Predicting ruminally undegraded and microbial protein flows from the rumen

摘要

The objectives of the present work were (1) toidentify the cause of the linear bias in predictions ofrumen-undegradable protein (RUP) content of feeds,and devise methods to remove the bias from predictionequations, and (2) to further explore the impactof rumen-degradable protein (RDP) on microbial N(MiN) outflow from the rumen. The kinetic modelused by NRC (2001), which is based on protein fractionationand rates of degradation (Kd) and passage(Kp), displays considerable slope bias (−0.30 kg/kg),indicating parameter or structural problems. RegressingKp by feed class and a static adjustment factorfor the in situ–derived Kd on observed RUP flowscompletely resolved the slope bias problem, and themodel performed significantly better than models usingunadjusted Kd and marker-based Kp. The Kd adjustmentwas 3.82%/h, which represents approximately a50% increase in rates of degradation over the in situvalues, indicating that in situ analyses severely underestimatetrue rates of protein degradation. The Kp forconcentrate-derived protein was 5.83%/h, which wasslightly less than the marker-predicted rate of 6.69%/h.However, the derived forage protein rate was 0.49%/h,which was considerably less than the marker-based rateof 5.07%/h. Compartmental analysis of data from asingle study corroborated the regression analysis, indicatingthat a 25% reduction in the overall passage rateand an 87% increase in the rate of degradation wererequired to align ruminal N pool sizes and the extent ofprotein degradation with the observed data. Therefore,one must conclude that both the in situ–derived degradationrates and the marker-based particle passagerates are biased relative to protein passage and cannotbe used directly to predict RUP outflow from the rumen.The effects of RDP supply on microbial nitrogen(MiN) flow were apparent when intakes of individualnutrients were offered but not when DM intake andindividual nutrient concentrations were offered, due tocollinearity problems. Microbial N flow from the rumenwas found to be linearly related to ruminally degradedstarch, ruminally degraded neutral detergent fiber(NDF), RDP, and forage NDF intakes; and quadraticallyrelated to residual OM intake. More complicatedmodels containing 2- and 3-way interactions amongnutrients were also supported by the data. IndependentMiN responses to RDP, ruminally degraded starch, andruminally degraded NDF aligned with the expectedresponses to each of those nutrients. Nonlinear representationsof MiN were found to be inferior to the linearmodels. Despite using unbiased predictions of RUP andMiN as drivers of AA flows, predictions of Arg, His, Ile,and Lys flow exhibited linear slope bias relative to theobserved data, indicating that representations of theAA composition of the proteins may be biased or theobserved data are biased. This is an improvement overthe NRC (2001) predictions, where bias adjustmentswere required for all of the essential AA. Despite thebias for 4 AA flows, the revised prediction system wasa substantial improvement over the prior work.