Økt robusthet og sykdomstoleranse hos gris gjennom identifisering av nye fenotyper og utvikling av avlsmetoder

In this project we have genetically investigated more already available traits in the breeding program in order to improve the models for more efficient breeding for robustness in pigs. This includes sow death loss, which is used for sows that are found dead or culled on farm, body condition score, and the new trait longevity previously generated from this project. The results from this shows that genetic improvement for body condition score at the nucleus farms in Norway has beneficial effects for the sow lifetime performance on commercial farms across the globe. A validation of the results based on corrected phenotypes and estimated breeding values was to be carried out. Unfortunately, the data and number of genotyped animals was not large enough to get reliable results, and the validation have to be post-ponded until more data are available. Some work on improving the statistical models have been carried out. Topigs Norsvin annual magazine goes out to employees and clients worldwide, and this years issue will feature an article based on results from this project. The results have also been presented at Topigs Norsvin annual meeting about health and robustness. In line with the activities above we also continued working on utilizing the gene by environment interaction in better models. Main objective is to improve the selection of sires for commercial farm environments. Specific objectives investigated were: a) to assess the usefulness GxE models for genomic evaluation of sires for longevity traits, b) to compare the results obtained from analysis using traditional (A) and genomic (G) relationship matrices, and c) to investigate the efficiency of genomic breeding value (GBV) prediction over different environmental conditions through cross-validation. Results obtained so far suggest that use of genomic relationship increases the accuracy of prediction. In case of lgy12 (ability to stay in herd to second farrowing), the use of GxE component in the models did not increase the accuracy. However, in case of lgy5 (ability to stay in herd in more than 5 parities), the use of GxE component increased the accuracy by 1.5% in the models based on traditional A matrix and by 6.5% in the model using genomic relationship matrix. PigAtlas is a map of pigs scanned in the CT scanner at the boar testing station in Norsvin, based on the average of several hundred pigs. This map makes it possible to automatically find landmarks in the images to, for example, define the size of the muscle, bone, fat etc. This map is developed in a separate project, but in connection with this project it is specifically developed a new version (version 2) of the PigAtlas for higher-level segmentation procedure in order to measure the size of organs that may be related to stayability/robustness. The changes are based primarily on going from a volumetric (voxel based) to surface based segmentation. This has several advantages, including increased speed in the segmentation process and, hopefully, a more precise segmentation. The PigAtlas project has already contributed with phenotypes that are used for robustness related traits. For segmented shoulder blades, a high genetic correlation has been found with, among other things, shoulder ulcers. We also found that the shape of the shoulder blade is very inheritable, which in turn means that it is possible to genetically select for it. Segmentation of important internal organs such as the heart and lungs is an integral part of PigAtlas version 2. Such data are currently not analyzed, but the experience of the shoulder blades makes us optimistic considering that you can find inheritable relationships between PigAtlas-based phenotypes for heart / lungs (volume, density, etc.). It will also be natural to use the methodology developed for shoulder blades for other bones that you know can cause disease or lameness in the pig. In the vision lab for collection and analysis of video data that was established last year, we have been working to make data collection of 3D images more efficiently. Instead of capturing many pictures of each animal, we can now capture fewer images of sufficient quality where most of them are good enough to use directly in image analysis. It is now also possible to collect 3D video (for example, from above the pen) with a pics of (per today) approx. 10-15 images per second. One of the purposes of this is to capture movements over time in the pen so that we can follow the animal and be able to analyze behavior, tail biting, etc., all related to the robustness and longevity of the animal.

Høy produktivitet har vært hovedfokus i avlsprogrammene over hele verden, selv om flere studier har vist at produktivitet alene har ugunstige konsekvenser for en rekke helsemessige egenskaper. Det er derfor et stort behov for å utvikle avlsprogrammer som optimaliserer produktivitet på tvers av en rekke miljøer uten noen kompromisser mhp dyrenes helse og robusthet. I dagens svineavl, også globalt, er seleksjonskandidatene oppvokst under svært gode miljøforhold, i motsetning til de mer utfordrende miljøforh oldene som er utbredt i kommersielle besetninger. Hovedmålet med dette prosjektet er derfor å øke robustheten til den norske grisen slik at den tilpasses både nasjonale og internasjonale markeder. Med robusthet mener vi her «griser som takler høy produksj on under en rekke forskjellige miljøforhold», inkludert sykdomstoleranse og livstidsproduksjon. Et av prosjektets viktigste utfordringer er å avdekke hvilke fenotyper som best beskriver robustheten til grisen. I prosjektet skal det derfor samles in feno typer fra referansepopulasjoner med kommersielle miljøforhold for å studere genotype-miljø samspill mhp robusthet. Det blir også tatt prøver for å analysere parametere som kan være potensielle indikatorer på sykdomstoleranse. Sikkerheten på avlsverdien fo r robusthet er svært lav på grunn av tidlig seleksjonstidspunkt (dvs. lite informasjon på dyret selv), egenskapenes natur, mm. Dette skal løses ved å implementere genominformasjon fra dyrene i referansepopulasjonene. Norsvin vil med resultater fra dette prosjektet kunne utvikle nye genombaserte avlsmodeller og egen internasjonal «robusthetsindeks» som skal sikre at vi selekterer avlsråner med det beste genetiske potensialet for robusthet, tilpasset markeder med høyere smittepress. Denne strategien vil s tyrke de norske svinerasenes posisjon i utlandet, men også gi mer robuste dyr i Norge som krever med mindre behov for behandlinger og gir reduserte rekruteringskostnader for bonden.