Post-doctoral position H/F

Climate change poses the risk of more variable forage production in terms of quantity and quality for feeding domestic livestock (particularly in the grass-based farming systems such as in the Massif Central). Indeed, in the event of drought or extreme heat, the quantity and quality of grass and forage produced will decrease. Adapting livestock farming to climate change therefore requires the selection of resilient animals capable of adapting to temporary variations in nutrient intake. When faced with a nutritional challenge, lactating cows reduce milk production and mobilise body reserves. These adaptations vary from animal to animal, depending on a number of factors (genetics, age, physiological stage, etc.) (Friggens et al. 2022). The responses observed may concern physiological and/or behavioural traits. An animal's resilience is usually considered in terms of a single trait (e.g. variations in milk production). However, the physiological responses are not independent of each other. A multi-trait approach is needed to address the complexity of adaptations. The aim of your work will be to develop multi-track analysis methods to understand the links between physiological and behavioural responses in order to identify the phenotypes that will enable us to select the most resilient individuals when facing nutritional challenges. We have a database (acquired as part of the European GENTORE project https://www.gentore.eu/) of meteorological, physiological (milk production, body mobilisation indicators) and behavioural (activity recorded by sensors) data, which was collected from 62 dairy and suckler beef cows (breeds characterised by contrasting milk potential and therefore potentially associated with different adaptive profiles). You will use this data to analyse the behavioural responses of cows to nutritional challenges and to explore the links between productive, physiological and behavioural responses in order to characterise the resilience of the animals. Time series modelling techniques (e.g. registration, wavelet transform, Markov chains) will be used to analyse the kinetics of the different responses of the animal in order to capture their complexity. These kinetics (recorded on different time scales) will then be combined to identify different adaptive profiles of cows with contrasting milk potential (Holstein, Montbéliarde and Charolais).