Appraisal of cowpea (Vigna unguiculata) hay as a replacement for noug seed (Guizotia abissynica) cake in the ration of Gumuz lambs in Ethiopia

Abstract Background The objective of this study was to evaluate the replacement value of cowpea (Vigna unguiculata) hay for noug seed (Guizotia abissynica) cake on dry matter (DM) and nutrient intake, nutrient digestibility, body weight change, carcass characteristics and economic feasibility of the feeding regime of Gumuz lambs. Methods The study was conducted using 25 yearling intact male lambs with an initial body weight of 18.26 ± 0.63 kg (mean ± SD).The lambs were assigned into five treatments: (T1) 272.3 g noug seed cake (NSC); (T2) 200.73 g NSC + 105.55 g cowpea Sewinet variety hay (CSH); (T3) 134.3 g NSC + 211.86 g CSH; (T4) 66.31 g NSC + 313.79 g CSH and (T5) 417.98 g CSH on DM basis. The experiment was conducted using a randomized complete block design (RCBD), and lambs were blocked based on their initial body weight. The feeding trial was conducted for 90 days followed by 7 days of digestibility trial. Natural pasture hay was treated with molasses solution for improvement of palatability and digestibility and offered for all experimental lambs in ad libitum. Results The crude protein (CP) contents of natural pasture hay (NPH), CSH and NSC were 4.99, 18.31 and 36.5%, respectively. The total DM intakes of lambs (P < 0.05) increased at higher levels of CSH supplementation. The CP intake decreased with increasing levels of CSH supplementation. Digestibility of DM, organic matter (OM), neutral detergent fiber (NDF) and acid detergent fiber (ADF) was improved by a higher level of CSH supplementation but CP digestibility was not affected by supplementation of CSH. The average daily gain (ADG) showed significant difference (P < 0.05) among treatments. Similarly, the feed conversion efficiency (FCE) was significantly different (P < 0.05) among treatments; higher FCE was recorded at T1 but lower was at T3 and T4. The hot carcass weight (HCW) showed significant difference (P < 0.05) among treatments; higher value was observed in T5 (9.36 kg), but the smallest value was recorded for lambs on the T3 (7.36 kg). The higher dressing percentage (DP) on empty body weight basis was recorded in three treatment (T1 = T4 = T5) groups compared with (T2 > T3) treatments. Significantly higher (P < 0.05) rib‐eye area was achieved at a high level of CSH and sole NSC supplemented lambs. The economic feasibility showed that the highest profit was achieved in T5. Conclusion From the present study, it was concluded that CSH could be used for supplementation of protein source feed for body weight gain for Gumuz lambs by replacing high‐cost concentrates in a native hay‐based basal diet. As a final point, CSH supplementation was recommended as replacement of NSC at 417.98 g CSH (T5) for better biological performance as well as economic value.

body weight basis was recorded in three treatment (T1 = T4 = T5) groups compared with (T2 > T3) treatments. Significantly higher (P < 0.05) rib-eye area was achieved at a high level of CSH and sole NSC supplemented lambs. The economic feasibility showed that the highest profit was achieved in T5.
Conclusion: From the present study, it was concluded that CSH could be used for supplementation of protein source feed for body weight gain for Gumuz lambs by replacing high-cost concentrates in a native hay-based basal diet. As a final point, CSH supplementation was recommended as replacement of NSC at 417.98 g CSH (T5) for better biological performance as well as economic value.

K E Y W O R D S
body weight, cowpea Sewinet variety hay, Gumuz lamb, hay, noug seed cake

INTRODUCTION
Ethiopia has huge livestock population with diversified functions such as food production, assistance in overall agricultural production and processing at the household level (CSA, 2019). The contribution of livestock in the national economy was estimated at 15%-17% of gross domestic product (GDP) and 35%-49% of the agricultural GDP provided through export commodities like live animals, hides and skins to earn foreign exchanges (Getu, 2015). The country receives annually on average about 105 million US dollars from the export sales of live sheep and mutton (ESPS-LMM 2011). Though sheep in the country have much relevance, their productivity is below their potential; for instance, the average carcass weight of Ethiopian sheep and goats is 10 kg, and an estimated meat production is 3.5 kg per sheep per year in the population (Hirpa & Abebe, 2008). These values are much lower than those of neighboring countries such as Sudan, Somalia, Djibouti and Kenya, which, respectively, produce 13, 13, 12 and 13 kg/head (Sebsibe, 2008). Inadequate feed supply and its poor quality are the major problems for low productivity of sheep (Deresse et al., 2014).
To tap the potential of sheep, feed resources must be improved in quality (CSA, 2019;Shapiro et al., 2015). Supplementation of sheep with protein sources is one of the strategies that improve the efficiency of utilization of available roughage feed resources (Hailecherkos et al., 2020;Obeidat et al., 2020). Noug seed (Guizotia abyssinica) cake is one of the commonly utilized protein sources for supplementation of sheep in Ethiopia. However, this supplement feed is becoming limited under smallholder farmers due to insufficient supply and escalating price, which calls to look for other cheap and accessible alternative protein supplements. Among the options, forage legumes which can be produced at local conditions is one of options to improve the productivity of sheep (Hailecherkos et al., 2020;Makuriaw & Asmare, 2018). In the study area, cowpea hay was indicated as a potential protein source used as supplementary feed for small ruminants (Gebrekidan et al., 2019). Agzaet al. (2012) suggested that cowpea (Vigna unguiculata) Sewinet variety is one of the best adaptable, short-duration, highforage biomass and nutrient-rich forage variety, but their nutrient utilization on performance of different animals is not yet determined.
However, the potential of cowpea Sewinet variety hay (CSH) as a feed supplement has not been utilized due to various reasons, such as the lack of awareness on the utilization of fodder crops. Assumption of the study was that CSH is substitute NSC as a protein source for fattening of Gumuz lambs. Therefore, this research was conducted to evaluate the replacement of NSC with CSH on nutrient utilization and the overall performance of Gumuz lambs.

Forage establishment and feed preparation
Cowpea (Vigna unguiculata) Sewinet variety was sown through rainfed on well-prepared land. The land plowing, sowing, fertilizer application, weeding, harvesting, preservation (haymaking) and storage were applied based on recommended practices for the forage (DIAS, 2011).
Cowpea Sewinet variety was harvested for hay at 50% flowering, chopped and stored under shade area to prevent nutrient loss due to solar radiation and rainfall. The NPH was harvested at 50% flowering (McDonald et al., 2010). CSH and NPH were chopped to 3-5 cm by the chopping machine to make ready for consumption of lambs and stored on a well-ventilated concrete floor shade to avoid spoilage and mold development. NSC was purchased from the local market. Molasses were purchased from the Wonji sugar factory and mixed thoroughly with natural grass hay. The experimental feed was mixed to make ISO-N and ISO-caloric basis properly before given to the experimental lambs.
A 72. 53, 63.47, 54.47, 45.18 and 40.65 g of molasses for T1, T2, T3, T4 and T5, respectively, was diluted in water at a ratio of 1:2, and diluted molasses solution was properly mixed with NPH (Valadares et al., 2002). The molasses solution was to improve the palatability and digestibility of feed.

Experimental animals and their management
Twenty-five intact yearling male Gumuz lambs with a mean initial body weight of 18.26 ± 0.63 kg (mean ±SD) were purchased from the local market. The age of lambs was determined by dentition and asking for birth information from the owners. The experimental lambs were quarantined for 21 days to observe their health condition. During this time, the animals were drenched with anti-helminthes (Albendazole 300 mg bolus) and sprayed with Amitraz 12.5% with a dose of 1.6 ml per liter water against internal and external parasites, respectively, and vaccinated against common diseases of the area peste des petits ruminants before the beginning of the actual experiment. After the quarantine, the animals were moved to an experimental pen and fed the experimental ration for 15 days of the adaptation period. Then animals were ear-tagged for identification purposes and due to the weight difference, the lambs were grouped into five blocks based on empty body weight (EBW) which was determined after two consecutive weighing after overnights fasting. Then lambs were randomly assigned to the treatments within the block and placed in an individual pen of the size 0.85 m width × 1.2 m length furnished with feeding trough and watering buckets. The feeding trial was conducted for 90 days followed by 7 days of digestibility experiment. NPH was weighed and offered ad libitum two times a day as a basal diet at a rate of 20% refusal. The supplement feed was offered separately twice a day at 8:00 AM and 4:00 PM in equal proportions. Water and salt lick were available to lambs at all the time. The refusal was collected and weighed every morning before offering the daily ration. After the growth trial, the digestibility study was conducted for seven consecutive days. In the digestion trial, lambs were allowed to adapted fecal bags for four days and a digestibility study was conducted for seven days following adaptation.

Experimental design and treatments
The experiment was conducted using a randomized complete block design (RCBD) with five blocks and five treatments. The lambs were blocked based on their initial body weight which was determined by weighing lambs after overnight fasting. The ration was formulated with different proportions of CSH and NSC, making to meet the requirement of growing tropical lambs given in Table 1. The proportion of CP obtained from CSH was substitution levels of 0%, 25%, 50%, 75% and 100% of NSC for treatments 1, 2, 3, 4 and 5, respectively.

Chemical composition of feed refusals and feces
Samples of CSH, NSC and natural pasture hay (NPH) offered and refused were collected daily and pooled over the experimental period for each feed and treatment. The samples of feed offered and refused were dried in an oven to a constant weight at 65 o C for 72 hours to determine nutrient content. The partially dried samples of feeds and feces were ground using a laboratory mill to pass through a 1-mm sieve screen size and taken to animal nutrition laboratory for chemical analysis. The dry matter (DM) content of feeds was determined after the drying of another sub-sample of feeds at 105 o C in an oven to constant weight. The samples of feed offered, refused and feces were analyzed for DM, ash and nitrogen (N) according to the procedures of AOAC (1990). Crude protein (CP) was estimated as N ×6.25. Neutral detergent fiber (NDF), acid detergent fiber (ADF) and acid detergent lignin (ADL) were analyzed by the method of Van Soest (1994).

Dry matter and nutrient intakes
The daily amount of DM and nutrients offered and the refusal were weighed for each animal and recorded to determine the amount of DM and nutrients consumed as the difference between the offered and refused. A sample of feed offered was taken from batches of feeds and refusals have been collected from each animal across the experimental period and finally pooled for each treatment basis and subsampled for laboratory analysis. Nutrient intake was determined as the difference between the quantity of nutrient offered and refused for each lamb.

Apparent digestibility of DM and nutrients
The apparent digestibility coefficient of DM, OM, CP, NDF and ADF was determined as a proportion of the nutrient intake not recovered in the feces using the following formula: The digestibility of the supplement feed was determined by difference, which is calculated after knowing the digestibility of the basal diet using

Body weight gain and feed conversion efficiency
The initial body weight of each animal was determined by taking the mean of two consecutive weights after overnight fasting. The weight of the animals was measured every 10 days afterward, after overnight fasting. Average daily body weight gain (ADBWG) was calculated as the difference between the final and initial BW divided by the number of feeding days. Feed conversion efficiency (FCE) is the measure of feed utilization how efficient the lambs were converting the feed into meat as indicated in the following formula: Bodyweight gain (kg) = Final body weight (kg) − Initial body weight (kg) ADBWG (g) = Body weight gain (g) ∕Number of feeding days (90) FCE = ADBWG (g) ∕Daily total DMI (g)

Carcass characteristics
The carcass characteristics of experimental lambs were evaluated by slaughtering all lambs in each treatment after overnight fasting at the end of growth and digestibility trials. Slaughter weight (SW) was taken right before slaughter. The lambs were slaughtered by severing the jugular vein and carotid artery with a knife. The blood of each lamb was drained into the bucket and weighed immediately after collection.
After the lambs were killed, the skin flayed carefully to avoid the adherence of fat and muscle tissue to the skin, and the skin was also weighed.
The entire gastrointestinal tract without the esophagus was removed and divided into two sections as stomach and intestine and weighed with gut fill. During removal of the gastrointestinal tract mesenteric fat was separated carefully. Internal organs, lung, heart, kidney, liver, spleen, pancreas, genital organ, gall bladder and other parts like head, visceral fat and feet have been removed and weighed. The weight of the hot carcass was measured after all the offal's properly removed from the carcass. Based on the feeding habit of the people in the area, edible and non-edible offal was identified and recorded. The total edible offal components were taken as the sum of blood, heart, head, liver, kidney, tail and visceral fat. The non-edible offal component was taken as the sum of skin with feet, genital organ, and lung with trachea, spleen, gall bladder and gut fill. The total usable product was taken as the sum of HCW, skin and total edible offal component. To measure the rib-eyearea of the carcass loin part was partitioned into fore and hindquarters between the 11th and the last ribs. The cut ribs were chilled for 12 hours in the deep refrigerator and the rib-eye area in cm 2 was measured at the 12th and 13th rib sites. The cross-section of the rib-eye muscle was traced first on transparency plastic paper after the loin part is cut between the 12th and 13th ribs perpendicular to the backbone.
Then the traced transparency paper was positioned on graph paper squares each having an area of 1 mm × 1 mm. The numbers of squares included within the mark were counted for the left and right sides, and the area was computed as the average of the two. Finally, EBW was calculated as the difference between SW and gut content. Dressing percentage (DP) was determined as the proportion of HCW to slaughter body weight (SBW) or EBW.

Statistical analysis
Data on feed intake, digestibility and carcass characteristics were analyzed by analysis of variance (ANOVA) using the GLM procedure of the statistical analysis system (SAS) version 9.3. The difference between treatment means was separated using the least significant difference test.
The statistical model was : where Yij is the response variable (an observation in i treatment and j block); μ is the overall mean; Ti is ith treatment (test diets) effects; βj is the jth block effect (initial body weight effect); and Ɛij is the random error.

Chemical composition of the experimental feeds and refusals
The chemical composition of experimental feeds and refusals of the experiment is shown in Table 2. The CP content (4.99%) of NPH used in the current study was below the threshold level of (7%) of CP for best microbial activity in the rumen that can support the maintenance TA B L E 2 Chemical composition of the experimental feed offered and refusals  requirements of the animals. Hence, the NPH used in this experiment can be characterized as lower quality feed in terms of its CP content.
The current NPH refusal had higher NDF, ADF and ADL values than the hay offered. This may be because experimental animals selected more edible portions of the basal diet and left the more fibrous parts (such as stems) of the grass which has higher fiber (NDF, ADF and ADL) fractions. Generally, DM, OM and CP content of all experimental feed offered was greater than feed refusal whereas in T2, T3 and T4 the NDF, ADF and ADL percentage was the reverse. Refusals mainly constituted stem parts of the feed because the stem of the CSH contained higher lignin content. The CP content of CSH offer was higher than refusal, but the NDF, ADF and ADL were contrary to CP content.

Dry matter and nutrient intakes
The DM and nutrient intakes of Gumuz lamb-fed molasses-treated NPH and supplemented mixture of CSH and NSC are given in Table 3.
Statistically, there was no significant difference on the basal diet DM intake (P > 0.05) among treatments. However, numerically higher basal diet DM intake was recorded by T1-, T2-and T5-supplemented groups compared with T3 and T4 treatments. Apart from T4 and T5 other treatments group were consumed all supplements offered feeds without any refusals. The total DM and OM intakes were significantly different (P < 0.05) among the supplemented groups in which the daily total DM intake increased as the level of CSH supplementation increased.
The OM, CP, NDF, ADF and ADL intakes of experimental lambs were significantly different (P < 0.05) among treatments. The daily CP intake was significantly higher (P < 0.05) in T1 and T2 and ranked increasing order in CP intake as T1 = T2 > T3 > T4 > T5. This is associated with the bulky nature of the CSH when it substituted NSC at a higher level (75% and 100%). The estimated ME intake was significantly different (P < 0.05) among the treatments. The ME intake ranked the supplemented group as T5 > T4 = T3 = T2 = T1.

Dry matter and nutrient digestibility
The values for DM and nutrient digestibility and digestible nutrient intake of Gumuz lambs fed on basal diet offered molasses-treated NPH and supplemented with CSH and NSC mixture are presented in Table 4.
A significant difference (P < 0.05) was observed among treatments in terms of DM, OM and ADF digestibility in the current study. However, there was no significant difference (P > 0.05) on CP digestibility as well as digestible CP intake among treatments. The higher DM and OM digestibility observed in sole CSH-supplemented sheep might be due to a high intake of the supplemented diet, and high concentration of nitrogen in legume forages improves the rate of roughage feed degradation in the rumen. There was no significant difference (P > 0.05) on CP digestibility among treatments, but higher apparent digestibility of ADF and NDF was observed in T5.

Body weight change, daily gain and FCE
The initial and final body weight, average daily weight gain and FCE of Gumuz lambs fed NPH and supplemented with NSC and CSH and its mixture are presented in Table 5. The current result showed a significant difference (P < 0.01) on ADG and FCE among treatment but non-

Carcass characteristics
Carcass characteristics were evaluated based on DP on SBW and EBW base, HCW, REA, internal fat deposits, and edible and non-edible offal.

Slaughter and EBW, HCW and DP
The present values of SBW, EBW, HCW, DP as a proportion on slaughter and EBW basis and rib-eye muscle area (REA) of the lambs are given in

Correlation between DM, nutrient intakes, digestibility and weight gain
The results of correlation among DM, nutrient intake and digestibility, weight gain and SBW are presented in Table 8. The finding elucidated DM intake was positively and strongly correlated (P < 0.001) with OMI, CPI and SBW. DMI was also significantly correlated (P < 0.01) with CP digestibility. In addition, DM, OM and NDF digestibility and TA B L E 8 Effect of replacement of NSC with different levels of CSH on correlation between dry matter and nutrients intake, digestibility and carcass characteristics of Gumuz yearling lambs fed MTNPH as a basal diet Abbreviations: ADFD, acidic detergent fiber digestibility; ADFI, acidic detergent fiber intake; CPD, crude protein digestibility; CPI, crude protein intake; DMD, dry matter digestibility; DMI, dry matter intake; NDFD, neutral detergent fiber digestibility; NDFI, neutral detergent fiber intake; NS, non-significant; OMD, organic matter digestibility; OMI, organic matter intake; SBW, slaughter body weight. ***Significant correlation at alpha 0.001. **Significant correlation at alpha 0.01. *Significant correlation at alpha 0.05.
ADG were significantly correlated (P < 0.05) with dry matter intake for Gumuz lambs. The positive correlation between these factors indicates the enhanced fermentation and passage rate, which leads to improved intake because of the dietary treatments. Besides, the molasses offered to lambs might have contributed to the improved intake of DM and enhanced rumen fermentation and passage rate of supplement feed and basal diets. But NDF intake and ADF digestibility were no significantly correlated, but ADF intake was negatively correlated with DM intake. The CP intake was positively and significantly (P < 0.001) correlated with DM, OM, CP, NDF digestibility and SBW, which might be associated with the fact that increasing CP intake linearly increases CP digestibility, enhance microbial population and facilitates rumen fermentation. The CP digestibility was positively (P < 0.05) correlated with DM, OM and CP intake, and digestibility of DM, OM and ADFD (P < 0.05). In the current study, DM digestibility was strongly (P < 0.001) correlated with OM, CP digestibility. Similarly, the DM digestibility was significantly (P < 0.01) correlated with CP, NDF and ADF digestibility. The CP digestibility was significantly (P < 0.05) correlated with NDF, ADF digestibility and SBW. This indicates that as CP digestibility increases, the NDF digestibility and ADG also increase.
This might be due to the effect of molasses, which cause improved fermentation in the rumen of animals by improving bacterial multiplication. The CP digestibility was highly significantly correlated (P < 0.001) with ADG and SBW, but negatively correlated with ADF digestibility.
This might be because lignification of plant material hampers rumen fermentation of roughage feeds.

Partial budget analysis
The result of the partial budget analysis for the performance of Gumuz lamb on replacement of NSC with CSH fed on NPH basal diet is indicated in Table 9. The prices of feed ingredients used in this research were 0.5, 1.5, 5 and 10 ETB/kg for NPH, CSH, molasses and NSC, respectively. There was a substantial variation among treatments on a partial budget of replacement of NSC with CSH for Gumuz lambs.
There was no loss of price per lamb in all treatments, and this might be due to the weight gain shown by experimental animals and better

Chemical composition of the experimental feeds and refusals
The DM content of NPH offered in this study was almost comparable with values 91.3 and 91.6% reported by Gulilat et al. (2018) andYigzaw et al. (2019), respectively. However, the current value is lower than the values 94.53 and 95.75% reported by Kibrom (2017) andGetent et al. (2019), respectively. Regarding findings in Ethiopia, the current CP value is higher than reported by Asmare et al. (2016) and Gelgelo et al. (2017), 3.5 and 3.73%, but lower than the value 9.81% reported by Gebregiorgis et al. (2017), respectively. The CP content of NPH used in the current study was below the minimum level (7%) required for maintenance in ruminant animals (Van Soest 1994). The differences in CP content among the various studies could be attributable to environmental factors in which the hay was grown, harvesting time and stage of growth at which the hay was harvested, drying process and grasslegume proportion of hay used for the experiment (McDonald et al., 2010;Wu, 2018).
The current NDF and ADF value of NPH was slightly comparable with the previous result of Yigzaw et al. (2019) and Gulilat et al. (2018), who reported 58.68 and 55.88% of NDF and 41.2 and 48.95% of ADF, respectively. However, these were lower than the values 71.23 and 83.85% of NDF and 46.43 and 48.98% of ADF contents reported by Asmare et al. (2016) and Kibrom (2017), respectively. In the feed refusals and offered composition, the NDF, ADF and ADL of refusal val-ues were higher than the hay offered. This might be because experimental animals selected more edible portions of the basal diet and left the more fibrous parts (such as stems) of the grass which has higher fiber (NDF, ADF and ADL) fractions (Faji et al., 2019;Mekonen et al., 2015). This entails that when sheep are allotted in roughage diets, selection can be ensured through provision of adequate amount of feed for producers in tropical regions of the world.
The OM content of NSC used in this experiment is higher than the result 88.1 and 84.6% reported by Kebede (2014) and Galgileo et al.  Lonsdale (1989), CSH used in the current experiment can be classified as a medium protein source feed for ruminant animals. In the present study, NDF content of CSH is lower than the value 57.77, 47.38% of NDF and almost similar value of ADF that 31.11 and 31.42% noted by Medikisa et al. (2016) and Tarekegn et al. (2019), respectively. This might be due to lower NDF content of CSH used in the current experiment showed that the hay is highly consumed and subsequent utilization by lambs (McDonald et al., 2010;Wu, 2018).
Generally, DM, OM and CP content of all experimental feed offered was greater than feed refusal while in T2, T3 and T4 the NDF, ADF and ADL percentage was the reverse. The reciprocal relationship between the protein and fiber content of plants was due to the maturity of the cell wall constituent along with lignin increase (McDonald et al., 2010). This might be due to the selective consumption behavior of the sheep.
Refusals mainly constituted stem parts of the feed because the stem of the CSH must have been containing higher lignin content. The CP content of CSH offer was higher than refusal, but the NDF, ADF and ADL were contrary to CP content.

Dry matter and nutrient intakes
The basal diet DM intake of T5 was higher than the value (358 g/day) reported by Ayenew et al. (2019) in the Gumuz sheep breed supplemented with Camel's Foot Tree (Piliostigma thonningii) leaf meal and lower than the value (586.39 g/day) reported by Getent et al. (2019) for NPH supplemented with toasted soybean grain for the same breed.
It has been learned that during the feeding experiment, the acceptance of CSH by experimental sheep was very high, which in turn elucidates the remarkable palatability of the hay. Moreover, the relatively low CP content of CSH compared with NSC that in turn forced lambs to con- leaves, cowpea hay and Camel's Foot Tree (Piliostigma thonningii) leaf meal, respectively. The variation of CP intake in the current study from the previous findings might be due to the bulky nature of diets, sheep breeds, feeding method and environment in which experiments were conducted. The estimated ME intake of sole CSH supplemented sheep disagree with the previous finding of Gebrekidan et al. (2019). The ME requirement for a 20 kg lamb gaining 50-150 g/day is 3.7-6.4 MJ/day for diets with 65% of metabolizability (McDonald et al. 2010). According to ARC (1980), ME requirement for the maintenance and growth (50-200 g/day) gain for the same weight lamb is 4.5-7.9 MJ/day. Thus, based on these assumptions, the estimated ME of the treatment diets (5.7-6.68 MJ/day) in the current study was assured not only the maintenance but also the energy requirement as well as the growth of the experimental lambs was met. The increase in ME intake might be asso-

Dry matter and nutrient digestibility
The DM and OM digestibility of the current study was lower than the value of Girmay and Amare (2017)  Higher apparent digestibility of ADF and NDF was observed in T5, associated with the lower fiber content of CSH, which was attained because of the optimum stage of maturity when forage was harvested.
A study by Jung and Allen (1995)  NDF and ADF digestibility observed for CSH supplemented groups in the current study might be due to high feed intake and legume forages are rich in nitrogen that facilitates fermentation of roughage feeds in ruminants. The study by Getent et al. (2019) indicated that DM intake was positively correlated with OM, CP and ADF intake and digestibility. McDonald et al. (2010) indicated that improvement in DM and OM digestibility associated with legume forage leaf meal is due to better nitrogen access in the rumen, thereby improving the rate of fermentation of ingested feed.

Body weight change, daily gain and FCE
The current finding in ADG is higher than the value 20.33 g reported by Abebe and Tamir (2016)

Slaughter and EBW, HCW and DP
The current finding agrees with the value 17.3 kg EBW and 9.25 kg HCW reported by Gelgelo et al. (2017) for black head Somali sheep supplemented with local brewery by-product and concentrate mixture.
However, these values were lower than those reported by Getent et al.

Rib-eye muscle area
The current study concurs well with the value (7.33 cm 2 ) reported by Getent et al. (2019)

Carcass offal
The amount of carcass offal obtained in the current result is higher than value 64.2 and 90 g reported by Adugna et al. (2020) andHailecherkos et al. (2020) for Gumuz and Washera lambs supplemented with treated local bamboo leaf hay and different levels of Tree Lucerne (Chamaecytisus palmensis) fed Desho grass as a basal diet, respectively. This variation might be due to the difference of NDF and ADF content, DM digestibility and the weight gain of lambs. The value of the current finding is in contradiction with the value reported by Gelgelo et al. (2017) for Somali sheep supplemented with local brewery by-product (Tata), and concentrate mix significantly improved the total weight of edible offal such as blood, heart, liver, tongue, reticulo-rumen and tail. The result was higher than the value (3.31 kg) described by Gulilat et al. (2018)

Correlation between DM, nutrient intakes, digestibility and weight gain
In the current study, as CP intake increased, DM intakeand digestibility also increased, which in turn enhanced ADG and SBW. This is similar to the result of Mekonnen et al. (2019) who indicated that the increased in DM intake was increased DM digestibility of the ration and improved body weight changes of animals. The current result was similar to the report of Adugna et al. (2020) and Hailecherkos et al. (2020), who indicated that CP intake was positively correlated with intake of DM, OM and CP and OM and CP digestibility and ADG of CP, but not correlated with NDF and ADF digestibility. The positive association of CP digestibility with NDF, ADF digestibility and SBW might be due to the effect of molasses, which cause improved fermentation in the rumen of animals by improving bacterial multiplication. White et al. (1973) reported that the addition of molasses (0-20%) to a straw-based diet significantly increases both DM and crude fiber digestibility. The CP digestibility was highly significantly correlated (P < 0.001) with ADG and SBW, but negatively correlated with ADF digestibility. This might be due to lignification of plant material that hampers rumen fermentation of roughage feeds.

Partial budget analysis
The growth of net return and ΔNR in a high level of CSH substitution for NSC could be partly attributed due to the difference in the sell- Regarding MRR, an increment was observed in T5 which might be due to the lower price of SH compared with the current high price of NSC. The current MRR result shows that each additional unit of 1 ETB (0.025 USD) per animal cost increment gain 0.5, 0.1 and 2 ETB profits for T2, T3 and T5 but lost −59.2 ETB in (T4), respectively. The higher negative value of MRR in T4 might be due to the addition of NSC in the treatment increased the total variable cost of the treatment and also lower selling price of lambs in T3 and T4. The current result was similar to Meaza (2012) for Tigray highland rams on supplemented with graded levels of mixtures of a sesame seed (Sesame indicum) cake and wheat bran marginal rate of return was decreased but the net return was inversed abovementioned parameter. As a result, based on the outcome of the current study, it is not necessary to spend additional expenses for high-price concentrate feeds T5 (417.98 g), CSH could be considered as an economically feasible supplement for growing lambs compared with T1, T2, T3 and T4.

CONCLUSION AND RECOMMENDATION
The current study revealed that inclusion of CSH in a native haybased diet helped to meet the CP maintenance requirement and for muscle development of lambs, indicating the need for supplementation of protein source feed for body weight gain. Therefore, based on the present result, supplementation of 417.98 g of CSH (T5) had increased the total DM and nutrient intake and carcass yield of lambs superior to sole NSC supplementation. In terms of economic returns and the biological performance of lambs, the dietary CSH supplementation in the treatment group (T5) was highly recommended. To confirm feasibility of the finding, it is vital to conduct an on-farm experiment under farmer's management for effective utilization of the results.

ACKNOWLEDGEMENT
The first author would like to acknowledge the Ethiopian Institute of Agricultural Research for funding this research.

CONFLICT OF INTEREST
The authors declare that they have no conflict of interest.

AUTHOR CONTRIBUTION
All the authors contributed in research conceptualization, experimental design, data collection and data curation, analysis and final draft manuscript writing.

Consent for publication
All authors agreed on the publication of this paper and assigned corresponding author responsible in charge for correspondence during manuscript publishing.

DATA AVAILABILITY STATEMENT
Data is available with the corresponding author upon request