Unfortunately the rumen outflow equations used by Rumen8 are non-linear so the many low cost linear optimisers available cannot be used. Non-linear optimisers are more expensive and usually require a license fee for every copy of Rumen8 which would mean we could not provide it free of charge. Fortunately, the Solver add-on in Excel can handle simple non-linear optimisation so Rumen8 copies the diet and cow into Excel, optimises it and then brings the solution back into Rumen8. It is a clumsy workaround but enables us to keep Rumen8 free for users.
It has been well recognised over the last 15 years or so that the AFRC (1993) feeding standard significantly underestimated the requirement for Metabolisable Protein (MP) in dairy cows. This underestimation has been corrected in the successor to AFRC (1993), known as Feed into Milk (Thomas, 2004).
Rumen8, being based on AFRC (1993) for predicting requirements for MP (as well as ME), has taken a pragmatic approach by adopting a "MP requirement correction" to bring AFRC predictions of MP requirements in line with our current understanding of MP demand by dairy cows. We have adopted a "MP requirement correction" value of 1.25.
As part of a review of 11 applied dairy nutrition models used in Australia (Little et al, 2009), requirements for MP by dairy cows were estimated using various Australian and international dairy nutrition models. The review assessed 7 diets for dairy cows at various stages of lactation, with diets ranging from grazed pasture with minimal amounts of supplement to full TMR diets without any grazed pasture. The review included the models CPM-dairy, AminoCow, Feed into Milk and Rumen8 as well as a number of others.
For the 7 diets assessed as part of the review, requirements for MP as predicted by Rumen8 (i.e. AFRC 1993 * 1.25) were close to the MP requirements predicted by Feed into Milk (mean 2.10 vs 2.03 kg/cow/day). The mean ratio for MP required according to Feed into Milk vs MP according to Rumen8 was 0.97 (range 0.92 - 1.04).
The US dairy nutrition model AminoCow predicted that MP requirements were significantly lower than that predicted by Rumen8 (mean 1.65 vs 2.10 kg) giving a ratio of Amino Cow vs Rumen8 of 0.78 (range 0.71-0.87). On the other hand, the US model CPM-Dairy predicted MP requirements significantly higher than that predicted by Rumen8 (mean 2.49 vs 2.10 kg) giving a ratio of CPM-dairy vs Rumen8 of 1.21 (range 1.09-1.37).
Our suggested "MP requirement correction" of 1.25 leads to 'middle of the road' estimates of MP requirements for dairy cows. Should experienced users of Rumen8 wish to increase MP requirements to be more in line with MP predictions made by CPM-Dairy, the "MP requirement correction" may be increased to a value of 1.4-1.5. However, we caution that this will lead to a significant increase in cost of the diet and risks a high rate of nitrogen loss in urine and manure, thus increasing potential loss of nitrogen into the wider environment.
References
Little S., King R., Lean I., Rabiee A. and Breinhild K. (2009) A review of 11 applied dairy nutrition models used in Australia. Summary Report. Grains2Milk project, Dairy Australia.
Thomas C. (2004) Feed into milk: a new applied feeding system for dairy cows. Nottingham University Press, UK.