The value of traditional milk products among smallholder farmers of Southern Ethiopia: handling, consumption and making of butter

Abstract

Although earlier studies reported on traditional dairying, there is limited information on variation in milk and milk products handling, production, and consumption across various dairy production systems (DPS) in Southern Ethiopia, which is why this study was initiated. This study assessed 360 smallholder dairy producer households across four DPSs in the Sidama and the former Southern Nations, Nationalities, and People’s (SNNP) Regional States of Ethiopia: cereal-based (CB) diversified crop-based (DCB), Ensete-based (EB), and cash crop-based (CCB). Over two-thirds of respondents in CCB use improved milk handling devices like aluminum cans and Mazzicans, compared to 21% in EB. Eucalyptus globulus were primarily used to clean milk equipment in the CB, DCB and EB DPSs, whereas enset plant residue (‘kacha’ in Amharic) was specific to CCB. Olea africana was commonly used to smock milk containers across all systems. More milk (18 liters) was required to produce one kg of butter from crossbred cows, while local breeds required 13 liters on average. In the CB system, crossbred cows needed a minimum of 15 liters of milk to produce one kg of butter. Higher annual milk production per household was reported in the CCB, with the lowest in the DCB. Whole milk consumption did not significantly vary among the DPSs (P < 0.05). On average, 38% of the produced milk was processed into traditional butter, with the lowest proportion (27%) in CCB and the highest (42%) in DCB. These findings suggest the need for tailored development options in milk handling, processing and butter making along with the introduction and adoption of improved handling technologies for the different DPSs in Southern Ethiopia.

IMPACT STATEMENT

This study explores the traditional practices of handling, consuming, and making butter from milk among smallholder farmers in Southern Ethiopia. By examining the methods used across various dairy production systems, including cereal-based (CB), diversified crop-based (DCB), enset-based (EB), and cash crop-based (CCB), we uncover the unique and resourceful ways farmers manage their milk products. The findings highlight significant variations in the use of different plant parts and methods for cleaning milk equipment and extracting butter, reflecting the cultural richness and diversity of these practices. Additionally, the study reveals differences in milk and milk product consumption, suggesting opportunities to improve household nutrition. Based on these existing practices, we propose tailored improvements in milk handling and processing to enhance the quality of milk produced, which could positively impact the nutrition and livelihoods of smallholder farmers in the future.

Keywords:

• Consumption behavior
• dairy production system
• handling
• washing plants
• smoking plants

REVIEWING EDITOR:

• Pedro González-Redondo, University of Seville, Spain

SUBJECTS:

• Science
• Food Science & Technology
• Dairy Science
• Food Chemistry
• Food Analysis
• Nutrition

1. Introduction

Milk and other dairy products have a great role in the diet of family nutrition. The contribution is primarily made through the production of dairy cattle. In Ethiopia, cow milk contributes around 66% of the milk production (CSA, 2022). Cattle production contributes to Ethiopia’s food and nutrition security, revenue generation, used as an organic fertilizer for agricultural productivity, and as a source of draft power (Amenu et al., 2019). Milk plays a vital role in fostering growth and development of children and preventing malnutrition (FAO, 2013; Tuba et al., 2022). It is a plentiful source of protein and calories, offering between 34 and 61 kcal per 100 g depending on the amount of fat present. Despite the significance of milk and other dairy products and the huge cattle population in the country, there are varying reports on the population’s per capita consumption of milk in Ethiopia. According to TRAIDE Ethiopia (2021), the per capita consumption was 19 kg per year. However, Yilma et al. (2017) reported a higher figure of 36.5 kg per year. Additionally, FAOSTAT (2020) indicates that the per capita consumption is 30 kg per year for Ethiopia, with comparable figures for Uganda, Djibouti, and Rwanda being 38, 26, and 14 kg per year, respectively. Looking ahead, TRAIDE Ethiopia (2021) projects that by 2030, the per capita milk consumption is expected to increase to 27 kg in Ethiopia, 65 kg in Uganda, and 115 kg in Kenya, whereas FAOSTAT’s report suggests even higher figures. This is considerably less than the average annual global consumption of 87 kg (FAO, 2023) and the average annual per capita consumption of 200 kg reported by Azage (2018) in the newsletter.

There have been a number of factors cited as limitations for Ethiopia’s low consumption of milk and other milk products. Unsatisfactory consumption of milk and other dairy products was caused by the growing human population, low productivity of indigenous cattle breeds, and low annual milk production. According to Dorp (2014) production directly affects consumption and alters the nutritional condition. The consumption of dairy products is on the rise due to reasons including economic expansion and rising income (FAO, 2013).

In Ethiopia, fresh whole milk is either consumed raw or processed into a variety of traditional products including ‘ergo’ (spontaneously fermented sour milk without adding defined starter culture, stabilizers and without pasteurization), ‘arera’ (defatted sour milk), butter, ghee, and ‘ayib’ (traditional cottage cheese). Traditional manufacturing techniques are mainly used to produce processed dairy products throughout the country. According to several study findings, almost half of milk produced at the household level in Ethiopia’s rural majority is processed rather than used as fluid milk (Abebe et al., 2020). The same finding revealed that around 44.6% of milk produced was used for domestic processing, and from this, approximately 75% and 25% also goes to the manufacturing of butter and making of Ethiopian cottage cheese, respectively. Similar to this, about 40% of the country’s total milk production is processed into butter (Alganesh & Tola, 2017).

Traditional techniques of handling and processing milk in Ethiopia involve manual operations from milking to processing, requiring particularly hygienic practices. The degree of processing varies across regions, resulting in a variety of milk products available throughout Ethiopia. The degree of processing varies across regions, resulting in a variety of milk products available throughout Ethiopia. One of the main elements determining the quality of dairy products according to Gonfa et al. (2001) is whether good milking procedures are carried out and clean milking utensils are used. According to Ayenew et al. (2009), water, shrubs, and smoking accounts for 73% of cleaning and disinfection materials and methods. Ethiopians usually smoke milk vessels perform work prior to milking and processing to disinfect to maintain the flavour, aroma, and to extend the shelf life of milk and milk products.

Different reports have confirmed that the ancient practice of smoking of milk containers was employed as a method of preservation and to enhance the flavour of milk and milk products (Alemu & Girma, 2018b; Asefa & Abrha, 2022; Wanjala et al., 2016). These studies also noted that storing raw milk in smoked containers reduces microbial load due to the inhibitory effects of smoke compounds, potentially contributing to the extended shelf life of milk and milk products. Additionally, fumigation has been shown to reduce the pH level of milk (Lemma et al., 2024).

Despite several studies on the handling, processing, and consumption of milk and milk products, information on the systemic level of dairy production was scarce. Therefore, this study aimed to assess ­butter production status, milk and milk product ­consumption behavior, and handling practices within different dairy production systems in Southern Ethiopia.

2. Material and methods

2.1. Study area

The research was carried out in the Sidama and the former Southern Nations, Nationalities, and People’s (SNNP) Regional States of Ethiopia, prior to their division in to South, Central and other regions. Sidama is located between 5°45’ to 6°45′N latitude and 38°5’ to 39°5′E longitude while the SNNP region spans from 4°26.03106′ to 8°27.4572′N latitude and 35°45.2214′E to 38°42.96108′E longitude. These regions are characterized by two primary agro ecological zones: midland and highland. The midland zone ranges from at 1500 to 2300 meters above sea levels, whereas the highland zone extends from 2300 and 3200 meters above sea level (Etana et al., 2020). The annual average temperature in Sidama ranges from 8.5° to 27.2 °C while in the SNNP region; it varies between 21 °C and 26 °C. These areas are renowned for their high potential for livestock production; with local breeds constitute approximately 97% of Ethiopia’s total cattle population (CSA, 2022).

2.2. Sampling of farm households

Based on a survey conducted between February and April 2022, 360 smallholder farmers from six districts (Melga, Wondogent, Damot Gale, soddo zuria, kedida gamela, and Domboya) in the Sidama and the former Southern Nations, Nationalities, and People’s (SNNP) Regional States of Ethiopia were categorized into four Dairy Production Systems (DPS): cash crop based (CCB), enset based (EB), cereal based (CB), and diversified crop based (DCB) (Tsedey et al., 2024). The study focused on gathering information about dairy products handling and consumption, with a purposeful selection of over 75 respondents who were primarily young females or women reflecting their predominant role in dairy product handling and processing in Ethiopia and similar study areas (Edemo, 2017; Ulfina et al., 2019).

2.3. Data collection and analysis

A semi-structured questionnaire was prepared and pretested before the actual survey. Both secondary and primary data sources were used for the study. The household heads and their spouses were interviewed face to face as the primary data sources. These interviews were conducted in person to ensure comprehensive data collection. Secondary data sources, we conducted face to face interviews with representatives from the agricultural offices in the respective selected districts within the regions. This interview provided valuable contextual and supplementary information that complemented the primary data. The questionnaire contained detailed sections on various topics, including household characteristics (sex, religion, level of education, and primary occupation), milk storage, milking and processing equipment, native plants used for cleaning and smoking of equipment, butter making practices, and annual milk production and consumption practices of milk and other milk products.

A one-way analysis of variance (ANOVA) was used to assess variation in yearly milk production and consumption, percentage of milk processed to make butter, and butter manufacturing practices across different dairy production systems. Qualitative data, such as the identification of milk storage and processing equipment and the consumption of milk and milk products across diverse dairy production systems, were analyzed using chi-square tests. The statistical package for social sciences (SPSS) version 25 was used for all analysis, with a significance level set at P < 0.05 to determine variations among variables across the production systems.

3. Results

3.1. Sample households characteristics

The results of characteristics of households across the dairy production systems in the study areas are shown in Table 1. There were more male household heads in the four dairy production systems, and women constituted the majority of study participants. Women respondents were interviewed to generate information on milk and milk product handling, processing, consumption, and butter production. With the exception of the enset-based (EB) dairy production system, more than 50% of the wives who were interviewed had completed primary school or higher education. Additionally, across all the four diary production systems, nearly 80% of study respondents depend on agriculture as their main occupation and source of income. In terms of religion, most respondents across all four dairy production systems were Protestant.

Table 1. Characteristics of sample households across the dairy production systems in the study area (frequency and percentage).

Description	CCB (N = 60)		EB (N = 101)		CB (N = 87)		DCB (N = 112)
	F	%	F	%	F	%	F	%
Sex of household heads
Male	59	98	92	91	78	90	98	88
Female	1	2	9	9	9	10	1	12
Sex of respondents
Female	57	95	98	97	82	94	108	96
Male	3	5	3	3	5	6	4	4
Religion of respondents
Orthodox	17	28.3	17	28.3	19	21.8	21	18.7
Protestant	39	65	79	78.2	63	72.4	84	75
Muslim	2	3.3	4	3.96	4	4.6	1	0.9
Catholic	2	3.3	1	0.99	1	1.2	6	5.4
Educational status of household wives
Illiterate	15	26	43	44	34	42	41	38
Read and write	10	18	10	10	3	4	4	4
Primary school	25	44	30	31	26	38	38	35
Secondary school	5	9	11	11	14	17	20	19
TVET/Diploma	0	0	3	3	3	4	2	2
Some university degree	2	4	1	1	2	2	3	3
Major occupation of the household
Government employee	1	1.4	7	5.5	11	10	4	3.03
Private sector	5	6.8	5	3.9	2	1.8	5	3.8
Agriculture	58	79.5	101	78.9	87	79.1	110	83.3
Off-farm activities	9	12.3	15	11.7	10	9.1	13	9.8

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based; TVET; Technical and Vocational Education and Training; F: frequency.

3.2. Milk and milk product handling

3.2.1. Milking, storage, and processing utensils

Table 2 shows various milking, storage and processing equipment used in the four dairy production systems. The use of equipment such as aluminum cans, mazzican, and conventional equipment (clay pot) varied significantly (p < 0.05) among the four dairy production systems. Accordingly, the majority of respondents in the CCB use food grade mazzican; whereas none of the respondents in the enset based dairy production system reported using such utensils. Additionally, aluminum cans are predominantly used in all systems except the enset-based system. In contrast to the majority of respondents in the enset-based dairy production system, none of the respondents in the cash crop-based (CCB) system reported using traditional equipment like clay pot for milking and storage. However, in all dairy production systems (DPSs) across the study locations, clay pots were the main equipment used for butter making. Respondents in all DPSs indicated that butter was stored and packaged mainly using enset plant leaves. In the CCB system, 35% of the study respondents used poly ethylene plastic bags for butter packaging.

Table 2. Handling materials for milk and butter in the study areas (frequency and percentage).

Description	CCB (N = 60)		EB (N = 101)		CB (N = 87)		DCB (N = 112)		χ^2/p-Value
	F	%	F	%	F	%	F	%
Milking, storage and transportation containers
Aluminum cans	30	47	7	21	9	38	14	35	15.26/<0.001
Mazzicans	17	27	0	0	4	17	4	10	18.93/<0.001
Plastic containers	17	27	15	46	8	33	12	30	0.17/0.982
Clay pot	0	0	11	33	3	13	10	25	15.74/<0.001
Butter processing material
Clay pot	57	97	96	92	87	100	112	98	10.11/0.023
Wooden containers	0	0	4	3.9	0	0	0	0	14.40/<0.001
calabash	0	0	3	2.9	0	0	0	0	7.81/0.054
Improved (manual churn)	2	3.4	0	0	0	0	1	1	2.476/0.480
Butter storage and packaging material
Poly ethylene plastic bags	12	35	3	3.6	6	7.9	2	2	36.88/<0.001
Enset leaves	22	65	81	96	70	92	88	98	64.34/<0.001

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based; F: frequency Calabash: a traditional processing device which is a round gourd from calabash tree.

3.2.2. Indigenous plant materials for cleaning milk utensils in the study areas

In the research areas, smallholder farmers have a long-standing custom of cleaning milk storage, milking, and processing equipment with various plant leaves. Table 3 details the specific types of plants used for this purpose and their prevalence among dairy productions systems. Plant leaves were crumpled up and used as sponges throughout the washing procedure. The top three plant leaves used for cleaning milking and storing utensils according to the interviewed smallholder farmers were ‘nech bahirzaf’ (Eucalyptus globulus Labill.), ‘koseret (Lippia abyssinica), and tenadam (Ruta chalepensis L). Among the plants, ‘nech Bahirzaf’ was reported as the primary choice for cleaning milking and storage equipment in all DPSs, except in the CCB, where most respondents indicated they do not use it for cleaning milk utensils. However, respondents in the CCB reported using kacha, the remnant of the ‘enset’ plant (Ensete ventricosum), for the same purpose. Among the different dairy production systems, the primary purpose for cleaning milk vessels and churns was reported to be increasing the efficiency of cleaning, with 54% in the CCB system, 40% in the EB, 39% in the DCB, and 36% in the CB. This was followed by other purposes, such as improving the aroma and taste of the milk. Additionally, a smaller proportion (5%) of study respondents in the EB indicated that using plant leaves for cleaning milk equipment enhances the efficiency of milk fermentation and increases butter recovery from buttermilk.

Table 3. Plant types used for cleaning milk vessels and churners across the dairy production systems in the study areas (frequency and percentage).

Description		Scientific names	CCB (N = 60)		EB (N = 101)		CB (N = 87)		DCB (N = 112)		Overall (N = 360)
			F	%	F	%	F	%	F	%	F	%
Types of plant leaves for cleaning milk vessels and churners
Nech Bahirzaf	(Eucalyptus globules Labill.)		14	21	53	28	64	34	85	38	216	60
Koseret	(Lippia abyssinica)		3	4	36	19	39	21	45	20	123	34.2
Kesse	(Lippia adoensis)		0	0	7	4	5	3	0	0	12	3.3
Tenadam	(Ruta chalepensis L)		4	6	31	16	26	14	39	18	100	27.8
Enset/kacha	(Ensete ventricosum)		29	43	33	17	5	3	12	5	79	21.9
Datata (Sidama)	(unidentified)		2	3	8	4	2	1	1	1	13	3.6
Tosigne	(Lhymus schimperi)		0	0	9	5	4	2	1	1	14	3.9
Miche  Medihanit	(Ocimum lamiifolium Hochst. ex. Benth.)		7	10	3	2	1	1	4	2	15	4.2
Tside	(Juniperus procera L.)		7	10	3	2	6	3	7	3	23	6.4
Nokota (Kembata)	(Unidentified)		0	0	3	2	7	4	1	1	11	3.1
Zigiba	(Podocarpus gracelior)		1	2	0	0	3	2	0	0	4	1.1
Bisana	(Croton macrostachyus)		0	0	0	0	5	3	0	0	5	1.4
Gulo	(Ricinus communis L.)		1	0	6	3	22	12	25	11	54	15
Farmers’ perception towards the importance of using parts of plant leaves for cleaning milk utensils
Improves taste			19	19	62	26	71	30	85	30	237	65.8
Improves aroma			26	27	74	31	76	32	87	31	263	73.1
Increases cleaning efficiency			53	54	96	40	84	36	110	39	343	95.3
Increases churning efficiency			0	0	11	5	4	2	2	1	17	4.7

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based; F: frequency.

3.2.3. Smoking milking and churning utensils

Table 4 illustrates practices of fumigation of milk containers and local churn across DPS in the study areas. Across all dairy production systems, Olea africana was the plant used most frequently (49%) for fumigating milk storage, milking, and processing utensils., In the DCB and CB dairy production systems, wetecha (scientific name unidentified) and Nech Bahirzaf (Eucalyptus globulus Labill) were the second and third most commonly used plants, respectively. In contrast, habesha tside (Juniperus procera L.) followed Woira (Olea africana) as the second most used plant in CCB and EB DPSs. These findings highlight significance difference in the choice of fumigating plants among the dairy production systems. Parts of the plant that serve as fumigation purpose have been reported during the interview. Plant leaves were often used first, then the steam section of the plant. Seeds of Eucalyptus globulus Labill. were used for fumigation of milk equipment. The primary purposes of fumigation using different parts of plants included improving the aroma, followed by enhancing the taste of milk and milk products. A smaller proportion of respondents indicated the bacteriostatic effect and the increased efficiency of coagulation as additional purposes of fumigation.

Table 4. Practice of fumigation of milk containers and local churns across DPS in the study areas (frequency and percentage).

Description	Scientific name		CCB (N = 60)		EB (N = 101)		CB (N = 87)		DCB (N = 112)			Overall (N = 360)
			F	%	F	%	F	%	F	%		F	%
Fumigating plants
Woira	Olea africana		45	63	44	38	36	36	53	41		178	49
Bamboo	Oxytenanthera abyssinica		0	0	3	3	3	3	4	3.1		10	3
HabeshaTside	Juniperus procera L.		13	18	44	38	10	10	19	15		86	24
Nech Bahirzaf	Eucalyptus globules Labill.		4	6	6	5	15	15	24	19		49	14
Wetecha (kembata)	unidentified		7	10	13	11	23	23	19	15		62	17
Gulo	Ricinus communis L.		1	1	3	3	7	7	5	3.8		16	4
Koso tree	Hagenia abyssinica		1	1	3	3	7	7	4	3.1		15	4
Tikur Enchet	Pygeum africanum		0	0	0	0	0	0	2	1.5		2	1
Parts of plant used for fumigation
leaves			0	0	17	16	23	24	29		22	69	19
Steams			58	97	84	79	63	66	78		59	283	79
Seeds			2	3	6	5	10	10	26		20	44	12
Purposes of fumigation
Bacteriostatic effect			2	2	12	7	5	4	6		4	25	7
Improves taste			56	49	63	38	51	45	72		43	242	67
Improves aroma			57	50	86	52	54	48	86		51	283	79
Increases the efficiency of milk  coagulation			0	0	3	2	3	3	4		2	10	3

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based; F: frequency.

3.3. Making of butter

The utilization of milk from local and crossbred cows, as well as the corresponding butter yield per churning and per kg of milk, exhibited notable variations across the four dairy production systems studied (Table 5). These differences highlight the distinct practices and outcomes associated with milk production and butter processing within each system. According to study respondents, to make one kilogram of butter 13 and 18 kg of milk of local and crossbred cows is required, respectively. According to the interview response, the CB production system used the highest quantity of milk (14.5 liter) from local cows to make a kilogram of butter, while the EB system used the least amount of milk (12.5 liter) to produce the same amount of butter. Furthermore, the milk from crossbred cows used to make a kilogram of butter showed no significant variation at P < 0.05 across the four dairy production systems. However, those respondents in DCB dairy production system reported that they used least volume of milk from crossbred cows for a single churning. The respondents in the EB system indicated that butter is made every 7 days and there is prolonged milk accumulation before making of butter. The proportion of milk that converted into butter varied significantly at p < 0.05 across the four dairy production systems. The CCB system had the lowest conversion rate, with only (27%) of the milk processed into butter, while the other three systems each had a conversion rate of approximately 40% (Figure 1).

Figure 1. Proportion of milk converted to butter across Dairy production systems.

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based.

Table 5. Smallholder farmers butter making practices across dairy production system in the study areas.

Butter making practice	CCB (N = 60)		EB (N = 101)		CB (N = 87)		DCB (N = 112)		Overall (N = 360)
	N	M + SE	N	M + SE	N	M + SE	N	M + SE	N	M+SE
Using local breeds
Milk used per churning (liter)	27	4.11 + 0.36^ab	67	4.28 + 0.28^a	44	3.32 + 0.33^b	62	3.90 + 0.26^ab	200	3.93 + 0.15
Butter produced per churning (kg)	27	0.63 + 0.29^a	67	0.37 + 0.03^b	44	0.24 + 0.03^b	62	0.31 + 0.02^b	200	0.36 + 0.04
Milk used to produce a kg of butter (liter)	27	12.78 + 1.1^ab	67	12.47 + 0.41^a	44	14.5 + 0.61^b	62	13.71 + 0.50^ab	200	13.35 + 0.29
Using Crossbred
Milk used per churning (liter)	33	8.12 + 0.68^a	36	7.06 + 0.72^a	43	7.2 + 0.65^a	53	6.37 + 0.48^b	165	7.09 + 0.31
Butter produced per churning(kg)	33	0.49 + 0.04	36	0.40 + 0.05	43	0.50 + 0.05	53	0.41 + 0.04	165	0.45 + 0.02
Milk used to produce a kg of butter(liter)	33	16.9 + 0.7^abc	36	19.47 + 1.08^ac	43	15.09 + 0.6^b	53	18.62 + 1.75^c	165	17.56 + 0.65
Butter making frequency (month)	58	9.50 + 0.71^b	98	7.22 + 0.32^a	86	8.84 + 0.53^b	112	9.32 + 0.50^b	354	8.65 + 0.25
Milk fermentation time before churning (days)	58	1.95 + 0.08^b	98	2.30 + 0.07^a	86	2.2 + 0.08^ba	112	2.03 + 0.07^b	354	2.13 + 0.4

Values with different letter across the indicated significant difference at p < 0.05; Variables with no superscript letter means there is no significant difference.

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based.

3.4. Consumption of milk and milk products

3.4.1. Consumption behavior

Smallholder farmers produce milk and milk products for home consumption from their own sources in all dairy production systems (Table 6). In addition to using their own cattle as the main source of milk and other dairy products for domestic consumption, 10% and 7% of the respondents in CCB and EB production systems, respectively, obtained milk and other dairy products from their neighbors or relatives either as gift or through borrowing. Borrowing in this context implies receiving milk and dairy products for free, with respondents typically returning the favor when their own pregnant cows give birth and produce milk. The respondents indicated that only a small proportion of them consumed purchased milk and milk products. Furthermore, a few respondents in the three production systems, except for the DCB system, sourced milk and dairy products from both their own cattle and through purchasing. Milk and milk products for home consumption were primarily sourced from indigenous breeds only in the EB (57%) and DCB (54%) dairy production systems. In contrast, those in the CCB and CB systems primarily consumed milk only from crossbred cows. More than half of the respondents reported a constant trend of decline in milk and milk products consumption in the DCB and EB dairy production systems, whereas the consumption of milk and milk products increased over the past five years during 2017–2021 in CCB and CB dairy production systems. Increased consumption of milk and milk products was observed across all dairy production systems, with more increment in the CCB system. In both the CCB and CB production systems, the primary reason cited for this increase was the ownership of crossbreeds. Additionally, a notable proportion of respondents indicated that both owning improved breeds and increasing the number of crossbreds contributed to the rise in consumption.

Table 6. Milk and milk products consumption behavior across dairy production system in the study areas (frequency and percentage).

Description	CCB		EB		CB		DCB		χ2/P-Value
	F	%	F	%	F	%	F	%
Sources of milk and milk products for home consumption
Own source (a)	51	85	90	89.1	82	94.3	105	94	11.09/0.272
Purchased (b)	2	3.3	2	2	0	0	4	4
Both (a & b)	1	1.7	2	2	2	2.3	0	0
Acquired by other meansa	6	10	7	6.9	3	3.4	3	3
Sources of milk for consumption by breed type
Indigenous breeds	25	41.7	58	57.4	41	47	60	54	7.99/0.244
Crossbreds	27	45	32	31.7	40	46	44	39
Both	8	13.3	11	10.9	6	7	8	7
Consumption status of milk and milk products during the last five years (2017–2021)
Increasing	30	50	38	37.6	42	48.3	48	43	4.80/0.573
Decreasing	16	26.7	38	37.6	22	25.3	34	30
Constant (no change)	14	23.3	25	24.8	23	26.4	30	25
Reasons for increased milk and milk products consumption during the last five year (2017–2021)
Ownership of crossbreds (a)	15	50	26	36.8	54	62	18	38	29.778/<0.001
Increased no. of crossbreds (b)	1	3.3	1	39.5	2	2.3	14	29
Both (a & b)	14	46.7	15	23.7	31	35.7	16	33

^a Acquired by other means is through a gifts and borrowing.

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based; F: frequency.

3.4.2. Production and consumption trend and frequency of milk and milk products

Table 7 presents the frequencies of consumption for milk and other milk products across dairy production systems. The types of dairy products consumed vary from location to location depending on cultural differences and the types of traditional foods consumed. The products those were intended for consumption in the study locations were butter milk, whole milk, spontaneously fermented milk (‘ergo’), cottage type cheese (‘ayib’), and local butter, although the frequency of consumption differed among dairy production systems. However, despite significant variation in milk production across the four dairy production systems (Figure 2), with the CCB system showing the highest annual production, consumption behavior showed no significant variation among these systems. Whole milk consumption is reported to be more frequent in the CCB system compared to the other three dairy production systems. In the CCB system, 35% of respondents consume whole milk on a daily basis, and 51% consume it twice per week, making a total of 86% for these two high frequency categories, followed by the EB system in this similar category. Frequency of consumption varied among the four dairy production systems for all milk and milk products except for ‘ergo’, with the variation observed being statistically significant at p < 0.05. More than 75% of the respondents across the four dairy production systems indicated that buttermilk is consumed on a daily to twice weekly. The frequency of consumption of ‘ayib’, local butter, and ‘ergo’, was generally on a weekly basis across the four dairy production systems, with occasional periods of more than a week without consumption.

Figure 2. Annual fluid milk production and consumption among dairy production system.

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based.

Table 7. Frequency of consumption of milk and milk products in the study areas (frequency and percentage).

Consumption frequency	CCB		EB		CB		DCB		χ^2/P-Values
	F	%	F	%	F	%	F	%
Whole milk
Daily	19	34.5	24	28.2	7	10.3	12	13.8	49.461/<0.001
Twice per week	28	50.9	20	23.5	21	30.9	21	24.1
Weekly	3	5.5	19	22.4	16	23.5	30	34.5
Twice per month	3	5.5	17	20	22	32.4	21	24.1
Every month	2	3.6	5	5.9	2	2.9	3	3.4
Yoghurt (Ergo)
Daily	0	0	0	0	1	1.9	0	0	14.506/0.488
Twice per week	9	20	11	16.9	6	11.1	8	12.3
Weekly	21	46.7	27	41.5	23	42.6	27	41.5
Twice per month	13	28.9	20	30.8	16	29.6	27	41.5
Every month	2	4.4	7	10.8	8	14.8	3	4.6
Cheese (Ayib)
Twice per week	0	0	3	6.3	6	9.2	1	1.3	30.399/0.002
Weekly	2	11.8	13	27.1	24	36.9	30	38.5
Twice per month	6	35.3	19	39.6	30	46.2	36	46.2
Every month	9	52.9	13	27.1	5	7.7	11	14.1
Butter
Daily	0	0	9	9.4	3	3.6	16	14.3	25.048/0.015
Twice per week	4	7.4	9	9.4	9	10.7	10	8.9
Weekly	28	51.9	58	60.4	47	55.9	50	44.6
Twice per month	12	22.2	15	15.6	20	23.8	26	23.2
Every month	10	18.5	5	5.2	5	5.95	10	8.9
Butter milk
Daily	27	49.1	39	43.8	20	22.7	50	45	31.075/0.002
Twice per week	19	34.5	45	50.6	49	55.7	43	38.7
Weekly	2	3.6	5	5.6	16	18.2	16	14.4
Twice per month	2	3.6	0	0	3	3.4	2	1.8

CCB: Cash crop based; EB: Enset based; CB: Cereal based; DCB: Diversified crop based; F: frequency.

4. Discussion

Smallholder farmers in the study areas consume milk and milk products together with other diets on a regular basis. In the current study, smallholder farmers within diverse dairy production systems (DPSs) were identified to use various materials for handling milk and butter, along with specific plant-based sanitation techniques (Tables 2 and 3). Smallholders in the cash crop based (CCB) production system showed better use of modern devices for milking, storage and transportation compared to those in the enset based (EB) system. The observed disparity in milking, storing and transporting device usage between smallholders in the cash crop based (CCB) dairy production system, known for their higher annual income (Tsedey et al., 2024), and those in the enset based (EB) system, characterized by lower annual incomes, may reflect differing access to modern technology, with the CCB system demonstrating superior utilization of modern devices compared to the EB system. The study by Barrett et al. (2012) highlights that increased farm income enhance the adoption of new agricultural technologies. The may also be a factor in the decision to purchase and own modern milking and storage equipment in the CCB system, compared to the EB system, may also be influenced by factors such as higher milk production volumes (Figure 2). Furthermore, Tsedey et al. (2024) noted that the EB production system’s distance from the market could restrict access to various technologies, a concern similarly highlighted in Andrade et al. (2019). Contrary to the adoption of advanced milking and storage devices, smallholders across dairy production systems have not upgraded their traditional butter churns or packaging material for butter.

Smallholder farmers use plants to wash and fumigate milking and processing equipment in all production systems in the examined locations. In all DPSs, it was common practice to crumple up the leaves and use them to clean storage, milking, and processing equipment. The plant leaves used for cleaning milk vessels and churns include ‘nech bahirzaf’ (Eucalyptus globulus Labill.) and ‘koseret’ (Lippia abyssinica), with the exception of the CCB dairy production systems, where enset plant residue known as ‘kacha’ was predominantly utilized (Table 3). These findings differ from those of Negash (2018), who studied in the highlands of South Tigray and identified Achyranthes aspera and Cucumis prophetarum as the main plants used. However, our findings align with those reported by Negash et al. (2012) in Ethiopia’s central Rift Valley and Temesegen (2018) in Oromia region. Lippia adoensis species which was used the least in the current study, was reported as main plant used for washing milk utensils in northern and northwestern Ethiopia (Alemu and Girma, 2018a; Seifu & Tassew, 2014). The majority of respondents across all dairy production systems indicated that using these plants for cleaning milk equipment and churns improved the efficiency of cleaning and enhanced the aroma of milk and other dairy products such as butter. This finding aligns with the study by Alemu and Girma (2018b), which reported that the use of plants for washing milk vessels not only extends fermentation time but also creates a good flavor for the milk. Similarly, Negash (2018) highlighted that plant-based cleaning methods serve multiple purposes, including the de-fragmentation of yogurt from rarely souring problems, providing multi-medicinal value to livestock, and acting as disinfectants. Furthermore, Alganesh and Tola (2017) noted that cleaning with plant parts imparts flavor into milk containers and churns, supporting our finding that plant use enhances cleaning efficiency and contributes to improved sensory qualities of dairy products.

In all dairy production systems, plant leaves were reported to be used not only for washing but also for fumigation of milk vessels. Olea africana was identified as the most frequently used plant for fumigating milk storage, milking, and processing equipment across these systems. Similarly, Juniperus procera was used equivalently for fumigation purpose in EB dairy production system, while in the CCB dairy production system, it served as the second most frequently used plant for fumigation. Many studies have also reported Olea africana as a major plant used for fumigating milk vessels and churns (Alemu and Girma, 2018a; Lemma et al., 2024; Teshome et al., 2019; Tsadkan & Belay, 2018). The purposes of fumigating milk vessels and churns were investigated in the current study, revealing that the primary aim was to improve the flavor and taste of milk and milk products. This finding aligns with Alemu and Girma (2018b), who noted that plant-based fumigation can extend fermentation times and enhance milk flavor. Similarly, Alganesh and Tola (2017) observed that fumigation with plant parts imparts a special flavor and odor to dairy products. These consistent findings underscore the traditional knowledge and practical benefits associated with using plant-based methods for enhancing dairy product quality through fumigation.

Although a larger volume of milk (14.5 liters) requirement from local cows in the CB production system was required to produce one kilogram of butter compared to the other three systems, slightly a smaller volume of milk (15 liters) from crossbred cows in the same system achieved the same yield (Table 6) This difference may be attributed to the higher presence of Jersey crossbred cows in the CB system, as noted in the study by Tsedey et al. (2024). Additionally, Yoseph et al. (2022) also highlighted that Jersey cows in rural mixed crop-dairy systems, with the access to a higher roughage diet, may exhibit diverse fat content that enhances butter production. Moreover, other studies have underscored that factors like breed, parity, lactation stage, and feed type significantly influence fat content, crucial for improving butter yield (Musema, 2022). The practice of butter making in the CCB was less compared to the other three systems, with a lesser proportion of milk converted to butter (28%) and an overall proportion of 39% across all four dairy production systems (Figure 1). The overall proportion converted to butter in this study was lower than the 62% reported by Berhanu et al. (2014) and comparable to the national rate of 40% as reported by Government of Ethiopia Livestock Master Plan as cited in Mebrate et al. (2019).

Smallholder farmers in the four DPSs primarily relied on their own sources for house hold milk and other milk products consumption. Recent findings indicated that crossbred cow milk was commonly consumed in CB and CCB systems, whereas local cow milk predominated in EB and DCB systems (Table 6). Despite producing more whole milk compared to the other three systems, the CCB dairy production system had the lowest consumption rate (15%) relative to its production, with the highest consumption rate of 33% reported in the EB system (Figure 2). This implies that increasing milk production alone cannot elevate consumption levels under smallholder farmer conditions. Remarkably, all systems, particularly CCB, significantly lagged behind the national figure, where 50% of milk production was consumed. In comparison, the SNNP region consumed 37% of its milk production, and the Sidama area consumed 44% (CSA, 2020). Despite the lower proportion of milk consumption in the CCB system, respondents reported consuming milk more frequently, with most consuming it daily to at least twice weekly. The most commonly used byproduct across all systems was buttermilk. Various studies have identified several factors that limit the consumption of dairy products. These factors include ownership of improved dairy cattle (Hoddinott et al., 2013; Melesse & Beyene, 2009), an increase in the number of lactating cows (Duguma, 2022), and the household’s income and level of education (Farrell, 2021; Kapaj, 2018). These findings also support the observed low consumption in the current study.

Additionally, this study found that the majority of respondents in all four dairy production systems were protestant, a religion that did not practice fasting from animal source foods. This contrasted with Ethiopian Orthodox believers, whose religious fasting leads to lower milk consumption (D’Haene et al., 2021; TRAIDE Ethiopia, 2021). Butter milk was more frequently consumed in all DPSs as compared to other dairy products. This might be because the staple diet in the study area (the southern part of Ethiopia), made from the enset plant, and was typically consumed with buttermilk.

5. Conclusion

Previous studies on milk and milk product handling, consumption among smallholder farmers, and butter-making in Ethiopia have frequently lacked comprehensive data systematically comparing these practices across diverse dairy production systems. The current study addresses this gap by providing such systematic comparisons. The use of improved milking, storage and transporting devices was significantly higher in the cash crop based (CCB) system, yet the enset based (EB) system lagged behind in adopting these technologies. Despite advancements in some areas, traditional methods like using clay pots for butter making remained prevalent across all systems. Notably, enset plant leaves were commonly used for storing and packaging butter, although some respondents in the CCB system also utilized polyethylene plastic bags. Cleaning practices for milk equipment varied regionally, with Eucalyptus globulus Labill. predominantly used in most systems and enset plant residue (‘kacha’) specific to the CCB system. Olea africana was widely used for smoking milk containers across all systems. Efficiency in butter production differed, with crossbred cows requiring more milk per kilogram of butter compared to local breeds, which indicates differences in milk quality between the breeds.

Milk production was highest in the CCB system, demonstrating its potential for higher yield, while the diversified crop-based (DCB) system showed the lowest production levels. Whole milk consumption patterns were consistent across all systems, yet the proportion of milk processed into traditional butter varied, with the DCB system processing the most.

Finally, the study underscores the diverse and multifaceted nature of milk handling and butter-making practices among smallholder farmers. These findings highlight the need for targeted interventions to improve milk handling and processing techniques tailored to the specific needs of each DPS. Enhancing these practices can significantly benefit smallholder farmers, promoting sustainability and efficiency in dairy production. It is also crucial to raise awareness about the consumption of milk and other dairy products. Smallholders should receive education and training on handling dairy products, enhancing traditional butter-making methods, and increasing dairy product consumption, alongside improving dairy cow management.

Ethics approval and consent to participate

Prior to the start of the survey, smallholder farmers were informed about the purpose of the study. The study respondents were informed about the importance of the information they would provide, assured of the confidentiality of their responses, and clearly told about their freedom to decline participation or withdraw from the research at any time. The farmers were reassured that there was no risk in participating in the study and were specifically informed about the anonymous handling of individual replies. Additionally, the study received ethical approval from Haramaya University with the approval code of SARS 103/22.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The datasets generated and analyzed during the current study are included in the body of this paper.

Additional information

Notes on contributors

Tsedey Azeze

Tsedey Azeze is a researcher at the Livestock Directorate of the Hawassa Agricultural Research Center in Ethiopia. She focuses on dairy productivity, milk quality and safety, and post-harvest handling of livestock products, implementing numerous research projects in these areas. She is also a PhD candidate at Haramaya University, specializing in Tropical Animal Production (Dairy Stream) within the department of Animal and Range Science.

Mitiku Eshetu

Mitiku Eshetu, with over 20 years of experience in teaching, research, and community development, is a specialist in Livestock production at Haramay University, Ethiopia. His expertise lies in livestock production, product processing, and technologies, with a major focuse on dairy.

Zelalem Yilma

Zelalem Yilma is deputy project manger for the ‘Building Rural Income through inclusive Dairy business Growth in Ethiopia’ (BRIDGE+) project under the Netherlands Development Orgazation (SNV), teaching, consultancy, and advisory services within livestock value chains, collaborating with various nional and international organizations.

Tesfemariam Berhe

Tesfemariam Berhe, a senior researcher at the Bio and Emerging Technology Institue in Addis Abeba, Ethiopia, specializes in dairy techology. With over 16 years of experince in research and teaching, he is currently focused on developing starter cultures for Ethiopian fermented foods.