Response to selection. Gene251/351 Lecture 7

Transcription

1 Response to selection Gene51/351 Lecture 7

2 Revision For individual parents we can predict progeny merit by Estimating breeding values of each parent EBV h P Averaging these to predict progeny merit Gˆ o EBV Sire + EBV Dam

3 Similarly for a selection policy For a given selection policy (breeding operation) progeny merit is predicted by Predicting the phenotypic superiority of selected parents Predicting superiority of progeny generation

4 Why is this useful? Predicting progeny merit for a selection policy (i.e. response to selection) is useful as it allows us to compare different selection policies For example, is there greater genetic gain if Breeding females are kept for three years only, requiring more replacements each year but quicker turnover OR Breeding females are kept for five years, requiring less replacements each year but slower turnover

5 Direct determination of phenotypic superiority of selected parents Assume that individuals are selected as parents only if their phenotypic value is greater than a truncation point, as shown below Truncation point Frequency Average phenotypic value of selected parents Phenotypic values Individuals selected as parents

6 Continued The selection differential (S) is the phenotypic superiority of selected parents (i.e. mean of selected parents population mean) Frequency Average phenotypic value of selected parents Phenotypic values S In this example S is ~9 units

7 Continued S is averaged if different proportions of males and females are selected For example Mean of selected males is 3kg, mean of all male candidates is 7kg. S male 5 kg Mean of selected females is 5kg, mean of all female candidates is kg, S female 3 kg Average S is 4 kg S S Male + S Female

8 A more useful way to determine S Rather than determine S directly (as in the previous slides) it is more useful to determine S from a knowledge of the selection policy

9 Determining S from knowledge of the selection policy First determine selection intensity (i)( Selection intensity (i)( ) is the number of phenotypic standard deviation units that selected parents are superior to the mean i is obtained from tables according to the proportion (P) of animals selected as parents selection intensity tables are provided in the back of both the course notes and prac manual

10 Continued Frequency P10% Standard deviation units i In this example i1.75 (from selection intensity tables for P10%)

11 Continued Note the relationship between proportion selected (P) and selection intensity (i)( Select few individuals low P and high i Select many individuals high P and low i

12 Continued Then determine S by multiplying i by the phenotypic standard deviation S i σ p The selection differential is equal to the selection intensity multiplied by the phenotypic standard deviation.

13 Continued Frequency P10% Standard deviation units Phenotypic units for σp5 In this example i1.75 and S8.75 as S i σ p 1.75 x 5

14 Continued i is averaged for males and females i i Male + i Female For example out of 100 males selected: P male % and i male out of 100 females selected: P male 80% and i female Average i is 1.38

15 Now that we have predicted the phenotypic superiority of the selected parents, we need to consider how much of this superiority will be passed onto the offspring

16 Predicting superiority of the progeny generation Only the heritable portion of the phenotypic superiority of selected parents will be passed onto the offspring Thus offspring mean superiority (over a no selection policy) equals the selection differential multiplied by the heritability

17 Continued Thus R gen Sh R gen iσ h p R gen i male + i female σ h p where R gen is response per generation. (you should memorise these equations they are important)

18 Another definition of heritability Regression of observed offspring performance on the mean of parental performance is equal to h b O P h Mean parent phenotypic superiority Offspring performance

19 From R to gen Ryear It is also useful to determine response per year (R year ) This requires calculation of the generation interval (L)

20 Generation interval (L) Generation interval (L) is the average age of parents when progeny are born L is calculated seperately for males and females and then averaged For example Equal numbers of and 3 year old bulls selected as parents L male.5 years Equal numbers of, 3 and 4 year old cows selected as parents L female 3.0 years average.75 years L average female

21 Continued Another example of calculating L Age structure of animals selected for breeding Age (years) Total Male Female L L L male female average (7x) + (5x3).4years (00x) + (150x3) + (100x4) + (50x5) years 3.0years

22 Continued Continued h L L i i R L L L and h i i R Given p female male female male year female male p female male gen σ σ L year gen where year gen x gen R year R Alternatively 1

23 Response to selection worked example Sheep breeder has 180 ewe flock, selecting for FW Rams first selected at years old, and mated for years Ewes first selected at years old, and mated for 4 years Each ram mated to 30 ewes, 90% lambing, 50:50 sex ratio No significant mortality in adults Trait heritability 0.5, and σ P 0.6kg What is R year?

24 Answer Age structure of animals selected for breeding Age (yrs) Total Male Female ewes, 90% lambing 16 lambs total (81 of each sex) Need to select 3 out of 81 males each year P3/813.7%, which corresponds to an i of.18 Similarly need to select 45 out of 81 females each year P45/8155%, which corresponds to an i of 0.7 L male.5years, L female 3.5years R R year year imale + i female σ ph FW is expected to increase by L + L male female x0.6x kg 0.07kg per year

25 Some revision problems For the following slides indicate which of the two breeding policy options would give the highest response to selection per year Recall imale + i female Ryear σ ph L + L male female

26 Continued 80% of females selected and % of males selected or 80% of females selected and 10% of males selected Hint consider the relationship between proportion selected (P), selection intensity (i) and response to selection (R)

27 Continued 5 males and 50 females selected, 1 progeny per female or 5 males and 50 females selected, progeny per female Hint consider what happens to P, i and R when the same number of animals are selected from an increasing number of selection candidates

28 Continued Selecting for a trait with V A 30 and V E 70 or Selecting for a trait with V A 10 and V E 90 Hint recall that h V A /V P, and V P V A +V E

29 Continued Selecting for a trait with V A 30 and V E 70 or Selecting for a trait with V A 3 and V E 7 Hint recall that σ P sqrt(v P )

30 Continued Males and females first selected at two years of age or Males and females first selected at one year of age Hint recall that generation interval (L) is the average age of parents when progeny are born