Previous Page Table of Contents Next Page



Roberto Garcia Trujillo


The humid and semi-humid tropical areas of Latin America comprise approximately 70% of the total area of the region and are where 60% of the cattle are concentrated. Four different sub-areas exist, these are: a) native savannas, b) the “cerrados”, c)tropical humid and semi-humid forests with fertile soils and d) tropical humid and semi-humid forests with non-fertile soils.

80% of the soils in these zones are barren, especially those of the savannas and “cerrados” where improved pasture establishment requires high levels of fertilizers and soil improvement. Rainfall is a limiting factor for pasture development, due to the unequal distribution throughout the year and the fact that the dry season can last for almost seven months, except in the humid forest area. There are dairy cattle in the region (3–11 cows/100 inhabitants) but, because of the low productivity (750–1700 litres/cow/year), large amounts of milk and milk products are imported.

The present situation in the foreign market, the general economic crisis and different aspects of agricultural policy in the countries of the region are against the development of milk production and cattle production in general. These problems were exhaustively analyzed in the First Meeting of Cattle Development in Latin America and the Caribbean, sponsored by FAO in September 1988 in Montevideo. There are many different milk production systems in the region, but the non-intensive pasture systems with low milk producing cattle predominate, while the most productive and the most intensive systems are mainly found in the high tropical areas or in production systems using mainly European breeds, where cattle remain housed or semi-housed and are fed forages, hay or silages and a high proportion of concentrates, supplemented with imported raw materials.

In Cuba, milk production is based on the utilization of pastures in the rainy season and green and preserved forages and sugar industry by-products in the dry season. Holstein dairy cows and their crossbreds with Zebu cattle are employed.

Although research on pasture production, feeds, production systems and other aspects is scarce, results are available which show that it is possible to considerably improve milk production in the tropics. The objective of this paper is to offer the results of some studies carried out in Cuba.


In the humid tropics, where rainfall is not limiting and declines to critical levels for only 2 to 4 months, milk production based on pastures is limited by the quality of pasture and the direct or indirect effects of the climate on the more productive animals. In the semi-humid tropics, the lack of rainfall during the dry period (lasting from 5 to 7 months), seriously limits pasture production and quality. Under these conditions, which prevail in the majority of the cattle breeding regions in the tropics, the development of a more productive milk production system should consider the equilibrium between the needs of the animals and feed production. The basic elements are pasture management, fertilizer use, stocking rate, forage production and supplementation, all being related to the animal potential.

Natural grasslands and their productivity

Natural or semi-natural, non-fertilized grasslands in areas with soils of low fertility allow a low stocking rate which, in general, does not exceed 1 cow/ha and is frequently about 0.5 to 0.8 cattle units per ha. Milk production recorded in the sub-humid area does not exceed 4 kg milk/cow/day and 300 kg of milk ha/year when Zebu cattle are used, as in the case of the western plains of Venezuela (Capriles, 1982), or from 5 to 7 kg milk cow/day and 1600 kg milk/ha/year when crossbreds are used and the animals are fed medium to low quality forage in the dry season, as reported by Blydestein et al. (1969) in Costa Rica or in certain dairy areas of the eastern provinces of Cuba (Table 1). In general, these herds are characterized by short lactations, poor calving rates and low percentages of cows in milk.

The low milk production of the animals in the grazing system with natural or semi-natural pastures is not only a consequence of this, but of the poor resources, operations and techniques with which these animals are managed and where supplementary feeds play an important role in their survival on account of the poor pasture quality or amount.

Under these conditions, legumes could be one of the most economic ways to improve the production level of these herds. In this sense, Monzote et al., (1985), with the introduction of Glycine into natural pasture in areas of low rainfall (800 mm/year), found an increase in milk production from 1041 to 1684 kg/lactation, compared to cows on natural pasture. In the rainy season, the association produced from 3 to 5 kg of milk/cow/day more than with natural pasture although, during the severe months of the dry season, this difference was only 0.3 to 1.2 kg of milk/cow/day.

Another alternative under these conditions could be the use of forage areas of sugar cane as a supplement during the dry season.

Table 1. Milk production results under different grazing systems in the tropics.
PasturesAreaIrrigationFertilization kg/ha/yearSupplementationBreedStocking rateMilk production kg
BativeSHnone savannahnonelownoneZebu0.5–0.8240–350120–300
NP+legume associationSHnoneP and Klownonexbred11650990
Improved+tree legumeSHnonemediumhighlowxbred2.5–2.72500–32006000–9600
Improved+tree legumeSHyeshighlowmediumHolstein2.5–3.03900–43007000–10500

SH - Semi-humid H - humid Fertilization: low 50–80, medium 130–180, high 240–350

Concentrate kg/cow/day low: less than 2 medium: 2–3 high: more than 3

Improved pastures

The utilization of improved pastures and fertilizers markedly increases stocking rate capacity, milk production per ha and individual yield, especially when breeds and crossbreds of medium to high milk potential are used.

In general, stocking rate is increased to 2.2–4.5 cows/ha in the sub-humid regions according to the type of pasture and the level of fertilizer. Milk production/cow is increased to 6–8.5 kg/cow/day and up to 2500 kg/lactation with crossbred animals and from 10 to 14 kg of milk/cow/day (3000–4000 kg/lactation) with European breeds. Milk production/ha could reach or exceed 12000 kg/ha/year (see Table 1).

In the humid regions, the stocking rate of the grassland could reach 6.5 cows/ha with 14000 kg of milk/ha/year and 8.5 to 9 kg/cow/day, as demonstrated in some areas of the region (Cubillos et al. 1975, Muñoz et al. 1988b).

The effectiveness of the systems based on improved pastures depends on the equilibrium between pasture selection, the level of fertilizer use and stocking rate. An imbalance in this sense could ruin the system and the stocking rate is a decisive factor in the productivity of dairy herds. Different trials have demonstrated that, when medium potential animals are used, the production per hectare can be employed as an indicator of productivity but when high potential cows are used and a higher production per hectare is obtained, the potential of the animals is only exploited to the extent of 50 to 60% (Figure 1), provoking a negative effect on reproduction and animal health.

The level of stocking rate and N fertilization are closely related. Our results show that from 50 to 100 kg N/ha/year, according to the stocking rate and system employed, are needed for soils of medium to low fertility per cow. A reduction in stocking rate, under Cuban conditions, has markedly improved individual milk production, without affecting milk production per ha, fertility, survival and health (Table 2).

The most productive pastures found are coastal bermuda grass with low stocking rates (2 to 2.5 cow/ha), common guinea grass and likoni for medium stocking rates and star grass with high stocking rates (4 to 5 cows/ha). Coastal bermuda grass requires a high level of fertilizer, irrigation in the dry season and careful management. Also,short grazing rotations in the rainy season (12 to 18 days), together with night grazing, allow ample selection, availability of pastures (50 kg DM/cow/day) and good milk production in European breeds. In the dry season, overgrazing should be avoided by restricting grazing time (3 to 4 hours/day) or housing some of the animals.

Figure 1. The effect of stocking rate and the type of pasture on milk production per cow and per hectare (from Pérez Infante, 1971 and others in Cuba).

Figure 1

Table 2. Stocking rate (SR) reduction and performance of various dairy herds in Havana Province (Martinez, 1980; Salinas, 1988, unpublished).
  SR cows/ha Improved pasture % Milk/cow Yield (kg)/hectare Fodder supplement DM/cow/d Animal losses % Birth rate %
Case A: Three dairy units
Unit A 2.84 33 1082 3078 4.2 3 58
Unit B1 2.26 38 1650 3741 3.8 6 67
Unit C 2.24 73 2102 4624 2.2 4 76
Case B: Dairy enterprise (15000 cows) then reduce stocking rate
Previous 2.9 - 2428 7043 - 11 68
After 2.5 - 2662 7037 - 9 72
Case C: Dairy district (3800 cows) then reduce stocking rate
Previous 3.3 - 2203 7279 6.4 13.8 -
After 2.2 - 2844 7580 4.4 10.0 -

1 Stocking rate reduced one year previously.


In the sub-humid tropics, large amounts of forage are used daily to make up for pasture deficiency in the dry season. In Cuba, various alternatives have been studied and used to this end. They can be classified as follows:

  1. Sowing areas of forage with irrigation so as to produce silage in the rainy season and green forage in the dry season.
  2. Conserving pasture surplus in the rainy season to produce silage and hay supplies for the dry season.
  3. Sowing forage areas with sugar cane which is harvested 12 months later during the dry season.
  4. Utilization of coarse by-products of the sugar industry (bagasse -molasses-urea, pre-digested bagasse).

The main forage used in Cuba is king grass, which under production conditions, yields from 97 to 130 t/ha/year when fertilizer (300 to 350 N kg/ha) and irrigation in the dry period are used. The application of manure (25 to 30 t/ha/year) favours forage stability, decreases the need for N fertilizer (100 to 150 kg H/ha/year) and the use of P and K. Because this forage has a low DM content (18%) when it has a good nutritive value (from 10 to 11% CP), it is necessary to pre-wilt it so as to obtain a good quality silage.

Star grass and guinea grass are also used as the main forage with good yields and supplementary forages such as sorghum and sunflower are used as pure crops or are intercropped with pastures, new sowings or forage areas during the dry period.

In general, green forage is not used as the sole source of roughage in the dry season, except for some high producing herds or high producing groups of cows and when there is great availability, being employed as a supplement to silage diets and sugar cane (30% green forage: 70 % sugar cane or silage) (Esperance and Perdomo, 1978, Muñoz et al., 1988a).

Medium quality silages, which are more abundant in commercial units, when supplied together with grazing on non-irrigated pastures, produce from 3 to 5 kg of milk/cow/day, but the animals lose weight and have short lactations. In order to obtain an average of 8 kg of milk, supplementation of 1.2–2kg of concentrate is required with the addition of 3–40% of total roughage as green forage. Above this level of supplementation, a response of 1.5 kg of milk/kg of concentrate could be expected (Gutierrez et al., 1988) (Figure 2).

Figure 2. Milk production from tropical silage or forage of medium quality (adapted from Gutierrez et al., 1988).

Figure 2

Milk production under non-irrigated conditions with animals fed silages is strongly related to the DM percentage of the silage and its consumption, an increase of 0.3 kg of milk and 0.37 kg of DM consumed per percentage unit increase in dry matter being reported.

Another difficulty with tropical silages is the loss of nutrients and materials produced during silage processing, when silages are not made with adequate techniques. Low capacity silos with long cut forage and high humidity could lose almost 60% of the material. However, when high DM and finely cut forages are used and the material is placed in adequate places or bunker silos, the losses could be diminished to 15% at the most.

Sugar cane is characterized by its high yields (up to 170 t fresh matter/ha with low N fertilization (100 to 150 kg N/ha/year) and is harvested during the dry period, thus there are no conservation expenses. Although used as forage, low consumptions are observed, particularly when it is not adequately mixed with other forage feeds and supplemented with NPN and minerals. When using sugar cane for animal diets, the material should be finely cut and 1% urea (Perez Infante and Garcia Vila, 1975), 0.12 to 0.15% sulphur (Ruiz, 1979) should be added and also it should be combined with 30% forages, pastures or both (Muñoz et al., 1988a), and the necessary concentrate supplementation to cover the animal requirements.

Experiments carried out in the Caribbean area (Garcia and Neckles, 1983; Perez Infante, 1975; Garcia Trujillo et al., 1981 unpublished; Senra et al., 1988 unpublished) show that as the percentage of sugar cane in the DM of the ration is increased, milk production decreases by approximately 1.46 kg milk for each 10% of sugar cane inclusion, yielding from 7 to 10 kg of milk/cow/day when 50% of sugar cane is added. This indicates that feeding factors and the animal potential could vary the response (Figure 3).

Figure 3. Sugar cane forage level in the diet and milk yield.

Figure 3

The sugar cane industry produces various fibrous by-products for animal feeding. The most widely used in Cuba is bagasse, pre-digested with 2–3% sodium hydroxide, to which 15% molasses and 10% urea are added, and which finally contains 8.37 MJ ME/kg DM and from 11 to 12% CP. This feed is used as forage in the dry season and consumed at a rate of 6–7 kg/cow/day.

At Indio Hatuey Branch Station, a comparison between the forage systems I, II and III (see above), crossbred (3/4 Holstein, 1/4 Zebu) cows on pasture without irrigation for 3 years showed higher milk production per lactation than on the sugar cane system (2565 vs 2427 kg), but the sugar cane system produced higher fertility (80 vs 67–72% cows pregnant per year), lower cullings (11 vs 22–25%) and thus the milk production/ha was higher (7704 vs 6640 kg) (Table 3).

In a later study where systems II and III included a protein supplement and stocking rate decreased to 2.5 cows/ha, milk yield per lactation was increased by 150 and 424 kg for the conservation and sugar cane systems, respectively. Protein supplementation resulted in calving intervals of 393 days on the sugar cane system and reduced both the use of concentrates and the cost of production in 10%.

The comparison of system I, III and IV with Holstein cows at ICA showed a higher milk production per cow on the silage system than on the sugar cane or bagasse pith systems (3553 vs 3100 kg/lactation and birth rate (87 vs 83%), but the production cost was slightly higher (0.22 vs 0.20 Cuban pesos/kg of milk).


Commercial concentrate supplementation is practised in dairy units using medium to high potential cows. The response to concentrates obtained, in terms of milk production, ranges from 0.2 to

Table 3. Comparison of three milk production systems (Garcia Trujillo, 1981, unpublished).
 With conservation areasKing Grass forage areasSugar Cane forage areas
Milk production
cows pregnant/year % 677280
Culled cows %
reproductive problems142211
Feeding cost
(pesos/kg milk)0.0880.0840.083

2.3 kg of milk/kg concentrate. This response mainly depends on the differential between the production potential of the basic diet and the cow's potential (Garcia Trujillo, 1988). Our experience recommends the use of limited amounts of the available concentrates 21 days before calving and in cows producing more than 12 kg of milk per day, although if basic diets of poor quality are used, economic responses to supplementation can be obtained but foreign currency will be wasted.

Legumes are the main crops available for total or partial substitution of concentrates in diets for dairy cows. Pereiro (1985) demonstrated, in a series of experiments, that with 30–38% of the total area in Glycine wightii with fertilized and irrigated coastal bermuda grass, irrigated in the dry season and grazed daily or on alternate days (3 hr daily), with 3.8 kg concentrate/cow/day, production of 13 kg milk with 2.8 cow/ha was achieved. Milk productions from 12 to 14 kg/cow/day were also obtained when Glycine was used as a protein supplement given to cows consuming diets of forage, silage and hay.

In the Institute of Animal Science in Cuba, Leucaena and Glycine have been used with 3.2 and 2.7 cows/ha respectively, with a 50% reduction in the concentrates needed to produce approximately 14 kg of milk/cow/day. Furthermore, good quality milk, a high calving rate (86%) and no health problems were observed in these cows.

Other alternative supplements have been developed using sugar cane by-products and sugar cane (Table 4). Muñoz (1982) worked with nitrogenous activator supplements (NAS) formed by filter cake mud (45%), molasses (40%), urea (11%) and minerals (4%), supplied to a dairy herd consuming fertilized pasture in the rainy season and non-irrigated pasture plus silage in the dry season, and produced 10.3 kg milk/cow/day with a calving rate of 83%. NAS increased the digestibility and consumption of low and medium quality basic diets.

Table 4. Some supplements produced from by-products of sugar cane industries or sugar cane.
Products Formulas %
1 2 3 4 5(NAS) 6
Saccharina 90 70 - - - -
Sun-dried sugar cane - - 50 - - -
Dried Filter Mud - - - 30 40 -
Molasses 5–6 5– 15 15 35 69
Cereals - 10 20 41 - 21
Proteins - 10 10 10 0–10 4
Urea 0–1 0–1 1 1 10 3
Minerals 4 3 4 3 5 3

NAS: Nitrogen activated supplement Supply: 350 g/kg milk Except NAS (1.5–2.5 kg cow)

Sugar cane dehydrated in the sun (50%) or dry filter cake mud (30%) have been used at lCA to produce supplements where the remaining components were molasses, cereals and minerals. These supplements are supplied at a rate of 300 g per kg of milk and allowed the production of 9 to 10 kg of milk during the dry season. More recently, a new product obtained from clean sugar cane, without leaves, enriched with protein, has been obtained by an aerobic fermentation process. This new product was developed by Dr. Arabel Elias at our Institute of Animal Science and has been named “Saccharina”. Saccharina can have between 9 to 11% of CP and approximately 10.4 MJ ME/kg DM for ruminants. It is used to supplement all animal species and from 10 to 12 kg of milk per cow per day have been obtained in dairy cows supplemented with this product.


In the system based on pastures, stocking rate adjustment is of vital importance so as to have a stable and productive system. 90% of the animal potential should be obtained from pastures and supplements. Stocking rates should also be adjusted to avoid surplus forage supply during the rainy period and the surplus conserved for the other months. Pasture availability of 50 kg DM/cow/day should allow the exploitation of the most productive pastures.

In the semi-humid areas of the tropics, the production of forages to cover pasture deficiency in the dry season is necessary. There are various options and one of them is to use sugar cane for crossbred cattle in small units. In large units or with cattle with a higher potential, silages and green forages should give better results, if these are of good quality.

Legumes are essential components of any system of dairying. The grass-legume pastures are better with low stocking rates, while with medium and highly stocking rates, forage legumes are preferred. Legumes save concentrates and fertilizers and improve the productive performance and health of the herd.

Commercial concentrates are very expensive thus these must only be used in those animals with the required physiological status (European breeds, end of the pregnancy, beginning of lactation).

Low producing crossbred cattle (10 kg milk/cow/day) can be fed with by-products obtained from sugar cane.


Blydestein, J., Louis, S., Toledo, J. and Camargo, A. 1968 Productivity of tropical pastures. l. Pangola grass. Journal of the British Grassland Society 24: 71–75.

Capriles, M. 1982 Sistemas de producción de leche y carne para los llanos occidentales de Venezuela. Informe Final IPA. Universidad Central de Venezuela. pp. 141–176.

Cubillos, S., Muñoz, H., Ruiz, M., Deaton, O. and Fuentes, G. 1979 Un sistema de producción de leche para pequeños productores. Resúmenes VII Reunión Asociación Latinoamericana de Producción Animal, Panamá, pp. 8–11.

Esperance, H. and Perdomo, A. 1978 Ensilaje y/o forraje para la producción de leche. Pastos y Forrajes 1: 415–436.

FAO 1988 Examen del sector ganadero de la región de America Latina y el Caribe. Informe de la Comision de Desarrollo Ganadero para America Latina y el Caribe. Primera Reunion. Montevideo, Uruguay. 20–30 Sept. 1988.

García, G. and Neckles, F.A. 1983 Feeding sugar cane for the productions of meat and milk. Feeding of Animals in the Caribbean. A workshop CARDI and SPC. Trinidad WI, pp. 88–100.

García-Trujillo, R. 1983 Potencial y utilización de los pastos tropicales para la producción de leche. In. Los Pastos en Cuba. Tomo 2. Utilización. J. Ugarte, R.S. Herrera R. Ruiz, R. Garcia, A.M. Vázquez and A. Senr. (Eds.). EDICA pp. 247–298.

García-Trujillo, R. 1988 A study of concentrate supplementation to dairy cows. Cuban Journal of Agricultural Science 22: 38–46.

Gutiérrez, A., Muñoz, 8., Esperance, M. and Michelena, J. 1983 Utilización del ensilaje en la producción animal. In. Los Fastos en Cuba. Tomo 2. Utilización. J. Ugarte, R.S. Herrera, R. Ruiz, R. Garcia, C.M. Vázquez and A. Senra (Eds.). EDICA pp. 437–476.

Monzote, M., Aira, Gómez, I., Rill, S. and Barbón, J. 1985 Comportamiento de la asociación pasto natural/glycine a nivel de extensión. Asociación Latnoamericana de Froducción Animal. Memoria. Volumen 20. p. 81.

Muñoz, E. 1982 Producción de leche y digestibilidad en vacas alimentadas con gramineas tropicales suplementadas con altos niveles de nitrógeno no proteico en secano. Thesis Candidate to Doctor in Veterinary Science. Agricultural Science Insti- tute for Higher Education, La Habana, Cuba.

Muñoz, E., González, R., Ruiz, R. and Galindo, Juana. 1988a Uso de la caña de azúcar para la producción de leche. Informe final de etapa de investigacióo. Instituto de Ciencia Animal. La Habana, Cuba.

Muñoz, H., Ramsammy, P. and Lallapachan, V. 1988b Searching for appropriate technology in dairy production. I. St. Stanislaus dairy unit milk production performance. Sugar Cane Fee Center Workshop. Trinidad and Tobago, Sept. 11–15.

Pereiro, A. 1985 Utilización del pastoreo restringido de glycine (Neonotonia wightii) como suplemento a vacas lecheras de mediano potencial suplementadas a base de pastos o forrajes conservados. Thesis Candidate to Doctor in Veterinary Science. Agricultural Science Institute for Higher Education, La Habana, Cuba.

Pérez-Infante, F. 1971 Efecto de diferentes especies de pastos y sus combinaciones en la producción de leche. Memorias Microestación de Pastos Niña Bonita. INRA, La Habana.

Pérez-Infante, F. and Garcia Vila, R. 1975 Uso de la caña de azúcar en la alimentación del ganado en época de seca. 1. Efecto de la adición de urea en el consumo y produccion de vacas lactantes. Revista cubana de Ciencia Agricola 9: 109–112.

Ruiz, E. 1979 Utilizaón de la caña de azúcar en la producción de leche. Thesis Candidate to Doctor in Veterinary Science. Agricultural Science Institute for Higher Education. La Habana, Cuba.

Previous Page Top of Page Next Page