SELECTION FOR MILK YIELD

In the last lecture we saw that selection for beefcharacteristics is fairly straightforward because most commercial beef characteristics are highly heritable. Selection can takeplace using the phenotype, or appearance, of the animals being selected for a breeding programme. Selection for milk yield in dairy cows is much more difficult because milk yield has a low heritability. When selecting animals for a breeding programme in a dairy herd, the following points should be considered.

As milk yield has a low heritability, there is a low correlation between the phenotype appearance and performance, and the genotype (genetic make-up) of an animal. It is impossible to look at a dairy heifer and to say that this animal will be a good milker. It is also impossible to look at adairy bull and to say that this animal will breed good milkers. In a bull, there is no link at all between the appearance of the animal and its ability to breed high yielding cows.

Many dairy cows never fulfill their potential for milk production because of poor environment. If the feeding, milking, housing and general management of the cows is poor, they cannot produce their maximum amounts of milk and the farmer will never know their true potential. As milk yield has a low heritability of about 20% this means that 80% of the milk production of a herd depends on the management of the cows; in other words, the environment.

PREPOTENT: Effective in transmitting hereditary characteristics to its offspring.    

The most important single factor in a dairy herd is the bull. Cows can produce good individuals, but a good bull can improve the whole herd because 50% of the genes of all the heifers in the herd come from the bull.

The breeder is looking for bulls that are Prepotent. This is the ability of an animal to pass its desirable characteristics onto his offspring.

GENE GROUPS

In the genotype of an animal, each gene has the main task of carrying a characteristic, or trait, and the secondary task of influencing or modifying the effects of other genes or groups of genes.

Single Genes are called the genes of major effect and they control the simple and relatively unimportant tasks like determining coat colour, horns etc.

Linkage Groups are groups of genes on the same chromosome, often lightly bound together. These groups do the most important jobs, such as controlling the formation of glands. The more important and complex the factor, the larger the number of genes and gene groups involved in that control. This acts as an insurance of nature, because the alteration of a few genes in a large group is less likely to alter the growth and development of the animal.

In practical breeding, particularly for milk yield, it is these large groups of genes, the linkage groups, which the breeder is trying to alter. With a simple trait, the breeder is looking for an animal, particularly a bull that is homozygous and dominant for that trait. Take the case of an Aberdeen Angus bull that has the phenotype and genotype for black coat colour and no horns – it is polled.

The phenotype of the animal is that it has a black coat colour and no horns and this can beseen by looking at the bull. The genotype would be:

COAT COLOURBBHOMOZYGOUS AND DOMINANT FOR BLACK COAT
 BbHeterozygous for black coat – not dominant
 bbRecessive for black – a red coat
HornsPPHomozygous and dominant for polled
 PpHeterozygous for polled – not dominant
 ppRecessive for polled – horned animal

In this case the Aberdeen Angus bull would have the genotype BB PP, being homozygous and dominant for both black coat colour and polled head. If this animal were mated to cows of the same genotype, all the offspring would always have black coats and be polled. With simple traits such as these, controlled by single genes, it is an easy matter to produce animals that are homozygous and dominant. Undesirable recessives such as red coat colour and horns can be identified and removed from the gene pool of the herd by culling the animals carrying such recessives.

In the case of a linkage group of genes, instead of having a single gene B for coat colour and P for polled, there might be a number of genes controlling the same trait or similar traits. These could be:

  •     A1: The normal gene because it occurs most frequently, but this might not be the most desirable gene from a commercial point of view.
  •     A2: An undesirable gene but one that is fairly common.
  •     A3: An undesirable gene but one that is uncommon, as in the production of lethal bulldog calves.
  •   A4: A highly desirable gene but one that is very rare.

The breeder is after the combination of A4A4, which is the desirable gene in the homozygous and dominant state. If he can achieve this in his bull, he has got a bull that is prepotent for that trait. However, what the breeder often gets is A1A4, an animal that performs well, but does not breed true. The only way to fix a rare factor like A4A4 in a herd is by practicing in-breeding, and this is discussed in the next lecture.

Once a breeding programme has been started by using Scientific Selection based on the genotype of the animals, the breeder tries to improve his herd by altering the gene content and gene arrangement in individual animals.

2.    REGRESSION

In all breeding there is a tendency to move towards the breed average. This is a natural factor and the reasons for it were discussed when we were talking about Natural Selection in the last lecture. This pull is towards a balanced, heterozygous state, and in any breeding programme it is about 50%. In the case of a herd that has been practicing in-breeding or line-breeding, the pull will be towards the herd average rather than the breed average. This pull towards breed average is the reason why the average milk yields for dairy breeds has risen so little over the years. In 1950, the breed average for Friesland cows in U.K was 950 gallons (4 300kg) per cow, and today it is 1000 gallons (4 600kg). This pull towards breed average is used to calculate the breeding value, or transmitting value as it is called, of a dairy bull, and it is done in the following way:

Milk Average of the Sire’s Dam5 500kg
Milk Average of the Dam4 500kg
Breed Average3 500kg
Transmitting Value = Sire’s Average + Breed Average + Dams’s Average 4
=5 500 + 3 500 + 4 500 + 3 500 4
=17 000 4
=4 250kg

This means that the dairy bull from the above parents would produce cows that would average     425kg of milk per lactation. This figure is below that of both parents and is taking into account the pull towards breed average. This pull takes place for both milk yield and for type of cow, and it will work both upwards and downwards. The milk yields of very good cows will be pulled down by random mating within the breed; and the milk yields of very poor cows will be pulled upwards, towards the breed average, also by random mating within the breed. The breeder is always looking for an exceptional combination of genes in an animal to overcome this pull towards the breed average.

3.    SELECTION USING FAMILY GROUPS

    In any selection programme for milk production, two factors have to be considered at all times:

  •     Milk production has a low heritability and 80% of the yield of a cow comes from that cow’s management. Therefore, it is no use embarking on a breeding programme unless herd management is very good, otherwise the breeder cannot know how much of an improvement in the performance of the herd is due to better management and how much is due to genetic improvement. The improvement in milk yield of a herd due to improved breeding can be measured only if the cows are already producing to theirabsolute genetic potential; feeding and general management must be absolutely first class; and
  •     The breeder is not dealing with simple, single genes, but with linkage groups of genes. Milk production depends on a large number of factors all of them interrelated and controlled by many gene groups. Selection has to be made from the cow families in the herd where these desirablegroups appear, rather from single individuals in the herd.

The easiest way to consider family groups is to look at animals that produce litters of young rather than single offspring – pigs, rabbits etc. In any one litter of pigs the following happens:

The whole litter is average for the breed for both appearance and performance.

However, in a good herd of pigs, you could get the following situation:

Figure 1: Good Pig Herd

The whole litter is above average for the breed for both appearance and performance.

With pigs it is an easy matter to assess a litter from their appearance, birth weights, live mass gain etc. With cattle this is not possible because a cow will produce only one calf at a time. The family group has to be assessed by using all the full and half-sisters in the family. Although, in any series of mating between the sameparents, both parents contribute 50% of the genes to the offspring, it is not the same 50% except in the case of identical twins.

When assessing an individual for breeding purposes, try to find out the performance of all of its relations i.e. the family group. Knowledge of the whole family is the most reliable way of measuring variations, but it must be knowledge of the whole family, worst member of the family as well as the best. You are looking for the family in which the best member is very good and the worst one is good, the whole family adding up to a performance which is above the breed average.

Looking at the best in a family without knowing the worst means that you cannot know if the whole family is breed-average or above breed-average, or even below breed average. If there is no consistency throughout the family group, no individual is likely to breed consistently.

If an individual being considered is the best of the family, it is an exception and will breed to the family average.

When estimating the effects of regression-and the pull towards the average, use a figure halfway between the family average and the breed average.

ESTIMATING THE TRANSMITTING VALUE OF A BULL FROM A FAMILY GROUP

Find the Family Average milk yield and, if possible, adjust this for management factors. This could mean adjusting yield figures to allow for the effects of drought on one year’s lactations, or the effects of disease on an individual lactation.

Example
Milk Average of the Sire’s Dam        5500kg
Milk Average of the Dam                  4500kg
Breed Average                                     3500kg
Family Average                                    4100kg
Transmitting Value = Family Average + Breed Average +Sire’s Dams’s Average + Dams Average 4
=5 500 + 4 500 + 3 500 + 4 100 4
=17 600 4
=4 400kg

This figure is below that of both parents, but it is above the Family Average and well above the Breed Average.

Finally, as a summary, when a farmer is trying to assess the value of ayoung bull for use in a dairy herd, he should find out the following information:

  •     The performance of the sire’s dam and the dam of the bull being assessed, and their position in the family groups;
  •     The yields, quality and breeding performance of the grand-dams and great grand-dams. It is enough to go back 3 generations, because if there is no evidence of any improvement over this period, this means that the animal being assessed is of no use for breeding purposes; and
  •     Evidence that in successive generations there have been consistent improvements in milk yields.

This information can be obtained only in herds where careful and accurate milk records are kept, but unless performance records are kept, there is no point in selecting animals for breeding from that herd. Another fact that makes life difficult for the breeder is that in most herds the family groups or the parents are quite small. A difficulty which arises when buying a young bull from another breeder is that few breeders will discuss the poorest performers in their family groups; the buyer hears all about the good news but nothing about the bad news.

The best way of assessing a mature bull for use in a dairy herd is by assessing the records of his daughter’s performance; this is Progeny Testing the bull.