Vachellia tortilis leaf meal improves antioxidant activity and colour stability of broiler meat

2.1 Leaf collection, processing, and diet formulation

V. tortilis leaves were hand-harvested during the post-rainy season at an advanced stage of maturity at Makhathini Research Station, Jozini Municipality, South Africa (27°23′42.5″S 32°10′47.7″E). The leaves were air-dried under a shade and sieved to separate leaves from thorns, pods, and twigs. The leaves were ground to pass through a 2 mm sieve using a Retsch GmbH mill (Model ZM200, Haan, Germany) to produce the VTLM. To determine the potency of AA and chemical stability of broiler meat produced from broilers fed on legume leaf meals, six diets were designed and formulated (Table 1) using the WinFeed (2018), feed formulation software (WinFeed Limited, Cambridge, UK). The six dietary treatments contained 0, 30, 60, 90, 120, and 150 g/kg of VTLM and formulated to meet the nutrient requirements for growth, as described earlier (Miya et al., 2019). Vitamins and minerals were added to meet the National Research Council [14] recommended specifications for broilers.

Samples from each diet were taken and ground to pass through a 1 mm sieve to determine the proximate chemical composition of the diets. Ash content was analysed using Method 942.05 according to the Association of Official Analytical Chemists [14]. Nitrogen content was determined by a Dumas Method using a LECO TruSpec N (LECO, Pretoria, South Africa) [15]. Crude protein content was calculated by multiplying the nitrogen content by 6.25. Crude fat content was determined using the BÜCHI 810 Soxhlet extractor (BÜCHI, Flawil, Switzerland) [14]. The neutral detergent and acid detergent fibre contents were analysed using the ANKOM Fibre Analyser (Ankom Macedon, NY, USA), as described by Van Soest et al. [16]. The VTLM contained 944 and 218 g/kg dry matter (DM) and crude protein, respectively. The respective contents of ether extracts and ash were 40.1 and 65 g/kg. The acid and neutral detergent fibre contents were 29.8 and 494 g/kg, respectively. The butanol–HCL method was used for the determination of condensed tannins [17]. Mineral contents were determined using a Varian 720 Inductively Coupled Plasma Atomic Emission Spectroscopy (Spettrometro ICP-AES, Vista MXP Rad Varian). The acid hydrolysis method [15] was used to prepare samples for the identification and quantification of amino acids using an amino acid analyzer (SY-KAM, Erising, Germany).

2.2 Experimental design

The study was conducted at Ukulinga Research Farm, University of KwaZulu-Natal, Pietermaritzburg, South Africa. A total of 180 Cobb-500-day-old chicks were fed on a commercial starter diet and water supplied ad libitum for a 14-day brooding period. The sample size used was based on the low levels of variation within the treatment on the variables tested [1]. After the 14-day brooding period, the chickens were randomly allocated to the six experimental diets in a completely randomized design. Chicks were fed on dietary treatments for 21 days. Each treatment was replicated three times, with 10 birds per replicate. The three replicates per treatment were a guideline to reduce sample size. Wood shavings were used as bedding and were changed weekly. At day old, room temperature was set at 30°C and gradually reduced by 1°C until the final temperature was maintained at 21°C (relative humidity = 35%). Birds were subjected to a 12L:12D lighting program throughout the study.

Birds were slaughtered at 35 days old by electrical stunning using 15 V and 50 Hz of pulsed direct current in a prestunner (model SF-7000, Simmons Engineering Co., Dallas, GA) and killed by exsanguination using a sharp knife. After bleeding for 5 min, carcasses were scalded and automatically picked in a 4-bank picker for 30 s (model D-8, Stork-Gamco Inc.). Carcasses were rinsed with water to remove blood and feathers. The Pectoralis major, and the thigh muscles, Peroneus (Fibularis) longus muscle, were cut off from the carcasses, skinned, deboned, vacuum packed, and chilled at a temperature of 4°C for 24 h. Breast and thigh meat samples were analysed for AA, chemical stability, and physico-chemical properties.

1. Ethical approval: The experimental procedures using animals were performed according to the ethical guidelines specified under the Certification of Authorization to Experiment on Living Animals provided by the University of KwaZulu-Natal Animal Ethics Committee (Reference No: 004/13/Animal).

2.3 Meat sample extraction

Approximately 5 g of meat sample was obtained from each of the vacuum-sealed samples for meat extraction. The samples were homogenized in 50 mL of 50% aqueous methanol (v/v) and placed in a sonication bath for 1 h. Phenolic compounds in the meat were extracted using methanol and ethanol. The extracts were filtered under vacuum using Whatman filter paper No. 1 and concentrated under vacuum using a BÜCHI Rotavapor (R-114, Flawll, Switzerland) at 35°C. The samples were dried at room temperature. The antioxidant activities in meat were measured using three assays: the radical scavenging (DPPH), ferric-reducing antioxidant power (FRAP), and the β-carotene–linoleic acid assays.

2.4 Meat quality characteristics

The methods described by Honikel [22] were used to determine meat pH, colour, water holding capacity, drip loss, and cooking loss (CL). The pH was determined using a pH meter probe at 24 h post-mortem. Three areas of the meat were poked with the probe. A thermometer pointer was pointed into the meat after being cleaned with distilled water and a cotton towel. Temperature was measured at 7 h post-mortem.

The Pectoralis major was cut off from the carcasses, deboned and meat colour (L*, a*, b*), pH, water holding capacity, CL, and drip loss were measured [22]. Fresh breast fillets were evaluated for lightness (L*), redness (a*), and yellowness (b*) colour readings using a portable colorimeter (Konica Minolta Chroma Meters CR-400, New Jersey, USA). Readings were taken on the ventral surface of the breast fillet to prevent any interference of scalding at an angle of 10°, using illuminant D65 with an illumination viewing system as 45°/0° (Commission International de l’Eclairage). The pH was measured in muscle samples in triplicates using a potable Crison pH-meter 25 (Crison, Alella, Barcelona, Spain).

Water holding capacity was determined by measuring water loss in meat when pressure was applied to the breast muscle [22]. Breast fillet cubes weighing 2 g were laid between two circular filter papers, and a 10 kg weight pressure was applied for 5 min. Samples were removed from the filter papers and reweighed. Water loss was calculated as the weight difference between the initial and final weight. Drip loss was measured by weighing steaks of breast meat and initial weights were recorded [22]. A small incision was made on the corner of the steak, suspended in a plastic bag, and placed in a refrigerator at 4°C for 24 h.

Caution was made to ensure that the meat did not touch the sides of the bag at any stage. The meat sample was removed from the bag and gently blotted dry with a paper towel, and final weight was measured. Drip loss was calculated as: % Drip loss = [(initial weight – final weight) ÷ (initial weight)] × 100. To determine CL, 10 g of freshly cut steaks were weighed (initial weight). The steaks were placed in thin-walled plastic bags and boiled at 80°C in a water bath for 60 min. Samples were removed from the water bath and chilled at 4°C until equilibrated. The meat was taken from the bag, gently blotted dry, and weighed (final weight). CL was calculated as: % CL = [(initial weight – final weight) ÷ initial weight] × 100.

Texture of the meat was measured on the raw meat samples. All breasts and thigh muscles were made identical. For breasts, a cylindrical of 14 mm was used to cut similar pieces, and for thighs, a scissor and a ruler were used to make samples of 10 mm in length. The texture of meat samples was determined using the Warner–Bratzler shear force, according to Honikel [22].

2.5 Proximate chemical and mineral composition of broiler meat

After freeze-drying the meat, moisture content was calculated as described by the AOAC Official Method 934.01 [14]. The DM was then calculated from the percentage of moisture content. Ether extract was evaluated using a Soxhlet method on the Soxhlet BÜCHI 810 (BÜCHI, Switzerland). Five grams of meat powdered sample was weighed into a thimble, dried in an oven for an hour, plugged with cotton wool, and placed into a Soxhlet extractor. A Buchi fat beaker with extraction stones was weighed, three-quarter filled with petroleum ether, and placed onto the heating place of the Soxhlet extractor. After 4 h of extraction, a beaker was left overnight on the bench for the residual solvent to evaporate and dried in an oven for an hour.

The weight of the beaker was measured, and the ether extract was calculated as described by the AOAC method 920.39 [14]. Nitrogen was analysed using the LECO TruSpec Nitrogen Analyser. The percentage of nitrogen was multiplied by a conversion factor of 6.25, as described by the AOAC Official Method 990.03 [15]. The weight of a crucible was measured, 1 g of powdered meat sample added and a crucible with a sample was placed in a furnace overnight at 550°C. The weight of a crucible with ash was measured, and the percentage ash was calculated. Organic matter was calculated from the percentage of ash as described by the AOAC Official Method 942.05 [15].

Calcium, sodium, iron, manganese, copper, zinc, and phosphorus were determined from the ash content. Ashed samples were placed into 100 mL conical flasks, and 5 mL of 6 M hydrochloric acid (HCl) was added and boiled to evaporate as described in the AOAC Official Method 984.27 [15]. Five millilitres of 6M nitric acid were added and boiled to dissolve the ash contents. The mixture was then filtered using Whatman’s filter papers no. 1 into 100 mL volumetric flasks and made up to the mark with deionized water. The ICP-AES (Spettrometro ICP-AES, Vista MXP Rad Varian) was used to analyse minerals using respective standards.

2.6 Statistical analyses

The responses of AA, ferric-reducing power, and physicochemical properties of breast meat, growth performance, blood biochemistry, and liver enzymes with VTLM inclusion levels were determined using the response surface regression procedure for Statistical Analysis System (SAS) (2011). Significance levels were considered at P < 0.05. The preliminary general linear models procedure showed that both storage time and VTLM inclusion influenced meat quality characteristics, but no interaction existed between inclusion level and storage period on meat quality characteristics. Water holding capacity, pH, and drip loss did not change over time and across inclusion levels. As such, responses in AA, ferric-reducing power, and physicochemical properties of breast meat to VTLM inclusion and storage time were determined separately using the response surface regression procedure for SAS (2011).

For quadratic relationships, a piecewise regression analysis was done using the non-linear model procedure to determine the threshold time at which the VTLM inclusion, proanthocyanidin content, and storage time caused AA, ferric-reducing power, and physicochemical properties to be constant, increase, or decrease.

The model: Yi = yo + y1 + y2 (Ixc) (xi − xc) + εi, using parameters (yo, y1, y2) and the xc, the two segmented simple regression functions;

where Yj = yo + y1 (xi), for xi ≤ xc and Yk = yo + (y1 + y2) xi, for xi ≥ xc,

where Yi is the response variable when VTLM inclusion, proanthocyanidin content, and storage time are constraining AA, ferric-reducing power, and physicochemical properties;

Yj is the response variable before VTLM inclusion, proanthocyanidin content, and storage time is constraining AA, ferric-reducing power, and physicochemical properties;

Yk is the response variable when VTLM inclusion, proanthocyanidin content, and storage time exceeds the maximal VTLM inclusion, proanthocyanidin content, and storage time;

Yo = yo − y2 xc; when xi = 0; yo is the intercept or minimum Yi when xc < 0; y1 is the rate of change in Yi when xi < xc; y2 is the rate of increase in Yi when xi > xc; xi is the VTLM inclusion, proanthocyanidin content and storage time; xc is the optimum VTLM inclusion, proanthocyanidin content and storage time beyond which AA, ferric-reducing power, and physicochemical properties were increased or reduced by increase in VTLM inclusion, proanthocyanidin content and storage time; and Ixc is a dummy variable with value 0 when xi < xc and 1 when xi ≥ xc. Significance levels were considered at P < 0.05.

3.1 AA of feeds and meat samples

The concentration (mg/mL) of meat extracts was plotted against the absorbance in the ferric-reducing power assay for each treatment (Figure 1). A concentration-dependent response of ferric-reducing power of meat extracts with increasing inclusion levels of VTLM was observed. All inclusion levels of VTLM showed similar patterns of increasing strength of reducing power as the concentration of meat extracts increased. Inclusion levels of 120 g/kg VTLM had the highest reducing power, comparable to that of ascorbic acid.

Figure 1

Ferric-reducing AA of different inclusion levels of VTLM.

A range of AA of between 4.3 and 49% based on the average rate of β-carotene–linoleic acid AA was recorded. The response of β-carotene–linoleic acid AA of diets increased with increasing inclusion level VTLM. The highest AA recorded was 39.96% at an inclusion level of 113.3 g/kg VTLM (Table 2 and Figure 2). There was a quadratic response in β-carotene–linoleic acid assay AA of breast meat with increases in VTLM inclusion levels (Table 2 and Suppl. Figure 3). AA of meat was highest at 40.8% corresponding to an inclusion level of 115 g/kg DM VTLM.

Figure 2

Responses in β-carotene–linoleic acid AA of diets containing increasing levels of VTLMs.

Figure 3

Responses in ferric-reducing power, β-carotene–linoleic acid AA, and DPPH radical scavenging activity of breast meat with increasing inclusion levels of VTLM.

The level of inhibition of the DPPH radical increased linearly in breast meat with increasing inclusion level VTLM using the DPPH radical scavenging activity assay (Table 2 and Suppl. Figure 3). The AA increased by 0.02% per 1 g/kg DM increase in V. tortilis in the diet. Ferric-reducing power of breast meat gradually increased by 0.001 per 1 g/kg DM increase in V. tortilis in the diet. The highest AA obtained for diets and meat samples was between 40 and 40.8% at VTLM inclusion levels of 113–115 g/kg DM.

There were significant quadratic relationships between DPPH radical scavenging activity, β-carotene–linoleic acid assay AA, ferric-reducing power AA of the diets and meat samples with increasing condensed tannin concentration in the diets (Table 2).

3.2 Responses of meat colour, physico-chemical properties of meat, blood metabolites, and liver enzymes to VTLM inclusion

There was a quadratic response in a* and b* values of the breast with increasing VTLM inclusion levels in broiler diets (Table 3). Increasing levels of VTLM up to 125 and 117 g/kg DM in broiler diets increased a* and b* to maximum values of 8.7 and 19.7, respectively, for a storage period lasting 28 days. The calcium content of the breast meat increased by 0.02, while the copper content gradually decreased at a rate of 0.01 per 1 g/kg DM increase in VTLM in broiler diets (Table 3). Mineral meat quality parameters such as phosphorus, potassium, zinc, manganese, magnesium, and iron were not significant for all treatments. The pH values ranged between 5.7 and 6.5 for all treatments. No responses in lightness (L*), shear force, drip loss, CL, proximate chemical composition, trace mineral content, and pH of breast meat with increases in levels of VTLM inclusion of breast meat was observed during the 28-day storage period.

There was a quadratic response in a* value of thigh meat with increasing VTLM inclusion levels in broiler diets (Table 4). Increasing levels of VTLM up to 75 g/kg DM in broiler diets increased a* to a maximum value of 7.7 for a storage period lasting 28 days. There was a linear increase in b* value and CL of thigh meat with an increase in VTLM in broiler diets. There were no responses in lightness (L*), shear force, drip loss, CL, proximate chemical composition, trace mineral content, and pH of breast meat, with increases in levels of VTLM inclusion of breast meat were observed during the 28-day storage period.

There was a quadratic response in alkaline phosphatase (ALP) and albumin concentrations with increases in VTLM inclusion levels (Table 5). The concentration of ALP and albumin was highest at 340 U/L and 2.2, respectively, corresponding to inclusion levels of 92 and 75 g/kg DM VTLM. There were linear increases in concentrations of alanine transaminase (ALT), aspartate aminotransferase (AST), and iron, while the concentration of urea decreased with increasing levels of VTLM in broiler diets.

3.3 Responses of meat colour and physico-chemical properties of meat to storage time

The L* value of breast meat decreased with storage time to a minimum of 45.8, after which an increase in storage time beyond 10.8 days improved lightness (Table 6). The texture of breast meat increased until 20 days of storage after which further storage resulted in breast of lower texture. At the inclusion level of 150 g/kg DM VTLM, L*, a*, and b* values of breast meat decreased and reached minimum values after 9, 10, and 7 days of storage, and further storage beyond these days improved these parameters. Our results show that treatments took 10.8 days to have a positive effect on the lightness of breast meat. The L* value of thigh meat decreased with storage time to a minimum of 40.7, after which an increase in storage time beyond 11.5 days improved lightness (Table 7).

The a* value, however, increased until 14 days of storage after which further storage resulted in deterioration in redness. The pH of thigh meat decreased gradually by 0.009 (30 g/kg VTLM) and 0.01 (120 g/kg VTLM) per day of storage and was lowest at pH 6.07 and 5.96 on day 21 of storage.