The sessions created an opportunity to showcase examples of successful optimization strategies, limitations for broader adoption of this approach, and opportunities for the future. In summary, the attendees were invited to share experiences and successes with formulating diets using various software platforms with linear and nonlinear optimization. Why all the fuss? The interest in optimization is to create better and lower cost rations that support equal milk production and cow well-being when compared to nonoptimized rations.
The keynote speaker featured Mark Hanigan from Virginia Tech University. With Hanigan’s recent leadership for the Protein and Amino Acids portion of the National Academies of Sciences, Engineering, and Medicine’s (NASEM) 2021 Dairy Nutrient Requirements of Dairy Cattle and longtime interest in ration optimization, his overview of the topic was a solid opening for the meeting. The convergence of the math and logic of biological modelling with the nutrient supply and animal requirements of animal nutrition were addressed in Hanigan’s session.
Putting this knowledge and interest to work in research was the topic of the second presentation. Having recently earned a Ph.D. from Virginia Tech, Leticia Campos, a postdoctoral research fellow at Colorado State University, presented recent findings from her doctoral project in Hannigan’s lab. In the soon-to-be-published trial, Campos described the project where cows in robot feeders were fed optimized diets to compare both production and economic results to a control diet. In this study, the use of the optimized diets had higher production and higher income over feed cost than the control diet fed on the same farm.
The majority of dairy ration formulation is based on the biology known as the Cornell Net Carbohydrate and Protein System (CNCPS), so the basics of the pools in CNCPS were presented to the group. Also emphasized was how a full understanding of this approach to formulation is essential before attempting to use current models to optimize rations. With a focus on nutrition formulation, including rates of passage, rates of digestion, and their synergy in supplying nutrients for milk production, growth, and maintenance, Jim Aldrich of CSA Nutrition made sure the basics were in place to let the model help optimize better and more profitable rations for dairy animals. Central to this session was the strong reminder to create a rumen environment that will maximize both carbohydrate fermentation and the production of microbial protein. Adding rumen undegraded protein to this strong microbial protein supply will result in maximum metabolizable protein supply for high-producing milking cows. Nutrition models based on the CNCPS complete this math for us to aid in building strong rations.
Following this, my team and I demonstrated two different commercially available programs to optimize diets for both growing heifers and lactating cows. Using more of a pragmatic field approach, we demonstrated some tricks of the trade to help the model use the optimizer to save feed cost. There was also a discussion regarding shadow prices and using optimization runs to evaluate ingredient options that can be used in ingredient purchase decisions. With many years of routine optimization use, we brought a real-world angle to the conversation and discussed occasional software frustrations when being committed to optimization as the default approach to routine ration formulation.
Another participant described different commercially available formulation platforms commonly used in feed companies. It is true that some parts of diets are built in more of a feed mill formulation environment as opposed to the field-focused platforms discussed to this point. In some cases, complete grain mixes and protein/fat concentrates, and in most cases, mineral-type products, are formulated in a feed mill-focused formulation platform that also has optimization capabilities. Examples were shown in both liquid and dry ingredient forms. Many of these mixes are supplied by the consulting nutritionist with ingredients and nutrients set. In other cases, only nutrient specifications are indicated, allowing the feed company to perhaps optimize these formulations to best manage cost.
Examples of optimizing fresh cow rations fed with an optimized partial mixed ration (PMR) and a robot supplement from outside the U.S. brought the expertise to an even higher level. Dialing in numerous nutrients to support high levels of milk production in Canada was demonstrated, with supporting dashboard production data indicating success. Some similarities and differences between competing ration formulation programs that both use CNCPS biology were explored. Even though the equations are the same behind the scenes, each software product had its own approach to various tasks.
An ingredient supplier in the industry presented a strong case, emphasizing their products are correctly described in the model’s ingredient library. This is necessary to allow it to compete correctly for space in a particular formulation. The prices and values for various ingredient options will go up and down with markets, but the correct nutrient detail in the model for a feed ingredient is critical to allow the optimizer to select the best feeds for each situation.
Another focus at the meeting was economic opportunities to reduce feed cost while supporting the same milk production. It was noted that since optimization savings go straight to the bottom line, considering a 10- or 20-cents per cow savings as a percentage of margin per cow could have a significant impact on farm profitability.
Follow up on the meeting has focused on the next steps to carry this message of optimization to other nutritionists. It is the feeling of the group, and most industry professionals surveyed, that the routine use of optimization in building diets has a big upside. Many nutritionists could be encouraged to take another look at the approach. With a savings potential of up to 20 cents per cow per day, this is a tool that should not be left on the shelf.
However, there are many challenges with boosting routine use. There is a poor comfort level with many nutritionists regarding this technique. In addition to losing a bit of control on ingredient feed rates, many nutritionists are frustrated by the quirky nature of the optimization steps and may abandon the approach and simply move things around in the ration until the nutrient supply meets the needs and goals of the animals being fed.
There is a significant unrealized opportunity here that can have a strong impact on dairy farm profitability. This isn’t your grandpa’s “least-cost calculator.” It is a dynamic nutrition model with both linear and nonlinear optimization modules that, along with an experienced nutrition technician, can simply solve for better rations than can be accomplished without it. Yes, it doesn’t always give you good answers, and at times, it may not have an answer at all. However, with certain tips and tricks, a patient nutritionist can adjust things to let the model give a better and cheaper solution. The math and science here are strong and proven. Perhaps this new effort to bring attention to these often-unused tools in our ration programs can help both nutritionists and dairy producers find success.
Stay tuned for more!
