Impact of a women's garden project on nutrition and income in Senegal Jacqueline Reynaud, Thierry Brun, and Tonia Marek During the last forty years, several development agencies such as the Food and Agricultural Organization of the United Nations, the United States Agency for International Development, and the Inter- national Development Research Center, have supported the introduction or extension of vegetable gardening in the tropics. The aim of most garden projects has been to improve the nutritional status or diversity of the diet of certain target groups and to raise the income of the vegetable growers through the sale of produce. Although a large number of such projects have been initiated (1), very few have been evaluated and it is not clear to what extent home gardens fulfill these expectations and have had a measurable impact on nutrition and income in the project areas. It appears to be generally assumed that there is a significant link between increased production of vegetables, increased income, and improved nutritional status. The complexity of the relationship between those potential changes, as well as many other aspects of garden projects, have often been overlooked. Recently, Brownrigg made an extensive review of the literature concerning home gardening (1). Sumberg con- ducted a study for Oxfam America on vegetable gardening projects in Gambia and Senegal (2). He observed that in recent years a growing number of garden projects were focused on women's groups. He stressed that the agencies promoting vegetable gardening usually have a limited capacity for research and evaluation. O'Brien-Place suggested a method for the economic evaluation of home garden projects (3), including direct and indirect nutritional indicators, but to our knowledge, this method has not been tested. The opportunity to attempt an evaluation of the nutritional impact of home gardening arose in 1980 when we were invited to conduct a survey in Senegal. The survey involved a vegetable gardening project that had been initiated in 1969 in two experimental units for agricultural development: Kumbija and Ciise-Kaymor. In the second unit, the project did not survive. But in the experimental unit of Kumbija, which includes several villages, the produc- tion and sale of vegetables proved to be quite successful commercially. It was part of a larger program that included the increased use of fertilizers, of improved seed varieties for the traditional crops of millet and groundnut, and the introduction of maize, combined with primary health care and nutrition education activities. We were able to use the results from seven separate food and nutrition studies that had been conducted in the project area between 1970 and 1982 (4-7). Their aim was to measure the food, energy, and nutrient intake of different ethnic groups of farmers and determine the changes occurring in food consumption pattems. Those surveys did not follow a strict evaluation of nutritional impact procedure; however, they include two baseline studies in 1970 and 1971. To assess the food and nutritional impact of the home garden project, we took advantage of the possibilities offered by the successive surveys to make both chronological and cross-sectional comparisons. The average dietary intake of 1970-71 can be compared to the average of the four surveys conducted ten years later in 1981-82. Additionally, the study conducted in 1980, using a different survey technique, provides data allowing cross-sectional comparison of dietary patterns between vegetable growers, non-growers, and urban dwellers. Description of Villages Geographic Location and Ethnic Groups. The village of Kumbija is located in the southeast of the Sine-Saloum region (Senegalese groundnut basin), in the department of Kaffrine, district of Kungel, about 10 km from Kungel. Kungel is a town of about 3,600 people located on the Dakar- to-Bamako (Mali) railway and the road between Dakar and Tambacunda. It is thus a local trade center, with a daily market that attracts the population of surrounding villages. Kumbija consists of several quarters where the ethnic groups Soce, Pel (Fulani), and Wolof are separated into homogeneous groups. The village gets a fair amount of rain (800 to 1,000 mm a year) and did not suffer excessively from the drought in the 1970s. The major crops are groundnut (mostly for sale) and millet (mostly for home consump- tion). There are only two seasons: the rainy season, from June to October, during which the fields are cultivated; and the dry season, the rest of the year. Only in recent years has home gardening been prac- ticed by a limited number of families during the dry season, using water from wells for irrigation. Food Habits. The staple food is traditionally home-grown millet but its consumption has declined progressively, especially in the cities, and it has been replaced by imported broken rice. The staple food is cooked with small quantities of ground green leaves or vegetables, and even smaller quantities of smoked or dried fish, which together constitute the sauce. Meat and poultry are consumed only on special occasions such as ceremonies snd visits. Home Gardening in Kumbija Home gardening as a dry-season activity was introduced in 1969 by the Senegalese Institute for Agricultural Research (ISRA), but only a small number of farmers were interested in practicing it. The Soce ethnic group was the most recep- tive to this innovation. During the first years young plants were grown in nurseries and distributed at no charge to the peasants. Then external assistance was limited to the sale of seeds and pesticides, and to some technical advice. A special stand was obtained at the market of Kungel where the vegetable producers from Kumbija could sell their produce. Extension agents had contacted men first. Male farmers appeared interested, started gardening, then stopped. This was in part because they were embarrassed to fetch water from the well, an activity specific to women. Soon women took over. Only one hamlet had a well equipped with a system of animal traction. In other hamlets, women got water from the wells manually. The cultivated acreage per household was small, 32 to 51 mZ The vegetable gardens were usually installed after the millet harvest on the plots of land closest to the huts. Men built fences and plowed, women cultivated and sold the products. The vegetables that were most frequently grown were (in descending order): tomato (sweet and bitter), onion, cabbage, lettuce, eggplant, vissap (Guinean sorrel), okra, turnip, cowpea, parsley, beets, peppers, and carrots. weeded frequently andThe gardens were watered twice a day. Water- ing and drawing water from the well were the most strenuous chores-"four women draw as much water as two oxen do," declared the men during an interview. Often women did not wait for the full maturation of their vegetables. For some plants, such as turnips, they sold the leaves instead of the root itself. This is related to the tradition of preparing sauces for couscous or rice from green leaves. Cultivated leaves were gradually replac- ing picked wild leaves. Women organized mar- 5 ketins at the Kungel market where they owned their own spot. To sell their products they had to rise early-5:00 a.m.-and walk eight miles with heavy tin bowls of vege- tables on their heads. Results 1970-71 Surveys. The 1970 and 1971 studies were con- ducted when vegetable gardening had barely started in Kumbija and long after the end of the gardening season, thus no direct impact of home gardening was expected to be found. These studies were used as basellne studies. Results from the successive nutritlon surveys (1970-82) are shown in Table 1. The average per capita energy intake in 1970 and 1971 was approximately 2,200 Kcal. Retinol equivalents intake was twice as high in 1970 as it was in 1971. Average per capita ascorbic acid (vitamin C) intake was adequate in Kumbija both in 1970 and 1971. In Kumbija in 1970 only 4 percent of the calories and 5 percent of protein intake came from vegetables and fruits, but 97 percent of retinol equivalents and 99 percent of vitamin C intake came from wild leaves and fruits. The only vitamin seriously deficient in the diet was riboflavin. This was due mainly to the low consumption of foods of animal origin (2.5 percent of the calories, 9.4 percent of the protein). Wild leaves and fruits provided 21.6 percent of the riboflavin intake and were consumed in greater quantities than in the later studies. Average energy protein and lipid intakes were higher in the Soce group than in other ethnic groups. 1980-81 Surveys. The study conducted in 1980 used a different technique based on home visits and interviews. Previous studies had used weighted dietary intake methods. Table 1 shows the per capita nutrient intakes in Kumbija between 1970 and 1982. Table 2 indicates the daily consumption of most foods by category of consumer. It is apparent that the consumption of cultivated vegetables was twice as high among town dwellers as it was among vegetable growers. In 1980, home gardeners consumed, on the average, almost four times as many vegetables as did non-gardeners. In 1981, home gardeners consumed the same cluantity of vegetables as non-gardeners did in 1980. (It should be noted that vegetable production declined between 1980 and 1981.) In comparison to the situation in the same villages in 1970, consumption of cultivated vegetables had increased markedly and that of wild fruits and leaves had declined markedly (Table 3). Most of the decline seems to be due to the lower intake of dimb fruit (Mon'nga pterygosperma) which was frequently used to prepare sauces. The study conducted in 1980 attempted to accurately measure women's income from vegetable sales and their expenditures during and after the cultivation season. The average income of the 64 women practicing home gardening for sale in 1980 was US$29.00 per season (which is approxi- mately 11 to 40 percent of the annual income of farmers). Their expenditures at the end of the cultivation season are indicated in Table 4. Only 12 percent of their income was spent during the marketing activities of the season, mostly for edible treats for their children. The rest of their income, US$25.00, was spent at the end of the vegetable cultivation season: half of it on clothes and cloth and the rest on miscellaneous supplies, as indicated in Table 4. It is note- worthy that no women mentioned spending any income on drugs or medicine for their children, themselves, or other family members. The total amount spent on food was very small. Satisfaction of Nutritional Requirements 1981 Survey. The average percentage of satisfaction of requirements for different nutrients was calculated. Protein intake, when compared to recommendations, was apparently adequate (8). No family was below the arbitrary critical intake threshold of 80 percent of the energy requirements. The energy recluirements of the household were calculated from estimates of activity levels of adult members. Using the FAO recommendations as a standard, the apparent intake of retinol equivalents, riboflavin, folate, and zinc were below 75 percent of the recommended dietary allowances. Increased consumption of vegetables could have markedly reduced those deficiencies. Discussion The successive surveys conducted in Kumbija in relation to vegetable cultivation were not organized in such a way as to elicit the effects of all the external factors that might have modified dietary patterns. However, the comparisons of the data presented in Table 1 suggest that nutrient intake- which would have been raised significantly by a marked increase in vegetable consumption (retinol equivalents, iron, calcium, ascorbic acid)-was not higher, on the average, in the 1980s than in the 1970s. A close look at household income and expenditures reveals that women were able to earn some income from vegetable sales as early as mid-December. Groundnut sales, a major source of income, did not take place before mid- January and usually lasted until mid-February. It appears that vegetables provided women with a modest but early income in addition to that provided by the sale of their own harvest of groundnuts and cereals. Unlike the rainy season income, obtained and kept by the husband, the income from garden- ing is kept by the wife. This change from the custom is ... appreciated by women, who are now able to spend money without asking their husband. This increase in purchasing power allows women to spend more and finally to perform better their duties in the family (6). As indicated in Table 4 only a small portion (1 percent) of the income from vegetable sales is spent on food and apparently nothing is spent on medicine or drugs. Therefore, it is not surprising that vegetable sales only marginally affect family food intake and nutritional status. Brownrigg suggested that a number of dietary surveys omit consideration of wild food sources, thus reaching the erroneous conclusions that people lack specific vitamins or that they need to eat more vegetables (1). According to May, "The bush foods are invaluable and it is a matter of doubt whether Africans would have survived without them, especially during the hungry season [just prior to harvest]" (9). Kuhnlein has stressed the importance of the contribu- tion of wild plant species in the diet of rural dwellers where there is a limited availability and variety of marketed foods (10). The consumption surveys reported here (Table 1) and the interviews of farmers show clearly that traditional vegetables as well as wild leaves and fruits have up to now played an essential role in the supply of minerals and vitamins. The nutritional value of recently introduced cultivated vegetables should be compared to the nutritional value of picked wild leaves and fruits. Fresh cultivated vegetables are rich in minerals and vitamins, but often not as rich as the wild leaves and fruits that they tend to replace. As was noted above, the consumption of cultivated vegetables increased markedly and the consumption of wild leaves and fruits decreased markedly from the 1970 survey to the 1980-81 surveys. The impact of gardening on the decreased consumption of wild foods is not known. It may have been that there were fewer wild foods available in 1980- 81 due to the drought and other environmental factors. Or it is possible that wild foods became less prestigious with the introduction of gardening and the urban example. The impact of gardening on consumption of wild foods should be furtherinvestigated. Conclusions Home gardening was first initiated in Kumbija with the purpose of improving the income and nutritional status of villagers. Whether it accomplished this purpose is not com- pletely clear. Both the direct and indirect impacts of garden- ing in achieving this improvement deserve to be considered. Vegetables were used only in minimal quantities in the villagers' diets, and gardening had no measurable impact on the food intake of children after weaning. Whereas there is no evidence of direct nutritional impact, gardening may have had an indirect impact on nutritional status by increasing the income of the female farmers. Although women spent only 7 percent of their income from vegetables on food, they might have saved part of their main income from groundnuts and cereals thanks to winter vegetable sales. To markedly improve nutritional status, home gardens should be productive throughout the year, and especially during the hungry season. It appears essential to develop local technologies of vegetable preservation that women could operate themselves. The production of vegetables seemed limited because of the lack of such processing tech- niques. Consequently, women were growing only the quantities of vegetables that they were able to sell on the market. Good training in gardening and nutritional education are essential. The Ilesha gardens in Nigeria are an example of a successful project with a strong training program. The majority of home garden produce from these gardens was consumed by households. Nutrition training and cooking demonstrations were given on site, and the children's diets were supervised in a Mothercraft Center. As a result, death from malnutrition was reduced from 10 percent to 6 percent in three years (1). The training provided in Kumbija was not successful and mothers here did not seem to understand that vegetables were good for their children; most of them stopped growing carrots because children, who liked them, would sneak through the fences into the gardens and eat them raw. Aid agencies are often more concemed with the nutritional contribution of home gardens to the well-being of families rather than with the economic impact of this activity. However, it is clear from the present study that the primary motivation of the women who engaged in horticul- ture was to increase their income. While some produce was consumed by the gardener's household, most of the vege- tables were grown for sale. Although the economic contribu- tion of vegetable gardens to the household income was small, it all»wed w«men to purchase items that were specifically important to the improvement of their social status in a society where men have a dominant social position. Even in the absence of measurable nutritional impact, household gardens may play an important part in promoting social change. Acknowledgments We are grateful to Dr. Simon Chevassus-Agnes for analyzing part of the data used in this study. Thanks also to the inter- viewers from Kungel and to the kind cooperation of the villagers. We are indebted to Jacyues Faye, Senegalese Institute for Agricultural Research (ISRA), and Marc Levy, Clroupe de Recherche et d'Echange Technologique (GRET) for inviting us to participate in this joint research. References 1. Brownrigg, L. 1985. Home gardening in international deuelo~ment: What the literature shows. League for International Food Education, Washington . 2. Sumberg, J. 1986. Consultant's report on vegetable gardening projects in West Africa, submitted to Oxfam America. Mimeograph. 3. O'Brien-Place, P. 1987. Eualuatinghome gardenpojects USDA. Office of International Cooperation and Development, Nutrition Economics Group, Washing- ton, D.C. 4. Hellegouarch, R.J., J. Pelé, and J. Toury. 1970. L'Enyuête de consommation alimentaire dans l'unité expérimentale de Kumbija. Organisme de Recherche sur l'i\limentation et la Nutrition Africaines. Mimeograph. 5. Pelé, J., R. Hellegouarch, and J. Toury. 1971. Enquête alimentaire dans l'unité expérimentale de Kumbija. Organisme de Recherche sur 1'Alimentation et la Nutrition Africaines. Mimeograph. 6. Groupe de Recherche et d'Echange Technologique (GRET). 1982. Développement d'une activité ma- raîchère villageoise dans le Sine Saloum, Sénégal. Mimeograph. 7. Chevassus-Agnes, S. 1983. Situation alimentaire et nutritionnelle dans les unités expérimentales. Enyuête. Organismee de Recherche sur 1'Alimentation et le Nutrition Africaines, Office de la Recherche Scienti- fique et Technique d'~utre-n/ier, Senegalese Institute for Agricultural Research. Mimeograph. 8. WHO. 1973. Energy andprotein requirenents: a re~ort of a joint FAO/WHO ad hoc expert commictee. Technical Report Series No. 522. World Health Organization, Geneva. 9. May, J.M. 1968. The ecology ofmalnucrition in French speaking counmes of West Africa and Madagascar. Hafner Publishing Company, New York and London. 10. Kuhnlein, H.V. 1985. Wild foods for modem diets. Pages 153-173 in J. Weininger and G.M. Briggs, eds., Nunition 29date, Volume 2. Eiley, New York. Jacqueline Reynaud is a researcher in human nucrition who has euorked in recent years with seueral leading institutions in Senegnl and France. Her affiliation at the time of this artick was oisiting fellow with the Dioision of Nutritional Sciences at Cornell Uniprersity . Thierry A, Brun is uisinng professor in Nuaition nnd Agriculture at Con~ell Uniuersity and coordinator of the Cornell-China-Oxford Proiects on Nutntion, Health nnd Enuironment. Tonia Marek is a UNICEF exl>ert in Nutrition and Public Health recently based in Mauntania and a consultant to seueral U.N. aRencies.