The Shaṭṭ al-ᶜArab floodplain in the Basra Governorate of southern Iraq is covered by a vast network of linear earthwork ridges, covering an area of more than 800km², far more extensive than the area under conservation today. These are thought to have been built between 630 AD, when the city of Old Basra was founded, and the late ninth century AD, when the city was engulphed in a period of political turmoil.
These earthworks are thought to have been built by a workforce of slaves known as 'the Zanj'. The term 'Zanj', in Medieval Arabic refers to the Swahili Coast of East Africa, and it has been suggested that the slaves of Medieval Basra may have been sourced from that area, however, 'Zanj' is also sometimes used for slaves from other origins (possibly making the East Coast of Africa the 'Slave Coast').
Thus the Zanj of Basra have been thought by various authors to have been either East African slaves, a mixture of East African and locally enslaved peoples, or possibly slaves from other parts of Africa (it is unclear if slave routes from East Africa had been established by the period when the earthworks were built, but slave routes from West and West Central Africa via North Africa are known to have been well established.
In 869 AD the slaves of Old Basra rose up in what is now known as the Zanj Rebellion. This destbalized the region leading to over a decade of civil strife before the Abbasid rulers of Iraq managed to regain control of the area. Two earlier, and smaller Zanj rebellions were also recorded in 689-90 AN and 694-695 AD, although these were apparently local affairs put down quickly, whereas the rebellion of 869 had a wider and more long-lasting political impact.
At the time of the rebellion of 869, the Zanj are recorded to have been spread out in a series of work camps in the agricultural zone to the east of Old Basra, on both sides of the Shaṭṭ al-ᶜArab River. This coincides with the area covered by the system of Earthworks, so presumably key events of the rebellion took place within this system. The Zanj are recorded to have established their headquarters at a location called al-Muḫtāra within the agricultural area where they once toiled, although the precise location of this is unknown today.
The Zanj Rebellion has been attributed to the harsh conditions under which the Zanj were forced to live, although details of this are lacking in records. If the earthworks of Shaṭṭ al-ᶜArab are the result of Zanj labour, as is generally assumed, then this could have been the exploitative labour which triggered the rebellion. However, the Earthworks have never been accurately dated, making the relationship between the earthworks and the rebellion essentially a matter of guesswork.
In a paper published in the journal Antiquity on 2 June 2025, Peter Brown of the Department of History, Art History and Classics at Radboud University, and the Department of Archaeology at Durham University, Jaafar Jotheri, also of the Department of Archaeology at Durham University, and of the Department of Archaeology at the University of Al-Qadisiyah, Louise Rayne of the School of History, Classics & Archaeology at Newcastle University, Nawrast Abdalwahab of the Department of Geology at the University of Basrah, and Eric Andrieux, again of the Department of Archaeology at Durham University, present the results of a study of the Shaṭṭ al-ᶜArab earthworks, including the first chronology of their creation and use.
The ridges which make up the Shaṭṭ al-ᶜArab earthworks are typically between 400 and 1200 m in length, 20-50 m wide, and rise up to 2 m above the surrounding ground level. The Earthenworks comprise a series of blocks of parallel ridges separated by 100-250 m of flat land. The ridges abut large canal features at right angles, with smaller irrigation canals running along the centre of the flat areas between the ridges. Ultimately, these canal systems connect to the Shaṭṭ al-ᶜArab river, with some still carrying water up to 5 km from the banks of the river.
Both modern satellite imgery and satellite photography from the 1960s suggest that there are about 7000 ridges across the Shaṭṭ al-ᶜArab floodplain, ranging from about 15 km south of the town of Al Qurnah in the north to the town of Abadan in Iran in the south. The total length of the ridges is about 3000 km, implying a significant amount of labour by a large number of workers was needed to create the systm.
There are several potential explanations as to how the ridge-and-canal system was created and used. The ridges could either have been formed as part of a process of clearing the land for agricultural use, or formed over time as canals were repeatedly cleared of sediment. In 1962 the American archaeologist H.S. Nelson published a study of the ridge system, in which he established that the soil of the ridges was twice as saline as the surrounding soil, which led him to hypothesise that the irrigation practices used led to the soil becoming saline over time, and the ridges were built up as this saline soil was repeatedly cleared. However, it is unclear how many soil samples Nelson took, nor from where in the system, making it difficult to assess how applicable his findings are to the whole system.
The ridge system clearly has a relationship with the Shaṭṭ al-ᶜArab river, and it appears that the canal-system once allowed the cultivation of large areas of the Shaṭṭ al-ᶜArab floodplain through tidal irrigation. This area is above the saline estuary of the river, but twice a day tides in the Persian Gulf push freshwater to flow upstream within the river, and this can be utilized to feed an irrigation system. Similar systems are utilised elsewhere in the world, including West Africa, where it has been practiced for at least a thousand years, and Spain, where the practice can be traced back to at least the sixteenth century.
A number of medieval authors mentioned a tidal irrigation system in use arround Basra. The tenth century Arab geographer al-Muqaddasī described a system of canals with mills at their mouths, which were driven by the ebbing tide as the water withdrew following each tidal flood. Two centuries later the writer al-Idrīsī described a canal system around Basra, driven by rising tides and used to irrigate crops. The fourteenth century writer Abū al-Fidāʾ described ships using the tide on the Shaṭṭ al-ᶜArab to reach Basra and its canal systems. However, in the early seventeenth century, the Portuguese traveller Pedro Teixeira described the area as infertile and covered with salt, although with evidence of former agricultural activity still visible.
The most northerly of the ridges are close to the town of Al Qurnah, which is also approximately the northernmost extent of tidal flow on the Shaṭṭ al-ᶜArab, which supports the idea that these are connected. The uniformity of the ridge-and-canal system throughout the Shaṭṭ al-ᶜArab floodplain also indicates that this system was intended for a single form of agriculrure.
However, what sort of crops were grown there is less clear today. H.S. Nelson thought that the spaces between the ridges would have been suitable for the cultivation of Winter Barley, whereas medieval authors associated the Basra area with Date production, and dates are still farmed along the Shaṭṭ al-ᶜArab today. Ceramic vessels from southern Iraq, which are believed to have been used to transport Dates, are common finds at archaeological sites as far away as China. Dates also form an important part of al-Ṭabarī.’s account of the Zanj rebellion, with the rebels described as hiding in Palm-groves on the salt flats to ambush Abbasid troops, and Abbasid cavalry being impeded by the densely packed Date trees. The famous fourteenth century travelled Ibn Baṭṭūṭa noted that '“no place on earth exceeds [Basra] in quantity of Palm-groves', and Pedro Teixeira also considered the area to be notable for date production.
Dates are a relatively salt-tolerant crop, although the levels of salt recorded by H.S. Nelson would be a serious problem, reducing crop-yields by about 50% in the least affected areas between the ridges. Furthermore, Dates are seldom, in traditional agriculture, grown on their own. Rather they tend to be intercropped with other produce, such as fruits, vegetables, Rice or Barley, which benefit from the shade provided by the Date trees. If crops require more light, the Dates can be planted further apart.
Medieval literature does not provide any direct descriptions of the system used on the Shaṭṭ al-ᶜArab floodplain. However, a number of sources may provide glimpses of how this worked. Al-Ṭabarī.’s description of the Zanj Rebellion makes several references to hills on the floodplains, including one hilltop which overlooked a Date Palm grove. Since there are no natural hills on the Shaṭṭ al-ᶜArab floodplain, it seems likely that this refers to the artificial ridge structures. An anonymous eleventh-century chronicle, the Kitab al-‘Uyun, refers to the Zanj creating hills to make the area cultivatable, but provides no further details.
Brown et al. first built up a map of the entire area, using a combination of modern sateellite imagary and CORONA satellite images from the 1960s. Based upon this they then chose sites for sample collection across the area. The samples collected were then subjected to radiocarbon and optically stimulated luminescence dating, with the later being particularly useful, since it can date how long mineral grains have been exposed on the surface (the technique was developed for use on samples returned to Earth by the Apollo Lunar Program) and does not therefor require the finding of organic objects, which can be difficult in environments such as the Shaṭṭ al-ᶜArab floodplain.
Many of the ridge structures show signs of recent damage, due to the construction of infrastructure for oil fields, and the First Gulf War of 1990-91. Brown et al. wished to collect samples from areas undisturbed since the abandonment of the system, and therefore avoided areas with clear signs of recent disturbance. They also wished to take samples from different parts of the system, to determine whether the whole network had been in use at once, or whether different parts date from different periods. Fieldwork was carried out in January 2022, with exploratory trenches dug into four ridge tops and one relict canal-bed.
No real stratigraphication was found on any of the ridges; each had an upper layer of topsoil with gypcrete crystals, between 50 and 200 mm thick, beneath which the ridges were composed of unstratified silty clay. The trench excavated in a canal bed was dug to 95 cm, and encountered five distinct layers of clay and/or silty clay. Two of the ridge-top trenches produced fragments of shell which were retained for carbon dating. Samples were collected from all sites for optically stimulated luminescence dating.
The first of the shells subjected to radiocarbon dating yielded a date of between 58 and 215 with a probability of 92.6%, while the second produced a date between 326 and 433 AD with a probability of 81.7%.
The optically stimulated luminescence dating results all yielded dates between the late ninth century and the mid thirteenth century AD, though a clear chronology could not be established as uncertainty intervals between samples overlapped and samples taken from deeper within the ridges were often younger than those found on the surface. The wide uncertainty values produced by the samples mean that each ridge could have been formed in a single episode, or built up over about a century. The oldest samples post-date the Zanj Rebellion.
Both shell samples yielded radiocarbon ages significantly greater than the optically stimulated luminescence dates from the surrounding sediment. This is not particularly surprising, as the process of excavating sediments and incorporating them into the ridges is likely to also excavate older shells and incorporate those.
Brown et al.'s excavations were limited, and all the samples come from within the homogeneous ridge-cores rather than reaching the undrlying sediments. This only establishes that the ridges were in existance in the century after the Zanj Rebellion (working on the basis that the ridges were a gradual accumulation, as sediments were continuously removed from the canals over time, a process without which the canals would have become blocked).
The sediments collected from the canal yielded slightly older dates than those from the ridges, with the oldest dates (between the eighth and the mid-eleventh centuries). Above this layer is a dark green layer about 10 cm thick, with the layer above this yielding a date of 1423-1503 AD. The dates above and below this green layer are before and after the dates obtained from the surrounding ridges, while th green layer itself appears to represent a period of low sediment deposition, consistent with the removal of sediment by Human activity.
Brown et al.'s results demonstrate that the ridges of the Shaṭṭ al-ᶜArab floodplain are of medieval origin, and that material was being deposited onto them between the ninth and thirteenth centuries. This is contrary to expectations for the region, where it is generally assumed peak agricultural activity occurred prior to the Zanj Rebellion, and that this was followed by a period of reduced economic activity. It is, however, in keeping with the accounts of visitors to the area at the time, who record that Basra was smaller and less populous than before, but still prosperous and well kept. These visitors also continued to record an active canal system in the area, suggesting the assumption that agricultural practices had been curtailed is probably erroneous.
Brown et al.'s work does not identify who built the earthworks of the Basra hinterland, but it does demonstrate that large-scale, labour intensive agriculture was still happening in the area after the Zanj Rebellion. It has been claimed that the rebellion led to a reduction in the number of slaves being used by the Abbasids, although this is not universally accepted, and the writer Nāsir Khusraw records large numbers of African slaves being used in agriculture at al-Ahsa in eastern Arabia (also under Abbasid rule) in AD 1050. If the Zanj Rebellion did not lead to the abandonment of large-scale agriculture on the Shaṭṭ al-ᶜArab floodplain, then some other event must have been responsible. Brown et al. suggest that this may have been the Mongol invasions of the mid-thirteenth century, which are known to have damaged the economic importance of Baghdad and the Gulf region, potentially also reducing the need for large scale economic production in southern Iraq.
It is also possible that climate change played a role in the abandonment of this system. The climate of the Gulf region, and in particular the Tigris catchment have been aridifying steadily since the mid tenth century. Since the Tigris flows into the Shaṭṭ al-ᶜArab, it is quite possible that a point was reached where water levels in the river were too low to sustain a large scale irrigation system on that river's floodplain.
Another possibility is that the agricultural system of the Shaṭṭ al-ᶜArab floodplain collapsed as a result of the plague outbreaks of the fourteenth century. These are estimated to have killed about 42% of the population of the Middle East, with Baghdad's rural hinterland known to have been particularly badly hit. If this was also the case in the Basra hinterland, then it would have been almost impossible to keep a large-scale, labour intensive agricultural system going, particularly one dependent on the constant maintenance of irrigation systems.
It cannot be determined, at this time, if any of these factors, or some combination of them, was responsible for the abandonment of the Shaṭṭ al-ᶜArab earthwork system. There was clearly a major shift in habitation in the area between the mid-fourteenth and early sixteenth centuries, with the decline and then abandonment of the city of Old Basra, in favour of the new city at the current location, on the banks of the Shaṭṭ al-ᶜArab, as well as a great reduction in the area under cultivation on the banks of the river.
Brown et al.'s work runs contrary to the traditional narative of the history of Basra, which is seen as having had its heyday in the Ummayad (661-750 AD) and early Abbasid (750-about 900 AD) periods, followed by a long period of decline. Instead it demonstrates that large-scale agriculture was still occurring in the ninth to thirteenth centuries, suggesting that the Zanj Rebellion did not have the long term economic significance it is sometimes ascribed.
However, Brown et al. only looked at four of the 7000 plus ridge systems which cover the Shaṭṭ al-ᶜArab floodplain, making it difficult to assert the extent to which the dates obtained reflect the chronology of the wider area. All of the ridges studied are on the west bank of the Shaṭṭ al-ᶜArab, and fairly close to the location of Old Basra. Potentially, this might have been the last part of the system as it was abandoned across the floodplain, being closest to the city dependent on it.
See also...
