Valuation of livestock population and national feed security to enhance livestock productivity in Ethiopia

Abstract

Background

Feed is a major input in the livestock industry and covers about 60%–70% of the total cost of producing meat, milk and eggs. Inadequate feed supply in terms of quality and quantity leads to lower production performance in livestock. However, the development of an appropriate livestock production strategy through efficient utilization of existing feed resources could raise the production and per capita consumption of livestock products. Efficiency of feed resource utilization can be measured as the ratio between input to production activities and output (e.g. kg of protein used per unit of meat, milk and eggs produced or hectare of land used per unit of milk produced).

Methodology

This study was designed with the objective of evaluating the livestock population and national feed security to enhance livestock productivity in Ethiopia. To achieve this objective, data were collected from the websites of the Ethiopian Central Statistical Agency from 2007 to 2021, FAO publications and websites, books and journals. The data obtained on different feed resources, livestock population and livestock feed requirement and balance were entered into an MS Excel spread sheet (Excel, 2010) and analysed using the general linear model (PRO GLM) procedure of SAS (2014) and multivariate analysis of covariance.

Results

The study results revealed that the livestock population had increased from 58.31 million tropical livestock units (TLU) to 81.10 million tropical livestock units (TLU), and the emission of entericCH4 had increased from 2511.08 Gg/year to 3661.74 Gg/year from 2008 to 2021. The study results also showed that the major available feed resources for ruminants are natural pasture and crop residues, which account for 56.83% (87.56 × 10⁶) and 37.37% (57.57 × 10⁶) of total feed production in the country, respectively. The contribution of concentrate and improved cultivated pasture and feed from permanent crops used as feed sources is very insignificant (3.05% and 1.96%, respectively). The estimated quantity of these feed resources was sufficient to meet the livestock feed requirement in the country in terms of dry matter (DM), digestible crude protein (DCP) and MEJ, which estimated about 153.31 × 10⁶ t, 4.56 × 10⁶ t and 1203.97 × 10⁹ MJ DM, DCP and MEJ, respectively. The estimated livestock feed requirements were 134.62 × 10⁶, 4.52 × 10⁶, and 918.83 × 10⁹ in DM, DCP and MEJ, respectively. The supply covered about 114.33, 100.04 and 131.33% of the DM, DCP and MEJ total annual feed requirements of livestock in the country. Hence, the current feed surplus obtained on feed requirements of ruminants and equines can support the nutrient requirements of 500 × 10⁶ broilers, about 5 × 10⁶ bulls, about 50 × 10⁶ small ruminants or 3 × 10⁶ crossbred lactating dairy cows, yielding 10 L of milk per day.

Conclusions

The findings of study indicated that natural pasture and crop residues cover a major proportion of the annual feed supply in the country. Therefore, proper grazing management, feed conservation practices, improving grazing land vegetation through clearing invasive species, replacing the grazing land with an improved grass and legume mixture, effective collection, conservation and proper utilization of crop residues, and other alternative options such as the use of chemical, physical and biological treatments to improve the nutritive value of fibrous feed should be practiced. More effective extension services and farmer training are also required to increase feed productivity and, hence, human development.

INTRODUCTION

Ethiopia has the most significant livestock population in Africa, according to the latest livestock census figures done on the African continent (Tanzania's National Bureau of Statistics [NBS], 2020). The country has about 58.43 million cattle, 6.91 million small ruminants, 8.15 million camels, 9.98 million equines and 1.38 million chickens (Central Statistical Agency [CSA], 2021) in the tropical livestock unit (TLU). Out of the total population of cattle, about 98.2% were unimproved indigenous breeds, 1.62% were crossbreeds and 0.18% were exotic pure cattle breeds (CSA, 2021), and about 82.5% were indigenous chickens, whereas the rest were 9.93% and 7.52% hybrid and exotics, respectively, which are mainly kept by smallholder farmers in scavenging environments (CSA, 2021). Livestock sector contributes to food production, security, improved crop yields, cash income generation for rural and urban people, fuel and transportation and the manufacture of items with added value that can have cascading consequences and increase the demand for services (Gebreegziabher, 2010). It also plays a significant role in the national economy, contributing about 40% of total GDP, 20% of agricultural GDP and 20% of foreign currency (World Bank, 2017).

Livestock and poultry are considered an industry that is able to convert inedible waste products and natural pasture (input) into valuable products or output (meat, milk and eggs) that can be consumed by humans. They can also contribute a significant source of foreign exchange earnings and provide essential goods and services to the national economy and the way of life for many Ethiopians (Belay & Negese, 2019). Globally, it is anticipated that demand for animal products will increase to around 70% in 2050 as a result of urbanization, rising income and an expanding global population (Food and Agriculture Organization of the United Nations [FAO], 2014). Despite this, the sector is severely constrained by reasons, such as the poor genetic composition of local animals, poor diet, the environment and inadequate veterinary services (Lamy et al., 2012). Animal productivity in Ethiopia is poor, even below the values recorded for the majority of the countries in eastern and sub-Saharan Africa (Gebreegziabher, 2010). According to Aynalem et al. (2011) and Getahun (2012), low productivity is primarily caused by ineffective nutritional and management practices, low genetic potential of livestock, a high incidence of disease and parasites, poor access to extension and credit services and a lack of knowledge to improve animal performance (Birhan & Adugna, 2014; Defar, 2018).

Ethiopia has an entire land area of about 113.624 million hectares (World Data Atlas [WDA], 2021). Grazing land is cover about 62% of total land mass and support livelihood of about 12–15 million pastoral and agropastoral communities (Coppock, 2011). According to FAO (2018), about 59.13 million hectares, or 61%–65% of the nation's total land area, are found in the lowlands of the country, which are located below 1500 m.a.s.l. and are home to about 26% of cattle and around 12% of the human population. The lowland areas have always been characterized by low human population concentrations and very erratic and unpredictable rainfall. Because of this, it is challenging for humans to live in pleasantly dense ways in the lowland regions of the country.

Animals depend mainly on natural pastures and crop residues for their feed requirements because they provide more than 90% of the livestock feed, yet the natural pastures are generally poorly managed due to overstocking, resulting in severe land degradation, losses of valuable species and dominance by unpalatable species (Flintan et al., 2013; Bizelew et al., 2016). Concentrate and cultivated fodder supplements contributed insignificantly to livestock feed in Ethiopia. Hence, by creating a livestock feed balance, it is possible to examine the needs of the existing livestock population as well as the feedstock that is readily accessible (Weisberg et al., 2006). After accomplishing this, it is feasible to pinpoint production-level constraints and calculate the amount of feed needed to boost output. Depending on legislative requirements and the amount of precision necessary, livestock feed balancing can be conducted at a local, regional or across-the-country level (Getahun & Tegegn, 2019). A feed balance, at its most basic level, is a comparison between the quantity of utilizable feed available at any given moment (supply) and the needs of livestock at any given time (demand) and so offers a ‘snapshot’ of the present situation (Getahun & Tegegn, 2019). A feed balance can help identify types of feed materials that may be needed if gaps are found, as well as possible feed shortages to fulfil the rising demand for food. Instead, if a feed balance shows a feed surplus, it may be used to calculate the potential increase in livestock output that the excess might sustain.

Although various research and development activities have been carried out in the past, these research projects have not led to productivity. Therefore, novel programmes are needed to improve herd and flock production and ruminant sector contributions to meet human needs. In order to achieve development goals, it is crucial to understand agricultural systems, prioritize issues and address unique local or regional obstacles. Ethiopia's livestock industry struggles due to inadequate feed quality, which accounts for 60%–70% of production costs (Birhan & Adugna, 2014; Defar, 2018). Suresh et al. (2012) stated that proper evaluations as well as efficient management of the feed supplies available are essential to increasing animal production and the socioeconomic growth of the nation. Therefore, this study was designed with the objective of evaluating the livestock population and feed security of the country to enhance livestock productivity.

MATERIALS AND METHODS Geographical location of Ethiopia

Ethiopia is a landlocked country in the Horn of Africa, bounded to the north by Eritrea, to the west by Sudan, to the south by Kenya and to the east by Somalia and Djibouti. It lies within the tropics between latitude of 3°24′ and 14°53′ N and 32°42′ and 48°12′ E. The country has about 113.624 million hectares of land, which possesses about 29.35% grazing and range land, 23.68% shrubs and bushes, 19.10% for cultivation, 16.62% forest land, 9.41% bare rock soils and settlement, 1.10% wet land and 0.72% water bodies (Moges et al., 2010). However, these land use systems changed during the last decade: The following land use systems, such as cultivation, water bodies, bare rock outcropping and settlement, increased by around 33%, 17%, 21% and 191%, respectively, whereas grazing and range land, forest land, shrubs and bush and wet land decreased by 18%, 19%, 7% and 32%, respectively (Regasa et al., 2021). Agro-ecologically, the upper highlands are suitable for sheep production, whereas the area within 1500–3000 m above sea level is suitable for both livestock and crop production. The lowland areas with elevations of less than 1500 m above sea level, for their part, have high potential for perennial crop and livestock production.

The study design and data collection methods

The model and study methodology used to assess the national feed availability, animal nutrient requirements as dry matter (DM), metabolizable energy (ME) and digestible crude protein (DCP), and feed balance trends from 2007 to 2021 were based on FAO (1987, 2012). The data on feed resources, such as grazing and range land, forest land, shrubs and bush land and wet land, were collected from journals, books and the FAOSTAT website. However, the cultivated land during Mehar and Belg and commercial land, fallow land and improved cultivated forage were collected from the record of the CSA Bulletin published from 2007/08 to 2020/21. Livestock species in number, age and sex were also collected from the CSA Bulletin published from 2007/08 to 2020/21. The national crop production, land use and livestock population data were used to evaluate feed security. The design of the study used to evaluate the national feed secrete is described below (Figure 1).

FIGURE 1. Study design.

Feed type and classifications

Feed can be classified as roughage and concentrate, or conventional and nonconventional, in Ethiopia. Roughage feed includes natural pasture obtained from shrub and bush land, grazed from grazing land, forest land, fallow land, wet land, crop aftermaths, cultivated pasture, cereal, pulse and oil seed crop residues, vegetable, root and tuber crop wastes and feed from permanent crops such as banana leaves and stems, enset leaves and stems and sugarcane tops. Generally, all feeds that have greater than 18% crude fibre, greater than 35% cellulose, and less than 60% TDN are categorized as roughage classes. Fruit and cash crop by-products, also called unconventional feed, sometimes play a great role in filling the feed deficit gap. Conserved roughage such as hay and silage were not included in this study because they are part of natural and cultivated pasture. Residues from cash crops like coffee, cotton, hops, chat, fruit land and other permanent tree crops, as well as leaves of tea dropped during harvesting, are all categorized as miscellaneous fodder sources in this paper. Concentrate feeds are types of feed that contain less than 18% crude fibre, less than 35% cellulose, and more than or equal to 60% TDN (Kartik et al., 2022). It includes broken and immature grains obtained during processing cereal grains, pulse hulls, oilseed cakes and molasses. Animal origin by-products, such as meat meal, fish meal, meat and bone meal, blood meal and dairy and poultry by-products, are part of concentrate, but due to the unavailability of the data, these feeds were not included in the current study. Mineral and vitamin supplements are also a part of concentrate feed but are not considered in this paper due to a lack of available data. According to CSA (2013), about 0.55% of the yield of cereal grain can be used as animal feed, and it is also difficult to exclude the grain from animal feed due to the increasing demand for commercial broiler and layer poultry production because more than 50% of the ration ingredient for commercial poultry production is grain, mainly maize.

Estimation of the quantity of available feed resource

The quantity of feed resource trends from 2007/08 to 2020/21 in Ethiopia was estimated using secondary data collected from CSA, FAO Bulletin, FAOSTAT DATA and articles published from 2007/08 to 2020/21. The amount of crop residues and by-products that are used as sources of animal feed was estimated using the conversion factors and multipliers developed by FAO (2012) and other researchers. The multiplier developed for wheat, barley and teff straw is 1.5 per unit weight grain yield, whereas the factors for maize and haricot beans are 2.5 and 1.2, respectively (Adugna, 1990; FAO, 2012) (Table 1). Crop aftermath grazing potential was estimated using a mean of 0.5 t/ha (Table 2). The quantity of crop residue on the basis of DM available and those actually available for livestock consumption was estimated by deducting 10% from the total estimated yield of t DM/ha, according to Tolera and Said (1994).

TABLE 1 Conversion factors used for estimation of the amount of cereal grains and crop residues used for animal feed.

Crop	Residue	CF	Sources
Teff	Straw	1.5	Adugna (1990)
Wheat	Straw	1.5	Adugna (1990)
Barley	Straw	1.5	Adugna (1990)
Barley	Grain	0.55	CSA (2013)
Oat	Straw	1.7	Adugna (1990)
Rice	Straw	1.3	FAO (2012)
Maize	Stover	2.5	FAO (2012)
Maize	Grain	0.55	CSA (2013)
Millet	Straw	2	FAO (2012)
Sorghum	Stover	2.5	FAO (2012)
Sorghum	Grain	0.55	CSA (2013)
Pulse crops	Haulms	1.7	FAO (2012)
Sugarcane	Tops	6.0 t/DM/ha/year	FAO (1987)
Sweet potato	Veins	0.3	FAO (1987)
Others root and tubers		0.3	FAO (1987)
Noug	Aerial part	4	FAO (1987)
Linseed	Straw	4	FAO (1987)
Vegetables	Vegetable waste	0.25	FAO (1987)
Banana	Leaves and other residue	8.0 t/DM/ha/year	FAO (1987)
Enset	Leaves and other residues	33 t/DM/year	FAO (2018)

Abbreviations: CSA, Central Statistical Agency; DM, dry matter.

TABLE 2 Conversation factors used to estimate biomass yield tones dry matter (DM)/h of animal feed from different land use type.

land use type	Conversion factors	Sources
Grazing land and range land	1.0 t DM/h/year	Alemayehu (2006)
Crop aftermath grazing	0.5 t DM/h	FAO (1987)
Fallow land grazing	1.0 t DM/h	FAO (2012)
Forest and wood land	1.5 t DM/h	FAO (2012)
Bush and shrub land	1.0 t DM/h	Alemayehu (2006)
Wet land	1.0 t DM/h	FAO (2012)
Cultivated forage	5 t DM/h	FAO (2012)
Miscellaneous sources	1 t DM/h	FAO (2012)

The conversion of by-products like oilseed to oil seed cake, cereal and pulse grain to their milling by-products and permanent crop residues were determined according to FAO (2012) and incorporated into the feed resources of each year. Feed resources from land use and land cover, such as grazing and range land, forest and wood land, bush and shrub land, crop aftermath grazing, fallow land grazing, wet land, cultivated forage and miscellaneous tree crops, were calculated according to the conversion factors developed by Alemayehu (2006), FAO (1987) and FAO (2012) Table 2.

Methods of calculating available nutrients DM, ME and CP from feed resources

The total available DM of feed resources was estimated by summing the DM of different feed resources. The total ME and crude protein (CP) were calculated by multiplying the available DM of feed resources by the ME and CP contents. The values of metabolizable and CP content of different crop residues and by-products were taken from various sources (Bogale et al., 2008; Feyisa et al., 2021; Feedipedia, 2023; Gudina et al., 2015; Gashaw & Defar, 2017; Getahun et al., 2018; ILRI, 2011; Seyoum et al., 2007). The green weeds and dropped grain during harvesting were taken into consideration, and the conversion factors of 76.5 g/kg and 8.23 MJ/kg CP and ME, respectively, used for natural pasture were used to estimate the values of DCP and ME of fodder obtained from aftermath grazing. The CP and ME of natural pasture were taken as 85 g/kg and 8.6 MJ/kg for the wet season and 51 g/kg and 7.1 MJ/kg for the dry season (Adugna et al., 2012; Keba Habtamu et al., 2013), and the average values of CP and ME of 76.5 g/kg and 8.23 MJ/kg were used to estimate the CP and ME produced for feed resources obtained from land use and land cover, such as grazing and range land, forest and wood land, bush and shrub land, fallow land grazing, wet land, cultivated forage and miscellaneous tree crops. The values of CP and ME of banana leaves were 77 g/kg and 8.7 MJ/kg, and 74 g/kg and 8.9 MJ values were taken for enset leaves and stem. The CP and ME values of cultivated forage were estimated based on the mean values of 142 g/kg DM and 9.13 MJ/kg DM, respectively, reported for different cultivated forage species (Feyisa et al., 2021). The DCP of the feed was calculated from CP according to the formula suggested by FAO (1986) (DCP = 0.929 × CP = 3.52, where DP is expressed in % DM). The overall estimated mean of DCP and ME of the entire feed sources of the current study was 29.71 g/kg and 7.56 MJ/kg, respectively.

Methods of calculating animal feed requirements Dry matter, metabolizable energy and digestible crude protein requirements

Before calculating the annual feed DM, DCP and MEJ requirements of different livestock species, they were converted to TLU. The TLU of adult cattle is 250 kg, and others are converted to TLU by conversion factors of 0.7, 0.1, 0.1, 1, 0.6, 0.8, 0.9 and 0.02 for cattle, sheep, goat, camel, donkey, mule, horse and chicken, respectively, developed by Kear (1982). The DM intake per TLU was estimated at 2.5% of the body weight (BW) or 6.25 kg/day (FAO, 2018). The net energy requirements of cattle and sheep were estimated using the equation developed by the IPCC (2006) Guidelines for National Greenhouse Gas Inventories. The total net energy was calculated based on the sum of maintenance, growth and lactation requirements. The metabolizable energy requirement (MER) was calculated according to an equation developed by CSIRO (2007): ME MJ/kg = NEm/km + NEg/kg + NE1/k_l. k is the efficiency with which the net energy used for maintenance, growth and lactation is related to the average forage ME. These values were calculated based on the following equations: metabolizability (q_m) = ME/GE; maintenance (k_m) = 0.35q_m + 0.503; growth (k_g) = 0.78q_m + 0.006; lactation (k_l) = 0.35q_m + 0.42. These predictions were done based on the mean ME concentration of all feeds (ME/GE). Metabolizability (q_m) was calculated by dividing the estimated mean ME content of the feed (7.56 MJ/kg) by its GE content, making the assumption that the mean gross energy value of feeds is 18.4 MJ/kg DM (Mac Donald et al., 2010). Therefore, the calculated metabolizability (q_m) used to determine the metabolizabled energy requirement of animal was 0.41. The MER for oxen assigned to grazing and daily work for 4 h was calculated according to an equation developed by the IPCC (2006).

The digestible protein requirement was calculated according to the formula suggested by FAO (1986). Assuming that 50% of TLU computes DCP for growth purposes according to FAO (1987), about 100 g of DCP was added on top of maintenance for production purposes according to FAO (1986) recommendation. A mean value of 0.21 g of DCP was used to estimate the total digestible crude protein requirement (DCPR) per TLU per year. The daily goat ME requirements for maintenance and growth were estimated based on equations of 0.452 (BW)^0.75 and 27.7 MJ/kg BW gain, respectively (Salah et al., 2014). The NE requirements for activity were calculated according to IPCC (2006) for goats <6 months assigned to moderate grazing, taken as zero; for goats of age >6 months and <1 year, it was taken as 25% of maintenance, and for adult goats, it was 50% of maintenance.

Daily camel MER for maintenance was calculated based on the suggested equation by FAO (2018): MJ = 0.435 (BW)^0.75 (Wardeh, 1997; Nagpal, 2016) and adding 40% on top of maintenance for camels assigned to grazing and daily work for 4 h. Similarly, the results of MEJ calculated for equines by using the equation suggested by FAO (2018) to calculate digestible energy (DE) of horses are as follows: DE = 33.3 kcal/kg BW (0.139427 MJ/kg BW) and the ratio of ME to 0.82 (NRC, 2007; Ralston, 2016) and NRC (1989) = [0.975+(0.021 × BW in kg)] × 4.187 for donkeys and mules, and DE was converted to ME by multiplying by a factor of 0.82.

Method of estimating enteric CH₄ emission

The enteric CH₄ emission from livestock was estimated based on equations developed by IPCC (2006). The value for each livestock species was calculated by multiplying the livestock number of head with the average emission factors associated with a specific animal category and summed up according to equation developed by IPCC (2006) then the values were expressed in Giga gram per year (Gg/year): [Image Omitted. See PDF]

where emissions (CH₄) (T) is the enteric CH₄ emissions for specific animal category T, Gg CH₄/ year; EF (T) is the emission factor for each specific livestock category T, kg CH₄ /head; N(T) is the number of heads for each specific livestock category in the country; T is the category of livestock. [Image Omitted. See PDF]

where total CH₄ enteric is the total CH₄ emission from enteric fermentation, Gg CH₄ /year; E_i is the emissions for the ith livestock categories.

The live weights of each livestock category explained to calculate enteric CH₄ emission for developing countries in IPCC (2006) were adjusted to TLU.

Data analysis and managements

Data on natural pasture from various sources, including grazing areas, forest land, wetlands, cultivated pastures and residues from crops, spanning from 2007 to 2021, were compiled from the CSA and additional references. These, details were organized into an MS Excel spreadsheet (Excel, 2010). Cultivated land and yields of cereal, pulse, oilseed, vegetables, root and tuber crops were analysed using the general linear model (PRO GLM) procedure of SAS (2014) and multivariate analysis of covariance (MANCOVA). However, crop aftermath grazing was not considered for the Belg season. If there was a significant difference between means, the Tukey honest significance test at α < 0.05 was considered a significant difference to adjust the mean separation. The model for the analysis was as follows: Yijkl = μ + Ci + Yj + Sk + εijkl, where Yijkl are the response variables; μ is the overall mean (intercept); Ci is the effect of crop type; Yj is the effect of year; Skis the effect of season; εijkl is the random error. The yield of unconventional feed used for animal feed was analysed using the same procedure, but seasons were excluded from the model. Yijk = μ + Ci + Yj + εijk, where Yijk is the response variables; μ is the overall mean (intercept); Ci is the effect of feed type; Yj is the effect of year; εijk is the random error. The available data on forage biomass yield obtained from grazing and range land, forest land, bush and shrub land, miscellaneous sources, wet land and fallow land grazing from 2007/08 to 2020/21 were subject to statistical data analysis (SAS, 2014) and MANCOVA. The model for the analysis was as follows: Yijk = μ + Li + Yj + εijk, where Yijk are the response variables; μ is the overall mean (intercept); Li is the effect of land use type; Yj is the effect of year; εijkl is the random error.

The balance was also analysed by using the general linear model (PRO GLM) procedure of SAS (2014) for MANCOVA after data on DM, metabolizable (ME) and DCP obtained from various sources were summarized for the years 2007/08 to 2020/21. The model for the analysis was as follows: Yijkl = μ + Ai + Yj + εijk, where Yijk are the response variables; μ is the overall mean (intercept); Ai is the effect of livestock type; Yj is the effect of year; εijkl is the random error.

RESULTS AND DISCUSSIONS Land use and land cover change in Ethiopia

The total area of grass and range land in the country was 61.75 million hectares (MoA, 2000). Currently, these grasslands and rangelands are diminished to 59.13 million hectares (FAO, 2018). Out of this, 19% is in the farming areas (with more than 600 mm of rainfall), and the rest is in the pastoral areas (FAO, 2018). The total forest land was estimated to be 18.52 million hectares (FAO, 2010), and now this quantity has diminished to 16.97 million (Regasa et al., 2021). Shrub and bush land was diminished from 26.40 million (FAO, 2010) to 24.54 million (FAO, 2018). However, cultivated land was increased from 20.05 million (Moges et al., 2010), and currently this area is at 30.48 million hectares (Regasa et al., 2021), and the wet land area was also increased from 0.802 million (FAO, 2010) to 1.23 million (FAO, 2018). The average area of fallow land was estimated at about 615,000 ha (FAO, 2021). These all-purpose land uses are the major feed resources for livestock in developing countries like Ethiopia because these land uses are covered with natural grass species, browsing vegetation and leaves and fruits from forest land, which are also additional sources of animal feed.

Estimated available feed resources

Livestock feed resources can be categorized as roughage, concentrates and by-products (which resemble roughage or concentrate) (Kartik et al., 2022). The roughage feed includes both green and dry roughage. Green roughage included forage obtained from different land uses, cultivated forage from residues of some permanent plants that have food value and are used for industrial processing (e.g. sugarcane), forages grown under miscellaneous tree crops and weeds from cereal crops (maize and sorghum), and stubble grazing, which also contains combinations of both green and dry roughages and dropped grain on the field during the harvesting period. Other crops, like vegetables and root and tuber residues, are also used as roughage sources. Cereals, pulses and oilseed crop residues are categorized under dry roughage, which are the major available feed resources in the highland areas of the country (Kartik et al., 2022). There are also agro-industrial by-products, such as coffee husk, coffee pulps and sugarcane bagasse, which are considered roughage feed due to their high cellulose, hemicellulose and lignin contents. Concentrate feed contains high nutritional value, especially in terms of DCP and energy. It includes grains and agro-industrial by-products obtained while processing cereal, pulse and oilseed crops (Kartik et al., 2022).

Estimated available feed from natural pasture and aftermath grazing

The quantities of dry matter yield (DMY), digestible crude protein yield (DCPY) and metabolizable energy yield (MEY) from grazing and range land, shrub land, forest land, aftermath grazing fallow land and wet land are presented in Table 3. The average land size used and forage biomass yields varied among land use types. Significantly (p < 0.001) higher values of forage biomass yield in terms of DM, DCP and ME were recorded for grazing and range land, followed by shrub land, forest land, aftermath grazing, miscellaneous sources, wet land and fallow land in decreasing order. There was also significant (p < 0.001) variation among natural pasture sources. The values of DMY were estimated based on the average value of the conversion factor of 1.0 t DM/ha/year developed by Alemayehu (2006) and FAO (2012) for poor grazing land. Contrary to this, Von Carlowitz (1984) recommended conversion factors of 3.0 t DM/ha for private grazing land, 1.8 t DM/ha for fallow land, 2.0 t DM/ha for communal grazing land and 1.2 t DM/ha for forest land, bush land and shrub land, which are higher than the current conversion factor. The DCP and ME of these feed resources were estimated according to the average values of CP and ME of 76.5 g/kg and 8.23 MJ/kg (Bayissa et al., 2022; Adugna et al., 2012), then the CP was converted to DCP using the formula (DCP = 0.929 × CP = 3.52, where DP is expressed in % DM) developed by FAO (1986). In agreement with these conversion factors, Bayissa et al. (2022) estimated the average value of CP obtained from wet and dry seasons at 76.6 g/kg for natural pasture grasses. However, the conversion used to estimate the CP value was below the findings of Feyisa et al. (2021), whereas the conversion factor used to estimate the ME energy value of natural pasture was consistent with Feyisa et al. (2021). The overall land used for grazing and forage biomass yields of DMY, CPY and MEJY in tones × 10³ were recorded in decreasing order from cropping year 2007/08 to 2020/21 (Figure 2). This is attributed to the increments in demand for crop cultivation that lead to a decrease in the grazing land size and forage biomass yield obtained from natural pasture.

TABLE 3 Land size in hectare, dry matter, crude protein and metabolizable energy yield for natural pasture and aftermath grazing.

Land use	LS × 103 h	DMY × 103 t	DCPY × 103 t	MEY × 106 t
Aftermath grazing	15,454.50c	7737.95d	278.57d	3382.90e
Fallow land	567.49e	567.49f	20.43f	4670.40e
Forest land	16,972.99c	16,972.99c	611.03c	13,9687.68c
Range land	34,549.75a	34,549.75a	1243.79a	284,344.44a
Miscellaneous	1117.64d	1117.64e	40.24e	9198.20d
Shrub land	25,561.88b	25,561.88b	920.23b	21,0374.25b
Wet land	1116.34d	1116.34e	40.19e	9187.46d
Average	13,620.08	12,517.72	450.64	94,406.48
SEM	237.90	214.57	7.72	1664.34
p-Value	<0.0001	<0.0001	<0.0001	<0.0001
CV	6.53	6.40	6.40	6.59

Note: Means of each parameter with different superscripts (a–f) within the same column are significantly different at p < 0.001.

Abbreviations: CV, coefficient of variance; DMY, dry matter yield; DCPY, digestible crude protein yield; LS, land size; MEY, metabolizable energy yield; SEM, standard error mean.

FIGURE 2. Forage biomass yield from different land use system; DMYFOL, Dry matter yield for forestland; DMYRL, Dry matter yield for rangeland; DMYSHL, Dry matter yield for shrubland; DMYAMG, Dry matter yield for aftermath grazing; DMYWL, DRy matter yield for wetland; DMYMSC, Dry matter yield for miscellaneous; DMYFL, Dry matter yield for fallowland.

Estimated available green fodder from permanent crops and cultivated pasture

Cultivated pasture and permanent crops such as banana, enset and sugarcane cultivated for human consumption will leave huge amounts of green fodder for ruminant production. Table 4 illustrates the annual forage biomass yield produced from permanent crops and cultivated pasture. The study revealed that there was significant variation in fodder obtained from cultivated pasture and permanent crops grown for human consumption. Significantly (p < 0.001) highest values of land size and forage biomass yield in DMY, DCPY and MEY were recorded for enset, followed by cultivated pasture, banana and sugarcane in decreasing order. DM, DCP and ME were estimated according to conversion factors of 6 t DM/ha for sugarcane, 8 t DM/ha for banana leaves and stem and 33 t DM/ha for enset leaves and corm (Adugna, 1990; FAO, 1987; FAO, 2018), and conversion values of CP and ME of banana leaves of 77 g/kg and 8.7 MJ/kg and 74 g/kg and 8.9 MJ for enset leaves and stem were used to estimate total DCP and ME yield from both permanent crops. However, these values were lower than the mean values of 94 g/kg and 13.50 MJ/kg of banana stem and leaf CP and ME, respectively (Feedipedia, 2023) and the mean values of 101.5 g/kg of enset corm and leaf and 120 g/kg of banana leaf CP (Feyisa et al., 2021). The mean values of the conversion factor of 27.7 g/kg DM and 7.62 MJ/kg were used to estimate the DCP and ME biomass yield of sugarcane tops (Getahun et al., 2018). The DMY of cultivated pasture was estimated according to a conversion factor of 5 t DM/h/year (FAO, 2012), which was lower than the conversion factor of 8 t DM/h developed by Von Carlowitz (1984). The CP and ME values of cultivated forage were estimated based on a mean value of 142 g/kg DM and 9.13 MJ/kg DM estimated for different cultivated forage species (Feyisa et al., 2021). Generally, the sizes of the land and yield obtained were not sufficient to ensure the animal with quality forage alone; this may be due to the competition of the land with crops grown, a lack of awareness among farmers of the benefits of cultivating forage crops and a shortage of forage seed and planting material.

TABLE 4 Land size in hectare, dry matter, crude protein and metabolizable energy yield annually from permanent food crop and cultivated pasture.

Permanent crops	LS × 103 ha	DMY × 103 t	DCPY × 103 t	MEY tones × 106
Cultivated pasture	154.59b	772.94b	74.74b	540.71a
Banana	68.25c	546.00c	52.80c	298.54b
Enset	183.32a	916.59a	88.63a	185.82c
Sugarcane tops	23.94c	790.00b	76.39b	541.38a
Average	107.52	756.38	73.14	391.61
SEM	21.40	134.85	13.04	55.81
p-value	<0.0001	0.28	0.28	<0.0001
CV	74.45	66.70	66.70	53.32

Note: Means of each parameter with different superscripts (a–c) within the same column are significantly different at p < 0.001.

Abbreviations: CV, coefficient of variance; DMY, dry matter yield; DCPY, digestible crude protein yield; LS, land size; MEY, metabolizable energy yield; SEM, standard error mean.

Crop, vegetable and root and tuber crop residues

Crop residue is the second major feed resource in the country. Ethiopia is a country known for its subsistence mixed crop-livestock production system, which is traditional and rain-fed, with very limited areas of irrigation. It has two cropping seasons, which are named Mehar and Belg. The average cultivated land, crop, vegetable, root and tuber yield and nutrients produced from residues of the current study for both seasons are presented in Table 5. The forage biomass obtained from crop residues gradually increased from 2007/08 to 2020/21 (Figure 3). The results of the study revealed that crop yield and nutrients obtained from residues significantly varied among seasons and crop types. The values of cultivated land, yield and residues recorded for Mehar season were significantly (p < 0.01) higher; this indicates that Mehar season production was practiced in all parts of the country as compared to Belg season. Likewise, DM, CP and ME yield in × 10⁶ t were significantly higher for Mehar season. Similarly, significantly (p < 0.01) higher values were observed among different crops in cultivated land size, yields in tones and residues in tonnes of DM, DCP and ME per year. The values of the current study showed that significantly (p < 0.01) higher land sizes in hectares were used for cereal crop cultivation, followed by oilseed and pulse crops in decreasing order. Significantly higher values of yield were recorded for cereal crops, followed by root and tuber, oilseed, pulse and vegetable crops in decreasing order. Conversely, significantly (p < 0.001) higher DM, DCP and ME values of residues were recorded for cereal crops, followed by oilseed, pulse and root and tuber crops, whereas the least value was recorded for vegetables. The higher value of crop residues recorded for cereal crops might be due to increased demand for the cultivation of cereal crops to ensure food security in the country (Aklilu et al., 2014; Kassahun et al., 2019). Furthermore, the variation among cultivated crops is attributed to the farming system, seasons, intensity of production, type of crop cultivation and availability of rainfall (Bedasa, 2012). Generally, we understand that residues obtained from cereals, oilseeds, pulses, root and tuber crops and vegetables of Mehar season are the major available feed resources for ruminants, whereas Belg season contributes a low amount of these residues.

TABLE 5 Land size, grain yield, residues dry matter, digestible crude protein and metabolizable energy yield of cultivated crops.

Variables	LS × 103 ha	Y × 103 t	DMY × 103 t	DCPY × 103 t	MEY × 106 t
Season
Mehar	2936.26a	7149.79a	9858.99a	104.88a	85,038.96a
Belg	432.15b	879.07b	785.76b	23.92b	7111.43b
Average	1814.42	4340.50	5794.18	90.39	50,127.43
SEM	250.36	900.23	1120.35	15.30	9208.10
p-Value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
Crops residues
Cereal crop residues	5650.65a	11,968.73a	21,658.46a	22.33c	183,749.69a
Oilseed crop residues	652.71c	692.34d	1699.25b	74.57b	11,290.79c
Pulse crop residues	948.46b	1362.12c	1984.15b	94.19b	17,039.19b
Root and tuber residues	643.18c	5951.04b	780.04c	128.95a	12,998.89c
Vegetable waste	525.86c	718.60d	161.68d	25.36c	3358.48d
Average	1814.42	4340.50	5794.18	90.39	50,127.43
SEM	98.25	1345.83	1694.08	13.02	14,438.67
p-Value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
CV	115.22	164.07	154.71	67.98	152.42

Note: Means of each parameter with different superscripts (a–d) within the same column are significantly different at p < 0.001.

Abbreviations: CV, coefficient of variance; DMY, dry matter yield; DCPY, digestible crude protein yield; LS, land size; MEY, metabolizable energy yield; SEM, standard error mean; Y, yield.

FIGURE 3. Forage biomass yield from different crop residues. DMYCCR, dry matter yield of cereal crop residues; DMYOSCKR, dry matter yield of oil seed crop residues; DMYPUCR, dry matter yield of pulse crop residues; DMYVGR, dry matter yield of vegetable crop residues; DMYRTCR, dry matter yield of root and tuber crop residues.

Concentrate feed

Although there is high competition between human food and animal feed owing to the high demand by livestock producers to incorporate grains in chicken rations, currently, commercial broiler and layer production is increasing at an alarming rate in Ethiopia. Grain, cereal bran and oilseed cake are sources of concentrate used to formulate the ration for these classes of chicken. The value of oilseed cake was estimated at 70% extraction rate after the quantity of exported seed was deducted from the entire annual yield, and the value of annual cereal bran was also estimated at 8% of total yields of wheat and rice grain (FAO, 2012). The amounts of these feed sources in DM, DCP and ME bases are presented in Table 6. The amount of concentrate annually produced has gradually increased in the last decade due to an increase in crop yields to ensure the increased human population with food security (Figure 4). Significantly (p < 0.001) higher DM and metabolizable energy yield were recorded for broken grain and grain, whereas significantly higher digestible protein was recorded for concentrate obtained from oilseed cake.

TABLE 6 Annual dry matter, digestible crude protein and metabolizable energy yield of concentrate feeds.

Year	DMY × 103 t	DCPY × 103 t	MEJY × 106 t
Cereal bran	356.34c	42.21c	3824.96d
Broken grain	2391.31a	119.43b	26,634.81a
Grain	780.48b	49.06c	11,360.43b
Oilseed cake	800.60b	257.38a	9482.76c
Pulse hulls	110.32d	10.91d	875.30e
Average	887.81	9.41	10,435.65
SEM	192.40	15.78	1856.53
p-Value	<0.0001	<0.0001	<0.0001
CV	81.10	61.64	66.57

Note: Means of each parameter with different superscripts (a–d) within the same column are significantly different at p < 0.001.

Abbreviations: CV, coefficient of variance; DMY, dry matter yield; DCPY, digestible crude protein yield; MEY, metabolizable energy yield; SEM, standard error mean.

FIGURE 4. Trends of feed biomass yield of different feed resources from 2007/08/ to 2020/21.

Estimated annual unconventional by-products

Nonconventional by-products can be used as sources of animal feed during feed shortage. Today, the use of these unconventional feeds in animal rations has increased from time to time due to a shortage of conventional feed in the animal industry. The estimated values of unconventional feed are presented in Table 7. The overall average 202.16 × 10³ ha of land, 302.02 × 10³ in tones of DM, 10.03 × 10³ in tones DCP and 2316.63 × 10⁶ in tones ME produced annually. However, there was no significant variation recorded among unconventional feed obtained from different sources except for significant (p < 0.05) lower values recorded for DCP obtained from sugar industry by-products.

TABLE 7 Land size, dry matter, digestible crude protein and metabolizable energy yield of some unconventional by-products.

Unconventional by-products	LS × 103 ha	DMY × 103 t	DCPY × 103 t	MEY × 106 t
Fruit by-products	647.13	406.37	12.39a	4136.58
Coffee plantation	1171.82	217.32	13.82a	2053.45
Sugar industry	606.94	524.69	1.49b	3016.96
Brewery industry	NA	59.70	12.42a	59.70
Average	202.16	302.02	10.03	2316.68
SEM	45,000.00	239.72	5.63	2170.30
p-Value	0.06	0.07	0.021	0.10
CV	158.85	158.75	112.42	187.36

Note: Means of each parameter with different superscripts (a, b) within the same column are significantly different at p < 0.001.

Abbreviations: CV, coefficient of variance; DMY, dry matter yield; DCPY, digestible crude protein yield; LS, land size; MEY, metabolizable energy yield; SEM, standard error mean.

Estimated annual total feed supply

The average feed supply in terms of DM in tones × 10³, DCP in tones × 10³ and MEJ in tones × 10⁶ annually produced are presented in Table 8. The study revealed that there was significant variation among feed sources.

TABLE 8 Mean total annual feed supply in terms of dry matter, digestible crude protein and metabolizable energy yield of different feed resources.

Feed sources	Land × 103 ha	DMY × 103 t	DCPY × 103 t	MEY × 109 t
Natural pasture	78,739.69a	87,552.60a	3140.38a	660.85a
Permanent crop	433.80d	3025.53d	292.57d	1.57e
Crop residues	13,351.61b	57,566.13b	345.40b	480.70b
Concentrate	NA	4700.15c	506.20c	54.20c
Unconventional feed	1667.24c	1208.08e	40.12e	9.27d
Average	23,548.09	29,520.10	864.93	234.29
SEM	4361.20	1870.13	48.25	16.50
p-Value	<0.0001	<0.0001	<0.0001	<0.0001
CV	23.80	23.70	18.95	26.35

Note: Means of each feed resources with different superscripts (a–d) within the same column are significantly different at p < 0.001.

Abbreviations: CV, coefficient of variance; DMY, dry matter yield; DCPY, digestible crude protein yield; LS, land size; MEY, metabolizable energy yield; SEM, standard error mean.

The forage biomass obtained from crop residues has gradually increased in the last decade, whereas the forage biomass yield obtained from natural pasture has gradually decreased from 2007/08 to 2020/21 (Figure 4). The highest significant value was recorded for natural pasture, followed by crop residues, concentrate, fodder from permanent crops and unconventional by-products in decreasing order.

Natural pasture shares about 56.83%, 47.67% and 54.77% DM, DCP and ME yield, respectively, followed by crop residues with a percentage of 37.37% DM, 7.99% DCP and 39.84% ME (Table 9). The results of the current study are in agreement with the findings of Chufa et al. (2022), who reported that natural pasture and crop residues have been covering about 56.23% and 30.06%, respectively, of the total feed supply of ruminants in Ethiopia. We concluded that natural pasture and crop residues are the major feed resources in Ethiopia.

TABLE 9 Overall mean percentage of dry matter, digestible crude protein and metabolizable energy obtained from various feed resources.

Feed category	DMY × 103	% DMY	DCPY × 103	% DCPY	MEY × 109	% MEY
Natural pasture	87,552.6	56.83	3140.38	68.73	660.85	54.77
Permanent crop	3025.53	1.96	292.57	6.40	1.57	0.13
Crop residues	57,566.1	37.37	612.58	13.41	480.70	39.84
Concentrate	4700.15	3.05	483.22	10.58	54.20	4.49
Unconventional feed	1208.08	0.78	40.12	0.88	9.27	0.77
Total	154,052	100	4568.86	100	1206.58	100

Abbreviations: DMY, dry matter yield; DCPY, digestible crude protein yield; MEY, metabolizable energy yield.

Trends of total animal feed supply

The total feed supply trends in terms of DMY, DCP and ME increased from 2007/2008 to 2020/21. However, the supply of these parameters decreased for natural pasture while showing increasing trends for crop residues (Figure 5a–c). In agreement with the current study, Feyisa et al. (2021) reported annual increment trends for biomass yield of crop residues for the past decade. The highest increment trends observed for crop residues are due to the high demand for crop cultivation to ensure increased human populations with food security.

FIGURE 5. (a) Trends of total dry matter yield from 2007/08/ to 2020/21. NPDMY, natural pasture dry matter yield; PCCPDMY, permanent crop and cultivated pasture dry matter yield; CRDMY, crop residues dry matter yield; UNCFDMY, unconventional feed dry matter yield. (b) Trends of total digestible crude protein yield from 2007/08 to 2020/21. NPDCPY, natural pasture digestible crude protein yield; PCCPDCP, permanent crop and cultivated pasture digestible crude protein yield; CRDCPY, crop residues digestible crude protein yield; UNCFDCPY, unconventional feed digestible crude protein yield. (c) Trends of total metabolizable energy yield from 2007/08 to 2020/21. NPDMEY, natural pasture metabolizable energy yield; PCCPMEY, permanent crop and cultivated pasture metabolizable energy yield; CRMEY, crop residues metabolizable energy yield; UNCFMEY, unconventional feed metabolizable energy yield.

Livestock population and feed requirements Livestock population

The overall mean of the livestock population in Ethiopia is estimated at about 59.42 million TLU (Table 10). It comprises about 46.60 million cattle, 2.08 million sheep, 3.04 million goats, 2.83 million camels, 3.87 million donkeys, 1.74 million horses and 2.60 million mules. Ruminants are the major livestock population and covers about 85% of the overall total TLU. The livestock populations showed increasing trends from 2007/08 to 2020/21 (Figure 6).

TABLE 10 Total annual livestock population in tropical livestock unit (TLU).

Year	Cattles	Sheep	Goats	Camels	Donkeys	Horses	Mules	Total
2007/08	39,668,278.26	1381,002.58	1225,226.87	11,009,040	1895,966.5	998,849.7	149,562	56,327,925.91
2008/09	41,488,588.41	1718,032	1545,860.14	759,696	2710,948	1608,489.9	261,463.3	50,093,077.75
2009/10	34,259,284.21	1813,550.4	1827,900.97	930,907.5	3042,563.3	1662,416.1	265,612.55	43,802,234.98
2010/11	44,712,788.51	1813,597.43	2109,941.8	1102,119	3374,178.5	1716,342.3	269,761.8	55,098,729.34
2011/12	43,690,450.75	1718,032.24	1572,208.54	979,318	4876,870	1765,754.1	258,146.7	54,860,780.33
2012/13	45,295,615.85	1808,007.29	1711,192.87	915,518	3374,178.5	1716,342.3	245,018.2	55,065,873.01
2013/14	46,163,067.2	1950,849.13	2054,861.01	1098,312	3476,538.5	1766,709	249,260.9	56,759,597.74
2014/15	47,450,266.17	2067,926.89	2054,861.01	8145,790	5395,948	1933,642.8	267,948.8	67,316,383.67
2015/16	48,420,684.32	2067,620.54	2109,941.8	1228,023	3940,697	1873,982.7	284,165	59,925,114.36
2016/17	49,615,704.67	2168,590.61	2141,077.11	1209,321	4219,610	1942,358.4	286,913.9	61,583,575.69
2017/18	38,400,132.77	2159,882.49	1614,368.03	1228,023	3940,697	1873,982.7	284,165	49,501,250.99
2018/19	51,025,070.46	2526,608.28	2782,169.34	1760,870	4827,720.5	1737,727.2	259,386.4	64,919,552.18
2019/20	49,703,018.41	2905,664.36	3011,030.5	1164,106	3714,018.5	1829,803.5	280,230.3	62,607,871.57
2020/21	58,432,274.06	3110,546.74	3804,114.41	8145,790	5395,948	1933,642.8	267,948.8	81,090,264.81
Average	45,594,658.86	2086,422.213	2111,768.171	2834,059.536	3870,420.2	1740,003.11	259,255.975	58,496,588.02

FIGURE 6. The overall trends of livestock population in tropical livestock unit (TLU) from 2007/08 to 2020/21.

Livestock feed requirements

Ethiopia has the largest livestock population in Africa and is among the top ten in the world. They contribute output and input functions for the owner. The average numbers of TLU and nutrient requirements are presented in Table 11. Significantly (p < 0.001) higher values of TLU, dry matter requirement (DMR), DCPR, and MER in tones were observed among livestock species. The highest TLU, DMR, DCPR and MEJR were recorded for cattle, followed by camels, donkeys, horses, sheep and goats in decreasing order. These indicate that ruminants have the highest DM, DCP and ME demands among other animal species in Ethiopia. The increment trends were observed in total TLU, DMR, DCPR and MER from 2007/08 to 2020/21 (Figure 7a–c).

TABLE 11 Mean tropical livestock unit (TLU), dry matter requirement, digestible crude protein requirement and metabolizable energy requirement of different cropping year and livestock species in tonnes.

Livestock species	TLU	TDMR × 103	TDCPR × 103	TMER × 109
Camels	2834,059.57b	217.23b	268.95b	28.60b
Cattles	46,226,187.64a	3543.24a	4386.87a	755.06a
Donkeys	3870,419.93b	296.66b	367.30b	48.95b
Goats	2010,806.00b	154.13b	190.83b	13.10c
Horses	1740,003.14c	133.37b	165.13b	30.55b
Mules	259,256.00d	19.87c	24.60c	3.09c
Sheep	2083,012.50b	159.66b	197.68b	40.95b
Average	8431,963.54	646.31	800.19	131.47
SEM	606,266.10	66.63	57.53	11.09
p-Value	<0.0001	<0.0001	<0.0001	<0.0001
CV	26.90	26.90	26.90	29.23

Note: Means of each parameter with different superscripts (a–d) within the same column are significantly different at p < 0.001.

Abbreviations: CV, coefficient of variance; SEM, standard error mean; TDMR, total dry matter requirement; TDCPR, total digestible crude protein requirement; TMER, total metabolizable energy requirement.

FIGURE 7. (a) Trends of dry matter requirements of different livestock species from 2007/08 to 2020/21, (b) trends of digestible crude protein requirements of different livestock species from 2007/08 to 2020/21 and (c) trends of metabolizable energy requirements of different livestock species from 2007/08 to 2020/21.

Enteric CH₄ emission from livestock

Enteric CH₄ emissions from different livestock categories and trends are presented in Table 12 and Figure 8. The average 2747.83 Gg of enteric CH₄ was emitted annually in the last two decades in Ethiopia. Cattle has largest enteric CH₄ production, contributing more than 90% of total emission of the country. The emission was annually increased at the rate of 2.8% and highest (3661.74 Gg) emission was recorded in 2020/2021. The increment trends (Figure 8) in enteric CH₄ emissions observed in the last two decades were attributed to the increasing population number of cattle, goat and sheep, which was caused by rapidly increasing demand for livestock products. Thorntona and Herrero (2010) and Federal Democratic Republic of Ethiopia (FDRoE) (2010) stated that the livestock population were gradually increased in the last two decades due to demand driven by population growth, increase income and urbanization.

TABLE 12 Enteric CH₄ emission of Ethiopia from different livestock categories in Gg/year.

Year	Cattle	Sheep	Goat	Camel	Donkey	Horse	Mule	Total
2007/08	2181.76	38.36	39.82	222.11	19.34	8.17	1.53	2511.08
2008/09	2281.87	47.72	50.24	15.33	27.65	13.16	2.67	2438.64
2009/10	1884.26	50.37	59.41	18.78	31.03	13.60	2.71	2060.17
2010/11	2459.20	50.38	68.57	22.24	34.42	14.04	2.75	2651.60
2011/12	2402.97	47.72	51.10	19.76	49.74	14.44	2.63	2588.37
2012/13	2491.26	50.22	55.61	18.47	34.42	14.04	2.50	2666.52
2013/14	2538.97	54.19	66.78	22.16	35.46	14.45	2.54	2734.55
2014/15	2609.76	57.44	66.78	164.34	55.04	15.82	2.73	2971.92
2015/16	2663.14	57.43	68.57	24.78	40.20	15.33	2.90	2872.34
2016/17	2728.86	60.24	69.59	24.40	43.04	15.89	2.93	2944.94
2017/18	2112.01	60.00	52.47	24.78	40.20	15.33	2.90	2307.67
2018/19	2806.38	70.18	90.42	35.53	49.24	14.21	2.65	3068.61
2019/20	2733.67	80.71	97.86	23.49	37.88	14.97	2.86	2991.43
2020/21	3213.78	86.40	123.63	164.34	55.04	15.82	2.73	3661.74
Average	2507.71	57.95	68.63	57.18	39.48	14.23	2.64	2747.83

FIGURE 8. Trends of enteric CH4 emission of different livestock categories from 2007/08 to 2020/21.

Feed balance between the amount of annual dry matter, digestible protein metabolizable energy required and supplied

The differences among DM, DCP and ME in joules available and required were measured by balancing the amount of feed required by the total livestock population with the amount of feed supplied. The total estimated DM, DCP and MERs and differences between supplies and requirements across cropping years and among different livestock species in TLU are presented in Table 13 and their trends are described in Figure 9a–c. The study revealed that there was no animal feed deficit in the last decade. The overall mean of 153.53 × 10⁶, 4.57 × 10⁶ and 1203.97 × 10⁹ t of DM, DCP and MEJ supply were recorded, respectively, in the last decade. Significantly (p < 0.001) highest variations in animal feed requirements were observed among cropping years and livestock species. The annual estimated feed requirements were 134.62 × 10⁶ t, 4.52 × 10⁶ t and 918 × 10⁹ J DM, DCP and ME, respectively. The mean surplus feed obtained was 18.91 × 10⁶, 37.57 × 10³ and 285.14 × 10⁹ t DM, DCP and ME, respectively. Similar to current findings, Sisay (2006) and Tesfaye et al. (2008) reported a surplus feed in total DM supply, which was higher than the total annual livestock requirements around the North Goundor and Metema areas. However, contrary to the current findings, FAO (2018) reported a negative annual livestock feed balance for the entire feed supply of the country, and other authors Ayela et al. (2021) for Lalo Kile district; Assefa et al (2013); and Wondatir (2010), Adamitulu Jildo Kobolicha district also reported a deficit feed supply, which was lower than the total livestock feed requirements. However, the insignificant negative feed balances were recorded for DM and DCP for the year 2014/2015, whereas only DCP was negative balanced recorded for the years 2015/16 and 2020/21 due to the higher number of livestock that demanded more than the total supply in this year.

TABLE 13 Tropical livestock unit (TLU), dry matter in tones × 10³, digestible crude protein in tones × 10³ and metabolizable energy in tones × 10⁹ supply, required and balance across different cropping years and livestock species.

Livestock spec.	TLU	DMY × 103	DCPY × 103	MEY × 109	TDMR × 109	TDCPR × 103	TMEJR × 109	∆DM × 103	∆DCP × 103	∆MEJ × 109
Camels	2834,059.50b	7197.91b	208.77b	55.50b	6465.20b	217.23b	28.60b	732.71b	−8.47d	27.39b
Cattles	46,226,187.70a	120,425.38a	3588.78a	945.81a	105,453.49a	3543.24a	755.05a	14,971.89a	45.54a	190.75a
Donkeys	3870,420.20a	10,111.38b	298.92b	79.46b	8829.40b	296.66b	48.95b	1281.98b	2.25b	30.50b
Goats	2010,806.00b	5264.02b	154.34c	41.30b	4587.15b	154.13b	13.10b	676.87b	0.21c	28.18b
Horses	1740,003.10c	4551.73b	135.99c	35.78b	3969.38b	133.37b	30.55b	582.35b	2.62b	5.23c
Mules	259,256.00d	678.00c	20.31b	5.33c	591.43c	19.87c	3.09c	86.57c	0.44c	2.24c
Sheep	2083,012.50b	5465.55b	161.77c	42.92b	4751.87b	159.66b	40.95b	713.68b	2.10b	1.97c
Total	58,906,710.10	153,531.15	4560.82	1203.97	134,620.53	4523.25	918.83	18,910.62	37.57	285.14
SEM	891,682.9	2854.60	45.21	21.22	3408.81	66.66	14.03	1603.40	51.87	18.20
p-Value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
CV	22.61	27.29	14.54	26.33	22.05	22.05	22.81	125.99	1737.17	95.10

Note: Means of each parameter with different superscripts (a–c) within the same column are significantly different at p < 0.001.

Abbreviations: CV, coefficient of variance; DMY, dry matter yield; DCPY, digestible crude protein yield; ∆DM, change in dry matter; ∆DCP, change in digestible crude protein; MEY, metabolizable energy in joule yield; ∆MEJ, change in metabolizable energy in joule; SEM, standard error mean; TDMR, total dry matter requirement; TDCPR, total digestible protein requirement; TMEJR, total metabolizable energy joule requirement.

FIGURE 9. (a) Trends of total dry matter yield, dry matter requirements and change from 2007/08 to 2020/21. DMY, dry matter yield; DMR, dry matter requirement; ΔDM, change in dry matter. (b) Trends of total digestible crude protein yield, requirements and change from 2007/08 to 2020/21. DCPY, digestible crude protein yield; DCPR, digestible crude protein requirement; ΔDCP, change in digestible crude protein. (c) Trends of total metabolizable energy yield, requirements and change from 2007/08 to 2020/21. MEY, metabolizable energy yield; MEJR, metabolizable energy in joule requirement; ΔMEJ, change in metabolizable energy in joule.

Generally, ruminants are highly dependent on the roughage feed annually supplied, which is able to cover about 94% of the total feed supply. The concentrate feed, which can supply high protein and energy, is available for monogastric animal and commercial livestock production. Therefore, the surplus feed can support the nutrient requirements of various classes of livestock according to the recommendations of the National Research Council: about 500 × 10⁶ broilers (NRC, 1994), about 5 × 10⁶ bulls (NRC, 1996), about 50 × 10⁶ small ruminants (NRC, 2007) or 3 × 10⁶ crossbred lactating dairy cows yield 10 litres of milk per day (NRC, 2001). These will increase per capita consumption of animal products and foreign currency obtained from livestock by 100%.

CONCLUSIONS AND RECOMMENDATIONS

The current study revealed that natural pasture and crop residues cover about 94% of the entire country's feed supply. Available feed resources are sufficient to meet the maintenance, production, activity, and work nutrient requirements of livestock, excluding feed required for commercial chicken production in the country in terms of DM, DCP, and ME. The overall mean estimated feed requirements by livestock covered about 114.04%, 100.83%, and 131.33% of the total DM, DCP and MEJRs, respectively. However, a higher number of ruminant species that are mainly fed on roughage feed is a crucial issue for environmental pollution, whereas surplus feed or concentrate can cover the annual ration required for 500 × 10⁶ broiler chicken. Based on these results, different strategies should be set up to encourage market-oriented ruminant and poultry production. Consequently, the following are recommended:

• The government should design a policy to increase commercial broiler production by reducing the number of unproductive ruminant species because ruminants can release more methane gas into the environment, which has the capability to increase global warming.
• As the major feed resource so far is natural pasture, it needs proper management and feed conservation practices (hay and silage making) for the area which there is surplus feed during wet season to, at least, support the extensive livestock production system practiced in the country.
• Crop residues are the most important feed resources for contributing to livestock feed, but they tend to be of low quality and waste a lot at the farm level during harvesting and feeding livestock. Hence, encouraging farmers to use proper collection, processing (physical, chemical and biological), transporting, storage and feeding practices in order to enhance the utilization of low-quality feed.
• Irrigation practices, clearing invasive species from grazing land, improved feeding, more effective extension services and farmer training are required to increase feed productivity and hence the overall development of the country.
• Improved forage cultivation is also a good strategy to improve the nutritional value of forages. Therefore, for the farmers who lack seeds and awareness of an improved forage development strategy, a joint extension and training service and the provision of seeds should be required.

AUTHOR CONTRIBUTIONS

Conceptualization; data curation; formal analysis; investigation; methodology; resources; software; supervision; validation; visualization; writing – review and editing: Regasa Begna. Conceptualization; data curation; formal analysis; methodology; resources; software; validation; visualization; writing – review and editing: Worku Masho.

ACKNOWLEDGEMENTS

Authors of this paper would like to acknowledge Mizan Tepi University for allowing access to library and internet facilities.

CONFLICT OF INTEREST STATEMENT

The authors declare that there are no conflicts of interest.

FUNDING INFORMATION

No special funding was obtained for this study.

DATA AVAILABILITY STATEMENT

Data used and analysed for this study available from the corresponding author on reasonable request.

ETHICS STATEMENT

None.

PEER REVIEW

The peer review history for this paper is available at https://publons.com/publon/10.1002/vms3.1415.