by Annabelle Vinois and David Meyer
The Site:
The site is on the southern shore of Mfangano Island, a small island on the Eastern side of Lake Victoria in Kenyan waters without electricity or running water. The only motor vehicles on the island are a fleet of motorcycle taxis, two tractors, and one land-rover; the primary mode of transportation is foot. The island has a total population of 19,000 people who work mainly fishing on the lake. The primary catch is Nile perch, an exotic species introduced by the British to help develop the local fishing industry, and which has itself severely damaged the local ecosystem as the perch feed themselves on other species of fish and have no non-human predators. The perch have also damaged the social system of the region, with the cash offered from the sale of freshly caught fish driving more people towards fishing and away from subsistence farming, and fueling a system of ‘fish for sex’ where impoverished locals are forced into prostitution in order to feed themselves. HIV/AIDS is more prevalent here than almost anywhere else in the world, and fish stocks on the lake have begun to decline as a result of over-fishing.

The view from the back door
The island also, like much of Kenya, suffers from severe deforestation. On Mfangano, this is primarily from cutting for cooking fuel (charcoal) both for use on Mfangano and for sale to other, smaller islands in the area where deforestation has already been complete, and where literally no trees grow, with small villages sustaining themselves on the cash offered by the international sale of the Nile perch.
The climate consists of two rainy seasons: the short rains which come around November and December, and the long rains, which come from April to July. Because of deforestation, however, neither rainy season has been reliable in recent years, and long periods of extreme drought are common.
The site itself is a rectangle of 4,000 square meters, stretching northwest to southeast from the peripheral road south of the village of Sena (the largest village on the island) to the shore of Lake Victoria. The border with the road is 23.5 meters long, that by the lake 18 meters; the length of the property stretches 177 meters (note: due to an ongoing argument with the northern neighbor, the land boundaries are, at the time of writing, not yet fixed and so all these numbers are approximate). The land slopes downward to the lake, leveling off to a flat beach for the remaining 30 meters.
The previous owner used the land to plant corn (used in making Ugali, the Kenyan staple food served with each meal), and then left it otherwise unused and unfenced and thus open to unrestricted grazing by the livestock which roam the island freely. There is little top-soil now remaining, with the possible exception of the western edge where a series of yellow oleander trees have provided an area of shade and mulch and where even earth worms were visible just below the surface. The soil at the road edge is a rocky clay mixture, transitioning in 20 meters to a fine clay which continues the length of the property until the land flattens out and becomes pure sand for the remaining 30 meters to the beach. The only trees on the property are a papaya, growing in the sand 5 meters from the lake shore, a small tree growing on a rock along the road, and a yellow oleander in the northwest corner. The northern neighbors, who dispute the land boundary, use their own land only to plant corn once a year; the southern neighbors live on the property but do not intensively manage it. Winds generally come from east to west, up the lake towards the road.
A priority for the site is clearly control of erosion and establishment of topsoil.
The client:
The clients are a family of five, Richard and Ruth Magerenge and their three young children. Richard works for a recently started local NGO called ‘Organic Health Response’ which seeks to repair the social network which has been so torn by HIV/AIDS and the Nile perch. Ruth also works with the NGO, as well as running a local salon / computer service in Sena. The family currently lives in a rented house not far from the site, and construction is underway on their new home, with the plan of moving in before the end of 2009.
Their priority for the site is to build their home and to establish a farm with an emphasis on fruit and vegetable production, with the goal of providing much of the family’s food, a surplus for sale in markets, and a demonstration site to encourage more ecological farming techniques on the island; tree-planting is another priority, and Richard is strongly committed to organic practices. Richard’s uncle currently runs an organic farm in the village of Kitawe, five kilometers south of Sena. Richard would also like to have a milking goat, chickens (for both eggs and meat), ducks, and rabbits.
The Design:
(The letters used in this section correspond to the overall image of the property and design available in a series of photos online here.)
Energy: Electricity for the farm and house is provided by a solar panel on the roof of the house and a small gasoline-powered generator. This is sufficient for lighting the house and television and the generator can occasionally be used for a water pump, though this is rare.
Water: The primary source of water is the lake, with water brought up by hand in buckets. A system to recover rain-water will be established on the house and toilets as funds become available. Because the water from the lake must be treated to be drunk, recovered rainwater will be used for human consumption rather than irrigation. When irrigation is necessary, there is a generator-powered pump which can be used to pump water from the lake; for smaller-scale irrigation, buckets can be carried up from the lake.
Swales: The primary feature of the design is a series of eleven swales cutting across the width of the property. These swales, heavily mulched, will serve to control soil erosion, to capture as much water as possible during the rainy seasons, and to establish topsoil (Mollison 1979, p. 179). Swales should be one meter wide and are spaced every ten meters. Another farmer in the area told us of having eighteen inches of earth build up behind a retaining wall in just two years, and so it can be expected that over time the swales will fill and the land will take on a terraced feel.
1: This swale is an already existing ditch which should be deepened to maximize water retention. The space below will include the goat pen, and so the swale should be planted primarily with ‘goat tree’ (the local name for a tree goats love) and luceana (legume). It is mulched from the goat pen, luceana cuttings, and lemon-grass cuttings from the nearby pathway.
2: Built up more as a mound actually from dirt removed during home-construction, this swale should be planted again with luceana for mulch and medicinal trees such as Neem which can repel mosquitoes around the house.
3: This swale serves also as the boundary of the children’s garden; it is mulched by the chicken pen hanging above it on the northern side (see G), as well as with lemongrass cuttings from the pathways. A chicken-wire fence runs along the top of the swale to confine this group of chickens to the uphill section of the property.
4 – 10: These swales are planted with tall wide-shading fruit trees and luceana, with the goal of shading adjacent areas while also providing fruit.
11: This is the end of the clay-soil, after this the property is flat and mostly sand.
A: Pathway: This pathway extends from the entrance down to the lake. Above the house it is wide enough to allow the occasional tractor or donkey-cart visit with mulch and manure from other farms. It is lined by lemongrass on the inside, and by the wind-break (see: S) on the outside. Below the house it is lined with lemongrass (both a barrier against erosion and a mulch for the gardens) and is usually one meter wide to allow wheel-barrows and loads up and down the property.
B: Rocky-soil beds: These beds are for plants preferring rocky soil – peanuts and pineapple in particular. It is mulched from lemongrass cuttings, the goat pen, and luceana from swale 2.
C: Crescents: Both above the house and below, large fruit trees are planted in small clay and stone walls molded in a crescent-shape to capture flowing water, aid in topsoil creation, and fight erosion. In the uphill case, a mango and avocado tree should be planted, capturing the runoff from the rock and from swale one. These two crescents are mulched from the goat pen, lemongrass cuttings, and luceana from swale 1. The lower crescents should also be fruit trees (papaya, mango, guava, fig, banana) mulched by wandering chickens (see G) and lemongrass cuttings. The lower crescents may also include a Neem tree to control mosquitoes around the house.
D: Flower / Medicinal beds: This area is planted with hardy flowers and medicinal plants of use both to the family and to the community. This includes aloe vera, sweet annie, ringworm bush, thorn apple, asthma weed, and moringa (more information on all these plants and their use can be found in Hire & M’Pia). These plants should not require much attention, but can be mulched from lemongrass cuttings as the need arises. There is also a kentank here which captures rainfall from the roof of the house (see: water).
E: Kitchen area: This is the outdoor kitchen on the eastern side of the house. A trellis will be constructed here, with passionfruit and grapes grown as a cover. This will provide both shade to cool the house, cover for the cooking area when the sun is strong, and an additional source of food. In the kitchen, there will be a solar cooker (see www.solarcooking.org), an earth stove (see www.surreycommunity.info), a wood fired oven (see www.mothereartnews.com/do-it-yourself) and a cooking basket (http://cookinginabasket.blogspot.com). It would also be useful for Richard’s organization to take time to look into bio coal (which is made from agricultural and forest waste residues into solid cylindrical form through briquetting and can be used as fuel). For an example, see www.biocoalindia.com.
E’ is an herb spiral (Mollison 1979, p.113) where all the useful household herbs can be planted.

The crescent retains the water and soil (to prevent erosion and drought).
F: Children’s garden: this is an area for the children to play in, shaded by fruit trees planted in crescents (see C). Chickens (see: G) roam this area as well and provide mulch for the trees. F’ is a pit divided in two parts for accumulating mulch and generating compost. Organic waste from the kitchen gets dumped here, as well as organic matter brought in from outside the farm (essential, especially during the first few years to rebuild topsoil). Mulch from this pile is used throughout the farm.
G: Chickens: There are two main chicken coops on the property (G and G’). G, closer to the house, is home for chickens wandering the area between the children’s garden (F) and the road, foraging for fallen fruit and providing pest control around the house. They are also fed from kitchen waste brought over to the coop, which is constructed hanging over swale 3 so the manure falls as mulch. The lower chicken coop, G’, is built over the fish pond (see M), with chicken droppings providing food for the pond. There is also the possibility of moving either coop to serve as a chicken tractor system, scratching up the soil, providing manure, and eliminating pests from beds to be planted (see L).
H: Gangamma Mandalas: These are two vegetable / fruit beds planted in the pattern of Gangamma Mandalas (Mollison 1979, p.281-285). These areas include banana, papaya, collards, tomatoes, eggplants, carrots, cowpeas, okra, peppers, cassava, lentils, beans, and palm trees. These are the primary gardens for home consumption. They are mulched from lemongrass cuttings from the borders (the boundary between the two mandalas is a lemongrass border as well), banana fronds, and the mulch pit (see F’).
I: Tree Nursery: This is the nursery for starting new trees for use both on the farm and to be sold / donated to other sites on the island. Given the extreme deforestation on the island, this is an opportunity for the farm to have a farther reach into the surrounding community.
J: Orchard: This is a small orchard of citrus fruits planted in crescents staggered to collect maximum of runoff and to control erosion. Chickens can be brought in occasionally to feed on fallen fruit and control pests.
K: Volunteer section: This is an area for housing volunteers (both WOOFers working on the farm and volunteers with Organic Health Response). There are 5 large trees shading the area, 3 individual two-room mud and thatch huts of 5m in diameter, and one smaller covered communal space.
L: Beds: These are ten beds and are the primary market gardens on the farm and also space for growing grains. They are planted in rotations, and in different patterns in function of yield (designs shown are suggestions and can be changed as results vary). L1 and L4 are spiral patterns for grain planting for maximum edge-effect. L2 is a Fukuoka-style plot (planted by throwing seeds rolled in clay balls). L3 and L5 are Mfangano Island rock-art inspired patterns. The island is home to some of the oldest rock art in Africa, and these bed designs use those shapes for a section with many separate beds and microclimates depending on how it is planted. L6 is a modification on a Mollison design (Mollison, pamphlet) that efficiently creates a series of microclimates. L7 is a series of keyhole beds.
M: Fishpond: To begin this would be a small pond of 75 square meters, of an average depth of 2 meters. The pond would be stocked with tilapia and lined with shade-giving trees and vegetable beds, particularly tomatoes and collards (see: Quilleré et al (1995) for some ideas on this; Mollison(1979)as well). This section would serve to complement the family’s food production, to bring in additional money as fish can be sold for relatively high prices in town, and to give a wider social goal to the project – allowing the farm to serve both as a demonstration site for alternative methods of fish production and as a break from the wider problematic of the fish for sex trade.
N: Beds: A series of beds related to the fish-pond and to be converted, if the fish-pond is successful, into an extension of the existing pond or as separate ponds to grow complementary species. As beds, it should be planted with other plants that complement the existing pond (tomatoes, collards, etc.).
O: Palm circles: These are a series of palm and banana / papaya circles planted with yams, cassava, and cabbage. Trees include papaya, banana, palm oil, date, and coconut.
P: Sandy bed: Small plot for sand-favoring plants.
Q: Banana / Papaya circle: This circle includes the existing papaya tree and the inside of the circle is laid with wooden boards to provide a secluded changing area for people using the beach. Eventually this could be equipped to fit a solar shower which would make it an ideal spot for a warm shower.
R: Dock: This is to be added over the long-term, and would allow both for a family boat to be kept on the property, as a place for the children to play, and as an aid in harvesting the water hyacinths which can choke the shores of the island and yet which make an excellent organic mulch (to be used in all beds as needed).
S: Wind-break: These are trees planted all along the edge of the property to provide a wind-break and privacy. Trees planted here should be varieties which can easily be coppiced for fire-wood or additional construction materials. Trees with strong roots which will resist erosion can also be selected here. Sissal, bamboo, and lemongrass may be inter-planted with the trees as well.
Bibliography:
Hire, Hans Martin & Bindanda M’pia, Natural Medicine in the Tropics I: Foundation Text,
www.anamed.net
Mollison, Bill (1979), Permaculture: A Designer’s Manual, Tagari Publications, Tyalgum Australia.
565 pp.
Mollison, Bill pamphlet, “Designing for Permaculture,” Permaculture Design Course Series, Yankee
Permaculture, Florida, USA
Quilleré et al (1995), “An Artificial Productive Ecosystem Based on a Fish / Bacteria / Plant
Association. 2. Performance,” Agriculture, Ecosystems, and the Environment 53 (19-30)
Installation Time-line:
Certain aspects of the design are funding-dependent and should be implemented as the money needed becomes available. This includes the purchase of different animals (chickens, ducks, rabbits, goats), and while chickens can be bread from existing stock, other animals should be bought in as money becomes available.
Of the larger funding-dependent investments, the first priority should be the rain-water recuperation system and kentank to provide potable water for the house. Next should be the volunteer huts, which themselves can be constructed using volunteer labor. Finally should be the fish-pond since, though it can have a huge impact on the productivity and social role of the farm, it will require additional expertise and knowledge before implementation, as well as the funds to dig, stock, and aerate the water.
For the rest of the design, the Mfangano climate seems to divide the year conveniently into two seasons: digging (dry) and planting (wet). The first priority should be digging: putting the swales, crescents, and lemongrass barriers on the land as soon as possible. These steps will serve to control erosion and store water on the land, and so even if they do not translate to immediate financial benefit for the farm, they are the most essential investment in the long-term viability of the project. All swales should be mulched with several inches of organic matter, and then planted with trees in time for the rainy season.
During the rainy season, when the clay soils should not be dug to prevent them losing their structure, planting should be the major focus of work. In particular this should begin with the Mandala beds by the house which will provide the family’s main food source. Later as the lower
beds are established between the swales, they should be planted in different rotations to best use the soil.
Through several cycles of digging and planting, it is reasonable to imagine the whole design implemented in three years time and bringing money to the family by then as well – keeping in mind of course that the fruit trees in particular will take time to mature.
Plant List (next page):
English
Very impressive! I am not able to judge its technical side, but it sounds like you have studied it well! I would love to see it in a few years from now ! I really like the balanced, intertwined and recycling side of this approach, but also the fact that it looks so beautiful with all the curved and round shapes!!!! Very inspiring! May it be copied everywhere! Thanks.