Download Heritability: The evolution of quantitative traits by artificial selection

Heritability: The evolution of quantitative
traits by artificial selection
After studying Drosophila Fly genetics we learned that some traits can be controlled by
simple Mendelian processes. But, what if the F1 generation in the initial cross completed in the
study was not all wildtype and the F2 generations was not 3:1? Does it mean that the trait under
consideration is not heritable? Are there other factors controlling heritable traits besides
Mendelian processes? What about traits that are not all or nothing in their expression? What
about traits that are thought to have a large environmental component like alcoholism? Is there a
way to determine how much of a given trait is under genetic control versus how much a trait is
influenced by the environment? And how do these traits evolve on a micro and macro-scale?
In the following activity, we will analyze some data produced over the last 4 years by
Bryn Mawr Intro Bio students. The study selectively bred Brassica rapa plants in order to
determine whether or not plant “hairiness” was a heritable trait and if so, what fraction of the
variation in “hairiness” is explained by genetic variation and how much is due to environment.
Assignment:
Read through the activity and answer the italicized questions in a new document you
create and submit to your “Portfolio” within 24hrs.
Quantitative trait variation
Unlike Mendelian traits that are controlled by only one gene, quantitative traits most
often results from many genes of relatively small affect whose exact number is unknown.
Description and analysis of variation and selection on such traits is based on statistical measures
and relations. Statistical variance quantifies variation around the average value of the trait. Such
phenotypic variance (VP) can be divided into a genetic component, the genetic variance (VG) an
environmental component, the environmental variance (VE) and a genotype by environment
interaction (VGE), thus;
VP = VG + VE + VGE.
Offspring will tend to resemble their parents both because of common environment and
because they share a common genetic background for the trait. However, selection operates only
on the genetic variance, thus a trait can only evolve if variation has a genetic component. Thus
assessing the genetic contribution to traits, VG is critical to understanding how traits might
evolve. The genotype by environment interaction term accounts for the fact that differences in
the environment do not have a uniform effect among all genotypes in the population. For
example, one genotype might grow better at 20°C but another better at 30°C. The environment
does not have a uniform effect among genotypes.
Natural selection and artificial selection
Natural selection stands as one corner stone of evolutionary biology and explanation for
the diversity of life on earth. Some have called natural selection a deceptively simple concept, or
instilled it with purposeful intent. However, natural selection is in essence a mathematical
process. It is simply differential survival and reproduction. Natural selection does not lead to
differential survival and reproduction, nor is it brought about by differential survival and
reproduction, natural selection is differential survival and reproduction. In this exercise with
artificial selection you will become familiar with the processes surrounding evolution and natural
selection. Fundamentally, artificial selection and natural selection are the same. Artificial
selection differs from natural selection primarily in that the reproductive success of the organism
hinges on a single character or small set of characters chosen by the human investigator rather
than by the organism’s overall survival and reproduction. Natural selection also lacks the
purposeful “directedness” of artificial selection.
Heritability and Selection
The rate of evolution by artificial or natural selection depends on the level of genetic
variation in the trait and the strength of the selection applied to the population. The process is
best illustrated with an example. Figure 1 shows the frequency distribution of a given trait, let’s
say beak size, in a population of finches. The distribution of individuals is bell-shaped (although
this need not be the case) and XPo indicates the mean value of the trait for the whole population.
Suppose that in the population, birds that have larger beaks are able to crack open a wider variety
of seeds and therefore store more fat and survive cold winters more often than birds that have
smaller beaks. During one winter, only finches with beaks of larger than a certain size survive,
so that the average beak size among the survivors is XS (Fig 1a). The differential survival
resulting from larger beak size is directional selection. If beak size is heritable, that is if the
variance in size of a bird’s beak is at least partially determined by genes,
a) Selection differential
frequency of individuals
XPo XS
40
30
20
10
0
S
frequency of individuals
b) Selection response
R
40
30
20
10
0
c) realized heritability
h2 = R/S
Fig. 1 Illustration of selection differential and
response for a population with normally
distributed trait values.
then the directional selection will produce a shift in average beak size in the next generation.
The difference between the average beak size among survivors and that in the original population
(which includes the survivors) is known as the selection differential (S: Fig. 1a).
We measure strength of selection as the selection differential (S). The selection differential is
simply the difference between the mean of the trait in the selected group of parents (XS) and that
in the entire base population (XP, Fig. 1a).
Selection response (R) is the change in the average value of the trait in the offspring
generation compared to that of the entire parental generation (not only the selected parents, Fig.
1b). A response to selection thus provides evidence for a genetic basis to the trait. Such a
response also represents an evolutionary change.
The relation between S and R can be used to provide an estimate of the genetic
component of the trait of interest. Selection response can be expressed as a proportion of the
selection differential, to yield h2 the realized heritability (Fig. 1c).
h2 = R/S
h2 = XP1-XPo / XS-XPo
(1)
(2)
Thus, if the realized heritability of a trait in a population = 1, then all the variation of that trait in
the population is due to genetic factors. If h2 = 0, then, R had to be zero and therefore, there was
no response to the selection. If the next generation shows no response to a selection force, then
that trait does not have a genetic component. That is, all the variation comes from the
environment or factors other than genetic make-up.
Selective Breeding
The plants you will be working with are the same species as cultivated turnip, pak choi, and
Chinese cabbage, each a product of selective breeding for specific phenotypic traits. Brassica
rapa forms part of a complex of related species that comprise many of the most important
vegetable crops world-wide, for example,

B. rapa: pak choi, Chinese cabbage, turnip, saichin

B. oleracea: kale, cauliflower, broccoli, head cabbage, Brussel sprouts, kohlrabi, collard
greens

B. juncea: brown mustard, mustard greens B. napus: canola
1. If you were trying to breed Brassica rapa to increase trichome number (hairiness),
describe in as much detail as possible the methods.
Results of First Round of Selective Breeding
In an attempt to increase trichome number or hairiness on Brassica rapa (thought to increase
resistance to drought as well as protect against herbivory), one hundred individual plants were
scored for trichome number on the first true leaf petiole (see Fig 2). Table 1 shows the
distribution of trichome number for the first original population (po) of 100 individuals. The
mean trichome number for the entire population was calculated as XPo = 3. Then the 10 most
hairy plants were selected and bred to generate a new offspring population 1 (p1). The mean
trichome number of the selected parents was calculated as XS = 16.
Petiole is between main
stem and leaf blade.
Figure 2. Photograph of trichomes on the first true leaf of B. rapa
Table 1: Trichome # of Original Populations
Trichomes
Gen 1 '05
0-4
5-9.
10-14.
15-19.
20-24.
25-29.
30-34.
35-39.
40-44.
45-49
>50
63
23
8
5
1
0
0
0
0
0
0
2. Given the properties of the original population and the mean trichome number of the 10
selected parents, predict and explain what the next generation will look like with regards
to trichome number. Hint: Think about what the distribution of the offspring generation
would be if h2 = 1.0 and h2 = 0.0. Also, what would the XP1 be for both cases? You may
want to open MS Excel and use the spreadsheet functions to graph and make
calculations.
Table 2 shows the distribution of the second generation (offspring of the selected parents) and
three additional rounds of selective breeding carried out. Table 3 gives the mean of each entire
generations and mean trichome number of the selected parents in each round of breeding.
Table 2: Distribution of Trichomes for the all rounds of
selective breeding. Gen 1-5 are the different
generations/populations used in the breeding. Each
subsequent generation is the offspring of the previous
generation.
Trichomes
0-4
5-9.
10-14.
15-19.
20-24.
25-29.
30-34.
35-39.
40-44.
45-49
>50
Gen 1
Gen 2
Gen 3
Gen 4
Gen 5
63
23
8
5
1
0
0
0
0
0
0
21
27
26
13
10
2
1
0
0
0
0
5
11
30
28
15
5
3
1
2
0
0
0
2
15
30
28
14
8
2
0
0
0
0
1
3
13
25
22
14
14
5
4
0
Table 3: Mean trichome # of entire populations
and selected parent populations. XPo is mean of
entire original population. X P1 is mean of entire
offspring population. Xs is the mean of only the
top 10 selected parents
Iteration of Selective Breeding
Means
G1-G2
G2-G3
G3-G4
G4-G5
Xp0
3
10
16
21
Xp1
11
15
23
28
Xs
16
18
29
33
Analyze the reported results from Table 2 and 3 and answer the following questions.
3. Is trichome number a heritable trait in Brassica rapa and if so, what portions of the
variation in trichome number is explained by genes? Show calculations.
4. Why does realized heritability of trichome number vary from generations to generation if
it is the estimated genetic contribution of a given trait? That is to say these data seem to
suggest that the genetic contribution to a trait can change from one generation to the
next. Explain.
5. If you were to do another round of selective breeding starting from Generation 5
population (G5-G6), exactly what would you expect the mean trichome number to be of
the next offspring generation (XP1 of G6) if the top 10 hairiest plants of G5 had a mean
trichome number of 51 (XS= 51)?Make a specific predication. Explain your assumptions
and show your calculations.
6. If microevolution is defined as change in heritable phenotypes within the original range
of variation of a trait and macroevolution is a change of phenotypes outside the original
range of variation, how would you classify this Brassica rapa breeding study? Explain
and support with evidence (Make a figure showing the distribution of trichome number
for the different populations and then explain).