Seasonality, balance and copying mechanisms of livestock feed in Northwestern Ethiopia

Abstract

This study aimed to determine the seasonality, feed balance, and copying mechanisms of livestock feed in the east Gojjam Zone of northwestern Ethiopia. To collect the data, the zone divided into three agroecologies (lowland, midland, and highland) and randomly selected two districts from each agroecology to represent appropriate agroecologies. The study involved 540 respondents – 45 from each peasant association – and utilized key informant interviews, focus groups, questionnaires, and direct observations for data collection. According to the findings, the study reveals that feed availability in all agroecologies fluctuates seasonally. The total dry matter and crude protein supplies per household per year in highland, midland, and lowland agroecologies were 8.69, 15.49, and 14.94 tons, and 0.05, 0.91, and 0.90 tons, respectively. The yearly dry matter and crude protein requirements for livestock in highland, midland, and lowland agroecologies were 11.9, 22.2, and 19.5 t/hh/yr, and 1.4, 2.6, and 2.5 t/hh/yr, respectively. The total amounts of dry matter and crude protein produced in the study areas satisfied only 73.3, 69.7, and 76.6%, and 32.1, 35.5, and 36.3% of the livestock requirements in the highland, midland, and lowland agroecologies, respectively. Producers in the study areas employ strategies such as purchasing feed, preserving crop residue, and reducing herd sizes to address livestock shortages of dry matter and crude protein. To address the issue, a new strategy should be employed, including managing natural pastures, increasing forage crop cultivation, providing protein-rich feed, and improving crop residue quality using effective microorganisms.

Keywords:

• Agroecology
• dry matter
• feed balance
• natural pasture
• seasonality

Reviewing Editor:

• Pedro Gonzalez-Redondo, Universidad de Sevilla, Spain

Subjects:

• Agriculture & Environmental Sciences
• Plant & Animal Ecology
• Botany

1. Introduction

Although Ethiopia has the largest livestock population in Africa and livestock plays a significant role in the country, the livestock industry has not developed to its full potential due to challenges such as inadequate feed, low genetic potential in native breeds, a lack of farmers’ knowledge regarding husbandry techniques, and the prevalence of disease (Alemneh & Getabalew, 2019; Alilo, 2019; Getabalew et al., 2019). The scarcity of high-quality feed is an obstacle to sustainable livestock production in Ethiopia (Ayele et al., 2021; Gashe et al., 2017).

Crop residues, natural pastures, crop stubble, agro-industrial by-products, and improved pasture and forage are the primary sources of feed in Ethiopia (Desta, 2023; Desta et al., 2023; Firew & Getnet, 2010; Gashe & Kassa, 2018; Tahir et al., 2018). Although natural pastures are the primary source of animal feed in the country, their production and size are declining due to poor grazing management, continual conversion of grazing land to cropland, urbanization, redistribution of grazing land to youth, the spread of aggressive weeds, and soil erosion (Desta et al., 2023; Fenetahun et al., 2021; Gurmessa, 2021; Yirga et al., 2019).

Crop residue accounts for approximately half of the feed used for ruminant animals in mixed crop and livestock production systems in highland Ethiopia. During the dry season, they can contribute up to 80% of the total feed (Firew & Getnet, 2010). Dependence on crop residues as livestock feed will continue to increase as more grazing land is cropped to satisfy the rapidly expanding demand of the human population for grain. The scenario regarding feeding livestock is made more difficult by the high cost of concentrates, shortage of feed concentrates, and their non-competitiveness with food crops (Mediksa et al., 2021).

The main obstacle preventing the growth and expansion of livestock production in Ethiopia is seasonal fluctuations of feed quality and quantity (Desta et al., 2023), which are mainly caused by seasonal variations in rainfall. As a result, especially during the dry season, animals had reduced body condition, increased vulnerability to diseases and physiological stress, negative impacts on reproduction and productivity, and longer calving intervals (Cheng et al., 2022). Therefore, to cope with feed shortages during difficult times, livestock producers in Ethiopia should use a variety of coping strategies. However, there is insufficient information on copying mechanisms in current research areas.

To ensure sustainable livestock production, it is crucial to ensure that the annual feed production from available resources meets the livestock requirements in a specific location. A balanced diet containing all the nutrients required to meet the objectives of production and considering the animal’s physiological state is essential for high and sustained productivity in livestock (Leitch et al., 2014). Therefore, determining the annual livestock feed balance is essential in designing research programs and potential interventions for livestock and feed production. Most studies have focused on assessing the availability of feed resources rather than evaluating the amount of dry matter and crude protein (Gashe & Kassa, 2018). However, such information is scarce in the east Gojjam Zone of Ethiopia. Therefore, the objectives of this study were to: (1) quantify the dry matter and crude protein yield from the main livestock feed resources; and (2) estimate the annual balance of livestock feed in the east Gojjam Zone, Ethiopia.

2. Materials and methods

2.1. Description of the study areas

The study was conducted in the east Gojjam zone, Amhara Region, Ethiopia. The zone is located in Ethiopia’s northwest highlands at latitudes of 10°1 46″ and 10°35’ 12″ N and longitudes of 37°12’ 45″ and 37°55’ 52″ E (Figure 1). The east Gojjam zone has different agroecologies owing to its varied elevations. The district’s elevation varies from 1500 to 3577 m above sea level. The East Gojjam Zone Agriculture Office reports that the average annual rainfall ranges from 900 to 2000 mm and that the average minimum and maximum temperatures fall between 7 and 25 °C. Figure 2 shows the contribution of the three agroecologies in the six sampled districts, in accordance with information acquired from the east Gojjam zonal agricultural office.

Figure 1. Location map of the study areas.

Figure 2. The proportion of the three agroecologies in the selected six districts.

2.2. Sampling techniques and data collection

The study was conducted in the east Gojjam zone, Amhara Region, Ethiopia. Based on agroecology, the districts in the east Gojjam Zone were divided into highland, midland, and lowland areas. Two districts were randomly chosen from each agroecology, namely Sinan and Machakel, Aneded and Enemay, Debre elias, and Basoliben, to represent the zone’s highland, midland, and lowland agroecologies, respectively (Figure 2). Two peasant associations representing appropriate agroecology were selected from each district. Thus, 540 respondents were purposefully selected from the study areas, with 45 respondents from each peasant association who had at least three animals selected for individual interviews. The study utilized Individual interviews were conducted to gather data on cropland size, livestock holdings, yield, and seasonality of feed resources. Field observations, focus groups, interviews with key informants, and discussions with livestock experts were conducted.

2.3. Seasonal availability of livestock feed

The rainfall patterns and seasonality of the major feed resources were collected through focus group discussions. The respondents were assigned rainfall scores ranging from 0 to 5, with 0 indicating no rainfall and 5 indicating heavy rainfall per month. Respondents were also given a scale to rate the seasonality of feed resources, with 10 indicating excess, 5 adequate, and 0 no feed available per month. The data were analyzed using feed assessment tools (FEAST) developed by the ILRI.

2.4. Copying mechanisms of livestock feed

The Garrett approach was used to rank the copying mechanisms of feed shortages at the study locations. Respondents were asked to rank all copying mechanisms of feed shortages, which were then converted into a score value using a formula: percent position = 100(Rij–0.5) Nj, Where R_ij = Rank given for the i^th variable by j^th respondents; N_j = Number of variables ranked by j^th respondents with the help of Garrett’s table, the percent position estimated is converted into scores. Then, for each copying mechanism, the scores of each individual were added, and the total value of scores and mean scores were calculated. The copying mechanism with the highest mean value is considered the most copying mechanism.

2.5. Estimation of annual livestock feed supply from the main feed resources

Crop yields were collected using questionnaires from sampled respondents or households. The quantity of crop residue was estimated by multiplying the total grain yield by the following conversion factors: maize stover (2.0), small cereals (1.5), pulse straw and oil crops (1.2), and potatoes (0.3) (Plan, 1987). For natural pasture land, stubble crops, bush and shrublands, and improved forage lands, conversion factors of 2.0, 0.5, 1.2, and 8 tons per hectare per year, respectively, were used (Plan, 1987). The amount of dry matter from the collected fodder and supplement feed was determined by asking farmers how much they supply each day as well as how often and how much they buy each month.

2.6. Estimation of the annual feed demand-supply balance

Individual interviews were used to estimate the average values of the feed resources available on a per household basis. Total available dry matter and crude protein were calculated by adding the feed supplied by each major feed resource category. The maintenance dry matter and crude protein requirements of the livestock were used to calculate the requirements for the livestock types. To determine the dry matter and crude protein requirements of livestock, livestock numbers were converted into tropical livestock units (TLU) (Jahnke & Jahnke, 1982). The conversion factors used for oxen, bulls, cows, heifers, calves, sheep, goats, donkeys, mules, and horses were 1, 1, 0.7, 0.5, 0.6, 0.1, 0.1, 0.5, 0.7, and 0.8, respectively. Then, using standard tropical livestock units of 250 kg dual-purpose tropical cattle with a dry matter requirement of 2.5% of body weight, or 6.25 kg dry matter per day or 2.28 tons per year for maintenance, the feed requirement for all livestock species was estimated (Jahnke & Jahnke, 1982). According to Plan (1987), a tropical livestock unit weighing 250 kg requires 160 g of digestible protein per day for maintenance. Digestible crude protein was estimated by summing the nutrients supplied by each feed resource category. According to Dana et al. (2020) a formula to calculate the quantity of digestible crude protein (DCP). DCP = 0.929*CP (g) - 3.48. The annual feed balance was determined by subtracting the annual requirement from the total amount of feed produced in a given year. There is either an excess or shortage of livestock feed depending on whether the annual production of feed exceeds the amount required for the maintenance of livestock.

2.7. Chemical analysis feed samples

Samples were collected from each feed resource to determine the crude protein content. The samples were weighed on an electronically sensitive balance before being placed in a paper bag with an appropriate label. The samples were ground in a Wiley mill to pass through a 1 mm sieve screen after being dried in an oven at 65 °C for 72 h. DM and CP were determined from feed samples using the AOAC method (Aoac, 1990). Crude protein was estimated by summing the nutrients supplied by each feed resource category.

2.8. Statistical analysis

Data from the household survey were analyzed using SPSS version 25 software. The least significant difference was used to test for significant differences between the mean comparisons. The following model was used for the analysis: y_ij = µ + A_i + e_j, where, y_ij = responsible variable, µ = overall mean, A_i = effect of agroecology, and e_ij = random error.

3. Results and discussions

3.1. Land Holding and utilization

The most important asset significantly relied upon by respondents in the study areas to sustain their households amid any form of crisis and to provide everything needed for their crops and livestock was the land. Table 1 shows the landholdings and utilization patterns of the respondents in the study areas. The average cropland size per household in the study area was 1.37 hectares (ha), which is greater than 1.07 ha in Enebsie Sar Midir district (Amsalu & Addisu, 2014a) and 1.14 ha in Ethiopian households (CSA, 2020). However, it is less than 2.18 ha in the north Achefer district, Ethiopia (Asmare & Mekuriaw, 2017). The average cropland size per household varied significantly (P < 0.05) among the agroecologies (Table 1). The average cropland size per household in lowland agroecology was significantly higher (P < 0.05) than that in highland and midland agroecology. However, the average cropland size per household was significantly higher (P < 0.05) in midland agroecology than in highland agroecology. This could be a result of the conversion of grazing land into cropland and the low density of people in lowland agroecology. The large amount of land per household in lowland agroecology promises a future for feed and livestock production.

Table 1. Average land size (ha) per household in different agroecologies.

		Agroecology			Contribution (%)
Land use system	Highland	Midland	Lowland	Mean ± SE
Cropland	0.86 ± 0.05^c	1.50 ± 0.05^b	1.75 ± 0.05^a	1.37±.05	68.26
Woodland	0.03 ± 0.04^b	0.09 ± 0.04^a	0.09 ± 0.04^a	0.07 ± 0.04	3.34
Improved forage land	0.01 ± 0.01	0.01 ± 0.01	0.01 ± 0.01	0.01 ± 0.01	0.10
Natural pasture land	0.43 ± 0.11^b	0.53 ± 0.11^b	0.71 ± 0.11^a	0.57 ± 0.11	28.30
Total	1.32 ± 0.10^c	1.59 ± 0.12^b	2.54 ± 0.12^a	1.82 ± 0.11^b	100.00

Means in the same row with different superscript letters differ significantly (P < 0.05) among agroecology; SE = standard error.

According to land-use pattern, cropland, pasture land, woodland, and improved forage land contributed approximately 68.3, 28.3, 3.3, and 0.1% of the total size owned per household, respectively. This indicates that the majority of cropland occupied 68.3% of the land, whereas natural grazing land constituted only 28.3% of the total area, and forage crops accounted for the smallest area. The amount of land allotted for forage production was much lower for all agroecologies than for the size of land allotted for the growth of food crops (Table 1). This might be because farmers prioritize growing food crops to produce animal feed in their available areas (Desta, 2022). The lack of comparative research on the economics of crop production and livestock per unit of land may be a problem with this ineffective forage development strategy in the country and the study areas (Desta, 2022).

3.2. Livestock holding size (TLUs) per household

Cattle, sheep, goats, and equines comprised the vast majority of tropical livestock units (TLUs) in the study area (Table 2). They are a crucial part of mixed farming systems, and have a strong connection with crop productivity. The overall average TLU of livestock holding per household in the study areas was 9.56, which is lower than 12.20 (Yadessa et al., 2016), while it is greater than 6.81 in the north Achefer district (Asmare & Mekuriaw, 2017) and 2.91 in the Gummara rib watershed, Ethiopia (Amsalu & Addisu, 2014a). The total tropical livestock unit holding per household was significantly higher (P < 0.05) in the lowland (8.54 ± 0.83) and midland (9.76 ± 0.83) agroecologies than in the highland (5.23 ± 0.83) agroecologies (Table 2). The disparity in livestock holdings could be because households in lowland and midland agroecologies had higher holdings of natural grazing land than those in highland agroecologies (Table 1). In general, cattle (49.30%), compared to other animals, contributed the most to the expansion of the herd in the research areas, confirming the findings of (Abera et al., 2014; Amsalu & Addisu, 2014a, 2014b; Birhan & Adugna, 2014; Leta & Mesele, 2014; Yadessa et al., 2016) who also found the same results in mixed crop-livestock production systems. Cattle are used to thresh crops and plow croplands, which explains the fact that there are more cattle in the study areas.

Table 2. Average number of classes of cattle (TLUs) and different livestock species (TLUs).

		Agroecology
Livestock	Highland	Midland	Lowland	Mean ± SE	Contribution (%)
Calves	0.14 ± 0.83^b	0.22 ± 0.83^ab	0.39 ± 0.83^a	0.25 ± 0.83	5.30
Heifers	0.38 ± 0.24^b	0.81±.0.24^a	0.48 ± 0.24^b	0.56 ± 0.24	11.84
Cows	1.04 ± 0.08^b	1.05 ± 0.08^b	1.91 ± 0.08^a	1.33 ± 0.08	28.31
Bulls	0.27 ± 0.43^a	0.52 ± 0.43^a	0.27 ± 0.43^b	0.35 ± 0.43	7.47
Oxen	0.68 ± 0.89^b	2.84 ± 0.89^a	3.13 ± 0.89^a	2.22 ± 0.89	47.07
Cattle	2.51 ± 0.38^b	5.44 ± 0.38^a	6.18 ± 0.38^a	4.71 ± 0.38	49.30
Sheep	0.50 ± 0.38^a	0.25 ± 0.38^b	0.24 ± 0.38^b	0.33 ± 0.38	3.45
Goats	0.00 ± 0.00^b	0.00 ± 0.00^b	0.10 ± 0.16^a	0.10 ± 0.16	1.05
Horses	1.69 ± 0.74^a	1.07 ± 0.74^b	0.00 ± 0.00^c	1.38 ± 0.74	14.44
Mules	0.53 ± 0.75^a	0.00 ± 0.00^b	0.00 ± 0.00^b	0.53 ± 0.75	5.54
Donkeys	0.00 ± 0.00^b	3.00 ± 0.02^a	2.01 ± 0.02^a	2.51 ± 0.02	26.23
Total	5.23 ± 0.83^b	9.76 ± 0.83^a	8.54 ± 0.83^a	9.56 ± 0.83	100.00

Means within the same row with different superscript letters significantly varied (P < 0.05) among agroecologies.

TLUs = tropical livestock unit, SE = standard error.

The average number of sheep kept per household was significantly higher (P < 0.05) in highland agroecology than in midland and lowland agroecology, whereas the average number of goats kept per household was significantly higher (P < 0.05) in lowland agroecology than in midland and highland agroecology, supporting the claim that there are more sheep in the highlands than in lowlands and midlands (Addisu et al., 2016; Yadessa et al., 2016). Owing to their adaptation to cold conditions and a portion of society’s preference for sheep over goats, sheep are more commonly used in highland agroecology. The availability of browsing species, the preference of goats for their feed and their adaptation to hot climates may be the cause of the highest number of goat holdings in lowland agroecology. Similar to the findings of Yadessa et al. (2016), the average number of horses and mules per household in highland agroecology was significantly higher (P < 0.05) than those in midland and lowland agroecologies. This area has Rocky Mountains and rough terrain; therefore, horses and mules are preferred because of their strength, physical fitness, and environmental adaptability. According to the respondents, equines are utilized in the study area for a variety of regular tasks, including the transportation of people, agricultural inputs, and products. To design livestock improvement plans in the study area, livestock development strategies should consider the number of livestock per household because the variation in livestock holdings across agroecologies in the area indicates that there is variation in livestock type preference.

3.3. Availability of livestock feed in relation to rainfall pattern

In the study area, the main sources of feed for livestock were ranked by respondents in descending order as natural pasture, crop residues, crop stubble, agro-industrial by-products, and cultivated forage (Figure 3). This finding is consistent with those of previous studies conducted by (Firew & Getnet, 2010; Jobir & Juhar, 2021; Miresa et al., 2019; Yadessa et al., 2016). In all agroecologies, natural pasture and crop residues served as the main sources of feed for livestock throughout the wet and dry seasons, respectively, which is consistent with observations (Belay & Negesse, 2018; Biratu & Haile, 2017; Desta et al., 2023), that reported the same situation in the major parts of Ethiopia. Approximately 93.1% of the feed supply for livestock in the study area came from natural pastures and crop residues (Table 3). Approximately 45.3% of the 93.1% of the feed supply for livestock came from crop residues. The remaining 6.9% was composed of stubble, agro-industrial by-products, and cultivated forage. This was comparable to the findings of Tolera, (2008), who reported that crop residues accounted for approximately 50% of the total feed supply for ruminant animals in a mixed crop and livestock farming system in the Ethiopian highlands.

Figure 3. Ranked major feed resources in the stud areas.

Table 3. Estimated annual dry matter (DM) and crude protein (CP) supply (t/yr/hh) from various feed sources.

Agroecology		Collected fodder	Crop stubble	Crop residue	Improved forage	Natural pasture	Total
DM	Highland	0.16 ± 0.10^a	0.43 ± 0.20^b	3.24 ± 0.24^c	0.01 ± 0.01^a	4.86 ± 0.25^c	8.69 ± 0.26
	Midland	0.20 ± 0.10^a	0.75 ± 0.21^a	8.50 ± 0.24^a	0.04 ± 0.01^a	6.00 ± 0.25^b	15.49 ± 0.26
	Lowland	0.27 ± 0.10^a	0.88 ± 0.21^a	5.98 ± 0.24^b	0.00 ± 0.00^a	7.81 ± 0.25^a	14.94 ± 0.26
	Mean ± SE	0.21 ± 0.10	0.69 ± 0.21	5.91 ± 0.24	0.02 ± 0.01	6.22 ± 0.25	13.04 ± 0.26
CP	Highland	0.02 ± 0.01^a	0.02 ± 0.03^b	0.07 ± 0.02^a	0.00 ± 0.01^a	0.35 ± 0.05^a	0.46 ± 0.31
	Midland	0.03 ± 0.01^a	0.03 ± 0.03^a	0.38 ± 0.02^b	0.01 ± 0.01^a	0.46 ± 0.05^a	0.91 ± 0.3^a
	Lowland	0.04 ± 0.01^a	0.04 ± 0.03^a	0.28 ± 0.02^b	0.00 ± 0.01^a	0.54 ± 0.05^a	0.90 ± 0.31
	Mean ± SE	0.03 ± 0.01	0.03 ± 0.03	0.25 ± 0.02	0.01 ± 0.01	0.45 ± 0.05	0.76 ± 0.31

Means within the same column with different superscript letters significantly varied (P < 0.05) among agroecologies; SE = Standard error.

The availability of feed had a seasonal correlation with rainfall patterns (Figure 4). In all agroecologies, the natural pasture supply peaked from June to September (the main rainy season) before declining as the approaching dry season began, which is consistent with the findings of (Ayele et al., 2021; Belay & Negesse, 2018). However, it gradually decreased and was completely lost during the dry season, which lasted from December to May. As a result, grass availability and quality decrease to a point where it may not provide enough nutrition for animals to maintain their body weight, resulting in an overall decline in livestock production. This is consistent with observations (Ayele et al., 2021; Desalew, 2008) that feed shortages started in late November and lasted until March, the driest month, during which the growth of natural pastures and fodder shrubs dropped. In the study area, grazing stubbles were the dominant feed resource at the end of the main rainy season (November to December). In addition to natural pasture and stubble grazing, the respondents also harvested forage (weeds) from croplands and roadways, which is consistent with earlier observations (Belay & Negesse, 2018; Yadessa et al., 2016).

Figure 4. Availability of feed resources in relation to rainfall pattern.

When grass from the natural pasture was unable to supply enough feed in the study area, crop residues such as maize stover, Eragrostis teff straw, wheat straw, barley straw, and others (Figure 4) were the main sources of feed from early December to early June, in descending order (Figure 4). Due to the depletion of all feed supplies in the study area from March to early May, this was a crucial period for livestock. For this reason, the only options were trees and unconventional feeding, which was not enough. This pattern of fluctuating feed availability was observed in all agroecologies. Furthermore, respondents and key informants reported that local brewer byproducts were used as additional and alternative feed sources (Desta et al., 2023).

Therefore, it should be recognized that the feeding schedule is a crucial part of controlling and utilizing the feed resources that are accessible in the study locations. The feed was especially plentiful from June to September because of the increased pasture growth, weed growth in annual crops, and crop residues from December to February. Therefore, proper harvesting, storage, and use of crop residues, as well as conservation fodder in the form of hay and silage, would increase the amount of feed that is easily accessible (Ayele et al., 2021). Furthermore, alternative feed quality improvement techniques, such as urea, effective microbes, and scaling-up of improved forage crops, should be used to enhance the nutritional content of the available feed during the dry season (Ayele et al., 2021).

3.4. Estimated annual available feed supply in the study area

The yearly feed supply per household was estimated using feed sources such as natural pasture, crop residue, collected feed from trees, shrubs, weeds within annual crops, and improved forage. The total mean dry matter production of the natural pasture throughout the growing season was 6.22 tons per household per year (Table 3). Similar to the results of Xu et al. (2017), the total dry matter yield per household from natural pastures was significantly higher (P < 0.05) in lowland agroecology than in highland and midland agroecologies. On the other hand, midland agroecologies had significantly higher (P < 0.05) dry matter yields per household than highland agroecologies, which is similar to the findings of Fenetahun and Yong-Dong (2020), who found that aboveground biomass declined as altitude increased. The lower production of dry matter per household in highland and midland agroecologies may be attributable to the ratio of pasture land to household size. In contrast, (Alemayehu, 1985; Fenetahun & Yong-Dong, 2020; Notenbaert et al., 2009; Sisay & Baars, 2002) have reported that the biomass yield of grass species significantly increases with altitude. This variation may be caused by environmental factors (terrain, grazing system, climate, soil variation, and topography) that affect the yield of pasture production in different agroecologies (Pulungan et al., 2019).

The type and quantity of crop residues varied among agroecologies (Figure 5). In contrast to the midland and lowland agroecologies, where Eragrostis teff, maize, and wheat residues were proportionally larger, the majority of crop residues in the highland agroecology were wheat, potatoes, and barley residues (Figure 5), similar to the findings of Desta (2023). The varied proportions of various crop residues in different agroecologies may be related to the different sizes of land allocated to different crops and the various ecological adaptations of crops in cold and hot climates. The residues from the pulse and oil seed crops were minimal compared with the cereal crop residues. This may be the result of insufficient land set aside for the cultivation of pulse and oil seed crops, similar to the findings of Desta (2023).

Figure 5. The proportion (%) of different crop residues.

The proportion of purchased feed, such as balanced diets, grains (maize and oats), wheat bran, and Noug cake, varied among agroecologies (Figure 6). In highland agroecology, maize, oat grain, and wheat bran were the main purchased feeds, whereas in midland agroecology, balanced diets, grains, noug seed cake, and wheat bran were proportionally larger, and in lowland agroecology, an extremely low amount of wheat bran was purchased feed (Figure 6). According to respondents’ reports, limited availability and high prices were the reasons for the varying proportions of different purchasing feeds in different agroecologies.

Figure 6. Contribution (%) of different purchased feed.

The overall, anticipated yearly contribution of dry matter per household was mostly made by natural pasture (6.22 t/hh) and crop residue (5.91 t/hh) (Table 3). Each household obtained 0.91 tons of dry matter annually from less plentiful feed sources such as crop stubble, collected feed, and improved forage. As a result, a total of 13.04 tons of dry matter were obtained annually from the major and minor feed sources in the research area for each household. This is comparable to the findings of Assefa et al. (2013), who reported that every household in Ethiopia’s Adami Tullu Jiddo Kombolcha district had a total dry matter availability of 11.72 tons, and 11.875 tons in the Lalo Kile district of Kellem Wollega zone (Ayele et al., 2021). However, the amount of utilizable dry matter feed produced annually per household varied across agroecologies. In the highland, midland, and lowland agroecologies, the annual sum up of utilized feed supply from the different feed sources per household was around 8.69, 15.49, and 14.94 tons of dry matter, respectively. It was found that the overall average annual amounts of crude protein per household were 0.756 tons (0.46, 0.91, and 0.90 tons in the highland, midland, and lowland agroecologies, respectively). This was significantly more than the amount that was determined by Ayele et al. (2021), who evaluated 0.12 tons of digestible crude protein per household in the Kellem Wollega zone’s Lalo Kile area. However, the amount of digestible crude protein (DCP) per household was considerably lower than the findings of Wondatir (2010), who investigated 21.3 tons of DCP per household in Ziway in Ethiopia’s central rift valley. This difference might be explained by variances in the poor-quality available feeds owing to their high fiber content and lack of protein for livestock (Desta, 2023; Desta et al., 2023) and landholding scenarios.

3.5. Annual estimated dry matter and crude protein requirements of livestock

Table 4 presents the total amounts of dry matter and crude protein required to maintain the total number of animals in each household. The overall average yearly dry matter and crude protein requirements for all livestock for every household for the study areas’ animals were 17.86 and 2.16 tons, respectively, which is comparable to the findings of Ayele et al. (2021), who reported that 17.95 tons of dry matter feed were required per household annually in Lalo Kile Kellem Wollega zone. However, a lower digestible crude protein amount (0.459 tons) per household per year is required in this region (Ayele et al., 2021). Cattle required more proteins (1.53 t/hh) and dry matter (10.72 t/hh) than any other kind of animal. In contrast, the dry matter and crude protein requirements for equines were comparable to those for cattle and were typically higher than those for small ruminants. In the study areas, equines were mostly used to carry agricultural items and people. Therefore, governments and other stakeholders should enhance access to modern transportation to improve feed availability and ruminant animal production.

Table 4. Estimated dry matter (DM) and crude protein requirement (CP) per household.

	Annual DM (t/yr/hh)				Annual CP requirement (t/yr/hh)
Livestock	Highland	Midland	Lowland	Mean	Highland	Midland	Lowland	Mean
Calves	0.32	0.49	0.89	0.57	0.06	0.07	0.13	0.09
Heifers	0.86	1.85	1.11	1.27	0.12	0.26	0.16	0.18
Bulls	0.61	1.18	0.61	0.80	0.07	0.08	0.09	0.08
Cows	2.35	2.38	4.37	3.03	0.36	0.35	0.61	0.44
Oxen	1.50	6.49	7.15	5.05	0.19	0.97	1.07	0.74
Sheep	1.15	0.56	0.55	0.75	0.08	0.04	0.04	0.05
Goats	–	–	0.23	0.23	–	–	0.03	0.03
Horses	3.85	2.43	–	3.14	0.46	0.29	–	0.38
Mules	1.22	–	–	1.22	0.10	–	–	0.10
Donkeys	–	6.84	4.59	5.72	–	0.51	0.34	0.43
Comparative total feed requirement of cattle, equine, and small ruminants (shoat) in the study areas
Cattle	5.64	12.39	14.13	10.72	0.80	1.73	2.06	1.53
Equines	5.07	9.27	4.59	10.08	0.56	0.80	0.34	0.90
Shoat	1.15	0.56	0.78	0.98	0.08	0.04	0.07	0.08
Total	11.86	22.22	19.50	17.86	1.44	2.57	2.47	2.16

The total annual requirements for crude protein and dry matter feed for livestock vary among agroecologies. In the highland, midland, and lowland agroecologies, the annual dry matter and crude protein maintenance requirements were estimated to be 11.86, 22.22, and 19.50 t/hh/yr, and 1.44, 2.57, and 2.47 t/hh/yr, respectively. Highland agroecology has a lower requirement for dry matter and crude protein per household than both midland and lowland agroecologies. This may be because highland agroecologies are likely to have fewer livestock than the midland and lowland agroecologies (Table 2). Compared to the data currently available, (Wondatir, 2010) found a higher quantity of crude protein per household per year of 21.3 tons in Ziway in Ethiopia’s central rift valley. This variation can be attributed to variations in landholding and feed quality.

3.6. Estimated annual feed balance between supply and requirements

Table 5 shows the total amount of dry matter and crude protein required, as well as any gaps between feed supply and livestock feed requirements. The overall average dry matter and crude protein requirement for all livestock per household in the study areas were roughly 17.86 and 2.16 tons per year (t/yr), respectively. The yearly yields of dry matter and crude protein per household were 13.04 and 0.76 t/hh/yr, respectively. Only 73.2% of the dry matter and 35.0% of the crude protein in the study area met the maintenance requirements of livestock. The annual dry matter supply is only approximately 49% in the central highlands of Ethiopia (Tahir et al., 2018), 67.49% in North Achefer (Asmare & Mekuriaw, 2017), 66.13% in the Lalo Kile District of the Kellem Wollega Zone (Ayele et al., 2021). The percentage of protein that met the maintenance requirement of livestock in the current study area was better than that in the Lalo Kile district of Kellem Wollega zone (Ayele et al., 2021), which reported that only about 25.81% of proteins satisfied livestock requirements. However, the percentage of protein that met the maintenance requirements of livestock in the current study area was lower than 62% in the central highlands of Ethiopia (Tahir et al., 2018).

Table 5. The gap between livestock feed demand and supply (t/yr/hh).

Livestock feed supply and demand	Annual DM yields in three agroecology				Annual CP yields in three agroecology
	Highland	Midland	Lowland	Mean	Highland	Midland	Lowland	Mean
Feed Supply (t/yr/hh)	8.69	15.49	14.94	13.04	0.46	0.91	0.90	0.76
Feed demand (t/yr/hh)	11.86	22.22	19.50	17.86	1.44	2.57	2.47	2.16
Deficit (%)	26.72	30.29	23.41	26.81	67.92	64.55	63.72	65.40

t/yr/hh = tons per year per household, DM = dry matter, CP = crude protein.

The annual dry matter supply per household in highlands, midlands, and lowlands was 8.69, 15.49, and 14.94 tons, respectively, while the total dry matter required for all livestock species was 11.86, 22.22, and 19.50 tons. The annual crude protein supply per household in highlands, midlands, and lowlands is 0.46, 0.91, and 0.90 tons, respectively, while the total annual protein needed for livestock maintenance is 1.44, 2.57, and 2.47 tons. Therefore, only 73.28, 69.71, and 76.59% of the dry matter and 32.08, 35.45, and 36.28% of the crude protein met the requirements of livestock for maintenance in highland, midland, and lowland agroecologies, respectively. The dry matter available from the different feed resources per household exceeded the maintenance requirement of livestock in the current study area compared to Debre Berhan, which only satisfied 4.2% (Wondatir, 2010), and Moyale district of Boran, which only satisfied 31.2% (Hassan et al., 2020). However, in the current study, the amount of dry matter available from the major feed resources per household was less satisfied than in the Amhara region, where the amount of feed produced fulfilled 83% of the maintenance requirements (Firew & Getnet, 2010).

3.7. Copying mechanisms of feed shortage in the study areas

The main obstacle for smallholder livestock producers in the study area was a lack of feed, which was mainly caused by seasonal variations in rainfall, expansion of cropland, poor management of natural pasture, the invasion of weeds in the natural pasture, and the low quality of crop residue (Desta et al., 2023). Green feed, which is frequently used as a stable diet during the rainy season, is scarce during the dry season (Figure 3). Therefore, all agroecologies could face a feed shortage as a result of moisture shortages, which is similar to the findings (Tassew & Seifu, 2009), which affected farmers in northeast Ethiopia during both dry and rainy seasons. Farmers in the study areas used the variety of strategies to deal with a shortage of feed during critical times, including conserving crop residues and hay, purchasing feeds, feeding livestock less frequently, reducing the size of their herds, and feeding them locally by product (Figure 7), similar to previous research conducted in the Burie district (Belay & Negesse, 2018). Unless livestock are sufficiently supplied, the poor availability and quality of feed during the dry season tend to make livestock more vulnerable to diseases and physiological stress, which has a negative impact on reproduction, productivity, and considerably longer calving intervals (Ongadi et al., 2020). This implies the necessity for involvement in the production and conservation of livestock feeding practices, particularly hay making, and that appropriate use of locally accessible crop residue is essential to enabling sustainable feed availability throughout the year.

Figure 7. Copying strategies to feed shortages in the study areas.

Conclusions

The results of this study showed that feed availability fluctuates seasonally and is usually plentiful from June to September and December to February. The results showed that the amount of dry matter and crude protein available was insufficient to satisfy the dietary requirements of the livestock in the research area. Therefore, it is suggested that proper collection, storage, and utilization of crop residues, as well as conserving pastures in the form of hay and silage, would improve the amount of feed that is readily accessible. It is also possible that scaling up improved forage crops, improving natural pastures with a variety of methods, and improving crop residues with urea or effective microbes will increase the amount and quality of feed in the study areas.

Consent to participate statement

All respondents provided oral consent to participate in the study.

Ethical approval

According to the local legislation, the study involving human subjects did not require ethical permission; all data were anonymized.

Acknowledgement

The author would like to express his gratitude to Debre Markos University’s research and community service offices for doing this work.

Disclosure statement

The author declares that there is no conflict of interest.

Data availability statement

The data for this study is available from the author upon reasonable request.

Additional information

Notes on contributors

Alemu Gashe Desta

Alemu Gashe Desta is a lecturer and researcher at Debre Markos University’s Department of Animal Sciences. He received his Bachelor of Science in Animal Sciences from Debre Markos University and his Master’s degree in Rangeland Ecology and Management from Haramaya University in Ethiopia. Since September 2011, he has taught various animal science courses at the Debre Markos University, Ethiopia, including forage production, pasture management, rangeland ecology and management, camel production, livestock production, sericulture, swine production, and animal power and technology. Research interests include forage and pasture production technology, livestock production, rangeland restoration, soil fertility management, animal nutrition and food security. Livestock is a key component of the food and nutrition security of Ethiopian communities. However, their production and productivity face challenges including the lack of animal breeds, capital outlay, imbalanced feed, animal disease, and climate change. Feed imbalance is one of the primary factors affecting livestock productivity and production. Therefore, Ethiopia’s livestock production can be enhanced by focusing on high-quality feed production and enhancing natural pastures and range production.