This project aims at determining the main genetic and environmental factors determining fine milk composition in fat and protein. It includes 3 tasks but is a part of a larger programme with six tasks, three of them being supported by non-ANR fundings : (1) adapt Mid Infra Red (MIR) spectrometers of the different milk analysis laboratories to export spectra, develop a dedicated data base ; (2) develop calibration equations to predict fatty acid composition of milk in the three dairy species through MIR and chromatography data, from samples collected in four INRA farms; (3=task1) develop a reference method for protein measurement and derive calibration equations by the analysis of both reference and MIR data ; (4=task2) collect MIR spectra, milk and blood samples, and diet and management information from 12,000 cows and 3,000 ewes, as well as complementary information from the national database ; predict fatty acid and protein concentration from MIR spectra ; (5=task3) genotype a part of these samples with a SNP chip and perform QTL fine mapping, determine the effect of management and feeding policies on fatty acids and protein profiles in milk. (6) develop management and feeding strategies for general recommendation. Steps (3), (4) and (5) are included in this Genomic project and are denoted TASK1, TASK2, and TASK3, respectively. Specific aims of TASK1 is first to establish a detailed qualitative and quantitative profile of the 12 main milk proteins: caseins, alpha-lactalbumin, beta-lactoglobulin, lactoferrin, lactoperoxydase, using highly resolvent reference steadfast (RP-HPLC) and innovative (UPLC) techniques combined with Mass Spectrometry (MS). The challenge will be to identify the known (and eventually unknown) genetic variants and different isoforms and to determine their relative proportions. The second aim will be to derive from the detailed phenotype descriptions a calibration equations allowing the prediction of the milk protein fraction composition on a large number of milk samples by MIR. TASK2 aims at collecting all the samples and informations required by the project in a large number of commercial farms, in cattle as well as in sheep. 12,000 Holstein, Normande and Montbéliarde cows and 3,000 Lacaune ewes are targeted, based on prior information in the national data base. In cattle, herds are defined according to their size, their geographic concentration around labs participating to the project, their breeds, and their genetic diversity. Four kinds of information are collected: (1) three milk samples per lactation, (2) MIR spectra from these samples, (3) blood samples from the corresponding cows, for DNA extraction, (4) diet and management information for a good interpretation of the results. TASK3 is the genotyping step. About two thirds of the blood samples are selected for DNA extraction and genotyping, on the basis of MIR results. Depending on the cost of the chips in 2010, two options are possible. The most conservative is proposed as a minimum. A dedicated 1536-SNP chip will be developed to target all regions of interest, defined on the basis of external knowledge (results from Le Pin QTL experiment, Sarde data, other literature data, QTL known to affect total fat or protein). The chip will be completed by additional markers from the remaining part of the genome at a lower density. Data will be analysed by combining linkage and linkage disequilibrium. The effect of all environmental and diet factors will be estimated.

MOUVE is is a research consortium integrated by INRA, CIRAD and Cemagref researchers in partnership with universities and research centres in Developing Countries. The purpose of the project is to produce knowledge to understand, model and summarize the necessary implantation and development conditions of Ecological Intensification (EI), and its possible implications in livestock farming systems at the local level. It aims at contributing to the scientific domain of "local development – livestock " interactions in relation with the ecological intensification and sustainable development. Livestock farming systems approaches, sociology of professions, landscape ecology and geography are the main scientific domains contributing to the pluridisciplinary of the project. From a methodological point of view, MOUVE will develop comparative analysis of 8 contrasting extensive livestock farming regions (2 in South America, 1 in African Sahel, 2 in Mediterranean basins, 3 in French mountains). The contrasts rely on i) landscape (rangeland or a mix with crops and forest); ii) livestock production stakes in relation to production or environment,; and iii) livestock dynamics (increasing or decreasing numbers, relocation). The project will treat simultaneously 1) what are the local stakeholders attempts in relation to livestock and sustainable development in times of ecological intensification. This will mobilize participatory methods, and will help to precise how EI is re- formulated at that local scale. 2 ) what are the characteristics of the "livestock – local development" interactions considering i) the diversity and complementarities of livestock farms, ii) the land use and the ecosystem services, iii) related people debates on what is a good farmer facing times of changes and uncertainties ; 3) the development conditions for livestock farming changes and collective initiatives that fits with EI, and the consequences on the ecosystems and farmers identities. 4) Scenarios methodologies will be utilised to integrate the knowledge with local actors’ one. Our final aim is to produce knowledge and development indicators at the local level to discuss the future of livestock and livestock farming within international arenas, such as FAO, LEAD (Livestock Environment and Development Group).