Comparison of Two Models Used to Estimate In Situ Nitrogen Disappearance

Abstract

Inclusion of lag time into a model describing in situ N disappearance influenced the parameter estimates describing N degradability and improved the goodness-of-fit of the model to data describing aeschynomene hay and alfalfa meal digestion. Ruminal N disappearance (%) was described using two models that assumed: 1) digestion of the potentially digestible but insoluble N fraction was initiated immediately upon exposure to ruminal contents, or 2) digestion of this fraction was preceded by a lag period during which hydration and microbial attachment is presumed to occur prior to the initiation of digestion. Digestion was assumed to occur from a homogeneous potentially digestible but insoluble N fraction by a first order process described by a rate constant, k. Incorporating a lag term resulted in an increase in the rapidly soluble N fraction, a decrease in the potentially digestible but insoluble N fraction, and a decrease in the total digestible N fraction. The two forages had different k values when fit to the nonlagged model but not when fit to the lagged model. The residual standard deviation was smaller for the lagged model, and the fit of predicted values to observed values was improved.