THE PRESENT STATUS OF THE SWINE INDUSTRY OF THE UNITED STATES AND STRATEGIES FOR FUTURE IMPROVEMENT

 

L. L. Christian

Iowa State University

Ames, Iowa,U.S.A.

 

Introduction

 

The pork industry of the United States continues to change, but at a rata that appears to be increasing.  The number of farms that raise pigs has declined dramatically in the past 20 years. Nationally, in 1991 there were only 30% as many farms with hogs as there were in 1970 (Figure 1).  In the midwestern states and in Iowa in particular, the traditional center of hog production, has fared better than the national average but have lost 60% of their producers over this period.  But, while the number of hog farms has declined, total numbers of hogs produced has remained stable.  The number of hogs per farm has more than doubled in Iowa and the other midwestern states.  In the state of North Carolina on the eastern seaboard an eight-fold increase in number of hogs per farm has occurred over this 20-year period.

 

The distribution of hogs per farm has also changed over this time frame.  In 1991 about 904 of U.S. hog farms had less than 500 head inventory.  But, these small farms accounted for only 32% of the national total (Figure 2).  Conversely, farms with over 1000 head inventory represent over 44% of the nation°¶s production and less than 5% of the hog farms.  It is expected that soon to be released 1992 Census of Agriculture data will reveal continued growth of farms marketing over 5000 head and a further Decline in the portion contributed by the small farm group.  It is anticipated that the over 5000 head group will account for over 25% of all the hogs marketed in our country.

 

Why are these changes occurring?  As was noted by Lawrence (1992),it is because the general structure of the marketing channel is changing. In the past farmers finishing pigs also farrowed them, raised replacement gilts and raised corn.  Then, there was a distinct break between production and the packing, processing, wholesale distribution  and retail segments. The industry relied on price signals sent through the various segments of the industry and an adversarial relationship existed between producer and packer.

 

But now the marketing channel is changing to a more totally integrated system where the separate entities are linked together permitting greater flow of information and direct signals between the segments. In many cases the segments are controlled by the same management. Several essentials are necessary for the producer - centered system to persist according to Ginder (1989).

 

He lists these foundation variables to be the ability of independent producers to finance their production,  for them to be the low-cost producers, for them to be able to sell in any market without substantial discounts and for them to have full access to technology. It is apparent that in the U.S. society today these essentials no longer exist to the extent they did in the past.

 

Figure 1

Source:  USDA Hogs & Pigs

 

Percentage Change In Number of

Hog Farms: 1970 = 100

 

Percent of Operations and Inventory

by Farm Size, United States 1991

°@

The Seedstock Industry

The seedstock industry of the U.S. has changed to service the large more specialized units.  Not only have private breeders increased the size of their units but they have taken on more than one breed to serve the entire needs of their commercial customers.  Large corporate breeding companies have developed and/or moved operations to the U.S. to enter this market.  Table 1 presents results of a comprehensive survey conducted by michigan State University in 1988. Approximately 304 of the paternal line seedstock was being supplied by the three major corporate seedstock companies operating'in the country.  Since that time many more have developed and expanded their production. Outside companies entering the market include Cotswold, National Pig Development, Newsham, and Seghers.  Of over 15DO market pigs voluntarily submitted by commercial producers to a state pork producer owned station at New Hampton, IA in 1992-93 approximately 454 were sired by company-produced boars, 384 were sized by purebred boars and 174 by crosses among the pure breeds. Hence, it is apparent that in the U.S. corporate-owned breeding companies are gaxnering an ever increasing share of the seedstock market.

 

Table 1.  Source for Acquiring Replacement Breeding Stock for Paternal (Terminal) Linesa

No. Head

Produced/year

 

PIC

Farmers

Hybrid

 

DeKalb

Babcock

Swine

 

------------------Percent of Market Share------------------

Less than 999

0.2

1.1

0.3

0.0

1,000 to 2,999

2.4

3.7

2.0

0.2

3,000 to 4,999

1.3

0.7

1.3

0.0

5,000 to 9,999

1.2

0.7

2.6

0.0

More than 9,999

2.8

1.2

1.6

0.0

TOTAL

7.8

7.4

7.7

0.2

a1988 MSU Survey of Swine Breeding Systems in the U.S.

 

Regardless of who supplies the seedstock t.o the coaaaarcial sector, the basic concepts necessary to create genetic change at the production level are the same. While modern techniques of genetic engineering may someday alter the genome used by the pork industry  in  a  practical  way,  there  are  presently  only  two approaches available to the pork producer to bring about genetic change--selection and choice of breeding system.  This discussion vill focus on each of these as they are being used to accomplish genetic improvement of swine in the United States.

 

Relative Importanae of Traits

 

Careful examination of the performance traits test exert the greatest impact on praduction costs of U.S. producers (Table 1) reveal that reproduction and maternal performance traits should be of greatest concern.  Carcass premiums offered by some companies for  laaner  pigs  have  increased  in  recent  years  and  have significantly  increased  the  relative  economic  value  for  body composition.  The growth rate traits of average daily gain and days to 105 kg have the least direct effect on net income.

 

The economic values used in this table are for a full unit of change, either one pig per litter, one kg of litter weaning weight.', one day change in days to 105 kg, one mm change in backfat probe, one kg per kg of gain change in feed efficiency and one kg per day change in daily gain.   These values multiplied by the standard deviation put the economic value for each trait on a weighted basis.

 

Table 2. Parameter estimates and economic values for swine traits. (U.S. dollars)

 

 

Trait

 

 

Heritability

 

Standard Deviation

 

Economic Value/unit

 

Economic Value

Relative

Economic Value

Number born alive (L)

.15

2.50

12.00

30.0

15.0

Adjusted 21-day

litter weight (W)

.15

9.00

1.5

13.5

6.7

Days to 105 kg (D)

.35

13.00

-.15

2.2

1.1

Backfat (B)

.40

.50

-6.00

3.0

1.5

Feed efficiency (F)

.30

.25

-13.00

3.2

1.6

Averaige daily gain (G)

.40

.09

22.0

2.0

1.0

 

New Analysis Procedures

 

The Mixed Model Approach

 

The year 1990 will be remembered as a time of major significance to swine seedstock improvement. The first national across herd genetic evaluation was completed for a major swine association, the American Yorkshire Club. This major event will soon be mimicked by all of the major breeds. Through January of 1993 over 60,000 pig records and 110,000 sow records had been processed by the American Yorkshire Club's across herd analysis (table 3). The Landrace breed has compiled its first across herd analysis of both maternal and pig data and in Juae of this year the Hampshire and Duroc breeds will release their first across herd analysis results which will include over 25,000 sow records and 20,000 pig records in each breed.

 

There were at least three major factors that contributed to this important historic development. First was the development of the Best Linear Unbiased Prediction (BLUP) procedure by C. R. Henderson at Cornell University (1973) and it adaptation to swine data. This methodology overcame the problems of unequal numbers per group (herd, sire progeny, etc.) and the difficulties of removing fixed effects (year, season, contemporary group, etc.) from the random genetic effects of concern. This approach permitted improved accuracy of selection by including family and prior statistical information in selection (Gibson and Smith, 1986). A second important development was the availability of new computing and data processing capabilities that reduced the time delay from the completion of the record to evaluation and selection. A third important development was the discovery and adaptation to farm animals of sophisticated ultrasonic procedures which permit accurate assessment of composition in the living animal.

 

BLUP is now recognized as the most effective method of genetic evaluation. The procedure uses not only the individual°¶s own record, but that of all of his relatives to predict an Expected Progeny Difference (EPD) for each trait evaluated. EPD is the predicted difference in performance of future offspring of a sire or dam relative to that of parents of average genetic merit. In short, EPD evaluates according to their genetic value as parents.

 

There are a number of features of BLUP that are responsible for its rating as the best genetic evaluation system. Also, there are a number of sources of bias that can lead to inaccurate genetic evaluations. Differences between herds and groups within a herd in environment, ration content, housing, and other feeding and management procedures will affect the way animals perform. The genetic merit of the animals to which an individual is mated will cause bias. The BLUP statistical procedures simultaneously estimate differences associated with herds and contemporary groups as well as adjust for the genetic merit of mates. The result is an unbiased estimate of the genetic value of sires, dams, and pigs. The multiple trait BLUP procedure used by the Swine Testing and Genetic Evaluation System (STAGES) includes all relatives in each participating herd. Through the use of this informatiqn and the genetic correlations between traits, the EPD is the most accurate possible.

 

Table 3. Yorkshire Across-Herd Analysis Summary, January 1993.

Number of

Growth Data

Maternal Data

Herds

51

664

Sires

952

6200

Active Sires@

773

2458

Dams

6273

23213

Pigs (or Sows)

63777

50686

Sow Records

--

110603

 

Caution must be exercised in conducting these programs in order to prevent bias. Preferential treatment of sire or dam progenies will permit them to appear better than they are genetically. A second concern is honesty and integrity in collecting records. This is of concern since operators may choose to collect records themselves or to have them obtained by farm employees in order to reduce costs. To ensure credibility these records must be obtained carefully, perhaps by special technicians. Premature culling of unpromising young animals of certain sires can also lead to bias. In spite of these limitations, this procedure still excels all other pmeedures used to date for effective genetic evaluation.

 

On-the farm Programs

 

The BLUP procedure adapted for use in swine is called the Swine Testing and Genetic Evaluation System (STAGES). It was developed by Purdue University in conjunction with the USDA, the National Association of Swine Records and others. It is a well developed on-the-orem pxogram that captures°• individual performance and combines it with information on miativea. °¶Ihe STAGE 6 step is now in place and involves across-herd evaluation (Stewart et aL, 1991). The merits of this and other on-the-hrm programs include:

 

1. Whole herd testing is possible.

2. Low test cost per animal.

3. Reproductive and maternal traits can be measured (Sow Productivity Index, SPI).

4. Large numbers of progeny per sire permit accurate sire evaluation.

5. The system permits selection priorities that are unique to the goals of the individual herd.

6. Indexes may be developed for specialized general, maternal or paternal lines.

 

Maternal Traits

 

In the past, it was believed that selection for lowly heritable, sex-limited sow productivity traits would not be effective. A summary of heritabilities of these traits appears in Table 4. Several recent findings, however, give us renewed optimism. Hyperprolific sow line development has shown that where population size is large, selection has been effective. Secondly, theoretical projections of Avalos and Smith (1987) show that annual changes of up to 1/2 pig per litter are possibles through family selection. STAGE 6 uses aU of the family in$wnation in an optimum manner and should be effective for the lowly heritable sow pmdvctivity traits (Belonsky and Kennedy, 1988).


°@

Table 4. Literature averages for heritabilities of reproductive traits.

Traits

No. Observations

Estimate

Age at puberty

8,119

.32

Ovulation rate

6,088

.39

No. born per litter

24,137

.10

No. born alive per litter

138,248

.07

Survival to weaning

78,738

.05

Rebreeding interval

135,569

.06

Litter birth weight

3,955

.29

21-day litter weight

76,335

.15

From: Lamberson (1990).

 

Evaluation of repeated records of the same sow have revealed a high genetic correlation (r=.96) between repeated records for SPI traits (Johansson and Kennedy, 1985), thus permitting the evaluation of these traits as sing1e rather thw multiple traits. In addition, the conclusion that maternal genetic effects are small for sow productivity traits (although slightly negatively associated with direct genetic effects) are highly variable and can be ignored with little loss of information (especially if litter sizes are balanced at birth). This permits a rather simplified approach to genetic evaluation of these traits. However, maternal and environmental factors common to littermates tend to make littermates more alike than expected from their genetic relationship alone (although small for most traits) and is accounted for in the STAGES model for these traits.

 

The STAGE 6 program computes EPDs for the two maternal traits: number born alive per litter (NBA) and 21-day litter weight (L%21). These EPDs are combined into a SPI that is expressed as a deviation from the average of the base population in dollar units. An EPD for NBA is expressed in number of pigs. Daughters of an individual with an EPD of .4 for this trait would be expected to farrow .4 more pigs than the daughter of the average sow. An EPD of +5 for LW21 would mean 5 more lbs of 21-day litter weight were produced by daughters.

 

The SPI is expressed as the sow°¶s EPD for sow reproduction. Former SPI programs were phenotypic values that only measured a sow°¶s individual reproductive records. By induding the genetic performance of relatives, the potential for rapid progress is improved dramatically. The EPD is 1/2 of the sow°¶s genetic superiority above her contemporaries (i.e., the amourit she will transmit on the average to her progeny). For example, if a sow°¶s EPD is 105, each of her daughters would be expected to produce litters worth $5 more at 21 days San those of a sow with an EPD of 100.


°@

Performance and Carcass Traits

 

These traits are of economic importance and have sufficient heritability to respond to selection (Tables 5 and 6). The genetic correlations among them are generally small or strongly favorable. The across herd STAGES program also involves the collection and analysis of on-farm postweaning growth and backfat information. The two traits involved are:

 

Adjusted Days to 105 kg (D105) - A measure of growth rate by using off-test age and weight. The recommended procedures are outlined by the National Swine Improvement Federation (NSIF, Weber, 1987). An individual°¶s own performance and that of his relatives are used to produce his EPD for days to 105 kg, a prediction of how this individual's progeny are expected to perform relative to the average of the genetic base population. For example, an EPD of -6 for D105 indicates that pigs produced by this individual would reach market weight (105 kg) 6 days earlier than those by the average of his breed.

 

Adjusted Backfat (BF) - The backfat of an animal adjusted to 105 kg. This procedure also follows NSIF Guidelines (Weber, l987). Adjusted values are converted to an EPD for backfat which is a measure of the expected superiority (or inferiority) of his progeny relative to that of an average boar. These EPDs are a function not only of the boar°¶s own backfat but alse that of all of his relatives both within and across herds. A boar with an EPD BF of -1.0 would be expected to produce progeny that average 1 mm less in backfat than those from the average boar of the breed.

 

New developments in ultrasonics have increased the accuracy of assessing backfat in live animals. Machines with the third fat layer measurement provision permit the operator to easily detect and accurately measure this additional layer that is commonly present. Real-time machines, though expensive, are extremely accurate in measuring backfat and are an improvement over previous devices for estimating loin muscle area. Correlations between live and carcass values of approximately .90 for backfat and .80 for loin muscle area have been reported with the latest versions of this equipment (Christian and Moeller, 1990).

 

Table 5. Heritabilities of growth, feed and backfat traits and their genetic (6) and phenotypic (P) relationships to percentage lean.

Trait

h2

% Lean (G)

% Lean (P)

ADG

.30

-.15

-.11

Days

.25

.10

.10

Backfat

.41

-.85

-.71

F/G

.30

-.43

-.25

AFI

.24

-.25

-.20

From: Stewart and Schinkel (1990).


°@

Table 6. Heritabilities of carcass characteristics and their genetic (G) and phenotypic (P) relationship to percentage lean.

Trait

h2

% Lean

% Lean

BF10

.52

-.87

-.81

LMA

.47

.65

.62

Dress %

.30

-.10

.00

Length

.56

.18

.10

% Lean

.48

.48

---

Fmm: Stewart and Schinkel (1990).

 

Measurement of feed conversion is considered to be optional in improvement programs because research reports indicate low realized heritability for this trait. Indirect selection for leanness and rapid growth in several studies has resulted in as much or more improvement in feed efficiency than has direct selection <able 7). This is fortunate since Sew breeders, because of the high costs involved, have facilities to measure feed intake on the small, genetically related groups necessary to yield meaningful data for selection.

 

Table 7. Phenotypic correlations of various traits with efficiency of lean gain (ELG).

Trait

ELG

G

.24*

F/G

.66**

% LC

.28**

LC Gain

.46**

*P < 0.05.

**P < 0.01.

 

From: Bereskin and Davey (1976)

 

Selection Indexes

 

Selection indexes reported in STAGES are optimized according to the economic values associated with costs of production and value of market pigs in a typical midwestern farrow-to-finish pork production facility. Four indexes are calculated. SPI (mentioned earlier) ranks individuals on NBA and LW21 only. Terminal Sire Index ranks the animals on D 105 and BF only. The General Purpose Index (GPQ ranks the animals on a combination of growth and maternal traits, and is recommended for breeds and/or lines used in rotational crossbreeding programs. Maternal Line Index (MLQ also utilizes the EPD of both growth and maternal traits but is optimized for a maternal line. All indexes are scaled so that the average parent has a value of 100 and superior parents have values greater than 100. Which index is most appropriate depends on the traits measured and the breeding objective. In most commercial pork production systems following a terminal or rotaterminal system, replacement gilts should be selected on the MLI and replacement boars selected on TSI.

 

Accuracy

 

An accuracy value is listed with each trait in the STAGES output. Accuracies range from .01 gow) to .99 (high). Accuracy is an indicator of precision with which the EPD is estimated (i.e., the level of confidence that the predicted value of the EPD is near the true genetic value of that sire).

 

When accuracy is high, the EPD is reliably predicted. If the sire is used, the breeder can be quite confident that the mean offspring performance will be near the predicted level. If the accuracy is lower, the mean performance of the offspring may vary more from the prediction. It is very important to understand that the error of prediction of EPDs is unbiased. That indicates that the EPD has an equal chance of being either over or under in predicting the actual performance, but on average is correct.

 

If, in deciding between bvo st with ami1ar EPDs, you want high predictability, then select the sire with the highest accuracy. You can be confident that the mean of his offspring will be nearer the estimate than will that of a sire with lower accuracy. In this case, you haved the possibility of selecting a sire whose offspring will perform much poorer than predicted. But, you have also negated the chance of selecting a sire whose offspring will be much better than expected. If, on the other hand, you are willing to take a greater risk in an attempt to select an outstanding sire, then choose the sire with the lower accuracy but higher EPD. This sire has a greater chance of producing offspring that will perform better than expected. Of course, there is an equally greater risk of failure.

 

Selection decisions should be based on the EPDs first, then use accuracy to decide among those sires with simBar EPDs. Accuracy is determined by the amount of information available to predict the EPD and is not affected by the level of performance of those records. (The EPD is the indicator of the level of performance.) As more records on more relatives are available, the accuracy of the prediction increases. But, it is not simply a matter of getting many records. Accuracy is also related to how many different herds, contemporary groups, relatives, and other sires and dams are included in the data set. Therefore, sometimes'sires with fewer offspring wi11 have a higher accuracy because his relatives are spread over more herds, groups and mates. In addition, a major determiner of accuracy is whether the individual itself has a recorded performance or not.

 

Central Testing

 

Since 1956 central testing has been a favorite selection technique used by U.S. breeders. Records have provided a benchmark for assessiqg change over time (much of which, we hope, has been genetic). This approach has aided in identifying superior herds and individuals and has provided the small breeder the opportunity to advertise his stock and assess the relative merit of his animals relative to that of his competition. Stations have provided a "show window" for the swine industry and have contributed greatly to the education of both seedstock and commercial producers concerning genetic principles and breeding programs.

 

Participation and interest in central testing has dwindled somewhat in recent years (table 8). Performance summaries comparing 1991 results with those from 10 years earlier reveal that progress has been made in growth rate and feed conversion but that backfat has remained rather constant (tables 9-12). No single reason is considered to be responsible, but surveys have mrealed a number of possible causes:

1.    Concern that pretest environment iafluences performance on test.

2.    Health concerns reduce the number of commercial producers wiBing to iatmduce animals to their herds from multi-source testing facilities thus buyer demand has been reduced.

3.    Larger sized commexeial herds desire more boars of a given bxeed or line than are available from a single station.

4.    Increased testing costs.

5.    Reduced faith in the accuracy of ultrasonic estimates of carcass traits.

 

A number of possibilities exist for overcoming the above concerns. These include running all pigs through a medicated early weaning program at a common site prior to being placed on test, testing larger groups of contemporary boars of a single breed at one time, and using real-time ultrasonics to evaluate backfat and loin musde area. It is hoped that faith and partici jation in central testing can be restored since it provides a basis for comparisons between herds, something especially important for the small herd, and that adds credibility to the on-the-farm records of participating producers. Recently BLUP procedures have been used to develop EPDs for animals tested in all stations across the U.S. Utilization of this information will improve the accuracy of selection and provide a valuable service to the private breeder in locating genetically superior stock.

 

Because of reduced interest in boar testing, several stations are now devoting some or all of their space to the testing of market pigs. The purpose of some of these tests has been to provide accurate carcass information on sib and progeny groups without the need of commingling the breeding animals themse1ves. Others are designed to simply evaluate the products of various terminal breeding programs. The Pork Challenge Test, conducted in each of the last three years in conjunction with the World Pork Expo, is an example of a large scale test for assessment of terminal market hogs. This test has provided a fair and unbiased comparison of germ plasm sources available to the commercial hog producer of the U.S. (table 10). Not only did this test compare terminal market hogs derived from purebred and corporate seedstock sources for production and quantitative carcass traits, other traits have been compared as well. The cholesterol and lipid content of muscle from these pigs not only differed considerably between the various crosses, but were much lower than values reported earlier. These findings point to possible genetic variation in these qualitative traits and to the nutritious and healthful aspects of lean pork. This test provides a model that could be copied in tests of a national scope to evaluate pure lines. Such a test is now being conducted under the supervision of the National Pork Producers Council.

 

Table 8. Boar Testing Survey Performance Summary

 

Number Tested

1991

1981

Berkshire

95

61

Chester White

143

152

Duroc

883

1,915

Hampshire

564

671

Landrace

215

281

Poland China

38

56

Spotted

177

461

Yorkshire

1,016

1,416

Total or Avg.

3,131

5,013

 

Table 9. Boar Testing Survey Performance Summary

 

Average Daily Gain

1991

1981

Berkshire

981

903

Chester White

981

854

Duroc

1044

953

Hampshire

985

935

Landrace

1003

917

Poland China

922

908

Spotted

958

922

Yorkshire

1035

953

Total or Avg.

1017

944

 

Table 10. Boar Testing Survey Performance Summary

 

Feed Efficiency

1991

1981

Berkshire

2.52

2.64

Chester White

2.49

2.63

Duroc

2.37

2.46

Hampshire

2.41

2.49

Landrace

2.38

2.64

Poland China

2.55

2.78

Spotted

2.48

2.64

Yorkshire

2.40

2.53

Total or Avg.

2.41

2.52

 

Table 11. Boar Testing Survey Performance Summary

 

Average Backfat

1991

1981

Berkshire

20.3

21.1

Chester White

20.3

20.3

Duroc

19.6

19.3

Hampshire

18.5

18.5

Landrace

18.8

20.3

Poland China

20.3

22.3

Spotted

20.0

19.8

Yorkshire

19.8

19.6

Total or Avg.

19.6

19.6

 

Table 12. Boar Testing Survey Loin Eye Estimates

 

Loin Eye Area

1991

1981

Berkshire

36.1

35.5

Chester White

37.5

34.1

Duroc

36.2

35.2

Hampshire

38.1

36.9

Landrace

37.0

34.6

Poland China

36.8

36.6

Spotted

35.9

34.9

Yorkshire

35.9

34.4

Total or Avg.

36.6

35.2

°@

Table 13. Efficiency of lean gain by breed of sire in the Pork challenge test 1988-90.

 

Breed of Sire

1990

No. of Pens

1990

Feed/Lean

Total

No. of Pens

Total

Feed/Lean

Standard

Error

Babcock

2

8.09

4

8.27

°”.181

Berkshire

3

8.57

6

8.56

°”.148

DeKalb 77

2

8.28

10

8.13

°”.114

DeKalb 88

2

8.25

 

 

 

Duroc

4

8.22

13

8.26

°”.100

Duroc-Hamp(F1)

2

7.99

8

8.34

°”.127

Farmers Hybrid

4

8.20

12

8.36

°”.103

GIS

2

8.32

 

 

 

Hampshire

5

7.82

14

7.94

°”.097

Lucie Hybrid

2

8.38

4

8.56

°”.187

PIC HY

3

7.68

7

7.78

°”.136

PIC L-26

2

7.58

9

7.72

°”.121

Poland China

 

 

4

8.66

°”.184

Yorkshire

2

8.48

8

8.67

°”.129

AVERAGE

 

8.12

 

8.23

 

 

This program utilizes Al and MEW:. lures (Figure 3) to reduce health risks associate with the commingling of animals from many herds. Cooperating producers will receive semen representing at least three different terminal lines each breeding period to breed to their saws. Litters resulting from these matings will be sampled (1 or 2 pigs). Selected pigs will be entered in Performance MEW facilities. Pigs will go from Performance MEW facilities to central test facilities for growth rate and feed efficiency evaluation. Carcass traits will be measured at a packing plant and loin muscle samples evaluated for muscle quality at a research laboratory.

 

Breed associations or breeding companies will nominate their breeds or lines and sires to the Program Director. A program goa1 is to have many sires per line but few progeny per sire. Soars miintained in facilities raeeting the "Health Guidelines for Boar Studs" will have semen collaAed in these approved locations. The GPC will approve semen collection, extension, and distribution practices. Boars sampled will be less than 15 months of age. Each boar will be given a coerfidenha1 code number.

 

All the sows during a breeding period within each cooperating herd must be genetically uniform, such as F1 cmssbreds, PIC Camborough, etc. If possible, parity of each sow used in this program should be recorded to account for litter parity effects. If sows are clearly identified within a herd by parity and genetic type, mene than one genetic type of female may be used in that herd. There must be females of each parity of each genetic type used in the program in that herd. Two doses of semen per sow wi11 be provided.

°@

Figure 3. Terminal Line Program of National Genetic Evaluation.

°@

The number of approved commercial producers willing to use artificial insemination sow breeding methods will be established and a schedule for semen distribution established for each breeding period. Records of mating and sow pregnancies will be reported within 60 days of breeding to the Project Manager. Cooperating producers will receive a $20 per sow recording fee when breeding records are reported to NPPC.

 

An impartial person designated by the Project Manager will select barrows and/or gilts from each litter at 10-20 days of age. Selected pigs should represent the average of their litter. One or two pigs per litter will be sent to a Performance MEW station depending on program needs. Producers will receive the price of a 18 kg feeder pig for each selected pig.

 

As pigs finish the MEW program, they will be sent to a testing facility for growth and feed efficiency measurement. Pigs will be penned by linage of sire. Structural soundness will also be recorded. These records will be reported to the Project Manager.

 

As pigs reach 109 kg they will be sent to a cooperating packer. Traits of backfat depth, him muscle area or trimmed musde weight, carcass weight and carcass length will be measured according to these procedures. Visual scores of loin muscle quality will be assigned. These records will he reported to the Project Manager.

 

Two loin chops between the 10th and 12th rib will be removed from each carcass and dehvered to a research laboratory. One chop will be used for laboratory tests of muscle quality and nutritional content (lipid, etc.) and the other chop will be cooked to 160 degrees and measured for tenderness, lipid content, sensory panel, cooking loss, moisture content, and any other trait approved by the GPC. These records will be reported to the Project Manager.

 

It is important that pork producer associations at both the state and national levels as well as university extension personnel continue to support performance and progeny testing programs. Such participation is justified since emphasis is only on the economically important traits and the services are available to all seedstock producers regardless of breed or company affiliation. The benefactors of these programs are the consumers who ultimately should enjoy a better quality product at a lower price, and the commercial pork raiser who is aided in his search for superior stock as a result of this service. Association and university leadership assures control over publication of results and permits smooth modification of testing procedures should it be warranted by research findings.

 

Population Size

 

Selection acmss herds is very important. This is especially true if herd sizes are relatively small. Population size has a major effect on rate of genetic gain. A fifty sow breeding herd can achieve only about half the genetic gain possible in a 300 sow herd (Kennedy, 1989). Most purebred seedstock producers have herds of much fewer than 300 sow', but numbers that large or )arger are common in pure lines of most of the major breeding companies. To remain competitive, individual breeders must be part of a population much larger than their own herd by having access to genetic evaluations that permit selection across herds, such as through EPDs. Although the number of recorded litters of the various breeds has dwindled in recent years, all eight of the major breeds present in the U.S. are of sufficient size (table 14) to create near maximumprogress if across herd selection is practiced. This requires a breeding structure that facilitates such comparisons. That is why use of test stations and AI are important to genetic gain in both the individual breeder°¶s herd and in the whale population. These provide most of the genetic links that can give the individual seedstock producer the advantage of large effective population size. Because of health concerns, a decreasing number of live animals are transferred among breeder herds. Some procedures such as embryo transfer and cesarean section are very expensive. AI provides a relatively economic means for breeders to link their herds together genetically so as to create a larger effective population size. This is important for improvement of all economic traits, but is particularly important for genetic improvement of sow productivity traits.

 

Table 14. Litter Recording Summary By Breed

No. of

Litter Recorded

1982

1987

1992

Berkshire

3,245

2,091

2,137

Chester White

6,436

6,439

6,358

Duroc

31,975

22,000

20,248

Hampshire

17,711

18,160

20,824

Landrace

6,170

3,215

4,589

Poland

3,010

2,411

1,766

Spotted

12,600

7,906

7,469

Yorkshire

25,581

22,246

26,659

TOTAL

106,728

84,468

90,050

 

Visual Traits

 

Caution should be exercised to ensure that breeders do not place too much emphasis on "indicators of performance" instead of measuring performance itself. Some traits, such as structural soundness, length of body, underlines, general conformation and presence of physical defects, can only be evaluated visually on the live animal. Appraisal of animals with outstanding performance will train the keen observer to make selections among animals lacking performance figures or to make meaningful selection among animals from different herds. Biased or erroneous measurements can sometimes be uncovered by a watchful eye. Performance records, in addition to "eyeball appraisal", should be better than either method by itself.

 

In addition to the maternal and growth traits, it is essential that the performance testing program include careful assessment of animals for structural soundness and freedom from stress susceptibility. Findings at Iowa State University and elsewhere reveal clearly that leg structure is moderate in heritability and will deteriorate if not constantly monitored (Rothschild and Christian, 1988).

 

Underline soundness is also of importance although probably overemphasized in selection. Teat number is moderately heritable in swine (most estimates suggest it to be approximately 30% inherited). Whether gilts with less than 12 prominent nipples are kept should depend on the level of production achieved in the herd. It seems rather senseless to discard the leanest probing, fastest growing gilt in the herd if only eight pigs are normally raised per litter. Number of functional nipples rarely constitutes the limiting factor in number of pigs raised.

 

Inverted nipple problems are inherited as a recessive trait but at least two pairs of genes seem to be involved. Although it is a rare exception when the number of inverts is greater than two on a gilt, invariably these are located in the prime locations and hence animals showing this undesirable condition should be culled. A boar with undeveloped or inverted nipples can spread this problem through a herd rapidly and reduce the selection pressure achieved in other economically important traits.


°@

Porcine Stress Syndrome (PSS)

 

PSS and its associated problem, pale, soft exudative muscle USE), continues to plague the pork industry. Detection methods are available that can accurately determine the various stress genotypes. These will likely be used by the industry if PSS become a more serious problem, and if muscle quality (PSE) detection devices are developed that can accurately determine potential PSE carcasses during the slaughter process and at line speed. If such carcasses can be identified and are discounted, greater producer interest in PSS and PSE elimination will develop,

 

Because susceptibility to the PSS is definitely under single gene, autosomal recessive contml, problem herds should be'carefully monitored by halothane screening and/or DNA testing to reduce the frequency or eliminate the PSS gene (Christian, 1989). The DNA test is considered tp be a major breakthrough that will have a tremendous impact on the use of this gene.

 

The DNA stress test

 

Through its use the presence or absence of this mutated gene is readily detected. Five to 10 ml of blood are drawn into a heparin or EDTA treated tube and transported to a diagnostic laboratory where DNA is extracted. The desiied portion of the extracted DNA is cut using primers and reverse primers. These pieces which contain the segment of interest are amplified by a technique cilled polymerase chain reaction (PCR). Two techniques are available to separate the three genotypes. One is called oligonucleotide hybridization where two specific probes are given the opportunity to bind to the PCR-amplified product. One probe binds to the mutant gene and the other to the normal. Both probes bind the heterozygote sequences. The probes are labelled so the results can be read off autoradiographs. In the second procedure the amplified DNA is incubated with a restriction enzyme that will clea°¶ve copies of the defective DNA but not the normal. The digested DNA is then separated by electrophoresis. In this process, fragments of smaller size move greater and longer pieces move lesser distances. Staining permits interpretation of the three possible banding patterns. Stress animals will show two small fragments, normal, NN animals will show only one large fragment and the Nn genotypes will be characterized by fragments of three sizes.

 

The diagnostic test has been patented internationally by the Universities of Toronto and Guelph and licenses are required for its use. These licenses may be obtained through Innovations Foundation, University of Toronto, 203 College Street, Suite 205, Toronto, Ontario, MST 1P9 Canada, (telephone 416-978-5117). The testing service is offered at $50 per sample through either Dr. P. J. O'Brien, Pathology Department, University of Guelph or through Dr. Charles Louis, Department of Veterinary Pathobiology, University of Minnesota,St. Paul, MN 55108 (telephone 612-624-4202).

 

Industry Application

 

At first glance it would appear that removal of the PSS gene from the pig population should be the immediate goal of the industry. That is probably the wise i. U.S. and Engbsh studies agree that the nn sow farrows and raises about one pig per litter less than NN females and that they produce litters that weigh about 10% less at weaning. Furthermore, these animals are somewhat smaller at birth and at early ages (table 15), grow no Inter and perhaps slower postweaning and produce shorter carcasses that are almast always PSE. In Iowa State studies over 95% of the an genotype was PSE and in Danish studies exactly 90% were either PSE or doubtful. In a large number of ISU studies intransit death losses of PSS positive arumals have exceeded 15%. The merit of these animals xeides in the fact that their ban composition exceeds that of their normal NN littermates by 2.74% (table 17) ~ result of their greater leanness particularly over their loin musde, and their greater Join musde area and larger ham development (table 16). They also tend to consume leas per day and be more efficient converters of feed to lean.

 

.Considerable interest has been generated in deaignillg a breeding program that produces either all or a high percentage of carriers. U the arw herd or breoh contzibuting to the production of commercial sow are tested free of the disorder they could be mated to nn or Nn boars and produce only NN or Nn pigs. None of the offspring should die. stress deaths and any advantage of the camer genotype would be achieved. Studiea dearly deraomtrate the Nn genotype to be intermediate between the homozygotes for lean compositioa. Hence a pig crop made up of all carrie could be increased in lean content by as much as 2% if stress boars were ased and by 1% if camer sixes are used (table17). Why should one not try this approach? The answer relates to the degree to which the carrier animal is misceptible to tbe production of PSE meat. Results presented in tablts18-20providesome insight into this question. Several studies confirm that the 45 min pH of carrier animals averages about 6.1. Judge {1991) suggests a value in excess of 6.3 is necessary for desirable musde quality, and that values of 5.8 to 6.3 produce intermediate values. Values below 5.8 are certain to produce carcasses with severe PSE. The reflectance values for the various genotypes presented in table19 nohow the intermediate status of the Nn group. Slightly less than half of these carcasses were considered to be of inferior rhuscle quality. Values in table20 are means of a test that measures the degree the muscle proteins are denatured. Average values of the Nn class are significantly below those for the nn dass but somewhat higher than the NN group in at least the large study conducted in 1985. In their large comprehensive study of the three genotypes the Danes found 27% of the Nn genotype to be PSE and an additional 11% fell into the doubtful rangt. Sixty-three percent of the earner group was reported to have produced normal muscle quality.

 

It seems obvious that one should be cautious about using the stress gene in any way. Its elimination from the pig population is advisable. In the short term producing carriers could have some merit since the resulting offspring will not die of stress and packers are not yet in a position to detect PSE at a time when carcasses are still traceable to their source. It clearly represents a "quick fix" to increase carcass leanness. Perhaps new methods of ante-mortem swine handling and post-mortem carcass handling will eliminate PSE produced by carcasses with intermediate 45 min muscle pH values. But,until such time it would seem to be in the best interest of the industry to eliminate the gene and concentrate on improving leanness by direct selection in populations free of this disorder. Perhaps molecular genetics procedures will one day separate the good and bad aspects of the gene so the benefits can be realized without endangering the product quality of our valuable industry.

 

Table 15. Performance traits pigs by stress genotype.

 

Item

Genotype

NN

Nn

nn

No.

60

107

53

Birth weight, kg

1.31

1.35

1.28

21-day weight, kg

5.64

5.82

5.0a

42-day weight, kg

11.5

11.8

9.6a

Feed consumption, kg/d

2.73

2.60

2.60

Feed efficiency, kg

3.60

3.50

3.45

Average daily gain. Kg/d

768

768

777

adiffers significantly from other genotypes.


°@

Table 16. Oueass traiD of pigs by stress genotype.

 

Item

Genotype

NN

Nn

nn

No.

52

84

32

Length, mm

78.2

78.0

77.5

10th rib backfat, mm

39.1

36.8

36.8

Avg. backfat, mm

39.9

39.1

41.1

Loin eye area, cm2

35.7

36.8

40.0a

Dressing percentage

73.6

74.1

74.7b

adiffers significantly from other genotypes.

bdiffers significantly from the NN genotype

 

Table 17. Percentage differences in separable lean between genotypes in Iowa State and Danish studies.

 

Genotype

 

nn-NN

Nn-NN

ISU

2.7

1.1

DMRI

4.0

1.7

ISU = Iowa State University (N= 28, 32 and 23 for NN, Nn and nn genotypes)

DMR = Danish Meat Research Institute (N = 37, 95 and 70 for NN, Nn and nn genotypes)

 

Table 18. Forty-five minute muscle pH in five studies by stress genotype.

 

Study

Genotype

NN

Nn

nn

Christian, 1981a

6.42

6.15

5.73

Denmark, 1985b

6.16

5.90

5.35

Skaggs, 1990c

6.39

6.16

5.80

Piedrafita, 1991d

6.55

6.28

5.75

Hawkins, 1992e

6.27

6.07

5.64

a 52, 84 and.32 pigs of each genotype, respectively

b 61, 167 and 111 pigs of each genotype, respectively

c 32 pigs of each genotype

d 16 pigs of each genotype

e 12 pigs of each genotype

 

Table 19. Muscle reflectance measurements of loin muscle by genotype in four ISU studies.

 

Study

Genotype

NN

Nn

nn

Christian

22.5

24.6

29.0

Skaggs

22.2

25.2

27.4

Piedrafita

19.7

21.1

31.8

Hawkins

23.0

25.5

30.2


°@

Table 22. Average percent heterosis for various swine traits.

 

Trait

First cross using Purebred female

Multiple Cross using Cross bred female

Litter size born alive

1.0

5.6

Litter size at 21 days

8.0

17.1

Lither weight at 21 days

11.3

24.3

Litter weight at 21 days per female exposed

11.3

29.0

Days to 220 pounds

6.5

7.7

Average daily gain

9.4

9.4

Peed per pound gain

2.3

2.3

 

The two-breed rotational scheme is still being used by a number of producers. A York x Duroc or York x Hampshire scheme is still very popular. These breeds complement one another well and generaQy produce a final product with good packer acceptance. The disadvantage of the two-breed system is that only 67% of maximum heterosis is achieved after six generations. One of its advantages is simplicity.

 

To capitalize on heterosis, the addition of a third breed increases heterosis to 86% of maximum. This is the most popular system with producers. It permits choice of breeds for maximum complementation. The Hampshire, Duroc and Yorkshire breeds are those most commonly used; the Hampshire for carcass desirability, the Duroc for superior growth and efficiency and the Yorkshire for maternal performance.

 

There are several disadvantages of the three breed rotational system. The genetic makeup of the sow herd is constantly undergoing change and boars of all three breeds must be kept w inventory and properly mated if maximum heterosis is to be realized. Because of the genetic diversity, there is a great deal of variation in the output. The second limitation is that the system doesn°¶t permit utilizing each breeds attributes to the best advantage. Ideally, oui producers would like to maximize the Hampshire carcass desirability in all market pigs rather than in one third of them.

 

Static or terminal systems of mating are becoming more popular since larger complexes. permit specific mating schemes for intraherd replacement gilt production and since Fl gilts art available from several commercial companies. These systems allow the positioning of brea where its best characteristics may be more fully realized and where it permits pigs of consistent breed background to be produced over a long period of time. Perhaps most importantly, the static cross can be designed to achieve 100% of maximum heterosis in both th~ pig and the dam. The most popular breed choices for this system have been a female produced from the mating of two white breeds. This F1 female is generally mated to Duroc, Hampshire Spotted or hybrid boars.

°@

Table 20. Percentage of light transmittance of loin muscle by genotype in three

ISU studies.

 

Genotype

Study

NN

Nn

nn

Christian

24.3

33.5

73.0

Skaggs

20.5

24.2

59.0

Piedrafita

5.6

8.9

71.1

 

Reduction in subcutaneous and intramuscular fat (marbling) has definite benefits in improving the image of pork in the eyes of the consumer. However, lowering marbling to values below 2.5% may result in reduced juiciness and flavor. Repartitioning agents will also reduce marbling, and hence, methods to detect marbling and other muscle quality traits in the live pig would be highly desirable. A shot biopsy procedure developed in Germany and Czechoslovakia shows promise for this purpose.

 

Crossing Systems

 

Considerable variation exists in crossbreeding systems employed by commercial producers (Table 21). Most are designed to capitalize on heterosis present in most traits (Table 22). Although most operations use some type of crossing scheme, variations in the breeds used and the crossing method are common.

 

Table 21. Type of Crossbreeding System used by Percentage of Farms*.

System

Percent of Farms

2-breed rotational

14.8

3-breed rotational

44.8

4-breed rotational

10.5

Rotaterminal 2-breed

5.3

Rotaterminal 3-breed

5.8

Terminal

18.8

 

100.0

*Michigan State University Survey 1988

 

A compromise system of mating that achieves the advantage of a specialized maternal female and 100% of maximum heterosis in the final pig produced is what has become known as the static-rotational or rotaterminal system. In this system the best producing sows from within the herd are identified and mated to boars of the maternal breeds (usually Yorkshire, Landrace or Chester Whites) to produce the replacement females. The balance of the sow herd is mated to a terminal sire and all of the offspring marketed. If 20% of the litters are sired by maternal line boars in a 3-breed rotational system, a replacement rate of one female from each such litter will allow the sow herd to remain of constant size if approximately 20% of the sows are culled twice each year.

 

The main advantages of this system are intraherd replacement of females, selection of female replacements from the best producing sows and nearly optimum heterosis (100% in the pig and 86% in the sow). Because smaller herds do not maximize use of maternal line boars, artificial insemination offers an excellent alternative in this system.


°@

Summary

 

Genetic improvement in commercial swine herds depends upon selection and optimum heterosis. Since most genes come from outside the herd, accurate and effective selection both within and across herds by the seedstock supplier is crucial if his commercial clients are to improve at an optimum rate. Participation in selection programs utilizing EPDs, such as STAGE 6, permit seedstock producers to maximize the rate of genetic change when results are used in choosing replacements and emphasis is placed on the economically important traits. Genetically linked herds provide small seedstock herds with a collective population size large enough for effective selection programs. This is particularly true for SPI traits. AI offers great potential to strengthen genetic ties between herds, facilitate across herd selection and transfer of genes between herds. This approach can permit the traditional seedstock producer to compete effectively with larger herds or breeding companies that have the advantage of large herd size. Survival of the small private purebred breeder could very well depend on adoption of these approaches.

 

The commercial producer can select outstanding animals of superior pure lines of breeds and combine them into programs that capitalize on heterosis. The optimum utilization of healthy, genetically superior stock that is well-managed, and fed to achieve their genetic potential, should result in the low cost production of a superior pork product.

 

Literature Cited

 

Avalos, E. and C. Smith. 1987. Genetic Improvement of Litter Size in Pigs. Anim. Prod. 44:153.

Belonsky, G. M. and B. W. Kennedy. 1988. Selection on Individual Phenotype and Best Linear Unbiased Prediction of Breeding Value in a Closed Swine Herd. J. Anim. Sci. 66:1124.

Bereskin, B. and R. J. Davey. 1976. Breed, Line, Sex and Diet Effects and Interactions in Swine Carcass Traits. J. Anim. Sci. 42:43.

Christian, L. L. 1989. A Genetic Overview of Pork Quality and Evaluation of Traits to Consider for Future Swine Breeding Programs. Proc. N. American Swine Improvement Conf. NSIF Vol. 14. U. of Minnesota, St. Paul.

Christian, L. L. and Steve Moeller. 1990. Serial Evaluation of Growing Pigs by Real-time Ultrasound. Proc. NSIF Conference and Animal Meeting. U. of Minnesota, St. Paul, MN.

Christian, L. L. 1992. Genetics/PSS/PSE. Proceedings Annual Meeting of Iowa Veterinary Medicine Association. Des Moines, IA.

Gibson, J. P. and C. Smith. 1986. Technology and Animal Breeding. Application in Livestock Improvement 3rd World Cong. Gen. Appl. Livest. Prod. XII:96.

Ginder, Roger G. 1989. Changing structure of the pork indudtry. Staff Paper Series, Depart. of Econ, Iowa State University, Ames, IA.

Harris, Dewey I, Donna L. Lofgren, Terry S. Stewart, Alan P. Schinckel and Mark E. Einstein. 19489. STAGES -National Decision Support System for U.S. Swine Breeders. Proc. N. American Swine Improvement Conf. NSIF Vol. 14. U. of Minnesota, St. Paul.

Benderson, C. R. 1973. Sire evaluation and genetic treads. In: Proceeding of the Animal Breeding and Genetics Symposium in Honor of Dr. Jay I Lush, Amer. Soc. Animal Sci. and Amer. Dairy Sci., Assoc., Champaign, IIIinois.

Jehansson, K. and B. W. Kennedy. 1985. Estimation of Genetic Parameters for Reproductive Traits in Pigs. Acta. Agric. Scand. 35:421.

Kennedy, Brian W. 1989. Concepts and Misconceptions on the Use of Estimated Breediag Values. Proc. N. American S Improvement Conf. NSIF Vol. 14. U. of Minnesota, St. Paul.

Lamberson, W. R. 1990. Genetic Parameters for Reproductive Traits. In: Genetics of Swine. Edited by Larry D. Young. USDA-ARS-USMARC, Clay Center, NB.

Lawrence, John D. 1992. The U. S. pork industry in transition. Staff paper no. 240, Dept. of ., Econ., Iowa State University, Ames, IA.

Miller, Dale. 1992. Testing station update. Nat. Hog Farmer, St. Paul, MN.

Rothschild, M. F. and L. L. Christian. 1988. Genetic Control of Leg Weakneas in Duroc Swine. I. Direct Response to Five Generation of Divergent Selection. Livst. Prod. Sci. 19:459.

Stewart, T., D. Lofgren, M. Einstein, A. Schinckel and D. Harris. 1991. STAGE 6 - Across - Herd Genetic Evaluation/National Sire Summary. Yorkshire Journal, Lafayette, IN.

Stewart, T. S. and A. P. Schinckel. 1990. Genetic Parameters for Swine Growth and Carcass Traits. 1990. In: Genetics of Swine. Edited by Larry D. Young. USDA-ARS-USMARC, Clay Center, NE.

Thulin, Andrew and Dale Miller. 19S8. U.S. seedstock survey. Nat. Hog Farmer, St Paul, NN.

Weber, G. M. 1987. Guidelines for Uniform Swine Improvement Programs. NPPC, Des Moines, IA.