Dabrowski, Bouchaud, Tengberg, Zazzo & Priestman, 2021: Archaeobotanical Analysis of Food and Fuel Procurement from Fulayj Fort (Oman, 5th-8th c. CE) Including the Earliest Secure Evidence for Sorghum in Eastern Arabia

Journal of Arid Environments 190 (2021) 104512 Contents lists available at ScienceDirect Journal of Arid Environments journal homepage: www.elsevier.com/locate/jaridenv Archaeobotanical analysis of food and fuel procurement from Fulayj fort (Oman, 5th-8th c. CE) including the earliest secure evidence for sorghum in Eastern Arabia Vladimir Dabrowski a, *, Charlène Bouchaud a, Margareta Tengberg a, Antoine Zazzo a, Seth Priestman b a Archéozoologie, Archéobotanique: Sociétés, pratiques et environnements (UMR 7209), Muséum national d’Histoire naturelle/CNRS, 43 rue Buffon – CP56, 75005, Paris, France b Department of Archaeology, Durham University, South Road, Durham, DH1 3LE, UK A R T I C L E I N F O A B S T R A C T Keywords: Archaeobotany Eastern Arabia Sasanian and Early Islamic periods Sorghum Oasis agriculture Fuel management The recent study of botanical macro-remains from the Late Sasanian and Early Islamic (5th to 8th century) fort of Fulayj (Batinah, Sultanate of Oman) provides a unique opportunity to discuss food and fuel acquisition strategies in an arid environment and to document periods that are little known from an archaeobotanical perspective in Eastern Arabia. Seed assemblages include the first well-identified and directly radiocarbon dated evidence of sorghum (Sorghum bicolor ssp. bicolor (L.) Moench.) in Eastern Arabia, which raises the question of whether the grain was imported from distant sources (for example Yemen, East Africa or India) or locally cultivated. In addition to sorghum, the food plant assemblage includes hulled barley (Hordeum vulgare), date (Phoenix dactylifera) and jujube (Ziziphus cf. spina-christi). Date palm gardens may have existed near to the site as they do today or food products may have been brought from date palm gardens on the Batinah coast where conditions for agricultural production are particularly favourable. Charcoal analysis reveals that the main taxa used for fuel (acacia, prosopis, jujube tree, tamarisk) were collected from local plant communities, occasionally supplemented with firewood gathered in the foothills and mountainous areas. 1. Introduction Direct archaeological evidence concerning agricultural practices and fuel collection strategies during the Sasanian and Islamic periods in Eastern Arabia remain very rare. The archaeobotanical study conducted at the Sasanian and Islamic period site of Kush (Ra’s al-Khaimah, UAE) represents the first study of its kind within the region but has been only partially published (Kennet, 1997, 2009; Tengberg, 2005). Publication of the whole corpus remains in progress (Dabrowski 2019: 159–200; Dabrowski et al. forthcoming; Tengberg et al. forthcoming). Recent archaeological investigations conducted at the Late Sasanian and Early Islamic fort of Fulayj on the Batinah plain (Sultanate of Oman), occupied from the c. early 5th to 8th century, provides a significant opportunity to increase our knowledge of the plant economy by studying macro-botanical remains (seeds, fruits, charcoals) from the only known site of this date in Oman (Priestman 2019; al-Jahwari et al., 2018). This study sheds light on food and fuel acquisition strategies in an arid environment during the Late Sasanian and Early Islamic periods. Moreover, it offers the earliest secure identification of sorghum in Eastern Arabia and allows us to raise the question of whether it was an imported commodity or incorporated into local oasis agriculture during this period. Of particular importance is the opportunity to securely document the process of diffusion by obtaining evidence for where the crop is attested for the first time. To properly assess its age, an additional date has been obtained directly from the available sorghum remains. The integration of sorghum into the general plant supply system is an issue of particular interest. Sorghum (Sorghum bicolor ssp. bicolor) is a drought-resistant and warmth-loving crop, frequently cultivated today in agrarian systems of the tropical and sub-tropical regions of the Old World (Africa, South Asia). It is used for human consumption (eaten as grilled and boiled grains, transformed into a flour for porridge, or fermented for alcoholic * Corresponding author. E-mail addresses: vladimir.dabrowski@mnhn.fr (V. Dabrowski), charlene.bouchaud@mnhn.fr (C. Bouchaud), margareta.tengberg-mongne@mnhn.fr (M. Tengberg), antoine.zazzo@mnhn.fr (A. Zazzo), seth.priestman@gmail.com (S. Priestman). https://doi.org/10.1016/j.jaridenv.2021.104512 Received 1 December 2020; Received in revised form 15 April 2021; Accepted 16 April 2021 Available online 2 May 2021 0140-1963/© 2021 Elsevier Ltd. All rights reserved. V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 beverages). It is also frequently used as fodder for livestock (Chantereau et al., 2013; Stenhouse and Tippayaruk 1996). Sorghum is a crop of African origin with its wild complex ancestor, S. bicolor ssp. verticilliflorum Steud. (Stapf.). The earliest evidence of cultivated sorghum undergoing domestication comes from contexts dated to the mid-to late 4th millennium BCE in eastern Sudan as indicated by sherd impressions of grains, spikelets and chaff (Barron et al., 2020; Winchell et al., 2017). Archaeobotanical investigations reveal the continuity of sorghum cultivation during the following millennia in Eastern Africa (Beldados 2019; Fuller and Stevens 2018; Fuller 2014; Walshaw 2010; Clapham and Rowley-Conwy 2007; Van der Veen and Lawrence 1991). The crop was introduced into the Indian sub-continent during the 2nd millennium BCE (Winchell et al., 2018; Fuller and Boivin 2009; Fuller 2003), where it was considered a marginal crop. Its importance in South Asia seems to have increased from the turn of the Christian era and in particular from the 8th century CE although archaeobotanical data is still generally lacking (Boivin et al., 2014). Genetic and archaeobotanical evidence allows us to document the evolution of this crop and its varietal differentiation during its diffusion in Africa and Asia. According to these researches, five main complexes of landraces have been defined. The original “bicolor” is characterised by hulled grains while the four others have evolved in the form of free-threshing and larger-grained specimens. “Caudatum” appeared in the Sahelian region, “durra” is probably originated from the Indian sub-continent and “guinea” from western Africa. The latter provides the origin of race “kafir” in southern Africa by the intermediate development of one more forest adapted “mageritiferum“ type (Smith et al., 2019; Fuller and Stevens 2018). The precise date of the introduction of sorghum into Eastern Arabia is not well known (see below). The earliest reported date of sorghum in the region is from the 3rd millennium BCE occupation of Hili 8 (Abu Dhabi, UAE) in the form of carbonized grains and mudbrick impressions (Cleuziou 1982; Cleuziou and Costantini 1982). However, several researchers have questioned this early discovery due to morphological considerations of the published material consisting mainly of photographs of plant impressions on mudbrick (Bouchaud et al., 2016; Tengberg 2012; Charbonnier 2008; de Moulins et al., 2003). In addition, caryopses (grains) and starch grains have been potentially identified at the site of Khor Rori/Sumhuram (Dhofar, Sultanate of Oman), occupied from the 2nd century BCE to the 5th century CE, but their state of preservation does not allow a secure identification (Bellini et al., 2020). The single other well-documented occurrence comes from the Islamic period coastal settlement of Qalhât in Oman (14th to 16th century) (Dabrowski et al., 2015). 2. Environmental and historical background 2.1. The Batinah plain and the al-Hajar mountains The site of Fulayj (24◦ 08′ 51′′ N – 56◦ 79′ 62′′ E) is situated in the central part of the Batinah plain, 13.5 km from the coastline, roughly halfway between the eastern foothills of the northern part of the al-Hajar Mountains, and the Gulf of Oman (Fig. 1). The plain is dominated by coarse gravels and rocks with limestone outcrops and low alluvial terraces formed during alternating wet and dry climatic episodes during the Pleistocene and Holocene (Sanlaville 2000: 141, Abrams and Chadwick 1994). The site is situated on the terraced edge of an interfluve overlooking the wide braided course of Wadi al-Mahmum extending towards the north and east. Today, the climate is arid with an average annual rainfall of around 100 mm and precipitation unequally distributed throughout the year (Kwarteng et al., 2009). Palaeoclimatological analyses indicate that there may not have been any substantial change during the last four millennia, except for some minor oscillations that still require further investigation (Parker and Goudie 2008), for example, a dry spell during the second half of the 1st millennium CE (Fleitmann et al., 2009). Wadis flowing from the mountains bring fine clayey loam onto the Batinah plain where they are deposited in a 2–5 km wide band near the shoreline. This phenomenon, known as khabra, as well as the presence of a higher water table, account for the high agrarian potential of the region (Sanlaville 2000: 140–141). A large continuous band of date palm gardens irrigated by runoff water and from wells stretch along the coastline of the Batinah. Date palms plunge their roots directly into the brackish water of the groundwater table. Other crops are cultivated in gardens 2–3 km behind the line of the palm trees protecting them from salty winds from the sea (Wilkinson 1977: 48–49). Today the Batinah is one of the most productive agrarian regions of the Sultanate of Oman with more than half of the irrigated areas of the country concentrated in this region (Abdelrahman et al., 1993). Some agriculture is also possible in the hinterland, for example near Fulayj where date palm gardens are maintained via water brought by underground galleries or aflāj running Fig. 1. Location of Fulayj fort and other contemporaneous sites in Southeast Arabia. Blue = sites occupied during the Late Sasanian period (5th-6th century), yellow = those which belong to the Early Islamic period (7th-8th century). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 2 Journal of Arid Environments 190 (2021) 104512 V. Dabrowski et al. 1 from the foot of the mountains. The vegetation cover of the Batinah plain has been heavily impacted by human activities. In particular, over-grazing by domestic herds prevents the regeneration of many fragile plant species (Ghazanfar 1998a). The velvet mesquite (Prosopis juliflora), an invasive species from Central and South America, has also been reported from the plain where it threatens the indigenous biodiversity (al-Abdali 2019 et al., Ghazanfar 1996). The flora belongs predominantly to the Nubo-Sindian phytogeographical region, characterised by dry sub-tropical plant communities (Ghazanfar 1992). Open thorn woodlands (Fig. 2) are composed of Acacia ehrenbergiana, A. tortilis, Prosopis cineraria and Ziziphus spina-christi, associated with shrubs such as Lycium shawii and Ochradenus arabicus. Among the annual plant species can be mentioned Zygophyllum simplex, Plantago ovata, Aizoon canariense and Cometes surratenis (Ghazanfar 1998b). Some species growing mainly in the foothills and in the lower parts of the mountains such as Rhazya stricta, Fagonia indica and Maerua crassifolia also occur on the plain (Ghazanfar 2003: 2). The vegetation of the northern al-Hajar mountains, situated at a distance of about 10–15 km from the site, is composed of different types of vegetation communities structured according to altitude and hydrological conditions. Between 650 and 1000 m, Acacia-Rhazya-Fagonia formations are, like in the Batinah Plain, heavily impacted by overgrazing leaving non-palatable, toxic species such as Rhazya stricta and Fagonia indica as dominant elements. Some characteristic species of seasonally flooded wadis such as Ficus cordata ssp. salicifolia and Tephrosia apollinea are also recorded. The vegetation cover between 1000 and 1500 m is dominated by Euphorbia larica shrub associated with acacias (A. tortilis, A. gerardii) and Periploca aphylla (Deil al-Gifri 1998). Higher slopes (c. 1100–2500 m) are dominated by formations with Olea europaea, Sideroxylon mascatense and Dodonaea viscosa associated with xeric shrub species such as Grewia erythraea, Barleria candida and succulents such as Caralluma aucheriana. Finally, the highest zone, situated between 2100 and 3000 m, is characterised by Juniperus-Ephedra-Teucrium vegetation-type (Ghazanfar 2003: 2, Ghazanfar 1991). Several of the mountainous formations cited above are also heavily impacted by over-grazing practices which deeply degrade the vegetation cover (Brinkmann et al., 2009). has been much debate among historians and archaeologists. Although Sasanian authorities from Persia appear to have shown interest in Eastern Arabia since the military conquest of Ardashir I, around 240 CE (Potts 1990: 232–234), the nature of this involvement remains difficult to determine precisely. During the Late Sasanian period, from the reign of Khusraw I (531–579 CE), Sasanian influence in Oman is more firmly attested (Ṭabarı̄, 1999: 237; 253, Potts 1990: 249; 335) but the nature and degree of the occupation are still not clearly defined. Some historians propose that the Sasanian occupation of the Oman Peninsula was extensive on the Batinah plain and that the establishment of numerous aflāj and the settlement of Persian soldiers as landowners indicate a substantial agrarian development. Military fortresses and quarters for garrisons are also thought to have been built in the region (respectively Rustaq, Sohar) (Wilkinson 2010: 57–60, Wilkinson 1979: 888–889, Wilkinson 1977: 130–133). Economic interest has also been advanced to explain the high degree of involvement of the Sasanians in Eastern Arabia either to secure maritime exchange routes within the Persian Gulf and the wider Indian Ocean (Yemen, India) (Daryaee 2003) or the supply of copper in response to the Roman interdiction to export it to the Sasanian Empire (Morony, 2001–2002). However, other historians have adopted a more critical approach to the Persian and Arabic chronicles written several centuries after the events upon which they report. They suggest that this economic development during the Late Sasanian period in Eastern Arabia may not have been as important as originally thought (Munt 2017; Ulrich 2011). Archaeological research tends to supports this hypothesis. A reassessment of Sasanian archaeological data in Eastern Arabia (Kennet, 2007) suggests a decline of settlement and economic indices compare to the previous period, although more data are needed in order to further validate this point. In general, archaeological discoveries securely attributed to the Sasanian period are scarce: they correspond to a few sites that are mostly limited in extent, such as Kush (Kennet, 1997, 2004, 2009), Khatt (Kennet, 1998) and Jazirat al-Ghanam (de Cardi et al., 1975; de Cardi 1972) as well as isolated graves and small finds (coins, seals, figurines, etc.) (Kennet, 2007; Simpson, 2019). The adoption of Islam in Eastern Arabia was not only an event of spiritual importance but also one with far reaching political and economic consequences. After a first agreement between the Prophet Muhammad and the main tribes of Arabian periphery in 630 CE allowing them to keep some privileges in exchange for their conversion to Islam and submission to the new state of Mecca, some tribes decided 2.2. The Sasanian and Early Islamic periods The extent of Sasanian influence in Eastern Arabia is a subject that Fig. 2. Today’s vegetation cover of the Batinah plain. We can see here an open thorny woodland composed mainly of acacia (Acacia tortilis (Forssk.) Hayne) 3 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 to rebel at the death of the Prophet in 632 CE during the episode known as the riddah (apostasy). But the defectors were quickly defeated by the remaining loyalist groups with the help of military troops from Mecca during a battle possibly near Dabâ (Dibba). According to textual references, Persian populations which were settled in Oman during this period refused to convert to Islam and were therefore targeted by Arab converts and defeated before leaving the area (Moez 2007). Shortly after, the region of Oman declared its independence from the central Caliphate between 656 and 692 CE. After several reconquest attempts, the independent state was eventually defeated by the Umayyads between 694 and 705 CE before a period of relative peace and prosperity ensued lasting until 750 CE, when the Abbasids overthrew the dynasty of the Umayyads (al-Rawas 2000: 61–67). After another attempt at establishing an independent Omani state by the Ibadis, the new dynasty took control over this region in 752 CE until the end of the 8th century (al-Rawas 2000: 111–133). Very few archaeological sites are securely attested dating to the initial period of conversion to Islam in Eastern Arabia. In addition to Kush, where the occupation continues until the 13th century, the extensive but short-lived settlement of Jazirat al-Hulaylah (Ra’s alKhaimah, UAE) and the harbour of Sohar (Sultanate of Oman) in the Gulf of Oman show clear evidence of occupation during the 7th and 8th centuries (Kennet, 2007, 2009, 2012). Some Christian settlements (churches, monasteries) such as Sir Bani Yas (Abu Dhabi, UAE) are also attested from this period, though their distribution appears to be restricted to the area of the Gulf (Simpson 2019; Carter 2008). The recent discovery of the Late Sasanian fort of Fulayj on the Batinah plain (Sultanate of Oman), occupied between the early 5th to 8th century, provides unique information about the nature of the Sasanian occupation. Moreover, with the exception of the tower from the site of Kush, this is the only site in Eastern Arabia to document the transition between the end of Antiquity and the beginning of the Islamic period (al-Jahwari et al., 2018). The sequence from Fulayj provides information on a period which is still poorly understood archaeologically, including archaeobotanical data framed by a robust absolute chronology. Fig. 3. Aerian view of the Fulayj fort with trenches opened during the two field campaigns in 2015 and 2016 (© Fulayj Project). domestic occupation. After the general abandonment of the fort, some irregular walls and later pottery indicate limited activity within the area dated to the Late Islamic Period (16th to 20th century, Phase 4) (al-Jahwari et al., 2018). Phases 1 and 4 have been mainly excluded from the archaeobotanical study because they are not directly linked to the occupation of the fort. The size of the building, the thickness and regularity of the walls and the presence of towers highlight its military and defensive functions. Thus, it constitutes one of a very limited number of attested military constructions dated to the Late Sasanian period in Eastern Arabia, and the only one facing the Indian Ocean within the main portion of Oman. Other potentially military structures include the tower from the site of Kush whose structure and function are not precisely determined and the ‘watch station’ from Jazirat al-Ghanam in Northern Oman, which although seemingly contemporary to Fulayj, lacks substantial defensive capabilities. The scarcity of domestic remains like pottery sherds (only 346 for phases 2 and 3) and animal bones within the building at Fulayj seems to further support the interpretation of a military function. Those pottery finds that do occur mostly originate from southern Mesopotamia and Iran, then under Sasanian domination, suggesting that foreign soldiers may have been provisioned directly from outside by the Sasanian state. Indian pottery has also been found and appears to indicate a degree of integration into long-distance exchange networks within the western Indian Ocean (al-Jahwari et al., 2018). This fort may have been part of a wider defensive network (Kennet et al., 2016) devoted to securing the fertile sectors and supply routes of the Batinah plain while maintaining strategic control of the coastal territory. These archaeological discoveries may be in position to comfort textual data about defensive strategy of the Sasanian Empire conducted by Khusraw I during the 6th century CE (al-Jahwari et al., 2018). 2.3. The archaeological site of Fulayj The archaeological site of Fulayj (FJ3. S3) corresponds to a complex dated mainly to the Iron Age (1300-300 BCE) among which a Late Sasanian fort has also been identified in the north-western part. The fort of Fulayj (Figs. 3 and 4) is 30 m square with neatly constructed stone walls measuring on average 2.66 m thick with projecting ‘U’ shaped corner towers, a narrow 1.62 m wide entrance in the east, and entrance flanking towers. It is likely that the stone base supporting a mudbrick superstructure coated with lime mortar. The construction and the planning of the fort suggest a high degree of experience, proficiency and planning consistent with the interpretation of the site being built by an external military force. Several trenches were opened during the excavation seasons in 2015 and 2016: some within the interior (Trenches A, F, G) and some outside (Trenches B, N) the fort as well as on both sides of the main gateway and entrance flanking towers (Trench E). To the south of the fort, several lime kilns probably used during the construction of the fort have been detected; a single test pit was excavated in this area (Trench C). Pottery sherds, small finds and a set of 25 radiocarbon dates obtained from selected carbonized macro-botanical remains have permitted the dating of several phases of occupation. Some Iron Age levels (1000-500 BCE, Phase 1) have been excavated below the level of the fort. The construction and first occupation of the fort are dated to the Late Sasanian period (between the early 5th to mid-6th century, Phase 2). Another occupation, either continuous or after an abandonment, has been assigned to the Early Islamic period (late 6th to 8th century, Phase 3). This phase is associated with the secondary insertion of mudbrick architecture inside the fort in Trench F and the later construction of an oven, possibly indicating a shift in site function from a military to a 4 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Fig. 4. Details of the Fulayj fort. A: The south-west projecting corner tower of the fort; B: The fort entranceway looking north-west with entrance-flanking towers; C: Mudbrick wall (F.021) of Phase 3 in Trench F; D: Fill of the oven consisting of a dense accumulation of animal bone within the upper layer (F.026) and charcoal concentrated towards the base (F.047) (© Fulayj Project). 3. Material and methods total number of remains by the volume expressed in litres. Ubiquity represents the number of samples in which taxa are attested. Because they were not botanical remains stricto sensu, coprolites were not included in the calculations. However, they were also considered and will be described where relevant. The botanical identification of charred wood fragments is based on the observation of the cellular structure (or anatomy) of the wood under a reflected-light microscope (Olympus), with magnifications from x50 to x800, and according to three wood-anatomical sections (transversal, longitudinal radial and longitudinal tangential) obtained by manually fracturing the charcoal fragment. As for the seed and fruit remains, we used a reference collection of modern specimens as well as anatomical atlases of wood (Neumann et al., 2001; Pajouh and Schweingruber, 1993; Schweingruber, 1990; Fahn and Werker, 1986). For the wood charcoal analysis and interpretation, a minimum number of fragments from each archaeological context is required (Chabal 1997). At Fulayj most of the samples did not contain enough charcoal fragments to be considered statistically representative and thus we chose to focus on five samples where the number of charred wood fragments was greater than 50, a number that can be considered as sufficient in an arid environment where the diversity of tree and shrub species is relatively low. For each sample, after the discovery of a new taxon, 50 more pieces of charcoal were identified. In order to compare the taxonomical diversity between these richer samples with samples containing fewer charcoal fragments, we included two smaller samples (A.016 and B.007) in our study. In total, seven samples were processed for the charcoal study: one coming from an Iron Age hearth (G.014), one from a Late Sasanian deposit (B.007) and five dated to the Early Islamic period. Among the latter, one came from an occupation level (F.030), one from a rubbish deposit (A.016), one from a hearth (A.014) and two from an oven (F.026 and F.047). Four selected contexts for charcoal analysis correspond to hearths and ovens; structures with high charcoal concentrations, which are not considered suitable for quantitative analysis of past vegetation reconstitution (Chabal et al., 1999: 62–63). Therefore, the ecological interpretation will be mainly based on qualitative considerations. 3.1. Macro-botanical remains Botanical macro-remains were collected from 41 contexts (occupation levels, dump deposits, pits, foundation trenches, hearths, ovens, mudbrick collapses, water- and wind-blown deposits) excavated in Trenches A, B, E, F, G and N (Table 1). Most samples come from Trench F (N = 13) while other trenches have provided from 3 (Trench N) to 8 sediment samples (Trench A). The Early Islamic period is most well represented (N = 15), followed by the Late Islamic period (N = 12), the Iron age (N = 7) and the Late Sasanian period (N = 7). We aimed at collecting at least 10 L of sediment whenever possible but according to the context. Sample volumes range from less than 1 L to 25 L. In total, 592.5 L of sediment have been processed at the archaeological base in Sohar during fieldwork in March 2016. Manual bucket flotation in which plant remains were recovered on a 0.3 mm mesh sieve was used for all of the samples except for the concentration of large-size charcoal pieces from an oven (F.047). This assemblage was sub-sampled before being dry-sieved with a 2 mm mesh and the fine fraction (less than 2 mm) floated. All processed samples were analysed at the National Museum of Natural History in Paris (France). Macro-botanical remains were extracted from the flotation residues with the help of a binocular stereomicroscope (Nikon SMZ645). The botanical identification of seeds and fruits was based on morphological and anatomical criteria and involved the comparison of the archaeological material with modern reference collections and illustrations in seed atlases (Cappers et al. 2006, 2009, 2012; Jacomet 2006). All samples contained seeds and fruit remains except one (B.007). The results are expressed as Number of Remains (NR) and Minimum Number of Individuals (MNI). The MNI was determined by counting the whole remains and adding the estimation of whole individuals made from the fragments available by reconstruction to the naked-eye. Percentages were calculated with the MNI including determined and undetermined remains. The density of remains per litre was defined by dividing the 5 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Table 1 List of archaeobotanical samples from Fulayj. Season Trench Context Phase Nature and Description Volume (litres) 2015 2015 2015 2015 2015 2015 2015 2015 2015 2015 2015 2015 2015 2015 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 A A A A A A A A B B B B B B E E E E E F F F F F F F F F F F F F G G G G G G N N N A.005 A.006 A.009 A.013 A.014 A.016 A.018 A.019 B.003 B.004 B.006 B.007 B.008 B.009 E.020 E.022 E.024 E.027 E.028 F.009 F.010 F.011 F.015 F.017 F.019 F.026 F.027 F.030 F.035 F.036 F.045 F.047 G.006 G.008 G.010 G.014 G.015 G.017 N.002 N.005 N.008 4 4 4 2 3 3 2 2 4 4 4 2 1 2 3 3 3 2 2 4 4 4 2 3 2 3 3 2 3 2 2 3 4 3 1 1 2 1 1 1 1 Wind-blown deposit Wind-blown deposit Wind-blown deposit Occupation layer Hearth Soft occupation deposit Occupation layer Surface layer Wind-blown deposit Water-flooded deposit Wind-blown deposit Occupation deposit Occupation layer Foundation trench fill Occupation deposit Occupation deposit Occupation deposit Surface layer Surface layer Edge of a fire pit Edge of a fire pit Windblown deposit with mudbrick decay Mudbrick decay with burning Mudbrick decay Occupation layer Upper fill of the oven consisting mostly of animal bone Windblown deposit above the oven Occupation layer Occupation layer Fill of post-hole Occupation layer Lower fill of the oven consisting of charcoal and ash Occupation deposit Occupation deposit Occupation deposit Hearth deposit Hearth deposit? Occupation deposit Occupation deposit Occupation deposit Occupation deposit 11 18 16 15 20 18 19 15 19 23 22 22 20 25 11 10 11 9.5 8 <1 9 10 20 8 19 20 9 3 20 10 18 11 10.5 10 20 10 3.5 20 10 22.5 20.5 Seed/fruit analysis + + + + + + + + + + + Charcoal analysis + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + 3.2. Radiocarbon dating IntCal20 atmospheric curve (Bronk Ramsey 2009; Reimer et al., 2020). One sample of sorghum (one fragmentary caryopsis of sorghum (Sorghum bicolor ssp. bicolor)) and one of tamarisk twig (Tamarix sp.) from the same context (E.024) were sampled for AMS dating. The latter sample was obtained together with 24 other high-precision AMS dates provided from selected carbonized botanical remains from across the excavations used to define the site chronology. Those from Trenches A, B and C were selected by I. Van Bergen Poole (2015) while the remainder from Trenches E, F, G and N were identified by the first author of the present paper. Another date made directly on sorghum was obtained in order to establish its age in relation to the context of its recovery. Indeed, later occupation dated to the Late Islamic period (16th-20th century) has been observed within the uppermost deposits (Phase 4), and the hypothesis of contamination, although unlikely, could not be completely ruled out. In addition, one sample of tamarisk wood has been dated from another context where sorghum is attested (F.036) as one of the other dates used to establish the chronology of the site. In this case, sorghum has not been dated directly because of the nature of the archaeological remains. Because of its very small size (1.6 mg) the sorghum sample was subjected to a gentle acid (HCl, 1 N) wash at room temperature for 1 h, then rinsed using Milli-Q water and dried overnight at 90 ◦ C. The 1 mgsample was combusted, providing 600 μg C and graphitized using an automated AGE 3 device. Radiocarbon measurement was performed using the compact AMS ECHoMICADAS at LSCE (Saclay, France). The radiocarbon age was calibrated using the Oxcal 4.4 software and the 4. Results 4.1. Analysis of seed and fruit remains Seed and fruit remains are relatively rare with a total of 434 items observed (MNI = 426) (Table 2). These elements correspond to whole and fragmented seeds and fruits as well as other vegetative parts of plants such as leaves and spines. Most of the material is preserved by carbonisation (71%) caused by charring occurring during daily-life activities or accidental fire. Some of them have been preserved by mineralisation (29%). The latter correspond to Boraginaceae nutlets. However, in such cases it can be difficult to ascertain whether they are ancient or modern remains. Indeed, they can survive in archaeological deposits by natural mineralisation (Messager et al., 2010) without any obvious morphological difference from modern specimens which makes it impossible to ascertain their age without direct dating. As a proportion may be part of the contemporary archaeological formation, we decided to take them into consideration, though we have to bear in mind their potentially problematic nature. In any case, the abundance of seed and fruit items recovered is very low, with the average density of remains per litre comprising between 0.1 and 8.2 (Fig. 5). Most of the samples (79%) have density values of less than 1. Samples with a density of remains per litre equal or more than 1 come from an oven (F.026, F.047), two hearths (A.014, G.014), an occupation deposit (E.024) and two levels associated with combustion structures (F.009, F.027). Two samples from 6 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Table 2 General results of seed and fruit analysis per phase. NR = Number of Remains; MNI = Minimum Number of Individuals; fg. = fragment; U = Ubiquity; % = Percentages. Ubiquity has been defined by taxon. Period Phase 1 Phase 2 Phase 3 Phase 4 Dating c.1000 - 500 BCE c. 5th-6thC CE c. late 6th - 8thC CE c. 16th-20thC CE Trenches B, G, N A, B, E A, E, F, G A, B, F, G A, B, E, F, G, N Number of samples 7 6 15 12 40 Total of volume sieved (litre) 123 Vernacular name Cereals Hulled barley, caryopsis Hulled barley, fg. caryopsis Probable millet, caryopsis Sorghum, caryopsis Sorghum, fg. caryopsis Sorghum, earth imprint Probable sorghum, caryopsis Free-threshing wheat, caryopsis Und. cereal, fg. caryopsis Pulses Und. pulse, cotyledon Fruit trees Date, fg. seed Probable date, fg. seed Probable date, perianthe Jujube, fg. of endocarp Probable jujube, fg. of endoarp Weedy/Wild plants Acacia/Prosopis, immature seed Probable acacia/prosopis, fg. of seed Amaranthaceae, seed Amaranthaceae, leave cf. Amaranthaceae, fg. of leaf Asphodel, seed Boraginaceae, nutlet Brassicaceae, seed cf. Brassicaceae, seed Barnyard grass, caryopsis Euphorbiaceae, seed Und. pulse, seed Und. pulse, cotyledon Probable pulse, seed Probable helianthemum, seed Lamiaceae, seed Mallow, seed Probable alfalfa, seed Probable alfalfa, fg. of seed Sweet clover, seed Millet, caryopsis Probable millet, caryopsis Probable millet, fg. of caryopsis cf. Panicum, caryopsis Graminae, caryopsis Graminae, caryopsis with chaff Graminae, fg. of caryopsis Salsola, seed Foxtail, caryopsis Solanaceae, fg. of seed Indeterminate Indeterminate, seed Indeterminate, remains in arc Indeterminate, fg. of pericarp Indeterminate, spine Organic amorphous remains Coprolites Coprolites, rodent dropping Coprolites, sheep/goat pellet cf. Coprolites Total of determined carpological remains 91.5 187.5 Total 168.5 NR MNI U NR MNI U NR MNI U – – – – – – – – – – – – – – – – – – – – – 1 – – – – – – 1 – – – – – – 1 – – – – – – 1 2 1 1 1 2 1 1 1 1 1 1 2 1 2 2 2 1 – – – – – – – – – – – – – 1 1 1 – – – 1 1 – – – 1 1 – – – 1 1 – – – – – – – – 14 – 28 – – – – – – – – – – – – – – – – – – – – – – – – – – – 14 – 28 – – – – – – – – – – – – – – – – – – – – – – – – – – – 2 – 7 – – – – – – – – – – – – – – – – – – – – – – – 10 – – – × 10 – – – × 1 – 1 18 592.5 NR MNI U NR MNI % 1 3 – – – – – – – – – 1 – – – – – – 1 – – – – – – 1 2 2 1 1 1 2 2 2 1 1 1 1 2 2 <1 <1 <1 <1 <1 <1 <1 – – – – – 2 2 <1 – – – 1 1 1 1 1 <1 1 – – – – – – 1 1 <1 1 – – – – 1 – – – – 1 6 1 1 5 1 3 1 1 5 5 – – 5 – 1 4 – – 1 1 2 – – 1 2 – – – 1 4 10 1 1 7 4 5 1 1 7 1 1 <1 <1 2 1 1 – 1 – – 23 – 1 – – – – 1 – – – – – – – – – – – – – 1 – – 1 1 – 1 – – 23 – 1 – – – – 1 – – – – – – – – – – – – – 1 – – 2 – – 1 – – 4 – 1 – – – – 1 – – – – – – – – – – – – – 1 – – – – 4 3 93 1 50 2 – 1 1 7 3 1 1 1 – 1 1 1 1 1 1 – 1 2 1 1 1 1 – – 4 3 93 1 50 2 – 1 1 7 2 1 1 1 – 1 1 1 1 1 1 – 1 2 1 1 1 1 – – 5 – – 1 13 2 – 1 1 4 – – 1 1 – 1 – 1 3 – – – 2 – – 1 1 1 – – 1 – 16 2 21 – – – – 1 – – – – 1 – – – – – – 1 – – 1 2 – – – – 1 – 16 2 21 – – – – 1 – – – – 1 – – – – – – 1 – – 1 2 – – – – 4 – – 1 7 – – – – 1 – – – – 1 – – – – – – 1 – – 1 1 – – 1 1 5 4 123 3 122 2 1 1 1 8 3 2 1 1 1 1 1 1 1 1 1 1 1 2 2 4 1 1 1 1 5 4 123 3 122 2 1 1 1 8 2 2 1 1 1 1 1 1 1 1 1 1 1 2 2 4 1 1 <1 <1 1 1 29 1 29 <1 <1 <1 <1 2 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 1 <1 <1 4 – – – 2 2 – – – × 2 – – – × 2 – – – 1 38 24 – 8 – 38 24 – 8 – 6 2 – 2 – 14 – 2 – × 14 – 1 – × 6 – 1 – 1 64 24 2 8 × 64 24 1 8 × 15 6 <1 2 – 1 – 1 1 – 1 – – – – – – – – – 1 4 1 1 4 1 1 2 1 – – – – – – – – – 2 4 2 2 4 2 – – – 18 7 32 32 5 204 199 15 54 52 10 336 329 77 Latin name Hordeum vulgare cf. Panicoïdae Sorghum bicolor ssp. bicolor cf. Sorghum bicolor ssp. bicolor Triticum aestivum/durum/ turgidum Cerealia Fabaceae Phoenix dactylifera cf. Phoenix dactylifera Ziziphus cf. spina-christi cf. Ziziphus sp. Acacia/Prosopis cf. Acacia/Prosopis Amaranthaceae cf. Amaranthaceae Asphodelus cf. tenuifolius Boraginaceae Brassicaceae cf. Brassicaceae Echinochloa cf. colona Euphorbiaceae Fabaceae cf. Fabaceae cf. Helianthemum sp. Lamiaceae Malva sp. cf. Medicago sp. cf. Melilotus sp. Panicoidae cf. Panicoïdae cf. Panicum sp. Poaceae Salsola sp. Setaria sp. Solanaceae (continued on next page) 7 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Table 2 (continued ) Period Phase 1 Phase 2 Phase 3 Phase 4 Dating c.1000 - 500 BCE c. 5th-6thC CE c. late 6th - 8thC CE c. 16th-20thC CE Total Trenches B, G, N A, B, E A, E, F, G A, B, F, G A, B, E, F, G, N Number of samples 7 6 15 12 40 Total of volume sieved (litre) 123 91.5 187.5 168.5 592.5 NR MNI U NR MNI U NR MNI U NR MNI U NR MNI % Total of undetermined carpological remains Total of carpological remains 10 10 5 2 2 3 70 70 8 16 15 8 98 97 28 28 28 7 34 34 6 274 269 15 70 67 12 434 426 100 animal bones, potsherds, including green coloured alkaline glazed ware (TURQ.G) and locally made coarse buff ware (COB), and two glass vessel rim fragments: one a narrow-necked bottle and the other a closed bowl. In addition, two earth impressions of sorghum grains have been found in the fill of a large post-hole (F.036) within the northeast corner of the fort interior. Two fragmentary sorghum have been found in two further contexts, one dated to the Early Islamic period (A.014) and the other to the Late Islamic period (F.011), but they are not so securely identified. The sorghum grain is widely ovate to round with a shallow scutellum (embryo cavity) covering approximatively half to two-thirds of the total length of the grain (Fuller 2017). Its big size and globular aspect allow one to identify it as a domesticate specimen belonging to subspecies S. bicolor ssp. bicolor according to Wiersema and Dahlberg (2007). Five main complex landraces have been defined on the basis of spikelet and inflorescence morphology and genetic features but it remains hard to differentiate them archaeologically, especially with only one whole caryopsis (Fuller and Stevens 2018). These grains constitute the first evidence of sorghum safely identified and dated in Eastern Arabia as we will see in further details below. Added to sorghum, three caryopses of hulled barley (Hordeum vulgare) are attested during Late Sasanian and Early Islamic periods, and two of free-threshing wheat (Triticum aestivum/durum/turgidum) are present only in one Iron Age period hearth. It is impossible to distinguish bread hexaploid wheat (T. aestivum ssp. aestivum) from tetraploid wheats (Triticum turgidum ssp. durum/turgidum) by the sole morphology of the caryopsis. One single Panicoideae caryopsis and one potential undetermined cereal are attested respectively in a sample dated to the Early Islamic (F.047) and Late Islamic (A.009) periods. Their general state of preservation did not allow us to provide a precise identification. Fruit trees are represented by date palm (Phoenix dactylifera) (2%) with fragments of date seeds occurring in all occupation phases; one potential perianth (remnant parts of flowers after the fruit formation) has been recovered in one Early Islamic context (A.014). Endocarp (fruit stone) fragments of jujube (Ziziphus cf. spina christi) (2%) have been found in samples from contexts dated to all phases, except the Late Sasanian period, and particularly during the Early Islamic period. Only one cotyledon of undetermined domesticated pulse has been found in one Late Sasanian context (E.028) but its mediocre state of preservation prevented us from obtaining a more precise identification. Weedy/wild plant remains are more numerous than crops (70%) with two having been identified to the species level, eight to the genus level and nine to the family or sub-family level. As these taxa are not precisely identified, it is in general difficult to determine their ecology. Their importance in the assemblage is mainly due to high proportions of Boraginaceae (29%) nutlets and possible fragments of Amaranthaceae leave (29%). The Amaranthaceae family includes many shrubs (formerly ascribed to the Chenopodiaceae family) often growing on saline and sandy soils in deserts, foothills and coastal areas in Northern Oman. This family is also attested by seeds from Salsola which correspond to shrubs growing mainly on sandy and rocky soils in coastal areas. The Fabaceae (4%) and Poaceae (3%) families are both minimally attested by several potential weeds such as barnyard grass (Echinochloa cf. colona), foxtail Fig. 5. Density of remains per litre for all archaeobotanical samples. the same oven (F.026, F.047) have important concentrations of possible Amaranthaceae leaves which seems to underline the use of this taxon as fuel resulting in an over-representation in the assemblage (29% of the total). In total, the seed and fruit remains have been grouped into 27 taxa, including 8 cultivated taxa and 19 weedy/wild plants (Fig. 6). Among the former, cereals and fruit trees are most frequently encountered (3% and 4% respectively). Among the cereals, two carbonized caryopses of sorghum (Sorghum bicolor ssp. bicolor) have been recovered from an Early Islamic context (E.024) located close to the base of the main wall outside the fort beyond the southeast entrance flanking tower. The context consists of a sandy-silt occupation deposit containing stones, 8 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Fig. 6. Seeds and fruit remains from Fulayj fort. A: Caryopsis of sorghum (Sorghum bicolor ssp. bicolor) in dorsal, lateral and ventral view; B: Earth impression of sorghum caryopsis, in dorsal, lateral and ventral view; C: Caryopsis of hulled barley (Hordeum vulgare), in dorsal, lateral and ventral view; D: Caryopsis of freethreshing wheat (Triticum aestivum/durum/turgidum) in dorsal, lateral and ventral view; E: Fragment of date palm seed (Phoenix dactylifera) in dorsal, lateral and ventral view; F: Caryopsis of barnyard grass (Echinochloa cf. colona) in dorsal, lateral and ventral view; F: Caryopsis of foxtail (Setaria sp.) in dorsal, lateral and ventral view; H: Immature seed of Acacia/Prosopis in front and lateral view; I: Seed of wild alfalfa (cf. Medicago sp.) on both side; J: Seed of asphodel (Asphodelus cf. tenuifolius) in dorsal, lateral and ventral view; K: Seed of Salsola sp.; L: Seed of mallow (Malva sp.); M: Fragment of leaf (cf. Amaranthaceae). (Setaria sp.), cf. Panicum sp., alfalfa (cf. Medicago sp.) and sweet clover (cf. Melilotus sp.) as is the case with other families with asphodel (Asphodelus cf. tenuifolius, Liliaceae) and mallow (Malva sp., Malvaceae). Barnyard grass, foxtail, alfalfa, sweet clover and mallow used to grow in irrigated fields (Ghazanfar, 1992, 2003, 2007). These taxa are attested in the assemblage during the Early Islamic period and thus, testify agrarian practices dated to this period. Seeds of acacia/prosopis (Acacia/Prosopis) have also been identified during the Late Sasanian period. Acacia and Prosopis are two common tree taxa amongst the current vegetation cover. Their seeds could have been accidently brought on site with wood collected as fuel or grazed by animals (Fagg and Stewart 1994). Finally, high proportions of badly preserved and fragmented plant remains (23%) could not be botanically determined. We can observe the presence of indeterminate remains under the form of arc that may belong to acacia/prosopis as suggested by the association with charcoals from these taxa although their nature remains to be determined. In addition, some fragments of coprolites have been noted amongst the samples. Four of them have been identified as sheep or goat pellets dated to the Early Islamic period and two of them as rodents’ droppings, one to 9 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 the same period and another to the Iron Age. (drought-tolerant) shrubs such as Grewia sp. (2%), Periploca sp. (1%) and cf. Pergularia tomentosa (<1%) (Ghazanfar 2007: 12–18, Ghazanfar 2003: 91–93). Some taxa are also attested as parts of hygrophilous formations growing in and near wadis or irrigated channels, like the Nile Acacia (Acacia cf. nilotica) (1%) and willow (Salix cf. acmophylla) (<1%) (Ghazanfar 2007: 11, Ghazanfar 2003: 129–130). For the latter, it is actually difficult to determine if it is indigenous or introduced since it is rarely observed today and mostly where it does occur, it is close to villages. 4.2. Charcoal analysis Among 591 analysed charcoal fragments, 551 could be identified to the species or genus level (Table 3). 40 others could only be identified to the larger Angiosperm group or remained indeterminate, including bark fragments. 102 charcoal fragments (17% of the total assemblage, indeterminate included) belong to one single Iron Age sample, 30 fragments (5%) come from one Late Sasanian sample, and 459 fragments (78%) from five contexts belonging to the Early Islamic Period. A total of ten taxa have been determined botanically among which six to the species level and four to the genus level. Two main categories of woody plants can be distinguished: indigenous wild trees and shrubs (N = 8) and cultivated fruit trees (N = 2), although the precise status – wild or cultivated? - of some taxa requires further consideration (Figs. 7 and 8). Tamarisk (Tamaris sp.) is the dominant taxon (20%), mostly concentrated in the Iron Age hearth (G.014), probably indicating that all the charcoals come from the same piece of wood associated with the last episodes of burning. Tamarisk is a hardy tree that can withstand both drought and flooding by growing in or on the edges of wadis. They are also salt-tolerant and thus found on saline soils in coastal areas and are planted as sand-stabilizers and wind-breaks near date palm gardens (Ghazanfar 2003: 120–122). Date palm (Phoenix dactylifera) is the second most represented taxon (19%), including petiole remains. They are mainly concentrated in the Early Islamic hearth (A.014) showing the use of petiole as fuel, as already seen in the Arabian Peninsula (Bouchaud et al., 2012). Several taxa are main components of the vegetation cover typical of plains and foothills of Southeast Arabia such as prosopis (Prosopis cf. cineraria) (19%), Christ’s Thorn jujube tree (Ziziphus cf. spina-christi) (10%), and acacia (Acacia sp.) (5%) (Kennet 2007: 6–11, 98–99). Jujube can also grow as cultivated fruit trees in gardens. The high concentration of prosopis and acacia charcoal in the Early Islamic oven is notable (F.047). Other elements of the vegetation belong to taxa mainly attested in the foothills and in mountainous areas but are present in the assemblage in minor proportions. These include xerophytic 4.3. Radiocarbon dating The sample of sorghum from context E.024 provided an age of 1175 ± 20 BP, corresponding to a calibrated range of between 772 and 950 Cal. CE, at 2 sigma or 95.4% probability (Fig. 9 and Table 4). The charred twig of Tamarix from the same context provides an AMS date of 1306 ± 22 BP, corresponding to a calibrated range of between 660 and 775 Cal. CE at 2 sigma or 95.4% probability. The piece of Tamarix wood from context (F.036) provided an AMS date of 1405 ± 28 BP with a calibrated range of between 601 and 664 Cal. CE at 2 sigma or 95.4% probability (calibration IntCal2020, Reimer et al., 2020). 5. Discussion 5.1. Trade versus local cultivation of sorghum If we exclude the claim for the 3rd mill. BCE presence of sorghum at Hili 8 (Cleuziou 1982; Cleuziou and Costantini 1982, see above) that has been questioned by several authors (Bouchaud et al., 2016; Tengberg 2012; Charbonnier 2008; de Moulins et al., 2003), the evidence from Fulayj represents the earliest securely identified and dated occurrence of sorghum in Eastern Arabia. The direct dating of sorghum from context E.024 is slightly later than the tamarisk twig from the same deposit and may testify to a longer occupation at Fulayj fort during the Early Islamic period than initially thought. In addition, the two glass fragments found in this context (E.024/FN234 and FN251) appear consistent with a date Table 3 Results of the charcoal analysis. Period Phase 1 Phase 2 Phase 3 Dating c. 1000 - 500 BCE c. 5th-6thC CE c. late 6th - 8thC CE Trench G B A A F F F Context G.014 B.007 A.014 A.016 F.026 F.030 F.047 7 Volume (litres) 10 22 20 18 15 3 11 99 Vernacular name Latin name Acacia Probable acacia Nile acacia Acacia/Prosopis Grewia sp. Probable Pergularia tomentosa Periploca sp. Date palm, petiole Date palm, indeterminate Willow Prosopis Probable prosopis Tamarisk Probable tamarisk Jujube tree Probable jujube tree Angiosperm dicotyledon Angiosperm dicotyledon, bark Indeterminate Total of determined charcoals Total of indetermined charcoals Total of charcoals Acacia sp. cf. Acacia sp. Acacia cf. nilotica Acacia/Prosopis Grewia sp. cf. Pergularia tomentosa Periploca sp. Phoenix dactylifera Salix cf. acmophylla Prosopis cf. cineraria cf. Prosopis sp. Tamarix sp. cf. Tamarix sp. Ziziphus cf. spina-christi cf. Ziziphus sp. 1 – 1 2 – – – – – – 2 – 21 – – – 2 – 1 27 3 30 – – – – – – – – – – – – 75 – 27 – – – – 102 0 102 10 – – – – – – – 35 70 – – – – – – – – – – 105 0 105 Total 2 – – – – – – 3 2 – – – 13 – 2 – 8 – – 22 8 30 – 11 – 38 – – 5 3 – – 43 – 1 – 1 – – – 3 102 3 105 3 – – 17 – 1 – – – 1 17 22 7 3 29 2 5 – 8 102 13 115 6 4 6 28 13 – 3 2 – – 20 9 – – – – 2 11 – 91 13 104 A, B, F, G NR % 12 15 7 85 13 1 8 43 72 1 82 31 117 3 59 2 17 11 12 551 40 591 2 3 1 14 2 <1 1 7 12 <1 14 5 20 1 10 <1 3 2 2 93 7 100 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Fig. 7. Charcoal remains from Fulayj fort (SEM photographs) (1/2). A: Acacia sp., transverse section; B: Acacia cf. nilotica, transverse section; C: Prosopis cf. cineraria, transverse section; D: Tamarix sp., transverse section; E: Ziziphus cf. spina-christi, transverse section; F: Phoenix dactylifera (petiole), transverse section. sorghum from Fulayj is clearly dated with the period immediately following the conversion to Islam. The evidence of sorghum from Fulayj raises the question about the origin of this crop; whether it corresponds to an imported trade item, or if it was locally acclimatized and cultivated. In the former case, both East Africa and India are relevant candidates as sources since sorghum is known to have been grown in both these areas at this time. Textual references, although scarce, mention trade activities between Oman and these regions in the Early Islamic period. The integration of Omani harbours, like Sohar, within the international trade networks of the western Indian Ocean therefore created a favourable context for the import of new food products to Fulayj. While it remains difficult to within the c. 7th/8th century. A later date within the 9th or 10th century, that the absolute dating of the sorghum material would potentially allow, seems unlikely on archaeological grounds as there is a complete absence of characteristic 9th or 10th century ceramics from anywhere within the excavations or across the surrounding landscape. Furthermore, it is relevant to consider the archaeological dating of sorghum within the fort in context F.036, which indicates the presence of the crop most likely within the first half of the 7th century according to the material found within this context including the AMS dating evidence. The context is certainly dated before the later 8th century since it is stratigraphically sealed by the construction of an oven (F.026) above dated to 1295 ± BP (664–774 Cal. CE at 94.5% probability). Therefore, 11 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Fig. 8. Charcoal remains from Fulayj fort (SEM photographs) (2/2). H–I: Grewia sp., transverse (H) and longitudinal radial (I) section; J–K: Periploca sp., transverse (J) and longitudinal tangential (K) section; Willlow (Salix cf. acmophylla) (L–M), transverse (L) and longitudinal radial (M) section. determinate the region of origin of sorghum, the presence of other cultural materials such as South Asian pottery at Fulayj (al-Jahwari et al., 2018), may point in favour of the Indian sub-continent. However, Yemen might also be considered as potential source of origin. Indeed, archaeobotanical evidence of sorghum grains and glumes in sheep and camel coprolites are found at Zabid (9th to 10th century) (McCorriston and Johnson 1998) and later Rasulid textual references mention numerous local varieties of sorghum in Yemen during the 14th century (Varisco and Umar Ibn Yusuf 1994: 165, Varisco 1991). Whatever the origin, the presence of sorghum at Fulayj sheds important new light on potential routes of supply and the dissemination of novel food commodities via processes of long-distance maritime exchange during the Early Islamic period. An alternative explanation for the presence of sorghum at Fulayj is local acclimatisation of an exogenous crop. This drought-resistant cereal is well-adapted to local climatic conditions in Oman, where the hot summer temperatures would have suited its ecological requirements. According to Ulbaydli (1993), the Jāmi’ of Ibn Ja’far (iii. 35), an author of the Omani Ibadi jurisprudence living in the 9th century, refers to the cultivation of sorghum in Oman during the Early Islamic period potentially adding to the available archaeological evidence. However, we were not able to get access to the original arabic mention of the Jāmi’ of Ibn Ja’far so far so we have to be cautious about its accuracy. Nowadays, sorghum has been recorded growing on the Batinah coast (Esechie 12 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Fig. 9. AMS dating of sorghum grain and other macro-botanical remains related to sorghum. Table 4 Detailed results of AMS dating included in this paper. Calibrated age, CE Context Find N◦ Lab N◦ Sample description 14C age Error 68.2% 95.4% E.024 E.024 F.036 – FN246 SN20 ECHo-3470.1.1 Poz-89865 Poz-89926 Sorghum caryopsis Tamarisk twig Tamarisk 1175 1306 1405 20 22 28 776–890 667–772 608–657 772–950 660–775 601–664 Emirate of Abu Dhabi (UAE) and dates to the late 6th to early 5th millennium BCE (Beech and Shepherd 2001). Date palm remains increase from the Bronze Age onwards (Tengberg 2012) and are well attested on several Iron Age sites (Cerro 2013; Bellini et al., 2011; Tengberg 1998; Willcox and Tengberg 1995; Costantini and Costantini-Biasini 1986). The earliest occurrence of jujube in archaeological contexts corresponds to an abundant find of about 300 fruit stones on the Neolithic sites of Ra’s al-Hamra H5 and RH6 (5th to early 3rd millennium BCE) in Oman showing the gathering of the edible fruits since prehistoric times (Biagi and Nisbet, 1992, 1999). Both dates and jujubes are attested in Sasanian and Early Islamic levels at Kush, the former as the most dominant fruit tree (Dabrowski 2019: 174–197, Dabrowski et al. forthcoming). The general composition of archaeobotanical assemblages together with climatic data may provide information on the organisation of local crop cultivation. The predominance of date palm remains (seeds and petiole fragments) indicates the existence of date palm gardens during all periods of occupation. Modern oases in the Middle East are vertically organised with date palms forming the upper level and providing shade to other fruit trees and annual crops cultivated in irrigated plots below the canopy (Battesti 2005; Munier 1973). The recurrent association of date and cereal remains in contexts where rainfed agriculture is excluded suggests that agriculture was practised in date palm gardens in Eastern Arabia since at least the 3rd millennium BCE (Tengberg 2012). Some of the weedy/wild taxa found at Fulayj associated mostly with the Early Islamic contexts correspond to weeds growing in current irrigated fields (Ghazanfar, 2003, 2007). Hygrophilous trees attested among the charcoal assemblages (tamarisk, Nile acacia, willow) may have been planted along irrigation channels in the date palm gardens; tamarisk could have also been established as wind-breaks or to stabilise the soil at the edges of plantations. The presence of one charcoal of willow dating to the Early Islamic period may represent further evidence for irrigation since, in the hot climate of Eastern Arabia, this tree would only grow along permanent watercourses or irrigation channels as observed today in the Sultanate of Oman (Ghazanfar 2003: 129–130). In the case of sorghum acclimatisation, it must have been incorporated in such an agrosystem where irrigation would have met its hydrological needs, as nowadays it is grown in low-altitude mountain oases in the northern part of the Sultanate of Oman (Gebauer et al., 2007; Guarino 1990). 1994). The two earth impressions found at Fulayj appear to favour the hypothesis of local acclimatisation. Indeed, these impressions may be explained by the accidental inclusion of grains together with by-product material in mudbrick architecture. The use of cereal chaff as a temper in earthen building materials is usually an indication of local cultivation, even though the trade in chaff cannot be entirely ruled out. In conclusion, the evidence of sorghum in Fulayj may thus testify to the introduction of sorghum agriculture in Eastern Arabia during the Early Islamic period although supplementary data would be needed in order to confirm this hypothesis. In either case, the import of sorghum as a foodstuff, or the translocation of a new crop and its addition to the repertoire of local agriculture during the Early Islamic period, would have depended on the vehicle of long-distance exchange to gain access to what appears to have been a new food commodity in the region. 5.2. Plant diet and oasis agriculture Archaeobotanical remains provide a crucial insight into certain aspects of diet and food production systems. In addition to the new evidence for sorghum, recovered staples include cereals with free-threshing wheat during the Iron Age and, hulled barley during the Late Sasanian and Early Islamic period. Free-threshing wheat and hulled barley are attested in Eastern Arabia since at least the Early Bronze Age (3rd millennium BCE) mainly in the form of chaff impressions on mudbrick (Willcox 1995; Willcox and Tengberg 1995; Cleuziou 1982). Well-preserved mudbrick impressions of free-threshing wheat chaff (glumes and rachis segments) correspond to the bread wheat type. In addition, the Iron Age site of Salut (Sultanate of Oman) has produced evidence of a single grain of Triticum sp. and pollen of Triticum and Hordeum groups (Bellini et al., 2011). A contemporary example of hulled barley is also attested at Kush (UAE) in the Sasanian and Early Islamic levels. Moreover, undetermined pulses are present at Fulayj during the Late Sasanian period. The first evidence of pulses in Eastern Arabia has been reported from the Bronze Age site Hili 8 as pea (Pisum sativum) (Tengberg 2012). The Sasanian contexts of the site of Kush have yielded some lentil (Lens culinaris ssp. culinaris) and grass pea (Lathryrus sativus) (Dabrowski 2019: 174–197, Dabrowski et al. forthcoming). Fruits constitute another part of the plant diet with dates represented throughout the occupational sequence and jujubes attested in particular during the Early Islamic period. The earliest evidence for date consumption in Southeast Arabia comes from the site of Dalma 11 in the 13 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 textual references, maybe by the time of Khusraw I during the 6th century (Wilkinson 2010: 57–60, Wilkinson 1979: 888–889, Wilkinson 1977: 130–133). Such a production system may have been complemented by more local initiatives closer to the fort as is suggested by the evidence discussed above. Food supply (for example, cereals) from more distant regions belonging to the Sasanian Empire is also possible. The predominance of imported pottery at Fulayj from sources in southern Mesopotamia and Iran could be taken as one potential indication of a regular system of external provisioning. The predominance of locally grown trees in the charcoal record indicates that most of the fuel wood was collected within the vicinity of the site, mostly from open dry woodlands dominated by acacia (Acacia sp.) and jujube trees (Ziziphus cf. spina-christi) as well as prosopis (Prosopis cf. cineraria) on deeper soils. Tamarisk wood (Tamarix sp.) may have been collected along wadis and field edges. The presence of large acacia and prosopis charcoal fragments in samples from the Early Islamic period oven within the northeast corner of the fort (F.026, F.047) may result from their selection for calorific properties or their suitability for charcoal production (Fagg and Stewart 1994; Leakey and Last 1980). The absence of Amaranthaceae among the charcoal assemblage, while some potential leaf fragments have been recovered, is quite surprising. Shrubs from the Amaranthaceae family should have grown near the site, and their wood elements might have been used as fuel. Methodological issues connected with sampling or taphonomic factors may explain their absence but this question deserves further investigation. That the catchment area for firewood extended somewhat beyond the immediate surroundings of the site is indicated by the presence in Early Islamic samples of shrub species such as Grewia sp., cf. Pergularia tomentosa and Periploca sp., that are more likely to have grown in the foothill zone or in the mountains. Periploca, for example, does not grow today below 500 m asl in the mountains of Oman. The woody parts of shrubs may also have been brought to the site as by-products of other exploitation processes. For example, we can cite the case of Grewia whose edible fruits are gathered for human consumption and its foliage as fodder or even Pergularia tomentosa and Periploca that can be used for medical treatments (Ghazanfar 2007: 12–18, Ghazanfar 2003: 91–93). However, the present vegetation cover is highly degraded and taxa that today grow only in the foothill zone possibly had a wider distribution in the past. 5.3. Local or regional plant acquisition strategies? Date palm gardens may have been located near to the fort as is the case today. As already noted above, groundwater tables on the Batinah plain are too deep to be reached through wells outside the coastal area (Wilkinson 1977: 48–49). Today’s date palm gardens surrounding the adjacent village of Falaj Al-Harth are fed with water provided via aflāj; underground galleries or irrigation channels, that capture the water upstream where the flow is still close to the surface. Initially, the Sasanian period was thought to have been a period of extensive development of irrigation systems with aflāj attested in the hinterland of Sohar but these have more recently proved to date to the Islamic period (Costa and Wilkinson 1987: 54, Wilkinson 1977: 130–133). Several open irrigation channels have been excavated in the vicinity of Fulayj fort (Fig. 10) and may have been used for agricultural purposes. OSL dating of selected channels indicates construction during the Late Islamic period (Snape Kennedy 2018), though the possibility of earlier irrigation systems cannot be excluded on the basis of this evidence. Similar open irrigation channels and aflāj are attested with certainty since the Iron Age in Southeast Arabia (Charbonnier, 2015, 2017), and their use for watering date palm gardens at Fulayj too appears likely. Moreover, the presence of woody tissues from date palm in the assemblage, notably in an Early Islamic hearth (A.014), indicates that its by-products such as petiole were commonly used as fuel as demonstrated in other charcoal analyses in the Arabian Peninsula (Bouchaud et al., 2012). So, the evidence indicates that date palm gardens, in addition to producing staple food crops, also generated valuable by-products in the form of fuel resources that were also regularly exploited during the occupation of Fulayj. While locally available plant resources clearly played a dominant part, it is also important to keep in mind the possibility of the use of wider regional supply strategies, particularly within the military context of the fort’s initial foundation. Indeed, some areas of the Batinah plain may be considered as more desirable for establishing agricultural systems. Today date palm gardens are mostly concentrated close to the coast (Fig. 11), which has edaphic and hydrologic advantages compared to Fulayj (Sanlaville 2000: 140–141, Wilkinson 1977: 48–49). This hypothesis of regional food supply systems could be further supported if the fort was part of a wider defensive and military network (Kennet et al., 2016). These ideas feed into the notion of increasing Sasanian involvement in agrarian activities on the Batinah suggested by some Fig. 10. Open irrigation channel close to Fulayj dated by OSL to the Late Islamic period. 14 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 Fig. 11. Satellite photograph with repartition of date palm gardens in the Batinah plain. Date palm gardens are mainly concentrated today in the coastline sector. 6. Conclusions cultivation and fixed irrigation infrastructure, in particular, suggests a significant degree of long-term planning and investment, and perhaps also, helps to explain the seeming longevity of occupation at Fulayj. At the same time, crucial evidence emerges for aspects of innovation and adaptation closely associated with the existence of long-distance maritime exchange networks. While the archaeobotanical evidence still remains relatively limited and partial, it does suggest important new avenues for investigation within the region. This includes potential alternative pathways to the transformation of society in Eastern Arabia during the later 1st millennium CE based on the development of innovative agricultural strategies capable of bolstering dietary resilience and diversity. The archaeobotanical analysis conducted at the site of Fulayj documents the food and fuel acquisition strategies in this arid environment during the Late Sasanian and Early Islamic periods which are still underrepresented in Eastern Arabia. The first direct dating of sorghum to the Early Islamic period makes it the earliest securely identified occurrence of this crop in the region. This warm and drought-resistant cereal may have been introduced to the site either in the form of an imported food commodity via long-distance exchange from areas such as the Indian sub-continent, Africa or Yemen, or have been introduced from these sources and acclimatized locally as is potentially indicated by earth impressions of sorghum grains. In addition to sorghum, the main components of the plant diet have been defined as cereals (hulled barley, free-threshing wheat during the Iron Age) and fruits (date, jujube). The composition of the whole assemblage shows that the agricultural system, which provided food and fuel products, is constructed around the use of date palm gardens; an oasis agrosystem typical of arid and semi-arid environments of the Middle East, attested in Eastern Arabia since the 3rd millennium BCE. In the case of acclimatisation, sorghum is likely to have been grown within this agrosystem as is seen nowadays in the Omani mountains. Date palm gardens were probably established close to the site, as they are today, as is suggested by the associated archaeobotanical remains. However, agricultural systems may also have been maintained in the shoreline sector of the Batinah plain where soil factors and hydrological conditions are better suited to agriculture, maybe within the framework of regional scale food supply systems organised by foreign political authorities. Fuel management practices at Fulayj include the exploitation of locally available species (acacia, prosopis, jujube tree, tamarisk) together with firewood gathering from the foothills and surrounding mountainous areas. The archaeobotanical assemblage obtained as part of the archaeological investigation of Fulayj provides substantial complementary information that helps us to understand the functioning of the site. Against the background of political and religious transformation between the 5th to 8th centuries, subsistence economies seem on the contrary to be underpinned by long-term factors of continuity and well adapted strategies to local ecological conditions. The establishment of palm garden CRediT authorship contribution statement Vladimir Dabrowski: Conceptualization, Methodology, Formal analysis, Investigation, Resources, Writing – original draft, Writing – review & editing, Visualization, Funding acquisition. Charlène Bouchaud: Validation, Writing – original draft, Writing – review & editing, Supervision. Margareta Tengberg: Validation, Writing – original draft, Writing – review & editing, Supervision. Antoine Zazzo: Formal analysis, Investigation, Writing – original draft, Writing – review & editing, Visualization. Seth Priestman: Resources, Writing – original draft, Writing – review & editing, Project administration, Funding acquisition. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements We would like to extent our particular thanks to the Ministry of Heritage and Culture of the Sultanate of Oman for supporting our work and especially to Sultan Al-Bakri, Director of the Department of Excavations and Archaeological Studies. We are also particularly grateful to the directors of the Fulayj Fort Project: Dr. Nasser Saïd al-Jahwari (Department of Archaeology, Sultan Qaboos University) who first 15 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 identified the site of Fulayj and helped in many ways to facilitate our work; Dr. Derek Kennet (Department of Archaeology, Durham University) for support and guidance throughout the study; and, Prof. Eberhard Sauer (School of History, Classics and Archaeology, University of Edinburgh) for integrating the investigation of Fulayj into the European Research Council funded Persia and its Neighbours Project. We would like also to thank Enki Baptiste (Université Lumière Lyon 2) and Dr. Harry Munt (Department of History, University of York) for their help with investigating textual mentions of sorghum in Ibadi sources. We would like to thank the doctoral college of Sorbonne Universités who granted the first author a PhD fellowship which permit the archaeobotanical analysis. Work was directed in the field by Dr. Seth Priestman. Processing of the archaeobotanical assemblage was undertaken by Dr. Vladimir Dabrowski and made possible by the efforts of the entire fieldwork team. Cappers, R.T.J., Neef, R., Bekker, R.M., 2012. Digital Atlas of Economic Plants in Archaeology. Groningen Archaeological Studies, 17. Barkhus and Groningen University Library, Groningen. Carter, R., 2008. « Christianity in the Gulf during the first centuries of islam ». Arabian Archaeol. Epigr. 19 (1), 71–108. https://doi.org/10.1111/j.1600-0471.2008.00293. x. Cerro, C. del, 2013. « Biological remains at Al-Madam (Sharjah, UAE) archaeological, archaeobotanical and archaeozoological studies in an Iron age farmingstockbreeding village ». Bioarchaeol. Near East 7, 21–32. Chabal, L., 1997. Forêts et sociétés en Languedoc néolithique final, Antiquité tardive l’anthracologie, méthode et paléoécologie.. Documents d’archéologie française, vol. 63. Maison des sciences de l’homme, Paris. Chabal, L., Fabre, L., Terral, J.-F., Théry-Parisot, I., 1999. « L’anthracologie ». In: Ferdière, A. (Ed.), La Botanique, Collection “Archéologiques”. Errance, Paris. Chantereau, J., Cruz, J.-F., Ratnadass, A., Trouche, G., Fliedel, G., 2013. Le Sorgho. Quae, Versailles. Charbonnier, J., 2008. « L’agriculture en Arabie du Sud avant l’Islam ». Chron. Yéménites 15, 1–28. https://doi.org/10.4000/cy.1627. Charbonnier, J., 2015. « Groundwater management in Southeast Arabia from the Bronze age to the Iron age: a critical reassessment ». Water History 7 (1), 39–71. https://doi. org/10.1007/s12685-014-0110-x. Charbonnier, J., 2017. « The genesis of oases in Southeast Arabia: rethinking current theories and models ». In: Lavie, E., Marshall, A. (Eds.), Oases and Globalisation. Springer Geography. https://doi.org/10.1007/978-3-319-50749-1_4. Clapham, A.J., Rowley-Conwy, P.A., 2007. « New discoveries at Qasr Ibrim, lower Nubia ». In: Cappers, R.T.J. (Ed.), Fields of Change: Progress in African Archaeobotany. Groningen Institute of Archaeology, Groningen, pp. 157–164. Cleuziou, S., 1982. « Hili and the beginning of oasis life in eastern Arabia ». Proc. Seminar Arabian Stud. 12, 15–22. Cleuziou, S., Costantini, L., 1982. « A l’origine des oasis ». La Recherche 13 (137), 1179–1182. Costa, P., Wilkinson, T., 1987. « The hinterland of Sohar ». J. Oman Stud. 9, 102–138. Costantini, L., Costantini Biasini, L., 1986. « Laboratory of Bioarchaeology ». East W. 36 (4), 354–365. Dabrowski, V., 2019. « Systèmes d’approvisionnement et gestion des ressources végétales en Arabie orientale aux périodes antique et islamique (IVème s. av. J.-C. - XVIème s. ap. J.-C.): approches archéobotanique et archéoentomologique ». Doctorat. Muséum national d’Histoire naturelle, Paris, France. Dabrowski, V., M. Tengberg, A. Parker, and D. Kennet. forthcoming. « Agricultural economy, plant use and trade on the Sasanian and Islamic site of Kush (Ra’s alKhaimah’s Emirate, U.A.E) ». Veg. Hist. Archaeobotany. Dabrowski, V., Ros, J., Tengberg, M., Rougeulle, A., 2015. « De l’origine et de l’utilisation des ressources végétales en Oman médiéval : première étude archéobotanique à Qalhât ». Routes de l’Orient 2, 1–13. Daryaee, T., 2003. « The Persian Gulf trade in late antiquity ». J. World Hist. 14, 1–16. de Cardi, B., 1972. « A Sasanian outpost in northern Oman ». Antiquity 46 (184), 305–310. de Cardi, B., Vita-Finzi, C., Coles, A., 1975. « Archaeological survey in northern Oman, 1972 ». East W. 25 (1/2), 9–75. de Moulins, D., Phillips, C., Durrani, N., 2003. « The archaeobotanical record of Yemen and the question of Afro-Asian contacts ». In: Neumann, K., Butler, A., Kahlheber, S. (Eds.), Food, Fuels and Fields: Progress in African Archaeobotany. Heinrich BarthInstitut, Cologne, pp. 213–228. Deil, U., al-Gifri, A.-N., 1998. « Montane and wadi vegetation ». In: Ghazanfar, S.A., Fisher, M. (Eds.), Vegetation of the Arabian Peninsula, Geobotany, 25. Kluwer Academic, Dordrecht London, pp. 125–174. Esechie, H.A., 1994. « Interaction of salinity and temperature on the germination of sorghum ». J. Agron. Crop Sci. 172 (3), 194–199. Fagg, C.W., Stewart, J.L., 1994. « The value of Acacia and Prosopis in arid and semi-arid environments ». J. Arid Environ. 27 (1), 3–25. https://doi.org/10.1006/ jare.1994.1041. Fahn, A., Werker, E., 1986. In: Wood Anatomy and Identification of Trees and Shrubs from Israel and Adjacent Regions. Israel Academy of Sciences and Humanities, Jerusalem. Fleitmann, D., Mudelsee, M., Bradley, R.S., Pickering, P., Kramers, J., Burns, S.J., Mangini, A., Matter, A., 2009. Megadroughts at the dawn of Islam recorded in a 2600-year long stalagmite from Northern Oman. In: EGU General Assembly 2009. Presented at the EGU General Assembly Conference Abstracts, p. 8174. Fuller, D.Q., 2003. « African crops in prehistoric South Asia: a critical review ». Africa Praehistorica 15, 239–271. Fuller, D.Q., 2014. « Agricultural innovation and state collapse in meroitic Nubia ». In: Stevens, C., Nixon, S., Murray, M.A., Fuller, D.Q. (Eds.), Archaeology of African plant use, Publications of the Institute of Archaeology, University College London, 61. Left Coast Press, Walnut Creek (Calif.), pp. 165–178. Fuller, D.Q., 2017. « A Millet Atlas: Some Identification Guidance », third edition. Unpublished manuscript. Fuller, D.Q., Boivin, N., 2009. « Crops, cattle and commensals across the Indian ocean. Current and potential archaeobiological evidence ». Etudes Ocean Indien 42-43, 13–46. https://doi.org/10.4000/oceanindien.698. Fuller, D.Q., Stevens, C.J., 2018. « Sorghum domestication and diversification: a current archaeobotanical perspective ». In: Mercuri, A.M., D’Andrea, A.C., Fornaciari, R., Höhn, A. (Eds.), Plants and People in the African Past. Springer, pp. 427–452. https://doi.org/10.1007/978-3-319-89839-1_19. Gebauer, J., Luedeling, E., Hammer, K., Nagieb, M., Buerkert, A., 2007. « Mountain oases in northern Oman: an environment for evolution and in situ conservation of plant References Abdelrahman, H.A., Lepiece, A., Macalinga, V., 1993. « Some physical and chemical characteristics of the Batinah soils ». Commun. Soil Sci. Plant Anal. 24 (17–18), 2293–2305. https://doi.org/10.1080/00103629309368956. Abrams, M.J., Chadwick, O.H., 1994. « Tectonic and climatic implications of alluvial fan sequences along the Batinah coast, Oman ». J. Geol. Soc. 151 (1), 51–58. https://doi. org/10.1144/gsjgs.151.1.0051. al-Abdali, S., Al-Dhuhli, A., Al-Reasi, H., 2019. « Preliminary investigations of allelopathic effects and herbicide-based eradication of mesquite (prosopis juliflora) ». Sultan Qaboos Univ. J. Sci. 24 (1), 11–17. https://doi.org/10.24200/squjs. vol24iss1pp11-17. al-Jahwari, N.S., Kennet, D., Priestman, S., Sauer, E., 2018. « Fulayj: a late sasanian fort on the arabian coast ». Antiquity 92 (363), 724–741. https://doi.org/10.15184/ aqy.2018.64. al-Rawas, I., 2000. Oman in Early Islamic History. Ithaca Press, Reading. Barron, A., Fuller, D.Q., Stevens, C., Champion, L., Winchell, F., Denham, T., 2020. « Snapshots in time: MicroCT scanning of pottery sherds determines early domestication of sorghum (Sorghum bicolor) in East Africa ». J. Archaeol. Sci. 123, 105259. Battesti, V., 2005. Jardins au désert: Evolution des pratiques et savoirs oasiens - Jerid tunisien. A travers champs. I.R.D, Paris. Beech, M., Shepherd, E., 2001. « Archaeobotanical evidence for early date consumption on Dalma Island, United Arab Emirates ». Antiquity 75 (287), 83–89. https://doi. org/10.1017/S0003598X00052765. Beldados, A., 2019. « Millets in eastern Sudan: an archaeobotanical study ». Azania 54 (4), 501–515. https://doi.org/10.1080/0067270X.2019.1691844. Bellini, C., Condoluci, C., Giachi, G., Gonnelli, T., Lippi, M.M., 2011. « Interpretative scenarios emerging from plant micro- and macroremains in the Iron Age site of Salut, Sultanate of Oman ». J. Archaeol. Sci. 38 (10), 2775–2789. https://doi.org/10.1016/ j.jas.2011.06.021. Bellini, C., Pavan, A., Pignotti, L., Gonnelli, T., Lippi, M.M., 2020. « Food plants in pollen records from ancient southern Arabia: the evidences from sumhuram (southern Oman) ». J. Arid Environ. 177, 104131. https://doi.org/10.1016/j. jaridenv.2020.104131. Biagi, P., Nisbet, R., 1992. « Environmental history and plant exploitation at the aceramic sites of RH5 and RH6 near the mangrove swamp of Qurm (Muscat—Oman) ». Bulletin de la Société Botanique de France 139 (2–4), 571–578. https://doi.org/ 10.1080/01811789.1992.10827129. Biagi, P., Nisbet, R., 1999. « The shell-midden sites of RH-5 and RH-6 (Muscat, Sultanate of Oman) in their environmental setting ». Archaeologia Polona 37, 31–47. Boivin, N., Crowther, A., Prendergast, M., Fuller, D.Q., 2014. « Indian ocean food globalisation and Africa ». Afr. Archaeol. Rev. 31 (4), 547–581. https://doi.org/ 10.1007/s10437-014-9173-4. Bouchaud, C., Dabrowski, V., Tengberg, M., 2016. « Etat de la recherche archéobotanique en péninsule arabique ». Routes de l’Orient H-S 2, 21–37. Bouchaud, C., Thomas, R., Tengberg, M., 2012. « Optimal use of the date palm tree (Phoenix dactylifera L.) during Antiquity : Anatomical identification of plant remains from Madâ’inSâlih (Saudi Arabia) ». In: Badal, E., Carrion, Y., Macias, M., Ntinou, M. (Eds.), Wood and charcoal. Evidence for human and natural history. Proceedings of the 5th International Meeting of Charcoal Analysis, Saguntum, extra 13. Universitat de Valencia, Valencia, pp. 173–186. Brinkmann, K., Patzelt, A., Dickhoefer, U., Schlecht, E., Buerkert, A., 2009. « Vegetation patterns and diversity along an altitudinal and a grazing gradient in the Jabal al Akhdar mountain range of northern Oman ». J. Arid Environ. 73 (11), 1035–1045. https://doi.org/10.1016/j.jaridenv.2009.05.002. Bronk Ramsey, C., 2009. « Bayesian analysis of radiocarbon dates ». Radiocarbon 51 (1), 337–360. Cappers, R.T.J., Bekker, R.M., Jans, J.E.A., 2006. Digitale zadenatlas Van Nederland. Groningen Archaeological Studies, 4. Barkhuis Publishing; Groningen University Library, Groningen. Cappers, R.T.J., Neef, R., Bekker, R.M., 2009. Digital Atlas of Economic Plants. Groningen Archaeological Studies, 9. Barkhuis and Groningen University Library, Groningen. 16 V. Dabrowski et al. 1 Journal of Arid Environments 190 (2021) 104512 genetic resources ». Genet. Resour. Crop Evol. 54 (3), 465–481. https://doi.org/ 10.1007/s10722-006-9205-2. Ghazanfar, S.A., 1991. « Vegetation structure and phytogeography of Jabal shams, an arid mountain in Oman ». J. Biogeogr. 18 (3), 299–309. https://doi.org/10.2307/ 2845400. Ghazanfar, S.A., 1992. « Quantitative and Biogeographic analysis of the flora of the sultanate of Oman ». Global Ecol. Biogeogr. Lett. 2 (6), 189–195. https://doi.org/ 10.2307/2997660. Ghazanfar, S.A., 1996. « Invasive prosopis in Sultanate of Oman ». Aliens 3, 10. Ghazanfar, S.A., 1998a. « Status of the flora and plant conservation in the Sultanate of Oman ». Biol. Conserv. 85 (3), 287–295. https://doi.org/10.1016/S0006-3207(97) 00162-6. Ghazanfar, S.A., 1998b. Vegetation of the plains. In: Ghazanfar, S.A., Fisher, M. (Eds.), Vegetation of the Arabian Peninsula Geobotany, 25. Springer Netherlands, pp. 175–190. https://doi.org/10.1007/978-94-017-3637-4_7. Ghazanfar, S.A., 2003. In: Flora of the Sultanate of Oman. Vol. 1: Piperaceae Primulaceae.. Scripta Botanica Belgica, 25. National Botanic Garden, Meise. Ghazanfar, S.A., 2007. Flora of the Sultanate of Oman. Vol. 2: Crassulaceae-Apiaceae. Scripta Botanica Belgica, p. 36. Guarino, L., 1990. « Crop collecting in the Sultanate of Oman in the context of the Arabian Peninsula ». FAO/IBPGRI Plant Genetic Res. Newslett. 77, 27–33. Jacomet, S., 2006. « Identification of Cereal Remains from Archaeological Sites ». Unpublished manuscript. Kennet, D., 1997. « Kush: a sasanian and Islamic-period archaeological tell in Ras alKhaimah (UAE) ». Arabian Archaeol. Epigr. 8 (2), 284–302. Kennet, D., 1998. « Evidence for 4th/5th-century sasanian occupation at Khatt, Ras alKhaimah ». In: Phillips, C.S., Potts, D.T., Searight, S. (Eds.), Arabia and its neighbours. Essays on prehistorical and historical developments presented in honour of Beatrice de Cardi. Brepols, Turnhout, pp. 105–116. Kennet, D., 2004. Sasanian and Islamic Pottery from Ras Al-Khaimah (eBook Version): Classification, Chronology and Analysis of Trade in the Western Indian Ocean.. British Archaeological Reports, 1248. Archaeopress, Oxford. Kennet, D., 2007. The decline of eastern Arabia in the Sasanian period. Arabian Archaeology and Epigraphy 18 (1), 86–122. https://doi.org/10.1111/j.16000471.2007.00274.x. Kennet, D, 2009. « Transformations in late sasanian and early Islamic eastern Arabia: the evidence from Kush ». In: Schiettecatte, J., Robin, C. (Eds.), L’Arabie à la veille de l’Islam. Bilan Clinique, Orient & Méditerranée, 3. De Boccard, Paris, pp. 135–161. Kennet, D., 2012. « Archaeological history of the northern emirates in the Islamic period: an outline ». In: Potts, D.T., Hellyer, Peter (Eds.), Fifty Years of Emirates Archaeology. Motivate Publishing, Abu Dhabi, Dubaï, London, pp. 188–202. Kennet, D., Deadman, W.M., Al-Jahwari, N.S., 2016. « The Rustaq-Batinah archaeological survey ». Proc. Seminar Arabian Stud. 46, 155–168. Kwarteng, A.Y., Dorvlo, A.S., Kumar, G.T.V., 2009. « Analysis of a 27-year rainfall data (1977–2003) in the Sultanate of Oman ». Int. J. Climatol. 29 (4), 605–617. https:// doi.org/10.1002/joc.1727. Leakey, R.R.B., Last, F.T., 1980. « Biology and potential of prosopis species in arid environments, with particular reference to P. Cineraria ». J. Arid Environ. 3 (1), 9–24. https://doi.org/10.1016/S0140-1963(18)31672-0. McCorriston, J., Johnson, Z., 1998. « Agriculture and animal husbandry at Ziyadid Zabid, Yemen ». Proc. Seminar Arabian Stud. 28, 175–188. Messager, E., Badou, A., Fröhlich, F., Deniaux, B., Lordkipanidze, D., Voinchet, P., 2010. « Fruit and seed biomineralization and its effect on preservation ». Archaeol. Anthropol. Sci. 2 (1), 25–34. https://doi.org/10.1007/s12520-010-0024-1. Moez, D., 2007. « Entre foi et compromis tribal: L’entrée de la région d’Oman dans l’islam ». Arabian Humanities. Int. J. Archaeol. Soc. Sci. Arab. Peninsula 14, 36–62. https://doi.org/10.4000/cy.1453. Morony, M.G., 2001-02. « The Late Sasanian economic impact on the Arabian peninsula ». Nāme-ye Irān-e Bāstān (International Journal of Ancient Iranian Studies) 1 (2), 25–37. Munier, P., 1973. Le Palmier-Dattier. G.-P. Maisonneuve et Larose, Paris. Munt, H., 2017. « Oman and late Sasanian imperialism ». Arabian Archaeol. Epigr. 28 (2), 264–284. https://doi.org/10.1111/aae.12102. Neumann, K., Schoch, W., Détienne, P., Schweingruber, F.H., 2001. Woods of the Sahara and the Sahel : an anatomical atlas. Haupt, Bern. Pajouh, D.P., Schweingruber, F.H., 1993. Atlas des bois du nord de l’Iran. Publications de l’Université de Téhéran, Téhéran. Parker, A.G., Goudie, A.S., 2008. « Geomorphological and palaeoenvironmental investigations in the southeastern Arabian Gulf region and the implication for the archaeology of the region ». Geomorphol. Geoarchaeol. Geomorphol.: Soils, Sediments, and Societies 101 (3), 458–470. https://doi.org/10.1016/j.geomorph.2007.04.028. Potts, D.T., 1990. The Arabian Gulf in Antiquity, vol. 2. Clarendon Press, Oxford. Priestman, S.M.N., 2019. « The archaeology of early Islam in Oman: recent discoveries from Fulayj on the Batinah ». Anglo-Omani Soc. Rev. 40–43. Reimer, P.J., En Austin, W., Bard, E., Bayliss, A., Blackwell, P.G., Bronk Ramsey, C., Butzin, M., Cheng, H., Lawrence Edwards, R., Friedrich, M., et al., 2020. « The IntCal20 northern hemisphere radiocarbon age calibration curve (0–55 cal kBP) ». Radiocarbon 62 (4), 725–757. Sanlaville, P., 2000. Le Moyen-Orient arabe: le milieu et l’homme. Collection U. Armand Colin, Paris. Schweingruber, F.H., 1990. Anatomy of European Woods. Haupt, Bern. Simpson, St J., 2019. « Nomads and monks, soldiers and sailors, farmers and fishermen: new archaeological insights into life in the Persian Gulf from late antiquity to the medieval period ». In: Zolotova, A.A. (Ed.), Ex Oriente Lux. Collected Papers to Mark the 75th Anniversary of Mikhail Borisovich Piotrovsky. State hermitage Museum, StPetersburg, pp. 288–335. Smith, O., Nicholson, W.V., Kistler, L., Mace, E., Clapham, A., Rose, P., Stevens, C., Ware, R., Samavedam, S., Barker, G., 2019. « A domestication history of dynamic adaptation and genomic deterioration in sorghum ». Nat. Plant. 5 (4), 369. Snape Kennedy, L., 2018. Wind, Water and Walls: Developing Geoarchaeological and Luminescence Dating Techniques for Archaeological Landscape Features. Unpublished PhD Thesis. Department of Archaeology, Durham University, Durham. Stenhouse, J., Tippayaruk, J., 1996. « Sorghum bicolor (L.) Moench ». In: Grubben, G., Partohardjono, S. (Eds.), Plant Resources of South-East Asia N◦ 10: Cereals. Prosea Foundation, Bogor, pp. 130–136. Ṭabarı̄, Muḥammad ibn Ǧarı̄r ibn Yazı̄d al, 1999. In: Bosworth, C.E. (Ed.), The History of al-Ṭabarı̄. State University of New York, New York. Tengberg, M., 1998. « Paléoenvironnements et économie végétale en milieu aride et semiaride: recherches archéobotaniques dans la région du Golfe arabo-persique et dans le Makran pakistanais (4ème millénaire av. notre ère e 1er millénaire de notre ère) ». Université Montpellier 2. Tengberg, M., 2005. « Les forêts de la mer. Exploitation et évolution des mangroves en Arabie orientale du Néolithique à l’époque islamique ». Paleorient 31 (1), 39–45. Tengberg, M., 2012. « Beginnings and early history of date palm garden cultivation in the Middle East ». J. Arid Environ. Ancient Agric. Middle East 86, 139–147. https://doi. org/10.1016/j.jaridenv.2011.11.022. Tengberg, M., V. Dabrowski, and D. Kennet. Forthcoming. « Vegetation history and wood exploitation at Kush (Ras al-Khaimah, UAE), 4th-17th/18th centuries AD. First results of the charcoal analysis ». Arabian Archaeol. Epigr.. Ulbaydli, A., 1993. « The agrarian economy of Oman (132–280/749–893) in Arabic sources ». J. Islam. Stud. 4 (1), 33–51. Ulrich, B., 2011. « Oman and Bahrain in late antiquity: the Sasanians’ Arabian periphery ». Proc. Seminar Arabian Stud. 41, 377–385. Van Bergen Poole, I., 2015. « Charcoalified wood analysis ». In: Report on the Archaeological Investigation of Fulayj Fort Near Saham on the Batinah Plain of Oman. Ministry of Heritage and Culture, Mascat, pp. 61–65. Van der Veen, M., Lawrence, T., 1991. « The plant remains ». In: Welsby, D.A., Daniels, C. M. (Eds.), Sobra: Archaeological Research at a Medieval Capital on the Blue Nile. British Institute in Eastern Africa Memoirs, London, pp. 264–274. Varisco, D.M., 1991. « A royal crop Register from Rasulid Yemen ». J. Econ. Soc. Hist. Orient 34 (1/2), 1–22. https://doi.org/10.2307/3632276. Varisco, D.M., Ibn Yūsuf, ʿD., 1994. Medieval Agriculture and Islamic Science the Almanac of a Yemeni Sultan. Publications on the Near East, University of Washington, vol. 6. University of Washington Press, Seattle. Walshaw, S.C., 2010. « Converting to rice: urbanization, Islamization and crops on Pemba Island, Tanzania, ad 700–1500 ». World Archaeol. 42 (1), 137–154. https:// doi.org/10.1080/00438240903430399. Wiersema, J.H., Dahlberg, J., 2007. « The nomenclature of Sorghum bicolor (L.) Moench (Gramineae) ». Taxon 56 (3), 941–946. Wilkinson, J.C., 1977. Water and Tribal Settlement in South-East Arabia: a Study of the Aflāj of Oman. Clarendon Press, Oxford. Wilkinson, J.C., 1979. « Ṣuḥār (Sohar) in the early Islamic period: the written evidence ». In: Taddei, M. (Ed.), South Asian Archaeology 1977. Seminario di Studi Asiatici. Naples: Instituto Universitario Orientale, pp. 887–907. Wilkinson, J.C., 2010. Ibâdism: Origins and Early Development in Oman. Oxford University Press, Oxford. Willcox, G., 1995. « Some plant impressions from Umm an-Nar island ». In: Frifelt, K. (Ed.), The Island of Umm An-Nar, vol. 2. Aarhus University Press, Aarhus, pp. 257–259. Willcox, G., Tengberg, M., 1995. « Preliminary report on the archaeobotanical investigations at tell abraq with sepcial attention to chaff impressions in mud-Brick ». Arabian Archaeol. Epigr. 6 (3), 129–138. https://doi.org/10.1111/ aae.1995.6.2.129. Winchell, F., Stevens, C.J., Murphy, C., Champion, L., Fuller, D.Q., 2017. « Evidence for sorghum domestication in fourth millennium BC eastern Sudan: spikelet morphology from ceramic impressions of the Butana group ». Curr. Anthropol. 58 (5), 673–683. https://doi.org/10.1086/693898. Winchell, F., Brass, M., Manzo, A., Beldados, A., Perna, V., Murphy, C., Stevens, C., Fuller, D.Q., 2018. « On the origins and dissemination of domesticated sorghum and pearl millet across Africa and into India: a view from the Butana group of the far eastern sahel ». Afr. Archaeol. Rev. 35 (4), 483–505. https://doi.org/10.1007/ s10437-018-9314-2. 17