Previous Page Table of Contents Next Page

Dairy cattle breeding in tropical South America

Lucia Pearson de Vaccaro

The tropical region of South America is characterized by an extreme range of climatic types. Of particular importance to its dairy industries are the modifying influence on temperature of the Andes mountains, which cover large areas of all the northern countries except Brazil, and the cooling effect of the Humboldt current on the climate of coastal Peru. The region's population of milked cattle includes cows of European breeds, distinct Criollo1 breeds such as the Colombian Costeño con Cuernos (horned Sinú), nondescript Criollo types, various zebu breeds and miscellaneous crosses between them all. The Criollo cattle have a history dating back to nearly 400 years on the continent. They are found at all altitudes, from the extreme lowlands to as high as 4 000 m in the Andes. In areas where climatic stress is not severe, most of the milk is produced by European breeds (now mainly North American Friesians) which are kept either on grassland in the mountains or in dry-lots around the cities. Many of them are imported or have been bred locally from imported semen, and they include animals of first-class genetic quality. As it is almost exclusively these cows that are performance recorded, statistics from the region relate mainly to them and are by no means representative of the total cattle population nor of the whole range of climatic zones. Apart from records from a few herds belonging to institutions or progressive farmers, there are relatively few published data for traits other than milk yield. Some examples from commercial farms are shown in Table 1, and it is evident that levels of performance which are fair by temperate zone standards are being achieved in these elite herds.

Lucia Pearson de Vaccaro is with the Departamento de Productión Animal, Universidad Nacional Agraria, Lima, Peru.

1 The term Criollo is used for the descendants of cattle originally imported by the Spaniards and Portuguese into Latin America.


It is to these herds that most research effort has hitherto been directed. In the field of breeding there is considerable interest in the comparative suitability of Friesian cattle of different origins for use in the region. From commercial herds near São Paulo, Brazil, it appears that the Dutch Friesian may have a slight advantage over the North American Friesian (Table 1) in areas where milk is sold according to solids content. This seems a rational basis for marketing throughout the region as malnutrition is widespread and transport difficult. Another comparison of Dutch and North American cattle is in progress under experimental conditions in Venezuela. In Peru, a comparison was made between the performance of locally bred Friesians and those imported from the United States and Canada (Sarmiento, 1970). The mean milk yields over the first six loctations were 4 924 kg, 5 125 kg and 5 263 kg for the three groups respectively, based on a total of 4 122 records. The mean age at first calving, 31 months, was the same for all groups and the interval between calvings about one month longer for the imported cattle. There is thus, predictably, little reason for the preference to import rather than breed herd replacements locally.

With regard to imports, interesting results were reported recently from Maracay, Venezuela (McDowell, 1974). Of a group of heifers brought in from New York aged between 10 and 14 months, 60 percent were still in the herd five years later, while none of a second group, moved at the same time but already pregnant, survived. The two groups left progency of, respectively, 2.5 and 1.2 female survivors in the herd per head. This adds to the evidence from elsewhere in the tropics that better adaptation results if heifers are imported well before breeding age.

There is also interest in the region as to whether dairy bulls should be selected specifically for the various ecological zones. So far the only published evidence comes from farms in Ecuador, ranging in altitude from 2 250 to 3 400 m, with mean annual (temperatures between 10 and 17°C (Roman, 1970). The sire × herd interaction accounted for only 2.9 percent of the total variation in milk yield. These results might have been more important if farms in the extreme lowlands had been included in the study, but they match with existing evidence that it is not necessary to select sires for the conditions under which their daughters will perform. Thus there is still no evidence that the best proven sires from temperate zones are not also the most profitable for tropical use.

Special problems

Two areas in the region present special problems to the milk producer: the extreme lowlands, which include vast tracts of jungle sparsely inhabited by cattle or man, and the Andean highlands.

Highland areas

In the highlands, European breeds and Criollo cattle are kept up to altitudes of about 4 000 m. Of the European breeds the Friesian and the Brown Swiss are the most popular, and the latter is traditionally supposed to be less susceptible to altitude sickness (brisket disease) than the former. However, problems with the Friesian seem to be less serious if they are really well fed and managed, and at least one Friesian herd is kept successfully as high as 4 000 m in Puno, Peru. The difficulties of feed supply are probably the most serious obstacle to increasing milk production. Concentrates are generally too expensive to use freely, and seem likely to remain so because of high transport costs. Present trends are to develop varieties of the improved temperate grasses, legumes, and forage crops such as oats. Under such a grazing regime at Huancayo, Peru (altitude 3 200 m), a mixed herd of Friesians and Brown Swiss cattle produced an average yield of 10 kg milk daily when stocked at five per hectare.

Table 1. Examples of dairy cattle performance in recorded herds in tropical South America

LocationYearBreedNumber of herdsLactation performanceAge at first calvingInterval between calvings
Number of milk recordsMilk yieldDaysFat
     Kg KgPercentMonths
(alt. 2 250–3 400 m)
1948–67Friesian (reg.) 7 3073 8882841313.3431.214.8
Friesian (grade) 26 2933 1182841063.4433.714.3
(São Paulo)
1943–66Dutch Friesian 2 5093 7852681443.80  
North American Friesian 5273 9582861383.47  
All Black & White Friesian 21 1443 6052671303.59330–3612.3
Red & White Friesian 3 2113 4542691273.663212.4
Jersey 2 4982 4762691224.90 12.4
Brown Swiss 1 7222 585264993.82  
Gir 8902 2152591084.89  
Pitangueiras (5/8 Red Poll, 3/8 Guzerat) 3632 8042501123.98  
Guzerat 861 8112451045.72  
Red Sindhi 212 0462461085.25  
Peru (Lima area alt. 500 m)41958–68Friesian (reg.)121 6354 931   3013.7
Friesian (grade)121 4014 917   3213.5
Peru (Lima area alt. 500 m)41958–68Friesian (reg.)121 6354 931   3013.7
Friesian (grade)121 4014 917   3213.5
     Mean daily milk yieldFatHerd in milk
(Lima & north alt. 500 m)51973Friesian (majority)335 83914.13.288
(Cajamarca, alt. 2 220 m)51973Friesian211 08911.73.670

1 Roman (1970).
2 Alves Netto et al. (1967).
3 Most frequent age class.
4 Sarmiento (1970).
5 Pallete (1974).


Dairy Criollo bull, Turrialba, Costa Rica.


Milk production in the lowlands is generally backward. Perhaps the only example of a thriving local industry is that of the Maracaibo region of Venezuela, where Criollo Lechero and crossbred (mainly with Brown Swiss) cows produce average yields of 1 500–2 000 kg per lactation, principally from grazing. The performances of Criollo, Costeño con Cuernos, Friesian, Brown Swiss and crossbred cattle are being compared at Turipaná in northern Colombia, where the annual mean temperature is 28°C and the average relative humidity 83 percent. From results so far published (Table 2), it appears that first-cross cattle gave as much milk as the exotic breeds but were easier to get in calf. The Criollo cattle produced less when milked alone than with their calves at foot and, although their conception rates were good, the relatively low proportion of animals pregnant again by 100 days after calving was due to difficulty in detecting their oestrus periods (Huertas, 1972).

Work in Peru suggests that the costs of preparing and maintaining jungle land are so high that cattle enterprises, at least by themselves, are of doubtful profitability. The only recent published dairy records from the region refer to Dutch Friesians taken at an average age of 19 months to Pucallpa, where the annual mean temperature is 25.4°C, with mean maximum and minimum values of 30.8°C and 18.4°C. The cattle were kept mainly on pastures of Hyparrhenia rufa and, because of prohibitive costs, only a little concentrate feed was given at milking time. Their performance (Meini, 1973) is summarized in Table 3. Their-reproductive performance was typically poor and is overestimated by the data shown, as conventional statistics such as these do not take into account the number of cows which do not conceive or calve. Here, long calving intervals were due to anoestrus under a system of inspecting the cattle twice daily for heat, irregular oestrus periods and repeat services (AI).

Table 2. Mean performance of European, Criollo and crossbred cattle at Turipaná. Colombia

BreedLactation performance1Reproduction2
Yield/dayDaysNumber of recordsPercentage pregnant by 100 days after calvingServices per conception
U.S. Friesian (F)11.0305216553.0
U.S. Brown Swiss11.030552672.6
Costeño con Cuernos (CCC)4.6180475651.7
1/2 F-1/2 CCC10.030090811.9
3/4 F-1/4 CCC10.0298---

1 Pineda (1971).
2 Huertas (1972).


Dairy Criollo cow, Maracay, Venezuela.

There is limited preliminary evidence from the same station that Criollo cattle (of highland origin) and Brown Swiss may be more productive than the Friesians because their yields of milk are similar but their reproductive performance is better. For instance, the mean calving interval of Criollo cattle at the station is 13.8 months.

Table 3. Performance of imported Dutch Friesian cattle at Pucallpa, Peru1

 Interval to first oestrus afterServices per conceptionAge at calvingLactation
Milk yieldDays
ArrivalFirst calving
Days MonthsKg 
First pregnancy/ lactation:      
Mean1061283.2381 258270
Coefficient of variation (%)866455122830
Number of observations252220201818
Second pregnancy/ lactation:      
Mean--2.7531 757300
Coefficient of variation (%)--5911247
Number of observations--22181818

1 Meini (1973).

Improvement through breeding

Improvement through breeding has been slower than it need have been throughout the region, and the reasons for this seem worth examining. First, there is widespread ignorance at government and farm levels as to what genetic quality for milk production consists of and how it is measured. The fact that AI and recording services exist in many parts of the region is frequently considered to be sufficient to solve all problems; also, the bulls used are often of inferior quality. An illustration of this comes from a study of Friesian sires used through AI in the Lima area in 1968. It was found that 42 percent of the bulls had negative progeny test ratings, that is, were sires whose future daughters would be expected to produce less milk than their contemporaries (Vaccaro et al., 1968).


Polled Sinú cow, Colombia.

Colombian Costeño con Cuernos (horned Sinú) cow, Turipaná, Colombia.

Costs are inflated and genetic progress hampered by the continued importation of bulls which, were they of first-class quality, would probably not be for sale. Instead, bulls for local use should be bred from the semon of the best foreign proven sires and selected cows. A great deal of emphasis is commonly given to “type,” which undoubtedly does have a high commercial value throughout the region, whereas greater publicity should be given to the fact that selection for physical characteristics does little or nothing to improve dairy productivity.

Second, selection is hindered in many parts of the region by a shortage of females. Instead of resorting to frequent importation, longer term benefits would result from investing in heifer calf subsidies and relevant technical assistance. Poor reproductive efficiency is an important contributing cause of the scarcity and, in this connexion, interesting results emerged from a survey carried out by Zemjanis and Sanint (1963) on cows on 32 Colombian farms. Most of the 4 122 cows in the sample were kept in the highlands around Bogotá, where heat stress is not a problem.

An average of 3.3 services were required for conception, but ranged from 2.0 where commercial (mainly frozen, imported) seme was used, to as many as 11.5 where herd bulls were used through AI. Furthermore, the difference between results obtained when the same bulls were used naturally and artificially emphasized the need to improve semen handling and insemination techniques. It is also significant that as many as 47 percent of the sample were classified as anoestrus and that as much of this was thought to be due to faulty detection of heat. In the lowland tropics particularly it seems likely that heats would be missed unless cattle were inspected frequently: a study at Maracay suggested that a minimum of two (at 0600 hours and 1700 hours) and up to four observations daily may be necessary (Fenton et al., 1972). The cattle in question were imported Friesians, and it was found that genuine anoestrus tended to persist as long as the animals were losing body weight after calving. Best conception rates resulted from breeding the cows within 12 hours after the detection of oestrus. By measures such as improved nutrition and culling, it was possible to reduce the average number of “days open” in this herd from 552 to 220 over a period of 18 months (McDowell, 1974).

Other considerations

It may be argued, however, that genetic progress has a comparatively minor role to play in the overall improvement of the region's dairy industries. The other ingredients are well known: they include better feeding and management, disease control, price and quality regulations, credit services and marketing facilities. What has happened in the past is that improvement efforts have too often been insufficiently comprehensive or, because of political changes, too short-lived. There is nevertheless a genuine lack of information about genotypes of cattle for profitable lowland farming systems.

The lack of useful performance data from the lowlands is notorious, and quite incommensurate with the local resources and foreign aid already spent there on dairy work. One reason is that policies have usually changed too quickly to permit the accumulation of sufficient data for the purposes of genetic study. Also, policy decisions are generally made away from the region and on the basis of insufficient knowledge of the local environment. This is especially true where the lowland experiment stations are difficult of access. Such stations have the additional problem of attracting and keeping the highly trained personnel that tropical animal science requires, as facilities for family living are often very limited. Dairying research in the lowlands has been insignated because large areas of land are unexploited and urban people need dairy produce, not because dairying is a significant part of local agriculture. In fact, few lowland areas in the region have any tradition of keeping cattle. Experiment stations have already seen how this fact affects the success of research programmes, and it raises serious questions about the type of dairying systems, and consequently breeding programmes, which are really applicable.

Despite the difficulties which face them, governments in the region are actively interested in improving milk production. Peru, for example, which now imports slightly over 50 percent of its requirements, expects to be self-sufficient in milk products by 1980. The frequent political changes which characterize the region make progress difficult, particularly in animal breeding, but real hope lies in the unprecedented number of highly trained professionals educated at home or returning from abroad, provided that they are increasingly involved in the planning and implementation of national livestock policies.


Alves Netto, F., Fang, I., de Melo Telles, J.D., Gubiotti Fonzari, W.M. & Kvarnstrom, O.R. 1967. Comportamento médio das vacas e rebanhos controlados pelo Servico de contrôle leiteiro da Asociacão paulista de criadores de bovinos. 1945–66. Revista dos criadores, 38: 18–108.

Fenton, F.R., Branton, C., McDowell, R.E. & Benezra, M.V. 1972. Reproductive efficiency of a Holstein herd in a tropical environment. Proc. VIII Int. Biom. Congr., Noordwijk, Aug. 1972.

Huertas, E. 1972. A system for the management of reproduction in cattle herds. North Carolina State University, Raleigh, Ph.D. thesis, 87 p. (Unpublished)

McDowell, R.E. 1974. Department of Animal Science, Cornell University, Ithaca, N.Y. Personal communication.

Meini, G. 1973. Adaptación al trópico de Pucallpa de un hato Hostein holandés. Paper presented IV Congreso Nacional de Medicina Veterinaria y Zootecnia, Huancayo, Peru, Nov. 1973.

Pallete, A. 1974. Servicio de control de productividad lechera: resultados oficiales 1969–73. 17p. Programa de Mejoramiento Ganadero, Univ. Nacional Agraria, Lima, Peru.

Pineda, J. 1971. Informe de progreso del programa nacional de ganado de leche. Inst. Colombiano Agropec. Quoted in: Gómez, F. 1973. Producción de carne con razas lecheras. Paper presented at: Reunión técnica de programación sobre desarrollo ganadero del trópico húmedo americano, Guayaquil, Ecuador, Dec. 1973.

Roman, J.O. 1970. Genetics of milk production in Ecuador. Florida, University of Florida, Gainesville. Ph.D. thesis, 213 p. (Unpublished)

Sarmiento, D.A. 1970. Estudio del comportamiento productivo de vacas Hostein nacionales en la importadas, registradas y no registradas en la cuenca lechera de Lima. Peru, Univ. Nacional Agraria, Lima. Ing. Zoo. Thesis, 94 p. (Unpublished)

Vaccaro, R., Pallete, A. & Jara Almonte, M. 1968. Pruebas de progenie de los toros Hostein ustados en la cuenca lechera de Lima. Programa de mejoramiento ganadero, Univ. Nacional Agraria, Lima. Bol. téc. No. 1. 31 p.

Zemjanis, R. & Sanint, D. 1963. Fertilidad de ganado en Colombia. Agricultura tropical, 19: 7–15.

Previous Page Top of Page Next Page