Population Genetics
Effects of a Bottleneck
-A bottleneck can be caused by drift or selection, and can make the population more or less fit -If it makes it more fit, it will do so by removing variation, which means that the organism will be less able to adapt if the environment changes, so the bottleneck may be good short-term but devastating longterm -When it is mainly caused by drift, we should see all genes moving together, but when selection causes it we will see certain alleles changing more than others
Heterozygote Disadvantage : Uverdominance
-Both s & t are positive because it is better to be a homozygote -There is an equilibrium point in the middle but you will more away from it, climbing up the side that you are already on -No great examples of this one because when it happens things diverge so quickly that it is tough to catch
Single Locus Selection Example: Biston betularia
-During the industrial revolution in Britain frequencies of melanic and peppered moths changed depending on location
Directional Selection
-Examples: Human brain size
Ne
-Genetic, effective size of a population = the number of individuals in a population where every individual reproduces and the rate of genetic drift is the same as in the actual population -The more time you spend at a lower population level = the more variation you loose (why we try to increase size quickly in conservation efforts) -Best estimated with the harmonic mean
Population Size and Allele Changes
-In small populations, selection will be ineffective, and drift will dominate, so good mutations can be lost and harmful ones can move in -In large populations, selection can act, and drift will be weaker
Genetic Effective Population Size
-Influenced by the breeding sex ratio -Means that we see an additional effect on the loss of variation in sexually selecting populations, because not all males mate
Disruptive Selection
-Leads to speciation *if* it is a non-randomly mating population -Examples: Black-bellied seedcracker
Causes of Linkage Disequilibrium
-Non-random mating -New mutation arises that is linked to other genes -New population is formed when two populations with different linkage disequillibriums merge -No recombination (Y chromosome) -Selection favors linkage disequilibrium Counteracted by recombination
Dominant Mutant Allele Equilibrium Frequency
-Qhat = the eq. frequency of the bad allele -As mutation rate (u) goes up, the eq. frequency of the bad allele will also -We would also expect to see lots of slightly deleterious alleles (small s) because that will make Qhat larger
Partially Recessive Mutant Allele Equilibrium Frequency
-Qhat = the eq. frequency of the bad allele -As mutation rate (u) goes up, the eq. frequency of the bad allele will also -We would also expect to see lots of slightly deleterious alleles (small s) because that will make Qhat larger
Recessive Mutant Allele Equilibrium Frequency
-Qhat = the eq. frequency of the bad allele -As mutation rate (u) goes up, the eq. frequency of the bad allele will also -We would also expect to see lots of slightly deleterious alleles (small s) because that will make Qhat larger
Hardy-Castle-Weinberg Assumptions
-Random mating -Infinite population -No selection -No mutation -No genetic drift -No migration
Stabilizing Selection
-Really just two forces of directional selection -Examples: Human Birthweight, Gall Fly
Why Evolution Occurs Fast
-Single allele (less loci that need to be mutated) -Dominant alleles -A lot of standing variation for selection to act on
Heterozygote Advantage : Overdominance : Heterosis
-There is no clear good or bad allele, and having both results in the highest fitness -This graph would be a bell curve but the maximum of the peak (frequency w/ the highest fitness) would not necessarily be in the middle -This is the only selection model that maintains genetic variation and polymorphism
An allele's ability to spread depends on its dominance
-This graph explains why selection cannot help us get rid of recessive deleterious alleles, because they are sheltered in heterozygotes (Unless they express some degree of dominance) -This means that a recessive gene with favorable fitness can become fixed, *but* a dominant gene with favorable fitness cannot -Also note that evolution occurs fastest at .5 allele frequencies, the point where heterozygosity is maxed out and there is the most variability
Four Effects of Selection on Population Mean Fitness
1) Dominant beneficial 2) Recessive beneficial 3) Overdominance 4) Underdominance
Ways to Reduce the Impact of Small Captive Populations
1) Genetic Augmentation = trading animals with other zoos for breeding to *avoid inbreeding* 2) Pedigree analysis = Tracking the reproductive success of each individual to make sure that all individuals are breeding and *maximize the effective population size*
Genetic Consequences of Small Population Size
1) Loss of additive genetic variance and heterozygosity (smaller populations will loose heterozygosity faster, and not having enough Va to adapt to a changing environment puts animals at risk of extinction) 2) Random genetic drift 3) Inbreeding and the reduction in fitness that follows due to deleterious, recessive alleles being expressed in homozygotes (can be seen in zoo animals that have a small effective population size) All of which leads to *long-term accumulation of deleterious mutations and eventual extinction due to "mutational meltdown"*
Significant Notes on Heretability
1) Morphological traits tend to have *relatively high heritabilities*, giving selection a lot of room to work 2) Highly artificially selected for traits of domesticated species have *very low heritability* because we have exhausted all the available genetic variation
How Selection Maintains Polymorphism
1) Overdominance 2) Varying selection spatially or temporally on different sub-populations 3) Frequency-dependent selection means that the fitness of a genotype depends on its frequency in the population, in other words, being rare helps (Ex: the Elderflower Orchids and the scale eating Cichlids)
2 ways to Quantify Genetic Variation
1) P = the proportion of polymorphic loci 2) H = the average heterozygosity (the proportion of loci where a random individual is heterozygous) Note: In H-W Eq both will be equal
Other Forms of Frequency Dependent Selection
1) Predator-prey interactions gives rare prey an example 2) Parasite-host interactions mean that parasites go for the most common host 3) Mimicry systems where being a successful mimic depends of the frequency of what you are mimicking
Mendel's Laws
1) Segregation = genes are on different chromosomes 2) Independent Assortment = they are randomly passed on to offspring
Criticisms of the Shifting Balance Theory
1) Very *difficult to test or falsify with empirical data* 2) While phases 1 & 2 are plausible, phase 3 seems improbable, because *migration would break up "co-adapted gene complexes,"* *higher fitness is not correlated to increased migration,* and *the specific situations needed for this to occur are highly unlikely*
Population Genetics definition of Evolution
A change in gene frequencies through time
Population
A freely interbreeding group of individuals
Fst
A measure to quantify population subdivision (from 0 - no division to 1 - complete division)
The Modern Synthesis
A revolution in the 1930s and 40s which re-united genetics and natural selection with the following points: -Populations have genetic variation that arises via random mutation (+ variation) -Populations evolve by changes in gene frequency -Gene frequencies can change through genetic drift (- variation), gene flow/migration (+/-), and selection (-) -Most adaptive variants have only small effects on the phenotype so changes are typically gradual -Diversification comes about through speciation
Population Genetics
A set of rules for how variation is maintained in a population (and the observation of what happens to Mendelian genes over time)
Hardy-Castle-Weinberg Law
A single generation of random mating under specific assumptions establishes equilibrium genotypic frequencies that will not change over time as long as all the conditions are stable. We use H-C-W Eq as a null hypothesis to test which forces are acting on a population.
Wright's 3 Models of Population Sub-division
A) Complete random mating in a population (panmitic) B) Complete isolation without migration C) Somewhere in the middle
W
Absolute Fitness = Total # of offspring produced
Wright's Island Model
Allows us to compare migration among populations
Genetic Drift
Alteration of allele frequencies due to chance, which will always happen in a finite population, and will be *stronger in smaller populations* *It results in a net loss of alleles over time which usually makes organisms less fit
Price's Rule
An alternative way to estimate the selection differential, using the covariance between phenotypic variance and fitness variance
Wright's Shifting Balance Theory
An enduring way to *explain how populations get from a lower fitness peak to a higher one* if there is a valley in the middle using 3 phases: 1) Selection cannot move us down, but drift can, so *genetic wobble caused by drift* causes a population to fall off a peak 2) Then *selection can drive you up an adjacent peak* (if you land closer to it) 3) Once on a new peak, the original population can send out migrants carrying the genes for higher fitness and bring the other population up through gene flow (but *this phase does not always happen and is highly contested*)
Population Genetics definition of Selection
Any consistent difference in fitness among phenotypically different biological entities
Loss of Heterozygosity in a Randomly Mating Population of Adults
As time goes on, if selection and migration do not balance out drift, we will love a lot of variation
H
Dominance Coefficient = Proportion of S applied to the heterozygous genotype: WAa = 1-hs *If A is completely dominant h=0 and vice-versa
Selection Differential (S)
Equal to *the difference in means of selected individuals and the base population*
Continent-Island Model of Migration
Eventually the island will look just like the continent
Recent Extinction Rates
Extinction rates were rising through the 1900s, but then they dropped when rounding the corner into 2000 This is not because we have put less pressure on organisms, but rather because *we have put all the vulnerable ones extinct* and only the tougher ones are left now
Fisher and Sex Ratio Evolution
Fisher argued that sex ratios should balance out to 50:50 because the rarer gender will have an advantage *This is tough to prove and non unanimously supported
Genetic Drift of Deleterious Alleles
For big mutations, s will be large and selection will win the game, but when when s is not as large, drift will win, and deleterious alleles can be advanced and even fixed in a population
Genome-wide Association Mapping
GWA is like QTL analysis over a whole genome, so it can only be done with species that we have tons of data about, including genomes and phenotypes
When Inbreeding is Good
If a species can survive via inbreeding for long enough to purge all the deleterious, recessive alleles from their population (because they were expressed as homozygotes which selection could act on) they will come out on the other side without any harmful variation (but mutation will likely continue re-introducing it, so this strategy is not perfect)
Estimating Heritability from the Breeder's Equation
If we can collect data for the selection differential and the difference in mean phenotype for offspring and parents, we can estimate how heritable a trait is
Formula for the Covariance
Looks at how much variance is shared (+,0,-)
Inbreeding
Mating between close relatives that leads to deviations from H-W Equilibrium by *causing a deficit of heterozygotes* *Note that in this case, allele frequency stays the same, but genotype frequency will change (whereas drift will actually change the allele frequency) *This also means that we can use inbreeding (like we do in drosophila experiments) to make pure homozygotes after a relatively small number of generations
Going from Covariance to Heritability
Midparent - Offspring is the most reliable estimation, but they are all relatively inaccurate *You cannot use monozygotic twins or full siblings for this because you have other forms of variability in their graphs*
Gene Flow
Movement of alleles between populations which homogenizes allele frequencies over time (and counteracts the effects of drift)
Migration
Movement of individuals from one subpopulations to another one followed by random mating and the exchange of gametes and fertilization *If migration were the only force at work, populations would eventually converge their allele frequencies
Source of Normal Distribution of Phenotypes
Occurs when you have many loci all interacting to determine a character, because the H-W-C Eq causes there to be a few individuals with the rare combinations and a lot in the middle with many heterozygous ones (there are also environmental effects for each trait, and the average of these curves and the bars from the figure is what produces the smooth normal curve for quantitative traits)
Inbreeding Depression
Often as F increases, fitness decreases, assuming that a population is sheltering some deleterious, recessive alleles (which almost all population are) *Finding closely related parents and a fitness cost is the signature of inbreeding depression
W bar
Population Mean Fitness = (p^2)(WAA) + 2p(1-p)WAa + (1-p)^2Waa *This is the fitness value for a phenotype
Gregor Mendel
Published the first work on heritability in 1866 (wasn't re-discovered until 1902)
Quantitative Trait Loci Analysis
QTLs are stretches of DNA that are correlated with variation in a phenotypic trait (hinting that these regions contain, or are linked to, genes that contribute to population differences in a phenotype) and analysis of them identifies regions of interest (which are then further confirmed with an Fst scan and investigated to try to learn more about the genetic component of variation in quantitative traits) For example: Pea Aphids, Sticklebacks, Monkey Flowers
Linkage
Refers to *how close two genes are*
Role of Drift in Population Divergence
Rule of thumb
Fisher's Fundamental Theorem of Natural Selection
Says that if there is any variation in fitness, selection will act on it, and *fitness will evolve at the same rate as there is additive genetic variation for it* (which will eventually erode genetic variance in fitness, relying on the variation by mutation, Vm, to add it back in) We can apply this to Price's Rule to show that the *directional selection differential*, S, on fitness itself is *equal to the phenotypic variance in relative fitness* and the *response to selection*, R, is equal to or less than the *additive genetic variation in fitness*
S
Selection Coefficient = Fitness disadvantage to homozygous genotype: WAA = 1-s
T
Selection Coefficient = Fitness disadvantage to homozygous genotype: Waa = 1-t At Equilibrium, Peq = t/(s+t)
The Extinction Vortex
Smaller populations (usually caused in the first place by demographic reasons like habitat destruction and fragmentation) will cause (1) more effective genetic drift and (2) a greater chance of inbreeding, which will create a *positive feedback loop leading to extinction*
Loosing Heterozygosity
So both drift and inbreeding remove heterozygosity, but drift will happen no matter what as long as the population is finite, and it will actually change allele frequencies, whereas inbreeding only happens when there is non-random mating and it only changes genotype frequencies
Rate of Polygenic Mutation
Testing for this has shown us that *mutation by itself cannot be the whole explanation for the variation that is maintained*, but if given time, mutation will build up variation in populations once again
Epistasis
The *interactions between alleles* that affect characters differently
Linkage Disequillibrium
The *non-random association of alleles*, because genes that are closer together have a lower chance of being separated during recombination This is rare in nature because unless selection acts to maintain it, it will eventually decay
Heritability (h2)
The *proportion of variance (from 0 to 1) that is due to genetic effects that will be passed on*, and therefore available for selection to act on Since it is a function of Ve (in the denominator), this is a *context-dependent measure* which means that it can vary depending on where and how you measure it (environment they are raised and measured in) *Not the same as inheritance*
Nm
The absolute number of migrant organisms that enter each subpopulation each generation *The larger this is, the smaller Fst will be
Critical Rate of Evolution
The amount of genetic variation for critical phenotypes a population must possess in order to be able to adapt to a changing environment
Gene Pool
The collection of all alleles in a population at a given time
Phenotype
The expressed characteristic (physiological, behavioral, biochemical, or morphological) of an individual organism
Quantitative Genetics
The field of developing predictive theories for evolutionary change, which effect plant & animal science, evolutionary biology, human disease research, and conservation biology, and was started by *Fisher* who *noticed that we needed to see these traits as normally distributed* *Different than population genetics* because it looks at the phenotypic variation and then moves towards the allele, as opposed to starting with the gene and working up
Genotype Frequency
The frequency of a genotype (p^2, 2pq, or q^2) in a population at a given time
Allele Frequency
The frequency of an allele (p or q) in a population at a given time
Genotype
The genetic date corresponding to that expressed characteristic (phenotype)
F
The inbreeding coefficient = the probability that an individual taken at random will be autozygous (two exactly identical alleles by descent)
Additive Effects
The more you have of an allele the more it effects you *w/o any conditions*
Univariate Breeder's Equation
The simplest equation for predicting *evolutionary change of phenotype from one generation to the next*, R, the response to selection (which will be *high when heritability is high*)
Mutational Heritability (Vm)
Usually determined by rearing clones
Adaptive Landscapes
Visualizations of the *many possible epistatic interactions between alleles* (which are simple with Mendelian traits but for too difficult to produce for many complex traits) with peaks and valleys Fitness increases up until it peaks out at a ridge, which is where selection will cause populations to rise to They can get stuck in between two higher peaks, however, because selection cannot push them down, so *where you start is very important* Also, know that the further you are from a peak, the stronger selection is to drive you back up one
Components of Phenotypic Variation
Vp = (Va+Vd+Vi) + Ve *Vg* = Va+Vd+Vi Va = Variation due to the additive effects of alleles (this is what selection can act on and is responsible for the resemblance between parents and offspring) Vd = Variance due to dominance relationships of alleles (may not always be inherited) Vi = Variance due to epistatic interactions between alleles *Ve* = The variance among phenotypes expressed by replicate members of the same genotype
Genetic Covariance between Relatives
We can use the phenotypic covariance between the two to calculate the heredity for that trait Know that *Midparent - Offspring is the most reliable*, because the *all the phenotypic variance is accredible to the additive genetic variance*
Coalescence Theory
We should be able to trace the descendants of a gene just like a haploid organism, so if we look back, all the genes shared a common anscestor
DNA Polymorphisms in Forensics
What it is: -Allows us to look at an individuals genotype at specific micro satellite loci with extremely high variation to identify if DNA from the crime scene is theirs Problems: -You have a higher probability of matching depending on how closely related you are the the criminal (including race) -There can be DNA contamination leading to false results -Lots of measurements (like the gels) where error can enter
Pleiotropy
When *loci affect two or more characters*
Polygenic
When a *character is affected by multiple loci*
Resemblance between Relatives
When there is genetic variance for a trait, you will expect resemblance, but just how much depends on how related the individuals are
Using Fst to Predict Selection
When we look at Fst for multiple genetic loci, we get some outliers, that have more subdivision that others and that have diverged faster than predicted by drift, which means they may be under selection
Antagonistic Pleiotropy and the Red Queen Theory
When you have traits that show negative correlation, selecting up on one will bring the other down, which constrains and maintains genetic variation (*The Antagonistic Pleiotropy Hypothesis*) In other words, selection can only go so far before it limits itself (because the traits it is selecting for have other epistatic interactions that will also fall victim to selection in the other direction) *The Red Queen Theory* applied here says that (as long as the environment continues changing and there are co-evolutionary relationships with the organisms at hand and other organisms) the fitness optimum will keep moving, so the population will perpetually be moving to reach it
Frequency of a Deleterious Mutant Allele at Equillibrium
deltaQ = loss due to selection + input by mutation = 0 *The Drosophila population bombarder with x-rays showed this to be true
Hardy-Castle-Weinberg Equations
p^2 + 2pq + q^2 = 1 p + q = 1