This article is part of our series of original articles on emerging featured topics. Please check here to see other articles in this series.
Most software used to formulate diets requires inputs for a specific dairy cow. For example, a diet might be formulated for a third-lactation, 1,450-lb cow producing 80 lb/day of milk with 3.8% fat and 3.1% protein, but most cows in the United States are housed and fed in groups. The nutrient specification goals, such as the concentrations of metabolizable protein, used in diet formulation for groups of cows can have a substantial impact on income over feed costs, and they should be considered carefully.
We have reasonably accurate estimates for the requirements of many nutrients; however, in practice, diets are usually overformulated. This is because if you feed the average cow in a group, approximately half the cows will be underfed and half the cows will be overfed. Milk production would be expected to decrease for the cows that are underfed, but milk yield would not change in those cows that are fed nutrients in excess of requirements.
A diet for a group of cows must be formulated for greater than the group average milk production; in other words, safety factors (i.e., the amount of overformulation) are essential to obtain maximal herd productivity. The question is not whether diets should be overformulated, but rather how much we should overformulate. The answer to that question is not a constant. It depends on milk price, feed costs, ingredients used, pen grouping system, and farm/feed management.
Some general relationships regarding safety factors include:
The relationship between the magnitude of the safety factors and milk and feed prices is basically risk versus reward. The almost certain risk associated with overformulation is higher feed costs, whereas the potential reward of overformulation is increased milk yields. When milk is expensive and feed is cheap, the potential reward of more milk is usually worth the risk of higher feed costs caused by greater overformulation. When feed is expensive and milk price is low, a larger reward (i.e., greater marginal response in milk yield) is needed to cover the cost of overformulation. Even with high feed costs, diets still should be formulated to exceed the nutrient needs of the average cow in the group; the question then becomes by how much over the average cow they should be formulated.
The more variable the nutrient composition of ingredients, the less confidence you have that the nutrient composition of a ration actually matches the formulated diet. On average, half the time, the diet will exceed your formulation goals, and half the time it will be below your goals. To reduce the likelihood that changes in the nutrient composition of a feed will cause a nutrient deficiency and reduce milk yield, safety factors should be increased when a diet is based on inconsistent feeds. Variation in nutrient composition can also be managed by limiting the inclusion rates of variable ingredients and using multiple ingredients to provide major nutrients.
Good feeding management increases the confidence that the ration delivered to the cows is the ration that was formulated and therefore diets can be formulated to more closely match actual requirements. On the other hand, poor feeding management or poor general management may or may not justify increased overformulation. With poor feeding management, you are less confident that the formulated diet is the diet delivered to the cows, which could indicate the need for greater overformulation. Poor general management means that factors other than nutrient composition may be limiting production, and overformulation will simply increase costs and not affect milk yield.
Another factor that needs to be considered is cow health. Overformulating diets with respect to energy can increase the risk of acidosis. Farms with poor management can be at high risk for acidosis, and feeding a diet with excess energy may lead to substantial health problems.
Pens that contain a diverse population of cows must be fed diets that substantially exceed the nutrient requirements of the average cow in the pen to allow the high-producing cows to maintain high production.
Cows factors that affect nutrient requirements and/or feed intake (intake will determine nutrient concentrations needed to meet requirements) include body weight, parity, stage of lactation, milk production, and milk composition. Of those factors, usually only milk yield and days in milk vary enough within a pen to cause substantial variation in nutrient needs among cows. At similar milk yields, cows in the first month of lactation can consume 10 to 15 lb/day less dry matter than cows in later lactation. This means very early lactation cows need to be fed diets with much greater nutrient densities than later lactation cows, and pens that include both very early lactation cows (days in milk less than 30) and later lactation cows need diets with greater safety factors. Keeping very early lactation cows in a separate pen from later lactation cows can reduce feed costs by reducing the amount of overformulation needed for the majority of cows (i.e., cows >30 days in milk) in the herd.
The degree of overformulation needed to account for variation in milk yield obviously depends on how variable milk yields are within a pen. If a pen does not contain early lactation cows, the normal relationship between milk yield and dry matter intake means that a diet balanced for the pen average milk yield will usually support 10% to 20% more milk than the pen average because higher-producing cows have higher intakes and consume more nutrients. However, if cows are in the pen that produce more than 1.2 times the average milk yield, a diet formulated for the average cow will not meet her requirements.
The most accurate method of determining the safety factors needed for a group of cows with variable milk yields is to calculate the standard deviation for milk yield within the pen and then formulate for approximately 1.3 standard deviation units greater than the mean milk yield. For example, if the average milk yield for a pen of cows is 80 lb/day with a cow-to- cow standard deviation of 6 lb, the diet should be formulated to support 80 + (6 x 1.3) = 88 lb. The dry matter intake used for formulation should either be estimated using pen average milk yields or actual pen average dry matter intake (do not use the formulation milk goal to estimate intake).
If you can obtain individual cow milk yields, the standard deviation is easily calculated using any computer spreadsheet program. If you cannot get individual cow milk yields, then formulating for 1.15 times the pen average milk yield should meet the requirements for about 95% of the cows in a pen, assuming milk yields follow a normal distribution.
When formulating diets, balance for the higher milk yield but use actual pen dry matter intakes or intakes estimated using actual pen average milk yields. If the pen includes very early lactation cows, that degree of overformulation will not be adequate to meet their needs because of low intakes. A safety factor of 1.2 to 1.25 times average milk yield will probably be needed for cows fresh less than 30 days.
Diets for groups of cows must be formulated for greater than the pen average milk yield. The degree of overformulation depends on several factors, but one of the most important factors is whether cows less than about one month in lactation (fresh cows) are in the pen. Because of low intakes by fresh cows, pens that include both fresh (<30 days) and later lactation cows must be fed diets with substantially higher concentrations of many nutrients than required by the average cow in the pen. Grouping cows less than about 30 days in milk separately should reduce feed costs by reducing the amount of overformulation necessary to maintain high production. If pens do not contain cows less than 30 days in milk, variation in milk yields within a pen (standard deviation) should be used to determine the safety factors. If individual cow milk yields are not known, diets should be formulated for about 1.15 times the average milk yield (assuming all cows are greater than 30 days in milk).
The Ohio State University