Numerous factors have been identifi ed that infl uence the effi cacy of exogenous
phytases, which is partially refl ected in the inconsistent responses to phytase
that have been reported in the literature. An exhaustive consideration of all
potential factors is simply impractical. To take one example, Leslie et al.
(2006) reported that reducing the lighting programme for broilers from 24 to
12 h increased dephosphorylation of IP6 by phytase. Presumably, this is a
consequence of longer digesta retention times in the crop, which would
facilitate phytase activity. Dietary phytate levels and their sources, the particular
type of phytase added and its inclusion rate are clearly important factors.
Nevertheless, dietary Ca levels, usually provided as limestone, have a
considerable infl uence on phytase effi cacy.
Calcium
The impact of Ca on phytase effi cacy was specifi cally considered in a review by
Selle et al. (2009a). The concept that high dietary Ca levels and/or ‘wide’
Ca:P ratios diminish responses to exogenous phytases is well accepted; the
likely genesis of this concept was a weaner pig study reported by Lei et al.
(1994). The addition of phytase 1200 FTU kg–1 to P-inadequate diets
containing vitamin D 660 IU kg–1 was associated with markedly enhanced
growth performance with 4.0 g Ca kg–1 as compared with 8.0 g Ca kg–1. For
example, from 21 to 30 days of age, weaners on the higher-Ca diets had a
daily gain of 303 g, a daily feed intake of 840 g and a gain:feed ratio of 367.
In contrast, the corresponding fi gures for the lower-Ca diets were 573 g, 1192
g and 480, which represents improvements of 89%, 42% and 31%,
respectively. The authors concluded that higher Ca levels, and wider Ca:P
ratios, depressed exogenous phytase effi cacy, which was attributed to Ca
progressively precipitating phytate in ‘extremely insoluble’ Ca–phytate
complexes in the intestine. However, a superior trial design would have included
non-phytase-supplemented diets to determine the impact of Ca per se in
this context.
However, Driver et al. (2005) subsequently reported confl icting results in
broilers, as 1200 FTU A. niger phytase kg–1 was more effective in maize-soy
diets containing 8.6 g Ca kg–1 than 4.7 g Ca kg–1. Predictably, these authors
concluded that much of the published data concerning the effi cacy of phytase
at different Ca:P ratios was misleading, that phytase effi cacy is a complex
function of dietary Ca, total P and phytate-P concentrations, and that Ca
reactions with inorganic P, which may lead to the fl occulent precipitation of
calcium orthophosphate (Ca3(PO4)2), merit more attention. While the Lei et
al. (1994) study (and similar studies) is open to criticism, the infl uence of Ca
on phytate degradation by phytase in pigs and poultry is an issue that has
been raised.
Thus Ca–phytate complex formation along the gastrointestinal tract,
where one phytate (IP6) molecule binds up to fi ve Ca atoms, assumes importance since approximately one-third of dietary Ca may be bound to phytate in digesta.
Consequently, phytate limits the availability of both P and Ca as a result of
insoluble Ca–phytate complex formation, the extent of which is driven by gut
pH and molar ratios of the two components. It is accepted that Ca–phytate
complexes are mainly formed in the small intestine, where they have a
substantial negative infl uence on the effi cacy of mucosal phytase. However,
exogenous phytases are mainly active in more proximal segments of the gut
and at lower pH levels, so their effi cacy should not be infl uenced by Ca–phytate
complexes in the small intestine. There are, however, data to indicate that Ca
and phytate interactions occur under acidic conditions with the formation of
soluble and insoluble Ca–phytate species, which could negatively impact on
exogenous phytase effi cacy. Also, limestone has a high acid-binding capacity,
which may raise the pH of the gastric phase. For example, McDonald and
Solvyns (1964) increased dietary Ca levels from 9 to 13 g kg–1 with limestone,
which elevated digesta pH from 5.6 to 6.1 in the small intestine of chickens.
Given that pepsin-refractory, protein–phytase complexes are formed in a
narrow pH range of 2.0–3.0 (Vaintraub and Bulmaga, 1991), any limestoneinduced
increase in gut pH could reduce complex formation and mute the
negative impact of phytate on protein digestibility.
Indeed, Pontoppidan et al. (2007) suggested that increasing Ca:phytate
ratios will counteract the precipitation of protein by phytate, and these workers
reported that Ca modestly increased the solubility of phytate and protein
between pH 2.0 and 5.0. Moreover, Prattley et al. (1982) reported that
additional Ca reduced the amount of bovine serum albumin bound to sodium
phytate by approximately 40% over the same pH range. Similarly, Hill and
Tyler (1954a) found that high Ca:phytate molar ratios from limestone addition
substantially increased the solubility of wheat gluten–sodium phytate complexes
and formed insoluble protein–phytase complexes at pH 3.0. Okubo et al.
(1974a,b, 1976) investigated the binding of glycinin by phytate where, at pH
levels below the isoelectric point of glycinin (pH 4.9), Ca decreased the stability
of protein–phytase complexes. These researchers found that suffi cient Ca was
able to dissociate glycinin–phytate complexes at pH 3.0, which was attributed
to Ca directly competing with basic protein residues for the negatively charged
P moieties of phytate. In fact, the capacity of Ca to release protein from binary
complexes at acidic pH has been adopted to prepare phytate-free soy protein
isolates (Okubo et al., 1975).
There is also evidence that Ca may react with soy protein directly, even
under conditions of acidic pH (Kroll, 1984; Gifford and Clydesdale, 1990).
The suggestion is that high dietary concentrations of Ca (relative to phytate
and protein) may reduce the extent of protein–phytase complex formation in
the stomach by reacting with phytate and/or protein at acidic pH. As a result,
increasing Ca concentrations may have the capacity to diminish binary protein–
phytase complex formation.
In this context, the study by Ravindran et al. (2000) is relevant, where A.
niger phytase 800 FTU kg–1 increased the AID of eight amino acids in broiler
diets based on wheat–sorghum blends by an average of 3.75%, with responses
ranging from 1.06 to 7.45%. However, this assay embraced a range of dietary concentrations of Ca (8.7–13.9 g kg–1), phytate (12.06–22.34 g kg–1) and
protein (213–221 g kg–1). It may be deduced from the analysed values that
there were signifi cant, negative correlations between phytase-induced
percentage increases in amino acid digestibility and both Ca:phytate ratios and
Ca:protein ratios. Taken together, the multiple linear regression equation (r =
0.74; P <0.001) is as follows:
Mean percentage phytase response = 15.0 – (10.1 × Ca:phytate) – (78.6 × Ca:protein).
This equation predicts that, as dietary Ca levels increase relative to phytate
and protein contents, amino acid digestibility responses to phytase diminish.
Interestingly, Agbede et al. (2009) subsequently determined the effects of P.
lycii phytase on amino acid digestibility in caecectomized layers on maizebased
diets with adequate (44.9 g kg–1) and low (38.5 g kg–1) Ca levels. At
adequate Ca levels, phytase increased the average digestibility coeffi cient of 13
amino acids by 0.70% (0.862 versus 0.856). However, at low levels of Ca, the
phytase response was a more robust increase of 2.25% (0.864 versus 0.845).
For example, phytase increased the digestibility of threonine by 2.7% in
adequate-Ca diets, but by 4.1% in low-Ca diets. The outcome prompted the
authors to conclude that interactions between dietary Ca and phytase may be
responsible for the variations reported in phytase amino acid digestibility
assays.
Phosphorus
Microbial phytase increases dietary non-phytate P levels, and it is to be expected
that the addition of phytase to diets that are inadequate in this respect drives
growth performance responses. Alternatively, the supplementation of diets
that are adequate or even contain a surplus of non-phytate P may generate
different outcomes, although there is the argument that high levels of inorganic
P will have a negative infl uence on phytase effi cacy (Lei and Stahl, 2000). For
example, Atteh and Leeson (1983) investigated the effects of increasing
available P levels in maize–soy broiler diets from 7 to 10 g kg–1. This increase
in available P signifi cantly depressed weight gain by 11.3% (468 versus 528 g
per bird), feed effi ciency by 3.4% (1.51 versus 1.48) and tended to increase leg
deformities in chicks to 21 days of age. Clearly, the implication is that the
addition of phytase to diets already containing relatively high non-phytate P
levels could generate a counterproductive P excess. This emphasizes the
importance of applying appropriate phytase matrix values to supplemented
diets with identifi ed non-phytate and phytate-P concentrations.
Feed processing
In recent years increasing attention is being paid to the effects of feed processing
on pig and poultry performance, with emphasis on grain particle size and temperatures at which diets are steam-pelleted. There are some initial
indications that these procedures may infl uence responses to microbial phytase.
Kasim and Edwards (2000) offered maize–soy diets to broilers in which the
grain component (532 g kg–1) was ground to three different sizes with geometric
mean diameters of 484 μm (fi ne), 573 μm (medium) and 894 μm (coarse).
Determined on a total-tract basis, retention of phytate-P increased (P <0.05)
with particle size from 0.389 (fi ne) to 0.426 (medium) to 0.457 (coarse). The
addition of 600 FTU phytase kg–1 further increased (P <0.01) phytate-P
retention to 0.558, 0.585 and 0.628, respectively, and there was no treatment
interaction. Similar fi ndings have been reported by Berwal et al. (2008), in
that increasing particle size of a maize-based diet was associated with higher
total P retention. Subsequently, Amerah and Ravindran (2009) offered broilers
maize–soy diets in which the grain was ground to medium (611 μm) and coarse
(849 μm) particle sizes. In this study, 500 FTU phytase kg–1 increased toe ash
of broilers offered medium-ground maize diets (11.65 versus 10.41%;
P <0.05). However, there was a treatment interaction (P <0.01) because
phytase did not signifi cantly infl uence bone mineralization (11.78 versus
11.42%) in coarse maize diets. The authors suggested that coarsely grinding
maize had benefi cial effects on P bioavailability. Therefore, it is interesting that
Gabriel et al. (2008) reported that offering broilers diets containing whole
wheat (200–400 g kg–1) signifi cantly increased alkaline phosphatase activity in
the duodenum and jejunum by approximately 16.5%. It may be that stimulation
of gizzard function by feeding whole or coarsely ground grain in turn stimulates
the development of small intestinal mucosa and alkaline phosphatase secretion,
which could enhance P bioavailability.
The addition of microbial phytase to broiler diets based on either ‘raw’
wheat or the same wheat that had been pre-pelleted (90°C) was compared
(Selle et al., 2007). More robust AME and growth performance responses
were observed following the addition of phytase to ‘raw’ wheat diets, but
treatment interactions were not signifi cant. However, phytase increased N
retention in broiler diets based on ‘raw’ wheat but depressed N retention with
pre-pelleted wheat, so that there was a signifi cant (P <0.01) treatment
interaction. There is some evidence to suggest that heat-treating wheat reduces
phytate and protein solubility (Ummadi et al., 1995) and, if so, it follows that
phytate may be less readily enzymatically degraded and the extent of protein–
phytase complex formation may be reduced. This suggests that high pelleting
temperatures of diets may depress responses to phytase supplementation.
Other enzymes
It has recently been reported that the benefi cial effects of exogenous xylanase
in poultry and swine diets are inextricably linked to the size of the undigested
portion of fat, protein and starch that leaves the ileum (Cowieson and Bedford,
2009). This observation, supported by some 19 peer-reviewed papers
published between 1998 and 2009, rules out, by defi nition, full additivity
between pro-nutrients. As phytase (whether credited or not) improves ileal protein, fat and starch digestibility by reducing endogenous loss and improving
dietary nutrient solubility, it thereby reduces the undigested fraction. Thus, in
this situation the energy matrix for xylanase should be reduced (by around
20%) in the presence of phytase to acknowledge the now reduced undigested
fraction. By defi nition then, only the fi rst additive of choice can carry its full
matrix when added to a ‘virgin’ diet, but subsequent additives should have their
matrices discounted to accommodate the infl uence of the current incumbents.
As theoretical (if not realistic) maximum ileal digestibility is 100%, digestibilityenhancing
pro-nutrients constantly move digestibility toward that fi xed
asymptote, so opportunity for further improvement declines with each new
addition. It is therefore recommended that, if the matrix values that a supplier
promotes were established in diets that do not contain phytase, antibiotic
growth promoters, coccidiostats and other commonly used additives, the
matrix be discounted proportionate to the benefi ts of the incumbents. For
example, an energy matrix of 100 kcal kg–1 for a xylanase may end up being
50–60 kcal kg–1 in a diet containing an array of performance- and digestibilityenhancing
therapeutics and enzymes.
Energy matrices and added fat
Conventionally, the energetic benefi ts conferred by exogenous enzymes are
captured by a reduction in the lipid concentration in the diet, i.e. removal of
vegetable or animal fat sources. However, it is important to note that enzymes
are not necessarily a suitable direct replacement for fats and oils, as extracaloric
effects of lipids will not be delivered through the use of enzyme
technology. Examples of extra-caloric benefi ts of fat include pellet quality,
essential fatty acids, fat-soluble vitamins (A, D, E and K), balancing gastric
emptying with protein and carbohydrate digestion, mill effi ciency (energy use
and throughput) and perhaps even heat increment. Clearly, xylanases and
phytases are not direct replacements for these important effects and so the
removal of fat to accommodate the energy matrices of enzymes should be
done with care. In a recent study (Cowieson, 2010), the removal of 2% soy oil
from a maize soy-based broiler diet resulted in a signifi cant decrease (~3%) in
ileal amino acid digestibility at day 21. Interestingly, this effect was not observed
by day 42 (change from PC to NC = approximately 0.4%), and furthermore
not all amino acids were similarly infl uenced. This observation supports a
previous report in piglets (Li and Sauer, 1994), where the removal of canola oil
resulted in a signifi cant reduction in amino acid digestibility. Presumably these
effects are mediated by changes in gastric empyting, which is driven in part by
dietary fat concentrations (Stacher et al., 1990; Gentilcore et al., 2006), i.e.
low-fat diets may reduce residency of feed in the proventriculus/gizzard, or
even residency of food in the intestinal tract per se (Mateos et al., 1982).
It is interesting that the amino acids most detrimentally infl uenced by the
removal of added fat are those that have been shown to be released last from
the sequence of endogenous proteolytic mechanisms (Low, 1990). Thus, the
removal of oil to accommodate the metabolizable energy advantages that enzymes confer may be unwise in young animals, as this strategy may
inadvertently compromise ileal amino acid digestibility, especially for threonine
which tends to be the last dietary amino acid to be exposed to exopeptidase
activity. Additionally, removal of fat may compromise the digestibility of nonlipid
energy sources such as glucose and fructose (Mateos and Sell, 1980),
another cause for constraint in application of bullish energy matrices in young
animals. It may be wise to employ moderation in fat removal in starter diets
and to capture the economic value of energy matrices in the grower and
fi nisher phases, when fat concentrations are higher and the animal is less
susceptible to gastric digestion constraints. Instructively, rapid gastric emptying
caused by the ingestion of diets with a low fat density does not persist, as
compensatory mechanisms are activated over time (Covasa and Ritter, 2000).
These deleterious effects may be transitory and restricted to neonates, a
contention that is supported by a previous report (Cowieson, 2010).
A further unforeseen consequence of reduced gastric residence time is that
the effi cacy of a phytase, if present, will also be compromised, since the
proventriculus/gizzard is thought to be the most relevant for phytase activity.
Thus a dietary modifi cation made in order to profi t from the energy-sparing
benefi t observed when a xylanase is used may result not only in direct losses in
amino acid and starch digestibility but also in phytate hydrolysis, with the
ensuing further losses in mineral, energy and amino acid benefi ts that were
attributed to phytate hydrolysis.
Conclusions
The gastrointestinal tracts of pigs and poultry differ structurally, physiologically
and functionally; therefore, it is not surprising that responses to the dietary
inclusion of phytases differ between the species. Somewhat paradoxically,
phytases appear to degrade phytate to a greater extent and liberate more
phytate-bound P in pigs than in broiler chickens, but the ‘extra-phosphoric
effects’ of phytases appear to be pronounced in broiler chickens. In a parallel
situation, growth performance and nutrient utilization responses to non-starch
polysaccharide (NSP)-degrading enzymes are typically of a greater magnitude
in broiler chickens than in pigs. Perhaps this is because grower-fi nisher pigs
are better able to tolerate the anti-nutritive effects of either phytate or NSP
than broilers. However, weaner pigs are probably more vulnerable to phytate,
as refl ected in feed effi ciency responses to phytase in relation to dietary phytate
levels (Selle et al., 2003a), which may refl ect the relative immaturity of their
gut development.
While microbial phytases have been used in practice for nearly two decades,
many advances could be made in their application in apparently fundamental
areas. The rapid and accurate determination of dietary phytate levels is one
example. Another is to establish the quantity of Ca actually released by phytase,
as it seems that this may be understated at present and further reductions in
dietary Ca levels are feasible, and that such reductions would enhance enzyme
effi cacy. The extent to which phytase increases ileal amino acid digestibility and/or protein availability in pigs and poultry still requires clarifi cation so that
full advantage of the ‘protein effect’ of phytase may be taken. This situation is
at least equally true for the possible phytase-induced enhancement of energy
utilization.
The likelihood remains that more effective exogenous phytases and/or
combinations with other facilitative enzymes will be developed. In this regard,
inherent phytate-degrading capacity, a broad pH spectrum of activity, resistance
to endogenous proteolytic enzymes, thermostability and the feasibility of higher
inclusion rates are all key factors. In this event, a better appreciation of how
best to manipulate diet formulations to take full advantage of higher phytate
degradation rates will be needed. Assuming that these advances take place,
exogenous phytases will be added to an even larger majority of pig and poultry
diets on a global basis. The growth in acceptance of feed enzymes in pig and
poultry production over the last two decades has been an extraordinary
development, as inclusions of NSP-degrading enzymes in wheat- and barleybased
poultry diets have already reached saturation point (Bedford, 2003). The
acceptance of exogenous phytases will also approach this point, with
appropriate scientifi c advances to the benefi t of sustainable pig and poultry
production.