The main objective of the current study was to compare
feed intake and milk production responses in continuous
and change-over design milk production trials with dairy
cows. Provided that marginal intake and production responses
are not different, experimental resources could be
used more efficiently using change-over rather than continuous
designs due to the better error control, higher statistical
power and smaller number of animals needed in changeover
trials. However, there are concerns about the residual
effects in change-over designs and that the short-term production
responses may not reflect the whole lactation responses
(e.g. Morris, 1999). On the other hand, confidence
intervals of the differences between treatment means can
be large in continuous trials, even with a large number of
cows per treatment