Genetic Improvement of Feed Utilisation in Cattle and Pigs

A large scale feces sampling experiment to measure feed digestibility in pigs and calibrate NIR predictions is planned. In a pilot study, lignin was tested as a naturally in the feed occurring marker to measure digestibility. However, the lignin based measurements of digestibility were imprecise probably because the pigs were given some hay directly in the pens (which is compulsory by law) next to their normal ration. Hence, we decided to use AIA (Acid Insoluble Ash), which is absent in hay. Calibrating predictions based on NIR spectra of faeces samples have been successful achieving an accuracy of up to 70%. For pigs, more than 900 faeces samples were collected in 2019. We have developed NIR prediction equations for the estimation of fat and nitrogen contents in these samples. A large scale faeces sampling experiment for cattle has been carried out during the winter 2018/2019. Here we tested a panel of two naturally in the feed occurring markers to measure total feed intake, and thus roughage intake (since intake of concentrates is known in practice). It was concluded that a larger panel of naturally occurring markers is needed to measure feed intake. We have calibrated MIR spectra of faeces and/or milk samples predictions for the recording of feed intake in practice. The faeces and feed samples were collected on 25 firstcalving cows during 5 periods, where each period consists of 2 days. Feed intake was also predicted using the NorFôr system, and both predictions are being compared for their accuracy. The accuracy of both predictions was between 65-75%. In order to further improve prediction accuracies, we are (i) extending the experiment in order to obtain more data points; and (ii) combining both prediction models. Prediction accuracies were further improved by the us of deep learning methods. Dry matter intake prediction was improved to an accuracy of 0.82 using the following parameters: milk yield, concentration intake, days in milk, metabolic body weight, feed type, and NE mobilization. Large scale application of this prediction model could aid the breeding value estimation for feed intake in cows. Genetic parameters have been estimated for apparent total tract digestibility of dry matter, organic matter and dry matter. The heritabilities ranged from (0.13 - 0.18). Apparent totalt tract digestibility of dry matter and nitrogen were both favorable correlated to feed efficiency already included in the breeding goal, but were not the same trait (correlation < 1). This suggested that the current breeding goal will improve digestibility of dry matter and nitrogen, and could give added genetic progress in the traits.

En prediksjonsmodell for prediksjon av fôrfordøyelighet hos gris er klar for bruk i stor skala og er klar for implementering av Norsvin. En prediksjonsmodell for storskala prediksjon av fôropptak hos kyr ved bruk av melkespektra og NorFor-parametere er også utviklet. Dette kan implementeres i GENOs avlsprogram for fôreffektivitet i kombinasjon med faktiske fôropptaksmålinger.

The project aims to improve the efficiency of feed utilisation in cattle and pigs in order to keep pace with the expected growth of the Norwegian population, to reduce our reliance on foreign food/feed ingredients, to reduce green-house-gas emissions, and to reduce feeding costs, which currently constitute ~70% of the total costs, and are spiraling upwards. To achieve this the project focusses on: (1) feed/energy intake since this is a simpler to improve trait than feed efficiency and its closer related to the bio-economy of animal production; (2) the large scale recording of feed/energy intake of animals under practical circumstances by Infrared and Near-Infrared spectroscopy; and (3) by modelling feed intake data over time, developing genomic breeding value estimates and optimizing feeding rations the project results will be directly applicable in practice. The project focusses on the genetic improvement of feed efficiency since genetic improvements are cumulative, in the sense that a genetic improvement of feed efficiency of 2% per year yields a 20% more efficient animal production sector over a period of 10 years. A multidisciplinary, international project team has been built that contains expertise on genetics, animal nutrition, the biology of animal production practices, and spectroscopy. The project incorporates the relevant industry partners to ensure that the research activities will have a commercial impact.