Previous Page Table of Contents Next Page

Some aspects of sheep breeding in the tropics

Helen Newton Turner

1. Nellore ram from India, typical of the hair sheep (with underdown) that occur in the wet tropics.

1. Nellore ram from India, typical of the hair sheep (with underdown) that occur in the wet tropics.

The area between the tropics of Cancer and Capricorn covers a wide range of climate and terrain, but only those sheep-breeding regions that have high temperatures and either light or heavy rain, falling seasonally or regularly during the year, are discussed. These are the so-called “dry” and “wet” tropics. This article is concerned with indigenous and exotic sheep breeds in these tropics, and genetic improvement techniques.

Indigenous tropical sheep breeds differ greatly from European breeds. Nearly all the latter are woolled and many are multipurpose. Fleeces range from primitive carpet-wool types (such as the Scottish Blackface), with a coarse, partly medullated outer coat and a finer inner one, to the supreme apparel-wool breed, the Merino, with its fleece of fine, relatively uniform fibre diameter.1 Indigenous tropical breeds, on the other hand, grow either coarse fleeces, usually containing medullated fibres and kemp, or a coat of short or long hair with little or no wool.2 They provide meat, milk, skins (sometimes all three) and occasionally their manure is also utilized, for fertilizer or fuel; the coarse-wool or long hairy fleeces are used for carpets, rugs, blankets, and even tents.

Helen Newton Turner is Honorary Research Fellow, Division of Animal Genetics, csiro, Australia.

1 A medullated fibre has a continuous or interrupted tube of air down the centre.
2 Kemp consists of shed fibres in the fleece, which are usually undesirable because of brittleness and different dye uptake.

Tropical and temperate sheep breeds

Differences between the woolled sheep of European origin and the coarsewoolled or hairy sheep of the tropics are found in skin structure. All sheep, like other pelaged animals, produce fibres from two types of skin follicle, primary (P) and secondary (S). Large breed differences occur in the total number of follicles per unit skin area (“population density”), and in the ratios of numbers (S/P) and diameters (dp/ds) of secondary and primary follicles. Table 1 shows the preeminence of the Merino, which has the highest population density, the highest S/P ratio, a low dp/ds ratio, and the finest fleece. There are various Merino strains (fine-, medium- and strong-wool) but all show this superiority over other breeds.

Table 1. Skin data for some European and tropical sheep breeds

Area of originCountryFleece typeBreedSkin dataSource
Follicle population densitySecondary/primary follicle ratio (S/P)Primary/secondary diameter ratio (dp/dS)Average fibreAverage medullated fibres
    f/mm2  μPercent 
EuropeSpain (origin) Australia (data)Fine wool (apparel)Merino      
Fine72191.118nilCarter (1955)
United KingdomDown wool apparel)Southdown2861.0251"
United KingdomLong wool (apparel)Border Leicester1641.436"
United KingdomCarpet woolScottish Blackface732.74913–48Carter (1955 von Bergen (1963)
Dry tropics light seasonal ain)Northwest IndiaCoarse woolBikaneri Chokla29.72.326Narayan (1960)
Carpet wool
    "         "
    "         "
    "         "Bikaneri (unspecified)–3353Krishna Rao et al. (1960)
Wet tropics (heavy seasonal rain)Southern IndiaCarpet woolBellary6.61.43.15635"
Hair (with underdown)Nellore6.01.75.35934"
NigeriaHairYankasa and Uda10.03.510.03430Burns (1967)
(rain every month)West MalaysiaCarpet wool4 similar breeds6.21.12.64842Smith and Clarke (1972)

1 (…) not available.
2 Per square millimetre taut skin.

The Bikaneri strains in Table 1 are from Rajasthan in northwest India, just outside the tropics; they are included for lack of other data. All Indian breeds are similar in having low population density, low S/P and high dp/ds ratios.

The hair breeds also have low S/P and high dp/ds ratios; the fibres from secondary follicles constitute a soft short undercoat of down.

Table 2 gives production data for European and tropical breeds. This information, particularly for tropical breeds, is scarce, and one can only state again the repeated plea of FAO (FAO, 1967) for complete documentation of productivity, so that the world's genetic resources can be used to their fullest extent.

Genetic improvement of tropical breeds

Looking at Table 2, it is not surprising that those trying to improve tropical sheep have often seen European breeds as the solution, either for crossbreeding or upgrading. But figures for the Merino in northern and southern Australia illustrate the effect on production of the very great differences in nutritional and climatic conditions prevailing between tropical and temperate areas. Clean wool weights in the north are only 1.4 to 1.6 kg, compared with 2.5 kg or more in the south, while lamb-marking percentages are only 20–40, compared with over 90 percent in the south (Dunlop, 1962; Brown and Williams, 1970).

An immediate question that arises in this context is whether low production in a breed under investigation is due to husbandry techniques, which may be changed, or to climate, which usually cannot. If husbandry were improved, would productivity respond as Demirüren (1972) has shown it to do with eastern Mediterranean breeds? Or, if genetic changes were made, could the changed genotype survive and reach its genetic potential under current conditions of husbandry and climate? In attempting to answer these questions it is necessary to define the objectives of sheep breeding in a local context, and examine inter alia the genetic techniques available for securing the desired improvement and the characteristics related to heat adaptation.

2. Woolled sheep from Java. They grow a poor carpet fleece but are3. Ewe and lamb of a hair breed, raised for meat production in
very prolific.Java.

Definition of objectives

Sheep produce meat, milk, fibre (wool or hair) or skins; more than one product may be important. Market requirements help to govern objectives. The main target could be one (or more) of three:

  1. Home consumption (by the producer or in his village).
  2. Domestic market, within the country of origin.
  3. World markets.

Table 2. Production data for some European and tropical sheep breeds

Area of originCountryFleece typeBreedProduction dataLamb-marking percentage Source
Greasy fleece weightBody weight (adult ewes)
Dry tropics Little rainSomaliaHairSomali (also called Blackhead Persian)-5293Mason and Maule (1961)
Light seasonal rainNorthwest IndiaCarpet-woolBikaneri (unspecified)1 5–3.027 3070–88Lall (1956)
SudanHairSudan Desert-70100Mason and Maule (1961) McLeroy (1961)
Wet tropics Heavy seasonal rainSouthern IndiaCarpet-woolBellary0.7–1.432Lall (1956)
HairNellore-38Said to lamb every 9 months  "        "
Europe (Spain - temperate)Australia      
dry tropics
Apparel-woolMedium-wool Merino2.9–3.832–5420–40Brown and Williams (1970)
  3.5–6.0 90+Dunlop (1962) Iwan et al. (1971)

4. Blackhead Persian ram.

5. Long-tail, haired Sudanese desert ram. Note cord to prevent indiscriminate breeding.

Market requirements will vary from place to place, and the demands of world markets are likely to differ from those at the home or village level. It is essential, however, that improving productivity for home consumption should not be overlooked in the desire for world trade.


In most developing countries meat is likely to become in increasing measure the most important single consideration in sheep breeding in the foreseeable future. This would cater to all three types of market, but with great qualitative variations between markets. Lean meat is often a requisite, but in some areas the fat tail is still a delicacy. There are markets requiring very young lamb, others lamb at 3–6 months, while others prefer mutton.


Sheep's milk in tropical countries is mainly for home consumption and could be an important item of diet.

Apparel wool

For apparel wool on the world market, average fibre diameter is the most important single processing characteristic. But for the coarser types of apparel or blanket wool grown by tropical sheep, requirements may be different, particularly if the wool is to be spun by hand rather than by machine. Local needs should therefore be investigated.

Carpet wool

Carpet wool requirements on the world market are not clearly defined. Traditionally, wool with a percentage of medullated fibres is used, sometimes blended with coarse wool containing little or none. Proportions of the two components vary from 100:0 in traditional carpetmaking countries such as India and Iran, to 30:70 in the United States.

The Wool Research Association in Bombay is currently planning a project in which wool samples from a number of Indian carpet-wool breeds will be processed separately into carpets and subjected to wear tests. Various measurements, including fibre diameter and medullation percentage, will be made. This project, combined with other, similar tests in the United Kingdom, should give a clearer picture of the processing requirements of carpet wool. Measurements are already available for wool traditionally used for carpets (von Bergen, 1963), but no comparisons of different wools have been made to assess processing loss or the life of the final carpet.


Much of this is used in local industry; some is sold on world markets. Requirements are not clearly defined.


Details of measured requirements for skins are hard to obtain. Some breeds (for example, the Sudan Desert) are traditionally known as sources of good glove leather, while the layered structure of Merino skins gives them a poor reputation for sheepskin products.

Genetic techniques

Selection within a breed

Selection must have operated in the tropics, as elsewhere, to produce the range of available breeds and strains. But genetic progress depends on the selection differential, which in turn depends on the numbers available. There is little scope for selection within small flocks.

The Australian Merino is an example of a breed which still responds to selection; various increases have been reported, including over 2 percent per year in clean wool weight (Turner et al., 1967) and from an additional 1.2 to 10 lambs born per 100 ewes mated per year (Turner, 1969). The 2 percent increase in clean wool weight per year would mean an increase of 20 percent in 10 years; that is, an initial weight of 2.5 would become 3.0 kg. With an initial weight of only 1.0 kg, as in some tropical flocks, a gain of 20 percent (= 0.2 kg) in 10 years might seem negligible. If the best animals were collected from a number of such flocks to form a ram-breeding nucleus, from which rams were later redistributed, an initial lift in production would be obtained which, added to subsequent gains from selection within the nucleus, would give a 10-year gain closer to that obtained in Australia.

The advantages of selection as a technique for genetic improvement are:

  1. The current type of product for the breed (carpet or apparel wool, for example) can remain the same, although quantity or quality can be changed.
  2. When selection is carried out in the environment where the animals are to be kept, selection is for high productivity in that environment, using adapted animals.
  3. Costs of selection are usually low.

The disadvantages are:

  1. Progress may be slow, unless special measures are taken (such as screening large numbers for a central nucleus).
  2. If a different product is the aim, rather than an increased amount of the same product, selection may not be able to achieve it. (Selection will not change an apparel- to a carpet-wool breed in a conceivable period.)

Carpet wool is in such demand in world markets that there is a strong case for keeping some of the tropical carpet-wool breeds and improving their productivity. However, if a greater supply of locally produced apparel wool is needed, both Lall (1956) and Narayan (1960) have pointed out that, at least among Indian breeds, types exist which could be selected toward finer wool production (such as the Chokla cited in Table 1).

Upgrading or formation of an intermediate breed

These two techniques, to increase productivity by using a superior breed or to change the product, share similar problems. Difficulties have arisen in the crossing of apparel-wool breeds (Merino, Corriedale) with carpet-wool or hair breeds.

Survival of introduced animals. Pure breeds from temperate areas have often suffered considerable stress when introduced to the tropics. The Australian Romney Marsh was taken to New Guinea some years ago, but quickly died out from parasite infestations. Many introductions of woolled sheep from temperate areas have been made in India. Most purebreds have suffered heat stress and contracted respiratory diseases; reproduction rates have been very low.

Pigmented lambs. A high incidence of pigmented lambs has sometimes occurred with crosses between Merinos and carpet-wool breeds, even from white parents. The countries involved are not strictly tropical and include Turkey, Egypt and India. No satisfactory genetic explanation has been advanced, but improvement programmes have been hindered as white wool is preferred.

Inhibition of medullation. The Merino and its crosses appear to carry genes which inhibit expression of medullation. This phenomenon is an advantage when using Merino types to upgrade carpet-wool into apparel-wool breeds.

Interference with wool growth. Genetic information was used very neatly by Burns (1967) in attempts to produce woolled sheep in Nigeria by crossing South African Merinos with northern Nigerian hair sheep (Uda and Yankasa breeds). The first-cross fleeces contained a high incidence of kemp, mixed with very fine wool, which was shed at intervals, leading to cotting. Fibre studies of the birth coat and fleece led to the conclusion that a breed with "central checked" fibre-type arrays might be more effective than the Merino in suppressing the kemp, derived from the hairy parent. The Wensleydale was recommended, and proved effective in the Wensleydale × (Merino × hair sheep) cross. Crossing apparel-wool breeds with those of very different skin structure may thus be fraught with difficulties. It has certainly been accomplished many times in the past, but it has been a slow process. Many of the current Russian Merino breeds were upgraded from coarse-or carpet-wool types, in a programme begun in the 1920s. In China also there are Merinos that have been similarly developed, while the Indian Hissardale was developed from Merino × Bikaneri, also possibly starting in the 1920s.

Exploiting heterosis

The technique of heterosis offers greater promise for meat production, in terms of more lambs born and higher survival rates, than it does for wool production. With the increased importance of meat production in tropical areas, several projects are in train. One was begun by the Ford Foundation in the Near East in the late 1960s, with the aim of crossing the highly prolific Chios with the Awassi to form a new breed. This project includes a comparison of contemporary F1 and F2 generations, from which a decision can be made about the value of continually forming the first cross.

Two other large-scale projects aimed at increasing meat production are now starting in the Libyan Arab Republic and Saudi Arabia, each involving tens of thousands of sheep. Details are not yet available; it would be interesting to know whether experiments to assess the value of exploiting heterosis are planned.

No work has been reported on estimating heterosis in crosses of some of the tropical hair breeds. For tropical meat production, crosses among the hair breeds seem well worth exploring; an analogy is the valuable level of heterosis exploited in beef-cattle production. Even a cross between the Merino and the Corriedale, itself half Merino, gave 13 percent heterosis in weight of lamb weaned per ewe mated, largely because of the increased survival rate of lambs (Iwan et al., 1971).

Characteristics related to heat adaptation

The low productivity of Merinos in northern Australia has led to a suggestion that tropical breeds might be tested there, when quarantine facilities permit their import. The possibility of developing Merinos adapted to heat stress has also been investigated.

6. Chokla ram. These sheep are noted for their coarse wool and occur in the drier areas of northwest India.

Vermont Merinos became popular in Australia in the late nineteenth century. This breed was characterized by an extensive skin development, the idea being that increased skin area would mean increased wool production. The Vermont sheep came from a temperate area of the United States, situated close to latitude 45°N. Australian research has shown that sheep with heavily wrinkled skin have a lower reproduction rate than plain-bodied animals, and that rams in particular are affected if mated in the hot season. Hot-room trials have indicated that they are less able to control their scrotal temperature than plain-bodied rams, with the result that semen quality is affected.

Work in progress by the Queensland Department of Agriculture at Julia Creek (approximately latitude 21°S, in the dry tropics) indicates that Merino ewes may differ in their ability to maintain temperature under heat stress, and that those able to do so may have a higher reproduction rate.

Protection of the growing fleece

Increasing attention is being paid in Australia to the question of protecting the growing fleece from damage. This problem is also of considerable importance for woolled sheep in the tropics. Hayman (1953) showed that various forms of fleece damage (bacterial stains not removed by scouring, cotting of fibres and even breakage) may occur when the fleece is soaked and the skin remains wet for some time. He demonstrated genetic differences in resistance between groups of sheep, the resistant animals having more wax in the fleece.

Dunlop and Hayman (1958) also found differences among Australian Merino strains in their resistance to fleece rot, the strain with the lowest fibre population density and the most open fleece being the most susceptible when exposed to rain.

Yellowing of the growing fleece is a long-standing problem in India, where in some areas sheep are shorn twice a year to obtain one white and one yellow clip. Some breeds are more prone to yellowing than others. However, although whiteness is desirable in carpet yarns on the world market, pigmented or yellow-stained wool can readily be used for weaving patterned carpets. As the tufted type of manufacture increases, self-coloured carpets will be more common, and white wool will be more in demand than ever. Research on the problem of weatherstaining is therefore of considerable importance.

7. Malpura ram from India, typical of the carpet-wool sheep occurring in the dry tropics.

Documentation and experimentation

FAO's plea for breed documentation (FAO, 1967) cannot be reiterated too often. It applies not only to tropical breeds but to those in temperate areas as well. Australia, for example, knows far too little about the comparative performance under the same conditions of sheep and cattle from its various breeds, strains and studs.

The FAO report also outlines experimental techniques for comparing the performance of indigenous breeds and their crosses with exotic breeds under the same conditions, as well as estimating the extent of heterosis. One can only hope they will be used, and indigenous breeds properly evaluated, before there is any further swamping by exotic breeds developed under different conditions. Introductions may be of value, but they must be tested against existing material and not automatically assumed to be superior because of their performance elsewhere.

One further point must be made. Selection may be slow, but so is the introduction of an exotic breed for upgrading or crossing. Some of the problems have been mentioned. Even without these, complete breed replacement or new breed development will seldom take less than 15 to 20 years, if a relatively small nucleus of exotic animals has to be used over a large existing population.


Bergen, W. von. 1963. Wool handbook. New York, Interscience Publishers.

Brown, G.D. & Williams, O.B. 1970. J. Aust. Inst. agric. Sci., 36: 182–198.

Burns, M. 1967. Trop. Agric., 44: 173– 192.

Carter, H.B. 1955. Anim. Breed Abstr., 23: 106.

Demirüren, A.S. 1972. Wld Anim. Rev. (FAO), 2: 21.

Dunlop, A.A. 1962. Aust. J. agric. Res., 13: 503–531.

Dunlop, A.A. & Hayman, R.H. 1958. Aust. J. agric. Res., 9: 260–266.

fao. 1967. Report of the FAO study group on the evaluation, utilization and conservation of animal genetic resources. Rome.

Hayman, R.H. 1953. Aust. J. agric. Res., 4: 430–468.

Iwan, L.G., Jefferies, B.C. & Turner, Helen Newton. 1971. Aust. J. agric. Res., 22: 521–535.

Krishna Rao, M.V., Schinckel, P.G. & Clarke, W.H. 1960. Aust. J. agric. Res., 11: 97–104.

Lall, H.K. 1956. Breeds of sheep in the Indian Union. Misc. Bull. Indian Counc. Agric. Res. No. 75.

McLeroy, A.B. 1961. Sudan J. Vet. Sci. Anim. Husb., 2: 101–165.

Mason, I.L. & Maule, J.P. 1961. The indigenous livestock of eastern and southern Africa. Edinburgh, Com. Bur. Anim. Breed. Genet. Technical communication No. 14. Farnham Royal, Bucks., England, Commonwealth Agricultural Bureaux.

Narayan, S. 1960. Aust. J. agric. Res., 11: 408–426.

Smith, I.D. & Clarke, W.H. 1972. Aust. J. Exp. Agric. Anim. Husb., 12: 479– 487.

Turner, Helen Newton. 1969. Proceedings, Reproduction Symposium, U.S. Sheep Development Programme, 67–94. (Editor: G. Scott)

Turner, Helen Newton, Dolling, C.H.S. & Kennedy, J.F. 1967. Aust. J. agric. Res., 19: 79–112.

Previous Page Top of Page Next Page