Effets du génotype de maïs ensilage sur les performances zootechniques de vaches laitières

French. Au cours de 4 années d’expérimentation avec des vaches laitières, différents hybrides de maïs choisis pour leur valeur énergétique mesurée sur moutons standards ont été comparés en ensilage de la plante entière. Les hybrides dont la valeur énergétique estimée sur les moutons était la plus élevée permettent des productions de lait standard supérieures de 1 à 2 kg et des reprises de poids pouvant atteindre 300 g de plus par jour. Les écarts de quantités de matière sèche d’ensilage ingérées ont également atteint 1,5 kg à teneur en matière sèche et teneur en grain constantes. Si les mesures faites sur moutons, comme les mesures de solubilités enzymatiques, permettent nettement de distinguer les hybrides à bonne valeur énergétique des hybrides à faible valeur énergétique, elles ne permettent en revanche pas de classer correctement les hybrides au sein de chacun de ces deux groupes. Enfin, aucune des mesures effectuées sur moutons ou in vitro ne permet de prédire la variabilité observée pour l’ingestibilité.

English. Genetic variation in the feeding efficiency of different maize genotypes was evaluated from experiments with dairy cows. The maize genotypes were chosen according to their variable digestibility range, determined in previous experiments with standard sheep. The variations in milk yield observed with dairy cows fed a low, respectively a high, digestible or energizing hybrid were between 1 to 2.0 kg per animal, per day. Moreover, the differences in body weight gain reached up to 300 g per day. Hybrids having similar energy values acted similarly on milk yield, but the intake could be very different when fed to dairy cows, even if no difference in intake was observed in sheep measurements. The 1.5 kg higher silage maize intake of DK265 was mostly converted into a body weight gain, and not in increasing the milk yield, probably because the milk yield potentiel of the animals was soon reached. It appeared that the hybrids which had a low predicted energy value from the sheep measurements also had a lower performance in cows. Dairy cow performances also appeared to concur with the enzymatic solubility values of the whole plant. However, only good hybrids could be distinguish from poor ones, but the variation within good, respectively poor, hybrids could not reliably be investigated with the enzymatic solubility test used. The variation in ingestibility could not be predicted. For maize breeders, the prediction of an hybrid efficiency for use in dairy cow rearing must include digestibility and ingestibility factors. Enzymatic processes could probably be used to avoid drift towards poor feeding values with hybrids bred for a higher stalk strength. The most important challenge for maize breeders will most likely be the determination of a criterium for the intake prediction. The stover digestibility and starch content can be reliably predicted through the use of NIRS (near infra-red reflectance spectroscopy) calibrations. For a stock-breeder, a pertinent choice of a maize hybrid might allow significant concentrates and money saving.