The feeding experiment was conducted at the Neudamm Campus of the University of Namibia, which is approximately 30 kilometres east of Windhoek in the Khomas region of Namibia. The campus is situated at 22°30'10.19" S latitude and 17°22'5.39" E longitude. All procedures conducted during this experiment were approved by the Animal Research Ethics Committee (ref: AREC/024/2020) of the University of Namibia.
Feed ingredients and experimental diets
The milled bush biomass of encroacher bush species used in this study was of the predominant encroacher species, namely Senegalia mellifera, Dichrostachys cinerea, T. sericea and R. trichotomum. Harvesting of the bushes was done between April and May 2019 and the harvested biomass was restricted to branches or twigs of ≤ 20 mm stem diameter. The fresh biomass was milled using a hammer mill to a particle size of 10 mm and air dried under shade until constant weight before being packed in bags and transported to Neudamm campus for storage until the feeding trial.
Apart from bush biomass, other roughage sources used in the preparation of experimental diets were grass hay (mixed veld grass species) and Lucerne hay, which were also milled to particle sizes of 10 mm. The experimental diets were formulated to constitute 40% of different roughage sources and similar 60% concentrate made from a combination of different feed ingredients (Table 1). Except for bush biomass, all other ingredients were purchased from a commercial feed supplier.
A conventional diet consisting of coarsely ground Lucerne (10%) and grass hay (30%) as roughage, was used as a control (T1), while each of the other four treatment diets (T2-T5) consisted of the selected bush species as roughage source at the same inclusion rate (40%) and the remaining portion (60%) consisted of the same combination of supplements. Batches of feed were mixed weekly and fed to the lambs. All treatment diets were formulated to meet nutrient requirements of growing sheep according to NRC (2007) recommendations.
Table 1
Ingredients composition of the five treatment diets
| | Treatment diets* | | |
| Feed ingredient (kg “as is”) | T1 | T2 | T3 | T4 | T5 |
| Coarsely ground grass hay | 30 | 0 | 0 | 0 | 0 |
| Coarsely ground lucerne hay | 10 | 0 | 0 | 0 | 0 |
| Milled Bush | 0 | 40 | 40 | 40 | 40 |
| Yellow maize meal | 22 | 20 | 20 | 19 | 20 |
| Molasses syrup | 5 | 5 | 5 | 5 | 5 |
| HPC 30 | 30 | 32 | 32 | 33 | 32 |
| Futterfos™ P14 | 1.5 | 1.5 | 1.5 | 1.5 | 1.5 |
| Coarse salt | 1.5 | 1.5 | 1.5 | 1.5 | 1.5 |
| Total | 100 | 100 | 100 | 100 | 100 |
* T1 control diet, T2 Senegalia mellifera-based diet, T3 Dichrostachys cinerea-based diet, T4 Terminalia sericea-based diet and T5 Rhigozum trichotomum-based diet; HPC 30 high protein concentrate with 30% crude protein; Futtterfos™ P14- Phosphate lick with 14% Phosphorus.
Experimental animals and management
Thirty (30) weaned 3-month old Damara lambs (15 males and 15 females) with a mean weaning weight of 16.7 ± 1.9 kg were used. The lambs were vaccinated with Multivax P™ (Intervet International B.V., The Netherlands) at the beginning of the trial to protect them against botulism, black quarter, pulpy kidney and clostridium. They were also treated with Dectomax™ (Pfizer Laboratories (Pty) Ltd, Sandton) against internal and external parasites.
Experimental design and data collection
The experiment was set up as a completely randomized design (CRD) with six (6) lambs per treatment (3 males and 3 females), to evaluate feed intake (FI) and the growth performance for 90 days, after an adaptation period of 14 days. The 30 lambs were randomly allocated to the 5 dietary treatments and housed individually in pens of 1 m2 with concrete floors in an open-sided roofed shed, where they received treatment diets and water on an ad lib basis.
After the adaptation period, the feed offered and feed refused daily were weighed and recorded to determine the feed intake. The daily feed was offered to each sheep in two portions, at 09h00 and 14h00. Initial body weights were obtained by weighing the lambs using an electronic scale (Micro T7E Scale; Premier Scale Services (Pty) Ltd) at the beginning of the adaptation period. Thereafter, the lambs were weighed weekly on the same day of the week in the morning before feeding, until the end of the experiment. Average daily gain (ADG) was calculated as difference between final and initial body weights divided by number of feeding days.
Sampling and chemical composition analysis of experimental diets
During the trial period, random samples of each treatment diet were taken once from a weekly batch, mixed and pooled separately in marked paper bags. At the end of the experiment period, all treatment diet samples were ground to pass through a 1 mm sieve (Retsch Mable mill; Retsch GmbH) and stored in plastic bottles for chemical analysis. The dry matter (DM) content of bush species samples was determined by drying the samples in a forced draught oven at 100°C for 24 h (AOAC, 2000). Ash was determined by incineration in a muffle furnace at 550°C for 6 h (AOAC, 2000). The crude protein (CP) method no. 978.04 (AOAC, 2005) was used to determine the total nitrogen content and CP was estimated by multiplying percentage of N content by a factor 6.25. Ether extract (EE) was determined using the AOAC method 920.39 (AOAC, 2000). Ash-free neutral detergent fibre (NDFom) and Ash-free acid detergent fibre (ADFom) were determined following the procedures of Mertens et al. (2002) with NDFom assayed without use of an alpha-amylase, but with sodium sulfite. Ash-free Acid detergent lignin (ADLom) was determined by placing dried ADF samples in 72% sulphuric acid (H2SO4) (Mertens, 2015). Calcium and Phosphorus were determined using inductively coupled plasma atomic emission spectrometry (Sah and Miller, 1992).
Slaughtering procedures
At the end of the feeding experiment, all sheep were individually weighed to obtain the final weight (FW), after being fasted overnight, with only access to water, to avoid digesta contamination. They were then transported to the Neudamm abattoir for slaughter. Sheep were slaughtered in accordance with the prescribed animal ethics and welfare regulations, at the registered abattoir facilities of the Neudamm campus, under the required conditions for local abattoirs in Namibia. Animals were stunned to render them unconscious and to ensure that they did not suffer pain during slaughter. After stunning, sheep were slaughtered by cutting the carotids, jugular vein, oesophagus and trachea to allow blood to drain from the carcass. The carcasses were hanged on the bleeding rail and the dressing operation began after the bleeding process was completed. All slaughtering procedures were done under the supervision of a veterinarian, specialized in Veterinary Public Health.
Carcass characteristics
After dressing and evisceration procedures, carcasses grading was done visually according to the South African Meat Classification System for beef, lamb, mutton and chevon, according to the Government Notice No. R.863 of 1 September 2006 (Republic of South Africa, 2006), which is also adopted by the Meat Board of Namibia. The carcasses were weighed to obtain hot carcass weights (HCW). The pH of the carcasses was measured 1 hour (pH1) after slaughter in the left Musculus longissimus dorsi at the area between 10th to 13th rib, using a portable digital pH meter with a penetration electrode and temperature probe. Subsequently, carcasses were chilled in a cooler under the temperature of 2–4°C for 24 hours to obtain the cold carcass weights (CCW). Cold carcass weight (CCW) was expressed as a proportion of the live weight at slaughter (SW) to determine the dressing percentage (Van der Merwe et al., 2020). The pH reading of the carcasses was taken again at 24 hours (pH24) in the Musculus longissimus dorsi on the same spot where pH1 was taken. After 24 hours of chilling, the carcasses were then split into two halves down the spinal column by a longitudinal cut on the vertebral column using a meat band saw. The left side of the carcass was used for further measurements such as external length (CEL) of each carcass, shoulder and buttock circumferences (SC and BC, respectively), which were measured with a flexible tape measure following the procedure described by De Boer et al. (1974).
The fat thickness was measured at ¼ (Fat Pos¼ ), ½ (Fat Pos½ ) and ¾ (Fat Pos¾) positions on the 12th rib from the chine bone end, with a vernier calliper and from these, average fat thickness was calculated (Lima et al., 2015; Araújo et al., 2017). The eye-rib depth and the eye-rib width of the Musculus longissimus dorsi exposed by cutting the carcass between the 12th and 13th ribs were measured with a Vernier calliper (Ferreira et al., 2012; Landim et al., 2015). As per the method described by Ferreira et al. (2012), the area of Musculus longissimus dorsi was determined by measuring the maximal width of the muscle, denoted as “A” and its maximal depth “B” and then applying the values into the following equation: (A/2 × B/2) × \(\pi\), where \(\pi\) equals to 3.1416. The three-rib cut was weighed and physically separated by blunt dissection into bone, lean and fat tissue which were weighed individually and expressed as a proportion of the cut as described by Van der Merwe et al. (2020).
Statistical Analysis
The chemical composition of the diets was subjected to analysis of variance (ANOVA) using Proc GLM (SAS, 2009). Similarly, FCR and ADG were subjected to analysis of variance (ANOVA) using Proc GLM (SAS, 2009) with the effects in the model being sex and treatment.
Intake is a function of the body weight (W) of the animal, hence analyses were based on metabolic body weight (W0.75). The model (1) included the effects: sex, treatment, week and treatment x week interactions. Data was analysed using Proc Mixed (SAS, 2009) which takes into account correlation between repeated measures on an individual and the Bayesian Information Criterion (BIC) which compares covariance structures based on goodness of fitness criteria, was used to select the appropriate covariance structure (Littell et al., 1998), which was simple. Estimate statements were used in Proc Mixed (SAS, 2009) to compare means and obtain standard errors. Effects were considered significant at P < 0.05.
Yijkl = µ + Fi + Sj+ Wk + FW ik + eijkl (1)
Where: Yijkl = feed intake; µ = overall mean; Fi = effect of treatment (i = T1, T2… T5); Sj = effect of sex (j = male, female); Wk = effect of week (k = 1,2,3…13); FW ik = interaction effect between treatment and week; eijkl = random error term explaining variation among experimental units (EU) on the same treatment.
A similar model to (1) was used to fit body weights, but with the additional interaction effect of treatment by sex. The best covariance structure for body weights was ante-dependence [ANTE(1)].
The data on carcass characteristics were subjected to analysis of variance (ANOVA) using the General Linear Model (GLM) procedure (SAS, 2009). The effects in the model were treatment (T1-T5), sex and their interactions. Data were checked for normality and transformation was carried out where necessary, before statistical analysis. Means were separated using the Turkeys’ Studentised range test.