The objective of this study was to compare feed intake and milk production responses in continuous
and change-over design dairy cow experiments with a meta-analysis. The dataset included
988 treatment means (354 in continuous and 632 in change-over trials) from 204
studies. Sub-datasets from studies investigating the effects of the amount (n=217) and
crude protein (CP) concentration of concentrate (n=336 diets) were analyzed separately.
Mixed model regression analysis with a random study effect was used evaluate the effects of
experimental design on feed intake and milk production responses. Concentrate dry matter
(DM) intake, dietary CP or metabolizable protein, total DM intake index and metabolizable
energy intake were used as independent variables in univariate models. In addition, multiple
regression models were developed from the continuous and change-over trial data and
cross-validated using the other data. Generally the intake and milk production responses did
not depend on experimental design, and the P-values for the interactions between the experimental
design and independent variables were high. When the data from studies displaying
large variation in the intake potential were analyzed separately, change-over designs tended
to underestimate milk production responses. However, up to the range of 5 kg/day in DM intake
potential within a study the feed intake and milk production responses to the changes
in nutrient supply were similar irrespective experimental designs. The analysis of five cyclicchange-
over studies including 60 diets indicated that there was no relationship between the
direct and carry-over effects of dietary treatments. The mean squares of carry-over effects
were on average similar to the error mean squares. The probability of the significance of
carry-over effects was not different from the random effects. It is concluded that the changeover
designs are as accurate as continuous designs in estimating feed intake and milk production
responses, excluding studies with extreme treatments resulting in extensive mobilization
of body fat and protein.
The objective of this study was to compare feed intake and milk production responses in continuousand change-over design dairy cow experiments with a meta-analysis. The dataset included988 treatment means (354 in continuous and 632 in change-over trials) from 204studies. Sub-datasets from studies investigating the effects of the amount (n=217) andcrude protein (CP) concentration of concentrate (n=336 diets) were analyzed separately.Mixed model regression analysis with a random study effect was used evaluate the effects ofexperimental design on feed intake and milk production responses. Concentrate dry matter(DM) intake, dietary CP or metabolizable protein, total DM intake index and metabolizableenergy intake were used as independent variables in univariate models. In addition, multipleregression models were developed from the continuous and change-over trial data andcross-validated using the other data. Generally the intake and milk production responses didnot depend on experimental design, and the P-values for the interactions between the experimentaldesign and independent variables were high. When the data from studies displayinglarge variation in the intake potential were analyzed separately, change-over designs tendedto underestimate milk production responses. However, up to the range of 5 kg/day in DM intakepotential within a study the feed intake and milk production responses to the changesin nutrient supply were similar irrespective experimental designs. The analysis of five cyclicchange-over studies including 60 diets indicated that there was no relationship between thedirect and carry-over effects of dietary treatments. The mean squares of carry-over effectswere on average similar to the error mean squares. The probability of the significance ofcarry-over effects was not different from the random effects. It is concluded that the changeoverdesigns are as accurate as continuous designs in estimating feed intake and milk productionresponses, excluding studies with extreme treatments resulting in extensive mobilizationof body fat and protein.
การแปล กรุณารอสักครู่..
