The inventory of camel feed resource and the evaluation of its chemical composition in south‐east rangelands of Ethiopia

Abstract Background Evaluations of available camel feed nutritive value are relevant to generate evidence on further camel feed improvements and find out the components to be supplemented. Objective This study aim to evaluate seasonal variations on chemical composition of selected camel feed in semi‐arid regions of south‐east Ethiopia. Methods Samples of edible portions from 15 browse species were collected during the dry and wet seasons, and their chemical compositions were analysed. Results The crude protein (CP), in vitro dry matter digestibility (IVDMD) and relative feed value (RFV) of evaluated browse species were higher (p < .01) in wet season than the dry season except for Acacia asak, Ipomoea donaldsonii and Acacia mellifera. Nonetheless, the neutral detergent fibre (NDF) and acid detergent fibre (ADF) contents were higher (p < .01) in the dry season except for A. asak and I. donaldsonii. Moreover, A. asak, I. donaldsonii and A. mellifera were the browse species with higher (p < .01) RFV, IVDMD and CP but lower NDF and ADF during the dry season than the wet season, and these species are qualified as good‐quality forage. Thus. Barleria spinisepala were better to use in both seasons, but browse species like I. donaldsonii and A. asak in dry season were ranked as best quality roughage. Conclusion At richest level on vital components (CP and fibres), these species can serve as well ruminant diets, like for camel. Further investigations based on animal trials are needed in order to confirm the classification standards of feed quality used in this study.


| INTRODUC TI ON
Animal feed resources are mainly derived from the natural pasture in Ethiopia (CSA, 2012). Agroecology and land use/cover types determine a potential contribution to palatable livestock feed resources (Bediye et al., 2001;Madsen et al., 2008). This potential contribution showed a declining trend in the last few decades, due to the expansion of cropland cover and invasion of unpalatable woody vegetation into the grasslands (Gebremedhin et al., 2009).
Thus, potential contributions of grassland cover to livestock feed resources have been gradually decreasing and became less indispensable in arid and semi-arid regions (Thornton, 2010). According to Konuspayeva (2007), the livestock feeds derived from bush/ shrubland vegetation cover are gradually increasing in arid and semi-arid regions of Ethiopia. Alemayehu et al., (2017) indicated that livestock feed quality is the major bottleneck for livestock production in semi-arid regions of Ethiopia. Following the results of McDonald et al., (1995), the quality of forage has been determined by its chemical and biological nutrients, which directly influence the digestibility and feed intake; consequently, milk and meat productivity were affected by consuming low crude protein (CP) and high contents of fibre. The seasonal variation of browse species, nutritional composition and forage quality is the primary concern (Chalchissa et al., 2014). As an example, the CP content of browse species was higher in the wet season and dropped down in the dry season (Melaku et al., 2010;Yayneshet et al., 2009). Quality and availability of browse species vary with agroecology, rainfall and temperature patterns, which limit biomass production and nutritive value (Melaku et al., 2010). Main browses species of natural pasture are useful for animal feeding in changing eco-environments of semi-arid regions. The primary contributors of browse feed resources such as bush, shrub and woody vegetation remain evergreen throughout the year with better forage quality when grasses dry out (Aregawi et al., 2008).
Temperature and rainfall affect forage quality through ecophysiological changes of plant species or direct influence on feed digestibility (Ball et al., 2001). Plants successively undergo ecophysiological changes in response to heat stress, low precipitation and soil water scarcity. Accordingly, climatic extremes lead to slow rate of plant maturation and decrease plants' water content and the leafto-stem ratio (high lignin and cell wall contents) which strongly affects feed digestibility (Collins, 1988;Rivera & Parish, 2010;Stone et al., 1960). High temperature increases plants' lignification process and decreases ruminants' voluntary feed intake as response to thermoregulation mechanism.
Studies related to the environmental effects on forage quality and availability in tropics in general and in Ethiopia in particular have been conducted mainly on few legume species such as Desmodium ovalifolium (herbaceous legume) and Calliandra calothyrsus (shrub legume) (Chou et al., 2008;Dumont et al., ,2014Dumont et al., , , 2015Hidosa & Guyo, 2017).
Furthermore, Madalcho et al., (2019) identified 50 species of trees and shrub plants that have potentially been used as camel feed resources in east and south-east rangelands of Ethiopia. Consequently, there are minimal information produced about quality parameters and nutritional composition of browse feed resources in arid and semi-arid condition. Melaku et al., (2010) have evaluated the quality parameters of few browse species that have potentially be utilized by dromedaries.
However, dromedaries tend to browse a wide variety of plant species in open rangeland condition (Mirkena et al., 2018).
Moreover, Moges et al., (2016) attempted to address the gap on low-quality camel feed resources in the rangeland through feed supplementation on top of free-ranging. The finding focused on feed supplementation of concentrates and urea-treated roughage on top of free-ranging because the available forage species have been depleting in quality. However, the findings lack the assessment and evaluation of seasonal nutritive variations of feed resources that have been potentially utilized by browser livestock animals in semi-arid region. Therefore, assessments of available camel feed resources and evaluation of its nutritive value are relevant to generate evidence on further camel feed improvements and find out the components to be supplemented to cope up with the impacts of declining feed quality. This study aim to evaluate seasonal variations on chemical composition of selected camel feed in semi-arid regions of south-east Ethiopia.

Hypothesis Seasons and browse species can influence chemical com-
position and forage quality of camel feed resources.

| Description of the study area
The study was conducted in the semi-arid area of East Guji Zone,

south-eastEthiopia.Itislocatedbetween4°38′55″Nand5°33′7″
Nlatitudeand39°9′25″Eand39°58′37″Elongitudeandcover about 742,644 ha. The locations are categorized as a pastoral and agro-pastoral region that belongs to the semi-arid lowland agroecological zone. The altitude of the study districts ranges between 1,370 and 1,650 m above sea level (m.a.s.l). The annual temperature of the area varied from 24 to 30°C with a mean annual rainfall of 526.75 mm. The pattern of the rain is bimodal with the primary wet season (Ghana) contributing about 60% of yearly rainfall which extends from March to May, while dry season ranges from December to February (Abate, 2016).

| Samplingproceduresanddesign
Three study districts were randomly selected from the five pastoral and agro-pastoral districts of East Guji Zone based on the drawing lots procedure indicated in Gomez and Gomez (1984). The study considered two traditionally classified geographical locations, namely, Golba (covers the altitude below 1,450 m.a.s.l) and Dida (the altitude up to 1,650 m.a.s.l). Three kebeles (the smallest administrative unit of Ethiopian Government) were randomly selected from each location. Accordingly, Hadhessa, Qoratti and Siminto kebeles were selected from Dida location, and Kalada, Gofi Ambo and Nura Umba were selected from Golba study location.

| Focusgroupdiscussion
Focus group discussion (FGD) was conducted to obtain a general overview of the camel feed resources and identify potential browse species that camel can have access in both dry and wet seasons.
Thirty-six participants from both sexes representing all groups of the community, locations, education level and the household heads with above 35 years were selected to identify available camel feed resources in the study area following Geilfus (2008) procedure, because camel raisers with more than 35 years old perceived as better in camel feeding experience. Six FGDs, one in each kebeles consisting of eight participants, were employed. The FGD were selected based on their experiences on camel raising, feeding and moving with a camel in the rangeland. Furthermore, the discussion was conducted with the local language (Afan Oromo) as the study locations are solely Oromo's ethnic group.

| Descriptionoftheselectedbrowsespecies
The choice of collected browse species (Table 1) depends on its availabilities in the area, contributions in camel feeding and preference by camels as indicated by FGDs.

| Samplecollectionandpreparation
The browse species samples were collected in the wet (March to May) and dry (December to February) seasons. This study is based on the identified flora of southern Ethiopian rangeland by Gemedo- Dalle et al., (2005). All edible portions of collected browse species samples were labelled and dried for the analysis of chemical composition. Samples of the same feed type were bulked together on a seasonal basis and then thoroughly mixed and subsampled following the method indicated by Herrman (2001) and Feeding-Stuffs (1988). The edible, healthy portions were sampled from 12 representative plants of the selected browse species, weighed immediately after collection with digital sensitive balance and oven-dried at 65 C for 72 hr. The dried samples were ground pass 1 mm Wiley sieve size and used for determination of chemical composition and in vitro dry matter digestibility (IVDMD). The sieved samples were kept in airtight containers pending analysis for chemical composition.

| Determination of nutritive value of browse species
Feed samples were analysed for dry matter (DM), ash and CP according to the standard procedures for feedstuffs analysis (AOAC, 1990). Neutral detergent fibre (NDF) and acid detergent fibre (ADF) and acid detergent were analysed by the method of Van Soest et al., (1991). The method of Tilley and Terry (1963) as modified by Van Soest and Robertson (1985) Reid et al., (1952), with 60% digestion coefficient. The digestible dry matter (DDM), dry matter intake (DMI) and relative feed value (RFV) were determined using the index recommended by Rivera and Parish (2010), Jeranyama and Garcia (2004) and Kiraz (2011). Relative feed value (RFV) = DDM × DMI∕1.29. The RFV was compared with full bloom alfalfa (reference feed), which is assigned an RFV of 100 (Rivera & Parish, 2010;Undersander et al., 2002).

| Statistical analysis
The analysis of variance (ANOVA) was conducted using the gen- where Y ijk is measurements of feed chemical composition in ith feed species at jth season; μ is the fixed effects of feed chemical composition in ith feed species at jth season; FS i is effects of feed species; S j is fixed effects of season; (S*FS) ij is interaction effects of season and feed species; and e ijk is residual.
Mean separation was employed using Duncan multiple range tests (Duncan, 1955). Moreover, Pearson product-moment correlation was used to measure the association between weather condition and camel physiological response (Chee, 2013). All results were presented as means ± standard error of means (means ± SE).

| Inventory of camel feed resources in southeastern Ethiopia
Focus group discussion have identified about 49 browse plant species that can potentially be utilized by camel during the dry and wet seasons in eco-environments of the study area. The identified species of feed resources were grouped into 24 families (Table 2).
Trees and shrubs such as Acacia asak, Acacia lahai, Acacia oerfota,  (Aregawi et al., 2008;Lu et al., 2012;Yagil, 1982). Cattle and sheep species do not readily utilize most of the preferred plant species by camel because they are bitter and thorny (Lu et al., 2012).

| Totalashandorganicmattercontent
The total ash content of the analysed forage showed a significant Dry matter intake (DMI) = 120∕NDF (basis % DM) .

| Crudeprotein
In the dry season, the CP content of the selected browse plant spe- All evaluated browse species CP contents at both seasons were higher than the required minimum level (7%) for ruminant feed intake and optimum rumen microbial functions (P. J. Van Soest, 1994). The minimum CP content of ruminant feed resources usually required for lactation and growth is 15% on a DM basis (Norton, 1982
According to Mosi and Butterworth (1985) criteria, A. mellifera, B. spinisepala, G. tembensis and I. donaldsonii, which contained >50% IVDMD on DM basis in both seasons, are qualified as a good-quality forage species the entire year. However, the IVDMD content of T. vogelii, E. divinorum and L. rivae were <50% in both seasons, which categorizes them as a low-quality roughage. The mean IVDMD observed in this study was comparable with the report of Mlay et al., (2006).
The IVDMD observed in this study was lower than value reported by Tufarelli et al., (2010) (56%). This variation might be due to the relatively higher contents of cell wall components in analysed browse species of this study. According to Warne et al., (2010) and Rust and Rust (2013), some plant species were adapted and produced quality forage under hot and dry conditions; some are adapted to cooler and moist conditions and powerless to maintain their nutritional quality. The forage plants adapted and survived under high ambient temperature conditions, and water scarcity was low in quality (Bellard et al., 2012).
Moreover, Sejian et al., (2016) reported that forage species adapted to hot and dried environmental conditions were more likely to have lower CP concentrations and high cell wall (lignin, cellulose and hemicellulose) contents, which qualify them as low quality and IVDMD.

| Etherextract
The EE value observed in this study varied from 2.71% in B. rotundifolia to 13.55% in A. mellifera during the dry season and varied from 4.91% in I. donaldsonii to 15% in G. tembensis during the wet season.
The crude fat contents varied significantly (p < .01) thought forage species. For most considered browse species, the EE content varied insignificantly (p < .01) with season. However, the EE content of B. rotundifolia, G. tembensis and R. ruspolii was found to be the highest in wet season. Mean EE content observed in this study at both seasons was higher than the mean, 1.5% and 3.3% of tropical grasses and legume browse trees, respectively, and most similar to concentrates (9.7%) as cited by Mlay et al., (2006). The crude fat contents of all the investigated browse species in both seasons fell within the range of 1%-20% EE on DM basis, the amount often found in livestock feedstuff (Galyean, 2009). was found to be the highest at the wet season (Table 3).

| Totaldigestiblenutrient
According to Rivera and Parish (2010) Ball et al., (2001), the quality of animal feedstuff is often affected by the differences in forage species, environmental temperature and maturity stage. However, it seems that E. divinorum in both seasons and A. bussei in the dry season are not preferred as roughage because their RFV fell to third level between 87 and 102 according to Rivera and Parish (2010).

| Relativefeedvalueofthebrowsespecies
The CP and IVDMD had shown a significant positive correlation with RFV (Table 4). Oppositely, the RFV and IVDMD of the browse species were negatively correlated with NDF, ADF and lignin contents of analysed feedstuff. Forage quality improved with increasing CP contents of feedstuffs (Rivera & Parish, 2010

| CON CLUS ION
This study showed that chemical composition and feed quality of examined browse species were found to be significantly varied ( animal trials are needed in order to confirm the classification standards of feed quality used in this study.

ACK N OWLED G EM ENT
We thank Mr Melaku for his assistance in feed chemical composi-

CO N FLI C T O F I NTE R E S T
No potential conflicts of interest to declare.

A N I M A L WELFA R E S TATEM ENT
The authors confirm that the ethical policies of the journal, as noted on the journal's author guidelines page, have been adhered to. No ethical approval was required as this is an original data with no animals used for scientific purpose.

PEER R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1002/vms3.471.

DATA AVA I L A B I L I T Y S TAT E M E N T
The authors declare that all data supporting the findings of this study are available within the article and its supplementary information files. Abbreviation: NS, non-significant **p < .05 ***p < .01