Influence of passage model, sampling site, and marker dosing time on passage of rare earth-labeled grain through Holstein cows.

Three duodenally cannulated, lactating Holstein cows were dosed with rare earth-labeled grain to evaluate effects of passage model, sampling site, and marker dosing time on digesta passage parameters. Cows were given ad libitum access to feed twice daily. Rare earth-labeled grain (applied by the 24-h immersion technique) was fed immediately before (Dy) or 2 h after (Yb) the morning feeding, and duodenal digesta and feces were sampled. Marker excretion curves were fit to a two-compartment, biexponential model, using curve peeling or to a series of two-compartment models, with one to six orders of gamma time-dependency in the fast compartment, using nonlinear regression. Passage estimates from the curve-peeled, biexponential model were similar to those from the best fit of the nonlinear models, which had three orders of gamma time-dependency. Ruminal passage rate of grain, averaged across models, sampling site, and dosing time, was .077/h. Estimates of time to first appearance of marker at the sampling site and mean retention times were longer (8 vs 1 h and 25 vs 17 h, respectively), but passage rates were similar, when determined from fecal compared with duodenal samples (P less than .05). Marker dosing time did not influence any of the parameters. It is concluded that both curve-peeling (linear regression) and nonlinear regression methods can be equally useful for evaluating passage kinetics of grain in dairy cows and that sampling site and time of marker dosing have little effect on passage parameter estimates.