Genetics and genomics of dairy cattle

C H A P T E R

6 Genetics and genomics of dairy cattle Francisco Pe~ nagaricano Department of Animal Sciences, University of Florida, Gainesville, FL, United States

O U T L I N E Introduction

101

The basics of genetic selection

102

Selection for traits that increase income

103

Selection for traits that reduce expenses

104

Selection for multiple traits

106

Genomic selection: the latest revolution

110

Effective use of genomics: sire selection

112

Introduction Genetic selection programs have transformed the dairy industry worldwide. The current dairy cow produces more than twice as much milk as the dairy cow of 50 years ago, and more than half of that improvement is due to genetic selection. The basic blocks for genetic improvement have been performance records and pedigree information. The development and widespread utilization of national milk recording systems, the introduction of artificial insemination, and

Animal Agriculture https://doi.org/10.1016/B978-0-12-817052-6.00006-9

Effective use of genomics: replacement heifer selection

113

Novel traits in the genomics era

114

Managing inbreeding and genetic diversity

116

Final remarks

117

References

118

the development of accurate genetic evaluation methods have enabled remarkable genetic improvement in dairy cattle populations. The success of these programs has been possible due to the close collaboration between dairy farmers, milk recording organizations, dairy records processing centers, breed associations, breeding companies, government agencies, and agricultural universities. Each organization has a key role in data collection and analysis, product development, education and outreach. It should be noted that the focus of selection

101

Copyright © 2020 Elsevier Inc. All rights reserved.

102

6. Genetics and genomics of dairy cattle

programs has evolved over time, from an initial emphasis on increasing milk yield and physical appearance, to a current interest in improving production efficiency, milk composition, and animal fitness. Recently, the advent of genomic selection has revolutionized dairy cattle breeding. This technology allows breeders to make accurate selection decisions at a much earlier age, reducing generation intervals, and thereby increasing the rate of genetic progress. In addition, genomic selection provides a mechanism for improving traits that were too difficult or expensive to measure in conventional progeny testing schemes, such as feed efficiency. This chapter provides an overview of different aspects involved in dairy cattle genetic improvement programs, including a concise description of economically relevant traits in dairy cattle breeding, a brief review about selection for multiple traits emphasizing the value of economic selection indices, and a discussion about effective use of genomics for sire and replacement heifer selection. The chapter is largely based on the structure and achievements of the US dairy industry, but the concepts described are applicable to dairy genetic programs worldwide.

The basics of genetic selection Genetic selection is a very powerful tool for achieving lasting gains in dairy cattle performance. Contrary to improvements in nutrition, management or cow comfort, changes achieved through selection are incremental, cumulative and permanent, which makes genetic improvement a very cost-effective strategy. Genetic selection works by identifying and selecting the animals with the highest genetic merit as parents of the next generation, resulting in the genetic (and also phenotypic) improvement of the population in each generation. The rate of genetic gain (R) in a given trait can be calculated as R ¼ h2$S, where S is the selection differential and h2 is the heritability of the trait. The selection differential

measures the superiority of the selected individuals with respect to the entire population, and the heritability captures the proportion of phenotypic variation 
due to additive genetic values h2 ¼ s2A s2P . Note that traits with higher h2 respond more rapidly to genetic selection. In general, type traits, such as stature and body depth, have high heritabilities (h2  35%), whereas production traits, such as milk yield and milk solids, have moderate heritabilities (15%  h2  30%), and fitness traits, such as fertility and health, have typically low heritabilities (h2  10%). Note that the selection differential (S) can be written as the product of selection intensity (i) and the phenotypic standard deviation (sP), and hence, the rate of genetic gain can be written as, R ¼ i$h$sA. More generally, the annual genetic gain is given by the famous breeder’s equation, i$rAC $sA L where DG is the annual genetic gain, i is the selection intensity, a measure of the superiority of the parents of the next generation, rAC is the accuracy of selection, a measure of the precision with which genetic merits are estimated (rAC is equal to the square root of the reliability), sA is the additive genetic standard deviation, and L is the generation interval defined as the average age of parents when their offspring are born. Note that i and L are properties of the population while rAC and sA differ from trait to trait. Animals are selected based on their genetic values, and the prediction of these genetic values occurs through the integration and analysis of multiple types of data, including phenotypic records (e.g., test-day milk yield, days open, health events), pedigree information, and more recently, genotypic data. The most important output of this process, known as genetic evaluation, is the estimate of genetic merit, commonly known in dairy cattle breeding as predicted transmitting ability (PTA). The PTA is an DG ¼

II. Lactation and management of dairy cattle

Selection for traits that increase income

estimate of the relative genetic superiority (or inferiority) that a particular animal will pass to its offspring for a given trait, and hence, represents the most important tool for making selection decisions. PTAs are exceptional tools for comparing and ranking animals, because the difference between the PTAs of two animals is an estimate of the difference expected to be observed in the performance of their progeny. Importantly, each PTA is accompanied by a value of reliability (REL), that measures the accuracy or degree of confidence in the PTA value. It is expressed as a percentage and ranges from 1 to 99. Technically, it is defined as the squared correlation between the true transmitting ability and the predicted ability of a given animal. REL is a function of the heritability of the trait and the amount of information available for the animal; basically, as heritability and the amount of information increases, REL also increases. Therefore, a bull has a more reliable PTA for protein yield than for daughter fertility because protein yield has a higher heritability. Similarly, a bull with many daughters has a more reliable PTA for any given trait than a bull with no or just a few daughters. Finally, another important output of the genetic evaluation is the percentile rank, a measure of the rank or position of the animal within the population evaluated for a given trait of interest. The interpretation of the percentile rank is very straightforward: if a bull ranks for a given trait at the 95th percentile, this means that the bull is genetically superior to 95% of all the evaluated bulls of its breed. In dairy cattle breeding programs, the rate of genetic gain can be described in terms of four paths of selection, namely selection of sires and dams of bulls, and selection of sires and dams of cows.1 The first path, sires of bulls, represents the most elite males in the population that are selected to be the sires of the next generation of bulls. This path is characterized by high accuracy and high selection intensity, top 3e5% of the bulls available in the market. The second path, dams of bulls, represents the group of elite

103

females that are mated to bulls from the group sires of bulls in order to produce bull calves. This path is characterized by high selection intensity, these females typically rank in the top 1% of the commercial cow population, but relatively low selection accuracy in the absence of genomic testing. The third path, sires of cows, represent the large group of bulls whose semen is used to generate the new generation of replacement heifers in commercial farms. This path is characterized by high selection accuracy and relatively high selection intensity. Finally, the forth path of selection, dams of cows, involves the large population of cows on commercial farms that are primarily used to produce milk. These cows are typically mated to bulls from the sires of cows group, in order to initiate a new lactation and produce female replacements on the farm. This selection path is traditionally characterized by low selection intensity and low selection accuracy, although recent herd management improvements coupled with the advent of genomic testing are modifying the potential contributions of this path to the overall genetic progress.

Selection for traits that increase income Milk yield. Dairy cattle selection programs have traditionally focused on increasing total lactation milk yield. Fig. 6.1 shows the average milk yield per cow per year between 1957 and 2016 in the US Holstein dairy herd. Notably, average annual milk production has increased from about 13,000 to 28,000 lb in the last 60 years (Fig. 6.1A). Interestingly, much of this improvement in productivity is due to genetic selection. Indeed, genetic improvement for milk yield averaged 146 lb per year, accounting for 57% of the total improvement (Fig. 6.1B). Note that genetic gains should be accompanied by improvements in cow nutrition and management, if not genetic selection would lead to an unrealized potential. Although milk volume remains

II. Lactation and management of dairy cattle

104

6. Genetics and genomics of dairy cattle

FIG. 6.1 Changes in milk yield for US Holsteins between 1957 and 2016. (A) Phenotypic trend. (B) Changes due to genetics or management: the orange (light gray in print version) area shows changes due to improved management while the blue (black in print version) area shows gains due to increased genetic potential. Source: Council on Dairy Cattle Breeding website (December 2018; https://www.uscdcb.com).

important in some markets, the emphasis placed on milk yield has decreased over time as fat and protein have gained more interest. Genetic merit for milk volume continues to increase because it is highly correlated with milk solids. Milk composition. In many markets, the vast majority of milk is used for making manufactured dairy products, such as cheese, ice cream, butter, and yogurt, among others, rather than for fluid milk consumption. In this situation, increasing fat and protein yield is more important than increasing milk volume. In theory, there are basically two ways to increase milk solids: (i) increasing fat and protein percentages while keeping milk yield constant; or (ii) increasing total milk yield while keeping fat and protein percentages constant. In practice, most breeding programs have increased fat and protein yield by direct selection for these traits, with little concern about whether this increase comes from milk volume or component percentages. In US Holstein cows, in the last 40 years, average annual fat yield increased from 624 to 1079 lb, whereas average annual protein yield increased from 541 to 873 lb. Notably, genetic

selection explains approximately 57% and 66% of the improvements in fat and protein yield, respectively. During recent years there has been growing interest in milk with specific nutritional value, such as specific protein composition (rich in A2 b-casein) or desirable fatty acid profile (high in unsaturated fatty acids), and improved manufacturing properties (coagulation time and curd firmness). If milk processors start to pay premiums for these traits, then farmers will have clear economic incentives to select for altered milk composition and manufacturing attributes.

Selection for traits that reduce expenses Fertility. Reproductive efficiency is a very important economic trait in dairy cattle. Reproductive inefficiency results in increased calving intervals, increased involuntary culling rates, decreased milk production, and delayed genetic progress, among other problems, causing significant economic losses.2 Genetic selection for improved cow fertility is a high priority

II. Lactation and management of dairy cattle

Selection for traits that reduce expenses

105

FIG. 6.2 Concomitant changes in milk yield and pregnancy rate in US Holstein dairy cattle. Source: Council on Dairy Cattle Breeding website (December 2018; https://www.uscdcb.com).

worldwide. It is well-documented that production and fertility are negatively correlated, and selection programs that have placed substantial emphasis on yield and have neglected fertility, have inevitably experienced a decline in reproductive performance.3 Fig. 6.2 shows the observed phenotypic trends in total milk yield and pregnancy rate in US Holstein cows in the last six decades. It is clear that concomitant with intensive selection for increased milk yield, pregnancy rate declined steadily until the beginning of the 2000s. Progress on improving reproductive performance began in 2003 with the introduction of daughter pregnancy rate (DPR), considered the primary trait for selection for cow fertility in US dairy cattle.4 An increase of 1% in DPR corresponds to a decrease of approximately 4 days open. Nowadays, besides DPR, two additional female fertility traits are routinely evaluated in US dairy cattle, namely heifer conception rate and cow conception rate. These traits reflect genetic potential in the probability of achieving conception when heifers of lactating cows are inseminated.

Health. Cow health directly impacts dairy farm profitability. Health events cause substantial economic losses, including losses due to onfarm death, premature culling, reduced milk production, and increased veterinary and treatment costs.5 Production and functional traits are negatively correlated, and the intense selection for milk yield in the last decades has compromised health and reduced fitness in dairy cattle.6 Genetic improvement of cow health and welfare is of paramount importance for the dairy industry worldwide. Traditionally, most breeding programs have focused on indirect measures of cow health and fitness, such as length of productive life or somatic cell count as an indicator of udder health. Indeed, genetic evaluations for somatic cell score were developed many years ago to facilitate indirect selection for mastitis resistance.7 However, direct selection for health traits is more effective than indirect selection using indicator traits. In this sense, the Nordic countries have led the development and implementation of genetic selection programs for cow health traits, including clinical

II. Lactation and management of dairy cattle

106

6. Genetics and genomics of dairy cattle

mastitis and metabolic disorders. Recently, the US dairy industry implemented genetic evaluations in Holstein for six health traits, including milk fever, retained placenta, metritis, displaced abomasum, ketosis, and clinical mastitis. These six health traits are considered the most common and most costly health events impacting US dairy herds. Longevity. The length of productive life, measured from first calving until culling, is commonly used to evaluate the longevity of a lactating dairy cow. This trait is arguably the most direct measure of a cow’s ability to survive on a commercial farm, i.e., cow’s ability to avoid dying on the farm or being culled. In the US, genetic evaluations for productive life have been available since 1994.8 Average length of productive life has increased approximately 0.25 months per year in the last decade in US Holstein cows. Despite this improvement, about 20% of dairy cows still die on the farm instead of being sold, approximately 7% per lactation, representing an economic loss of about $2 billion. In order to alleviate this problem, the US dairy industry recently introduced cow livability, a new longevity trait defined as the probability of a lactation not ending in death or on-farm euthanasia.9 Note that livability measures a cow’s genetic ability to stay alive while on the farm, while productive life measures a cow’s genetic ability to avoid dying on the farm or being culled. Livability is positively correlated with productive life (0.70) and daughter pregnancy rate (0.45), and negatively correlated with somatic cell score (0.25). Calving ability. Calving performance is considered an important functional trait in dairy cattle. Calving difficulty (dystocia) increases labor and veterinary costs, increases mortality in both cows and calves, decreases milk production, and leads to impaired female fertility.10 Genetic evaluations for calving ability include calving ease (direct and maternal), stillbirth rate (direct and maternal) and gestation length. Calving difficulty and stillbirth rate are highly

correlated, and both are related to differences in gestation length.11

Selection for multiple traits There are a large number of traits, including production traits (such as milk yield and milk composition) and functional traits (such as fertility, health, longevity, and calving ability), that directly impact the profitability of any dairy production enterprise. One simple method of multiple-trait selection is the use of independent culling or rejection levels. In this method, first minimum standards or cut-off values are chosen for each of the traits undergoing selection, and then only animals that meet simultaneously all the criteria are selected. For example, one might decide to use only bulls with PTAs that are at least þ35 for protein yield, þ4.5 for productive life, and þ2.8 for daughter pregnancy rate. Although this method is quite popular and allows one to select simultaneously for multiple traits using simple rules, it has some important limitations. First, the threshold values are in general chosen arbitrarily without using any formal approach. In addition, these cut-off values may vary over time due to genetic progress and changes in the definition of the genetic base, and therefore, cut-off values that are appropriate today may be too restrictive or too liberal in the near future. Second, this method ignores the genetic relationships between traits of interest; this adversely impacts the efficiency of selection when we want to select for traits that are genetically correlated, such as production and fertility. Finally, the effectiveness of the independent culling levels decreases rapidly as the number of traits under selection increases; as more traits are considered, fewer bulls meet simultaneously all the criteria, and more importantly, these bulls are probably only marginally superior for each trait. The best approach for selecting animals considering multiple traits is the use of an economic selection index.12 The overall breeding

II. Lactation and management of dairy cattle

Selection for multiple traits

goal (H) is to improve multiple economically relevant traits by selecting animals using a selection index (I). The selection objective is represented as H ¼ a1$G1 þ a2$G2 þ.þ am$Gm where Gi are breeding values and ai represent economic weights. These economic values are based on prices for both inputs (e.g., feed and veterinary costs) and outputs (e.g., milk prices, calf prices) of a dairy production enterprise. The selection index is computed as I ¼ b1$P1 þ b2$P2 þ.þ bk$Pk where Pj are measured phenotypes and bj represent index weights. These weights are calculated as b ¼ P1Ga, where P is the phenotypic (co)variance matrix for the traits included in the selection index, G is the matrix of genetic (co)variances among the traits in the selection index and the breeding objective, and a is the vector of economic values. The correlated response to selection in the breeding objective due to selection based on the selection index is maximized when the traits in the index are accurately measured and highly correlated to the traits in the breeding objective. Contrary to the method based on independent culling levels, selection indices perform well regardless of the number of traits under selection, and even more importantly, these indices allow for selection of animals that are highly superior for one trait and slightly deficient in other traits, which leads to the maximization of the selection response. Economic selection indices are updated periodically in order to include new traits and to reflect price trends.13e15 Table 6.1 shows the evolution of USDA Lifetime Net Merit (NM$) index, probably the most popular index in the US dairy industry. The first USDA index, Predicted Difference Dollars (PD$), included only milk and fat yield. The NM$ was developed in 1994 combining five traits, namely milk yield, fat yield, protein yield, productive life and somatic cell score. Three functional type traits, udder composite, feet and legs composite, and body weight/size composite, were included in NM$ in 2000. Subsequent updates included the

107

incorporation of daughter pregnancy rate (2003), calving performance (2006), heifer and cow conception rates (2014), and livability (2017). The last updated of the NM$ index was in August 2018 with the inclusion of six health traits, clinical mastitis, ketosis, retained placenta, metritis, displaced abomasum, and milk fever. These six health traits were added to the index in the form of a health trait sub-index (HTH$). Overall, the emphasis on yield traits has declined over time as health and fertility traits, commonly grouped as fitness traits, were introduced. Nowadays, most economic selection indices include both production and fitness traits.16 Fig. 6.3 shows traits included in selection indices from 13 different countries. Common trait groups include production (e.g., milk volume, protein and fat yield), fertility (e.g., pregnancy rate, calving interval), longevity (e.g., productive life, survival rate), health (e.g., somatic cell counts, clinical mastitis, postpartum disorders), type (e.g., udder conformation, feet and leg score), calving (e.g., calving ease, stillbirth, dystocia), and others (e.g., milking speed, feed efficiency). All these total merit indices include production, fertility, longevity and health traits, although with different emphasis; for instance, for production, the BPI in Australia and the ICO in Spain have a relative weight of 51% while the NVI in the Netherlands has a relative weight of only 26%. Not all the indices include type traits and only few incorporate calving traits. US lifetime merit-based selection indices. The NM$ is the flagship index in the US dairy industry, and it is probably the most appropriate breeding goal for the vast majority of US dairy farms. Fat and protein yield receive the highest relative weights in NM$, representing 27% and 17%, respectively. Longevity traits, female fertility traits, health traits, and functional type traits receive relative weights of 19%, 10%, 6%, and 15%, respectively. Calving ability (CA$), a sub-index that includes both service-sire and daughter calving ease and stillbirth, receives a relative weight of 5%. Overall, current NM$

II. Lactation and management of dairy cattle

108 TABLE 6.1

6. Genetics and genomics of dairy cattle

Evolution of the USDA lifetime net merit (NM$) index.

Traits

PD$ (1971)

MFP$ (1976)

CY$ (1984)

NM$ (1994)

NM$ (2000)

NM$ (2003)

NM$ (2006)

NM$ (2010)

NM$ (2014)

NM$ (2017)

NM$ (2018)

Milk

52

27

2

6

5

0

0

0

1

1

1

Fat

48

46

45

25

21

22

23

19

22

24

27

Protein

e

27

53

43

36

33

23

16

20

18

17

Productive Life

e

e

e

20

14

11

17

22

19

13

12

Somatic Cell Score

e

e

e

6

9

9

9

10

7

7

4

Body Weight Composite

e

e

e

e

4

3

4

6

5

6

5

Udder Composite

e

e

e

e

7

7

6

7

8

7

7

Feet & Legs Composite

e

e

e

e

4

4

3

4

3

3

3

Daughter Pregnancy Rate

e

e

e

e

e

7

9

11

7

7

7

CA$ (calving trait sub-index)

e

e

e

e

e

e

6

5

5

5

5

Heifer Conception Rate

e

e

e

e

e

e

e

e

1

1

1

Cow Conception Rate

e

e

e

e

e

e

e

e

2

2

2

Livability

e

e

e

e

e

e

e

e

e

7

7

HTH$ (health trait sub-index)

e

e

e

e

e

e

e

e

e

e

2

Data obtained from the Council on Dairy Cattle Breeding website (December 2018; https://www.uscdcb.com).

has relative weights of 45% for production traits, 40% for fitness traits, and 15% for functional type traits. The USDA-ARS Animal Improvement Programs Laboratory has also developed three alternative selection indices for producers with special milk markets or production systems. For dairy producers who are paid mainly for milk volume, i.e., markets where the incentives for components are insignificant, the Fluid Merit Index (FM$) is probably the most appropriate breeding goal. FM$ has relative weights of 18% for milk yield, 27% for fat yield, and 0% for protein yield. For dairy farmers who are paid

mainly for milk components, Cheese Merit Index (CM$) is probably the most appropriate economic selection index. Compared to NM$, CM$ places more emphasis on protein yield, and milk volume is more penalized indicating that the selection for more milk solids should be achieved by improving fat and protein percentage rather than improving total milk yield. Pasture-based dairy producers may find the Grazing Merit Index (GM$) index as the most convenient economic selection index; GM$ places roughly the same emphasis on production and health traits as NM$, but more

II. Lactation and management of dairy cattle

Selection for multiple traits

II. Lactation and management of dairy cattle

FIG. 6.3

Economic selection indices for dairy cattle in 13 different countries.

109

110

6. Genetics and genomics of dairy cattle

emphasis on female fertility traits and slightly less emphasis on productive life and livability.

Genomic selection: the latest revolution Genomic selection refers to selection decisions based on genomic-estimated breeding values. These genomic breeding values are calculated using genetic markers across the entire genome.17 This technology has revolutionized dairy cattle breeding worldwide because it allows breeders to make accurate selection decisions at a much earlier age, even when neither the animal nor its offspring have been assessed for the phenotypes of interest. Three major developments allowed the widespread use of DNA information in dairy cattle breeding: the identification of many thousands of single nucleotide polymorphism (SNP) markers spanning the entire bovine genome,18 the development of SNP-chip genotyping technologies that allow the genotyping of thousands of SNP markers in a (very) cost effective manner,19 and the development of suitable statistical methods where genome-wide SNP effects are estimated simultaneously without any significance testing.20 Moreover, several factors make dairy cattle improvement programs ideal for implementing genomic selection, including (i) nearly all relevant traits are sex limited and cannot be measured until the females begin lactating, (ii) individual animals have sufficient value to easily offset the genotypic costs, (iii) massive historical phenotypic data for building large reference populations, and (iv) access to data processing and evaluation infrastructure. Indeed, genomics has undoubtedly caused the most remarkable change in dairy cattle breeding since the introduction of artificial insemination. Hundreds of thousands of animals have been genotyped worldwide, including nearly every potentially elite young animal, and this genomic information is fully integrated into national genetic evaluations. Young bulls and potential elite

females are typically genotyped using mostly medium-density (roughly 50,000) or even highdensity (roughly 700,000) SNP genotyping arrays, while most heifers in commercial farms are genotyped with low-cost, low-density genotyping arrays with roughly 10,000 to 20,000 SNP. Fig. 6.4 shows the number of genotyped animals included in the US Holstein genomic evaluation since January 2011. The first official US genomic evaluation for Holsteins was realized in January 2009, and since then over 2.3 million genotypes have been received. Notably, the vast majority among the genotyped animals are heifers genotyped with low-density SNP chips. The effective use of low-density genotypes for predicting genomic breeding values has been made possible due to the development of efficient imputation algorithms that allow the prediction of the genetic merit of a heifer calf with almost the same accuracy as using mediumdensity SNP data but for a fraction of the cost.21 Genomic selection has the potential to increase genetic gain considerably by reducing generation intervals and increasing selection intensity and selection accuracy. Progeny testing, the basis of dairy cattle breeding programs, is a very expensive and time-consuming process. At least 4.5 years are required for collecting semen of a potentially elite bull, rearing his offspring, and finally predicting his genetic merit based on his offspring’s performance. If the bull is good enough to use in the entire population, then his first sons and daughters will be born when he is about 5.5 years of age. This long generation interval limits the rate of genetic progress. However, genomic testing allows breeders to identify superior bull calves within a few weeks of age, and hence, instead of waiting a minimum of 4.5 years, breeders can used genomic-tested young bulls before 1 year of age. This drastically reduces the generation interval. Similarly, genomic testing of heifer calves allows one to make accurate selection decisions at an early age, and superior females can eventually enter into in vitro fertilization programs,

II. Lactation and management of dairy cattle

Genomic selection: the latest revolution

II. Lactation and management of dairy cattle

FIG. 6.4 Number of genotyped animals included in the US Holstein evaluation since 2011. Low density (LD; less than 40 k) or high density (HD; more than 40 k) SNP chips. Source: Council on Dairy Cattle Breeding website (December 2018; https://www.uscdcb.com).

111

112

6. Genetics and genomics of dairy cattle

even before they reach sexual maturity. Moreover, for young bull calves and heifers, genomic testing provides more accurate PTA estimates than traditional parent averages, with average gains in reliability around 30%.22 Finally, greater selection intensity can be achieved using genomics because a large number of selection candidates can be screened in search of elite animals. Overall, by shortening the generation interval and increasing the accuracy and intensity of selection, genomic selection in dairy cattle can at least double annual genetic gains for economically important traits. Fig. 6.5 shows the average net merit (NM$) of marketed US Holstein bulls that entered artificial insemination service in 2005 and later. Interestingly, the rate of genetic improvement in NM$ has increased dramatically since the implementation of genomic evaluation. Note that the benefit of genomics is greatest for lowly heritable traits such as fertility, and traits that can be measured only late in life such as longevity. Indeed, genomic selection in US Holstein cattle has doubled the annual rates of genetic gain for production traits, but has increased from 3-fold to 4-fold for fitness traits, including female fertility, udder health, and productive life.23

Effective use of genomics: sire selection Dairy sire selection has dramatically changed with the implementation of genomic selection. Nowadays, dairy farmers have basically two main options when they make sire selection decisions: use proven, progeny-tested bulls or use young genomic-tested bulls, i.e., young bulls with no progeny that have been evaluated using only their own genomic data. In the US, the National Association of Animal Breeders (NAABs) distinguishes these two groups of bulls as the active (A) bulls, progeny-tested bulls with performance information from at least 10 daughters, and the young genomic-tested (G) bulls, young bulls that do not have offspring yet with milk records. It is important to remark that the number of young genomic-tested bulls currently in the market far exceeds that of progeny-tested bulls. For instance, of the 3,270 Holstein bulls available in the US market in December 2018, 2,735 (84%) were young genomic-tested bulls. Similarly, 403 (79%) of the 510 available Jersey bulls had G status. The key concept regarding young genomictested dairy bulls is that, on average, these young bulls have greater predicted genetic merit values than the proven bulls. For instance,

FIG. 6.5 Average net merit (NM$) of US Holstein bulls by year of entry into the market. Source: National Association of Animal Breeders website (December 2018; https://www.naab-css.org).

II. Lactation and management of dairy cattle

Effective use of genomics: replacement heifer selection

considering the bulls available as semen donors to US dairy farmers in December 2018, the average NM$ of young bulls was $245 and $126 greater than for proven bulls in Holstein and Jersey breeds, respectively. It is worth noting that the changes achieved through genetic selection are cumulative and permanent, and hence, it is expected that the new generation of bulls (G bulls) have (on average) greater genetic merit than the older bulls (A bulls). Now, in the case of the young genomic-tested dairy bulls, higher genetic values are accompanied by lower reliability values. Indeed, considering NM$ of the bulls available in the US market in December 2018, young bulls had 15% and 18% lower reliability values than proven Holstein and Jersey bulls, respectively. This is not surprising considering that the young genomic-tested bulls do not yet have progeny. The question is how dairy farmers should proceed in this scenario, i.e., farmers should use young genomic-tested bulls because they have greater PTA values, or, instead, farmers should use proven bulls because they have more reliable PTA estimates. At this point, it is important to remark that sire selection decisions should be always based on PTA values, and the reliability should be used as a guide to decide how intense to use a bull. Therefore, in this scenario, the best strategy is to use a group or team of young genomic-tested bulls. The advantage of using a group of young bulls is that reliability of the average genetic merit of the team is considerably greater than the reliability of each individual bull.24 The formula for calculating the reliability of a team of young genomic-tested bulls is given by team REL ¼ 1  (1  average RELi)/n, where average RELi is the average REL of individual bulls and n is the number of bulls in the team. For instance, if the reliability of individual young bulls is 70%, the reliability of the genetic merit for a team of three young bulls is about 90%, and if the team increases to six or even twelve young bulls, then the reliability values for the group average between 95% and 98%.

113

Effective use of genomics: replacement heifer selection The selection of replacement heifers in commercial dairy farms has been traditionally characterized by very low intensity of selection, because, in general, farmers retain nearly every heifer calf as a future herd replacement. However, recent improvements in herd management and cow comfort have reduced culling rates and improved reproductive efficiency, which has led to the ability to produce an excess of heifers. In addition, sexed semen is now commonly used is dairy farms, generating a considerable surplus of heifer calves. In this context, the selection of replacement heifers is feasible, and genomic testing allows the identification of superior or inferior heifer calves accurately and at an early age. What are the advantages of using genomics for selecting heifer calves? The key point is trying to estimate as precisely as possible the genetic merit of a heifer at a young age. In the absence of genomic information, the selection or culling of a given heifer calf is based on the average genetic merit of her parents, also known as parent average. The reliability of parent average typically ranges from 0 to 0.40 depending on the completeness and accuracy of the pedigree data. Now, if genomic testing is used, then the reliability of the genomic-predicted genetic merit of the heifer calf ranges from 0.60 to 0.75 depending on the trait. Interestingly, this genomic prediction early in life is generally more reliable than the traditional PTA estimated using several lactation records of both the cow and her daughters. Therefore, genomic testing allows farmers to make accurate selection (culling) decisions at an early age; and these decisions are more reliable than those than can be achieved using pedigree information alone. Genomic information on individual heifer calves can be used to reduce feed costs and improve the genetic level of herd replacements.

II. Lactation and management of dairy cattle

114

6. Genetics and genomics of dairy cattle

The identification of genetically inferior heifer calves allows early culling of these animals, significantly reducing the cost of rearing replacements. Alternatively, these genetically inferior heifers can be inseminated with beef semen to produce high-value crossbred beef calves. Note that cows inseminated with beef semen are in fact removed as parents of the next generation. On the other hand, the identification of superior heifers through genomics can be combined with the use of advanced reproductive technologies to rapidly propagate these animals and generate superior replacements. For instance, highgenetic-merit heifers can be used as donors in either an in vitro fertilization program or an embryo transfer program. Instead, these superior heifers can be inseminated using sexed semen from top sires. It is worth noting that genotyping replacement heifers has extra benefits other than making proper selection and mating decisions, including parentage verification, controlling inbreeding, and avoiding the spread of genetic disorders through genomic-enhanced matings. Arguably, these benefits add value to genomic testing. One of the key points regarding the use of genomics for selecting herd replacements is to demonstrate that the results of the genomic testing are highly correlated with future phenotypic performance. As such, early genomic predictions were compared to subsequent production, udder health, and reproductive performance in the first lactation of Holstein cows.25 Cows were ranked based on their own genomic PTA values (predicted at 12 months of age), and these alternative quartile rankings (from top 25% to bottom 25%) were then compared with the actual phenotypic performance in the first lactation. The 305-day mature equivalent milk yield, average monthly log somatic cell counts, and days open were evaluated as production, udder health, and fertility traits, respectively. If there is an association between genomic testing and future performance, then it is expected that the best heifers in terms of genomic values show

greater phenotypic records. Indeed, for milk production, the observed difference between the top and the bottom quartiles was equal to 4,800 lbs. (Fig. 6.6A). For udder health, the difference in log SCC between the top and the bottom quartile was equal to 0.82 (Fig. 6.6B). For female fertility, the actual difference in days open between those heifers classified as top 25% and those classified as bottom 25% was equal to 21 days (Fig. 6.6C). Therefore, these findings clearly show that early genomic predictions (performed on calves or yearling heifers) can be effectively used as predictors of future performance. In other words, genomic testing can be used to make accurate selection decision at a young age.

Novel traits in the genomics era Genomics has created opportunities to improve traits that are critically important, but too difficult or expensive to measure on the entire population. These relevant phenotypes can be measured only on a relatively small group of genotyped animals, and this reference population can then be used to predict genomic breeding values for the entire population, including young selection candidates.26 Examples of these important traits include feed efficiency,27 methane emission,28 milk progesterone profiles,29 thermoregulation,30 adaptive immune response,31 susceptibility to bovine leukemia virus,32 and resistance to bovine respiratory disease.33 Among these, feed efficiency is probably the most important, as well as the most challenging. Feed represents more than 50% of the total production costs. Hence, improving the efficiency with which dairy cows convert feed into milk has a large economic value. At the same level of production, cows with reduced feed intake requirements are more profitable. It has been suggested that the US dairy industry could save $540 million/year with no loss in milk production by breeding for cows that are more feed efficient. Residual feed intake, the difference

II. Lactation and management of dairy cattle

Novel traits in the genomics era

115

FIG. 6.6 Phenotypic performance of Holstein cows in their first lactation according to genomic potential. (A) Milk production. (B) Udder health. (C) Reproductive performance. Genomic PTA values were obtained through genomic testing at 12 months of age. DPR, daughter pregnancy rate; SCS, somatic cell score. Adapted from Weigel KA, Mikshowsky AA, Cabrera VE. Effective use of genomics in sire selection and replacement heifer management. In: Paper Presented at: Proc. Western Dairy Management Conference, Reno, NV; 2015.

II. Lactation and management of dairy cattle

116

6. Genetics and genomics of dairy cattle

between actual intake and intake predicted based on body weight and production level, has been proposed as a selection criterion for improving feed efficiency.34 Interestingly, the selection for lower residual feed intake (improved feed efficiency) has the potential to not only reduce feed costs, but also reduce significant sources of greenhouse gas emissions, such as enteric methane and manure. To date, measures of residual feed intake are limited to research facilities that can precisely determine individual cow feed intake, body weight, body condition score, and milk energy output. Measuring residual feed intake on larger populations, including commercial farms, seems infeasible due to cost and labor constraints. Genomics is an attractive approach for improving feed efficiency because feed intake phenotypes can be collected for a small group of lactating cows and genomicbased breeding values predicted for the entire population.35 In addition, new technologies have been developed that may help predict feed intake. These include sensors for monitoring body temperature, feeding behavior, and physical activity, as well as infrared spectral profiles of milk. These low-cost phenotypes may be combined with direct observations of feed intake to increase the accuracy of genomic evaluations. Dairy bull fertility is another trait that has gained much attention recently. Semen from one service sire bull is used to inseminate hundreds of cows and, thus, one sub-fertile bull could have a major impact on herd reproductive performance. Bull fertility has been evaluated traditionally in the laboratory using different semen attributes, such as sperm morphology, sperm concentration, and sperm motility.36 Unfortunately, these semen traits explain only part of the differences observed in fertility among bulls. Alternatively, bull fertility can be directly evaluated using conception rate records. Since 2008, the US dairy industry has had access to a phenotypic evaluation of bull fertility called Sire Conception Rate (SCR), that is based on a

large, nationwide database of confirmed pregnancy records.37 Interestingly, there is a remarkable variation in SCR among sires, more than 10% conception rate difference between highfertility and low-fertility bulls, and part of this variation is explained by genetic factors.38e40 Recent studies have revealed promising results for predicting SCR values using genomic data.41 Note that SCR records are available only after the bulls are in the market, and hence, early genomic predictions can help the dairy industry make accurate genome-guided selection decisions, such as early culling of predicted sub-fertile bull calves.

Managing inbreeding and genetic diversity Balancing rapid genetic progress and maintenance of adequate genetic diversity has become one of the major challenges of the dairy industry worldwide.42 The loss of genetic diversity can be monitored using the inbreeding coefficient, defined as the probability that the two alleles at any locus in an individual are identical by descent, i.e., the two alleles come from the same ancestor. Inbreeding results from the mating of related individuals e an animal’s inbreeding coefficient is equal to half of the additive genetic relationship between its parents. The mating of related individuals is unavoidable in populations of finite sizes, but this is especially exacerbated in dairy cattle populations due to intense selection and heavy use of reproductive technologies, such as artificial insemination and embryo transfer. As an example, the average inbreeding coefficient for US Holstein cows increased from 0.33% in 1968 to 7.60% in 2018. Inbreeding increases the proportion of loci that are homozygous throughout the genome, some of which causes homozygosity of recessive alleles that negatively impact an animal’s performance. This phenomenon is commonly known as inbreeding depression and tends to be most

II. Lactation and management of dairy cattle

117

Final remarks

pronounced on fitness traits, but undesirable effects are observed in most traits including production. For instance, in US Holstein cows, lifetime net income decreases about $23 per 1% increase in the inbreeding coefficient.43 In addition, inbreeding increases the chances of the expression of lethal or sub-lethal recessive alleles. Examples of known genetic defects in dairy cattle include bovine leukocyte adhesion deficiency (BLAD), complex vertebral malformation (CVM), deficiency of uridine monophosphate synthase (DUMPS), mulefoot (syndactyly), cholesterol deficiency (HCD), and an increasing list of recessive haplotypes that lead to impaired fertility, early embryonic losses, and abortions.44 Haplotype and SNP tests are now routinely used to identify carriers and track the inheritance of these genetic defects. Inbreeding in the short-term can be controlled in the herd using computerized mate selection programs.45 Given a cow and a list of potential bulls, these programs control the inbreeding of the hypothetical offspring either by (i) minimizing the inbreeding, (ii) maximizing the expected genetic merit subject to a fixed inbreeding threshold, or (iii) maximizing the expected genetic merit after adjustment for anticipated costs of inbreeding

depression. It is possible to control inbreeding without affecting genetic progress. Fig. 6.7 compares the genetic trends in lifetime net merit and the trends in inbreeding of North Florida Holsteins, a large and very progressive US commercial farm, versus the entire US cow population. North Florida Holsteins has an annual genetic trend of $61 while the annual trend for the US cow population is $35 (Fig. 6.7A). This farm has achieved remarkable genetic gains by using genomic testing for selecting heifers combined with embryo transfer and in vitro fertilization for rapid propagation of the best females. Interestingly, these remarkable genetic trends have been achieved while keeping the inbreeding at the same rate as the rest of the US cow population (Fig. 6.7B).

Final remarks Dairy cattle genetic programs have achieved remarkable progress, mainly in production traits. The success of these programs is based on the collection and analysis of massive databases of performance records and pedigree information, widespread use of assisted reproductive

FIG. 6.7 Genetic trend of Lifetime Net Merit (A) versus trend of Inbreeding (B) in Holstein dairy cows. US population of Holstein cows (US; blue [light gray in print version]) and North Florida Holsteins (NFH; orange [black in print version]).

II. Lactation and management of dairy cattle

118

6. Genetics and genomics of dairy cattle

technologies, and more recently, genomic data. Indeed, genomics has transformed dairy cattle breeding programs because breeders can select young bulls and heifers with sufficient accuracy, reducing generation interval, thereby increasing the rate of genetic gain. Selection objectives have evolved over time, from increasing milk yield to improving milk solids and enhancing health and fertility traits, following the needs and concerns of producers, milk processors, and consumers. The widespread implementation of on-farm sensors and monitoring systems, such as activity and rumination monitors, automated calf feeders, and in-line milk sensors, among others, provide huge amounts of data, generating opportunities to incorporate new traits into genetic selection programs. In addition, new genomic technologies, such as whole-genome sequencing and genome editing, will provide new tools for genetic improvement. In the future, health and fertility traits, as well as environmental sustainability traits, such as feed efficiency and methane emission, will be very important given the increasing concerns of society about dairy cow welfare and the environmental impacts of dairy farming.

References 1. van Tassell CP, van Vleck LD. Estimates of genetic selection differentials and generation intervals for four paths of selection. J Dairy Sci. 1991;74(3):1078e1086. 2. Inchaisri C, Jorritsma R, Vos PLAM, van der Weijden GC, Hogeveen H. Economic consequences of reproductive performance in dairy cattle. Theriogenology. 2010;74(5):835e846. 3. Royal MD, Flint AP, Woolliams JA. Genetic and phenotypic relationships among endocrine and traditional fertility traits and production traits in HolsteinFriesian dairy cows. J Dairy Sci. 2002;85(4):958e967. 4. VanRaden PM, Sanders AH, Tooker ME, et al. Development of a national genetic evaluation for cow fertility. J Dairy Sci. 2004;87(7):2285e2292. 5. Liang D, Arnold LM, Stowe CJ, Harmon RJ, Bewley JM. Estimating US dairy clinical disease costs with a stochastic simulation model. J Dairy Sci. 2017;100(2): 1472e1486.

6. Egger-Danner C, Cole JB, Pryce JE, et al. Invited review: overview of new traits and phenotyping strategies in dairy cattle with a focus on functional traits. Animal. 2015;9(2):191e207. 7. Shook GE, Schutz MM. Selection on somatic cell score to improve resistance to mastitis in the United States. J Dairy Sci. 1994;77(2):648e658. 8. VanRaden PM, Klaaskate EJ. Genetic evaluation of length of productive life including predicted longevity of live cows. J Dairy Sci. 1993;76(9):2758e2764. 9. Wright JR, VanRaden PM. Genetic evaluation of dairy cow livability. J Anim Sci. 2016;94:178. 10. Dematawewa CM, Berger PJ. Effect of dystocia on yield, fertility, and cow losses and an economic evaluation of dystocia scores for Holsteins. J Dairy Sci. 1997;80(4): 754e761. 11. de Maturana EL, Wu XL, Gianola D, Weigel KA, Rosa GJ. Exploring biological relationships between calving traits in primiparous cattle with a bayesian recursive model. Genetics. 2009;181(1):277e287. 12. Hazel LN, Dickerson GE, Freeman AE. The selection index–then, now, and for the future. J Dairy Sci. 1994; 77(10):3236e3251. 13. VanRaden PM. Invited review: selection on net merit to improve lifetime profit. J Dairy Sci. 2004;87(10): 3125e3131. 14. Shook GE. Major advances in determining appropriate selection goals. J Dairy Sci. 2006;89(4):1349e1361. 15. Cole JB, VanRaden PM. Symposium review: possibilities in an age of genomics: the future of selection indices. J Dairy Sci. 2018;101(4):3686e3701. 16. Miglior F, Muir BL, Van Doormaal BJ. Selection indices in Holstein cattle of various countries. J Dairy Sci. 2005; 88(3):1255e1263. 17. Goddard ME, Hayes BJ. Genomic selection. J Anim Breed Genet. 2007;124(6):323e330. 18. Gibbs RA, Taylor JF, Van Tassell CP, et al. Genomewide survey of SNP variation uncovers the genetic structure of cattle breeds. Science. 2009;324(5926): 528e532. 19. Matukumalli LK, Lawley CT, Schnabel RD, et al. Development and characterization of a high density SNP genotyping assay for cattle. PLoS One. 2009;4(4): e5350. 20. Meuwissen THE, Hayes BJ, Goddard ME. Prediction of total genetic value using genome-wide dense marker maps. Genetics. 2001;157(4):1819e1829. 21. Weigel KA, Van Tassell CP, O’Connell JR, VanRaden PM, Wiggans GR. Prediction of unobserved single nucleotide polymorphism genotypes of Jersey cattle using reference panels and population-based imputation algorithms. J Dairy Sci. 2010;93(5): 2229e2238.

II. Lactation and management of dairy cattle

References

22. Wiggans GR, Cole JB, Hubbard SM, Sonstegard TS. Genomic selection in dairy cattle: the USDA experience. Annu Rev Anim Biosci. 2017;5:309e327. 23. Garcia-Ruiz A, Cole JB, VanRaden PM, Wiggans GR, Ruiz-Lopez FJ, Van Tassell CP. Changes in genetic selection differentials and generation intervals in US Holstein dairy cattle as a result of genomic selection. Proc Natl Acad Sci USA. 2016;113(28):E3995eE4004. 24. Schefers JM, Weigel KA. Genomic selection in dairy cattle: integration of DNA testing into breeding programs. Anim Front. 2012;2(1):4e9. 25. Weigel KA, Mikshowsky AA, Cabrera VE. Effective use of genomics in sire selection and replacement heifer management. In: Paper Presented at: Proc. Western Dairy Management Conference, Reno, NV. 2015. 26. Calus MP, de Haas Y, Pszczola M, Veerkamp RF. Predicted accuracy of and response to genomic selection for new traits in dairy cattle. Animal. 2013;7(2):183e191. 27. VandeHaar MJ, Armentano LE, Weigel K, Spurlock DM, Tempelman RJ, Veerkamp R. Harnessing the genetics of the modern dairy cow to continue improvements in feed efficiency. J Dairy Sci. 2016;99(6):4941e4954. 28. Wall E, Simm G, Moran D. Developing breeding schemes to assist mitigation of greenhouse gas emissions. Animal. 2010;4(3):366e376. 29. Sorg D, Wensch-Dorendorf M, Schopke K, et al. Genetic analysis of new progesterone-based fertility traits in dairy cows measured on-farm. J Dairy Sci. 2017; 100(10):8205e8219. 30. Dikmen S, Cole JB, Null DJ, Hansen PJ. Genome-wide association mapping for identification of quantitative trait loci for rectal temperature during heat stress in Holstein cattle. PLoS One. 2013;8(7):e69202. 31. Thompson-Crispi KA, Sewalem A, Miglior F, Mallard BA. Genetic parameters of adaptive immune response traits in Canadian Holsteins. J Dairy Sci. 2012;95(1):401e409. 32. Abdalla EA, Pe~ nagaricano F, Byrem TM, Weigel KA, Rosa GJ. Genome-wide association mapping and pathway analysis of leukosis incidence in a US Holstein cattle population. Anim Genet. 2016;47(4):395e407. 33. Neibergs HL, Seabury CM, Wojtowicz AJ, et al. Susceptibility loci revealed for bovine respiratory disease complex in pre-weaned holstein calves. BMC Genomics. 2014;15:1164.

119

34. Connor EE. Invited review: improving feed efficiency in dairy production: challenges and possibilities. Animal. 2015;9(3):395e408. 35. Yao C, Zhu X, Weigel KA. Semi-supervised learning for genomic prediction of novel traits with small reference populations: an application to residual feed intake in dairy cattle. Genet Sel Evol. 2016;48(1):84. 36. DeJarnette JM, Marshall CE, Lenz RW, Monke DR, Ayars WH, Sattler CG. Sustaining the fertility of artificially inseminated dairy cattle: the role of the artificial insemination industry. J Dairy Sci. 2004;87(Suppl.): E93eE104. 37. Kuhn MT, Hutchison JL. Prediction of dairy bull fertility from field data: use of multiple services and identification and utilization of factors affecting bull fertility. J Dairy Sci. 2008;91(6):2481e2492. 38. Nicolini P, Amorin R, Han Y, Pe~ nagaricano F. Wholegenome scan reveals significant non-additive effects for sire conception rate in Holstein cattle. BMC Genet. 2018;19(1):14. 39. Han Y, Pe~ nagaricano F. Unravelling the genomic architecture of bull fertility in Holstein cattle. BMC Genet. 2016;17(1):143. 40. Rezende FM, Dietsch GO, Penagaricano F. Genetic dissection of bull fertility in US Jersey dairy cattle. Anim Genet. 2018;49(5):393e402. 41. Abdollahi-Arpanahi R, Morota G, Pe~ nagaricano F. Predicting bull fertility using genomic data and biological information. J Dairy Sci. 2017;100(12):9656e9666. 42. Howard JT, Pryce JE, Baes C, Maltecca C. Invited review: inbreeding in the genomics era: inbreeding, inbreeding depression, and management of genomic variability. J Dairy Sci. 2017;100(8):6009e6024. 43. Smith LA, Cassell BG, Pearson RE. The effects of inbreeding on the lifetime performance of dairy cattle. J Dairy Sci. 1998;81(10):2729e2737. 44. VanRaden PM, Olson KM, Null DJ, Hutchison JL. Harmful recessive effects on fertility detected by absence of homozygous haplotypes. J Dairy Sci. 2011; 94(12):6153e6161. 45. Weigel KA, Lin SW. Use of computerized mate selection programs to control inbreeding of Holstein and Jersey cattle in the next generation. J Dairy Sci. 2000;83(4): 822e828.

II. Lactation and management of dairy cattle