The effect of maize bran or maize bran mixed with sunflower cake on the performance of smallholder dairy cows in urban and peri-urban area in Morogoro, Tanzania

P S Mlay, A E Pereka, S T Balthazary, E C J Phiri, T Hvelplund*, M R Weisbjerg* and J Madsen**

Department of Physiology, Biochemistry, Pharmacology and Toxicology, Sokoine University of Agriculture,
P. Box 3017, Morogoro, Tanzania
panejo@yahoo.com
^* Department of Animal Science and Animal Health, The Royal Veterinary and Agricultural University, Copenhagen,
Grønnegårdsvej 2, 1870 Frederiksberg, Denmark
^**Department of Animal Nutrition and Physiology, Danish Institute of Agricultural Sciences, Research Centre Foulum, 8830Tjele, Denmark

Abstract

A study was conducted to determine the effect of feeding milking cows a supplement of maize bran alone (MB) or maize bran mixed with sunflower meal (MBS) during the dry season. Eighteen smallholder farms in urban and peri-urban areas of Morogoro practising complete zero grazing with not less than two milking cows participated in the trial. Forty-eight cows were used with ranges of body weight 232-556 kg, previous milk yield 3-13 litres/day, body condition score  2.5-5, parity 2-5, and 3-6 months post calving.  For each farm, both MBS and MB treatments were randomly distributed to the cows. Body weights and body condition scores were taken before, at 6^th and 12^th week of treatment. Daily milk production was recorded three weeks before, 12 weeks during and 3 weeks after the withdrawal of treatments. Milk samples were collected during the 12^th week of treatment and analysed for butterfat (BF), crude protein (CP) and total solids (TS). The economic viability of the dairy enterprises in relation to supplementation was also assessed.

MBS fed cows had significantly higher (P<0.001) milk yield compared to MB fed cows. No differences were observed in milk composition parameters, body weights and body condition changes. The economic return for sunflower meal incorporation in MBS was found to be very high.

It is concluded that using sunflower meal mixed with maize bran was effective in increasing milk yield during the dry season and was economically profitable compared to maize bran alone.

Key words: Dairy cows, maize bran, milk yield, sunflower cake, supplementation

Introduction

 

Smallholder dairy production in urban and peri-urban areas is a common phenomenon in Tanzania as is the case in most tropical countries (Mlozi et al 92; Mosha 1991). Dairy production offers household income-generating and employment opportunities (Mlozi et al 1992; Mlay et al 2001). More importantly, the urban and peri-urban dairy industry supply urban dwellers with dairy products that are highly needed since rural based dairy production is very low (Anonymous 1994).

 

One of the major production constraints in the smallholder dairy sector in the tropics is the dwindling both in quantity and quality of the roughage during the dry season thereby leading to drastic reduction in animal productivity (Leng 1990; Plaizier et al 1999). In order to sustain modest production as well as maintaining body condition, it is important to provide additional sources of energy and protein during the dry season (Odham 1982; Preston and Leng 1987; Jelantik 2001).

 

Dairy farmers in Morogoro feed maize bran, which is mainly energy rich, to milking cows. However, rarely do they provide protein sources due to either lack of awareness of the importance of nitrogen supplementation to animals, or scarcity and high cost of protein supplements, or a combination of the these factors. In Morogoro, oilseed cakes from sunflower and cotton seeds are relatively cheaper than animal protein sources like fish, meat, bone and blood meals (Mlay et al 2001). Sunflower cake is widely available in Morogoro as most people grow sunflower seed crop for sale to the oil processing factory in Morogoro town as well as extracting oil for domestic use using the hand press (Hayman 1992; Shayo et al 1997; Temi 1999). However, the potential of sunflower meal as a protein supplement to improve ruminant productivity in urban and peri-urban areas of Morogoro has not been fully explored.

 

Therefore, it was the aim of this study to: (i) Investigate further the suitability of sunflower meal as a source of protein that is cheap, effective and acceptable by farmers in Morogoro. (ii) Find out the effect, if any, of added sunflower meal on milk yield, milk composition and body condition of cows during the dry season

 

Materials and methods

 

Study area

 

The study was carried out in Morogoro urban and peri-urban area at an altitude of 528m above sea level with an average rainfall of 800-1200 mm per annum. Heavy rains fall between April and May. The mean maximum and lowest temperatures are 32.4 and 14.8 ^oC respectively. Morogoro town is located at about 250 km West of Dar-es-Salaam along the Tanzania-Zambia highway. The study was carried out between November 1999 and March 2000. A total of 18 farms distributed in various wards in Morogoro (Table 1) with a total of 48 milking cows participated in the trial.

Table 1 Distribution of farms and number of cows fed either maize bran-sunflower meal concentrate mixture (MBS) or maize bran alone (MB)
Ward	No. of Farms	Supplement
		Cows on MBS	Cows on MB
Bigwa	1	2	2
Forest	2	3	2
Kihonda	3	4	3
Kichangani	2	2	2
Kilakala	3	5	3
Mazimbu	2	3	2
Mbuyuni	3	5	4
Mlimani	2	3	3
Total	18	27	21

Animals and treatments

 

The participating farmers owned at least two milking cows, kept under a zero grazing system. The selected cows were those with the following criteria: 2-6 months post calving, previous average yield of 3.0-13.0 litres per day, parity of 2-3, and body condition score 2.5-5.0. The type of animals kept by farmers in this study included various crosses between the local Tanzanian short horn zebu cattle and Boran with Holstein-Friesian, Ayrshire and Jersey.

 

Treatments were either maize bran alone (MB) with crude protein (CP) of 10.9% DM (control) or maize bran mixed with sunflower meal (MBS) with CP of 15% in DM. The composition of MBS (% DM basis) was: 31, 68 and 1 of sunflower meal, maize bran and mineral powder, respectively; and that of MB was 99 and 1 maize bran and mineral powder, respectively. The mineral powder (Cooper Kenya Limited, Nairobi Kenya) contained the following minerals with the % contribution of each given in brackets; NaCl (27.0), Ca (18.5), P (11.0), Mg (3.0), Fe (0.50), Cu (0.16), Mn (0.40), Zn (0.50), S (0.40), Co (0.02), I (0.02), Se (0.0015) Mo (0.0002) and Ca:P ratio of 1.68:1.

 

Each cow was given 4 kg/day (as fed basis) of the respective supplement with half the amount given during morning milking time (around 6-7am) and the other half during the evening milking time (5-6pm). For a given farm, treatments were equally distributed to the milking cows so that both MBS and MB treatments were present.

 

Bodyweight and condition scoring

 

Heart-girth measurement of the animals was taken before, at 6^th and 12^th week of treatment in order to predict the live weights.  Concurrent measurements of body condition score were done by the method of Edmondson et al (1989) where 1 is severe under-condition and 5 severe over-condition.

 

Roughage intake

 

Once every week, one farm per ward  was used in estimating forage intake (Table1). Weights and samples of forage offered were taken each morning and of refusals the next morning.

 

Feed analyses

 

Dry matter (DM) of samples was determined by hot oven drying at 105 ^oC for 24 hours. Neutral detergent fibre (NDF) was determined according to the method of Van Soest et al (1991). Mineral analysis including Ca, P, Na, K and Mg were determined using inductively coupled plasma spectrophotometer (AOAC 1990). In vitro organic matter digestibility was determined by the method of Tilley and Terry (1963). Nitrogen (N) contents were analysed by the Kjeldahl method (AOAC 1990) using a semi-automated N analyser (Kjeltec system 1002, Tecator AB, Hoganas, Sweden). Crude protein (CP) was derived from the N content by multiplying with a factor of 6.25.

 

Rumen degradability of dry matter (DM) and nitrogen (N) from MB and MBS were determined by the nylon bag technique (Ørskov and McDonald 1979; Madsen et al 1997) using ruminally fistulated mature crossbred heifers fed grass hay and 3kg/day of MBS. The incubation times were 0, 2, 4, 8, 16, 24, 48 and 72h. The residues in the bags were analysed for N by Kjeldahl method. Nitrogen degradabilities were fitted to the exponential equation p = a +b (1 + e^-ct) where P =degradation at time t, a =the zero time intercept of the fitted curve, a +b = the asymptote of the curve at infinite time t and c = the degradation rate constant. Corrections for particle loss from the bag were made by the equation described by Weisbjerg et al (1990).

 

The Metabolisable energy of the supplements was determined by multiplying the calculated values of digestible energy (Hvelplund and Weisbjerg 1998) by a conversion factor of 0.82.

 

Milk yield and composition

 

Daily milk produced was recorded and weekly averages calculated.  Yields were recorded for three weeks before imposing the treatments, for 12 weeks during treatment and three weeks post-treatment. Milk samples were taken in bottles (50 ml) in ice packed cool boxes in the 12^th week of treatment. The samples were deep frozen in the lab and later analysed for milk butterfat (BF) percent by Gerber method while milk protein (CP) was calculated from nitrogen content from Kjeldahl method analysis (AOAC 1990) multiplied by a factor of 6.25. The percent total solids (TS) was determined by the reference method of oven drying at 105 ^oC in acid-washed sand for 12 hours. The solids not fat (SNF) were calculated as the difference between %TS and %BF (i.e. %TS -%BF).

 

Cost-benefit analysis

 

The production costs were computed from the major costs of feeds and labour charges. The cost-benefits of using sunflower meal was calculated based on the extra expense incurred when it was incorporated in the diet in relation to the revenue realised from sale of the extra milk obtained due to the supplements. No attempt was made to estimate the overall profitability of the dairy enterprises.

 

Statistical analyses

 

Analysis of variance was carried out using the SAS (1985) general linear model procedure with initial milk yield as a covariate with cow and farm as independent variables according to the model shown below.

Y = I + T+ F + M + e (Model 1)

Where: Y = Dependable variable

            I = Intercept

            T = Treatment (MB, MBS)

            F = Farm

            M = Mean initial yield for the 3 weeks pre-treatment

            e = Random error

Means for milk yield during the 3 weeks prior and after treatment, body weight and condition scores were calculated using the above model but without the covariate statement i.e.

Y = I + T + F + e (Model 2)

Where each statement in the model is as explained in Model 1 above.

 

Results

 

Chemical composition and rumen degradability of the two supplements

 

The chemical composition of sunflower (S), maize bran (MB) and maize bran mixed with sunflower meal (MBS) used in this trial is shown in Table 2.

Table 2 Chemical composition of sunflower meal (S), maize bran (MB) and maize bran mixed with sunflower meal (MBS)
Parameter	Type of feed
	S	MB	MBS
Dry matter, %	94.2	91.5	92.2
Components, %DM basis
Ash	4.3	5.1	5.4
Organic matter	95.7	94.9	94.6
Crude protein	23.6	10.9	14.8
Ether extract	4.5	10.7	9.2
Neutral detergent fibre	59.1	31.9	38.0
Carbohydrates	67.6	73.3	70.6
Potassium	NA	0.82	0.72
Calcium	NA	0.14	0.20
Magnesium	NA	0.32	0.29
Phosphorus	NA	0.76	0.72
Sodium	NA	0.23	0.22
IVOMD, %	58.2	64.4	62.0
NA: not analysed; IVOMD: In vitro organic matter digestibility

There was slightly higher NDF content in MBS compared to MB due to higher NDF content of sunflower meal. MBS had higher CP and less ether extract (EE) compared to MB (14.8 Vs 10.9) and (9.2 Vs 10.7%), respectively.

 

Table 3 shows the rumen degradability constants and effective DM and N degradation of MB and MBS. Both supplements had high potential degradability of DM and N (83-90%). Effective N degradation at 2 and 5% h^-1 passage rates was higher in MB compared to MBS.

Table 3 Degradation constants and the effective degradation of dry matter (DM) and Nitrogen (N) in MB and MBS supplements. Values expressed as %
		Degradation constants				Washing loss, % DM	Effective degradation
		a	b	a+b	C (% h-1)		2% h-1	5% h-1
MBS	DM	33.3	50.5	83.8	3.02	44.3	58.8	47.5
MBS	N	56.3	29.3	85.7	6.76	55.2	63.7	54.1
MB	DM	36.4	49.5	85.9	5.37	34.3	66.9	54.2
MB	N	37.6	52.2	89.9	6.55	35.2	73.5	63.4
DM; dry matter, N; nitrogen

MB had higher energy while, MBS had a higher nitrogen contribution (Table 4). Energy supplied from MB was slightly higher compared to MBS. The daily amount of N escaping rumen degradation was higher with MBS than MB.

Table 4. The energy and nitrogen contribution from supplements given during the on farm trial
Parameter	Supplement
	MB	MBS
ME content, MJ/kg DM	10.7	10.2
DM, %	91.5	92.2
Amount given, kg as fed	4.0	4.0
Amount given, kg DM	3.7	3.7
Energy supply, MJ	39.6	37.6
N content, %	1.8	2.4
N supplied, g/day	64.1	87.4
Effective N degradation, %	73.5	63.4
Rumen degraded N, g/day	47.1	55.4
Rumen undegradable N, g/day	17.0	32.0

 

Milk yield and composition

 

Milk yield trends 3 weeks before, 12 weeks during and 3 weeks post treatment are portrayed in Figure 1 as raw means and differences in milk yield between MBS and MB. From an initial mean value of 0.7 litres/day in the pre-treatment period, the difference in milk yield between the two groups increased to 1.5 litres in the first week to 2.2 in the 12^th week of treatment. Deduction of the mean difference in yield for the pre-treatment period from the differences in yield during the 12 weeks of treatment gave the extra yield due to MBS supplement. The extra yield increased from 0.84 litres/day in the first week to a maximum value (1.68 litres) in week 7 of treatment. There appeared to be a residual effect of MBS as the last three weeks post treatment showed extra daily milk yield of 1.28, 1.46 and 1.12 litres per week, respectively.

Figure 1. Mean milk yield and the difference between MBS and MB supplements three weeks before (b), 12 weeks of treatment and three weeks post treatment (a)

There was a higher (P<0.001) milk yield in cows fed MBS compared to MB for the whole treatment period and also for the three weeks post treatment (Table 5).

Table 5 Least square means milk yield (litres/day) during the on farm feeding trial with maize bran mixed with sunflower meal (MBS) or maize bran alone (MB)
Time	Supplements				Difference	P-Value
	MBS		MB
	Mean	SE	Mean	SE
3 Weeks before	7.2	0.44	6.3	0.52	0.9	**0.2
12 Weeks of treatment	8.1	0.2	6.6	0.20	1.5	*0.001
3 Weeks after treatment	7.2	0.43	5.1	0.51	2	**0.01
* Model 1; ** Model 2

The composition of milk collected at the 12^th week of treatment showed no differences between treatments (Table 6).

Table 6. Milk composition during the on farm feeding trial
Parameter	Supplements				*P-value
	MBS		MB
	Lsmean	SEM	Lsmean	SEM
Butterfat, %	4.01	0.187	3.84	0.208	0.5
Solids not fat, %	9.01	0.210	9.03	0.234	1.0
Total solids, %	13.02	0.254	12.86	0.283	0.7
Crude protein, %	4.04	0.160	3.77	0.179	0.3
*Model 2

Body weights and conditions scores

 

All the cows showed average improvements in live weight and body condition score during the 2 weeks of treatment, with no differences between supplements (Table 7).

Table 7 Body weight and body condition scores for animals during the experimental period
	Supplement				*P-value
	MBS		MB
	Mean	SEM	Mean	SEM
Body weight (BW), kg
Initial BW	381.9	13.71	358.1	15.93	0.3
BW at six weeks	384.4	13.79	361.4	16.02	0.3
BW at 12 weeks	387.1	13.71	363.2	15.93	0.3
BW change at 6 weeks	2.40	0.656	3.4	0.763	0.3
BW change at 12 weeks	5.15	1.272	5.10	1.478	0.6
Weight gain, kg/day
First 6 weeks	0.06	0.016	0.08	0.018	0.3
12 weeks of treatments	0.07	0.030	0.06	0.035	1.0
Condition score (BCS)
Initial BCS	3.50	0.064	3.38	0.075	0.2
BCS at 6weeks	3.52	0.062	3.39	0.072	0.2
BCS at 12 weeks	3.56	0.059	3.41	0.069	0.3
Change in BCS at 6 weeks	0.02	0.013	0.018	0.0148	0.6
Change in BCS at 12 weeks	0.06	0.017	0.030	0.0191	0.2
*Model 2

Feed intake

 

Due to the inherent practice of group feeding by smallholder farmers, it was not possible to determine roughage intakes for individual cows. The average DM intake was estimated to be 10.9 ± 1.04 (Mean ±SEM) kg per day, based on intakes of all mature animals in the farm including those that were involved in the experiment.

 

Economic implications due to supplementation

 

The cost of MB and MBS was estimated to be 59 and 68Tsh/kg. The additional cost to the farmers for feeding 4 kg of either MB or MBS per animal per day was calculated to be 234 and 270Tsh respectively. Therefore, the extra milk due to inclusion of sunflower had to cover the extra cost above MB that was 36Tsh (Table 8). That was considered the break-even point for including sunflower meal in MBS.

Table 8 Cost of individual ingredients and overall cost of preparing 100 kg (as fed) of MB and MBS supplements
Ingredient	Price/kg	Amount for 100 kg		Cost for ingredients
		MBS	MB	MBS	MB
Maize bran	50.00	68.70	99.00	3435	4950
Sunflower meal	80.00	30.30	0.00	2424	0.00
Mineral powder	900.00	1.00	1.00	900	900
Totals		100.00	100.00	6759	5850
Cost per kg, Tsh				67.59	58.50
Cost/day/cow, Tsh				270.36	234.00
Break even for MBS, Tsh				36.36
Note: No attempt was made to calculate the overall profitability of the dairy enterprises since to do that, other factors like investments in animals, houses, feeding costs of dry animals, bulls and calves would have to be considered

The major production costs in an already established dairy unit and the income from milk sales for two cows; one fed MBS and another fed MB were estimated (Table 9). . Based on the mean yields obtained in this study, milk sales from both would have earned the farmer some 3,411Tsh per day showing that having a dairy cow in milk was highly profitable. The extra income due to the use of sunflower meal was 390 Tsh per animal per day. Again, compared to the extra cost due to sunflower meal (36 Tsh), it was highly economic to use sunflower meal.

Table 9 Estimates of production costs and income from milk sales of a farmer owning two lactating cows; one fed MBS and the other MB. Costs expressed in Tanzanian Shillings (Tsh)
Item	Supplement
	MBS	MB
Mean yield, l/day	8.07	6.65
*Income from sales	2421.00	1995.00
Major production costs
Supplements	270.36	234.00
**Labour (350 Tsh/day)	175.00	175.00
Drugs 50 (Tsh/day)	25.00	25.00
1Forage (100Tsh/day)	50.00	50.00
Total expenses	520.36	484.00
Net income	1900.64	1511.00
Extra income due to sunflower inclusion	389.64
*Based on the price of 300Tsh per litre of fresh milk **Based on the highest pay level during that time (10,000 Tsh/month) 1Forage price estimated to be 4.55Tsh/kg DM

Discussion

 

Previous studies (Kidunda 1988; Wilson and Hatfield 1997; Mlay et al 2001) have shown a decline in nutritional value of forages with advancing dry season in Morogoro. During the dry season, forages are low in essential nutrients such as nitrogen, energy, minerals and vitamins required for optimal rumen microbial growth. Thus it is very important to give supplements that will supply some of the deficient nutrients in the poor quality forage during the dry period.

 

In Morogoro, sunflower cake and maize bran are important sources of supplemental protein and energy for ruminants. The CP contents of sunflower meal and maize bran used in this study (23.6 and 10.9 % respectively) was nearly of the same order as those used by Temi (1999) which were 21.9 and 11.6%.  The ether extract of the sunflower meal (4.5%) was within the range of reported values for solvent extracted sunflower meal. (Göhl 1981). Surprisingly, the maize bran used in this trial had a very high EE content (10.7%). The chemical composition of maize milling by-products can vary depending on the proportions of the outer skins, germ (fat rich) and the endosperm in the products. Göhl (1981) refers to the maize grain milling by-product containing the outer skins and germ as horminy feed, that with the outer skin without germ as maize bran and a product with high ether extract (9.2%) as maize oilcake. Its is possible that the by-product used in this study and the one used by Temi (1999) with an ether extract content of 8.5% had larger amounts of the fat rich germ and could possibly be close to maize oilcake (Göhl 1981).

 

The observed higher effective N degradation from MB compared to MBS was most likely due to the shielding of the structural carbohydrates that were higher in MBS. Rumen N degradability was 74 and 64% for maize bran and the concentrate mixture, respectively; therefore the two supplements were good sources of rumen degradable protein to yield ammonia that is required by rumen microbes for protein synthesis (Madsen and Hvelplund 1988). However, it appeared that there was a large particle loss from the dacron bags with the concentrate mixture than for those with maize bran. The sunflower meal was brittle and therefore may have yielded a relatively higher number of smaller particles well below the sieve size during grinding thereby increasing high chances of escape from the dacron bags during washing.

 

In most instances, cows respond immediately to an increased energy and protein level in diets by increased milk yield (Spörndly 1989; Shayo et al 1997; Schingoethe 1996) or weight gain or both depending on the lactation phase. The fact that most of the animals in this study maintained their body conditions proved that the supplements supplied additional energy and protein that were deficient in the poor quality roughage.

 

The increased milk yield in the MBS fed cows compared to MB fed cow can be attributed to the higher ammonia levels in the rumen or/and the N that bypassed  rumen degradation and was digested in the small intestine. N supplementation in a rumen environment that is deficient in N leads to increased DM digestibility and voluntary feed intake (Leng 1990; Mlay et al 2001) thereby leading to increased availability of the end products of fermentation (VFAs and microbial protein). MBS had almost double the amount of rumen bypass N (32g) compared to MB (17g) per day. The extra amino acids from digestion of the by pass protein could have led to an overall change in the pattern and efficiency of utilisation of absorbed nutrients (Oldham 1982) and this positively affected milk yield. Temi (1999) supplemented dairy cows with a sunflower-maize bran mixture (15.8% CP) and obtained a significant increase (P<0.05) in milk yield from 3.6 to 5.9 litres/day/animal in cows fed 3kg of the concentrate mixture compared to those supplemented with the same amounts of maize bran alone. In this study mean milk yield increased from 6.6 litres per day (MB) to 8.1 litres per day (MBS) in animals fed 4 kg/day of either supplement.  The greater response (64%) in the study of Temi (1999) compared with our data (23%) may have been because the research reported by Temi (1999) was carried out on station under controlled conditions and with an almost homogeneous group of animals. Urassa et al (1999) reported a 50% increased milk yield from smallholder dairy cows in the Tanga region that were supplemented with a mixture of maize bran-cottonseed cake  compared to animals given maize bran only.  All these studies show that there is a fairly good response in increased milk yield when the crossbred dairy cows in Tanzania are given supplements containing protein-rich ingredients such as sunflower cake or cottonseed cake.

 

Feed costs are considered to represent up to 45-60% or even more of the total cost of producing milk depending on management. Good dairy management practices must maintain low feed costs while at the same time maintaining nutrient and feed intake levels that will support an optimum level of production. The extra yield of 1.4 litres per day per animal (valued at Tsh 390) exceeded by a factor of 10, the extra cost of giving sunflower meal (Tsh 36). Therefore, the use of sunflower meal was found to be highly profitable.

Conclusions

• This study proved that sunflower meal; a protein-rich byproduct, available in Morogoro, is an effective and economically profitable supplement for use in dry season feeding of dairy cows in Tanzania.

References

 

Anonymous 1994 Tanzania Agriculture. A Joint Study by the Government of Tanzania and the World Bank. The World Bank, Washington D.C. pp 244.

 

AOAC 1990 Associations of Official Agricultural Chemists. Official Methods of Analysis.15th Edition AOAC Washington D.C. Volume 1, pp 62.

 

Edmondson A J, Lean I J, Weaver L D, Farver T and Webster G 1989 Body condition scoring chart for Holstein dairy cattle. Journal of Dairy Science 72, 68-78.

 

Göhl B 1981 Tropical feeds. Feed information and nutritive values. FAO, Rome. pp 529.

 

Hayman E L 1992 Local agro-processing with sustainable technology. Sunflower seed oil in Tanzania. Gatekeeper series No. SA 33 1992

 

Hvelplund T and Weisbjerg M R 1998 In vitro techniques to replace in vivo methods for estimating amino acid supply. In: In vitro techniques for measuring nutrient supply to ruminants. Deaville E R, Owen E, Adesogan A T, Rymer C, Huntington J A and Lawrence T L J (editors). British Society of Animal Science Occasional Publication No. 22, pp131-144.

 

Jelantik I N G 2001 Improving bali cattle (Bibos banteg Wagner) production through protein supplementation. PhD thesis, Department of Animal Science and Animal Health, The Royal Veterinary and Agricultural University, Copenhagen, Denmark. pp120.

 

Kidunda S R 1988 The yield and nutritive value of some tropical grasses and legumes at different stages. MVM thesis. Sokoine University of Agriculture. pp 131.

 

Leng R A 1990 Forage utilisation by ruminants. Nutrition Research Reviews 3, 277-303.

 

Madsen J, Hvelplund T and Weisbjerg M R 1997 Appropriate methods for the evaluation of the tropical feeds for ruminants. Journal of Animal feed Science and Technology 69, 53-66.

 

Madsen J and Hvelplund T 1988 The influence of different protein supply and feeding level on pH, ammonia concentration and microbial protein synthesis in the rumen. Acta Agricultura Scandinavica 38, 115-125.

 

Mlay P S, Madsen J, Pereka A E, Hvelplund T and Weisbjerg M R 2001 Smallholder Livestock Production Trends in Urban and Peri-urban areas of Morogoro. Tanzania Veterinary Journal Volume 21 (1) 15-26.

 

Mlozi M R S, Lupanga I J and Mvena Z S K 1992 Urban Agriculture as survival strategy in Tanzania. In: The rural urban interface in Africa. Jonathan Baker and Paul Ove Perdersen (editors). Seminar proceedings No. 27. Nordic Afrikainstitutet, Uppsala, Sweden pp284-294.

 

Mosha A C 1991 Urban farming practices in Tanzania. Review of rural and urban planning in southern and eastern Africa No.1, 83-92. Proceedings of a Workshop held in Kadoma Ranch Hotel, Zimbambwe, 20-23 July 1992. Mubi S (editor). ILCA. Addis Ababa, Ethiopia pp 157-164.

 

Oldham J D 1982 Protein-Energy Interrelationship in Dairy Cows. Journal of Dairy Science 67(5), 1090-1114.

 

Ørskov E R and McDonald I  1979 The estimation of protein degradability in the rumen from incubation measurements weighed according to rate of passage. Journal of Agricultural Science (Cambridge) 92, 499-503

 

Plaizier J C B, Nkya R, Shem MN, Urio N A and McBride B W 1999 Supplementation of dairy cows with nitrogen molasses mineral blocks and molasses urea mix during the dry season in Central Tanzania. Asian-Australian Journal of Animal Science 12:735-741.

 

Preston T R and Leng R A 1987 Matching ruminant production systems with available resources in the tropics. Penambul Books, Armidale, NSW, Australia pp 245.

 

SAS 1985 Statistical Analysis System, Institute Users Guide Statistics, SAS Inst. Inc, Gary, NC.

 

Shayo C M, Ogle B and Uden P 1997 Comparison of water melon (Citrullus vulgaris) seed meal, Acacia tortilis pods and sunflower-seed cake supplements in central Tanzania. 2. Effect on hay intake, milk yield and composition of Mpwapwa cows. Tropical grassland 31(2), 130-134.

 

Schingoethe J D 1996 Dietary influence on the protein level in milk and milk yield in dairy cows. Journal of Animal feed Science and Technology 60, 181-190

 

Spörndly E 1989 Effects of diet on milk composition and yield of dairy cows with special emphasis on milk protein content. Swedish Journal of Agricultural Research 19, 99-106.

 

Temi A W M 1999 Utilisation of local feed resources by cross-bred dairy cattle in a sub-humid area of Central Eastern Tanzania. M. Sci. thesis. Swedish University of Agricultural Sciences, Uppsala, Sweden pp 48.

 

Tilley J M A and Terry R A 1963 A two stage technique for the in vitro digestion of forages. Journal of the British Grassland Society 18,104-111.

 

Urassa A, Kurwijila L R and Aboud A A 1999 Feeding practices of smallholder dairy farmers and effect of improved supplementation on milk yield of dairy cattle in Tanga, Muheza and Lushoto districts. Proceedings of the 26^th Scientific conference held in LITI, Tengeru, Arusha, Tanzania. TASP Conference Series, Vol. 26, (1999), pp11-56.

 

Van Soest P J, Robertson J B and Lewis B A 1991 Methods for dietary fibre, neutral detergent fibre and non-starch polysaccharides in relation to animal nutrition. Journal of Animal Science 69, 973-987.

 

Wilson J R and Hatfield R D 1997 Structural and chemical changes of cell wall types during stem development: consequences for fibre degradation by rumen microflora. Australian Journal of Agricultural Research 48 (2), 165-180

 

Weisbjerg M R, Bargava P K, Hvelplund T and Madsen J 1990 Use of degradation profiles in feed evaluation. 679 Beretning Statens Husdyrbrugsforsog pp33