AGBU
Menu:
PIGBLUP Proof
Predicting Differences in Performance
EBVs can be obtained for an animal that has no records for a trait using information from relatives for that trait.
Differences in EBVs represent the expected difference in phenotypic performance for these animals. How well do PIGBLUP EBVs predict phenotypic differences in litter size, growth rate and backfat?
The Idea
- Obtain EBVs for young animals, which have no records available yet, using PIGBLUP.
- Group these animals based on their EBVs (ten percentile classes).
- Compare differences in mean EBVs with differences in mean phenotypic performance.
Material
- Data for all the studies was provided by Aztec Farms comprising 16 years (1980-1996).
- The animal's own records did not contribute to its EBVs.
- Litter size EBVs for gilts were calculated using only pedigree information - with a total of 30,018 gilts having EBVs as a result of the genetic evaluation. Of these gilts, 1862 were selected as breeding sows and consequently had their own performance record.
- Growth rate was subsequently recorded on 22,225 gilts and 22,450 young boars.
- Of these animals, 5,615 gilts and 3,773 young boars had backfat records.
Results
Litter Size
Mean EBVs of the top and bottom percentile groups differed by:
| - 0.96 piglets for litter size |
These differences in EBVs are reflected by the differences in mean phenotypic
performance of:
| - 0.91 piglets |
These results indicate that genetic improvement in litter size is possible due to the difference of approximately one piglet between the top and bottom percentile groups. This difference is due to the large variation in litter size and not a high heritability (Figure 1).
 |
Figure 1. Phenotypic Performance in Litter Size (Hermesch, et al., 1997a)
Growth and Backfat
Mean EBVs of the top and bottom percentile groups differed by:
| - 49 grams for growth rate in gilts and boars |
| - 2.20mm for backfat in gilts |
| - 2.30mm for backfat in boars |
This difference in EBVs corresponded well to the difference in mean phenotypic performance of:
| - 49 grams in gilts and 48 grams in boars for growth rate |
| - 2.40mm in gilts and 2.20mm in boars for backfat |
In addition, the slope of mean EBVs over all percentile classes is generally parallel with the slope of mean growth rate and backfat (Figure 2 and Figure 3). This indicates that EBVs do predict real differences in phenotypic performance as a result of genetic differences between animals.
 |
Figure 2. Mean EBVs for BF of gilts and young boars including mean phenotypic performances for each percentile group (Hermesch, et al., 1997b).
 |
Figure 3. Mean EBVs for ADG of gilts and young boars including mean phenotypic performances for each percentile group (Hermesch, et al., 1997b).
Conclusions
- PIGBLUP reliably predicts genetic difference between animals.
- Genetic improvement in litter size is possible using PIGBLUP. These genetic differences were reflected in phenotypic performances.
References
Hermesch S., H.-U. Graser, M. Mingay. 1997a. Lifetime Reproductive Performance of Gilts with High and Low EBVs for Litter Size. Proceedings of the 12th Association for the Advancement of Animal Breeding and Genetics Conference, Dubbo, April, 499-504.
Hermesch S., H.-U. Graser, M. Mingay. 1997b. Phenotypic Performance of Gilts and Young Boars can be Predicted from their EBVs for Growth Rate and Backfat. Manipulating Pig Production, edited by P.D. Cranwell, Australasian Pig Science Association, December, 169.
