2.2. Experimental Sheep and Feed Preparation

Ethiopia is known for its high population of domestic livestock, including sheep. The sheep production system in Ethiopia is characterised as an extensive system of management based mainly on grazing with little supplementation of concentrates and agricultural by‐products. Bonga sheep is one of the largest sheep breeds in the country. It is characterised by long and wide fat tailed with tapering and twisted ends. Both males and females are polled, have short and smooth hair, and mainly convex facial profiles of males and all types of colour but predominantly light red. The breed is known for its docile temperament, fattening potential, meat quality, fast growth and prolificacy (Mamo 2015). These sheep are commonly fed under grazing systems on pastures where they freely graze on open fields.

Natural pasture hay consisting of mixed pasture containing predominantly grass species was obtained from nearby farmers. It was chopped manually into 5–10 cm to increase feed intake and stored under shelter that avoids direct sunlight to maintain its nutrients. It was used as a basal diet and was offered daily as far as a 20% refusal rate. The commercial concentrate wheat bran (WB) and Noug seed cake (NSC) were purchased at the local market. The concentrate mix (CM) (WB, NSC) was a 1:1 ratio. The CH was purchased from the Awurada multipurpose cooperative. The CH was processed with an adaptation of the traditional experience of coffee preparation for an average of 20 min of boiling at 100°C and roasting at 144°C (brown spot formation in ranged 5–8 min) in the community. The total amount of supplement feeds was 400 g/day based on recommendations of CM supplementation to Washera sheep fattening (Abebe 2008). The processed CH (200 g) was mixed with CM (200 g) to maintain the nutrient requirements of the growing tropical sheep.

2.3. Experimental Design and Treatments

A total of 24 rams uncastrated with an average age of 1.02 years were tagged using an ear tagger for identification and quarantined for 21 days. During this period, each ram was vaccinated against common diseases such as ovine pasteurellosis, sheep pox, blackleg and anthrax using 1 mL ovine pasteurellosis, 1 mL sheep pox and 1/2 mL anthrax, respectively. Albendazole and fasinex were used for deworming against the internal parasite (flat/roundworms), and diazinon was sprayed for ticks and mange. An average of three initial weights were taken after overnight fasting using a spring balance scale (50 kg × 200 g). The rams were blocked based on the initial body weight (21.5 ± 1.01 kg) into six blocks with four rams per block and assigned randomly into one of the four experimental feeds in a randomised complete block design (RCBD). The basal diet (ad libitum), common salt (1%) and water were freely accessed by experimental rams. The treatments were conventional natural pasture hay (C) and 400 g conventional CM (T1), C + 200 g CM + 200 g boiled CH (T2), C + 200 g CM + 200 g roasted CH (T3) and C + 200 g CM + 200 g raw CH (T4). The rams were placed in an individual pen with a dimension of 2 m² smooth earth floor, which had partitioned feed troughs. The size of the natural pasture hay trough had a width of 30 cm, a height of 40 cm and a depth of 15 cm. The water and common salt troughs were installed on the left side by the right side at the centre natural pasture hay trough. The supplement feed was offered alone with a mobile bucket for each ram. The four experimental feeds were assigned randomly with a lottery system.

2.4. Feeding Trial

The natural pasture hays were offered once a day in the morning by measuring the required amount to ad libitum level for the rams. The supplement feeds were CM boiled CHCH, roasted CH and raw CH. The daily supplements were offered in two equal portions at 10 AM and 4 PM. Tap water and common salt (1%) were freely accessible. Data on experimental feed offered and leftovers were collected and weighed every day per individual ram, and feed intake was calculated with the difference between them. The average daily weight gain of the rams was calculated from the difference of final and initial body weight divided by the number of feeding trial days (90 days). The body weights of the rams were measured within the interval 7 days after overnight fasting of the rams. Feed conversion efficiency was computed with the ratio of average daily weight gain (g) to average daily feed intake (g) (Brown, Hindmarsh, and McGregor 2001).

2.5. Digestibility Trial

The digestibility trial was continued for 7 days after the feeding trial, using all rams under the feeding trial. The rams were harnessed with faeces collection bags and adapted for 3 days. The representative samples were prepared using 20% of sub‐samples of fresh faecal collection every day. It was preserved at a temperature of −20°C for 7 days. The faecal samples were thawed at room temperature. The samples were mixed to prepare composite samples and dried at 60°C for 72 h. The apparent digestibility coefficient of current feed nutrients as a percentage was estimated from the difference of nutrient intake and nutrient in faecal (out) divided by total nutrient intake.

2.6. Chemical Composition Analysis

The dried sample of offered feeds and collected natural pasture leftovers and faecal samples from each treatment during this trial were milled and passed through a 1 mm sieve size. The proximate analysis procedures were employed for the chemical composition analysis of representative samples for dry matter (DM), crude protein (CP) using the conventional Kjeldahl method, and ash content (AOAC 1990). The content of ash was obtained from overnight (24 h) ignited samples using a muffle furnace at 550°C. The neutral detergent fibre (NDF), acid detergent fibre (ADF) and acid detergent lignin (ADL) content of the samples were analysed with the procedure of the Van Soest method (Van Soest, Robertson, and Lewis 1991). The metabolisable energy (MEI) intake of experimental sheep was estimated from digestible organic matter intake using the formula ME (MJ/kg) = 0.0157 × DOMI g/kg (AFRC 1993), where DOMI = digestible organic matter intake per kg of DM.

2.7. Data Analysis

The data collected were organised with Microsoft Excel 2010. Before the analysis of variance (ANOVA), the normality test of data was undertaken using the Shapiro–Wilk test. ANOVA was performed for these data at a significance level of p < 0.05. The least significant difference (LSD) test was used to separate treatment mean differences (p < 0.05). The general linear model (GLM) was employed to data analysis which is deployed in R software v. 4.3.2 through packages agricolae version 1.3‐5 (R Core Team 2022). It was fitted to the following formula:

where

3.2. Dry Matter and Nutrient Intakes

The DM and nutrient intakes of Bonga rams offered (CM, processed and raw CH) are shown in Table 2. In the current study, high total DM intake (p < 0.05) was recorded when rams fed on experimental feed assigned in treatment T1 and followed by T2. The maximum MEI (p < 0.001) was recorded in the boiled CH (6.970 MJ/kg DM) and CM (6.875 MJ/kg DM). The minimum MEI of the supplement feed was obtained in the raw CH (4.811 MJ/kg DM). Also, rams responded highly differently to the intake of CP, ADF and ADL when offered processed and unprocessed CHs. However, ram intake of organic matter was comparable between treatments (p < 0.06). The rams assigned in the treatment (T1) showed a maximum intake of CP (120.026 g/day) and a minimum intake of ADF (99.887 g/day) and ADL (47.850 g/day). It was followed by treatment (T2) with the intake of CP 113.403 g/day. On the other hand, the rams assigned in treatment (T4) showed a minimum intake of CP (104.525 g/day) and a maximum level of ADF (162.544 g/day). But ADL intake was maximised to 63.249 g/day when rams were supplemented with roasted CH.

3.3. Apparent Digestibility

The apparent digestibility of nutrients in the current experimental feeds is presented in Table 3. The level of digestibility of each nutrient was strongly (p < 0.01) affected when rams were fed on processed and unprocessed CHs. About 57.351% apparent digestibility of DM and 59.781% organic matter were obtained from boiled CH (T2) and less than 50% from raw CH (T4). The apparent digestibility of NDF was statically similar for processed CH and higher than raw CH (28.860%). The ADF was digested more (p < 0.05) in the boiled CH (43.126%) and followed by the roasted CH (31.408%). But the minimum apparent digestibility of ADF was recorded from raw CH (20.343%).

3.4. Body Weight Changes and Feed Conversation Efficiency

The body weight changes and feed conversion efficiency of Bonga rams supplemented with CM and processed and raw CH are presented in Table 4. The initial body weights of the rams were comparable (p < 0.92) before the rams were exposed to different feed treatments, whereas the performance of final body weight, body weight change, average daily weight gain and feed conversion efficiency of rams were highly (p < 0.001) affected among rams fed on control (T1), raw CH (T2), broiled CH (T3) and roasted CH (T4) under supplementation of CM. The mean final body weights of the rams were 31.280, 30.442 and 28.666 and 25.860 kg, obtained with supplementation of boiled CH, CM and roasted and raw CHs, respectively. The mean body weight gains of ram per day were improved at the rate of 108.296 g/day when half of the CM (200 g) was replaced with boiled CH (200 g). It was also comparable with the CM (T1) in which it was increased with an average body weight gain of 97.960 g/day. The minimum body weight gain (51.777 g/day) was recorded from treatment (T4). The feed conversation efficiency was maximum (0.047 g) for treatment T2 and followed by T1 (0.043 g), whereas it was 0.023 g for raw CH (T4).

4.2. Dry Matter and Nutrient Intakes

The DM intake was improved by more than 16.751 g/day in comparison to raw CH when the boiled CH was offered to rams (Table 2). This might be due to the effectiveness of the boiling method for loosening the digestible fraction of the plant material and lignin fraction and reducing anti‐nutritional factors of the CH. Furthermore, metabolised energy intake of CH was significantly influenced by the types of processing. It was high when rams were supplemented with concentrate and followed by a boiled CH. These results might be due to the loss of anti‐nutrients caffeine, an alkaloid (trigonelline) and hydrolysis of polysaccharides such as hemicellulose, pectin and cellulose at high temperatures during boiling (Cangussu et al. 2021). This result showed a similar metabolised energy intake of Doyogena sheep fed steam‐processed sweet lupin (Tilaye et al. 2022).

In the current feeding trial, organic matter and NDF intakes were not affected when rams were fed either processed or unprocessed CH. On the other hand, when the treatment feeds were changed from raw to boiled and roasted CHs, the intake of ADF, ADL and CP by the rams was significantly affected. This indicates that processed CH enhanced the intake of protein more than raw CH, but ADFI and ADLI were inconsistent for roasted (T3) and raw (T4). It might be due to the effect of processing on ADF and ADL. These results confirmed that the intake of nutrients was significantly affected when ram fed processed CH than raw, which is comparable to the nutrient intake of processed sweet lupine (Tilaye et al. 2022). It is similarly reported as an opposite trend between low protein and high NDF, ADF and ADL intake levels for different supplement reports (Mekonnen et al. 2019; Sileshi et al. 2021; Tilaye et al. 2022). Moreover, the rams fed raw and roasted CH characterised a high level of ADF, ADL intake, and the low level of protein might be due to the filling of reticulorumen space and then minimised the intake of protein. This is in line with the nutrient intake of cows, which can be limited due to reticulorumen capacity receiving high‐fibre diets (Dado and Allen 1995).

4.3. Apparent Digestibility

In the present study, rams fed with boiled CH (T2) had maximum apparent digestibility of DMD and OMD in comparison with raw and roasted CH (Table 3). This might be due to the boiling process of CH minimising the content of caffeine, tannins and phenolic acid contents, which enhanced the digestibility of the husk in rams. Boiling as a treatment on mangrove leaves reduces these compounds and increases nutrient digestibility (Basyuni et al. 2023). Moreover, CH blanching also caused hydrolysis of polysaccharides, proteins and alkaloids (Cangussu et al. 2021), which might be a cause for the increment of the apparent digestibility of it in this study, whereas rams fed with roasted CH had minimum apparent digestibility. This could be due to excessive heat treatment during roasting, which caused a change in the texture and retrogradation of crude fibre and decreased the size of the washable fraction (at 136°C), which reduces degradability within the rumen microbes (Ahn et al. 2019; Goelema et al. 1999; Iommelli et al. 2022; Sajilata, Singhal, and Kulkarni 2006). Moreover, the apparent digestibility of ADF significantly deferred between raw and processed CH and CMs. The minimum apparent digestibility of raw CH might be attributed to the high content of NDF and ADF yet not treated as the processed CH.

The apparent digestibility of Bonga ram in this experiment was in line with the apparent digestibility of Washera (Alemu, Animut, and Tolera 2014) and Simada sheep (Dessie et al. 2010) fed a basal diet of natural pasture hay supplemented with Millettia ferruginea leaf hay (dried Bribra) and graded levels of CM, respectively. However, it was lower than the report on Doyogena sheep fed a basal diet of natural pasture hay supplemented with steamed sweet lupine (Tilaye et al. 2022). Thus, the apparent digestibility of nutrient differences in the diet of rams might be due to the difference in the breeds of sheep, source of basal diet and supplement feed resources.

4.4. Body Weight Changes of Rams

The maximum body weight change was comparable with the body weight change of Doyogena sheep fed on processed sweet lupine (soaked) in the same region in south Ethiopia (Tilaye et al. 2022). However, the mean daily weight gain of ram was 108.296 g/day (Table 4), which is lower than the mean weight gain of Doyogena sheep per day fed on a basal diet supplemented with processed sweet lupine (145 g/day) (Tilaye et al. 2022) The daily average body weight gain performance in the current study is higher than the value (56.0–87.7 g/day/h) reported by Mengistu, Assefa, and Tilahun (2020) who performed supplementary feeding of different protein sources for the same breed of sheep in Ethiopia. The discrepancy of reports might be related to the ages of lambs and feeding management practices during the experiments. Moreover, the present finding is higher than Sidama sheep (51.1 g/day), Washera sheep (71 g/day) fed on a basal diet supplemented with 350 g CM (Dessie et al. 2010) and 25% tree lucerne dried leaves and 75% CM (Hailecherkos, Asmare, and Mekuriaw 2021), respectively, in the different parts of Ethiopia. The Bonga sheep breeds with high rate of growth and weight gain with better secondary metabolites (tannins) reduction strategy responses might have improved the performances of this particular breed (Tarekegn, Shitaye, and Gafaro 2022; Yisehak et al. 2016). The feed conversion efficiency was maximised for rams under the treatment of boiled CH. However, it had lower feed conversion efficiency than Washera and Doyogena sheep (Hailecherkos, Asmare, and Mekuriaw 2021; Tilaye et al. 2022). This might be due to genetic differences between sheep breeds and the type of supplement feeds. Hence, based on the maximum body weight changes of rams, boiled CH had the potential to replace 50% of the commercial CM and be used as an alternative supplement feed containing essential nutrients alike. Essential nutrients are significant for rumen microbe and provide nutrients at the tissue level for the host ruminant (McDonald et al. 2002).

The authors have nothing to report.