Modeling post absorptive amino acid metabolism in dairy cattle
MetadataShow full item record
The purpose of this research was to evaluate four objectives: 1) update and evaluate predictions of essential amino acid (EAA) outflows from the rumen, 2) predict EAA use and release by the portal drained viscera (PDV) and liver (LIV) of dairy cows, 3) predict EAA use by the mammary (MAM) and non-splanchnic, non-mammary (OTH) tissues, and 4) predict milk protein production from MAM use. To evaluate the first objective, a model was constructed using previously derived equations for ruminally undegraded (RUP), microbial (MiP) and endogenous protein (EndP) flow from the rumen and refit to literature data. Corrections were included in the model to address recovery of EAA during 24-h acid hydrolysis. Upon initial evaluation, all EAA, except Leu, were over predicted and slope bias (P < 0.01) was present for all except Met and Leu. Because of the bias, residuals were regressed on the EAA from each protein flow and adjustments were made to the protein flows. The added adjustments removed all mean bias for the EAA; however, a small slope bias was introduced for Lys and Thr. To evaluate the second objective, equations of Hanigan et al. (2004b) were tested and modifications were made to determine which equation form best represented EAA use by the tissue. Upon initial evaluation of the PDV model of Hanigan et al. (2004b), significant slope bias was present and addressed by deriving alternative forms of the equation. Initial predicted EAA use displayed a mean bias ranging from 0.15 to 45 % and a slope bias ranging from 0.02 to 76% mean square error. The alternative equation forms derived reduced the overall mean and slope bias and improved other fit statistics (RMSE, CCC). To evaluate the third objective, previously derived equations from Hanigan et al. (1998b) were tested using literature data and modifications were made to address deficiencies for each EAA. Upon initial evaluation of the MAM model, significant mean and slope bias was present and was further addressed by derivation of alternative equation forms. Initial evaluation of the OTH model displayed significant mean and slope bias for majority of the EAA ranging from 0.3 to 26 % for mean and 46 to 61 % for slope. For the last objective, several models, both linear and non-linear were evaluated to determine which EAA have a significant impact on milk protein. All models derived has prediction errors below 18-20 % which is comparable or a s light improvement as compared to previous literature data (Moraes et al., 2018). Overall, the equations evaluated show promise in accurately predicting dietary EAA from the time of absorption to their use within the tissues (PDV, LIV, MAM, and OTH) and further impact on milk protein production.
- Masters Theses