Lessons from Predicting DDGS Quality, Part 2: Prediction Equation
In my previous blog, I discussed data from a recent publication on variation in the quality of DDGS (Distillers Dried Grains with Solubles), a by-product of the corn ethanol industry. This is important because DDGS is used to replace other ingredients in livestock diets – namely corn, soybean meal and phosphorus sources – and is so highly-variable in quality that the authors devised a prediction equation to determine the value of each batch of DDGS.
First off, it says something when a complex prediction equation must be devised and revised (see discussion in the article on the various equations) by several researchers for one ingredient. It certainly indicates a lack of standardization of procedures in the ethanol plants, which consider DDGS to be a by-product.
In order to develop the most-accurate prediction equation, the authors looked at all of the nutrient data (some of which I outlined in Part 1, and determined which sets of data, when combined, could predict the energy available to the chick for growth. Simply put, they answered questions like the following: If I included data on crude protein, ether extract (oil), and starch contents from the DDGS samples, how much of the variation in energy could this explain? So, from a practical point-of-view, if someone bought a rail car of DDGS and knew crude protein, ether extract (oil), and starch, could they use the equation and have a good idea of how this load would support chick growth?
With trial and error, the authors concluded that an equation with gross energy, crude protein, fiber (NDF), and starch would best predict the available energy to support performance in the chick. Gross energy (total energy in each DDGS sample) and starch (a readily-digested energy source) would seem to be useful predictors of the available energy to the chick. Fiber, or NDF, is not easily digested by chicks, so this would also seem to make sense as an important part of a prediction equation. Crude protein is a major component of DDGS, and exhibits variation, so this is also important for predictions.
What is very surprising and telling is that ether extract (oil) was determined to NOT be an important predictor of energy availability in the chick – even though ether extract values ranged from 3.2 to 13.2% in the DDGS samples, and that oil is a concentrated source of energy.
I determined the contribution of DDGS to the total amount of fat/oil in the experimental diets, assuming that corn had 4.8% oil, solvent-extracted soybean meal had 1% and poultry fat was 100%. Knowing that DDGS samples were included at 15% of the diet, and also knowing the range of residual oil contents in the DDGS samples, the following contributions of DDGS to total dietary fats/oil were calculated:
Note that, at the extreme end, DDGS contributed nearly 45% of energy-rich oil to the total fats/oils in the complete diet, but this was NOT found to be important when predicting the energy available to the growing chick.
We can conclude from this that the residual oil in DDGS from corn ethanol plants is of low value, and is essentially unimportant when fed to chicks. The authors picked the 15% DDGS inclusion rate because that is typical in the broiler industry.
Perhaps this should not be surprising because high-quality energy sources, such as extruded, full-fat soy, contain highly-digestible energy to support performance – even more so when compared with two types of DDGS. Please feel free to contact myself, or Dr. Nabil Said, to discuss the benefits of using high-quality ingredients, such as extruded, full-fat soy and extruded, partially defatted (ExPress®) soy meal, as compared with the drawbacks of using lower-quality ingredients, such as DDGS.
So, to conclude, what the recent work on DDGS prediction equations shows us is that great care must be used when selecting ingredients to support animal performance – especially those with significant nutrient variation, and low-quality residual oil.