The taxonomic revision of Maréchal et al. (1978) has brought within the confines of a single genus crops which had previously been distributed over three genera: Vigna itself, the genus Voandzeia (now merged with it) and Phaseolus (re-defined). The genus, in the revised sense, is distributed in the warmer parts of both the Old World and the New; all the cultigens are of Old World origin although the cowpea and the mungbean are now more widely distributed.

The members of the genus commonly cultivated are the cowpea itself (Vigna unguiculata), the groundbean or Bambara groundnut (Vigna subterranea) and the group of Asiatic forms (commonly known as grams) Vigna radiata, V. mungo, V. angularis, V. umbellata, V. aconitifolia, V. glabrescens and V. trilobata. The last two are of peripheral interest; V. glabrescens is a polyploid and V. trilobata can perhaps best be regarded as a semi-domesticate.

The extent to which these different species are exploited has been determined largely by economic factors. The cowpea is the subsistence legume par excellence; it is grown extensively in Africa and India in the Old World, and in Brazil in the New (Rachie, 1985). It is only in the USA that the cowpea is primarily a commercial crop, where it is grown to meet the demand for ‘black-eyed peas’ (dry mature seed). Green mature seed is also used both canned and frozen.

At the present time there is a very lively interest in oilseed crops as a whole. This arises from a number of causes but principally the fact that the supply of animal fats and oils has been reduced as a consequence of conservation measures taken in the interests of whale populations. Perhaps more importantly, the role of saturated animal fats in human diets and nutrition has been called in question. This has led to an increased demand for polyunsaturated plant edible oils, which the supply of long established traditional oilseed crops has not been able to meet. This has meant that oils which twenty years ago were used almost entirely in industry are now extensively used in the food trade. The longerestablished traditional oils, from the olive, groundnut and sesame, have been substantially replaced in part at least by a later generation of oilseeds: sunflower, soyabean, rapeseed, mustard and cottonseed, for example. Production of the traditional oils could not keep up with increased demand; there seems to be little prospect, for example, of a substantial increase in olive oil production. The very high culinary quality of oils such as olive and groundnut oils and their high cash value has led to problems of adulteration. Adulteration is not always easy to detect and control, with the increased tendency for oilseeds to be expressed in the country of origin.

The grain legumes which evolved in the Mediterranean basin have a particular claim to the attention of students of crop plant evolution. Not only have they played an important supporting role to that of the cereals in sustaining the development of the classical civilisations of the area, but it is arguable that the scientific study of crop evolution began here. The crops of the Mediterranean region were among the most familiar to Linnaeus (1753) and to de Candolle (1886), the father of the scientific study of crop origins. De Candolle appreciated that non-biological disciplines could contribute valuable information on the evolutionary history of crop plants. For example, records of contemporary crops in classical writings such as Virgil's Georgics are readily accessible and useful. Representations in art and artefacts are equally valuable. The evidence shedding light on crop histories varies widely from crop to crop in quantity and quality. This depends on the significance of the crop in economic, social and religious life, and also on what records or materials have survived and been discovered. The evidence, although from a wide variety of sources, can only be fragmentary but it can nevertheless be very informative. In any event we have, perforce, to do the best we can with it.

In the present treatment I propose to treat each crop individually in the first place and then conclude with a brief comparative consideration of this group of pulses.

Research on grain legumes is now coming more to the forefront in the world as a whole than at any time in the past. This is reflected in the number of monographic treatments of individual legume crops which have appeared in the past decade. Forty years ago one would have been able to unearth rather little information on grain legumes generally and very little indeed specifically on tropical grain legumes. In present circumstances it still can be argued that the market is not as yet oversupplied with literature on the legumes. Be this as it may, it nevertheless behoves any author setting forth his labours before the public to define his aims and objectives as clearly as possible for the benefit of the potential reader. Most writers on technical subjects perhaps feel the urge to write the book they would themselves need if they were embarking on work in the field it covered. There has been a tendency in all this new writing for treatments to become very detailed and specialised. The rise of the multi-authored tome has been irresistible, with all this convergent thought; perhaps there is scope and a need for thinking on a broader front, in a more lateral vein.

My own contact with legumes as research material goes back to my student days, some thirty-five years ago. I found them to be problematic but fascinating then as I still do. They can be rewarding to work with but the rewards are often hard-won.

Although the grain legumes have shown remarkably similar patterns of evolutionary response to the selection pressures which have operated under domestication (Smartt, 1976a, b, 1978a, 1980a), the genetic resources which are available for future conscious, man-directed evolution of pulse and legume oilseed crops are very different in both their nature and extent for the different species. The present time is crucial for collection and conservation of crop genetic resources; in the case of grain legumes no comprehensively consistent or coherent strategy has as yet evolved, but might well do so following the guidelines of Ford-Lloyd and Jackson (1986). There could be a considerable gain in the effectiveness of this effort if there were to be more overt rationalisation, co-ordination of conservation activity and the adoption of consistent procedures.

The work of Harlan and de Wet (1971) gives a very sound foundation on which to base genetic resource conservation strategies. Some modification of their approach may be necessary for grain legumes owing to the distinctly different pattern of biosystematic relationships found in the Leguminosae as compared with the Gramineae on which their work was largely based. On the whole, inter-specific and even inter-generic hybridisation is more common in the grass family than in the legumes, and development of polyploidy is much greater in the grasses, even though both major oilseed legumes (groundnut and soyabean) are in fact polyploid (tetraploid).

The major stimulus to detailed studies of the biosystematics of the genus Arachis, to which the groundnut belongs, came directly from practical interest in the crop. This work, initiated in the United States and Argentina, is associated with the names of Gregory, Krapovickas and their co-workers. Although it has not yet reached a definitive stage, sufficient has been published to provide a sound and effective biosystematic framework within which to consider the origin, evolution and germplasm resources of the crop. Indeed, few legumes have been more effectively investigated by what is a surprisingly small number of individual workers. Data and observations from a variety of sources have been integrated to produce a workable taxonomic scheme for the groundnut and its relatives. Morphological evidence is of course paramount, but this has been very effectively supplemented by studies of experimental hybridisation, comparative cytology, cytogenetics and biochemistry. Arguably this has been more effective for the groundnut and its allies than for any other grain legume; as a result, a very satisfactory taxonomic synthesis is emerging.

Biosystematics of Arachis

The present state of Arachis taxonomy

The genus Arachis is morphologically well defined and clearly delimited from its closest relatives by the development of a peg and its geocarpy.

If one were writing a school report on grain legumes the observation ‘could do better’ might very well be apposite, with all that this remark implies. The major preoccupation of many engaged in legume research is to find out why these crops in their performance so often fall far short of our expectations. There is some encouraging evidence that, in favourable conditions, some grain legumes at least can in fact perform very well indeed. The nadir of grain legume performance, as far as many people are concerned, is undoubtedly the notorious fiasco of the Overseas Food Corporation's East African Groundnut Scheme of the late 1940s and early 1950s. Curiously enough, it is also the groundnut which has given the clearest indication of what might lie in the realms of future grain legume production. The highest recorded groundnut yields (9.6 t ha–1 pods, equivalent to 6.41 ha–1 kernels) have been obtained in Zimbabwe with the cultivar Makulu Red (Hildebrand and Smartt, 1980). Interestingly, this cultivar was not the product of a long-drawn-out and expensive breeding programme but obtained as a single plant selection from a Bolivian landrace ‘maní pintado’. Agricultural improvement depends as much on recognising opportunities such as this and exploiting them efficiently, as on the execution of complex research programmes. Unfortunately it is the more grandiose schemes, not the most cost-effective, which make the headlines and attract the attention of the politicians.

Constraints peculiar to forest canopy habitats helped shape epiphyte natural history, but the selective forces are not always apparent. Some characters and lineages appear to have been affected more than others; for instance, iteroparity is nearly routine in the group, whereas breeding mechanisms and modes of pollen and seed dispersal are much more diverse. Multiple paths to similar ends complicate the search for a common theme. Enough data are available, however, to offer tentative judgments on several aspects of the plant life cycle that permit success in tree crowns. This chapter deals with such aspects and the factors responsible for their existence.

Breeding systems

Pollination: identity of vectors

If breeding systems differentiate terrestrial from canopy-based pteridophytes, the fact remains unreported. Comparisons of angiosperms are easier, and pollination has been studied in numerous epiphytic flowering plants, especially neotropical Orchidaceae. Pollinators of these taxa tend to be more species-constant and specialized than those serving nearby terrestrials, although sharing of pollinators is sometimes possible. Avians are especially important in northern South America, where nectar-feeding birds and the flora they serve reach unparalleled diversity. Large, heavily ornithophilous families include Bromeliaceae, Ericaceae, Gesneriacese, and Loranthaceae; birds frequently pollinate Cactaceae, Marcgraviaceae (Norantea), and Rubiaceae (Ravnia, Manettia) as well. The relationship between epiphytism and avian pollination is particularly apparent in families with diverse floral syndromes. Ornithophily in Bignoniaceae is rare except in two epiphytic genera: Gibsoniothamnus is entirely pollinated by birds, Schlegelia partly so (Gentry and Dodson 1987b).

Plant nutritionists have paid little attention to epiphytes, concentrating instead on plants rooted in earth soil. Consequently, consideration of mineral cycling in tropical woodlands has seldom taken into account the potentially major impact of arboreal vegetation. Reports from a scanty but developing literature on epiphytes, together with relevant aspects of plant physiology in other groups, provide a basis for the following chapter. Epiphytes will be portrayed as plants that not only tap a variety of nutrient sources (at times with novel absorptive organs) but also are significant players in the nutrient and energy economies of many tropical forests.

Nutritional categories

All higher plants require at least six macronutrients and seven trace elements for growth (Table 4.1). Some taxa supplement these basic 13 with others that support out-of-the-ordinary functions. For example, many grasses produce silicon-containing granules that reduce palatability to vertebrates; selenium (Se) can act as a sulfur (S) analog, helping to ward off herbivores; some halophytes and all C4 and CAM taxa require sodium (Na); and so on. Descriptive epithets are applied in certain cases: halophytes versus glycophytes, to describe occurrence on hyperosmotic media; calcifuges versus calcicoles, to specify calcium (Ca) content of native soil; eutrophs and oligotrophs, to distinguish quantities of key macronutrients needed.

Eutrophs are characterized by several features. Critical concentrations of foliar N and P (those levels required to maximize growth) tend to be elevated (Table 4.1). Shoot/root ratios are generally high, life cycles and longevity of leaves, brief. Mature size and vigor are tied closely to nutrient supply.

This final chapter concerns epiphyte occurrence in three contexts: global, taxonomic, and ecological. First, global and taxonomic patterns and the question of why epiphyte floras are unevenly developed throughout the tropics are addressed. Second, the effects of climate, topography, and soil fertility on species range and abundance are considered. Finally, hypotheses are offered to explain why certain plant lineages have been more successful than others in forest canopy habitats.

Distribution: taxonomic and geographic

At the higher taxonomic levels, epiphytes are diverse; excluding the mistletoes, 84 families, including 69 in Magnoliophyta, contain qualifying taxa. But rather few major clades account for most of the species; just 23 families harbor about 98% of the total flora in 87% of the epiphytic genera (Tables 1.1, 1.2). Fifteen families include but a single epiphyte; 52% of the 871 epiphytic genera contain five or fewer species, and about half of those contain only one. Heaviest contributors are Araceae, Bromeliaceae, Ericaceae, Gesneriaceae, Melastomataceae, Piperaceae, Orchidaceae, Rubiaceae, and several fern families. Forty-three genera each contain more than 100 epiphytic species (Table 1.2): of the 43, 22 are orchids, 8 are ferns, 4 are bromeliads, 3 are from Aracaceae, and the remaining 6 are contributions from five additional families.

Geographic asymmetry is also considerable, especially in the more advanced taxa (Madison 1977). Of the 86 canopy-adapted fern genera listed by Madison, approximately two-thirds are pantropical; the remainder are divided about equally between the two subregions – 16 exclusively neotropical and 14 paleotropical.

There are about 25,000 vascular species sharing the peculiar habit of rooting in tree crowns rather than on the ground, yet only an occasional epiphyte – the wide-ranging Spanish moss, for example – has attracted much scientific curiosity. Uncounted thousands of animal populations (mostly insects) regularly associate with these plants, sometimes because there are no alternatives for lodging, food, or other critical resources. Vascular epiphytes remain best known to horticulturists and systematists; the how and why of their growth in nature under such novel conditions have been mostly ignored. Other ecological groups such as carnivores, halophytes, mangroves, and parasites have been thoroughly covered in monographs despite their smaller numbers, more restricted distribution, and limited literature base. Also underrepresented is information on epiphyte-dependent fauna and effects on supporting trees. But times are changing. Improved climbing techniques allow extended observation and collection of representative fauna. Portable equipment for measuring such plant phenomena as gas exchange has opened the upper canopy to sophisticated analyses. Clearing land for roads, while destroying woodland, has fostered research in the field; so has establishment of permanent field stations, particularly in the neotropics. Results are heartening; what for many years was only a trickle of papers on nontaxonomic aspects of epiphyte biology and forest canopy fauna now approaches a flood. Three international symposia devoted to epiphytes have been held in just the past four years.

Historical notes

Columbus is credited with the first recorded comment on canopy-adapted vegetation; he wrote that tropical trees “have a great variety of branches and leaves, all of them growing from a single root” (Gessner 1956). The earliest known picture of an epiphyte – or, for that matter, reference to American botany – appears in The Badianus Manuscript, a Mexican herbal of 1552, written and probably illustrated by the Aztec Indian physician Martinus de la Cruz and translated into Latin by his Indian colleague Juannes Badianus at the College of Santa Cruz (Emmart 1940). The subject was Vanilla fragrans, a vining hemiepiphytic orchid. The fruit of this species (tlilxochitl in Aztec, meaning “black flower”) was an ingredient in the doctor's prescription for “The Traveler's Safeguard,” a mixture of pulverized herbs wrapped in a magnolia leaf and hung around the neck so that the voyager could “catch and inhale the very redolent odor …”

By the eighteenth century, ships' captains and explorers the world over were carrying ornamental plants, epiphytes included, back to Europe. Within decades, a brisk trade had developed; many additional aroids, bromeliads, cacti, orchids, and ferns were imported. Showiness, small size, and easy culture encouraged fads that drove prices to exorbitant levels and prompted more than one collector to lie about where he found his specimens. But scientific interest in these plants did not keep pace; other groups such as carnivores, halophytes, ruderals, and succulents are far better known today.

Mistletoes are unique enough among canopy flora to merit separate treatment. Certain relic terrestrial forms parasitize roots of other plants, but they will be mentioned only in passing; the principal focus will be on aerial mistletoes which are here defined as shrubby hemiparasites growing attached to branches. These unusual plants deviate from “true” epiphytes in form, diversity, physiology, and impact on hosts. Most mistletoes belong to Santalales, a sizable, predominantly tropical, order. Xylem rather than phloem supply is reputed to be the usual consequence of santalalean parasitism, but, as noted later, advanced forms as well (e.g., Arceuthobium) take host substrates. Mistletoes have long occupied a place in European folklore and continue to figure prominently in certain holiday rituals of the Western world. Their destructive qualities are widely recognized. Fortunately, enough scientific curiosity has been aroused by these remarkable organisms to encourage a hard look at their biology. In fact, vegetative and reproductive activity is better known for these plants than for any other like-size assemblage of forest-canopy residents. In this chapter, that information is summarized and aerial mistletoes are contrasted with the true, fully autotrophic, epiphytes.

Systematics and biogeography

The mistletoe habit is polyphyletic, having arisen at least three or four times in Santalales and again in Laurales. The largest mistletoe family is santalalean Loranthaceae with some 900 species distributed unevenly among about 65 genera. Second in size and much more uniform in floral structure is Viscaceae, a group of perhaps 400 species in just seven genera.