Development of maintenance energy requirement and energetic efficiency for lactation from production data of dairy cows

A total of 139 treatment mean data of lactating dairy cows, offered mixed diets of concentrates and silages of grass (n=33) and maize (n=5) ad libitum, were obtained from 12 long term feeding experiments (8–10 weeks/period). In these experiments metabolisable energy (ME) concentrations in diets were either directly measured or estimated from in vivo digestibility data. Milk energy output (El) was calculated from milk composition and live weight change (LWC) was estimated from linear regression. Energy balance (E(g)) and then El((0)) (E(l) adjusted to zero E(g)) were calculated using different energy contents per unit of LWC and energetic efficiencies as suggested in energy feeding systems [Livest. Prod. Sci. 5 (1978) 331; Ruminant Nutrition-Recommended Allowances and Feed Tables (1989); Nutr. Abstr. Rev. 60 (1990) 729; Energy and Protein Requirements of Ruminants (1993); Feeding Standards for Australian Livestock-Ruminants (1990); Nutrient Requirements of Dairy Cattle (2001)]. The objectives of the present study were to use these treatment mean data to derive the ME requirement for maintenance (ME(m)) and the efficiency of ME use for lactation (k(l)) for dairy cows and then validate these ME(m) and k(l) values using published calorimetric data. In the present study, ME intake, milk yield and LWC ranged respectively from 116 to 286 MJ/d, from 10.0 to 41.4 and from −0.405 to 0.825 kg/d. In the various feeding systems the energy content per unit of LWC and the energetic efficiencies are different, thus the mean calculated E(l(0)) values varied marginally (from 0.707 to 0.721 MJ/kg(0.75)). Consequently, the MEm and kl values, derived from the linear regression of ME intake against El(0), ranged respectively from 0.59 to 0.62 (MJ/kg(0.75)) and from 0.64 to 0.67. However, a parallel test on the linear relationships between ME intake and E(l(0)) revealed that the above difference in El(0) between these systems had no significant effects on either the intercepts or the slopes, indicating that there were no significant differences in ME(m) or k(l )values derived from these systems. The mean ME(m) and k(l )values across the feeding systems therefore were 0.60 (MJ/kg(0.75)) and 0.65 respectively and are within the range of ME(m) or k(l) values reported recently from the calorimetric data of lactating dairy cows. The ME(m) and k(l)values derived from the present production data set were then used to validate calorimetric data of lactating dairy cows (experimental mean, n=49) published since 1976. This demonstrated that the predicted ME requirement was comparable to the actual ME intake.