4th Biennial

W. E. Petersen Symposium

Crossbreeding of Dairy Cattle:

The Science and the Impact

Presented by the Department of Animal Science University of Minnesota, St. Paul

April 2, 2007

Continuing Education & Conference Center University of Minnesota

St. Paul Campus

Department of Animal Science University of Minnesota

305 Haecker Hall 1364 Eckles Avenue St. Paul, MN 55108

Tel: 612.624.2722 Fax: 612.625.5789

Email: south005@umn.edu Web: www.ansci.umn.edu

The University of Minnesota is committed to the policy that all persons shall have equal access to its programs, facilities and employment without regard to race, color, creed, religion, national origin, sex, age,

marital status, disability, public assistance status, veteran status or sexual orientation.

Table of Contents W. E. Petersen Symposium .............................................................................................................. i Event Organizers............................................................................................................................... i Program Agenda ............................................................................................................................. ii About the Speakers ....................................................................................................................... iii Mechanisms of Inbreeding Depression and Heterosis for Profitable Dairying ...........................................................................................................................1 – Bennet Cassell, Virginia Tech, Blacksburg Impact of an Old Technology on Profitable Dairying in the 21st Century...................................................................................................................................... 7 – Brad Heins, University of Minnesota, St. Paul Experience with Crossbreeding–From Headaches to Happiness .......................................................................................................................................20 – Kevin Prins, Dairy Producer, Modesto, CA Genetic Evaluation Using Combined Data from All Breeds and Crossbred Cows ....................................................................................................................23 – Paul VanRaden, USDA, Beltsville, MD Crossbreeding–An Important Part of Sustainable Breeding in Dairy Cattle and Possibilities for Implementation ....................................................................29 – Morten Kargo Sørensen, Danish Agricultural Institute, Foulum, and Danish Cattle Federation, Aarhus

i

W. E. Petersen Symposium

Professor W. E. Petersen was a highly influential dairy scientist at the University of Minnesota in the mid-1900s. In honor of his memory, Dr. Petersen's family established a fund in the Department of Animal Science at the University of Minnesota to sponsor dairy-related symposia. The first biennial symposium was held in 2001.

The 4th biennial symposium addresses the science and impact of crossbreeding of dairy cattle. Crossbreeding is the hottest topic at this time in dairy genetics, and the University of Minnesota has taken a lead in conducting new research. Furthermore, the University of Minnesota has provided international leadership in educating dairy producers on implementation of crossbreeding systems.

Event Organizers

Dr. Les Hansen Morse Alumni Distinguished Teaching Professor University of Minnesota, St. Paul Dr. Dennis Johnson Professor West Central Research and Outreach Center Morris, MN Mr. Jim Dickrell Editor Dairy Today Dr. Dana Allen Dairy Producer Eyota, MN Mr. Joe Molitor Dairy Producer St. Cloud, MN

Supplemental financial support provided by MN Select Sires

ii

Program Agenda Monday, April 2, 2007

12:30 p.m. Welcome and Introductions Dr. Jim Linn, Interim Head and Professor Department of Animal Science, University of Minnesota, St. Paul 12:35 p.m. Mechanisms of Inbreeding Depression and Heterosis for Profitable Dairying Dr. Bennet Cassell, Virginia Tech, Blacksburg 1:20 p.m. Impact of an Old Technology on Profitable Dairying in the 21st Century Mr. Brad Heins, University of Minnesota, St. Paul 1:50 p.m. Experience with Crossbreeding–From Headaches to Happiness Mr. Kevin Prins, Dairy Producer, Modesto, CA 2:20 p.m. Refreshment Break 2:45 p.m. Genetic Evaluation Using Combined Data from All Breeds and Crossbred

Cows Dr. Paul VanRaden, USDA, Beltsville, MD 3:30 p.m. Crossbreeding–An Important Part of Sustainable Breeding in Dairy Cattle

and Possibilities for Implementation Dr. Morten Kargo Sørensen, Danish Agricultural Institute, Foulum, and Danish Cattle

Federation, Aarhus 4:15 p.m. Panel Discussion of Minnesota Dairy Producers Moderated by Dr. Les Hansen, Department of Animal Science, University of

Minnesota, St. Paul ▪ Dana Allen, Eyota ▪ Joe Becker, Eden Valley ▪ Joe Molitor, St. Cloud

5:00 p.m. Adjourn

iii

About the Speakers Dr. Bennet Cassell was raised on a SW Virginia farm unique from today’s commercial agriculture – 18 dairy cows, a 60-head cow-calf beef operation, and 90 commercial ewes. Most were crossbred to some degree. Dr. Cassell obtained BS and MS degrees from Virginia Tech, and a PhD in Animal Breeding from North Carolina State. He has been Extension Dairy Scientist-Genetics and Management at Virginia Tech since 1982. In 2002, Dr. Cassell implemented a joint Virginia Tech/U of Kentucky/North Carolina State crossbreeding project utilizing pure Holsteins and Jerseys, and reciprocal crosses of the two breeds. Mr. Brad Heins, a native of Lake City, MN, received his MS in dairy cattle genetics from the U of Minnesota, and is working on a PhD with Dr. Les Hansen. Mr. Heins is researching the effects of crossbreeding on production, fertility, and longevity using data from 7 California dairies. He is also studying the effects of crossbreeding in U of Minnesota herds. Mr. Kevin Prins operates a 560-cow dairy near Modesto, CA. Cows are grazed on 130 acres for 7 months and are confinement fed the remainder of the year. The Prins have been crossbreeding their cows for 7 years. Their dairy currently has a replacement heifer for every cow in the herd. Kevin indicates crossbreeding has resulted in them having "cows coming out of our ears". Dr. Paul VanRaden grew up on an Illinois dairy and became interested in genetics while working as a DHI supervisor. He completed a BS at the U of Illinois and a PhD at Iowa State. Dr. VanRaden joined USDA’s Animal Improvement Programs Laboratory at Beltsville, MD, in 1988. He introduced genetic evaluations for Productive Life and Daughter Pregnancy Rate, and has combined all available traits into the Net Merit index published since 1994. He currently conducts research on international evaluations, genetic markers, inbreeding, and cross-breeding. Dr. Morten Kargo Sørensen grew up on a dairy farm in SW Denmark and completed an MS at the Danish Agricultural University in 1992. He worked as a dairy breeding adviser for 3 years before returning for a PhD. Dr. Sørensen is jointly employed by the Danish Agricultural Institute, Foulum, and the Danish Cattle Federation, Aarhus. His research emphasizes inbreeding, cross-breeding, and sustainable breeding goals for dairy cattle, and he works closely with the dairy breeding industry in Denmark and throughout Europe. His efforts are focused on implementing optimum crossbreeding systems for Danish dairy producers.

1

Mechanisms of Inbreeding Depression and Heterosis for Profitable Dairying

Bennet Cassell

Extension Dairy Scientist, Genetics and Management Virginia Polytechnic Institute and State University, Blacksburg

Genes in diploid organisms operate singly, in pairs, and in conjunction with genes at other locations throughout the nuclear DNA. Inbreeding depression and heterosis arise from the effects of gene combinations, that is, the effects of pairs of genes. Gene pairs are unique characteristics of individuals that are broken down and reformed each generation. This basic biological fact, introduced to many of us in high school biology, is the foundation of everything there is to say about this topic. Very few topics are so well rooted in a simple process. Simplicity suffers, however, under real-world challenges of combining breeds or selecting to improve within a pure breed under a multi-trait breeding objective. This paper attempts to explain in a rudimentary way the ways in which genes interact, transmit, and recombine and the implications of those processes to breeding options available to dairy farmers. Mechanisms of inbreeding Inbreeding results from matings between related parents. Because breeding populations have finite size and long pedigree histories, mild inbreeding always exists by this definition. A more practical working definition of inbreeding is mating of parents more related than one would expect by chance alone. High levels of inbreeding are difficult to achieve in species where “selfing” is not possible. A 35-year project at the Beltsville Agriculture Research Center between 1912 and 1949 produced one dairy cow with an inbreeding coefficient of over 75%, the highest ever recorded for bovines under experimental conditions. A single generation of mating between this cow and an unrelated sire would break down all of the inbreeding in the dam. Extreme inbreeding is difficult to achieve and easy to eliminate – if you have access to an unrelated mate. Selection toward a single breeding objective can increase inbreeding, even in large populations. U.S. Holsteins have been under effective selection pressure for higher production and improved type since mid-1960. From 1982 until 2004, average inbreeding in a pedigree-recorded population of over 1,000,000 Holsteins increased from 1% to 5%. These figures may well understate actual inbreeding, as estimates are relative to a 1960 pedigree base and some pedigree information is missing in this population grade and registered animals. Consequences of inbreeding Inbreeding increases homozygosity. More gene pairs become identical because they are copies of the same ancestral genes. Consequences of inbreeding include an increase in uniformity of offspring of inbred individuals through reduced variation in genetic material between germ cells. Offspring face higher frequencies of deleterious recessive gene combinations, increased inbreeding depression, and greater variation in response to environmental stress. The last three effects of inbreeding are undesirable in commercial animals, while the first is of insufficient advantage to commercial producers to justify organized inbreeding programs. While inbred animals are expected to express undesirable recessive characteristics more often than out bred animals, inbreeding does not cause “bad” alleles. Such alleles already exist in populations, almost exclusively as complete recessives, where dominant alleles suppress their expression and elimination through natural selection. Inbreeding increases probabilities that an individual will inherit two copies of such alleles from related parents who inherited the undesirable gene from one common ancestor. While

2

unfavorable under commercial production systems, this feature of inbreeding can be used to “purge” an inbred line of such alleles. Inbreeding has been used to great advantage in plant species where generation intervals tend to be short, “selfing” allows rapid build-up of high levels of inbreeding, and many highly inbred lines can be developed and sustained simultaneously. Crosses of inbred lines produced hybrid corn – arguably one of the most successful advances in agriculture production in the last century. The role of inbreeding in animal breeding, however, is much more restricted and less successful. Outbreeding – the mating of individuals less related than the average pair of animals in a population – produces opposite effects of inbreeding. Outbred individuals are less likely to express deleterious recessive alleles. Outbred animals may bear little resemblance to their parents due to breakdown of homozygous gene combinations and dominance. Outbred animals tend to be less subject to environmental stress than inbred animals. These characteristics have made outbred animals very useful under certain production systems, particularly those where reproductive efficiency is highly valued or environmental stress is not moderated by human intervention. Two sources of homozygous gene pairs Homozygosity is common throughout animal genomes, particularly for genetic regions related to fitness and survival. Natural selection has driven unfavorable genes from breeding populations at many loci of genes. Random drift also contributes to homozygous combinations. A limited number of ancestors, perhaps survivors of a genetic bottleneck, may have, by chance, been homozygous at specific loci of genes where genetic variation once existed. A professor in my academic past, Dr. O. W. Robinson, called this a “founder effect”. [Bottlenecks also contribute to inbreeding and homozygosity at gene locations where genetic variation existed in founder animals.] If the possible ancestors are all homozygous for certain alleles, alleles passed to future generations will be alike, regardless of whether animals are closely related or not. Identical alleles are referred to as “alike in state” when they were inherited from unrelated parents. Alleles that are alike because they were inherited from related parents, both of whom received a copy of the same allele from their common ancestor, are “identical by descent”. Formulas for calculation of inbreeding coefficients rely on probabilities of genes being identical by descent. Genes that are “alike in state” are ignored. Genetic effects of homozygous gene combinations are the same, regardless of whether genes are alike in state or identical by descent. How does inbreeding happen in dairy cattle? Most dairy farmers, given an economically acceptable choice, would avoid close inbreeding. Acceptable choices aren’t always readily available. Dairy cattle in U.S. herds are either direct descendants of bulls widely used in AI service or a generation removed through use of sons of A.I. bulls in natural service. Bulls in A.I. service are used in many herds, which increases relationships between females that may be separated by time and distance. Prior to widespread use of A.I., relationships between females within a herd were likely higher than today, because many females were half sibs sired by the same bull. However, half sibs seldom existed in different herds. Such relationships are common today. Furthermore, selection toward a generally similar and stable breeding objective by A.I. studs caused the same animals to appear as parents of bulls in A.I. Two sires in the U.S. Holstein population from the 1960’s, Round Oak Rag Apple Elevation and Pawnee Farm Arlinda Chief, became highly influential through extensive use as sires of sons, sires of bull mothers, and through heavy use of successful progeny tested descendents. The February 2007 Holstein Red Book reported that Elevation and Chief were responsible for 15.2% and 14.7% of all the genes

3

carried by Holstein bulls likely to be used as A.I. sires this year. Matings of these current A.I. bulls to cows and heifers in the current Holstein population (daughters of these bulls themselves or of their sires) will certainly produce many gene combinations that are homozygous by reason of “identity by descent” from Elevation and Chief. It would be very difficult to construct a mating between truly unrelated animals in U.S. Holsteins provided remote generations were probed for relationships. Elevation and Chief appear very rarely in the first four generations of today’s A.I. bulls, but appear again and again in generations seven through ten. Options to avoid inbreeding Avoiding inbreeding entirely is highly restrictive, if it is possible at all. Dairy cattle of pure breed origin must serve a functional purpose in production of dairy products. Profit oriented herd managers want productive cattle, have the tools (progeny testing, sire evaluations, etc) to differentiate between more and less productive choices, and will logically favor a subset of the purebred population for breeding purposes. Rather than “avoiding” inbreeding, a more workable option would be to reduce and/or control major undesirable effects of inbreeding. Relationship between pairs of prospective parents in a breeding population should be constrained to a target compatible with reasonable genetic progress. Several steps can be taken to reduce relationships without unduly compromising genetic progress toward economically justifiable goals. The key concept would be to recognize that maximum genetic progress comes at a high price in inbreeding depression and in restriction of future options to maintain some genetic diversity. The most important decisions to control inbreeding and obtain most (but not all) benefits of selection are made by managers of A.I. young sire sampling programs. These managers respond to demands from semen purchasers, so the family of important decision makers to moderate effects of inbreeding is inclusive. I suggest that the following approaches be considered:

• Limit the number of sons of the “top” sire of sons in any given time period. Use an expanded list of sires of sons, with particular interest in bulls with divergent pedigrees.

• Limit the use of the most popular bull mothers or bull-mother families. This objective parallels the expanded list of sires of sons above. E.T. and other reproductive technologies give us the power to use the best females too much.

• Diversity of breeding objectives stimulates diversity of genetic background. Separate selection indices for intensive management conditions and grazing herds are an example of such diversity and should be pursued by dairy producers.

• The industry, and in particular, individual dairy farmers through semen purchases, should embrace with enthusiasm inclusive selection indices directed towards improvement of lifetime economic merit. Such indices offer opportunity for prospective parents with divergent pedigrees to find a role in genetic improvement programs.

Success begets success (and increases relationships) through use of pedigree information in genetic evaluations. Yet, Mendelian segregation remains a powerful tool for genetic improvement. Sampling divergent pedigrees (with the unavoidable reduction in genetic merit) must be rewarded in the marketplace. Ultimately, dairy producers will demand, and pay for, pedigree diversity. Payment may be through higher semen prices for bulls with unique pedigrees (but potentially lower genetic merit than “mainstream” genetics) or through use of unrelated parents from another breed in a crossbreeding program. Independence of dairy producers in the U.S. will lead to a variety of approaches in the years to come. I am optimistic that the trend will be away from the small and restrictive set of bull mothers and sires of sons that has dominated the breeding industry until very recently.

4

Options to utilize heterosis through crossbreeding Heterosis arises from favorable gene combinations. Gene combinations are not equally important to all traits in dairy cattle or other species. Furthermore, more genetically divergent breeds are more likely to generate more heterosis than breeds with more similar genetic backgrounds. Traits subject to small amounts of inbreeding depression (perhaps somatic cell score is one such example) are expected to show less heterosis, as inbreeding depression results from the breakdown of favorable gene combinations. Producers should be realistic about the benefits of heterosis for any trait and for specific pairs of breeds before finalizing expectations for crossbreeding. We know very little about heterosis for specific breed combinations for most traits in dairy cattle. For some important traits such as productive life and fertility, even general heterosis is not well established. Dairy producers should anticipate that benefits of additive genetic merit of individual breeds are more important to performance of a particular cross than is heterosis. This is not to discount the benefits of heterosis. However, heterosis alone should not be expected to overcome breed weaknesses in individual traits. The best rule for planning crossbreeding programs is to choose breeds carefully. Producers who don’t like one of the breeds they use in a cross probably won’t like the crossbred, either. Breed additive merit for different traits in dairy cattle is better established than heterosis for specific crosses, but we have few breed comparisons in the scientific literature where cows of different breeds performed at the same time in the same herds. The table that follows is one comparison of breeds for mature body size and production characteristics, two performance areas that will affect breeder approval of crosses. The averages shown are certainly subject to adjustment, but were selected to represent breeds realistically rather than to favor any breed. Changes may affect conclusions regarding breeding plans in important ways and should always be made based on better information.

Breed Mature body wt Milk yield Fat yield

Protein yield Fat % Protein %

Holstein 1600 23,300 840 700 3.6 3.0 Jersey 1100 17,600 810 640 4.6 3.6 Brown Swiss 1600 20,700 830 680 4.0 3.3 Swedish Red 1300 20,000 840 700 4.2 3.5 Normande 1100 16,000 700 580 4.4 3.6 Montbeliarde 1450 18,000 680 610 3.8 3.4

The choice of breeds to include in such a table is problematic in itself, as the listing omits choices that may appeal to others. The six breeds in the table either have long histories of performance as dairy breeds under U.S. conditions or are from European breeding programs with considerable recent success in genetic improvement towards dairy breeding objectives. The list does not include other viable populations such as New Zealand Jersey and Friesian strains or numerous dairy breeds/strains in Europe that may ultimately play important roles in U.S. crossbreeding programs. Predicting performance of breed crosses The “merit” of any combination of two or more breeds is the average of breed means, weighted by the percent of that breed in a cross, plus the benefits of heterosis for the trait, adjusted for any recombination loss generated by the breeding program that produced the cross. For instance, the expected performance for a 3-breed cross where Holstein-Jersey cows were mated to Brown Swiss sires would include 25% of the Jersey and Holstein breed average, plus 50% of the Brown Swiss average. This particular cross includes no loss of heterosis from recombination, because breeds represented in the female line are not

5

represented in the sire line. All gene combinations produced arise from different breed origins. Expected milk yield, then, with 5% heterosis would be (1 + retained heterosis) X [(Holstein contribution) + (Jersey contribution) + (Brown Swiss contribution)] or (1.05) [(.25)(23,200) + (.25)(17,600) + (.50)(20,700)] = 21,604 lbs A Jersey bull could be used as a backcross on the same Holstein-Jersey dam, creating quite different results. Heterosis would be reduced. Half of the genes contributed by the dam would be of Jersey origin, and would combine with Jersey genes of the sire to eliminate 50% of the heterosis in the Holstein-Jersey dam. Additive breed merit would also be affected. In this case, expected milk yield with 5% heterosis would be (1.025) [(.25)(23,200) + (.75)(17,600)] = 19,501 lbs The 3-breed crossbred exceeds the 75% Jersey backcross for milk yield by over 2,000 lbs per lactation because of more heterosis, but more importantly because of replacement of genes from a Jersey sire with a Brown Swiss sire. Brown Swiss exceed Jerseys by 3,100 lbs of milk. Loss of heterosis cost the backcross 475 lbs of milk yield but the replacement of Brown Swiss genes with an additional 50% Jersey genes reduced yield by 1,550 lbs. An additional advantage to the 3-breed cross, in this case, is that full heterosis is applied to higher breed additive merit. 2-breed versus 3-breed crossbreeding systems with purebred sires The complexity of planning a crossbreeding program arises from

1. the breed choices available 2. several traits contribute to merit, with breed differences for each 3. heterosis for different traits (and breed combinations) 4. retained heterosis of the specific crossbreeding system

Each additional breed included in a system increases mating complications and increases the difficulty of managing a semen inventory. Dairy producers should be grateful, however, for the relative ease with which A.I. can be incorporate into a dairy breeding program. Beef breeders who use natural service need separate pens and perhaps even multiple pens for different breed groups! At some point, multiple pens will be required for bulls involved. Semen tanks are a most reasonable alternative. Two-breed (and 3-breed) rotational systems using purebred bulls operate by mating each individual to a bull of the breed least represented in that individual’s pedigree. The Jersey backcross in the example above would be mated to a Holstein bull, producing an individual that is 5/8 Holstein, 3/8 Jersey, with 75% retained heterosis. The next generation, sired by another Jersey bull, halves the Holstein proportion to 5/16, leaving 11/16 Jersey blood and 63% retained heterosis. At equilibrium, two breed rotational systems produce two groups of animals with 2/3 alleles from one breed and 1/3 from the other, with 67% retained heterosis. The advantage of the 2-breed system is simplicity. The system also incorporates the two most favorable breeds for a particular producer’s purpose. If a third breed is considered to be inferior for important traits, the 2-breed system can be quite attractive. The disadvantages of the 2-breed system are that one-third of heterosis is lost and that breed additive merit changes considerably between breed groups. For a Holstein-Jersey 2-breed system, mature body weights (using figures in the table) differ by almost 200 lbs for Jersey sired versus Holstein sired cattle.

6

The 3-breed system adds complexity, but also adds the opportunity to utilize a third breed in a rotational system. Heterosis is an advantage. At equilibrium, 86% of full crossbred heterosis is retained. The shift in percentages of genes from the breed of the sire is also reduced from 66% to 57%. However, for a trait like mature body size, a mixture of two large breeds like Holstein and Brown Swiss with one much smaller breed like Jersey still generates large size differences between Jersey-sired crosses and crosses sired by the other two breeds. From a management perspective, such issues should be resolved as much as possible by choice of pure breeds at the initiation of the project. Cow management is an alternative to a genetic approach, as large herds could solve most size problems by cow grouping. Other systems Breeders have options to use systems other than those based on purebred sires. In dairy cattle breeding, however, availability of genetic evaluations from progeny tested and highly selected A.I. bulls is a tremendous impediment to any advantages arising from crossbred bulls. Other species have taken advantage of crossbred males, but those species, for the most part, have sought to improve traits expressed in both sexes relatively early in life. Milk production, fertility, and survival are not so kind to dairy producers. Selection of sires based on progeny-test results will remain an important consideration in dairy crossbreeding systems. This writer sees little advantage to use of crossbred bulls for the foreseeable future. New “breeds” created as composites of existing breeds are another option that simplify mating systems and eliminate generation to generation variation in performance. Composites are subject to the same problems of inbreeding depression as pure breeds. Composites, however, can be “recreated” from the original pure breeds from time to time to take advantage of genetic progress within those pure breeds and to correct inbreeding buildup. Progeny testing programs would have to be developed for composites to allow for selection to operate. Limitations of time and expense of such programs may be prohibitive in light of potential benefits. Conclusions Selection within pure breeds remains a viable option for producers wishing to improve traits that can also be modified through crossbreeding. Results, however, will be more slowly attained than for crossbreeding, and perhaps MUCH more slowly attained for traits where breed additive merit differs greatly and heterosis is substantial. Selection, however, imparts a permanent advantage that accumulates and builds over generations. Benefits from favorable gene combinations convey only to the individual and must be recreated each generation through mating plans. Inbreeding in pure breeds motivates interest in crossbreeding but perhaps a more powerful force has been deterioration in health and fertility traits. Had breeding plans that included fitness traits been in effect (and effective) for the past 40 years, current interest in crossbreeding might have been greatly reduced, if existing at all. Under current conditions, however, breed differences in size, calving ease, fertility, and production traits encourages many producers to consider crossbreeding programs for commercial milk production. Breeders devoted to their favorite pure breeds are encouraged to implement and to carefully follow selection plans that improve lifetime economic merit of the dairy cow. This effort requires selection pressure on the lowly-heritable, slow-to-change, difficult-to-measure fitness traits. Breeders opting to utilize crossbreeding programs should choose breed combinations carefully, use progeny-tested, purebred A.I. bulls and use the same selection for lifetime economic merit as purebred enthusiasts. Finally, crossbreeding programs should follow plans that maintain favorable combinations of breed additive merit and minimal recombination loss.

7

Impact of an Old Technology on Profitable Dairying in the 21st Century

Bradley J. Heins

University of Minnesota, St. Paul

INTRODUCTION Crossbreeding is an old technology; however, when used in today’s dairy systems, crossbreeding can produce profitable results for dairy producers. Interest in crossbreeding of dairy cattle has become a topic of great interest in the last five years and has developed in response to concerns dairy producers have about fertility, calving difficulty, and stillbirths in today’s genetically improved Holstein cows. The commercial pig, beef cattle, and sheep production have relied on crossbreeding to improve mortality, fertility, growth, and disease resistance for 50 years! There have been many research studies documenting the role of crossbreeding in the dairy industry, but many are quite old and dated. Old research indicated heterosis is greatest for traits related to mortality, fertility, health, and survival. The first scientific trials using crossbred dairy cattle date back as early as 1906 in Denmark and used the Jersey and Danish Red breeds. In the 1930s and 40s, experiments with dairy cattle were conducted to determine heterosis for milk and fat production resulting from crossbreeding (Touchberry, 1992). Crossbreeding has not been studied in research herds in the U.S. for many years. Earlier studies with experimental herds indicated that crossbreds were at least as profitable as pure Holsteins at the University of Illinois (Touchberry, 1992) and Agriculture Canada (McAllister et al., 1994). A crossbreeding project involving the Holstein and Guernsey breeds was conducted at the Illinois Agricultural Experiment Station from 1949 to 1969 (Touchberry, 1992). Heterosis for first-lactation milk and fat production was 4.3% and 4.1%, respectively; however, heterosis was considerably higher (12.0% for milk, 12.8% for fat) in second lactation. Heterosis for days open was 9.4%. When evaluating total performance of purebreds and crossbreds, Touchberry (1992) combined measures of survival, growth, production, and reproduction into an index to calculate the total income produced per cow per lactation and reported heterosis of 14.9% for total income produced per lactation. A Canadian study was conducted in five research herds during the 1970s and 1980s and heterosis of 16.5% was observed for lifetime milk production and 20% and 17.2% for lifetime fat and protein production, respectively. In the same crossbreeding study at Agriculture Canada, McAllister et al., (1994) reported greater than 20% heterosis for lifetime performance in crossbreds of Holstein and Ayrshire. This paper will report current results from studies of crossbreeding Holsteins with US Jersey and Brown Swiss sires, as well as sires from European dairy breeds. A recent crossbreeding study from New Zealand will also be discussed.

THE CALIFORNIA EXPERIENCE The decline in fertility and survival of pure Holsteins led the managers of seven large dairies in California to mate Holstein heifers and cows with imported semen of the Normande and Montbeliarde breeds from France, as well as the Swedish Red (SRB) and Norwegian Red (NRF) breeds. Some cows continued to be bred to Holstein A.I. bulls for a period of time in these dairies. The Swedish Red and Norwegian Red breeds share similar ancestry and exchange sires of sons; therefore, the breeds were collectively regarded as “Scandinavian Red” for this study.

8

Production Crossbreds and pure Holsteins that calved for the first time from June 1, 2002 to January 31, 2005 were studied for production. A total of 1,447 cows calved for the first time during this period, and these cows were followed throughout their lifetimes to gauge production. Actual production (milk, fat, and protein) for 305-day lactations was calculated with the Best Prediction technique used by USDA for national genetic evaluations in the USA. Results for 305-day actual production during first lactation are in Table 1. Fat (lb) plus protein (lb) was used to gauge the overall production of the pure Holsteins versus crossbreds. The Scandinavian Red-Holstein crossbreds (-3%), Montbeliarde-Holstein crossbreds (-5%) and the Normande-Holstein crossbreds (-9%) were all significantly lower (statistically speaking) than the pure Holsteins for fat (lb) plus protein (lb). These results for first lactation production are slightly different than those reported earlier from this study (Heins et. al, 2006c), because all cows now had the opportunity to complete their 305-day lactations. Table 1. First lactation production (actual 305-day with 2X milking).

Holstein

Normande-Holstein

Montbeliarde-Holstein

Scandinavian Red-Holstein

Number of cows 380 245 494 328

Milk (lb) 21,801 18,926 ** 20,305 ** 20,499 ** Fat (lb) 777 711 ** 743 ** 756 Protein (lb) 677 611 ** 645 ** 655 ** Fat (lb) + Protein (lb) 1454 1322 ** 1388 ** 1411 *

% of Holstein -9% -5% -3%

* Statistically significant difference from pure Holsteins (p<.05). ** Statistically significant difference from pure Holsteins (p<.01).

Table 2 has results for production for second lactation. Production of the pure Holsteins climbed substantially from first to second lactation. The three crossbred groups also greatly increased in production from first to second lactation, but not at quite the rate of the pure Holsteins. Consequently, the pure Holsteins continued to have a statistically significant advantage for fat (lb) plus protein (lb) production, and the difference from pure Holsteins increased from 9% to 12% for the Normande-Holstein crossbreds, from 5% to 7% for the Montbeliarde-Holstein crossbreds, and from 3% to 6% for the Scandinavian Red-Holstein crossbreds. Table 2. Second lactation production (actual 305-day with 2X milking).

Holstein

Normande-Holstein

Montbeliarde-Holstein

Scandinavian Red-Holstein

Number of cows 285 204 381 243

Milk (lb) 26,194 21,863 ** 23,547 ** 23,683 ** Fat (lb) 941 826 ** 885 ** 894 ** Protein (lb) 817 714 ** 752 ** 762 ** Fat (lb) + Protein (lb) 1758 1540 ** 1637 ** 1656 **

% of Holstein -12% -7% -6%

** Statistically significant difference from pure Holsteins (p<.01).

9

Average production of Swedish Red cows versus Swedish Holsteins in Sweden suggests that the production of Swedish Red-Holstein crossbreds should be very near the production of pure Holsteins if heterosis of 5% for production traits is assumed. Perhaps, less than 5% heterosis for production was realized is this study, because the Swedish Red and Holstein breeds share distant ancestry and because they were developed in the same general region of northern Europe. On the other hand, the Montbeliarde and Holstein breeds share little ancestry, even distantly; therefore, Montbeliarde-Holstein crossbreds might express a higher average level of heterosis than crosses strictly among dairy breeds of the plains or islands of northern Europe, which include Holstein Calving Difficulty and Stillbirths Results for calving ease and stillbirths are the same as previously reported (Heins et. al, 2006a). Calving difficulty was measured on a 1 to 5 scale, with 1 representing a quick and easy birth without assistance and 5 representing an extremely difficult birth that required a mechanical puller. Scores of 1 to 3 were combined and regarded as no calving difficulty, and scores of 4 and 5 were combined and represented calving difficulty. Stillbirths were recorded as alive or dead within 24 hours of birth. Calving difficulty and stillbirth are traits of both the sire and the dam. Table 3 provides the number of births, calving difficulty rate, and stillbirth rate by breed of sire for first-calf pure Holstein dams. Inadequate numbers prevented the evaluation of Normande sires; however, some Brown Swiss semen was used by these dairies. Scandinavian Red sires had both significantly less calving difficulty and significantly less stillbirth than Holstein sires when dams of calves were first-calf pure Holsteins.

Table 3. Calving difficulty and stillbirths for breed of sire for first-calf pure Holstein dams.

Breed of sire

Number of births

Calving difficulty Stillbirth

--------------- (%) --------------- Holstein 371 16.4 15.1 Montbeliarde 158 11.6 12.7 Brown Swiss 209 12.5 ** 11.6 Scandinavian Red 855 5.5 ** 7.7 **

** Significant difference from Holstein sires (p<.01). To estimate differences in breed group of dam for calving difficulty and stillbirths, breeds of sire were limited to Brown Swiss, Montbeliarde, and Scandinavian Red, because numbers of births by sires of other breeds were small and were not well distributed across breed group of dam. Table 4 has number of births, calving difficulty rate, and stillbirth rate for 1,572 first births of cows. All crossbred cow groups had significantly less calving difficulty than pure Holsteins (17.7%) at first calving. Stillbirth rates tended to follow the averages for calving difficulty respective to breed group of dam, and Montbeliarde-Holstein dams (6.2%) and Scandinavian Red-Holstein dams (5.1%) had significantly lower stillbirth rates than pure Holstein dams (14.0%).

10

Table 4. Calving difficulty and stillbirths for breed group of dam at first calving.

Breed of dam

Number of births

Calving difficulty Stillbirth

-------------- (%) --------------Holstein 676 17.7 14.0 Normande-Holstein 262 11.6 * 9.9 Montebeliarde-Holstein 370 7.2 ** 6.2 ** Scandinavian Red-Holstein 264 3.7 ** 5.1 **

* Significant difference of crossbreds from pure Holsteins (p<.05). ** Significant difference of crossbreds from pure Holsteins (p<.01).

Survival First-lactation cows that calved from June 2002 to May 2005 in six of the seven California dairies were compared for survival to 30 days, 150 days, and 305 days post-calving. Because one of the dairies participated in the whole-herd buy-out program (heifers were retained to continue dairying), cows from that dairy were removed from the analysis of survival. Table 5 has the survival rates for 724 pure Holsteins and 1,792 crossbreds. Pure Holsteins left these dairies sooner than all crossbreds groups, with 86% of pure Holsteins surviving 305 days post-calving compared to 93% to 96% of crossbreds. To put this in context, pure Holsteins were 3.5 times more likely to leave these dairies before 305 days after first calving than the Montbeliarde-Holstein crossbreds.

Table 5. Survival rates during first lactation.

Breed

Number of cows 30 days 150 days 305 days

------------------------ (%) ------------------------ Holstein 724 96 93 86 Normande-Holstein 437 98 97 * 94 ** Montbeliarde-Holstein 806 99 97 * 96 ** Scandinavian Red-Holstein 549 98 96 93 **

* Statistically significant difference of crossbreds from pure Holsteins (p<.05). ** Significant difference of crossbreds from pure Holsteins (p<.01).

Cows that had an opportunity to calve a second time (Table 6) were compared for three thresholds for calving interval by breed group – within 14 months of first calving, within 17 months of first calving, and within 20 months of first calving. All crossbred groups had significantly higher percentages of cows calving a second time within the fixed windows of opportunity than the pure Holsteins. From 16% to 20% more crossbred cows calved a second time within 14 months of first calving compared to pure Holsteins. When cows were provided more time to calve a second time (20 months – which is an ideal 12-month calving interval plus an additional 8 months), the difference of the crossbred groups from the pure Holsteins narrowed (10% to 16%); yet, the differences remained substantial and highly significant statistically.

11

Table 6. Percentage of cows that calved a second time after first calving within fixed windows of opportunity.

Breed

Number of cows 14 months 17 months 20 months

------------------------ (%) ------------------------ Holstein 565 44 61 67 Normande-Holstein 392 62 ** 76 ** 79 ** Montbeliarde-Holstein 561 64 ** 78 ** 83 ** Scandinavian Red-Holstein 389 60 ** 73 ** 77 **

** Statistically significant difference of crossbreds from pure Holsteins (p<.01). Fertility Fertility of the pure Holsteins and crossbreds was measured as actual days open for cows that had a subsequent calving or had pregnancy status confirmed by a veterinarian. To be included in the analysis, cows were required to have at least 250 days in milk, which meant the pure Holsteins had an advantage because they were a more highly-selected group compared to the crossbreds – a smaller percentage of pure Holsteins than crossbreds survived to 250 days postpartum. Cows with more than 250 days open had days open set to 250. The 677 pure Holsteins in these dairies had average days open of 156 days (Table 7) during first lactation, and all of the crossbred groups had significantly fewer days open than the pure Holsteins. The difference from the pure Holsteins ranged from 14 days for the 529 Scandinavian Red-Holstein crossbreds to 23 days for the 421 Normande-Holstein crossbreds. These results agree with most other recent research on fertility of pure Holsteins versus F1 crossbreds involving Holstein, which have typically reported two to three weeks fewer days open of crossbreds versus pure Holsteins.

Table 7. Days open during first lactation with a maximum of 250 days.

Breed

Number of cows

Number of sires Days open

Holstein 677 79 156 Normande-Holstein 421 24 133 ** Montbeliarde-Holstein 805 33 137 ** Scandinavian Red-Holstein 529 14 142 **

** Statistically significant difference of crossbreds from pure Holsteins (p<.01) 3-breed versus 2-breed crossbreds All first generation (F1) crossbreds in the seven California dairies are bred to bulls from a third breed; however, these dairies were no longer calving first-lactation pure Holsteins by the time the 3-breed crossbreds began to calve. Therefore, the comparison of 3-breed crossbreds versus contemporary pure Holsteins is not possible in these dairies. On the other hand, comparison of 2-breed and 3-breed crossbreds that calved during the same 4-month herd-year-seasons is possible. Table 8 has first lactation production for the 2-breed versus 3-breed crossbreds. The production of 2-breed and 3-breed crossbreds was very similar, and differences were not statistically significant. A reduced Holstein content might be expected to lower the production capability of 3-breed crossbreds (25% Holstein) versus 2-breed crossbreds (50% Holstein). However, preliminary results comparing 2-

12

breed and 3-breed crossbreds in these seven dairies suggest the production of 3-breed crossbreds is extremely similar to the production of 2-breed crossbreds.

Table 8. Actual 305-day production during first lactation of 3-breed and 2-breed crossbreds. Type of crossbred

Number of cows

Number of sires Milk Fat Protein

Fat plus protein

-------------------- (lb) -----------------------2-breed crossbreds 607 66 20,533 769 661 1430 3-breed crossbreds 173 27 20,258 761 660 1421

No differences were statistically significant.

CROSSBREEDING WITH BROWN SWISS Researchers at Penn State compared Holstein, Brown Swiss, and crosses of Holstein with Brown Swiss for milk, fat, and protein production, somatic cell score, and days open. This study also estimated the effects of heterosis for these traits. Data from this study was collected from 19 herds that were using Dairy Comp 305 or PCDART as management software programs. The results from this study will be published in more detail in the Journal of Dairy Science in 2007 (Dechow et al, 2007). The total number of cows, average milk, fat and protein production, mature equivalent (ME) for milk, fat, and protein production, days open, and somatic cell score and heterosis for Holstein (HO), Brown Swiss (BS), and Brown Swiss-Holstein crossbreds (SH) across all lactation groups are in Table 9. Sires and maternal grandsires of cows were required to have NAAB-assigned sire codes. Pure Holsteins were not different from Brown Swiss-Holstein crossbreds for daily milk production, mature equivalent milk and fat production, and somatic cell score. The crossbreds had higher daily fat and protein production, as well as higher mature equivalent protein production. For fertility, the Brown Swiss-Holstein crossbreds had 11 fewer days open than pure Holsteins. Across lactations, heterosis estimates ranged from 6.7% to 10.4% for daily production traits. For mature equivalent production, heterosis was lower and ranged from 5.6% to 8.5%. For days open, heterosis was 7.3% and was lower than reported in earlier studies of Brown Swiss and Holstein crossbreds (Brandt et. al, 1974; McDowell and McDaniel, 1968) which found heterosis estimates for days open that ranged from 11.6% to 31%. Table 9. Total number of cows, daily and ME production, days open, and heterosis across lactation groups.

N

Milk (lb)

Fat (lb)

Protein (lb)

MEM (lb)

MEF (lb)

MEP (lb)

Days Open SCS

HO 1773 74.3 2.7 2.2 24,747 874 725 156 2.75 BS 805 62.1* 2.5* 2.1* 21,695* 833* 699* 156 2.82 SH 132 73.2 2.9* 2.3* 24,520 915 772* 145* 2.57

HET 6.7% 10.4% 7.1% 5.6% 7.2% 8.5% 7.3% 7.8%

* Statistically significant difference of crossbreds from pure Holsteins (p<.05). Table modified from Dechow et. al (2007)

13

Table 10 has results and heterosis estimates for daily fat plus protein production and days open for Holstein, Brown Swiss, and Brown Swiss-Holstein crossbreds for first, second, and third lactation.

Table 10. First, second, and third lactation results for combined daily fat plus protein production, and days open.

Lactation number 1 2 3 Fat + Protein (lb) HO 4.6 4.9 5.0

BS 4.2* 4.7* 4.9

SH 4.6 5.1 5.6*

Heterosis 4.4% 5.5% 13.4%

Days Open HO 130 152 194

BS 137 143 188

SH 113* 140 188

Heterosis 15.0% 4.9% 1.4%

* Statistically significant difference of crossbreds from pure Holsteins (p<.05). Table modified from Dechow et. al (2007)

The Brown Swiss-Holstein crossbreds were at the same level of fat plus protein production in first and second lactation; however, the Brown Swiss-Holstein crossbreds were higher in third lactation. Heterosis estimates for fat plus protein production were greater in third lactation (13.4%) versus first and second lactation (4.4% and 5.5%), respectively. The pure Holsteins in this study average days open of 130 days during first lactation and the Brown Swiss-Holstein crossbreds had 17 fewer days open than the pure Holsteins. The Brown Swiss-Holstein crossbreds had numerically fewer days open in second (12 days) and third lactation (6 days), respectively. Heterosis estimates for days open were significantly higher for first lactation (15%) compared to second and third lactations. The benefits of crossbreeding with Brown Swiss could be influenced by difficult calf management practices (Dechow et. al, 2007). The common complaint by dairy producers with Brown Swiss calves is their inability to drink from buckets. In spite of this, a crossbreeding system with Brown Swiss does have advantages in calving difficulty and stillbirths. In the California study, Brown Swiss-sired calves born from first-calf Holstein heifers had significantly less calving difficulty and numerically lower stillbirth rates than Holstein-sired calves (Heins et al, 2006a). The results from this Penn State study look promising for the Brown Swiss breed; however, further research should be done to determine if Brown Swiss is a feasible breed for crossbreeding systems.

CROSSBREEDING OUTSIDE OF NORTH AMERICA Crossbreeding of Holstein and Jersey is common in New Zealand, where crossbreds comprise nearly one quarter of milk-recorded cows (Harris, 2000). Crossbreeding has grown substantially in popularity, and numerous studies have been performed to assess the benefits of crossbreeding in pastoral production systems. Ahlborn-Breier and Hohenboken (1991) analyzed New Zealand field data of Holstein, Jersey and various crosses for additive and nonadditive genetic effects for milk production, fat production, and fat percentage. Heterosis of 6.1% for milk production and 7.2% for fat production of crossbreds of Holstein and Jersey compared to the pure breeds was observed.

14

New research from New Zealand (Bryant et al., 2007) reports heterosis for milk, fat, and protein production. Production records from over 180,000 first-calf heifers were used in this analysis. The objective of this study was to determine if environmental conditions in New Zealand influenced the expression of heterosis. Figure 1 has heterosis levels for breed groups for milk, fat, and protein production by varying levels of fat plus protein production. Heterosis was highest for Holstein-New Zealand Jersey crossbreds for milk, fat, and protein production and ranged from 5.0% to 9.5%. The study also reports the largest effects of heterosis occur in herds with intermediate production levels. The study concluded that crossbreds of Jersey and Holstein had higher fat and protein production that pure Holsteins due to the expression of heterosis.

NEW RESEARCH AT UNIVERSITY OF MINNESOTA Crossbreeding with Jersey sires Crossbreeding was initiated in 2000 with two research herds of Holsteins. The herds were the St. Paul campus herd and the herd at the West Central Research and Outreach Center at Morris. The cows at St. Paul are housed in a tie-stall barn and the herd at Morris is managed as a low input grazing system. From 2000 to 2002, one-half of pure Holsteins were bred to Holstein AI sires and one-half to Jersey AI sires. In 2003, the mating system was slightly changed. One-third of pure Holsteins were continued to be bred to Holstein sires and ⅔ were bred to Montbeliarde sires. The F1 Jersey-Holstein crossbreds were bred to Montbeliarde sires as the third breed in the crossbreeding rotation. Jersey-Holstein crossbreds and pure Holsteins that calved for the first time from September, 2003 to May, 2005 were studied for production. A total of 149 cows calved for the first time and had production records during this period were compared for milk, fat, protein, fat plus protein production, and somatic cell score. Results for 305-day actual production for Jersey-Holstein crossbreds versus pure Holsteins during first lactation are in Table 11. Pure Holsteins produced 1230 more pounds of milk than the Jersey-Holstein crossbreds. There was no difference in fat production; however, the pure Holstein had more protein pounds than the Jersey-Holstein crossbreds. Fat (lb) plus protein (lb) was used to gauge the overall

Figure 1. Heterosis from Holstein-New Zealand Jersey crossbreds (blue ◊), New Zealand Jersey-New Zealand Holstein crossbreds (red □), and Holstein-New Zealand Holstein (green ∆) for milk, fat, and protein production in relation to production level of fat plus protein. Figures reproduced from Bryant et. al (2007)

15

production of the pure Holsteins versus crossbreds. The Jersey-Holstein crossbreds (-3%), were significantly lower than the pure Holsteins for fat (lb) plus protein (lb). There was no significant difference in somatic cell score, but the Jersey-Holstein crossbred had numerically higher somatic cell scores.

Table 11. First lactation production (actual 305-day with 2X milking) of pure Holsteins versus Jersey-Holstein crossbreds.

Holstein Jersey-Holstein Difference Number of cows 73 76 Milk (lb) 16,986 15,756 -1230 ** Fat (lb) 610 605 -5 Protein (lb) 524 492 -32 ** Fat (lb) + Protein (lb) 1134 1096 -38 * Somatic cell score 2.95 3.21 +.26

* Statistically significant difference of crossbreds from pure Holsteins (p<.05). ** Statistically significant difference of crossbreds from pure Holsteins (p<.01).

Fertility of the pure Holsteins and Jersey-Holstein crossbreds was measured as first service conception rate and actual days open for cows that had a subsequent calving or had pregnancy status confirmed by a veterinarian. Only 145 cows were compared for fertility because two cows from each breed group were culled before they had an opportunity to be bred. First service conception rates between the pure Holsteins (41%) and Jersey-Holstein crossbreds (39%) were not different. The 71 pure Holsteins had average days open of 150 days (Table 12) during first lactation, and the Jersey-Holstein crossbreds had significantly fewer days open than the pure Holsteins. The difference from the pure Holsteins was 23 days. These results agree with most other recent research on fertility of pure Holsteins versus F1 crossbreds involving Holstein (Heins et al, 2006b), which have typically reported two to three weeks fewer days open of crossbreds versus pure Holsteins.

Table 12. First service conception rate and days open during first lactation.

Breed

Number of cows

First Service Conception Rate

Number of cows Days Open

Holstein 71 41% 67 150 Jersey-Holstein 74 39% 70 127 **

** Statistically significant difference of crossbreds from pure Holsteins (p<.01). Pure Holsteins and Jersey-Holstein crossbreds were also compared for percent pregnant by 90, 120, 150, 180, 210, and 250 days post calving (Table 13). There were no differences between breed groups for percent pregnant by 90 and 120 days; however, more Jersey/Holstein cows were pregnant by 150, 180, and 210 days. By 250 days postpartum, 13% more crossbreds were pregnant than pure Holsteins. Body measurements were also recorded on 145 first-calf heifers. Body weights, hip heights, and heart girths were recorded within 24 hours post-calving. Thurl with, foot angle and length, body condition score, udder clearance and front teat placement were measured within the first 150 days of lactation. Udder clearance was measured from the floor to the bottom on the udder and front teat placement was the distance between the front teats. Table 14 has results of body measurements of pure Holsteins compared to Jersey-Holstein crossbreds.

16

Table 13. Percentage of cows that were confirmed pregnant after first calving within fixed windows of time.

Breed group

Number of cows 90 days 120 days 150 days 180 days 210 days 250 days

--------------------------------------- % ---------------------------------------- Holstein 73 26 51 59 61 64 68 Jersey-Holstein 76 33 57 75* 77* 79* 81

* Statistically significant difference of crossbreds from pure Holsteins (p<.05).

Table 14. Body and udder measurements for first-calf heifers. Holstein Jersey-Holstein Difference Number of cows 73 76 Body weight (lb) 1153 1021 -132 ** Hip Height (in) 56.1 52.6 -3.5 ** Heart Girth (in) 74.7 70.1 -4.6 ** Thurl Width (in) 19.9 18.3 -1.6 ** Body Condition Score 2.71 2.80 +.09 * Foot Angle 44.4 42.6 -1.8 * Foot Length (in) 2.92 2.99 +.07

Udder Clearance (in) 21.5 18.8 -2.7 ** Front Teat Placement (in) 5.5 6.2 +.7 **

* Statistically significant difference of crossbreds from pure Holsteins (p<.05). ** Statistically significant difference of crossbreds from pure Holsteins (p<.01).

The pure Holsteins weighed 132 pounds more than Jersey-Holstein crossbreds immediately after calving. The Holsteins were also taller at the hips, had larger heart girths, and had more width between the thurls. The Jersey-Holstein crossbreds had significantly higher body condition scores, which may indicate why they became pregnant faster than the pure Holsteins. The crossbreds had lower foot angles, but there was no statistical difference in foot length between breed groups. For udder traits, the Jersey-Holstein crossbreds had 2.7 inches less in udder clearance, which indicates that they had more udder depth. The Jersey-Holstein crossbreds also had more distance (+0.7) between the front teats. Many dairy producers in the U.S. are breeding their Holstein virgin heifers to Jersey sires to capitalize on the reduction in calving difficulty and stillbirths. When using Jersey sires, calving difficulty almost disappears. The F1 Jersey-Holstein crossbreds are appreciated; however, in future generations the variation in cows size from the use of Jersey sires might not be very welcome. The current study indicates that the Jersey-Holstein crossbreds are 132 lbs. smaller and have less udder clearance than pure Holsteins; however, the Jersey-Holstein crossbreds had similar fat production to pure Holsteins, had 23 days fewer days open, and were confirmed pregnant quicker than pure Holsteins. Crossbreeding with Montbeliarde sires During the fall of 2005 (October to December), 15 Montbeliarde crossbreds and 12 pure Holstein calved for the first time in the campus herd in St. Paul of the University of Minnesota. Of the original 27 cows, 13 Montbeliarde crossbreds and 10 pure Holsteins have begun their second lactations from September

17

2006 to January 2007. One pure Holstein died and one Montbeliarde-Holstein crossbred was culled for reproduction in first lactation. One pure Holstein and one Montbeliarde-Holstein crossbred are currently dry and will calve again in May 2007. Twenty-three Montbeliarde crossbreds and 14 pure Holsteins calved for the first time during the fall of 2006. Of the 24 Montbeliarde crossbreds, 12 are F1 crosses out of pure Holstein dams, and the other 11 are 3-breed crosses out of Jersey-Holstein crossbred dams. All pure Holsteins and all of the Montbeliarde/ (Jersey-Holstein) crossbreds are bred to Holstein A.I. sires. Also, all of the Montbeliarde-Holstein crossbreds are bred to Jersey A.I. sires Results for 305-day actual production during first lactation are in Table 15. The Montbeliarde-Holstein crossbreds had significantly lower milk, fat, and protein production than pure Holsteins. The Montbeliarde/(Jersey-Holstein) crossbreds were not different from pure Holsteins for production. The Montbeliarde-Holstein crossbreds (-6%) were significantly lower than the pure Holsteins for fat (lb) plus protein (lb); however, the differences were small for the Montbeliarde/(Jersey-Holstein) crossbreds (-2%) compared to the pure Holsteins.

Table 15. First lactation production (actual 305-day with 2X milking).

Holstein Montbeliarde-Holstein

Montbeliarde/ (Jersey-Holstein)

Number of cows 26 26 12 Milk (lb) 20,871 19,341 ** 20,020 Fat (lb) 735 696 * 722 Protein (lb) 646 606 ** 633 Fat (lb) + Protein (lb) 1382 1303 * 1355

% of Holstein -6% -2%

* Statistically significant difference of crossbreds from pure Holsteins (p<.05). ** Statistically significant difference of crossbreds from pure Holsteins (p<.01).

Table 16 has results for production for second lactation. Not all cows in the original data file (Table 15) had an opportunity to calve a second time; consequently, these results are somewhat preliminary. Production of the pure Holsteins and crossbred groups greatly increased from first to second lactation. The difference from pure Holsteins decreased from 6% to 1% for the Montbeliarde-Holstein crossbreds, and increase from -2% to +1% for the Montbeliarde/(Jersey-Holstein) crossbreds. The pure Holsteins (3.05) had numerically higher lactation averages for somatic cell score, but did not differ from the Montbeliarde-Holstein (2.31) or Montbeliarde/(Jersey-Holstein) crossbreds (2.41), respectively. Pure Holsteins and Montbeliarde-Holstein crossbreds were compared for blood levels of circulating progesterone (P4) from the 27th to the 55th day after calving during first lactation (Table 17). The Montbeliarde-Holstein crossbreds tended to have fewer days to a critical level of 1 nanogram of P4, tended to have higher average levels of P4, and tended to have higher peak levels of P4. All of the differences for measures of circulating progesterone (indicating the formation of a corpus luteum) were substantial, but were not statistically significant possibly because of the small sample size.

18

Table 16. Second lactation production (actual 305-day with 2X milking).

Holstein Montbeliarde-Holstein

Montbeliarde/ (Jersey-Holstein)

Number of cows 10 12 1 Milk (lb) 23,684 22,815 22,936 Fat (lb) 832 821 887 Protein (lb) 733 722 757 Fat (lb) + Protein (lb) 1564 1544 1646

% of Holstein -1% +5%

Table 17. Circulating progesterone (P4) of Holstein versus Montbeliarde-Holstein crossbreds during first lactation.

Breed

Number Of cows

Days to 1 nanogram of P4

Average level of P4

Highest peak for P4

Holstein 11 49.9 0.6 1.97 Montbeliarde-Holstein 14 45.9 1.0 3.23

No differences were statistically significant.

CONCLUSIONS Crossbreeding should be regarded as a mating system that complements genetic improvement within breeds. Continuous use of progeny-tested and highly-ranked A.I. bulls is critical to genetic improvement regardless of mating system (purebreeding or crossbreeding). Unfortunately, some dairy producers might interpret the merit of crossbreeding as justification to use natural service bulls rather than A.I. That would be an unfortunate consequence of dairy producers’ interest in crossbreeding. Heterosis is a bonus that dairy producers can expect in addition to the positive effects of individual genes obtained by using top A.I. bulls within breed. The bonus from heterosis is about 5% for production and at least 10% for mortality, fertility, health, and survival, and heterosis comes on top of the average genetic level of the two parent breeds. Therefore, the impact of heterosis on profit should be substantial for commercial milk production. However, some dairy producers might need to get beyond the notion that level of milk production is the only measure of profitability of dairy cows. For the study of seven dairies in California, production of the Montbeliarde-Holstein crossbreds and the Scandinavian Red-Holstein crossbreds was slightly reduced (about 5% for fat plus protein production across the first two lactations) compared to pure Holsteins. Mating Holsteins to Scandinavian Red, Montbeliarde, and Normande A.I. sires resulted in fewer stillborn calves, as well as cows with less calving difficulty, enhanced fertility, and improved survival compared to pure Holsteins. Crossbreeding systems that use Jersey or Brown Swiss have shown to have similar fat production to pure Holsteins; although milk volume is lower in the crossbreds compared to pure Holsteins. The fertility of Jersey or Brown Swiss-Holstein crossbreds demonstrates an advantage of 2 to 3 weeks fewer days open. Heterosis for production ranges from 2 to 15% and heterosis for days open is about 8%. Crossbreeding systems should make use of three breeds. Preliminary results in California and the University of Minnesota show no loss in production by adding a third breed into a crossbreeding system.

19

Individual dairy producers should carefully choose three breeds that are optimum for conditions unique to their dairy operations (facilities, climate, nutritional regime, reproductive status, level of management, and personal preferences).

REFERENCES

Ahlborn-Breier, G. and, W. D. Hohenboken. 1991. Additive and nonadditive genetic effects on milk production in dairy cattle: evidence for major individual heterosis. J. Dairy Sci. 74:592-602.

Brandt, G. W., C. C. Brannon, and W. E. Johnston. 1974. Production of milk and milk constituents by Brown Swiss, Holsteins, and their crossbreds. J. Dairy Sci. 57:1388–1393.

Bryant, J. R., N. Lopez-Villalobos, J. E. Pryce, C. W. Holmes, D. L. Johnson, and D. J. Garrick. 2007.

Short Communication: Effect of Environment on the Expression of Breed and Heterosis Effects for Production Traits. J. Dairy Sci. 90:1548–1553.

Dechow, C. D., G. W. Rogers, J. B. Cooper, M. I. Phelps, and A. L. Mosholder. 2007. Milk, fat, protein, somatic cell score, and days open among Holstein, Brown Swiss, and their crosses. J. Dairy Sci. (Accepted).

Harris, B. L., and E. S. Kolver. 2000. Review of Holsteinization on intensive pastoral dairy farming in New Zealand. J. Dairy Sci. 84(E. Suppl.):E56-E61.

Heins, B. J., L. B. Hansen, A. J. Seykora. 2006a. Calving difficulty and stillbirths of Pure Holsteins versus crossbreds of Holstein with Normande, Montbeliarde, and Scandinavian Red. J. Dairy Sci. 89: 2805 – 2810.

Heins, B. J., L. B. Hansen, A. J. Seykora. 2006b. Fertility and survival of Pure Holsteins versus crossbreds of Holstein with Normande, Montbeliarde, and Scandinavian Red. J. Dairy Sci. 89: 4944 - 4951.

Heins, B. J., L. B. Hansen, A. J. Seykora. 2006c. Production of Pure Holsteins versus crossbreds of Holstein with Normande, Montbeliarde, and Scandinavian Red. J. Dairy Sci. 89: 2799 – 2804.

McAllister, A. J., A. J. Lee, T. R. Batra, C. Y. Lin, G. L. Roy, J. A. Vesely, J. M. Wauthy, and K. A. Winter. 1994. The influence of additive and nonadditive gene action on lifetime yields and profitability of dairy cattle. J. Dairy Sci. 77:2400-2414.

McDowell, R. E., and B. T. McDaniel. 1968. Interbreed matings in dairy cattle. I. Yield traits, feed efficiency, type and rate of milking. J. Dairy Sci. 51:767-777.

Touchberry, R. W. 1992. Crossbreeding effects in dairy cattle: The Illinois Experiment, 1949 to 1969. J. Dairy Sci. 75:640-667.

20

Experience with Crossbreeding – From Headaches to Happiness

Kevin Prins Prins Dairy, Modesto, California

The Prins Dairy, owned by John and Kevin Prins, is located near Oakdale, CA. Our dairy has been in operation since 1971. Today, we milk 570 cows, and cows are grazed in the summer and free stall housed in the winter. All cows are fed a TMR once per day. Prior to 2000, our herd was pure Holstein. For the most part, the Holstein breed served our operation well. However, as the years went by, frustrations began to increase with fertility and calf survival. The list of problems kept growing! Our local A.I. technician, along with a few other dairy producers, came to the conclusion that many of the frustrations were results of inbreeding within the Holstein breed. A short time later, crossbreeding was implemented on our dairy. In the first years, several breeds were used, and none failed. But, as time progressed, some breeds rose to the top as our favorites based on functional type and production. We are now in our seventh year of crossbreeding, and we have had the opportunity to observe improvement in functional traits. Observations 1. Colostrum Prior to crossbreeding, colostrum was fed by a tube feeder. Many of our calves had lost the newborn instinct to suck on a nipple, much less to get up and find the teat on their own. At present, an estimated 75% of our calves nurse on their own. Because the instinct of calves to nurse has improved dramatically, it has become more important than ever to keep our close-up cows clean and dry. Colostrum quality has also improved. We have tested colostrum with the use of a BRIX meter and found that the Holsteins will range from 11-16, while the crossbreds will range from 18-25. It appears that this has nothing to do with the volume of colostrum produced. Crossbreds simply have a higher nutrient density in their colostrum. This could be fueled by a stronger immune system in the cow, perhaps an outcome of hybrid vigor. Is it any wonder why Holstein calves are recommended to be fed as much as 1.5 gallons of colostrum at birth? Our observation is our largest calves desire no more than one gallon. This might suggest that many calves, when force fed, receive more volume than their stomach can actually hold. We are sure we killed calves with the tube feeder in the past. 2. Calving ease and survival During the final years with Holstein calves, we experienced a very high stillbirth rate. This was especially true with first-calf heifers. From a management standpoint, we felt like we were doing a good job watching the close-up cows. Currently, our stillbirth rate is less than 5%. The number of cows that require assistance at calving has also dropped. We are often amazed that even the larger calves require less assistance. Perhaps, they are formed differently than the Holstein calves. It might be selection for Holstein cows to walk “up-hill” has something to do with this. The crossbred cows seem to have more room in the cervix as well. 3. Transition cows Our close-up cows continue to receive a DCAD ration. In the past, there was little room for error in this diet. Today, our crossbred cows transition very well, and we have virtually eliminated milk fever. Our crossbreds haven’t gotten dislocated abomasums. Crossbred cows calve very easily. No post-calving protocol is used.

21

4. Body condition The average body condition score is much, much improved with our crossbreds. In some cases, this is a result of hybrid vigor, while in other cases it is because some breeds simply have more body condition. In general, crossbreds are genetically conditioned, not nutritionally conditioned. Therefore, a crossbred cow that “looks” heavy doesn’t have health problems related to over-conditioning. We find that crossbreds keep much better body condition while they are grazing. Our Holsteins have struggled with body condition, especially while on grass. 5. Heat detection Heat detection has always been done with tail chalk. Over the years, more and more hormone shots had been used to keep groups of cows cycling. As crossbred cows started to enter the herd, fewer and fewer shots were needed. Presently, no shots are used. In fact, many cows come in heat at less than 3 weeks fresh, and cows then cycle at 21-day intervals thereafter. The “extra” heats exhibited in early lactation are a big help in finding heats in general. Because the natural reproductive cycle of cows has returned with crossbreeding, pregnancy checks go very well. Cows are pregnancy checked at 40+ days. Once confirmed, no further confirmation is done. 6. Conception rates In a slow, subtle way, our Holstein conception rates had dropped to under 30%. This, combined with the inability to cycle, caused huge delays in calving intervals. Our calving interval had been more than 14 months; today, with the herd 90% crossbreds, it is 12.8 months. As we started to breed Holstein cows to other breeds, a slight increase in conception was noticed, and these Holstein cows seemed to have a lower rate of early embryonic death, too. A much larger difference was noticed as we started breeding our F-1 cross to a third breed. We now have greater than 45% conception year around. Over all, abortions in our herd are minimal. Reproduction is the number one economic factor on a dairy – not production! If a cow peaks at 140 lbs of milk per day but lacks reproduction, she means nothing to us. 7. Mastitis and SCC Many say that mastitis and somatic cell counts are mostly management related! Yes, management is important, and we are serious about providing a clean environment. However, our data shows that these traits are much controlled by genetics. We have observed a significant reduction in SCC in just two generations of crossbreeding, which is not likely to be possible within a pure breed. We have fewer cases of mastitis. Our crossbred cows seem to respond to mastitis treatment much better than our pure Holsteins. No treatments are used at dry-off. 8. Feet and legs Concrete will always be a challenge for cows. However, we seem to have fewer lame cows among our crossbreds than before with pure Holsteins. The typical hoof seems to be harder for crossbred cows. We select for a straight leg and a steep foot angle. A black hoof is also a desired trait in our program. Unfortunately, we still battle the hairy foot wart. 9. Production At the start of our crossbreeding adventure, we recall saying we were willing to give up some production just to get function back in the cows. To our surprise, production has been maintained very well. In fact, there is no detectable loss in milk volume. The components appear to rise with each generation. Reproduction is a major factor when looking at our daily tank average. The more fresh cows, the more milk!

22

10. Replacements Just a few years ago, the number of cows in our dairy was dropping, as was the number of replacement heifers. Because crossbreeding has enhanced survival and reproduction so much, we have a surplus of cattle in 2007. WARNING: It takes years to climb out of a hole! 11. Identification Identification is important when crossbreeding. Every cow and calf in our operation carries her complete (breed) parentage in her ear tag. We must know her past in order to make the best decisions for her future. This information is also kept in our office. 12. Sire selection We do not use a mating program. Within each breed, we select bulls with positive evaluations for udders. Legs must be straight and the feet steep. Generally, we use the top bulls from each breed. Our Plan As stated earlier, a number of breeds were used for crossbreeding during the first couple of years. Today, our preferred plan is to cross a Holstein cow with a Montbeliarde sire. The F-1 Holstein x Montbeliarde is then bred to Swedish Red. The selection process within both the Montbeliarde and Swedish Red breeds is superior, and the cattle show it. Then, our 3-breed crossbreds are bred to either Holstein (to restart the 3-breed rotation) or to Danish Red (as a 4th breed), and we are doing some of each. Data in the future will tell us which is best. When crossbreeding, the use of three or more breeds is VERY IMPORTANT, as is use of the best sires from each breed. We will NOT consider a two-breed cross. Reflections Looking back over the past seven years, we feel very blessed! We have met many fine people from all over the world because of our switch to crossbreeding, and we have learned much from them. Also, America doesn’t always have all the answers. As our cattle continue to gain “vigor,” our day-to-day quality of life improves, too. We are truly “having fun”. A superb cow is a cow that takes care of her owner. To get many superb cows, we are selecting for the sort of cow we want, and NOT for the sort of cow we don’t want.

23

Genetic Evaluations Using Combined Data from All Breeds and Crossbred Cows

Paul VanRaden and Melvin Tooker

Animal Improvement Programs Laboratory, USDA Beltsville, MD

Summary National genetic evaluation programs were modified to include data from crossbred animals. An all-breed evaluation system was compared with previous within breed evaluations. Genetic differences among breeds seemed to be estimated well, and convergence was fairly rapid, which indicated sufficient within-herd connections among purebred and crossbred groups. Joint evaluation of all breeds and crossbred animals can provide more information but does not greatly change rankings for animals that have herd mates and most relatives from the same breed. Changes were largest for breeds with small populations. Additional herd mates of another breed can add accuracy but can also cause bias if they are managed differently or if genetic effects are not modeled correctly. Breeders can compare breeds and design crossbreeding programs using information from the all-breed evaluation. Implementation is expected in May 2007. Background Genetic evaluations in the US have compared animals within each breed. The USDA animal model is a series of programs used since 1989 that calculates Predicted Transmitting Ability (PTA) by comparing records of cows within the same breed, lactation group (first vs. later), registry status, and management group (two or more month calving period) within each herd and year. With that system, contemporaries groups, pedigree files, and genetic bases were separate for each breed. Holsteins were compared to Holsteins, Jerseys with Jerseys, and so forth. Crossbred animals, unless they were part of a breed association grading up program, were not included. Daughters of other breed sires such as Swedish Red, Montbeliarde, and Normande also were excluded from being evaluated. Rather than exclude these animals from evaluations, the Animal Improvement Programs Laboratory (AIPL) has researched several methods to include all breeds and crossbreds together in routine genetic evaluations. Crossbreeding is of increasing interest to dairy producers and dairy geneticists. The number of first-generation (F1) crossbred dairy cows with usable yield records was about 10,000 in 2001, the latest birth year with complete data. This exceeds the numbers of purebred Brown Swiss, Guernsey, Ayrshire, or Milking Shorthorn cows. Holsteins became popular in many countries because of superior milk production, but some crossbreds have economic merit that is comparable with purebred Holsteins and may exceed Holstein merit if calving ease, calf livability, cow fertility, and cheese yield pricing are considered. AIPL began including data from crossbred and purebred Brown Swiss and Holstein calves in US calving ease evaluations in 2005. Most countries use the animal model for within breed evaluations of dairy animals with the exceptions of New Zealand and the Netherlands where all-breed evaluations are used. An all-breed model has also been used to evaluate US dairy goats since 1988 and beef cattle in Ontario since 1994. Inclusion of data from crossbred animals can lead to more reliable evaluations of purebred relatives, more accurate comparisons of genetic merit among all potential mates, and improved breeding programs that identify the best gene combinations. Goals of this research were to compare methods for evaluating mixed-breed populations and then to apply the best methods for routine evaluation of US dairy cattle.

24

All-breed animal model The all-breed animal model uses lactation records from 1960 to the present, including crossbred records. The total number of cows with records in the national database as of 2006 ranged from 10 to 22 million for milk, fat, protein, somatic cell score (SCS), productive life (PL), and daughter pregnancy rate (DPR). The all-breed model developed was similar to that used for US goat evaluations with the main difference being the computation of general heterosis or hybrid vigor. Estimates of heterosis for individual traits were previously estimated and were not recalculated. Pedigrees for over 46 million dairy cattle were traced to the earliest ancestors recorded electronically, with a lower birth year limit of 1950 because earlier ancestors were not stored. Most animals (99%) had ancestors of only 1 breed, but 431,000 had ancestors of more than 1 breed. Of those, more than 350,000 had breed compositions with less than 94% of 1 breed and greater than 6% of another breed because the crossbreeding occurred within the most recent 4 generations of the pedigree. Beginning in November 2005, the percentage of primary breed was reported for bulls and cows with pedigrees that contain more than 1 breed. Breed composition of the cows with first lactations in 2004 included 90.9% Holsteins, 6.2% Jerseys, 0.8% Brown Swiss, 0.4% Guernseys, 0.3% Ayrshires, <0.1% Milking Shorthorns, 1.2% F1 crossbreds (coefficients of heterosis >50%), and 0.3% backcross cows (coefficients of heterosis >25%). Nearly all F1 cows had Holstein as one parent breed, and contributions from the other breeds were proportional to population size. The number of F1 crossbreds doubled in the last 3 yr. For bulls born since 1997, only 4 Jerseys and 1 Brown Swiss had >25 crossbred daughters, and each of these bulls had >200 purebred daughters. More recently, semen from Scandinavian Red and French breeds was imported and the resulting daughters are nearly all F1 crossbreds. Since 1987, over 5,000 herds had at least 1 crossbred cow, and currently 1,377 herds were coded as mixed-breed herds containing >25% crossbreds or cows of a different breed. Unknown-parent groups in the animal model were separated by breed, pedigree path (dams of cows, sires of cows, and parents of sires), national origin (US or foreign), and birth year. Groups were formed when they included at least 500 animals within a time period and at least 2,000 animals across all years. The grouping pattern was similar to that for Dutch evaluations except that it required only 40 animals per group. Larger numbers are needed for traits with lower heritability. Crossbred ancestors with no records and only one progeny were kept in the relationship matrix and treated as known so that the system of equations could link animals with records back to purebred ancestor groups. Heterogeneous variance adjustments were modified for all-breed analysis of production traits and DPR. For mixed-breed herds, variance within herd would be overestimated if no account were taken of breed differences. Variance adjustments for milk, fat, and protein were previously based on ratios of milk variances, but variances of fat yield were used in the all-breed analysis. Variance adjustments were not used for all-breed PL and SCS evaluations because they had not been used previously in official within-breed evaluations. Data for other breeds were adjusted to make genetic variance equal to Holstein base cows. Age-parity-season adjustment factors have adjusted yield traits to mature equivalence. However, economic comparisons are more precise if records are adjusted to younger or more central ages, because more cows have records at those ages and maturity differences may be inherited. As a result, adjusted yields were lower by about 5% for Guernseys; 10% for Holsteins, Jerseys, and Ayrshires; and 15% for Brown Swiss and Milking Shorthorns. Sire breed was used to adjust crossbred records. Holstein factors were applied if the sire was crossbred or if the cow’s breed was not listed above. Additional age-parity-region-time factors were included in the

25

animal models to account for gradual changes that might occur after the multiplicative pre-adjustments for age-parity were developed in 1995. These were estimated separately in the within-breed animal model from the data for each breed, but were estimated as uniform effects across breeds in the all-breed model. Recent age effects indicated that cows of all breeds are more productive at early ages relative to mature ages when compared with estimates from past decades. Management groups in the within-breed evaluation were separate for registered and grade Holsteins if at least 5 cows of each type were present, whereas cows within the other breeds were grouped together regardless of registry status. In the all-breed evaluation, crossbreds were grouped together with registered or grade cows to allow estimation of breed differences. Crossbred cows sired by Holstein bulls were treated as grades, but all cows sired by bulls of other breeds were treated as purebreds and grouped with purebred cows. For within breed evaluations, heritability of yield traits for Jerseys and Brown Swiss was higher (0.35) than for other breeds (0.30). For the all-breed model, the higher heritability for daughters of Jersey and Brown Swiss sires was accounted for by adjusting their lactation-length weights. Conversions between all-breed and within-breed bases involved both a mean and the standard deviation ratio for traits with variance adjustment that differed by breed:

within-breed PTA = (all-breed PTA − breed mean) (breed SD/Holstein SD); all-breed PTA = (within-breed PTA × Holstein SD/breed SD) + breed mean.

Results Evaluations from an all-breed model can be reported with different genetic base options and including or excluding heterosis. An all-breed base was calculated using the mean of all cows born in 2000. Within-breed bases were calculated from the PTA of cows whose coefficients of heterosis were <50% (i.e., F1 and backcross cows were not included). The PTA for each breed was adjusted to the within-breed base, as is done for goat evaluations and for current dairy cattle evaluations. Evaluations for crossbred animals with breed code XX will be reported on the base of the sire breed, which may cause some confusion because evaluations of animals from reciprocal crosses will be on different bases. All 3 estimates are in Table 1 to provide confidence that estimated breed differences are reasonable. For PL and SCS, current estimates were more similar to phenotypic breed differences than to previous estimates. Reasons may be that previous PL estimates were based on a previous definition of PL and cows that were born before 1990. The largest changes in PTA were for bulls and cows with pedigrees that included more than one breed, and reliabilities also increased for those animals. Gains in reliability were small for sires of crossbred cows because most already had hundreds or thousands of purebred daughters. Only 25 Jersey and Brown Swiss bulls born since 1997 had ≥10 crossbred daughters. Because many purebred animals have no crossbred progeny, changes in their PTA might also be explained by changes in the grouping of unknown dams and the addition of other breeds and crossbred cows to the management groups in mixed-breed herds. Those additional herd mates should increase accuracy but might also cause some bias if management of different breeds is not the same within herd. Genetic trend Genetic trends for each breed and trait in the all-breed system are presented in Figures 1 to 6. Three trend validation tests were performed for each of 5 breeds (excluding Milking Shorthorn) and 5 traits. Interbull requires trend tests to be within 2 standard errors of 0.01 genetic standard deviations per year. Few biases were detected. For 64 of the 70 tests, 95% confidence intervals included the range of −0.01 to +0.01 genetic standard deviations per year. Trends were then converted back to within-breed scales and

26

compared with the previous official estimates. Most changes in Holstein trends were accounted for by a coding error in the within-breed animal model that was corrected during development of the all-breed software. This affected all traits except productive life. Estimates of trends in the other breeds were also improved by proper accounting for crossbred animals that had been treated as purebred animals in the within-breed model. Most trend estimates changed by < 0.01 genetic standard deviations per year. Changes for SCS and DPR seem large relative to trends because all breeds had small SCS and DPR trends during the last 10 years. Brown Swiss had the largest changes in trends, but all new trends seem reasonable. (For a more detailed review of the all-breed animal model, please see “Genetic Evaluations for Mixed-Breed Populations”; P. M. VanRaden, M. E. Tooker, J. B. Cole, G. R. Wiggans, and J. H. Megonigal, Jr.; Journal of Dairy Science, Accepted December 18, 2006. The paper will be available upon publication online: http://aipl.arsusda.gov) Table 1. Breed differences from Holstein base estimated from an all-breed model, previously, and from national phenotypic means adjusted to 36 months of age and previously estimated heterosis.

Difference from Holstein

Source of estimate Breed Milk,

kg Fat, kg

Protein,kg SCS

PL, mo.

Daughter pregnancy

rate, %

Ayrshire −2390 −61 −59 −0.16 0.3 2.4 Brown Swiss −1911 −36 −32 −0.10 0.8 1.1 Guernsey −2776 −37 −62 0.07 −8.5 0.8 Jersey −2962 −34 −47 0.19 3.2 5.5

All-breed EBV

Milking Shorthorn −3230 −111 −90 −0.07 −2.2 4.5

Ayrshire −2118 −54 −53 −0.24 −1.0 1.8 Brown Swiss −1914 −33 −29 −0.14 −0.6 0.2 Guernsey −3014 −46 −70 −0.10 −6.0 2.0 Jersey −3096 −33 −53 0.04 1.6 4.6

Previous1 EBV

Milking Shorthorn −2403 −83 −66 −0.12 −4.8 4.2 Phenotypic difference Ayrshire −2988 −94 −80 −0.11 3.3 0.8

Brown Swiss −2066 −44 −37 −0.15 1.9 −0.6 Guernsey −3305 −65 −81 0.29 −1.3 −1.1 Jersey −3115 −45 −54 0.26 4.9 5.0

Milking Shorthorn −3819 −145 −109 0.12 1.5 3.3 Phenotypic mean Holstein 10480 382 315 3.07 28.1 25.5 Previous heterosis — 317 16 12 0.02 0.3 1.8

27

Figure 1. Genetic trend for milk (kg/lactation) on the all-breed base.

Figure 2. Genetic trend for protein (kg/lactation) on the all-breed base.

Figure 3. Genetic trend for fat (kg/lactation) on the all-breed base.

28

Figure 4. Genetic trend for SCS on the all-breed base.

Figure 5. Genetic trend for productive life (mo) on the all-breed base.

Figure 6. Genetic trend for daughter pregnancy rate (%) on the all-breed base.

29

Crossbreeding – An Important Part of Sustainable Breeding in Dairy Cattle and Possibilities for Implementation

Morten Kargo Sørensen

Danish Cattle Federation and Faculty of Agricultural Sciences, Aarhus University, Denmark

Crossbreeding can improve the profit for most dairy producers if economically similar breeds are used. However, it is important to stress that crossbreeding cannot replace pure breeding. Pure breeding is a prerequisite for crossbreeding. The heterosis obtained from crossbreeding is an added bonus on top of the genetic gain created by pure breeding. The size of the bonus depends on the number and types of breeds involved in the breeding program. Most studies report at least a 10% increase in total economic gain per cow among F1 crosses between “unrelated” breeds. Introduction During the last century, dairy cattle breeding improved markedly. From initially being based on phenotypic selection with very few measurements, dairy cattle breeding now involves high-tech breeding schemes based on extremely large data files. These data files, in combination with optimized breeding schemes based on systematic progeny testing, have increased genetic gain with ever increasing speed. The breeding goal, however, has changed from being primarily focused on milk production and conformation, just a few years ago, to a much broader breeding goal that includes functional traits, such as fertility, health, and calving ease, in most of the western dairy countries. The reason for this change is mostly because of the deterioration of functional traits of cows, which results from the antagonistic genetic correlations between functional and production traits (Rauw et al., 1998; Miglior et al., 2005; Mark, 2004). At the same time, rate of inbreeding has increased within most breeds. Because of the increased need for robust cows in dairy herds with increasing herd sizes, crossbreeding seems very appealing to many. Sustainable breeding The genetic level for numerous functional traits has been reduced within many dairy breeds. Therefore, animal welfare of cows and economics of dairying have been adversely impacted. Long term, the genetic change for the functional traits is not sustainable in regard to economic loss or to animal welfare, because dairy producers are unable to adequately compensate via improved management for the decreased genetic level for the functional traits. Breeding goals and definition of breeding goals are, therefore, very important parts of sustainability for all species of livestock. If the breeding goals for a breed are not defined based on future economic circumstances, then commercial dairy producers will avoid that breed, and the breed will diminish in importance. Therefore, if the aim of a breed is survival, then an economically sustainable breeding goal is essential. In addition to an economic component, the definition of a sustainable breeding goal can include an animal welfare component. The reasons for considering animal welfare are not only based on moral grounds, but also on the assumption that consumers, in the future, will pay more attention to animal welfare issues related to dairy production. Selling products from breeds with sustainable breeding goals will eventually become easier. The extra weight that can be placed on top of purely economic values for some traits is called “non-market” economic weights (Olesen et al., 2000). Another important issue related to sustainable breeding is inbreeding. The breeding programs of dairy breeds have been successful in improving production. The cost has been high rates of inbreeding. With existing pedigree information in the Danish cattle database, the level of inbreeding in Danish Holstein

30

was 3.9% for calves born in 2003 with pedigree completeness in five generations greater than 90% (Sørensen et al., 2005). This level of inbreeding is slightly below corresponding estimates for U.S. Holsteins (VanRaden, 2005). Inbreeding leads to inbreeding depression, to reduced genetic variation, and to higher frequencies of recessive lethal diseases (Kristensen and Sørensen, 2005). For dairy cattle, inbreeding depression has been reported for production traits in several populations (Miglior et al., 1995); however, inbreeding depression also occurs for the functional traits (Smidt et al., 1998; Sørensen et al., 2006). For recessive lethal diseases such as BLAD and CVM of Holsteins, inbreeding increases the negative consequences of these diseases. Reduced genetic variation from inbreeding will result in a future reduction in genetic gain even with the same breeding scheme. Results from simulation analyses show a reduction in genetic gain of 20% over a 25 year period due to reduced genetic variation from increased inbreeding (Sørensen et al., 1999). Therefore, managing the rate of inbreeding clearly is an important part of sustainable breeding. Requirements For sustainable breeding goals, data recording for the traits of interest is required. Within the Nordic countries, data recording for the functional traits, such as veterinary treatments, reproduction, calving ease, and calf size, has been done for the past 25 years (Heringstad et al., 2000). To be usable for calculation of PTA, easy access to the data is essential; therefore, storage of data in a common database is very valuable. In Denmark, recorded data are collected in the Danish cattle database as shown in Figure 1 (Bundgård and Høj, 2000). Quality of data is even more important for calculation of PTA. Dairy producers must appreciate the importance of high quality data. Making dairy producers and veterinarians, who are responsible for data recording, aware of the importance of high-quality data has been a long process in Denmark. For many years, the percentage of Danish herds with valid data recorded for health by veterinarians was approximately 70%; however, within the past few years, this has increased to more than 90% of herds for all breeds. A high percentage of health data must be recorded to achieve acceptable accuracy for PTA. Two methods exist to obtain adequate amounts of data for the functional traits – either deliberate data recording or contracting a large number of herds to do the data recording. The first method is the most efficient, but cooperative thinking among the dairy producers is required. The second method is more expensive, because many dairy producers must be under contract to provide enough data – especially when functional traits are included in the breeding goal. To date, the Nordic countries have used deliberate data recording, and we hope to be able to continue this approach into the future; therefore, improved online forms for data recording have been developed. Data recording for the functional traits is, in itself, not enough. The breeding scheme needs to be optimized in accordance with the breeding goal, which under most circumstances means larger daughter groups than exist today (Christensen, 1998). Another requirement for sustainable breeding schemes is appropriate control of the increase in rate of inbreeding. The rate of inbreeding is greater than 1% per generation in many populations, which has increased the need to monitor the actual rate of inbreeding. Also, tools are needed to control future rates of inbreeding, such as optimal genetic contribution selection within populations (Meuwissen, 1997; Grundy et al., 2000). With dynamic tools for maximising genetic gain, while constraining the future rate of inbreeding (Meuwissen, 1997; Grundy et al., 1998; Meuwissen and Sonesson, 1998; Grundy et al., 2000), the rate of inbreeding can be kept under control by assuring that the parents of future breeding animals are not too closely related. Such methods have been tested in large dairy cattle populations (Weigel and Lin, 2002; Kearney et al., 2004; Sørensen et al., 2007). In Denmark, the computer program referred to as “EVA” (Berg et al., 2006) is used for inbreeding control (Sørensen et al., 2006).

31

Cow data -base

Calvings , Purchase / Culling /Slaughter

Milkrecording ,production

Milk -analysis

AI-serviceMatings ,

Fertility service

ET Slaughterdata

Disease reportsby vets

Disease reportsby farmers

Linearassessment

BreedingEvaluation

Basis forManagementDairy Farm

Health , basis for- Preventive measure

- Package of measures

Feeding plans

Basis forManagement

MilkproductionBreeding plans

- Population- Herd

Identity- Pedigree

Cow database

R&DCow data -base

Calvings , Purchase / Culling /Slaughter

Milkrecording ,production

Milk -analysis

AI-serviceMatings ,

Fertility service

ET Slaughterdata

Disease reportsby vets

Disease reportsby farmers

Linearassessment

BreedingEvaluation

Basis forManagementDairy Farm

Health , basis for- Preventive measure

- Package of measures

Feeding plans

Basis forManagement

MilkproductionBreeding plans

- Population- Herd

Identity- Pedigree

Cow database

R&D

Figure 1. Illustration of the Danish Cattle data base. When genetic gain is the major focus for selection of sires of sons and bull dams, the result should be substantial genetic gain in the next generation; however, an increase in the genetic relationship between the selected young bulls will also likely result. Closer relationships results in more inbreeding in future generations. If a small decrease in genetic gain can be accepted among the selected bulls, then the degree of relationship will be reduced in the next generation. Figure 2 shows the schematic connection between genetic relationship and maximum genetic gain in the next generation. In the short term, a little genetic gain is lost when genetic relationship in the next generation is considered – for example, by choosing strategy B instead of strategy A (Figure 2). By choosing strategy B, more sires of sons will be used, leading to less of an increase in average relationship and to maintenance of more genetic variation. In the long term, it pays to choose strategy B rather than strategy A (Figure 3). Furthermore, a population using strategy A will suffer more inbreeding depression than the population using strategy B. The time when the two lines intersect depends on where B is placed on the curve in Figure 2 – the further to the left B is placed, the later the curves will intersect. Therefore, as more weight is placed on relationship, more time is needed to obtain the genetic gain lost in the short term.

Figure 2. Schematic connection between average genetic relationship and genetic gain.

• A Focus only on genetic gain • B Focus on both genetic gain and average

relationship in next generation

Figure 3. Genetic gain with the two different selection strategies in Figure 2.

Inbreeding / Relationship

Gen

etic

gai

n

B A

Number of generation

Tota

l gen

etic

gai

n AB

32

Possibilities Crossbreeding is another way to increase sustainability for dairy cattle breeding. Inbreeding problems are removed by use of crossbreeding strategies within herds. As mentioned in other presentations in this symposium by Brad Heins and by Kevin Prins, heterosis has substantial impact in dairy cattle. The scientific literature also has many reports on the meaningful influence of heterosis in dairy cattle. For dairy producers who focus on functional traits, crossbreeding is of special interest, because heterosis effects tend to be greater for functional traits, which have low heritability compared to production traits. In addition to heterosis, the degree to which breeds compliment one another needs to be considered to evaluate crossbreeding systems. By choosing breeds for crossbreeding with higher genetic levels for traits of importance than the original breed, rapid improvement might result for these traits. For example, when Nordic Red breeds are used for crossbreeding with Holstein cows, the Nordic Red breeds contribute a higher genetic level for the functional traits. More or less economically equal breeds must be used for crossbreeding; however, with respect to this, the Nordic Red breeds probably are a good fit with Holsteins. A new Swedish investigation, based on economic information from dairy herds, has shown that total economic gain for Swedish Red and Swedish Holstein is equal using Swedish economic values (Lidfeldt, 2006) – Swedish Holsteins had a slightly higher income, and Swedish Reds had slightly lower costs. Genetic gain Many studies on the effects of various breeding schemes have been done, but few with more than a single trait. Therefore, the Ph.D. project, “Stochastic simulation of breeding schemes for dairy cattle”, studied all traits within the Danish breeding goal for dairy cattle (Sørensen et al., 1999). Given Danish circumstances, the simulation study clearly showed the importance of selecting for total merit compared with selection for yield alone (Table 1). Results are in agreement with an Austrian study (Willam et al., 2002). Denmark has used a total merit index since the early 1980s; although, in the early 1990s, the total merit index was expanded to include udder health. Therefore, functional traits of dairy cattle have been genetically improved in Denmark, as well as in the other Nordic countries. Progress for the functional traits has not been as substantial as it could have been in Denmark, and the major reason for this was the use of sire of sons from countries without PTA for the functional traits. The genetic trends for female fertility are in Figures 4 and 5 for the Nordic Holstein and Nordic Red populations in Finland, Denmark and Sweden. The standard error for the index is 10, and 10 index units approximately correspond to 14 days open.

Table 1. Genetic gain ($ per cow per year) by selection for total merit and yield respectively (Sørensen, 1999).

Selection for total merit Selection for yield Production traits 20.0 28.0 Conformation 6.5 -1.0 Functional traits 1) 1.5 -9.0 Total economic gain 28.0 18.0 1) Female fertility, calving ease and health resistance.

For Holsteins, additive genetic level drops substantially for female fertility, which is similar to other Holstein populations. For the Nordic Red populations, additive genetic level has been very stable for female fertility, which is impressive compared to most dairy cattle breeds. The reason for the breed difference is because the Nordic Red breeds actually followed the breeding goals for the “Nordic breeding

33

profile” and used only sires of sons with known genetic level for female fertility. Also, undue emphasis on type traits in Holsteins has been a problem according to Shook (2006). Because of changes in the selection strategy, sires of sons and bull dams within Nordic Holsteins are now selected more in accordance with the total merit index. Therefore, the trend for genetic progress in the future is expected to be more like those in Table 1.

Figure 4. Genetic trend for female fertility among Holstein AI bulls in Denmark, Sweden, and Finland (Nordic breeding value estimation, 2006).

Figure 5. Genetic trend for female fertility among Red AI bulls in Denmark, Sweden, and Finland (Nordic breeding value estimation, 2006).

90

100

110

120

130

140

1981 1983 1985 1987 1989 1991 1993 1995 1997 1999

Birth year for bulls

Index for female fertility

Denmark Sweden Finland

90

100

110

120

130

140

1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Birth year for bulls

Index for female fertility

Denmark SwedeFinland

34

“EVA” results Accounting for inbreeding in breeding schemes results in lower average genetic relationship within the population, which leads to a reduction in future rate of increase in inbreeding. Somewhat more weight on average relationship and somewhat less on genetic merit will result in more dispersion of sires of sons and maternal grandsires. Table 2 has the distribution of sires of sons for Danish Holsteins with different weights on genetic relationship versus genetic merit. Twenty potential sires, 2167 potential dams, 1421 A.I. sires, 754 contract matings, and all pedigrees were analyzed (Sørensen et al., 2006). Table 2. Total merit index for potential sires of sons, their relationship with the populations, and the distribution of sires of sons given different weights on average genetic relationship compared to merit.

Distribution of sires with low (L), high (H)

and only (O) cost on average relation

Sire Total merit

index

Average relationship with

population L H O V Groovy 139 0.095 100 22 4 F Halvor 136 0.101 6 6 3 V Gottorp 136 0.087 59 28 2 T Krarup 135 0.093 - 4 4 L Martin* 135 0.085 10 14 11 D Stilist* 134 0.074 25 38 27 V Eron 134 0.108 - 2 3 R Murphy* 132 0.105 - 3 4 Alves* 131 0.073 - 29 40 T Kargo 130 0.101 - 10 4 RGK Esne 129 0.089 - 6 9 T Katborg 129 0.090 - 1 4 T Audi 129 0.089 - 5 4 Amador* 129 0.096 - 3 5 V Ejlif 128 0.073 - 8 37 Var Gress 128 0.079 - 8 28 K Potter* 128 0.113 - 1 2 Burt* 128 0.102 - 4 3 V Force 128 0.139 - 2 3 H Bo* 127 0.084 - 6 3 * Imported sires of sons. Use of sires of sons is more diverse with less weight on merit compared to average relationship in next generation, and bulls less related to the population are used to a larger extent. Crossbreeding The reason for inclusion of sustainable breeding, including breeding goals and inbreeding, in a paper on crossbreeding is because “healthy” breeding programs within the pure lines are a prerequisite for crossbreeding. If crossbreeding is used in a population at the expense of genetic gain in the pure breeds then, in the long run, crossbreeding will harm the overall economics of milk production. Used properly, heterosis is a bonus on top of the gain from traditional dairy cattle breeding programs.

35

One of the most important things in regard to efficient breeding schemes is size of the test capacity for young bulls. With large numbers of crossbred cows in the population, the test capacity might be reduced if crossbred offspring cannot be used to calculate PTA of A.I. bulls. Methods, such as those presented by Paul VanRaden at this symposium and by Lidauer et al. (2006), which include crossbred cows in the calculate of PTA, must be in place if systematic crossbreeding becomes routine. Systematic crossbreeding could also reduce genetic gain in the pure breeds if the number of purebred cows is reduced, which would lower the selection intensity for bull dams. As long as the proportion of crossbred cows is less than 50%, this should not be a problem. Calculations from New Zealand have shown that the reduction in genetic gain will be 10% for Jerseys and Holsteins in a systematic 3-breed crossbreeding program compared to the present gain, if 90% of the New Zealand dairy producers turn to crossbreeding (Lopez-Villalobos et al., 2000). For Ayrshire, the third breed in New Zealand, an extra genetic gain of 10% results because of more progeny-tested bulls than in the present situation. If new technologies such as genomic selection become important contributors to dairy cattle breeding schemes, the importance of progeny testing and bull dam selection within the whole population will decrease. In that case, the negative side-effects of crossbreeding would be eliminated. The Næsgård experiment Because of knowledge from other species and other experiments with dairy cattle, a large crossbreeding experiment (The Næsgård Experiment) was conducted in Denmark from 1972 to 1985. To my knowledge, this crossbreeding experiment was the largest carried out under research conditions, and three pure breeds were maintained, as well as crosses between the breeds. The breeds were Holstein, original Danish Red, and Ayrshire, and the experiment included more than 3,000 lactations of cows. Unfortunately, the results are published only in Danish (Christensen and Pedersen, 1988). The main conclusions from the experiment were:

"F1 heterosis for total economic merit (expressed per live born female calf from birth to first calving or culling) was 9.9% when estimated by the dominance model. The obtained heterosis by 3-breed rotational crossing estimated by the recombination model was 19.4%. The total merit for cows was expressed per heifer in a 3 year period from first calving. F1- heterosis was 21.2% (dominance model), and the obtained heterosis by 3 breed rotational crossing was 30.4% (recombination model). The estimates for total merit were only slightly dependent on the prices used. A major part of the heterosis for total merit was due to good stay ability and high survival rate of crossbreds. The high survival rate among crossbred cows could not be explained by favorable heterosis for yield, reproduction and resistance to diseases but was rather due to general superiority in constitution (robustness). It was concluded that crossbreeding of dairy cattle breeds can be expected to produce a considerable amount of economic heterosis and that crossbreeding is particularly beneficial in herds with sub optimum environmental conditions”

Based on the results from this experiment, one might have expected large numbers of dairy farmers to start systematic crossbreeding. That was not the case, and some have argued the results from this experiment came 15 years too early, because dairy producers today think more in terms of economics compared to 15 years ago. Approximately 10 to 15 herds that started crossbreeding at that time have continued using systematic crossbreeding. Among those are Ann and Anders Grosen.

36

A case study - long term experience with systematic crossbreeding Ann and Anders Grosen took over their dairy farm in 1991. At that time, the former owner had started crossbreeding with Danish Red sires in the Jersey herd. They have continued crossbreeding and have done so for many years with a 3-breed rotational system using Jersey, Danish Red, and Holstein. In the meanwhile, they have become organic dairy producers with their 110 cows. Their income per cow is well above average for organic dairy producers, and the reasons are good production of cows combined with excellent health of their cattle. The production level is 1,418 lb fat plus protein per cow per year, with average SCC of 296,000 and veterinarian costs per cow per year of $44, which is approximately one half of normal veterinary costs. Furthermore, they have had only five stillborn calves during the past 12 months. One nuisance is variation of cow size with Jersey included in the 3-breed rotational system, which causes some problems in the milking parlor and for stall sizes. Ann and Anders have addressed this by using Jersey bulls with high PTA for body size and Holstein bulls with low PTA for body size.

Danish situation in 2007 The Danish dairy industry consists of 510,000 dairy cows enrolled in milk recording out of a total of approximately 550,000 dairy cows (92% of cows are enrolled in milk recording). Average herd size has increased steadily in recent years. Average herd size in March 2007 was 110.6 cows, and the average annual production of cows enrolled in milk recording was 829 lb fat and 664 lb protein. The breed distribution is 73% Holsteins, 12% Jerseys, 8% Danish Red, and the remaining 7% are crosses among dairy breeds; however, most of the crossbreds are the result of grade-up programs or unsystematic crossbreeding. During the late 1990s and early years of this decade, crossbreeding was discussed very little; however, increased concern about inbreeding has resulted in more discussion of crossbreeding. In 2004, crossbreeding was a theme at the Danish Cattle Congress, and many dairy farmers became interested. Dansire, the Danish A.I. company, decided to actively provide advice on crossbreeding programs. Two of Dansire’s breeding advisors were appointed as crossbreeding specialists, and a crossbreeding support group was established. Furthermore, informational material was produced, which stated heterosis effects are specific to breed, meaning that crosses of two breeds might not contribute the same heterosis as crossings of two other breeds. The general rule is crossing breeds that are more genetically distant results in higher heterosis. Nevertheless, Dansire provided general estimates of heterosis for important traits. The estimates, in Table 3, are based on the scientific literature available in 2004, and effects of epistasis were ignored.

Table 3. General estimates for heterosis effects for important traits within dairy cattle production.

Obtained heterosis

Trait F1-animals

2-breed crisscross at equilibrium

3-breed rotational at equilibrium

Production traits Ca. 3% Ca. 2% ca. 2,5% Fertility Ca. 10% ca. 7% ca. 9% Calving ease (maternal) Ca. 15% ca. 10% ca. 13% Stillbirth (maternal) 2-3% 1-2% 2-3% Longevity 10 –15% 7-10% 9-13% Income per cow per year Min. 10% Min. 7% Min. 9%

37

More dairy producers have started systematic crossbreeding in Denmark because of the impact of the two breeding advisors at Dansire, who advise 72 herds that use systematic crossbreeding on at least some of their cows. Beyond that, a few herds are using breeding advisors other than from Dansire to implement crossbreeding systems. The average herd size of herds using crossbreeding is approximately 150 cows. A common argument is more heterosis is realized for herds with low management level. However, new research from New Zeeland reports largest effects of heterosis in herds with average production (Bryant et al., 2007). Breeds The actual breeds selected for crossbreeding are critically important for the success of systematic crossbreeding. In Denmark, most dairy producers initiating systematic crossbreeding have Holstein herds of cows; however, a few Danish Red and Jersey herds have also started crossbreeding. A 3-breed rotational system for crossbreeding is recommended. For Holstein herds, the first breed of sire to use in a 3-breed rotational system, obviously, is one of the Nordic Red breeds. The Nordic Red breeds are Swedish Red (SRB), Norwegian Red (NRF), Finish Ayrshire (FAY), and Danish Red (RDM). The approximate population sizes of the breeds are 160,000; 270,000; 210,000; and 45,000, respectively (Sonesson, 2005). The breeds are related to some extent, because of substantial semen exchange. Somewhat different breeding strategies have been implemented for each of the four breeds, which has resulted in slightly different genetic levels for the various traits. These differences will be reduced in the future, because of increased cooperation among the breeds. Danish informational material reviews the strengths of the different populations of most interest for Danish dairy farmers, and a summary is in Table 4.

Table 4. Possible breeds for crossbreeding, and the more stars – the better.

Breed Mil

k pr

oduc

tion

Mil

k

cont

ent

Mea

t pr

oduc

tion

Size

Fee

t an

d

legs

Udd

er

Mil

king

sp

eed

Still

bir

th

Fer

tili

ty

Hea

lth

Holstein ***** ** ** ***** *** **** *** ** ** ** Jersey *** ***** * * **** **** *** *** *** ** Nordic Red: Danish Red *** *** *** **** **** *** ** *** *** *** Swedish Red **** *** ** *** *** ** *** **** **** **** Norwegian Red ** *** **** *** *** ** *** **** **** **** Finnish Ayrshire **** *** ** ** ** ** *** **** **** **** Montbéliarde *** *** ***** ***** - - - - - - The stars in Table 4 say nothing about economic profit using one or another breed, but there is little reason to expect large differences for total profit from these breeds (e.g., Lidfeldt, 2006). In Denmark, Swedish Red sires are used most often for crossbreeding on Holstein females. Danish Red sires with little or no Holstein genes are, of course, also used on Holsteins. In the future, the percentage of Holstein genes in the Danish Red breed will be reduced. As a third breed, Jersey is an obvious choice for Danish dairy producers, because Denmark progeny tests more than 60 Jersey young sires each year. However, dairy producers need to be aware of the variation of cow size and milk composition that results

38

when Jersey is used in a 3-breed rotation; therefore, the Montbéliarde breed is preferred by some Danish dairy producers over the Jersey breed. General dairy farmer opinions As a consequence of the growing interest in crossbreeding in Denmark, a survey was conducted to assess dairy producer attitudes toward crossbreeding (Laursen, 2005). The survey was send to 475 dairy producers, representing nearly 10% of Danish dairy producers. Sixty percent of the dairy producers responded to the survey, which is a very high response rate compared to the survey on crossbreeding by Weigel and Barless (2003) in the U.S. Forty percent of respondents revealed a positive attitude toward crossbreeding, while 34% did not like crossbreeding. Also, 24% thought crossbreeding was a potentially rewarding breeding strategy, but 40% would not consider using it. The future In both the U.S. and Europe, dairy herds are increasing in size, and dairy producers spend less time with each cow. Therefore, robust cows will probably become more important in the future. Because of the increased focus on functional traits in dairy cattle breeding, systematic crossbreeding of dairy cattle is expected to increase substantially in the future. The introduction of sexed semen could accelerate this trend, because sexed semen facilitates breeding schemes with F1 crossbred cows in production. Conclusions Crossbreeding can provide dairy producers increased economic output and improve the welfare of animals at the herd level; therefore, crossbreeding can be an important part of sustainable breeding. However, it must be stressed that crossbreeding is not “the total solution” for herds with low management levels or with fertility problems. On the other hand, crossbreeding can be an important contributor to “the solution”, along with other management tools. Furthermore, crossbreeding is not a substitute for sustainable breeding schemes in the pure breeds, which require broad breeding goals based on quality data recording for the functional traits and proper control of increases in inbreeding. The combination of sustainable breeding within the pure breeds and systematic crossbreeding at the herd level could provide optimal results for commercial milk production. The gain is expressed as more profit for dairy producers and as improved animal welfare for dairy cows. References Berg, P., Nielsen, J., and Sørensen, M.K. 2006. Computing realized and predicting optimal genetic

contributions by EVA. 8th WCGALP, Belo Horizonte, Brazil.

Bryant, J.R., Lopez-Villalobos, N., Pryce, J., Holmes, C.W., Johnson, D.L., and Garrick, D.J., 2007. Short Communication: Effect of Environment on the Expression of Breed and Heterosis Effects for Production Traits. J. Dairy Sci. 90: 1548-1553.

Bundgaard, E. and Høj, S. 2000. Direct access to the Cattle Database with Livestock Registration. Annual report 1999, National Committee on Danish Cattle Husbandry, Aarhus, Denmark.

Christensen, L.G. and Pedersen, J. 1988. Crossbreeding of Dairy Cattle. 650. Beretning fra Statens Husdyrbrugsforsøg. 229 pages. In Danish with English summary and subtitles.

Christensen, L.G. 1998. Possibilities for genetic improvement of disease resistance, functional traits and animal welfare. Acta Agric. Scand. Suppl. 29:77-89.

Grundy, B.,Villanueva, B., and Wooliams, J.A. 1998. Dynamic selection procedures for constrained inbreeding and their consequences for pedigree development. Genet Res. 72, 159-168.

39

Grundy, B.,Villanueva, B., and Wooliams, J.A. 2000. Dynamic selection for maximizing response with constrained inbreeding in schemes with overlapping generations. Animal Science 70, 373-382.

Heringstad, B., Klemetsdal, G. and Ruane, J. 2000. Selection for mastitis in dairy cattle: a review with focus on the situation in the Nordic countries. Livest. Prod. Sci. 64:95-106.

Kearney, J.F., Wall, E., Villanueva, B., and Coffey, M.P. 2004. Inbreeding Trends and Application of Optimized Selection in the UK Holstein Population. J. Dairy Sci. 87, 3503-3509.

Kristensen, T.N. and Sørensen, A.C. 2005. Inbreeding – lessons from animal breeding, evolutionary biology and conservation genetics. Animal Science, 80:121-133.

Laursen, M.V. 2005. Krydsningsavl i fremtidens kvægbrug (Crossbreeding in future dairy farming). Bachelors thesis, The Danish Agricultural University, Copenhagen. 53 pages, in Danish.

Lidauer, M., Mäntysaari, E.A., Stranden, I., Pösö, J., Pedersen, J., Nielsen, U.S., Johansen, K., Eriksson, J.Å., Madsen, P., and Aamannd, G. 2006. Random Heterosis and Recombination Loss Effects in a Multibreed Evaluation for Nordic Red Dairy Cattle. 8th WCGALP, Belo Horizonte, Brazil.

Lidfeldt, M. 2006. Lönsamhetsjämförelse mellem SRB och SLB. 18 pages. In Swedish.

Lopez-Villalobos, N., Garrick, D.J., Blair, H.T., and Holmes, C.W. 2000b. Possible effects of 25 Years Selection and Crossbreeding on the Genetic Merit and Productivity of New Zealand Dairy Cattle. J. Dairy Sci. 83:154-163.

Mark, T. 2004. Applied genetic evaluations for production and functional traits in dairy cattle. J. Dairy Sci. 87: 2641-2652.

Meuwissen, T.H.E. 1997. Maximizing the response of selection with predefined rate of inbreeding. J. Anim Sci. 75, 934-940.

Meuwissen, T.H.E. and Sonesson, A. 1998. Maximizing the Response of Selection with a Predefined Rate of Inbreeding: Overlapping Generations. J. Anim. Sci. 76, 2575-2583

Miglior, F., Burnside, E.B. and Kennedy, B.W., 1995. Production traits of Holstein cattle: Estimation of nonadditive genetic variance components and inbreeding depression. J. Dairy Sci. 78:1174-1180.

Miglior, F., Muir, B.L., and Van Doormaal, B.J. 2005. Selection indices in Holstein cattle of various countries. J. Dairy Sci. 88:1255-1263.

Olesen, I. Groen, A.F., and Gjerde, B. 2000. Definition of animal breeding goals for sustainable production systems. J. Anim. Sci. 78, 570-582.

Rauw, W. M., Kanis, E., Noordhuizen, E.N., and Grommers, F.J. 1998. Undesirable side effects of selection for high production efficiency in farm animals: a review. Livest. Prod. Sci. 56:13-33.

Shook, G.E. 2006. Major Advances in Determining Appropriate Selection goals. J. Dairy Sci., 89:1349-1361.

Smidt, L.A., Cassell, B.G. and Pearson, R.E. 1998. The effects of inbreeding on the lifetime performance of dairy cattle. J. Dairy Sci., 81:2729-2737.

Sonesson, A.K. 2005. Assessing total profit of alternative levels of co-operation between Nordic cattle populations. Project report “Sustainable Breeding in the Nordic Red Dairy Breeds”. Page 6-22.

Sørensen, A.C., Sørensen, M.K., and Berg, P. 2005. Inbreeding in Danish Dairy Cattle Breeds. J. Dairy Sci. 88, 1865-1872.

Sørensen, A. C., Madsen, P., Sørensen, M.K., and Berg, P. 2006. Udder health Shows Inbreeding Depression in Danish Holstein. J. Dairy Sci. 89:4077-4082.

40

Sørensen, M.K., Berg, P., Jensen, J., and Christensen, L.G., 1999. Stochastic simulation of breeding schemes for total merit in dairy cattle. GIFT Seminar on Genetic Improvement of Functional Traits in Cattle, Wageningen, The Netherlands. Interbull Bulletin No. 23, 183-192.

Sørensen, M.K., Sørensen, A.C., Baumung, R., Borchersen, S., and Berg, P. 2007. Optimal genetic contribution selection in Danish Holstein is very dependent on pedigree quality. Submitted for publication in Livestock Science.

Sørensen, M.K., Sørensen, A.C., Borchersen, S., and Berg, P. 2006. Consequences of using EVA software as a tool for optimal genetic contribution selection in Danish Holstein. Proceeding 01-83 of the 8th World Congress in Genetics Applied in Livestock Production. Belo Horizonte, Brazil.

VanRaden, P.M. 2005. Inbreeding Adjustments and Effects on Genetic Trend Estimates. Interbull Bulletin no. 33, 81-84.

Weigel, K.A. and Barless, K.A. 2003. Results of a producer survey regarding crossbreeding on US dairy farms. J. Dairy Sci. 86: 4148-4154.

Weigel, K.A. and Lin, W. 2002. Controlling Inbreeding by Constraining the Average Relationship Between Parents of Young Bulls Entering AI Progeny Test Programs. J. Dairy Sci. 85, 2376-2383.

Willam, A., Egger Danner, C., Sölkner, J., and Gierzinger, E. 2002. Optimization of progeny testing schemes when functional traits play an important role in the total merit index. Livest. Prod. Sci.