logo 2
   
08 Sep2003

Overstory #128 - Wild Foods in Agricultural Systems

Written by Ian Scoones, Mary Melnyk and Jules N. Pretty.

Introduction

Throughout the world, agricultural systems are potential sites for a great diversity of managed and collected plant and animal foods. Conventional agricultural research and extension, by focusing only on the main food crops, chiefly cereals, roots and domesticated livestock, have long ignored the range of other plants and animals that also make up agricultural systems.

Studies of this diversity and attendant complexity are demonstrating the importance of understanding the full range of products harvested. For instance, in the apparently maize-dominated agricultural system of Bungoma in Kenya, people consume at least 100 different species of vegetables and fruits drawn from 70 genera and belonging to 35 families (88; 729). Similarly, the agropastoral Tswana people in Botswana use 126 plant species and 100 animal species as sources of food (261). Similar patterns are shown in SE Asia (179; 132; 152; 188), Himalayan areas (160), Central America (682), Latin America (2; 3; 126), northern Europe (866; 867) and elsewhere in Africa (47-49; 82; 273; 400a; 493; 494; 770).

Studies of settled agriculture worldwide, whether in semi-arid, temperate or humid settings, on the plains or in mountain areas, show that hunting and gathering remains an important component of the livelihoods of agricultural peoples. There is no progressive evolutionary trend of 'development' from hunter-gatherer to small-scale settled agriculture and livestock keeping to intensive agricultural systems. Livelihood strategies in all social, economic and ecological settings encompass a wide range of activities. These include the exploitation of a hidden harvest of wild food sources. Diversification of livelihood strategies, combining agricultural sources of food and income with that derived from wild resources, is particularly important for the poorest households, and for women and children (48; 49; 608-630).

Agroecological change

As agroecosystems change through expansion of cultivated areas or changes in cropping patterns, so the availability of wild foods alters. As woodlands are cleared, new edible weeds and pests linked with agricultural lands appear as other foods associated with the woodland ecology disappear (170; 171 for southern Zimbabwe). The simplification of agroecosystems, such as in the conversion of forest areas to cattle pastures in Brazil and the intensification of small-scale agriculture through Green Revolution technologies worldwide, has the greatest impact on the poor, as key sources of food are lost (76).

Although the greatest diversity of wild foods are found in multi-layered, complex agroforestry systems and home gardens, wild foods are still important in apparently simple, monoculture systems. For instance, canals feeding extensive rice areas are habitats for fish, frogs and plants associated with excess irrigation water (4; 69; 152). Similarly, other forms of intensive agriculture may harbour particular pests such as rats, mice and locusts, which may be eaten. Intensification of agriculture, with increased use of pesticides and fertilisers, may have a negative effect on the wild food crop by killing off the potential foods 'pests' or 'weeds'. However fertiliser inputs may increase the prevalence of certain species (152).

Wild foods are not only associated with undisturbed systems that replicate the ecological diversity of the uncleared forest, they are also found in degraded sites. Sometimes disturbance increases the diversity of wild products (173). For instance, in Kenya and Tanzania, the greatest prevalence of wild vegetables was found in gullies caused by erosion on farmland (90; 173). Pathways, roadsides, home sites and field edges are also potential sites for wild products; sites which otherwise might be considered valueless. Areas that are logged within forests may become the site for mushroom fruiting (861).

More often, the diversity of wild foods declines during the conversion of complex woodland to simplified cropped land. For instance, in three Tanzanian villages there is a correlation between the diversity of edible plants being eaten and the degree of deforestation (400a). Similarly, twenty years of agricultural change in Kenya have had a major impact on Mbeere wild food collection and use strategies, causing them to use fewer wild food sources, because of reduced access to bush land as a result of land privatisation (28-30; 137).

As agricultural systems change there are new pressures on wild food resources. One response is actively to domesticate the wild foods. Vegetables and fruit trees, formerly harvested in the forest or grazing areas, are increasingly protected or planted (1). In north-west Uganda valuable weeds in the diet of farmers have started to be cultivated within the home compounds (156; 729 for western Kenya). Similarly bush-fallow systems may be transformed by enrichment planting (108), such as the planting of fruit trees in fruit-poor Acacia fallow areas (39).

Agroforestry systems

Farmers have always incorporated trees into farming systems. An increased interest in agroforestry during the past decades has resulted in a more thorough documentation of agroforestry practices from Africa (1; 24; 56; 89; 105; 114; 138; 139), home gardens in SE Asia (40; 62; 102-104; 116; 151a; 151b; 168), kitchen gardens in the West Indies (27) and agroforestry systems in Latin America (2; 3; 9; 53; 58; 99; 117-119). The complexity and diversity of many managed agroforestry systems is immense (23).

The inclusion of trees as a component of the agricultural system increases the structural complexity of the field environment, provides a degree of complementarity in seasonal and interannual production patterns with annual crops and changes the labour commitments to an area of land. Retaining trees on farm land during clearance for agriculture or subsequent planting of trees produces a range of ecological habitats and seasonal niches ideal for wild food production.

During the selective clearing of land for agriculture, farmers usually retain particular tree species. These are often fruit trees, the providers of seasonally important wild foods. In Zimbabwe, clearing of Julbernadia globiflora dominated miombo woodland reduced canopy cover from 52% to 8%. But the canopy cover of fruit trees was only reduced from 7% to 1%. Some favoured fruit tree species, notably Diospyros mespilliformis, Strychnos cocculoides and Azanza garkeana, showed no reduction in cover when forested areas and cultivated areas were compared. Despite widespread deforestation for land clearance in one site, patterns of fruit use in two agricultural areas were similar because trees had been selectively retained in cleared fields during woodland clearance (36).

In other settings, trees within farming systems have been enriched by planting. The complex multi-storey home garden systems found in Indonesia (104), Mexico (7), Tanzania (63), Kenya (110), tropical America (165) and elsewhere (56; 65; 101; 102) are examples of intensively managed multi-species systems. Within such gardens wild foods can be found occupying a diversity of ecological niches.

Wild fruit trees may be collected from forest areas and planted in agricultural land to enrich on-farm tree species. This has been recorded in many instances (89; 126; 132). The fruit producing potential of wild cultivars can also be upgraded by simple grafting and breeding techniques (115).

Home gardens

Complex home gardens have been described in West Sumatra (104) and Java (102; 151a) in Indonesia. Here a range of annual and perennial crops are grown together, complementing the main rice crop derived from other fields. In Java, home gardens containing 500 species are found within a single village (102). There are several different types of garden, including the intensively managed home garden, the village/forest gardens and the forest fringe gardens. The importance of wild foods increases in gardens towards the forest fringe; these gardens resemble more closely the ecological conditions of the forest itself. In Western Sumatra a range wild foods including 22 fruits, 8 vegetables and spices and 3 fern species are protected and harvested in different garden types, while in migrant communities' gardens in Mexico some 338 species are found (7). Home gardens are also important as a site for experimentation with new varieties, domestication attempts and evaluation of different cropping rotations and patterns (128; 188; 243).

Weeds as food

Trees are not the only component of the agricultural system that are potential sources of food. Along with the major crops planted by the farmer, a range of plant material can be found in agricultural fields that represent potential food (25; 70; 90; 111-113; 175; 703b). These wild foods (vegetables, tubers, grasses) may be potential competitors with the major crop, but whether they are weeds depends on the observer. To many agronomists anything but the major crop itself is regarded as a weed, and so the monocrop ideal (or at most, simple intercropping) is preached by agricultural extension workers throughout the world. Yet many plants deemed 'weeds' may have a variety of uses to local people. To a woman attempting to find cheap and nutritious ingredients for relish, the wild food resource found in agricultural lands may be critical.

Many studies document the importance of wild vegetables in local diets; many of these are available from farmlands. A particularly extensive set of literature exists for east and southern Africa (e.g. 79; 86; 135; 143 for Zambia; 111-113 for Swaziland; 171; 172 for Zimbabwe; 98, 139 for Kenya; 90 for Tanzania; 50; 148; 149 for southern Sudan; 96 for Zaire). Most studies note that it is women who are primarily engaged in the collection and management of wild vegetables in Africa (139). Similar findings are reported from elsewhere (876 for Assam; 152, 153 for SE Asia; 528 for Mexico).

Plants collected from fields may be either managed or simply left to grow. Four categories of weed have been identified in southern Sudan: self-sown species, wild species whose seeds are collected and scattered in the fields, those collected as they appear and those eaten only when under severe food shortage (148; 149).

Pests as food

Arable lands also attract certain pest species. Rodents tend to be at a higher density in fields compared to surrounding areas (51). The abundance of rodent species thus often changes with agricultural clearance (e.g. 57 for the New Guinea Highlands). Rodents of various sorts are an important source of food in agricultural communities (317; 747). In Zimbabwe, roasted mice fetch a high premium as a snack food at beer parties (172). The African giant rat (Cricetomys gambiani) is also important in southern Nigeria (382; 286). Changes in bushmeat availability have occurred as a result of land clearance in West Africa with increasing rodent hunting possibilities, e.g. of the grasscutter rodent, Thyrononmys swinderianus (292).

Insects also represent an important dietary component in many agricultural societies (21; 54; 140; 547). Such insects may also be crop pests, but their role as supplementary food is well documented. For instance, termites are an important fat and protein supplement across southern Africa (97; 109; 172) and in southern Sudan (148; 149). Caterpillars (e.g. Gonimbrasia belina - the mopane worm) are also widely eaten and marketed (26; 106). In Zambia there is a wide range of edible insects in the diet (159). This has a seasonal dimension. The importance of caterpillars may rise to 40% of relish items in the period November to January (159). Locust or cricket swarms can also provide important additions to local diets (45; 100; 155).

Hunting and gathering

There are few groups who can sustain livelihoods solely on the basis of hunting and gathering from wild resources (19). Most 'hunting and gathering' communities have some food inputs from arable agriculture or livestock, either from their own plots (48; 49; 65) or through trade and exchange with farming communities (55; 212; 236).

Hunting and gathering is dependent on a diverse source of products that can offset seasonal and interannual variability in wild food production (11; 49; 67; 174). Alternatively a highly reliable and plentiful food source must be available. For instance, the !Kung San rely on mongongo (Ricinodendron rautenenii) as their major food source, with two to three days' supply gathered at any one time (20; 93; 94; 154; 415). Together with the hunting of a variety of wild animals and the collection of a range of wild plants, this is a highly labour and energy efficient survival strategy for the harsh environment of the Kalahari (92-94; 415). Similarly, in the past, Australian aborigines were able to collect a day's food from the bush in two to four hours (259; 260). In the Fertile Crescent of the Middle East, highly nutritious wild wheat and barley provided food for people before the origins of agriculture some 10,000 years ago (134; 263; 279b, 264). With such an abundance of wild produce, the alternative investment of land, labour and capital into the domestication of plants and animals with settled agriculture and livestock keeping appears to make little sense.

Patterns of use: seasonality, regulation and sustainability issues

Wild foods in agricultural systems often fit a particular seasonal niche. They may provide green vegetables early in the rainy season, or can be dried and stored for the dry season. They may also provide counter-seasonal food with fruit bearing in the dry season when little else is available. Wild foods may be particularly important in years when harvests fail (49; 170; 400a-d; 484; see 374-452).

Wild foods may only be available occasionally. For instance, insect outbreaks may only be sporadic (or cyclical) or mushroom fruiting dependent on particular (rare) conditions. Diets changed rapidly in a Sudanese village following locust swarms (45), and in Kenyan agricultural areas after rat outbreaks (157). However when such events do occur, as in extensive mushroom fruiting, labour may be diverted away from normal agricultural activities to collection and marketing (121 for southern Zaire; 123 for Zambia; 861 for northern Thailand; 146 for north India).

Collection and consumption of wild foods is often differentiated between socio-economic groups and gender (see 608-630). Women are primarily engaged in wild food management and harvesting, particularly of green vegetables (90; 113). Wild foods are also important nutritional supplements for children (139; 623 for Kenyan case material; 617 for forest edge communities in Sierra Leone; 125 for central Indian tribal areas). These may be eaten as snack foods or as main meals. The importance of wild foods is greatest amongst poorer households, where main field crops are often insufficient to provide food for the family for the whole year. In a dry miombo area of Zimbabwe poor households use fruits as the alternative to grain for a quarter of all dry season meals (172).

The use of wild products may be regulated by local rules and institutions (see 569-607). Large fruit trees in farmland are often protected (170 for Zimbabwe; 129 for West Africa) by local communities. Rights over wild products may change as land is cleared for agriculture. In Zimbabwe, fruit trees retained on farm plots effectively become individually owned by farmers during the cropping season, although they may revert to common property in the dry season (but still are protected from cutting by community rules). In other settings all products are privatised by the process of conversion of land to agriculture, reducing the access of those without control over land. Recently, Malawian farmers have sold the rights to the collection of wild resources to Mozambican refugees (171).

Few studies, however, have addressed the degree of dependency on these food sources or their economic value as part of agroecosystems and peoples' livelihoods (see 631-702b). As a consequence, it is difficult to assess the impact of patterns of land-use and land tenure change on different groups of people.

The sustainability of wild food use has also received relatively little attention in the literature. Sustainable harvesting levels for different plant and animal populations remain largely unknown. A number of studies report that wild foods are diminishing with the clearance of forest areas (e.g. 152 for NE Thailand; 90; 400a for Tanzania) or the heavy harvesting of wild animals (108; 284; 353). The consequences of agricultural intensification on wild food production are also poorly studied (152).

Unlike large game animals (342), small animals, especially rodents, may be heavily harvested without affecting the viability of the population (170; 353). They are also less susceptible to changes in agricultural land-use and agronomic practice. Indeed rodent populations increase with arable land expansion (292). The same applies to weedy plant species which quickly regenerate following collection and can survive in ephemeral environments on field edges or degraded lands (170; 172; 628). This is in contrast to many fruit tree species which are less resilient to agroecological change, as they may take many years to regrow to maturity.

The study of wild foods in agricultural systems requires an interdisciplinary approach that can examine the role of wild foods in the context of local people's livelihoods.

Literature cited

Due to the extremely large number of references cited in this article (well over 200), the reader is referred to the original source for cited literature. The citation numbers in the text correspond to the numbers in the bibliography of the original source. See below under Original Source for the citation and information about where to purchase the book.

Original source

This article is excerpted with the generous permission of the publisher and authors from:

Scoones, I., M. Melnyk and J. Pretty. 1992. The Hidden Harvest: Wild Foods and Agricultural Systems, A Literature Review and Annotated Bibliography. IIED, London.

To order this publication contact:

EARTHPRINT Ltd. P.O. Box 119 Stevenage Hertfordshire SG1 4TP England Telephone: +44 1438 748 111 Fax: +44 1438 748 844 Email: customerservices@earthprint.com

About the authors' organization

This article was authored by Ian Scoones, Mary Melnyk and Jules N. Pretty, and published by the Sustainable Agriculture and Rural Livelihoods (SARL) Programme of the International Institute for Environment and Development (IIED). The SARL Programme seeks to promote sustainable, equitable, decentralised agri-food systems based on local diversity and participatory democracy, thereby contributing to improved livelihoods and entitlements, poverty reduction, and long-term ecological and economic sustainability. By working to develop more effective and equitable forms of agriculture and natural resource management, the SARL Programme helps different interest groups to understand trade-offs relating to their livelihood strategies, identify common ground, and negotiate pathways to positive actions that support rural regeneration.

Established in 1986, the aim of the SARL Programme was to provide key policy makers, project designers, and rural development practitioners with concepts, tools, and methods to put into practice the challenges facing sustainable agricultural development. Since then, the SARL approach has evolved, and has emerged out of a growing recognition that sustainable agriculture and rural development cannot be treated in isolation from broader ecological, economic, social, and political processes. It is these broader processes, particularly counterproductive and inappropriate policies, and weak and ineffective institutions, which are limiting the spread of sustainable agriculture and the regeneration of rural economies.

Sustainable Agriculture and Rural Livelihoods Programme International Institute for Environment and Development 3 Endsleigh Street London WC1H 0DD United Kingdom Tel: +44 (0) 20 7388-2117; Fax: +44 (0)20 7388-2826 Email: sustag@iied.org.

Related editions of The Overstory