CHAPTER 17: Ancient Irrigation

Irrigation

The first people to grow plants had to learn three main tasks: how to concentrate desirable plants into a manageable area; how to prevent weeds from growing there; and how best to encourage the plants to flourish. In short, people learned to plant, weed, and water or drain crops. In the English climate , there is enough natural rainfall to grow staple crops. But in a dry climate, most crop plants need water, and it must be provided by the farmer. Irrigation has been the technology underlying many of the world's greatest civilizations.

But irrigation is not easy. Usually, irrigation schemes require a large investment of money or labor or both. Water supplies must be safe and reliable. In each growing season, water of the right quality must be delivered at the right time and place in the right quantity at the right price, and irrigation water must later be drained off the crops. Even within one growing season, irrigation systems need continual maintenance.

The long-term hazards of irrigation are even greater. Investment in irrigation projects pays best in dry areas where evaporation is high. Water is never pure, but has mineral salts dissolved in it. Evaporation will therefore make it saltier still. Rivers flowing through dry or desert areas lose water by evaporation, and become salty. For example, the water in the lower Colorado contains over a ton of salt per acre-foot of water.

Many areas in dry climates have natural brackish or salty lakes, or even dry salt pans. Soils are often laden with lime (calcium carbonate) or salt (sodium chloride), and many sedimentary rocks contain natural salts. Promising irrigation areas thus may have natural salts in rocks or soils that will easily be transferred into fields as soon as irrigation water is applied, and even that water may come from rivers that have become saltier from evaporation along their courses.

As water is used on crops, it spreads out as a thin sheet, exposed to the surface. Much of it may evaporate, making it more saline. It may dry up altogether, leaving a thin layer of salts on and in the soil. Even under normal circumstances, plants absorb moisture from the soil, leaving behind excess salts. Eventually salts build up in the surface soils until they become infertile. Over time, therefore, soils in dry irrigated areas tend to become salinized.

The only way to deal with this problem is to apply enough water so that salt is flushed off or flushed through the soil. The flushing must remove salts from the area altogether, along natural or artificial drainage. In well-drained areas with a dry season and a wet season, natural flushing takes place each year. But in poorly-drained areas, over-watering simply mobilizes the salt while the water table rises to ground level. Capillary action draws the saline water to the surface, where the salt dries out as a surface deposit, and the problem is made worse rather than better. Once the soil is saturated, with water up to the surface, there is no way to leach salts out of the soil, and the fertility of the region is destroyed unless major drainage channels are built to carry away the salt. Even flushing may not be a net environmental plus: flushing simply delivers salt somewhere else, perhaps to downstream users, or into groundwater supplies. Flushing also leaches away soil nutrients with the salts.

Therefore, irrigation can only be maintained on a long-term basis in the following conditions. Water is applied in such a way that salt is not allowed to build up in the soil. Usually, this means that a lot of good-quality water is applied, and that drainage is rapid and efficient. Soils need a large infusion of fertilizer, to balance the flushing that is required to keep them salt-free.

A region that can be irrigated on a long-term basis thus has

If any of these conditions fails, the system will eventually fail. Such failures have brought down civilizations that solved the engineering and logistic problems of designing, building, and maintaining irrigation systems, but neglected the long-term effects of salinization or nutrient depletion. Long-term problems of irrigation may not appear for a long time: today, for example, the valleys and basins of the San Joaquin, Rio Grande, Indus, Nile, Murray-Darling, Jordan, and Tigris-Euphrates are being irrigated, with progressive and visible increases in salinization and water-logging, and no remedy in sight. Only a few civilizations based on irrigating dry country have lasted for any length of time: sensible civilizations should not try to grow wetland crops in arid climates.

The major success stories for civilizations based on agricultural irrigation are Egypt and China. The major stories of failure are happening right in front of us. In present-day California, a giant industry is trying to maintain an irrigation economy with a diminishing supply of poor-quality water, on clay soils with very poor natural drainage, in an almost landlocked plain with poor or non-existent regional drainage, applying water that has been stripped of its natural load of silt.

Ancient Irrigation: Egypt and the Nile

Ancient irrigation styles depended very much on the physical geography and geology of the area, and the engineering skills available. Four different styles of irrigation were developed very early in agricultural history. All irrigation systems depend on taking water from natural sources and diverting it to artificial channels or ponds where it is applied to crops.

The Nile valley is rainless and extremely fertile. Herodotus wrote more than 2000 years ago, "Egypt is... the gift of the river." Egypt depends on the Nile in a way that no other nation does. 97% of Egyptians live on 2.5% of its area. The prosperity of the Nile valley civilizations has depended throughout recorded history on the efficiency with which the central government has organized the best use of the river water. Crops could be stored after years of abundance, for example, and irrigation schemes could be both built and maintained.

The Nile receives its water from the tropical highlands of Africa. The river receives no tributaries at all for the last 1500 km of its course across the Sahara Desert to the Mediterranean. In Egypt, far from its sources of water, the Nile has no sudden flood-wave crests. The annual flood starts in June as snowmelt and summer rain flow down the river. It rises gently to its peak in late September and early October, then gently subsides by the end of December. The Nile is one of the most predictable rivers in the world, and its "flood" period averages more than a hundred days, rather than being very short-lived like those of other rivers.

At first, Egyptian agriculture along the Nile was based on growing winter crops after the annual floods had subsided. Egyptian irrigation was based on several facts. There was only one water source (the river) which was too powerful to control. Irrigation works therefore had to be passive in construction, and built relatively high along the river bank so that they dealt only with the peak of the flood. The river valley is flat-floored, but narrow and steep-sided, never more than 25 km wide until it reaches the Delta below Cairo. Irrigation schemes could therefore not carry water any great distance away from the river.

The ancient Egyptians built large flat-bottomed basins for growing crops along the river banks, and simple sluices that diverted water into them at the peak of the flood. It was easy in engineering terms, if not in labor, to arrange for good water flow through several basins in succession, controlled by simple gates. Water was allowed to stand in the fields for 40 to 60 days, then was drained off the crops at the right moment in the growing cycle, downstream back into the river. There was always plenty of water, so salts never built up in the soil; and the flow in the canals and ditches was strong enough to avoid silting. (Silt that settled in the basins was beneficial in two ways: it made the floors of the basins evenly flat, and it brought a lot of nutrients with each year's flood.) Ditches and canals were short, and the typical irrigation scheme was very local.

The design of the irrigation system depended critically on knowing in advance the height of the annual flood, and the Egyptians developed a system of "Nilometers" at various points along the valley. Rapid communication and early warning of the height of the flood as it rolled downstream from the south made a great difference to the size of the harvest. Herodotus wrote that the Egyptians "get their harvests with less labor than anyone else in the world."

Early irrigation was rather local and primitive, and food was not stored efficiently, so the early civilizations were vulnerable to long-term fluctuations in the Nile floods. There was no significant attempt at water storage: since all the water came from the Nile, any storage would have meant damming the river, which was far beyond the capability of the ancient Egyptians. Therefore their irrigation system was passive, and early Egyptian civilization depended largely on one winter crop per year. After it was harvested in the spring, the land lay fallow until after the next flood. Only in a few places with very wet soil was there any chance of a second crop, and among these areas were Abydos, Memphis, and Thebes, the great centers of ancient Egyptian civilization. They lay along the river, upstream from the Delta.

The Old Kingdom, the Middle Kingdom, and the New Kingdom were periods in Egyptian history when strong central government flourished in times of prosperity, followed by periods of stagnation in economy and population, often accompanied by social, military, and artistic decline. It's not clear whether strong central government resulted in effective irrigation and good crop production, or whether strong central government broke down after climatic changes resulted in unstable agricultural production.

Diversion Dams

Irrigation problems are completely different in dry hilly country. Rivers and springs are smaller, and do not have the perennial flow of the major rivers in Egypt and Mesopotamia. Water storage must be built into any sophisticated system, but some simple irrigation can still be practised. Even so, irrigation systems must be active, rather than passive on the Egyptian pattern. Most likely, irrigation systems began with small springs and in small drainages, and as engineering techniques were developed, they could be applied to larger rivers and to larger areas of plains.

Perhaps the earliest successful technique was to build a diversion dam. The structure is built right across the bed of the stream, and is not merely a cut in the river-bank that takes off some of the water into a canal. Stream valleys with only seasonal flow are the best places to learn dam-building. The dam can be built during the dry season, allowing engineers to learn how to build a successful dam, without having to master also the technique of diverting the stream. The dam can also be maintained or repaired during dry seasons. If the dam fails in flood, it does no more than restore the old flow, and is not catastrophic. It's probably not a coincidence that all the most ancient dams are built across intermittent streams, or wadis. Perennial rivers are much more difficult to dam because there is no access to the dry river bed unless the river is diverted.

One of the most impressive dams of ancient times is near Marib, the ancient Sabaean capital in the Yemen, and was built about 600 BC. It is 500 m long, and is built of carefully worked masonry, strengthened by copper fastenings. The dam was not meant to be a reservoir filled with usable water. It was more a giant weir, to hold back some of the annual flood waters coming down the valley, and to divert some of that water, under control, out of sluices and into a canal system. The same sort of diversion dam system evolved independently in the Hohokam culture of Arizona.

Mesopotamia: Canals on the Plain

Irrigation has been an important base for agriculture in Mesopotamia (what is now Iraq and part of Iran) for 6000 years. But Mesopotamia is very different from Egypt. Mesopotamia has low rainfall, and is supplied with surface water by only two major rivers, the Tigris and the Euphrates. Although they are much smaller than of the Nile, they have much more dramatic spring floods, from snowmelt in the highlands of Anatolia, and they carry more silt. Furthermore, the plains of Mesopotamia are very flat, and poorly drained, so that the region has always had persistent problems with poor soil, drought, catastrophic flooding, silting, and soil salinity.

Mesopotamian engineers had to worry about water storage and flood control as well as irrigation. Silt built up quickly in the canals, threatening to choke them. This could be overcome by constant dredging as long as organization and manpower were available. The other problem was more insidious, and could not be overcome by the engineering available at the time. It was difficult to drain water off the fields, and there was always a tendency for salt to build up in the soil.

Although the plain of Mesopotamia is very flat, the bed of the Euphrates is higher than that of the Tigris; in fact, Euphrates floods sometimes found their way across country into the Tigris. Engineers used this gradient as soon as irrigation schemes became large enough, using the Euphrates water as the supply, and the Tigris channel as a drain.

Mesopotamia has had times of successful irrigation, and times of silt and salinity crises: the latter around 2000 BC, 1100 BC, and after 1200 AD. The first crisis may have been caused by water politics. In any irrigation system, the farmers most downstream are those most likely to be short of water in a dry year, or to receive the most polluted water. In Sumeria, the city of Lagash was rather far downstream in the canal system based on the Euphrates. Apparently Entemanna of Lagash decided that he would instead cut a canal to tap Tigris water, but the addition of poor-quality water led to rapid salinization of the soil.

Sumeria
The earliest city-states of Mesopotamia, those of Sumeria, lay in the lowest, most water-rich areas of what is now southern Iraq. Irrigation could be fairly simple in this region, with each city-state probably building one irrigation system. The cities may have originally been administrative centers, marketing centers, and defensive centers related to local irrigation schemes: in other words, they were "irrigation cities".

From time to time catastrophic floods overwhelmed the region. At Ur there is a well-known band of 1.5 m of clay between two layers of pottery. This is evidence of a major flood, and this event was probably the basis for the flood story in the Sumerian Epic of Gilgamesh and for the much later Biblical story of the Flood.

Mesopotamian engineers built very large weirs and diversion dams, to create reservoirs and to supply canals that carried water considerable distances across the flat countryside. The scale of their irrigation was larger than in Egypt, and Mesopotamian irrigation was interventionist and active. Almost certainly the idea of diversion dams was brought to Mesopotamia from the hills, since the rivers are mostly perennial. Mesopotamian tradition suggests so: Sargon of Assyria, probably learned it from the ancient nation of Urartu. The scale and ambition of early Iron Age Mesopotamian projects was matched only in China and Egypt.

The Abassids

After the wave of Moslem expansion broke over Mesopotamia, the Abassid Caliphate was based on Baghdad from 762 AD until its demise in 1258. Existing irrigation schemes were renovated and greatly extended in very large projects. Abassid engineers drew water from the Euphrates at five separate points, and led it in parallel canals across the plains, watering a huge area south of Baghdad. This system provided the basis for the enormously rich culture of Baghdad, which is still remembered in legend (Scheherezade, the Caliph of Baghdad, and the Arabian Nights) as well as history. But it required a lot of physical maintenance, and there was a lot of salinization in the south. As central government began to fail in the 12th century (mostly from extravagant overspending), the canals became silt-choked, the irrigation system deteriorated, and the lands became more salinized. The deathblow to the system was natural: massive floods about 1200 AD shifted the courses of both the Tigris and the Euphrates, cutting off most of the water supply to the Nahrwan Canal and wrecking the whole system. The Abbasids were too weak (or bankrupt) by now to institute repairs, and the agricultural system collapsed. By the time the Mongols under Hulagu devastated Iraq and Baghdad in 1258 AD, they were finishing off a society that was already a wasteland. Iraq has remained a desert for more than 600 years.

Page last updated May 1999.

Return to Geology 115 main page.

Return to Geology Department Web page