The feed enzyme market today is worth approximately $1 billion USD and is dominated by NSP’ases and phytases, with proteases also playing a minor but significant role. The principal reason for their employment is that they improve digestibility of a number of nutrients, specific to the enzyme employed, and as a result enable the use of less digestible, cheaper and most frequently more sustainable ingredients in the ration employed. In the case of phytase the nutrient of greatest interest is phosphorus, although ancillary savings in energy and amino acids, if taken into account, are also of significant benefit. NSP’ases principally spare energy, with some evidence that amino acids are also spared, particularly in viscous grain diets. The link between the use of such enzymes on performance is clearly self evident, whereas that with health are more difficult to grasp immediately. Since the effect of each enzyme class is quite distinct, each will be dealt with in turn and then the effect of a combination of the two will be considered. 

Early NSPase work focused on the anti-nutritive effects of barley and established that the principal problem was its content of soluble, viscous B-glucans which markedly reduced the ability of the broiler to digest the diet efficiently (Burnett, 1966; Fry et al., 1958; Moran and McGinnis, 1968) and resulted in viscous, sticky droppings. Application of B-glucanases, which depolymerised and thus removed the anti-nutritive capacity of these NSP, quickly restored performance and removed the problem of sticky droppings at the same time. A similar mechanism, but different culprit, arabinoxylans rather than B-glucans, was identified in rye, and to a lesser extent triticale and wheat based diets with the solution being the application of an endoxylanase which effected the same response as the glucanase in barley based diets (Annison, 1992; Bedford et al., 1991; Campbell et al., 1983; Campbell et al., 1991; Grootwassink et al., 1989). Although the responses to the use of an appropriate enzyme in rye and barley based diets were almost entirely beneficial and that in wheat and triticale based trial most often beneficial, it was noted that there is a huge variation in the response of the animal to such enzymes (Rosen, 2002a). Many nutritional and environmental factors impinge on the ability of an NSP’ase to elicit a response including. 

Although the factors seem unrelated in fact there is a reasonably consistent thread which ties them all together. Essentially any factor which contributes to increasing intestinal viscosity (items 1, 2, 3 and 4) results in slower digestion and hence poorer performance. Potential health benefits result from the fact that slower digestion enables a greater proportion of nutrients to escape absorption and reach the large intestine.  Under such circumstances the ensuing proliferation of the microflora can lead to detrimental fermentation patterns which can result in increased microbial degradation of bile acids, for example (hence item 9) and increased likelihood of disease (Riddell and Kong, 1992) (hence item 6). Generally, microbial pressure is greater in larger scale, pen trials, (hence item 7) which only leaves items 8 and 5, the latter to be addressed below. For reasons which hopefully become apparent, the effect of NSP’ase enzymes on both performance and health needs to be split out into viscous and non viscous diets. 

The effect of NSP’s and thus NSP’ases very much depends upon the principal cereal in the ration. Cereals such as rye and most barley varieties exhibit such extreme intestinal viscosities that this mechanism overwhelms and renders insignificant any secondary effect of “encapsulation” as a potentially anti-nutritive effect of the NSP’s. With increasing intestinal viscosity the bird perceives a reduction in nutrient density and compensates by eating slightly more until a threshold is reached, which is likely in the region of 20 mPas, at which point the digesta is too viscous for further compensatory increments in passage rate. The consequence is that when viscosity increases from 1 to 20 mPas, the principal problem is a deterioration in FCR with little loss in gain as intake is able to compensate, whereas above 20 mPas both gain and intake fall with further viscosity increments and FCR is doubly compromised. 

The graph above is adapted from work where a range of intestinal viscosities were generated by blending wheat and rye in various proportions and a xylanase used in each diet at one of 6 levels (Bedford and Classen, 1992).  It is self evident from the graph above that diets which generate viscosities in excess of 20 mPas will have far more detrimental effects on performance than those that are below 10. 

The physical constraints on digesta flow rates above 20 mPas not only increases digesta residence time, hence increasing the time available for the intestinal microbiome to replicate, but it also reduces diffusional rates of oxygen from the villi to the digesta. As a result the environment in the ileum becomes much more anaerobic, favouring the growth of opportunistic pathogens such as C. perfringens as has been suggested by earlier work (Riddell and Kong, 1992). In the case of high viscosity diets, the effects of NSP’ases on both performance and health are therefore much more self-evident than in the case of low viscosity diets. In that regard, maize based diets rarely result in viscosities in excess of 10 mPas, and thus ut is likely that viscosity plays only a minor role in the response to NSP’ases in such diets. Moreover, in recent years our experience has been that the average wheat based diet is nowhere near as viscous as it was in the 1990’s, which in the EU has been due to the removal of the 1B1R translocation gene from all wheat varieties (originally bred in from rye to increase yield and resistance to yellow rust, but which also resulted in higher viscosity). As a result many wheat based diets are not that much more viscous than their maize counterparts. Nevertheless there is clear evidence that such low viscosity maize and wheat based diets do respond to NSP’ases, suggesting a different mechanism may be dominant as will be discussed briefly below. 

With viscosity being of only marginal importance in such diets, the direct effects on health in particular are difficult to envisage. The cell wall or encapsulation effect, also known as the cage effect, is routinely used to explain the mode of action of NSP’ases used in low viscosity diets. It is suggested that the NSP’ase breaks down the cell walls which encapsulate the nutrients in those cells that have remained intact after feed processing, such that the nutrients within are exposed to digestive enzymes more rapidly. In this regard the nutrients entrapped within the intact cells are digested and absorbed prior to exposure to significant microfloral populations in the large intestine and hence bacterial overgrowth is avoided. Indeed, microscopy of jejunal contents has indicated significant cell wall destruction on use of an NSP’ase suggesting encapsulation is a real effect. This may be too simplistic a theory however, since in vitro work purporting to support this mechanism utilised doses well in excess of 10 times and close to 100 times that employed commercially before significant increments in cell wall destruction were noted (Le et al., 2013; Tervila-Wilo et al., 1996). Also, given the short time taken for digesta to reach the jejunum where such destruction is noted, and the less than favourable pH for commercial NSP’ases en route through the gastric phase, it is unlikely that the cell wall effect is fully explained by direct enzymatic depolymerisation. Further challenges to the cell wall effect come from digestibility data. Empirically it has been shown that the use of a xylanase in both pig and poultry rations leads to a constant 16% improvement in digestibility of the undigested fraction of each amino acid (Cowieson and Bedford, 2009).  If this is the case then the NSP’ase is improving digestibility of all amino acids in a consistent manner regardless of whether they are derived from the cereal fraction of the diet or from other protein sources, including animal based proteins. This is clearly inconsistent with the cell wall mechanism and leaves low viscosity diets with no clear mechanism of response to NSP’ases.

However, recent mode of action work suggests that the benefits of NSP’ases in low viscosity diets may be intrinsically linked with the microfloral populations in the lower gut and thus influence not only performance but also the health of the bird. The theory has been proposed back in the 1990’s but recently gained support. NSP’ases produce undigestible but fermentable oligosaccharides from their activity on the cell walls of the cereal. These are fermented to VFA’s which are monitored by the L cells in colon and caeca of the animal, and in response to elevated levels they release entero-hormones including peptide YY. Indeed the inclusion of a xylanase in a corn based diet was shown to result in elevated levels of PYY in the blood (Masey O'Neill et al., 2012a; Singh, 2012). This hormone has been shown to delay gastric emptying, enabling more complete gastric digestion and in the case of the chicken perhaps greater chance to physical breaking open of the endosperm cells which would account for the microscopic evidence of cell wall destruction in the jejunum. Such a mechanism would also account for the observed improvement in digestion of all undigested amino acids to a similar degree, and not just those that predominate in the target cereals. Given the crux of this mechanism is the production of fermentable CHO in the large intestine, and that the balance between fermentable CHO and N is critical in maintaining intestinal health, this suggests that any activity which increases the amount of fermentable CHO will likely reduce the risk of an intestinal upset provided the CHO is beneficial and the animal is capable of supporting such a population of organisms. With regard to the latter, the age of the animal will clearly influence the potential response as the more mature the animal, the larger and more complex the large intestinal microbiome, and thus the more likely it is to be able to respond to the provision of fermentable CHO. Thus, even in low viscosity diets, the microflora likely are involved in the response to an NSP’ase and as a result will influence the general health of the animal. 

When used as simply a P and Ca replacement strategy it would only be expected to restore performance if employed in a P deficient diet. The likelihood that phytase would be viewed differently from inorganic P with regards to its effect on performance or health was and probably still is not considered. However, when phytase is considered as an enzyme that removes phytate, an antinutrient, rather than an enzyme that simply provides P, it becomes evident that performance and perhaps even health may benefit on use of this enzyme. It should be noted that for this benefit to be realised the vast majority of phytic acid and its lower phospho-esters need to be hydrolysed, an end point that is not achieved with the more traditional dosages of phytase. The benefits of almost quantitative phytate hydrolysis accrue from the fact that phytate and its lower esters are thought to significantly impair digestion, particularly in the gastric phase, such that the animal hypersecretes both acid and pepsin in an attempt to maintain optimum digestion rate (Cowieson et al., 2011). Additional mucin is secreted to protect the gastric environment from self-digestion, the net result being that the efficiency of amino acid and energy digestion is compromised, with endogenous amino acids featuring heavily in the excreted amino acids. Thus on use of sufficient phytase to remove the vast majority of phytate from the diet, there is considerable sparing of endogenous inputs and a reduction in the amount of undigested but fermentable material, particularly N, which can enter the large intestine and presumably support putrefactive growth of pathogens. Certainly the use of “superdoses” of phytase, ie dosages capable of releasing at least 80% of all phytate P, has resulted in consistent and significant improvements in the FCR of broilers and piglets (dos Santos et al., 2013; Walk et al., 2013). It is interesting to note that use of phytase in broilers for phosphorus replacement does not usually influence FCR, only intake and gain (Rosen, 2002b) further suggesting that these effects are not related to phosphorus release.

With regards to anti-oxidant status, there is some evidence that low P diets reduce the stores of antioxidants in the liver of broilers and use of high doses of phytase can restore such deficits (Karadas et al., 2010), and recent work has suggested that Se stores in the egg are increased with use of very high doses of phytase in P adequate diets (Santos, 2013, pers communication). If high dosages of phytase can indeed improve the anti-oxidant status of the animal then it may prove of value for the health of animals subjected to oxidative stresses. 

Efficacy of enzymes if variable and depends upon the undigested fraction

The response to a given enzyme will depend upon the performance of the control animal to which it is offered (Rosen, 2002a; Rosen, 2003), and whether the constraints on performance, if any, are related to the nutrients which will be unlocked by the enzyme. For example, an animal deficient in P will most commonly benefit from the inclusion of a phytase in its diet, but the scale of the response will depend upon the degree of the P deficiency and the ability of the phytase to release the required P. P deficiency can be exacerbated significantly be increasing the Ca content of the diet and reducing vit D levels for example, and the efficacy of a phytase can be significantly increased by use of low Ca diets, use of diets rich in phytate, inclusion of fat and fibre which encourage gastric retention and hence increases the time of exposure of dietary phytate to phytase. NSP’ase enzymes are expected to be more effective if the cereals are of poor quality and the enzyme is able to address this quality issue. For example larger responses to NSP’ases are routinely found when xylanases are applied to poor as opposed to good quality wheat and barley based rations (Barrier-Guillot et al., 1995; Bedford et al., 1998; Dusel et al., 1998; Scott et al., 2001). However the data for maize tend to suggest the response is constant regardless of initial quality although it has to be noted that the range in performance between samples was small compared with that observed for wheat and barley samples (Masey O'Neill et al., 2012b).

The greater the constraints on digestion, the greater the impact the enzyme will have on performance. Indeed Rosen (2002) noted that the single most important factor in describing the scale of response to any feed enzyme was the performance of the control, which is a composite of the quality of the neonate, the ingredients, the nutrient specifications and management. The likelihood of small intestinal health problems is increased when digestion is impaired for whatever reason and large quantities of nutrients enter the distal regions of the tract where they can be fermented. In such cases the benefits of an appropriate enzyme may manifest not only in improved growth rate and efficiency, but also in terms of improved intestinal health.

It should be noted that such effects are highly variable and dependent upon the circumstances that each flock are exposed to. Those flocks fortunate enough to be composed of chicks from 34week old broiler breeders, who suffer no heat stress or limitations in water/feed, have optimal nutrition as a result of accurate formulations using ingredients that are all highly digestible are likely to benefit very little from the use of an enzyme. Those flocks exposed to the combination of the worst of all variables however will likely benefit the most. The use of an enzyme can in such cases be viewed as a tool to improve the average performance of the producer by radically reducing the number of very poor flocks without necessarily improving performance in the best – in this regard they can be seen as somewhat of an insurance policy. 

If a phytase improves nutrient digestibility, and an NSP’ase likewise, then the combination of individual enzymes may be synergistic, additive, sub additive or even detrimental. If the matrices of each enzyme are taken at face value, then mistakes in their employment will accrue unless their combination is shown to be additive as is assumed.  

A recent review of the literature indicated that the whilst xylanases (19 studies)  and phytases (16 studies) both improved ileal digestibility of amino acids when applied in isolation, when the combination were used the effects were subadditive (5 studies) suggesting that simple addition of the amino acid matrices of both would over-estimate the value achieved (Cowieson and Bedford, 2009).  Given that NSP’ases recover a fixed proportion of the undigested amino acids present in the diet, and that co-administration of a phytase effectively reduced the content of undigested amino acids, the use of a phytase effectively reduces the substrate on which the NSP’ase can act. This work therefore suggests sub-additivity of the two enzymes.

A confounding factor to consider is that in much of the literature, when phytases and xylanases are combined factorially, this is done so in diets which are deficient in P in order to guarantee a response to the phytase. Low P diets would be expected to respond to a phytase moreso than xylanase since the former specifically releases P and the latter assumed to release energy. The response to the xylanase in the absence of a phytase may be compromised as it will be releasing nutrients which are not growth limiting. Commercially, however, the xylanase will be employed in a diet containing a phytase which is not limiting in P, and thus the literature may not represent commercial reality. 

Such a review of the literature regarding the effect of a combination of xylanases and phytases on animal performance was conducted by Rosen in2004 inwhich he commented on the points raised above (Rosen, 2004). At that point in time only 11 publications provided a total of 17 2x2 factorial combinations of the two enzymes. He found that in general the effects of xylanase and phytase were at best additive as far as FCR was concerned, although this was only true in wheat based diets which constituted the bulk of the trials investigated (15 out of 17 tests) and sub additive for gain.  He also noted that the extent of the response was dependent upon the performance of the control diets, the poorer the control performance, the bigger the response to both enzymes. Given the control performance was often dictated by the AvP level of the diet, ie P deficient diets resulting in poor control performance, it suggests that even the xylanase may be improving performance in low P diets. This counter-intuitive consideration was not addressed in the work of Rosen. 

In the present work, 13 publications were collected between 2005 and the present day where NSP’ases and phytase were factorially combined and fed to broilers, 12 of which provided performance data. From these 12 papers there were 41 tests where the NSP’ases were added to a diet in which a phytase was either present or not such that the effect of the presence of phytase on the NSP’ase response could be determined.  The average age of birds used was 23d with the vast majority of trials being 1-21d, with only 1 trial going out to 35d of age. Ca and P levels averaged 0.84 and 0.30% respectively and ranged from 1.67-0.64% and 0.45 to 0.15 respectively with ME averaging 3017 kcals/kg but ranging from 3140-2868. All but two of the trials used males (the others being mixed sex and undeclared!) and all but 4 studies fed mash diets, the rest cold pellet (2),70Cpellet (1) or not declared (1). Of the 41 data points, 24 were maize based diet with the balance being wheat.  The average responses are shown in table 1 below. 

Table 1. Percentage response of broilers to the addition of a NSP’ase overall, in the presence and absence of a phytase, compared with the relevant control 

The average response to NSP’ase across the whole data set was a 2.5% response in gain and a 2.3% reduction in FCR. When the xylanase was used in isolation, the gain response was marginally larger, at 3% and FCR approximately the same. This was surprising given the fact that in such diets P was most often the limiting nutrient. This suggests that xylanase can release some P and thus improve performance, but it must be noted that in such P deficient diets the response to phytase was more than twice that recorded for xylanases. In the presence of the phytase the effect of the NSP’ase on FCR and gain was weakened, suggesting sub-additivity and thus some cross over in mechanism. It must be noted, however, that when the NSP’ase was placed in a diet containing the phytase, the weights of the “control” (ie the diet containing the phytase alone) were greater than that of the control diet containing no phytase (555g ave vs 523g ave). Thus the final weights of the birds fed the combination of enzymes was on average heavier than those fed either enzyme in isolation. 

Given the data presented here it seems that the combination of xylanases and phytases will overlap to a degree and as a result there may be a need to reduce the matrices of each enzyme when both are fed in combination. These conclusions are based on the average of the responses from the literature but clearly in some trials there is no additivity at all (Wu, 2004). In such cases it may simply be a case of the diet being relatively nutrient dense such that either of the two enzymes fed in isolation can release the rate limiting nutrients. Such observations are also evident for single enzymes when employed in nutrient rich diets and highlights the need to use appropriate diets to exploit the benefits of the enzymes employed. 

Mechanistically, the synergy between the two enzymes may well be due to the proposal that NSP’ases delay gastric emptying, and as a result allow the phytase to work longer in an environment conducive to phytate degradation – ie low pH. More work is needed to confirm such hypothesis but if found to be correct, the use of these two enzymes in combination may prove to be as beneficial commercially as has been suggested from the literature.

Conclusions 

The fact that exogenous enzymes influence the performance of the animal is almost without debate, although the scale of response is variable for the reasons noted above.  The link with animal health seems more tenuous but further consideration of the mode of action of these products suggests that this should not be the case. Better nutrient status enables the animal to mount more effective immune strategies when challenged and direct manipulation of the intestinal flora towards less putrefaction results in less challenge in the first place. Whilst feed enzymes can play a role in raising performance and the barriers to disease, they are not a panacea, however, and cannot compensate for poor hygiene or overwhelming challenge. Nevertheless their use does reduce variability between flocks and individuals and hence susceptibility to intestinal disorders, delivering better and more consistent performance. 

Annison, G., 1992. Commercial Enzyme Supplementation of Wheat-based Diets Raises Ileal Glycanase Activities and Improves Apparent Metabolisable Energy, Starch and Pentosan Digestibilities in Broiler Chickens. Anim. Feed Sci. Technol. 38, 105-121.

Barrier-Guillot, B., Bedford, M.R., Metayer, J.P., Gatel, F., 1995. Effect of xylanase on the feeding value of wheat-based diets from different wheat varieties for broilers. Proc. WPSA 10th European Symposium on Poultry Nutrition, 324-325.

Bedford, M.R., Classen, H.L., 1992. Reduction of intestinal viscosity through manipulation of dietary rye and pentosanase concentration is effected through changes in the carbohydrate composition of the intestinal aqueous phase and results in improved growth rate and food conversion efficiency of broiler chicks. J. Nutr. 122, 560-569.

Bedford, M.R., Classen, H.L., Campbell, G.L., 1991. The Effect of Pelleting, Salt, and Pentosanase on the Viscosity of Intestinal Contents and the Performance of Broilers Fed Rye. Poult. Sci 70, 1571-1577.

Bedford, M.R., Scott, T.A., Silversides, F.G., Classen, H.L., Swift, M.L., Pack, M., 1998. The effect of wheat cultivar, growing environment and enzyme supplementation on digestibility of amino acids by broilers. Can. J. Anim. Sci. 78, 335-342.

Burnett, G.S., 1966. Studies of viscosity as the probable factor involved in the improvement of certain barleys for chickens by enzyme supplementation. Br. Poult Sci. 7, 55-75.

Campbell, G.L., Classen, H.L., Reichert, R.D., Campbell, L.D., 1983. Improvement of the nutritive value of rye for broiler chickens by gamma irradiation-induced viscosity reduction. Br. Poult. Sci.  24, 205-212.

Campbell, G.L., Teitge, D.A., Classen, H.L., 1991. Genotypic and environmental differences in rye fed to broiler chicks with dietary pentosanase supplementation. Can. J. Poult. Sci.71, 1241-1247.

Cowieson, A.J., Bedford, M.R., 2009. The effect of phytase and carbohydrase on ileal amino acid digestibility in monogastric diets: complimentary mode of action? Worlds Poultry Science Journal 65, 609-624.

Cowieson, A.J., Wilcock, P., Bedford, M.R., 2011. Superdosing effects of phytase in poultry and other monogastrics. Worlds Poultry Science Journal 67, 225-235.

dos Santos, T.T., Srinongkote, S., Bedford, M.R., Walk, C.L., 2013. Effect of high phytase inclusion rates on performance of broilers fed diets not severely limited in available phosphorus. Asian Australasian Journal of Animal Sciences, 26: 227-232.

Dusel, G., Kluge, H., Jeroch, H., Simon, O., 1998. Xylanase supplementation of wheat-based rations for broilers: influence of wheat characteristics. J. Appl. Poultry Res. 7, 119-131.

Fry, R.E., Allred, J.B., Jensen, L.S., McGinnis, J., 1958. Influence of enzyme supplementation and water treatment on the nutritonal value of different grains for poults. Poult. Sci 37, 372-375.

Grootwassink, J.W.D., Campbell, G.L., Classen, H.L., 1989. Fractionation of crude pentosanase (Arabinoxylanase) for improvement of the nutrtional value of rye diets for broiler chickens. Journal of the Science of Food & Agriculture 46, 289-300.

Karadas, F., Pirgozliev, V., Pappas, A.C., Acamovic, T., Bedford, M.R., 2010. Effects of different levels of  dietary phytase activities on the concentration of antioxidants in the liver of growing broilers. Journal of Animal Physiology & Animal Nutrition 94, 519-526.

Le, D.M., Fojan, P., Azem, E., Pettersson, D., Pedersen, N.R., 2013. Visualization of the anticaging effect of ronozyme WX xylanase on wheat substrates. Cereal Chem. 90, 439-444.

Masey O'Neill, H.V., Haldar, S., Bedford, M.R., 2012a. The role of peptide YY in the mode of action of dietary xylanase. Poultry Science Abstracts.

Masey O'Neill, H.V., Liu, N., Wang, J.P., Diallo, A., Hill, S., 2012b. Efect of xylanase on performance and apparent metabolisable energy in starter broiler diets containing one maize variety harvested in different regions of China. Asian-Australasian Journal of Animal Sciences 25, 515-523.

Moran, E.T., McGinnis, J., 1968. Growth of chicks and turkey poults fed western barley and corn- based rations: Effect of autoclaving on supplemental enzyme requirement and assymetry of antibiotic response between grains. Poult. Sci 47, 152-158.

Riddell, C., Kong, X.-M., 1992. The influence of diet on necrotic enteritis in broiler chickens. Avian Dis. 36, 499-503.

Rosen, G.D., 2002a. Exogenous enzymes as pro-nutrients in broiler diets. In: Garnsworthy, P.C., Wiseman, J. (Eds.), Recent advances in Animal Nutrition 2002, Nottingham University Press, Nottingham, pp. 89-104.

Rosen, G.D., 2002b. Microbial Phytase in broiler nutrition. In:Recent Advances in Animal Nutrition, Ed Garnsworthy, P.C., Wiseman, J. (Eds.), Recent advances in Animal Nutrition 2002, Nottingham University Press, Nottingham, pp. 105-118.

Rosen, G.D., 2003. The effects of genetic, managemental and dietary factors on the efficacy of exogenous microbial phytase in broiler nutrition. Br. Poult. Sci. 44, S25-S26.

Rosen, G.D., 2004. Admixture of exogenous phytases and xylanases in broiler nutrition. In:  Worlds Poultry Congress, Istanbul, Turkey.

Scott, T.A., Leslie, M.A., Karimi, A., 2001. Measurements of enzyme response with hulless barley-based diets full-fed to Leghorn and broiler chicks or restricted-fed broiler chicks. Can. J. Anim. Sci: 81, 403-410.

Singh, A., 2012. Effects of xylanase supplementation on performance, total volatile fatty acids and selected bacterial populations in caeca, metabolic indices and peptide YY concentrations in serum of broiler chickens fed energy restricted maize-soybean based diets. Anim. Feed Sci. Technol. 177:194-203

Tervila-Wilo, A., Parkkonen, T., Morgan, A.J., Hopeakoski-Nurminen, M., Poutanen, K., Heikkinen, P., Autio, K., 1996. In vitro digestion of wheat microstructure with xylanase and cellulase from Trichoderma reesei. J. Cereal Sci. 24, 215-225.

Walk, C.L., Bedford, M.R., Santos, T.S., Paiva, D., Bradley, J.R., Wladecki, H., Honaker, C.F., McElroy, A.P., 2013. Extra-phosphoric effects of superdoses of a novel microbial phytase. Poult. Sci 92, 719-725.

Wu, Y.B., 2004. Influence of phytase and xylanase, individually or in combination, on performance, apparent metabolisable energy, digestive tract measurements and gut morphology in broilers fed wheat-based diets containing adequate level of phosphorus. International J. Poult Sci. 3:450-455