The coupling of urban and agriculture into a single expression may seem contradictory, but in fact agriculture as a basic urban function is nothing new (Mlozi, 199?: 105-106; Lee-Smith and Memon, 1994: 70).
Archaeological fieldwork and aerial imagery have unveiled massive and ingenious earth and waterworks, within and on the edge of the urban settlements constructed by ancient civilizations. Many of these facilities and infrastructure were used wholly or in part to produce food, feed, and fodder crops; to provide wood for fuel, building, shade, fencing, and windbreaks; to grow shrubs, ornamental, medicinal, and other useful plants; and raise livestock for food, materials, traction, transport, and trade, sacrifices and status.
Uruk, the most important city in fourth-millennium Mesopotamia (with possibly 50 000 people), extended over 1100 acres, a third of which was covered with palm groves; the large majority of Uruk's working adults were engaged in primary agricultural production on their own holdings, on allotments of land from temples or as dependent retainers on large estates; they also had other occupations (Adams, 1994: 18). The Neolithic Egyptian settlement of Knossos developed mixed farming (wheat, barley, lentils, sheep, goats, pigs and some cattle); the Minoan town spread over 75 hectares with pop. 12 000; Knossos had isolated farms on its edge (Rodenbeck, 1991: 124, 129). Minoan palaces had a central court around which were grouped storage and production areas; rulers probably controlled much of the agriculture in the surrounding region, about 1000 hectares (Warren, 1994: 46, 51).
Under Persian emperor Darius, walled gardens or pairidaeze ( paradises ) were associated with hydraulic facilities, thereby exploiting water resources more fully. In Thebes, capital of the New Kingdom 1500 BC, walled gardens of prosperous Egyptians provided fresh fruit for the household (including indigenous vine, pomegranate, imported apple and almond), sycamores, date and down palms, fresh fish from lotus-covered pools; larger gardens with water tanks had ducks (Jellicoe, 1989: 25). In the capital city of Akhenatan, Egypt, gardens were everywhere, with additional spaces reserved to storage, underground cellars, breweries and animal keeping (Courtlandt and Kocybala, 1990: 126).
Water shortages may have curtailed urban horticulture in ancient Greece, but ingenious use was made of it wherever some was available. On Crete, the large inland city of Eleutherna was important until the Late Roman period and had a vaulted aqueduct taking water from cisterns under the acropolis and to extensive fields for crops, terraced down along the limestone spur on which the city was erected; some of these terraces are still cultivated (Rodenbeck, 1991: 91). Greek city-states were self-supplied with goat milk and olive-oil fuel for house lighting.
Vast agricultural drainage schemes were revealed on the Roman imperial sites of Timgad in Algeria and of Volubilis in Morocco. Near the mouth of the Tiber River, in the densely settled ancient Roman port of Ostia, a planned complex of garden houses, surprisingly similar to contemporary counterparts, was erected in approximately 128 AD. The complex was more likely built for middle and lower classes, with 40-100 apartments which probably housed 400-700 people; all that remains of its original gardens are the six fountains from which residents drew their water (Watts and Watts, 1994: 86, 88, 89).
The Roman coastal city of Cosa, 140 km north of Rome, at its height in 100 BC, had its harbour linked by artificial and natural channels to a commercial fishery in a lagoon. The catch was dried, pickled or salted and shipped in amphoras. The fish farm had tanks more than 100 m long, covering about one hectare at the west end of the adjacent lagoon. Some of the catch of eels, grey mullet, sea bass, gilthead and sole would have been eaten by the local population. A modern lagoon fishery is in operation at the nearby town of Orbetello (McCann, 1994: 95-96). Elsewhere, Andalusian cities had houses surrounded by gardens and orchards. Cities on the Indus River, such as Harappa and Mohenjo Daro, discovered under the shifting mud of the Indus, once were specialized agro-urban centres.
Of Medieval Europe, Susan Reynolds (1984: 200) writes: "The provision of food in sufficient quantity, sufficiently fresh, and at a reasonable price, was a constant anxiety...". Crop-rotation systems were being tested in farms and fields of monasteries, walled cities, and castles. Medieval tapestries suggest ladies' castle gardens included herbs on raised beds and rabbitries (Jellicoe, 1989: 34). A fifteenth-century College of the Vicars Choral in the city of York, England, had buildings surrounding a garden; behind them lay orchards (Addyman, 1994: 117-118). Nearly perfectly preserved medieval Novgorod in Russia is depicted in a seventeenth-century icon, showing well-spaced housing, gardens and orchards, within its outer and inner walls(Yanin, 1994: 123).
In North America's Mississippian culture (peak 1050 1250 AD), intensive riverine horticulture supported what Burland (cited in Coe et al. 1986: 57) qualifies as true pre-industrial cities in the rich alluvial valleys of the Mississippi, Ohio, Tennessee, Arkansas, and Red rivers and tributaries. One of them, the 10 000- people city of Cahokia in Illinois, was the largest pre-Columbian urban centre north of Mexico. Also in the middle course of the Mississippi, the Moundville site (population 3 000) in Alabama contains borrow-pits apparently used to store live fish, part of the food needed to support its people (Coe et al. 1986: 56). In Central America, authority was exerted from major centres to terrace steep hills and drain swamps into fields on the edge of Nohmul; this was a large late pre-classic city near the Belize-Mexico border. At the city of Edzna, on the coastal plain of Campeche, waterworks of staggering proportions (2.25 million m3 of water storage) supported a highly organised agricultural economy (Hammond, 1994: 132).
Four thousand years ago in the pre-Olmec Valley of Mexico, small towns on stone-faced terraces, such as Tlatilco and Ticoman, farmed vegetables and raised dogs and turkeys, (Burland 1976: 15 18). The Aztec state was partly dependent on food production within and on the periphery of the metropolises of Teotihuac n and the capital city of Tenochtitl n, south-west of the former and on a man-made island built on Lake Mexico (Anton 1993: 116). In 1519, Diaz marveled at the agricultural nature of the city he discovered, an island capital extending over 20 square miles, with five times the population of Henry VII's London at the time (Redclift, 1987: 109). Teotihuacan itself at the height of its power (500 BC) was larger (more than 4000 buildings and 50-100,000 people) than imperial Rome (Millon, 1994: 138); Millon's maps of Teotihuac n (population 125,000 250,000) clearly indicate chinampas in one section of the city: these were rectangular raised-beds anchored with planted fences of willows, filled in and periodically fertilised with piles of marshy vegetation removed through canal-cutting, topped with canal-bottom mud.(Coe et al. 1986: 104).
Chinampas carried fruits, vegetables, trees and houses, supplied most of the
produce consumed in the city at the time of discovery and still supplied some
vegetables as late as 1900. Three harvests were possible, with transplanting from
reedbeds. Animals were kept and their manure and that of humans applied to
organic gardens (Redclift, 1987: 109-110). Highly fertile and productive
chinampas were found in Xochimilco (surviving to the present), towns on
southern shores of Lake Xochimilco, and in most of the island of Tenochtitl n
Tlatelolco. A plan to recover the Xochimilco region has prompted new interest for
the chinampa economy which has survived to the present (Millon, 1994: 139; see
also Canaval, 1992). A 15 km-long dike, built across Lake Texcoco,
protects chinampas from rising saltwater in the rainy season (Coe et al. 1986:
144, 146, 149). An aqueduct raised water to irrigate a hilltop orchard in the
northeast of Tenochtitlan (Haas, 1993: 22). The well-spaced layout of outer house
mounds probably enabled each home to have its own garden (Burland, 1976: 40). In
Haas' Gardens of Mexico (1993: 22) a nineteenth-century painting depicts a woman
in her central Mexico City rooftop garden, attended by her mestizo and
native maids, a water seller approaching the group. Haas finds that secure
rooftop potted plants are an enduring phenomenon in provincial urban and rented
houses of Mexico.
At Tairona's Buritaca 200 site in the Colombian Sierra Nevada, an elaborate landscape of retention walls, canals, and drainage systems afforded in-city cropping (Coe et al. 1986: 166 167; Burland 1976: 162). In the Peruvian Andes, central plazas of U-shaped structures might have been irrigated or flooded and crops possibly grown; large ceremonial complexes were usually adjacent to cultivated fields (guinea pig remains found earlier than 1800 BC at Culebras, halfway between Trujillo and Lima) (Coe et al., 1986: 197). At Cuzco and Machu Picchu, extensive retention walls, terrace gravel beds and stone-lined drainage afforded intensive farming of steep slopes. In our times, the Yoruba of western Nigeria maintained sizeable cities that were largely self-sustaining because "most of the productive inhabitants were full-time farmers" (Adams, 1994: 15).
The above overview suggests that food production in the more advanced urban settlements of ancient civilisations, was not a rare, temporary activity sited haphazardly in the urban fabric. UA was not socially demeaning or technically primitive when it was practised. Quite contrarily:
Being more vulnerable to supply disruption or insufficiency, malnutrition or famine, food provision throughout history has been a pervasive concern of city populations. Under the Islamic empire, the Abbasids even turned a postal service into an intelligence system, through which postmasters kept the capital city informed on food prices in their postal districts, so that supplies could be sent wherever shortages threatened. Cities have always had to ensure that a reasonable share of their food needs be supplied from within a controllable hinterland.
Throughout most of mankind's history and in different civilisations, urban populations have engaged to variable extents in producing at least some of the food they require, close to or at their own residence, within or just outside the city. Reliable, minimal amounts of food and nonfood items were needed to ensure subsistence and trade by what were then unprecedentedly large human agglomerations; this may explain the frequent coupling of elaborate earth- and waterworks with food production in ancient cities.
Archaeological remains and artistic depictions highlight the living conditions of ancient urban elites more than any other human group. Clearly, food production was not only carried out by better-off individuals; those in authority also commissioned, built and managed massive food-producing systems, even making room for food production when designing lower-class living quarters.
Urban food production took a variety of forms, making ingenious use of space, site conditions, closeness to equipment, facilities and markets, natural light, water, soil and waste resources. The scarcer these resources, the more valuable have been the products chosen to be grown or raised by city dwellers. Many cities probably did offer the incentives, conditions, market and testing grounds for more intensive and productive farming systems to be experimented, perfected, and disseminated. Technological breakthroughs included sun reflectors; water collection, storage, and distribution; frost protection; wetland drainage and slope terracing, multi-cropped and layered chinampa-style raised beds. In his review of recent archaeological findings on ancient cities, Adams concludes (1994:15) that large-scale canal networks clearly seem to have followed the advent of fully established cities.
Go back to Table of Contents: Urban Food Production by Luc Mougeot
Go back to Urban Agriculture Notes
revised, June 12,1995
Send comments here