Lec 6 Flashcards
What happens to allele frequency change when we have finite (instead of infinite) population size?
When we talk about evolution, we often talk about natural selection in the same breath, or use the terms interchangeable - as if natural selection is the only mechanism of evolutionary change over time
Natural selection is just ONE mechanism
The TWO most important processes that cause allele frequencies to change through time (evolution) are:
Natural selection and genetic drift
Genetic Drift
The RANDOM change of allele frequencies in a population of FINITE size
Beetles squashed randomly for no genetic reason
If someone happened to step on green beetles more than brown beetles, the amount of brown beetles will increase
Differential survival of individuals is due to extrinsic factors, NOT to genetics or natural selection
Mendel’s law of segregation tells us that alleles are transmitted RANDOMLY to gametes
Random transmittal of alleles to offspring by heterozygotes in small populations means that the frequency of transmission of each allele may not be equal in each generation
Genetic drift happens because Mendelian inheritance is about ____________
Averages
What percentage of offpring are ___________?
Is it possible that two individuals with __ genotype produce four ________________ offspring? (Concept important, details not)
We can think about the effects of genetic drift by starting with the analogy of flipping a coin. Imagine you flip a normal coin. What’s the probability of getting heads?
50%
Flip it again. What’s the probability of getting heads?
50%
Flip again. What’s the probability of getting heads?
50%
Even though the chance of getting heads is 50%, if you got 3 heads in a row would you think it was weird?
No
How about if you flipped the coin 3000 times and got 3000 heads. Would that be weird?
Yes
We can think about the effects of genetic drift by starting with the analogy of flipping a coin
At large sample sizes, OBSERVED frequencies are similar to EXPECTED frequencies - but this is often NOT the case at small sample sizes
At small sample sizes, we often see deviations between expectations and observations
Wright-Fischer Model
Most commonly used model to describe genetic drift
Important for making predictions
Assumes a HAPLOID population
NO sexes
Discrete generations (every individual replaced each generation)
Constant size of 2N
-Simulate behavior of diploid population even though we are assuming haploid
No mating required for reproduction
No other evolutionary processes at work (no selection, mutation, migration)
-Baseline model, there WILL be other processes at work
Use this model to predict HOW allele frequencies will change under genetic drift
Discuss reasons why a population wouldn’t conform to this model
Have allele frequencies in our population changed in generation 2?
Yes
Was natural selection operating?
NO
We RANDOMLY picked alleles, there was no reasoning
Genetic drift is the ______ change in allele frequencies from one generation to the next, in the ________ of natural selection
RANDOM; ABSENCE
We want to come up with some sort of _______ about allele frequency changes over time due to drift
Prediction
Calculate EXPECTED allele frequencies in generation 2 given allele frequencies present in generation 1
Assume random sampling
Whether or not a particular allele “survives” to generation 2 DEPENDS ONLY ON ITS FREQUENCY IN GENERATION 1
Alleles at higher frequencies are more likely to be “chosen” (survive) than alleles at low frequencies
EXPECTED changes in allele frequencies
Expectations are bout averages
Roll a 6-sided die 6 times, you will on average roll a 2 1/6 rolls
However, this number will vary across trials - sometimes it will be above average, sometimes below average
If we start over with a new population with the exact same allele frequencies and do our random draws again, will we get the same result?
NO, we do NOT get the same outcome
Many small changes from generation to generation = big, somewhat random changes in allele frequencies over time
Consequences of genetic drift for allele frequency change
Over time, ALL alleles will eventually become fixed or lost
Chance of fixation = chance that allele persists to next generation = frequency of the allele in current generation
Drift is blind to allele identify or dominance - all that matters is frequency
If the starting frequency of the A allele in a population is 0.6, what is the probability of that allele going to fixation?
60%
If an allele is at a frequency of 10% in a population, it has a _______ chance of being FIXED and a _____ chance of being LOST
10%; 90%
Over time, genetic drift will purge ______ from populations
Variation
How fast can allele frequencies change under genetic drift?
Consider a population at HWE with alleles A and a
Both alleles are at 50% in the population (equal frequencies). How fast will allele frequency change?
How fast allele frequencies change depends on starting population size
Will alleles fix faster due to drift in large or small populations?
Small
Allele frequencies change due to drift MUCH MORE QUICKLy in small populations
Remember how averages work
At large sample sizes, observed frequencies are similar to expected frequencies - but this is often not the case at small sample sizes
If we have 10 individuals and are randomly drawing A and a alleles for the next generation, can we pretty easily draw 9A and 1a alleles?
-We can! There is then a high chance that we draw 10As and 0as in the next generation, leading to the loss of a and the fixation of A
With a population of 1000 individuals, if we are randomly drawing A and a alleles for the next generation, are we likely to get 999A and 1a>
-NO - we expect we will get close to 500As and 500as
So - drift is a much more powerful evolutionary force in SMALL populations
Based on these figures showing allele frequency change over time, which population would you hypothesize is the smallest?
Population A
Heterozygosity
The presence of different alleles at one or more loci on homologous chromosomes
This is our key measure of the genetic variation in a population
-There can be NO adaptation/natural selection without variation
Observed heterozygosity
H0
The fraction of individuals that are heterozygous at a given locus in our population
-What we actually observe
Expected heterozygosity
He
The fraction of heterozygotes EXPECTED under Hardy-Weinberg equilibrium
-What we EXPECT to see
Departures of observed heterozygosity from expectations suggest other processes are removing genetic variation
Often find FEWER heterozygotes than expected
What effect does genetic drift have on heterozygosity (genetic variation) in populations?
Decrease heterozygosity
What effect does genetic drift have on genetic variation in populations?
Genetic drift removes variation from populations - over time all alleles will either be fixed or lost
Loss of variation by drift is random - which allele is fixed or lost only depends on its frequency in the population
Genetic drift is always operating in the “background” of populations, regardless of what is going on with selection/migration/assortative mating
Thus all populations are constantly losing genetic variation over time due to drift
Drift, variation, and population size
Small populations experience stronger genetic drift
Small populations therefore lose genetic variation (heterozygosity) more rapidly than large populations
Why is genetic drift stronger in small populations?
The chances of a large change in allele frequency due to drift are greater when population size is small
Genetic bottlenecks
Population experiences a REDUCTION in size
-Much LESS variation after being restricted
Bottlenecks lead to a loss of genetic variation
Bottlenecks and allele frequency change
This figure shows 10 simulated populations
We start at the same allele frequency (A = a = 0.5)
Populations experience a bottleneck in size during the period indicated by the shaded region and return to the original size of 1000 individuals afterward
Allele frequencies fluctuate much more during the bottleneck than before or after
When population experiences a reduction in size, we observe LARGE and UNPREDICTABLE changes in allele frequencies among populations
The bottleneck causes divergence between populations. Before the bottleneck, allele frequencies are similar in all populations. After the bottleneck, allele frequencies differ greatly from one population to the next
Look at the highlighted section of this figure. What happens to variation in allele frequencies across generations?
Allele frequency variation increases in magnitude
Bottlenecks are _________ in nature
COMMON
Island Diversity
An order of magnitude lower than mainland diversity
Why did foxes undergo a bottleneck?
- Populations of golden eagles migrated to Channel islands
- Bald eagles LEFT (previously kept golden eagles away)
- Golden eagles eat foxes
Nucleotide diversity
Average proportion of nucleotide differences between sequences
If there are differences between observed and expected heterozygosity at a locus, that likely means:
The conditions of Hardy-Weinberg Equilibrium are not met in the population
Founder effects also reduce genetic diversity
Lost some alleles by colonizing new population
Leads to a decrease in genetic diversity
New populations are colonized by a small number of individuals
Serial founder effects in humans
Humans evolved in Africa
This migration is known as the Great Human Expansion
Founder effects can result in DECREASED heterozygosity/variation
The vast majority of human genetic variation originated in Africa
In which scenario would you expect to find the lowest heterozygosity?
An island population that has recently recovered from intensive hunting
Genetic drift can be deceptive…
Snappers are thought to be a relatively healthy fishery
However, large effects of genetic drift were observed in snapper populations despite large population size - over 3 million individuals
Why might this be?
In a snapper, only a few large individuals in each generation produce most of the offspring - this is common in pelagic fish
Without understanding the biology and geography of a system, we may mistakenly predict the effects of genetic drift
Why might we see low heterozygostiy in a population with millions of individuals?
a) This population was started by a founder event
b) This population has recovered from a bottleneck
c) All of the above
c) All of the above
Genetic drift leads to loss of heterozygosity during population bottlenecks
We see a larger effect of genetic drift in populations that have experienced bottlenecks or founder effects in the past
Even though population size is large, genetic variation is low due to past bottlenecks
Effective population size
Langurs live on steep cliffs, used to be joined by jungle; now separated by rice paddies
Hardy-Weinberg equilibrium assumes random mating
Can all the adult Delacour’s langurs mate with each other?
The total population surviving in the above region is estimated to be between 200 and 250 individuals, surviving in 19 isolated subpopulations; the species is believed to be extirpated form 3 additional sites, and some important populations, including those in Cuc Phuong National Park and Pu Luong Nature Reserve, have decreased by 20% in the last 5 years
Say we measured heterozygosity in the langurs and compared it to expectations under a Wright-Fisher model with population size = 250. Would we expect H0 to be higher or lower than He?
Lower
Effective population size (Ne)
Say we measured heterozygosity in the langurs and compared it to expectations under a W-F model with N = 250. Would we expect H0 to be higher or lower than He?
-Probably pretty low! Because the populations are isolated, drift is playing a larger role than would be expected for a population of 250 individuals. We might instead see the amount of drift expected from a population of 10 individuals
Ne is the number of individuals in a Wright-Fisher model that would produce the amount of genetic drift observed in the real population
Other factors can affect Ne
1) Change in actual population size
2) UNEQUAL SEX RATIOS: All offspring need to inherit one allele from each parent. If there are many males and few females, the population will ACT smaller because the probability of drawing alleles from the same mother is large. A population with 100 males and 10 females will act like a population of N = 36, not N = 110 (has the same heterozygosity as a population of 36)
3) Variance in offspring number: If some individuals have 10,000 offspring and others have none, the population will ACT like a small population even if there are large numbers of individuals
Wright-Fisher Expectation
Unequal offspring
-Effect on heterozygosity will be similar to a much smaller population
Mating disproportionately with one individual instead of equal mating between all individuals
Variation, drift, and risk of extinction
Reduced heterozygosity is often associated with reduced ability to adapt - small isolated populations are more at risk and often considered “threatened”
The smaller the populations, the less the genetic diversity, the less heterozygosity, the less allelic and phenotypic polymorphism
Small populations are more vulnerable to chance events
Extinction vortex
Populations are reduced and fragmented by habitat loss and hunting
Heterozygosity declines in small populations due to genetic drift - alleles become fixed or lost
Small population size increases inbreeding, which leads to inbreeding depression (recessive alleles exposed to selection)
Genetic diversity is irreversibly lost
-Loss is much more rapid than introduction of new variation by breeding
Populations are more vulnerable to chance events
Populations go extinct
From a conservation perspective, intervening only when populations are already small means its often too late
Small population -> inbreeding or random genetic drift -> loss of genetic variability -> Reduction in individual fitness and population adaptability -> lower reproduction and high mortality -> Smaller population
When population sizes get so small that they resort to inbreeding, they are usually beyond saving
Drift can also cause population divergence and (perhaps) the formation of new species
Imagine we have a group of islands that can each sustain 10 individuals
No migration, selection, or mutation - just drift operating
We seed each island with 10 Aa heterozygotes
We then look at how alleles are distributed across islands over time
Eventually, each island will fix for a different allele
Divergence = different allele frequencies among populations
If allele frequencies are different enough across many loci, we end up with new species
Isolation and drift: The Galapagos lava lizard
Lava lizards are poor dispersers
They maintain small populations on dry rocky areas in Santa Cruz and surrounding islets
Drift in Lava Lizards
During the Late Pleistocene glacial event sea levels were substantially lower, allowing lava lizards to disperse between islands
Sea level rose, cutting lizards on different islands off from each other
As predicted by theory, there are now allele frequency differences at the same loci in lizards on different islands
Lizards on smaller islands also had less genetic variation, again as predicted
Effective population size is reduced when populations are fragmented and can lead to an extinction vortex because:
Individuals are not able to mate with each other due to isolation, making the potentially breeding population smaller than total population size
Interactions between selection, drift, and mutation
If drift alone was at work, all populations would become homozygous at all loci over time
In reality, mutation introduces new mutations
Selection, drift, and mutation are always operating simultaneously in populations
How important are selection vs. drift to allele frequency change (and thus evolution)?
The effectiveness of natural selection at fixing alleles depends on population size and strength of selection
At small population sizes, drift is a stronger evolutionary force
At larger population sizes, selection is stronger than genetic drift
Figure shows probability that a novel mutation fixes in the population at different selection coefficients
Note that here the selection coefficients give selection against the other allele, so s = 0.5 means the new mutation is 50% better
We see that at small population sizes, the only way for a new mutation to fix is if it is very, very strongly selected
When population size is larger, the chance of fixation increases even if selection is weaker
Allele is less likely to be lost when population size is large; if population size is small, alleles with low frequencies are likely to be LOST unless there is very strong selection for it
At low frequencies much MORE likely to be LOST
