When it rains some water enters the soil, the rest runs off. As it gathers strength it becomes the floodwater that runs through valleys, streets, arroyos and wadis after a rain storm. Sparse vegetation, little soil development, soil crusts, and low infiltration/absorption contribute to fast peaking flows in arid regions[1]. Storms lead to floods that may continue for minutes, hours or days depending on the rainfall. Flood depths of 5-10 meters may occur in extreme events. Residents in arid and semi-arid lands have developed a wide range of strategies to hold and capture critically needed water. In many cases floodwater has been the primary source of drinking water for people and livestock, water to irrigate trees and crops, and to replenish groundwater.
Floodwater also carries organic materials and fine sediment that can be captured to improve soil fertility, soil tilth and structure. Tohono O’odham families in the American SW sought out places where moist, nitrogen rich litter had accumulated beneath mesquite trees and would dig up the top soil and spread it on farm fields[2]. Nitrogen fixation may take place at 5-8 meters depth so these deep roots can provide little competition for shallow-rooted crop plants grown nearby[3]. Floodwaters in these desert ecosystems can carry large amounts of rodent dung, leaves from nitrogen fixing trees and shrubs, litter, and twigs. Enough material may come to floodwater irrigated fields in these floods to add an inch of rich soil and organic matter a year. Over hundreds of years the deposits grow quite deep. Studies suggest sediment deposits up to 30 m or more deep in the Marib area resulted from irrigation[4].
Floodwater can be destructive when flows are so big they overwhelm the infrastructure, damage homes and farms, and put people at risk. The ancient city of Petra was hit by a massive flood in the 4th or 5th century[5]. This was a rare, catastrophic event[6]. Water rushing down wadis would have created a fast moving 5-8 meter deep flood rushing down the main street. It would have left boulders, heaps of gravel and pebbles and a road covered with 2 to 5 meters of sediment. Capturing more water higher up in the landscape with rock lines, terraces, dams and cisterns reduces the risk from floods, but would not prevent these mega floods.
Water retention
The first goal is capturing and holding water so it can sink into the soil or fill a pond or cistern. Techniques range from a simple line of rocks on the contour to complex sets of dams in the wadis sized to capture just enough water to maximize soil water holding capacity in irrigated fields[7, 8, 9].
The oldest versions of run-off farming probably consisted of water spreading and/or small rock or brush check dams to slow and collect run-off. Small rock dams are found throughout the SW. One or two rainfall events with runoff can be sufficient to grow a crop. In areas with very limited rainfall the people of the SW would plant seed only when the soil was wetted[10]. Much work was done while the floods were in progress. An everyday sight during showers was the irrigators at work with hoes or sticks, or even with their hands[11]. Cultivars were selected that allowed for deep burial of the seeds, up to 15 cm or more. This provides soil moisture for a longer period of root growth and enables the crop to withstand another surface flow flooding event.
Terraces retain runoff and transform slopes into flat areas that are easier to plant and harvest[12]. Most people are familiar with the terraced rice paddies of Asia, but equally impressive terraces for agroforestry, grains and crops can be found in the MENA countries and American SW. The trees and shrubs on terraces provide food, fodder, fuel and wood while helping to stabilize the terraces and, in many cases, providing a source of income. They help conserve soil and protect it from runoff and erosion. Agricultural terraces require regular maintenance and if it lapses, it can lead to the collapse of retaining walls accompanied by increased soil erosion.
Microcatchments of various kinds are also effective. These can be crescent shaped, rectangular, or square. Some have ridges on all four sides. Microcatchments have been used for millennia in the Midles East, Africa, and the Americas. Microcatchments are built at low gradients[13, 14, 15].The area inside the microcatchment is steeper with flow leading to the lowest spot. About 10% of the rain that falls on the catchment may flow. The desired area can be calculated with an understanding of the crop, the weather, and the soil. In the Negev the catchments were 17-30 times the planted plot[16]. This could give an augmented rainfall equivalent to 300-500 mm with just 100 mm of precipitation. This is enough to grow most crops.
Soil pits (zai) are smaller but can be effective. Pitting improves water infiltration and retention, reduces evaporation, and increases surface storage and the time available for infiltration to occur. Pits capture rainfall directly and get a minor boost from runoff. Zai can collect up to 25% or more run-off[17]. Not as much as water as the microcatchments, but still of value. They increase surface water storage and water capture, allowing water to seep deeper into the soil. In one study moisture penetration reached 61 cm on pitted soil but only 12.7 cm on unpitted soil[18].
Systems that collect runoff in larger catchments such as hillsides with long slopes may be called macro-catchments[19]. The larger open reservoirs of MENA may be called hafirs, tabias and limans. In arid Tunisia, the tabia system is a traditional macro-catchment with a runoff area that occupies two-thirds of the slope and is traditionally used for grazing; with one to five cropped plots within U-shaped soil banks arranged in a cascade in the third downstream area.
These run-on areas accumulate and store the occasional runoff. In Southern Sudan, hafirs provide water to livestock, agriculture, humans and, to some extent, wildlife[20.21]. Small field plots were leveled and terraced to ensure the efficient distribution of water as well as the conservation of both water and soil[22]. Research with various fruit and fodder trees in the Negev has shown that during and after a rainfall event significant amounts of soil moisture in the topsoil between tree rows on terrace fields were not fully utilized and could be used for an intercrop.
In Northern Kenya runoff capture irrigation made it possible to double crop[23]. Normally, local Turkana farmers won’t risk a second crop during the short rains in August/September. But with rainwater capture the second crops were more reliable and could be further improved with mulching and tree pruning. A variant of these are the hillside conduit systems found in the Negev[24]. There ridges and channels run down slope collecting water to augment wadi flow to crop fields.
Water for people and animals can be stored in cisterns. A remarkable variety and number of cisterns in Petra were filled by runoff from rocks acting as the catchment. Small grooves or ridges would direct water into a sediment basin and then into the cistern. Rainwater can also be captured and stored in bigger cisterns or ponds. The most impressive cisterns are those designed to capture floodwater during flash floods. The Nahal Zin cistern is filled only after the flow in the arroyo is 1.5 meters deep[25]. The full cistern can hold 1,400 cubic meters of water. A large cistern in Resafa, Syria was filled with floods from a wadi west of the city to a full capacity of 18,000 cubic meters. These floodwater filled cisterns would need to have sediment removed periodically.
Dams used by ancient communities were typically small and used to hold water for drinking and animals. For example, the transhumant pastoralists of the Sheeb region in Eritrea build small water diversion structures[26]. Larger systems of wadi bed floodwater harvesting involved a series of stepped dams built across wadi beds. These shallow ponds would partially or completely fill with flood water and recharge the soil moisture so a crop could be grown. A flood event we monitored in the SW desert recharged the soil moisture to 7 meters and it remained high for months. Permanent and temporary dams of various kinds were used to divert water from wadis onto adjacent fields[27.28]. This could be small scale or large.
The Maʾrib Dam in Yemen, the largest, was built around two thousand years ago to regulate the waters of the Wadi Sadd[29]. It was about 550 meters long built of fine stone-and-masonry construction, with sluice gates to control the flow of water. It irrigated more than 1,600 hectares and was used and repaired for hundreds of years. A Sabaean inscription from 449 CE records acquisition of, “14,000 camels, 200,000 sheep (seems high), 217,000 pounds of flour as well as 630 camel loads of beverages” to supply the needs of workers mobilized to repair the Ma’rib dam.
To better understand the options for runoff agriculture some of the imaginative and effective techniques developed to use runoff and flood water in the Mediterranean and North Africa (MENA) and the American Southwest (SW) are described in Part 2.
References
1. Shanan, L. 2000. Runoff, erosion, and the sustainability of ancient irrigation systems in the Central Negev Desert. IAHS 261. 75–106 pp.
2. Nabhan, G. P. 1982. The Desert Smells Like Rain: A Naturalist in Papago Indian Country. North Point Press.
3. Virginia, R. A., Jenkins, M. B. and Jarrell, W. M. 1986. Depth of root symbiont occurrence in soil. Biology and Fertility of Soils. 2:127-130.
4. Brunner, U. 1997. Geography and human settlements in Ancient Southern Arabia. Arabian Archaeology and Epigraphy. 8(2):190-202.
5. Paradise, T. 2012. The Great Flood of Petra: evidence for a 4th-5thAD Century catastrophic flood. Annual of the Jordan Department of Antiquities 56(1):143-158.
6. One of my field research sites in the Mojave Desert of the SW was hit with 30 cm of rain in 4 hours. The flash flood from this 1,000 year event was epic.
7. Beckers, B., Berking, J. and Schütt, B. 2013. Ancient water harvesting methods in the drylands of the Mediterranean and Western Asia. eTopoi. Journal of Ancient Studies. Volume 2 (2012/2013), pp. 145–164.
8. Bainbridge, D. A. 2007. A Guide for Desert and Dryland Restoration. Island Press.
9. Bainbridge, D. A. 2024. Go Big! The Challenge of Large Scale Restoration of the Badiya. EcoMENA July 18.
10. Nabhan, G. P. 1997. Cultural adaptations to the desert’s bounty. Sonorensis. 17(1).
11. Gregory, H. E. 1916. The Navajo country: A geographic and hydrographic reconnaissance of parts of Arizona, New Mexico, and Utah. US Geological Survey. Water-Supply Paper 380. pp. 104-105.
12. Bainbridge, D. A. 2002 [1986]. Self-reliant Agriculture for Arid Lands. Sierra Nature Prints. https://works.bepress.com/david_a_bainbridge/16/
13. Shanan, L. and Tadmor, N. 1979. Microcatchment system for arid zone development. Hebrew University.
14. Edwards, F., Bainbridge, D. A., Zink, T. and Allen, M. F. 2000. Rainfall catchments improve survival of container transplants at Mojave Desert site. Restoration Ecology. 18(2):100-103.
15. Oweis, T. nd. Planning and design of micro-catchment system. CGIAR. https://mel.cgiar.org/ reporting/download/hash/1407b174dc2d34cbeb51c927a96081fd see also In Arabic. Micro- catchment rainwater harvesting: planning, design and implementation: Training Manual May 2022. ESCWA Publication: E/ESCWA/CL1.CCS/2021/MANUAL.3. ESCWA is one of five regional commissions under the jurisdiction of the United Nations Economic and Social Council.
16. Evenari, M., Shanan, L. and Tadmor, N. 1982 [1971]. The Negev: The Challenge of a Desert. Harvard University Press. page 104.
17. Malesu, M. M., Sang, J., Oduor, A., Odhiambo, O. and Nyabenge, M. 2006. Rainwater Harvesting Innovations in Response to Water Scarcity: The Lare Experience. World Agroforestry Centre.
18. Slayback, R. D. and Cable, D. R. 1970. Larger pits aid reseeding of semi-desert rangeland. Journal of Range Management. 23(5):333-335.
19. Beckers, B., Berking, J. and Schütt, B. 2013. Ancient water harvesting methods in the drylands of the Mediterranean and Western Asia. eTopoi. Journal for Ancient Studies. Volume 2 (2012/2013), pp. 145–164
20. Nasri,S. Albergel, J, Cudennec, C. and Berndtsson, R. 2004. Hydrological processes in macrocatchment water harvesting in the arid region of Tunisia: the traditional system of tabias/ Hydrological Sciences Journal. 49(2):261-272.
21. Public Water Corporation. 2009.Technical Guidelines for the Construction and Management of Improved Hafirs. A Manual for Field Staff and Practitioners. April. MIWRGONU/MWRIGoss. UNICEF.
22. Hillel, D. 1996. Chapter 2. Negev: land, water, and civilization in a desert environment. United Nations University website. The UNU Global Environmental Forum V on Freshwater Resources in Arid Lands13 June.
23. Lehmann Lab. Section of Soil and Crop Sciences. Cornell University web page. https:// lehmannlab.cals.cornell.edu/research/agroforestry/
24. Evenari et al. p103.
25. Evenari et al. p163.
26. Harrower, M. J. 2006. Environmental versus Social Parameters, Landscape, and the Origins of Irrigation in the Southwest, Arabia (Yemen). PhD Thesis. The Graduate School of The Ohio State University.
27. Bainbridge, D. A. 2007. A Guide for Desert and Dryland Restoration. Island Press.
28. Bainbridge, D. A. 2025. Acequias for More Sustainable Irrigation: Rediscover an Ancient Technology. EcoMENA. March 8.
29. Brunner, U. and Haefner, H. 1986. The successful floodwater farming system of the Sabaeans, Yemen Arab Republic. Applied Geography. 6(1):77-86.