Introduction of the CVB 2015 net energy model for swine in NutriOpt

The Centraal Veevoederbureau (CVB) has released a new version (NE 2015, CVB) of the model used to estimate the Net Energy (NE) of feed ingredients, which will replace the current version (NE 2007, CVB) available in NutriOpt. The guidelines issued by CVB are the standard for feed ingredient evaluation in NutriOpt. The two versions of the NE model (CVB, 2007; CVB, 2015) use slightly different equations. As a result, the predicted NE value of feed ingredients is expected to change slightly after the update.  Therefore, the objective of this communication is to: 1) Inform of changes in nutrients associated with NE; 2) offer an insight of the new features of the NE model to be implemented 3) briefly discuss the practical implications of introducing a new version of the NE model.

After implementation is finalized at the end of Q4 2021, a new NE 2015 nutrient (MJ and kcal) will be introduced, and the NE 2007 nutrient will be phased out from NutriOpt (Table 1).

Table 1. List of NE nutrients and their status in NutriOpt after completion of NE 2015 implementation

In the old NE 2007 and new NE 2015 models by CVB, equations (1) and (2) respectively, the concentrations of the following nutrients are used for prediction of NE in “dry” feed ingredients: 1) crude protein (dCP); 2) crude fat (dCF); galactooligosaccharides (GOS); starch (STA), free sugars (SUG), and non-starch polysaccharides (dNSP).

(1)

NE_2007(MJ) = 10.8 * dCP + 36.1 * dCFAT + 13.7 * (GOS + STA) + 12.4 *  SUG + 9.6 * (dNSP)

(2)

NE_{2015 (MJ)} = 11.70 * dCP + 35.74 * dFAT + 14.14 * (GOS + STA )+12,73 * SUG+ 9.74 * (dNSP)

To account for energy contribution from products of fermentation in liquid ingredients, the NE 2015 also incorporates the following:

(3)

Eq. 2 + 10.61 * Ac. Acid + 14.62 * Prop. Acid + 19.52 * But. Acid + 20.75 * Ethanol + 12.02 * Lac. Acid + 13.83 * Gly

But, what are the major differences between these two versions?

• Digestible nutrients are calculated by regression

The letter “d”, accompanying some nutrients (Eq. 1 and 2), is the digestibility coefficient (apparent total tract) associated with that nutrient. Nutrients with no “d” are assumed to be 100% digestible. Traditionally, digestibility coefficients (d) come from feed composition tables and change with each version. Therefore, changes in the NE value of ingredients are expected after updating versions in the CDB.

The greatest difference is, however, that the NE 2015 update deviates from using digestibility coefficients, and estimates digestibility of CP, FAT, and NSP with regression equations (not shown). This change is motivated by reports in the scientific literature of negative effects on digestibility of a nutrient by the concentration of another nutrient in the same feed ingredient.  The equation also corrects digestibility estimates for endogenous losses using a factor. This approach is used by default, but new digestibility coefficients are also available if inputs for correct operation of regression equations are “not” present.

• Energy factors are greater for CP, Starch, Sugars and NSP

The factor accompanying each nutrient represents the NE contribution per unit of the nutrient (Eq. 1 and 2). In the NE 2015, the NE factor is greater for CP, STA (and GOS), SUG, and NSP. Therefore, after the update, the NE value of ingredients is expected to increase with the concentration of CP, GOS and STA, SUG, and NSP. In contrast, the NE factor of FAT is lower in the NE 2015, which may also lower the NE value of fat-rich feed ingredients (Table 2).

• The NE 2015 uses the fatty acid composition of fat sources to better predict their NE value

 The NE 2015 accounts for the negative effect of the concentration of free and saturated fatty acids on the digestibility of fat in that ingredient. This will directly affect the NE estimate. The concentrations of short and medium chain fatty acids are added to the unsaturated fraction.

Table 2. Average difference in NE estimates between NE 2015 and NE 2007 by type of feed ingredient

The new features of NE 2015 result in different NE estimates for feed ingredients in CDB. In table 2, the averages of differences per type of ingredient are summarized. The new NE 2015 predicts approximately 3.5% greater NE, on average, than the current value in CDB (NE 2007) for most ingredients. As discussed before, these differences are expected because concentrations of CP, STA (and GOS), SUG, and NSP received a greater NE factor in the equation. The average NE 2015 value for fat sources is -1.1% inferior than of NE 2007, which is expected due to the lower NE factor assigned to FAT in NE 2015 (Eq. 1 and 2).

The average difference between NE 2015 and NE 2007 for 108 wean to finish diets is 93 ± 13 kcal/kg NE (Graph 1). The plot shows that the average kcal/kg difference of diets between NE 2015 and 2007 follow a constant pattern, which does not seem affected by phase and dietary NE concentrations.

Graph 1. NE  values calculated with NE 2015 or NE 2007 for wean to finish diets  (n=108 diets) formulated with different NE concentrations

Graph 2 depicts the difference between the NE 2015 and NE 2007 of wean to finish diets, formulated with different NE density, number of constraints and availability of ingredients. The plot shows that the difference between the two models appear to be slightly affected by energy density of diet, number of constraints, and availability of ingredients.

Graph 2. Difference between NE 2015 and NE 2015 for 108 wean to finish diets formulated with different NE content

Although the difference between dietary NE 2015 and 2007 values is somewhat constant across NE densities and phases, it is critical to acknowledge that there is still a difference in NE value for ingredients and diets.  

Conclusions

1. The introduction of the regression approach, to estimate digestible nutrients, is aimed at improving the accuracy of the NE estimate
2. Digestibility coefficients of nutrients are variable based on that nutrient or on other nutrients, so based on the ingredient quality.
3. There is good correlation between NE 2015 and 2007 estimates for cereals, cereal co-products, fiber and vegetable protein sources
4. Use of fatty acid composition of fat sources better explains the NE value of each source in the NE 2015
5. On average, the difference between NE 2015 and NE2007 is 93 (± 13) kcal for wean to finish diets
6. The difference between NE 2015 and NE 2007 is slightly affected by energy density of diet, number of constraints, and availability of ingredients
7. Correlation equations between NE and Effective Energy (used in the swine model) will be updated to fit the NE 2015
8. Nutrients “NE swine_2015 (MJ)” and “NE swine_2015 (kcal)” will be soon available in the NutriOpt portal
9. Net energy values of complete diets will likely increase after the update. Therefore, an update of recommendations will come at implementation.