Milking the data: A new reproduction model for Holstein Dairy Cows
The animal Article of the Month for August is entitled ‘ An individual reproduction model sensitive to milk yield and body condition in Holstein dairy cows’
To face rapid changes in economic or climatic conditions and to maintain their competitiveness, dairy producers increasingly need to anticipate the consequences of their management decisions. Individual-based herd simulation models are a valuable and widely-used solution to predict milk and meat production responses to management practices. Reproduction is a key-driver of milk production and feed intake, at the animal level owing to physiological changes, as well as at the herd level owing to demographics changes (calving dates, culling rate). Its influence has been magnified by the decrease in reproductive performance over the last decades. At the animal level, management strategies, like feeding strategy or milking frequency, have an effect on milk yield and body reserves. Both of these factors are known to influence reproductive performance. Our objective was therefore to build an up-to-date animal reproduction model sensitive to both milk yield and body reserves.
A dynamic and stochastic individual reproduction model of Holstein dairy cows was built, mainly from data of a single recent long term experiment. This model covers the whole reproductive process and is composed of a succession of discrete stochastic events, mainly calving, ovulations, conception and embryonic loss. Each reproductive step is sensitive to milk yield or body condition score levels or changes. The model takes into account recent evolutions of reproductive performance in dairy cows, concerning the calving-to-first ovulation interval, the cyclicity pattern (length of normal cycles, prevalence of abnormal cycles such as prolonged luteal phases), the oestrus expression and the pregnancy establishment (conception, early and late embryonic loss). A sensitivity analysis of the model to milk yield and body condition score at calving was performed. To validate the model, the simulated global reproduction performance was compared i) to observed data from the database used to build the model and ii) to average values reported in the literature.
The animal reproduction model, built with an analytic approach of the successive reproductive steps, succeeded in simulating consistent aggregated reproduction outputs. It was able to well predict the overall reproductive performance observed in commercial farms in terms of both success rate (recalving rate) and reproduction delays (calving interval).
This model has the purpose to be integrated in dynamic individual-based herd simulation models to usefully anticipate and assess the consequences of management strategies, in terms of reproduction or feeding practices, on dairy cows’ reproductive performance and thus on calving patterns, culling rates and milk production at the herd level.
Authors: L. Brun-Lafleur, E. Cutullic, P. Faverdin, L. Delaby and C. Disenhaus
The animal Article of the Month is selected by the Editor-in-Chief and is freely available for one month