Maximising enzyme response in poultry diets

 

 

One of the major challenges in feed enzyme research is that in vivo data,
be it performance or digestibility, is often equivocal. There are literally
thousands of peer reviewed publications reporting a vast range of responses to
feed enzymes from negative to highly positive (Rosen, 2002). A major
reason for the variable responses that have been reported is that the response
to exogenous enzymes is dependent on a large number of interacting factors such
as bird age, health status, cereal type, cereal quality, nutrient balance, feed
processing conditions and the environmental conditions under which the birds are
reared (Bedford, 2002). With this background variance in animal
performance, demonstrating a response to a feed enzyme is difficult unless the
sources of variance can be minimised or the magnitude of the response is so
great that the background ‘noise’ ceases to become a mitigating factor. In the
case of xylanases and cellulases for the so-called ‘viscous grains’ and also for
phytase the latter is true, with mean responses of a magnitude that improves
consistency merely due to the very poor quality control diets that are fed.
However, with non-phytase enzymes for diets that are based on the non-viscous
cereals such as corn and sorghum the mean responses are generally of a lower
magnitude and thus are at risk of being lost in the animal production ‘noise’
(Cowieson et al., 2006a). In a commercial production
system it is impossible to reproduce the controlled conditions within a
University research facility and thus demonstrating a significant response to an
enzyme cannot be achieved by minimising variance. It is precisely for these
situations that empirical and mechanistic models can be invaluable as they seek
to predict the response to a feed enzyme based on key dietary, environmental and
animal-related factors.

 

Predicting the response
Although the results of one
experiment can be useful to market an enzyme product or to shed light on a
particular mechanism they are not particularly informative regarding the
mitigating factors in the measured response and may be an over or under-estimate
of the efficacy of that product under a range of commercial systems. Variance
associated with bird age and species, gender, feed type, ingredient quality,
environmental conditions
and bird health status can be relatively easily
controlled in a randomised block experiment under research conditions and may
deliver a result that is not subsequently reflected in economic benefits to the
end user. Thus, marketing of enzyme products based on a handful of strategically
selected data sets that tell a convenient story do not greatly assist end users
in predicting the response under their specific production systems (Bedford,
2002; Cowieson, 2005; Cowieson I., 2006a). It is to this end that empirical
and mechanistic models can be invaluable in that they can highlight influential
factors in the response to a feed additive that may not be apparent to a
reviewer of one or more scientific publications. Rosen (2002; 2006) has adopted an
empirical approach to animal modelling and has successfully
emphasised the critical factors in enzyme response against a background of
considerable performance variance. This empirical approach (termed
‘holo-analysis’) to the review of published feed additive effects is not only
scientifically enlightening but can also be used to create nutritional tools to
assist in the marketing of enzyme products. Tools such as AvicheckTM Corn and
PhycheckTM (Danisco Animal Nutrition) are based on large databases of in vivo
data and have been developed to give predictions to end users of Avizyme 1502
and Phyzyme XP based on key dietary and animal factors (Cowieson,
2005). The arbitrary addition of enzymes to poultry diets without the use
of such tools to maximise economic benefits is crude and recommendations based
on a handful of trial results cannot be expected to reflect response in a more
complex production system. An example of this is that the magnitude of the
response to phytase is determined to a large degree by the concentration of
dietary phytate. This is intuitive as phytate is the primary substrate for
phytase in poultry diets, especially for the bacterial 6-phytases which have a
particularly high affinity for the fully phosphorylated myoinositol
hexakisphosphate ester (Wyss et al., 1999). Although a handful of
scientific publications on the scale of response to phytase in diets differing
in phytate content may allow certain assumptions to be made, a full
meta-analysis of the data is required across a large number of data points in
order to establish the relationship between phytase bioefficacy and dietary
phytate concentration. These kind of statistical approaches to large
animal-derived datasets are extremely useful when generating matrix values for
enzyme products as differences in response can be credited to a variety of
important dietary, environmental or husbandry criteria and can subsequently
allow the end user to ascribe robust matrix values to the product based on their
own dietary constraints.

 

Enzyme combinations
Whilst the use of single enzyme
products such as xylanase or phytase can be optimised using predictive models
the situation is further complicated when both products are used simultaneously.
In this situation, assumed additivity may lead to an under or overestimation of
the scale of the response depending on overlap or synergy in the mode of action,
and so lead to control diets that have been inappropriately formulated. As the
global market for phytase becomes increasingly competitive and environmental
legislation enforces its use, phytase addition to poultry diets by many
nutritionists is becoming the norm. Thus the efficacy of non-phytase enzymes
must be demonstrated in a diet that already contains phytase as a background
activity, with proof of additivity of matrix values for all enzyme products in
the diet. Recent work has attempted to elucidate the additivity of matrix values
for phytase and carbohydrases in corn/
soy-based diets for broilers
(Cowieson & Adeola, 2005; Cowieson et al. 2006b,c). In this work
phytase and a carbohydrase/ protease enzyme cocktail (AvizymeTM 1502,
Danisco Animal Nutrition)
were added to a corn/soy diet that had been strategically formulated to allow
for the predicted improvements in ME and
the retention of Ca, P and amino acids. This removal of nutrients was based on
predictive models (PhycheckTM and AvicheckTM Corn) that have been
developed based on a large number of broiler performance and digestibility
trials. The conclusion from this work was that combinations of enzymes are
extremely effective in order to achieve a particular nutritional goal and both
the scale and consistency of the response is improved through this strategy
(Table 1). However, as the magnitude of response increases so does the
opportunity to produce nutrient imbalance within the diet and so empirical tools
are critical to allow for the strategic removal of oil, amino acids and
inorganic phosphate sources, creating a control diet that maximises return on
investment.

 

Conclusions
The observed response
to exogenous enzymes is dependent on a large number of complex interactions
between the animal, the diet and the environment. It is therefore inadequate to
arbitrarily add an enzyme product to a diet and expect to achieve the same
result that has been demonstrated under controlled research conditions. Holistic
least square models are useful in this regard to shed light on key mitigating
factors in the response to allow the end user to estimate the economic value
under specific environmental, dietary or husbandry conditions. Exogenous enzymes
are not a panacea but they are an extremely potent biotechnological tool when
wielded with the necessary knowledge to maximise return on investment.

The response to exogenous enzymes depends on the animal, environment and the diet.

Mechanistic and
empirical models equip nutritionists with the knowledge they require in order to
formulate a diet strategically to allow for the expected improvements in
nutrient retention following addition of a supplemental enzyme. It is likely
that the acceptance by the industry of novel enzyme technologies in the future
will be determined to a significant degree by the ability of the feed additive
company to demonstrate a holistic knowledge of the product to allow response to
be predicted under a range of dietary conditions, precluding the arbitrary
addition of enzymes based on minimum cost in use.

 

References are available on request.