2 Assessing the tools

Transcription

1 2 Assessing the tools Successful breeding programmes do not require a detailed knowledge of animal genetics. A broad understanding of the way genetic merit is assessed and expressed in modern dairy breeding, however, is essential for anyone wishing to utilise the large amount of information available as reliably and cost-effectively as possible. It allows breeding programmes to be better planned; sire selection and semen buying to be better managed; and, mating decisions to be better organised and controlled. What s in this section? Understanding the breeding tools available Establishing their value and limitations Appreciating how they can be used for sustained improvement. 2 Assessing the tools More than anything else, understanding the contents of the genetic toolbox helps all concerned avoid the pitfalls of misunderstanding that can be so costly in breeding decision-making. Contents Summary Page 2:2 Genetic evaluations Page 2:3 Production traits page Page 2:9 Type traits page Page 2:10 Management traits page Page 2:11 Heritability and correlations page Page 2:17 Genetic defects page Page 2:20 Inbreeding and outcrossing page Page 2:22 Selection indexes page Page 2:24 Improving through breeding 2:1

2 A summary of the section For any trait, only a bull with a higher genetic merit than the cow to which it is mated will, on average, produce progeny of a superior genetic merit to that cow Genetic merit ratings will always be different from an animal s performance, because an animal receives half its genes at random from each parent and performance is markedly affected by differences in feeding and management Providing a sufficient number of daughters are recorded in a sufficient number of herds, progeny testing produces accurate predictions of a bull s genetic merit The lower the reliability of a genetic merit rating, the more likely it is to change over time as more information becomes available Production, type and management trait ratings allow animals to be ranked to establish the best individuals from which to produce the next generation for the greatest progress Avoiding matings between carrier parents or only using non-carrier bulls will prevent recessive genetic defects appearing As a rule of thumb, commercial producers should avoid matings leading to an inbreeding level of greater than 6.25% Used with an awareness of their limitations, PIN and PLI selection indexes can be a very useful way of screening animals for improvement value. The science of genomics has now entered the breeding arena; genomically tested sires are more reliable than young unproven sires but they are still a long way off replacing the high reliability proven sire See also... Section 3: Section 4: Section 6: Section 7: Section 8: Section 9: Section 10: Planning your approach Selecting your sires Breeding your cows and heifers Making the most of specialist strategies Valuing modern breeding techniques Factsheet 1: Dairy breeding definitions Factsheet 2: MACE Factsheet 3: Linear type assessment Factsheet 4: Dairy trait correlations Factsheet 5: Inbreeding level assessment Worksheet 1: Inbreeding level calculator 2:2 Improving through breeding

3 Genetic evaluations Genetic evaluations estimate how much of a cow s performance or appearance is genetic in origin and so possible to influence by breeding rather than due to differences in environment, feeding or management. They also estimate the extent to which cow characteristics are transmitted (without being expressed) by the AI bulls on which so much of dairy progress through breeding depends. Genetic evaluations in the UK are undertaken by DairyCo, in association with the milk recording organisations and the dairy breed societies, through a number of increasingly sophisticated computer programs. These evaluations compare large amounts of information from bulls and cows to remove the complicating effects of different environments, including age at calving, month of calving, feeding and management regime. For production traits (which involve evaluations over several lactations) adjustments are also made for lactation number and days in milk. For type (where measurements are only made in the heifer lactation), adjustments are made for stage of lactation. The accuracy (or reliability) with which genetic evaluations predict the genetic value or merit of dairy cows and bulls depends on both the amount and quality of the information used to produce them. As the science of genomics develops, genetic evaluations will be significantly improved by the increasing availability of direct measurements of the genetic make-up of individual animals (Section 8). Factsheet 1 defines the key terms used in UK dairy breeding. Understanding genetic merit Predicted Transmitting Abilities (PTAs) and Breeding Values (BVs) are the basic units of breeding currency used in all evaluations. Both are estimates of genetic merit, calculated from an animal s own performance and appearance (in the case of cows) and information from parents, progeny and all known relatives. A PTA indicates the amount of a particular trait an animal is, on average, predicted to pass on to its offspring. A Breeding Value indicates the overall genetic value an animal is, on average, predicted to have for a particular trait. Since each individual transmits a random sample of half its genes to its progeny, the PTA of an animal for a particular trait is half its Breeding Value. As a result, predictions of every individual s genetic merit continually change, becoming increasingly more reliable and less likely to vary as more and better information becomes available. For any trait, only a bull with a higher genetic merit than the cow to which it is mated will, on average, produce progeny of a superior genetic merit to that cow (Figure 2.1). Improving through breeding 2:3

4 Figure 2.1: Breeding values and PTAs Mating 1 Mating 2 Breeding Value PTA Breeding Value PTA = = = = Superior to her mother Inferior to her mother Table 2.1: Current genetic base (PTA2010) heifer equivalent production values of cows born 2005 by breed Breed Milkkg/yr Fat kg/yr Protein kg/yr Fat% Protein% Holstein Shorthorn Ayrshire Jersey Guernsey Friesian Montbeliarde Brown Swiss All PTAs and BVs are expressed as deviations from a fixed Genetic Base, up-dated every five years to keep the values meaningful to the current population in the face of genetic progress. The current zero point for production in the UK for most breeds is PTA 2010, the average of cows born in 2005 (Table 2.1). Because of their different bases, genetic values must never be compared between breeds or countries. Nor should current values ever be compared with those from a previous base period. 2:4 Improving through breeding

5 Genetic merit ratings will always be different from an animal s actual performance because: It receives a random sample of half its genes from each parent Performance is markedly affected by differences in environment, feeding and management. Understanding progeny testing Progeny testing schemes are the primary way of predicting the genetic merit of AI bulls. They involve independent assessments of the production, type and management traits of the daughters of young bulls against those of other bulls in milk-recorded herds both pedigree and nonpedigree. Production information is collected from these herds as part of their normal milk recordings with type and other data collected by breed society classifiers. Type evaluations are breed specific with each breed assessed in comparison to a true type model of that breed. All the available information is analysed by DairyCo Breeding+ to produce official UK bull proofs which predict how much of the recorded differences between individuals come from genetics rather than feeding, management and other environmental influences (Section 4). Multiple Ovulation and Embryo Transfer (MOET) schemes with genetic predictions based on the performance of sisters rather than daughters were originally felt to offer major advantages over progeny testing in increasing rates of genetic progress. However, practical experience has led most MOET schemes to be integrated into progeny testing with nucleus herds used to generate potentially higher merit young bulls for evaluation through their daughter performance. Progeny testing schemes involve a continual process of young bull breeding, test insemination, daughter recording and information analysis to produce proven sires which then join AI studs for widespread semen marketing (Figure 2.2). Figure 2.2: Typical dairy progeny testing process Promising young bulls primarily bred through contract matings or in nucleus herds Semen from young bulls inseminated into cows in milk recorded herds Daughters reared and brought into milk recorded herds as replacements Young bulls temporarily retired to lay-off units to await progeny test results Production, type and other records collected from daughters and herd mates PTAs and Breeding Values calculated from initial progeny test results Highest rating young bulls returned to studs as proven sires for widespread AI use All other young bulls culled Proven sire PTAs and Breeding Values updated regularly with records from more daughters in more herds, progressively increasing proof reliability Improving through breeding 2:5

6 Typically less than 10% of bulls that enter progeny testing are returned to stud as proven AI sires. Since the 1970s, UK progeny testing has been expanded considerably to a current level of around 150 Holstein Friesian bulls/year (Cogent and Genus ABS each test about 60 per year, with other companies accounting for roughly another 30). The global nature of the AI industry, however, means UK studs have access to a progeny test base of more than 4000 Holstein Friesian bulls each year from all the major dairying nations. Providing a sufficient number of daughters are recorded in a sufficient number of herds, progeny testing produces accurate predictions of a bull s genetic merit. In spite of an evaluation system designed to account for bulls with daughters in a limited number of herds, a good spread of herds remains essential to a reliable bull proof. While globalisation has also been valuable in giving dairymen using other breeds greater access to proven sires, the large population of milk-recorded cows required for reliable progeny testing seriously restricts the supply of proven bulls in many cases. Herds based on breeds with small gene pools may need to adopt modified breeding strategies to make the most of their more restricted genetic resources (Section 7). Although still in its infancy the science of genomics is looking like it may significantly alter sire testing and may, longer term, depending on future results, replace progeny testing (Section 8). Understanding international evaluations Multiple-trait Across Country Evaluations (MACE), undertaken at the INTERBULL Centre in Sweden, allow the most accurate predictions of genetic merit to be published in UK values for all bulls, regardless of where in the world they were progeny-tested. MACE takes account of all known between-country correlations to combine bull proofs from other countries on their own systems with any domestic information, producing a single proof for each animal in each country. Information from bulls proven in many countries is, thus, combined into a single proof for the UK, incorporating every recorded daughter (Section 4). Bulls with a first proof in a single exporting country will have a proof that includes the daughter information in the exporting country, combined with parental information from any UK evaluations. This enables better evaluation of overseas bulls with limited proofs but a large amount of UK-evaluated family information. For production evaluations MACE works extremely well, as all the traits are measured in a similar way across countries. Although great efforts have been made in recent years to harmonise the linear traits measured by each country, some significant differences still occur. In all cases, MACE evaluations are better than the conversion system they replaced. Even so, UK progeny-tests involving UK cows managed under UK conditions are still considered the best predictor of a bull s merit for the UK. Factsheet 2 gives details of the Multiple-trait Across County Evaluation (MACE) system. 2:6 Improving through breeding

7 Appreciating reliability estimates In its simplest form, a Predicted Transmitting Ability (PTA) or Breeding Value (BV) for any trait is the average genetic merit of an animal s sire and dam, known as the Pedigree Index. As information on the animal itself (if a cow), its sisters and then its progeny increases, progressively less weight is placed on its pedigree in the calculation. The greater the proportion of information derived from the progeny and the broader the base of that information (from more daughters milking in more herds) the more reliable it is as an estimate of an animal s true genetic merit. A spread of daughters across a large number of herds is particularly important in this respect, creating stability from the figures as it shows the ability of a bull to work across a range of farm types and systems. Reliabilities are published for all genetic values as a guide to the degree of confidence that can be placed in them. Shown as a percentage, reliabilities commonly vary from 10% to 99% (Table 2.2) and are particularly important in evaluating sire proofs (Section 4). Table 2.2: Typical production PTA reliabilities Reliability Comment Information contributing to PTA 10-29% Extremely low Some Pedigree Indexes involving estimates from relatives % Very low 41-55% Low 56-65% Low to moderate 66-75% Moderate Most Pedigree Indexes involving estimates from parents with reasonable reliabilities. Some Pedigree Indexes. Usually the sire is well proven and the dam is very high reliability. Cows with one or two lactations. An early genomic PTA may have a reliability of 40-60%. Bulls with officially published PTAs (minimum 50%). Cows with three lactations. General maximum for most cows (3+ lactations). May be exceeded if a cow has many ET daughters in the UK % Moderate to high Bulls with an initial progeny test through AI % High Proven bulls with a large number of daughters from a wide cross section of herds. 99% Very high Widely proven and used AI bulls. The lower the reliability of a PTA or BV the more likely it is to change over time as more information becomes available (Table 2.3). Improving through breeding 2:7

8 Table 2.3: Variation in true genetic value for production at different reliabilities Reliability (%) Milk (kg) Fat (kg) Protein (kg) 30 +/ /- 17 +/ / /- 15 +/ / /- 11 +/ / /- 7 +/ / /- 5 +/ /- 66 +/- 2 +/- 2 The true genetic value for production of a bull with a PTA of 500kg at 50% reliability could be as low as 30kg or as high as 970kg. At 99% reliability the range in which the true PTA almost certainly lies narrows to between 434kg and 566kg. Up-to-date reliability estimates for the different traits should always be taken into account when interpreting genetic evaluations. Do not purchase a bull without asking for and looking at his reliability. Also make sure that reliability is a UK figure, not from the country of origin, if he does not have UK daughters, as this will be significantly higher. MACE evaluations adjust reliabilities from the country of origin to the UK, illustrating that a bulls figures will be more accurate if tested under UK systems. A UK MACE reliability on a foreign tested sire will be lower than his country of origin reliability. 2:8 Improving through breeding

9 Production traits Predicted Transmitting Abilities (PTA) for key production traits are widely available for both bulls and cows of all the main UK dairy breeds. Production PTAs allow animals to be ranked within breed to establish the best individuals from which to produce the next generation for the greatest progress in the chosen trait. PTAs are produced for five main production traits (Table 2.4): Weight of milk Weight of fat Weight of protein Fat percentage Protein percentage. They are derived from official milk records adjusted to account for the effects of lactation number, days in milk, age, calving interval and month of calving. The Test Day Model (TDM) is now used for all UK production trait evaluations. This method increases the accuracy with which PTAs are calculated from daughter records, especially for young bulls and cows with lactations in progress. As well as more effectively removing non-genetic effects by applying all adjustments to each individual milk recording instead of the completed lactation, the TDM system: Accounts for animals with non-standard lactation curves, appreciating that all cows do not perform to a formula Copes better with flexible milk recording regimes Enables bull proofs for lactation persistency to be calculated. Table 2.4: A typical production PTA 384kg milk 24.1kg fat +0.13% fat 17.4kg protein +0.07% protein Reliability 78% This bull s daughters are expected to give, on average and in their first lactation, 384kg more milk, 24.1kg more fat and 17.4kg more protein than cows whose sire has a PTA of 0kg milk, 0kg fat and 0kg protein. The reliability figure is a reflection of the number of daughters that have contributed to the bull s proof and their distribution across herds. When published for bulls, reliability ranges from 50% to 99%. The higher the reliability, the more likely is the proof to be an accurate reflection of the bull s true transmitting ability. It is therefore recommended that bulls of low reliability are not used on a widespread basis within any one herd. As with all genetic indexes, it is important not to directly compare PTAs from more than one country. These should be expressed on a UK scale where they can be easily compared. Improving through breeding 2:9

10 Type traits Some 17 separate linear traits assessed on a 1-9 scale are used to describe the degree of type shown by cows and heifers of the main dairy breeds (Section 3). Figure 2.3: Type trait breeding value distribution At the same time, most breeds classify females against an ideal standard for so-called composite type traits and overall type score (Section 3). Heifer type assessments and classifications from progeny testing are used to produce a range of type Breeding Values in the proofs of bulls of the main UK dairy breeds (Section 4). Type Breeding Values allow animals to be ranked within breed to establish the best individuals from which to produce the next generation for the greatest progress in the chosen trait. The values express the degree of extremity of an individual in the population, not what the daughters actually look like. Both linear and composite type Breeding Values are presented in a standardised form on a scale of 3 to +3, calculated by a statistical technique (Figure 2.3) is the breed s average; 68% of animals fall within the range 1 to +1 95% of animals fall within the range 2 to +2 99% of animals fall within the range 3 to +3 For any trait, bulls with type Breeding Values of less than 2 or more than +2 are within the most extreme 5% of the breed for the trait 2.5% at either end. Equally those rated as less than 3 and more than +3 are in the most extreme 1% of the breed 0.5% at either end. Higher Breeding Values are always preferable in Composite Traits, although by their nature these are more subjective than individual linear traits (Sections 3 and 4). Because Linear Traits relate to appearance rather than desirability, in some cases mid-scoring Breeding Values are preferable (Sections 3 and 4). Factsheet 3 provides a practical guide to linear type assessment. 2:10 Improving through breeding

11 Management traits A variety of additional traits, neither strictly production nor type-based, are measured in UK progeny testing and published in bull proofs to assist herd improvement efforts (Section 4). The most significant of these are: Somatic Cell Counts (SCC) Locomotion Lifespan Fertility Temperament Ease of milking Body condition score Calving ease. Many of the management traits are low heritability and reliability. With this in mind the information they provide should not be used in isolation in bull selection, unless they are very extreme (which may render a bull unmarketable). However, they should be given some consideration in building up an overall picture of a bull s strengths and weaknesses. Interpreting SCC evaluations Since SCCs are a good indicator of Udder Health, Somatic Cell Count PTAs offer the opportunity to breed animals with an inherently greater resistance to udder infections. PTAs for Somatic Cell Count are calculated for all progeny-tested bulls marketed in the UK. Derived from official milk records, the PTAs are expressed as a percentage, the overall range being from +30% to 30%. In practice, 95% of all bulls have been found to lie in the range +/- 10%, with every 1% change in PTA predicting a 1% change in cow SCC. Unlike most other traits, negative SCC % PTAs are desirable because they indicate a potential reduction in mastitis. On average, daughters of a bull with a SCC % PTA of 10 will have a 20% lower Somatic Cell Count than those of a bull with a SCC % PTA of +10. Somatic Cell Count PTAs tend to be lower reliability than equivalent Production PTAs. Improving through breeding 2:11

12 Interpreting locomotion evaluations The heritability of SCC is lower than that of production traits, indicating that environmental factors can have a large effect. Alongside improved breeding for SCC, attention to farm management is still required. However, as with any trait with lower heritabilities, genetic change is an effective long term solution to helping to improve these traits. Due to the lower heritability, the reliability of bull PTAs tend to be somewhat lower. Caution is therefore required in interpreting the SCC PTAs of bulls, since they could changes with more daughter information. Linear traits for feet and legs are valuable in assessing the form of these vital aspects of a cow s constitution. However, they provide little, if any, information about an animal s ability to walk. In contrast, locomotion scoring on a scale of 1-9 (where 1 is acutely lame and 9 is a perfect gait in which the back feet are placed into the marks left by the front ones), assesses soundness in all the components needed to walk properly. Locomotion should not be confused with mobility which is not a genetic trait but a management assessment (now being requested by some milk buyers and milk contracts). The importance of locomotion as an indicator of lameness is underlined in its inclusion in evaluations as a predictor of Lifespan. All UK-proven Holstein Friesian bulls have a Breeding Value for locomotion, expressed on the same 3 to +3 statistical basis as type traits (Section 4). Locomotion values are unfortunately not available for bulls proven overseas and there is no conversion or composite formula to calculate locomotion for foreign tested sires. The UK was the first country in the world to introduce locomotion scores and again this illustrates the benefits of using a sire whose daughters have been tested under UK conditions. 2:12 Improving through breeding

13 The reliability of a locomotion proof will always be significantly lower than a production proof from the same number of daughters, so locomotion proofs from newly-proven bulls should be viewed with due care. Interpreting lifespan evaluations Lifespan PTAs are produced for the majority of progeny-tested bulls marketed in the UK, allowing positive selection for longer-living cows (Section 4). For UK-proven sires, the PTA calculations use a combination of the type traits shown by research to be most strongly associated with a longer herd life, namely: Feet and Legs composite Mammary composite Fore Udder Attachment Locomotion Somatic Cell Count. Most importantly they also involve actual daughter survival information from completed lactations. Lifespan PTAs for Holstein Friesian bulls tested overseas with no recorded UK milking daughters receive UK equivalent PTAs based on their home country genetic evaluation. This conversion is performed through the previously described MACE evaluation. The overall range of Lifespan PTAs is from +1.0 lactation to 1.0 lactation, with the vast majority of bulls lying within the range +/- 0.5 lactation. On average, daughters of a bull with a Lifespan PTA of +0.5 lactation will last one lactation more than those of a bull with a Lifespan PTA of 0.5 lactation. Lifespan PTAs tend to be of much lower reliability than equivalent Production PTAs. In view of their lower reliabilities, particular caution is required in interpreting the Lifespan PTAs of most bulls, since they could change markedly with more information. As more information becomes available on actual daughter survival, the weight given to this in the PTA calculation increases, while that given to type information decreases, substantially increasing the reliability of the proof. Improving through breeding 2:13

14 Interpreting fertility evaluations Although levels of herd fertility are more dependent on management than breeding, a variety of measurable traits have been combined into a UK Fertility Index (FI) for use in selecting bulls as part of a determined effort to improve fertility. The index comprises a total of six separate traits: The Calving Interval The number of days in milk at the first recorded insemination The number of inseminations required to get a cow in calf The Non-Return Rate at 56 days Yield at 110 days in milk The Body Condition Score. Based on the above traits, plus information gathered on family members, this index provides a practical tool for selection on daughter fertility. The Fertility Index is published as a financial figure. The higher the proof the more financial benefit results from improved fertility. A bull with an above average Fertility Index will breed cows with improved calving intervals and better non-return levels. As a rough guideline, every point increase in Fertility Index, for example from 3 to 2 will decrease calving interval by half a day and improve non-return rates by 0.5%. The range in Fertility Index is +15 to 15. The low heritability of the components of the Fertility Index means it should be used with caution in selecting bulls, especially young animals with relatively few daughters in their proofs which will have very low reliabilities for the character. Interpreting temperament evaluations The key to a quiet and contented herd is undoubtedly sensitive handling. Nevertheless, anyone who has spent time in the parlour will recognise a clear genetic component to temperament in cows. Holstein Friesians are scored for temperament on a 1-9 scale (from 1 being nervous to 9 being very quiet) in the opinion of those who manage them and Breeding Values calculated on the same 3 to +3 statistical basis as type traits (Section 4). Temperament is a more subjective trait than most (people s perceptions and tolerance levels vary greatly) and evaluations are of a lower inherent reliability. Although unlikely to be a major breeding consideration for most herds, temperament ratings can form a useful guide to bulls likely to produce the most manageable animals and conversely bulls with extremely low (negative) scores for temperament may well be avoided. They may be particularly valuable in herds with temperament problems. The characteristic is likely to become increasingly sought-after with an increasing herd size and decreasing availability of skilled labour. 2:14 Improving through breeding

15 Interpreting ease of milking evaluations Like temperament, the ability to milk-out rapidly and efficiently has long been considered a particular virtue, especially in high-yielding animals. However, cows that milk too easily can run-out large amounts of milk before milking and tend to be more prone to both udder infections and teat-end damage through over-milking. Ease of milking is recorded on a 1-9 scale (from 1 very hard to 9 runs milk) and Breeding Values calculated on the same 3 to +3 statistical basis as type traits (Section 4). As with temperament, ease of milking records are based on the subjective opinion of those who manage the animals, so care is essential in using them in bull selection. Although unlikely to be a major breeding consideration for most herds, ease of milking ratings can form a useful guide to bulls likely to produce the most manageable animals. They may be particularly valuable in herds with ease or speed of milking problems and also in herds where animal throughput is vital, such as farms with robotic milking systems. Like temperament, ease of milking is likely to become increasingly sought-after with an increasing herd size and decreasing availability of skilled labour. Improving through breeding 2:15

16 Interpreting calving ease evaluations Independent calving ease indexes are now being produced in the UK. Their introduction follows an industry-wide collaboration and the analysis of some 400,000 calving records. The information has been gathered by British farmers as either part of their regular milk recording or as part of their participation in progeny testing. The indexes are expressed on a scale of 4 to +4 around a breed average of zero, with positive figures indicating that calvings are predicted to be easier than average and negative figures predicting more difficult calvings. There are two calving ease measurements; Direct Calving Ease (dce) gives a prediction of the ease with which a calf by that sire will be born and Maternal Calving Ease (mce) predicts the ease with which a daughter of that sire will give birth. Direct Calving Ease is likely to be of most interest and is very important when deciding what bull to use on maiden heifers. Calving ease, however, should not be ignored in older cow matings either. Again, the advice would be to avoid extremes, give bulls which are likely to produce very difficult calvings a wide berth. Calving ease ratings have now been introduced with industry standardisation, they cover both direct calving ease and maternal calving ease. They should be considered most when selecting sires for maiden heifers. However, bulls with extreme negative figures are prone to harder calvings and may well be avoided for use on older cows too. Calving ease should also be a consideration when carrying out ET work, flushing to a hard calving sire and implanting those embryos into maiden heifer recipients can have serious implications. There are also considerations with calving ease and sexed semen. By their nature bull calves are bigger, so sexed semen should render easier calvings with heifers being born. This may facilitate the use of bulls with a lower dce on maiden heifers if using sexed semen. Attention should also be paid to Maternal Calving Ease, as long-term selection for dce without any regard to mce could set up problems for the future. Farmers tend to associate easy calvings with smaller calves but those smaller calves sometimes go on to have difficult calvings themselves, creating a vicious circle with escalating problems and obvious cost implications. 2:16 Improving through breeding

17 Heritability and correlations While individual animals may be clearly superior to the average for a particular trait, the value of this depends on the extent to which the superiority is transmitted to their progeny. This fundamentally depends upon the trait s heritability, which indicates the relative importance of heredity compared to environmental influences in the physical expression of a trait. Evaluating heritability estimates Heritability figures for the various different dairy traits in the UK are published as percentages (ranging from 0% to 100%) or decimals (0.0 to 1.0). Calculated from extensive national data, they vary widely between different traits (Table 2.5). Provided there is sufficient variation in the population, a trait with a high heritability implies a greater degree and speed of improvement progress is possible through breeding. Low heritability traits, on the other hand, are more affected by the environment and may, consequently, be more easily improved through management. Breeding is also complicated by the fact that, rather than working in isolation, there are many genes controlling individual traits and they interact with one another either positively or negatively. Wherever traits are correlated in some way, selection for one will result in a change in the other, the extent and direction of the change will depend on the degree and value of the correlation. This will either help or hinder the breeding process, depending upon the combination of characteristics sought. Improving through breeding 2:17

18 Table 2.5: Current UK dairy trait heritability estimates Trait Heritability Production Milk Yield 55% (0.5) Fat Yield 47% (0.47) Protein Yield 51% (0.51) Fat Percentage 68% (0.68) Protein Percentage 68% (0.68) Linear type Stature (ST) 41% (0.41) Chest Width (CW) 25% (0.25) Body Depth (BD) 33% (0.33) Angularity (ANG) 34% (0.34) Rump Angle (RA) 30% (0.30) Rump Width (RW) 26% (0.26) Rear Leg Side (RLS) 20% (0.20) Foot Angle (FA) 10% (0.10) Fore Udder Attachment (FUA) 22% (0.22) Rear Udder Height (RUH) 23% (0.23) Udder Support (US) 19% (0.19) Udder Depth (UD) 35% (0.35) Teat Placement Rear (TPR) 29% (0.29) Teat Placement Side (TPS) 29% (0.29) Teat Length (TL) 29% (0.29) Composite type Legs and Feet 16% (0.16) Mammary 27% (0.27) Type Merit/Type Score 32% (0.32) Management Somatic Cell Count 11% (0.11) Locomotion 10% (0.10) Lifespan 6% (0.06) Fertility 3% (0.03) Temperament 11% (0.11) Ease of Milking 21% (0.21) Body Condition Score 27% (0.27) Direct Calving Ease 7% (0.07) Maternal Calving Ease 4% (0.04) High heritability traits (such as yield, fat and protein percentages) are relatively easy to improve through breeding, as are a number of linear and type traits. Those with a lower heritability (such as SCC, Locomotion, Lifespan and Fertility Index), require far more emphasis to achieve the same degree of progress. Lower heritability traits also need information from a larger number of daughters in a wider range of herds to achieve the same degree of reliability in bull proofs (Section 4). 2:18 Improving through breeding

19 Evaluating genetic correlations The genetic correlation between traits always lies between 1.0 ( 100%) and +1.0 (+100%). A reasonably strong negative genetic correlation ( 0.30 or greater) between two traits means that breeding for one on its own will lead to a reasonably strong negative effect on the other. Equally, a reasonably strong positive correlation (+0.30 or greater) between two traits means that breeding for one on its own will lead to a reasonably strong positive effect on the other. Where the correlation between two traits is fairly weak (between 0.10 and +0.10), breeding for one will have little or no effect on the other. Dairy breeders need to be aware of a number of particularly important genetic correlations most of which are entirely logical in the light of experience (Table 2.6). Table 2.6: Particularly important genetic correlations in dairy breeding Traits Correlation Consequences Milk yields and solid percentages Milk yields and angularity Protein % and fat % Direct CE and maternal calving ease 0.37 to to Breeding for yield alone will lead to serious reductions in fat and protein percentages Breeding for yield will also automatically increase angularity Breeding for one component will improve the other Breeding for direct calving ease can lead to a reduction in maternal calving ease Increased stature and angularity are further considered by many to have a negative relationship with fertility, health and longevity. Factsheet 4 summarises the published correlations between the main dairy traits. Improving through breeding 2:19

20 Genetic defects A number of genetic defects are known to cause problems in dairy breeding. Most prominent among these are: Bovine Leukocyte Adhesion Deficiency (BLAD) causing a fatal deficiency of the immune system which cannot be overcome even by high levels of antibiotics Complex Vertebral Malformation (CVM) causing stillbirths or, more commonly, abortion or foetal re-absorption before 260 days gestation with serious effects on herd fertility Mule Foot (MF) causing the two claws of the hoof to become fused, most commonly seen in the front feet Brachyspina causing embryonic death, still births and other deformaties. Because these defects are caused by recessive genes, masked by the presence of dominant partners, an animal must inherit a copy of the gene for the character from each parent to be affected. Equally, since animals receive a random selection of half their genes from each parent, matings between stock carrying the gene only result in the defect in one in every four progeny, on average (Figure 2.4) Figure 2.4: Genetic defect expression from a recessive gene Parents Sire Carrier (Aa)* Dam Carrier (Aa)* Genes Progeny Defect A a A a AA Aa Aa aa No No No Yes Genetic testing allows carrier animals to be identified but the expense involved generally only makes this worthwhile for AI bulls and ET dams. The genetic status of Holstein Friesian animals for the main conditions is indicated by suffixes after their names (Table 2.7). Table 2.7: International coding for main breed genetic defects Defect Carrier code (WHFF) Tested-free code (WHFF) BLAD *BLC *TL (BLF) CVM *CVC *TV (CVF) Mule Foot *MFC *TM (MFF) Brachyspina *BYC *BYF Only a tested-free coding indicates the animal is definitely not a carrier; no coding simply means no notification of a test result has been received so the animal may or may not be carrying the defect. From April 2009 the WHFF (World Holstein Friesian Federation) has attempted to bring some standardisation to genetic codes and with the aim of advancing breeding want all information, both desirable and undesirable, made available to the interested parties. The aim is to generate a simple, easy to use and interpret code, which is based on an expression code of F = tested and non carrier, C = carrier. Other codes cover such genetic defects as DUMPS, Factor XI and Citrullinaemia, although these are less common. If any abbreviations appear after a bulls name which you are unsure about ask the semen representative for a full explanation. A: Dominant gene for character a: Recessive gene for character *: carrier of recessive gene 2:20 Improving through breeding

21 Coat colour can also be identified by a genetic test. Suffixes associated with coat colour include: Coat colour carrier gene Red Black/red Varriant red Black Description Red gene Black/red gene Variant red gene Black gene Gene and expression code RDC = carrier of red gene RDF = tested noncarrier of red gene BRC = carrier of black/red gene VRC = carrier of variant red gene BKC = carrier of black gene Breeding animals with red coat colour genes has become popular in recent years. However there is need for caution, breeding decisions should not be based on coat colour at the expense of other important traits, as this can be detrimental to overall profitability, especially in non-pedigree situations where there is no scope for the added revenue of marketing red and white genetics. Avoiding matings between carrier parents will prevent any risk of expressing a recessive genetic defect. Only using bulls tested and confirmed free of the gene will prevent any risk of the defect entering the herd. Improving through breeding 2:21

22 Inbreeding and outcrossing Artificial Insemination (AI) on the one hand and, to a far lesser extent, Embryo Transfer (ET) on the other, are valuable tools in speeding-up genetic improvement by concentrating breeding on only the very best bulls and cows. AI, in particular, also inevitably concentrates the breeding pool around fewer and fewer bloodlines, raising the level of inbreeding. In exactly the opposite way to inbreeding, mating two animals with different ancestries extends the range of genes in their progeny. A single out-cross mating with an unrelated animal will remove all inbreeding in the resulting progeny. Assessing inbreeding A degree of inbreeding always occurs when animals related to one another are mated as a result of a concentration of the same genes (Table 2.8). Table 2.9: Inbreeding levels Mating Inbreeding % Sire/Daughter 25% Full brother/full sister 25% Half brother/half sister 12.5% Grandsire/Grand-daughter 12.5% Grandson/Grand-dam 12.5% Uncle/Niece 6.25% Son/Grand-daughter 6.25% Daughter/Grandson 6.25% Full cousins 6.25% Grandson/Grand-daughter 3.13% Half cousins 3.13% A certain degree of inbreeding has historically been considered desirable in pedigree breeding to fix a trait so it is consistently expressed down the generations. This has been identified as linebreeding. Too much concentration of the genetic make-up, however, is detrimental, leading to significant losses in production, fertility, longevity and general vigour. Production losses per 1% inbreeding are calculated to be: 15kg milk yield 0.6kg fat yield 0.5kg protein yield Inbreeding also makes genetic defects more likely to occur, as there is an increased chance of undesirable recessive genes coming together. Rule of thumb Commercial producers should avoid matings leading to an inbreeding level of greater than 6.25%. Globalisation of breeding on the one hand and small gene pools on the other make it difficult to avoid some level of inbreeding. It is important to appreciate that only ancestors common to both parents contribute to inbreeding. The fact that a bull or a cow are themselves inbred has no effect on the inbreeding level of their progeny unless they share common ancestors. The extent to which common ancestors contribute to inbreeding, of course, depends on their distance in the pedigree and is often far less than might be supposed at first sight. Reducing inbreeding is often a perceived benefit of computer mating programmes (Section 4-7). 2:22 Improving through breeding

23 Calculating inbreeding co-efficients Dairy farmers can readily access inbreeding coefficients for animals within their own herd by logging into the relevant section of the DairyCo Breeding+ website. The inbreeding coefficients are calculated after each proof run. As previously mentioned, the maximum level of inbreeding recommended by geneticists before animal health, fertility or productivity levels may be affected is 6.25%, therefore, any mating that would result in a coefficient over 6% should be avoided. It is also possible to calculate the inbreeding coefficients that would result from mating your cows to certain AI sires this can be done through the WebMate programme available to all Holstein UK members free of charge. This gives the breeder a valuable guide to the level of inbreeding that may occur as a result of any proposed mating. Inbreeding is prevented within the program to three generations but commonalities regularly occur beyond this. Again, it is important not to become obsessed with a single aspect of breeding but trends within the herd should be monitored as should extremes (those above 6%) to prevent the problem being compounded and the subsequent potential loss in profitability in future generations. The genetic list can be ranked according to inbreeding coefficients enabling a snap shot of the herds position to be easily seen. Assessing outcrossing Outcrossing can be beneficial in increasing production, fertility, longevity and vigour, in particular, as a result of hybrid vigour (or heterosis). It essentially involves crossing two completely unrelated parents. Taken to its extreme, outcrossing becomes the cross-breeding of totally different dairy populations. Such as Holstein Friesians and Jerseys, which has found such favour in New Zealand and it has become far more common in the UK over recent years. New Zealand experience suggests such crosses can show hybrid vigour for yield and fitness traits ranging from 5-20% which may be extremely valuable in commercial herds. Studies in other parts of the world are also highlighting the potential value of cross-breeding to improve longevity and fertility in come cases (Section 7). Cross-breeding could lead to a loss of uniformity and consistency, as an increasing variety of genetic characteristics are expressed. Factsheet 5 sets out a standard inbreeding level assessment. Worksheet 1 provides a pro forma for calculating inbreeding levels. Improving through breeding 2:23

24 Selection indexes Selection indexes have been developed to help UK producers achieve progress towards specific improvement goals by selecting for a balance of different traits. The Profit Index ( PIN) predicts the additional margin over food and quota costs per lactation a bull or cow is expected to pass on to its progeny on the basis of future milk market requirements. Developed from PIN and using the same milk, fat and protein weightings, the Profitable Lifetime Index ( PLI) incorporates additional Lifespan, Fertility, Somatic Cell Count, Locomotion and udder components to account for the extra value of longer lived cows with better genetics for health and welfare. The standard values used in the PIN and PLI calculations are likely to be sufficient for many producers looking for a simple initial screening of bulls and cows for replacement breeding. Evaluating PIN There is considerable debate over the value of index breeding within the dairy industry. The principal benefit of PIN undoubtedly lies in bringing together three production traits in a single, simple measure of genetic merit that allows bulls and cows to be shortlisted for herd replacement breeding relatively easily. The key limitation of PIN is that it only involves production traits and uses standardised economic weightings. While production improvement remains the Number One objective in most programmes, selecting on PIN alone has been shown to reduce longevity. The large number of different milk contracts and prices available, diversity of transport arrangements and charges and fluctuations in feed costs further mean the standard economic weights of the PIN formula are unlikely to be completely appropriate for any individual herd at any one time. Used with an awareness of its limitations, PIN can be a valuable tool, especially as an initial screen of an animal s value in improving production. PIN, however, should not be used as the sole breeding decider, as there is no consideration given to the importance of fitness traits. Evaluating PLI PLI further builds on the PIN index with the added advantage of an increased emphasis on fitness traits, achieved through the inclusion of Lifespan, Fertility, Somatic Cell Count, Locomotion and Udder Composite. PLI gives an increased emphasis on health and welfare traits. Around 45% of the index is now based on production, while the remaining 55% concentrates on traits associated with fitness. The updated PLI is expressed on a lifetime basis and indicates the extra margin a bull or cow is expected to pass on to its progeny per lifetime. PLI can be valuable tools, especially as initial screens of an animal s overall improvement value. Overreliance on PLI without adequate consideration of specific individual production and type traits can be dangerous. 2:24 Improving through breeding