The genetic makeup of maternal (sow) lines and terminal sire lines used to produce market hogs has a considerable bearing on how efficiently their progeny convert feed ingredients to lean meat protein.
This article will explain the heritability of growth and feed conversion traits and genetic correlations among the traits, plus provide some guidelines to ensure your genetic suppliers are applying appropriate selection pressure to the economically important traits.
Furthermore, it is important to understand your genetic base as you build diets that will maximize sow performance as well as the pigs' performance potential and their ability to hit carcass targets and maximize packer premiums.
Feed conversion ratio (FCR) reflects the rate at which an animal converts feed to meat. This ratio is calculated by dividing the amount of feed used by the total weight gained. The genetic contribution of both the sire and the dam lines must be considered for all production traits, including FCR.
The sire lines contribute half of the progeny's ability to convert feed ingredients to lean meat protein. Dam lines contribute the other half of the progeny's genetic potential, plus their ability to convert feed ingredients to reproduction and milk production.
Heritability, often defined as the resemblance between relatives, ranges from 0 to 1. High heritability indicates a high level of resemblance between parents and progeny. Effectively selecting progeny to become parents based on a particular trait will create a new generation of higher-performing animals.
Genetic correlations are the proportion of variability that two traits share due to genetic causes and, theoretically, may range from -1.0 to +1.0.
Table 1  shows the heritabilities and genetic correlations among production traits, feed intake and feed conversion, plus the traits often used to predict or improve the accuracy of estimating feed conversion.
The yellow portion of Table 1  shows the traits of average daily gain (ADG) in grow-finish, backfat depth (BF) and loin muscle area (LMA). All three traits have a moderate to high heritability and moderate genetic correlations. The sign (direction) of the correlations must be evaluated in addition to the magnitude. The 0.25 correlation between ADG and BF indicates that as ADG increases, backfat will also increase. This is not a favorable result and is indicative of an antagonistic correlation.
In comparison, the 0.36 correlation between ADG and LMA is beneficial and indicates that as ADG increases, so will muscle mass.
Finally, the -0.33 correlation between BF and LMA is also advantageous because a decrease of BF and an increase of LMA are both desirable changes.
Antagonistic correlations, such as the one between ADG and BF, can be controlled by utilizing selection indexes as long as both traits are measured and included in the indexes. Goals for either trait can be adjusted in the formulation of an index.
Heritability and Correlations
A few decades ago, feed efficiency was directly estimated by recording feed consumption of an animal placed in an individual crate. Selecting sires with the best feed efficiency using this method resulted in future generations with better feed efficiency for those animals “if” they were raised in crates. However, FCR progress was often disappointing for various reasons, so that method was dropped.
The research data generated by this method built an excellent case for indirectly selecting feed efficiency via its component traits, namely ADG, BF and LMA. By building the correlations of ADG, BF and LMA with FCR, a new selection index can be developed that will allow improvement in all four traits but only measure three. This is made easier by the advantageous correlations in all three traits with FCR.
Returning to Table 1 , the green genetic correlations have the correct sign to allow improvement in FCR and are of a moderate to large magnitude. This selection method has contributed to most of the genetic improvement made in FCR in the 20th century.
Beginning in the 1990s, electronic feeding stations in pens of up to 20 head were used. Feeding stations were an improvement over the individual crates, but they suffered from occasional mechanical breakdown and erroneous data production.
Continue on Page 2
More recently, mathematical calculations have been developed to adjust for feed data errors and allow for much more accurate predictions of feed consumption. Consequently, FCR heritability and correlations are more accurate.
The use of feeding stations for direct measurement is often prohibitive, but from a research standpoint, results can be generated over a period of years to compute the correlations necessary to do indirect (component) selection based on growth, backfat and muscling. Continued use of feeding stations will also contribute to our understanding of the mechanisms that affect feed conversion.
Ad libitum feed intake (FI) is a major component in FCR. Previous discussions showed how growth and composition can be used to predict FCR. Similarly, those traits can be used to predict FI. One half of the genetic variation of FI and FCR are not explained by those production traits.
Over the past 20 years, research has explored the remaining genetic variation by measuring the residual feed intake (RFI) trait. RFI is used to estimate the feed consumed over or under expected requirements for production.
Variation in RFI is caused by maintenance requirements, feeding behavior, nutrient digestion and energy homeostasis and partitioning. Methods of measuring FI have evolved much like those described previously for FCR. The mathematical methods of calculating RFI continue to evolve.
RFI has been shown to be a moderately heritable trait similar to FCR. Continued improvement in FCR will require methods beyond the correlated response from ADG, BF and LMA. RFI would serve as such a tool. Genetic correlations with RFI are found in pink in Table 1 . The correlation with ADG is small and in the wrong direction. The remaining correlations of RFI with BF, LMA and FCR are all large and in the correct direction. The correlation with FCR is also quite large and shows the potential for improvement in FCR.
Unfortunately, measuring RFI suffers from the prohibitive costs mentioned with FCR improvement when using individual feeding stations. RFI will become a better research tool than FCR by providing data for identifying new gene markers. The gene markers would then be an economical method for many seedstock suppliers.
Finally, the insulin-like growth factor-I (IGF-I) is a naturally occurring polypeptide produced in the liver, muscle and fat tissues. IGF-I is associated with growth and development during the postnatal period.
Research began over 10 years ago to compare the juvenile IGF-I blood test taken at 3-5 days postweaning and before 35-42 days of age with subsequent postweaning production traits. After several years of testing at sites in five countries, heritabilities and genetic correlations were summarized by Kim Bunter of the University of New England in Australia. Results are presented in blue in Table 1 .
All genetic correlations are in the correct direction. The large correlation with FCR is especially promising, and only the genetic correlation with ADG is small. The use of IGF-I has two advantages over RFI. One is the much lower cost of testing, and second is its use at an early age. Test results at a young age allows for early culling. Early castration of males based on IGF-I results and estimated breeding values of production traits provide substantial increase in cull value. Second, fewer animals need to be processed through more expensive test procedures, such as direct FCR or RFI measurements because of early culling.
Genetic markers are the final measurement to be discussed under the general selection for FCR. A genetic marker may be a gene or a DNA sequence with a known location on a chromosome, and is associated with a particular trait. It can be described as a mutation or alteration in the genomic loci that can be observed.
A genetic marker may be a long DNA sequence or a short sequence, such as a single nucleotide polymorphism (SNP). This discussion will be limited to the SNPs, which are commonly referred to as “snips”.
SNP usage in the swine industry began in the early 1990s with the hal-1843 SNP for meat quality (pale, soft and exudative, or PSE) and the porcine stress syndrome (PSS). Markers have been most often developed for traits that have economic value, are difficult to measure in live animals or are lowly heritable. Another early and well-known marker is the RN- (Rendement Napoli [negative]) gene for meat quality. The potential that the tests hold can be significant, such as the improvement in meat pH shown by the use of the PSS and RN- tests.
Continue on Page 3
SNPs and other gene markers were originally very difficult to locate. In early gene marker work, it was not uncommon for it to take years between when a gene marker was described and when an accurate test was developed. New developments in sequencing, better understanding of the underlying biology and the accumulation of good trait data will help greatly in the development of future markers.
The number of SNPs affecting FCR is not known. This is largely because of the unknown number of markers controlled by private industry. Today, there are three SNPs readily available for testing by laboratories in the United States and Canada (GeneSeek and DNA Landmarks).
The melanocortin-4 receptor (MC4R) is associated with ADG, FI, BF and lean meat yield. The allele G delivers lean growth, less BF and lower FI. Allele A increases ADG.
In spite of the antagonistic correlation with ADG, allele G is expected to provide lower FCR because of the beneficial correlations of FCR with BF and FI. The A allele is nearly fixed in the Hampshire breed, making selection for the G allele ineffective in that breed.
The cholecystokinin type A receptor (CCKAR) associates with FI and growth traits. Selection of the G allele can be used to select animals with higher FI and faster growth. Selection of allele A is used to lower FI and thus FCR.
The high-mobility group A (HMGA1) is highly associated with BF and lean growth. Selection for the T allele will reduce BF and improve lean percentage. Reduction in FCR will also occur because a reduction in fat deposits reduces the feed required to add weight to the animal. This marker is best used only in sire line selection because of the emphasis on reduction of fat deposits and its potential impact on lactation performance.
Lactation Feed Intake
The importance of feeding sows correctly during lactation has been recognized for many years. The increased productivity of sows increases the risk of a more pronounced negative energy balance — loss of body condition — during lactation.
Research that would help estimate the heritability and genetic correlations of lactation feed intake (LFI) with other traits remains limited. Research on LFI by Susanne Hermesch at the University of New England in Australia has shown it to be a heritable trait with estimates ranging from 0.14 to 0.30. Genetic correlation estimates were moderate for prolonged wean-to-service interval (0.18) and number weaned (0.24), high for litter weight gain (0.48) and very high for the number of piglets and sow parities achieved over their lifetimes (0.67) and (0.73).
Another trait used to reduce the risk of negative energy balance is lactation efficiency (LE). LE is an estimate of energy output (energy deposition in piglets) divided by energy input (energy in feed above sow maintenance plus energy out of body tissue). Work by R. Bergsma and others in the Netherlands has shown that the heritability is low (0.12). Genetic correlations were low for total number born (0.09) and prolonged wean-to-service interval (0.1), and moderate for piglet mortality (-.24), litter weight gain (0.23), stayability as defined as the first-litter survival of sows (0.3) and LFI (-0.38).
Inclusion of LE in the breeding goal will improve stayability without negative consequences on other economically important traits. The rather high and negative genetic correlation with LFI is a long-term concern. A revised breeding goal of increasing LE and holding LFI constant is probably more realistic and achievable.
There are no gene markers currently available that are directly related to LFI or lactation feed efficiency. But if we view the reproductive efficiency as a ratio of output divided by input — much like LE — there are two gene markers that have potential for increasing output.
The first gene is the estrogen receptor gene (ESR) test, which has been shown to be effective in increasing litter size in Large White and Yorkshire lines and crosses including either breed. The second marker is the erythropoietin (EPOR). This genetic variant has been associated with uterine capacity and litter size. Selection for EPOR has been effective in two swine populations at the U.S. Meat Animal Research Center, including the industry relevant BX population, which is a Yorkshire, Landrace, and Duroc crossbred line. Improvement in litter size using one or both markers should improve gestation feed efficiency.
Maternal Line Evaluation
Overall, information about genetic relationships between LFI of sows and performance of pigs were positive for ADG (0.44 to 0.60) and FI (0.33 to 0.59), and small and negative for BF (-0.24 to -0.02). Unfortunately, the direction of the genetic correlation between LFI and FI is antagonistic if the breeding goal is to increase LFI while decreasing FI. Dam line selection must take this into consideration.
Continue on Page 4
Limited data is available on the relationship between finishing pig performance and sow reproductive performance. B. Holm and others at the University of Norway published genetic correlations between finishing FI and age at first mating, number born alive at first parity and number born alive at later parities, which were 0.2, 0.23 and 0.2, respectively. All three correlations are relatively small and indicate that selection for FCR by reducing FI can have a detrimental effect on litter size. This must also be considered in the development of dam line selection.
Dam line selection sustains a high reproductive output (requiring high body reserves and high FI), and also supplies half of the genetic potential for lean, fast-growing and feed-efficient (low FI and low FCR) slaughter pigs. Selecting only for reproduction has often become the focus for dam line selection with slaughter pig traits left to the sire line. With a better understanding of the genetic correlations between slaughter pig traits and maternal line traits, we should be able to develop new methods for dam line selection.
A more realistic breeding goal for dam lines should include grow-finish traits such as ADG or days to 250 lb., BF and, perhaps, FCR or correlated consumption traits; reproductive traits such as litter size, piglet mortality, weaning-to-mating interval and sow mortality; and lactation traits such as litter weight gain, sow body tissue loss and LFI.
Including lean growth traits in a maternal-line evaluation using a multiple-trait model to estimate inputs into selection indexes should increase the accuracy of the genetic evaluation for litter traits. Emphasis can be varied on the grow-finish side — from holding performance constant — to mild improvements without reductions in reproduction. The genetic correlations between traits need to be evaluated periodically.
Sire Line Selection
Sire line selections are to include some combination of the traits found in Table 1 . Meat quality traits are beyond the scope of this discussion but can be antagonistically correlated to feed traits.
A potential sire selection program could begin with culling of young males at an early age using factors such as IGF-I, gene markers and estimated breeding values based on the performance of relatives. This would allow for a smaller number of boars to be tested for RFI or FCR utilizing feeding stations. Gilts would not be tested in feed stations, but estimated breeding values from their male relatives could be used in their selection.
Different sire lines should be tested using different methods because of their intended use in the marketplace. It is also important to understand the genetic response from combinations of these various methods. This knowledge will allow producers to ask the right questions of their genetic supplier and know if they are applying the appropriate selection pressure on economically important traits.