Chapter 18: Genetic Variation in Populations

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome back to the Deep Dive.

Today we're tackling, well, a really fundamental question in evolution.

What actually is the basic unit that evolves?

Yeah, it's a big one.

We're going to dig into Strickberger's evolution, looking at where genetic change comes from, how it works in populations.

And really clarify that core concept, that apparent paradox.

It's definitely a distinction that trips people up sometimes.

If you have to get this clear, natural selection.

That process acts hard on individuals.

Right.

It's about my survival, your ability to reproduce here and now in this generation.

Exactly.

Who survives, who reproduces more, that's individual fitness.

But, and here's the twist, individuals don't actually evolve during their lifespan.

I mean, I'm not evolving right now.

No, you're not.

Evolution is that heritable change, the response you see across generations.

So the population is the entity that truly evolves over time.

Okay, so the individual is kind of the test case, the vessel.

Precisely.

Selection pressure hits the individual, but the lasting change is measured in the population.

So if the population is our unit, how do we measure its potential for change?

What's the actual stuff evolution works on?

The raw material is genetic variation.

Simple as that.

Without different versions of genes,

selection has nothing to choose between.

No variation, no evolution.

Right.

And to measure it, we talk about the gene pool.

That's all the alleles.

Every version of every gene in all the individuals within that population.

But we need something more specific to track change, right?

Not just the whole pool.

Yeah.

The most basic unit we track in population genetics is allele frequency.

How common is this version of a gene versus that version?

That's the bottom line number.

Is the frequency of allele A changing?

If yes, evolution might be happening.

If no, maybe things are stable.

Exactly.

That frequency is our core metric.

Okay, so where does this essential variation come from?

Let's start with the classic source.

Mutation.

Right, mutation.

Often taught as completely random, but it's a bit more nuanced than that.

Yeah, that's important.

Random in terms of benefit, a mutation doesn't arise because it's helpful but not random in where it happens in the genome.

Exactly.

Some parts of the genome are just, well,

they're more fragile or prone to errors.

We call them hot spots.

Hot spots.

They can mutate maybe 10 times, even 100 times more often than other regions.

So the genome's own structure kind of guides where new variation is more likely to pop up.

But even with hot spots, the overall rate for any specific gene is still super low, isn't it?

Oh, incredibly low.

You might be looking at one mutation per, say, 100 ,000 copies of a gene passed on.

Maybe even lower.

So relying purely on new mutations to drive significant change would take, well, ages, geological time, basically.

It really would.

And theoretically, you could reach a point called mutational equilibrium.

Where the rate of forward mutation, say A turning into A, perfectly balances the rate of back mutations, A turning back into A.

Precisely.

The math works out so that the net change in allele frequency, delta Q, would be zero.

Equilibrium.

But you're saying this is mostly theoretical.

For the most part, yeah.

Because mutation is so slow, other forces like natural selection almost always kick in long before this kind of equilibrium is ever reached in the real world.

It's just not fast enough to be the main driver, usually.

So if new mutations are too slow, what is the main source of variation that selection acts on quickly?

It's the variation that's already there.

Genetic polymorphism.

Polymorphism, meaning just having multiple forms, right?

Multiple alleles already present in the gene pool.

Exactly.

A population showing two or more genetically distinct forms for a trait.

And the amount of this stored variation is actually staggering.

How much are we talking?

Well, studies show that for many species, something like two thirds to three quarters of all their protein -coding genes, their loci, are polymorphic.

They have multiple alleles floating around.

Wow.

So there's this huge existing library of options just waiting.

A massive reservoir, yeah.

Ready for selection to act upon when the environment changes.

And we see this happen fast.

Like the DDT resistance examples?

Perfect example.

Think about houseflies after World War II, when DDT was introduced.

The amount needed to kill them shot up incredibly fast between 1945 and 1951.

That wasn't waiting for a lucky new mutation.

No way.

Selection was rapidly favoring flies that already had resistance alleles, likely scattered across several chromosomes, that were already present in the population at low frequency.

It's sorted through the existing deck, basically.

Exactly.

Or the classic textbook case,

the peppered moth, piston -bitule area.

Right, the shift to darker forms in industrial areas.

The dark, melanic alleles were already there, but rare.

Then industrial pollution changed the environment, made dark bark common, and suddenly selection heavily favored the dark moths.

The allele frequency shifted dramatically.

Evolution by sorting, not just waiting for novelty.

Okay, that makes sense.

Existing variation is key for rapid response, but you mentioned something molecular about hidden variation.

Ah, yes.

This is fascinating.

It involves a molecule called heat shock protein 90, or HSP90.

It's a molecular chaperone.

Chaperone, meaning it helps other proteins.

Yeah, it helps them fold into their correct shapes and stay stable.

It's like a quality control manager for proteins.

Okay, so how does that relate to hiding variation?

Well, HSP90 is so good at its job, it can sometimes compensate for slight defects in other proteins caused by mutations.

It basically papers over the cracks.

So a mutation might be there in the DNA.

But its effect on the organism's phenotype, its observable traits, is masked or buffered by HSP90 working correctly.

The variation is genetically present, but phenotypically hidden.

Like potential energy stored away.

Exactly.

It's accumulating silently.

Then imagine the organism faces a big environmental shock.

Sudden heat wave, maybe another stressor.

What happens to HSP90 then?

It can get overwhelmed or its function might be compromised by the stress.

And when HSP90 stops buffering effectively, poof.

All that hidden variation is suddenly revealed.

Yes.

A whole suite of previously masked phenotypes can appear all at once, exposed to natural selection.

It can suddenly open up completely new avenues for adaptation, almost instantaneously.

Whoa.

So HSP90 isn't just about surviving heat stress, it's potentially a mechanism for storing up evolutionary potential and releasing it under pressure.

It acts like a capacitor for evolutionary change.

A huge potential accelerator for adaptation in stressful times.

Okay, so we've got single gene effects, hidden variation.

But what about traits that aren't just on or off?

Things like height or weight,

continuous variation.

Right.

These are the traits influenced by many genes, often called polygenic traits.

Each gene contributes a small additive effect.

And when you measure these traits across a population, you get that classic bell curve, the normal distribution.

Yep.

Like human height, most people are average, fewer are very short or very tall.

Exactly.

And this is really the kind of gradual,

heritable change that Darwin focused on, the cumulative power of many small variations.

We see this power really clearly in artificial selection, don't we?

Oh, absolutely.

Look at milk yield in dairy cows.

Humans have been selecting from more milk for centuries, but the really intense selection over the last, say, 60 years has been dramatic.

How dramatic?

The average dairy cow today produces something like two and a half times more milk than his ancestor did just 60 years ago.

That's a massive shift driven by selecting on existing quantitative variation.

But there are trade -offs, right?

You mentioned pleiotropy before.

Yes.

Pleiotropy is key here.

Genes rarely do just one thing when you select intensely for one trait, like milk yield.

You get bigger udders, sure.

Right.

But the genes involved often affect other things too.

So high -yield cows often have associated issues, like reduced fertility or more difficulty giving birth.

There's a cost to pushing one trait to an extreme.

Evolution involves compromises.

Always.

These traits involve potentially hundreds of genes.

How on earth do researchers pinpoint the genetic basis?

That seems incredibly complex.

It is.

But we have tools.

We use mapping techniques to find quantitative trait loci or QTLs.

QTLs.

So that's not a single gene, but a region of a chromosome.

Exactly.

A region of the genome that's statistically associated with variation in a specific quantitative trait like body size or bristle number on a fly or milk yield.

And finding a QTL tells you what?

That somewhere in that stretch of DNA, there are genes affecting the trait.

Pretty much.

It narrows down the search.

Initially, the thinking was that these QTLs would contain many genes, each with a tiny effect, adding up.

The additive model.

Right.

But sometimes, research turns up surprises.

The three -spined stickleback fish is a fantastic example.

Ah, the sticklebacks.

They vary a lot, don't they?

Especially between marine and freshwater forms.

Hugely.

Big differences in things like the number of bony armor plates on their sides, spine length, gill rakers.

Traits clearly linked to their environment.

Marine ones have lots of plates, freshwater ones often have very few.

So researchers used QTL analysis to find the region's controlling plate number.

They did.

And they pinpointed a major QTL responsible for a large chunk of the variation in plate number.

And inside that QTL, lots of little genes.

Nope.

They zoomed in and found that a single gene within that region, called ectodisplasin, was largely responsible.

Specifically, a derived allele, the low allele, causes the dramatic reduction in plate seen in many freshwater populations.

Wow.

So one gene mutation has this massive morphological effect that's strongly selected for in a specific environment.

Exactly.

It showed that sometimes, major adaptive shifts aren't just the slow accumulation of tiny changes.

Sometimes, it can be down to a single gene of surprisingly large effect, a precise molecular switch matching an environmental pressure.

That's a powerful finding.

It moves us beyond just saying it's complex to actually identifying the specific molecular players in adaptation.

And discoveries like the stickleback work, linking genetics clearly to observable evolution, really cemented the neo -Darwinian synthesis back in the 1930s.

Right.

The modern synthesis.

Bringing together Darwin's selection ideas with Mendelian genetics.

The consensus became clear.

Evolution is a population -level phenomena.

It's fundamentally about changes in allele frequencies over time, driven by forces like mutation, selection, migration, and genetic drift.

OK.

So if evolution is changed in allele frequencies, we need a way to measure no change, a baseline.

Exactly.

A null hypothesis.

And that's the elegance of the Hardy -Weinberg principle, or Hardy -Weinberg equilibrium.

The cornerstone of population genetics.

It truly is.

Developed independently by G .H.

Hardy and Wilhelm Weinberg.

Their core insight was this.

In a large,

randomly mating population, what we call a deem, where there are no other evolutionary forces acting.

Like no mutation, no selection, no one moving in or out.

Right.

Under those ideal conditions, the allele frequencies and the genotype frequencies will remain constant from generation to generation.

They stay in equilibrium.

And that equilibrium is predictable, right?

The famous 2p taller plus 2pq plus q2 equals a 1.

That's the one.

If you have two alleles, say t and t, with frequencies pot of taller sen, or pot of plus q must equal 1, then after one generation of random mating, the genotypes will settle into the frequencies 2 tallers for tt, 2p tallers for t, and 2 tallers for tt.

And it stays that way indefinitely, unless something disturbs it.

Precisely.

And it gets to that equilibrium state fast, just one generation of random mating does it.

But those assumptions are pretty strict.

Large population, random mating, no selection.

That doesn't sound like most real populations.

It doesn't.

And that's the point.

Hardy -Weinberg isn't meant to perfectly describe nature.

It's the baseline we compare nature to.

If a real population's genotype frequencies don't match the TPRH 2pqu2 expectation.

Then we know one of those assumptions is being violated.

We know some evolutionary force is acting.

Exactly.

It's a tool for detecting evolution in action.

Let's pick one of those assumptions.

Random mating.

What happens if mating isn't random?

Say through inbreeding.

Ah, inbreeding mating between related individuals.

That definitely violates the random mating assumption and messes with Hardy -Weinberg equilibrium.

How does it mess with it?

Well, here's a key point that's often misunderstood.

Inbreeding by itself does not change the overall allele frequencies that are in the population.

Really?

It doesn't make one allele more or less common overall?

No.

It doesn't create or destroy alleles.

What it does do is change how those alleles are packaged into genotypes.

Specifically, it increases the frequency of homozygous genotypes, Tt and T, and decreases the frequency of heterozygous genotypes, Tt, compared to the Hardy -Weinberg expectation.

So fewer heterozygotes, more homozygotes, but the total count of T and T alleles stays the same.

Exactly.

We measure the extent of this effect using the inbreeding coefficient, symbolized by $5.

It quantifies the probability that two alleles at a locus in an individual are identical by descent from a common ancestor.

And this increase in homozygosity, that's where inbreeding depression comes in, right?

That's the consequence.

Most populations carry rare recessive alleles that are harmful or deleterious when homozygous.

Alleles that are usually hidden in heterozygotes.

Right.

But inbreeding increases the chances that an individual inherits two copies of the same rare, deleterious recessive allele from that common ancestor.

Suddenly, the harmful trait is expressed.

Leading to reduced fitness, lower survival, lower fertility,

the depression.

Yes.

And the effect can be dramatic, especially for very rare alleles.

The text gives a striking statistic.

If a harmful recessive allele has a very low frequency, mating between first cousins can increase the chance of having an affected homozygous recessive offspring by something like 126 times compared to random mating.

Wow,

126 times.

So inbreeding doesn't change the gene pool's allele frequencies, but it drastically changes the expression of the variation within it by altering genotype frequencies.

Precisely.

It reshuffles the deck in a way that exposes hidden negative traits.

Okay, let's try and synthesize this deep dive.

We started with the idea that populations, not individuals, are the fundamental unit of evolution.

And evolution is measured as a change in allele frequencies within that population's gene pool.

The necessary fuel for this change is genetic variation.

It comes partly from new mutations which aren't entirely random in location and are often buffered by things like Hsp90, but mostly from the vast pool of existing polymorphism.

Right.

Selection can act very quickly on that existing variation, as we saw with DDT resistance and peppered moths.

Then we looked at continuous variation, traits influenced by many genes, and how QTL analysis helps us dissect their genetic basis.

And the sickleback example was key there, showing that sometimes major evolutionary shifts linked to the environment can be driven by changes in single genes with large effects, like ectodisplacent.

Not always just tiny additive steps.

And finally, we measure these changes against the backdrop of the Hardy -Weinberg Principle, our theoretical baseline of no evolution.

Which highlights how forces like selection, mutation, drift, and significantly non -random mating like inbreeding disrupt that equilibrium.

And drive real -world evolutionary change by altering allele or genotype frequencies.

So the really provocative thought here, maybe, is how far we've come.

We can now use tools like QTL analysis to go from observing a large -scale adaptation like sicklebacks losing plates in freshwater directly to pinpointing a specific gene, even a specific allele, responsible for that change in response to a known environmental pressure.

It's really quite remarkable.

It moves evolutionary biology from sometimes abstract theory to incredibly precise molecular detective work.

We're increasingly able to connect the dots from environmental challenge to molecular change to organismal evolution with real clarity.

It shows adaptation isn't always this mysterious slow process.

Sometimes the right molecular switch is just sitting there in the population's toolbox, waiting for the environment to demand it be flipped.

Something to think about.

Definitely.

Well, that brings us to the end of this deep dive.

Thanks, as always, for joining us.