Livestock Research for Rural Development 2 (1) 1990

Citation of this paper

Effect of supplements of balanced concentrates and cottonseed cake on milk production in mauritian villages

A A Boodoo, R Ramjee, B Hulman, F Dolberg* and J B Rowe**

Animal Production Division, Ministry of Agriculture, Fisheries and Natural Resources, Reduit, Mauritius

*University of Arhus, Denmark;
**Department of Agriculture, Western Australia


An experiment was carried out to evaluate effects of two supplements (a formulated concentrate and cottonseed cake) for cows kept in confinement by smallholders in villages in Mauritius. The basal diet consisted mainly of sugar cane tops (from July to December) and grasses harvested from roadsides and common lands. The supplements were fed for 3 months prior to calving and throughout lactation. The amounts of commercial feed were 2, 3 kg/day for the 7th and 8-9th month of pregancy and 0.5 kg/litre of milk. The quantities of cottonseed cake were half the amount of concentrate.

Lactation yields (litres) for concentrate and cottonseed cake were 3023 and 2871 (wet area) and 2358 and 2649 (dry areas). There were no differences in milk yield due to supplement but yields were higher in cows kept in the wetter areas. Breed effects (local "Creole", Friesian and crosses) were not significant.

Key words: Milk production, cattle, sugar cane tops, cottonseed cake.


The annual production of fresh milk in Mauritius is approximately 9 million litres which is equivalent to 12% of total consumption of milk and milk products. In its continued effort towards maximum self-sufficiency the Ministry has set a target of 25% self sufficiency in terms of milk and meat by 1991.

In the villages of Mauritius milk production can be summarized in the following way.

On the basis that the small breeders make a major contribution to national production of fresh milk the work reported here was designed to investigate the extent to which productivity in this sector could be improved.

As far back as 1956, Bennie reported that the local Creole cattle could double their milk production with improved feeding. In 1971 an FAO project on Milk and Meat Production suggested that the most important factor limiting milk production was the supply of a protein concentrate to the cow. This FAO project also demonstrated that milk yield could be increased considerably by better feeding and management.

More recently Dolberg and Rowe (1984), in reviewing experimental work done by the Mauritian Ministry of Agriculture on milk production, concluded that under local conditions greatest responses in milk production would be expected from protein supplementation. They referred to the work of Mapoon et al (1977) which showed that ground nut cake was more efficient then either a balanced concentrate feed, or a molasses/urea mixture, as a supplement for milk production; and to that of Gaya et al (1982) who reported that supplementation with cottonseed cake increased milk production more effectively than the formulated concentrate "cowfeed". In fact, similar increases in milk production were achieved with about half the level of cotton seed cake as commercial concentrate. A second advantage of cottonseed cake as a supplementary feed is that it requires no mixing.

The project described here was designed to investigate and compare the effect of two types of supplements: the commercial concentrate "cowfeed" and the protein-rich cotton seed cake. In addition to the measurements of milk production the study provided the oportunity to investigate the nutritive value of the most commonly used feed resources (see Boodoo et al 1990).


Location and treatments

The study was carried out in two locations:

In each area 22 cows received each of the two supplements (cotton seed cake or cowfeed) during the 3-months precalving and the following 10-months of lactation.

Supplementary feeds and feeding

"Cowfeed" is a locally compounded concentrate feed for dairy cows, with about 17% crude protein on fresh matter basis. The formulation of is as follows: wheat bran 5%, rice bran 11.5%, maize 20%, cottonseed cake (or groundnut cake) 30%, molasses 30%, common salt 1% and calcium carbonate 2.5%. Only approximately 45% (by weight) of these ingredients is produced locally - half of the maize and all of the molasses, salt and lime. The cottonseed cake was purchased either from Zimbabwe or India and was fed together with a mineral supplement of 15 g sodium chloride and 50 g calcium carbonate per day.

The amounts and the pattern of feeding the supplements are summarized in Table 1. During lactation the amount of concentrate fed was adjusted each week according to the previous week's milk production and the predicted production for the following week.

Table 1: Daily levels of supplementation during the last 3 months of pregnancy and during lactation.
  Cowfeed Cotton seed cake
Pregnancy (kg/head)    
7th month 2 1
8th and 9th month 3 1
Lactation (kg/litre milk) 0.5 0.25


Selection of cows and allocation to treatments

A group of participating farmers was identified by first explaining the objectives and benefits of the project to milk producers in the two areas. For those participating, the supplements were provided free of charge through the UNDP project (MAR 82/007) and on this basis farmers voluntarily registered their names.

Cows that were 7-months pregnant were accepted to participate in the project and these were allocated alternately to each of the two treatments - cowfeed and cotton seed cake. A total of 44 cows was selected in each climatic area.


The animals were representative of the following breeds: Creole, Friesian and Friesian-Creole crosses at various levels of crossbreeding. The Creole cow, which is the local breed, is of the Bos taurus type. It is of medium size (in the range 300 - 400 kg adult liveweight), polled and humpless. It is predominantly white or white brown with dun, black or brown characteristic spots.

Milking and milk recording

The cows were hand-milked twice a day, in the early morning at about 5.30 a.m. and in the afternoon at about 6 p.m. Milk production was recorded daily by extension officers in each area travelling on a bicycle. These field officers visited the cowkeepers at dawn and dusk, to measure the amount of milk produced. A volumetric measure called a "quart" (5 quarts to a litre) is traditionally used in Mauritius. A record book was kept permanently at the cowkeeper's place for the recording. In addition, a weekly record sheet was used to transfer the milk records to the master file. All calves were bucket-fed during the experiment. This was a pre-requisite for participation in the project in order to obtain a precise measurement of milk production.

Forage and forage consumption

The cowkeeper was responsible for the collection of fodder according to normal practices. During the crop season (June to mid December) the main source of forage is canetops. In the intercrop season a variety of grasses, shrubs, twigs, creepers and vegetable crop residues are usually collected from the road verges, common land and adjacent fields. The forages were fed to appetite at all times. The amounts of fodder offered and consumed were recorded. In addition, samples of forages were taken weekly for chemical analysis and nylon bag degradability. These data are reported in a separate paper (Boodoo et al 1990).

Chemical analysis and nylon bag digestibility

Samples of the supplements were taken at intervals of approximately 14 days and sent to the Agricultural Chemistry Laboratory, Reduit, for analysis (dry matter, crude protein, crude fibre, ether extract and ash.)


Milk production and composition

Effect of supplement and area

The mean milk production of cows in the two areas on the two types of supplement is summarised in Table 2. There was considerable variation in the age and lactation number in the cows. However, there were no significant differences in these parameters between treatment groups and milk production data have been averaged for all animals. In the same way, milk production data have been calculated for all breeds together.

Table 2: Mean milk production and composition during the full 301 day lactation according to area and type of supplement
  Vacoas Mapou
  Cowfeed CSC Cowfeed CSC
Milk production (kg)        
301d lactation 3023 2871 2538 2649
SE 146 104 139 129
Milk composition (%)        
Fat 4.08 4.57 4.31 4.61
SE 0.222 0.164 0.183 0.192
Protein 3.41 3.58 3.47 3.40
SE 0.149 0.118 0.061 0.089


When averaged over the whole trial the concentration of milk fat was not significantly higher in the animals fed cottonseed cake and there was little effect of area or supplement on protein concentration.

The effect of breed

All animals were classified on the basis of their phenotypic appearance into one of three categories: (i) Friesian; (ii) Friesian x Creole; and (iii) Creole. The milk production of the animals in these three categories is summarised in Table 3 for the two areas.

Seasonal effects on milk production

There were differences between the two regions in the pattern of milk production with time. Fig 4 summarises the mean daily milk production for each month of the year in the Vacoas and Mapou areas. It is clear that from the start of May until the end of August there was little difference between the pattern of milk production in the two regions. The milk production in the Vacoas area remained high from August to January and fell from February to May. On the other hand the milk production on the Mapou area reached a peak around August and from then until May was significantly lower than that measured in the Vacoas area.

Table 3: Milk production by breed type in the Vacoas and Mapou areas. Data are summarised for the total milk production over the 301 d lactation and for the "peak daily production" for the second month of lactation.
  Creole Cross-bred Fresian
Vacoas area      
Total (litres) 2788 2958 2899
SE 232 115 176
"Peak" (litres/d) 13.0 14.5 13.8
SE 0.2 0.5 0.8
Mapou area      
Total 2889 2536 2459
SE 216 124 156
"Peak" 12.1 12.4 10.3
SE 0.4 0.6 0.7

There were no significant differences in total or "peak" milk production due to breed.


The average milk production of cows calving during the crop season (June to November) was 2950 kg and was significantly (P <0.001) higher than the average for cows calving outside the cropping season (December to May) which was 2639 kg.

Table 4: Effect of season (month) on mean daily milk yield (litres) of cows in the wet (Vacoas) and dry (Mapou) area
Month Vacoas Mapou SEdif Month Vacoas Mapou SEdif
Jan 10.4 7.7 0.8 Jul 9.7 9.3 1
Feb 9.6 7.7 1 Aug 10.7 10.6 1
Mar 9 7.8 1.1 Sep 11 10.2 1
Apr 8.3 8.5 0.8 Oct 11.3 9.5 1
May 8.5 8.2 1.2 Nov 10.3 8.5 0.9
Jun 9 9 0.8 Dec 10.5 7.8 0.8


Analysis of feed


The results of the chemical analyses carried out on the supplements are shown in Table 5. The crude protein content of the cottonseed cake (41.1%) is slightly more than twice that of the cowfeed (17.4%).

Table 5: Chemical composition (% air dry) of the concentrate and cottonseed cake
  DM CP Ash EE CF Ca P
Cowfeed 84.2 17.4 10.2 1.13 5.8 1.10 0.4
SE 0.5 0.2 0.2 0.2 0.1 0.06 0.056
N 47 48 48 21 21 12 12
CSC 90.5 41.1 6.3 5.9 11.3 0.17 1.1
SE 0.3 0.4 0.1 1.2 0.6 0.01 0.11
N 37 38 38 15 15 12 12



The quality and the amounts of fresh forage consumed during the lactation period are described in a separate paper (Boodoo et al 1990).


The findings in this village-based trial, that cottonseed cake was more efficient (250 g/litre of milk) than a "balanced" concentrate (500 g/litre) as a supplement for basal diets of moderate digestibility (about 55% degradability in situ) and low nitrogen content (1-2% N in DM), confirm the earlier reports by Mapoon et al (1977) and Gaya et al (1982) carried out under experiment station conditions. They provide strong support for the hypothesis that "bypass" nutrients, especially protein are the principal limitations to ruminant production on tropical forages (Preston and Leng 1987).

The quantity of milk produced by the village cows in this experiment (9.2 kg/d during 300 days) is higher than the average production of between 3.5 and 9.2 kg/d reported for the Government stations where cowfeed is fed at the rate of 0.5 kg/kg milk. It is also relevant to compare it with the milk production of unsupplemented village cows (4 to 5 kg/d for a lactation period of around 225 days).

Although there were only 23 (about 25%) Creole cows in the study their milk production potential appeared to be equal to the more exotic genotypes with a daily mean of 9.6 and 8.3 kg per head for a 301- day lactation in the Vacoas and Mapou areas respectively. This indicates that under these village conditions the Creole breed has a similar performance to the imported Friesians or their crosses.


Acknowledgement is made to the UNDP for funds made available through Project MAR 82/007; the Extension Division, the Chemistry Division and the Division of Veterinary Services of the Ministry of Agriculture, Fisheries and Natural Resources for their collaboration. Statistical help was provided by the Agronomy Division. We are also grateful to the cowkeepers who agreed to participate in the project.


Bennie J S S 1956 The Mauritius Creole breed of milch cattle. Empire Journal of Experimental Agriculture 24:95.

Boodoo A A, Ramjee R, Rowe J B and Dolberg F 1990 Evaluation of the basal forage diet of village cows. Livestock Research for Rural Development 1:in press

Dolberg F and Rowe J B 1984 Research, Demonstration and Training in the Use of Agricultural By-products for Livestock. UNDP Project MAR/82/007. Consultancy Report. UNDP: Mauritius

Gaya H, Hulman B and Preston T R 1982 The value for milk production of different feed supplements: Effect of cereal protein concentrate, poultry litter and oil seed meal. Tropical Animal Production 7: 134-137.

MaPoon L K, Delaitre J C and Preston T R 1977 The value for milk production of supplements of mixtures of final molasses, bagasse pith and urea, with and without combinations of maize and groundnut cake. Tropical Animal Production 2:148-150.

Preston T R and Leng R A 1987 Matching Ruminant Production Systems with Available Resources in the Tropics and Subtropics. PENAMBUL Books Ltd: Armidale NSW, Australia