The effect of pregnancy on the rate of decline in milk yield from month 5 of gestation onwards has been documented for some time, (Brody et al., 1923), and has been a key factor in choice of calving interval. There has been renewed interest in the possibility of extending the traditional 365 day calving interval, because of higher genetic merit cows and difficulties over live calf exports. The objective of the current work was to look for an interaction between pregnancy status and concentrate supplementation.

Sixty-five autumn-calving Holstein-Friesian cows (26 non-pregnant (B) and 39 pregnant (P)) were selected for the experiment; prior to the experiment they were managed on a flat-rate concentrate allocation with forage available ad libitum. The pregnancy-status groups were randomly allocated to receive ad libitum grazed grass and either 0 or 3 kg of 18% crude protein dairy concentrate. The animals were grazed and milked together as one group. The non-pregnant cows were deliberately not inseminated for management reasons (e.g. failure to use Calan gates). During the 10 weeks before imposition of the concentrate treatment, there was no significant difference in milk yield between treatment groups (21.9 (s.e.=1.71) kg/day). The main recording period ran from 10 June until 16 September 1996. Milk yield was recorded daily, milk composition, live-weight and condition scores fortnightly. Data were summarised as means (for milk yield and composition) or slopes (rates of change; for live-weight and condition score (CS; 0-5 scale)) over the period from 3 to 13 weeks after the imposition of the concentrate treatment.