The effect of diluting diets with ground and pelleted or with whole wheat on the performance of growing turkeys *

The objective of this study was to determine the effect of moderate dilution (gradually up to 22.5%) of standard diets (C) with wheat in different physical forms, ground and pelleted (GW) or whole (WW), on growth rate, feed conversion ratio (FCR), carcass yield, and incidence of footpad dermatitis (FPD) in growing heavy-type turkeys. Dietary treatments in which part of the basal diet (control, C) was replaced with a low, medium, or high wheat content (8.2%, 13.3%, or 18.2% on average in groups WL, WM, and WH, respectively) were compared. Seven experimental groups consisted of six replicates each (n=30 birds) fed from 4 to 18 weeks of age. After four weeks of experimental feeding, groups GWL and WWL, relative to group C, were characterized by higher coefficients of dry matter digestibility (P=0.003), apparent nitrogen retention (P=0.096), and energy metabolizability (P=0.003), as well as by a higher metabolizable energy content of the daily ration (P<0.001). The beneficial influence of wheat, both ground and whole, was lower at higher levels of dietary dilution. In 8-week-old turkeys, the symptoms of FPD were significantly less severe in groups GWL and WWL (P=0.039). In birds aged 18 weeks, no significant differences were found among groups. Neither the amount of wheat nor its physical form had a significant effect on the average body weight of growing turkeys. Throughout the experiment, only in groups WWM and WWH were FCR values lower than in group C (P<0.001). An increase in the wheat content of diets led to a linear decrease in FCR (P<0.001), and average FCR values were significantly (P=0.004) lower in WW than in GW groups. Whole wheat, compared with ground pelleted wheat, contributed to an increase in gizzard weight (P=0.003), although it had no influence on the proportions of breast muscles, leg muscles, and fat in turkey carcasses. * Supported by the Polish Committee for Scientific Research, Project No. NN 311 400 339 3 Corresponding author: e-mail: dariusz.mikulski@uwm.edu.pl Journal of Animal and Feed Sciences, 21, 2012, 735–747 736 WHOLE WHEAT AND PERFORMANCE OF TURKEYS


INTRODUCTION
Recent years have witnessed the growing popularity of feeding whole grain to poultry.This tendency results from economic reasons, the beneficial influence of whole grain on the development and health of the digestive tract of birds, and greater consumer acceptance of a more 'natural' feeding system (Gabriel et al., 2008).Originally, the purpose of such treatment was to reduce the transport and processing components of feed costs (Majewska, 1996;Bennett et al., 2002b).Studies have also shown that this mode of feeding improves the growth performance of birds and feed efficiency (Ravindran et al., 2006), most probably due to better development of the gastrointestinal tract of broilers and an increase in relative gizzard weight (Amerah et al., 2007).A large and more active gizzard can improve starch digestibility by serving as a mixing compartment for digestive juices and nutrients (Hetland et al., 2002).The results of recent studies (Amerah and Ravindran, 2008;Biggs and Parson, 2009) justify the expectation that feeding whole grain will improve not only the function of the gizzard, but also the health of birds.A number of studies, extensively reviewed by Svihus (2011), have shown improvements in nutrient utilization and bird performance as a result of including whole grains or coarse feed and fibre particles in diets for broiler chickens.
In contrast to numerous experiments on chickens, studies investigating the effects of whole grain in growing turkeys remain scarce, and their results are inconclusive.Without dilution, turkeys have been successfully fed diets containing up to 50% whole wheat or barley (Bennett et al., 2002b).At the highest level of dilution, where whole wheat dilution was 50% of the diet, body weight and the weight of breast meat per bird were reduced by 15% and 20%, respectively (Bennett and Classen, 2003).Such results, reported for turkeys with final body weight below 14 kg, suggest that the degree of diet dilution with whole wheat should be decreased, particularly in heavy-type birds.
The objective of this study was to determine the effect of moderate dilution of standard diets with wheat in different physical forms (ground and pelleted or whole) on growth rate, feed conversion ratio, carcass yield and incidence of footpad dermatitis in growing heavy-type turkeys.

Birds and housing
The experiment was carried out at the Research Laboratory of the Department of Poultry Science, University of Warmia and Mazury in Olsztyn, according to the guidelines of the Local Animal Experimentation Ethics Committee.A total of 1260 four-week-old heavy-type BIG-6 turkey males were kept in pens (30 turkeys per pen, 10 m 2 ) on litter in a building with a controlled environment.The temperature and lighting programmes were consistent with the recommendations of Aviagen Inc. (2007).The birds had free access to feed and water.The trial was conducted to 18 weeks of age.

Diets
The four-phase feeding programme applied in the study followed commercial recommendations.Pelleted control diets (group C), whose composition is given in Table 1 and whose nutritive value was consistent with British United Turkeys recommendations (BUT, 2005), were used in each of the four successive feeding periods (three 4-week periods and the final 2-week period).In experimental groups, part of the basal diet (low, medium and high, respectively) was replaced with ground pelleted (GW L ,GW M, GW H ) or whole (WW L ,WW M, WW H ) wheat grain, according to the experimental design (Table 2).Experimental groups consisted of six replicates, 30 birds each.

Growth trial
At the end of each period, in week 8, 12, 16, and 18 of the experiment, the birds were weighed and feed intake was recorded.The feed conversion ratio was calculated for each period.At 8 weeks of age and at the termination of the experiment, the excreta dry matter (DM) was analysed and foot-pad dermatitis (FPD) scores were determined according to the method described by Hocking et al. (2008).

Balance trial
After 4 weeks of feeding, 14 birds were randomly selected from each treatment (7 replications, 2 birds each) and transferred to battery cages for balance trials.Cages were thermostatically controlled, and turkeys had free access to water from nipple drinkers.The birds received the experimental diets ad libitum.
All diets contained TiO 2 at a rate of 5 g/kg as an indigestible dietary marker for the calculation of nutrient digestibility coefficients.After a two-day preliminary period, a collection period was continued for 5 days; the excreta were collected daily, freeze-dried and ground through a 0.5 mm screen.

Sample collection
Eight birds representing the average body weight of the control group and groups with the lowest and highest addition of wheat (40 birds in total) were sacrificed by cervical dislocation.The carcasses were scalded, defeathered, and eviscerated.Following evisceration, whole carcasses were air pre-chilled at 12°C for 30 min, air chilled and stored at 4°C, and then hand-deboned on a cone 24 h post-mortem.Abdominal fat, giblets, breast meat (including M. pectoralis major and M. pectoralis minor), and leg meat (including thigh and drumstick) were weighed.The weights of the muscles, giblets and fat were determined relative to live body weight.

Analysis
Samples of feed and dried excreta were analysed in triplicate for dry matter and nitrogen (AOAC, 2005).Ti was determined by inductively coupled plasma-optical emission spectrometry with excitation (ICP-OES) using the Ultima 2 sequence camera with vertical plasma by Horiba Jobin Yvon (procedure developed in the GBC Poland, Niezdrowice Laboratory).The gross energy value (GE) of feed and excreta was determined with the use of a Parr adiabatic oxygen bomb calorimeter (KL-10, Precyzja, Bydgoszcz, Poland).

Statistical analysis
For performance parameters and FPD scores, a pen was considered a replicate experimental unit for the statistical analysis.Other results were analysed with each turkey as a replicate.The model assumptions of normality and homogeneity of variance were examined by the Shapiro-Wilk and Levene tests, respectively.The percentage data (e.g., excreta DM content, digestibility coefficients and carcass traits) were transformed to arcsine of the square root before analysis to achieve homogeneity of variance.Because no statistical differences were observed between the transformed and original (untransformed) data, the statistics presented in this paper were calculated for the untransformed data.All data were subjected to one-way ANOVA according to the GLM procedure for Statistica 8.0 software (StatSoft Inc., 2007) according to the following model: where: Y ij -the observed dependent variable; μ -the overall mean of all the observations; α i -the mean effect of the i th level of factor relative to µ; ε ij -the random error.
When a significant treatment effect was noted, the post-hoc two-sided Dunnett's test was used to determine differences between the control and each of the experimental groups.Differences between the forms of wheat (GW vs WW) were determined using the planned comparisons of trend analysis.In addition, linear and quadratic polynomial contrasts were used to evaluate the effect of different dietary levels of wheat, separately for each form of wheat, including the control group.Treatment effects were considered to be significant at P≤0.05.All data were expressed as mean values with pooled SE.

RESULTS
Due to the similar energy value of the control diets (Table 1) and added wheat (3070 kcal/kg), the energy content in all of the experimental diets was comparable.In diets with low, medium, and high wheat inclusion, the average protein concentration decreased by 3.4%, 5.8%, and 8%, respectively, in comparison with the control diet.
The coefficients of dry matter digestibility, apparent nitrogen retention and energy metabolizability, as well as the metabolizable energy content of the dietary treatments, determined after four weeks, varied depending on the amount of wheat added to the basal diet, but were not affected by the physical form of the wheat (Table 3).Statistical analysis revealed that in groups of turkeys fed diets with an increasing wheat content, the values of the above coefficients, relative to the control group, were represented by a quadratic function, with the highest effects in the groups with a low addition of wheat.The coefficients of dry matter digestibility noted in groups GW L and WW L were significantly (P=0.003)higher than in the control group, whereas in the remaining groups (GW and WW), the influence of wheat inclusion levels was lower.The effects of ground wheat (groups GW) and whole wheat (groups WW) on dry mater digestibility were expressed as quadratic equations, significant at P<0.001 and P=0.058, respectively.The same dependencies were noted for apparent nitrogen retention (C vs GW, quadratic contrast, P=0.019; C vs WW quadratic contrast, P=0.072); energy metabolizability (C vs GW, quadratic contrast, P<0.001; C vs WW quadratic contrast, P=0.052), and the metabolizable energy content of the daily ration (C vs GW, quadratic contrast, P<0.001; C vs WW, quadratic contrast, P=0.010).
After four weeks of feeding, the experimental factors were found to exert different effects on the dry matter content of excreta and FPD occurrence (Table 4).The differences in the excreta dry matter content among groups were minor, but certain trends were observed relative to the control group: the decrease in the dry matter content of excreta was a linear trend in groups GW (P=0.098) and quadratic in groups WW (P=0.002).The lowest dry matter content of excreta was noted in group WW L , and a considerably higher value was observed in group WW H . Wheat particle size did not significantly influence excreta dry matter content.In 8-week-old turkeys, the symptoms of FPD were significantly less severe in groups GW L and WW L (P=0.039).In the remaining groups, the effect of wheat inclusion was statistically non-significant.In birds aged 18 weeks, no significant differences were found among groups with respect to the dry matter content of excreta and FPD incidence, regardless of the experimental factors.
Neither the amount of wheat nor its physical form had a significant effect on the average body weight of growing turkeys (Table 5).Statistical analysis of the body weights of turkeys aged 8 and 12 weeks showed, however, that in the GW groups, body weight increased in response to low levels of ground wheat and decreased in response to high wheat inclusion, relative to the control group.the post-hoc two-sided Dunnett's test was used to determine differences between the control and each of the experimental groups.Means with different superscripts within the same column differ significantly (P<0.05);GW -diets with ground and pelleted wheat, WW -diets with whole wheat, W L , W M and W H -low, medium and high wheat content, respectively A similar trend was found for 12-week-old birds from groups WW, whose body weight increased in response to low and medium levels of whole wheat, and decreased in response to high wheat inclusion.In the first and second stages of the experiment (weeks 5-8 and 9-12), there were no significant differences in feed conversion ratio (FCR) among groups.In the third stage, FCR was significantly lower in groups GW H , WW M , and WW H than in the control group.FCR decreased linearly as the inclusion levels of ground or whole wheat in basal diets increased (P=0.003 in both cases), but throughout the experiment, only in groups WW M and WW H were FCR values lower that in group C (P<0.001).The average FCR values between week 5-18 were significantly (P=0.004)lower in groups WW than in GW.
In birds aged 18 weeks, the yields of breast muscles, leg muscles (thigh and drumstick), and abdominal fat were similar in all groups (Table 6).Significant (P=0.026)differences were noted for gizzard weight, which was lower in the GW groups and higher in the WW groups (P=0.003).Turkeys fed diets with a high wheat content had a lower liver weight (P=0.013) in comparison with those fed diets with a low wheat content.

DISCUSSION
In our study, the dietary inclusion levels of wheat grain were relatively low, averaging 8.2%, 13.2%, and 18.2%.Thus, in diets with low, medium, and high wheat contents, the average protein concentration decreased by 3.4%, 5.8%, and 8%, respectively, in comparison with the control diet.The experimental diets met the protein requirements of growing turkeys (NRC, 1994).
In the first stage of the experiment, low (5%) inclusion of wheat in the diets had a beneficial influence on dry matter digestibility, and on nitrogen and energy utilization.Irrespective of the level of wheat inclusion, wheat particle size had no significant effect on nutrient digestibility and utilization.Different results were reported for broiler chickens fed diets containing 10%-20% whole wheat, where metabolizable energy concentrations and amino acid digestibility were found to increase (Biggs and Parson, 2009).A positive effect of feeding whole cereals to broilers on the digestibility of nutrients, including starch, has also been observed by other authors (Hetland et al., 2002;Wu et al., 2004;Svihus et al., 2010), and could result from improved gastrointestinal motility (Amerah et al., 2007).
Compared with this study, relatively lower or similar results of body weight values were observed when turkeys were fed diets with soyabean protein products (Jankowski et al., 2009) and rapeseed meal (Mikulski et al., 2012).Diet dilution with wheat had no significant effect on the final body weight of turkeys.Despite certain differences in the growth rate of birds observed during the experiment, their final body weights were similar, regardless of the inclusion levels and physical form of wheat.In a study by Bennett et al. (2002b), total weight gain was also unaffected by feeding diets with 20% barley to turkeys with final body weights of 11 kg.In another experiment by the same authors (Bennett and Classen, 2003), dilution of turkey diets with wheat (at 21%-29% on average) for 3 to 16 weeks markedly decreased final body weight and feed utilization.The results of our study and the findings of other authors indicate that the levels of diet dilution with wheat should be lower in turkeys than in broiler chickens, since in the latter even high inclusion levels of wheat (up to 30%) do not decrease the final body weight.
In the current study, diet dilution with whole wheat vs ground and pelleted wheat, was found to improve FCR over the entire experimental period.These results differ from those reported by Bennett and Classen (2003) who found that dilution of turkey diets with whole wheat (21%) caused a progressive decline in feed utilization in comparison with undiluted diets.The results of previous experiments on broiler chickens are also inconclusive: in some of them diet dilution with wheat improved FCR (Bennett et al., 2002a;Ravindran et al., 2006), while in others such an effect was not observed (Svihus et al., 2004;Amerah and Ravindran, 2008).
In our study, FPD incidence varied among young turkeys, but at the end of fattening (at the highest level of diet dilution with wheat) there were no significant differences among groups in this respect.Similarly, diet dilution with wheat had no significant effect on the dry matter content of excreta, and high excreta moisture is correlated with FPD occurrence in poultry (Wang et al., 1998).
Neither level nor form of wheat affected the proportions of major muscle groups (including breast muscles) in turkey carcasses.The experimental factors had no effect on the final body weights of birds, either.This suggests that the addition of wheat to diets, which reduced their total protein content, was well tolerated by growing turkeys.In a different experiment (Bennett and Classen, 2003), a high level of diet dilution with whole wheat (from 21% to 29%) significantly reduced the final body weight and the weight of breast meat per bird, by 15% and 20%, respectively.We did not note a decrease in carcass fatness in response to feeding wheat-diluted diets, which was reported in an experiment involving broiler chickens (Amerah and Ravindran, 2008).

CONCLUSIONS
The results of our study indicate that moderate (below 20%, on average) dilution of standard turkey diets with wheat has no effect on the final body weights of birds or muscle yield, and improves feed utilization, in particular when whole wheat is fed instead of ground or pelleted grain.

Table 2 .
Experimental design 1 ground and pelleted wheat (GW diets) and whole wheat (WW diets); W L , W M and W H -low, medium and high wheat content, respectively

Table 3 .
Coefficients of dry matter and energy digestibility and apparent retention of nitrogen 1 energy metabolizability = (gross energy -energy of the excreta) : gross energy, %; GW -diets with ground and pelleted wheat; WW -diets with whole wheat; W L , W M and W H -low, medium and high wheat content, respectively; a-b the post-hoc two-sided Dunnett's test was used to determine differences between the control and each of the experimental groups.Means with different superscripts within the same column differ significantly (P<0.05)

Table 4 .
Dry matter (DM) content of the excreta and incidence of foot-pad dermatitis (FPD) in turkeys aged 8 and 18 week

Table 5 .
Body weight (BW) and feed conversion ratio (FCR) of turkeys Indices

Table 6 .
Effect of dietary treatments on meat yield and abdominal fat content in turkeys, % of live body weight