Lawrence Kaplan. Cambridge World History of Food. Editor: Kenneth F Kiple & Kriemhild Conee Ornelas. Volume 1. Cambridge, UK: Cambridge University Press, 2000.
On Sunday, November 4, 1492, three weeks after his first landing in the New World, Christopher Columbus saw lands planted with “faxones and fabas very diverse and different from ours [those of Spain] and two days afterward, following the north coast of Cuba,” he again found “land well cultivated with these fexoes and habas much unlike ours”(Hedrick 1931: 3).
In a transcription (Dunn and Kelley 1989: 132) from Columbus’s diary, the Spanish phrase faxones y favas has been translated as “beans and kidney beans” (Morrison and Jane-Vigneras, cited by Dunn and Kelley 1989: 133). But considering what Columbus might have seen in the markets or kitchens of the fifteenth-century Iberian–Mediterranean world, faxone probably refers to the African-Asian cowpea (Vigna unguiculata), and fava surely means the fava (= faba), or broad bean (Vicia faba), long known in Europe and the Mediterranean-Asian world. Columbus’s brief record presaged the long confusion and debate over the names and origins of some important food grain legumes. Had herbalists and botanical authors of the succeeding three centuries taken account of Columbus’s recognition that these New World legumes were different from those of Europe, some of the confusion might have been avoided.
The beans, peas, and lentils (pulss, or food grain legumes) discussed in this chapter are legumes, treated in technical botanical literature as members of the very large family Fabaceae (= Leguminosae), sub-family Papilionoideae (having butterflylike flowers), although some taxonomists accord this group family status (Papilionaceae).The names of the species, however, are not changed by the differing positions taken by plant taxonomists on family nomenclature. The flowers of papilionaceous legumes have five petals consisting of one broad standard, two lateral wings, and two keel petals that clasp the 10 stamens and single ovary (which becomes the pod).
The European herbalists of the sixteenth century repeatedly attempted to reconcile their plant world with that of the ancients as represented by the fragmentary remains of the works of Theophrastus and Dioscorides. Their efforts, however, were destined to fail inasmuch as many of their subjects were novel plants newly introduced from the New World and from the reopened contact with Asia and coastal Africa.This nomenclatural dilemma is illustrated by an attempt to reconcile the name of a bean with the appropriate plant. J. P. Tournefort (1656-1708), a notable pre-Linnaean French botanist, sought to clarify the use of boona or baiana for the broad bean or fava by a well-known sixteenth-century herbalist:
Dodonaeus [said Tournefort] called this kind of pulse “boona” in Latin, who [Dodonaeus] relying on a Germanism abuses his own language in order to appear learned but our Boona or Bean seem rather to be derived from the Italian word Baiana which Hermolaus says is the word used by those that sell new BEANS all over the state of Milan and along the Appenine mountains … Garden beans are common and universal in Europe and are a great supply in dearth of Provisions in the spring and whole summer season … The ancients and Dodonaeus believed that beans are windy and the greener the more so. (Tournefort 1730: 386)
Tournefort then disagreed on the suitability of beans in the diet and said that he would leave the hard dry beans to “the laboring men who can better digest them, but [even] those of delicate constitution and sedentary life digest green beans well enough if they eat them with butter and pepper [especially if they will] be at the pain to take off their skins” (Tournefort 1730: 386). Inasmuch as American Phaseolus beans had entered Europe by his lifetime, one could wonder whether Tournefort meant Phaseolus or Vicia “beans.” However, his remark concerning the removal of “skins [seed coats]” should end any doubt.
“Skins” refers to the seed coats. Where the broad bean and American Phaseolus beans are concerned, only the broad bean has its skin or testa customarily removed—probably to eliminate substances that are toxic for individuals having an inherited enzyme (glucose-6-phosphate dehydrogenase) deficiency. Even in a contemporary and remote Quechua-speaking community, located in the southern Andes in the vicinity of Cuzco at an altitude of 3,000 to 5,000 meters, which is unfavorable for Phaseolus cultivation, cooked Vicia faba seeds are always peeled before eating (Franquemont et al. 1990: 83). Because this enzyme-deficiency sensitivity to fava bean components evolved in human populations native to certain malarial regions of Eurasia and Africa (Strickberger 1985: 738), the custom in the Andes was probably introduced along with the fava bean.
Tournefort further assumed the responsibility for ending the quandary of post-Columbian botanists concerning the identity of the fava. He recognized that there was
… much controversy among the botanists as to whether our bean be the bean of the ancients … that of the ancients was small and round [according to] Theophrastus, Dioscorides and others. But it is strange that a pulse so common should have come close to disuse and been replaced without anybody’s knowing anything of this matter. (Tournefort 1730: 386)
The reason for the difference (and confusion), he went on, could be that “their faba was not arrived at the bigness that our Garden Bean now is.” But however intriguing the evolutionary explanation, the writers of the classical period, whom he cites, may have been referring to the African-Asian cowpea, Vigna unguiculata, rather than one of the small-seeded races of the fava bean.
Contemporary linguistic sources (Webster’s Third New International Dictionary 1971) derive the familiar English word “bean” from a root common to Old English, Old High German, and Old Norse that apparently once referred to the fava or faba bean, a staple of the Romans. Over the centuries, however, the word has grown to encompass seeds of other plants, including a multitude of legumes besides the fava, most of which belong to other genera—a terminological tangle that botany attempts to avoid through the use of scientific names.
The distinct identities of these two groups of food crops, favas and Phaseolus beans, were being established as seventeenth- and eighteenth-century botany ceased the attempt to reconcile the known species of that period with the fragmentary records of the classical authors. These advances were only the beginning of the solutions to the geographic, temporal, and cultural problems surrounding the origins of these foods.
The Search for Geographic Origins
The origins of most domesticated plants, meaning the time span, the wild relatives, and the conditions (both natural and human-influenced) under which divergences from wild ancestral stock took place, are remote in time and are matters of continued botanical and genetic inquiry.
For Europe and the Mediterranean Basin, sixteenth-century European herbalists turned for information to tradition, the observations of travelers, and the surviving books of classical authors. Linnaeus assimilated the writings of the herbalists and added his contemporary experience with eighteenth-century introductions and collections. Alphonse de Candolle (1964: v) in 1855, and especially in the 1886 edition of his Origin of Cultivated Plants, brought to the attention of botanists the utility of archaeological findings for supplementing plant morphology and taxonomy in adducing evidence for the geography of domestication.
In the twentieth century, N. I. Vavilov, following Candolle’s pioneering work in geographical botany, embarked on his global, decades-long projects on the origin and genetic variation of cultivated plants. In 1926, Vavilov (1992: 22-7) organized a comparative chart of the morphology of cultivated species of the papilionaceous legumes and presented the rationale for botanical-genetic determination of the centers of origin of these and other crop plants. His geographic centers of origin for crop plants—which have been highly influential and much discussed in the literature of crop plant geography—included eight major and several minor centers (Vavilov 1992: 324-53). The lentil, the pea, and the broad bean were all traced to the Inner-Asiatic Center: northwestern India, Afghanistan, Tadzhikistan, Uzbekistan, and western China. The lentil and pea were also assigned by Vavilov to the Asia Minor Center (the Middle East, Iran, and Turkmenistan). The American common bean and lima bean were both assigned primary centers in the South Mexican-Central American Center, and the lesser-known species, scarlet runner beans and teparies, were also located there. The origin of the little-known polyanthus bean (Phaseolus polyanthus) has been documented in the same area (Schmit and Debouck 1991).
New World Beans: Phaseolus
The Pathway of Domestication in Phaseolus
The species of domesticated Phaseolus beans have shown parallel changes in structure and physiology and share some of these characteristics with the Old World grain legumes. J. Smartt (1990: 111) summarized his own findings and those of others on the nature of evolutionary changes in the Phaseolus cultigens. These are gigantism (increased size of seed and other plant parts); suppression of seed dispersal mechanisms (decreased tendency of pods to twist and discharge seeds); changed growth form (especially the loss of rampant vining); loss of seed dormancy; and other physiological and biochemical changes. The genetic bases for seed size, dispersal mechanisms, and growth form are partly understood, and some are even observable in archaeological specimens, in which wild Phaseolus beans can be readily distinguished from domesticates by both seed size and nondehiscent pod structure.
The common bean. It is clear in the writings of sixteenth-century herbalists, and later in the works of Linnaeus and Candolle, that the original home of the common bean was unknown to them. It was archaeological excavation on the arid coast of Peru in the last quarter of the nineteenth century that convinced Candolle (1964: 341-2) that Phaseolus vulgaris and Phaseolus lunatus had been cultivated in the Americas since pre-Columbian times.
At the time of contact with Europeans, varieties of the common bean were grown by Native Americans as far south as Chile and Argentina, and as far north as the valleys of the St. Lawrence and upper Missouri rivers. Edward Lewis Sturtevant (Hedrick 1972: 424) has noted that beans were observed to be in cultivation among Florida Indians by at least three explorers from 1528 to 1562, including Hernando de Soto (in 1539), who said that “the granaries were full of maes and small beans.” The Natchez on the lower Mississippi grew beans as a “subsidiary crop” (Spencer et al. 1965: 410), and there is a 1917 description of traditional Hidatsa-Mandan cultivation of beans in hills between the rows of maize or occasionally “planted separately” (Spencer et al. 1965: 343).This observation suggests the planting of erect or semierect beans. Such beans are intermediate between strong climbers and truly dwarf nonclimbing beans. In California, outside of the lower Colorado River drainage where the Mohave grew tepary beans (as did other Yumans, along with maize, cucurbits, and other crops), bean agriculture began with the introduction of Mexican and European crops when the earliest Spanish missions were established. G. W. Hendry (1934) found a bit of seed coat of the common bean cultivar ‘Red Mexican’, or ‘Pink’, in a Spanish adobe brick dated 1791 at Soledad.
R. L. Beals (1932) mapped pre-1750 bean cultivation in northern Mexico and the adjacent southwestern and Gulf Coast United States using historical documents and reports by Spanish explorers. Bean distribution coincided with maize cultivation, extending from the Colorado River east to include the Rio Grande Pueblos, Zuni, and Hopi. The area of the eastern Apache in the Pecos River drainage, and of the Comanche, was nonagricultural. Beans were grown from eastern Texas, beginning with the Waco and the Kichai, eastward to the Atlantic.
Southwestern bean horticulture derives from Mexico. P. A. Gepts’s (1988: 230) mapping of bean dispersal routes by means of the beans’ protein structure corroborates this generally accepted view. Where water is the limiting factor in growth and production, as on the Hopi mesas, varieties having the dwarf or bush habit are planted. Where surface water is available or can be supplied, vining beans are planted with corn or, if planted separately, are provided with poles to climb upon. Except for the pinto or garapata group, the common beans collected during the years 1910 to 1912 by G. F. Freeman (1912) among the Papago and Pima do not appear in the archaeology of the southwestern United States. Instead, these beans represent introductions from the Mexican Central Highlands and may well have arrived during the Spanish colonial period.
According to E. F. Castetter and W. H. Bell (1942), the Papago planted teparies in July and harvested in October; the Pima planted twice, when the mesquite leafed out (late March to mid-April) and after the Saguaro harvest (July). The first planting was harvested in June, the second in October. The harvest of teparies and common beans was women’s work. The plants were pulled, dried, and threshed on a clean, hard-packed soil floor in the open, in the field, or near the house. Different varieties were planted, harvested, and threshed separately. After threshing they were sealed in baskets or pots for protection from pests. Castetter and Bell (1942) and Freeman (1912) reported only bush beans in which the “vines” (sprawling plants) are grown separately from corn.
Vavilov (1992) and C. B. Heiser (1965) speculated on multiple American origins for the common bean, and substantial evidence has been adduced to show that common beans were domesticated independently in at least two distinct areas: Mesoamerica and Andean America (Kaplan 1981; Delgado Salinas 1988; Gepts 1988).
Lima beans. The lima and sieva beans were recognized by Linnaeus to belong to the same species, which he called P. lunatus to describe the “lunar” shape of the seeds of some varieties. The small-seeded or sieva types are natives of Mexico and Central America. These are distinct from the large-seeded South American lima beans, but can be crossed with them, and for this reason are considered to be of the same species. The sievas appear in the archaeological records of Mexico about 1,200 years ago, but do not occur in the known records of Andean archaeology. The large limas of South America, conversely, do not appear in the Mesoamerican or North American record. Seeds of the South American group have been found in Guitarrero Cave, the same Andean cave as the earliest common beans of South America, and have been 14-Carbon dated at the National Science Foundation, University of Arizona Accelerator Facility, to approximately 3,500 years ago (Kaplan 1995).The archaeological evidence of geographic separation coincides with contemporary observations of the distribution of both the wild and cultivated types. It seems clear that the two groups were domesticated independently. Both vining and bush forms are known, but the vining forms predominate in indigenous horticulture.
On the desert north coast of Peru, remains of the large-seeded lima group, dating to about 5,600 years ago (the preceramic period) (Kaplan 1995), were well preserved by the arid conditions and were so abundant that they must have constituted a major part of the diet. Ancient Peruvians even included depictions of beans in their imaginative painted ceramics and woven textiles. Painted pottery from the Mochica culture (A.D. 100-800) depicts running messengers, each carrying a small bag decorated with pictures of lima beans (pallares). Rafael Larco Hoyle (1943) concluded that the lima beans, painted with parallel lines, broken lines, points, and circles, were ideograms. Some of the beans that Larco Hoyle believed to be painted were certainly renderings of naturally variegated seed coats. Other depictions show stick-limbed bean warriors rushing to the attack. Textiles from Paracas, an earlier coastal site, are rich in bean depictions.
Scarlet runner beans. The cultivated scarlet runner bean (Phaseolus coccineus L., for the scarlet flowers) is not known in the archaeological record north of Durango, Mexico, where it was grown about 1,300 years ago (Brooks et al. 1962). It has undoubtedly been cultivated for a much longer period in the cool central highlands of Mexico and Guatemala, but in that region, archaeological specimens securely dated and identified as older than a few hundred years are wanting. Runner beans, both purple-seeded and white-seeded, have been collected among the Hopi in historic times, especially by Alfred Whiting during the 1930s.
Tepary beans. The tepary is one of the two cultivated species not to have been named by Linnaeus. The wild type was called Phaseolus acutifolius by the nineteenth-century Harvard botanist Asa Gray. However, it was not until the early years of the twentieth century that the cultivated types were recognized by the Arizona botanist, Freeman, to belong to Gray’s species rather than simply being varieties of the common bean.
Teparies, now little-known as commercial beans, were cultivated in central Mexico 2,300 years ago (Kaplan 1994), and in Arizona, teparies were being grown 1,000 to 1,200 years ago (Kaplan 1967). Despite their antiquity and ancient distribution, teparies have been absent from village agriculture in historic times, except in the Sonoran Desert biome of northwestern Mexico, Arizona, and New Mexico, and in the area around Tapachula in the Mexican state of Chiapas and in adjacent Guatemala. They are grown and eaten by the Pima, Papago, and peoples of the lower Colorado River and by some “Anglo” enthusiasts for dryland-adapted crops. Because of their drought tolerance, they have been tested in many arid regions of the world.
Polyanthus beans. A fifth cultivated bean species, P. polyanthus, has long been known to be distinct from the other, better-known species, but only recently have its identity and distribution been documented (Schmit and Debouck 1991). To the best of my knowledge this bean of high elevations in Mexico and Central America has never entered into Old World cultivation.
The five Phaseolus beans are distinct species. They have different botanical names, applied by plant systematists on the basis of their structural differences. They have the same number of chromosomes (2 n = 22) but do not freely hybridize. However, the domesticates do hybridize with some wild-growing populations that are regarded as their ancestral relatives.
The Antiquity of Phaseolus Beans: New Evidence
Uncovering the botanical and geographic origins of domesticated crops includes the search for their temporal origins. Candolle (1964), as noted previously, brought to the attention of plant scientists the utility of archaeological evidence in the quest for the temporal as well as the geographic origins of crop plants. The presence of Phaseolus beans on the arid coast of Peru in pre-Conquest graves of indigenous peoples did not indicate a specific calendar date for the remains, but it was convincing evidence that Phaseolus beans were present in the Americas before contact with European cultures. With the development of radiometric dating by the middle of the twentieth century, it became possible to determine the age of archaeological organic materials with significant precision. However, many of the published dates for the earliest crop plant remains, including beans (Kaplan 1981), are now being questioned because the 14-Carbon determinations of age are “contextual dates,” meaning they are based on organic materials in the same strata with the bean remains but not on the beans themselves (Kaplan 1994).
Because of the tendency of small objects, like seeds, to move downward in archaeological contexts, some of the dates now in the literature are too early. The development of 14-Carbon dating by Atomic Mass Spectrometry (AMS), which measures very small samples, has allowed the dating of single bean seeds or pods, and in some instances has produced dates that disagree with the contextual dates. For example, a single bean-pod found in Coxcatlan Cave in the Tehuacan valley, in a content 6,000 to 7,000 years old, was AMS-dated to only 2,285 ±60 years ago (Kaplan 1967, 1994). An early date for beans in South America comes from Guitarrero Cave in the Peruvian Andes, where radiocarbon dates of plant debris are as old as 8,000 years (Kaplan, Lynch, and Smith 1973; Lynch et al. 1985). But the AMS 14-Carbon date for one seed embedded in this debris is 2,430 ±60 years old (Kaplan 1994). Disagreements over the accuracy of AMS dates (unpublished data) versus contextual 14-Carbon dates are being aired, and the debate will continue. However, an AMS date from New England (Bendremer, Kellogg, and Largy 1991) supports a contextual 14-Carbon date, which suggests the entry of common beans into northeastern North America (Ohio) about 1,000 years ago (Kaplan 1970). In the southwestern United States, AMS dates (Wills 1988: 128) of beans from a New Mexico cave agree with contextual dates (Kaplan 1981).
The wild types of all Phaseolus bean species are vining types, as the earliest domesticates must also have been; the dwarf-growing types came later. The earliest evidence now available for the presence of the dwarf, or bush, growth habit comes from an accelerator radiocarbon date of 1285 ±55 years ago for bean-plant remains from an archaeological cave site in the northeastern Mexican state of Tamaulipas.The vining or pole beans of each species were planted with corn so that they could depend on the stalks for support.The sprawling and dwarf types could be grown independent of support.
Tracing Bean Migrations
Molecular evidence. Phaseolin is the principal storage protein of Phaseolus bean seeds. Gepts (1988) has used the variation in phaseolin structure to trace the dispersal of contemporary common bean cultivars within the Americas and in the Old World. He has shown that the majority of the present-day cultivars of Western Europe, the Iberian Peninsula, Malawian Africa, and the northeastern United States originated in the Andes. The ‘C’ phaseolin type is found in highest frequency in the Americas in Chile and in the Iberian Peninsula among those Old World regions sampled.
Gepts has applied the phaseolin-structure method to questions of dispersal within the Americas, such as that of the traditional beans of the northeastern United States. There, a majority of the cultivated bean varieties of historic times are of the ‘T’ phaseolin type, which is the type that is most frequent in Western Europe and in the Andes south of Colombia. ‘T’ phaseolin is common elsewhere in South America but not in Mesoamerica. Indeed, in Mesoamerica and the adjacent southwestern United States, ‘T’ types make up only 8 percent and 2 percent, respectively, of the cultivated bean varieties. The archaeological record of crop plants in the northeastern United States is limited because of poor conditions for preservation (humid soils and no sheltered cave deposits), but those beans that have been found, although carbonized, are recognizable as a southwestern United States type (Kaplan 1970).This common bean type, which was dispersed from northwestern Arizona along with eight-rowed corn, must have been of the ‘S’ phaseolin type, which Gepts has found characteristic of 98 percent of contemporary southwestern common beans. It seems clear that historic-period northeastern bean cultivars are primarily South American, which could have reached the northeastern United States by way of sailing ships, directly from Peruvian and Chilean ports during the late eighteenth and early nineteenth centuries, or from England and France along with immigrants, or through seed importation.
In the foregoing, we see a dispersal pattern that was probably common in much of the New World, and especially in semiarid places in Mesoamerica and the greater Southwest. In dryland prehistoric sites, organic remains are well preserved in the archaeological record, and we see that prehistoric bean cultivars have often been eliminated, or reduced in frequency, by better-adapted (biologically, culturally, economically), introduced cultivars. Such a pattern suggests that Columbus’s introduction of beans to Europe from the Caribbean Islands was soon augmented by later introductions from Andean agriculture bearing the ‘C’ phaseolin type.
Historical evidence. Success in tracing the dispersion of beans from their regions of origin rests to some extent on historical records. But such records are likely to be strongly biased. One of the richest sources of evidence is the body of data available from seed catalogs, magazines, and newspapers devoted to agriculture and horticulture. Such publications, however, are unevenly representative of the larger dispersion picture. The United States, parts of Latin America, and Western Europe may be better represented in this respect than are some other parts of the world. Specialized libraries and archives have preserved some of this published material in the United States for about 200 years. In the United States, the earliest sources for named varieties of garden plants are leaflets or advertisements for seeds newly arrived from England. In Portsmouth, New Hampshire, over a period from 1750 to 1784, 32 varieties of beans, including common beans, both vining and erect types, scarlet runner beans, possibly small-seeded limas, and fava beans, were listed for sale in the New Hampshire Gazette (documents courtesy of Straw-bery Banke Museum, Portsmouth, N.H.). Earlier still, but less informative, are lists prepared for the guidance of colonists heading for English North America in the seventeenth century, who were advised to supply themselves with peas and (broad) beans, in addition to other vegetable seeds, garden tools, and weapons.We do not begin to detect evidence for the ingress of Phaseolus bean cultivars from England and France to the United States until the early nineteenth century.
The Old World: Broad Beans, Peas, and Lentils
As in much of the Americas, the Mediterranean world’s combination of cereal grain and food grain legumes has been the foundation for both agriculture and the diet of settled farming communities.
William J. Darby and colleagues (Darby, Ghalioungui, and Grivetti 1977), in tracing food use in Pharaonic Egypt, found papyrus texts and archaeological remains to contain much evidence of food grain legumes in the daily life of the kingdom. Rameses II spoke of barley and beans in immense quantities; Rameses III offered to the Nile god 11,998 jars of “shelled beans.” The term “‘bean meal’ [medicinal bean meal], so commonly encountered in the medical papyri,” could apply to Vicia faba, to other legumes, or to the “Egyptian bean” (Nelumbo nucifera Gaertner, the sacred lotus), all of which have been found in tombs (Darby et al. 1977: II 683).
Fava beans were avoided by priests and others, but the reasons are not clear. The avoidance of favas by Pythagoras, who was trained by Egyptians, is well known, and it may have been that he shared with them a belief that beans “were produced from the same matter as man” (Darby et al. 1977: II 683). Other ancient writers gave different reasons for the taboo, such as self-denial by priests of a variety of foods, including lentils. And more recently, as already mentioned, favism, a genetically determined sensitivity (red blood cells deficient in glucose-6-phosphate dehydrogenase) to chemical components of fava beans, has suggested still another explanation for the avoidance of this food (Sokolov 1984).
Structural change under domestication in fava beans, peas, and lentils is much like that in the Phaseolus beans, but there are important differences in what can be determined from archaeological samples. Pods of pulses in Middle Eastern sites are seldom found; hence, the loss of pod dehiscence has not been traced in archaeological materials from that region as it has in the Americas (Kaplan and MacNeish 1960; Kaplan 1986). In an effort to find good characters for distinguishing between wild and domesticated types in Middle East pulses, A. Butler (1989) has studied their seed coat structure but has found that the loss of seed dormancy (the impermeability of seed coats to water, which separates the wild types from the domesticates) cannot readily be detected by examination of the seed coats with scanning electron microscopy, or, at best, there is no single structural change that accounts for the difference between permeability and impermeability. Butler (1989: 402) notes that the surface of testa has been used to distinguish wild from cultivated taxa in Vicieae. With the exception of testa thickness in seeds of Pisum, no characters recorded in the seed coat of Vicieae can be associated directly with the presence or absence of hard seed coats. Evidence from seed anatomy of dormancy, and therefore of wildness, is lacking in Vicieae.
The origins and domestication of broad beans, peas, and lentils have been the focus of extensive research by plant geneticists and by archaeologists seeking to understand the foundations of agriculture in the Near East. D. Zohary (1989a: 358-63) has presented both genetic and archaeological evidence (not uncontested) for the simultaneous, or near simultaneous, domestication in the Near East of emmer wheat (Triticum turgidum ssp. dicoccum), barley (Hordeum vulgare), and einkorn wheat (Triticum monococcum), “hand in hand with the introduction into cultivation of five companion plants”: pea (Pisum sativum), lentil (Lens culinaris), chickpea (Cicer arietinum), bitter vetch (Vicia ervilia), and flax (Linum usitatissimum). V. faba may also have been among these earliest domesticates (Kislev 1985).
Lens culinaris (Lens esculenta), the cultivated lentil, is a widely grown species in a small genus. Archaeo-logical evidence shows that it or its relatives were gathered by 13,000 to 9,500 years ago in Greece (Hansen 1992) and by 10,000 to 9,500 years ago in the Near East (Zohary and Hopf 1988). Candolle (1964) in 1885 wrote of its presence in the Bronze Age sites (the so-called Swiss Lake dwellings) in Switzerland. Lentil cultivation, by historic times, was important and well established in the Near East, North Africa, France, and Spain (Hedrick 1972), and had been introduced into most of the subtropical and warm temperate regions of the known world (Duke 1981: 111).With a production of about 38 percent of the world’s lentils and little export (Singh and Singh 1992), the Indian subcontinent—a region in which food grain legumes are an especially important source of protein in the population’s largely vegetarian diet—has made this crop one of its most important dietary pulses. Traditionally, two subspecies are recognized on the basis of seed size: L. culinarisssp. microsperma, which are small-seeded, and L. culinaris ssp. macro-sperma, which are large-seeded. The large-seeded form is grown in the Mediterranean Basin, in Africa, and in Asia Minor. The small-seeded form is grown in western and southwestern Asia, especially India (Duke 1981: 110-13).
According to G. Ladizinsky (1989: 377, 380) the genus Lens comprises L. culinaris and four wild species: Lens nigricans, Lens ervoides, Lens odemensis, and Lens orientalis. The same author, however, notes that the genus may be reclassified to reflect the genetic relationships of the species, thus: L. culinaris ssp. odemensis, ssp. orientalis, ssp. culinaris; L. nigri-cans ssp. ervoides, ssp. nigricans. Many populations of ssp. orientalis and odemensis are sufficiently close to ssp. culinaris to be used for breeding purposes. However, the chromosome structure within these subspecies indicates that the cultivated lentils evolved from the principal cytogenetic stock of ssp. orientalis. Because this cytogenetic stock ranges from Turkey to Uzbekistan (Ladizinsky et al. 1984), there is little evidence of where in this vast region domestication might have first occurred.
Lentil domestication. Ladizinsky (1987) has maintained that the gathering of wild lentils, and perhaps other grain legumes, prior to their cultivation could have resulted in reduced or lost seed dormancy, a factor of primary importance in the domestication of legumes. As an operational definition of cultivation, Ladizinsky accepted the proposal of Jack R. Harlan, who emphasized human activities designed to manage the productivity of crops and gathered plants. As part of the gathering or foraging process, such management could lead to changes in genetic structure and morphology of wild plants.
High rates of seed dormancy observed by Ladizinsky in wild L. orientalis populations demonstrated that they are ill-adapted to cultivation. Dormancy, a single-gene recessive trait, is absent from most of the lineages of domesticated L. culinaris. Nondormancy is determined by a single dominant allele. A loss of seed dormancy in lentil populations resulting from gathering practices, Ladizinsky (1979, 1987, 1993) has argued, is the result of establishing this mutant dominant allele in small populations where most of the lentil seeds produced are removed by human gathering. A patchy distribution of small populations of annual legume plants, low population density, and intensive gathering would be, under the conditions he defines, advantageous for the increase of the seed-dormancy-free trait. He has concluded that this process would have taken place in wild, noncultivated lentil (and probably in wild pea and chickpea) populations, and would have predisposed these populations to domestication. This, he maintained, was an evolutionary pathway different from that of the cereal grains for which dormancy of seeds is not a barrier to domestication.
Ladizinsky’s views have been disputed by Zohary (1989b) and by M. A. Blumler (1991), who both criticized Ladizinsky’s assumptions and conclusions. Blumler investigated Ladizinsky’s mathematical model for the loss of dormancy in lentils and concluded that fixation of alleles for nondormancy, as a result of intensive human gathering, did not depend upon gathering at all. He further concluded that Ladizinsky’s model would predict loss of dormancy under any circumstances and therefore was not tenable as a hypothesis for preagricultural domestication. He went on to propose that lentils, and possibly other legumes that were gathered and brought to camps, might inadvertently have been introduced to campground soils, but that the poorly competitive introductions would have had to be encouraged (cultivated) in order to survive. Blumler thus contested Ladizinsky’s proposal that legumes could have been domesticated prior to actual cultivation and agreed with Zohary that the pathways of domestication in cereals and legumes were similar, and that legume cultivation followed upon the beginning of cereal cultivation.
As noted earlier in this chapter, the virtual absence of grain legume pods from archaeological sites in the Near East contrasts with the record of Phaseolus in the Americas. This circumstance makes it difficult to judge the merit of Ladizinsky’s view that the forceful dissemination of seeds following pod dehiscence—a process resulting in the loss of much of a valuable food resource—can be circumvented by pulling whole plants. This would be one sort of gathering technique that could be considered predomesticative management. Accordingly, Ladizinsky placed a low value on the loss of pod dehiscence, or shattering in legumes, while agreeing with most of those concerned that the loss of shattering in cereal grains is of high value in the process of domestication.
Zohary, in response (1989b), accorded the suppression of the wild type of seed dispersal in grain legumes equal weight with the parallel process in the cereal grains. In so doing, he reinforced the argument that domestication in the Near Eastern cereal grains and grain legumes were similar processes. These provocative models for the transition of legumes, and other economic plants, from the wild to the domesticated state provide the basis for both field and theoretical tests. As more evidence is gathered, these theories will be revised repeatedly.
Because archaeological legume pods are seldom found among the plant remains of early Near Eastern societies, dehiscence versus nondehiscence of pods cannot be determined as it has been from New World bean remains. The appearance of seed nondormancy as a marker of the domesticated legume has not been made on the basis of archaeological materials for reasons noted previously in connection with Butler’s study of seed coat structure. The archaeological record, however, does reveal changes in legume seed size in the Near East, such as small-seeded (diameter 2.5 to 3.0 millimeters) forms of lentil in early (seventh millennium B.C.) aceramic farming villages. By 5500 to 5000 B.C., lentils of 4.2 millimeters in diameter were present in the Deh Luran Valley of Iran, as reported by Hans Helbaeck (cited by Zohary and Hopf 1988: 89). Zohary and M. Hopf regarded the contrast in size between lentils of the early farming villages and those of 1,500 to 2,000 years later as a change under domestication.
Lentils spread into Europe with the extension of agriculture from the Near East in the sixth millennium B.C., where records of this crop and other legumes during the Bronze Age are less abundant than they are in the earlier Neolithic and later Iron Ages (Zohary and Hopf 1988: 87-9). Only their difference in size from the wild types (L. orientalis or L. nigricans) gives evidence for the development of domesticated lentils, which probably occurred in the Near East (Zohary and Hopf 1988: 91-2). These larger lentils appeared in the Near East as late as the end of the sixth millennium B.C., about 1,500 years after the establishment of wheat and barley cultivation (Zohary and Hopf 1988: 91). Although the archaeological and morphological evidence do not disclose the place of origin of the lentil, it is in the Near East where it occurs early in the large-seeded form and where L. orientalis is to be found. The lentil, however, is not mentioned as a food crop by F. J. Simoons (1991) in his comprehensive work on food in China.
Ever a source of confusion for readers of crop plant history and ethnobotany, common names often serve less to identify than to confound. “Pea” (P. sativum), as well as “bean,” illustrates this problem. A name with Latin and Greek origins, “pea,” in English, was formed from “pease,” which took on the meaning of a plural; the “-se” was dropped by A.D. 1600 and the current form of the word was established (Oxford English Dictionary 1971). Pea is used combined with descriptives to form the names of numerous plants. In the context of this discussion of food grain legume origins, the important distinction is that between P. sativum and the grass pea or chickling pea (Lathyrus sativus L.), and the cowpea, black-eye pea, or black-eye bean Vigna unguiculata.
The grass pea is a minor crop in the Mediterranean Basin, North Africa, the Middle East, and India, and has been documented archaeologically in parts of this broad region since Neolithic times (Zohary and Hopf 1988: 109-10). In India, it is a weedy growth in cereal grain fields and is the cheapest pulse available (Purse-glove 1968: 278). During times of famine, it has been consumed in significant amounts even though its consumption has caused a paralysis of the limb joints (Aykroyd and Doughty 1964).
The cowpea, an important crop in the tropics and subtropics, especially in Africa (Purseglove 1968: 322) and Southeast Asia (Herklots 1972: 261), appeared in the Mediterranean Basin in classical times (Zohary and Hopf 1988: 84). A plant mentioned in Sumerian records about 2350 B.C., under the name of lu-ub-sar (Arabic lobia), may have been a reference to the cowpea (Herklots 1972: 261).
Wild peas include Pisum fulvum, Pisum elatius (P. sativum ssp. elatius), and Pisum humile (P. sativum ssp. humile, = Pisum syriacum). Hybrids between P. sativum and P. elatius or P. humile are usually fertile, whereas hybrids between the cultivated pea and P. fulvum are fertile only when P. fulvum is the pollen parent (Ladizinsky 1989: 374-7). Zohary (1989a: 363-4) reports that similarities in chromosome structure between cultivated peas and P. humile populations in Turkey and the Golan Heights point to populations of humile having a particular chromosomal configuration (a reciprocal translocation) as the “direct”ancestral stock of the cultivated peas.
Peas were present in the early Neolithic preceramic farming villages (7500 to 6000 B.C.) of the Near East, and in large amounts with cultivated wheats and barleys by 5850 to 5600 B.C. They appear to have been associated with the spread of Neolithic agriculture into Europe, again together with wheat and barley (Zohary and Hopf 1988: 96-8). In central Germany, carbonized pea seeds with intact seed coats have been found dating from 4400 to 4200 B.C., and these show the smooth surface characteristic of the domestic crop. In Eastern Europe and Switzerland, pea remains have been found in late Neolithic or Bronze Age sites (Zohary and Hopf 1988: 97-8).
The date at which the pea entered China is not known, but along with the broad bean, it was called hu-tou or “Persian bean,” which suggests an origin by way of the Silk Road in historic times (Simoons 1991: 74-5).
Ladizinsky (1975) has investigated the genetic relationships among the broad bean and its wild relatives in the section Faba of the genus Vicia. He concluded that none of the populations of wild species, Vicia narbonensis, Vicia galilaea, or Vicia haeniscyamus, collected in Israel, can be considered the ancestor of the fava bean. Differences in the crossability, chromo-some form, and chromosome number separate the broad bean from these relatives, as do differences in ranges of seed size, presence of tendrils, and epidermal hairs on the pods. Ladizinsky was careful to note that these characters may have evolved under domestication, but given the absence of evidence for derivation of the fava bean from its known relatives, he concluded that the place of origin of the broad bean could not have been in the Middle East.
Ladizinsky further drew attention to what he regarded as a paucity of broad bean remains only “circumstantial[ly]” identified in Neolithic sites of the Middle East and was led by these lines of evidence to conclude that the origin of V. faba had to be outside of this region. He further speculated that the occur-rence of a self-pollinating wild form of the species in Afghanistan and areas adjacent might be the region in which the now cross-pollinating broad bean originated. Zohary (1977), however, interpreted the archaeological record quite differently, asserting that the distribution pattern of archaeological broad bean remains points to its domestication by the fourth or fifth millennium B.C. in the Mediterranean Basin.
The dates for the introduction of broad beans into China are not secure—perhaps the second century B.C. or later—but it has become an important crop in “many mountainous, remote or rainy parts of China at the present time especially in western China” (E. N. Anderson, cited by Simoons 1991: 75). The earliest introduction of broad beans into North America appears to have taken place in the early seventeenth century when Captain Bartholomew Gosnold, who explored the coast of New England, planted them on the Elizabeth Islands off the south shore of Massachusetts (Hedrick 1972: 594). Less than 30 years later, records of the provisioning and outfitting of the supply ship for New Plymouth list “benes” (and “pease”) along with other seeds to be sent to the Massachusetts colony (Pulsifer 1861: 24).
The domestication of food grain legumes in the Americas and in the Mediterranean-West Asian region reveals important parallels. Both groups suppressed, at least in some varieties, the tendency to vine or grow rampantly. The seeds of both have markedly increased in size over their wild counterparts. Both groups have suppressed pod dehiscence, projectile seed dissemination, and seed dormancy, and in both regions, the seeds have functioned primarily as protein sources in predominantly cereal grain diets. Studies in genetics, molecular structure, and archaeology have contributed to an understanding of the origins of species and races within species. Nonetheless, uncertainties over important aspects of the origins and evolution under domestication remain, and are the subject of active multidisciplinary research.