Livestock Research for Rural Development 21 (7) 2009 Guide for preparation of papers LRRD News

Citation of this paper

Handling, processing and marketing of milk in the North western Ethiopian highlands

Yitaye Alemayehu Ayenew*,***, M Wurzinger*,****, Azage Tegegne**and W Zollitsch*

* University of Natural Resources and Applied Life Sciences (BOKU), Department of Sustainable Agricultural Systems, Division of Livestock Sciences, Vienna, Austria
** International Livestock Research Institute (ILRI,) Addis Ababa, Ethiopia
*** Andassa Livestock Research Centre, Bahir Dar, Ethiopia
**** International Livestock Research Institute (ILRI), Nairobi, Kenya
and   BOKU-University of Natural Resources and Applied Life Sciences, Department of Sustainable Agricultural Systems, Division of Livestock Sciences, Gregor-Mendel-Strasse 33, A-1180 Vienna, Austria


In order to describe the ways of handling, processing and marketing of milk in urban and peri-urban production systems of the North western Ethiopian highlands, 256 and 54 dairy farms were selected for survey and monitoring data collection, respectively.


Due to poor market access and high transaction costs and the perishable nature of raw milk, the amount of sold milk (products) and prices were significantly lower for peri-urban producers. Prices were also low during fasting periods and during the wet season; when milk production was reactively high. Traditional ways of milk processing at household level include practices which may contain risks for product quality affecting consumers’ health. Providing basic handling and health education for producers therefore is likely to result in improved milk (product) quality on the markets. Direct delivery to the nearby consumers was the largest primary milk outlet for producers of both systems, while retailers and milk cooperatives were the second most common outlets for the urban and the peri-urban system, respectively.


Therefore, to enhance dairy production and marketing, milk co-operatives that could be able to market larger volumes and sufficiently reduce transaction costs should be supported by governmental and non-governmental organisations through providing technical and infrastructural support.

Key words: dairy, Ethiopia, peri-urban, production systems, urban


Urban and peri-urban dairy production systems involve production, processing and marketing of milk and milk products that are channelled to urban centres. It plays a vital role in the lives of the urban and peri-urban poor by providing a source of subsistence through household nutrition (milk and meat), supplementary income and generating employment opportunity. The dairy industry in the country is constrained by several technical and economic factors (Fekadu 1994) and the national milk production remains among the lowest in the world, even by African standards (Zegeye 2003). One of the necessary conditions for increased milk production is the provision of assured marketing outlets that are sufficiently remunerative to producers. Experience from countries like Uganda and Kenya pointed to marketing outlet being a key initiator of milk production by smallholders. Most development and research projects in dairying were conducted within and/or around the Addis Ababa milk shed (Sintayehu et al 2008). Currently, dairy development projects such as ILDP (Integrated Livestock Development Project), IPMS (Improving Productivity and Market Success of Ethiopian farmers) and Land O’Lakes (USAID Dairy Development Program) have been launched in the areas addressed herein. Although different sectors of the regional government offices have been also providing technical support for these and similar projects, no studies were conducted on the impact of all these projects on the dairy producers and the current status of processing and marketing the milk produced by smallholders. Therefore, the objective of this study was to explore the dairy processing and handling practices and to evaluate the ways of marketing in the urban and peri-urban dairy production systems of the North western Ethiopian highlands.


Materials and methods 

Study areas and data collection


The study was conducted between July 2006 and March 2007 in order to characterize the milk handling, processing and ways of marketing of the urban and peri-urban dairy systems in the milk shed areas (Bahir Dar and Gondar) of North western Ethiopian highlands. The urban production system deals with primary dairy supply chains in the inner cities and major towns, while the peri-urban system includes farms found in the periphery of cities and secondary towns. The area lies on an elevated plateau ranging from 1720 to 3000 m above sea level. The average annual temperature and rainfall range from 10 to 30 °C and from 1000 to 1500 mm, respectively (NMSA 2008). According to the CACC (2003) report, the urban and rural areas of these districts comprise about 2.28 and 0.09 million households, respectively.


In a first step, a rapid survey was conducted at district level with the Agricultural Development Offices in order to identify dairy farmers who keep different cattle breeds and represent different farming systems and milk production potentials in varying agro-ecological zones; information was collected from 19 districts. Based on this, a total of 256 dairy farmers (57 from urban and 199 from peri-urban areas), representing 17 districts, were chosen randomly. The questionnaire was designed in a way to obtain information on milk production and utilization, including types of milk products, ways of handling, processing and marketing. Moreover, to verify survey information like the amount of milk production and utilization, 54 farmers out of 256 respondents were selected randomly and monitored once per season. For seasonal classification, the amount of rainfall and temperature distribution information from NMSA (2008) was used and the year was divided into two broad categories, the dry season (October to April) and the wet season (May to September). With the objective of evaluating the contribution of milk cooperatives/producers as option for market entry point for smallholder dairy farmers, a semi-structured questionnaire was also developed and information about their status was collected from a total of 13 milk cooperatives or producer groups found within 17 districts covered in the study areas.


Data analysis


Fixed factors such as dairy production system (urban and peri-urban) and season (wet and dry) were used as classification variables for most dependent variables such as amount of milk produced, milk sold, home consumed, bucket fed to calves and processed, shelf life (refers the interval between the time of production to possible period of stay being accepted by the end users) and the price of dairy products. Preliminary analysis showed that interaction effects of the fixed factors were not significant and hence were excluded from the model. An attempt was made to evaluate the influence of sex of the household head, household size, farming system and religion of dairy farmers on the amount of milk production and utilisation. However, no significant effect was observed. Hence, these fixed factors were excluded from the model. Data were analysed using General Linear Model (GLM) procedures of the Statistical Analysis System (SAS 2002). Chi-square-test was used to determine differences in relative frequency of ordinal data. The differences between different levels of factors were analysed utilizing the Bonferroni-Holm test procedure (EßI 1987). For all analysis, the level of significance was set at an alpha of 0.05.


Results and discussion 

Milk production and utilization


In the study areas, the total annual milk produced was estimated to be about 1.4 million litres from 256 farms. By production system, it was about 0.6 and 0.8 million litres from 57 and 199 farms of urban and peri-urban dairy systems, respectively. To verify the information about the milk utilization collected during the survey (Figure 1ab), actual measurements were made of the amount of milk produced, sold, bucket fed to calves, home consumed and processed (Figure 2ab).

a) Urban dairy production system


b) Peri-urban dairy production system

Figure 1ab.
  Proportion (%) of milk utilization in 256 dairy farms
 in the North western Ethiopian highlands (survey results)

The monitoring data revealed that significantly (p<0.001) more milk was produced by urban than by peri-urban farmers (43 and 14.1 litres/household/day, respectively; Figure 2ab) which is mostly due to the presence of a greater number of crossbred (local zebu type x Holstein Friesian) cows in the first system (11 TLU) than the second systems (1.9 TLU).

a) Urban dairy production system

b) Peri-urban dairy production system

Figure 2ab.  Average daily amount of milk produced and utilized (litre/farm) and proportion (%) of milk utilization
in 54 dairy farms in the North western Ethiopian highlands (results from monitoring)

A similar conclusion was given by Yoseph et al (2003) and Sintayehu et al (2008). From the survey data, a comparison of the two systems showed that the proportion of milk sold was significantly (p<0.001) higher in the urban system than peri-urban system (47% versus 21%), while the opposite was true for the proportion of milk processed (15% versus 25%; Figure 1ab). Similar results, but with an even greater difference between production systems, were obtained from the monitoring data (Figure 2ab). For urban farmers, the relatively high proportion of milk sale probably was a result of access to a better market which might be related to the increasing urbanization (Azage 2004; Anthony 2002). Both in the monitored and the surveyed farms of the urban system, the amount of milk bucket-fed to calves was significantly higher than in the peri-urban farms (Figure 1ab and 2ab). Even though, more (9%) urban farmers were observed feeding the calf by bucket than peri-urban farmers which was 3%, generally the practice was not common in the study areas. Comparable results (8.2 and 4.9% for urban and peri-urban areas, respectively) were reported by Anthony (2002). The survey data analysis showed that in the peri-urban farms a significantly (p<0.05) higher proportion of milk was consumed at home as compared to urban farms (51 versus 31%). Anthony (2002) also confirmed the higher rate of home consumption of milk on peri-urban (41.3%) as compared to the urban farms (10%). The higher consumption of fluid milk in peri-urban as compared to urban areas is most probably due to poor market access for milk by peri-urban producers. Therefore they use more of the fluid milk for home consumption and process it into butter which has a better shelf-life and sell it on distant markets.


As shown in Figure 1ab and Figure 2ab, the proportion milk processed and bucket-fed to calves was nearly the same form both the survey and monitoring data analysis. Unlike the survey information, the monitoring data revealed that the proportion of milk used for the household consumption ranked least, while from survey information home consumption had taken the second lead. Such difference might be by chance in which more farmers selling than using for household use selected during sub-sampling and/or the sub-sample size (54 farms) selected might not as representing as main sample size (256 farmers).


Handling of milking processing equipments


Surveyed dairy farmers were observed using different milk utensils for collecting, storing and processing milk. The majority (62%) of the peri-urban producers used gourd and the rest (48%) used clay pot utensils. Besides, few (10%) of the total peri-urban producers had access to the aluminium milk utensils.  In urban farms more (83%) used plastic utensils and the rest (17%) used aluminium utensils. I addition to these, 33 and 21% of the urban farmers had access to use gourd and clay pot utensils.


One of the major factors affecting the quality of dairy products is related to adequately perform milking procedures and cleanness of the milking utensils (Gonfa et al 2001). In both production systems about 94 % of the farmers cleaned the udder before milking, 6.7% (urban) and 5% (peri-urban) producers did so before and after milking. While urban farmers equally used warm and cold water for this purpose, the majority of peri-urban farmers (68%) used warm water. Except one farmer, all producers cleaned their milk utensils at least once a day. The predominant practice (73.4% of farmers) was the use of water and leaves of shrubs (such as Combretum molle (Abalo), Ocimum suave (Dama Kessie) and Buddleja polystachia (Anfar)) twice a day, followed by drying and smoking with plants specifically used for this purpose (e.g. Rosa abyssinica (Qega), Osyris quadripartite (Keret), Otostegia integrifolia (Tinjut), Olean Africana (Woira), Thymus vulgaris (Tosgne) and Juniperous procera (Tid)). Procedures of cleaning and disinfection of milking utensils prior to milk collection reported herein, were similar to previous results from Ethiopia (Gonfa et al 2001; Yousuf Kurtu 2003; Sintayehu et al 2008). According to the local understanding, the practice of smoking the vessel by burning wooden chips of specific trees and shrubs has an advantage of imparting special taste and odour to the product, and to disinfect the vessels, thus reducing the numbers of micro-organisms and thereby extending the shelf life of the product. The report of Ashenafi (1996) supported this assumption, as greater numbers and a faster development of aerobic mesophilic micro organisms occurred in milk kept in non-smoked as compared to smoked containers.


The sources of water which the farmers had access to, were significantly different for the two production systems: while almost all (94.7%) of the farmers had access to pipe water in the urban areas, water from bore wells (18.1%) and rivers (17.6%) was also used in the peri-urban locations. However, the water from bore wells and especially from rivers used for cleaning was probably of doubtful quality, mainly at peri-urban farms. This might jeopardize the effect of cleaning the udder and milking utensils.


Types, processing systems and shelf life of dairy products


In the 256 dairy farmers studied, only traditional household utensils were used for collecting or milking, storing and processing. As a result of such technical and economic constraints, the major dairy products found were only yoghurt–like fermented/sour milk (Ergo), traditional butter (Kibe), traditional ghee (Neter Kibe), cottage cheese (Ayib), traditional hard cheese (Metata Ayib), buttermilk (Arrera), and whey (Aguat). The typical processing scheme for these products is shown in Figure 3.

Figure 3.  Flow scheme for milk utilization by smallholder dairy farmers
in the North western Ethiopian highlands

Similar information was published by Gonfa et al (2001), Yousuf Kurtu (2003) and Sintayehu et al (2008). Churning was observed in 58% of the surveyed urban households. Forty two percent of these were involved in further processing of buttermilk into cottage cheese and whey. Churning and subsequent production of cottage cheese and whey from butter milk were even more frequent in the households of peri-urban producers (73.4% and 52.8%, respectively).


The traditional milk processing is generally time consuming, verities of products was limited and less milk fat recovery turned into butter per unit of milk processed (ILCA 1992). If the farmers could not produce greater varieties of products, it is most likely that farmers could not get the full value-added products from milk production. This is probably one reason for the relatively low proportion of urban dairy producers processing milk into butter. The current observation is in agreement with the report of Sintayehu et al (2008). Moreover, the proportions of peri-urban producers involved in further processing of buttermilk into cottage cheese was slightly higher than of urban producers. The latter more frequently used buttermilk for bucket-feeding of calves. Another reason for urban producers not practicing this processing might be the lacking availability of fire wood for cooking buttermilk into cottage cheese and whey. Using electric energy for the cooking of buttermilk is economically less favourable than the utilization of buttermilk for household consumption and/or for rearing calves. Similar results were observed by Holloway et al (2000) who stated that a vast majority of dairy farmers who live far from urban centres in Ethiopia processed milk into different by-products.


Fresh milk


The survey results showed that the milk producers used different techniques to preserve fresh milk without clotting, such as smoking of the container and boiling of fresh milk before collection, or refrigeration. In both production systems, smoking was the predominant practice, performed by 54 and 62.4% of urban and peri-urban producers, respectively. Sixteen percent of the urban farmers used a refrigerator, an option which was almost not present in the peri-urban areas. Cooling by putting the container with milk into a cold-water bath was practiced by about 4.8% of the peri-urban producers. It is obvious therefore, that for technical and economic reasons technologies in fluid milk processing such as steam-pasteurization, sterilization and aseptic packing are not common on Ethiopian smallholder farms (Gonfa et al 2001; Yousuf Kurtu 2003; Sintayehu et al 2008). The average shelf life of milk and milk products showed no significant (p>0.05) difference between the production systems.


Fermented/sour milk (Ergo)


Twenty one percent of the milk produced was consumed fresh or allowed to ferment naturally. The product typically was semi-solid and in smallholder dairy farms it was produced from whole milk, while in milk cooperatives or other producer groups it was produced from skim milk. On average, milk was accumulated in a clay pot or a gourd over a period of 1 to 4 days and allowed to develop acidity. The mean shelf life of fermented milk was 3.8 days. According to the information collected from the respondent farmers, fermented milk was the main product used as basis for further processing of various fermented milk products such as traditional butter, ghee, cottage cheese, butter milk and whey (Figure 3).


Traditional butter (Kibe)


The proportion of interviewed farmers using different means of preservation for traditional butter and cottage cheese is depicted in Table 1.

Table 1.  Proportion of farmers using different means of preservation for traditional butter and cottage cheese by production system


Traditional butter

Cottage cheese

Urban (N=32)

Peri-urban (N=164)

Urban (N=18)

Peri-urban (N=113)






Cooling in refrigerator















* Spicing in this context includes further processing of traditional butter and cottage cheese into ghee and traditional hard cheese (Metata Ayib), i.e. after removal of water and spicing.

**Others include consumption and/or selling immediately (within a day) after production and traditional cooling in water bath.

All surveyed producers, except one NGO-farm, produced traditional butter from fermented whole milk, but not from cream. The traditional churning utensils were on average filled with about 11 litres of fermented milk, followed by agitation with a wooden stick to break the curd before churning. After this, the churning device was typically covered with a piece of skin or leather which was stretched over the mouth and was securely tied. The final point of the churning process was detected by a sequence of indicators, such as observing the sound of the churned milk and the visual judgement of the surface of a piece of straw inserted into the churn. The butter was then kneaded in cold water and washed to remove visible residual buttermilk (Arerra). The average churning time reported was about 3.5 hours. According to the information from the interviewed farmers, the relative amount of whole milk converted into butter and cottage cheese was about 4.3 and 16%, respectively.


Buttermilk (Arerra)


As described above, buttermilk is a by-product of the formation of butter from fermented milk. At household level, part of this by-product was reported to be consumed by the household members and by suckler calves. The rest was processed in to cottage cheese.


Cottage cheese (Ayib)


Cottage cheese is a soft curd-type cheese made by the churning of buttermilk mainly at household level, while in one farm and in milk cooperatives or other producer groups it was produced from skim milk. Cottage cheese was prepared by heating the buttermilk in an iron or clay pot until a curd-mass formed; this was followed by cooling to coagulate the curd. Then, the curd was separated from the whey through a fine-meshed cloth or a sieve. It was observed that about a kilogram of cottage cheese could be obtained from about 11 litres of buttermilk or skim milk. Part of the cottage cheese was further processed into traditional hard cheese (Metata Ayib). Putting the cottage cheese in a sieve container and squeezing was done in intervals for about 3 days until the water content was sufficiently reduced. After the cheese was drained sufficiently and became dry enough, spices were added to give a desirable flavour. The spices mainly used were the aerial parts of Ruta chalepensis (Tena Adam) and Ocimum basilicum (Beso Bila), the rhizomes of Zingiber officinalis (Zigibl), Turmeric (Erd) Allium sativum (Netch Shinkurt), red shallots (Key Shinkurt) and the seeds of Aframomum corrorima (Corerima). The traditional hard cheese was reported to be able to store averagely about 2.1 years before consumption with out losing its desired flavour and taste by the local consumers.


Traditional ghee (Neter Kibe)


Traditional ghee was made by evaporation of the water from butter by heating and melting of butter in an iron or clay container until bubbling ceases. The ghee was decanted into another container leaving the curd material in the pan. The procedure and plants used therein to spice ghee for flavour and preservation purposes were similar as described for traditional hard cheese above. The consistence of this product was described as semisolid at room temperature. It could be stored for about 2.8 years without losing the quality desired by the local consumers.


Generally, to increase the shelf life of milk products, the current practice of further processing of traditional butter and cottage cheese into traditional ghee and Metata Ayib was typical for most smallholder dairy farmers of Ethiopia (Ashenafi 1990; Gonfa et al 2001). However, most of the milk processing was done by use of traditional household utensils under unsanitary conditions; to deliver a desirable flavour, plant materials were added to the finished product (traditional ghee and Metata Ayib) might have contributed to the high microbial load. This is in agreement with the report of Ashenafi (1990), Gonfa et al (2001) and Sintayehu et al (2008). Therefore, giving basic handling and health education for producers is likely to result in a better milk quality on the market, including shelf life and aspects of consumers' health.


Marketable dairy products and price determinants


Like in most parts of the Ethiopian highlands, mainly raw milk, butter and ghee were marketed by the primary dairy producers. Furthermore, 100, 33, 89 and 78% of the milk cooperatives and other producer groups were selling traditional butter, cheese, skim milk and yoghurt, respectively. Averagely, each milk cooperative sold about 22 litres of yoghurt, 18 kg of cheese, 8 litres of buttermilk and 7 kg butter per day. In the urban system it was not common to sell butter (7% of producers), while about 42% of the peri-urban farmers did so. In the urban system, only one NGO farm produced and sold hard cheese; about 4.5% of peri-urban farmers marketed cottage cheese.


Factors affecting the prices of milk and dairy products included season, fasting and non-fasting days, and access to urban sites; to some extent the quality and sources of dairy products also influenced their price. The price of dairy products was affected by the totally 250 fasting days per year practiced by the followers of the Ethiopian Orthodox Church in which these refrain from eating food of animal origin.


In the urban system, the average daily fresh whole milk sale was significantly (p<0.001) higher (50.6 litre) than in the peri-urban production system (8.8 litre). The price of fresh whole milk per litre was significantly (p<0.01) higher for urban than for peri-urban producers (2.4 versus 2.0 Ethiopian Birr (i.e. 0.28 versus 0.24 USD). Fresh milk could not be kept for a long time before consumed. As a result, dairy producers living distant from urban centres could fetch a relatively lower selling price than farmers in or close to urban centres. This view is supported by Sintayehu et al (2008) who stated that the distance from market sites (mainly to urban centres) had a major prohibiting effect on farmers from selling fresh milk to urban consumers. The current findings are also in line with those of Yousuf Kurtu (2003).


Furthermore, the amount of milk sold was significantly (p<0.05) higher during the wet season when cows produced relatively more milk and on non-fasting days as compared to the dry season and fasting days. Related to this, significantly (p<0.01) higher prices were reported for dry and non-fasting periods as compared to wet and fasting periods (2.4 versus 2.1 Birr, i.e. 0.28 versus 0.25 USD per litre). Yousuf Kurtu (2003) and Sintayehu et al (2008) derived suggestions related to this specific market situation for the Eastern and Southern parts of Ethiopia. In times of low demand for raw milk, farmers were strategically processing of milk into butter and cottage cheese to sale in times of better market. A similar result was drawn by Sintayehu et al (2008).


Neither location nor season significantly affected the average amount of butter sold per farm per week (3.2 and 4.5 kg in the peri-urban and urban area and 3.7 and 4.0 kg in the dry and wet season, respectively). Proximity of the dairy producers to the urban consumers also showed no significant effect on mean butter price, but the mean price for a kg of butter was significantly (p<0.001) higher (24 Birr, i.e. 2.82 USD) in non-fasting periods and during dry season as compared to other seasons of the year (19.5 Birr, i.e. 2,29 USD).


In general, compared with the world, costs of milk production in Ethiopia are low (Staal, 1997), but high transactions costs for households and processors most likely prevent from a potential market entry (Holloway et al 2000). In the tropics, the existence of relatively high transaction costs coupled with perishable nature of milk play a central role in limiting dairy production and marketing (Staal et al 1997). He further stated that under such conditions, milk co-operatives have an advantage as they are able to market larger volumes and sufficiently reduce transaction costs. This underlines the importance of enhancing dairy production and marketing in a coordinated way. Supporting milk cooperatives by provision of technical and infrastructural support may therefore be a specifically efficient for increasing the income of farmers and for providing consumers with more high-quality dairy products.


Organization of dairy marketing


The proportion of dairy producers who marketed milk and butter at respective primary outlets and market places is depicted in Table 2.

Table 2.  Characteristics of relevant points of sale for milk and butter and percentage of producers using these

Product / point of sale

Customer / site of delivery

Urban production system

(N=37 for milk; N=4 for butter)

Peri-urban production system

(N=152 for milk; N=83 for butter)

Milk outlets







Cooperatives/producers group



Milk market place

Farm gate (homestead)



Open local market



Delivery to buyer place






Butter outlets




First rural assemblers






Butter  market place

Farm gate (homestead)



Open local market



Delivery to buyer's place



* Retailers in this context includes hotels, shops, and tea or coffee houses

As it is also common in most parts of Africa, informal dairy marketing was the sole marketing system in the study areas; these involved direct delivery of raw milk and milk products by producers to consumers in the immediate neighbourhood and sales to itinerant traders, milk cooperatives or individuals in nearby areas. The prices of milk and milk products from the point of production to the next outlets were determined by the producers. Similar findings were reported by Thorpe et al (2000), Yousuf Kurtu (2003) and Nigussie (2006). In respective order, 78.8, 10.6 and 5% of the farmers were transported milk and milk products on foot, by bicycle and public transport and spent up to 6 hours and 13.6 Birr or 1.6 USD per round trip. Only very few dairy producers (2.8%) used a private car for transport of dairy products to outlets. Milk producers, cooperatives/producer groups, first rural assemblers, wholesaler, retailers and consumers were observed participating in dairy marketing of the study areas (Figure 4).

Figure 4.  Flow scheme for dairy marketing chains in the North western Ethiopian highlands

This showed that dairy marketing in the study area involve more intermediaries, each of whom luckily adds some delivery or transformation service to the product. The informal dairy marketing and especially butter marketing found in the study areas involves more intermediaries; this is likely to affect the producers’ price and thereby may eventually influence market supply.


Dairy products such as whole raw fermented and skim milk had short marketing channels which is most likely due to their short shelf life. On the contrary, traditional butter, ghee, cottage cheese and traditional hard cheese showed the longest marketing channels. Hard products such as ghee and traditional hard cheese (Metata Ayib) could stay for about 2.8 years without loosing their desired quality. Traditional ghee is a more convenient product than traditional butter in the Tropics, because of its better shelf life even under warm conditions (O’Connor and Tripathi 1992).


The largest primary milk outlets for producers’, comprising some 67.3% (urban production system) and 46.1% (peri-urban system) of marketed milk, consists of direct sales of raw milk by producers to consumers, typically through farmer delivery to nearby. For urban producers, the second most important players in the informal markets were retailers (hotels, shops, tea and coffee houses), who handle another 28.6% of marketed milk, and who deliver milk to consumers, while milk cooperatives handled only 4.1% of the marketed milk and sold this to the consumers or other retailers. Contrary to this, the second most important primary outlets for peri-urban producers were milk cooperatives and other producer groups which comprised 44.7% of marketed milk; they mostly processed milk into butter which they sold to consumers or retailers. Only a small proportion of raw milk (9.2%) was sold by peri-urban producers directly to retailers.


Currently the milk cooperatives and other producer groups on average had 48 member farmers, the actual numbers ranging from 25 to 121. The milk cooperatives buy milk mainly from members (95 %), process and sell it to traders and local consumers. These groups exclusively focussed on the processing and sale of dairy products; they were neither providing additional services (such as credits, feeds and veterinary services) to the member farmers nor to the buyers. A total of 13 milk cooperatives and other producer groups were found within 17 districts covered in the study areas. At the time of data collection, two of them were not functional due to damage of milk processing equipments. 8 out of the 13 cooperatives and other producer groups were in their initial stage, as they were only certified and started working not until 2006 and thereafter. The average capacity of the milk cooperatives and other producer groups to handle fresh whole milk was 123 and 187 litres per day on fasting days and non-fasting days of the Orthodox Christian Church, respectively. The cooperatives and other producer groups mainly used hand-operated cream separators. Consequently, they produced butter from cream milk. Partially, skim milk was sold to the local consumers; some was fermented to yoghurt, while the rest was further processed into cottage cheese and sold through different outlets. A similar result was reported by Yousuf Kurtu (2003). Due to infrastructural constraints (lack of suitable traffic ways, cooling and other devices), losses were reported of 5-50% and of 15-75% of cottage cheese and yoghurt, respectively. All milk cooperatives and other producer groups had only one milk collection site each, which increases marketing costs for smallholder producers who thereby were forced to travel up to 10 km (round trip) per day to market their milk.


In the current study, poor infrastructure and poor logistics for collection, transport and sale of milk and milk products coupled with seasonal fluctuation of market prices due to relatively more supply and refrain of more people from taking milk and milk products during fasting are observed to form a bottle neck in the further development of the dairy sector. Therefore, under difficult market conditions like these, the processing of dairy products into ghee and Metata Ayib and selling those in times of a greater seasonal demand is an option.



The authors are very grateful to the Austrian Exchange Service (ÖAD) and Amhara Regional Agricultural Research Institute (ARARI, Ethiopia) for their financial support. The support of Dr. Getachew Alemayehu and Dr. Eshete Dejen (ARARI, Ethiopia) were particularly important. We would like also to express our gratitude to the livestock owners for their cooperation during data collection and to the staff of Andasssa Livestock Research Centre for their technical and logistical assistance. Dr. Roswitha Baumung's (BOKU) help in all statistical aspects is highly acknowledged.



Anthony I 2002 Comparison of urban and peri-urban dairying in Awassa, Ethiopia (unpublished MSC thesis, Southern University Awassa, Ethiopia)


Ashenafi M 1990 Microbial quality of Ayib, a traditional Ethiopian cottage cheese, International Journal of Food Microbiology 10: 263--268


Ashenafi M 1996 Effect of container smoking and incubation temperature on the microbial and some biochemical qualities of fermenting Ergo, a traditional Ethiopian sour milk, International Dairy Journal 6: 95-104


Azage Tegegne 2004 Urban livestock production and gender in Addis Ababa, Urban Agriculture Magazine 12: 30-31


CACC (Central Agricultural Census Commission) 2003 Ethiopian agricultural sample enumeration 2001/02 statistical report on socio-economic characteristics of the population in agricultural households and land use for Amhara region, Addis Ababa, Ethiopia, Part-I


EßI A 1987 Statistische Methoden in der Tierproduktion, (Österreichischer Agrarverlag, Vienna)


Fekadu B 1994 Present situation and future aspects of milk production, milk handling and processing of dairy products in Southern Ethiopia, (unpublished PhD thesis, Agricultural University of Norway)


Gonfa A, Howard A F and Wilhelm H H 2001 Field survey and literature review on traditional fermented milk products of Ethiopia, International Journal of Food Microbilogy 68: 173-186


Holloway G, Nicholson C, Delgado C, Staal S and Ehui S 2000 Agro-industrialization through institutional innovation: Transaction costs, cooperatives and milk-market development in the east African highlands, Agricultural Economics 23: 279-288


ILCA (International Livestock Centre for Africa) 1992 Alternative milk processing and preservation techniques and the quality of market butter and cheese. ILCA Annual Program Report, Addis Ababa, Ethiopia, 40 pp


Nigussie G 2006 Characterization and evaluation of urban dairy production system in Mekelle city, Tigray region, Ethiopia, (unpublished MSc thesis, Hawassa University, Ethiopia)


NMSA 2008 National Meteorological Services Agency of Bahir Dar Station; unpublished data


O’Connor C B and Tripathi B R 1992 Milk processing techniques - sour milk, ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia, No. 2, 20 pp


SAS (Statistical Analysis System) 2002 User Installation Guide for the SAS System Version 9.1, (SAS Institute Inc., Cary, NC, USA)


Sintayehu Y, Fekadu B, Azage T and Berhanu G 2008 Dairy production, processing and marketing systems of Shashemene–Dilla area, South Ethiopia. IPMS (Improving Productivity and Market Success) of Ethiopian Farmers Project Working Paper 9, ILRI (International Livestock Research Institute), Nairobi, Kenya. 62 pp


Staal S J, Delgado C and Nicholson C 1997 Small-holder dairying under transactions costs in East Africa. World Development 25: 779-794


Thorpe W, Muriuki H G, Omore A, Owango M O and Staal S 2000 Dairy development in Kenya: the past, the present and the future. Paper prepared for the Annual Symposium of the Animal Production Society of Kenya, March 22nd-23rd, 2000, KARI Headquarters, Nairobi, Kenya


Yoseph M, Azage T, Alemu Y and Umunna N N 2003 Evaluation of the general farm characteristics and dairy herd structure in urban and peri-urban dairy production systems in Addis Ababa Milk Shed.  In: Yilma Jobre and Getachew Gebru (editors), Proceedings of the 10th annual conference of the Ethiopian society of animal production, Addis Ababa, Ethiopia, August 21-23, 2003: 139-144


Yousuf  Kurtu M 2003 Certain aspects of the dairy systems in the Harar milkshed, Eastern Ethiopia, (unpublished PhD thesis, University of the Free State Bloemfontein)


Zegeye Y 2003 Imperative and challenges of dairy production, processing and marketing in Ethiopia. In: Y Jobre and G Gebru (editors), Challenges and opportunities of livestock marketing in Ethiopia. Proceedings of the 10th annual conference of the Ethiopian Society of Animal Production, Addis Ababa, Ethiopia, August 21—23, 2003: 61-67

Received 8 February 2009; Accepted 10 March 2009; Published 1 July 2009

Go to top