The principles of diet formulation are simple and straightforward: we assume we know exactly what the nutrient requirements of a given group of animals are, we know the nutrients each potential ingredient can deliver, we assume all nutrients are additive and linear, we know the cost of the materials, we apply any production constraints and we then combine all that knowledge in our linear programming software so we can get the lowest cost solution to meet the nutrient requirement given the available raw materials. With regards to feed additives, defined as low inclusion products that have an impact on nutrient availability of the diet, there are many opportunities to be confused when applying these in feed formulations. The main reason for this is confusion about matrix values, how best to apply them and how to combine matrix values from different types of additives. This paper is meant to clarify some of these issues and give food for thought to get to the best strategies for applying feed additives in formulations. Enzymes are one of the major categories of additives with matrix values but several other product categories use the same approach and in principle should be considered together.
The problem is that most of the ‘we know’ phrases in the previous paragraph are really more accurately described as ‘we think we know’. And some of the assumptions are incorrect, especially when we look at matrix values for additives: These are unlikely to be linear and nor are they additive. For instance, if we look at an enzyme like phytase the dose response is normally described as logarithmic, which in effect means that a doubling of the dose will give 30% more nutrient release. It is possible to deal with this in the formulation system, either through specific enzyme modules in the software or by having different versions of the same material for different inclusion levels. Similarly, nutrient requirements are normally based on a dose-response curve, with the target value set based on the economic situation as well as the biological response. This means that supplying more nutrients than targeted may well have some positive impact on performance but supplying less nutrients such as the level of substrate in the diet, whether it is possible to than targeted will have a relatively bigger negative impact.
Additivity means, in fact, that there is no interaction between any of the ingredients in the formulation, which is much less likely than them having some sort of interaction. This could of course be both negative (resulting in sub-additivity) or positive (synergy). In the case of most additives it is probably fair to assume some level of sub-additivity, simply because each subsequent additive has less opportunity to improve the availability of nutrients as availability can’t get better than 100%! A simple rule of thumb would be to assume that a combination of two additives that affect the same nutrient would deliver 80% of the combined nutrient release, which is based on typical diet compositions reviewed and responses observed, rather than the 100% expected with full additivity. This approach effectively deals with the expected sub-additivity of additives whilst still opening up substantial opportunity to save diet costs.
A formulation exercise using a broiler grower diet as an example nicely shows the potential benefits of using (and combining) full recommended matrix values of phytase at various dosing levels combined with an NSP enzyme using the 80% concept (Figure 1). It can be seen that whilst the common strategy of just using mineral release values for the phytase gives worthwhile cost reductions from increasing the dose up to 2000 FTU/kg, much greater cost benefits result from taking all nutrients into account using the full matrix. While not taking account of the matrix values offers a way to improve performance, provided the extra nutrients can be utilised by the animal, one caveat would be the fact that we may be unbalancing the diet and thus not extracting the ideal cost-benefit. Reformulating the diet in order to take account of the expected nutrient release is the most cost-effective approach as nutrient balance would be maintained and nutrient requirements will be met exactly, provided they were met in the original diet of course.
Next, we need to consider whether there is room in the diet to apply the provided matrix values. There are several factors to consider, reduce nutrient levels and whether expected nutrient release values are realistic. The level of substrate is probably easiest to take into consideration when we are looking at releasing phosphorus from phytate, as it is easy to understand that you can’t release more phosphorus than the level of phytate-bound phosphorus in the diet. It is normal to assume that no more than 90% of phytate-P can be converted to ‘available P’ (defined as equivalent to P from MCP), even at high phytase doses of a product that is capable of finding and breaking down most of the phytate to the last phosphate group. This means that for a nutrient release of 0.24 % avP (as expected from 2000 FTU/kg Quantum Blue) the level of phytate-P in the diet should be above 0.267 %, which is a realistic value for a typical UK broiler diet. For the non-phosphorus elements of the matrix the situation is different, as these depend not on the amount of phytate present but more on how fast the phytate concentration can be lowered and how low it can get. This means that extra-phosphoric benefits, translated into amino acid and energy matrix values, can be achieved even in lower phytate diets. A prime example are piglet diets which typically are around 0.15 % phytate-P, yet the benefits from using high levels of phytase on performance have been shown repeatedly.
Whether there is room to take account of the expected nutrient release values depends of course on any specific constraints being applied, such as fat minimum for product quality or minimum weighing quantities etc. It is always sensible to check for this sort of thing to avoid formulations ‘filling up’ with materials like limestone, there could be an opportunity to redesign diets to optimise efficient use of the additives and matrix values. In general, the use of appropriate matrix values should result in reduced feed costs and equal animal performance.
For more information visit: www.abvista.com or contact email@example.com
Rob ten Doeschate,Technical Director EMEA, AB Vista