Milk processing requirements for satisfying the demand for various dairy products in Tanzania

Introduction
Milk quality and Marketing
Fermented milk
Butter
Ghee
Milk/Blood mixture
Organisational set-up and training
Conclusions
References

Ruth Ryoba and R.L. Kurwijila,
Dairy Technology Laboratory,
Department of Animal Science and Production,
Sokoine University of Agriculture,
P.O. Box 3004, Chuo Kikuu,
Morogoro, Tanzania.

Abstract

Milk production by the smallholder farmers in Tanzania is increasing. However, only a minor portion of the locally produced milk enters the commercial sector owing to the marketing constraints and lack of processing techniques suitable for smallholder dairying. In order to sustain milk production to satisfy the demand, efforts to increase milk production should go hand in hand with efforts and knowledge to dispose milk surpluses above local requirement in the milk producing villages. The manufacture of stable marketable products including butter, ghee, low moisture cheese and fermented milks will provide smallholder producers with additional source of cash, facilitate reinvestiment in the interprise, yield by products for home consumption and enable the conservation of milk solids for future sale or consumption.

This paper discusses the major traditional milk processing and preservation methods in use in Tanzania and gives some possible improvement on traditional milk processing and preservation. Advantages and disadvantages of each method of milk processing and preservation are also discussed. The paper recommends that the smallholder farmers start milk collection centres where milk will be bulked or manufactured into different products. This will ease the organization of the processing, manufacturing and marketing of milk and milk products to all parts of Tanzania to satisfy the demand.

Introduction

Estimates of 1993/1994 shows that 555 thousand metric tons. of milk was produced in Tanzania where 75 % was from traditional sector, 25 % was from small holder farmers and 5 % was from large scale farmers. The decline of milk processing by TDL has been from 1979 where it collected and processed 14 million liters to 2.8 million by 1988. This indicates that most of the milk produced by the farmers was consumed at home or near home. Kurwijila (1993) reported that a subsequent FAO project URT/86/013 1986 assitance to Small Holder Dairy Development (SHDD); Swiss funded Southern Highlands Iringa/Mbeya project, Dutch Government funded project in Kagera, SECAP project funded Tanga and FAO/DANIDA initiated three small scale milk processing units operated by Women groups in Kilimanjaro. These projects have led to increase in milk supply ie Kagera produced 31,031 litres in 1982 which increased to 2,126,390 litres by 1989 (de Wolf, 1990) and in Morogoro milk production have increased from 37 million in 1989 to 43 million litres in 1994 (RLADO report 1994) while the TDL processing is decreasing the milk produced increases meaning that is there will be a lot of milk losses in future.

In Lushoto, 45% of the small holder farmers indicated that they were unable to sell their milk because of lack of markert (Scheinman et at 1992 and Kurwijila, 1993). Also Mdoe (1993) showed that only 17% of the small holder farmers in Moshi/Arusha preserve their milk by refrigeration. Moreover Mdoe (1993) found out that one of the 120 small holder farmers 104 or 87% had surplus milk and the reasons for the surplus milk included poor roads, shortage of cooling facilities, high transport costs and delayed payments by TDL. These reasons seems to be common to many smallholder farmers in the whole country such as Mwakaleli (Personal communication with the milk cooperative Chairman 1992) and parts of Dodoma.

The marketing of fresh milk presents considerable problems in the remote rural areas in Tanzania, thus small holder farmers consume some milk, sell some to milk vedors and the surplus is processed into some dairy products such as butter, ghee, sour milk and some cheese which have better keeping quality than the fresh milk. This traditional milk processing gives low yield, poor quality products whose shelf life is not as good as those produced by TDL. There is a need to develop milk preservation systems which are simple and more adaptable in an environment characterised by scattered small scale milk production and low technological know how (Kurwijila, 1984; 1986 and Ryoba and Kurwijila, 1992).

Therefore research into traditional techniques can yield important insights for appropriate interventions to improve processing efficiency and products shelf life, not only for pastoralists but also for other smallholder farmers having less cultural experiences in handling milk. Investigations must start with understanding the indigenous or traditional processing systems used by farmers. This will form a base for future research into traditional milk processing and possibilities for technical innovational milk processing. This paper will therefore describe traditional milk processing technology in Tanzania and point at possible areas of improvement.

Milk supply

Kurwijila (1988), reported that indigenous zebu cows milk is the most important source of milk in Tanzania as well as in many parts of Africa, with goats next in importance. Most pastoral farmers keep a lot of cattle but the average milk production is only 0.5 litre per cow per day. As pointed out earlier milk production is increasing in the country because of the introduction of improved dairy stocks; for example, a Mwakaleli farmer reported to be getting 24-30 litres from Friesian cows and 8-12 litres from crossbred cows.

Milk quality and Marketing

Milk is sold as sour milk, butter or ghee and some cheese. Very little milk is sold per household due to low productivity or lack of markets. The little milk sold directly to consumers and some to milk vendors who sell to consumers but such milk is very much prone to milk adulturation by water, cow's urine, coconut milk or cassava milk. There is a lot of milk wastage especially during the rainy season when the supply is high and marketing conditions are poor.

In recent years there has emerged a strong argument advocating the use of decentralised rural based milk processing as an appropriate system of milk marketing in a situation where smallholder milk production is thinly scattered and situated far away from urban markets (Bachmann 1979; 1983; Schulthess, 1987; Kurwijila, 1984, 1986, 1987, 1990).

Liquid milk

Apart from boiling of milk before consumption there are no known traditional technologies for preservation of liquid milk through heat processing in Tanzania. Indeed is has been reported that in some pastoral communities it is a taboo to heat milk (Shallo, 1987). Table 1 gives the technical possibilities of applying conventional milk technology processing in rural milk processing.

It is obvious that for technical and economic reasons, technologies involved in liquid milk processing (refrigeration, steam generation, pasteurization, sterilization/aseptic packing) are not immediately accessible to the small scale milk processor. It follows then that the small scale manufacturer must undertake the processing of preserved dairy products with long shelf life at ambient temperatures such that marketing is done only once or twice a week.

Some simple equipment could be used in decentralized rural dairies in developing countries (Bachmann, 1979; Kurwijila, 1986; Schulthess, 1990). The setting up of such rural industries would create employment in the rural areas, promote local skills, as some of the equipment could be made locally, and stimulate growth of production of milk by providing an assured market for local milk.

Table 1: Possibilities for application of conventional dairy technology unit processes in rural dairy processing (Adapted from Kurwijila, 1990).

Process	Rural Dairy Industry Application
	"Status"	"Potential"
		- In-can cooling with running water (highlands) = 0.5°C/100 m rise
Cooling/refrigeration	Natural air cooling	- Solar refrigeration - Surface coolers (highlands) - non existent in hot/humid areas - batch pasteurization
Heating	Boiling	- in-container sterilization - in-plastic sachet pasteurization - solar heaters, biogas plants, drum water broilers
Separation	Gravitational	- hand cream separator - ILCA's internally agitated clay pot
Churning	Traditional vessels	- Wooden churns - Milk can churns
Fermentation	natural fermentation	- Controlled fermentation using own culture; lyophilised culture
Coagulation	Natural acidification plant latice	- organic acids-e.g. lemon juice - bovine pepsin, porcine pepsin
Drying	Sun drying	- atmospheric drum drier
Air conditioning	Altitude dependent	- highland areas - evaporative cooling charcoal coolers?
Chemical preservation	Smoking	- LP system

Reduction of the bulk weight of milk to be transported is yet another important factor to be considered. Table 2 shows the relative merits with regard to technical and transport requirements of some of the dairy products.

Table 2: Requirements and Input Availability of Various Milk Products Made Under Rural Dairy Processing Conditions (Adapted from Olsen, 1988)

	Level of Technology	Shelf Life	Energy Consumption	Yield %	Transport* % saving
Fresh milk	+	1 day	-	100	0
Fermented milk	+	1 week	-	100	0
Butter	++	2 weeks	-	5	95
Ghee	+	2 months	++++	4	95
Cheeses
White picked (1)	++	6 months	+	12-14	80
Pasta filata (2)	+++	2 months	+++	8-10	90
Mould (3)	+++++	1 month	+	8-10	90
Whey (4)	++	1 week	++++	2
Cottage	+++	1 week	+	15-20	80
Semihard (5)	+++++	2 months	+	9-11	90
Hard (6)	+++++	1 year	++++	7	90

Notes: Level of technology: + simple +++++ advanced
Energy consumption - no energy consumption +++++ high energy consumption
Yield is estimated in percent of milk input.
+ Transport saving is related to weight of remaining product and storing material.

(1) Feta, Domiati, Gybna Beyda
(2) Provolone, Mozzarella
(3) Roquefort, Gorgonzoba, Brie, Camembert
(4) Ricotta, Myseost
(5) Gouda, Tilsiter, Cheddar
(6) Grana, Emmental, Gruyere

Fermented milk

Production of fermented milk is widely spread in Tanzania. In most tribes, milk is allowed to ferment naturally without addition of starter culture. Milk is accumulated over a number of days in a clay pot vessel of 2 - 4 litres volumetric capacity or a bottle gourd (O'Mahony and Bekele, 1985) and allowed to develop acidity up to 1% lactic acid. In cold weather, the container is kept near the fire to keep it warm. Soured milk is consumed as a side dish. If not consumed, sour milk is churned into butter. The shelf life of naturally fermented or sour milk is only 72 hours at ambient temperature (Ali, 1990 and Ryoba and Kurwijila, 1992). The disadvantage of this type of milk is that since it is not boiled there is always a danger of people getting TB or Brucellosis.

Among the pastoralists like the Maasai and others, milk vessels are smoked using specific wood which produce embers in the fire (Shallo, 1987; Kurwijila, 1988b). Smoke application appears to improve the colour, flavour and extend the shelf life up to 20 days to 3 months (Bekele and Kasaye, 1987 and Ryoba and Kurwijila, 1992). Preparation of fermented milk and smoking of milk vessel have been discussed in details by Shallo and Hansen, (1987). Fidder (1966) and Knowles (1975) reports that the smoke contains primarly phenols of about 21 components as identified by the gas chromatography which have the ant bacterial effect in smoked milk thus prologing the shelf life of milk (Ryoba and Kurwijila, 1992).

Apparently no research has been done in identifying, characterizing and classifying the plants and no analysis of the smoke components and its effects on the milk components has been done. Fidder (1966) reports that methods of generating the smoke, collecting the phenolic fractions and identifying the components have not been uniform or reproducible. The correlation of results from the investigations to obtain a meaningful information about the desirable phenol composition of smoke and the best parameters for its production is difficult. It is therefore a big challenge to the African researchers.

Natural fermentation of milk exploits lactic acid producing Streptococci and Lactobacilli which are present in any milk and which generally suppress spoilage and pathogenic organisms. The processing steps of heating milk followed by cooling and inoculation of starter culture to achieve a controlled fermentation is not done traditionaly in Tanzania. Heating kills undesirable organisms and also concentrates the milk" giving the fermented milk a heavier body (Fox, 1967). The use of pure cultures enables different acid flavours to be developed leading to a range of fermented (cultured) products. Since cultured milk has better keeping quality than fresh milk and better product consistency than naturally soured milk, their introduction to smallholder processing systems should be considered. Standard technology of fermented milk manufacture is first to boil at 80 - 90°C for 5-30 minutes to reduce the microbial population. It is then cooled, inoculated with starter culture or previously fermented milk and incubated at 30°C to 38°C for 4-16 hours. Higher temperatures cause wheying off while lower temperatures result in slow fermentation and may lead to the establishment of undesirable micro-organisms. Fermented milk can be made in simple containers of any size.

Ali (1990) working at SUA, Morogoro, undertook a study to compare the keeping quality of naturally fermented milk with sour milk made by use of standard manufacturing methods and inoculated with B-CH: 40, B-CH: 253 mesophilic starter cultures and yoghurt culture. Samples were stored at 19°C to simulate conditions in highland areas where smallholder dairying is predominant and 30°C to simulate conditions in coastal and semi-arid areas. Sensory evaluation, acidity development, pH measurement, gassing and wheying off were assessed daily for 5 days. Table 4, 5 and 6 give a summary of the main results.

Table 4: Summary evaluation scores comparing the keeping quality of milk under different cultures after 1, 3 and 5 days (Ali, 1990).

Temperature	Samples	Days 1	Day 3	Day 5
19°C	NF	8	2	7
	LC	7	3	4
	CH: 01	6	2	2
	CH: 40	2	0	2
	CH: 253	0	0	6
	YC	8	2	8
		13% too sour	66% too sour	64% too sour
30°C	NF	6	3	6
	LC	4	3	5
	CH: 01	3	6	7
	CH: 04	0	2	5
	CH 253	0	2	5
	YC	2	5	5
		47% too sour	66% too sour	64% too sour

Key: NF - Natural fermentation, LC - Local culture, YC - Yoghurt culture.

Overall results showed that yoghurt incubated and stored under ambient temperature leads to either lack of characteristic flavour if temperatures are low (19°C) or harsh acid flavour if temperatures are high (30°C). Thus with high storage temperature the keeping quality of yoghurt was very much reduced by the 5^th day. With natural fermentation of milk under ambient temperature (19°C) the process took longer time to complete and defects such as wheying off and gas production were encountered. An improvement to the traditional method i.e. using local culture to inoculate into preheated milk resulted in a product of high acidity after five days at 30°C.

Table 5: pH, % lactic acid of fermented milks during storage for 1, 3 and 5 days (Ali 1

Sample	Day 1		Day 3		Day 5
	pH	% LA	pH	%LA.	PH	%LA
NFE19	4.6	0.73	4.3	0.85	4.2	0.88
NF30	4.2	0.93	4.1	1.08	3.8	1.15
LC19	4.8	0.76	4.4	0.82	4.3	0.87
LC30	4.0	0.96	3.8	1.07	3.8	1.06
CH: 0119	5.0	0.65	4.2	0.75	4.2	0.85
CH: 0130	3.7	1.16	3.4	0.85	3.4	1.60
CH: 4019	4.2	0.76	4.2	0.87	4.1	0.88
CH: 4030	4.4	0.84	4.2	0.89	4.0	1.01
CH: 25319	4.5	0.82	4.3	0.83	4.2	0.86
CH: 25330	4.4	0.84	4.2	0.96	4.0	0.96
YC19	5.3	0.40	4.4	0.86	4.1	0.93
YC30	4.0	1.00	3.5	1.37	3.4	1.56

Key: %LA = % Lactic acid, LC = Local culture; YC = Yoghurt culture
19'30 = temperature: 19°C, 30°C.

Yoghurt .was found not to be a suitable product under local conditions since the control of fermentation is critical for a good quality product. The cultured milk made by use of mesophilic lactic cultures, had better keeping quality than both naturally soured milk and milk made with local culture. These results obtained by Ali (1990) are in agreement with the findings of Kurwijila et al. (1981).

Therefore the use of mesophilic lactic cultures in smallholder processing of fermented milks should be considered favourable. The local culture can be made by choosing two to three healthy cows, stripping about 150 ml milk into the sterile glass bottles during the middle of milking of each cow into a separate bottle and covering immediately. The milk is then left at room temperature until it coagulates. From the three bottles with coagulated milk the best culture is the one which has no air bubbles no wheying off but is smooth and compact. It is important to remember that the milk is not pasteurized and that is why the cow must be in good health before the milk is drawn from it. This culture can be used just like any other mesophilic culture.

Table 6: Whey separation and gas formation in 1, 3 and 5 days (Ali, 1990)

Sample	1st day		3rd day		5th day
	Gas prod.	Wheying	Gas prod.	Wheying	Gas production	Wheying
NF19		+	++	++	+++	+++
NF30	+	+	+++	+	+++	+++
LC19	-	-	+	+	++	++
LC30	-	-	+	-	+++	++
CH: 1019	-	-	+	-	+	-
CH: 1030	-	-	+	-	+	-
CH: 4019	-	-	+	-	+	-
CH: 4030	-	-	+	-	+	-
CH: 25319	-	-	-	-	+	-
CH: 25330	-	-	-	-	+	-
YC19	-	-	-	+++	-	+++
YC30	-	-	-	+	-	++

Schulthess (1987) suggested that further improvement on fermented milk should involve efforts to train rural dairymen in the manufacture of high quality cultured and milk products resembling the fermented milk products prepared and consumed traditionally. This therefore leaves a special challenge to the research work in warm countries.

Also good sanitary procedures and adequate heat treatment prior fermentation help to prevent spoilage described as " unclean", putrid and bitter which are caused by the coliform and putrefactic organisms and sharp acid flavour accur if the fermentation is too vigrous (O'Mahony, 1987).

Mala cultured milk has found acceptance in Tanzania where it is made by one group of Womens Cooperative Society at Nnronga near Moshi (Enemark, 1989). The reason for the successis that Mala is made by using the mesophilic starter culture which producess flavour that is much closer to traditionalsour milk comsumed by most adult Africans. Further more it has an advantage of not requiring incubation during its manufacture or refrigeration during retailing of the product.

O'Mahony (1987) reported that in Ethiopia the Borana pastoralists in Sidamo make concentrated fermented milk by removing the clear whey about 1/6 of the totalmilk volume from coagulated cow's milk and product can be stored for 20 days. This method of preserving the coagulated milk is also done among the Ndali, Safwa, Nyiha and Nyakyusa, Sukuma and Gogo tribes in Tanzania (personal experience)

Butter

Traditionally, butter is made from sour milk using a gourd or earthnware jar (O'Mahony and Bekele, 1985; Kurwijila, 1988). Sour milk is normally churned by shaking the vessel until the butter separates out. Among the Maasai, the vessel with sour milk is hung on a tripod and women or children swing the vessel to and from until the butter separates. Churning is normally done in the morning or late evening when the temperatures are normally low (Kurwijila, 1988b).

Butter is normally over-churned and not washed, factors which favour a high moisture content in the butter with a consequent short shelf life. Since the smallholder farmers do not have refrigerators, butter is prone to deterioration (O'Mahony 1987). Many people in the rural areas do not use butter as a spread on bread or for baking. Most of the butter is converted to ghee (personal experience). Butter is normally not salted and is kept in small gourds or earthware jars.

During processing there are many losses which are unavoidable. Webb (1983) reported that wastages should not exceed 5 and 3% of total butterfat when butter is made from sour milk and cream, respectively.

Improvements which can be done are to shorten the churning time and increase butter yield or butter fat recovery. Time can be reduced by gathering cream from the top of the vessel after souring and thus churn only the concentrated creamy portion of the sour milk. Yields will be high since the churning will take shorter time and temperature rise which tend to melt the fat, leading to high fat loss in the butter milk, is minimised (O'Mahony, 1987). Churning efficiency can also be improved by using an agitator developed at ILCA (O'Mahony and Bekele, 1985; Mbaga et al 1990), provided the churning temperature is held below 16°C. For larger volumes of cream more than 300 litres especialy in co-operative farms band driven centrifugal separation of cream followed by agitators like the wooden churn and ILCA agitator figure 1 are more appropriate to churn cream with 35-40 BF of proper temperature < 16°C producing butter and 0.1 % BF buttermilk. Fermenting the cream 2-3 days to develop the serum acidity to 0.5 %, fat solidifies, becomes viscous and influence speed and effectivenes churning and aromatic flavour develops due to diacetyl, propionic and acetic acid during fermentation (O'Mahony and Peters, 1987). The process enables the efficient recovery of milk fat as butler of about 96) % of the total fat processed. Churning temperature can be reduced by evaporative cooling (Dule, 1991).

Cream separation offers more processing options than are available with sour milk. Therefore a hand cream separator can be used to separate whole milk into skimmed milk and cream. Churning cream to butter takes a very short time and gives high butter yield.

Mwakapala (1990) did compare churning efficiency and fat recovery by using traditional bottle gourd, ILCA's clay pot churn (O'Mahony and Bekele, 1985; Mbaga, et al 1990), a wooden butter churn and an electric butter churn as a control to churn sour whole milk and uncooled sour whole milk. He obtained the results shown in Table 7.

Table 7: Average churning time, fat recovery and fat content in buttermilk from different types of butterchurns (Mwakapala, 1990).

Type of churn	Cooled sour whole milk (in %)			Uncooled sour whole milk
	Churning time (min)	%Fat in butter milk	%Fat recovery	Churning time (mill)	% fat in butter milk	% fat recovery
Traditional gourd	93	0.83	78.3	60	1.18	69.6
ILCA-clay pot	73.8	0.3	92.8	38.8	0.38	90.2
Wooden butterchurn¹	65.3	0.3	92.9	35	0.5	87.2
Electric (control)	705	0.15	96.2	35	0.35	91.0

¹Mbaga et al, 1990

Adoption of ILCA' clay pot churn with an internal agitator and a wooden churn could be best and most appropriate option for traditional small scale butter making provided at that the quantity of the milk to be processed is reasonably high (at least 50 - 100 litres/day). Also churning sour whole milk at low temperature resulted in increased fat recovery. Cooling to 3 - 4°C below ambient dry bulb temperature can be achieved by evaporative cooling in charcoal lined metal cabinets (Fig. 2) (O'Connor, ILCA, personal communication).

Salting butter with 2% salt increases its shelf life and yield. Salted butter should contain 80% milk fat, 2% milk solids-not-fat and 15% moisture.

Buttermilk as a by-product of traditional butter making is usually consumed without further processing i.e. consumed in a similar way to fermented milk (Shallo and Hansen, 1973; Bekele and Kassaye, 1987). Processing larger volumes of milk into butter and skim milk gives the processor more efficient recovery of butterfat and provides more options for the disposal of separated milk, but also requires more equipment (O'Mahony, 1987).

Ghee

Ghee, which is also known as butter oil or dry butter fat, consists of fat which is almost completely free from water, protein, milk sugar and mineral substances (McDowell, 1953). Most smallholder dairy farmers make ghee from butter. Butter is melted in a saucepan usually made of clay or stainless cooking pan, "sufuria", heated over a slow fire until all the moisture .has evaporated. It is then left to cool and is packed in normal bottles. The Maasai keep it in the gourds just at room temperature.

Normally during slow heating of butter, as the temperature rises, a foam forms which tends to rise and overflow, and a lot of wastage occurs. It also needs a lot of stirring as the liquid heats or else burning at the bottom of the vessel occurs and the product acuires a burnt flavour and the colour changes.

Therefore, it is good to remember that the vessel should be big enough i.e. double the volume of the butter to be processed. To avoid a burnt flavour, a thermometer (which small scale dairies can afford) may be used so that the butter boils up to 125-130"C only; above that ghee will burn. O'Mahony (1987) suggested that in order to get a good yield of butter oil the butter should be melted in an equal volume of water at 60°C followed by centrifugal separation. The product is free from non-fat milk solids and contains no more than 1.5% moisture. According to Madam Pal and Royorhia (1975), quoted by O'Mahony (1987) the residual moisture can be removed by further heating in butter oil manufacture.

Ghee can keep for 1 - 3 years (Bekele and Kasaye, 1987) at room temperature because the heat treatment, low moisture content and salt addition in ghee prevents development of hydrolytic rancidity and growth of micro-organisms and has a preservative effet (O'Mahony 1987). Oxidative rancidity will occur but can be minimised by packing ghee in opaque, air tight containers (McDowell, 1953) such as coloured orange juice or wine bottles with tight fitting lids and stored in a cool place or just in a dark corner of the house or cupboard where light cannot go through (personal experience at Sokoine University of Agriculture (SUA), Morogoro, Tanzania).

Milk/Blood mixture

This is common among the Maasai. They produce a popular food product of milk/blood mixture in a ratio of 4: 1. This is either during the milk shortage periods or simply to make a desired food product. Blood is drawn from a healthy animal, is beaten with a forked stick to remove fibrinogen (the ability to clot) then added to a specific quantity of milk, at a maximum of 20% by volume. This milk/blood mixture may be consumed immediately or stored and consumed after souring, or it may be boiled to coagulate and stored for future use. This product has a longer shelf life than milk alone. It is given to children or mothers who have recently give birth. It was reported that consumers of this product do not feel cold even when they walk or work in cold environment (Shallo and Hansen, 1973).

Cheese

Cheese is a good concentrated source of fat and protein (Scott, 1981) and as such it has a high nutritional value. Cheese is not much known in Tanzania but is becoming popular among the educated especially those who have lived abroad. TDL produces very little amount of cheese (Lohay, 1977; 1988).

Many cheese varieties require considerable technical skill and equipment in their manufacture. An important aspect of cheese is the treatment of milk with starter culture to produce controlled amount of lactic acid in the curd while many cheese varieties require special storage facilities for ripening. Common cheese made at a smallholder dairy in Moshi/Arusha is Alpine and Pasta filata (Ulicky, 1989). A detailed description of their manufacture is given by FAO-RDDTT (1988). They are simple to make and one women's group in Hai district produce them (Ulicky, 1989; Kurwijila, 1990).

Five cheese varieties: Queso Blanco, Halloumi, Feta, Domiati and white cheese are relatively simple to make under small holder conditions in Africa using small scale processing techniques, have good yield, stable under tropical conditions and the product proves acceptable to consumers (O'Mahony, 1987),

Rennet availability is a limiting factor since it must be imported and needs foreign currency. Studies at SUA on the use of Bovine pepsin as a rennet substitute were carried out by Wigenge (1989) who extracted pepsin from the abomasa of adult cattle. Mghenyi (1990) produced Past filata and alpine cheese using the crude pepsin extract, with rennet cheeses as control. The results obtained are shown in Table 8.

Table 8: Comparative characteristics of crude pepsin and rennet made cheese

Type of cheese and enzyme	% Total Nitrogen (TN)	% WSN	Ripening indices (RI)	Moisture content at 4wks1	Mean sensory difference2	Bitterness scores/5pts at 6 wks³
Pasta filata:
Pepsin	7.54*	2.05	27.19	37.09	2.30	1.08 NS4
Rennet	10.37*	0.91	8.80	36.18		0.92 NS
Alpine:
Pepsin	7.59*	1.75	23.05	41.24	1.67	1.50 NS
Rennet	6.65	1.64	24.66	37.48		1.25 NS

* dry weight basis; WSN = water soluble N; RI = WSN/TN
¹ = initial moisture content 15.3% and 52.6% in Pasta filata and Alpine, respectively.
² = Sensory difference scale: 1 = slight; 2 = moderate; 3 = much; 4 = extreme
³ = Bitterness scale: 0 = no bitternes; 1 = trace; 2 = slightly bitter; 3 = bitter; 4 = very bitter; 5 = extremely bitter
⁴NS = Non significant < ¹ 0.05.

The trials show that pepsin dues not adversely affect the quality and yield of the cheese. These results agree with those found at ILCA (O'Mahony, 1987). These results were too few to give any recommendation to the small scale cheese processors in Tanzania. More trials on the extraction, purification and testing of the bovine pepsin are going on at SUA by a Ph.D student.

Also at SUA in the Department of Animal Science and Production, a Latin American cheese made by coagulating milk with lemon juice - "Queso Blanco" has been introduced and found to be an acceptable cheese. Although it needs refrigeration, no rennet and starter culture is needed in the manufacturers shelf life can be increased by curing or smoking. Since it is a low fat product, can be eaten fresh, is cheap because only lemon juice diluted to equal amounts of water; acidification imparts pleasant flavour to the cheese and product yield is good, the cream separated can be made into butter and ghee and the whey left can be fed to animals or consumed by humans (O'Mahony, 1987) as is the case with the Losaa womens group in Moshi, Manufacture of cheese is difficult with small quantities of milk. However, a village cooperative could make cheese and thus widen its products range just as the Losaa womens group in Hai district, Moshi have done. Preliminary observations indicate that cheese making may be a suitable milk marketing option for rural cooperatives that are far from all weather roads. Use of deep freezer technique in the preservation of lactic cultures can extend the usefulness of limited supplies of lactic cultures (Kurwijila, 1990).

Cheese could be coated so as to: increase the protection against mould/years contamination during storage/ripening; improve quality of the ripened cheese; protect the rind against chemical damage during storage and improve the appearance of the cheese (Kuyper, 1990).

Organisational set-up and training

People are the most important asset of food and dairy processing plants. Ulicky (1989) reported that so far 6 training sessions at the Livestock Training Institute (LITI), Tengeru have been done on the manufacture of cheese. Also an expert consultant from FAO was brought to the Lossaa womens group in Hai district to teach them how to make Pasta filata type of cheese. Later on they were successfully introduced to semi-arid, Alpine type of cheese by local experts from SUA. Also selected members attended a course on small scale industrial management and accounting at the cooperative college at Moshi. The womens of Lossaa, Nronga and Ng'uni have advanced in tackling the problem of milk marketing. We can learn that cooperatives are really a tool for self sustenance and self reliance. The members appear determined, motivated to actualise themselves, utilize their own power and try to develop themselves. They have solved problems of finding feed, transport, walking for long distances to sell milk and to get a market of their products (Ulicky, 1989; Phelan, 1990; Kurwijila, 1990). Such developments and training need to be extended to other regions in Tanzania.

Also dairies attached to diploma colleges must be provided with equipment and layout resembling those which students are expected to work with in rural areas later on. It is advisable to integrate such dairy operations in the training curriculum of the students on a day to day basis. Ideally such practical training facilities should be run by students only (Schulthess, 1987). Follow up visits should be done to reveal if the trainees were passing on their newly acquired skills to farmers and that the equipment was being used properly (O'Mahony, 1987).

Conclusions

At present only a small fraction of the milk produced by smallholders in Tanzania enter the commercial market owing to the lack of milk collecting centres in rural areas, low producer prices for milk and scarcity of small scale processing techniques. Yet making this milk available to consumers would improve both the economic status of the smallholders and the nutritional status of the population. Dairy products that can be manufactured by small scale dairies include fermented milk, butter, ghee and some cheese. Some technologies for their improvement have been adopted to suit local conditions and some of the modified processing techniques have been tried successfully in Kilimanjaro area and SUA. These need to be extended to other parts of the country. Therefore more research, training and extension work is needed to ensure that the smallholder milk producers in various farming systems countrywide acquire the improved milk handling and processing techniques.

It is suggested that sponsorship should be aquired to train the small holder farmers on milk processing preferably at SUA where there are the facilities and experties for training.

References

All, K.O. 1990. Comparison of the keeping quality of naturally fermented sour milk and milk made by use of standard manufacturing methods. Animal Sci. Spec. Project, B.Sc. Agric. Sokoine University of Agriculture.

Bachmann, M.R. 1979. Adapted techniques for collection, processing and marketing of milk in developing countries. 2: Milk collection and processing techniques. Milchwiss, 34 (9): 538 - 541.

Bachmann, M.R. 1983. Milk processing in rural areas to support dairying in developing countries. J. Dairy Sci. 68: 2134 - 2139.

Bekele, E. and Kasaye 1987. Traditional Borana Milk processing efficient use of subtle factors needs further research work. lLCA Newsletter Vol. 6 No. 4.

Dule, 1991. The use of evaporative cooling and lactoperoxidase treatment of milk under simulated small holder farmer conditions. B.Sc. Special Project. Department of Animal Science and Production, Sokoine University of Agriculture.

Enermark, L. 1989. Assistance to womens groups in Kenya and Tanzania. TSAP Proceedings Vol. 16: 209 -218.

FAO-RDDTT, 199. Rural milk processing: A practical guide to small scale cheese making in E. Africa. Prepared by FAO-RDDTT, Naivasha, Kenya (unpublished).

Fox, P.F. (1967). Fermented milks. Technical Bull. Dairy Series 6: An Foras Taluntais (The Agricultural Institute - Dublin, Ireland). Jennes, P. and Patton, S. 1959. Principles of dairy chemistry London; Chapman and Hall Ltd.

Jennes, P. and Patton, S. 1959. Principles of Dairy Chemistry. London. Chapman and Hall Ltd.

Kurwijila, R.L.N., W. Schulthess and M.I. Gomez 1981. The influence of viscosity and of flavour metabolites in the acceptability of fermented skim milk by adult Kenyans. Milchwiss. 36 (10): 624 - 626.

Kurwijila, R.L. 1984. A review of research and development of the dairy industry in Tanzania: Milk processing and marketing. Proc. Tanzania Soc. Anim. Prod. Vol. 11: 100 - 113.

Kurwijila, R.L. 1986. Flame sterilisation and marketing of milk in 20 litre containers. Ph.D. thesis ETHZurich Diss. No. 8037.

Kurwijila, R.L. 1987. Possibilities and limitations of application of research and development in small scale dairy processing in developing countries. In: Dairy development in E. Africa Proc. IDF Seminar, Nairobi, 9 - 13th March 1983. IDF Bull. No. 221: 106 - 109.

Kurwijila, R.L. 1988a. Some reflections on milk supply and consumption statistics in Tanzania with reference to the role of the traditional cattle herds. Proc. 15th Sci. Conf. Tanzania Soc. Anim. Prod. Vol. 15: 216 - 242.

Kurwijila, R.L. 1988b. Traditional Milk Products in Eastern and Southern Africa. In FAO Monograph on Traditional Milk Products Worldwide (in Press).

Kurwijila, R.L. 1990. Rural dairy technology appropriate to rural smallholder production: The Tanzania experience. Paper Presented at Dairy Marketing Symposium, 26 - 30 Nov. 1990, ILCA, Addis Ababa, Ethiopia.

Kuyper, A. 1990. Plasicoat Cheese coating. Danish Dairy and Food Industry worldwide, 1990.

Knowles, M; Gilbert, J. and Mc Weeny, 1975. J. Sci. Fd. Agric. 26 189-196.

Lohay, A.B. 1977. Dairy production and processing in Tanzania. Proc. 4th Sci. Conf. Tanzania Soc. Anim. Prod. Vol. 4: 271 - 309.

Lohay, A.B. 1988. Contribution of the milk processing industry towards the National Nutritional in Tanzania. Proc. 15th Sci. Conf. Tanzania Soc. Anim. Prod. 15: 243 - 256.

MALD, 1987. Preliminary Review of the Ruminant Livestock Industry. Marketing Development Bureau, DSM 1987 R7/87. Ministry of Agriculture and Livestock Development - Tanzania (unpublished).

Mbaga, S.H., C.V. Shayo, R.L. Kurwijila and R. Ryoba (1990). Development of butter making techniques for small holder milk processing. Proc. 20th Sci. Conf. Tanzania Soc. Anim. Prod. Vol. 17: 15 - 27

Mdoe, N. 1993. Constraints to Milk Marketing in the Kilimanjaro Highlands of Hai District. Proc. 20th Sci. Conf. Tanzania Soc. Anim. Prod. 20: 247.

McDowell, F.H. 1953, The Butter Makers Manual. New Zealand Univ. Press. Wellington.

Mghenyi, F.B.D. 1990. Performance of crude pepsin from cattle Abomasa in cheese making. Special Project in Food Science and Technology, Sokoine University of Agriculture (unpublished).

Mwakapala, M.A. 1990. Efficiency of traditional method of butter making on butter fat recovery. Special Project in Animal Science, SUA (unpublished).

O'Mahony, F. and Bekele, E. 1985. Traditional butter making in Ethiopia and possible improvements. ILCA Bulletin 22: 9 - 14.

O'Mahony, F. 1986. Milk processing by Borana pastoralists. Mission report. ILCA Addis Ababa, Ethiopia.

O'Mahony, F. and Peter, J. 1987. Sub-Sahara Africa options for small holder milk processing. FAO World Animal Review Dull. No. 62.

Phelan, J.A. 1990. Support to rural dairy processing in Tanzania. Proc. 23^rd Intern. Dairy Congress, Montreal, Canada, Oct. 7-12th 1990 (in press).

Schulthess, W. 1987. The role of the dairy enterprise in training of technical personnel. IDF/FAO Seminar, Nairobi: 9 - 13^th March 1987. IDF Bull. No. 221.

Schulthess, W. 1990. Appropriate production of cultured milk in Zimbabwe. A report to appropriate technology international, Washington, DC and to Africa Now. London (unpublished).

Shallo, P.L. and Hansen, K.K. 1973. Pastoral method of handling and preserving milk. Egerton College, Kenya.

Shallo, P.L. 1987. Pastoral Methods of handling and preserving milk. In: Dairy Development in E. Africa. Proc. IDF Seminar, Nairobi 9-13 IDF/FIL Bull. 221.

Ulicky, E.W. 1989. A case study of three women's cooperatives for milk collection in Tanzania. Paper pre sented at FAO Regional Seminar on Dairy Extension for Rural Smallholders, Zimbabwe, 1989.

Webb, B.H.; Johnson, A.H. and Alford, J.A. 1983. Fundamentals of dairy chemistry. The AVI Publishing Company, Westpart Connect. USA.

Wigenge, R. 1989. Extraction and activity testing of crude preparations of pepsin enzyme solutions from abomasa of adult cattle. Special project in Food Science and Technology, Sokoine University of Agriculture.

DISCUSSION

Q. Mwakatundu, G

Traditional milk accounts for 70% of the milk in the country. How does SUA (Dept. of Animal Science) assist the traditional milk successor market their product in the open market?

Response: Mrs. Ruth Ryoba

It is difficult to make intervention in traditional milk since products made are for domestic consumption.

Q. Dr. Elizabeth Kyewalabye

Do you think there is enough demand of the traditionally processed products so that their increased production will assist the farmers? If so, how would you compare the income from these products and that from fresh milk on part of the farmers.

Response:

If the quality is good the demand is very much available. The comparison in terms of income from different products is difficult to tell.

Q. W.G. Mteti

I suggest SUA to go ahead and make specific standards on the types of products they are talking about. This could be of interest to consumers in the market.

Response:

This goes to the commercial aspects of the products and needs assessing consumer preference on the types of sour products etc. mentioned.