproduction from having yellow, blue, or any other allele. - What happens when we relax this assumption? - What happens when some force causes one allele to preferentially be passed on to the next generation relative to another? - What happens when genetic evolution is not just random?

Crossing Over

- But chromosomes contain thousands of genes, typically. - Does that mean that all the alleles on a particular chromosome are bound together forever as one inherited, assorted unit? - Not at all. Sets of alleles linked on a chromosome are broken up by a process called crossing over. - Crossing over occurs during the meiotic phase when the four homologous chromatids are in the progenitor cell. - At this stage, the chromosomes may come into physical contact at a random homologous site on the chromosomes and they exchange DNA segments at that point of contact. - What this means is that rather than inheriting your parent's chromosome whole, you inherit a chromosome that has parts of each of the two chromosomes. - Where did those chromosomes come from? - From your parent's parents. - So, rather than inheriting a chromosome directly from your grandmother or grandfather, crossing over would generate a new chromosome that has parts of the chromosome from each grandparent. - How often does crossing over occur? - In humans, the approximate rate is once per chromosome per meiotic event. - The place at which crossing over occurs is "effectively random," which means that different parts of the chromosome and genome may recombine more or less often but not necessarily in a generally predictable way.

Variable Phenotypic Effect Size

- Finally, we should note that not all additive genes contribute the same amount to a trait. - The usual terminology is that alleles of major effect contribute more than alleles of minor effect, although the designation of these effect size classes is situational for each trait. - We will discuss this more in the next tutorial when we talk about environmental effects and genome-wide association. - Key points: Epistatic genes can also be additive when their effects on a trait are cumulative (sum together to determine trait level). By definition, all additive genes are epistatic, but not all epistatic genes are additive.

Meiosis

- Gametes form by process called meiosis. - Meiosis has little to do with mitosis. - But these two processes are happening to different types of cells for different functions entirely. - Mitosis happens to somatic cells (non-gamete forming cells) and replicates one cell into two somatic cells. - This helps tissues grow, maintain, develop, and generally function . - In sexual eukaryotes, meiosis is for forming new gametes, and is restricted to the "germ line". - Meiosis does not maintain somatic tissues or help immune function or heal a paper cut. - It's for making gametes only. - The genetic processes that initiate meiosis and mitosis are separate genetically and it's best not to think of them as particularly related. - The original cell is called a gametogenic cell and undergoes DNA replication. - Then the cell divides, and then divides again. - This creates four gametes, each with one copy of each chromosome. - Let's say that you have a genome with just one pair of chromosomes (it's "diploid") and we will label the chromosomes 1a and 1b. - When you replicate these chromosomes you will make a copy of each chromosome and end up with four gametes: two with chromosome 1a and two with chromosome 1b. - But what if you have two pairs of chromosomes? - Now we have 1a and 1b and 2a and 2b. - Chromosomes do not stay in the same sets, but instead end up in the gametes in random combinations by a process called independent assortment. - So, with two pairs of chromosomes, you could have one gamete that contains 1a and 2a, or 1a and 2b, or 1b and 2a, or 1b and 2b. - You have to have one representative from each pair of chromosomes in the gamete, but which member of each pair will end up in a gamete, and in a haploid combo with the representative from other chromosome pairs, is a random process called "independent assortment". - If we extrapolate this to the 23 pairs of chromosomes in a human chromosome, then this means there are 223 (≈ 8.3 million) possible gamete combinations from each parent. - Since then you have the chance fertilization of two gametes, you have multiply the two parents' gamete combinations together so 223 ✕ 223 = 246 ≈ 70 trillion.

Adaptation (and Exaptation)

- Generally, an adaptation is specifically a trait that has evolved under natural selection and has a demonstrable relationship between the adapted trait and a particular external selective factor/force/challenge/problem. - Calling something an adaptation is more of a conceptual distinction than a biological distinction. - If we can show evidence that, as a result of this selective process, the change in phenotype increased the fitness of the organism in response to some specific external selective factor/problem/challenge, then we can confidently label it an "adaptation." - For example, giraffe necks are an old (and often misused or misconstrued) trope of an adaptation. - If giraffes with longer necks can get more food, then they might produce more offspring because they have more energy resources (or avoid starvation, in extreme cases). - Since we can easily see with our human perspective that (1) selection for long necks has "solved" a particular "problem" of the tall trees (2) this trait change enabled a higher relative fitness phenotype, then we can confidently label this an "adaptation." - A counter-example will be helpful, and we can also define "exaptation". - Consider the effect of a fungal pathogen on a population of plants. - If the plants (prior to the infestation of the fungus) have variation for resistance to the fungal pathogen, then some will survive better. - But not because they have an adapted resistance to the pathogen. - The variation in the trait was already there. - Selection just acted upon the existing population's phenotypes. - If the next generation has higher resistance to fungus on average, that is likely just the result of propagation of individuals with the current level of resistance, rather than the mutation and adaptation of individuals in the next generation to have even more resistance than their parents. - Mutations needed for high levels of resistance would need time to occur at random, spread in the population, and be selected. - So the immediate change in the resistance of a population is called "exaptation" because it's the result of selection that happens when the external selective factors change. - A change in the environment has resulted in directional evolution of a trait by creating a relative fitness differential between resistant and non-resistant plants, rather than a mutational change that makes a new high-fitness trait in the organism.

Modes of Selection

- Generally, we classify three "modes" of phenotypic selection based on their different effects on the phenotype. (1) In directional phenotypic selection, the individuals with trait values that are at the margins within the current distribution of trait values have higher fitness. - This means over time the distribution of trait values will (directionally) shift toward the new trait fitness optimum. (2) The second type is stabilizing phenotypic selection, where selection becomes stronger around the mean value of the current distribution in the population. - This means the variance in the trait will decrease (with more extreme individuals becoming less frequent), but the mean of the distribution remains roughly the same. (3) Finally, a rarer mode called disruptive phenotypic selection occurs when selection favors both margins of a distribution and the population evolves to have a bimodal distribution around the two new fitness optimal values. - What's really happening is understanding that when the relative fitness of phenotypes changes, the distribution of phenotypic values will move to center around the phenotype value (or values) with highest fitness. - What causes relative fitness to change? - Could be many things, but environmental change is a major cause.

Interactions of alleles at a single locus

- Genes occur in discrete units of DNA. - Every diploid organism has two alleles at homologous loci, and genes encode the gene products that act in concert to generate a trait or traits. - The alleles of genes in a genome interact in two key ways: (1) interaction of alleles of the same gene affecting a trait (2) interaction of different genes affecting a trait. - These two types of interaction also frequently overlap, so we will try to learn them together here. - For alleles of the same gene, let's start with a single gene (called A) in a made-up fish that encodes an enzyme necessary for catalyzing production of a pigment molecule that makes black stripes on the fish. - You discover that the fish is homozygous for the wild-type allele, so genotype A1A1. - One day you find a fish that has no stripes, and discover that it is homozygous for a different allele A2. - So its genotype is therefore A2A2. - You take the fish to your friend the biochemist, and she tells you that the A2 allele does not make a functional gene product protein enzyme. - So the pigment is not being made. - This makes A2 a loss-of-function allele (because it's an allele that lost its wild-type function due to a mutation). - Next, you bring a striped fish and no-striped fish into the lab. - You know that if you cross a striped A1A1 with a no-stripe A2A2, then 100% of the offspring should be A1A2. - When you look at the offspring that are A1A2, they all have stripes. - What does that observation mean on a molecular level? - It means that the fish only needs one functional A1 allele to have the stripes trait. - A single gene copy of the enzyme is sufficient to produce the same trait as a homozygous A1A1 in this case. - If you have an allele present where a heterozygote and homozygote have the same trait appearance, then we can say that the A1 allele is "dominant with respect to A2 for the black stripes trait". - We could also have said A2 is "recessive with respect to A1 for the black stripes trait." - Why did I say "with respect to"? - Because "dominance" is not a universal property of the A1 allele. - Dominance is a property of A1 relative to A2 in reference to a particular trait. - Another example will clarify: - You discover another fish in the same population with orange stripes due to being homozygous for yet another allele, so its genotype is A3A3. - Here the new allele is a change-of-function allele. - You cross A3A3 (orange stripes) with A1A1 (black stripes) and all offspring (A1A3) have orange stripes. - Why? - The A3 must be dominant w.r.t. A1 for the orange stripes trait. - A homozygous A3A3 and heterozygous A1A3 look the same for a given trait. - If A3A2 also has orange stripes then A3 is also dominant w.r.t. A2 for the orange stripes trait. - Finally, there are two special cases of dominance that are rare, but worth mentioning. (1) If you discover another allele A4 where A4A4 individuals have bright blue stripes and A4A2 heterozygotes have pale blue stripes, then half the functional enzyme only makes half of the phenotype when functional A4 is paired with a loss-of-function allele (A2). - A situation where a heterozygous individual with one loss-of-function allele only has a partial phenotype for the given trait means that A4 is incompletely dominant w.r.t. A2 for the blue stripe trait. (2) If an A1A4 individual (heterozygous for two different functional alleles) expressed the traits of both functional alleles, then we say that A1 and A4 are co-dominant w.r.t. each other. - Human blood types (A, B, AB, and O) is a good example of this very rare phenomenon in nature. - Key point: Dominance/recessivity are relative properties of the two alleles at the same locus that determine how the combination of alleles affects a trait.

Genotype and Phenotype

- Genotype is all or part of the genetic constitution of an individual or group. - We can refer to the genotype of individual genes, chromosomes, or the genome as a whole. - "Genotype" must refer to some part or the whole of the inherited chromosomes, not the gene products or any other molecules or processes. - Phenotype is the observable properties (i.e., traits) of an organism that are produced by some combination of genotype (G) and environment (E). - The traits that organisms exhibit involve a combination of their inherited genes, which are expressed into gene products that combine and function in the cell to eventually lead to what we observe as traits. - But phenotypic traits might also be modulated by environment effects. - Some traits are more strongly affected by genetic factors, some more strongly by environmental, and others are a mix of the two. - For some traits, all phenotypic variation can be explained by genotypic variation, and the environment can be considered irrelevant to the trait's expression. - The reverse is also true. - Traits can be at the molecular or cellular level, or at the organismal level.

Heritability:

- Heritability for a given population and a given trait measures how reliably parents pass the trait genetically to their offspring. - The most common way that heredity is measured is by measuring the trait value for parents and their offspring and checking their relationship. - In the plot on the left, you can see that offspring height correlates strongly with average parent height. - In this population, the height of the offspring is highly predictable based on the parent height. - This means that much of the variation in the trait comes from the genetic component from the parents and therefore the trait is "highly heritable." - Consider by contrast the plot on the right. - There is no correlation between parents and offspring in height. - Offspring are a completely random height compared to their parents. - Instead, the phenotype is either random and/or influenced by the environment instead of genetics. - So the trait is not heritable in this second scenario. - What does that mean for selection if the trait is not heritable? - If a trait is not heritable, then a parent who has a high-fitness trait variant might have many children, but those children do not inherit the phenotype value for that trait. - Instead they have a random phenotype value based on environmental factors or other chance events. - High-fitness parents do not give birth to more offspring that also carry the high-fitness alleles. - Similarly, parents with the low-fitness version of the trait might have fewer children, but they could have any random value of the trait. - So the trait does impact fitness, but it's not heritable. - So the frequency of the trait cannot increase or decrease directionally because the bias in fecundity does not translate to bias in the offspring phenotype distribution. - Random drift is still evolution (frequencies of the trait variants changing in the population), but in order for natural selection to directionally change the frequency of individuals with the high-fitness trait must increase over time. - If the trait is heritable then parents with high-fitness trait values will also have offspring with high-fitness trait values. - Since the parents with the high-fitness trait variant are having more offspring than the low-fitness variant, there are more individuals in the next generation that have the high-fitness variant of the trait. - So the population now can change directionally. - For example, if you have a population of fish with (1) variation in body size (small and large) (2) the large body size is higher fitness than small body size (larger fish produce more offspring) (3) larger fish parents have larger fish offspring compared to small fish (body size is heritable), then the population will evolve over generations to have larger average body size. - Selection of a heritable trait phenotype causes evolution of the trait phenotype. - Perhaps you doubt that actual traits could have zero heritability. - Consider the results of the selection practiced by chicken breeders to improve egg-laying rate. - In each case, breeding only the top egg-layers causes a rapid phenotypic response in the intended direction. - This implies a genetic basis for the trait under selection: early in the breeding effort, hens vary in egg-laying phenotype, and that variation is due in part to allelic variation at the many loci affecting this complex trait, and helpful alleles are being concentrated more and more in each successive generation. - The farmer doing the artificial selection is delighted. - Profits increase by 50% in roughly 15 years. - But trouble looms. - After a period of continual, unrelenting directional selection favoring the very best alleles for egg-laying, your flock has only those alleles. - All of the other alleles have been purged, and the curve flattens. - The flock of hens continue to show phenotypic variation in egg-laying rate, but the phenotypic variation no longer has a genetic basis - Why? Simple. - The flock has no more genetic variation at the loci that affect egg-laying). - In other words, heritability changed from a pretty large value at the beginning of the artificial selection effort to zero when the flock of hens became genetically homogeneous. - This extremely important "component of fitness" had no heritability, and so no more ability to respond to selection. - Response to selection can be dramatic. - Other indications of this fact are the wide diversity achieved by artificial selection on dogs, cats, guinea pigs, chickens, pigeons, cows, horses, goats, pigs, goldfish, wheat, ornamental shrubs and annuals, and any of a multitude of microbes. - Another is that the response to the selection was remarkably predictable: each of the four independent efforts yielded basically the same outcome on the same time course (too bad that we cannot see the initial data for the Cornell flock, but still). - A third is that selection is not guaranteed to yield a response to selection. - After the curves flattened, each farmer continued to see variation in the egg-laying phenotype, but the selection procedure no longer induced a response.

Trait variation:

- If there is no trait variation then there cannot be trait fitness differences. - Without different relative fitness, there can be no selection.

What a "Punnett square" literally means

- It means meiosis. - In the sense that what it is actually telling us are the combinatorial possibilities of getting certain chromosome combinations during meiosis. - The rows and columns represent the possible gamete chromosomal combinations from each parent, respectively. - Then the cells in the inner part of the table represent potential zygotes resulting from the fertilization of the gametes in their rows and columns.

The Fallacy of Blending Inheritance

- It was accepted in the past, and still provides an informal model for how many people think day-to-day about inheritance. - From early biologists in the 1700s until the early 1900s, many scientists (including Charles Darwin) modeled inheritance as Blending Inheritance. - We still see mutation and alternative alleles (since there are red and white flowers), but all inheritance is viewed as a blend of the parental traits directly. - In the absence of new mutations, this leads eventually to a blending towards intermediate phenotypes as shown. - For reasons that will become clear by the end of this tutorial, this model was consistent with observations of many traits. - However, it failed to explain a common observation that frequently offspring would have traits completely different from, or even more extreme than, the parents.

Dialog 6.8: organism relationships

- JEAN: How do we figure out the relationship among organisms? - DR. LOCUS: What do you mean by "relationships?" - JEAN: What speciated away from what and in what order...? - DR. L: Right idea, let's formalize a bit. How about: "What organisms are related to each other through common ancestor populations"? Also, one species doesn't necessarily split away from the other; we usually say the two new species mutually split away from each other. - J: Makes sense. A whale and a dolphin are more closely related to each other than to a cat because there was in the past a common ancestor population from which whale and dolphin descended, and cats were not involved. But aren't they all mammals? - DR. L: Yes, at some point all mammals are related through the ancestor of all mammals, but whales and dolphins have a "most recent common ancestor" with each other that is much more recent in time than the much older ancestor of cat, whale, and dolphin. What data would you use to test a hypothesis that whales and dolphins share a common ancestor before cats? - J: I mean... you can just kinda tell by looking at them? - DR. L: That statement happens to be true in this case. Is it always? Thinking of counterexamples is a good way to test a model or method. - J: Bats look like birds, but are not. Koalas are not bears, they are marsupials. Elephants and rhinos are not closely related, I think. And what about bacteria vs. other microbes? - DR. L: All true. Relying solely on certain phenotypic characters or our human aesthetic intuition can be misleading. What other data could you collect? - J: What about DNA sequences? Organisms with more recent ancestors should have more similar DNA sequences. - DR. L: A strong idea. Let's see where it takes us.

Dialog 6.6.1: the hen paradox

- JEAN: I get that the gene pool of this hen population will have only the top alleles after all of that selection. Selection can be very effective in removing alleles that get in the way of top performance. But isn't it a paradox that allelic variation for egg-laying is required to get a response to selection, and that variation was available in the beginning, and then disappeared? - DR. LOCUS: It has that look, yes. - JEAN: Umm... I was hoping you would have a little more to say. - DR. LOCUS: Very well. As you say, selection must have variation in the selected trait in order to select. If the phenotypes are all identical, then no phenotypic attribute can give some a competitive advantage over others. And if the phenotypes do differ, that variation must be due, at least in part, to heritable genetic variation, if the farmer wants the selection to continue to cause improvement in future generations. - JEAN: Yes! The hens continued to vary in egg-laying rate, and the farmer became frustrated when the top-laying hens did not have top-laying offspring any more. And I am frustrated about the hens showing any phenotypic variation at all in egg-laying when they had no more genetic variation at loci affecting egg-laying. How can we explain this phenotypic variation once the chickens had their genetic variation reduced through selection? - DR. LOCUS: I have one equation to mention: VP = VG + VE - JEAN: Oh. Environmental variation can cause phenotypic variation as a process completely independent of the effect of genetic variation. After so much selection, these hens were as similar in egg-laying alleles as clones would be, but still varied in phenotype because of their environments? Like some were stressed by a cat looking at them all the time, or were at the drafty end of the henhouse? - DR. LOCUS: Yes. - JEAN: But doesn't VP = VG + VE + VG✕E ? You left off the genotype-environment interaction. - DR. LOCUS: Yes. But note that the VG of the interaction would be zero when the time course curve had gone flat... meaning that variation of type VG does not exist to condition variation in type VE. That genotype x environment interaction has no effect in this case. The genotypes were all the same. - JEAN: This complete loss of genetic variation from directional selection certainly has peculiar aspects. Allelic variation is required for a response to selection, and the selection then takes place successfully in the beginning, and eventually makes further selection impossible by eroding that allelic variation. Would this always be the case, with any mode of selection? - DR. LOCUS: Good question. Think about the effect of long-term stabilizing selection on the allelic variation for a trait under selection. And also think about the same in frequency-dependent selection. And also think about the same for directional selection which reverses direction before it purges all of the not-top-performing alleles. And consider why mutation is not producing new allelic variation fast enough to keep the response to selection going. And then get back to me.

Dialog 6.7.1: Reinforcement

- JEAN: Returning to the now reunited antelope populations, what happens now? - DR. LOCUS: Well, if mating happens at random, then some matings involve one parent from each population, and these produce hybrid offspring. If we say there is postzygotic reproductive incompatibility then these hybrid offspring are either inviable or sterile. - JEAN: But natural selection should restrict mating to homotypic (intrapopulation) mating! Because offspring of homotypic mating project a parent's alleles into future generations so much better. The mating should be non-random, restricted to each subpopulation. - DR. LOCUS: But in our scenario, the populations came into secondary contact without any prezygotic isolating mechanism in place. Unless the populations developed a physiological incompatibility or behavioral mate preference, there's no prezygotic isolation mechanism once the river is removed. - JEAN: But are the phenotypes under selection now? Surely, the sterile offspring would be low relative fitness? - DR. LOCUS: Definitely, but there's selection happening to the reproductive mating adults as well. Let's use frogs as another example. In a certain species, frog males attract frog females with loud calls and females select mates based on their calls. Suppose that, when two reproductively incompatible populations come back into contact, a small difference in the calls has evolved already, during the time in allopatry. The difference is too slight for females to show a strong preference and prefer the calls from their own population over the other. But, individuals who have more viable intrapopulation matings and fewer interpolation matings will produce more offspring. So any small trait variation in females that gives them even a slightly stronger ability to detect intrapopulation male calls or any slight trait variation in males that makes their calls more differentiable will be selected. - JEAN: So the female preference for a more easily identifiable male call causes selection on males to call as differently as they can from males of the other population. The cost of heterotypic matings induces selection on females, and the new discrimination by females induces selection on male's calls! - DR. LOCUS: Exactly. We're talking about the rapid evolution of a prezygotic reproductive isolating mechanism, rooted in the for-no-good-reason evolution of a postzygotic isolating mechanism before they came into secondary contact. - Jean: So the calls would not have differentiated as quickly without the selection induced by the secondary contact, after genetic incompatibility formed. - DR. LOCUS: Far less likely or slower without the secondary. What do you think the term "reinforcement" means? - JEAN: I will guess that it is short-hand for "reinforcement of the little first step toward a prezygotic isolating mechanism". In the frog case, the slight difference in male calls. - DR. LOCUS: Exactly. The postzygotic isolation was reinforced by development of a prezygotic isolation mechanism that further entrenched reproductive isolation AND phenotypic differentiation.

Dialog 6.1: What is an allele?

- JEAN: When I learned the Punnett square before, it was done as individual genes connected to traits, not whole chromosomes. - DR. LOCUS: Where and what are genes? - JEAN: A piece of a chromosome...ah right. So this is just shorthand to refer to a whole chromosome by a single gene of interest. - DR. LOCUS: Yes. Those letters stand for the allele of a particular gene on a chromosome involved in meiosis and fertilization. - JEAN: What is an allele? - DR. LOCUS: An allele is merely one occurence of something. In the context of a gene in a diploid, each of the two chromosomes has its own occurrence of a particular gene. - JEAN: What do you mean occurrence? Like its own version? - DR. LOCUS: Picture two bookshelves with the same books on them and the books are in the same order. You might have a different edition of Moby Dick on the two shelves or the same edition, or one copy might have notes written in the margin and the other is unmarked. But it's undeniable that those two books are both occurrences of Moby Dick. Each chromosome has its own occurrence of some gene, whether or not the gene's sequences are different or the same. - JEAN: So the upper and lower case letters denote different alleles, but even if an organism has two upper-case A alleles they still have two alleles, two occurrences. - DR. LOCUS: Exactly. - JEAN: Is this related to homology? - DR. LOCUS: Definitely, two genes descendant from the same ancestral gene will be homologous alleles of the gene. The "two copies of Moby Dick" are homologous. - JEAN: So alleles just two occurrences of the same thing? Can that apply to more than just genes? - DR. LOCUS: Definitely, can refer to chromosomes, genes, loci, and variants at a single homologous nucleotide site as alleles. Even sometimes the states of a trait can be called "trait alleles." - JEAN: Sounds confusing. - DR. LOCUS: It is. But if you strive to be specific about your context for using the word "allele," and be sure to directly what are you talking about occurrences of, then you can avoid much confusion. Try to always say "alleles of a chromosome" or "alleles at a single nucleotide site," for example. - JEAN: Makes sense. Alleles are always occurrences of "something" so best to mention what you're referencing to stay clear. - Whole chromosomes are inherited.

Trait Fitness Effect:

- Just having variation in a trait is required, but not sufficient in itself for natural selection. - The traits with variation must have variable consequences for reproduction, or else they will not create variation in fitness. - For example, you might have blue and red plumage on a bird, but these two types might have equal relative fecundity (reproductive output). - So there is variation in color, but the trait variation does not translate into relative difference in reproduction. - Without a fitness differential due to the trait, then there cannot be selection.

Allele inheritance

- Let's say we have a diploid organism with a karyotype that has two pairs of chromosomes. - Chromosome pair 1 has a gene called A that has two alleles: A1 and A2. - Let's say that the father has two copies of the A1 allele on the two homologous chromosomes. - Same for the mother. - Each parent is therefore "homozygous for the A1 allele." - If two homozygous parents mate and produce offspring, then 100% of their offspring will have a diploid genotype of the A gene of A1A1. - Why? - Because if the parents only have the A1 allele, then all their gametes can only have the (haploid) genotype A1. - This means the only possible combination in offspring is the fertilization of two A1 gametes. - However, if one parent is A1A2 ("heterozygous for the A1 and A2 alleles") and the other is homozygous for A1 (A1A1), then their offspring have a 50% chance to be homozygous for the A1 and 50% chance of being heterozygous for A1 and A2. - Remember this does not mean that a pair of fish that give birth to 100 offspring fish will have exactly 50 homozygous and 50 heterozygous. - This only refers to the chance of each offspring to be homozygous and heterozygous. - You could have 100 homozygous offspring, all with the same probability of flipping heads on a coin 100 times in a row. - Finally, if each parent is heterozygous for A1 and A2, then their offspring have a 25% chance of being homozygous for A1, a 25% chance of being homozygous for A2, and 50% chance of being heterozygous for A1 and A2: - What about one gene found on Chromosome pair 1 and another, different gene found on a different pair of chromosomes (call them the Chromosome 7 pair)? - We know that two pairs of chromosomes can now form four possible combinations in gametes. - Let's add another gene to Chromosome pair 1 called B, again with two alleles B1 and B2. - Let's just look for now at two parents who are heterozygous for two alleles on gene A and two alleles on gene B (on the separate pairs of chromosomes). - You can see there's a 25% chance the offspring will be the same as the parents: heterozygous at both A and B (highlighted green). - Then there's a 12.5% chance each of A1A2B1B1, A1A1B1B2,, A1A2B1B1, or A1A2B2B2 (one gene homozygous, these are highlighted by color in pairs). - There's a 6.75% chance of A1A1B1B1, A2A2B1B1, A1A1B2B2, or A2A2B2B2 (both genes homozygous). - The point is that this is merely the consequence of meiosis and fertilization. - Different pairs of chromosomes make random combinations in gametes and these gametes randomly combine during fertilization to create zygote genotypes. - What if the A and B genes are on the same chromosome pair? - Then, we say that they are linked genes or genes under linkage. - What happens to the patterns of alleles for linked loci? - They are inherited as if they were one gene as far as independent assortment goes. - Because the genes are linked physically on the same chromosome, so when chromosomes assort during meiosis they are literally stuck together on the same chain of DNA. - So now you have a 25% chance of A1A1B1B1, a 25% chance of A2A2B2B2, and a 50% chance of A1A2B1B2. - These are the same ratios as a single gene on a single chromosome pair because it's the chromosomes that are being assorted and not the genes. - Since the genes (and specifically these alleles) are linked here and don't assort independently, they are inherited together. - Could we have had A1B2 or A2B1 alleles be linked? - Sure! - It changes the outcome for the genotype chances in the zygote, but not the general process.

Recombination

- Let's say you have an individual that is homozygous for alleles A1 and B1 that are linked on the same chromosome. We can represent them in a crude text format like this: -----A1------B1----- -----A1------B1----- - Here we have two chromosomes (the lines) with two loci (A and B) and the individual is homozygous. This individual will only produce A1B1 gametes. The two chromosomes are different colors so that we can show the next step. Let's say crossing over occurs at a point (the "\") between A and B and look at the results. -----A1------B1----- -----A1--\---B1----- -----A1------B1----- -----A1--\---B1----- This individual will still only produce A1B1 gametes even with crossing over. Crossing over did happen, but because each locus was homozygous, no new combinations of alleles were made. Let's look at a second example, where now the individual is heterozygous for A, but still homozygous for B: -----A1------B1----- -----A1--\---B1----- -----A2------B1----- -----A2--\---B1----- Before crossing over the gametes would be 50% A1B1 and 50% A2B1. After crossing over, the gamete genotypes would be... exactly the same. Again crossing over happened, but we did not change the outcome of meiosis in terms of the gamete allele combinations for the two loci. Let's look at a third case, now where each locus is heterozygous. -----A1------B1----- -----A1--\---B2----- -----A2------B2----- -----A2--\---B1----- Before crossing over the gametes would be 50% A1B1 and 50% A2B2. After crossing over, the gamete genotypes would be 50% A1B2 and 50% A2B1. Now we actually changed the combinations of the A and B alleles in the resulting gametes. Now crossing over produced a recombination event. - A recombination event means that the combinations of alleles changed on a pair of chromosomes, so the combination of genetic information in the alleles has been altered. - As we saw in the first two examples, it's possible to have crossing over without recombination. - Crossing over is the physical exchange of chromosome segments. - Recombination requires a change in the "combinations" allele information (a "re"-"combination"). - The third example illustrates the first condition that is necessary for recombination: having two heterozygous loci. - You cannot have recombination unless you have two heterozygous loci to form new combinations. - To get the second condition of recombination, we need to look at a fourth example: -----A1------B1----- -----A1------B1-\--- -----A2------B2----- -----A2------B2-\--- - Here we have two heterozygous locus and we have a crossing over event, but recombination did not occur. Why? Because the crossing over event was not between the two heterozygous loci. - For recombination to happen you have to have crossing over occur between two heterozygous loci. If the (1) either of two loci is homozygous (2) crossing over does not happen between the two loci, then you will not have recombination occur. - No new allele combinations will be generated. - Recombination is the mechanism by which a pair of linked alleles might form new combinations. - Crossing over happens just before independent assortment in meiosis and creates a chromosome with a mix of alleles. - Crossing over can happen multiple times on the same chromosome, although generally each recombination event is considered an independent chance. - So, if the probability of one recombination event is r then the recombination of two crossing over events on the same chromosome is (effectively) r times r (or r2). - This is the same as the odds of flipping heads on a coin twice in a row. - The first time the odds are ½ and the second time the odds are ½ so the odds of two heads in a row is ½ times ½ or (½)2=¼.

Genetic Drift

- Let's start with a simple simulation of a haploid asexual population for a single allele. - Start by dealing a vertical row of 6 cards, three yellow and three blue. This will represent our first generation with six individuals a polymorphic allele (yellow and blue are the alleles). Y Y Y B B B - Now make a pile of cards on the side that has twice as many cards by color as the previous generation. - So 2x3 = 6 yellow and 2x3= 6 blue go in the side pile. - Now shuffle and draw six cards randomly from the pile of twelve. - Lay them out in order with blue on the bottom. - For my result, I got two blue and four yellow. - Now we repeat the procedure. - This time our side deck will have 8 yellow (2x4) and 4 blue (2x2). YY YY YY BY BB BB - Now I draw 6 again: YYY YYY YYY BYB BBB BBB - What happened? - Well, by chance, I drew 3 blue and 3 yellow, so the allele frequency of blue randomly "drifted" back to 3 and 3. - Let's repeat the procedure a few more times. YYYYYYYYY YYYYYYYYY YYYBBYYYY BYBBBBYYY BBBBBBYYY BBBBBBBBY - At first blue rises up to 4/6, holds for a generation, then declines to 3/6, then 1/6, holds at 1/6, then zero. - These ups and downs are no more than the outcome of random "sampling errors", in which the frequencies of blue and yellow in the parents were not reproduced reliably in the offspring. - We learned a few things here: - Random means there are no trends of up and down. - If we go up, we still have a 50/50 chance of going up or down. - If we go down, we are not more likely to keep going down, it's just random. - We don't have to go up or down by 1 at a time. - But, because we use the previous generation to draw the next generation, having frequencies similar to the previous generation are the most likely outcomes. - So changes tend to be small, but they do not have to be small. - We could draw any possible outcome x/6 at *any* time. - Some outcomes are just more likely than others. - What happens now that yellow is 6/6? - Our draw pile will be 12 yellow cards. - So we can only draw 6 yellow cards, for the next generation and now for all future generations. - We can end the simulation. - In the term of population genetics, the frequencies of yellow and blue bounced around for a while until yellow became fixed. - This means that the last chromosome carrying the last blue allele simply failed to be inherited (there could have been many reasons). - This simulation by cards mirrors the process by which alleles are inherited. - Imagine that yellow and blue represent two alleles of a given gene. - The blue "allele" is passed to children at a frequency proportional to its frequency in parents. - If most of the parents in a population have some allele, then under genetic drift most of the offspring will also tend to have that allele. - But over time, an allele's frequency might increase and become common, or be lost entirely and go to zero frequency. - Once lost to the next generation, it's gone! - Forever? - Not necessarily, but for a very long time until the exact mutation occurs again by chance. - If your genome is very large and the mutation rate is slow (like humans) having the same mutation recur is extremely unlikely. - So genetic drift is the general term for the process of change in allele frequencies due to stochastic effect (mating, chromosomal assortment, recombination), and not dependent on the actual effects of alleles on the phenotype. - When one allele goes to 100% frequency, then we call that a "fixed" allele (the other allele is "lost"). - When there is variation in the allele in a population, then the site/gene/trait (whatever the alleles are alternatives of) is called polymorphic (many forms). - The imaginary yellow-blue gene is polymorphic until yellow becomes fixed. - Given infinite time, drift will always result in fixation of one allele or the other. - Why? - Most simply because even if allele frequencies were completely random (instead of based on the current generation) eventually (if you draw infinite times) you will eventually get all one allele. - Just by random chance. - The key is this is "eventually" and can take a very long time. - What controls how strongly drift acts and how fast fixation happens? - The size of the reproductive population. - If we do the same simulation with only two individuals (1 blue, 1 yellow) we arrive at the fixation of blue or yellow in the first generation 50% of the time. - If we did the simulation with 100 yellow and 100 blue, then it would take many more generations for random chance to cause either allele to go extinct through random fluctuations.

Penetrance and Genome-wide Association Studies

- Most traits are not on-off switches that only depend on a single gene. - Most traits are controlled by a complex network of genes that produce variable effects and interactions with the environment. - How do we disentangle genetic and environmental effects on various phenotypes. What about human medical conditions? Can we use this framework to disentangle these effects? - The most common framework for these questions at present is called a genome-wide association study (GWAS). - In a GWAS study, you collect genomic sequence data from (1) a set of individuals with a particular condition (affected population) and (2) a set of other individuals without a particular condition (control populations). - Then we compare the positions that have variation in the nucleotide sequence. - These individual nucleotide sites that are variable in the human population are called single-nucleotide polymorphisms. - A polymorphism means that within a population there is variation in the genomic sequence at a single nucleotide position (for example, some individuals have A and some have G). - A fixed site means everyone in a population has the name allele. - The next step is to do a statistical test on every SNP to see if the affected population has a certain allele more often than the control population. - In an extreme case, if all of your affected individuals for a given genetic condition have a "A" allele at Chromosome 12, position 14315, but all of your control individuals have "C", then you would have strong inferential support for the statement that the "A" allele at Chr12:14315 is "associated" with the phenotype. - The end result of the GWAS study is a lot of statistical tests. - These are typically plotted in a "Manhattan plot". - The plot shows each position on the genome on the x-axis. - The y-axis is the logarithm of the P-value, which is a probability that we can reject a "hypothesis of no effect" that control and affected individuals have the same allele frequencies. - The points highest on the y-axis have frequencies that are so enriched that we are 99.9999% sure that we can safely say the alleles are significantly enriched in the affected population relative to the control population. - For GWAS to work, we need a lot of individuals in both the control and affected populations. - Why? - Because of the environmental effects. - If the effects of the environment are really strong for a given trait (often we think this is the case for cancer, heart disease) then we need to design our study to control for all these effects. - The other problem is what if the trait of interest is highly polygenic with many alleles of minor effect. - You might sequence 100 affected individuals who have 100 completely different mutations that all result in the same genetic condition, but this test would not work because we need to see the same allele repeatedly happening in the affected population. - In this case, the allele is said to have low penetrance. - Penetrance refers to the degree to which having a particular genetic variant is associated with having the phenotype. - Alleles of major effect would have high penetrance. - For example, cystic fibrosis is known to be the result of a specific mutation that confers the symptoms fairly directly. - This CF-associated allele has high penetrance. - Genes that make minor contributions to heart disease would be low-penetrance. - You also need a lot of affected individuals in order to have the power to test these patterns statistically with confidence, so the genetic basis of rare genetic conditions can be difficult to determine these methods.

Epilogue: Humans are still fascinating

- Much of this tutorial has focused on explaining how modern humans fit into the scope of the biological diversity and history of the earth in ways that explain how the universal principles of evolution apply to us the same as any other organism. - This explanation can be misinterpreted as saying that humans are not biologically unique, which would be taking this kind of thinking to the other extreme. - Humans have amazing and apparently unprecedented capacity for language, memory, cultural rituals, complex social organization, and technology that has transformed the planet. - Our unique traits are undeniable. - So we can close by cautioning against both extremes of thought. - Scientific evidence tells us that humans are neither (1) so different and removed from the rest of nature that the processes of evolution and biological inheritance operate in a fundamentally different way (2) are they so commonplace that we should fail to appreciate the unique ways we do interact with Earth and the other organisms living here.

Drift, Selection, and Speciation

- One last important point is that speciation does not require natural selection. - New reproductive incompatibilities whether genetic or phenotypic (encoded by genes) enter in the population through mutation and recombination and can rise to high frequency under drift alone. - The antelope example was entirely speciation under mutation and drift in the absence of phenotypic selection. - Selection could accelerate the process substantially and has in the past produced some dramatic examples of adaptations that evolved under selection which also caused reproductive isolation. - For example, populations of flowering plants adapting to open their flower and produce pollen at different times of year to maximize their contact with their local insect pollinators have been shown to adaptively evolve to have different flowering times. - Each population has locally adapted to maximize their fitness in regards to the local pollinators, but then the populations with different pollinators have adapted to flower at different times. - Since they have different pollinators at different times, this creates a prezygotic barrier as the result of adaptive selection. - So we can disentangle speciation and selection, since sometimes the reproductive isolation (or genetic differentiation in asxeuals) will be driven by selective effects and sometimes just the background of drift and mutation.

Bottleneck and Assortative Mating

- So far we have only considered the effects of mutation-drift-selection balance on genetic diversity in a population where the size of the population has not changed. - However, consider a scenario where habitat destruction has reduced a population of bobcats from 100,000 to 1000. - That means for a diploid bobcat population, the number of chromosomes randomly mating in a population has reduced from 200,000 to 2,000. - Assuming the starting population is fairly genetically diverse, having 99% of the individuals going extinct means a lot of the genetic diversity that has built up over time through mutation and recombination will go extinct too. - When a severe reduction in population size leads to a reduction in genetic diversity too, this is called a genetic bottleneck. - Even if the habitat is restored and the census size (total number of individuals) of the population rebounds to 200,000 again, those 200,000 individuals will have lower genetic diversity because they are more recently related genetically to the 2,000 individuals that represent the founding population at the bottleneck minimum. - So their genetic diversity at the end is much lower than the original 200,000. - This is also similar to a founder effect, where a subset of organisms that move to a new environment might have unusually high frequencies of certain alleles and a small population size that might make these previously rare alleles become fixed in the newly established population. - In both bottlenecks and founder effects, the allele frequencies of a population are rapidly shifted and the overall diversity is reduced by the founding (or re-founding) of a population by a small subset of the original population. - One practical problem that genetic bottlenecks present is that by making the population smaller and more genetically homogenous, they become more sensitive to chance extinction effects like pathogens or perhaps from food competitors that might expand their populations. - This is why even severe reductions in the sizes of endangered species (even if they do not go fully extinct) can leave the species in a precarious situation for the survival of the species. - This highlights that the strength of the genetic drift is inversely proportional to population size

Selection and Allele Masking

- One last item to mention with respect to selection relates to dominant and recessive alleles. - Recall that in a diploid organism, one allele may be dominant with respect to the other allele. - This means that only one allele affects the phenotype. - Let's say we have a squirrel with grey fur phenotype that has higher relative fitness to a recessive red fur phenotype. - The grey phenotype can be either a GG or Gg individual and the gg individual is red. - That means that both homozygous and heterozygous grey individuals will have higher relative fitness compared to gg red squirrels. - The Gg individuals will reproduce just as well as the GG ones, and both Gg and GG are higher fitness than gg. - At some point, the number of gg red individuals might have become very small, and they pass fewer g alleles to the offspring generation, but there are still Gg individuals reproducing (which can produce gg offspring). - This means that even though a recessive allele is selectively deleterious or disadvantageous in a homozygous individual, the dominant allele masks the recessive allele and so it is maintained at low frequency in the population even in the face of natural selection against it. - This is the case for many recessive alleles associated with many health-related human genetic conditions. - If the condition-causing alleles are rare, then it's very unlikely that two heterozygous humans would meet and have children that would be homozygous recessive and thus have the genetic condition. - More often, the alleles will be masked because heterozygotes probably mate with homozygous dominants when the recessive allele is rare. - Although the recessive allele is deleterious, it will not drop all the way to zero frequency in the population under this negative selection.

Genes and Environment

- One of our most important points about biology is that phenotypes are always some combination of genetic and environmental effects. - Even when the environmental effect is weaker, there is always some effect of the environment on traits. - This manifests sometimes as just small adjustments on the baseline genetic effect on a trait, and sometimes to completely change the relationships of genes and their associated traits.

Additive Traits plus Environmental Effects

- Previously, we have discussed how additive epistatic genes can combine their effects to produce phenotypes. - This answers part of the question of how the effects of discrete genes can combine to form the distributions of quantitative traits that we observe all around us (for example, human height). - However, for the distributions we simulated in the last Tutorial, the phenotypes are still in relatively discrete units. - There would have to be a large number of genes with various effect sizes to get the kind of continuity we see among populations of real organisms. - So what's the missing piece? - What's missing is that along with the phenotypic variation caused by genotype, environmental and random effects also add variation as they impact the phenotype. - It's important to conceptually distinguish consistent environmental effects from just random chance variation. - Populations living in a habitat will all experience the consistent environmental effects they share, but each individual in that population may still have random chance variation during their lives. - For example, all sunflowers in a field will experience a similar temperature cycle, but if one sunflower grows in a patch of soil that happens to have fewer nutrients then it might not grow as well. - Here, the temperature is the consistent environmental effect and the chance of soil is a "random" effect. - What happens with environmental effects is that they create an additional source of phenotypic variation on top of the genetics. - A plant might have the genotype that would normally confer the tallest phenotype, but if it is growing in an environment that is providing insufficient resources for maximal growth, then the actual phenotype will be shorter. - Conversely, you might have an environment with particularly abundant resources that allow a plant with a medium-height set of alleles to grow taller. - The bottom chart below shows what happens when we take our baseline measure of 5cm, add the distribution of four genes with 2cm additive alleles, and then add a 1cm random effect. - Each individual randomly gets 0cm or 1cm from the environmental effect. - You can see now our discrete graph from 6.3 (no top) gets the missing values "filled" in by the random addition of +1 to some of the individuals. - Some individuals with values of 7 on the top will become 8 due to the environmental bump, 9 becomes 10, etc. - You can see we get much closer to an actual continuous normal distribution of values. - To simplify the data sheet, we're using integers, but in real biological systems the effects of environment will be continuous and we will have a more smoothed distribution. - You can also try out some more extreme values of random environmental variation. - Let's turn the effect up to a range of 0-20. - We can see that the distribution is starting to flatten out and become a uniform distribution where all values are equally likely. - However, many environmental effects also have a normal distribution with a central tendency, so non-random environmental effects will reinforce the normal shape of the phenotypic

Does phenotypic selection lead to genetic selection?

- Recall from the previous tutorials that most traits are affected by many genetic and environmental factors in complex relationships, rather than single Mendelian genes. - In the rare case where a single allele of a single gene is responsible for the phenotype, then selection of the particular variant form of a phenotype will lead to selection of its associated genetic allele. - In other words, if a particular allele has high penetrance (strong correlation between the presence of the genetic variant and phenotypic variant), then genetic selection will accompany the phenotypic selection closely. - However, consider a highly polygenic trait with variants in hundreds of genes affecting the phenotype. - If selection is acting on the phenotype, then any combination of genetic alleles that leads to the high-fitness phenotypic variant would be selected instead of a specific gene or nucleotide allele. - So the effect of selection might be more diffuse at the genetic level. - This is generally true of highly additive traits influenced by many genetic alleles. - The phenotypic selection selects for a random subset of the hundreds of genetic alleles in various phenotypically selected individuals, meaning the combination of selected alleles at the genetic level will be variable even as the phenotypic selection is consistent. - Also recall that heritability is required for directional selection to occur. - This means the variation in the trait under selection must have some genetic basis, rather than being controlled entirely by environmental factors. - So in the extreme case of a trait with strong environmental factors and weaker genetic factors, selection on the genetic traits may be lessened.

Reproductive isolation

- Reproductive isolation occurs in sexually reproducing organisms when two individuals cannot merge their genes, or, in other words, when they can no longer make offspring that jointly inherit their genes. - Mechanism of genetic isolation can be prezygotic isolation mechanisms or postzygotic isolation mechanisms. - Prezygotic isolation occurs "before the zygote," meaning any factor that might prevent fertilization. - This can be geographic (organisms never meet), mate choice (organisms never mate), physiological incompatibility (mating cannot occur), or biochemical/cellular biological factors that inhibit fertilization (anything biochemical that prevents gamete fusion). - Postzygotic isolation occurs if the genotype of the fertilized zygote is inviable or develops into a sterile adult. - Either way, fertilization is successful, but this particular genotype combination cannot be passed to offspring. - Consider a scenario where a large population of antelope inhabits grazing lands that span a river bed that is usually dry. - The river flow changes and now the river is deep all year and impassable by the antelope. - This is called vicariance where the two creatures have been physically separated in their habitat rages. - Vicariance, in this case, means there is no migration between the populations. - Since migration is inhibited, this is a prezygotic reproductive isolation mechanism. - Once separate, the two populations will mate only within their own populations. - Over time, drift will cause their allele frequencies to differentiate and any new mutations will be restricted to only one subpopulation (the same random mutation occurring in both populations is unlikely). - Selection might also influence the two habitats differently, selecting different traits or different phenotypic variants of the same trait. - Since they are reproductively isolated, the frequencies of genotypes and phenotypes can differentiate because they are no longer randomly mating as a unified population. - After a time, we will be able to detect distinct patterns of alleles in the two populations, meaning they are now genetically differentiated. - After a couple of million years, the water flow changes again and the river bed returns to it's mostly dry state. - Now the antelope can freely move between the previously isolated population ranges and find mates in the other subpopulation. - However, if the process of genetic differentiation has caused genetic incompatibility between individuals from the two populations, then they will still be reproductively isolated. - The biogeographic prezygotic isolation of the river has led to the development of a genetic postzygotic isolation mechanism in the form of a genetic incompatibility. - What does this mean on a cellular level? - It means that any combination of gametes from the two populations does not form a viable zygotic genotype, perhaps by a combination of alleles that is epistatically lethal.

Aneuploidy

- Sometimes during meiosis a gamete might receive more or fewer than the usual copies of a certain chromosome. - This often occurs by non-disjunction: the failure of replicated chromosomes to separate during meiosis. - As a result, the fertilized zygote might have too many or too few of any particular chromosome (or chromosomes). - In humans, most aneuploidies are lethal at very early developmental stages, with two notable exceptions. - First, individuals with Trisomy 21 or Down Syndrome carry three copies of chromosome pair 21. - Second, roughly 0.5% to 1% of human individuals have a karyotype of sex chromosomes other than XX or XY. - The most common are Turner syndrome (45,X; missing an X or Y) and Klinefelter syndrome (47,XXY; two X and one Y). - There are many other rarer forms (XXX, XXXY, XXXX, XYY, etc). - Some forms have few consequences on traits, while others may cause mild to severe infertility, changes in height, intersex characteristics, or other developmental changes. - Thus, the idea that all humans are XX or XY and are either biologically male or female is unsupported by genetics.

Interactions of separate genes for a trait

- The models above are quite oversimplified, as we have discussed earlier when we talked about regulation. - You know now that biochemical pathways and genetic regulatory networks in cells are often monstrously complex and involve dozens of genes. - In humans, very few traits are Mendelian (one-gene-to-one-trait). - Most traits are genetically epistatic (affected by several genes) and many genes are pleiotropic (affect several traits). - Traits affected by multiple genes are polygenic traits. (1) Scenario 4.1 (epistatic redundant): Consider a situation where you have two genes A and B and each, all by itself, causes red coloration in a bird. - This means you can be homozygous for loss-of-function alleles in either gene without losing the red coloration trait. (2) Scenario 4.2 (epistatic non-redundant): Consider the mice shown below. You can see that C is dominant w.r.t. c for the brown coat. - However, B is dominant w.r.t b for the black coat color, but only produces the black coat trait when the dominant C allele is present. - So, the B allele is epistatic with the C allele in producing black coat color, since black coat color depends on both B and C being present (which are alleles at two separate loci). - BB or Bb with cc in a genotype results in white mice. - Key Point: In addition to alleles at the same locus interacting (dominant/recessive), alleles at different loci can interact epistatically w.r.t. to a single trait.

Models of speciation

- The scenario described above is called an allopatric speciation model (Greek: allo=other, patria=homeland). - Allopatric speciation requires the physical separation of two related populations on "other homelands." - This physical separation means the divided populations evolve independently, and, given enough time and differentiation, will form reproductive incompatibility. - This reproductive incompatibility makes them separate species under the definition of the Biological Species Concept. - Two other models of speciation are commonly discussed in evolution. - Peripatric speciation also involves spatial isolation, but not so extreme. - In peripatric speciation, the peripheral population experiences a different set of evolutionary processes, often strong selection in its peripheral environment. - So this does not involve a physical separation like allopatry, but rather a strong difference in selective factors at the fringes of a population's range that promote differentiation. - Parapatric speciation is essentially the same model, just there is no physical separation of habitats, one population is merely on the fringes. - Sympatric speciation involves the establishment of genetic incompatibilities without physical separation or selective environment disparity. - This by far the rarest model in nature (some researchers even question its existence entirely). - The possible mechanisms are many, but commonly proposed genetic causes are changes in chromosome number (either fission or fusion of chromosomes), changes in ploidy (especially in plants), or establishment of new sex chromosomes.

An Unadorned Example of Natural Selection

- The struggle for existence in the peppered moth (Biston betularia) is canonical knowledge for all students of evolution. - This population lives on the island of Great Britain. - It comes in two colors: a wild-type that is basically white with speckles and a melanic, all-black form. - Before the Industrial Revolution, the wild-type predominated. - During the day, the nocturnal moths rest quietly on rock surfaces and hope that bird predators do not notice them. - The wild-type look blends well with rocks speckled with lichens (the coloration is "cryptic"). - Melanics were conspicuous on most available surfaces and were removed quickly by birds, thus failing to have many melanic offspring. - In the struggle for existence, wild-types had a big advantage, and most offspring were wild-type offspring of wild-type parents. - Then the Industrial Revolution and complete lack of emission controls colored all surfaces near industrial centers black. - Suddenly the selection mediated by birds switched to favor melanics, and during the Industrial Revolution period the melanics had the advantage around Liverpool, Birmingham, and London, but not in the rural Scotland, Wales, and the southwest. - Now in modern times, the forests are not black, and the melanics are rare again, due to their conspicuousness to birds. - We see directional selection favoring the wild-type, always keeping the melanic form right at the extinction level, eroding genetic variation in color as soon as it arises. - Mutation keeps injecting the melanic allele; directional selection keeps kicking it out. - Then the directional selection reverses for a time. - And then the directional selection reverses back.

Preface

- The topic of human evolution is a contentious and sensitive one in contemporary culture. - Humans have many personal, cultural, philosophical, and religious beliefs both about the origins of humans and the human experience of life. - In this tutorial, we are going to focus only on addressing the scientific models and some pseudoscientific misconceptions of evolution. - Pseudoscience consists specifically of statements, beliefs, or practices that are claimed to be both scientific and factual, but are incompatible with the scientific method. - Pseudoscientific claims assert the authority of the scientific method to state facts that are, in reality, not supported at all by rigorous science. - We are specifically not going to dissect or analyze any person's or group's philosophical or religious beliefs, folk wisdom, legends, stories, aphorisms, adages, or personal beliefs and opinions. - We're going to focus entirely on pseudoscientific arguments and common scientific misconceptions. - The historical debate over human evolution (and evolution in general) is a fascinating topic about which dozens of books have been written, but not a topic we will investigate here.

Molecular evolution

- There are other questions in evolution that can be answered with phylogenetic approaches to molecular sequence information. - One example is the study of the evolution of gene families, which can yield interesting insights about the evolution of genes themselves (not just the species relationships). - Gene families arise when genes are duplicated (as a replication error, for example) and then the two copies can sometimes change to have different functions. - Figure 6.8.6 shows the evolution of the globin family of genes (as in hemoglobin). - Mammalian adult hemoglobin is made up of two alpha-globins and two beta-globins. - Hemoglobin is the molecule in red blood cells responsible for carrying oxygen. - Note that the duplication events ("the dots" in Fig 6.8.6) created several new globin genes in the genomes of mammals (compared to amphibians and sauropsids). - The gamma-globin specific to eutherian mammals (placental mammals that carry offspring in a uterus) is particularly noteworthy. - In fetal blood cells, hemoglobin is made of two alpha- and two gamma-globins (instead of beta-). - The gamma-globin binds oxygen more strongly and so can "steal" oxygen away from the maternal blood cells. - This capability arose through a duplication and slow mutational refinement of the epsilon-globin, which is the hemoglobin active in amniotic yolks! - Another example comes from beta-defensin genes (Figure 6.8.7). - These are normally genes used in the immune system to fight bacterial pathogens. - However, in snakes, lizards, and platypuses, copies of these genes have been duplicated and adapted into venom proteins. - Yes, platypuses have venom, and what's more that defense against predators evolved from a duplicated gene originally for defense against microbial pathogens.

Extinction

- There is abundant research on the process by which new species formed, but considerably less on the many reasons they might go extinct. - In part, this is a bias of "the now," meaning that we can much more easily study extant species than extinct ones. - However, the vast majority of species in the Earth's history are already extinct, and many models of evolution try to simultaneously consider the "birth" of a new species (speciation) and the "death" of a species (extinction) in terms of the phenotypic, genetic, and/or environmental factors that might promote the continued survival of lineages over hundreds of millions of years or their extinction after only a few generations. - Extinction is the most powerful, and yet in many ways the most mysterious, force in evolution because the ways of survival are few but the ways of extinction can be many and the evidence is buried underground instead of scurrying underfoot.

Evolution is just an extension of the principles of heredity

- There is no special dividing line between where simple inheritance stops and evolution takes over. - Evolutionary patterns are the bigger picture of how hereditary processes play out over many generations or millions of years in time. - But historically, "evolution" has acquired a complex social dialog around it as something separate and special. - The complexity of the subject, and uncertainty inherent in historical inference, creates opportunities for misinterpretations and misconceptions to arise. - From the earliest formal considerations of heredity in a modern scientific context (more than century before Darwin), we struggled to understand how small changes could render the vast biodiversity we see in the present day. - The next two sentences, if you keep them firm in your mind, will prevent a lot of misconceptions about biology: (1) The random process of mutation creates the variation we observe, while forces like selection and drift can only shape it after the fact; (2) evolution does not proceed in a particular direction or trajectory and each step is determined by the circumstances of the moment.

Recombination Mapping

- There's another application of recombination that is useful to think about and formed the early technique (before genome sequencing) by which the relative locations of genes on a chromosome were estimated. - Let's consider an individual with a chromosome with three loci A, B, and C that are all heterozygous and specifically in the combination where one chromosome has A1B1C1 and the other has A2B2C2 . - This is exactly what the researchers studying fruit flies in the 1910s created. - They specifically bred fly offspring that had three heterozygous loci linked in the same configurations. - When they genetically crossed (bred artificially, by design) two triple-heterozygous fruit flies and then observed the offspring, what they noticed was that some recombined offspring genotypes were more common than others. - Let's say, for example, that they observed that A1B1C2 and A2B2C1 each occurred ~4% of the time, but A1B2C2 and A2B1C1 only are observed ~2% of the time in offspring. - What would cause this discrepancy? - Assuming crossing over is effectively random across all points in a chromosome, then what would cause these different recombination rates? - The answer is that loci are not all equidistantly arranged on the chromosome. - Some pairs of loci are farther apart than others. - If you imagine the chromosomes as a piece of rope with beads on the rope marking where the genes are. - Then imagine you take a pair of scissors, randomly cut the two ropes at the same position and then swap segments. - If you randomly cut with your scissors on the rope, then you have twice as much chance to cut between two genes that are 12 inches apart as you do to cut between two genes that are 6 inches apart. - This means the rate of recombination of two genes is proportionate to their physical distance apart. - If you observe A1B1C2 and A2B2C1, these are each the result of a recombination of A+B relative to C (A1+B1 and A2+B2 stay linked, and C crosses over). - A1B2C2 and A2B1C1 are the result of a crossing over for A relative to B+C, which remain linked. - Since recombination between C and A+B occurs twice as often as between A and B+C, this means that the B-C distance is twice as far as A-B. - Perhaps something like this: -----A1--------B1--------------C1----- -----A2--------B2--------------C2----- - Since there's twice as much DNA on the chromosome between B and C as A and B, then recombination rate should be twice as often. - Random crossing over events should happen twice as often on a segment of DNA that's twice as long. - You make one more observation, which is that A1B2C1 and A2B1C2 occur extremely rarely at 0.008% of the time. - Does this make sense? - How would these haplotypes occur? - The process would have to look something like this: -----A1--------B1--------------C1----- -----A2--------B2--------------C2----- -----A1--\-----B2------\-------C1----- -----A2--\-----B1------\-------C2----- - Producing a recombinant chromosome with the "middle" (B) allele recombined relative to the two one either side requires two crossing over events simultaneously. - Since both pairs of loci are heterozygous in this case, this leads to a double-recombination event. - Double-recombinations require a lot of coincidences to happen. - Two crossing over events have to happen at points exactly between the loci of interest at the same time. - Then you have to randomly select the chromosome during independent assortment. - So, you're multiplying together several random draws of fairly low frequency. - However, if you can detect these low-frequency recombinants in our offspring, then you know which locus in the center locus that requires two recombinations to be recombined relative to the other two loci. - By doing this experiment repeatedly for various trios of loci, you can slowly map out the locations of various genes. - This is incredibly painstaking work requiring artificially breeding of lots of organisms, which is only possible for a precious few model systems. - With the advent of genome sequencing, we are now able to produce much more high resolution maps of chromosomes.

Selection

- To answer this, let's do the same simulation again. - But, this time, every time we draw a yellow card then we also flip a coin. - If the coin lands heads, then we discard the yellow card and draw again. - Blue cards do not have to pass a coin flip, we just accept them automatically. - This process gives blue an advantage. - Blue cards have a relatively higher chance of making it into the offspring draw pile and yellow have a relatively lower chance. - Let's play this out over a few generations. BBBBBB BBBBBB BBBBBB YBYBBB YYYYBB YYYYBB - What happened? - Without the coin flips, my first six draws were four yellow and two blue. - So without the disadvantage of the coin the final frequencies would have been 4Y and 2B. - But I threw away three yellow on "tails" flips of the coin. - So the inherent relative advantage of the blue allele over the yellow meant that blue increased in the actual pool. - In the second round, yellow survived more coin flips and increased slightly. - Then blue increases again and finally yellow is lost. - If we run this over and over again, blue should fix a lot more often than yellow. - Yellow still could fix by chance due to genetic drift (randomness is always operating even when there is implicit relative advantage), but it has to be randomly drawn and pass through the coin challenge as well. - So yellow is at a disadvantage to fix, while blue is at an advantage for fixation. - If we have two alleles and one is more often passed on to the offspring than the other due to intrinsic advantage to reproduction of the allele itself, then this is called selection (or allele selection or genetic selection). - So unlike genetic drift (where allele frequencies just change at random) selection has occurred when one allele had intrinsic properties that made it more or less likely to be passed to the next generation relative to another allele. - How could a genetic allele affect its own inheritance probability? - Let's think about the processes that we said are random in drift: mating success, number of offspring produced, and independent assortment. - Independent assortment is a cellular process, so we will assume that it stays random. - If we just say unsuccessful mating is just "zero offspring produced", we can now lump together mating success and number of offspring into just "number of offspring" or fecundity. - When two individuals produce different numbers of offspring in the next generation, we say they have different relative fitness. The individuals contributing more offspring (and thus passing on more alleles) to the next generation have higher relative fitness, while the individuals leaving fewer offspring (and thus fewer of their alleles) are lower relative fitness. - So what is selected is certain individuals with an overall phenotype that caused them to have more offspring. - The effects of the phenotype on relative fitness are key here. - We can have changes in the frequencies of a given phenotype by chance under random drift. - But if the phenotype itself is correlated with having more offspring, then it has the implicit advantage we've been discussing. - This means selection operates directly on phenotypes, and only indirectly on genotypes and alleles. - For example, a blue fish with better camouflage against predation might be able to mate more and produce more offspring than a yellow fish that is easily found by a predator. - In this case, we say that the blue fish has higher relative fitness than the yellow fish. - If blue and yellow were equally camouflaging to the predator, then either might still increase randomly. - But the advantage in reproduction of the blue phenotype creates a differential in relative fitness. - The actual number of offspring it leaves behind is higher. - These offspring have the same advantage in reproduction in the next generation, and so will generally make even MORE blue offspring (proportionately). - What does this mean over several generations? - If the blue trait has higher relative fitness, then it will increase in the population over time. - The genetic alleles associated with the blue trait also increase over time. - This is called "positive genetic selection" of the blue trait, because a positive effect of the trait leads it to have higher relative fitness and be selected to higher frequency over time. - The yellow trait and its associated alleles is (by definition and by default) under "negative genetic selection" and thus should decrease over time. - If we think back to the card-plus-coin flip simulation we did, we might make the coin flip stand in biologically for getting eaten by a predator, and you can see how advantage that provides. - Selection operates at the phenotypic level, and occurs when some phenotype has an inherent advantage in producing more offspring over another phenotype. - The first generation that inherit higher fitness alleles should also express the same higher-fitness phenotype, so they will in turn produce more offspring than the inheritors of low-fitness alleles. - Under this logic, the offspring generation must have a higher representation of these helpful alleles than the parent generation did. - If the allele with high relative fitness became so common that it was fixed, then relative fitness disappears because all individuals have the same phenotype (until mutation adds new alleles leading to new phenotypes). - This underscores another key point, only traits with variation can "be selected." - There must be two or more trait variants to exist in a relationship before there can be relative fitness between them.

Genetics, Environment, and Interaction Effects on Phenotypes

- VP=VG+VE+ Equation 6.4.1 - Equation 6.4.1 is one of the more important conceptual equations in genetics and evolution. - This equation is stated: "the total variation in phenotype is equal to the variation in phenotype from the inherited genetic factors plus the variation in phenotype coming from the non-inherited environmental factors plus random error". - We have already said the distribution of trait values is dependent on the genetics, environment, and random chance. - One of the useful implications of this equation is that some traits might have a strong genetic component where most of the variation in phenotype comes from genetic variation. - Other traits might have very little genetic basis and so the environmental effect dominates the variation. - Does this make biological sense? - Sure. - Hair color has a strong genetic component, but stress can make people's hair fall out or go grey. - Some genetic conditions cannot be modulated or alleviated by environmental effects (like therapies or drugs), and so they have a very strong genetic component. - In practice, biologists can study these properties experimentally by putting organisms with the same genotypes in different environments and observing their trait variation. - Let's look at an example: (1) Scenario 1: Let's say we have several populations of soybeans and we want to know whether they will maintain their yield in the face of climate warming. - Let's start by merely sampling three populations of soybeans from different farms in North Carolina. - We will call these "strains" that have different genotypes: NC1, NC2, and NC3. - We set up an experiment in three temperature-controlled greenhouses set to hot, medium, or cool. - We grow many soybean plants in these environments and then measure their average soybean seed yield per plant. - The data show us that environment (temperature) does have an effect on the phenotype (seeds). - So we have an environmental effect (VE). - Is there a genetic effect? Not really. - There is a little bit of variation in the three strains but nothing consistent. - So this would be a case of strong environmental effect on variation but weak genetic effect. (2) Scenario 2: Now we select strains that we know to be heat tolerant strains that were developed separately in different areas of the world called HT1, HT2, and HT3. - We repeat the experiment and see this result. - Now the three strains show no effect of the heat on the phenotype, so environment is not a source of variation. - But the unrelated strains have different inherent yields from their genetics. - This is a case where the phenotypic variation is coming mostly from genetics, and not from environmental variation. (3) Scenario 3: Finally, we look at two strains of soybean. - One comes from very hot growing areas in Vietnam (VNM), where the strain has adapted over many thousands of years. - The second strain comes from Southern Argentina (ARG) where the weather is quite cool during the growing season. - We run the same experiment. - In the VNM strain, we can see temperature does affect the phenotype. - The ARG strain also has an environmental effect. - However, in this case, the direction of the effect of temperature depends on the genotype. - The VNM has a positive phenotypic association with heat, and the ARG has a negative phenotypic association with heat. - This matches our expectation, since we would assume that something growing well in hot or cold environments might have the strongest agricultural phenotype in those contexts. - This situation is known as a G×E interaction [said "G by E"], which stands for Genotype-Environment interaction. - This means that not all genotypes have the same reaction to the environmental stressor, or (equally true to say) not all environmental effects have the same reaction in all genotypes. - The general shape of the profile of reactions to an environmental factor, known as the norm-of-reaction (equivalently, "reaction norm"), depends on the genotype. - We could look at another example with a soybean strain from Turkey (TRK). - Now we have one that has a positive association with heat, and one with no association with temperature at all. - This is also a G×E interaction because the relationship between G and E has changed from "no relationship" to "positive relationship". - So, the slope does not have to invert to become G×E interaction, only has to substantially change. - The three NC and HT strains did not exhibit substantially different norms-of-reaction under the different conditions. - What does this mean? - It means that sometimes reactions of phenotypes to the environment are genotype-dependent. - This might mean that sometimes these effects might not be predictable and that different genotypes of various organisms might react completely differently to an environmental stressor.

Inheritance of alleles and traits

- We can now think about the scenario that caused blending inheritance to be dubious. - If we cross A1A2 with A1A2: - The offspring produced will be 1/4 no stripes and 3/4 stripes. - You can see how rather than all striped fish (matching the parents) or some blending between the extremes (intermediate greyish stripes), the offspring are either all stripes or no because A1 is completely dominant w.r.t. A2 for black stripes (yes, I want you to always say the w.r.t. and the trait, every time). - Immediately, we can see that this blended inheritance does not work at all. - Some traits seemed to appear with weirdly regular frequency of presence-absence patterns, instead of being a blended greyscale. - This is what Mendel noticed (and others when they discovered his papers decades later). - At this point, we should define one more type of allele. - Let's add an allele A5 such that A5A5 individuals do not survive, but an A1A5 individual is healthy. - This makes A5 a recessive lethal allele w.r.t. A1. - If two A1A5 individuals were to mate, then 1/4 of their offspring would be A5A5 and would not survive. - Often, healthy individuals with a recessive lethal allele are called carriers. - Mutational probability says that we are all carrying a few extremely rare (perhaps unique) very deleterious or lethal recessive alleles. - However, unless the recessive lethal alleles are moderately common in a population, we are unlikely to meet and have offspring with another individual who carries exactly the same recessive lethal allele. - If there's another allele A6 such that A1A6 and A6A6 individuals both do not survive, then A6 is a dominant lethal allele. - These are extremely rare in nature, because only one copy is needed to cause mortality. - Most of these individuals do not survive to be born or have their own offspring, except in rare cases where other genetic or environmental effects act to counteract the lethality.

Some applications of phylogenies

- We can use phylogenies to answer some interesting questions. - Figure 6.8.4 shows a phylogeny of human mitochondrial genomes with the tree drawn onto a map showing the sampling locations of the mitochondrial sequences from indigenous populations. - From the rooted phylogeny we can draw conclusions about the migration of human populations over time. - Models like this have been hugely informative alongside other historical, linguistic, and archaeological evidence toward understanding human migration history. - Another example of the uses of phylogenetics is shown in Figure 6.8.5. - This phylogeny shows the relationships among strains of ebolavirus from various outbreaks. - The conclusion here is that the 2014 strains (red highlighting) form a separate clade from other outbreaks from 1994-2007. - What does that mean biologically? - It means the 2014 outbreak does not share a recent common ancestor with any of the previous outbreaks, and this means its genotype, pathology, and reaction to treatment could be quite different from past strains.

Inferring Phylogenies

- We can't directly observe extinct creatures. - With rare exceptions, most fossils are not even the actual biomatter of organisms but rather just the impressions left in stone. - This means that when we try to figure out a phylogenetic model of how organisms diverged, we have (very) incomplete information. - For phenotypic evidence, we have only the observations and measurements we can make from extant organisms that are alive in the present and the few fossils that have been found from extinct organisms. - For comparing DNA, we are (almost) entirely limited to living creatures (a very few DNA samples from recently extinct animals like mammoths and extinct hominids exist and have been sequenced). - It means that phylogenetics as a science relies a lot on statistical inferential modeling. - We know that we cannot sample the entirety of every organism or species that has ever lived, so instead we have sample a smaller and incomplete set of information from living organisms and fossils and draw "inferences" based on this incomplete evidence, keeping in mind that we are drawing conclusions in the face of a huge amount of uncertainty. - In other words, we can't directly observe what happened 100 millions years ago, so we have try to interpolate what happened based on what evidence we do have, while being really honest about how certain or uncertain we are about various conclusions. - An example will help us understand the incompleteness of knowledge. - You can see now that at a certain homologous genome position, gorilla, human, and mouse have a "T" nucleotide, and frog, elephant, and gibbon have a "G" nucleotide. - Remember, all we know as actual evidence are the data at the bottom and our best guess overall about what the phylogenetic tree is. - So, what is the actual timing and order of the nucleotides, and what nucleotide did the ancestor of the three primates (on the right) have? - Two simple models are possible. - On the left, we can construct a model with two G-to-T substitutions on the human+gorilla ancestor and mouse branches. - On the right, we could have a G-to-T happen in the ancestor of rodents+primates and then a reverse mutation from T-to-G on the gibbon branch. - There are also many other possible models, but these two require the fewest number of changes while still explaining all the nucleotide states. - Which is most probable? - We would have to have the actual DNA sequences from the ancestor of primates to find out. - But since we don't have that, we have to make an inference based on whatever modeling rules we select (looking for a model with the fewest number of changes is a common one). - Finally, we must also accept that we might never know, but we can develop some degree of confidence in our conclusion. - So it goes when studying the ancient biological past: much of the direct evidence is lost and so we have to rely on inference and accept the uncertainty. - Let's consider another issue raised by Figure 6.8.3. - In the absence of other evidence, we might also infer that mice are more closely related to humans and gorillas than gibbons are. - After all, they all have a T nucleotide? - This would be based on the idea that a single mutation occurred in a hypothetical common ancestor of mouse, gorilla, and human that diverged from gibbons. - There are a host of reasons why much phenotypic evidence (and probably many other DNA homologous genome sites) would contradict this model of an ape-mouse ancestor, but helps point out that in addition to lots of missing evidence there can be much contradictory evidence when doing inference. - We should be careful, though, not to say "well, that means it's all just guessing and we don't know anything". - We cannot directly observe the past, but unless the rules of nucleotide mutation have changed drastically over the Earth's history, these models make some reasonably confident assertions about the relationships among organisms.

Fossils

- We discussed before that one cannot directly observe the evolutionary past, but also that fossil evidence can offer some really important clues about what extinct organisms looked like, their past geographic distribution, and sometimes clues about how they behaved or even their chemical composition. - A fossil is any remains, impressions, imprints, or trace evidence of something once living in a past geological time period. - Generally, they are divided into body fossils from the actual body of the organism and trace fossils from footprints, burrows, nests, chemical deposits or other indirect evidence or organismal presence or activity. - Fossils provide an opportunity to add information to the internal ancestral nodes of the phylogeny about the organismal phenotypes or habitat information. - This can be really important to more clearly determining when certain changes in organismal physiology first arose and also in helping to calibrate the timescale of phylogenies. - Because fossils occur in particular geological strata (layers of rock that correspond to particular time periods in geological history) we can place the fossil organisms in absolute time by dating the surrounding rock using isotope dating, although the estimated time periods can be quite large (millions to tens-of-millions of years).

Evolution is more than just "change"

- What is "evolution"? - We could say evolution simply means "change over time". - This is sort of true, but incomplete. - More specifically the word comes from Latin evolvere, which has two parts, a prefix ex/e meaning "out of," and a verb volvere meaning "turn/roll/rotate/proceed". - Evolution is a specific kind of change: something unfolding out of itself in cycles through time. - Evolution is a spiral not a circle, since it both proceeds in many cycles but never retreads the same path. - Language, environments, social customs, and just about everything that exists for more than a moment can (and usually does) evolve. - But there is a generative quality of evolution too. - A sense of becoming and of renewal through the turns of a cycle, rather than just a static continuation of something though time. - "Change over time" is nice and useful in many quantitative ways that we might analyze evolution, but "becoming over time" is nearer the truth. - Also, populations of organisms do not have to "change" to evolve, in the sense of big changes from one organismal form to another or one genetic state to another. - Evolution is happening even when no changes are visible, through the renewal of the generations and all the mutation, assortment, and recombination processes we just discussed.

Species and hybridization

- What is a "species?" - There is not a universal definition of this term. - The Biological Species Concept is probably the most widely referenced, which defines species as populations of individuals that are mutually reproductively compatible but incompatible with other related populations. - The BSC uses reproductive isolation as the line between species. - The BSC is not usable with asexual organisms, since they do not have mating to be incompatible or compatible. - For asexual organisms, the Phylogenetic Species Concept is more common, and defines species through genetic relatedness. - Species contain individuals that are highly genetically similar to each other, and highly differentiated genetically from other species. - This can be used for both sexual and asexual organisms. - What about extinct species that we only know through fossils? - We have no idea about their reproductive compatibility or genetic similarity. - In this case, the Morphological Species Concept (or "Morphospecies") is generally used, where species are defined as clusters of individuals with phenotypic similarity within the species and often particular diagnostic traits that distinguish them from near relative species. - What is a hybrid organism? - The terms "species" and "hybrid" are linguistically mutually defined. - A hybrid offspring has parents from two different species. - Now, you may be saying to yourself: "I thought species were reproductively incompatible by definition. So how does hybridization happen?" - In reality, many closely related species can maintain facultative reproductive compatibility, meaning that rarely particular individuals that are particularly genetically compatible and/or given the right environmental circumstances will be reproductively compatible enough to form a rare hybrid offspring. - How rare is hybridization? - Not that rare probably, though usually the hybrids do not form their own distinct fully hybrid population or species. - Mostly they are sterile or mate with an nonhybrid individual from one parental population. - The effect of this is an increase in genetic diversity. - Ultimately, both the terms "species" and "hybrid" rely on a lot of context and convention to be defined. - The reality of reproductive compatibility is more complex in biological reality.

Populations

- What is it that "evolves"? - An individual? - Simply, no. - An individual does change over its life, but that is called "development." - There is definitely a sense of unfolding through temporal cycles in development as well. - But what is it that is turning over when we talk about evolution? - For genetic evolution, what is evolving is the genetic composition of a population. - A population is a group of organisms. - We could call any group of organisms a "population." - But most often when we talk about a population in the context of genetic evolution, we mean a group of organisms that live in a habitat together and reproductively are related to one another. - In a sexually reproducing organism, individuals in the same population are generally reproductive compatible and part of the same mating pool. - In asexual organisms, a population might refer to the individuals derived from a shared clonal ancestor. - Populations could also be defined by geography, shared habitat, morphology, or other factors. - A single individual does not undergo genetic evolution. - How could it? - An individual acquires its particular genetic alleles by the process meiosis and fertilization (or clonal replication for asexual organisms) and the alleles in its (germline) genome are set at fertilization or cloning. - So genetic evolution instead refers to the changes in frequencies of genetic variants of the many individuals in a population over time. - This is the most common form of genetic evolution that we discuss in "population genetics": genetic evolution as measured by changes in allele frequencies in a population over time. - The changes in gene frequencies are the result of the turnover of individuals through the generations, during which assortment, mutation, recombination are all happening. - These processes (and others that we shall discuss further here) cause the frequencies of genetic alleles of a population to fluctuate over time. - We have already seen how random processes drive changes to the chromosomes and the combinations of alleles. - Specifically, mutation can randomly create new alleles, while recombination and assortment can create new combinations of those alleles. - Even operating alone, these forces would be sufficient to create changes in the allele frequencies in a population (population genetic evolution). - However, there's another important random force that causes allele frequencies to change, and these are a set of factors collectively known as genetic drift. - Genetic drift comes from many sources, but collectively refers to chance factors that affect how many offspring an individual contributes to the next generation. - Individuals might randomly not survive to reproduce. - Or they might not mate successfully or less frequently. - Some matings might produce more offspring, and some fewer. - The processes of chromosomal assortment and recombination during meiosis adds further randomness to what alleles and allele combinations are passed on. - This affects both sexual and asexual organisms. - So drift is the evolution of the allele frequencies in a population that just "happens" because of random fluctuations and can have causes endogenous (like chromosomal assortment) or exogenous (like a tornado knocks over a bunch of trees). - This is still genetic evolution of a population because it changes the frequencies of the alleles in a population.

Inheritance

- When you inherit chromosomes, they are replicated copies of the chromosomes, and, due to inherent error in the replication process, they are not precise copies. - So what you inherit is something very much like your parents' chromosomes, but not literally the chromosomes belonging to your parents. - However, we would not call these chromosomes different, either. - They still have approximately the same genes in approximately the same order. - The technical word for this is "homologous"; the chromosomes that you inherit are homologous. - Two chromosomes are homologous if they are related to each other through a chromosome in a common ancestor. - All human X chromosomes are homologous because all human X chromosomes are related through an ancient common ancestor's X chromosome at some point way back in the evolutionary history of humans. - In fact, human, gorilla, and chimpanzee X chromosomes are still homologous, because they are all modified versions from a common ancestral Great Ape X chromosome. - Why is homology important for studying evolution and heredity? - Because we are concerned with the study of things changing through time. - In order to track changes in something (a gene, a trait, a genome) through time, we have to be sure we are talking about the "same thing" through all that genealogical time, so that we can track the changes. - We cannot observe how things change through time unless we are sure that we have been consistently talking about the same entity coming from an ancestor to a descendant and observing changes in that entity through time. - The target of our inquiry must be consistent through time. - Thus, in evolution and heredity, we study "homologous" properties. - A classic example of homology is the homologous bones of forelimbs of several tetrapod vertebrates. - The humerus bone (light grey) in all of these organisms is homologous because the ancestor tetrapod had a humerus too and the fossil record indicates that each humerus in all its descendant species represents evolution from this common ancestral bone. - By contrast, flight in tetrapods from pterodactyls, bats, and birds - However, the common ancestor of these three organisms did not fly! - Flight in these three groups is analogous, which means that they have a trait that is similar in function or form but not as the result of all of these organisms inheriting it from a shared ancestor. - Instead, flight evolved separately in each lineage; this is also sometimes called "convergent evolution". - Nearly anything in biology that can be inherited can also be homologous, not just traits and physiological features like bones. - This includes chromosomes and all the genes on those chromosomes. - This also includes the sequences of individual genes and their protein products. - Even more complex properties like genetic pathways, regulatory systems, and developmental patterns could be inherited from your ancestor and so we can study these homologous properties under evolutionary models too. - We are studying something homologous when we discuss heredity and evolution so that we are confident we are studying "the same entity" in different individuals as it changes through the generations.

Recombinations and Genotype Frequencies

- You can see that we've added frequencies to the alleles. - When everything was just 50%/50% these numbers were not necessary. - But now let's add 1% recombination between A and B. - [Q: Wait a minute. Didn't we say before that crossing over happens about once per chromosome per meiosis in humans? 1% seems way too low. Remember that the single crossing over event could happen anywhere on the chromosome. So 1% is the recombination rate between A and B specifically because the crossing over event must occur on a spot on the chromosome between A and B. So the recombination rate will always be lower than the crossing over rate.] - Now that we've cleared that up, let's add recombined gametes to the table with appropriate frequency for a recombination rate of 1%. - This is the result, highlighting the same zygote genotypes with the same colors again. - Why are the recombinant genotypes only 0.5%? - Because the recombination event occurs 1% of the time, and produces two recombinant chromosomes. - Recall that the chromosomes swap, so a crossing over event produces both of the recombined chromosomes at once. - Once you produce this pair of recombined chromosomes, you will end up with one or the other 50% of the time. - So the odds of recombination itself between A and B is 1% and then you have to flip the independent assortment coin once more to pick one of the two recombined chromosomes. - So each recombined chromosome has a 0.5% (half of 1%) chance of being involved in a fertilization event. - This leaves half of 99% for the parental genotypes (non-recombined), which is 49.5%. - Then the frequency of the possible zygotes is just the product of their parental frequencies (multiply the top of the row and column. - You can see the most common outcomes will be the combinations of parental genotypes, a fertilization of two non-recombined allele combinations. - These events are shown in the upper left quadrant of the table. - Collectively, these genotypes make up a combined 98.01% chance for offspring to have inherited just the non-recombined A and B combinations. - The scenarios with a fertilization of one recombined and one non-recombined set of A and B alleles are shown in the upper right and lower left quadrants. - This means either the maternal or paternal meiosis event had a recombination between A and B but not both parents. - Finally, the lower right quadrant is where both parents independently have a recombination event. - How often does that happen? - 0.01% of the time. - 0.01% is 1% times 1%. - Meaning we have to have two independent events happen with 1% frequency, so their coincidence is 1% times 1% or 0.01%. - This would be very rare. - It also happens that having two recombinations produces offspring genotypes identical to the parental double-heterozygous genotypes (green shaded). - The configuration of the alleles on the chromosomes is different, but the total allele content is the same, like this: -----A1------B1----- -----A1--\---B2----- -----A2------B2----- -----A2--\---B1----- - Two recombinations end up back at the same place (at least for A and B). - This is only the result for a recombination rate of 1%. - When alleles are on different chromosomes, then they assort independently. - You have a 50%/50% chance of inheriting two alleles on different chromosomes together in the same gamete. - When alleles are on the same chromosome, then they are linked together. - If the recombination rate is really low (r is approximately zero), then you have effectively a 99.999..% percent chance of inheriting the alleles together. - If the recombination rate is really high (maybe 20%), then you have a relatively strong chance of inheriting some combination of the two alleles.