The effects of feeding diets with ensiled maize grain on the production efficiency and meat characteristics of broiler ducks

In the experiment, 180 one-day-old male Cherry Valley broiler ducks were kept for 49 days. The birds were purchased from a commercial duck hatchery (DROB, Tulce, Poland). The birds were divided into three groups of 60 and each group was divided into six replicates. The size of each pen was 1.44 m², and each was made of stainless steel frames and mesh. Wheat straw was used as bedding. On the first day, the temperature in the building measured at the height of the ducklings was 28°C. Then, it was lowered gradually to fall to 20°C in the 4^th week of rearing. During the first four weeks of rearing, the ducks had access to an additional heat source (30–32°C). Other microclimate parameters were consistent with those established by Biesek et al. (6). The ducks had access to water and ad libitum feed throughout the experiment. The birds were provided with nipple drinkers (two per 10 ducks) and feeders (with a length of 7 cm per duck). Feeding was divided into two periods: starter feed was provided from day 1 to day 28 and grower feed from day 29 to day 49. Three feeding groups were reared: the CON group, which received 100% granulated commercial diet ad libitum; the CSMIX group, which received a mixture of 80% granulated commercial diet and 20% ensiled corn grain, administered ad libitum; and the CD+SG group, which received a limited amount of granulated commercial diet along with ensiled corn grain ad libitum. The ensiled corn had a smaller grain size than the commercial diet. The CD+SG group were left a free choice between the two offered feeds. For this purpose, double feeders were used, one of which contained commercial diet, and the other of which contained ensiled corn grain. The proportion reduction in the daily amount of commercial diet was established based on data previously reported by Zaremba et al. (64). In the CD+SG group, only the commercial feed was given for the first three days. Subsequently the following amounts of commercial feed were given per pen (10 birds): 251.75 g (days 4–7), 528.00 g (days 8–14), 1,020.00 g (days 15–21), 1,735.71 g (days 22–28), 2,201.86 g (days 29–35), 2,337.14 g (days 36–42) and 2447.71 g (days 43–49).

Based on the manufacturer’s declaration, the starter feed contained maize, wheat, soybean extraction meal, wheat bran, sunflower extraction meal, hulled sunflower seeds, barley, rapeseed extraction meal, wheat gluten feed, calcium carbonate, animal fat, monocalcium phosphate, vegetable oil and fat (raw sunflower), sodium chloride, and sodium sulphate. The crude protein proportion was 19.30%, that of crude fat was 3.90%, crude fibre was 4.40%, crude ash 5.20%, lysine 0.94%, methionine 0.44%, calcium 0.60%, phosphorus 0.61% and sodium 0.16%. The grower feed contained maize, wheat, wheat bran, soybean extraction meal, sunflower extraction meal, from dehulled sunflower seeds, triticale, rapeseed extraction meal, animal fat, calcium carbonate, monocalcium phosphate, sodium chloride and calcium bicarbonate. Crude protein made up 17.20%, crude fat 3.90%, crude fibre 4.90%, crude ash 4.70%, lysine 0.85%, methionine 0.36%, calcium 0.60%, phosphorus 0.56% and sodium 0.15%.

Corn grain of the Pioneer 8500 variety with a moisture content of 34% was ensiled. Twenty-four hours after harvesting, the grain was ground in Gruber machinery (Gaspoltshofen, Austria). The Agro-Sil Corn ensiling additive (Agrifood, Bojanowo, Poland) was added in a 3 kg mass per 1,000 kg of grain. The additive contained propionic acid, formic acid, lactic acid, ammonium formate, propionate, anti-corrosive substances and a carrier. The prepared mixture was next tipped into plastic bags with a capacity of 1,000 kg and compacted by rapid lifting and lowering, and the bags were tightly sealed after 24 h to limit access to oxygen. Ensiled corn grain was given to the ducklings 32 weeks after the bag was closed.

A POL-EKO dryer (Wodzisław Śląski, Poland) was used to calculate the dry matter content with the technique stipulated in the Polish Animal feeding stuffs - determination of moisture and other volatile substances standard PN-ISO 6496:2002 (35). The analysis was carried out using the weight method, similarly to the analysis of crude ash content (10). The crude fat content was assessed using a Soxtec System HT 1043 extraction apparatus (FOSS Tecator, Hilleroed, Denmark), in observance of the Polish Animal feeding stuffs - determination of fat content standard PN-ISO 6492:2005 (37). The Kjeldahl method determined the crude protein content using the Kjeltec 8400 analyser and the Kjeltec 8420 sampler (FOSS). This analysis followed the method in the PN-EN ISO 20483:2014-02 standard, considering a factor of 6.25 (43). The content of acidic detergent fibre (ADF) was assessed by weight, including residual ash and acid-detergent lignin (ADL) (40) and neutral detergent fibre (NDF) (39) using an ANKOM 220 fibre analyser (Ankom, Macedon, NY, USA). Gross energy was determined, using a KL-21 PLUS isoparabolic calorimeter (Precyzja-Bit, Bydgoszcz, Poland) and adhering to the Polish Animal feeding stuffs, animal products, and faeces and urine - determination of gross energy – bomb calorimetric method standard PN-EN ISO 9831:2005 (38). The above analyses were carried out for commercial diets, commercial diet mixtures with ensiled corn grain (80:20 ratio) and ensiled corn grain, and for both rearing periods. Samples of starter feed were taken on day 1 and samples of grower feed were collected on day 29 (grower feed) – three samples of each type of feed/silage in two replicates. Gross energy was calculated in MJ/kg of dry matter and other nutrient values in kg of dry matter. The chemical composition of feed and ensiled corn grain is shown in Table 1.

For the determination of vitamins E (α-tocopherol and β+γ tocopherol) and A (retinol), samples were prepared according to Polish standards (36, 42) with modifications. The duck meat and feed samples were homogenised and saponified. The next step was extraction using a mixture of n-hexane and ethyl acetate. After separating layers, the supernatant was collected into chromatographic vials and evaporated to dryness under nitrogen. The residue was reconstituted with 1 mL of solvent and filtered through a 0.45-μm PTFE filter disc into two chromatographic vials for simultaneous analysis by liquid chromatography with UV and fluorescence detector (UFLC HPLC System; Shimadzu, Canby, OR, USA). Peak areas were manually integrated with LCsolution v. 1.25 (Shimadzu). Vitamin content was estimated using a standard curve obtained from various concentrations of standard solutions. The total vitamin content in the samples was calculated as the amount of vitamins in mass units of the sample.

Lipid extraction from feed and meat was carried out according to Folch et al. (14), with adaptations. Lipids were extracted from samples using chloroform and methanol (2:1, v/v). The mixture was homogenised, ultrasonicated and centrifuged. The chloroform layer was removed under a nitrogen stream. The obtained residue was transesterified and centrifuged, and the extract was collected. Finally, hexane with the fatty acid methyl esters was subjected to chromatography analysis using a 7890B gas chromatography system (Agilent Technologies, Santa Clara, CA, USA) equipped with a flame ionisation detector, autosampler and split/splitless injector. The retention times of fatty acid methyl ester (FAME) compounds for meat and feed samples were identified by comparison with a standard (FAME Mix-37; Sigma-Aldrich, St. Louis, MO, USA). Peak areas were manual integrated with Agilent ChemStation F.01.00.1903 (Agilent Technologies). The results were presented as a percentage of the total area of the chromatograms using the correction factors of the flame ionisation detector (5). The fatty acid profile and vitamin content in diets are presented in Table 2.

Birds were weighed on days 1, 28 and 49 (Radwag, Radom, Poland) with an accuracy of ±0.01 g. The amount of feed and leftovers was controlled. Feed intake (FI), body weight gain (BWG), growth rate (GR), feed conversion ratio (FCR) and European broiler index (EBI) were calculated. Bird deaths were recorded in each group, and viability (%) was calculated. The formulas for the parameters mentioned above are as follows: BWG = final body weight (g)−initial body weight (g)GR= final body weight (g)−initial body weight (g)0.5×( initial body weight (g)+final body weight (g))×100%FCR=FI( kg)BWG( kg)EBI=viability (%)×average daily gain(chick¯gday)FCR(kg feed kg gain )×10\matrix{ {{\rm{BWG }} = {\rm{ final body weight }}({\rm{g}}) - {\rm{initial body weight }}({\rm{g}})} \hfill \cr {{\rm{GR}} = {{{\rm{ final body weight }}({\rm{g}}) - {\rm{initial body weight }}({\rm{g}})} \over {0.5 \times ({\rm{ initial body weight }}({\rm{g}}) + {\rm{final body weight }}({\rm{g}}))}} \times 100\% } \hfill \cr {{\rm{FCR}} = {{{\rm{FI}}({\rm{kg}})} \over {{\rm{BWG}}({\rm{kg}})}}} \hfill \cr {{\rm{EBI}} = {{{\rm{viability }}(\% ) \times {\rm{average daily gain}}\left( {{{\mathop {\overline {chick} }\limits^g } \over {day}}} \right)} \over {{\mathop{\rm FCR}\nolimits} \left( {{{{\rm{kg feed }}} \over {{\rm{kg gain }}}}} \right) \times 10}}} \hfill \cr }

On the last day of rearing, 12 selected males from each group were weighed, and 2 from each pen with body weight close to the average of their penmates were selected. They were sacrificed by stunning with an electrical device following the recommendations (13) and then decapitation. The birds were plucked with a mechanical plucker (Soda Plus, Krępice, Poland). Then, the carcasses were placed in melted wax approved for contact with food for 3 s (Polwax, Jasło, Poland) and cooled in cold water. Wax and feather remains were removed, and the carcasses were eviscerated. The heart, liver and gizzard were collected. Carcasses and offal were cooled at 4°C in a refrigerator (Hendi, Poznań, Poland) for 24 h. Before dissections were performed, the carcasses and offal were weighed. The following were dissected and separated: neck without skin, pectoral muscles (m. pectoralis major and minor), leg muscles (drumsticks and thighs without bones), skin with subcutaneous fat (including neck skin), abdominal fat, wings with skin and carcass remains (trunk and leg bones) (65).

The digesta viscosity was analysed based on a modified method presented in the study by Pestana et al. (34). After the carcasses had been eviscerated, six samples of jejunal and ileal digesta (approximately 7.5 mL each) were taken from each group into 15 mL Falcon tubes. The samples were separated in a laboratory centrifuge (Eppendorf, Hamburg, Germany) at 4,200 rpm for 10 min. The supernatant (1.5 mL) was collected with an automatic pipette (Chemland, Kraków, Poland) and placed in Eppendorf tubes of the same capacity. A Brookfield DVNext viscometer (Middleboro, MA, USA) was used to analyse the digesta viscosity. The supernatant (0.5 mL) was poured onto a plate, and viscosity was measured at a speed of 6 rpm. The final result was the average of all the results obtained during 30 s of measurement. Each measurement was repeated giving 12 results per group. Results were recorded in centipoise (cP).

The pH of the pectoral muscle (major) was measured using a pH meter equipped with a dagger electrode (Elmetron, Zabrze, Poland). The device was calibrated before it gave measurements using standard buffers with pH 4.00, 7.00 and 9.00. The colour of the pectoral and leg muscles was measured from their inner side. A colorimeter was used (Konica Minolta, Tokyo, Japan) with a three-point colour scale: L*–lightness, a* – redness, and b*–yellowness (11). The right pectoral muscle was weighed (M1) and placed in a cut string bag, which was itself placed inside a larger string bag. After 24 h at a temperature of 4°C, the samples were weighed again (M2), and the drip loss was determined. The water holding capacity of the pectoral and leg muscles was assessed after they were comminuted with a meat grinder (Hendi, Poznań, Poland). Samples weighing 0.300 ± 0.005 g (in M1) were measured. The muscles were placed between two sheets of Whatman tissue paper (4 × 4 cm) and compressed with a 2 kg weight for 5 min. The weight was then removed, the samples were removed, and the samples were weighed again (making M2), according to Grau and Hamm (18). Chemical composition analysis was performed in the pectoral and leg muscles to determine the content of protein, collagen, salt, intramuscular fat (IMF) and water. This was performed on three samples weighing approximately 500 g from each group in three replicates). Near-infrared transmission spectrophotometry was used with a FoodScan instrument (FOSS), working to the Polish Standard for Meat and meat products - determination of fat, protein, and water content - Near Infrared Transmission Spectrometry (NIT) using Artificial Neural Network calibration PN-A-82109:2010 standard (41).

The mean values, SEM, homogeneity of the sample and normal distribution were analysed. A unidirectional ANOVA was conducted to evaluate the data, and Tukey’s test was used to identify statistically significant differences between groups at a significance level of P-value < 0.05. The data were analysed using Statistica software v. 13.3 (TIBCO, Palo Alto, CA, USA). The statistical model was Ya = μ + Da + ea, where Ya was the dependent variable, μ was the overall mean, Da was the effect of nutrition (a = CON, CSMIX or CD+SG) and ea was residual error.

Significantly lower body weight on the 28^th day was noted in the CD+SG group compared to the CON and CSMIX groups (P-value < 0.001). On the 49^th day of rearing, significantly higher body weight was measured in the CSMIX group than in the CD+SG group, while the CON group did not differ significantly from either group. The lowest GR and BWG (in days 1–28) were found in the CD+SG group and this deficit was significant (P-value < 0.001). In the second feeding period, the values of GR and BWG were significantly higher in the CD+SG group than in the other groups (P-value < 0.001). Birds in the CSMIX group gained significantly more weight for the entire rearing period than birds in the CD+SG group (P-value = 0.047). The CON and CD+SG groups had significantly lower feed intake throughout the rearing period than the CSMIX group (P-value = 0.023). There was a significantly higher FCR throughout the entire rearing period in both experimental groups than in the control group (P-value = 0.048). A significantly higher EBI value was returned for ducks in the CSMIX group compared to those in the CD+SG group (P-value = 0.004) (Table 3).

Ducks in the CD+SG group had significantly higher jejunal viscosity than ducks in the CSMIX group (P-value = 0.005). It was noted that ducks in the CON and CSMIX groups had significantly lower ileal viscosity than their counterparts in the CD+SG group (P-value = 0.003) (Table 4).

Significantly higher slaughter yield with offal was found in the CSMIX group compared to the CON group, while the CD+SG group did not differ significantly from either group (P-value = 0.036). The highest liver weight occurred in the CD+SG group, and this surplus was significant (P-value = 0.005). Significantly higher pectoral muscle weight (P-value = 0.031) and total muscle weight (P-value = 0.029) were found in the CON group than in the CD+SG group. Skin weight with subcutaneous fat (P-value = 0.030) and total fatness (P-value = 0.046) were significantly higher in the CD+SG group compared to the CON group. In contrast, the CSMIX group did not differ significantly from the other groups (Table 5).

A significantly higher protein content was measured in the pectoral muscles of control group ducks than in these muscles of experimental group ducks (P-value = 0.001). In the pectoral muscles of the CON and CD+SG groups, significantly higher salt content (P-value < 0.001) was found. In CD+SG group birds’ pectoral muscles, significantly greater IMF content (P-value < 0.001) was present, and in the CSMIX group birds’ pectoral muscles significantly higher water content (P-value < 0.001) was observed. The leg muscles of ducks from the experimental groups were characterised by a significantly lower protein content (P-value = 0.033). More collagen was contained by leg muscles in the CON group than by these muscles in the CSMIX group, and this surplus was significant (P-value = 0.003) (Table 6).

In the pectoral muscles of ducks from the CD+SG group, statistically significantly different and the highest contents of C10:0 (P-value = 0.002), C12:0 (P-value < 0.001), C14:0 (P-value = 0.008), C14:1 (P-value = 0.032) and C15:0 (P-value = 0.015) were demonstrated. Statistically significantly lower contents of C:18:1n-9t were found in CON and CSMIX than in CD+SG (P-value = 0.003). Compared to the other groups, the control group had higher contents of C18:2n9t (P-value = 0.003) and C18:2n6c (P-value = 0.001). In the pectoral muscles of CON and CSMIX ducks, there were significantly higher contents of C20:2 (P-value < 0.001), C20:3n3 (P-value = 0.001) and C22:2 (P-value < 0.001). This tissue from the CD+SG group contained significantly more monounsaturated fatty acids (MUFA) (P-value = 0.007) and trans fatty acids (TFA) (P-value = 0.003). In the CON and CSMIX groups, a higher content of polyunsaturated fatty acids (PUFA) was noted (P-value < 0.001). Duck muscles from the CON and CSMIX groups were characterised by a significantly higher content of α-tocopherol (P-value = 0.039) and those from CON by content of β+γ tocopherol (P-value = 0.002) compared to the other groups (Table 7).

In the leg muscles of ducks from the CD+SG group, significantly higher contents of C10:0 (P-value < 0.001), C12:0 (P-value < 0.001), C14:0 (P-value < 0.001), C14:1 (P-value < 0.001), C15:0 (P-value < 0.001), C16:0 (P-value < 0.001) and C17:1 (P-value = 0.005) were found compared to these fatty acid contents in the leg muscles of ducks from the CON and CSMIX groups. Similar results were obtained for C20:5n-3 (P-value < 0.001), C22:2 (P-value < 0.001) and C24:0 (P-value =0.001). The lowest content of C18:2n-6t was detected in the CSMIX group (P-value = 0.005), and the lowest contents of C18:2n-6c (P-value < 0.001), C18:3n-3 (P-value < 0.001) and C20:2 (P-value = 0.008) in the CD+SG group. This tissue from the CON and CSMIX groups was characterised by a significantly lower saturated fatty acid (SFA) content than the same tissue from the CD+SG groups (P-value < 0.001). The opposite situation occurred in the case of PUFA (P-value < 0.001). In the leg muscles of the control group, there was a significantly higher content of β+γ tocopherol than in those of the 80:20-proportioned feed group (P-value = 0.049) (Table 8).

The costs of feeding ducks in the first rearing period in the CD+SG group were significantly lower than in the other groups (P-value < 0.001). In the second feeding period and throughout the whole rearing period, feeding costs were significantly lower in both experimental groups compared to the control group (P-value < 0.001). (Table 9).