Evolution Ch 7

Discrete vs Quantitative Traits

Discrete: qualitative; only a few, distinct trait options (ex: colors)

Quantitative: continuous; many possible values (ex: height, weight)

Quantitative traits are typically controlled by more than one gene

Why is male fitness harder to quantify than female fitness

Males can mate with multiple females, so since they aren’t directly producing the offspring it’s hard to know how many they produce over their life time

Fitness Components

viability, mating success, fecundity, survivorship, size, age at maturity

positive/negative directional selection

one extreme phenotype has highest fitness

mean shifts toward one direction

if a lot of genes contribute to the trait, new phenotypes can arise in the direction of selection so variation increases

stabilizing selection

intermediate phenotype has the highest fitness

the mean trait value doesn’t change

variance is greatly reduced (fewer individuals with either extreme)

disruptive selection

both extreme phenotypes have the highest fitness

the mean doesn’t change

variance could increase or be maintained

positive beta

positive, linear relationship

positive directional selection

negative beta

negative linear relationship

negative directional selection

negative gamma value

parabola that looks like a rainbow (ends are lower than middle)

stabilizing selection

positive gamma value

parabola is like a bowl (ends are higher)

disruptive selection

***Even though a trait might be an important adaptation for a population, why might natural selection not be detected?

• variance: there might not be enough variance in a trait to show selection

• sample size: depending on the individuals you observe, different selection can be occurring

• population mean/adaptive landscape

Correlational Selection

combinations of traits are more advantageous/better fit than others

leads to genetic correlation

Breeder’s Equation

R = h² * s

R → response to selection (the change in mean from one generation to the other)

h² → heritability

s → selection gradient (strength of selection)

A population of plants has an average height of 100 cm.

Researchers select the tallest plants, with an average height of 120 cm, to reproduce.

The next generation has an average height of 110 cm.

Find S, R, and h²

S = 120-100= 20

R= 110-100 = 10

h² = 10/20= 0.5

A population of fish has an average weight of 2.0 kg.

Scientists select fish with an average weight of 2.6 kg to breed.

The offspring generation has an average weight of 2.3 kg.

Find S, R, and h²

S = 2.6-2.0= 0.6

R= 2.3-2.0= 0.3

h² = 0.3/0.6 = 0.5

heritability in terms of different types of variance

V_a/ V_p

numerator: additive genetic variance

• Additive effects are passed directly from parents to offspring

denominator: phenotypic variation

• a combination of environmental variance, dominance variance (allele interactions at same gene), interaction/epistatic variance (genes interacting across loci) AND additive variance

What are reasons why a trait may have a heritability of zero?

If V_a = 0, no variation in that trait: still genetically controlled, but no variation exists for that trait (like 2 eyes in humans)

if the environment controls phenotype (like additional limbs in forms, which was caused by trematodes)

If a trait has a heritability of zero, does it mean that the trait is not controlled by genes? Why or why not?

No, just because heritability is zero, that does not mean the trait is not controlled by genes. This just means that there is no additive genetic variation contributing to the difference in this trait. Difference can be controlled by environment. It could also mean that the trait in controlled by genes, but there is no variation within the population.

Measuring heritability by Mid-Parent / Mid-Offspring Regression

If a trait is heritable, offspring should resemble their parents. By taking the average trait value of the parents (x) and the average trait value of offspring (y), the slope of the line represents the heritability.

Slope of 1 → highly heritable

Slope of 0 → not heritable (no correlation)

Measuring heritability by artificial selection

If a trait is heritable, selecting certain individuals should change the next generation. Start with mean trait value for a population, selection some extreme individuals and find average trait value (S), let those reproduce a new generation and calculate new mean of the offspring (R)

heritability = R/S

Explain circumstances where offspring-parent regression might give you higher estimates of h² than the true heritability

Offspring–parent regression can overestimate heritability when parents and offspring resemble each other due to shared environment, maternal effects, non-additive genetic effects, or non-random mating, rather than purely additive genetic variance.

Maternal/Paternal effects

traits/behaviors/environments of mom or dad that can influence traits in offspring nongenetically

ex: if a female eats more during pregnancy, offspring will be bigger due to better nutrition (maternal), sperm quality (paternal)

Can bias estimates of heritability because we assume physical resemblance = additive variance, but just because they look the same doesn’t mean the trait is genetically controlled and heritable.

How can offspring-parent regression still be used while getting a less-biased estimate of h²?

Father - Offspring regressions (when the father does nothing) or Mother - Offspring Regressions (when the mother does nothing) can eliminate parental effects. For this single parent regressions, h² = the slope x 2

Cross fostering (where the offspring are raised by someone else) also eliminates parental effects

if h² = 0, how will populations evolve?

Populations will not evolve and trait values will not change because there’s no additive genetic variance and trait difference are mostly environmental.

if h² is low, how will populations evolve?

Populations can evolve, but slowly, and the response to selection is small

if h² high, how will populations evolve?

This means most variation is genetic, so populations will evolve quickly and the response to selection will be high.

Pleiotropy

One gene controls multiple traits, positive or negative correlation

Can lead to genetic correlation because any allele that increase trait A will also increase trait B, so the traits become genetically linked in their effects.

Linkage Disequilibrium

Non-random association of alleles at different loci

Creates genetic correlation by physically linking alleles

Linkage equilibrium, vs Linkage disequilibrium

Equilibrium: alleles at different loci are completely random

disequilibrium: alleles at different loci associated non-randomly (come allele combinations are over or underrepresented)

How to use date to find linkage disequilibrium

Compare observed frequency of genotype to expected. If they are no equal, then linkage disequilibrium is occurring

Why do loci show linkage disequilibrium

If loci are extremely close together, it’s very unlikely that recombination would occur in-between them.

Correlational selection: if combinations of traits are selection for, correlations could arise

Explain why genetic correlations between two traits can constrain adaptive evolution

Opposing selection pressures could mean that Trait A is favored, but Trait B is not. Trait B would be “dragged along” and the population can’t meet optimal value for both traits

Trade-offs could occur. So, improving one trait could cause another to worse