Water Buffalo

Robert Hoffpauir. Cambridge World History of Food. Editor: Kenneth F Kiple & Kriemhild Conee Ornelas. Volume 1. Cambridge, United Kingdom: Cambridge University Press, 2000.

When the first created man saw the animals that God had made, it is said that he presumptuously, over-rating his powers, asked that he too might be given the creative power to fashion others like them. God granted his request and man tried his prentice hand. But the result was the buffalo, and man seeing that it was not good, asked in disgust that the creative power might be taken back again from him for ever.The buffalo, however, remained as the only living handiwork of man. (Bradley-Birt 1910: 115)

Although of limited value as a clue to the origins of the domesticated water buffalo (Bubalus bubalis), this tale from India does reflect the rather low opinion of this bovine held by many, including, it seems, scientists who have shown relatively little interest in it. Considering the large size of its population, its widespread distribution, and its essential role in the economic lives of millions of people, especially in southern and eastern Asia, it is remarkable that so little is known about the water buffalo. Bovine admiration and scholarly attention have been reserved for those more highly regarded distant relatives of the buffalo, the taurine and zebu cattle.

Admittedly, some admirable efforts have been made within the last few decades to remedy this situation. Most of the work that has been done has focused on the present-day conditions and future potential of buffalo husbandry, with breeding, management, and productivity being of central concern. Research on the cultural and historical aspects of the buffalo, however, has been very limited. Certainly, in the discussion of animal domestication, the buffalo has been largely ignored.

This chapter presents the results of a preliminary investigation of the original domestication and history of the water buffalo. A perusal of the available literature reveals a frustrating scarcity of information and very sketchy evidence. Consequently, interpretations must be highly speculative. At this stage in the research definitive answers are well beyond reach, and the best conclusions we can now offer are merely those that might guide future research.


Although there is general agreement that the domesticated water buffalo represents a single species, its relationship to other bovines has been the subject of much disagreement. It received the name Bos bubalis from those naturalists who, beginning with Linnaeus in 1758, believed that bovines were all sufficiently similar to warrant classification under the same genus (Bos). Later taxonomists, using elaborate anatomical criteria, argued that bovine differences should be recognized by subgenera or genera status. In the many multiple genera and subgenera schemes that were proposed, a Bubalus genus or subgenus was always included for the classification of the water buffalo (and usually the other related “buffalo”—the anoa, tamarau, and African buffalo).

The origin of the Latin appellation bubalus (or bubalis) is unknown.The learned ancients of Greece and Rome used the term, but seemingly in reference to an African antelope, most likely the bubal hartebeest (Alcelaphus buselaphus). From Roman through medieval times it was also applied to the European wild bison and the wild auroch (Buffon 1812: 304-8, 316-18; Pliny 1855: 262-3). When a Benedictine monk, Paul the Deacon, recorded the introduction of the domesticated Asian buffalo into Italy during the late sixth century A.D., he used the term bubali (Paul the Deacon 1907: 158-9; White 1974: 203-4). Thereafter, the accepted usage for bubalus was in reference to the Asian water buffalo.

The word “buffalo” was undoubtedly derived from the Latin bubalus. The Old English forms, buffle and buff, probably evolved through the French form buffle, but the present form seems to have entered English from the Portuguese buffalo. After some early European travelers in India erroneously used the word “buffalo” to refer to the zebu, the term “water buffalo” came into use to differentiate the true buffalo (Yule and Burnell 1903: 122; Murray et al. 1933: 1157). The confusion continued—as part of what Lynn White, Jr. (1974: 204), has called “the colonial transmission of antique and medieval perversity to North America”—when English speakers mistakenly used the word “buffalo” for the American bison. Use of the term “water buffalo” has been perpetuated as a means of avoiding this further confusion.

Biological and Ecological Characteristics


The buffalo has a broad and massive body with short muscular legs ending in large splayed hoofs. Its shoulder height averages about 5 feet, and its length from nose to tail averages about 10 feet. The weight for adult buffalo varies from about 600 to 2,000 pounds. Although their massive bodies do not appear particularly graceful, they do reflect the animals’ great strength.

Both sexes carry very distinctive transversely grooved and triangular cross-sectioned horns. These horns normally are crescent-shaped and incline upward, outward, and slightly backward following a single plane extending above and in the same plane as that of the forehead. Many variations in size and shape are found, from the short and tightly curled horns of the Murrah breed of northwestern India to the very long and nearly straight horns of the wild buffalo in Assam.

Shades of slate gray to black are the most common colors for the buffalo’s skin and hair. Frequently the animal has light-colored chevrons under the jaw and on the chest and white or gray stockings. Although buffalo are born with a thick coat of long coarse hair, this hair generally becomes sparse as they age. In temperate or high-altitude locations, however, where cold weather is encountered, a heavy coat of hair is retained in the adult stage. Hair whorls are found in all breeds, and their individual distinctiveness has caused them to be used in some countries as a criterion for legal recognition and registration (Cockrill 1974a).

Adaptation and Habitat

Usually described as a denizen of the tropics, the buffalo is not capable of high heat tolerance. Its heat-regulating mechanisms are less efficient than those of European and Asian cattle, and under conditions of exposure to direct solar radiation, the buffalo exhibits a greater rise in body temperature, pulse, and respiration rates, and general discomfort. The buffalo is quicker, however, to show signs of comfort with a lowering of body temperature as a result of shade, artificial showers, or natural rain. As a means of cooling themselves and ridding themselves of insect pests, buffalo, unlike cattle, show a natural inclination to wallow.

The animal’s preference for shade and semi-aquatic conditions is at least partially explained by the color and structure of its skin. Like most tropical animals, the buffalo has black skin that absorbs ultra-violet light, thus preventing inflammation and damage to the deeper skin layers. But unlike most tropical mammals that have a coat of light-colored hair that reflects infrared waves, the buffalo has only a sparse coat of dark hair on its dark skin. Although this black surface facilitates the absorption of heat in direct sun, it is also advantageous in shade because of its high heat-radiating powers. The buffalo also suffers from the absence of an efficient perspiring mechanism. The number of sweat glands per unit area of buffalo skin is less than one-sixth that of Western cattle and an even lesser proportion of that of zebu.

Thus, the buffalo is well adapted to a tropical environment provided that environment includes sufficient cooling systems in the form of shade and water. In the wild state these conditions are met with an environment of high rainfall, heavy vegetation cover, and streams, lakes, swamps, or marshes for wallowing. Husbandry practices attempt to duplicate the conditions of the animal’s natural habitat as much as possible by providing artificial shade, nearby water bodies for wallowing, and frequent splashing when needed.

The nature of the buffalo’s feet, and its dietary habits, further indicates a semiaquatic, humid/tropical adaptation. The exceptional flexibility of its fet-lock and pastern joints and its large splayed hoofs enable the animal to move easily in mud and water (Cockrill 1967: 124). Buffalo also reveal the remarkable ability to subsist entirely on those crude fibrous plants that form so much of tropical vegetation (especially around water courses), which other livestock will not touch. Experimentation has shown that compared to the zebu (its major economic rival in South Asia), the buffalo will consume more low-quality roughage, digest it more efficiently, and thereby maintain itself in better condition with greater strength and productivity (Whyte 1968: 218; Whyte and Mathur 1974: 555-6).

The “Swamp” and “River” Buffalo Types

Roderick MacGregor (1941) was probably the first to recognize a significant subdivision within the B. bubalis species. Based initially on his observations of differences in bathing habitat preferences, he distinguished the “Swamp buffalo,” which was the native domesticated buffalo of East and Southeast Asia, from the “River buffalo,” which was the type generally found in India and farther west.

A number of differences separate these two types. Swamp buffalo prefer mud-bottomed swamps and marshlands in contrast to the River buffalos’ preference for rivers with firm bottoms and clear water. Swamp buffalo throughout their range seem to be morphologically very similar, with little specialization having occurred. They appear to be much closer to the wild buffalo than are the River buffalo. The latter, in contrast, show a great deal of regional variation, with many specialized breeds having been developed, especially for improved milk yield. The Swamp buffalo tends to have a shorter body, larger girth, shorter and thinner legs, and a shorter face than the River buffalo. Swamp buffalo horns are fairly uniform in size with the common widespread crescent shape; River buffalo manifest great variety in horn shapes, often with a spiraling curve close to the head.

Population, Distribution, and Uses

Before we venture into the distant past in search of clues to the buffalo’s original domestication, it is valuable to view the present-day consequences of that domestication. As the result of many millennia of human-buffalo interaction, the buffalo currently has an extensive geographical range and is involved in a great variety of husbandry contexts. Recent estimates place the total world population of domesticated buffalo at about 142 million. Although these animals are reported from 31 countries, a high degree of concentration characterizes this distribution: Ninety-six percent reside in the countries of South, Southeast, and East Asia, with 80 percent in just the three countries of India, Pakistan, and China. India leads all other countries with 54 percent of the world total (about 77 million buffalo). Relatively small numbers are found in Southwest Asia, southern and Eastern Europe, and South America. The only countries within this peripheral zone that possess fairly significant numbers (over 1 million) are Egypt and Brazil.

East and Southeast Asia

In the area east of India, the Swamp buffalo dominates and is essentially a single-purpose animal, contributing primarily to the traction needs of the agricultural system. It is in the wet rice fields of southern China and Southeast Asia that the buffalo distinguishes itself as the preferred draft animal.The humid climate, the mud, and the water suit the animal’s constitution and habits. No other domesticated animal can equal the buffalo’s ability to carry out the plowing, harrowing, and puddling tasks of wet rice cultivation. Its considerable weight causes it to sink deep in the mud; the flexibility of its fet-lock and pastern joints, its large splayed hoofs, coarse limbs, and bulky body allow it to maintain a balanced traction in mud and water; its enormous strength enables it to pull the plow deep into the soil.

In China, the cattle to buffalo ratio is about 4:1, with the buffalo distribution limited almost entirely to the rice-growing area in the southeast. Cattle are preferred in the drier and more temperate northern and western regions (Phillips, Johnson, and Moyer 1945: 62-6; Epstein 1969: 26-32; Cockrill 1976). But in mainland and island Southeast Asia, where wet rice cultivation dominates the agricultural economies, the buffalo becomes more important. The average cattle to buffalo ratio throughout Southeast Asia is about 2:1. Three countries—Laos, Brunei Darus, and the Philippines—actually possess more buffalo than cattle.

Eastern and southeastern Asia is located beyond the traditional limits of milking and milk use in Asia (Simoons 1970).The peoples in this area generally do not include milk or milk products in their diet and exhibit high incidences of primary adult lactose intolerance. A limited market for buffalo milk, however, has been found in some cities and in the small Indian and European communities scattered throughout the region.

Buffalo are rarely raised specifically for meat production or for any other purpose. Retired work animals, after having provided 10 to 20 years of service, are either slaughtered or allowed to die a natural death, at which time they contribute meat, horns, and hides to the local economy. While alive, the buffalo produces a very important by-product for the farmer—manure, a valuable alternative or addition to scarce and expensive chemical fertilizers.

South Asia

The buffalo is found in its greatest concentration in South Asia and there exhibits its greatest biological diversity and attains its greatest utility as a dairy animal.The distributions of the River and Swamp buffalo meet, and in some places overlap, in South Asia, and, although River buffalo predominate, Swamp buffalo are found in the Assam and Tamil Nadu states of India, and in Bangladesh, Sri Lanka, and Nepal (Porter 1991: 290-1). Selective breeding of River buffalo, primarily in India and Pakistan, has resulted in the formation of several distinct dairy breeds and even a few draft breeds. Probably between 15 and 20 percent of the buffalo in this region can be identified as representatives of defined breeds; the rest are nondescript local varieties (Hoffpauir 1974: 141-51).

As in eastern Asia, the buffalo in South Asia performs an important service as a work animal, especially in the wet rice fields, but this role is subordinate to that of milk production. In India, for example, only about 15 percent of all buffalo are classified as work animals, and they account for only about 10 percent of the total bovine workforce.The more versatile zebu bullock is the generally preferred work animal.

Even though the zebu outnumbers the buffalo in India by about 2.6 to 1, and buffalo account for only about one-third of the bovine population that is milked, over 60 percent of the milk produced comes from buffalo (Hoffpauir 1982: 223). In Pakistan and Nepal, fully two-thirds of the milk produced comes from buffalo (Cockrill 1974b: 605; Khan 1974: 612). Nature (presumably aided by selective breeding) has provided the buffalo with certain milk-producing abilities that are distinctly superior to those of the zebu—namely, the ability to yield more milk (usually between two and three times more), the ability to produce milk with higher butterfat content (between 6.5 and 8.5 percent, compared with 3.0 to 5.0 percent in zebu milk), and the ability to produce this milk while subsisting on poor-quality fodder (Hoff-pauir 1977).

In Hindu India the slaughtering of the buffalo and the eating of its flesh is not generally practiced.This is probably explained by the general ahimsa (the Hindu/Jain/Buddhist principle of noninjury to living things) and vegetarian proclivities of the Indian culture, rather than by any special status similar to that of the sacred zebu. The Moslems in Pakistan and Bangladesh have no prohibition on the slaughtering of buffalo, but buffalo meat is unpopular because of the poor quality that is marketed. Most Pakistanis and Bangladeshis prefer beef (Cockrill 1974b: 535; Khan 1974: 611).

The population of buffalo in Bangladesh is far less than might be expected, given the predominance of wet rice cultivation throughout this Ganges-Brahmaputra delta area. In fact, it appears that the disasters this countr y has suffered in recent decades (typhoons, floods, earthquakes, and warfare) have decimated the buffalo population. Most of the buffalo still there are of the Swamp type and are used mainly for work. Small numbers of dairy buffalo have been imported, but milk production remains low (Cockrill 1974b: 533-5).

The adaptability of the buffalo to high altitudes is clearly demonstrated in Nepal, where this tropical animal is maintained in village settings as high as 9,000 feet and is taken for summer pasturage as high as 15,000 feet. In Nepal, the buffalo is truly a multipur-pose animal: It provides work power, milk, manure, and meat. In the tropical lowlands of the terai (the northern extension of the Indo-Gangetic plain in southern Nepal), it is the preferred draft animal for wet rice cultivation; it is seldom used for work elsewhere. In the lower and middle altitudes, the buffalo is the primary source of milk. Buffalo have been introduced into Nepal from India for the purpose of improving milk production. Whether the native Nepalese buffalo are of the Swamp or River type has yet to be determined. In many mountainous locations, the buffalo is used exclusively for the production of manure, an extremely valuable resource given the low fertility of the mountain soils. Nepalese law permits the killing of male buffalo, and considerable numbers are slaughtered for ritual sacrifice and for the urban meat market (Rouse 1970, 2: 905-6; Cockrill 1974b: 603-9; Epstein 1977: 38-46; Hoffpauir 1978: 238-41).

In many ways Sri Lanka follows the typical Southeast Asian pattern of buffalo use. Its indigenous buffalo is of the Swamp type and serves primarily as a work animal in the rice-growing areas. Some buffalo milk is produced and sold in the form of curd, but generally buffalo owners are not interested in milk production. Frederick Simoons (1970: 561-3) has commented on the survival of a nonmilking attitude among the Sinhalese, the dominant group in Sri Lanka. With a population that is predominantly Buddhist and Hindu, Sri Lanka has a limited demand for meat. Only old, sick, or infertile buffalo are allowed by law to be slaughtered; so the buffalo meat that reaches the markets tends to be tough and without taste (Cockrill 1974b: 629-35; Porter 1991: 291).

Southwest Asia and Egypt

There are few environments in the dry southwestern end of Asia and northern Africa that are suitable for the buffalo. Consequently, those in Iran, Iraq, Syria, Turkey, and Egypt account for only about 2.4 percent of the world’s total. The only country in this region where buffalo make a significant economic contribution is Egypt. For the Southwest Asian countries, the cattle to buffalo ratio is 26:1; for Egypt it is 1.4:1.

All of the buffalo in this region appear to be of River buffalo ancestry, having been introduced from South Asia during the first millennium A.D.According to historical records, Sassanians brought domesticated buffalo and their herders from Sind in Pakistan to the marshes of the lower Tigris and Euphrates Delta during the fifth century A.D. Buffalo were established in the Jordan Valley by the eighth century and in Anatolia by the ninth century (White 1974: 204-5). They were not known in ancient Egypt and do not seem to have appeared in the Nile Valley until about the ninth century A.D. (Epstein 1971: 567-8; Cockrill 1984: 57).

The 2.5 million buffalo in the Nile Valley of Egypt today are fully utilized for milk, work, and meat. Egyptians have thoroughly researched their productivity and have developed an efficient husbandry system. Buffalo are highly valued as milk animals and produce far more milk than cattle. Although cattle are more important than buffalo for draft purposes, the latter are used for plowing and harrowing in rice fields, for raising water from wells for irrigation, and for threshing. John E. Rouse (1970: 621-2) reported that only buffalo cows are used for work.The common practice is for male calves to be slaughtered at a very young age for veal, which brings a high market price. Female buffalo are also slaughtered for their meat, but only if they are old, infertile, or injured (Epstein I: 1971: 564; El-Itriby 1974; Porter 1991: 299-300).

Not surprisingly, given the water needs of this bovine, the major concentrations of buffalo in Southwest Asia are in swampy areas and along coasts. The buffalo in Southwest Asia is esteemed only by small numbers of peasant farmers and herders who recognize the animal’s unique adaptability to environments that are unable to support other livestock. Throughout this area the buffalo is valued primarily as a milk animal and used very little for work, except in Turkey, where in some areas buffalo are employed in plowing and road haulage. Buffalo meat is merely a by-product from old retired animals (Rouse 1970, 2: 794, 804, 854; Cockrill 1974b: 510-32; Porter 1991: 297, 298).

Europe and Transcaucasia

The existence of the buffalo in the temperate latitudes of Italy, Greece, the Balkan peninsula, and Transcaucasia in the former Soviet Union attests to the animal’s ability to tolerate cold winter temperatures. But their numbers are small—the total population being only about 775,000—and care must be taken to keep these animals warm during the winter; cold winds and sudden drops in temperature can cause fatal illnesses, including pneumonia (National Research Council 1981: 45-6).

All European and Transcaucasian buffalo are of the River type. Historical accounts are sketchy but indicate that domesticated buffalo may have entered Europe by a route north of the Black Sea, from Persia to southern Russia to the Danube Valley and, finally, to Italy by the end of the sixth century A.D. Another likely route could have been from Turkey to Greece, the Balkans, and Italy. They were established in Transylvania by the eleventh century and were numerous in the area around Rome by the twelfth century (Bökönyi 1974: 151;White 1974: 203-6).

The discontinuous and spotty distribution pattern that characterizes the buffalo’s existence in Southwest Asia is also found in Europe. In Greece, for example, 90 percent of the buffalo are found in Macedonia and Thrace. In Romania there are two concentrations, one in the central regions near Cluj and Fagaras,and another in the Danube Valley of the south.The Italian buffalo are found mainly in the southern part of the country. The only buffalo in the former Soviet Union are in the Transcaucasus region, where they are most numerous in the Kura Valley and along the Caspian coast in Azerbaijan.

Throughout most of its European and Transcaucasian distribution, the buffalo has served as a triple-purpose animal. Its use as a work animal, originally of primary importance, has declined in recent times with the mechanization of agriculture, especially in Italy, Greece, and Bulgaria. Meat production is growing in importance, although buffalo meat still comes mostly from old animals whose usefulness for work and milk production has been exhausted.The buffalo is highly valued for its rich milk, which is often made into yoghurt and cheese. In Italy, where probably the best buffalo in Europe are found, the emphasis has long been on milk production, with the highly popular mozzarella cheese the major product (Cockrill 1974b: 708, 731-5, 748-54; Polikhronov 1974; Salerno 1974; Porter 1991: 297-9).

Tropical America

Introductions of buffalo into various Caribbean and South American countries during the past 100 years have met with mixed results. Those into Bolivia, Colombia, French Guiana, Guyana, Peru, and Venezuela were apparently unsuccessful, and either the animals have entirely disappeared or their numbers have become too insignificant to be reported in recent censuses. Possibly a thousand buffalo can still be found in Surinam, derived from an 1895 introduction of Swamp buffalo from Southeast Asia, via French Guiana, to work on the sugar plantations.They are used today in logging operations and for meat production. Indian River buffalo were brought into Trinidad and Tobago during the early 1900s, also for the sugar plantations.A sizable population of about 9,000 animals has developed, and they are highly valued for their work power and meat (Cockrill 1974b: 676, 692-7, 705-7; Mahadevan 1974; Porter 1991: 300-2).

The most successful establishment of buffalo husbandry in the Western Hemisphere has been in Brazil. Numerous introductions of both Swamp buffalo (from eastern and southern Asia) and River buffalo (from India and Italy) have occurred over the last century. The nearly 2.5 million Brazilian buffalo today are concentrated in two areas, with about 75 percent in the Amazon Basin and the rest in and around the states of São Paulo and Minas Gerais. The buffalo are used primarily for milk in the latter areas, with meat production secondary although growing in importance. In the Amazon area the buffalo has found a paradise, with ideal environmental conditions. On Marajó Island, large herds of free-ranging buffalo thrive on the tropical grasses and aquatic plants. Buffalo meat and milk command a ready market in the towns of Amazonia, and the animal is also used for road haulage and riding (Gade 1970; Cockrill 1974b: 677-91; Porter 1991: 301).

The Wild Ancestor

Unlike the progenitors of many other domesticated animals, the wild form from which the domesticated buffalo originated is not only known but still survives in a few locations.Wild buffalo differ from the domesticated form chiefly in their greater size, larger horns, and generally more robust and sleek appearance.

The Plio-Pleistocene Buffalo

The evolutionary history of the wild buffalo appears to have begun about 3.5 million years ago in the northwestern corner of the Indian subcontinent. The earliest fossil to show uniquely Bubalus characteristics is Proamphibos from the upper Dhokpathan Formation (Middle Pliocene) of the Siwalik Hills (Groves 1981: 270, 276).

Within South Asia, when the conditions in the northwest became unfavorable on account of Pleistocene glaciation, the Siwalik fauna, including the buffalo, apparently migrated southward and eastward. The Middle Pleistocene form Bubalus palaeindicus has been discovered in the Potwar Plain of northern Pakistan as well as in the Narbada Valley of peninsular India. By the Upper Pleistocene the buffalo was found as far south as Tamil Nadu and as far east as Bihar. As revealed by bone finds and cave paintings over a wide range of the Indian sub-continent, the buffalo was undoubtedly hunted by Paleolithic and Mesolithic peoples throughout the Pleistocene and into the Holocene.

The buffalo appeared in China as early as the Middle Pleistocene (possibly between 1 million and 500,000 years ago), and its remains are often found associated with early hominids/humans in Paleolithic sites.Although most Bubalus finds from Chinese sites are reported as being of indeterminate species, some distinct species have been identified. A dozen Pleistocene sites in South China (south of the Qin Ling Mountains) have yielded buffalo remains.

The Pleistocene buffalo that appear in North China are interpreted as immigrants from the south, and their existence as far north as 40° north latitude attests to their adaptability to cooler and drier environments. The buffalo apparently extended its range eastward from North China into Korea during the Middle Pleistocene. Bubalus has been uncovered at two cave sites, Sangwon Komunmoru and Tokchon.As in North China, this presence appears to represent a remarkable adaptation to what must have been a rather harsh Pleistocene environment.

Buffalo ventured out onto the land masses of the Sunda shelf probably as early as they reached China. Although no Pleistocene buffalo have been firmly identified as yet from mainland Southeast Asia, they must have existed there, or at least passed through on their way to Sundaland. A very large form of a now extinct buffalo Bubalus paleo kerabau in habited Java from the early Middle Pleistocene to the early Holocene.The only other indication of the buffalo in island Southeast Asia comes from the discovery of Bubalus in association with the Liwanian Flake Tool Tradition in the Cagayan Valley on the Philippine island of Luzon. Robert Fox (1978: 82, 1979: 229-30) has suggested that Paleolithic hunters, along with the buffalo and other typical Middle Pleistocene mammals, migrated from South China into the Philippines across the then-existing land connections via Taiwan.

A long-distance westward migration of Pleistocene buffalo is evidenced by the finding of Bubalus at three sites in Europe (not shown on Map II.G.23.2). A form called Bubalus murrensis is known from the two sites at Steinheim and Schönebeck in Germany, both dated from the Mindel-Riss interglacial, about 250,000 years ago, and Riss-Würm interglacial site of possibly 100,000 years ago in the Peneios Valley of Greece. The buffalo seem to have been able to exist in Europe only during the interglacial periods and had disappeared by the beginning of the last (Würm) glaciation (Zeuner 1963: 246; Kurtén 1968: 187; Bökönyi 1974: 149-50).

Holocene Survivals

The wild buffalo probably maintained a broad range throughout southern, eastern, and southeastern Asia well into the Holocene. Evidence of B. bubalis has been reported from many archaeological sites in this area, dating from between 10,000 and 2,000 years ago. But whether these findings reflect locations before or after domestication is, in many cases, difficult to say because of the morphological similarity between wild and domesticated buffalo. (The significance of these Neolithic and early Metal Age findings to the question of earliest domestication is discussed later in this chapter.) Consequently, we do not have a clear picture of the distribution of the wild buffalo several thousand years ago when it probably was first domesticated.

Reported sightings during the last two centuries indicate that the wild buffalo’s range has generally retreated to east of the 80° meridian in southern and southeastern Asia. In South Asia, the wild buffalo are known in the terai of Nepal, the plains of Bengal and Assam, and lowland areas in Orissa, Madhya Pradesh, and Andhra Pradesh.They are described as generally inhabiting moist lowland riverine tracts where dense and high vegetation cover, usually tall grass jungles, and nearby swamps are available.

The distribution of the wild form of a domesticated species has often proved a valuable indicator of that animal’s original place of domestication.This, unfortunately, is not the case with the buffalo.All that can be concluded from the available evidence is that the wild buffalo’s distribution has responded to changing environmental circumstances that have occurred during and since the Pleistocene.As conditions under which the buffalo can exist have appeared and disappeared, its range has changed accordingly. Generally, the animal’s extremely broad range during the Pleistocene, especially during the interglacial stages, has been shrinking throughout the Holocene. Its recent limited distribution is best seen as merely its last remaining refuge area and not necessarily as indicative of an ancestral hearth of domestication.


Archaeological Evidence

The most convincing and satisfying evidence for the place and time of domestication of an animal species would be well-preserved fossilized bones that can be clearly identified as to species and domesticated status. Unfortunately, such evidence is relatively rare for the buffalo because the warm and moist environments in which buffalo usually exist make the chances of good skeletal preservation poor. Even when B. bubalis bones can be identified, as they have in many Neolithic and Metal Age sites, the ability to distinguish between wild and domesticated forms is hindered by the absence of clear morphological differences. Consequently, fossil evidence must be very cautiously interpreted.

South Asia. The earliest indication of a possibly domesticated buffalo in South Asia comes from the Indus Valley civilization (about 2300-1750 B.C.). The evidence, collected from the two sites of Mohenjo-daro and Harappa, consists of only a few fossils (teeth, horn cores, and bones), depictions on a small number of seal-amulets, and two small figurines. Many authorities have accepted the notion that the Indus Valley buffalo was domesticated, and some have even suggested that these finds represent the original domestication. Such a conclusion, however, seems to be based solely on the fossils’ structural resemblance to the domesticated form and on the fact that a few seals show the buffalo standing in front of an object that has been interpreted as a feeding trough.

Evidence that the buffalo existed in the Indus Valley only in the wild state is far more convincing.There are four scenes from seals and prism-amulets that seem to depict the buffalo being hunted. In three scenes a man is shown with a foot on the animal’s head, one hand grasping a horn and the other hand holding a spear about to be thrust into the beast. In another scene the buffalo appears to have attacked a number of people who are sprawled around the ground.

When the sex of the animal can be detected, it is always male. As with the depictions of all other bovines, the Indus Valley seals seem to emphasize the power and strength of the bulls rather than the services derived from them or the cows. There are no scenes of the buffalo being milked or pulling a plow or cart, and the so-called food trough also appears on other seals in front of tigers and rhinoceroses. In fact, the troughs are not shown with those animals that were most likely domesticated, such as sheep, goats, elephants, and zebu. Sir John Marshall (1931, 1: 70), the original excavator of Mohenjo-daro, suggested that the troughs represent offerings of food to worshiped animals, both in the wild and in captivity. Others have interpreted the troughs as cult objects or symbols connected to sacrificial ritual (Hiltebeitel 1978: 779).

Interpreting religious content from artifactual remains is, at best, a dangerous endeavor. With the Indus Valley culture in mind, Sir Mortimer Wheeler (1968: 108) has reminded us of “the notorious incapacity of material symbols to represent the true content and affinity of a religion or belief” and of “the indivisibility of religious and secular concepts in ancient times.” Much of the attempt to decipher the religion of the Indus Valley people has focused on the interpretation of one particular seal scene, called the “Proto-Siva” scene, in which the buffalo plays a prominent role. The significance of the buffalo and other animals shown around the seated godlike figure and the meaning of the buffalo-horn headdress worn by the seated figure are but two of the many concerns scholars have had about this picture. Some writers have assigned an important role to the buffalo in the religious symbolism and ritual life of these people (Sastri 1957: 6-13; Sullivan 1964; Srinivasan 1975-6; Hiltebeitel 1978). If true, the implications could be profound for the question of domestication, suggesting, for example, a ceremonial motivation.

Within the riparian environment of the Indus River and its tributaries, the wild buffalo could have found the marshy and grassy habitats that it prefers. The safest conclusion that can be reached at this point is that the wild buffalo existed in the Indus Valley during the third millennium B.C.; it was possibly hunted for its meat and may even have played some role in the religious life of the people. But the case for the existence of the domesticated buffalo is weak and unconvincing.

The buffalo appears in later archaeological sites in India, but the domesticated status continues to be indeterminable. Small numbers of buffalo bones have been uncovered in Neolithic and Copper Age sites, dated from the second and first millennia B.C., south and east of the Indus Valley. But the majority of the bones at these sites belong to the domesticated zebu (Bos indicus), and the number of buffalo bones is very small, sometimes no more than a single bone at a site. Some of the bones are charred and have cut marks on them, a clear indication that the animals were cooked and eaten. An occasional successful hunt of the wild buffalo could easily explain these remains.

The north Indian site of Hastinapura (1100 B.C. to A.D. 300) is said to reflect the expansion of the Indo-Aryan peoples from their original homeland in the Punjab region into the Ganges Valley of Uttar Pradesh. Although the fossil evidence (12 buffalo bones, some charred and with cut marks) is no more convincing than at other sites, it is probably safe to say that these were domesticated buffalo.

We can draw this conclusion because, although archaeological evidence fails us, literary evidence, at least for early historic northern India, throws some light on the question.The oldest literary work, the Rig Veda, which is believed to reflect Indo-Aryan culture as early as 1500 B.C., makes reference to “buffaloes yoked in fours,” the eating of buffalo flesh by the gods, and the slaughtering of from 100 to 300 buffalo at a time as sacrificial offerings to the god Indra (Griffith 1963, 1: 489, 575, 2: 123, 133, 226).

The Yajur Veda, dating from the early part of the first millennium B.C., mentions buffalo “reared for purposes of cultivation” (Chand 1959: 234). Although no earlier references exist, it appears that at least by the time of the writing of the Sutras (about 800-300 B.C.), the use of the buffalo for milk was well established (Bühler 1882: 73, 1896: 63; Jolly 1900: 166, 167).

This record probably represents the earliest use of the buffalo as a dairy animal anywhere. Although literal interpretations of this ancient literature are not always possible, it would seem safe at this point to conclude that the buffalo was a domesticated animal in northern India at least 3,000 years ago. For the wetter eastern parts of India, where environmental conditions suit the buffalo particularly well, archaeological and literary evidence is notably scarce. Consequently, the antiquity of the domesticated buffalo in eastern India cannot, as yet, be clearly shown.

Southwest Asia. Mesopotamia presents us with a situation very similar to that of the Indus Valley. The buffalo was undoubtedly well known in ancient Mesopotamia, but its domesticated status is disputed. One site in northern Iraq (Grai Resh), dating from the Uruk culture period (about 3500 B.C.), has yielded a horn core of a buffalo. Most of the remaining evidence is representational, coming from depictions on cylinder seals from the third millennium B.C.Arguments in favor of the buffalo’s domesticated status rely mainly on scenes showing the animal in the company of humans (or possibly gods or mythical heroes), in some cases being watered and fed by the human figures. Some supporters of this argument propose that the buffalo was actually domesticated by the Sumerians or Akkadians, and others suggest that the animal was imported from the Indus Valley already in a domesticated state. These seal scenes might well reflect, as suggested by Henri Frankfort (1939: 85-94), religious and mythical symbolism, or maybe just artistic imagination or descriptive design. Wild buffalo probably existed at that time in the swamps of the Tigris and Euphrates Delta and could easily have inspired the seal depictions (Duerst 1908: 360-2; Ward 1910: 414-5; Bodenheimer 1960:50, 102;White 1974:202).

Whether wild or domesticated, buffalo seem to have disappeared from Mesopotamia possibly by the end of the first millennium B.C., although they may have survived longer in other parts of Southwest Asia. Frederick Bliss (1894) reported that fossil water buffalo teeth, dated between 1700 and 500 B.C., were found at Tell el Hesy, near Gaza in ancient Palestine. Buffalo bones were also found in post-Hittite layers of Boghazköy in north central Turkey, dated between the twelfth and seventh centuries B.C. (Bökönyi 1974: 151). A Sassanian seal depiction of a buffalo (illustrated in Bivar 1969: 83, plate 15 EM1) indicates that the buffalo still existed (or at least was still remembered by artists) in northern Iraq or Iran during the first half of the first millennium A.D.

China. Because of the close association of the domesticated buffalo with wet rice cultivation throughout southern and eastern Asia, some authors have suggested the possibility that the buffalo was originally domesticated as a work animal in the rice-growing area. They usually point to southern China or the mainland of southeastern Asia (Zeuner 1963: 251; Bökönyi 1974: 150-1; Clutton-Brock 1981: 140). Until recently there was no archaeological evidence to support this contention. But that has all changed with the astonishing archaeological finds in China during and since the 1970s.

Between about 6,500 and 4,500 years ago, two closely related Neolithic cultures coexisted near the mouth of the Yangtze River; the Ma-chia-pang culture was located north of Hang-chou Bay, and the Ho-mu-tu culture was located to its south. This was a “subtropical, fresh-water wetlands environment of the coastal plain” that was “crisscrossed by rivers and streams and dotted with large and small lakes, ponds, and freshwater marshes”(Smith 1995: 124).

Large quantities of refuse reveal that these people were rice cultivators who probably kept domesticated pigs, dogs, and buffalo. Although the buffalo bones found there are no easier to distinguish as domesticated than at other sites, the unusually large number of bones does indicate a heavily exploited resource. Buffalo shoulder blades were lashed to wooden handles and used as spades, which were the main cultivation tools; no plows were used. Since these people evidently were skilled hunters and fishermen, it could be reasoned that large numbers of wild buffalo existed in the area and were intensively hunted.

If, however, the buffalo was domesticated, which seems to be the unanimous opinion of the excavators and analysts, it undoubtedly was kept for its meat (and shoulder blades); other uses are impossible to determine at this time. It is possible that even without the plow the buffalo might have been used to prepare the rice fields. As is practiced by many rice farmers in South China and Southeast Asia today, the soil can be thoroughly puddled merely by driving one or more buffalo around in the flooded field (Ho 1975: 72). However, the finding of buffalo in association with cultivated rice does not necessarily imply that the animals were actually used in the cultivation process.

Radiocarbon dates place these buffalo in the fifth millennium B.C.Thus, if domesticated, they are the earliest ones known anywhere in the world (Pearson 1983; Chang 1986: 196-201, 208-11; Zhao and Wu 1986-7; Smith 1995: 124-7). Domesticated rice and buffalo have also been found as part of the Liang-chu culture (4000-2000 B.C.), which, although known since the 1930s, is now considered an outgrowth of the Ma-chia-pang culture in the same Yangtze Delta area.

The next appearance of the buffalo in a Neolithic context is in northern China at sites dated from the third millennium B.C.The earlier Neolithic cultures of North China (between 5500 and 2500 B.C.) appear to have been based on the hoe cultivation of millets and the keeping of domesticated pigs, dogs, and chickens. Although a few bovine bones have been found, their identity and domesticated status are uncertain (Ho 1975: 91-5; Smith 1995: 133-40). Bovine remains, claimed as domesticated, are more frequently encountered at later sites, with most identified as Bos sp. and only a few as Bubalus sp. (from Shaanxi Province).

By early historic times in northern China (second millennium B.C.), the domesticated buffalo seems to have been well established, as the remains from the Shang Dynasty sites near An-yang suggest. Bones identified as belonging to the species Bubalus mephistopheles were found in great abundance. Authorities believe that buffalo were able to exist this far north because the climate, as surmised from paleontological and palynological evidence and oracle bone inscriptions, was warmer and wetter than it is today. There seems to be little doubt that these buffalo were fully domesticated, along with cattle, pigs, dogs, sheep, and goats.

Southeast Asia. Archaeology has not been able to offer much support for the rather popular assertion that peninsular Southeast Asia was the place of origin for the domestication of rice and, by association, the buffalo. The oldest presumably domesticated rice in Southeast Asia (from the Non Nok Tha and Ban Chiang sites in Thailand) dates only from the fourth millennium B.C., which makes it at least 1,000 (and possibly 3,000) years younger than the earliest Chinese rice. Demonstration of a great antiquity for the domesticated buffalo has also eluded archaeologists.

One early site in northern Vietnam has yielded buffalo remains dating from about 8000 B.C. Vietnamese archaeologists have suggested that this site, Bolum, and the many others belonging to the Bacsonian culture could be associated with the beginnings of agriculture (fruit trees, tubers, leguminous plants, and hill rice) and animal husbandry (dog and buffalo). But this claim appears to be based solely on the assumption that polished stone tools and pottery, which are found at the Bacsonian sites, imply an early Neolithic agricultural society. It is with greater confidence that we can approach the assertion that wet rice cultivation with buffalo-drawn plows existed in northern Vietnam by the time the Bronze/Iron Age Dong Son culture appears (600-400 B.C.).This claim is based on the finding of rice remains and bronze plowshares (Bellwood 1985: 273-5).

The most significant Southeast Asian evidence relating to the buffalo comes from the site of Ban Chiang in the Sakon Nakhon Basin of northeastern Thailand. An analysis of the faunal remains by Charles Higham and Amphan Kijngam (1979, 1985) has shown that the buffalo made its appearance here about 1600 B.C., coinciding with the first appearance of iron metallurgy. Remains of rice have been found throughout the site’s occupation period, going back to about 3500 B.C.

Based on the size of the buffalo bones, which are considerably smaller than those of wild buffalo, and on the fact that the buffalo phalanges showed stress areas (indicative of plowing) similar to those found on modern work buffalo, the authors concluded not only that the Ban Chiang buffalo was domesticated but that it was used for plowing. No prehistoric plows have been found at this or any other site in Thailand. Higham and Kijngam reasoned that before 1600 B.C. the people of Ban Chiang were probably collecting wild rice and practicing swidden rice cultivation.With the introduction of the buffalo and iron, the agricultural system changed to an inundation system of wet rice farming. Two other sites in the Sakon Nakhon Basin show similar patterns, with buffalo absent from the earlier levels and appearing only in the later periods, apparently about the same time (during the second millennium B.C.) as they do in Ban Chiang.

Ethnographic Evidence

When archaeological findings are unsatisfyingly fragmentary, as is certainly the case with the buffalo, we search for any other possible indicators of the prehistoric past.The use of ethnographic evidence is based on the shaky assumption that primitive people living today or in the recent past can provide us with a glimpse, via extrapolation, of ways of life that may have existed long ago. Drawing conclusions from such evidence must be done with great caution. Sometimes several different, yet equally reasonable, scenarios can be postulated.

A cursory investigation of the position of the buffalo among the primitive tribal societies of southern and eastern Asia reveals a surprisingly widespread pattern of use that might suggest an early motive for keeping and domesticating the buffalo. Most of these tribes inhabit hilly regions, practice the plowless shifting cultivation of dry rice, and do not drink milk. Consequently, they usually do not keep domesticated animals for milk production, plowing, or other work.

Yet in spite of the absence of any of these economic uses, many groups do keep domesticated bovines. These animals are slaughtered and their meat eaten, but the motive is not simply the secular dietary desire for meat, which is met by the hunting of wildlife in the nearby forests. The major purpose for keeping these animals is for sacrifice, and their meat is eaten only at ceremonial feasts. Within the southern and eastern Asian realm, the buffalo is the bovine most frequently used for this purpose. Only a few tribal groups prefer other domesticated bovines as sacrificial victims, such as the zebu among a few tribes in India and the mithan among the groups inhabiting the eastern Himalayas and the India-Burma border region. Tribes practicing buffalo sacrifice are scattered throughout the Indian subcontinent, across mainland Southeast Asia, in parts of southern China, and in the hilly interiors of the islands of the Philippines.All of the occasions for buffalo sacrifices are broadly related, in one way or another, to the desire to enhance or maintain the well-being of an individual or community—fertility in its broadest sense.

Economic or Religious Motivation

Given the obviously useful contributions of domesticated bovines to the economic life of humans, it is not surprising that most attempts to explain their origin seek rational and practical motives. Eduard Hahn (1896) was one of the earliest scholars to propose a thesis on cattle domestication that did not conform to the principles of materialistic rationalism. He was the first in a line of speculators who argued that cattle were originally domesticated not for milk, meat, or traction, but to provide sacrificial victims for the worshiped deities.

Those who followed with sympathetic support, similar premises, and further research included, most notably, Carl Sauer (1952: 88-94), Erich Isaac (1962), and Frederick and Elizabeth Simoons (1968: 234-58). These scholars have pointed out the widespread Eurasian distribution and antiquity of a complex of traits that link bovines, sacrificial practices, and fertility rites. Simoons and Simoons (1968: 261-2) postulated that taurine cattle were the first to be domesticated in this ceremonial context, followed by the buffalo in imitation of cattle, and then by the mithan in imitation of the buffalo.

In the case of the domestication of the water buffalo, we have very little concrete evidence and can offer only speculations that seem to accommodate the fragmented information we have. Nature has provided a massively built and formidable bovine that is ideally suited to tropical and swampy lowland environments. Paleontological evidence places the ancestor of this creature in southern and eastern Asia.Wild buffalo have seemingly adapted to a wide range of habitats as their distribution has expanded and contracted throughout the course of the Pleistocene and Holocene. Until the next discovery, currently available archaeological evidence points to southern China during the fifth millennium B.C. as the most probable place and time for the earliest domestication of the buffalo.

With the invention of the plow, the buffalo’s contribution to the wet rice cultivation process was fully realized. Thus far, the earliest indication of the buffalo-plow-wet rice complex comes from northern Thailand about 3,600 years ago at the Ban Chiang site. It is postulated that this complex diffused into the Indo-Malaysian archipelago roughly 1,000 years later (Bellwood 1985: 205, 241-4). This line of thinking, which focuses on the evolving economic benefits accrued from the domesticated buffalo, would fulfill the requirements of materialistic rationalism. If ethnography can be transposed to the Neolithic, an alternative scenario can also be envisioned. Buffalo might have been initially tended for their religious and social value, rather than their economic value.A belief that the sacrificing of buffalo would be repaid with fertility and well-being could have motivated people to begin domestication.To conform with the archaeological evidence, it could be hypothesized that this buffalo sacrificing fertility cult initially developed in South China. Buffalo sacrificing has been reported among the early Thai (Eberhard 1968: 183-93, 216) and the traditional Miao peoples (Mickey 1947: 78-80; Beauclair 1956: 29-31) of that area.This sacrificial complex could have diffused southward into mainland Southeast Asia and from there westward into India.

By historic times, both buffalo systems—the primitive hill farming system with the sacrificial buffalo and the sophisticated lowland rice paddy system with the plow buffalo—were well established and widely distributed in southern and eastern Asia. Archaeology has not, as yet, been able to sort out the chronological relationships between the two systems. The suggestion that the sacrificial buffalo system is antecedent to the plow buffalo system is based primarily on ethno-historical traditions that commonly identify the hill peoples of southern and eastern Asia as being the indigenous inhabitants and indicate that settlement and cultivation of the lowland plains came later.Without evidence to the contrary, we further assume that the recent cultural practices of these tribal peoples represent survivals from a much earlier time that have been preserved by the groups’ relative isolation. Obviously, caution must accompany our speculation.

The popular archetypal image that many Westerners have of the water buffalo is as the lumbering beast of burden patiently pulling a plow and a frail Asian farmer through a flooded rice paddy. Although this may be an accurate depiction of the animal’s economic value to humans, it may not reflect the original value conferred upon the water buffalo when it was first brought under domestication. We must consider an additional image of the buffalo as a powerful, virile symbol of fertility that is tied to a sacrificial post and about to be ceremonially slaughtered as an offering to the supernatural.

Biological, archaeological, and ethnographic evidence provides us with only a few fragments with which to construct images. As we try to fit the pieces together, we hope that the images that emerge, as preliminary and speculative as they may be, will be the result of objective, multidimensional analysis that has considered economic, religious, and all other relevant factors.