SWINE GENETICS NSIF National Swine Improvement Federation Purdue University Cooperative Extension Service West Lafayette, Indiana Fact Sheet Number 15 NSIF-FS15 Selection Programs for Seedstock Producers Authors: Erik R. Cleveland, University of Hawaii at Hilo; Charles M. Stanislaw, North Carolina State University; and Charles Smith, University of Guelph. Reviewers: Rodger K. Johnson, University of Nebraska; William T. Ahlschwede, University of Nebraska; Lynn A. Cole, Ohio seedstock producer; Allan Schinckel, Purdue University. Introduction One goal for seedstock producers is to maximize genetic progress per year within their herds. Genetic progress is maximized when the following practices are used: 1. Maintaining contemporary groups of adequate size and minimum age range, 2. Using equipment and techniques that allow collection of accurate records on all animals in the herd, 3. Timely processing of records through a program that allows the most accurate evaluation of the animals, 4. Using culling policies to shorten the generation interval to an acceptable value and allow the replacement of breeding animals with superior individuals, 5. Selecting animals based on an index of estimated breeding values (EBVs) or expected progeny differences (EPDs), and 6. Using mating policies that will minimize rates of inbreeding. Many of these concepts have been covered in other National Swine Improvement Federation (NSIF) fact sheets. This fact sheet ties together many of these concepts by providing recommendations for seedstock producers interested in setting up effective selection programs. Topics covered in this fact sheet related to within herd selection include setting up contemporary groups, culling breeding animals, selecting replacements and developing policies to help minimize inbreeding. Other topics include selecting outside breeding stock, developing association breeding programs and planning matings for purchased semen. For discussion related to selecting and culling, it is assumed that producers are using selection indexes based on EBVs or EPDs. When using STAGES or the SWINE-EBV program (as discussed in NSIF fact sheet #12), producers can compare index values for animals from different contemporary groups. With the Nebraska SPF Performance Testing program, the index values are not directly comparable for animals from different groups. Thus, comparisons should only be made within a given contemporary group. Because of the differences among indexes, culling and selection strategies will be discussed separately for the Nebraska program. Any mention of boar or gilt selection in this fact sheet will refer to choosing animals with the highest index scores. However, a certain percentage of higher indexing animals will be eliminated in most groups due to disposition, underlines, feet & legs or other physical problems. Using a large number of physical characteristics as culling criteria, will limit the rate of improvement in performance traits. Thus, producers should cull animals based on only a few traits that have a definite, detrimental effect on animal function. The calculations made in this fact sheet does not take this culling into consideration so estimated and actual selection values can differ somewhat. The actual number of animals to cull or select can vary somewhat from estimates provided by fact sheet equations due to environmental factors. For example, it may be necessary to select a greater number of replacements in the summer due to the effects of heat stress on reproduction. Some seedstock operations consist of both nucleus and multiplier herds. Nucleus herds send semen, off-test animals, and/or cull sires and sows to their multiplier operations for breeding purposes. Several breeding practices can be used to increase the rate of genetic progress in these nucleus-multiplier systems. Seedstock producers are encouraged to use artificial insemination (A.I.) since it allows a greater use of elite sires in both nucleus and multiplier herds. Furthermore, producers should include performance data from both nucleus and multiplier herds in across-herd genetic evaluations to obtain more accurate genetic merit estimates. Contemporary Group In setting up a within herd selection program, it is important to decide which group of animals will make up the contemporary group. A contemporary group should consist of litters that have farrowed during a period of 3 weeks or less to reduce environmental differences within the set of pigs. Ideally, each contemporary group should consist of at least 15 to 20 litters to allow for more meaningful comparisons. Smaller producers should aim for at least 10 litters from at least 3 sires in each contemporary group. Two examples will illustrate how to set up contemporary groups. These examples are based on a 90% farrowing rate. Example #1. In this herd about 28 sows are bred every month. To reduce the time period within a farrowing group, each month's breeding period will be limited to 14 days. From each breeding, approximately 25 sows will farrow. Thus, each contemporary group will consist of the 25 litters that farrowed during the 14 day period. Example #2. In this operation the producer breeds 17 sows every week with about 15 litters farrowing. To meet the stated guidelines, the producer decides that a contemporary group will consist of 30 litters that are born during the two week period. Culling Sows STAGES and SWINE-EBV. For each contemporary group of sows, the producer should keep a list of animal ear notches and index values. Notations can be made on the list for those sows that will be culled due to health, reproductive or feet & leg problems. After weaning, these individuals will be culled while the remaining sows in the group will be bred. During the assigned breeding period, gilts will be bred as replacements for the cull sows. If additional gilts can be bred during this period, they will serve as candidates to replace the lower indexing sows. To serve as a replacement in the group, a gilt must have a higher index score than the sow that will be culled. After pregnancy testing the bred gilts, the low indexing sows can be removed from the herd. Nebraska SPF. With this program the index values for sows and gilts are not directly comparable so some type of culling policy is needed to help achieve an acceptable female generation interval. In some herds a policy may not be needed if heavy culling of sows is already occurring. Let's calculate the female generation interval and parity distribution of sows in a contemporary group with four culling policies that might be used by herds enrolled in the Nebraska program. The culling policies and calculations were based on the assumptions of Johnson (1990): 1. 90% of selected gilts are cyclic and will express estrus, 2. gilts are bred during a 14 day period so 2/3 of the cycling gilts will mate, 3. 90% of all bred females will farrow, and 4. 80% of the sows that farrow will be mated again. ____________________________________________________________________________ Table 1. Distribution of sows by parity within a contemporary group and generation intervals for different culling policies. ____________________________________________________________________________ Sow Culling Policy % of Sows in each Parity Generation ____________________________________________________________________________ 1 2 3 4 5 Interval 1. 20% parity 1 all parity 2 58 42 1.17 2. 20% parity 1 & 2 all parity 3 45 32 23 1.32 3. 20% parity 1, 2 & 3 all parity 4 38 28 20 14 1.44 4. 20% parity 1, 2, 3 & 4 all parity 5 35 25 18 13 9 1.56 ____________________________________________________________________________ A brief description of the culling policies and the results of the calculations are shown in Table 1. For each system 20% of the sows in assigned parities are culled between farrowing and breeding. These systems differs in which parity classes this 20% culling is used. For example, in System #1 the 20% culling applies to only parity 1 sows. However, in System #3 the 20% value applies to parities 1, 2, and 3 sows. Keep in mind that the 20% value is a suggested goal but the actual culling percentage can vary somewhat. The four systems also differ in regard to the final parity in which all remaining sows are culled. For example, in System #1 all sows are culled after farrowing their second litter. In System #3 all sows are culled after farrowing their fourth litter. Based on Table 1 it is possible to make some comparisons among the four systems. System #1 results in the shortest female generation interval and the lowest selection intensity due to the numbers of gilts that are needed to replace the large number of cull sows. With this system one can expect to wean less pigs per sow due to the high percentage of first parity females. System #4 results in the longest generation interval and the highest selection intensity. From a practical standpoint, System #3 is the most attractive. System #3 produces a female generation interval of less than 1.5 years and a relatively high selection intensity as shown by examples in the section on gilt selection. Selecting Gilts STAGES and SWINE-EBV. Producers should select a certain percentage (such as 20%) of the higher indexing gilts from each contemporary group. These females will go into the gilt pool as candidates for breeding. Within the pool, gilts showing estrus during an assigned breeding period can be bred to serve as replacements for cull sows in the group. For further details on this topic, see the section on sow culling. Nebraska SPF. The number of gilts that should be selected from each contemporary group will depend upon the sow culling policy and the length of the breeding interval. Based on the four sow culling policies listed in table 1, the estimated number of gilts to select from each contemporary group can be calculated by using the following equation: NG = (CP1 x NFB / MR) / (BID / 21) where: NG = no. of gilts selected from a contemporary group, CP1 = coefficient for parity 1 sows (Table 1), NFB = no. of females normally bred within a contemporary group, MR = the coefficient for proportion of selected gilts that are cyclic and will mate (examples will use .90), and BID = the breeding interval in days for the contemporary group. Two examples will illustrate the use of this equation. Example #1. For this herd about 28 females are bred during a 14 day period for a contemporary group. About 25 litters farrow in each group. Based on culling policy #3 the producer will sell all sows after their fourth litter while 20% of the females are culled between farrowing and breeding in each of the other parity classes (1 to 3). With this system the coefficient for parity 1 sows is #3). The number of gilts that should be selected from each contemporary group is: NG = (.38 x 28 / .9) / (14 / 21) = 18 gilts. Thus, the 18 top indexing gilts from each contemporary group should be selected. If 18 top animals are selected from 100 gilts that may be tested within this contemporary group, the selection rate will be 18%. Obviously additional gilts will be needed if some animals are culled due to health, reproductive or feet & leg problems. Producers should aim for a gilt selection rate of less than 25% in each contemporary group. Example #2. The contemporary group consists of 34 females bred during a 14 day period out of which 30 litters farrow. Since culling policy #3 is used, the coefficient for parity 1 sows is .38 (Table 1). The number of gilts that should be selected from each contemporary group is: NG = (.38 x 34 / .9) / (14 / 21) = 22 gilts. Thus, the two top indexing gilts are selected from each contemporary group. If two top animals are selected from 120 gilts that may be tested within this contemporary group, the selection rate will be 18%. Culling Sires STAGES and SWINE-EBV. For each contemporary group, compare the index values of young off-test boars to those of current sires. To serve as a replacement, a boar must have a higher index score than the sire that will be culled. Once the new boar is ready for breeding, the lower indexing sire can be remove from the herd. Nebraska SPF. Since index values for sires and young boars are not directly comparable, a culling policy must be developed that is easy to use, allows for equal use of boars and results in a rapid turnover of sires. To meet these objectives, producers should cull sires based on the number of females bred. Thus, each sire would be culled after reaching a set breeding quota. If a producer selects a quota of 20, then a sire would be culled after the 20th female is bred or inseminated. A quota between 15 to 25 bred females per sire seems reasonable for most herds. However, some herds may want to use even lower quotas (such as 10 to 15 females per boar) to help increase the number of sires used during the year. With larger numbers of sires, one can expect lower rates of inbreeding in a closed system. Number of Sires per Contemporary Group Research has shown that at least three sires should be used in each contemporary group. The number of sires needed per group is also affected by the number of matings that must be performed, the type of breeding system used (hand mating versus A.I.) and by the length of the breeding period. The estimated number of sires needed in each contemporary group can be calculated by using the following equation: NBC = ((NSC x NMS) / BIW) / NMBW where: NBC = number of sires per contemporary group, NSC = number of sows bred per contemporary group, NMS = number of matings per sow, BIW = breeding interval in weeks, and NMBW = number of matings (or inseminations) allowed per boar per week. An example will illustrate the use of this equation. Assume the herd under consideration has 28 sows bred in each contemporary group. The breeding interval of each group is two weeks. Assume that each sow is bred twice and that each boar is allowed four matings per week. The number of sires needed in each contemporary group is: NBC = ((28 x 2) / 2) / 4 = 7 boars. Number of Boars to Select per Contemporary Group STAGES and SWINE-EBV. For herds using these programs, producers do not need to select a set number of boars from each contemporary group. From each group producers should select boars that excel in index scores compared to current sires. In other words, a young boar must have a higher index score than the sire that will be culled. Nebraska SPF. The number of boars to select will depend upon the number of sires used and the number of contemporary groups bred during the year. First, the estimated number of sires used in a year can be calculated based on the following equation: NBY = (NCG x NSC) / BQ where: NBY = number of sires used in a year, NCG = number of contemporary groups bred in a year, NSC = number of sows bred in a contemporary group, and BQ = breeding quota (number of bred females allowed per boar). The following equation can then be used to determine the number of boars to select from each contemporary group: NBS = NBY / NCG where: NBS = number of boars to select from each contemporary group, NBY = number of sires used in a year, and NCG = number of contemporary groups bred in a year. Two examples will illustrate the use of these equations. In both examples, each boar is culled after breeding 20 females. Example #1. For this herd 12 contemporary groups of 28 sows each are bred during the year. The number of boars used during a year is: NBY = (12 x 28) / 20 = 17. Now it is possible to calculate the number of boars to select from each contemporary group as: NBS = 17 / 12 = 1.4 or 1 to 2. Thus, one or two top indexing boars should be selected from each contemporary group. If one or two top animals are selected from 100 boars that may be tested within this contemporary group, the selection rate will vary from 1 to 2%. A boar selection rate of less than 5% within each contemporary group is recommended. Example #2. In this herd 24 contemporary groups of 34 sows each are bred during the year. The number of sires used in a year is calculated as: NBY = (24 x 34) / 20 = 41. Next, the number of boars selected per contemporary group is calculated as: NBS = 41 / 24 = 2. Thus, the two top indexing boars would be selected from each contemporary group. If two top animals are selected from 120 boars that may be tested within a contemporary group, the selection rate will be 2%. Minimizing Inbreeding in Closed Herds All seedstock herds should use within herd selection of boars and gilts to achieve genetic improvement. Within herd selection is the key to effective seedstock improvement programs. However, some inbreeding will eventually occur when only within herd selection is used. High rates of inbreeding can easily offset the progress possible from within herd selection. Thus, seedstock producers need to have an understanding of what factors affect the rate of inbreeding so they can take the appropriate action if they decide to close their herds. In a closed herd the genetic relationship among sires is important. If all selected replacement boars are sired by one individual, the herd can expect a more rapid increase in inbreeding buildup. Consequently, it is important to start a closed herd breeding program with as many unrelated sires as possible. For example, it would be desirable to use at least 15 unrelated sires during the year prior to closing the herd. Once the herd is closed to outside sires, the number of replacement boars selected each year will become a very important factor affecting the rate of inbreeding. The greater the number of sires, the lower the rate of inbreeding. It would be desirable to use at least 15 sires each year in a herd. Another factor that affects the rate of inbreeding, but to a lesser extent, is the size of the sow herd. The expected rate of inbreeding is lower in operations with larger sow numbers. The relationship among mates is another factor that affects the rate of inbreeding. Mating closely related animals results in a greater degree of inbreeding. To help minimize inbreeding in a closed herd, it would be beneficial for a seedstock producer to use computer software that calculates the relationship coefficients among all selected animals. Based on these values, one could plan matings among individuals that have smaller relationship coefficients. Small and moderate size seedstock herds can expect faster rates of inbreeding when they adopt a closed herd selection program as compared to large operations. Thus, these herds may need to use some outside breeding to help minimize inbreeding. One method of obtaining outside breeding is to purchase semen or individual animals. Another method is to participate in an association breeding program. Association Breeding Programs To minimize inbreeding and to maximize the rate of genetic change in operations using within herd selection, seedstock producers should consider forming or participating in an association breeding program. In other countries these associations are called cooperative or group breeding programs. Herds participating in an association should be involved in intense within herd selection based on the same selection index. Involvement in an across-herd genetic evaluation program is highly desirable for participating herds to allow the comparison of animals at different farms. Breeding associations can consist of any number of herds. The number of herds involved in the association is of less importance compared to the total number of sows in the system. Associations with more sows can support greater number of sires which will help minimize inbreeding in the system. For this reason it is desirable that an association breeding program contain at least 200 sows. A variety of associations could be developed where semen, boars, gilts and/or sows are exchanged among herds. From a health standpoint, it would be desirable if only semen or lab pigs were transferred among herds. Several methods of exchanging semen could be developed but only two examples will be discussed in this fact sheet. In both examples A.I. sires will be selected from herds involved in the association. Association #1. Each herd has set up it's own selection program using the procedures discussed in the previous sections. In this association, hand mating is used for most sows. In a given herd, semen can be collected when boars have rest periods between contemporary groups or after reaching their breeding quotas. Collected semen can be distributed to other herds in the system that have sows ready for insemination. Each herd should use semen from at least one outside boar in each contemporary group. This practice may bring about a small change in the planned boar allotment, selection and culling policies. However, this procedure will help provide links between operations for across-herd evaluations. Association #2. In this association the participating herds use A.I. for most matings. Since A.I. boars are used across herds, a sire list should be developed which might include information such as EBVs (or EPDs), index values and pedigrees. It would be desirable for an association to have at least 15 sires on the list at a given time as a means to help minimize inbreeding in the system. The association should develop a selection policy for boars since these sires will be shared by all herds. For herds using STAGES or SWINE-EBV, selection can be based on index values by comparing the scores of young off-test boars and current sires. If the Nebraska SPF program is used, producers might select the top 1 to 2% of the boars from each contemporary group. If 100 boars are performance tested in a contemporary group, then the one or two top indexing boars would be selected from each set. Next the association should develop a sire culling policy. For herds using STAGES or the SWINE-EBV program, sires can be culled when higher indexing boars are available for breeding. With the Nebraska SPF program, the association should rank animals numerically on the boar list based on their date of entry into the pool of A.I. sires. Older sires would be at the top of the list and young boars would be added to the bottom. The first individual on the list would be the next sire culled once a new service age boar is ready to be added to the listing. With this method the rate of culling is dependent upon the boar selection policy. It would be desirable that each herd inseminate sows with semen from both home-raised and outside sires. Furthermore, semen from at least 3 sires should be used in each contemporary group for accurate evaluations. Selecting Outside Breeding Stock For contemporary groups it is desirable to have at least 15 to 20 litters born within a time interval of three weeks or less. Small herds should aim for at least 10 litters in a group. With STAGES and SWINE-EBV, low indexing sows can be culled when higher indexing gilts are available. For the Nebraska SPF program, culling 20% of the sows after weaning within parities 1, 2 and 3 plus total culling after the fourth litter is a practical compromise between maximum productivity and minimum generation interval. The culling rate will determine the number of gilts to select from each contemporary group. For STAGES and SWINE-EBV, sires can be culled when higher indexing boars are available. With the Nebraska SPF program, a culling policy should be developed to remove a boar from the herd after approximately 15 to 25 females are bred or inseminated. The culling rate will determine the number of boars to select from each contemporary group. The use of some outside breeding stock warrants consideration by producers since it is a way to help prevent inbreeding and provide links for across-herd genetic evaluations. In closed herds, rates of inbreeding can be minimized by practices that increase the number of sires used yearly. Some of these practices are difficult for smaller seedstock producers to follow. Therefore, they should consider forming a breeding association and exchange semen among herds to help minimize inbreeding and to increase the rate of genetic improvement. Selecting Outside Breeding Stock Using outside breeding has both advantages and disadvantages. Using semen from A.I. centers provides genetic links among operations which is useful in across-herd genetic evaluations. For operations that plan to close their herds to outside introductions within one or two years, the purchase of boars or semen is a way to establish more pedigree lines. With more pedigree lines, one can expect lower future rates of inbreeding. One disadvantage of using outside breeding stock is the risk of introducing diseases into an operation. This disadvantage can be minimized by purchasing semen instead of buying boars. Another disadvantage is the lower selection accuracy compared to within herd selection. If a producer is selecting outside animals, across-herd genetic evaluations should be to improve the selection accuracy. Based on these evaluations, select high ranking sires or their sons. In operations using intense within herd selection, home-raised boars can be expected to be superior to many outside boars that are available for sale. The number of sires needed during a year in the cooperative can estimated from: NBYBC = (TNS x NIS) / ENI where: NBYBC = number of A.I. sires used during the year in the breeding cooperative TNS = total number of sows bred during the year in the cooperative NIS = number of inseminations per sow ENI = expected number of inseminations per boar across all herds. After calculating NBYBC, it is possible to calculate the number of boars that should be selected from each contemporary group as: NBS = NBYBC / NCG where: NBS = number of boars selected from each contemporary group NBYBC = number of A.I. sires used during the year in the breeding cooperative NCG = the number of contemporary groups bred in a year for the cooperative. Let's use the following example to illustrate the use of the two equations. For the example the cooperative has a total of 550 sows bred during the year. Each sow receives two inseminations while each boar provides semen for about 40 inseminations during his allotted stay as an A.I. sire. Assume that the cooperative has 20 contemporary groups of sows bred during the year. The estimated number of A.I. sires needed is: NBYBC = (550 x 2) / 40 = 28. Based this calculation, the number of boars to select from each contemporary group is: NBS = 28 / 20 = 1.4 or 1 to 2 boars. Thus, producers should select 1 to two top indexing boars from each contemporary group. Development of an A.I. boar culling policy requires a different plan than used in a closed herd breeding program. Cooperative herds will differ in contemporary group size and breeding interval. Thus, it is probably not practical to cull animals based on number of inseminations since a given boar might removed before he has a chance to be used in all member herds. Instead, it is more practical for producers to develop a culling policy based on the time period the boar is in the pool of A.I. sires. The cooperative might choose a maximum time period of 1 to 4 months. For this system to work, producers could develop a list of A.I. boars with pertinent data including index values and date of entry into the pool of sires. Each sire would be culled after reach the specified time limit. When a new boar is ready for service, it would be entered on the list and would replace the individual that had been used for the longest period as a sire. Summary In setting up a breeding program, producers should decide what will make up a contemporary group. A contemporary group should consist of at least 20 littters that are born during a time interval of three weeks or less. The producer must develop a sow culling policy. We suggest culling 20% of the sows in each parity class (1 to 3) between farrowing and breeding. With this policy, all sows would be culled after farrowing their fourth litter. The number of gilts to be selected from each contemporary group can be determined based on an equation provided in the factsheet. For within herd selection, the boar culling policy should be based on a quota of no more that 30 to 50 matings per sire. The number of sires and selected boars per contemporary group can be determined based on provided equations. For a within herd breeding program, at least 15 sires should be used per year to help minimize rates of inbreeding. Small seedstock producers should consider forming a breeding cooperative with the exchange of semen among herds to help minimize inbreeding and to increase the rate of genetic improvement. References Johnson, R.K. 1990. Closed herd breeding programs. Record of Proceedings North American Swine Improvement Conference. 15:55. NEW 12/92 (3M) ______________________________________________________________________ Cooperative Extension Work in Agriculture and Home Economics, State of Indiana, Purdue University and U.S. Department of Agriculture Cooperating. H.A. Wadsworth, Director, West Lafayette, IN. Issued in furtherance of the Acts of May 8 and June 30, 1914. It is the policy of the Cooperative Extension Service of Purdue University that all persons shall have equal opportunity and access to our programs and facilities. .