Evaluation of two feeding strategies with a herd model integrating individual variability

Abstract

Individual variability is a key element to understand and predict herd response to management practices. To study how individual variability is generated within the herd, an individual-based herd model was developed. It was applied to intensive dairy goat systems. The model combines a decisional with a biological sub-model. The decisional sub-model represents technical decisions relative to reproduction, feeding and replacement management. The biological sub-model represents each individual within the herd and is based on a dairy goat model. This latter simulates body weight and milk production dynamics throughout each goat’s productive life. Individual goat performance depends on its production potential, physiological regulations and responses to diet. As each goat model has its own individual performance scaling and its own operation driven by management decisions, individual variability is an output of simulation. In the present work, the herd model was used to simulate two feeding strategies involving two-step (2S) or five-step (5S) feeding sequences. The 5S feeding sequence is based on a regular adjustment of the quantity of concentrate feedstuff throughout lactation whereas the 2S feeding sequence is based on a single adjustment in late lactation. Results show that increasing the number of steps within the feeding sequence did not lead to a major difference in terms of milk production and feed cost. When focusing at individual level, results show that the 2S feeding sequence led to a greater proportion of goats whose body weight was lower than their potential body weight. For the 2S feeding sequence, production and efficiency relied on goats which tended to lose more weight than those fed with the 5S sequence. Hence, the same level of herd production and efficiency can be achieved with different biological underlying processes. This result stresses the importance of understanding the mechanisms behind overall performance to achieve a sound evaluation of management effects.