Know more

Estimating body reserve of dairy cows

The body reserves are a vital and dynamic resource in dairy cows that enables them to face the variations of energy intake. However, their status and changes during lactation determine production performance but also indirectly reproduction and health status of the cow. The difficulty lies in the precise, objective and frequent measurement of these phenotypes. Body condition scoring (BCS) is the most typical method used: it is simple, inexpensive but subjective and not sensitive. The development of new imaging technologies offers new exciting prospects for phenotyping of animal traits. The project objective was to develop and validate a method estimating BCS from the comprehensive information of three-dimensional (3D) image from the dorsal area of the cow. This work has been done in collaboration with the 3DOuest company specialized in the acquisition and processing of 3D images.

Precise, objective and frequent measurement with imaging technologies

The work has shown proof of concept by proposing a methodology with a prototype and evaluating the properties of this technic. The extraction of four anatomical points of the basin has been used to standardize 3D surfaces. Validation results and repeatability of the 3D-imaging BCS estimation via manual location of these points on pictures have demonstrated the ability to predict the BCS. This extraction has recently been automated. The 57 3D-surfaces used for calibration covering a wide range of BCS and animal size were projected in the coordinate system defined by the principal components characterizing all these 3D surfaces. The BCS is estimated by multiple linear regression of the coordinates of surfaces on the principal components.

The two methods - one with manual extraction, the other with automatic extraction points - were compared for external validation and repeatability. External validation was performed on a population of 25 cows. A control population was scanned several times in one day to estimate the repeatability of the method. The two methods have similar external validation results but are more repeatable than the reference BCS (σ = 0.20 versus 0.28 for the 3D BCS). The first results of phenotyping BCS with this method are encouraging automation and opens the possibility to increase the number of measures to improve the estimate. These techniques offer prospects of high throughput phenotyping body condition, essential for animal robustness and adaptation studies.

Characterize more aspects of animal morphology

Currently, a simplified prototype that scans more rapidly cows is being tested, including reducing the number of unusable 3D-areas related to movement of cows during the scan. The interest in 3D image processing for animal phenotyping seems demonstrated and will be developed for evaluating other phenotypes related to other aspects of morphology in collaboration with 3DOuest and Livestock Institute. An application project to the body condition scoring of dairy goats is also underway.

New perspectives for livestock precision farming

This prototype will be developed and used in other experimental sites to better assess changes in body reserves and standardize the measures. Moreover, these methods are fast and cheap. They therefore offer potential applications in breeding. Valuation projects for the breeding council are under discussion with the relevant partners. These tools could eventually provide new perspectives for livestock precision farming and objectification of the animal welfare.

Beyond the European scientific community (EAAP) and American (ADSA), the method was presented to Paris international agricultural show, the Prairiales show (video, in french), the morning of Rennes Atalante and to the international exhibition for animal productions.

References

• Fischer, A., Luginbühl, T., Delattre, L., Delouard, J. M., Faverdin, P. (2015). Rear shape in 3 dimensions summarized by principal component analysis is a good predictor of body condition score in Holstein dairy cows. Journal of Dairy Science, 98 (7), 4465-4476. [DOI]
• Fischer, A., Luginbühl, T., Delattre, L., Faverdin, P. (2015). Analyzing the rear shape of dairy cows in 3D to better assess body condition score. In: 2015 Joint annual meeting abstract book (p. 814). Journal of Dairy Science, 98. Suppl. 2. Presented at ADSA-ASAS Joint Annual Meeting, Orlando, USA (2015-07-12 - 2015-07-16). Nouvelle Orléans, USA : ADSA - ASAS. [lien]
• Fischer, A., Luginbühl, T., Delattre, L., Delouard, J., Faverdin, P. (2014). Analysing the back of dairy cows in 3D imaging to better assess body condition. In: Book of Abstracts of the 65th Annual Meeting of the European Federation of Animal Science (p. 127). Annual Meeting of the European Association for Animal Production, 20. Presented at 65. Annual Meeting of the European Federation of Animal Science (EAAP), Copenhague, DNK (2014-08-25 - 2014-08-29). Wageningen, NLD : Wageningen Academic Publishers. [lien]
• Fischer, A., Lunginbuhl, T., Delattre, L., Delouard, J., Faverdin, P. (2014). Améliorer la mesure de l’état des réserves corporelles des vaches laitières en analysant la surface 3D du dos de la vache. In: 21èmes Rencontres autour des Recherches sur les Ruminants (p. 23-26). Rencontres autour des Recherches sur les Ruminants, 21. Presented at 21. Rencontres autour des Recherches sur les Ruminants, Paris, FRA (2014-12-03 - 2014-12-04). Paris, FRA : Institut de l'Elevage - INRA. [lien]