Encyclopedia of Geoarchaeology

2017 Edition
| Editors: Allan S. Gilbert

Canals and Aqueducts in the Ancient World

Reference work entry
DOI: https://doi.org/10.1007/978-1-4020-4409-0_141


People of the ancient world understood that geophysical and climatic anomalies could alter the environments that permitted the growth of comestible agricultural resources for urban and rural populations. When their technical capability proved adequate, they were able to modify water supply systems to sustain agricultural productivity through times of environmental change. When technological solutions or adaptations to other resources were not possible, societal transformation and/or collapse followed, leaving archaeological remains that now testify to the lack of appropriate technology, management, or manpower to overcome the deteriorating resource base. Water for urban and agricultural use is vital to sustainability. When the collapse of agricultural systems is manifest in the archaeological record, remains of canals, aqueducts, water storage, and transport systems provide vital geoarchaeological clues detailing how and why failure occurred. These clues often point to long-term drought that limited water availability for farming, floods that incurred changes in agricultural landscapes through soil erosion or aggradation, seismic/tectonic effects that disrupted canal and aqueduct systems, river downcutting (rejuvenation) that stranded canal inlets, and aeolian soil transport that led to landscape inflation or deflation processes. The influence of these geoarchaeological processes on water supply and distribution systems is basic to (1) understanding the fate of ancient sites and cultures and (2) interpreting the processes of societal collapse and transformation.

Geoarchaeological effects on canals in ancient South America

Research on ancient canals of pre-Columbian Peru has shown the influence of seismic distortion and tectonic uplift on once-functional canals. Many ancient canals built during the Late Intermediate Period (LIP 800–1480 CE) Chimu occupation of North Coast Peru (Moseley, 2001) show layered silt deposits indicative of the functional, negative-declination slopes necessary to conduct water from rivers through irrigation canals onto field systems. Several major irrigation and water transport canals were subsequently rendered nonoperational by tectonic/seismic effects during the Chimu occupation within the Moche Valley heartland (Ortloff et al., 1985; Ortloff, 2009), thereby severely compromising the intravalley canal network supporting the agricultural field systems vital to maintaining large population centers. A further effect from tectonic uplift was incisive downcutting (rejuvenation) of the Moche River, the main water source for the Moche Valley irrigation network. In addition to uplift, erosion from torrential rains deriving from episodic El Niño events also accelerated the downcutting processes. The river gradually deepened its bed relative to the adjacent land surface that contained the agricultural field systems (Whitten and Brooks, 1982). Figure 1 shows the south-side riverbank of the steeply downcut Moche River layered with Early Intermediate Period (EIP 300 BCE–600 CE) cultural remains indicating the effects of river downcutting episodes that began in earlier EIP and continued through the later LIP Chimu occupation. Fed by flow from seasonal Andean mountain rainfall runoff, the Moche River deepened its bed in step with tectonic uplift, leading over time to lowered water levels and the stranding of inlets to irrigation canals originating in valley neck areas far from the coastline. These inlets served field systems on the coastal plains and were eventually abandoned when they became nonfunctional. As canals with valley neck inlets were abandoned, new canals were constructed further downriver toward the coast in places where downcutting was as yet minimal. These new constructions required short canals with shallow slopes that were built into riverbank sidewalls to conduct water from the level of the river to slightly lower canal inlets serving lower elevation field systems. With ongoing tectonic uplift and river downcutting over many centuries, and with new canal inlets being constructed as higher elevation inlets were abandoned, agricultural field area was compressed into ever-decreasing areas downriver from the older abandoned inlets. This process is illustrated by the inlet/canal sequencing shown in Figure 2. Thus, geophysical processes acting on canal-based irrigation agriculture were a factor contributing to the ultimate collapse of the Chimu Moche Valley irrigation system in the tenth to eleventh centuries CE.
Canals and aqueducts in the ancient world, Figure 1

Downcut riverbank sidewall from the effects of Moche River rejuvenation; the riverbank profile contains artifacts from the Early through Late Intermediate Period (300 BCE–1400 CE) (From Ortloff, 2009: Figure 1.1.12, p. 30).

Canals and aqueducts in the ancient world, Figure 2

Graphic representation of downcutting evolution in the Moche River: frames A, B, and C represent sequential riverbed lowering producing loss of agricultural land as canal inlets are stranded and tapping of river water progressively upriver limits the acreage that can be brought under cultivation (From Ortloff, 2009: Figure 1.1.13, p. 32).

Coincident with these ongoing geophysical effects, long-term drought (Thompson et al., 1985, 1995) in Peru from ∼1000 to 1300 CE lowered river flow rates and thereby compromised the water supply available for agriculture. A typical Chimu Moche Valley canal serving the Pampa Huanchaco field system (Ortloff, 2009) exhibited cross-section profiles that showed continual modification to accommodate decreasing water flow rates (Figure 3). Low Moche River water supplies impacted agricultural production severely, which then failed to support the large Chimu population at the Moche Valley Chimu capital city of Chan Chan. A remedial measure was attempted in order to overcome the decline in agricultural production: the 75 km Intervalley Canal was constructed prior to ∼1000 CE to bring water from the higher flow rate river in the adjacent Chicama Valley to resuscitate failing Moche Valley field systems (Ortloff et al., 1982). Another remedial measure was the extensive excavation of deep pits dug to the phreatic zone of the water table near the Pacific coast margin of Chan Chan, also to maintain limited agricultural production. The Intervalley Canal’s hydraulic design was ingeniously crafted to produce a maximum flow rate to match and reactivate the Moche Valley’s main intravalley Vichansao distribution canal and provide water to field systems along its route to the Moche Valley (Ortloff et al., 1982; Ortloff, 2009); however, it stretched the drought-limited water supply of the Chicama River and ultimately provided only a short-lived solution to reactivate desiccated branch canals and maintain Moche Valley agriculture. Evidence of path modifications to restore functionality to the Intervalley Canal attest to seismically induced landscape distortions affecting canal slopes during its period of use; ultimately the combination of long-term drought and seismic distortions rendered the canal unusable. While the cumulative effects from centuries of climatic and geomorphic change steadily contracted the heartland of Moche Valley agriculture, the Chimu expanded into north coast valleys with richer water and land resources in later times.
Canals and aqueducts in the ancient world, Figure 3

Cross-section profiles of a canal from the Pampa Huanchaco area of the Moche Valley; profiles indicate continuous infilling and contraction of cross sections as tenth- to eleventh-century drought reduced water supply to canals (From Ortloff, 2009: Figure 1.1.8, p. 25).

The Late Preceramic (2600–1800 BCE) society at Caral, located in the Supe Valley of coastal Peru, is a further example of geophysical landscape changes influencing both agricultural productivity and the marine resource base underwriting their economies. With intensifying ENSO (El Niño-Southern Oscillation) floods occurring at this time (Sandweiss et al., 2009), flood sediment transport produced large offshore beach ridges that trapped flood sediments, infilled bays with marshes, and prograded earlier shoreline locations by flood sediment deposition and aeolian sand transport (Ortloff, 2009; Sandweiss et al., 2009; Ortloff, 2012). These geophysical processes gradually buried extensive coastal agricultural field systems in the valley delta area and caused agriculture to be transferred to narrow and distant up-valley farming areas that were insufficient to support the large valley population. 14C dating of near-shoreline-habitat mollusk deposits, now far inland from the present-day shoreline (Figure 4), indicates that large-scale sediment transport and littoral infilling had occurred over centuries (Ortloff, 2012). These episodes buried coastal agricultural fields beneath non-fertile sands and gravels that derived from concentrated ENSO flood events, preventing agriculture from being reestablished. Supe Valley excavation pits ∼3 km inland reveal 4–5 m of sediment deposits covering what was previously Holocene beach sands. As sediment deposits and clay banks accumulated behind the beach ridges, decreased coastal zone hydraulic conductivity resulted, creating limited groundwater drainage. This caused the water table to back up and rise, producing up-valley bottomland springs in narrow valleys. These springs were canalized to support limited, narrow-valley bottomland field systems and ramped aqueduct canals built along the south-side Supe Valley canyon walls supplied by bottomland springs from far up-valley sources. The ramped canals led water to plateau field areas that added limited agricultural field acreage. Up-valley agricultural bottomlands were subject to episodic El Niño erosive flood events that reduced land areas suitable for agriculture and further destabilized the economic base of valley population centers. The major loss of large coastal fields buried by flood sediments, however, made it difficult to sustain all 19 Supe Valley sites, and their populations as geophysical landscape change progressed. Again, geophysical processes played a prominent role in influencing site abandonment of Preceramic societies in the Supe and adjacent coastal valleys. All sites were eventually abandoned by 1800 BCE, presumably due to the large-scale geophysical effects that compromised their economic base founded upon trade of marine resources from coastal sites for agricultural resources from inland sites (Shady Solís, 2000, 2007). As the marine resource base (primarily shellfish gathering) was compromised by bay infilling and the agricultural area underwent progressive contraction, the sustainability of coastal Late Preceramic society’s economic base decreased.
Canals and aqueducts in the ancient world, Figure 4

14C dates from mollusk layer concentrations far inland from the present-day Pacific Coast shoreline indicating aeolian/flood sediment transport producing coastal infilling over millennia; mollusk species sampled are known to thrive in ∼1 m seawater. Samples obtained surrounding the salt pans of Salinas de Huacho, north of Lima, Peru.

The Middle Horizon (300 BCE–1100 CE) site of Tiwanaku (Bolivia) also underwent a collapse due to long-term drought in the tenth to eleventh centuries. A dropping water table and declining spring flow ultimately stranded 100,000 km2 of raised field systems adjacent to Lake Titicaca that supported agriculture for the Tiwanaku capital city (Ortloff and Kolata, 1993); no 14C dates indicating occupation after 1100 CE are recorded for the capital city or satellite centers.

Ancient South American societies experienced a variety of climatic effects that induced geophysical landscape changes affecting their agricultural and marine resource base. The disappearance and/or transformation of major societies, and their resurrection under different social structures in different areas when drought conditions relaxed in the twelfth to thirteenth centuries CE, depended largely upon exploitation of more sustainable agro-systems that benefitted from increasing water resources. Colonies and satellite settlements characterized many major Andean societies (Murra, 1962). This strategy expanded agriculture into different ecological zones and applied different farming techniques, employing local water resources to lessen the dependence upon a single agro-system type. This approach proved valuable in diversifying the agricultural resource base of Andean societies.

Geoarchaeological effects on societies of the ancient Middle East

While climate change and severe weather events characterize coastal and highland societies of ancient (and modern) South America due to ENSO El Niño and La Niña drought and flood effects, ancient (and modern) Mediterranean societies had the advantage of milder climate and fewer significant weather fluctuations. For the most part, Roman, Greek, and Levantine civilizations experienced fewer climate-related environmental challenges than their South American counterparts. While drought and flood events certainly occurred, colonies and captive areas under central state authority that encompassed vastly different ecological zones guaranteed a resource base available through trade and tribute to sustain the large populations of capital cities. Under stable climate/weather conditions, canal and aqueduct construction exhibited a degree of permanence that reflected the monumental labor input dedicated to their construction. This is in contrast to ancient South American societies whose survival and continuity depended upon water transport and agricultural systems that had to be modified to accommodate changing ecological conditions.

As an example, consider the case of Roman Ephesus (Turkey) and the engineering that characterized the permanent water distribution structures constructed without major interferences from the effects of climate or other geophysical change (Bammer, 1988; Scherrer, 1995; Ortloff and Crouch, 2001; Crouch, 2004; Ortloff, 2009). Water supply to the city was by means of canals and aqueducts from distant spring systems (Figure 5); the terminal point for aqueduct flows was a multi-chambered holding tank structure (castellum), each of whose chambers held water at different depths, a condition that determined the hydraulic pressure and flow rate from each chamber into pipes that emerged from the base of individual chambers. The water distribution system consisted of joined segments of terracotta piping running from individual castellum chambers to city destination points, each with given water requirements. Other water distribution components consisted of open channels that led directly from a castellum to reservoirs and then to pipe systems that supplied baths, elite housing compounds, administrative structures, marketplace areas, fountains, and latrines within the city. When open channels were impractical due to access or unstable landscape paths, buried multiple pipeline bundles were used. Pipeline systems buried at shallow depths had the advantage of being conveniently reparable, as broken segments could easily be replaced. Open channels likewise had easy repair access. When earthquake damage occurred, repairs could be readily conducted. Thus, few elaborate renovation mechanisms were needed as repairs were made by virtue of the simplicity and accessibility of near-surface water piping networks. Drainage from built structures and baths was through subterranean channels that drained into the nearby bay. Rainfall runoff was collected through gaps in street paving slabs and conducted directly to the bay. Thus, both water supply and drainage facilities could be readily repaired due to the nature of the design. Large aqueduct structures were able to resist seismic loading pulses due to the flexibility provided by non-mortared, stone block construction that distributed and dissipated energy through sliding friction between the many individual blocks forming the aqueduct. While this type of construction limited earthquake damage, it is questionable whether this was a thoughtful consideration in Roman construction methodology. It produced a fortuitous outcome nonetheless. Redundancy was built into the water supply to the city by five different aqueducts from different springs (Figure 5). Although most components of the Ephesian water system were resistant to moderate seismic loading due to their flexibility and ease of repair, Ephesus nevertheless was vulnerable to some geophysical risk brought about by its location adjacent to major rivers that carried the large volumes of rainfall runoff (silt and sand) that are characteristic of the mountainous coastal zone of western Turkey. Silt from runoff erosion of mountain soils carried by the Meander (Büyük Menderes) River over many centuries gradually infilled the bay adjacent to the city (Kraft et al., 1999) and reduced the maritime trade and commercial importance of Ephesos. These river silts deposited ∼10 m of sediments and engulfed an earlier Greek settlement below Roman Ephesus, thereby covering sacred processional paths associated with the earlier Greek settlement and burying springs at the base of the largely karst composition mountain (Crouch, 2004) adjacent to the Roman city precincts.
Canals and aqueducts in the ancient world, Figure 5

Water supply and drainage lines for Ephesus in the Roman occupation period (From Ortloff, 2009: Figure 2.3.8, p. 313).

At Ephesus, canal and aqueduct systems leading from mountain spring sources distributed water into the urban core through complex pipeline systems that supplied water for 250,000 inhabitants at 150 gal/day/person, including baths, fountains, nymphea, latrines, elite housing compounds, public buildings, a coliseum, gymnasia, and a theater, all of which required a continuously running water supply (Figure 5). The nearby Temple of Artemis, originally ∼3 km inland from the Mediterranean shoreline, was served by multiple underground pipelines of different designs for different ceremonial uses (Ortloff, 2009). This site, originating around 800 BCE, became inundated after many centuries of operation as a result of coastal subsidence and progradation, uplift of inland mountains, and sediment deposits interfering with springs that supplied water to the temple. Here, the geophysical effects were so gradual and subtle that compensatory structural engineering considerations made in advance of construction were apparently not a major concern.

The site of Petra in Jordan (Bourbon, 1999; Taylor, 2001; Guzzo and Schneider, 2002) is a further example of the creative use of intermittent water supplies from rainfall and springs to maintain city activities over centuries. Spring systems within tens of kilometers of Petra provided water to reservoirs from which terracotta pipes guided the flow to inner urban precincts for agoras, fountains, theater, water gardens, temples, public buildings, and domestic housing. Piping systems of different hydraulic designs were necessary given how distant these springs were from distribution hubs. Known from ancient times, and verifiable from modern computer calculations, is that a linear increase in supply hydraulic head does not result in a corresponding linear flow rate increase in long pipelines (Ortloff, 2009) due to nonlinear, cumulative water-internal pipe wall friction effects. This design constraint, together with how landscape-governed slope variations place constraints on pipe flow rate, results in a catalogue of hydraulic designs (Ortloff and Kassinos, 2003; Ortloff, 2009) that were utilized at Petra for the Siq, Jebel el Kubtha, and Zurraba water supply systems. Such urban core water supply systems possessing different hydraulic solutions for different geophysical constraints demonstrate that the ancient engineers possessed a wide knowledge base, approaching in many cases that of modern hydraulic design practice. Rainfall catchment basins and reservoirs provided additional water supplies and limited runoff into the urban center; some 250 such basins have been located in the mountainous areas surrounding Petra. The predictability of rainfall periods in this area of Jordan was well understood in antiquity and served to provide city reservoirs with water through many catchment basins and springs. Defensive water diversion channels and dams limited water damage to the urban core of Petra; here knowledge of water control was key to the permanence of the city for many centuries.

For the sites mentioned, the permanence of construction of fixed water supply elements (canals, pipelines, aqueducts, and reservoirs serving city and agricultural systems) indicates that geophysical threats were minimal outside of occasional, but reparable, earthquake damage. Thus, with regard to water supply systems, the advantage of the Mediterranean world with its stable climate and weather norms is apparent compared to New World cities and settlements.

Geophysical effects on the water systems of the Khmer Kingdom city of Angkor

The site of Angkor (800–1450 CE) in central Cambodia (Laur, 2002; Coe, 2003) reveals a long history of innovative water management that supported agricultural resources for a vast population centered about the central city core. A series of moats, channels, reservoirs, dams, and ritual water healing centers characterized the city’s precincts (Figure 6). Of interest are two large reservoirs (barays), the largest of which is the West Baray with an 8 × 5 km footprint. A wide-barrier dike enclosed the baray on all sides, and water from monsoon rainfall and the Puok and Siem Reap rivers provided water to fill the reservoir during the rainy season. While some release points along the dike led to irrigation canals, the large reservoir had a profound reason for its existence. Groundwater flow modeling (Ortloff, 2009) indicates that water stored in the reservoir during dual monsoon seasons was slowly released by groundwater seepage during the dry season to maintain a constant groundwater height throughout the year in areas south and east of the baray. The area between the baray and the edge of Lake Tonlé Sap was primarily for rice cultivation in sunken pits dug below the water table. The East Baray served a similar function by maintaining groundwater height under the city’s urban core and keeping the water level in moats and ceremonial pools constant year-round. Without dry season groundwater recharge from the barays, a permanent collapse in subsoil porosity would have occurred causing ground subsidence and structurally compromising the Angkor temples. Thus, two major barays, together with reservoirs, pools, and moats within city precincts, served to maintain the structural integrity of the many temples of Angkor, permitted extension of rice cropping on a year-round basis, and provided aesthetic embellishment to the Khmer version of the celestial capital of the gods. Through captured monsoon runoff, groundwater seepage systems, and surface transfer canals, an elaborate three-dimensional water control system (subsurface and surface) gave prosperity and continuity to Angkor over many centuries of occupation. Thus, Khmer knowledge of geophysical effects related to groundwater movement was a vital element in their city’s prominence over many centuries.
Canals and aqueducts in the ancient world, Figure 6

Site feature map of Angkor (Cambodia) (From Ortloff, 2009: Figure 3.1.1, p. 359).

Southwestern Native American societies: geophysics of canals and aqueducts

Over geologic epochs spanning millions of years, the Colorado Plateau has been etched by the deepening and headward extension of innumerable small valleys opened during periods of intermittent heavy rainfall. These valleys are characterized by floodplain incision from rain runoff producing areas of unconsolidated sediment deposits within the valleys that limit water control for irrigation agriculture (Longwell and Flint, 1962; Cooke and Reeves, 1976). Heavy rains lead to sediment deposition over bedrock, creating arable land for irrigation agriculture but not in areas prone to periodic erosion. Farther south in the Basin and Range country, deeper alluvial valleys containing sandy desert soils limit agricultural productivity due to limited moisture retention, as well as climate/weather conditions characterized by high desert temperatures and more frequent drought conditions. Across the Southwest, many alluvial valleys are prone to stream entrenchment (arroyo cutting) that lowers water tables and restricts the amount of arable land that can be irrigated (Cooke and Reeves, 1976). Thus, as a result of heavy flood runoff and periodic droughts, agriculture was limited by both climate and geomorphic processes that placed constraints on water control.

Yet despite difficulties with unstable farming terrains in these geographic zones, Spanish settlers coming into the area post-1540 CE found land being productively farmed by indigenous peoples (Doolittle, 2000) who ingeniously modified the landscape to capture and store intermittent rainfall and snowmelt to sustain crops (Anschuetz, 2001, 2006; Plog, 2008). On the Colorado Plateau, for example, most of the agriculture noted by the Spanish was floodwater farmed. Periods of drought that deteriorated grasslands and amplified erosion during heavy rainstorms, together with periods of light, but more frequent, rains caused continuous infilling of valleys with alluvium. Farming required adaptive responses to prehistoric climate variability influenced by El Niño and La Niña rainfall and drought periods, respectively (Dean and Robinson, 1977; Fish and Fish, 1984; Doolittle, 1992; Damp et al., 2002). Several of the major prehistoric (pre-Columbian) Indian societies of the Colorado Plateau (e.g., Anasazi and Mogollon) and southern Basin and Range areas outside of the Plateau (e.g., Hohokam and Patayan) farmed floodplains watered by melting winter snow and summer rains. Additional hill-slope terracing was used to stabilize planting surfaces, and flood diversion dams were built to limit erosional/depositional effects on field plots (Doolittle, 1992; Lightfoot and Eddy, 1995; Doolittle, 2000; Anschuetz, 2001).

Diverse technical innovations founded upon highly evolved indigenous cultural knowledge allowed for successful crop production in distinct geomorphic settings (Woosley, 1980; Doolittle, 1992, 2000). For example, the prehistoric Puebloans (Anasazi) located on Mesa Verde in Colorado (Ferguson, 1996) constructed a series of four reservoirs (Box Elder, Morefield, Far View, and Sagebrush) that were operational from 750 to 1180 CE and captured rainfall runoff to redistribute water for agricultural and domestic use (Leeper, 1986; Wilshusen et al., 1997; Wright, 2003). Ethnographically, the Tewa of north-central New Mexico employed bermed terraces to capture rainfall, together with stone-lined transport canals, dams, and spreaders to capture (or divert) runoff. The Tewa could exploit a combination of direct precipitation, intermittent runoff, groundwater, and canal extraction from springs and rivers to water their fields (Doolittle, 1992, 2000; Anschuetz, 2001, 2006). Stone-mulched and stone-bordered sunken pits were also used by Pueblo society to trap precipitation. Anschuetz (2001, 2006) describes anecdotal evidence that winter snow was rolled into balls and deposited in these pits in order to store water and amplify soil moisture for later agricultural use. The Hohokam (600–1350 CE) of south-central Arizona utilized extensive canal networks drawn from rivers and springs to irrigate vast field areas. The Salt River Valley contained as many as 400 km of main and distribution canals (Howard, 1987; Doolittle, 2000; Plog, 2008) with the Gila and Verde Valleys containing yet more irrigation canals estimated to be on the order of 600 km in cumulative length.

South-central Arizona employed the greatest extent of canal irrigation compared to all other southwestern indigenous societies. Aerial photography of these prehistoric canal and field system complexes taken 80 years ago (Judd, 1930) has proven indispensable in discovering and documenting trace canal, and field system remains now obliterated by erosion, sediment deposition overlays, and modern agriculture proceeding from urban expansion. Canal water transport technologies practiced by the Pima (Akimel O’odham) along the Gila River involved long, low-slope, open channels that supplied field systems. While similar water control systems were used elsewhere in the Southwest, canals originating from smaller river tributaries to major rivers were a preferred strategy due to easy water control practices. Other agricultural practices depended upon floodwater farming, especially in areas lacking large, perennial rivers. For example, the Papago (Tohono O’odham) were known for their ak chin (floodwater) farming along ephemeral streams, while the Navajo and Hopi planted fields in drainage areas where floods and runoff occurred during heavy rains (Plog, 2008). Further innovative indigenous agricultural strategies practiced in the Southwest are summarized by Doolittle (2000) and Plog (2008).

Modern scientific techniques integrated into archaeological studies add greatly to our understanding of complex geomorphologic processes and the response of indigenous societies to challenges posed by climate and landscape limitations. For example, 14C and luminescence dating (Berger et al., 2009; Watkins et al., 2011), as well as pollen and biometric analysis of sediment layers in canals and reservoirs, provides insight into age, use history of water control features (Huckleberry, 1999; Wright, 2003; Wright et al., 2005; Wright, 2006), and an understanding of crop types farmed by different societies. Additionally, much has been learned about prehistoric canals and fields through analysis of the physical-mechanical properties of sediments and alluvial deposits (hydraulic conductivity, porosity, stratigraphy). These studies provide insight into rain infiltration and seepage rates, as well as details illuminating the formation processes of canals and reservoirs. For example, sedimentological and stratigraphic analyses of Anasazi mesa top and valley water storage reservoir systems and canals were essential to understanding their role in sustaining local farming communities (Rohn, 1977; Wright, 2003; Wright et al., 2005; Wright, 2006). Application of concepts from fluvial geomorphology (Knighton, 1998) (e.g., erosion initiation, sediment transport and deposition) has proven useful in recognizing the impacts of floods and climate change on indigenous farming in the American Southwest (Cooke and Reeves, 1976; Bettess and White, 1983; Abrahams, 1987), as well as post-abandonment weathering of agricultural landscapes. When combined with dendrohydrological studies (Dean and Robinson, 1977), fluvial geomorphic analysis provides insight into how indigenous Southwestern societies changed their irrigation strategies (which are detectable from archaeological studies) as an adaptation to climate and landscape changes.


A survey of urban/agricultural water supply systems of major New and Old World societies on four regions of the world reveals exploitation of different varieties of water sources available in different ecological zones. Dams, reservoirs, canals, aqueducts, pipelines, open channels, and groundwater resources served to collect, transport, and distribute water to urban centers and agricultural fields. Each water system type with its selection of water transport and storage systems exhibited vulnerabilities when subject to climate and geophysical landscape changes. When system modifications were not possible due to insufficient technology, labor shortage, or lack of management expertise, societies underwent collapse, transformation, and altered societal and cultural trajectories as observed in the archaeological record. Differences exist between water transport and distribution systems employing different construction techniques and materials by New and Old World societies. Where the effects of climate and geophysical landscape change were minimal over long time periods, construction was permanent and alterations remedial in nature; where climate and weather patterns were changeable and affected the stability of water transport systems, flexibility of design and modification is evident to guarantee sustained use of these water systems. Examples discussed reveal this basic strategy difference between Old and New World societies. The many different water usage strategies employed by these societies constitute a virtual library of solutions tailored to different ecological and geomorphic conditions and provide insight into the creativity and resourcefulness of ancient engineers to maintain their communities despite changes in environmental conditions affecting their agricultural resource base.


  1. Abrahams, A. D., 1987. Channel network topology: regular or random? In Gardiner, V. (ed.), International Geomorphology, 1986: Proceedings of the First International Conference on Geomorphology, Part II. Chichester: Wiley, pp. 145–158.Google Scholar
  2. Anschuetz, K. F., 2001. Soaking it in: Northern Rio Grande Pueblo lessons in water management and landscape ecology. In Weinstein, L. L. (ed.), Native Peoples of the Southwest: Negotiating Land, Water, and Ethnicities. Westport: Bergin & Garvey, pp. 49–78.Google Scholar
  3. Anschuetz, K. F., 2006. Tewa fields, Tewa traditions. In Price, V. B., and Morrow, B. H. (eds.), Canyon Gardens: The Ancient Pueblo Landscapes of the American Southwest. Albuquerque: University of New Mexico Press, pp. 57–74.Google Scholar
  4. Bammer, A., 1988. Ephesos: Stadt an Fluss und Meer. Graz: Akademische Druck- u. Verlagsanstalt.Google Scholar
  5. Berger, G. W., Post, S., and Wenker, C., 2009. Single and multigrain quartz-luminescence dating of irrigation channel features in Santa Fe, New Mexico. Geoarchaeology, 24(4), 383–401.CrossRefGoogle Scholar
  6. Bettess, R., and White, W. R., 1983. Meandering and braiding of alluvial channels. Proceedings of the Institution of Civil Engineers, 75(3), 525–538.CrossRefGoogle Scholar
  7. Bourbon, F., 1999. Petra: Art, History and Itineraries in the Nabataean Capital. Vercelli: White Star.Google Scholar
  8. Coe, M. D., 2003. Angkor and the Khmer Civilization. New York: Thames and Hudson.Google Scholar
  9. Cooke, R. U., and Reeves, R. W., 1976. Arroyos and Environmental Change in the American South-West. Oxford: Clarendon.Google Scholar
  10. Crouch, D. P., 2004. Geology and Settlement: Greco-Roman Patterns. Oxford: Oxford University Press.Google Scholar
  11. Damp, J. E., Hall, S. A., and Smith, S. J., 2002. Early irrigation on the Colorado Plateau near the Zuni Pueblo, New Mexico. American Antiquity, 67(4), 665–676.CrossRefGoogle Scholar
  12. Dean, J. S., and Robinson, W. J., 1977. Dendroclimatic variability in the American Southwest, A.D. 680–1970. Final Report to the National Park Service, Department of the Interior. Tucson: Laboratory for Tree-Ring Research.Google Scholar
  13. Doolittle, W. E., 1992. Agriculture in North America on the eve of contact: a reassessment. Annals of the Association of American Geographers, 82(3), 386–401.CrossRefGoogle Scholar
  14. Doolittle, W. E., 2000. Cultivated Landscapes of Native North America. New York: Oxford University Press.Google Scholar
  15. Ferguson, W. M., 1996. The Anasazi of Mesa Verde and the Four Corners. Niwot: University Press of Colorado.Google Scholar
  16. Fish, S. K., and Fish, P. R., 1984. Prehistoric Agricultural Strategies in the Southwest. Tempe: Arizona State University. Anthropological Research Papers, Vol. 33.Google Scholar
  17. Guzzo, M. G. A., and Schneider, E. E., 2002. Petra. Chicago: The University of Chicago Press.Google Scholar
  18. Howard, J. B., 1987. The Lehi Canal System: organization of a Classic Period community. In Doyel, D. E. (ed.), The Hohokam Village: Site Structure and Organization. Glenwood Springs: American Association for the Advancement of Science, pp. 211–222.Google Scholar
  19. Huckleberry, G., 1999. Assessing Hohokam canal stability through stratigraphy. Journal of Field Archaeology, 20(1), 1–18.Google Scholar
  20. Judd, N. M., 1930. Arizona Sacrifices her Prehistoric Canals. Washington, DC: The Smithsonian Press. Explorations and Field-Work of the Smithsonian Institution in 1929.Google Scholar
  21. Knighton, D., 1998. Fluvial Forms and Processes: A New Perspective. London: Arnold.Google Scholar
  22. Kraft, J. C., Brückner, H., and Kayan, I., 1999. Paleogeographies of ancient coastal environments in the environs of the Feigengarten Excavation and the ‘Via(e) Sacra(e)’ to the Artemision at Ephesus. In Scherrer, P., Taeuber, H., and Thür, H. (eds.), Steine und Wege: Festschrift für Dieter Knibbe zum 65. Geburtstag. Vienna: Österreichisches Archäologisisches Institut. Sonderschriften 32, pp. 91–100.Google Scholar
  23. Laur, J., 2002. Angkor: An Illustrated Guide to the Monuments. Paris: Flammarion.Google Scholar
  24. Leeper, J. W., 1986. A computer model of the Mummy Lake water collection system in the Mesa Verde National Park. In Proceedings of the 6th Annual American Geophysical Front Range Branch Hydrology Days. Colorado State University.Google Scholar
  25. Lightfoot, D. R., and Eddy, F. W., 1995. The construction and configuration of Anasazi pebble-mulch gardens in the Northern Rio Grande. American Antiquity, 60(3), 459–470.CrossRefGoogle Scholar
  26. Longwell, C. R., and Flint, R. F., 1962. Introduction to Physical Geology. New York: Wiley.Google Scholar
  27. Moseley, M. E., 2001. The Incas and Their Ancestors: The Archaeology of Peru. London: Thames and Hudson.Google Scholar
  28. Murra, J. V., 1962. Cloth and its functions in the Inca State. American Anthropologist, 64(4), 710–728.CrossRefGoogle Scholar
  29. Ortloff, C. R., 2009. Water Engineering in the Ancient World: Archaeological and Climate Perspectives on Societies of Ancient South America, the Middle East and South-East Asia. Oxford: Oxford University Press.Google Scholar
  30. Ortloff, C. R., and Crouch, D. P., 2001. The urban water supply and distribution system of the Ionian city of Ephesos in the Roman Imperial Period. Journal of Archaeological Science, 28(8), 843–860.CrossRefGoogle Scholar
  31. Ortloff, C. R., and Kassinos, A., 2003. Computational fluid dynamics investigation of the hydraulic behavior of the Roman inverted siphon system at Aspendos, Turkey. Journal of Archaeological Science, 30(4), 417–428.CrossRefGoogle Scholar
  32. Ortloff, C. R., and Kolata, A. L., 1993. Climate and collapse: agro-ecological perspectives on the decline of the Tiwanaku State. Journal of Archaeological Science, 20(2), 195–221.CrossRefGoogle Scholar
  33. Ortloff, C. R., and Moseley, M. E., 2012. 2600–1800 BCE Caral: environmental change at a Late Archaic Period site in north central coast Peru. Journal of Andean Archaeology (Ñawpa Pacha), 32(2), 189–206.Google Scholar
  34. Ortloff, C. R., Moseley, M. E., and Feldman, R. A., 1982. Hydraulic engineering aspects of the Chimu Chicama-Moche Intravalley Canal. American Antiquity, 47(3), 572–595.CrossRefGoogle Scholar
  35. Ortloff, C. R., Feldman, R. A., and Moseley, M. E., 1985. Hydraulic engineering and historical aspects of the Pre-Columbian Intravalley Canal System of the Moche Valley, Peru. Journal of Field Archaeology, 12(1), 77–98.Google Scholar
  36. Plog, S., 2008. Ancient Peoples of the American Southwest, 2nd edn. New York: Thames and Hudson.Google Scholar
  37. Rohn, A. H., 1977. Cultural Change and Continuity on Chapin Mesa. Lawrence: The Regents Press of Kansas.Google Scholar
  38. Sandweiss, D. H., Shady Solís, R., Moseley, M. E., Keefer, D. K., and Ortloff, C. R., 2009. Environmental change and economic development in coastal Peru between 5,800 and 3,600 years ago. Proceedings of the National Academy of Sciences, 106(5), 1359–1363.CrossRefGoogle Scholar
  39. Scherrer, P., 1995. The city of Ephesos from the Roman period to late antiquity. In Koester, H. (ed.), Ephesos Metropolis of Asia: An Interdisciplinary Approach to its Archaeology, Religion, and Culture. Valley Forge: Trinity Press International, pp. 1–25.Google Scholar
  40. Shady Solís, R., 2000. Sustento socioeconómico del estado pristino de Supe-Perú: las evidencias de Caral-Supe. Revista Arqueología y Sociedad, 13, 49–66.Google Scholar
  41. Shady Solís, R., 2007. The Social and Cultural Values of Caral-Supe, the Oldest Civilization in Peru and the Americas, and Their Role in Integrated and Sustainable Development. Lima: Instituto Nacional de Cultura.Google Scholar
  42. Taylor, J., 2001. Petra and the Lost Kingdom of the Nabataeans. Cambridge: Harvard University Press.Google Scholar
  43. Thompson, L. G., Mosley-Thompson, E., Bolzan, J. F., and Koci, B. R., 1985. A 1500-year record of tropical precipitation in ice cores from the Quelccaya Ice Cap, Peru. Science, 229(4717), 971–973.CrossRefGoogle Scholar
  44. Thompson, L. G., Mosley-Thompson, E., Davis, M. E., Lin, P.-N., Henderson, K. A., Cole-Dai, J., Bolzan, J. F., and Liu, K.-B., 1995. Late glacial stage and Holocene tropical ice core records from Huascarán, Peru. Science, 269(5220), 46–50.CrossRefGoogle Scholar
  45. Watkins, C. N., Rice, G. E., and Steinbach, E., 2011. Dating Hohokam canals: a methodological case study. Journal of Arizona Archaeology, 1(2), 162–168.Google Scholar
  46. Whitten, D. G. A., and Brooks, J. R. V., 1982. A Dictionary of Geology. New York: Penguin.Google Scholar
  47. Wilshusen, R. H., Churchill, M. J., and Potter, J. M., 1997. Prehistoric reservoirs and water basins in the Mesa Verde region: intensification of water collection strategies during the Great Pueblo Period. American Antiquity, 62(4), 664–681.CrossRefGoogle Scholar
  48. Woosley, A. I., 1980. Agricultural diversity in the prehistoric Southwest. The Kiva, 45(4), 317–335.CrossRefGoogle Scholar
  49. Wright, K. R., 2003. Water for the Anasazi. Kansas City: Public Works Historical Society. Essays in Public Works History, No. 22.Google Scholar
  50. Wright, K. R., 2006. The Water Mysteries of Mesa Verde. Boulder: Johnson Books.Google Scholar
  51. Wright, K. R., Bikis, E., Wiltshire, R. W., and Pemberton, E., 2005. ASCE recognizes Mesa Verde prehistoric reservoirs. US Society on Dams Monthly Newsletter, 136(July), 8–12.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2017

Authors and Affiliations

  1. 1.CFD Consultants International, Ltd.University of ChicagoLos GatosUSA