Computer Models Can’t Replace Observation and Experience
Oct 10, 2011
Although modeling programs certainly have helped us assimilate research, the real work is still up to the nutritionist, the dairy producer, dairy manager and feeders.
By Rick Lundquist, Ph.D.
Nutrition computer models are becoming as large and complicated as our tax code. As we accumulate more data and knowledge about the cow, scientists can incorporate it into formulation models to help predict milk production.
But Dr. Mike Allen of Michigan State University cautions that these models still have many limitations, due to the unpredictability of the ruminant digestive system. Allen presented a paper on this subject at the 2011 Tri- State Dairy Nutrition Conference. As a practicing nutritionist, I was encouraged to hear that the push of a laptop button won’t replace practical knowledge, observation and experience in the field anytime soon.
Modeling programs certainly have helped us assimilate research to understand what’s happening in the digestive system of a cow. But Allen states that these models should not be used for routine diet formulation. They are another tool to check our work.
Models do a pretty good job of predicting protein requirements and have helped us target and reduce the amount of supplemental protein needed in the diet. But one of the biggest limitations is predicting energy intake and partitioning, according to Allen. This can be extremely dynamic and affected by the type of ration, the physiological state of the cow, concentration and digestion characteristics of the NDF, starch fermentability and type of fat in the diet.
So, the real work is still up to the nutritionist, the dairy producer, dairy manager and feeders. “Boots on the ground,” as they say.
Minimizing variability in rations is key to maximizing production
, and will also help computer models be better predictors. Rumen pH affects NDF digestibility and consistent rations minimize pH fluctuations. Routine dry matter analyses as well as protein, NDF and starch are needed to minimize fluctuations in nutrient composition of the ration consumed by the cow. I refer you to my May 2011 Agweb nutrition column
about how long it takes to re-establish rumen microbial populations after a ration change.
As Allen states, “Production response has little to do with the program used for diet formulation.” It does, however, depend on high-quality feeds and forages, good silage management, proper mixing and feeding procedures, clean water, proper grouping of cows, facilities to reduce stress, etc. The cow is still the ultimate model. Evaluating cow responses to diet and management changes will provide feedback to optimize diets.
Reference: Mind Over Models. Michael S. Allen. 2011 Tri-State Dairy Nutrition Conference Proceedings.