Back about 25 years ago when the first personal computers were becoming affordable, I discovered the power and utility of the electronic spreadsheet. With the aid of feed analysis data that I found in dairy science textbooks – moisture, crude protein, TDN – I was able to construct tables that would calculate rudimentary nutritive values for dairy rations. I could substitute one feedstuff for another and have an instant answer.

I could conduct price sensitivities and know in a moment if one ingredient was a better value than another. I could create tables for the high-producing strings and the low-producing strings and save them on a 5 ¼-inch floppy disk. It wasn’t long before I was printing batch sheets that could be hung in the cab of the loader. Talk about making good use of new technology. It was great!

Well, as it turned out, I wasn’t the only genius who had figured out that dairy ration formulation could be easily computerized. Many others in both academia and private industry were developing models for formulation and evaluation of dairy rations that were significantly more sophisticated than any of my simple spreadsheets. No one will argue that the computerization of dairy nutrition has certainly been one of the primary contributing factors to the increase of milk production and feeding efficiencies for the dairy industry.

Today the sophisticated computer ration model relies on a host of different criteria when balancing a dairy cow’s diet. To accurately predict milk production for any given cow, we need to know many more things than just the type and quality of feed she’s eating. We must consider the cow’s size, the environment in which she’s living (including temperature and humidity), the type of housing she’s in and the hours that she’s standing or resting. It’s not just about milk production targets, it’s also about butterfat and milk protein levels, as well.

In order to maximize milk production and feeding efficiencies, the central focus of dairy nutrition is on the cow’s rumen. But because the function and physiology of the rumen is so dynamic, constantly changing and influenced by the feedstuffs that a cow consumes, attempting to describe how a rumen functions by way of mathematical equations has been a challenge. The dairy industry can consider itself fortunate to have many mathematicians, engineers, technicians, researchers and scientists who have dedicated their careers to developing computer models that accurately describe how a rumen will function, and therefore how much milk a cow will give.


For many years computers and their operating systems were not fast enough to satisfactorily process the gargantuan amounts of data that must be analyzed to reach a meaningful solution for a dairy ration. Today processor speeds, memory capacity and data storage are more than adequate to handle ration modeling software.

Thanks to the ever-growing computer industry, ration programs are now able to analyze a host of nutrient fractions including amino acids, fatty acids and various carbohydrate fractions in any analyzed feedstuff. Protein degradability and solubility, along with bypass fractions, are used to predict milk production. Feed analysis laboratories have been instrumental in archiving thousands of feed analyses, which have been incorporated into feedbanks in the ration programs. And the programs keep getting better, more sophisticated and more user-friendly. The best programs today are doing better jobs of predicting rumen microbial growth and fermentation efficiencies. As research into amino acid and fatty acid metabolism continues to move ahead, so do the models’ abilities to predict milk production based on that research.

The power of the computer ration balancer today is exhibited through the least-cost optimization capabilities that are incorporated into the programs. If a computer can do nothing else, it can manipulate numbers by the millions in just a few seconds. Nutritional models take a host of feed ingredients, each one with dozens of nutritional attributes that must be considered, analyze them based upon a cow’s size, environment and desired milk production and milk components. They calculate the “optimum” or most economical solution that meets all the user-specified requirements at the lowest price based on the information supplied. Properly executed, ration balancing software can quickly determine if that “good deal” on a grain mix is really that good of a deal. There is now enough data on every aspect of a feedstuff – how it ferments in the rumen, what the amino acid content is, what the fatty acid profile is – to allow nutritionists and dairy farmers to make quick and accurate decisions as to whether a feedstuff should be used in a ration. Whether a nutritionist optimizes several rations on a single dairy or uses the model to analyze an existing ration from the competition, computerized ration modeling is a powerful management tool on any dairy farm today, whether it be big or small.

Today the dairy industry finds itself in unprecedented economic circumstances requiring dairy farmers to feed cows and heifers as efficiently as possible. Efficiently doesn’t necessarily mean cheaply. There’s no longer any excuse to overfeed or underfeed protein in a diet. Dairy producers have the means available to quickly determine if a ration is too heavy on starch and low on effective fiber. They can quickly make changes based upon prices while still meeting nutritional parameters. There’s no longer any need to be playing guessing games when it comes to selecting feeds for a dairy diet and finding out too late that a change destroyed milk revenue for the next month. Computer ration modeling allows any dairy farm to analyze rations thoroughly and in a very timely manner. PD

John Hibma