Inclusion of NSP-hydrolysing enzymes in diets for broiler chicks containing increasing contents of distillers dried grains with solubles (DDGS) / Einsatz von NSP-spaltenden Enzymen in Futterrationen für Broiler mit unterschiedlichen Trockenschlempegehalten (DDGS)

In total, 360 one-day-old Ross 308 chickens were purchased from a commercial hatchery (Helmut Wohlin, Klagenfurt, Austria). Considering initial body weight (BW), one-day-old chicks were distributed equally among 24 pens of 3 m² and six dietary treatments (four replicates, 15 birds per pen). Animals were housed in the poultry trial station of the institute (Wimitz, Austria). Each of the 24 pens was equipped with wood shavings as litter material, an automatic waterer, infrared warming lamp, and size-adjusted feed troughs. Mash feed and water were offered ad libitum.Nutrient contents of diets met or exceeded the recommendations of the society of nutrition physiology (GfE 1999). The ambient temperature was gradually decreased from 29 °C at the beginning of the study to 20 °C at the end of the experiment.

The trial was divided into starter (day 1–14), grower (day 14–28), and finisher (day 28–36) period. Starter (12.4 AME_N kg⁻¹; 22.0% CP), grower (12.8 AME_N kg⁻¹; 21.0% CP), and finisher (12.7 AME_N kg⁻¹; 20.0% CP) diets were mainly based on corn and soybean meal. Experimental diets (Table 1) were formulated to contain rising amounts of DDGS at the expense of corn and soybean meal. Thus, DDGS originating from wheat and corn was included into the diet at 8%, 16%, and 24% levels. Wheat-corn-DDGS (wheat-to-corn ratio: 44:56) was obtained from an Austrian bioethanol plant (AGRANA, Pischelsdorf, Austria). Analyzed composition of the DDGS (as-fed basis) is given in Table 2. Digestibility values for DDGS were taken from Kluth and Rodehutscord (2010). To improve ethanol extraction, hydrolyzing enzymes of the categories endopeptidases (IUB number 3.4.23.18), cellulases (IUB number 3.2.1.4), β-glucanases (IUB number 3.2.1.6), α-amylases (IUB number 3.2.1.1), and gluco-amylases (IUB number 3.2.1.3) were added in the production process of the applied DDGS. Aliquots of the three DDGS basal diets (8%, 16%, and 24% DDGS) were supplemented with an NSP-hydrolyzing enzyme (Roxazyme G2G, DSM Nutritional Products Ltd, Basel, Switzerland) using a dose of 200 ppm. The enzyme dose was carefully blended with a portion of the diet before being mixed with the entire feed mix. The NSP-hydrolyzing product is stated by the manufacturer as providing major activities of cellulase, β-glucanase, and xylanase. NSP-hydrolyzing enzyme was added on top without taking the potential energy contribution into account. Titanium dioxide (0.25%) was added to all finisher diets as an indigestible marker. Experimental diets within each growth phase were formulated to contain similar concentrations of total AME_N, digestible (d) lysine (1.16% starter, 1.05% grower, and 0.86% finisher), d methionine (0.44% starter, 0.42% grower, and 0.37% finisher), and d threonine (0.73% starter, 0.68% grower, and 0.62% finisher). To achieve isocaloric conditions, the fat content in the diets was increased linearly with rising DDGS concentration, applying a blend of vegetable oils (Unifrutol®, Garant, Pöchlarn, Austria).

Feed samples were obtained from each batch of feed, vacuum packed and stored at −20 °C until analysis. All diets were analyzed for dry matter (DM), ash, CP, ether extracts (EE), starch, sugar, neutral detergent fiber (NDF), and acid detergent fiber (ADF) according to standard procedures (Naumann and Bassler, 2012). AA composition was analyzed by applying the methods of Altmann (1992). P was determined photometrically (U5100-Spectrophotometer-Hitatchi, Metrohm, Vienna, Austria) using the vanadomolybdate method to measure color intensity at 436 nm after wet ashing of feed samples in HNO₃ and H₂O₂ via microwave (MLS-ETHOS plus Terminal 320, Leutkirch, Germany). Ca and Na were determined by flame atomic absorption spectrophotometry (AAnalyst 200, Perkin Elmer, Brunn am Gebirge, Austria). The concentration of TiO₂ was determined photometrically following sulfuric acid digestion, as described by Jagger et al. (1992). An adiabatic oxygen bomb calorimeter (IKA-C400, Staufen, Germany) was used to determine the gross energy (GE). Broiler chickens’ BW and feed consumption were evaluated per pen on days 1, 14, 28, and 36. Average daily gain (ADG), average daily feed intake (ADFI), and feed conversion ratio (FCR) were calculated for the starter, grower, finisher, and the whole fattening period.

On days 32 and 33 of fattening, clean excreta (free from feathers and feed) were collected twice daily from steel liners placed in the collection trays underneath each pen of birds. The excreta from the four collections from each cage were pooled and then frozen for storage. Nitrogen and DM content (Naumann and Bassler, 2012) was determined in thawed digesta. Before the analysis of GE (Naumann and Bassler, 2012) and TiO₂ (Jagger et al., 1992), the samples were freeze dried (TELSTAR LyoBeta 15, Terrassa, Spain), followed by grinding.

After analysis, the energy lost in excreta was determined and the AME_N was calculated as follows: AME_N (kJ kg ^{− 1} of diet) = GE_diet − [GE_excreta × (marker_diet /marker_excreta ) − 34.42 × (N_diet − N_excreta × (marker_diet /marker_excreta ))] where GE is the gross energy [kJ kg⁻¹ of sample (diet, excreta)] and marker is the concentration of TiO₂ (Hill and Anderson, 1958;GfE, 1999).

At the end of the experiment, broilers were stunned and killed by bleeding in the slaughter house of the poultry trial station. At the day of slaughter and the subsequent day, dressing, eviscerated carcass (weight of the slaughtered animals without blood, feathers, uropygial gland, viscera, abdominal fat, and giblets), chilled carcass (weight of eviscerated carcass after 16-h storage in a cooling chamber at +3 ° C), carcass for grilling (weight of chilled carcass without head and neck and legs at the hock joints), and giblets (weight of empty gizzard, liver without gall bladder and heart immediately after slaughter) were assessed.

From 72 representing broilers (6 male and 6 female per treatment), whole bone-less breast meat samples were removed 24-h post mortem and stored at −20 °C until analyses.

For the sensory analysis of chicken breast meat, a consumer-based sensory panel of six testing persons experienced in sensory evaluation was used. Chicken breasts were thawed to 2 °C for 24 h before sensory testing. Untreated 1-cm thick breast meat pieces were grilled well-done for 4 min on each side in individual dishes at maximum power (Raclette Gourmet Grill Deluxe, Tefal S.A.S., Rumilly, France). Water and unsalted crackers were provided, and panelists were asked to expectorate and rinse their mouths between each sample. Each panelist was asked to evaluate six coded chicken breast samples per session (i.e., 12 sessions in total, performed on 2 days), one sample from each treatment, for tenderness, juiciness, and flavor, using a six-point hedonic scale, in which 1 = very tough/very dry/dislike extremely and 6 = very tender/very juicy/like extremely.

Chemical analyses in meat samples: The remaining raw breast meat samples without skin were freeze dried to determine the DM content. Subsequently, lyophilized meat samples were homogenized and the amount of CP and EE was analyzed by applying standard methods (Naumann and Bassler, 2012). For the determination of fatty acid profile, the one-step methylation method by Sukhija and Palmquist (1988) was applied. Briefly, 0.5 g of freeze-dried meat samples was extracted and converted to methyl esters for 2 h at 70 °C with toluene and 5% fresh methanolic HCl using nonadecanoic acid (C19:0, Sigma Aldrich, Munich, Germany) as internal standard. Subsequently, 5 ml of 6% K₂CO₃ was added followed by another 2 ml of toluene. After centrifugation, 1 μl of the organic layer (split 50:1) was analyzed on a gas chromatograph (Agilent Technologies 7890A, Waldbronn, Germany) equipped with an automatic sampler 7693, an Agilent HP-88 capillary column (100 m × 0.25 mm of internal diameter and 0.2 μm film thickness), and flame ionization detector (FID). The flow rates of the carrier gases (hydrogen and synthetic air) were 35 and 350 ml min⁻¹, respectively. Conditions: The injection port temperature and the detection temperature were 250 °C. Oven temperature was ramped to 170 °C for 1 min followed by an increase to 200 °C at 2 °C min⁻¹. After 1 min at 200 °C, temperature was increased by 8 °C min⁻¹ to 230 °C, which was held for another 10 min. Commercial standard fatty acid methyl esters (FAME) mixtures (Supelco 37 Component FAME Mix, Supelco, Pennsylvania, USA) were used for the identification of individual fatty acids.

All experimental data were statistically analyzed by the general linear model (GLM) procedure of SAS 9.1.3 (SAS, Inst., Inc., Cary, NC, USA) by applying the following model: y_ijk = μ + α_i + β_j + α*β_ij + e_ijk (y_ijk is the dependent variable, μ is the overall mean, α_i the effect of DDGS concentration (i = 1, 2, 3), β_j is the effect of NSP-hydrolyzing enzyme supplementation (j = 1, 2), α *β_ij is the effect of interaction of DDGS concentration and NSP-hydrolyzing enzyme supplementation, e_ijk is the residual experimental error). Treatment effects were determined by analysis of variance (ANOVA) using the Tukey–Kramer test in case of a significant interaction. For performance, energy parameters, and digestibility, the pen served as experimental unit, whereas for carcass characteristics and meat quality parameters, the individual animal was the experimental unit. Tables 4–9 present the least squares (LS)-mean (except for organoleptic parameters, where means were used) values of the different dietary treatments and the pooled standard error of means (SEM). For the statistical evaluation of the organoleptic tests, the Wilcoxon range test was applied. Significant differences among means (p < 0.05) are indicated by different superscripts; trends are reported at p < 0.1. Pearson correlation coefficient was determined for selected parameters.

Analyzed nutrient concentrations of starter, grower, and finisher diets are presented in Table 2. The results of fatty acid profile of the diets as well as the applied pure DDGS and vegetable oil are presented in Table 3. On an average, the EE content and accordingly the absolute amount of fatty acids increased by 9% and 30%, with the dietary rise in DDGS to 16% and 24% (and concomitant oil supplementation), respectively, compared to diet DDGS 8%. Values were within the expected range.

Unlike the results of Heincinger et al. (2012), in which the relative amount of dietary linoleic acid increased by 5.1% when increasing DDGS content by 15% (replacing soybean-meal and wheat), increases from 8% DDGS to 24% DDGS in our study only resulted in a relative increase of 0.9% in finisher diet (data not shown).

The performance results of broilers during the experimental period are shown in Table 4. Starting with homogeneous distribution of BWs, BW, ADG, and ADFI did not differ among treatment groups (p > 0.05) in the starter period. Diets containing 16% DDGS showed an improved FCR compared to diets including 8% DDGS (−4.4%) (p < 0.05). In the grower period, enzyme supplementation improved ADG about 3.1% (p < 0.1) and FCR about 3.8% (p < 0.05). However, increasing DDGS content showed no effect on performance in the grower phase. In the finisher period, a high DDGS supplementation increased the FCR up to 4.0% compared to the diet with the lowest DDGS addition (p < 0.05). The NSP-Enzyme increased BW about 2.5% (p < 0.05), ADG about 4.8% (p < 0.05), and ADFI about 2.6% (p < 0.1) and decreased FCR about 1.5% (p < 0.1). For ADG and FCR, significant interactions between DDGS and enzyme supplementation were observed (p < 0.05), indicating a more pronounced enzyme effect in the low-DDGS diets. Over the whole fattening period, performance was unaffected by DDGS and enzyme supplementation.

Our results show that an increasing DDGS content had no effect on ADG, ADFI, and FCR during the whole fattening period, provided that a sufficient supply with digestible amino acids (dAA) to cover requirements is warranted. Other studies based on the similar study design confirm the present observations (Oryschak et al., 2010). Nevertheless, other authors reported decreases in ADG and ADFI by increasing DDGS content (Lumpkins et al., 2004, Schedle et al., 2010a). A possible reason for the decline in performance of those studies might be the formulation of AA on basis of total AA without consideration of their digestibility (Whitney and Shurson, 2004).

Raising amounts of DDGS by expense of corn and soybean meal enhances the NSP content in feed (Widmer et al., 2008;Schedle et al., 2010b). Supplementation of NSP-degrading enzymes to diets containing high amounts of NSP or rather fiber is of interest to increase the potential for use of fiber-rich industrial by-product such as DDGS in the diet of monogastric animals (Zijlstra et al., 2010). In the present study, enzyme supplementation led to an enhanced ADG in the finisher phase and final BW. Additionally, the NSP-hydrolyzing enzyme complex improved FCR in the grower (p < 0.05) and finisher phases (p < 0.1). It is well established that NSP-hydrolyzing enzymes can improve FCR in fiber-rich diets as reviewed by Slominsky (2011) and Woyengo and Nyachoti (2011). However, especially young broilers seem to tolerate or even require moderate amounts of fiber for gastrointestinal tract development (Jiménez-Moreno et al., 2009). This may be the reason for the lack of performance improvements in the starter phase with NSP-hydrolyzing enzyme supplementation. Interestingly, an interaction between DDGS and NSP-hydrolyzing enzymes was observed in the finisher phase as well as for heart and liver weight, as described below.

No differences between treatments were detected regarding the AME_N content of diets (Table 5). The energy content reached 11.5 MJ AME_N kg⁻¹, on an average. Nevertheless, these values are lower than the calculated energy contents. Similar effects were observed for the parameters energy intake per day and energy/gain. Neither DDGS nor enzyme supplementation affected N retention and total tract apparent retention of DM.

The replacement of soybean meal and corn by dietary DDGS inclusion demands an increase in vegetable oil addition in the broiler diet to keep the energy level constant (Thacker and Widyaratne, 2007;Schedle et al., 2010a). It is well known that increasing oil levels, altered oil/fat composition of diets, as well as fiber content can affect nutrient digestibility in broilers. This may result from alteration in transit time, mucosal architecture, and activity of digestive enzymes, especially proteases (Ketels and DeGrote, 1989;Jiménez-Moreno et al., 2009;Honda et al., 2010). However, AME_N content, DM digestibility, and apparent N retention did not differ between treatments. Similar observations were made by Crespo and Esteve-Garcia (2002). Other authors reported a decrease in protein digestibility by increasing oil content (Leytem et al., 2008;Honda et al., 2010). The dietary fatty acid profile has a strong impact on energy, nitrogen, and fatty acid deposition in broiler chickens (Crespo and Esteve-Garcia, 2002). Hence, both the increase in dietary oil and the concomitant raise in unsaturation of dietary fatty acids might be a reason for the unaltered apparent N retention in the finisher phase of our study.

Furthermore, Oryschak et al. (2010) found interactions between triticale DDGS and NSP-hydrolyzing enzymes in broiler diets. In their study, apparent ileal digestibility of CP and several AA increased with enzyme supplementation in diets containing 15% DDGS but not in diets with 30% DDGS inclusion. The difference between the treatments with and without enzyme in the low DDGS diet (8%) in our study might indicate an improved effect of the enzyme on the non-DDGS diet components in our study. The inclusion of similar NSP-degrading enzymes in the production process of the applied DDGS product may be an explanation for the absence of distinct effects of the NSP-hydrolyzing enzyme supplemented in the diet on performance in high DDGS groups.

Carcass characteristics are shown in Table 6. Diets containing 16% DDGS had significantly higher dressing percentage compared to treatments with 8% DDGS (p < 0.05). Gizzard weight was significantly higher in the treatment with the highest DDGS concentration compared to the low DDGS diet (p < 0.05). It is well described in literature that the inclusion of insoluble fiber sources in the diet increases the retention time of the digesta in the upper part of the digestive tract which in turn might stimulate gizzard development (González-Alvarado et al., 2010). Increasing fiber content or 24% DDGS in diet reduced remainder of carcass percentage compared to treatment including 8% DDGS (p < 0.05). The trend of heavier BW affected by enzyme supplementation was also observed for carcass bodies (p < 0.1). Furthermore, livers of enzyme-supplemented broilers were significantly heavier than those of non-supplemented animals (p < 0.05). However, heart and liver weights showed a significant interaction between DDGS and enzyme (p < 0.05). Thus, enzyme supplementation enhanced weight of liver and heart predominantly in the low DDGS group (+5.8%; p < 0.05). González-Alvarado et al. (2008) and Gracia et al. (2009) observed higher relative liver weight when broiler were fed heattreated cereals. The authors hypothesized that this increase may either result from a higher digesta viscosity or improved availability of dietary fat, which in turn may stimulate the liver in secreting biliary juices. Hence, the improved FCR and, therefore, also the energy per kilogram demand in the finisher phase can be associated with the liver weights. This suggestion is further supported by the strong negative correlation between the FCR and energy per kilogram, respectively, and the liver weight (correlation coefficient: −0.72; −0.65; p < 0.0001) in the finisher phase.

Increasing DDGS and vegetable oil contents had no influence on crude nutrient contents of raw breast meat (Table 7). However, enzyme addition resulted in a significant reduction of EE and a concomitant increase in CP (% DM). Thus, enzyme addition resulted in less intramuscular fat content in breast meat, indicating a potentially preferential utilization of dietary energy for protein retention. However, while apparent N retention was only numerically increased with enzyme addition, CP (%DM) content in breast meat was significantly higher (p < 0.05). Based on these results, the improved FCR affected by enzyme supplementation without effect on digestibility likewise seems to be due to intermediate metabolism processes. As described by Fernandez et al. (1999), fat is responsible for organoleptic properties of meat and has a distinct effect on juiciness (Wood et al., 2003). However, we could not show that lower intramuscular fat content correlates with a decrease in the sensation of juiciness (correlation coefficient 0.14, p > 0.1).

In this respect, sensory analyses showed a trend of high DDGS diets to increase meat tenderness (p < 0.1, Table 8), whereas the enzyme addition led to lower juiciness of breast meat samples (p < 0.1). In general, mean organoleptic determination of flavor in meat samples resulted in a mean value of 3.44 without the influence of experimental factors, DDGS or enzymes.

The influence of DDGS and vegetable oil on fatty acid composition (Table 9) was primarily seen in a significant increase in the amount of polyunsaturated fatty acids (PUFAs) in a linear manner. Furthermore, the significant increase in PUFA-to-saturated fatty acid (SFA) ratio shows that this increase goes along with a significant decrease in SFA (p < 0.05). The relative amount of monounsaturated fatty acid (MUFA) was significantly reduced by increasing DDGS and vegetable oil in broiler diets. In detail, increases resulted from a significant increase in total n6-fatty acids with only minor influence on total n3-fatty acids (p < 0.1). Thus, especially, linoleic (C18:2 n6) acid was significantly increased in high DDGS diets. Consequently, the n6-to-n3 ratio augmented as well by 11.3% because of the DDGS and vegetable oil increase in diet (p < 0.0001). The reduction in MUFA concentration derives primarily from a significant reduction of oleic acid (C18:1 n9) in the 24% DDGS fed broilers. Like with DDGS increase, a significant decrease in MUFAs and an increase in PUFAs (n6 but not n3) (p < 0.05) were also observed due to the addition of NSP-hydrolyzing enzyme. However, as the percentage of SFA remained uninfluenced, no significant changes in PUFA-to-SFA ratio could be detected for enzyme supplementation (p > 0.1). Even though several significant changes because of DDGS or enzyme addition were observed for fatty acid profile, no significant interaction between the factors occurred.

Unlike in our findings, where high DDGS diets affected a more tender breast meat (p < 0.1), Schilling et al. (2010) showed increased shear force values for broiler breast meat fed up to 24% DDGS. Higher PUFA concentrations in fat go in line with increased softness of the fat in meat (Wood et al. 2003); however, the amount of PUFA as well as total fat content in breast meat may have been too low to correlate with the tenderness sensation (correlation coefficient of parameter PUFA and tenderness: 0.13, p > 0.1) in our study. However, sensory analysis revealed that flavor and overall consumer acceptability was uninfluenced by DDGS addition (Schilling et al., 2010). Studies showed that the absolute fat content of breast meat remained uninfluenced by the increase in DDGS (Schilling et al., 2010) or PUFA concentration (Cortinas et al., 2004) in broiler diet. These findings are in accordance with our study.

Former trials showed that dietary PUFAs—in contrast to SFA or MUFA—decrease abdominal fat in broilers (Crespo and Esteve-Garcia, 2002). However, absolute increases in dietary PUFA content (+ 10.2% and + 27.7% in 16% DDGS and 24% DDGS compared to 8% DDGS, respectively, Table 3) in the present diets did not affect abdominal fat tissue. This may be caused by the concomitant absolute increase in dietary SFA content of diet in a comparable manner (+ 10.2% and + 26.0%, respectively) resulting in similar dietary PUFA-to-SFA ratios with a mean value of 1.94 ± 0.02. Furthermore, the energy intake as well as the N metabolism was similar between the treatments.

The composition of fatty acids in chicken tissues results from direct deposition of dietary fatty acids and endogenous fat synthesis (Cortinas et al. 2004). A general lipogenesis inhibition by dietary fat addition but a comparably higher endogenous synthesis is reported in broilers fed diets rich in PUFAs (Crespo and Esteve-Garcia, 2002). In our study, an increase in linoleic acid and total PUFA content because of increasing dietary DDGS supply up to 24% (and increased vegetable oil concentrations for isocaloric diets) in breast meat was observed. Schilling et al. (2010) and Heincinger et al. (2012) reported similar findings in broiler thigh and turkey breast, respectively. Furthermore, Corzo et al. (2009) also found increases in eicosadienoic acid (C20:2) with DDGS. However, unlike in the study of Schilling et al. (2010), we did not observe reductions in stearic acid in our breast meat samples. This discrepancy may result from differences in sampling site or origin of DDGS and dietary oil.

Although the PUFA-to-SFA ratio remained quite constant in feed despite increases in DDGS and fat content, only the PUFA family was incorporated in a higher manner in the intramuscular fat content. This is mirrored in a 14.6% increase in PUFA-to-SFA ratio in meat (from 8% DDGS to 24% DDGS, p < 0.0001). Lopez-Ferrer et al. (2001) also showed that PUFAs are more likely changeable via dietary intervention than SFA and MUFA. This may result in part from the preferential inclusion rate of PUFAs in phospholipids (Hulan et al., 1988), of which the concentration is high in breast meat fat (Ratnayake et al., 1989). As reviewed by De Smet et al. (2004), the absolute intramuscular fat level has a major impact on the fatty acid profile, showing higher PUFA contents in leaner meat samples. This was also observed in our study feeding broiler with diets supplemented with NSP-hydrolyzing enzymes. Generally, an increase in PUFAs without concomitant increases in SFA and MUFA is a positive development concerning physiological fat quality. However, as the increase is more pronounced in the n6 fraction, the balance of n6 to n3 may be impaired. Furthermore, increases in intramuscular PUFA concentration pose greater potential for lipid peroxidation (Kouba and Mourot, 2011). Higher TBARS values in 5d-refrigerated thigh meat because of broiler diets with DDGS concentrations higher than 12% were described (Schilling et al., 2010). Thus, future feeding trials should concentrate on increasing antioxidative capacity in meat, for example, with adapted tocopherol concentrations in diets, when increasing DDGS content. In general, it has to be stated that changes in dietary oil source may have a major impact on the fatty acid profile.