Veget Hist Archaeobot DOI 10.1007/s00334-013-0394-2 ORIGINAL ARTICLE Climate, human palaeoecology and the use of fuel in Wadi Sana, Southern Yemen Masoumeh Kimiaie • Joy McCorriston Received: 30 July 2012 / Accepted: 5 February 2013 Ó Springer-Verlag Berlin Heidelberg 2013 Abstract This study integrates analysis of wood charcoal assemblages with climate proxies, palaeoenvironmental and archaeological data sets in hyper-arid Wadi Sana, Yemen, to address the availability and use of wood fuels by South Arabian hunter-herder groups from the Early Holocene (8000–7700 cal. B.P.) to Middle Holocene (6000–4800 cal. B.P.) periods. The Early Holocene environment, regulated by a stronger Southwest Asian monsoon, was moister than the present, providing a marshy winter grazing area for cattle herders, whose construction of hearths and food preparation strategies changed over time. This study provides an insight into long term stability of land cover and use as well as the dynamics of human contributions to landscape change. We suggest that complex environmental and cultural processes affect species availability, fuel choice and land use management. Despite environmental and economic changes in Wadi Sana, our dataset does not show changes in fuel choice from the Early to Middle Holocene. Keywords Holocene
Palaeoecology
Wadi Sana
Charcoal analysis
Fuel choice
Human impact Communicated by G. Willcox. M. Kimiaie (&)
J. McCorriston Department of Anthropology, The Ohio State University, 4034 Smith Laboratory, 174 W. 18th Ave., Columbus, OH 43210, USA e-mail: kimiaie.1@osu.edu Introduction During the Holocene period, the local environments of South Arabia were regulated by a stronger Southwest Asian monsoon than today and were generally moister than the present (Davies 2006; Fleitmann et al. 2007; Parker and Goudie 2008). While the climate history and regional moisture balances of the Holocene are broadly understood across Arabia (Lézine et al. 2010), local ecosystem histories and especially human-environment interactions are poorly documented, and the dynamic of human contributions to landscape change is poorly understood. This study explores a record of archaeological wood charcoals once used as fuel to assess whether economic and landscape management changes in the Early-Middle Holocene correspond to changes in fuel choices and ultimately in woody vegetation cover. Fuel use is a major source of charred macro-botanical remains. Where populations lived in permanent settlements, such remains may accumulate in site middens as the result of numerous hearth and cooking episodes (Miller 1984; McCorriston 1998), but in the Wadi Sana, no settlement occurred prior to about the 3rd century A.D. (McCorriston et al. 2005; McCorriston 2006). Archaeological charcoal came from numerous, widely distributed camp hearths, and each archaeological charcoal assemblage is the product of the last use episode of a hearth. These hearths may be safely associated with mobile peoples and show a distinct shift in construction techniques over time, possibly corresponding to changing cooking practices. Furthermore, these assemblages may be dated, either with radiometric dates on charcoal (14C) or in stratigraphic context with other absolutely and relatively dated features. The composition of fuels in hearths provides a sequence of cooking events that contribute overall 123 Veget Hist Archaeobot to our understanding of human use of natural resources throughout the Early Holocene wet period and across important changes in human economy. Although our sample size is small and might not be suitable for palaeovegetation reconstruction (Chabal 1997), it can be a good indicator of human management of landscapes and economical patterns. Palaeoclimate and human adaptations in Wadi Sana Researchers in Africa have argued that Early Holocene cattle-centred mobility may have retarded people’s commitment to agriculture and sedentism (e.g. Marshall and Hildebrand 2002; McIntosh 1994; Wengrow 2006; Haaland 1992), a cultural framework also plausible for Southern Arabia, where cattle herding preceded agriculture by at least 3,000 years (McCorriston 2006; McCorriston et al. 2012). Field research and analysis over a decade in the highland arid zone of Southern Arabia (Hadramawt, Southern Yemen) by the RASA Project (Roots of Agriculture in Southern Arabia) has yielded integrated palaeoecological and archaeological data that allows us assess early landscape management and environmental impacts by humans and their animals. The Wadi Sana which drains northward through the Southern Arabian highlands, today flows only episodically with discharge from summer rainfall (Fig. 1). The modern channel is deeply incised through well-dated silt terraces laid down during an Early Holocene phase of aggradation when summer flooding created wetlands suitable for winter pasture (Harrower et al. 2012). Well-documented, largeregion environmental proxies like pollen, stable isotopes and dune formation (e.g. Lézine et al. 2010; Parker and Goudie 2008), indicate that major climate change—the weakening of the southwest monsoon—would have aridified the Middle-Late Holocene Wadi Sana environment. Wadi Sana palaeoenvironmental responses to this climate change have been inferred from geomorphologic sequences (Harrower et al. 2012; McCorriston et al. 2012) and hyrax middens (McCorriston et al. 2002) that suggest that the uplands have had no soil (and therefore almost no cover) for all of the Holocene. Sparse parkland with xerophytic trees such as acacias and myrrh clung to the slopes, and most vegetation, including grasses, rushes, and shrubs was concentrated along channels and the seeps from springs. GIS-calculations suggest that pasture could support up to 60 herders (about 7–8 households) dispersed along the Wadi Sana marshlands during the period 8000–4800 B.P. Regional surveying and excavations have provided an Early-Middle Holocene cultural history (Crassard et al. 2006; McCorriston et al. 2002, 2005). The earliest Holocene dates are from the seasonal, re-visited campsite of Manayzah, adjacent to a natural spring and pool. Broad exposures 123 Fig. 1 Map of the modern Arabian Peninsula. The oval on the far right shows the location of Wadi Sana revealed multiple contemporaneous hearths, pits, post-holes and stone tools (Crassard 2008), and suggested a mixed hunting-herding economy of ungulates, sheep, goats and the earliest domesticated cattle recorded in Arabia (McCorriston and Martin 2009; Martin et al. 2009). Several hearths included in this study come from the sequence at Manayzah dating back to 8000–7700 cal. B.P. The seasonally visited site of Shi’bKheshiya, some 1,500 years later, is significant in Arabian prehistory for the preservation of an adjacent ring of cattle skulls representing the ritual aftermath of a sacrifice, gathering and feast, which united pastoral groups in supra-household alliances across broad territories (McCorriston et al. 2012). Although this site lacks much economic data, it is significant in its cultural and social constitution (McCorriston et al. 2011), and is consistent with a wider pattern of cattlehuman symbolism across the Old World tropics (Wengrow 2006; Di Lernia 2006; Boivin 2004). Several of the surrounding hearths were sampled for wood charcoal. Regional surveying has detected a wide array of stratified hearths with preserved charcoal. People remained mobile in small groups, even as they adopted new vegetation management. New practices included anthropogenic firing of local vegetation (McCorriston et al. 2005) and subsequent smallscale water management channels and check-dams (Harrower 2008). Over time, there was a shift in the manner in which people constructed hearths. In the Early Holocene hearths were built into a steep pit lined with small slabs of limestone (Fig. 2). Cooking relied on the even heat released from cooking stones, smooth limestone cobbles brought from the active Wadi channel. In modern Bedouin practice, a hot fire is built atop the hearth then cinders and charcoal are stirred into the cobbles to ensure a long-lasting heat. The top of such a hearth can be blown free of ash and cinders to provide a searing surface for grilling meat, while the oxygen-poor interior may be used for cooking tubers or dough without fear Veget Hist Archaeobot Fig. 2 Illustrating three different Holocene hearth types in Wadi Sana: Hearth type I (lower), hearth type II (middle), and hearth type III (upper). The analyzed samples in this paper only come from hearth types I and II (drawn by Jennifer Everhart) of combustion. Such a hearth also produces hot stones used to boil milk or liquids in perishable leather-lined baskets. By the Middle Holocene, hearth construction practices had changed; archaeologists find numerous hearths directly or stratigraphically dated to 6800–4700 B.P. built as a shallow scoop filled with cooking cobbles (Fig. 2). Some hearths contain charred animal bone fragments, a few of which have been definitively assigned to cattle (McCorriston et al. 2002). Finally, people began constructing hearths on surfaces by erecting a ring of small boulders to retain hot cooking stones and charcoal. This style continues to be built and used today, and surface examples contain charcoal that may be very recent (Fig. 2). Although ethno-archaeological observations show that modern Bedouin bury dough to cook (gritty) bread in ashes, and also make charcoal in buried pits containing no stone, there is no archaeological evidence for use of multiple types at any one time in the past. Hearth types were sequential, and apparently replaced earlier types throughout the Holocene (Table 1). Data and methods The data generated for this paper consists of 20 charred wood samples from hearth types I and II in Wadi Sana excavated from 1998 to 2005. Archaeologists recovered charcoal pieces greater than 5 mm length to reach a minimum sample size of 20-plus intact specimens. (It was impractical to recover every [5 mm-sized piece). Where charcoal crumbled, as is normal in hyper-arid conditions, a few fragments were selected and an aluminium foil wrap ensured that these could travel together to minimize multiple identifications on one plant (pre-excavation fragmentation could not be addressed in this study). These samples were sorted by 109 magnification stereomicroscope. Within each sample (one per hearth) twenty pieces of charred wood greater than 2 mm in size were chosen for identification, where available. Flotation samples, which were collected to look for seeds, also yielded some identifiable charcoal fragments. In total, 331 wood fragments were identified. In small assemblages (less than 20 fragments within the sample) every attempt was made to identify as many wood fragments as possible, regardless of fragment size. While it would be desirable to examine much larger assemblages, preservation, small deposits (hearths) and field collection constraints (water for limitedsize flotation samples, fragility of charcoal that pulverizes in sieving in arid environments) meant that samples were small. The procedure for identifying charred wood involved matching microscopic cellular patterns with modern reference material, charred and mounted on glass slides in three fracture planes: tangential, radial and transverse. These different sections were identified under a Leica compound forensic microscope with 50–1,0009 magnification and a dual stage with bridge prisms for simultaneous viewing of archaeological and reference specimen. All wood taxa are identified to the genus level. The key characteristics and terminology used for identification of these taxa are drawn from wood anatomy reference manuals and keys (Asouti and Fuller 2008; Fahn et al. 1986; Jagiella and Kürschner 1987). The modern reference collection is housed in McCorriston’s Archaeobotany Laboratory at The Ohio State University. To explore quantitative differences across sets of samples, we generated a Shannon Index of Diversity for each sample. Next we performed a two-tailed Monte Carlo test to assess the statistical significance of differences in taxa representation in the two groups of hearths. A Monte Carlo randomization resampling test can be used in conjunction with any measure of population based on observed species abundances and runs for up to 1,000 iterations, which makes this test a robust procedure with a high degree of confidence regardless of the size of the population (Zhu 2005). When the assumption of a normal distribution is not satisfied or the sample size is too small, a Student T test is not valid. In this case, a two-tailed test should be used to test the hypothesis that both diversity means were generated from sampling the same population. A two-tailed test allocates the 5 % chance of an error under the null 123 123 Table 1 Details of all samples presented in this paper including hearth type, context code, date, absolute counts of identified taxa, Shannon Diversity Index, and percentage of determined fragments Hearth type Context 2 SU110-6-3 4800 22 2 SU37-3-C-012 BAG 5 6500 20 2 SU37-3-0-D-002 BAG 1&2 6500 2 SU 37-3-C-14 BAG 6 6500 2 SU 037-3-C-12 6500 20 2 WS 1 Hearth 001 FS1112 6610 6 2 SU154-1-2 6610 20 2 WS Hearth 8 FS1103 6610 19 18 2 WS Hearth 9 FS1104 6610 14 12 2 WS 1 Hearth 002 FS1111 6692* 20 2 WS Hearth VI FS1102 6700 19 19 2 2 SU80-4 Hearth 2-2 SU151-1-H3-1 6800 6861* 13 10 2 2 WS Hearth 14 FS1107 6933* 8 2 WS Hearth 16 FS1101 6968* 5 1 SU155-2 K9 Hearth1 FS1413 7742* 20 1 SU155-2-M11 QuadB-007 FS1399 7755 17 3 1 SU155-2-L11 QuadB-005 FS1404 7755 27 5 1 SU155-2 K9 Niv14 FS1416 7829 20 1 SU155-2-K9-18 8044 11 1 331 59 Total (n) (%) Date (ca.l B.P.) Number of ident. fragm. Acacia Anogeissus Cadaba 1 Ficus 13 Tamarix Ziziphus 8 Shannon div. index 0.82 19 1 0.20 20 8 12 0.67 20 20 5 1.04 0 0.47 5 1.03 20 5 1 0 10 0.45 5 2 20 13 3 8 2 3 0.67 20 4 10 0.96 1 21 0.63 20 18.0 10 30 9.2 33 10.1 145 44.3 0.30 16 4.9 44 13.5 Veget Hist Archaeobot All dates calibrated using online Calpal (indicated by *). Other dates have been interpolated from stratigraphy and stratified dates on other features that could be stratigraphically related to the hearth Veget Hist Archaeobot Table 2 List of modern Wadi Sana vegetation (left) and analyzed archaeobotanical taxa (right) Modern woody taxa Archaeobot. taxa Acacia ramulosa W. Fitzg (Pedley) Acacia sp. Acacia odorata Desv. Acacia sp. Acacia mellifera (M. Vahl.) Benth. Anogeissus bentii Baker. Anogeissus sp. Boswellia sacra Flueck. Commiphora kataf Engl. Cadaba heterotricha Stocks. Cadaba sp. Calotropis procera (Aiton) W.T. Aiton Commiphora gileadensis (L.) C. Chr Cyphostemma crinitum (Planch.) Desc. Delonix elata Gamble. Ficus salicifolia Vahl. Ficus sp. Indigofera s.l. Lycium shawii Roem. & Schult. Maerua crassifolia Forssk. Moringa peregrina Fiori. Tamarix L. Tamarix sp. Ziziphus leucodermis (Baker) O. Schwartz Ziziphus sp. hypothesis equally to both tails so each tail has the probability of 0.025, and the null hypothesis is rejected if the test statistic is too extreme in either direction. In other words, the two groups of hearths would be significantly different if the p value was below the 2.5th percentile or above the 97.5th percentile of the distribution. Fig. 3 Distribution of taxa (percentages of species by count) in hearth type I, 8000–7700 B.P. (top) and hearth type II, 6900–4700 B.P. (bottom) Results A Monte Carlo test with 1,000 iteration runs testing the null hypothesis that the means of Shannon Diversity Index values of the two groups of hearths are drawn from the same population resulted in a p value of 0.76. This suggests no significant change in the composition of wood fuels preserved as charcoals in hearths of type I and type II. The composition of fuel in these hearths is dominated by a few woody species still present in the region today (Tables 1, 2). Modern taxa distribution, including a single Ficus salicifolia tree in Wadi Sana, is known from the collections and identifications by McCorriston and Ms. Catherine Heyne in conjunction with widespread archaeological fieldwork. Inferences about regional palaeovegetation come from the Middle Wadi Sana work on hyrax middens (Cole in McCorriston et al. 2002) and the palaeoenvironmental reconstructions from sediment sequences and modern analog plant communities (McCorriston et al. 2012; McCorriston and Martin 2009). Figure 3 shows the percentage of charred wood recorded by counts from different hearths in Wadi Sana. The identified charred woods include Acacia sp., Anogeissus sp., Cadaba sp., Ficus sp., Tamarix sp. and Ziziphus sp. All of these are the result of fuel use during the dry season of Wadi Sana when there was human aggregation in the marshy oxbows as the grassy flushes rimming the upland vernal pools were depleted (McCorriston et al. 2012). All of the species found in archaeological hearth structures are present in modern day Wadi Sana (Table 2). Acacia sp. This wood has homogeneous uni- to multi-seriate rays, procumbent cells, simple perforation plates, and storied paratracheal parenchyma. Its foliage is used today as good source of fodder, especially for confined goats and is well known among herders as a livestock fatness enhancer and milk enricher (Ghazanfar 1994; Miller and Morris 1988). Acacia is an excellent fuel providing good charcoal (Miller and Morris 1988), and preferentially collected in the recent past for charcoal production destined for urban consumption. 123 Veget Hist Archaeobot Anogeissus sp. Diffuse-porous wood and heterogeneous rays with upright and tile cells, radially arranged pores, and paratracheal parenchyma characterize this wood (Asouti and Fuller 2008; Fahn et al. 1986). While the foliage is never browsed by cattle, the wood provides an excellent fuel and yields good charcoal (Arbonnier 2004). Modern residents of Wadi Sana pollard the trees to provide fodder for camels and goats, which also browse the trees in modern Dhufar. Cadaba sp. This wood has diffuse-porous vessels, heterogeneous rays with procumbent and upright cells and paratracheal parenchyma (Fahn et al. 1986). Ethnographic and ethnobotanical research show that Cadaba foliage is fed to herds when there is no graze, and its firewood produces an unpleasant flavor. The ash from Cadaba mixed with butter is used as a remedy on livestock skin where hair has fallen out. It is also important as treatment of parasitic skin infestation in goats and camels in modern day Oman (Ghazanfar 1994; Miller and Morris 1988). Since the shrub is a common woody plant of the arid rain shadow of the Southern Arabian highlands and Wadi Sana today in particular, its presence in ancient hearths may reflect its common occurrence. Ficus sp. This taxon has diffuse-porous vessels with wide parenchyma bands, and bi- or multi-seriate heterogeneous rays (Fahn et al. 1986). Ficus produces much smoke, so local people of the Southern Arabian highlands find it a desirable quality wood for fires lit for livestock—especially those inside cattle byres, where such fires have a deterrent effect on pests and insects (Ghazanfar 1994; Janzen 1986; Miller and Morris 1988). The high percentage of Ficus in the samples suggests that early inhabitants of Wadi Sana predominantly used a fuel wood that was not a provider of browse. Ficus trees are present in modern day Wadi Sana but with a different distribution than suggested by the representation of Ficus wood in Early-Middle Holocene hearths across a 40 km region. For example, F. salicifolia Vahl is still present today but very sparse and limited to a small number of plants in the entire Wadi channel. There is a strong possibility that the identified charred Ficus recovered from Holocene samples is F. salicifolia, but other potential Ficus species would include F. vasta or F. sycomorus L. which are today limited to Dhufar in the Southern Arabian highlands. Tamarix sp. The most important characteristic of this taxon is diffuseporous vessels with simple perforation plates, vasicentric 123 parenchyma storied with vessels, and heterogeneous multiseriate rays (Fahn et al. 1986). The ash from Tamarix is effective in treating animals and it was used widely in the Arabian Peninsula (Ghazanfar 1994; Miller and Morris 1988). It is an extremely hard wood, resistant to termites, and a desirable firewood with long, slow-burning properties. Ziziphus sp. This taxon is characterized by a diffuse-porous wood, simple perforation plates, scanty to vasicentric parenchyma, radially-arranged pores, and mostly uniseriate narrow heterogeneous rays (Fahn et al. 1986). Ziziphus foliage is browsed by herds especially during the dry season; moreover it bears an edible fruit (Miller and Morris 1988) and its hard wood is a popular fuel. Charred Ziziphus seeds are also occasionally recovered from hearths in Wadi Sana (perhaps discarded after eating or possibly introduced with branch wood) and are one of the few edible archaeobotanical taxa found commonly throughout Arabia. Discussion Wood fuel remains, as indirect evidence of ancient land use and vegetation, help us understand not only ancient environmental conditions, but also the behavioral patterns that impose a selective filter on the preservation of charred remains. This study provides an insight into long term stability of land cover and use as well as human interactions with the surrounding landscape. We suggest that understanding the archaeological context of archaeobotanical materials helps to reconstruct pastoral land use, and secondly we suggest an approach to understand the weight of environmental and cultural variables in fuel and land use management. Wood charcoal data reported here are best understood within the context of Wadi Sana palaeoenvironmental data constructed from sediment sequences, hyrax middens, and modern vegetation analogs across the Southern Arabian highlands. These data provide insight into the availability of fuels preferred by humans. In Wadi Sana aridification has changed the density of cover, but not the woody taxa represented. In wood charcoal data, this is reflected in Ficus sp., which has high ubiquity across assemblages, suggesting that this taxon was once readily obtainable and not limited to the single plant growing as seen today. There is also good evidence for landscape manipulation by humans. From about 8000 cal. B.P. archaeologists have dated regional brush fires that were human-set (McCorriston et al. 2002, 2005). While it is unknown whether these were set for game drives or to increase grazing and browsing resources, this strategy represents a modest labour investment by humans Veget Hist Archaeobot in landscape change. In the archaeological sequence, burning precedes the onset of ritual sacrifices, linked to greater territorial defence of limited resources by mobile pastoralists (McCorriston et al. 2012). Finally, there appear water management channels and check-dams that represent an intensification of labour input to resource production and landscape manipulation (Harrower 2008). Yet the data available from a sample of hearths suggest that Early-Middle Holocene landscape manipulation thus practiced from 8000 to 4700 cal. B.P. did not impact the availability of preferred fuel to a point where people were forced to shift their acquisition of it. From the identified woods we see no change, despite change in construction technique and possible changing use of hearths. Although the growing population was using new practices such as burning and territorial defence marked by communal sacrifices, the human population did not over time exceed the capacity of Wadi Sana to provide the preferred fuels. In an arid landscape occupied only by mobile herders, the amount of wood used as fuel depends on the following factors: the quality of the species as a fuel source, the population of people using wood, and the demand of herds for browsing. The low percentage of some taxa recovered from the hearth structures may be interpreted as economic choice of those species which do not provide good charcoal but instead offer a very good quality browse. Based on our wood analysis and statistical results we argue that economic changes, especially the growing population pressure documented in Wadi Sana through archaeological and geomorphological data, did not entail changes in fuel choice from the Early to Middle Holocene. Not only did economic changes such as the emergence of territorial defence of resources (McCorriston and Martin 2009; McCorriston et al. 2012) not alter fuel choice and land use by Wadi Sana inhabitants, but the data suggest that changes in lithic technologies (Crassard 2008), pastoral camps (McCorriston et al. 2002), or even hearth types did not entail changes in plant resource availability and fuel use strategies employed by Early-Middle Holocene Wadi Sana pastoralists. References Arbonnier M (2004) Trees, shrubs and lianas of West African dry zones. CIRAD, Montpellier Asouti E, Fuller D (2008) Trees and woodlands of South India: archaeological perspectives. Left Coast Press, Walnut Creek Boivin N (2004) Landscape and cosmology in the South Indian Neolithic: new perspectives on the Deccan Ash Mounds. Camb Archaeol J 14:235–257 Chabal L (1997) Forêts et sociétés en Languedoc (Néolithique final, Antiquité tardive), L’anthracologie, méthode et paléoécologie. Documents d’Archéologie Française, Paris Crassard R (2008) La préhistoria du Yémen. BAR International Series. Archaeopress, Oxford Crassard R, McCorriston J, Oches E, Bin’Aqı̄l A, Espagne J, Sinnah M (2006) Manayzah, early to mid-holocene occupation in Wadi Sana (Hadramawt, Yemen). Proc Sem Arab Stud 36:151–173 Davies C (2006) Quaternary paleohydrology and past climates of the Dhamar Highlands, Yemen. Quat Res 66:454–464 Di Lernia S (2006) Building monuments, creating identity: cattle cult as social response to rapid environmental changes in the Holocene Sahara. Quat Int 151:50–62 Fahn A, Werker E, Bass P (1986) Wood anatomy and identification of trees and shrubs of Israel and adjacent regions. Israel Academy of Sciences and Humanities, Jerusalem Fleitmann D, Burns S, Mangini A, Mudelsee M, Kramers J, Villa I, Neff U, Al-Subbary A, Buettner A, Hippler D, Matter A (2007) Holocene ITCZ and Indian monsoon dynamics recorded in stalagmites from Oman and Yemen. Quat Sci Rev 26:170–188 Ghazanfar SA (1994) Handbook of Arabian medicinal plants. CRC Press, Inc., Boca Raton Haaland R (1992) Fish, pots, and grain: early and Middle Holocene adaptation in the Central Sudan. Afr Archaeol Rev 10:43–64 Harrower M (2008) Hydrology, Ideology, and the origins of irrigation in ancient pastoral/agro-pastoral landscapes along Wadi Sana (Yemen). J Arid Environ. doi:10.1016/j.jaridenv.2011.11.020 Harrower M, Oches E, McCorriston J (2012) Hydro-geospatial analysis of ancient pastoral/agro-pastoral landscapes along Wadi Sana (Yemen). J Arid Environ. doi:10.1016/j.jaridenv.2011.11.020 Jagiella C, Kürschner H (1987) Atlas der Hölzer Saudi Arabiens: die Holzanatomie der wichtigsten Bäume und Sträucher Arabiens mit einem holzanatomischen Bestimmungsschlüssel. Reichert, Wiesbaden Janzen J (1986) Nomads in the sultanate of Oman: tradition and development in Dhofar. Westview Press, Boulder Lézine A-M, Robert C, Cleuziou S, Inizan M-L, Braemer F, Saliége J-F, Sylvestre F, Tiercelin J-J, Crassard R, Méry S, Charpentier V, Steimer-Herbet T (2010) Climate change and human occupation in the Southern Arabian lowlands during the last deglaciation and the Holocene. Glob Planet Change 72:412–428 Marshall F, Hildebrand E (2002) Cattle before crops: the beginnings of food production in Africa. J World Prehist 16:99–143 Martin L, McCorriston J, Crassard R (2009) Early Arabian pastoralism at Manayzah in Wadi Sana, Hadramawt. Proc Sem Arab Stud 39:271–282 McCorriston J (1998) Landscape and human interaction in the Middle Habur drainage from the Neolithic period to the Bronze Age. In: Fortin M, Aurenche O (eds) Espace naturel, espace habité en Syrie du Nord (10e-2e millénaires av J-C.). Canad Soc Mesopotamian Stud Bull 33, Toronto, Lyon, p 43–54 McCorriston J (2006) Roots of agriculture (RASA) project 2005: a season of excavation and survey in Wadi Sana, Hadramawt. Bull Am Inst Yemeni Stud 47:23–28 McCorriston J, Martin L (2009) Chapter 17. Southern Arabia’s early pastoral population history: Some recent evidence. In: Petraglia MD, Rose JI (eds) The evolution of human populations in Arabia. Springer, New York, pp 237–250 McCorriston J, Oches E, Walter D, Cole K (2002) Holocene paleoecology and prehistory in highland Southern Arabia. Paléorient 28:61–88 McCorriston J, Harrower M, Oches E, Bin’Aqil A (2005) Foraging economies and population in the Middle Holocene highlands of Southern Yemen. Proc Sem Arab Stud 35:143–154 McCorriston J, Steimer-Herbet T, Harrower M, Williams K, Saliège J, Bin‘Aqı̄l A (2011) Gazetteer of small scale monuments in prehistoric Hadramawt, Yemen: a radiocarbon chronology from RASA project research 1996–2008. Arab Archaeol Epigr 22:1–22 123 Veget Hist Archaeobot McCorriston J, Harrower M, Martin L, Oches E (2012) Cattle cults of the Arabian Neolithic and early territorial societies. Am Anthropol 114:45–63 McIntosh SK (1994) Changing perceptions of West Africa’s past: archaeological research since 1988. J Archaeol Res 2:165–198 Miller NF (1984) The interpretation of some charred cereal remains as remnants of dung cake fuel. Bull Sumer Agric 1:45–47 Miller AG, Morris M (1988) Plants of Dhofar (The southern region of Oman: Traditional, economic and medicinal uses). The office of the adviser for conservation of the environment, Diwan of Royal Court, Sultanate of Oman 123 Parker AG, Goudie AS (2008) Geomorphological and palaeoenvironmental investigations in the Southeastern Arabian Gulf region and the implication for the archaeology of the region. Geomorphology 101:458–470 Wengrow D (2006) The archaeology of early Egypt: social transformations in North-East Africa, c. 10000 to 2650 B.C. University Press, Cambridge Zhu L-X (2005) Nonparametric Monte Carlo tests and their applications. Springer, New York