In response to pork producers needs for comprehensive evaluation of sow productivity, the National Pork Producers Council (NPPC) and allied industries recently concluded the most complete maternal line genetic evaluation program ever undertaken in the U.S.
The project, funded by the National Pork Board, evaluated lifetime productivity of six genetic lines, represented by 3,559 females through four parities. Growth, carcass, meat quality and eating quality traits were also measured on 1,261 progeny from those sows.
Maternal Line Data Lacking Terminal crossbreeding using specialized sire and dam lines is the predominant breeding system used by commercial producers. Results of the National Genetic Evaluation Program (NGEP) terminal sire line program provided unbiased comparisons of sire lines for 40 growth, carcass, meat and eating quality traits (see National Hog Farmer Special Report, June 1, 1995). However, maternal and reproductive traits such as age at puberty, conception rate, sow longevity, litter size and rebreeding interval are also extremely important.
Breeding companies often use proprietary data to illustrate performance of commercial lines, however unbiased comparisons of maternal line performance are not readily available. And, because genes from dam lines make up 50% of the genetics of the market pig, it is even more important than sire line information in making sound breeding decisions.
The new National Genetic Evaluation Maternal Line Program (MLP) data complements the NGEP sire line evaluation data, valuable to pork producers because it provides:
* Unbiased, accurate maternal line performance data from development and conception rate of gilts through fourth-parity production;
* Economic values of traits to maximize production efficiency; and
* Application of results in seedstock herds to further enhance efficiency of commercial customers.
Genetic Line Sampling All 3,559 gilts that entered the program were F1 crosses representing products of five breeding organizations (Table 1). The project continued through four parities. The last litters were weaned in August 1999.
Four of the genetic lines, designated as Danbred USA, Dekalb Monsanto DK44, American Diamond Genetics and Newsham Hybrids USA, are commercially available product lines. Each is an F1 cross, selected and developed for maternal traits from the company's lines, in a closed breeding program. In most cases these lines originated from the Landrace and Large White/Yorkshire breeds, but may include other breeds.
A fifth line, entered by the National Swine Registry, is an F1 Yorkshire-Landrace cross gilt produced from purebred Yorkshire boars bred to purebred Landrace females and the reciprocal cross.
The sixth line, Dekalb Monsanto MXP200, is a cross between the Dekalb Monsanto DK44 with an experimental line developed at the University of Nebraska. The Nebraska Index Line (NIL) is a composite population of Large White and Landrace that was closed to outside lines in 1981, then selected for increased ovulation rate, embryonic survival and litter size at birth for 16 generations.
The Dekalb Monsanto MXP200 gilts were produced by crossing NIL boars with Dekalb Monsanto DK44 sows. The company purchased samples of the index line from the University of Nebraska and maintains this line at its nucleus farm.
Gilts entered the test program in three groups beginning in February 1997 (see Table 1). Vasectomized boars were brought in with the gilts and used to stimulate sexual development using appropriate boar-exposure procedures.
The NPPC Genetics Program Committee established sampling procedures to ensure unbiased comparisons. The number of gilts sampled was set to ensure reliable estimates of differences in sow longevity could be made. And, to ensure results were repeatable, the pure lines used to produce the F1 gilts came from essentially closed populations.
Gilt Development The test program was accomplished in cooperation with Pork Technologies, a commercial enterprise located in southern Iowa. Facilities included two 1,600-sow breeding-gestation-farrowing units and three 1,100-pig wean-to-finish barns.
Gilts entered the wean-to-finish barns between 7 and 20 days of age, averaging 9.4 lb. (2.6-19 lb. entry range). Gilts were grouped by genetic line, approximately 25/pen. They received a Naxcel injection on days 1, 2 and 3, electrolytes through water medicators on days 2-6. All gilts were vaccinated for porcine reproductive and respiratory syndrome (PRRS) and swine influenza virus (SIV). Ivomec was used at entry.
Diets fed at different phases are shown in Table 2. Gilts were fed a 23% protein (1.7% lysine) diet until they weighed 20 lb. From 20 to 30 lb., a diet with 1.5% lysine was fed. At 30 lb., the corn-soy starter diet with added fat contained 1.35% lysine and was fed to 70 lb. These diets contained Mecadox. From 70 to 150 lb., gilts were fed a corn-soy grower diet that contained 1.2% lysine with added fat. The gilt developer diet (150-210 lb.) was corn-soy with 0.8% lysine.
Heavier gilts at entry into segregated early weaning (SEW) units had better performance and greater survival rates. Gilts weighing less than 7 lb. at entry had a survival rate of 85% while those over 7 lb. had a 94% survival rate. For each increase of 1 lb. in entry weight there was a 2-lb. advantage at 50 days on-test and a 3-lb. advantage at 150 days on-test. Survival rate was calculated as the percentage of gilts that entered the barn that were moved to the sow units for breeding.
Gilt performance was excellent, although it varied among lines. The Dekalb Monsanto DK44 gilts grew the fastest. Danbred USA and Newsham Hybrids USA gilts were leanest. The Dekalb Monsanto MXP200 gilts grew the slowest and American Diamond Genetics gilts were fattest.
The attending veterinarian, unit manager and NPPC program manager jointly evaluated gilts for health and abnormalities after approximately 150 days in the wean-finish unit (roughly 165 days of age). About 3% of the gilts were culled for umbilical hernias, chronic illness and severe injuries; 4.6% died. Growth or backfat were not considered.
Ninety-two percent - 3,283 gilts - passed the inspection, entered the MLP program and data recording began. It is the lifetime production of these 3,283 gilts that was evaluated.
Gilts were eartagged and randomly allocated to one of two identical, 1,600-sow breeding-gestation-farrowing (BGF) units with approximately equal numbers of gilts of each entry group and genetic line at each farm (Table 3).
Gilts were vaccinated for leptospirosis and parvovirus as they entered the BGF unit. Tenth rib backfat thickness was measured off the mid-line and loin depth was measured using A-mode ultrasound. These measurements were used to calculate gilt performance in the wean-finish buildings (Table 4).
Gilt, Sow Management Gilts moved to the BGF units were placed in 17 x 72 in. gestation stalls. They were exposed to vasectomized boars daily and checked for estrus. Test protocol required gilts 205 days of age or older to be mated at their second or later estrous period.
Age of the gilts at first observed estrus varied by line (Table 5) with the Dekalb Monsanto MXP200 gilts being approximately 14 days younger than the other lines.
A random sample of females of each line were artificially inseminated (AI) with semen from Danbred USA Duroc-Hampshire crossbred boars to produce progeny test pigs that expressed 100% heterosis. All other program matings were by AI to an unrelated sire line regularly used at the BGF units, also to ensure 100% heterosis in the progeny.
Gilts were bred when found in estrus and every 24 hours thereafter while in estrus.
The MLP program protocol required individual sow management to optimize performance of all sows. Daily gestation rations were adjusted to sow size and condition. The National Research Council (NRC) Feed Requirements Model was used as a guide.
Small gilts requiring additional feed for development were put in pens of six and fed ad lib. When sufficiently developed, they were placed on the reduced feed regime. The gestation diet fed from entry into the BGF units, and throughout each gestation period, is shown in Table 2.
Sows were weighed and last rib backfat thickness measured as they entered a farrowing crate and again when they left at weaning.
Sows were fed ad lib from weaning until mating. After mating, all gilts and sows were limit fed 4 lb./day for 14 days to enhance embryo survival. Backfat and weight data were used to calculate gestation feed requirements from day 15 of gestation until farrowing. Very thin sows or fat sows had their daily rations adjusted to match their needs during each gestation period. All sow feed intakes were recorded. Electronic eartags identified each sow, greatly improving the efficiency and range of information captured in this project.
Culling Procedures Gilts were removed from the program only if they failed to express estrus by 300 days of age, or they expressed estrus by 300 days of age but failed to conceive during a 60-day breeding period or, of course, if they died. Gilts that expressed estrus were given a maximum of three opportunities to conceive.
Unmated gilts (433 total) were slaughtered and their reproductive tracts evaluated by Don Levis, University of Nebraska reproduction specialist. Another 240 gilts either died or were culled before farrowing a litter.
Levis' evaluation indicated the BGF technicians did an excellent job of estrus detection. Of the 433 culled gilts, 372 were found to have abnormal and/or non-cycling tracts. Only 27 of the culled gilts had normal tracts that had expressed estrus regularly but went undetected; 60 identified as non-cycling gilts actually had one cycling date recorded but subsequent estrous periods went undetected. Four gilts with no recorded estrous date were actually pregnant when slaughtered.
Sows were checked for estrus daily after weaning. Sows were mated 12 hours after estrus was first observed, then every 24 hours thereafter while still in estrus. Sows were culled if they failed to conceive within 50 days of weaning their litter. No culling on the basis of litter size or litter weight was done. A female's record of lifetime productivity included total output from when she entered the BGF unit until she weaned her fourth litter, or was culled or died.
Progeny Test Pigs Progeny test pigs, both barrows and gilts, were selected randomly from the litters sired by Danbred USA Duroc-Hampshire boars. A total of 1,261 pigs were tested for growth, 935 had feed intake recorded, and carcass traits were recorded in 1,253 pigs. (Integrity screening eliminated feed intake data on some progeny pigs.)
Approximately 210 females per line produced a litter from which a progeny test pig was used. These pigs were weaned at 10-21 days of age and transported in one of three intake groups to an SEW station.
After completing the SEW program, they were allocated, by age, to pens for growth evaluation. The test facility was equipped with electronic feed intake recording equipment (FIRE). Initial weight, final weight and individual feed intake were recorded from approximately 50 lb. to an off-test weight of 250 lb. An 18% crude protein (0.95% lysine) corn-soy diet with 2% added fat was used. As progeny test pigs reached 250 lb., they were sent for carcass evaluation. A loin muscle sample of each pig was collected for meat quality and eating quality evaluation. The number of progeny test pigs per line is shown in Table 6.
MLP Data Analyses The traits evaluated in the MLP test includes measures of fertility and expression of estrus in gilts, sow productivity through four parities, and growth and carcass traits of a sample of their progeny. A list of the major traits appears in a special gatefold table, pages 48-51.
The main objective of the project was to evaluate lines for longevity and sow productivity. The time clock began on the day the gilt was moved to the BGF unit and continued until she died, was culled or weaned her fourth litter.
The output variables recorded for each female included her age at first observed estrus, all breeding dates, age at each farrowing, interval between litters, total number and number of live pigs at each parity, and litter birth weights and weaning weights.
The input variables recorded were feed intake each day of the female's life and the number of days she was in the herd. Other costs, such as gilt replacement costs and non-feed fixed and variable costs, were estimated for economic analyses.
Measures of longevity and total sow production included total number of live pigs born and weaned in the life of a sow, total weight of pigs produced, percentage of sows reaching fourth parity, pounds of pigs/sow/day of life and pounds of pigs produced/ sow/pound of sow feed.
Cohort Analysis When measured on individual sows, measures of longevity have poor statistical properties for data analysis. Every female has input variables of either time or feed, but some females produced no litters and have zero output; others have output of either one, two, three or four litters. When individual sow records with zeros were analyzed the distribution of records was not normal. Normal distribution is needed for testing differences between lines and parities.
To avoid the statistical problem caused by non-normal data, 130 cohort groups of females were established. A cohort is defined as a group of animals which share one or more events in common within a defined time period. Females were assigned randomly within genetic line/gilt intake date/BGF unit to a cohort group.
Because there was not always an equal number of each line on each gilt intake date, the number per cohort group ranged from 21 to 29 females, with an average of 25 per group.
For measures of longevity and lifetime production only, variables analyzed were lifetime output and input of females in each cohort group. Traits are expressed as output per gilt or output per group of 25 gilts. This is a standard statistical procedure to normalize data. All traits expressed on a cohort basis were checked and found to have normal statistical distributions.
Producers should think of each cohort group as 25 replacement gilts of a particular line entering the breeding herd at 180 days of age. One can then follow the lifetime output of these 25 females and compare lines for percentage that farrowed, the percentage that had one, two, three and four litters, total number and weight of pigs produced by these 25 females, and sow replacement rates and costs of purchasing and developing replacement gilts. Further analysis of these 130 cohort groups begins on page 42.
Presentation Of Results There is a very large amount of data from this project. The most important findings are summarized in the following articles. Economic analyses were performed to determine the net differences among maternal lines considering differences in reproduction and longevity and differences in transmitted effects on progeny performance. Pork producers will find the information valuable in making breeding decisions to enhance profit.
Rodger Johnson, University of Nebraska; Rodney Goodwin and David Boyd, National Pork Producers Council