IMTO Italian Mission to Oman University of Pisa PRELIMINARY REPORT (FEBRUARY MARCH 2007)

Transcription

1 IMTO Italian Mission to Oman University of Pisa PRELIMINARY REPORT (FEBRUARY MARCH 2007)

2 The archaeological excavations on the site of Salut were carried out by the team of the Italian Mission to Oman - IMTO - directed by Alessandra Avanzini, from till The members of the team were: Carl Phillips, Muhammad al-jahfali, Chiara Condoluci, Mauro Cremaschi, Marco Iamoni, Riccardo Santoni. The 2007 excavations at Salut have provided further information about the Iron Age of South East Arabia (c.1300 to 300 BC), the period for which the site is best known. In addition to this, the first architectural evidence for a Bronze Age presence at the site has been uncovered, as well as further details about the occupation of the site in Islamic times. Bronze Age remains at Salut. Results fom one of the first archaeological surveys in the Sultanate of Oman showed that the Wadi Bahla is rich in Bronze Age sites and espcially in the area near Bisya (Hastings, Humphries and Meadow 1975). Several large circular towers and the distinct profile of tombs along the crests of adjacent mountains create a distinct Bronze Age landscape that dates from the third millennium BC. Because the hill on which the site of Salut is situated occupies such a prominent position in the landscape, it would hardly be surprising to find that it was also incorporated in the Bronze Age cultural landscape. Architectual evidence suggesting that this was indeed the case has now been revealed during the 2007 excavations. On what would have been the highest point on the hill of Salut, the foundations of what appears to have been a large circular structure were found beneath Iron Age deposits which in turn had been disturbed by the later Islamic occupation of the site (Fig. 1). The evidence for a circular Bronze Age structure is far from complete and comprises the remains of four concentric walls that appear to form an arc, which, when complete, would have formed a circular structure with a maximum diameter of approximately 14 metres (Fig. 2). The four curved walls are built of large stones with smaller angular stones used to fill the space between the walls. It should be emphasised that this remnant structure lies beneath later Iron Age deposits and its curved plan is not at all consistent with the lay-out of the Iron Age building. Unfortunately no finds can be related to the early structure but the probable circular plan and method of construction, especially the concentric nature of the walls, present a striking resemblance to the plan of some Early Bronze Age tombs. The best published example with which the plan can be compared is of a tomb excavated some years ago at Tawi Silaim (de Cardi 1977: 20, fig, 2). The remains are also similar to the arrangement of concentric walls that form the structure of the characteristic bee-hive tombs that are frequently found on hill-tops and ridges, and as seen in the area surrounding Salut. It must be added, however, that if the remains excavated at Salut are the foundations of a similar type of tomb then it would have been an extremely large example of its sort. That said, the location on top of the hill at Salut would be entirely consistent with the siting of the monuments that form such a distinctive element of the Bronze Age landscape. If the remains described above are accepted as being part of the foundations for a tomb, then the likely date of its construction would have been at sometime during the first half of the third millenium BC. Presumably the tomb remained a prominent feature until the Iron age period when, if not already in a state of disrepair like the tombs still to be seen on the mountain ridge east of Salut, it was razed to the ground and covered by Iron Age foudations / mud-bricks. That this appears to have benn is suggested by the dense accumulation of small angular stones, characteristic of the fill between the concentric walls, which have become dispersed over the summit of the hill. Having been overlain by Iron Age mud-bricks, this layer of stones has later been interupted by pits and hearths dating from the Islamic period.

3 Additional Bronze Age finds from Salut Occupying such a prominent position within the surrounding Bronze Age landscape it is hardly surprising that a few artefacts from this earliest period should also be found at the site objects picked up fom the surrounding plain and mountains and brought to the site during the Iron Age or later times. An example of this is provided by a typical block of ashlar masonry that would have origially formed part of the facade of an Umm an-nar tomb (c to 2000 BC). Removed from its original location (such tombs are usually located on the plain, near to a contemporary settlement), this particular example (Fig.3 ) was found amongst the collapsed stones of an Iron Age wall at the southern end of the site. Amongst the thousands of Iron Age and Islamic potsherds collected at the site, both from the surface and excavated contexts, there are only two examples that might possibly be of pre-iron Age date. With fragmentary soft-stone vessels it is possible to be more certain and so far five fragments from Salut can be dated to the Bronze Age. One of these (fig. 4a ) dates from the latter part of the Early Bronze Age (c to 2000 BC), whilst the remaining four fragments (Figs 4b and 5a-c ) can be dated to the Middle-Late Bronze Age (c. 2000to 1300 BC). The main period of occupation evident at Salut thus remains the Iron Age (c to 300 BC), but before considerig some of the main features of the site during this period, it is worth reviewing what is known (archaeologically speaking) of the Islamic occupation at the site. Islamic Salut The Islamic re-occupation of Salut is marked by a wide range of archaeological features and a varied ceramic assemblage, as well as a number of interesting small-finds. It is clear that the massive external wall that defined the Iron Age site has remained largely intact until quite recently - at least along the northern and eastern sides of the site. So, even if the original intention of this wall, in Iron Age times, was not defensive, then for the later Islamic occupants of the site it would have certainly served this purpose if and when necessary. Over the whole of the summit of the hill Islamic hut-circles have been recovered (Fig. 6). These are evident as stone foundations or, in some instances, an arrangement of post holes, usually forming a circular structure about three metres in diameter (Fig. 7). No evidence for the superstrucure of these buildings has been found, but it possibly comprised pise or wattle walls and a thatched roof. Such a reconstruction is suggested by the vernacular architecture still to be seen in some regions of Oman. Evidence of small fire-places inside the huts is sometimes present, whilst evidence for more substantial ovens and elaborate hearths has been revealed, especially on the highest part of the site which seems to have acted as a focal point for the Islamic occupants. In addition to hearths, ovens and pits, the settings for rotary querns have also been recorded. The foundations of the Islamic buildings and associated installations are sometimes found directly on top of the underlying Iron Age surface, whilst just as frequently the foundations and installations have been cut into the Iron Age layers and sometimes even down into the underlying bed-rock. The most abundant artefact category from Islamic contexts is, of course, pottery. There is a wide range of domestic wares that includes water jars, cooking pots and large storage vessels as well as a wide range of finer glazed wares. A comprehensive study of this pottery assemblage has still to be done. It can be stated, however, that amongst the glazed wares, there is a relatively very small amount of blue-glazed Abbasid wares that might date to the 10 th century or even earlier (Fig. 8). The bulk of the glazed wares have, however, been compared and dated to types more typical of the

4 14 th 15 th Century (Figs 8 and 9) and this would appear to represent the main period of Islamic occupation at the site (Whitcomb 1975). In addition to pottery, other finds from Islamic contexts have included coins (Fig.10), utilitarian bronze objects (Fig 10b), decorative bronze objects such as belt and dagger fittings (Fig. 11), a small ivory container (Fig. 12a), a variety of small glass vessels (Fig.12b-f) and various beads and bangles (Fig.12g). The manufacture of soft-stone vessels and untensils continued in Oman until recent times and a few of these products have been found in the excavation at Salut (Fig. 13). A striking aspect of the Islamic archaeological evidence from Salut is its overwhelmingly domestic nature, by which is meant, it indicates all the activities one normally associates with daily-life; shelter, food preparation, cooking, storage facilities and a range of pottery and stoneware adapted to these needs. In addition there are a few probable luxury or personal items, such as small glass vessels and bronze ornaments. This picture of domestic-life is in marked contrast with the emerging picture of the Iron Age occupation at Salut. Iron Age Salut With an increasing amount of excavation the general lay-out of the Iron Age occupation at Salut is becoming clearer (Fig. 14). With this, it becomes more evident that the site does not constitute a regular Iron Age settlement; there is no concentration of houses and evidence of all the domestic activities that one might associate with these is absent. Rather, the site appears to have had some special purpose which is reflected in its architecture and the assortment and quantity of certain artefacts that have been recovered. In the 2007 excavations, the extent of the outer wall on the north and east sides of the site has been more fully exposed. The wall has been traced over a distance of 60 metres, curving from the north west corner of the site around to the east side. On the east side there is a large bastion or tower which projects onto the plain below for a further 25 metres. In some places this outer wall is preserved to well over four metres in height. Given that the foundations are also built mostly on the upper slopes of the hill, then, from a distance, the site presents a most impressive view. Whether the function of this monumental wall was defensive or just intended to impress, or perhaps both, is difficult to say, but so far it is impossible to say to what extent the wall might have projected above the large flat platform that was constructed on top of the hill. The mud-brick platform which crowned the top of the site in the Iron Age period has now been fully uncovered and seen to cover an area of approximately 30 metres, along a roughly north-south axis, by approxiamtely 20 metres, along a roughly west-east axis. Unfortunately, much of the southern half of the Iron Age platform has been all but destroyed by later Islamic features. The main walls that define the platform do, however, remain intact. Where the Iron Age platform is well preserved there still needs some further detailed excavation to determine what features there might have been on it. So far, there is very little evidence for any standing architecture apart from two, perhaps three, small rooms on the westerside. It is also clear that there was a basement room (rooms?) on the north side and partly incorporated in the outer wall, but for how long this space remained to be used and whether its construction precedes the main Iron Age phase of occupation also requires further investigation. The only other Iron Age features built on the mud-brick platform are two ovens, one of which had a bronze cauldron associated with it (see Report for 2005). On the east side of the platform there appears to be a number of terraces that follow the contours of the underlying hill-slope. So far, three terraces have been defined and it remains to be seen how far

5 this pattern is continued down the slope and how the upper part of the site relates to the bastion, or tower, that projects on to the plain at the base of the hill. Work continued in this area in 2007 and revealed that one moves down the slope there is a considrable build-up (approximately 2 metres) of later deposits that comprise collapsed walls and accumulated debris which has subsequently been disturbed by deep pits and hut-foundations from the Islamic period. A similar pattern of terraces appears to exist on the southern side of the mud-brick platform and it appears that at some point these become articulated with the outer wall as it continues around to enclose the southern end of the site. In the south west corner of the site, the continuation of the outer wall, which clearly appears to run along the western side of the mud-brick platform, has been further excavated in In this area it appears that the make-up of the wall is characterised by a massive build of mud-bricks with internal divisions that have the appearance of rooms, sometimes with stone foundations. There is an absence, however, of any recognisable floors and instead the rooms/spaces were filled with mudbricks, and form part of the internal structure of the outer wall. This same characteristic construction technique has been investigated along the western side of the mud-brick platform. In this instance it was fortunate that one of the rooms/spaces had been left vacant for some time (perhaps during further construction work) and a quantity of rubbish was deposited on the exposed surface. This thin accumulation of rubbish has proved to be of great interest and was further excavated in The rubbish deposit (US 74 and 75) contained at least one basket containing dates. As well as informing about the probable oasis environment near to the site, the preserved date stones will provide ideal samples for C14 dating and give a clear indication of when this part of the wall was constructed. The actual baket was also made from date-palm and gives a unique indication of the kind of containers that were made from organic materials and thus rarely preserved in the archaeological record. The basket remains show that it was woven using a technique whereby the strips of palm create a distinctive pattern known as twill (Fig. 15a ) Also, associated with the basket remains were the fragments of an S- spun fibre (probably date fibre) that has been used to make a loosely woven strap in which the strings of fibres are grouped in fours and plaited in the same manner as the fibres of the basket, so as to form a twill pattern (Fig. 15b). In addition to the organic remains, the rubbish deposit contained three pots, one a complete bowl (Fig. 16 a), part of a jar (Fig. 16b) and part of a long-handled bowl (Fig. 16c). More specifically, the latter comprises the handle which is decorated with an applied snake. More surprising still was the recovery of two bronze ladles (Fig. 17) conatained in the rubbish deposit along with a bronze snake (Fig. 18a) and a small bronze axe (Fig. 18b). The bronze axe is clearly not a functional axe and pehaps it had a decorative, possibly amuletic purpose. The combination of items recovered from US 74 and 75 provides a good reason to consider the range of Iron Age objects that have been found at Salut and what these might indiacte about the site. Long-handled bowls

6 A distinct type of pottery vessel found at Salut is the long-handled bowl (Fig. 19a). As the name implies, this comprises a small bowl protruding from which is a long handle. Its form might evoke some alternative description such as brazier, censer or incense-burner because of the similarity with objects of more certan function. It is very clear from the numerous examples found at Salut that burning took place on the inside of the bowls. A further characteristic of these vessels is that the handles are usually decorated. The decoration might comprise a simple chevron pattern, but other examples have applied decoration in the form of a snake. In some cases the body of the snake rises towards the rim of the bowl, and some examples appear to have had snakes shown in relief on the inside or ouside surface of the bowl. In addition to snakes, there is a particularly fine example of a handle decorated with fish and a ram s head. Similar long-handled bowls do not appear to have been found at the majority of Iron Age sites excavated in South East Arabia. There is, however, at least one exception, al-qussais near Dubai, where they have been reported, though whether in such great a number as at Salut is not kown. But interestingly, the site of al-qussais also produced a number of small bronze snakes (from the socalled Mound of Serpents ), smaller but nonetheless comparable with examples that have been found at Salut. Snakes of Bronze To date no fewer than six bronze snakes have been found at Salut. This includes the example referred to above from US 74/75 which was found in The latter example is distinct from the other snakes found at Salut in that it has clearly been forged and is comparable with the bronze snakes found at al-qussais. The remaining snakes are more substantial and appear to be made of cast bronze and have sometimes then been decorated with the application of numerous spherical pieces of bronze, giving a very distinctive spotted appearance. The bronze snake found in 2007 is the largest one found so far and measures approximately 26 centimetres in length (Fig. 19b). It is difficult to imagine that such objects served a regular daily purpose perhaps they had an apotraic function. Either way, it is surprising to see that bronze, presumably an expensive material, should be used for such objects unless they were deemed to be of some importance. Snake-decorated pottery The theme of snakes is also found on other pottery vessels. The pottery assemblage at Salut includes various types of storage jars and lids and in several instances these are seen to be decorated with snake designs (Fig. 20a). Whilst this is not totally unique since examples of vessels with snake decoration have been found elsewhere, the frequency with which these vessels occur at Salut appears to be far greater than at other sites. Small carinated bowls A further category of pottery that is abundant at Salut is that of small carinated bowls (Fig, 20b). Such bowls have long been taken as a key indicator of the Iron Age in South East Arabia. Frequently these are found with burials, in collective tombs, where painted and un-painted examples usually form the most common type in the tomb s pottery assemblage. They are also found on settlements and no doubt acted as every-day drinking cups. The quantity found at Salut, however, appears disproportionately high and especially in some of the sealed Iron Age contexts that have been excavated. It might be that in some situations (e.g. in burials and at Salut) this

7 particlar type of vessel took on some special significance distinct from being simply a cup or drinking vessel. Comments The combination and frequency of the first three object categories described above, long-handled bowls, snakes of bronze and snake decorated pottery, and possibly the abundance of small carinated bowls, alongside the absence of the usual remains of the sort manifest at other Iron Age sites (e.g. dwelling places, hearths, pits etc), suggest that Salut was an extra-ordinary place which served some sort of ceremonial or religious role. As such, it is unique amongst Iron Age sites excavted in the Sultanate of Oman and the United Arab Emirates, with the possible exception of the Mound of Serpents at al-qussais which unfortunately was never fully explored and has sadly now been destroyed. Clearly, more detailed excavtion and comparative studies need to be done in order to substantiate some of the views expressed above, but whatever the true picture might have been, the impression that Salut was an important place in Iron Age times is quite clear. The architecture of the site, lacking all the features of an everyday settlement, nevertheless is the product of a massive undertaking. The extent and grandeur of the outer wall, as shown in 2007, indicates that the site was meant to be monumental it dominates the Salut plain. The masive amount of mud-bricks and the systematic lay-out of the site also implies some sort of central authority or co-ordination of the enterprise. This of course begs some more questions, in particular, if this is not a settlement occupied by people engaged in the more mundane aspects of daily life, then where did the people who built Salut live? In addition to this, one then needs to explain how such a presumably large community could sustain itself. It will be seen from the accompanying reults of the 2007 geoarchaeological survey that there is possibly an Iron Age settlement located not far west of Salut. It is also possible that, if a site is confirmed in this location, then it could be related to field systems that were irrigated with water derived from a falaj system. Whatever the case might be, it is clear that further work at Salut is justified and it is also time to envisage an extended goal which will combine the work at Salut with fresh investigations focussed on the surrounding plain. This will provide a fuller picture of Iron Age society in this part of Arabia. Conclusions The excavations at Salut in 2007 have helped to confirm a number of details about the site during the Iron Age period and how it compares and contrasts with contemporary sites found elsewhere. It is possible that Salut had some special ceremonial purpose. As such it will have occupied a dominant position in the Iron Age landscape this would have been determined by its physical position and further enhanced by the monumental architectural edifice built at the beginning of the Iron Age. It has been shown that Salut possibly formed part of an earlier Bronze Age cultural landscape and whether this provided any sort of legacy which led to its being a specially chosen location in later times can only be conjectured. But, in a similar sort of way, it is well knowm that pre-islamic Salut still plays an important part in the oral, and early written history of the region. The relationship betwen local history and wider regional history is a potentially very interesting research topic and one where studies at Salut can make an important contribution. From a more practical, but still complementary perspective, it is of some importance to locate the settlement or settlements that the site at Salut served. If such settlements could be located and excavated this would provide for a

8 more comprehensive picture of Iron Age society and provide important information about how the settlements were sustained, their irrigation and husbandry practices and their contacts with the wider world. Suggestions The work at Salut should continue for the foreseeable future. We are planning a shorter season for about four weeks. This would provide an opportunity for detailed inspection of stratigraphy at Salut and importantly, an archaeological survey of the surrounding area. The detailed examination of stratigraphy etc. would require a minimal number of labourers no more than 10. The production of an archaeological map, and the possible sounding of any sites found, is also one of the desired requirements of any further geoarchaeological survey. For a second season of mainly excavation 30 labourers would be required and the continued employment of local national Omanis should be further encouraged. To make most of this opportunity a more disciplined regime will need be implemented, a positive aspect of which would be the acquisition of skills that local people might be able to use in finding longer term employment opportunities at a time when heritage related expertise is required for projects in Bahla and elsewhere. At Salut the excavations should, in the meantime, be aimed at completing the excavation on the summit of the hill and connecting this with the tower that extends on to the plain on the east side of the site. Now that a large section of the outside wall at Salut has been exposed, it is an appropriate time to obtain advice about how this can be best preserved and possibly restored along with other parts of the site. A programme of conservation and restoration should also be considered, in detail, alongside plans for the presentation of the site to the public. There further remains the obligation of providing information about the site at both academic and popular levels. The first of these can be fufilled by further publications in appropriate journals, conference proceedings and mongraph formats. Careful consideration should be given to what would best serve the public whilst further work is still on-going. Needless to say, any provisions will need to cater for both local Omanis and the foreign tourist trade.

9 Fig 3. Ashlar facing stone from an Umm an-nar tomb, found in a reused context at Salut.

10 Figure 4 Bronze Age soft-stone vessels found at Salut.

11 Figure 5 Bronze Age soft-stone vessels found at Salut.

12 Fig. 7 Islamic hut-circle (Structure 29) excavated in 2007.

13 Fig. 8 A selection of Islamic glazed-wares found at Salut. The top row consists of Early Islamic Blue-glazed-wares. The remainder are probably dated to the 14 th -15 th Centuries AD.

14 Fig. 9 Further examples of Islamic glazed-wares found at Salut.

15 Fig. 10a Islamic coins found at Salut. Fig. 10b Utilitarian bronze object found in 2007.

16 Fig. 11 Bronze objects of Islamic date.

17 Fig 12 Ivory and glass objets from Islamic contexts at Salut.

18 Fig 13 Soft-stone items of Islamic date.

19 Fig. 15 Organic materials from US 75: a) Detail of basket made from woven palm-leaf. b) Detail of material woven from twisted palm-fibres.

20 Fig. 16 Pottery items from US 75: a) Bowl; b) part of a jar; c) part of a long-handles bowl with applied snake decoration.

21 Fig.17 Bronze ladles from US 75.

22 Fig. 18 Bronze objects from US 75: a) Bronze Snake; b) Bronze Axe / amulet.

23 Fig. 19 Characteristic finds from Salut: a)long-handled bowl. b) Bronze snake found in 2007.

24 Fig.20 Characteristic finds from Salut: a) Applied snake decoration on a large storage vessel. b) Carinated bowl found in 2007.

25 Fig. 21 Salut viewed from the surrounding plain.

26 LIST OF FIGURES Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 The location of probable Bronze Age remains at Salut. Part of the foundations of a Bronze Age tomb. Ashlar facing stone from an Umm an-nar tomb, found in a repused context at Salut. Bronze Age soft-stone vessels found at Salut. Bronze Age soft-stone vessels found at Salut. Plan of the Islamic settlement. Figure 7 Islamic hut-circle (Structure 29) excavated in Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 A selection of Islamic glazed-wares found at Salut. The top row consists of Early Islamic Blue-glazed-wares. The remainder are probably dated to the 14 th -15 th Centuries AD. Further examples of Islamic glazed-wares found at Salut. a) Islamic coins found at Salut. b) Utilitarian bronze object found in Bronze objects of Islamic date. Ivory and glass objets from Islamic contexts at Salut. Soft-stone items of Islamic date. Plan of Iron Age Salut. Figure 15 Organic materials from US 75: a) Detail of basket made from woven palm-leaf. b) Detail of material woven from twisted palm-fibres. Figure 16 Pottery items from US 75: a) Bowl b) part of a jar c) part of a long-handles bowl with applied snake decoration. Figure 17 Bronze ladles from US 75. Figure 18 Bronze objects from US 75: a) Bronze Snake b) Bronze Axe / amulet.

27 Figure 19 Figure 20 Characteristic finds from Salut: a) Long-handled bowl. b) Bronze snake found in Characteristic finds from Salut: a)applied snake decoration on a large storage vessel. b) Carinated bowl found in Attachment 1 Report on the Geoarchaeological survey (M Cremaschi) List of figures Figure 1a Figure 1b Figure 2 Irrigation ditch Mounds located west of Salut Pottery from mounds west of Salut (Site 68): Blue-glazed sherds of Early Islamic date and small bowl of Iron Age date.

28 AFLĀJ, CANALS AND IRRIGATION DITCHES Preliminary report of the geoarchaeological survey in the surrounding of Salut February Mauro Cremaschi, Dipartimento di Scienze della Terra dell Università di Milano 1 - Foreword The 2007 geoarchaeological survey at Salut was focused on the irrigation systems, dominated by the falaj system which constitutes an important part of the archaeological landscape. The purpose was to collect data suitable for dating the system and reconstructing its significance in the relationship between human cultures and environment during the late Holocene. Methods were geomorphological and geoarchaeological field survey supplemented by remote sensing observation and GPS positioning of archaeological sites and irrigation structures. Samples of sediments, soils and travertine were collected for laboratory analyses and radiocarbon and U/Th dating 2 - Geological background The area investigated extends from Bisyah to Nattalah to the South and the North respectively, by the Wadi Sayfan in the West and by Wadi Bahla in the East. From the geological point of view the area is included in the sedimentary apron at the southern margin of Gebel Al Hajar al Gharbi, mainly consisting of gravely alluvial fans indicated on the geological map as Qf and dated to the late Pleistocene. On the satellite images, on the eastern bank of the Wadi Sayfan, there is an area of fine light coloured silty and fine sand deposits, which are indicated as Kabrha formation on the geological map and dated to the Holocene. Most of the archaeological evidence and the irrigation features recovered so far are located in this formation. 3 - Before irrigation The Kabrha formation appears to be an essential component of the agricultural exploitation of the area and of the concentration of irrigation features in it. This is due to its pedological characteristics: silty texture and light structure, good water retention and easy to be plough. These conditions, provided that water availability exists, makes the Kabrha formation the most suitable soil for agriculture in a region where soils are in general stony and shallow and consequently mostly unproductive. The distribution of this formation is easily identified on the satellite imagery due to its whitish colour, which is in contrast with the dark colour of the surrounding alluvial fans. It is delimited to the West by the present course of Wadi Sayfan, lying upon its terrace, and to the East by the distal fringes of the alluvial fans descending from the north. In-situ outcrops of this formation were discovered south of Bisyah in the alluvial valley of the Wadi Sayfan, where they are about one metre thick and are covered by a gypsum crust. At its base, some clusters of Neolithic flint artefacts have been found at some locations. Sedimentological characteristics indicate that the Kabrha formation consists of wind blown particles and the lithic industry found mainly at its base, indicates for it an early Holocene age. Its formation is to be related to a wetter climatic phase during which the area was covered by bush grassland which trapped the dust brought by the northern winds. A similar phenomenon has been observed elsewhere in Oman and in particular along the southern margins of the Nejd desert. Several sites, particularly from the Bronze Age, have been recovered in the investigated area, mainly at the margin of the Kabrha formation, on outcrops of bedrock and on small gravel bars

29 inside the Salut plain. In several cases the Bronze Age monuments appear to have been in part covered by alluvial sediments indicating that the Kabrha deposits have undergone some colluvial redistribution after the middle Holocene. Environmental information about the middle Holocene and particularly about the Bronze Age settlement pattern, is badly needed to understand the beginning of irrigated agriculture in the area. 4 - The irrigation systems The area occupied by the Kabrha formation (here indicated as the Salut plain) appears to have been used for intensive agriculture in the past since it is crossed by a dense network of now inactive field ditches. These are clearly seen on satellite images and are recorded on the topographic map as enclosures. The cultivated fields appear to have been abandoned for long ago, and today the area is dry and almost desert, with palm groves and cultivated fields being restricted to limited places where water is pumped from deep wells. However, an impressive amount of irrigation systems which were active in the past: (about 49 kilometres of aflāj and canals) have been observed and mapped during the present survey. According to local people the falaj ceased definitively to carry any water supply during 1970s. Some wells were excavated during the nineties inside the falaj shafts, with the purpose to reach the water table, but also they are now dry. The northern part of the Salut plain is reached by a main canal 8 km long, which comes from the area of Jabrin fort deriving water from the upper reach of the alluvial fan of Wadi Bahla. From the satellite photograph, it can be observed that minor ditches, in the shape of comb pins diverge from the main canal to bring water the surrounding areas which have now remained uncultivated for a long time. Connected to the irrigation network there are the ruins of an Islamic village and a stone building associated with Islamic pottery. This canal has been built in several phases: the aqueduct existent on it, made of stone masonry cemented by plaster, represents only the last and more recent one. The canal includes thick travertine deposit which has been sampled for dating and isotopic analysis. The following individual aflāj have been located all along the Salut plain: Falaj Salut ( 700 metres) Falaj Farud ( 5 km ) Falaj Bisyah 1 and 2 ( 5 km, 4 km respectively ) Falaj Azabi ( 8 kilometres not surveyed) Falaj al Hayl 1 and 2 (in total 3 kilometres) It is very possible that they are incomplete and that in some cases their beginning and their ends have been obliterated by colluviation, collapse and subsequent erosion of the spoil heap. Also, most of the shafts are totally or in part collapsed and their depth cannot be assessed exactly. Also the underground tunnel is often inaccessible and only rarely preserves its original shape. Two main types of aflāj can be distinguished on the basis of their state of preservation: A) very degraded falaj whose shaft is in general filled up and spoil heaps are flat and smooth B) better preserved falaj whose shaft is in general opened and reach a depth from 10 to 2 metres, In this case the spoil heap is up to two metres high and very distinct. The sediments in which the shafts of the falaj were excavated are clearly exposed and their stratigraphy is rather homogeneous and consists of the following units : 1- top unit, about one metre thick reworked Kabrha silts, which includes mostly pottery of the Islamic period and some sherds of the Iron Age. This unit is most developed along the Bisyah Falaj 1.

30 2 - gravel unit; about five metres thick of cemented gravel dominated by limestone and ophiolite lithotypes. At the top of this unit some calcite concretions are found, which indicate the persistence of the water table. One sample collected during the season 2006 has been dated by U/ Th to calendar years BP. This dating is quite meaningful for the history of water availability in the region and will be discussed below in more detail. 3 - caliche bedrock, planar laminated limestone or deeply cemented fine sediment and gravel, up to ten metres in thickness. Most of the falaj shafts cross the first three units and the underground tunnel (in most cases not accessible) is excavated in the hard cemented bedrock. In the whole area a strict functional connection between aflāj and canals has been observed and they appear to be two aspects of the same irrigation system. The largest part of the aflāj which have been surveyed end in open channels which have the purpose to bring the water to the ground surface and distribute it at field level. In several cases (Falaj Farud and Falaj Shaww), however, a complex relationship between falaj and channel has been observed. In these cases the aflāj are shallow but large canals (up to forty metres), delimitated by straight levees from 1 to 1.8 metres above the ground surface. The shafts of the middle part of the Falaj Farud have been excavated inside the western levee of the adjoining canal and therefore the falaj has to be considered as more recent than the canal itself. In some other cases, the falaj has been excavated inside the canal, confirming the fact that the opening of the falaj follows the canal construction and also that between these two phases there is a clear substantial change in the water table. Only in one case the section of a canal has been directly observed due to a modern bulldozer cut. The section is located close to the beginning of the irrigation canal originating from the Bisyah falaj: it is cut in the bedrock up to the depth of fifteen metres and filled by water laid sediments. Included in the lower part are exclusively Iron Age pottery sherds. Archaeological survey around aflāj, done in order to judge their relations with archaeological settlements, indicate that clusters of Islamic materials are often in relation with them (the intermediate part of the Falaj Farud, the whole Falaj Bisyah 1) but in other cases, in general in the case of the more eroded falaj, the archaeological material is completely absent. Therefore the association with the Islamic material does not indicate the period of construction but probably a later period of re-use or restoration. 4 - The ruins surrounding the site of Salut as a Central Place of the irrigated area. West of the hill of Salut, on which the archaeological excavations of the Italian mission currently focussed, are more than twenty small mounds, which occupy wide area between some Islamic ruins and the Salut site. The mounds have not been interpreted as mounds of fertile soil, resulting from field construction, due to the fact that locally the soil is still in place and has not been excavated in so large quantity. Furthermore the survey of the mounds revealed that they actually are in most cases remains of collapsed buildings and evidence of architectural structures (mud brick and stone walls, plaster concentrations etc.) outcrop from their slopes which are littered with sherd fragments The pottery assemblages include glazed sherds of XIV XV century type and sherds of the early Islamic era (glazed Abbasid pottery). Iron age pottery is also quite abundant (an almost complete cup was found together with other fragments at site S7/68).

31 Furthermore, all the area in between the mounds and the base of the Salut hill is completely covered with a squared pattern of relict agrarian ditches, aqueducts in masonry, with travertine deposits and stone walls to retain the fertile soil and prevent its erosion. All these features are clearly seen in the field and are evident from the remote sensing. They have to be referred to gardens and irrigated cultivation. While further research is needed, it is quite probable that the mounds described are the remains of an extensive settlement, related to a former oasis: it is relevant to note that the concentration of mounds occupies a central position in the Salut plain and occupies focal point in the irrigation system. 5 - A lost oasis: age of the irrigation system and its relation to climatic change. On the basis of the present research, during its apogee the oasis of Salut may have reached the size of about 75 square kilometres of cultivated land, sustained by water provided by irrigation facilities, fed mainly by canals and aflāj. The present surface occupied by palms and cultivated fields is reduced to about 6 square kilometres, suggesting a scenario of dramatic aridification and consequent collapse related to inactivation of the irrigation system. However the reconstruction of the history of the land use history in the plain of Salut is difficult to reconstruct at present because of the scarcity of data. Most of the irrigation features appear to have been in use during the Islamic era up to the XIV XV century, but it is also clear that the Islamic era re-used and probably expanded an irrigation system already existent and well organized, which, on the basis of fragile archaeological indications, may be referred to the Iron Age. The large fossils canals of the central part of the Salut plain and probably some of the most degraded aflāj, both to the northern and the southern fringes of the area, might be representative of the Iron Age. The date of 2077 y. BP for the calcitic concretions indicates a high water table in the central part of the Salut plain at the end of the first millennium. For this reason the canals were probably fed by surface run off and shallow water, at that time. The late excavation of aflāj inside the former canals may, in contrast, indicate the lowering of the water table and an adaptation of irrigation strategies to increasing aridity. It is still unknown when the irrigated agriculture began in the area. The present knowledge of archaeological sites suggests that attention should be concentrated on the Bronze age, which is still close to the wet early Holocene and may have enjoyed greater availability of water, rendering effective and profitable the onset of intensive agriculture to complement cultural dynamics, more demanding of resources, the exploitation of soil including therefore agricultural potential. 6 - Project for future research From the present research it is clear that the plain of Salut is a case study of the highest interest for reconstructing the prehistoric background and historical development of irrigated agriculture in Omani region. Furthermore, the local problem appears to be strictly linked to a more general scenario of climatic change and cultural reaction and adaptation to increasing aridity. For these reasons the research deserves to be developed. This should include - Aerial photographs from the 1970s or earlier, if available, should be studied for information prior to modern works (bulldozering, quarrying, modern cultivation, road construction ) since they may preserve evidence of irrigation systems which at present may have been erased.

32 - Geoarchaeological survey and mapping, while more than one archaeological team is at work, and the area is quite rich of archaeological evidence, a systematic map of the archaeological evidence does not exist at present. However it is a necessary requirement to any development of the research. - A selected number of falaj shafts should be investigated through drilling or excavation in order to collect direct evidence for dating. The same applies to the underground tunnels of the falaj which may contain deposits (mainly travertine) suitable for dating and further palaeoclimatic investigation such as oxygen and carbon stable isotopes. - There are few possibilities that a foreign team may understand by itself the nature of the falaj system, whose tradition in deeply rooted in the local people. A campaign of interviews with people which have still memory of the construction and use practice of the falaj may be of the outmost importance for the location of sites where concentrate further research. - It is clear from the present knowledge that the oasis of Salut has been in the past large, rich of water and crops, with important irrigation structures and may have sustained a large population, as this situation extended to historical times, a memory of it should be preserved in written sources which should be checked in search of historical records.

33 Fig. 1a. Irrigation ditch Fig.1b. Mounds located west of Salut

34 Fig. 2. Pottery from mounds west of Salut (Site 68): Blue-glazed sherds of Early Islamic date and small bowl of Iron Age date.