Last update: January 14, 2011 10:33:35 AM E-mail Print

 

INTERPRETATION AND APPLICATION OF PERFORMANCE TESTING DATA

 

W.J. Olivier# & M.A. Snyman

Grootfontein Agricutural Development Institute, Private Bag X529, Middelburg (E.C.), 5900

#E-mail: Willem Olivier

 


INTRODUCTION

Livestock farming is a very important sector in the South African agricultural industry, especially when it is considered that 70% of agricultural land is suitable for livestock farming only.  This sector was also responsible for 45% of the gross income being generated by the agricultural industry (Abstract of Agricultural Statistics, 2008).  It is therefore important that this land be utilised to its full capacity.

To reach this goal, farmers or breeders must make decisions based on scientific principles to ensure that it is successful and most profitable.  It is important to realise that the interpretation of information will have a vital effect on the success or failure of the business.  One must also realise that the technological age that we are living in, makes it possible that information regarding market trends and production results is easily accessible.  This information can then be implemented in the daily decision making on the farm.

Information that has an effect on production and reproduction of a flock can also be used to improve the genetic abilities of the flock through correct selection.  It is, however, important that the selection made, is of economical importance for the flock.   Step one to achieve genetic improvement in the flock is to draw up a list of selection goals that will improve the production potential of the flock and therefore profitability.  Selection criteria should then be identified to ensure that selection goals are achieved.

The purpose of this paper is to enable producers to understand and use information on production and reproduction to genetically improve their flocks.  Implementing the principles of genetic improvement is of great importance to both stud and flock farming in improving profitability.  Stud farmers will therefore have to ensure that genetic improvement in a breed is achieved and that the genetically improved breeding material is available to flock owners.  Flock owners should ensure that the breeding material they buy, is genetically better than their own animals.  This will result in production and reproduction improvement and subsequent improvement in profitability.    

There are two tools available for producers when doing selection, namely indices and breeding values.  It is important to be familiar with the differences in indices and breeding values before it can be applied successfully.

 

Indices and BLUP breeding values

An index is an indication of the phenotypic value (actual measurement) of an animal as a deviation from the group average in which the animal was tested.  This means that if the indices of an animal for body weight (BW), clean fleece weight (CFW) and mean fibre diameter (MFD) were 112%, 107% and 89%, respectively, that the animal was 12% heavier, shore 7% more wool and was 11% finer than the group average for the respective traits.

It is also important to note that the distribution of phenotypic values of different traits is different and that some have narrower distributions than other traits.  This is because the scales between the minimum and maximum values of different traits are different.  For example, BW can vary by 20 kg between the lightest and heaviest animals, and CFW only 3 kg between the heaviest and lightest fleeces.  It is clear that BW will have a broader distribution than CFW.  This means that an index of 110 for BW is not the same as an index of 110 for CFW.  An index of 110 for BW (upper third of the data) will roughly be equal to an index of 116 for CFW, 105 for MFD and 112 for staple length.  For an animal to be in the top 5% for BW, CFW, MFD and staple length (STPL), its indices for the respective traits should be 119, 133, 89 and 124.

An estimated breeding value (EBV) for a trait is an indication of the genetic value of an animal as a parent.  In other words, how the offspring of an animal will be different from the average of the population or breed.  Figure 1 shows a schematic representation for estimating a breeding value for weaning weight of a ram.  Breeding values are more accurate than indices because all available information about the performance of an animal, its parents, ancestors, half-siblings and the performance of his descendants (if any) are used for an EBV estimate.  Owing to the fact that a male or female passes only half its genes to its offspring, the breeding value of an animal must be divided by two to give an indication of the contribution of the animal to its offspring.

Differences between the accuracy of selection when only the performance of an animal is used during selection, in contrast to the inclusion of data from all related animals, are summarised in Table 1.  It is clear from the table that guesswork will be taken out of selection by making use of breeding values, because information from related animals were taken into account.  This will increase the accuracy of selection.  As shown in Table 1, the accuracy of selection for BW and CFW improved by nearly 30% and MFD by about 10%.

Despite the improvement in selection accuracy, the EBV has a further advantage over indices.  Animals born in different seasons, years and even flocks can be compared directly with one another.  Indices on the other hand, can only be used to compare animals evaluated in the same contemporary group. 

 

Figure 1. Schematic representation of the estimation of a breeding value for weaning weight of a ram

 

Table 1. Accuracy of selection when using only own performance versus own performance plus data from related animals

Trait

Accuracy of Index

Accuracy of Estimated breeding value

Body weight (BW)

0.57

0.84

Clean fleece weight (CFW)

0.51

0.83

Mean fibre diameter (MFD)

0.74

0.85

 

In Table 2, the weaning data of two Afrino rams and their progeny are shown.  From the table it is clear that the index for weaning weight of the two rams differ by 12%.  The breeding values for weaning weight also differs by more than 1 kg, but the difference was reversed.  This means that the largest animal on its own performance will not necessarily always be the animal with the best genetic ability.  It supports the fact that if estimated breeding values are used during selection, it will increase accuracy.  Therefore, as estimated breeding values becomes available, it should be used during selection.  If the progeny of the two rams were compared, it is clear that the progeny of the ram with the higher EBV, were slightly larger and that their average EBV for weaning weight was also 0.40 kg higher.

 

Table 2. Weaning data of two Afrino rams and their progeny

 

007-045

007-109

Own performance

 

 

Waning weight index

125

113

Weaning weight breeding value (kg)

1.43

2.50

Progeny performance

 

 

Weaning weight (kg)

28.5

29.1

Weaning weight index

99.9

101.6

Weaning weight breeding value (kg)

0.09

0.50

 

The main aim of selecting replacement animals is to ensure that the breeding objectives of a flock or breed are met.  In order to achieve this, breeders must have a rough idea of the genetic potential of their flocks for a specific trait and know how to use the selection tools to achieve the breeding objectives.

The genetic trends for body weight in four flocks are presented in Figure 2.  If the genetic level of the flock of a breeder is the same as that of Flock 4, animals from any of the other three flocks will improve his flock, because they are genetically superior to his flock.  If it is the same as Flock 2, only animals from Flock 1 must be considered as the other flocks are genetically stronger, and genetic improvement in the flock may be hampered.  If a breeder knows the genetic potential of his flocks it can also be used to determine the minimum requirements for achieving the objectives when selecting male animals.

 

Figure 2. Genetic trends for body weight in four flocks

 

Take for example a flock with the genetic potential for BW, CFW and MFD of 2.00 kg, 0.60 kg and -0.80 μm, respectively, and the selection objectives of the flock are to increase BW, maintain CFW and decrease MFD.  This means that the average breeding values of rams selected should be better than these values to ensure genetic progress.

The genetic potential of a flock, the breeding values for eight possible rams and the averages of best and poorest four rams are summarised in Table 3.  The breeding values of these eight rams in Table 3 will be used to illustrate the selection of breeding sires.  If four rams from the group have to be selected, it is important to ensure that the rams that best meet the selection objectives of the flock, based on their production data, are selected.  Two of the rams (1 & 2) meet all three requirements and will therefore have a positive effect on the selection objectives.  Only two of the remaining rams (4 & 6), meet two of the three requirements, while the other four rams only meet one of the requirements.

 

If the four best rams (rams 1, 2, 4 and 6) are selected, it is important to determine whether the rams will meet the requirements of the selection targets.  From the table it is clear that the average breeding values of the four ram for BW, CFW and MFD, (3.00 kg, 0.88 kg and -1.95 µm) was better than the minimum requirements.  The use of these rams will result in the improvement of the BW of the flock, while CFW remain constant and MFD decrease.  If the information was not available, the wrong rams could easily have been selected, which could result in the selection targets not been achieved, as is the case with the four worst rams (rams 3, 5, 7 and 8) in the group.  The same principle applies to the purchase or selection of rams when only indices for the different traits are available.

 

Table 3. Genetic potential and breeding values of eight rams available for selection as sires

 

Rank

BW

CFW

MFD

BW

CFW

MFD

Genetic potential

 

2.00

0.60

-0.80

 

 

 

Selection target

 

2.40

0.80

-1.60

 

 

 

1. 06-3868

2

2.6

1.2

-1.8

U

U

U

2. 06-3874

1

2.4

1.0

-2.1

U

U

U

3. 06-3905

8

1.9

1.5

-0.6

N

U

N

4. 06-3968

6

3.0

0.6

-2.0

U

N

U

5. 06-3975

4

1.6

1.0

-1.4

N

U

N

6. 06-4056

3

4.0

0.7

-1.9

U

N

U

7. 06-4080

7

2.0

0.6

-1.5

N

N

N

8. 06-4168

5

1.6

1.2

-1.5

N

U

N

Average: Best 4 rams

 

3.00

0.88

-1.95

 

 

 

Average: Poorest 4 rams

 

1.78

1.10

-1.25

 

 

 

Average: All 8 rams

 

2.39

0.98

-1.60

 

 

 

 

When rams are selected, the indices or EBVs of rams must be used firstly to identify the rams that will improve the selection objectives of a flock.  Thereafter, the rams are visually assessed.  It is important that the process is not reversed, as a male with good visual traits, but with poor performance records, can be selected.  It happens too often that a group of rams purchased or selected have a negative effect on selection objectives, as the performance data were not available or incorrect information was used.

 

Standard deviation (SD) and Coefficient of variation (CV) OF FIBRE DIAMETER

The fibre diameter of fibres in a wool sample with a MFD of 19 µm could range from 7 μm to 37 μm.  It is important that the distribution is quantified as the width of the distribution curve has an effect on the staple structure and consequently the processing traits (Lamb, 1992).  Two parameters are used to quantify the width of the distribution curve, namely coefficient of variation (CV) and standard deviation (SD).

Standard deviation (SD) is the mathematical measurement of variation, defined as the average deviation from the mean.  Figure 3 shows an illustration of a normal distribution of fibres within a wool sample.  If the observations are normally distributed, 68% of all observations are within one standard deviation on either side of average, 95% of all observations within two standard deviations from the mean and 99% of the observations within three standard deviations.  In the example (Figure 3) the average MFD and SD are 19 µm and 1.5 µm respectively.  Therefore, 68% of the measurements will be between 17.5 and 20.5 µm, 95% of the measurements between 16 and 22 µm and 99% of the measurements between 14.5 and 23.5 µm.

 

 

Figure 3. The normal distribution of fibre diameter in a wool sample

 

Standard deviation relates only to a specific wool sample and can not be compared with the SD of other animals.  CV however, can be used to compare animals and is a function of SD.  The formula used to calculate CV:

CV (%) =

Standard deviation  * 100

Mean fibre diameter

CV is an indication of the normal distribution of the fibres within a wool sample.  The lower the CV of an animal, the higher and narrower the normal distribution curve.

 

Comfort Factor (CF)

Fibres with a diameter of more than 30 μm cause a scratchy feeling by a garment (Dolling et al., 1992). CF is the measure that is used to indicate how many of the fibres are finer than 30 μm and is expressed as a percentage.  If the CF of an animal is 99.6%, it means that only 0.4% of the fibres are over 30 μm.

 

Reproduction traits

Reproduction is probably the most important trait in any livestock enterprise on which selection should be based.  This is because reproduction has, in the first place, a direct impact on the income of a flock, and secondly an influence on the selection intensity.  The more animals available for selection, the more stringent selection can be applied and the faster the selection progress will be.

The reproduction information obtained from the National Small Stock Improvement Scheme includes the following production traits of a ewe:

The number of production years of a ewe is equal to her number of lambing opportunities.  The number of lambs born and weaned is the number of lambs the ewe produced in her productive life.  It is desirable for ewes not to miss a lambing opportunity and that she weans at least the same number of lambs as production years.

The weaning weights of all the lambs of a ewe per lambing opportunity are used to calculate her total lamb weight weaned for each lambing opportunity.  The total weight of lamb weaned of each lambing opportunity of a ewe is added up to calculate the total weight of lamb weaned over her lifetime.  This weight is used to calculate a ewe productivity index within a production year group.  For example, all the ewes with five production years, forms a group.  Hence, from this index the productivity deviation of a ewe is calculated to determine whether a ewe is better or worse than the average of the production-year group.  The mean lamb index of a ewe is calculated from the weaning weight index of each lamb weaned and should serve as an indication whether or not the ewe weaned above average lambs.  For some breeds a breeding value for total weight of lamb weaned is also estimated.

 

RELATIVE ECONOMIC VALUE

The relative economic value (REV) is a selection index that predicts the contribution of individual animals to the profitability of a sheep farming enterprise (Herselman & Olivier, 2010).  The REV is an estimated breeding value for profitability.  The breeder receives two REVs; the first value is selection for reproduction indirectly through body weight, while with the second value direct selection for reproduction can be performed.

The contribution of estimated breeding values on the relative economic values calculated from the wool and mutton prices from 2006 to 2010 is summarised in Table 4.  It is evident from this table that reproduction contributes almost 75% to the estimated breeding value for profitability and it is therefore the most important trait for small stock farmers.

 

Table 4. Contribution of estimated breeding values on the relative economic values

 

Contribution of each trait (%)

Breeding value

Relative economic value

(Excluding reproduction)

Relative economic value (Including reproduction)

Body weight (BW)

26.15

4.19

Clean fleece weight (CFW)

30.81

10.06

Mean fibre diameter (MFD)

27.89

7.76

Staple length (STPL)

14.48

4.32

Total weight of lamb weaned (TWW)

0.00

73.44

 

Conclusion

It is imperative that stud and flock farmers manage their livestock according to scientific principles to ensure success and profitability.  Therefore, the selection of replacement animals must be done in the most efficient way possible, to ensure genetic progress.  Whether by auction or flock selection, available information should be used to make the correct selection decisions.  If no information is available when buying rams, the buyer should demand that the information be made available.  It is also very important that animals, not meeting these economic selection criteria, are culled, without the breeder being influenced by the visual traits.

 

REFERENCES

Abstract of Agricultural Statistics, 2008.  Directorate: Agricultural Information Services, Private Bag X144, Pretoria, 0001.

Dolling, M., Marland, D., Naylor, G.R.S. & Phillips, D.G., 1992.  Knitted fabric made from 23.2 µm wool can be less prickly than fabric made from finer 21.5 µm wool.  Wool Tech.  Sheep Breed. 40, 69 – 71.

Herselman, M.J. & Olivier, W.J., 2010.  Description of a model for the calculation of breeding values for profitability.  Grootfontein Agric. 10, 67 – 75.

Lamb, P.R., 1992.  The effect of fibre diameter distribution on yarn properties.  Wool Tech. Sheep Breed. 40, 65 – 68.

 

Published

Grootfontein Agric 11 (1)