Working with a recipe, every time that the diet composition changes, the sub-model is automatically invoked and calculates the individual responses for all the parameters described above. The total response is predicted through a summative approach. Then, starting from the current milk fat concentration as conventionally handled in Animal Inputs, it is corrected according to the total response and the final milk fat concentration and yield are predicted.
The figure shows a numerical example of the predictions:
Current MFC: 3.95%
Current MFY: 2.960 lbs/day
Predicted summative response for MFC: +0.046% (it could also be a negative value)
Predicted summative response for MFY: +0.034 lbs/day (it could also be a negative value)
Predicted MFC: 4.00% (3.95 + 0.046)
Predicted MFY: 2.994 lbs/day (2.960 + 0.034)
It also shows the milk fat concentration adapted from Zebeli et. al. (2008), just for comparisons purposes.
In this version of the sub-model we paid attention to the quantitative responses on milk fat and it is not able to predict the responses of milk fatty acids profile and composition.
CONCLUSION
The concentration and yield of milk fat is driven by the nutrition of the dairy cow; therefore, diets that allow an improvement in milk fat output would potentially be economically advantageous.
The NDS Milk Fat Sub-model seems to suggest that it is possible to develop an additive model of prediction of changes in MFC and MFY from the digestive flow of nutrients. It should be a first step towards a more robust model able to estimate variations in milk fat based on modifications in the supply of nutrients predicted from the dairy cow diet, which act as precursors or inhibitors of mammary fat synthesis.
Given the complexity of the biology related to milk fat yield and the variability of the factors involved in its prediction, no warranty or guarantee of animal performance is given or implied. The prediction represented by the NDS Milk Fat Sub-model is based upon the information provided by the user and RUM&N cannot be held responsible for misuse and/or misrepresentation of data.