Tobias Schrag part Flashcards
Describe the augmented design and how to compute the adjusted value for the entries. (Slide
Twenty of Field designs).
-An augmented design is a type of unreplicated trial: genotypes are unreplicated entries and checks
are replicated.
-Only the replicated checks are analysed to estimate block effects.
-The unreplicated entries are then adjusted using the block effects.
Mention the advantages of the alpha design compared to the balanced square lattice design.
Alpha design is a type of partially incomplete block design, it is more resource efficient, flexible as there is no block size restriction, informative and robust compared to the balanced square lattice design
Mention differences between the factorial and the diallel design. (Slide 5 matting designs)
Factorial
Two sets of parents used as male and female
Estimates: 1. Genotypic value of crosses
2. GCA
3. Variance components
Diallel
One set of parents used as both male and female
Estimates: 1. Genotypic value of crosses
2. GCA and SCA
3. Variance components
Which are the requirements to carry out a factorial or diallel? (Slide 9 of matting designs)
Each female must be able to be mated with several males
*Genetically identical plants are available: Clones, inbred lines, DH lines
*A plant has several female inflorescences: Ears of a maize plant
Hybrid mechanism
*Mechanical
*Cytoplasmic male sterility (CMS)
What is the meaning of a fixed or random model in a diallel design?
Fixed model
*Parents are genotypes of interest: as fixed factors
*Conclusions refer to the tested parents only
*Estimation of general and specific combining ability
Random model
*Parents are a sample from a population of interest: as random factors
*Conclusions refer to the population
*Estimation of variance components
Define covariance and correlation and how those two concepts are related. (Slide 7 of variance
components).
The Covariance between two random variables x and y measures if large values of x often occur together with the large values of y, whereas correlation is a statical measure (expressed as -1 to +1) to describe the size and the direction of the relationship of two or more variables. Correlation between two random variables is derived through normalizing the covariance with the standard deviation
Explain how the partition of variance helps to estimate variance components. (Slides 11 y 14 of
variance components)
Partitioning the variance helps to estimate variance components as the variation between group (due to treatment variance σ2α) and within group (due to error variance σ2e) can be derived. SSTm=SSA+SSE
Define entry-mean heritability and explain how the breeder could influence it.
OR: Which heritability can be directly influenced by the breeder, and how is this heritability
defined? Which factors have an influence? (Slide 23 of Heritability)
h2=σ2g/σ2*p
Entry mean heritability is the proportion of the phenotypic variation in a group of plants which is due to the genetic difference among them. It can be influenced by changing σ2P, eg. In an experimental design, if Location, Year and Replications are higher then σ2m masking variance would be small due to which σ2p will be small. Furthermore if Location, Year and Replications are infinite then σ2m masking variance would be 0 and σ2p would be negligible or close to 0. σ2p can also be equal to σ2g this making h2=1.
What is the meaning of pre-test estimation of response to selection and provide two equations
that could be used to compute it. (Slides 21, 23 y 24 of Response to selection)
OR: Describe the purpose of estimating the response to selection in the pre-test situation.
OR: Give two equations for response to direct selection (pre-test) and name the factors
In a pre-test estimation of R-to-S, variance parameters and heritability estimates from previous
experiments are used to compute expected selection differential (S), from which R-to-S (R) can be
calculated. Expected R-to-S can be used to: determine the required size of the field trial to obtain a
desired R to S, and to compare expected response to selection for alternative scenarios.
R=h2S=h2(¯xs-¯x) Check slide for formula
How many replications in a trial are needed for optimum allocation? Why? And why is that
number not always considered under practice conditions? (slide 6 of optimal allocation)
To reduce masking variance, the best approach is to have a single replication and allocate the resources towards expanding the number of locations or extending the duration of the study. Nevertheless. Practical limitations, such as restricted access to certain locations or time constraints, may make it necessary to conduct research with two or more replications.
How to estimate indirect response to selection, please mention the equation. (Slide 19 of
indirect selection)
OR: What is indirect selection? Give the equation for the expected gain of indirect selection.
Name the factors used in the equation (similar to question 11 above).
Selection is carried out for the phenotype x1 of a trait 1 in order to improve the genotypic value y2 of
trait 2. Prerequisite: the genotypic values of both traits are highly correlated.
i1 selection intensity on trait 1
ρg genetic correlation between traits
h1 heritability of trait 1
σg2 genetic variance of trait 2
Check pdf
How to estimate the efficiency of indirect selection (Slide 20 of indirect selection)
OR: Give the equation for the efficiency of indirect selecti
R1,2/R2=i1h1ρg/i2h2
Where R1,2 is the indirect R-to-S and R2 is the direct R-to-S.
Indirect selection is more efficient than direct selection if: h1 ρg > h2 (where i1 =i2)
What is index selection and which parameters are required to compute it? (Slides 2 and 7 of
index selection)
The selection index is a multiple regression predictor of progeny performance or breeding value for
an objective, typically economic merit, from a set of observations, on an individual or its relatives.
Parameters per trait: heritability and progeny performance
a. The selection index is a method of predicting the performance or breeding value of progeny for a specific objective, usually related to economic merit. It is based on a multiple regression analysis of observations collected from an individual or its relatives. The index incorporates two parameters per trait. The heritability and the progeny performance.
Which procedure needs to be followed when the variables are correlated? OR Index selection: which approach provides the optimum index weights for correlated traits? Which information is required for the determination of the weights?
For correlated traits in a selection index, apply the Hazel-Smith index. The Hazel-Smith index uses information on 1. Phenotypic means, 2. phenotypic and genotypic covariance matrices P & G and 3. economic weights. The assumption is that the correlated traits follow a multivariate normal distribution. The Smith-Hazel index represents a multiple regression approach. Economic merit of each trait is required for the weights.
Define multi stage selection. (Slide 2 of multistage selection)
Selection in a Y year breeding program is carried out with decreasing G and increasing L and/or R.
Goal is to evaluate the best genotypes more precisely and not to waste testing capacity in later years
on genotypes with low performance.
b. Multi-stage selection is a plant breeding method where, genotypes are selected at multiple stages of the breeding programme hence focusing on the best genotypes. This way the resources can be used efficiently on increasing the number of locations and or replications for the breeding programme. This is a good approach as it saves times from testing low performing genotypes.
