A number of factors combine to determine whether homegardens are an appropriate strategy for improving the livelihood of poor families. Among these, access to suitable land - i.e. a land plot that is large enough and sufficiently fertile for a family to establish a homegarden, and to which the family has ownership or ownership-like rights - is perhaps the most fundamental factor. The importance of access to land is occasionally mentioned in the homegarden literature, but does not appear to have been addressed in any detail. Where access to suitable land is not a constraint to establishing homegardens (or once planners have arranged to overcome the issue by providing secure access to land), other important factors become relevant, including access to water, access to know-how, and access to stocks of appropriate plants and animals. Cultural acceptance of homegardening and access to sufficient capital and labour are equally important. This section reviews these factors.
We are interested in how homegardens can benefit the poorest of the poor, which usually includes families without adequate access to land. A number of commentators have noted that lack of access to land is a serious constraint to homegardening (Hoogerbrugge and Fresco 1993, Marsh 1998, Vasey 1985). Across cultures, ownership of land appears to be a significant inducement to gardening and landlessness, and tenant status have been identified as constraints that effectively eliminate homegardening as a viable development strategy (Brownrigg 1985). Indeed, for families without adequate and secure access to land, lack of land is the single most important barrier to homegardening.
"Arrangement of adequate access to land and security of tenure should be a primary consideration in the design of projects which promote home or community gardens, as is true for other kinds of agricultural production development." (Brownrigg 1985: 111).
The poor sometimes have adequate access to land, as for example in Russia and most other countries that emerged from the Soviet Union. In these countries access to land for homegardening has generally not been an issue, and all village families have access to plots used for homegardening. Such land is usually located immediately adjacent to the village house, and vegetable plots arranged on arable fields near the village. During the harvest season, owners take turn guarding the vegetable plots against theft.
However, inadequate access to land remains a significant problem for poor families worldwide, even in areas that commentators commonly associate with homegardens, such as Java. On crowded Java, the great majority of homegarden plots (known locally as "pekarangan") are smaller than 200 m2 (Prosterman and Mitchell 2002).[13]
Fig. 1. Size of household pekarangan plots on Java.
Source: Computed using 1995 Housing and Settlement Statistics (Indonesia Statistic Center Bureau 1996), cited in Arifin (2002), App. 2.
Even where a family has nominal access to land, the insecurity of their rights to such land may dissuade the family from making any long-term investments in improving the land, such as by planting trees, improving drainage, installing fencing or building a fishpond. Squatters and families who occupy land merely with the permission of a landlord (often the head of household’s employer) may even worry that their improvement of the land they occupy may lead to eviction as others seek to cash in on the investment. Thus, it is not only the quality and sufficiency of the land itself, but the quality of the family’s right to control the land which are critical.
If lack of access to land is a constraint in establishing a homegarden, how much land do families need to establish a homegarden? The size of homegardens varies considerably across cultures, and even within the same community. For example, Hoogerbrugge and Fresco (1993) survey a number of studies that report variations in size from 10 to 120 m2 in one Zambian study to between 5000 and 20,000 m2 in another Zambian study and variations of 172 to 500 m2 in one Java study to between 200 and 1700 m2 in another Java study. In Papua New Guinea, houseplots of 300 - 400 m2 are often too small to meet the ambitions of household gardeners, who extend their gardens beyond the allotment or establish second gardens away from the house (Vasey 1985).
While the average size of existing homegarden plots is instructive regarding social and economic norms in a given community, there is evidence that distribution of even relatively small amounts of land to land-poor families can provide a base for important improvements to the household nutrition. In a study of Kerala wage-earning families who cultivate homegarden plots occupying a fraction of an acre, Kumar (1978) found that the value of homegarden production was the most consistent positive predictor of child nutrition, and was an especially strong predictor during the slack employment season, as well as in households in which the mother is not employed outside the home.[14] Kumar goes on to report:
Results of the study indicate that the proceeds of the produce from even small family plots of land, if intensively cultivated, lead to large increments in child nutrition. This is especially true when maternal labour force participation is absent. Additionally, during slack employment seasons, availability of garden produce, or income thereof, seems to provide a buffer against reductions in child nutrition during this period. Even though the exact mechanisms by which this occurs are not clear, the conclusion is that provided intensive cultivation of distributed land is possible, even small family plots can have an important nutritional impact for young children. (Kumar 1978: 60 - 61).[15]
The larger role of women in Kerala homegarden production, as well as the fact that women can more easily and better care for children at home, may contribute to the fact that better child nutrition is strongly associated with more productive homegardening (Kumar 1978). Given the strong correlation between homegardening and the presence of mothers not engaged in wage labour, distribution of land to landless and land-poor families may be a particularly fruitful strategy where employment opportunities for women are limited for societal or economic reasons (Kumar 1978).
In a study of 62 rural homegardening households in Karnataka, India, researchers found that intensity of tree growing increased markedly once homegarden plot size reached 1800 sq. ft. (See Fig. 2). Of 14 households that grew 5 or more trees, only one had a homegarden plot smaller than 1800 sq. ft., while of the 16 households with plots 1800 sq. ft. or larger, 13 households (72 percent) grew 5 or more trees. This suggests that there is likely to be a critical minimum plot size above which households will begin planting more trees. The footprint of the house itself, which commonly occupies 500 sq. ft. or more of the land plot, is presumed to reduce greatly the number of trees that can be planted on the smallest parcels.
In a separate study, researchers identified Karnataka households that appeared to be making very intensive use of land. The researchers found that, after taking into account the size of the parcels, families who receive land from the government appear to be as likely to plant trees or raise animals as are families who inherit or purchase the homegarden plot. None of the interviewed families had received government extension advice or assistance with planting trees or raising livestock (Hanstad et al, forthcoming). This data suggests that at least for some poor families in Karnataka, access to land is a primary barrier to tree-planting and animal-raising.
Fig. 2. Trees planted by Kamataka homegardening households.
Source: Hanstad and Lokesh 2003 (unpublished data).
In a sample survey of 97 land poor households that had received houseplots under government housing schemes in West Bengal, researches found that the productive value of plots increased significantly with plot size until plot size reached approximately 3000 sq. ft. (270 m2) (Hanstad and Lokesh 2002). More detailed interviews of 45 similarly situated West Bengal households revealed similar results. In the smaller sample, homegardens smaller than 1000 sq. ft. (90 m2) were found to provide the fewest benefits. Productive value and reported benefits increased significantly for homegardens of 1000 - 1999 sq. ft. (90 - 180 m2), and increased further for homegardens of 2000 - 2999 sq. ft. (180 - 270 m2), but then plateaued or even decreased for plots larger than 3000 sq. ft. (270 m2) (Hanstad and Lokesh 2002).
In Papua New Guinea, one study suggests that an area of only 150 m2 is sufficient to supply adequate fresh vegetables for a family of four, although another researcher argues that it would require 3,100 m2 (Thaman 1990). In a study of Philippine urban homegardens, Miura (2003) found that size of homegarden plot did not affect the number of varieties of fruits and vegetables grown in the garden. Out of 103 households, 54 (52 percent) with plots larger than 100 m2 had planted more than five varieties of fruits and vegetables, while 26 of 49 households (53 percent) with plots smaller than 100 m2 had planted more than five varieties. Of course, the quantity of production of the larger and smaller plots is likely to differ.
Another important consideration is the cost of distributing small plots of land to land poor households. Assuming the government acquires land at market prices, can governments afford to distribute small parcels? Again, affordability cannot be considered in the abstract, but must be evaluated by looking at the cost and effectiveness of other programmes designed to address the needs of land poor families.
Hoogerbrugge and Fresco (1993) suggest that homegardens generally occupy "marginal" plots; that is, plots that are too small to be used for field cropping or grazing, or are on land that is too steeply sloped to be used for field cropping. Of course, to the extent that this is so, it is likely a consequence of locating rural housing, and the housing of the urban poor, on hillsides and other lands that are less suitable for field crops and urban development. To the extent that homegardens are cultivated by poor households, it is natural to suppose that the gardens and housing will be located on economically marginal land.
Another consideration is whether land is available near the families who need the land. Vasey (1985) describes experience in Papua New Guinea in which a government programme to allocate gardening land away from the house failed since it was too difficult for the cultivators to guard against theft and vandalism. In Indonesia, government planners express concern that in some cases, land located near rural villages on Java is prime rice paddy land, the terracing of which has been undertaken at great social cost. Strict policies are in place to prevent conversion of such land to other uses, though the conversion of rice paddy land to residential land continues to occur.
At least one state in India (Karnataka) is initiating a programme that will establish several-acre colonies of house-and-garden plots with 10 plots per acre, and allocate the plots to landless and land-poor families. The land will be located within one kilometer or less of the village in which the recipients currently reside, and such land will either be existing government land or land that the government purchases.
Although homegardens are primarily rainfed, it is common for homegardeners to irrigate during the dry season. Watering depends on the type of crop and can vary from twice daily to twice annually (Hoogerbrugge and Fresco 1993). Several studies have found that drawing, transporting and hand irrigating the homegarden are the most onerous and time consuming gardening tasks. For example, Russian women expend considerable effort and time carrying water to irrigate home production gardens (tho Seeth et al 1998), although Russian gardens are much larger on average than tropical home gardens and are thus likely to require much more water. Homegardens that require even a few gallons of water per day during the dry season may require too much labour to be worthwhile (Brownrigg 1985).
In some areas, lack of water may be the major factor limiting homegardening (assuming the household has access to land). In Papua New Guinea, potable, piped water for irrigation is the most expensive input for urban homegardeners, and water is especially expensive during the dry season (Vasey 1985). In the United States, subsidies for water delivered to agriculture, including homegardening, is the most important government subsidy that agriculture receives (Brownrigg 1985, citing Cleveland 1982). Ninez (1985) reports that in Lima, Peru, urban homegardeners used waste water from the kitchen to irrigate their garden, but sometimes found they had to purchase small amounts of additional water. In 1983 the cost of water to irrigate a garden of 200 m2 was approximately US$3.50 per month, but in the poorer neighborhoods water was sometimes not available during the warmer months, and gardens suffered greatly.
It can be prohibitively expensive for the household to install a system for bringing water to the house, and there is a large social cost for providing irrigation water to households, especially in urban areas. Homegarden irrigation is rarely a consideration in the design of a reticulated water system, and it may be too expensive to make capital improvements to an existing system necessary to accommodate demand for irrigation water (Vasey 1990). For urban homegardens in Papua New Guinea, use of unmetered water for irrigation during the dry season exceeded the value of the crops produced, while metered households found it uneconomic to use water for irrigation (Vasey 1990).
However, even if water costs exceed the value of production, low cost irrigation water might be justified as an appropriate subsidy for poor neighborhoods since it contributes to social safety net while reducing moral risks associated with government handouts. The cost-benefit analysis of subsidizing irrigation water for homegardens should not consider purely the economic value of homegarden production, but should take into account the broader social benefits of helping poor families to gain a degree of economic independence and the accompanying enhancement of self image and social status. Development projects that bring potable water to the village could free up labour spent hauling water, allowing the household to devote saved labour to homegardening (Brownrigg 1985).
Households may address the need for irrigation water using various strategies. In drought prone areas of the African Sahel, simple wells and artisinal irrigation are used. In West Bengal, rural communities are sometimes sited near a rainfed natural or artificial community pond that families can draw from for irrigation and other household water needs. Rainwater harvesting may be an affordable means of capturing, storing and applying water for homegardens (Agarwal and Narain 1999). International Development Enterprises - India (2004) has developed and successfully marketed several low-technology, low-cost micro-irrigation systems in India that are appropriate for homegardens ranging in size from 20 to 1000 m2 and cost from US$5 to US$90. Households may also use household wastewater (Marsh 1998).
Water conservation strategies can reduce homegarden demand for water. Such strategies include terracing, trenching, deep mulch and surface mulch (including living mulch and ground cover creepers). Plant spacing and mulch may be used to conserve moisture (Vasey 1985), and drought-tolerant plants can reduce the impact of water shortfalls (Marsh 1998).
In areas of high rainfall, canopy layers, raised beds and drainage canals may help to prevent flooding (Brownrigg 1985), while use of water-loving plants and plastic coverings can help to reduce the effects of water abundance (Marsh 1998).
Most homegardeners are unwilling or unable to invest much capital in their homegardens (Vasey 1985, Mendez 2001). Where households have access to capital, it is not surprising to find that homegardens are more productive. For example, one reason that homegardens cultivated by rural households in Russia are more productive than those cultivated by urban households is that the rural households have easier access to inputs and implements that originated from the former collective farms, some of which they may receive as a part of their wage for work on agricultural enterprises that replaced the collective farms, and some of which they may "divert" from such enterprises (tho Seeth et al 1998).
Homegardening materials need not be prohibitively expensive. In India, a "kitchen garden kit" developed for a pilot activity containing seven varieties of tree seedlings, high-quality vegetable seeds and five hybrid chicks currently costs 600 Rupees, or about US$12 at current exchange rates.
An important limiting factor discussed in relation to capital is the use of inputs to improve soil fertility. Lack of capital is likely to be more of a barrier to homegardening where soils do not contain sufficient nutrients to support gardening and where households are not familiar with composting and other methods for improving soil fertility. In such cases extension advice should include instruction on accelerated composting and the possibility of gardening in containers while establishing better soils (Brownrigg 1985).
In addition to composting, households may construct terraces and plant leguminous trees to improve soil fertility (Marsh 1998). These methods may be labour intensive, but do not require large amounts of capital. Where gardening tools and other inputs are absolutely necessary, the government may find it useful to establish small local stores to sell simple tools and supplies at affordable prices.
Where fencing is required to reduce foraging by animals or theft of homegarden production, live fencing can be used to reduce costs. Plants used in fencing can also provide additional products for household use or sale (Marsh 1998).
Distribution of land for homegardens is most likely to lead to successful establishment of homegardens if homegardening is already successfully practiced in the vicinity where the new homegardens are to be established. The presence of viable homegardens in the immediate vicinity is the best predictor of success since such homegardens demonstrate that homegardening is socially and culturally acceptable, and is valued by households (Marsh 1998). Although projects to promote gardening often prefer to work with communal organizations, household-level food production must be a family undertaking since labour, space and time are valuable resources to poor households and "cannot be risked on the uncertain participation of a number of individuals," even on a small scale (Ninez 1985).
Cultural preferences may inhibit households from taking up homegardening. Households may associate homegardening with poverty and therefore decline to establish gardens (Miura 2003). In Nepal, dark green leafy vegetables are often considered low-status foods, which might help to explain why Shankar (1978) found that consumption of vitamin A-rich foods did not increase along with homegarden size in the studied groups. In Bangladesh, dark green leafy vegetables are widely believed to be bad for young children, and this belief does not depend upon the household’s income or access to land (Cohen 1985).
Noting that the smaller size of homegardens in some western and eastern areas of Java, as compared to larger homegardens in the central part of the island, cannot be explained ecologically, Terra (1954) concluded that the differences must therefore have an ethnographic explanation. He opined that the difference has to do with the tendency of lands in matriarchal societies of the region to evolve into family-owned plots, while lands in patriarchal societies of the region tend to remain controlled by the head of the dominant clan (Terra 1954).
Lifestyle preferences can limit the ways in which households use their house plot. For example, although it is common to combine gardening and small livestock production in Indonesia and other South East Asian cultures, small livestock are said to be the main ravagers of gardens in other areas (Brownrigg 1985). Households may thus feel compelled to choose between gardening or raising small livestock. Where local culture places a high value on free access to house sites by neighbors, households may be unwilling to construct fences necessary to keep out animals (or keep in animals) (Brownrigg 1985).
Cultural values may evolve over time to cause changes in the way homegardens function. Soemarwoto (1985) reports a trend of moving the household toilet from above the fishpond to inside the house in order to improve aesthetics and comply with ideas of modernization, even though this means that human waste is no longer be recycled in the fishpond where it had been converted into protein, but is instead be flushed into streams where it will contribute to harmful algae blooms.
Public information campaigns can play an important role in encouraging families to plant and consume more vegetables. Many families are not aware that vegetables and fruits are nutritional. In a study of Philippine urban homegardens Miura (2003) found that mothers generally had no knowledge of vitamins and iron in foods until informed by community health workers. Before they learned about the nutritional value of vegetables, some Filipino families mistakenly believed that micronutrient tablets distributed by the government were more desirable than consumption of vegetables, and that vegetables were the poor man’s substitute for tablets. Brun (1989) recounts the case of a village studied in west Senegal, where mothers cultivating homegardens did not seem to understand that vegetables were good for their children and most mothers stopped growing carrots when their children snuck through homegarden fences to eat them raw.
Miura (2003) found that urban Filipino families who obtained health information from NGO-funded local health improvement programmes and community health workers had planted more fruits and vegetables, including dark green leafy vegetables, than families who had not received such information, regardless of socio-economic status, size of the homegarden plot and general knowledge of nutrition. Miura concluded that the fact that socio-economic status did not affect the degree to which families adopted advice received from NGO-funded local health improvement programmes and community health workers indicates that home gardening is a suitable strategy for poor communities. Even where families are already consuming vegetables and animal proteins, educational campaigns can help to increase their consumption by mothers and children. One such campaign in Central Java, conducted through radio, banners, billboards, posters and face-to-face communication, resulted in an increase in consumption of eggs and vegetables by mothers and young children across all socio-economic groups, leading to an improvement in serum retinal levels (de Pee et al 1998).
It is common to find that government agricultural extension programmes ignore homegardens in favor of working with commercial field crop producers. This should not be surprising given that most agriculture ministries are focused on production of staple foodstuffs and export commodities rather than subsistence agriculture. On the other hand, it is rather strange that governments that spend large sums on "household food security" programmes - in India these take the form of redistributing food resources through public distribution systems, fair price shops, food for work programmes, etc. - should so neglect improvement of homegardening plots that would not only add additional food to the system, but would deliver that food to households that most need it.
Agricultural extension can contribute significantly to homegarden production. At the beginning of the Helen Keller International pilot homegardening project in Bangladesh, 50 percent of households reported having a garden with mean size of 61 m2 and growing an average of 3.1 varieties of vegetables, whereas after two years with the project 100 percent of households reported having a garden with mean size of 138 m2 and growing an average of 17 varieties (Marsh 1998). This demonstrates that families can learn to cultivate homegardens more intensively.
The background of extension agents can play an important role. Brownrigg (1985) recounts the experience of a project in Ghana in which senior extension agents were young women with graduate degrees in agriculture and training in nutrition, while junior agents were local young women and men from the target communities. She reports that the local extension agents were accepted much more easily than either foreigners or non-local national staff.
One common approach to providing extension advice for homegardening is for a project to establish a "demonstration garden." Brownrigg believes this can be a poor strategy, especially if the gardener in charge is either from the village elite (which causes people to view homegardening as an elite activity) or is a hired labourer from the lower rungs of the village social ladder (which causes people to view homegardening as a low status activity).[16] If a demonstration garden is used, implementers may want to convince at least one of the better local home gardeners to work with the project to improve an established homegarden (Brownrigg 1985).
Gardening promotion projects are likely to be most successful where homegardens already exist. Homegardens must be introduced gradually in areas where they do not exist; an overnight "revolutionary" approach has proved problematic in many homegarden campaigns (Ninez 1985).
Interventions that seek to advise families on appropriate techniques for improving homegarden production may find that homegardening families do not have adequate access to seedlings and other necessary materials. This is more than a question of access to information or even access to capital. The materials necessary for homegardening may not be locally available.
The introduction of improved inputs can make a significant contribution to homegardening productivity. A useful example is a project undertaken by Heller Keller International to improve animal husbandry in rural Nepal, Cambodia (poultry and eggs) and Bangladesh (poultry, eggs and fish). The projects targeted household egg production by introducing improved breeds of birds that produce more eggs, along with vaccinations and assistance with proper housing and feed, targeted livestock production by promoting improved grass fodder and deworming tablets for milk cows, and targeted fish household fish production by introducing fast growing fish cultivars and plant sources of fish feed. In all three countries, ten to twelve months after the improved breeds were introduced, households in which chicken liver had been consumed within the past 7 days rose from 21 percent of households to 35 percent of households. In addition, the number of eggs consumed in the household rose from a weekly average of 5 eggs to a weekly average of 12 eggs, while the number of eggs consumed by all household children rose from a weekly average of 2 eggs to a weekly average of 3 eggs (HKI/AP 2003).
It is important for planners to avoid the unsustainable practice of importing planting materials and breeding stock and instead establish "sustainable local mechanisms for the procurement, production, and distribution of seeds and other planting materials" (Brownrigg 1985: 104). Dozens of successful gardening projects have independently concluded that the solution lies in establishing nurseries, seed multiplication units and seed beds under local control (Brownrigg 1985). Perhaps the largest homegardening improvement project ever undertaken is the NGO Gardening and Nutrition Education Surveillance Project, in which Helen Keller International works with local NGO’s to establish village nurseries to provide seedlings and seeds to rural homegardeners in Bangladesh. The programme was scaled up over ten years and was reaching 800,000 households as of 2001 (HKI/AP 2001).
Before a homegardening project identifies which plants and livestock to promote, planners should study the preferences and capacities of local homegardening families. The choice of appropriate plants and livestock must depend upon what local people prefer to eat, as well as what plants and livestock are well suited to the local environment and resource profile of households (land, labour, capital), especially during the "hungry season" when stores of foods are scarce (Brownrigg 1985).
Homegardens typically are cultivated using "marginal" labour, which is marginal in the sense that it is flexible and its use reflects low opportunities for alternative employment (Hoogerbrugge and Fresco 1993). Although there is not much data on volume and timing of labour inputs (Hoogerbrugge and Fresco 1993), commentators tend to agree that most traditional gardening practices involve only a few days of preparation and less than an hour per day for maintenance and harvesting. During the five-month growing season, it was estimated that homegardens in urban Lima required an average of 50 minutes per day to prepare soil, plant, cultivate, water and harvest (Ninez 1985).
Poorer families may garden more than better off families. In the urban capital district of Papua New Guinea, squatter settlements and government housing estates, which have a high proportion of unemployed and underemployed workers, contain homegardens averaging 469 - 523 m2, which is 29 - 35 percent larger than the average for the district as a whole (Vasey 1985).
However, even poor families may rationally decide that the opportunity cost of homegardening is too steep. For example, it is reported that many poor families in Java must engage all household members in wage labour and therefore have no time to spare for gardening, even though the family is aware that they could obtain additional income or nutritious food from the homegarden (Soemarwoto 1987).
In this respect it is important for planners to appreciate that the opportunity cost of spending time on gardening is not zero, and labour intensive technologies may not be appropriate in many contexts since household members do not have unlimited time available for gardening (Brownrigg 1985). Homegardens are more sustainable if labour requirements are low and somewhat flexible (Marsh 1998). Planners contemplating homegardening projects should consider the availability of marginal labour among households targeted for assistance.
|
[13] In a study of a lowland
rural village in Central Java, Rajagukguk (1989) found that of 1002 village
families, 44 families (4 percent) owned no pekarangan or household plot, 347
families (35 percent) owned 50 m2 or less, 328 families (33 percent)
owned 50 - 100 m2, 259 owned (26 percent) 100 - 500 m2 and
only 24 families (2 percent) owned more than 500 m2. The size of
rural pekarangan plots is declining over time, which Arifin (2002) attributes in
part to the widespread practice of parents allowing their children to build
houses on the pekarangan plots, which are then divided among the children upon
the death of the parents. [14] Presumably, mothers who do not work outside the home - whether by choice or because such wage labour is not available - have more time to devote to managing homegarden production. [15] Urban homegardening was found to improve the diet of urban Filipino families by increasing the varieties of fruits and vegetables consumed (Miura 2003). [16] Brownrigg also notes that school teachers are a particularly poor choice as local homegarden extension agents since they either have no experience in gardening or tend to approach the garden as an academic exercise rather than focusing on actually producing plants and animals. |
