Modelling feed intake and milk yield responses to different grass ley harvesting strategies

Abstract A meta‐analysis of feeding trials using grass silages was conducted to predict production responses for dairy cows fed grass silage. They were divided into two subsets: 69 diets from 11 studies were used for comparison of silages made from primary growth and regrowth grass (harvesting subset), and another 157 diets from 24 studies were used for comparison of digestibility influenced by the maturity of grass ensiled (D‐value, digestible organic matter in dry matter) (maturity subset). The minimum prerequisite for an experiment to be included in the data set was that milk production, feed intake, silage characteristics and concentrate ingredients were reported. Both subsets were analysed using the mixed model procedures of SAS. The mean response in dry‐matter intake (DMI) and silage DMI to improved silage D‐value was 0.0175 and 0.0161 kg per unit D‐value (g/kg DM) respectively. The average increase in milk and energy‐corrected milk yield was 0.30 and 0.37 kg per 10‐unit increase in silage D‐value respectively. Milk protein concentration increased, and fat concentration tended to increase with enhanced silage D‐value. Each 10‐unit increase in D‐value reduced milk yield by 0.092 kg at a given dietary metabolizable energy intake (MEI), suggesting that the ME concentration of high D‐value silages was overestimated. Cows fed regrowth silage produced 0.55 kg/day more energy‐corrected milk than those fed primary growth silage at a given dietary MEI. The prediction models can be used to improve ration formulation systems or incorporated into economic models for optimizing milk production in various farming systems.