After 36 years of building dairy diets, I am looking for places where I might become the old guy with old ideas. I must avoid this risk. I was asked in a recent podcast interview what I thought was a key characteristic that separates successful dairy professionals from those who never achieve a high level of success. What sets these people apart? My answer was twofold. First, you must have a passion for dairy cows and our industry. Secondly, and connected in a way, you must strive to keep learning.
I remember a specific instance in graduate school at Texas A&M when it occurred to me that I really knew nothing about practical dairy production. With a strong family background in beef production, my only experience with anything dairy related was FFA dairy judging. Our state-winning livestock judging team, unable to compete again the subsequent year in that event, made the quick switch to dairy judging. How hard could it be? Since we were good at talking reasons, we won the state dairy contest the following year. Remember, I am from Alabama, not Wisconsin. Other than that, I knew nothing about practical dairy farming.
The revelation
About two years ago, I made a change in the nutrition modeling software I use. It can’t be underestimated how significant of a disruption this change is, even if making the change is the right choice. It has been a good change and one of the significant differences I made in the process was allowing the software to import the forage analysis. My previous software also had this capability, but I never trusted it. Instead, I input the nutrients in a manual fashion. I decided this was a situation where I was acting like the “old guy” and it was time to get with the program and trust the import process. I was ready to learn again and to feed cows better.
So, why was I so nervous about this auto-import? I wasn’t scared of errors or glitches in the process, but going through the manual process of typing a new nutrient value on top of the current one gave me some feel for what was changing and what I thought about it.
Some of these key nutrients including protein, fiber, and starch values carry a lot of weight in how the diet should or shouldn’t be adjusted. If the auto-import quietly changed the crude protein on an alfalfa haylage from 19 to 24, how do I feel about that? Or if a corn silage being fed upward of 25 pounds of dry matter changes from 39% to 42% starch, how will I respond to that change that has been made? These questions of art and experience must be considered extremely carefully.
Good and bad
If I had some type of “automatic reformulation” process or AI was handling the adjustments, the diets would be tweaked to accommodate new values, and ration targets would be maintained through an adjustment. This may or may not be good. What if the haylage protein boost coincided with a notable drop in milk urea nitrogen (MUN)? Or if the starch increase correlated to a jump in butterfat and more manure stacking than in recent weeks? If the “auto-formulate” function did its math perfectly, both of these negative trends wouldn’t be carefully fixed but perhaps made even worse. I am sure rules and algorithms can be put in place to consider much of this just as a skilled nutritionist will do. But in either case, the fact that the protein in the haylage could be quietly changed, along with the starch in the silage with the import process gives me pause.
Finding a happy medium
I decided to stay with the auto-import of the forage analysis since the model needs to keep so many details updated. Just like I need to be sure to update any changes on the dynamics of the cow I am formulating for, the model needs many details on the forages; more than just a few key nutrients like protein neutral detergent fiber (NDF) and starch. The automation has created a convenient process, but I must still remain very aware of where these forage nutrients are trending and how those trends feel compared to the contemporaneous performance data. One must have the other to do this correctly.
Sure, if you are just building an academic ration as an example, or in a very infrequent situation where you are building the very first diet for a group where you don’t have any information on how the current diet is being used by the cow, just make sure the diet is adjusted to meet the target nutrients. In a real world example, you must see how these real forage results are being used in a ration that has performance results to evaluate the current situation and then make any needed changes for the next iteration of this ration. No one ration is an entity unto itself.
Instead, every ration is just one point in a continuum evaluated by current results. The influence of the previous ration depends on how long it has been in place, and all of this informs the important process of building the next ration. The perfect ration report might have three rations on the page with the corresponding nutrients and performance results included. Seeing the current ration in the bunk flanked on the page by the previous and the next ration captures the essence of this principle. When the current ration is updated with actual nutrients based on real-time ingredient analysis results, it makes this even better.
To respect all these principles, let the software auto-import the analysis results. Add a process by which the formulator can see key nutrient changes that may impact potential tweaks in formulation. Print the PDF and mark it up, use a spreadsheet that can show trends of key nutrients, or better yet, see if the forage lab has some online tools to help. You can avoid getting lost in the minutia by concurrent consideration of some type of weekly report including key metrics such as intake, milk flow, components, and manure observations.
To make the best “next ration,” it is critical to understand how the current ration is performing compared to the goals set when it was built. Ration analysis can be helpful in this process, but I always wonder if this data point could misinform us due to many variables in the process. More samples to the lab are better, but there is certainly a point of diminishing returns on that cost. Combining these pieces result in the most successful rations. Doing it right is not for the faint of heart.
