Feeding whole grain with pelleted diets to growing broiler chickens

In three trials with growing broilers (aged 12-8 days) the proportion of whole grain (WG) was increased step by step up to 30% (Trial 1) along with commercial starter (S) and grower (G) pellets or up to 40% (Trial 2) and 50% (Trial 3) with special balancer (BA) pellets (crude protein 250 g/kg). WG accounted for about 20%, 31% and 38% of the broilers’ total feed intakes, respectively. In Trial 1, the grain was exclusively wheat (W). In the other trials, also mixtures of wheat and barley (W:B 70:30, 50:50) were used. Crude protein analyses of feed remains revealed that birds did not to any considerable extent select between pellets and grain. Mixing wheat with S or G pellets reduced feed intake significantly (P<0.05) and, as a consequence, the growth rate decreased by 3-5%. However, the feed efficiency (FCR) of the SW birds was significantly (P<0.05) better than that of the control and GW birds. In Trial 2, the BAWG treatments reduced feed intake (P<0.05) and weight gain, but there were no significant differences in FCR. In Trial 3, the growth rate was similar with all the regimens, and birds fed the BAW diet consumed feed less (P<0.001) but more efficiently (P<0.001) than those fed the other diets. In each trial, the WG feedings increased markedly (50-90%) the empty gizzard weight. Fatness of broilers decreased significantly (P<0.05) when whole wheat was fed with the S diet (Trial 1) but not so when fed with the BA diet (Trials 2 and 3). The present results suggest that during the growing phase, broiler chickens can be fed efficiently with mixtures of a pelleted starter or a pelleted balancer diet and whole wheat which can partly be replaced by barley.

decreases.Therefore the addition of wheat apparently tends to reduce the weight gain of broilers (Chwalibog and Eggum 1986, Petersen 1990, Elwinger et al. 1992, Jensen 1992).If the use of whole wheat is cautious or the nutrient contents of the complete diet is raised, the growth rate is similar to that of the controls (Tiller 1989, Covasa and Forbes 1994, Ristic et al. 1994).
Other feeding systems which include the use of whole grain are choice feeding and sequential feeding.These methods base on the use of a special balancer diet which in choice feeding systems is provided with whole grain ad libitum in separate feeders.In sequential feeding time-lim- ited ad libitum access to whole grain alternates with a balancer diet.Choice feeding of broilers has been studied extensively using whole wheat but it has not been applied in practice because there is a risk ofimbalanced selection of the components, and the system requires extra equipment.Forbes and Covasa (1995) concluded that birds need a period of learning or adjustment to choice feeding.However, the life of a broiler may be too short for this.According to Gumming  (1994) the learning period is 7-10 days for chickens, broiler chicks needing a longer period than layer-type chicks.Sequential feeding could be easier to apply in practice, but research into this regimen has been very limited.The length of the feeding period is important.Rose et al. (1995) reported the highest growth rates in the 8-h periods.Mixtures of crumbled or pelleted concentrates and whole grain have been studied previously but the results were not very encour- aging (Kiiskinen 1992).The objective of the present study was to examine feeding of whole grain mixed with a complete diet or with a spe- cial balancer diet in order to increase the pro- portion of grain and to replace part of the wheat with barley.

Material and methods
Ross broiler chicks were kept in floor pens (4 m 2) in a temperature controlled windowless room    S = starter, G = grower, F = finisher, SW = mixture of a pelleted starter and whole wheat, GW = mixture of a pelleted grower and whole wheat, SWP = pelleted mixture of a starter and wheat meal under a daily light period of 23 h.Peat was used as litter material.Each pen contained 60 sexed birds (15/m 2 ) which had free access to feed and water.Half of the pens contained male birds and the other half females.The treatments were allotted between the successive pens containing the same sex.Three trials were conducted.Trial 1 included five treatment groups with the feed- ing regimens shown in Table 1.The industrial complete diets which were used either as the only feed (Group 1, control) or mixed with whole wheat (Groups 2-4) were based on soybean meal, pea meal, wheat, barley and oats.According to the declarations, the starter diet contained crude protein (CP) 23.0%, metabolizable energy (ME) 12.0 MJ/kg, lysine 1.4% and methionine 0.6%.The corresponding values for the grower diet were 20.0%, 12.5 MJ/kg, 1.2% and 0.5%.All birds, excluding those of group 3, were given the same crumbled starter diet (S) up to the age of 11 days.Group 3 was fed with the starter diet up to the age of one week.Thereafter groups 2 and 3 were given a mixture of pelleted starter diet and whole wheat (SW) and group 4 a mixture of grower pellets and whole wheat (GW).
The proportion of whole wheat was gradually increased up to 30% (Table 1).Group 5 served, in fact, as a second control which was given a pelleted mixture of the starter diet and ground wheat in a proportion of 80:20 (SWP).
In Trial 2, all birds were fed the same indus- trial crumbled starter diet during the first 11 days.
The number of replicates (pens) for a treat- ment was 6, 8 and 12 in Trials 1, 2 and 3, re- spectively.The live weight of chicks was meas- ured as a total weight per pen at one day, 11 days and 37-38 days of age.Feed consumption was measured between the weighings.Total slaughter weight and mortality per replicate were re- corded.Abdominal fat pad, including gizzard fat and gizzard, was removed and weighed for 24 female birds per treatment in Trial 1, for 16 females per treatment in Trial 2 and for 20 male birds per treatment in Trial 3, excluding group 3 in Trial 1.In addition, six carcasses (3 males, 3 females) per treatment (1,2 and 4) in Trial 1 were taken in the slaughter house for determining the fat content of thigh and breast meat.Subcutane- ous fat was included in the meat samples.The fat content of the meat samples was determined by eluting the fat with dichloromethane-meta-  (1980).At the end of Trial 2, litter samples near the feeders were taken from six pens per treat- ment for dry matter determination.In statistical analysis the analysis of variance was applied using a SPSS program.The comparisons between treatments were performed using Tukey's test (Steel and Torrie 1960).

Results and discussion
The analyzed crude protein (CP) content for the starter and grower diets of Trial 1 were 225 and 204 g/kg, respectively (Table 3).The CP of the corresponding whole wheat containing diets de- creased to 192 (SW) and 181 g/kg (GW) at the end of the growing period.According to the CP analysis of feed remains, birds did not select between pellets and whole wheat.In Trial 2, the analyzed CP for the grower was 209 g/kg and that of the balancer 248 g/kg.In the BAW diets Table 3. Analysed crude protein (CP) content in the diets, grain and feed remains (in parentheses).
However, the weight gain of the wheat -fed broilers, can be regarded as satisfactory.Also the slaughter weights of the wheat -fed groups de- creased by 1-4% and there was a significant (P<0.05)difference between the GW and the control groups.Compared with the control and the GW feeding treatments, each SW feeding regimen improved significantly (P<0.05) the feed efficiency (FCR) and the FCR of group 2 was significantly (P<0.05)better than that of group 5 which was given ground wheat in the pellets.This implies that whole wheat feeding has a beneficial effect on the digestion, apparently due to the enlargement of the gizzard.
It seems that better results can be achieved when whole wheat is used with the starter pellets than with the grower pellets.This is under- standable, because the starters contain higher concentrations of several nutrients (protein, minerals, vitamins) than the growers and the effect of dilution with whole wheat is not so marked.This is clearly shown in the calculated total intakes per bird (Table 5).The CP consumption of the GW birds was 570 g and that of the SW birds 607-608 g.The uptake of methionine was 13.4 g whole wheat, GW = grower + whole wheat, SWP = pelleted starter (80%) + wheat meal (20%) 21 SEM = standard error of mean, * P<0.05, ** P<o.ol, *** P<o.ool, NS = non-sinificant 3) FCR = feed conversion ratio " Means marked with a common letter do not differ at P<0.05.and 15.6 g and that of lysine 31.4 g and 35.6-35.7 g, respectively.Concerning reduced growth rate, the present results are in agreement with earlier studies (Chwalibog and Eggum 1986, Petersen 1990, Elwinger et al. 1992, Jensen 1992).Elwinger et al. (1992) also report- ed lower feed intake and FCR when a pelleted complete diet contained on average 20% whole wheat.The similar results of treatments 2 and 3 suggest that whole wheat feeding can be started at the age of one week.
In Trial 2, the proportion of WG was around 31% of the total feed intake.Even in this trial the feed intake was significantly (P<0.05)re- duced when whole grain was fed with the bal- ancer.The reduction was 5-6%.This resulted in 3.5-4% lower weight gain during the period of 12-37 days.However, only group 4 (BAWB2), which was given the highest proportion of bar- ley (50%) differed significantly (P<0.05) from the control group.In this trial, feed intakes and growth rates were poorer than in Trial 1.The cal- culated supplies of ME, CP and amino acids per bird were also clearly lower in Trial 2 (Table 5).
The BAWG birds received less CP and amino acids than the control birds, which explains the differences in weight gain.FCR was non-significantly lower in the whole grain -fed groups, and feed consumption per kilo slaughter weight decreased by 4% (P<0.05) in groups 2 (BAW) and 4 (BAWB2) compared with the control group.Rose et al. (1995) used in their sequentical feeding study balancer -whole wheat mixtures (loose-mix feeding) as a control.They found that the broilers given the loose-mix feed had growth rates and feed intakes similar to those given complete diets.
in Trial 3, the final body weights and weight BAW = balancer + whole wheat BAWBI = balancer + whole grain (W:B 70:30) BAWB2 = balancer 0 whole grain (W:B 50:50) gains during the growing period (12-38 days) were very similar for all treatments.However, the BAW birds consumed feed less (P<0.05) and more efficiently (P<0.05)than those on the other regimens.Reduced feed intake and lowered FCR of the BAW group was obviously due to the higher dietary energy content, at least compared with the diet containing whole barley.Intakes of ME, CP and amino acids were comparatively similar, which is in agreement with the growth results (Table 5).Feeding whole barley with the balancer did not affect feed intake or FCR compared with the control diet.The good growth results with barley may seem exceptional be- cause barley contains more fibre and less energy than wheat, reducing the growth rate and feed efficiency (Jeroch and Dänicke 1995).The BAWB2 diet contained, however, wheat and barley in the same proportion as the control diet, and older broilers tolerate barley better than young ones.This is explained by an adaptation of the microflora in the gastrointestinal tract to produce entzymes (p-glucanases) which decom- pose antinutritive carbohydrates, p-glucanes.
The proportion of WG was 38% of the total feed consumption.
No significant treatment x sex interactions concerning growth rate, feed intake or FCR were found in the trials.
There were no significant differences in any of the trials between the treatments in mortality during the experimental period.As expected, whole grain either with the complete diets or with the balancers increased significantly (P<0.05) the empty gizzard weight.On average, the absolute weight of the gizzard was 18 g in the birds fed complete diets and 27 g in those fed whole wheat (Table 4).Expressed as a percentage of live weight, the values were 1.0% and 1.65%,respectively.In Trials 2 (Table 6) and 3 (Table 7), the corresponding values for the con- trol birds were 19-20 g or 0.9-1.3% and for those fed whole grain 28-36 g or 1.7-2.1 %.This shows that whole grain kept the gizzard working.This can have a beneficial effect on the bacterial population in the intestine, and Gumming (1992) suggested that an active gizzard may contribute to the development ofresistance against coccid- iosis in the chicken.The strengthened gizzard may have improved the FCR of some whole grain-fed group.Mclntosh et al. (1962) reported that unground wheat yields about 5-10% more energy.This may be due in part to the fact that the finely ground and pelleted complete diets are much faster removed from the gizzard than whole grain, possibly resulting in incomplete absorption of some nutrients.
Using whole wheat with the starter pellets reduced the abdominal fat of female birds from 47 to 38 g or, given as a percentage of live weight, from 2.7% to 2.3% (P<0.05,Table 4).The other treatments did not affect significantly the abdominal fat content compared with the control.The fat content of the thigh and breast was also decreased in the carcasses of the SW birds.The difference from the control was significant (P<0.05) in the thigh and this was ob- BAW = balancer + whole wheat, BAWBI = balancer + whole grain (wheat: barley 70:30), BAWB2 = balancer + whole grain (W:B 50:50)) 2) see Table 4 31 FCR = feed conversion ratio ""h see Table 4 Table 7. Effects of mixtures of a balancer and whole grain in Trial 3.  4 served especially among male broilers.Inversely, feeding the GW diet increased but not significantly the fat content of the thigh and breast.The difference (P<0.05) in fatness between the SW and GW birds can be explained by the differ- ence in protein: energy ratio of the diets.Appar-ently for the same reason, the SW broilers con- tained less fat than the controls.In trials 2 and 3 there was a tendency towards lower abdominal fat in the whole grain-fed broilers but the differ- ences were not significant.Tiller (1989) andRistic et al. (1994) did not find any significant effect on the abdominal fat when using mixtures ofa complete diet or a balancer and whole wheat.In an earlier study (Kiiskinen 1992), the fatness of broilers given mixtures of a concentrate (CP 40-43%) and whole wheat, decreased significantly, but the growth rate, too, was considerably reduced.Litter moisture seemed non-significantly to increase in the whole grain fed groups in Trial 2 (Table 6).This has been reported also in earlier studies (Elwinger et al. 1992, Kiiskinen 1992, Rose et al. 1995).
In conclusion, the pelleted starter diets are more suitable than the grower pellets for mixing with whole wheat.Feeding of whole wheat can be started at the age of one week.A whole grain content of more than 20% of total feed intake will yield a satisfactory weight gain and efficient feed conversion if a suitable pelleted balancer diet is used.Part of the wheat can be replaced with barley.
Feeding regimens in Trial 1.

Table 2 .
Composition of the experimental diets in Trials 2 and 3.
nol according to the method of Maxwell et al.

Table 4 .
Effects of whole wheat feeding in Trial 1.

Table 5 .
Calculated consumption of ME, CP and amino acids during the growing period.

Table 6 .
Effects of mixtures of a balancer and whole grain in Trial 2.