Midterm

(1) [2 points] True or False: "Evolution by natural selection requires variation in fitness among individuals, but evolution itself does not."

True, evolutionary change can occur through mutation, drift, and segregation distortion, among other processes, which do not necessitate differences among individuals in fitness.

(2) [2 points] True or False: The Chain of Being describes the modern evolutionary perspective on the relationships among species.

False, the Chain of Being refers to Aristotle's linear hierarchy of organisms. The modern evolutionary perspective on the relationships among species is not hierarchical, but tree-shaped with all living species on the tips of this tree of life.

Circle the best answer to the following questions. [Four points each]

(3) Of the following people, who did not support the concept of biological evolution?

(4) Lamarck's view that organic progression occured along a hierarchy of more advanced forms led him to believe in which of the following concepts?

(5) The graph to the right shows the results from Robertson's 1955 selection experiment on thorax length in Drosophila. Focus only on the bottom half (the down-selected line). The selected lines appear to reach a selection limit at about 0.85 mm. There are several possible reasons why the response to selection may have tapered off. When the direction of selection was reversed (dashed line), however, thorax length did not increase rapidly. This provides evidence that the selection limit of 0.85 mm was reached because:

(6) Inbreeding depression is a term used to describe the common finding that inbred individuals (which are more homozygous than average) often have a lower fitness. Inbreeding depression could be caused by all of the following forms of selection except which one?

(7) The graph on the right shows the change over time in the frequency of the A allele in a diploid one-locus model under four different fitness schemes. On the graph, label each curve with a letter corresponding to the appropriate fitness regime. [Note: The starting allele frequency is 0.05 in each case.]

(8) [15 points] William Bateson and colleagues reported one of the first examples of physical linkage. They studied flower color and pollen shape in sweet peas. They bred two different lines, one fixed for the dominant alleles causing purple flowers (P) and long pollen (L), and the second one fixed on the recessive alleles causing red flowers (p) and round pollen (l). They crossed pure-breeding parents to generate hybrids (PL/pl). They then back-crossed these hybrids with red-round individuals (pl/pl). The progeny of these back-crosses were then typed, with the following results:

Genotype: PL/pl Pl/pl pL/pl pl/pl
Frequency:
0.44
0.06
0.06
0.44

(9) [15 points] Travisano et al (1995) describe their first experimental design as follows: "A single genotype of E. coli strain B was cloned and used to found 12 replicate populations, which were serially propagated for 2000 generations at 37C in glucose-limited medium...One genotype from each of the 12 replicate populations was cloned and used to found 3 new replicate populations. These 36 populations were then propagated for another 1000 generations under the ancestral conditions, except that an equal concentration of maltose replaced glucose." Describe how you might change their experimental design to:

(a) increase the relative contribution of "adaptation" to the fitness variation among lines.

"Allow more generations in the new medium (maltose) to allow for more adaptation to occur"

(b) increase the relative contribution of "chance" to the fitness variation among lines.

"Reduce the amount of media to grow on which will decrease the population size and genetic drift can have a greater effect."

(c) increase the relative contribution of "history" to the fitness variation among lines.

"Grow the 12 replicate populations at 37C in glucose-limited medium for much longer than 2000 generations - the more generations at this stage the greater the contribution of history"

[NOTE: There are several possible answers to the above question. This student got full marks for the above answers.]

(10) [20 points] Dr. Stephanie Schrag and colleagues observed evolutionary changes in a strain of Escherichia coli that recently became resistant to streptomycin (an antibiotic). They called the new streptomycin resistant allele Strr, whereas the original strain carried a streptomycin sensitive allele Strs. After growing the new streptomycin resistant strain for 180 generations in the absence of the antibiotic, Dr. Schrag noticed that the strain was still streptomycin resistant, but it appeared to grow much faster. She concluded that a second mutation had appeared that compensated for the fitness costs of Strr without eliminating resistance.

Dr. Schrag then competed the various strains, used the haploid selection formula for changes in allele frequency over time, and estimated the fitnesses of the various strains as:

Genotype:
Strs not compensated
Strr not compensated
Strs compensated
Strr compensated
Strain name:
A
B
C
D
Fitness:
1
0.84
0.79
0.93

If the D strain (Strr compensated), which evolved from the B strain by the appearance of the compensatory mutation, were cultured for a period of time in the absence of streptomycin, would you expect the resistance to disappear rapidly or not? Use the above fitness information and justify your predictions.

"Resistance would not disappear rapidly. If a mutation occured to make Strs compensated or Strr not compensated, both of these have a lower fitness and would disappear. Only if a double mutant oocured to make Strs not compensated (with a higher fitness) would resistance start to disappear, but a double mutant is much less likely to occur and the difference in their fitness (Strs not compensated vs Strr compensated) is not that great."

[NOTE: Many students said that 0.93 and 1 were very similar and that the fitness differences is small. Actually, a difference in fitness of 7% is quite large. If strain A appeared, it would very rapidly sweep through the population (in a few hundred generations -- about a week or so). What takes time is for strain A, with two mutations, to appear. Before it does, further compensatory mutations could occur making Strr even less costly.]

(11) [20 points] Dr. Gromko et al (1991) performed a selection experiment in Drosophila on copulation duration (= length of time that the flies have sex).

(a) Previous work had shown that the correlation in copulation time between fathers and sons was 0.23, while the the correlation in copulation time between mothers and daughters was 0. Estimate the narrow-sense heritability based on the father-son correlation. Estimate the narrow-sense heritability based on the mother-daughter correlation. Average these two heritabilities estimates to get an overall heritability estimate.

Corr[parent-offspring] = 1/2 h2, so h2 = 0.46 based on the father-son correlation and 0 based on the mother-daughter correlation. Averaging 0.46 and 0 gives 0.23.

(b) Over the first three generations of selection, the authors applied a cumulative selection differential of 25. The mean copulation duration of the selected flies went from 19 minutes to 23.5 minutes during these three generations. Estimate the narrow sense heritability from this response to selection. The authors conclude that the realized heritability is "in very good agreement with predictions based on heritabilities estimated from parent-offspring designs" (i.e. within a few percentage points).

0.18

(c) In the above experiments, copulation duration was timed to the nearest five seconds. Describe how heritability (measured from parent-offspring correlations or from the response to selection) would change if instead the scientists timed copulation duration to the nearest minute. Justify your answers.

There were two sorts of answers that we accepted, as long as the argument made was logical and coherent.

(A) By increasing the measurement error, the measured phenotype of an individual would be more subject to "random" environmental perturbations (whether its copulation time is rounded up or down). This introduces more environmental noise into the experiment, which would act to increase VE. This in turn would reduce the heritability, VA/VP by inflating the denominator. We would also expect parent-offspring correlations to decrease when the phenotype is less accurately measured, because more of the variation in phenotype will be due to random measurement error and less due to shared genes. Similarly, when a cohort is selected for increased copulation time, if the scale of measurement is coarse, then those chosen to reproduce may not actually be the longest copulating individuals, so the response to selection will be less strong, and heritability would be estimated to be lower if it were estimated from response/selection differential.