Patricia J O’Brien. Cambridge World History of Food. Editor: Kenneth F Kiple & Kriemhild Conee Ornelas. Volume 1. Cambridge, UK: Cambridge University Press, 2000.
The sweet potato (Ipomoea batatas, Lam.) and the yams (genus Dioscorea) are root crops that today nurture millions of people within the world’s tropics. Moreover, they are plants whose origin and dispersals may help in an understanding of how humans manipulated and changed specific types of plants to bring them under cultivation. Finally, these cultivars are important as case studies in the diffusion of plant species as they moved around the world through contacts between different human populations.
This chapter reviews the questions surrounding the early dispersals of these plants, in the case of the sweet potato from the New World to the Old, and in the case of yams their transfers within the Old World. In so doing, the sweet potato’s spread into Polynesia before European contact is documented, and the issue of its penetration into Melanesia (possibly in pre-Columbian times) and introduction into New Guinea is explored. Finally, the post-Columbian spread of the sweet potato into North America, China, Japan, India, Southeast Asia, and Africa is covered. In addition, a discussion of the domestication and antiquity of two groups of yams, West African and Southeast Asian, is presented, and the spread of these plants is examined, especially the transfer of Southeast Asian varieties into Africa.
The evidence presented in this chapter can be viewed fundamentally as primary and secondary. Primary evidence consists of physical plant remains in the form of charred tubers, seeds, pollen, phytoliths, or chemical residuals. Secondary evidence, which is always significantly weaker, involves the use of historical documents (dependent on the reliability of the observer), historical linguistics (often impossible to date), stylistically dated pictorial representations (subject to ambiguities of abstract representation), remanent terracing, ditches or irrigation systems (we cannot know which plants were grown), tools (not plant specific), and the modern distribution of these plants and their wild relatives (whose antiquity is unknown).
The Sweet Potato
In general, studies of the origin of domesticated plants have first attempted to establish genetic relationships between these plants and some wild ancestor. In the case of the sweet potato, all evidence employed by previous archaeological, linguistic, and historical investigators establishes its origins in the New World.
The remains of tubers excavated from a number of archaeological sites in Peru provide the most persuasive evidence for this conclusion. The oldest evidence discovered to date is from the central coast region at Chilca Canyon where excavated caves, called Tres Ventanas, yielded remains of potato (Solanum sp.), of jicama (Achirhizus tuberosus), and of sweet potato (Ipomoea batatas) (Engel 1970: 56).The tubers were recovered from all levels, including a level in Cave 1 dated to around 8080 B.C.
These plants were identified by Douglas Yen who could not determine if they were wild or cultivated species, although he observed that wild tuber-bearing sweet potatoes today are unknown (Yen 1976: 43). Whether they ever existed is another matter, but at least the findings in these cases suggest the consumption of “wild” sweet potato at 8000 B.C. in Peru, or more radically, a domesticated variety (Hawkes 1989: 488). If the latter is the case, sweet potatoes would be very ancient in the New World, raising the possibility that perhaps sweet potatoes were the earliest major crop plant anywhere in the world.
Sweet potato remains were also present at a Preceramic site, Huaynuma, dating around 2000 B.C., and at an Initial Period site, Pampa de las Llamas-Moxeke in the Casma Valley, dating from around 1800 to 1500 B.C. (Ugent, Pozorski, and Pozorski 1981: 401-15). In addition, remains have been found at the Early Ceramic Tortugas site in the Casma Valley (Ugent and Peterson 1988: 5).
Still other sweet potato remains in Peru were discovered at Ventanilla, dating from around 2000 to 1200 B.C., the Chillon Valley (Patterson and Lanning 1964: 114), the central coast in the third phase of the Ancon sequence dating 1400 to 1300 B.C. (Patterson and Moseley 1968: 120), and the third Colinas phase dating 1300 to 1175 B.C. (Patterson and Mosely 1968: 121).
Thus, archaeological evidence gives a date of at least 2000 B.C. for the presence of the domesticated sweet potato in the New World, while suggesting a possible domesticated form as early as 8000 B.C.
The Botanical Data
In the past, a precise identification of the ancestor of the sweet potato was hampered by a lack of taxonomic concordance. However, Daniel Austin’s proposal (1978: 114-29) of a Batatas complex with eleven closely related species represents a significant revision. With other species it includes I. trifida, which is often identified as the key species for the origin of Ipomoea batatas (Orjeda, Freyre, and Iwanaga 1990: 462-7). In fact, an I. trifida complex, to include all plants of the section Batatas that could cross with I. batatas, has been proposed (Kobayashi 1984: 561-9).
In the early 1960s, a wild Mexican species of Ipomoea (No. K-123) that was cross-compatible and easily hybridized in reciprocal crosses was identified (Nishiyama 1961: 138, 1963: 119-28). Though it resembled the sweet potato, it lacked domesticated traits (Nishiyama 1961: 138, 1963: 119-28). Cytological studies showed K-123 had a chromosome number (n = 45) similar to the sweet potato, and it was identified as I. trifida (Nishiyama, Miyazakim, and Sakamoto 1975: 197-208). One concern with K-123 was that it might be a feral sweet potato (Austin 1983: 15-25). But other research proposed that I. leucantha was crossed with a tetraploid I. littoralis to produce the hexaploid I. trifida, with the sweet potato selected from this hexaploid (Martin and Jones 1986: 320).
Critical to the debate is the discovery of the natural production of 2 n pollen in 1 percent of diploid I. trifida (Orjeda, Freyre, and Iwanaga 1990: 462-7). The 2 n pollen is larger than the n pollen, and the diploid populations of I. trifida exhibit gene flow between diploids and tetraploids (Orjeda, Freyre, and Iwanaga 1990: 462). Fundamentally, crosses between n and 2 n pollens make various 2x, 3x, 4x, 5x, and 6x combinations of theI. trifida complex possible and could result in 6x combinations leading to the sweet potato (Orjeda, Freyre, and Iwanaga 1990: 466). The plants exhibiting this feature come predominantly from Colombia in northwest South America (Orjeda, Freyre, and Iwanaga 1990: 463).
While this new evidence shows how the sweet potato could have arisen from I.trifida, the present evidence fails to account for the enlarged storage tuber, nonclimbing vines, red periderm color, and orange roots of the sweet potato (Martin and Jones 1986: 322). It is interesting to note the report of Masashi Kobayashi (1984: 565) that some Colombian varieties of I. trifida produce tuberous roots at high elevations. Typically, these plants are found at much lower elevations of between 5 to 20 meters (m) above sea level in Columbia, although some occur at about 1000 m.
Given these observations, it should be mentioned that it is often assumed, after a new species arose through sexual reproduction, that it reached a static stage and that vegetative multiplication has nothing to do with the origin of new forms (Sharma and Sharma 1957: 629). In the various researches into the origin of the sweet potato, investigators have assumed the species arose through sexual reproduction, but karyotypic alterations in somatic cells are common in vegetative reproducing plants, and speciation can occur at those points (Sharma and Sharma 1957: 629). The sweet potato is known to send out roots from the nodes, which will bear small potatoes (Price 1896: 1); therefore, it is possible that karyotypic alterations occurred in the daughter forms, giving rise to new forms, as in some species of Dioscorea. This is important because spontaneous mutations occur quite often, and the sweet potato mutates easily using gamma and X rays (Broertjes and van Harten 1978: 70).
In summary, 20 years ago, the singling out of any one species of Ipomoea as the ancestral form of the sweet potato represented no more than an educated guess (O’Brien 1972: 343;Yen 1974: 161-70). But today the evidence points strongly to the I. trifida complex of plants found in northwestern South America.
Present evidence shows that the sweet potato was introduced into Europe, Asia, Africa, and Australia after Christopher Columbus reached the New World. There is no data to indicate that the plant was known to the ancient civilizations of China, Egypt, Babylon, Persia, Rome, Greece, or India (Cooley 1951: 378), but there is evidence of a pre-Columbian introduction into Oceania. Therefore, in this section, a pre-Columbian and a post-Columbian spread of the sweet potato is outlined, starting with the post-Columbian transfer.
The Post-Columbian Spread
Europe. The sweet potato was introduced into Europe via Spain at the end of the fifteenth century by Christopher Columbus and Gonzolo Fernandez de Oviedo (de Candolle 1959: 55). From this beginning, it spread to the rest of Europe and was called batata and padada (Cooley 1951: 379).
United States. The sweet potato was brought to Londonderry in New Hampshire by the Scotch-Irish in 1719 (Safford 1925: 223). Yet sweet potatoes are mentioned as growing in Virginia in 1648, and perhaps as early as 1610. Further, they are mentioned in 1781 by Thomas Jefferson (Hedrick 1919: 315). They were also reportedly introduced into New England in 1764, and in 1773 the Indians in the South were reported to be growing them (Hedrick 1919: 315-16).
China. Ping-Ti Ho writes that two theories exist for the introduction of sweet potatoes into China. The first involves an overseas merchant who brought the plants from Luzon, which were given to the governor of Fukien in 1594 to alleviate a famine (Ho 1955: 193). The second claim suggests that sweet potatoes arrived via the southern port of Chang-chou, but no specific date of this alleged introduction is known (Ho 1955: 193). Ping-Ti Ho indicates that whereas the former story may be true, the sweet potato was already in China by 1594, having been observed by 1563 in the western prefecture Ta-li, near Burma (Ho 1955: 193-4). Ho concludes that the introduction could have been either overland or by sea via India and Burma, well before the generally accepted date of 1594.
Japan. An English factory at Hirado was allegedly responsible for first introducing the sweet potato to Japan about 1615. It did not, however, “catch on,” and the plant was reintroduced from China in about 1674 to stave off a famine (Simon 1914: 716, 723-4).
India and Southeast Asia. The sweet potato was introduced into India by the Portuguese who brought it to Macão via Brazil (Zavala 1964: 217). Moreover, a Portuguese influence in the spread of the plant to Ambon, Timor, and parts of the northern Moluccas is indicated linguistically, since names for the plant are variations of the word batata (Conklin 1963: 132).
In Malaysia the sweet potato is called Spanish tuber (Conklin 1963: 132). In the Philippines it is called camote (Merrill 1954: 161-384), whereas in Guam it is called both camote and batat (Hornell 1946: 41-62; Conklin 1963: 129-36). In the Moluccas and on Cebu it is called batat (Merrill 1954: 161-384). The names themselves indicate the source of the plant, since the Portuguese used the Arawak term (batata), whereas the Spanish employed the Nahuatl one (camote).
Africa. Harold Conklin (1963: 129-36) reports that the terms for the sweet potato take the form “ batata, tata, mbatata, and the like” from much of the African continent south of the Sahara. The other most widespread term is bombe, bambai, bambaira, or bangbe, with the Indian trade center of Bombay the etymological source of the word and its pronunciation. Names for sweet potato not falling into these categories have a small geographic distribution or are found only in closely related languages, but as lexical entries, they refer to indigenous yams or other root crops. From these data Conklin concludes that the sweet potato was introduced into Africa by the Portuguese from Brazil, probably early in the slave-trade period of the sixteenth century (Conklin 1963: 129-36). In addition, its introduction into West Africa, specifically Angola, probably coincided with Paulo Diasde Novais’s charter of colonization in 1571, which included provisions for peasant families from Portugal with “all the seeds and plants which they can take from this kingdom and from the island of São Tomé” (Boxer 1969: 30).
The Portuguese ports of Mozambique probably saw the introduction of the batata from India as well, although the presence of the word batata cannot always be taken as indicative of a Portuguese association. Their language became the lingua franca along the coasts of Africa and Asia (Boxer 1969: 55).
The word bambai, by contrast, is obviously linked to the city of Bombay, but that port was not significant in the India trade network until the British acquired it in 1662. Consequently, for the word Bombay to be associated with the sweet potato in Africa suggests either a British connection or possibly Indian clerks and merchants involved with British colonization. Consequently, Bombay’s involvement in introducing the sweet potato to Africa could have been as early as the last quarter of the seventeenth century or as late as the nineteenth (O’Brien 1972: 347).
Thus, the evidence seems to suggest that the sweet potato was introduced into Africa by the Portuguese from Brazil and Lisbon in the sixteenth century. A later spread of the plant possibly occurred via British influence in the late seventeenth through nineteenth centuries.
The Pre-Columbian Spread
Polynesia. The traditional view is that the sweet potato was introduced into Polynesia by the Spanish, who brought it to the Philippines in 1521 (Dixon 1932: 41). Moreover, its name in the Philippines, camote, is generically related to the Nahuatl camotl and camotili (Merrill 1954: 317-18). Another point linked to this theory is that the earliest European Pacific explorers, Alvaro de Mendaña and Pedro Ferñandez de Quiros, did not mention the plant by 1606, although they had visited the Marquesas, Santa Cruz, the Solomons, and New Hebrides (Yen 1973: 32-43).
Yet scholars have also argued that the sweet potato was introduced into Polynesia long before Ferdinand Magellan’s 1521 voyage, based on the fact that sweet potatoes were found to be a major part of the economies of the islands located at the points defining the triangle of Polynesia, at the time of their discovery—these being New Zealand in 1769, Hawaii in 1778, and Easter Island in 1722 (Dixon 1932: 45).
Further support for the antiquity of the sweet potato in Polynesia has to do with the very large numbers of varieties found in the South Seas: 48 in New Zealand (Colenso 1880: 31-5), 24 in Hawaii (Handy 1940: 133-5), 16 in the Cook Islands, and 22 in New Guinea (Yen 1961: 368, 371).
Twenty years ago the best evidence to document an early introduction of the sweet potato to Polynesia was historical linguistics that had reconstructed the word for sweet potato in Proto-Polynesian to kumala (O’Brien 1972: 349-50). Over the years, other scholars have scrutinized the proposed antiquity of the word for sweet potato, and believe that a Proto-Polynesian origin of the word is plausible (Pawley and Green 1971: 1-35; Biggs 1972: 143-52; Clark 1979: 267-8).
Such linguistic evidence establishes a base line for the antiquity of the sweet potato in Polynesia, and when combined with archaeological information about the peopling of the Pacific, it is possible to hypothesize the approximate time of entry of the plant to the region. Jesse Jennings (1979: 3) suggests a Polynesian presence on Tonga and Samoa around 1100 and 1000 B.C., respectively, with an initial thrust east into the Marquesas by A.D. 300. This early appearance was probably associated with the Lapita penetration of western Polynesia at around 1500 B.C. from Melanesia (Bellwood 1978: 53).
And finally, in the past ten years, another line of secondary evidence has been investigated in New Zealand, where prehistoric storage facilities and man-made soils had been discovered (Leach 1979: 241-8, 1984, 1987: 85-94).
However, much primary evidence also exists to indicate a pre-Columbian introduction of the sweet potato into Polynesia.
Hawaiian Islands. Archaeological evidence for antiquity of the sweet potato in Hawaii has been found in the form of a carbonized tuber from a fireplace within a “middle” phase domestic structure at Lapakahi. The fireplace is dated A.D. 1655 ± 90 with a corrected date of A.D. 1635 ± 90 or A.D. 1515 ± 90, giving a range of A.D. 1425 to 1765 (Rosendahl and Yen 1971: 381-3). James Cook visited Hawaii in 1778, and so it would seem that this tuber was incinerated at least 13 years prior to his arrival and potentially as many as 263 years before.
Easter Island. The sweet potato was the major crop plant on Easter Island when it was discovered by Jacob Roggeveen in 1722. Charred remains of the plant were recovered there from a fireplace dating A.D. 1526 ± 100 (Skjolsvold 1961: 297, 303). This gives a range between A.D. 1426 to 1626, making the plant remains pre-European by at least 96 years.
New Zealand. In New Zealand, Maori traditions of reconstructing lineage genealogies back to A.D. 1350 recount the arrival of some mysterious “fleet” with the sweet potato and other domesticated plants and animals aboard (Golson 1959: 29-74). Archaeological evidence for the early presence of the sweet potato may exist in the form of ancient storage pits. Jack Golson, for example, (1959: 45) has argued that pits excavated at the fourteenth-century Sarahs Gully site, may have been storage pits for sweet potatoes (kumara). To R. Garry Law (1969: 245) as well, sites like Kaupokonui, Moturua Island, Skippers Ridge, and Sarahs Gully give evidence of widespread kumara agriculture by A.D. 1300.
Primary archaeological evidence was furnished by Helen Leach (1987: 85) with her discovery of charred sweet potato tubers in a burned pit at a pa site from (N15/44) the Bay of Islands. Locally called “Haratua’s pa,” the site is prehistoric, as are the charred sweet potatoes, a point that seems to confirm that these pits were used to store them (Sutton 1984: 33-5).
In addition to the charred tubers at the Bay of Islands, a single charred tuber was discovered at Waioneke, South Kaipara (Leach 1987: 85), a “classic” Maori site 100 to 300 years old (Rosendahl and Yen 1971: 380). Helen Leach (personal communication, letter dated 13 Feb. 1992) notes that no European artifacts were present, and therefore she considers “these kumara pre-European in origin.”
Central Polynesia. The most exciting recent evidence dealing with the antiquity of the sweet potato in Polynesia is the discovery of charred kumara tubers at site MAN-44 on Margaia Island in the Cook Island group dated at A.D. 1000 (Hather and Kirch 1991: 887-93). The presence of charred remains this early seems to establish beyond doubt a pre-Columbian introduction into Polynesia.
Micronesia. In the Carolinas, infrared spectroscopy analyses of organic residues found on pottery has documented the presence of the sweet potato (and taro) at the Rungruw site in the southern part of Yap at about A.D. 50 (Hill and Evans 1989: 419-25). The presence of rice and banana at about 200 B.C. at the same site was also established (Hill and Evans 1989: 419-25). Yap and the Carolinas are near the northern fringe of Melanesia.
Melanesia. The spread of the sweet potato into Melanesia appears to be the result of Polynesian and European introduction, with the former probably ancient. When the Solomons were discovered, there was no mention of the plant, although taro and yams were reported (Mendana 1901: 212). Because Polynesians were present in the Solomons, it is possible that they brought the plant, since the word kumala is found in Melanesian pidgin on the islands (O’Brien 1972: 356).
The term kumala is used in New Caledonia and may be pre-European (Hollyman 1959: 368). The sweet potato was in this area in 1793 (Hollyman 1959: 368), and was grown, in lesser quantities than present, in precontact Fiji (Frazer 1964: 148). Finally, there is evidence of the plant’s presence in the New Hebrides at the time of discovery (Dixon 1932: 42-3).
New Guinea. New Guinea is the one region of Oceania where the sweet potato is of profound economic importance today. It is more widely grown in the western part of the island than in the east (Damm 1961: 209) and is of great importance in the highlands of Irian New Guinea (Damm 1961: 209). Among the inhabitants of Wantoat Valley in highland Papua, the sweet potato is the only important cultivated food (Damm 1961: 212-3).
Dating the introduction of the sweet potato into New Guinea, however, is a problem. Some historical data point to a late entry. For example, it was introduced into the Morehead district by missionaries (Damm 1961: 210) and was even more recently introduced to the Frederick-Hendrick Island region (Serpenti 1965: 38). Moreover, a survey of plants and animals in 1825-6 revealed no sweet potatoes on the islands west of New Guinea, on the island of Lakor, on the Arru and Tenimber islands, and the southwest coast of Dutch New Guinea (Kolff 1840). The plant ecologist L. J. Brass has suggested that the sweet potato came from the west some 300 years ago, carried by birds of paradise and by hunters and traders in the Solomons region (Watson 1965: 439), which may point to a European introduction.
Introduction to the South Pacific
The primary evidence available today suggests that the sweet potato had a prehistoric introduction into Polynesia and Micronesia at around the time of Christ, while the linguistic evidence points to its presence during Proto-Polynesian times. If Proto-Polynesian was the language of the Lapita culture populations, then the sweet potato was present in Oceania possibly as early as 1500 B.C. Given these new data, the next question must be about the mechanism that facilitated its transfer into Oceania at this early date, since the plant is definitely a New World species.
In attempting to answer that question, a number of researchers over the years have been struck by the similarity between the Polynesian words for sweet potato (kumala, kumara) and the word cumara found in some Quechua language dictionaries (Brand 1971: 343-65). This, in turn, has led to the suggestion that the sweet potato came to Polynesia from Peru, with Quechua speakers playing some role.
Alternately, Donald Brand (1971: 343-65) argues that the word was Polynesian and introduced into the Andes by the Spanish. He notes that archaeologists, historians, and philologists consider coastal Ecuador, Peru, and Chile to have been occupied by non-Quechuan and non-Aymaran people until shortly before the arrival of the Spanish. The languages spoken would have been Sek, Yungan, and Chibchan, and their terms for sweet potato were chapru, open, and unt. The Quechua word is apichu and is reported along with the words batatas, ajes, and camotes in the literature and dictionaries of Peru, whereas the word cumara is found only in dictionaries, and cumar proper occurs only in the Chichasuyo dialect of Quechua (Brand 1971: 361-2).
If it is true that the Spanish introduced the word, then one need not explain its presence in Polynesia as the result of people going or coming from the New World. And if the word kumala is Proto-Polynesia, then the term was created by the Polynesians for a plant in their cosmos.
But this still leaves the question of how it might have entered that cosmos. Since the tuber cannot float at all, let alone the thousands of miles separating Oceania and northwest South America, only two explanations appear possible: Transference was accomplished by either a human or a nonhuman agent.
A human agency might have involved a vessel with sweet potatoes aboard drifting from the New World and being cast upon one of the islands of western Polynesia, like Samoa. If any members of the crew survived, they might well have passed along the South American name for the plant. On the other hand, if an empty vessel reached an inhabited island, it would have been examined along with its cargo, and the sweet potato, looking a great deal like a yam, might have been treated like one until its particular features were known. Finally, during a vessel’s long drift, rain water might have accumulated within it, in which case the tubers would have begun to grow, taking hold first in the vessel and then in the soil of some uninhabited island, ultimately becoming feral. Later people finding both the island and the plants would have redomesticated and named them.
An alternative possibility would be transfer by a natural agent. Sweet potato tubers cannot float, but its seeds are more mobile, making birds a likely vehicle. Indeed, Douglas Yen (1960: 373) has suggested the possibility of birds as an agent, and Ralph Bulmer (1966: 178-80) has examined the role of birds in introducing new varieties of sweet potato into gardens in New Guinea by dropping seeds. Bulmer observed that the golden plover, a bird that ranges over Polynesia, is a casual visitor to the west coast of the Americas as far south as Chile. These birds are strong fliers and could have carried the small, hard sweet potato seeds either in their digestive tracts or adhering to mud on their feet.
Another potential nonhuman agent was proposed by J. W. Purseglove (1965: 382-3), who noted that some species of Ipomoea are strand plants and are distributed by sea. He points out that dried sweet potato capsules with several seeds can float. Because the Polynesian sweet potatoes are very distinctive, he suggests that this distinctiveness is the predictable result of an introduction by a few seeds. Purseglove also observes that introduced crop plants have a considerable advantage if major pests and diseases have not been transferred.
At present, all of these scenarios are only speculative, but an accidental introduction would explain how the plant reached the area early, and yet account for the absence of other useful New World products (manioc, maize, and so forth), which might have been transferred if any sustained exchange between living people had been involved.
Although a number of wild members of Dioscorea are edible, there are four domesticated yams that are important to agricultural development: D. alata and D. esculenta from Southeast Asia, and D. rotundata and D. cayenensis from West Africa. A fifth domesticated yam, D. trifida, is found in the New World (Hawkes 1989: 489), but was not especially significant because of the presence of the sweet potato, manioc (cassava), and potato (Coursey 1975: 194) and, thus, has not been a specific focus of research.
The Southeast Asian varieties are interesting because aspects of their spread into Polynesia can be linked to the spread of the sweet potato, whereas African varieties are significant for the role they played in the development of the kingdoms of West Africa. Like the sweet potato, there is no evidence of the use of yams in classical antiquity, but historical data point to their presence in China in the third century A.D. and in India by A.D. 600 (Coursey 1967: 13-14).
The Botanical Data
The family Dioscoroeaceae has hundreds of species, and the Dioscorea is its largest genus. In general, the New World members have chromosome numbers that are multiples of nine and the Old World species multiples of ten (Ayensu and Coursey 1972: 304). The section including the food yams typically has 2 n = 40, but higher degrees of polyploidy do occur (Coursey 1967: 43-4). For example, D. alata has 2 n = 30 to 80; D. esculenta has 2 n = 40, 60, 90, 100; D. rotundata has 2 n = 40; and D. cayenensis has 2 n = 40, 60, and 140.D. trifida, the New World domesticated yam, has 2 n = 54, 72, and 81 (Coursey 1976a: 71).
The two yams domesticated in Southeast Asia have been major constituents (along with taro, plantains, bananas and breadfruit) of root crop agriculture in the region, and throughout Oceania before European contact. According to D. G. Coursey (1967: 45), D. alata, or the “greater yam,” is native to Southeast Asia and developed from either D. hamiltonii Hook. or D. persimilis Prain et Burk. It is unknown in the wild state and today is the major yam grown throughout the world (Coursey 1967: 45-6). D. esculenta, or the “lesser yam,” is a native of Indochina, has smaller tubers than the “greater yam” (Coursey 1967: 51-2), and occurs in both wild and domesticated forms (Coursey 1976a: 71). The two native African yams, probably ennobled in the yam belt of West Africa, are D. rotundata Poir., the white Guinea yam, and D. cayenensis Lam., the yellow Guinea yam (Coursey 1967: 11, 48, 58). The most prominent English-speaking scholars to work on the genus Dioscoreahave been I. H. Burkill and D. G. Coursey (1976b, 1980; see also Coursey 1967: 28 for further discussion). Indeed, Coursey, the preeminent yam ethnobotanist, has developed a detailed theory of their domestication and antiquity in Africa.
Nonetheless, African yams have not received the attention of plant scientists that they need and deserve, especially in terms of cytological research and breeding programs (Broertjes and van Harten 1978: 73-4). This omission is particularly regrettable in light of their ancient importance, but is doubtless the result of yams being displaced by New World cultivars like maize, sweet potato, and manioc in many parts of Africa.
The lack of botanical research, however, allows plenty of room for controversy. For example, some botanists separate West African yams into two species, whereas others argue that there are insufficient criteria (basically, tuber flesh color) to separate them. They suggest the existence of a D. cayenensis-rotundatacomplex under the rubric of one species, D. cayenensis (Miege 1982: 377-83).
D. G. Coursey, as mentioned, identifies the two yams as D. rotundata Poir., and D. cayenensis Lam. (Coursey 1967: 11, 48, 58). He suggests that the former is unknown in the wild, being a true cultigen, and it may have developed from D. praehensilis Benth. (Coursey 1967: 59). The latter, however, is found in both a wild and domesticated condition (Coursey 1967: 48), which may indicate that the wild D. cayenensis is the ancestor of the domesticated D. cayenensis.
J. Miege (1982: 380-1) states that D. cayenensis is a complex cultigen most probably made up of several wild species: D. praehensilis Benth. for the forest varieties; and D. sagittifolia Pax., D. lecardii De Wild., D. liebrechtsiana De Wild., and D. abyssinica Hochst. ex. Kunth for the savanna and preforest types. An implication of the argument that these two domesticated species are but subspecies of D. cayenensis is that both the white and yellow Guinea yams could have risen from wild forms of D. cayenensis.
Clearly, such uncertainties will only be resolved by concerted research focused not only upon taxonomic issues but especially on cytological ones. A whole series of breeding and cross-breeding studies are essential, and it would be particularly useful to determine whether Dioscorea polyploidy is related to 2 n pollen as it is in Ipomoea.
Transformation and Dispersal
As we noted, the four major domesticated yams come from Southeast Asia and West Africa, respectively. This section examines data, primary and secondary, for their antiquity and their movement throughout the world.
Southeast Asia. In the late 1960s, charred and uncharred botanical material was recovered from excavations at Spirit Cave in Thailand. It was associated with the Hoabinhian complex, dated to around 10,000 to 7000 B.C., and was used to argue for the early development of agriculture in Southeast Asia (Gorman 1969, 1970). Later, however, these materials were critically reexamined by Yen (1977: 567-99), who concluded that most of the remains were not domesticates. Yen thought that early yam domestication could not be inferred from these remains, but that it was probably reasonable to suspect that wild yam was being eaten at that time (Yen 1977: 595).
The fundamental evidence for the antiquity of domesticated Southeast Asian yams and other cultivars is linguistic and lies within words for the whole assemblage of plants and animals making up Southeast Asian root crop agriculture. Robert Blust, a linguist, notes (1976: 36) that Proto-Austronesian speakers had pigs, fowl, and dogs and cultivated a variety of root and tree crops including taro, yams, sago, bread-fruit, sugarcane, and banana (Blust 1976: Table II.B.6.1). The linguist Ross Clark reports that words for all the crop plants important in Polynesia horticulture—yam, taro, bananas, sugarcane, and sweet potato—reconstruct to Proto-Polynesian (Clark 1979: 267-8). In relation to this, it should be mentioned that a Lapita site on Fiji, dating between 1150 to 600 B.C., has primary evidence for aspects of this economy in the form of bones of dogs, chickens, and pigs (Hunt 1981: 260).
Helen Leach (1984: 20-1) believes that a series of 21 roundish holes about 35 centimeters (cm) in diameter and some 60 cm deep excavated within a 33 square meter area at Mt. Olo in Samoa implies yam cultivation, for she reports that large yams on Fiji and in other parts of Melanesia are planted in individual holes 60 cm in diameter and 45 cm deep. She also argues for the antiquity of root crop agriculture at Palliser Bay in New Zealand through indirect evidence such as storage pits, garden boundaries, old ditches, and “made-soils” (Leach 1984: 35-41). Susan Bulmer (1989: 688-705) makes these same points, but emphasizes the importance of forest clearance, which in New Zealand appears as early as A.D. 400. Indeed, the antiquity of root crop agriculture in New Guinea is documented by this same type of indirect evidence, and Jack Golson outlines a five-phase model of agricultural development and intensification based upon a whole series of field drainage systems that can be dated as early as 7000 B.C (Golson 1977: 601-38).
In sum, the evidence, though more indirect than direct, supports the notion that the domestication of the Southeast Asian yams, D. alata and D. esculenta, is very ancient, maybe as early as 4500 B.C. This being the case, what of their dispersal?
The first dispersal is clearly associated with its transfer by Proto-Austronesian-speaking peoples throughout the Southeast Asian tropical world. However, the diffusion of these people is in some dispute. For example, Peter Bellwood (1985: 109) argues that the original Pre-Austronesians were located in Taiwan, whence they moved to the Philippines and from there to parts of Indonesia like Borneo, Java, Sumatra, and Malaya, then into the Moluccas, and finally into New Guinea and Oceania (Melanesia, Micronesia, and Polynesia). But Wilhelm Solheim (1988: 80-2) suggests that Pre-Austronesians developed around 5000 B.C. in Mindanao and the northeast Indonesia regions. He argues against Taiwan as a homeland because of the difficulties posed by winds and currents for sailing south to the Philippines. William Meacham (1988: 94-5), however, considers the languages of south China to have been Mon-Khmer, not Austronesian, and argues that these people could not have migrated to Taiwan and from there south into the Philippines. Rather, Meacham suggests, the homeland of the Proto-Austronesians is somewhere in the triangle defined by Taiwan, Sumatra, and Timor, basically encompassing modern Indonesia.
Regardless of which theory of a Proto-Austronesian homeland is correct, once the Proto-Oceanic languages of that family began to differentiate, they also began to provide linguistic evidence for yams and other cultivars. Thus, yams were in Melanesia by 2000 B.C., in Micronesia by 2000 to 1000 B.C., and in eastern Polynesia by 1500 B.C. The bulk of western Polynesia received yam horticulture (depending on when a specific island group was occupied) sometime between A.D. 1 and 1000 (Bellwood 1985: 121).
In addition, the transfer of Southeast Asian yams with Austronesian speakers to regions outside this early core area is documented. They were present in China in the third century A.D. and in India by A.D. 600 (Coursey 1967: 13-14). To the west, yams were introduced (principally D. alata into Madagascar, probably between the eleventh and fifteenth centuries A.D. By the end of the sixteenth century, D. alata was grown in West Africa, from whence it was transferred to the New World by a Dutch slaver in 1591 (Coursey 1967: 15-6).
Africa. The student of African yams, D. G. Coursey, argues (1967: 13; 1975: 203; 1976a: 72) that the use of D. cayenensis and D. rotundata is ancient, and he proposes the following scenario for the process of yam domestication in West Africa (1980: 82-5). He suggests:
- That hunter-gatherers, before 60,000 B.P. (before the present), utilized many species of wild yam;
- That the Sangoan and Lupemban Paleolithic stone industries, c. 45,000 to 15,000 B.P., developed hoes or picks to excavate hypogeous plants, including the yams, and at this time started to develop ritual concepts and sanctions to protect these and other plants;
- That sometime around 11,000 B.P., with the contraction of West Africa forest and savanna environments and appearance of proto-Negro people, microlithic industries developed which point to new human/environment interactions; these interactions involved selection and protection of favored species, particularly nontoxic yams; and this greater control led to population increases, movement into forest environments and a planting of wild plants—a “protoculture” with a final result being the understanding that one could replant stored tubers;
- That by 5,000 to 4,000 B.P. Neolithic grain-crop people from the Sahara belt, influenced by Middle Eastern agriculturalists, moved south and interacted with yam “protoculturalists” and from this relationship yam-based agriculture developed;
- And finally, around 2,500 B.P., with the advent of ironworking, West Africa people could expand deeper into the forest which ecologically favored yam over grain crops, and yam growing populations could achieve numerical superiority over grain farmers and create complex culture systems.
Although this model seems reasonable for the most part, the problems of documenting the domestication of West African yams are similar, and in some cases identical, to those associated with Southeast Asia. Here, too, primary evidence is lacking (Shaw 1976: 108-53).
Preliminary research on yam ennoblement, which was begun in 1977 in Nigeria, has led to the discovery that digging wild yams even with modern tools like machetes, shovels, and spades, let alone digging sticks, was arduous work (Chikwendu and Okezie 1989: 345). Wild yams could not be excavated like domesticated ones. They have long, sharp thorns all over their roots, and in addition to cutting through the yam roots, one has to cut through the tangled roots of the forest itself. A pick-like tool would only get caught between the roots. Trenching around a yam patch was the best procedure, but it still took several days just to dig up the first yam (Chikwendu and Okezie 1989: 345). This finding in turn casts some doubt on Coursey’s proposal (1967: 13, 1975: 203) that the pick-like stone tools and Lupemban “hoes” of the Sangoan period were used for grubbing yams.
As with research on the Southeast Asian yams, indirect evidence like forest clearance and linguistics is our main avenue of inference. M. A. Sowunmi (1985: 127-9) reports significant changes in pollen counts from a Niger Delta soil core occurring around 850 B.C. He notes a sudden increase in oil-palm pollen and an increase in weed pollens associated with cultivated land, accompanied by a decrease in pollen of rain forest components, such as Celtis sp. and Myrianthus arboreus. Because there is no evidence of environmental change at that time, he concludes that humans artificially opened the forest for agricultural purposes. Because oil palm and yams are the main cultivars of aboriginal West African agriculture, he believes that these data document their appearance on a large scale.
It should be noted that, on the one hand, iron hoes, axes, and knives appeared in Nigeria (with the Nok complex) only about 300 years later, around 550 B.C. On the other hand, the site of Iwo Eleru has polished groundstone axes, dating as early as 4000 B.C., that could have been used in forest clearance, and Coursey (1967: 197-205, 1976b: 397) argues that yams were grown before the development of an iron technology because many of the peoples of West Africa have strong prohibitions against the use of iron tools in their important New Yam festivals.
Linguistics, as mentioned, is another source of information. Kay Williamson’s study (1970: 156-67) of names for food plants within the Kwa Branch of the Niger-Congo family (spoken in the Niger Delta region) isolated “three main layers of names; ancient West African plants, crops of the Malaysian complex introduced long ago, and more recent introductions over the last five hundred years” (Williamson 1970: 163). The oil palm and the yam (D. cayenensis-rotun-data complex) belong to the oldest layer; the banana, plantain, and water yam (D. alata) occurred in the Malaysian layer; and such plants as maize and manioc (cassava) are more recent introductions from the New World some five hundred years ago.
Williamson does not assess the antiquity of the words for yam and oil palm in calendar years, but P. J. Darling (1984: 65) proposes that the Proto-Kwa language dates from between 4,000 and 10,000 years ago. Although he calls these Proto-Kwa speakers Late-Stone-Age hunter-gatherers, it seems clear that as they had words for major domesticated plants, they must already have been farmers. It is interesting that the more recent end of this date range matches Coursey’s model for yam “protoculturalists” quite well. Finally, Proto-Niger-Congo not only has the word for yam (and cow and goat) but also the root meaning “to cultivate,” and Proto-Niger-Congo may date back to at least 6000 B.C. (Ehret 1984: 29-30).
Thus, the evidence, though indirect, does point to the existence of yam usage and the concept of cultivation at around 6000 B.C. and forest clearance at about 850 B.C., presumably for the purpose of producing oil palms and yams on a wider scale. All of this in turn suggests an antiquity for agriculture in Africa far greater than believed by many scholars, which probably can best be explained in terms of an independent agricultural development in West Africa. Yet the West African yams had no dispersal beyond their region of origin until they were transferred to the tropical New World in association with the slave trade.
Three main waves of dispersal are associated with the spread of the sweet potato, in what Yen (1982: 20-1) calls the kumara, kamote, and batatas lines of transfer. The best-known and documented transfer was the post-Columbian spread via Europeans associated with the latter two lines. The Spanish, or kamote line, introduced the sweet potato into Europe, the Philippines, Guam, and Malaysia. From the Philippines it was then carried to China and from China ultimately to Japan. English immigrants transmitted it to the United States, English traders brought it to Japan (though it was not accepted), and English missionaries introduced it in parts of Melanesia and Australian New Guinea.
The Portuguese, or batatas line, introduced the sweet potato into India and Africa, Ambon, Timor, the northern Moluccas, and Cebu. The African introduction was from the Portuguese into Angola and Mozambique, as well as to Africa via Bombay through English associations with that trade center in India. Apparently the plant was carried from Burma to China after the Indian introduction.
The kumara line, the earliest, is associated with the appearance of the sweet potato in Oceania. This transfer has intrigued scholars for years. New primary evidence, combined with linguistic and historical data, point to a pre-Columbian spread somewhere into eastern Polynesia or even into northern Melanesia by the time of Christ. From this region the plant spread to all points of the Polynesia triangle. It then moved to parts of Melanesia via the Polynesians, and traveled from Melanesia into New Guinea. The transfer into New Guinea was probably accomplished by Melanesians, possibly bird of paradise hunters or migrants settling on the southeast coast. Though some specific areas of New Guinea received the plant from Europeans, in general it was first spread by Melanesians and then by Papuans from the coast into the highlands, probably through the Markham Valley. The way in which early sweet potatoes reached New Guinea cannot presently be determined, but in the light of the Yap data it could be earlier then generally supposed.
The establishment of the sweet potato in many areas of Micronesia, parts of central Polynesia, and sections of Dutch New Guinea, including Lakor and the Arru and Tenimber islands, was prevented by ecological conditions unsuitable to its growth.
Yams also had several waves of dispersal. The Southeast Asian yams moved through the region beginning about 4500 B.C, and on into Oceania by 1500 B.C. They arrived in India and China in the first millennium A.D., and early in the second millennium entered Madagascar. From East Africa they moved to West Africa by the sixteenth century, and at the end of the sixteenth century came to the tropical New World.
This survey on the problem of the origin and dispersal of the sweet potato and of yams indicates the following. First, the sweet potato originated in northwestern South America around 8000 B.C., in association with the initial development of tropical-forest root crop agriculture. The actual botanical ancestor is probably the result of various n and 2 n crosses within the I. trifida complex.
Primary evidence of the pre-Magellan introduction of the sweet potato into central Polynesia is established at around A.D. 1000, and even earlier, A.D. 50, in Micronesia on Yap. When combined with the archaeology of Oceania, these data suggest, conservatively, that the plant arrived in eastern Polynesia, maybe in the Fiji area, by about 500 B.C. Alternatively, the plant was dispersed by Lapita people sometime between 1500 to 500 B.C. during their movement through Melanesia. From Melanesia it was carried to New Guinea by Melanesians at an unknown date, but this could have taken place prior to the arrival of the Europeans.
The transference between Polynesia and the New World would seem to have been the result of either human accident or natural causes. An introduction by the casting up of a vessel upon some island of eastern Polynesia is possible, but it is equally possible that the plant was spread by natural agents, such as birds carrying seeds or by floating seed capsules. Both these hypotheses need further examination.
The post-European introduction of the sweet potato into Africa, North America, Europe, India, China, Japan, the Philippines, the Moluccas, and other islands in the Indonesian area was the result of Spanish, Portuguese, and English trade, exploration, colonization, and missionization.
The five ennobled yams were domesticated in Southeast Asia, West Africa, and tropical America, respectively, although the last region is not especially important to this study. Southeast Asian yams were probably domesticated before 4500 B.C., whereas the West African yams could be as old as 6000 B.C. but were probably domesticated by the first millennium B.C. The possible botanical ancestors of these yams are a subject of debate, and considerable cytological and taxonomic research is needed before this issue will be resolved. Needless to say, these ancestors will be found to have been native to each respective area.