87SR/86SR Analysis As a Method to Explore Human Ecology and Forest Resilience in Ancient Meroe, Sudan

87SR/86SR Analysis as a Method to Explore Human Ecology and Forest Resilience in Ancient Meroe, Sudan

Item Type text; Electronic Thesis

Authors Herrick, Hannah Marie

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 26/09/2021 15:07:12

Link to Item http://hdl.handle.net/10150/628127

87SR/86SR ANALYSIS AS A METHOD TO EXPLORE HUMAN ECOLOGY AND FOREST RESILIENCE IN ANCIENT MEROE, SUDAN

Hannah M. Herrick

A Thesis Submitted to the Faculty of the

DEPARTMENT OF GEOSCIENCES

In Partial Fulfillment of the Requirements

For the Degree of

MASTER OF SCIENCE

In the Graduate College

THE UNIVERSITY OF ARIZONA

2018

Acknowledgements

Lengthy documents such as this are seldom truly completed through the efforts of just one person. So many individuals have helped this research come to fruition through their own, particular brand of support. More so than anything I’ve ever done, this thesis requires a substantial section of acknowledgements.

My first thanks go to the supervisor of this research, Charlotte Pearson. She has encouraged me even in the most hand-wringing of times, and unfailingly helped to make my most far- fetched ideas into starbursts of clever, testable hypotheses. It has truly been a privilege to call her my mentor.

I must also thank with great gusto Jane Humphris, director of the University College London Qatar’s Research in Sudan. Jane’s willingness to share charcoal samples and new data from the excavations at Meroe made this thesis possible, and I very much look forward to our continued collaboration.

Kathryn Howley is also deserving of a myriad of thanks. I am so grateful that she introduced me to the fantastic history and culture of ancient Nubia, and made a dream come true by bringing me to Sudan. Her enthusiasm constantly makes me hungry for more of this beautiful place, both the ancient vestiges and the modern culture.

To the following individuals and entities, I offer a sincere “thank you” for their questions, advice, inspiration, assistance, and work on my behalf:

David Killick, Malcolm Hughes, AlHassan Ahmed Mohamed, Jason Kirk, David Frank, Todd Lange, Greg Hodgins, Jacob Martin, Tomasz Wazny, John “Mac” Marston, Rennan Lemos, Jessie Pearl, Eleni Hasaki, Vance Holliday, Jay Quade, David Dettman, Stephanie Kukolich, Derek Runge, Arin Haverland, the Laboratory of Tree-Ring Research (UA), the Mediterranean Dendroarchaeology Lab (UA), the Arizona Laboratory for Emerging Contaminants (UA), the National Corporation for Antiquities and Museums (Sudan), UCL Qatar Research in Sudan, the Qatar-Sudan Archaeological Project, the British Institute in Eastern Africa, the Sanam Temple Project, and the Accelerated Mass Spectrometry Laboratory (UA).

Finally, to my family and wonderful friends: thank you for your acceptance of my two-year obsession with acacia.

Table of Contents List of Figures ...... 6 List of Tables ...... 7 Abstract ...... 8 Introduction ...... 9 Charcoal in Context ...... 14 Charcoal Formation & Chemistry ...... 14 Using Charcoal as a Multi-Disciplinary Proxy ...... 16 Chronological Data ...... 17 Species Identification ...... 18 Provenance ...... 19 Paleoclimatic & Ecological Data...... 26 Human Dimensions ...... 27 Summary ...... 28 Study Context & Research Overview ...... 29 Regional Environment & History ...... 29 Site & Study Period ...... 37 Research Overview ...... 39 Methods ...... 41 Field Collection ...... 41 Meroe ...... 41 Al-Ghazali ...... 41 Laboratory Methods ...... 44 Experimental Charcoal Production...... 44 87Sr/86Sr Analysis ...... 48 Carbon Isotope Analysis ...... 48 Results ...... 50 Experimental Charcoal Analysis ...... 50 Archaeometallurgical Charcoal:87Sr/86Sr ...... 52 Archaeological Charcoal: δ13C ...... 54 Discussion ...... 57 87Sr/86Sr Stability during Acacia Charcoal Production ...... 57 Carbon Isotope (δ13C) Stability during Acacia Charcoal Production ...... 58

Archaeological Charcoal: 87Sr/86Sr Analysis ...... 59 Archaeological Charcoal: Carbon Isotope (δ13C) Analysis ...... 62 Implications for Ancient Meroe ...... 64 Wood Fuel Procurement ...... 64 A Reduction or Cessation of Iron Production ...... 65 Environmental Considerations ...... 68 Conclusions ...... 72 Future Work ...... 73 References ...... 76

List of Figures

Figure 1. Homogenization of different cell wall components during wood carbonization...... 15 Figure 2. Map of Nubia, with several key locations...... 29 Figure 3. Map of the Nile Basin in Eastern Africa...... 31 Figure 4. Alluvial terrace on the bank of the Nile at Aswan, Egypt...... 32 Figure 5. Jebel Barkal near Karima, Sudan...... 33 Figure 6. Ferruginous sandstone outcrops near al-Ghazali, Sudan...... 33 Figure 7. Vachellia nilotica growing in the Wadi Abu Dom at al-Ghazali, Sudan...... 36 Figure 8. Paleochannel at al-Ghazali, Sudan...... 43 Figure 9. Acacia at al-Ghazali...... 43 Figure 10. Laboratory experiment ...... 45 Figure 11. Charcoal production in the "hearth" simulation...... 46 Figure 12. Experimental materials ...... 47 Figure 13. Results from Sr isotope analysis on experimental material ...... 50 Figure 14. Results from carbon isotope (δ13C) analysis on experimental materials ...... 52 Figure 15. Results from Sr isotope analysis on Meroe material ...... 54 Figure 16. Results from carbon isotope (δ13C) analysis on Meroe material...... 55 Figure 17. Upper Nile Valley 87Sr/86Sr data...... 60 Figure 18. Radiocarbon dates from Meroe and Hamadab...... 67

List of Tables

Table 1. Hearth temperature during charcoal production...... 47 Table 2. Experimental charcoal strontium isotope results...... 51 Table 3. Carbon isotope (δ13C ) results for experimental charcoal...... 52 Table 4. Strontium isotope results for archaeological material ...... 54 Table 5. δ13C (VPDB) values for archaeological material from Meroe...... 56

Abstract

Archaeologists frequently apply 87Sr/86Sr geoprovenance to human and animal bone to answer questions surrounding foodways and migration. Utilizing this method to source archaeological timber has recently expanded, enriching the archaeological record and enhancing studies in wood geochemistry. Due to the effects of time and post-depositional alteration, a considerable portion of archaeological wood is found in the form of charcoal. However, the impacts of carbonization and chemical contamination on wood strontium resilience—and therefore, the applicability of 87Sr/86Sr geoprovenance for anthropogenic charcoal—have not yet been examined. Experimental analysis comparing 87Sr/86Sr ratios and δ13C values in wood and experimental charcoal (heated at ~400°C and 900°C) from two Vachellia nilotica trees sourced in modern Sudan illustrated a consistency between untreated wood and both forms of charcoal sufficient to suggest that pyrolysis alone does not significantly alter 87Sr/86Sr ratios or δ13C values in V. nilotica. 87Sr/86Sr values ranged from 0.71715 – 0.71740 (GZL 1) and 0.71678 – 0.71692 (GZL 2). Carbon isotope values ranged from -26.4±0.06؉ (VPDB) to -28.4±0.06؉ (VPDB). Tests on V. nilotica charcoal remnants from late Holocene iron smelting contexts in Meroe, Sudan show that it is possible to produce 87Sr/86Sr ratios from anthropogenic charcoal. Further, these preliminary indications suggest that 87Sr/86Sr data from charcoal supported by carbon isotope data from the same material can be used as a reliable indicator for wood provenance. 87Sr/86Sr data from the Meroe material ranged from 0.70724 - 0.70737, with δ13C values ranging from -26.9±0.06؉ to -24.6±0.06؉ (VPDB). Using this knowledge, I posit that iron smelting fuel charcoal at Meroe may have come from one region between the Atbara confluence and the confluence of the Blue Nile and the White Nile from 400 BCE – 400 CE. Carbon isotope analysis supports this hypothesis. These results suggest that the preference of local V. nilotica charcoal for iron smelting, combined with intensive production ca. 400 – 200 BCE, may have been a catalyst for a break in iron smelting activity at Meroe due to resource exhaustion or climatic variability evidenced throughout the region. Additional improvement in 87Sr/86Sr and δ13C methods for charcoal analysis will enhance this and similar studies in the future.

Introduction

Though it is no longer the fuel of choice in the industrial age, charcoal nevertheless persists in modern settings around the world. Burning charcoal gives heat to cooking surfaces, kilns, and campfires1, although its elemental carbon composition and absorptive qualities are advantageous as part of filtration systems2, medical pigment3, cosmetics4, and toothpaste.5

Even so, charcoal is a product brought from flame, and fuel remains its most prominent purpose in both domestic and industrial contexts. Because of this worldwide ubiquity, charcoal has become a useful proxy in a variety of fields, including fire ecology6, paleoclimatology7, woodland ecology8, biogeochemistry9, and archaeology.10

The incorporation of charcoal as a proxy in studies with environmental components has been swiftly increasing as scholars develop novel methods for its analysis. More concretely,

Quaternary International published three entire volumes on anthracological (charcoal-related) studies during late 2017 and early 2018.11,12,13 These special issues include articles which utilize charcoal in a variety of ways, including in paleoforest reconstruction and analysis of

1 Namaalwa et al 2009 2 Hayashi et al 2016 3 Chami et al 2015 4 Kulkarni et al 2017 5 Brooks et al 2017 6 Hoffman et al 2016 7 Xiao et al 2017 8 Marston et al 2017 9 Pingree 2017 10 Barakat 1995 11 Ludemann and Nelle 2017a 12 Ludemann and Nelle 2017b 13 Ludemann and Nelle 2018

9 historical fire regimes.14,15 However, charcoal studies involving marginal environments such as deserts still require expansion.

The study of charcoal in semi-arid or arid landscapes has exceptional potential. Species in these regions have a high preservation potential. For example, living Acacia varieties can withstand termites and moisture damage, as well as salinity levels up to 100 mM.16,17 This resilient biomass is then more likely to be carbonized through anthropogenic or natural processes. In these dry regions, wood is a precious resource. Wood is often carefully selected by species or other factors to be then fashioned into useful tools and materials, which may remain in their place of origin or translocated should their maker move locations.18 While charcoal production can sometimes be economically lucrative, the finished product is often considered waste after its use.19 The lack of value placed on charcoal in concert with its inert characterization results in high availability of charcoal on the landscape even when the wood products have gone, sometimes thousands of years ago.20 Dry climates are favorable to organic preservation, though charcoal often persists in soils in wet climates unsuitable to pollen preservation.21 Perhaps due to their general lack of well-developed soils and perceived absence of diverse plant life, semi-arid and arid-regions typically boast few (if any) studies particular to trees or wood.

14 Ludemann et al. 2017 15 Hoffman et al. 2016 16 Wickens 1997 17 Nabil & Coudret 1995 18 Springuel & Mekki 1994 19 Springuel & Mekki 1994 20 Barakat 1995 21 Hoffman et al 2016

The fields of paleoclimatology22, dendrochronology23, and wood science24 are advancing in such a way that allows extreme environments like deserts and polar regions to contribute to anthracological study. For example, new chemical methods in determining annual ring boundaries in species which do not display conventional annual growth (many of which can be found in semi-arid and arid regions) can allow the application of dendrochronology to new species.25 Further, developments in understanding the carbonization process has allowed for biogeochemical techniques such as light isotope-based paleoclimate analysis to be incorporated in charcoal studies.26

The modern country of Sudan is one such region where environmental studies (and especially those of wood or charcoal) had historically been ignored until the last few decades.

Breakthroughs in understanding climate change and human impacts on the landscape have altered this trajectory, bringing in a small number of studies on the modern Sudanese environment.27,28,29 Changes in this region are especially critical to monitor due to the extreme aridity present throughout the country and possible threats against economic livelihoods and food security. Further, Sudan is an example of sustaining life on an arid landscape—a reality more of the globe may have to embrace in a warming world.

22 Francis & Poole 2002 23 Blondel et al 2018 24 Xu et al 2017 25 Witt et al 2017. 26 Baton et al 2017 27 Sulieman 2018 28 Abaker et al 2017 29 Fahmi 2017

Fortunately, there is a great deal of raw data to work with if one knows where to look.

Sudan’s past is reflected in a myriad of archaeological sites throughout the country, ranging from peppered evidence of mobile Neolithic communities30 all the way to capital cities of

African empires boasting luxurious infrastructure and intensive industry.31 At sites like these, archaeologists frequently uncover charcoal that, while capable of informing anthropological questions, may also impact climate studies. Investigating the biogeochemistry of charcoal from contexts such as these also lends new information about the resilience of plant life, especially in stressed environments.32 Continuing to develop methods to decode more information from these charcoal fragments is key to advancing anthracological evidence in desert regions such as Sudan.

Through combined methods in the earth sciences and archaeology, the purpose of this study is to contribute to the growing body of knowledge on charcoal. It seeks to explore the feasibility of establishing the provenance of archaeological charcoal through 87Sr/86Sr ratios.

This novel approach can inform a range of studies and expand the nature of multi- disciplinary studies involving charcoal and, to a certain degree, wood. The present study will focus on the application of the strontium geoprovenance method on wood charcoal from

Nahr al Nil State, Sudan. First, charcoal formation, chemical processes, and proxy applications and their associated methods will be discussed. The study determining the suitability of charcoal for 87Sr/86Sr analysis will follow. This technique will then be applied to

30 Barakat 1995 31 Humphris 2014 32 Pingree 2017

12 archaeological charcoal from Meroe, Sudan. Finally, potential relationships between these results and other environmental proxies, as well as future applications, will be explored.

Charcoal in Context

Charcoal Formation & Chemistry

The carbonization of both floral and faunal material under oxygen-free conditions results in the creation of the material we know as “charcoal”. This work will specifically discuss carbonized woody material, such as tree stems or branches. While most “wood” charcoal comes from dicotyledon trees, carbonized remains of monophyte plants that produce somewhat woody material, like palms and bamboo, may also be considered wood charcoal.

Charcoal is created when woody material is heated (with 260°C being an optimum temperature33) in an anoxic environment. Instead of combusting, a process which would need a ready oxygen supply, the plant material undergoes pyrolysis.34 The conversion of cellulose-rich biomass via pyrolysis thermodynamically favors the production of carbon in this process35:

C6H12O5  3.74C + 2.65H20 + 1.17CO2 + 1.08CH4

The pyrolysis reaction serves to decompose larger molecules into smaller components by application of heat.36 Though chemical spectra confirm that wood exposed to heat ≤250°C maintains unaltered chemical characteristics, cellulose- and lignin-based components

33 Wilk et al 2016 34 Antal & Gronli 2003 35 Ibid 36 Kung 1972

14 eventually are replaced by carbon, carbon dioxide, and benzene by 400°C, via aromatic products as an intermediary.37

The process of carbonization also impacts the physical characteristics of the parent material.

In some cases, this is due to the destructive nature of high temperatures. For example, the structural integrity of fiber cells degrades at temperatures higher than 300°C, making the carbonized secondary cell walls indistinguishable from those of the compound middle lamella (Figure 1).38 In deciduous species, the proportion of tyloses (and ability to determine if a specimen contains heartwood or sapwood) may alter if the wood is heated above

800°C.39

Figure 1. Homogenization of different cell wall components during wood carbonization at different temperatures. From Kwan et al 2009.

37 Braadbaart & Poole 2008 38 Xu et al 2017 39 Dufraisse et al 2018

Other physical changes during carbonization do not destroy or degrade the wood, but rather alter its measurable properties. Inertite macerals are organic, mineral-like compounds made from the production of coal and other materials such as charcoal; the particular temperature during carbonization of the woody material determines the macerals’ final physical and optical characteristics.40 In other words, through reflective microscopy, one can work backwards from charcoal composition to elucidate the temperature at which it was charred.41

However, not all changes from pyrolysis are advantageous to anthracological study. The mass loss occurring during carbonization and the related wood shrinkage is variable across radial sections of samples, this process is not yet well-understood.42,43

Using Charcoal as a Multi-Disciplinary Proxy

As discussed above, charcoal is recovered from a variety of contexts. Whether by intentional charcoal production or accidental formation, carbonized woody material offers some information even beyond that of its raw counterpart. The ways to derive data from charcoal to understand the complete environment and human behaviors are numerous. Here, some of the primary types are discussed, as well as typical and novel practices across several disciplines. The three key pieces of information gained from charcoal analysis are chronological data, species identification, and provenance—these three can be combined and applied to explore paleoclimatology, ecology, and human-landscape interactions.

40 Scott & Glasspool 2007 41 Caricchi et al 2014 42 Paradis-Grenouillet & Dufraisse 2018 43 Blondel et al 2018

Chronological Data

One of the most important types of data obtained from wood charcoal is chronological dates. Radiocarbon (14C) analysis is commonly used to determine dates on small charcoal fragments. For example, archaeologists often use charcoal in stratified sequences to determine dates of occupation44, understand production episodes45, and date sociopolitical collapse.46 Researchers focused on the physical aspects of the environment, such as forest ecologists, fire historians, paleoclimatologists, and forest managers may wish to know the date of historic or paleoforest wildfires, and volcanic activity.47,48,49 These chronological data can be used to create hazard assessments, or become incorporated into models projecting future events. Annual ring-forming species are particularly advantageous, as they can be dated by both dendrochronology and radiometric methods.50 This double-edged chronometry allows for the creation of an annually-resolved time series which can be used to calibrate radiocarbon dates and improve local chronologies.51

However, the fragile nature of charcoal leads to a preponderance of fragmentary samples.

This, in turn, results in short tree-ring sequences, complicating the dendrochronological analysis.52 Even if the sample is of considerable size, the effects of shrinkage-related carbonization are still poorly quantified.53 Further, radiocarbon dates from charcoal can be

44 Beyin et al 2017 45 Humphris & Scheibner 2017 46 Inomata et al 2017 47 Hoffman et al 2016 48 Robin et al 2014 49 Okuno et al 2017 50 Rosler et al 2012 51 Xu et al 2013 52 Marguerie 2011 53 Blondel et al 2018

17 complicated by the “old wood problem”, wherein the decay of the young outer edges of wood or charcoal results in an artificially “old” radiocarbon date in terms of its application to site stratigraphy.54

Species Identification

Another metric obtained from charcoal is the species of its parent material, information that is relevant across a wide variety of disciplines. Although the cellular structure of wood may be altered on the microscopic scale due to pyrolysis, many of the anatomical features necessary to elucidate genus and often species remain intact.55 Primarily, identification of species in charcoal can be used to understand facets of the local ecology in both past and present environment.

Archaeobotanists analyze the parent material species of charcoal to develop an understanding about past humans’ relationship with the wooded landscape. For instance, species identification of charcoal can aid in reconstructing past flora assemblages near human settlements,56 as well as identifying human behaviors surrounding consumption of landscape resources related to activities such as ceramic production and metallurgy.57 This application of species identification is not exclusive to archaeology, however—recreating

54 Schiffer 1986 55 Leney & Casteel 1975 56 Marston et al 2017 57 Iles 2016

18 burned forests from charcoal remains can add great temporal depth to the ecological record, extending the forest history in some cases up to 13,000 years.58

Provenance

The contextual meaning of a charcoal fragment in some instances is dependent upon the knowledge of the parent tree’s origin. Determining the geographic provenance of charcoal fragments in cases where this information is unknown is becoming an increasingly important part of charcoal studies. When determining charcoal origin, several methods may be utilized.

One proven method of charcoal provenance determination—dendroprovenance—is discussed in detail below. New avenues of inquiry, including strontium isotope ratios, light isotope analysis, and DNA analysis, are posited thereafter.

Dendrochronology uses several metrics (e.g., ring-width) to chronicle climatic sensitivity of annual ring-forming trees.59 In regions with relatively sensitive climatic responses, such as the

American Southwest, tree-ring chronologies can be used as tools in determining the geographic origin of archaeological timber as well as charcoal.60,61 This method is known by the portmanteau “dendroprovenance”.62 In dendrochronology, understanding the regional climate signal is paramount: dendroprovenance, then, is an integral component of tree-ring dating.63

58 Hoffman et al 2016 59 Fritts 1976 60 Guiterman et al. 2016 61 Marguerie and Hunot 2007 62 Speer 2010 63 Fritts 1976

Dendroprovenance studies hinge on the unique properties of regional tree-ring records resultant from regional climatic variations.64 Investigators utilize site-specific ring-width chronologies for a series of forest areas where the wood (or charcoal) may have originated, and compare against the chronology of the sample in question.65 Statistical analysis applied to ring-width measurements (via platforms such as Tellervo) are then used to determine the most likely match of the unknown sample to a regional chronology network.66

Dendroprovenance techniques have yielded successful results in a number of contexts, including in archaic structures of the American Southwest,67 historic churches in the Near

East,68 Flemish panel paintings,69 and ancient French hearth charcoal.70 Dendroprovenance, having basis in techniques honed for nearly a century,71 is likely the most frequently attempted method used in sourcing charcoal. However, other methods all show potential for this purpose.

Strontium (Sr) isotope analysis is a technique with developing roots in wood provenance.

Though frequently used in studies to determine the geologic origin of rocks, minerals, or water, 87Sr/86Sr analysis may also be applied to calcium-containing compounds, such as bone or vegetal material, wherein strontium atoms replace calcium atoms due to the similarity in size of their atomic radii.72 The ratio between two stable isotopes of strontium, 87Sr and 86Sr,

64 Fritts 1976 65 Haneca et al. 2006 66 Brewer 2016 67 Guiterman et al. 2016 68 Bernbei & Bontadi 2012 69 Baillie et al. 1985 70 Marguerie and Hinot 2007 71 Douglass 1929 72 Pollard et al. 2007

20 is shared between a geological source material and the products it impacts through transportation in water, including soil, sand, bone, and trees.73 The unique properties of

87Sr/86Sr values have led to applications in several fields, including the reconstruction of river systems74 and archaeological migration studies.75

Trees (and especially archaeological wood, given the stability of the strontium atom within the tissue) also contain some of this strontium-calcium substitution, and 87Sr/86Sr values have been used to trace calcium bioavailability.76 The application of strontium isotope analysis for provenance of woody material is in its infancy. Nevertheless, this method has been used to create indices to determine the origin (and possible building location) of ship timbers77,78 and provenance archaeological structural elements.79,80,81

Using Sr isotope analysis requires a set of conditions to be satisfied. First, the suspected origin areas must have sufficient variability in strontium uptake, such that the sources can be statistically differentiated.82 This differentiation is not as simple as ascertaining variable geologic bedrock (for which data is sometimes unavailable)—water catchment, erosion, and aeolian sediment deposition all play a part in determining the terrestrial distribution of Sr

73 Graustein & Armstrong 1983 74 Woodward et al. 2015 75 Buzon & Simonetti 2013 76 Aberg et al. 1990 77 Rich et al 2012 78 Rich 2013 79 Drake et al 2014 80 English et al 2001 81 Reynolds et al 2005 82 Pollard et al 2007

21 from rock to growing tree.83 If the uptake variability is satisfied, taphonomic processes must also be considered, especially in the case of archaeological objects. A wooden artifact may have been buried in an area displaced from the grove where its wood was sourced, which may lead to contamination by the local soil, and ambiguous results.84 Further, the wood may be submerged underwater. A final caveat of 87Sr/86Sr provenance is the tendency for these isotope ratios to vary over time. For example, 87Sr/86Sr isotope ratios in the Dongola Reach area of the Nile display a significant decrease from 6000 Cal BP to 4000 Cal BP, with an increasing trend thereafter due to shifts in river sediment source.85

Studies investigating the processes of carbonization affecting wood Sr retention or modification have not yet been published. It is possible that carbonization of wood may cause a small fractionation of Sr compared to raw wood, as it does with 13C in charcoal.86

Preliminary 87Sr/86Sr data combined with x-ray fluorescence suggests that some Sr is preserved during carbonization events, but more analysis must be undertaken.87 If successful, this method could be applied to a range of situations, including provenancing archaeological fuel sources and confirming the origin of wildfire kindling.

Carbon, oxygen, and hydrogen isotopes (often called “light” or “stable” isotopes collectively) also have the potential to elucidate origins of wood. Light isotopes are excellent proxies for

83 Graustein 1989 84 Pollard et al 2007 85 Woodward et al 2015 86 Poole et al. 2002 87 Herrick, this study

22 water distribution and atmospheric condition.88 In much in the same way as tree-ring data, light isotope ratios are influenced by unique, microenvironment-specific conditions resultant from climate, ecological niche, and anthropogenic alteration.89 Meteoric water records vary in oxygen (δ18O) and hydrogen (δ2H) isotope composition relative to a number of factors, including temperature, elevation, amount of precipitation, and site position relative to coastlines.90 Changes in carbon isotopes, or δ13C, reflect the ability of the tree to uptake

91 water via the presence of CO2 in the atmosphere at the time of tissue formation. By matching these isotope values against extant climate data, wood can be mapped to a geographic location.

Carbon, oxygen, and hydrogen are abundant in organic tissues present in trees, though in varying quantities and gradients throughout the stem.92 For example, variations as large as 1-

,3؉ δ13C, 1-4؉ δ18O, and 5-30؉ δ2H have been demonstrated to exist within a single tree creating unique challenges in sampling strategy—further, these values are, of course, variable by species.93 Additionally, components of the woody tissue (such as lignin or cellulose) fractionate these isotopes differently, resulting in varying values based on the specific compound analyzed.94 Together (or in some cases, individually), light isotope values can be derived from woody material, and compared to existing isotope chronologies or other paleoclimate indicators.

88 Sharp 2006 89 Ibid 90 Rozanski et al. 1993 91 Bert et al. 1997 92 Leavitt 2010 93 Ibid. 94 McCarroll & Loader 2004

One avenue for applying light isotope chronologies to properly elucidate the provenance of wood is through comparison with long-term tree-ring records from the same region. This method has been tested on wood from the American Southwest: well-studied tree-ring records were used to construct 13C isotope chronologies, which were used in concert with ring-width values to appropriately source the wood to their regions of growth.95 This technique also has potential to assist in confirming calendar years assigned to wood via tree- ring dating in ring-forming species, as the carbon isotope chronologies cross-date in much the same way.96 However, using light isotopes as source origin indices does not require that the tree produce annual growth rings—indeed, construction of paleoclimate records for species from regions such as the tropics that do not form rings on a yearly basis may be possible through creating isotope chronologies for provenance purposes.97

Although much of the hydrogen and oxygen leaves the wood cellulose during charcoal formation, a substantial amount of elemental carbon remains. As charcoal may possess both tree-ring data and carbon isotope information, a combination of these techniques would be complementary. However, there is some evidence that different stages of pyrolysis of wood favor either the 12C or the 13C carbon isotopes, rendering the carbonized wood slightly changed in δ13C value from its parent material.98,99 To advance carbon isotope provenance techniques, greater examination of the effects of carbonization must be performed.

95 Kagawa & Leavitt 2010 96 Leavitt et al 1985 97 Kagawa & Leavitt 2010 98 Poole et al 2002 99 Jones et al 1993

A final suite of methods used in the assessment of wood origin are those involving DNA analysis. Chloroplast DNA (or cpDNA) is inherited from the “female” plant, and has proven to be an effective molecular marker for the tracing of wood populations, and especially of

European oak.100,101,102,103 New studies using single nucleotide polymorphisms (or SNPs) have shown potential to act as less-expensive alternative measures of genetic material—this is especially pertinent in species such as iroko (Milicia excelsa), whose genetic composition primarily consists of nSSRs (nuclear simple sequence repeats), short sequences of genetic material which are commonly transferred across taxa, complicating cpDNA analysis and making SNP analysis more accessible.104,105 Phylogenic relationships between species must be well-studied for DNA analysis to be applicable to a particular regions’ trees. Nevertheless, if an investigator is fortunate enough to work in such an area, current DNA extraction and analytical technologies are inexpensive and often produce quick results.106

DNA analyses are somewhat restricted by the quality of the wood, casting doubt on the applicability of this technique to charcoal. Dried wood or old tissue (such as one might find in fossil settings or at archaeological sites) often do not retain sufficient quantities of DNA to perform analysis, and much of what may be left after drying is often degraded.107 An exception comes with oak species: chloroplast DNA in these dried specimens continues to

100 Ibid 101 Dumolin-Lapegue et al 1999 102 Deguilloux et al 2003 103 Speirs et al 2009 104 Blanc-Jolivet et al. 2017 105 Billotte et al. 2004 106 Lowe & Cross 2011 107 Ibid.

25 be of appropriate condition for analysis.108 This is evidenced by the development of

European oak genetic sequences to the level of the individual tree.109 The extraction efficiency of DNA from biochar has been assessed, and obtaining genetic information from such materials is indeed possible.110 Additional study on wood charcoal specifically may determine if the same methods used on woody tissue are feasible for carbonized specimens.

Paleoclimatic & Ecological Data

In combination with charcoal sample date, data such as species, and charcoal provenance, paleoclimatic and ecological records (such as ice cores, lake sediments, or pollen data) can be used to reconstruct, or at least inform, studies of past climate. The above data categories

(dating, species identification, and provenance) each include applications of their analyses for this purpose. This status is fitting, as the investigation of past environments is by necessity a trans-disciplinary field. However, paleoclimatology is somewhat unique in that analyses may range in date from recorded history111,112 up to tens of thousands of years.113,114,115 The equally-unique inert properties of charcoal make it a powerful proxy in paleoclimatology as well.

Ecology as influenced by climate change is a prime intersection of these applied techniques.

108 Dumolin-Lapegue et al. 1999 109 Deguilloux et al. 2004 110 Dai et al 2017 111 Rosler et al 2012 112 Drake et al 2012 113 Yost et al 2018 114 Hoffman et al 2016 115 Enache et al 2009

Ecological landscape reconstruction can be accomplished by the identification of species in an assemblage, followed by an analysis of the plant interactions of those types. Especially in combination with other vegetation proxies such as phytoliths and fungi, charcoal remains offer an insight to plants present on the landscapes of the past.116,117,118 These results can also be paired with stable isotope data (as above) to further support paleoclimatic hypotheses.119

One of the most powerful uses of charcoal in past landscape studies is its application in fire ecology. Charcoal, as an indicator of past burning, can be integrated in soil stratigraphic layers120 and dated (often by radiocarbon) to form chronological sequences in fire history.121

In turn, these records of past fire events can be used as tools in understanding local vegetation or moisture changes, their relationship to wildfire occurrence, and resulting alterations to the landscape.

Human Dimensions

Finally, charcoal holds information relating to human behavior in the landscape. Human behaviors surrounding trees and wood, such as species selectivity or resource management, can be examined through interpretations of their woody resource acquisition.122 The species of charcoal from different contexts provides insight into species selection for a specific purpose. For example, potters in the ancient Egyptian city of Karanis preferentially selected tamarisk for fueling their kilns due to its high heat capacity,123 and modern materials

116 Innes et al 2003 117 Yost et al 2018 118 Pessenda et al 1998 119 Baton et al 2017 120 Robin et al 2013 121 Hoffman et al 2016 122 Marston 2009 123 Marston et al 2017

27 scientists have selected bamboo charcoal as particularly effective in chemical adsorption.124

Anthropogenic impacts on the landscape can be assessed through charcoal remains.

Metallurgy, one of the most consumptive craft technologies, has been tied to deforestation in several cases across history,125 a phenomenon which can continue to be explored through further charcoal research. Another human facet that charcoal can elucidate comes from the assessment of fire temperatures. Reflectivity in charcoal fragments can be used to reconstruct the temperature at which the parent material was carbonized (as discussed in

“Charcoal Formation & Chemistry”), which translates into new understanding of technological processes or response to fire events such as volcanic eruptions.126

Summary Overall, charcoal’s vast potential to reveal several components of the past renders its use as a powerful multi-proxy. Further investigation into the chemistry of charcoal, as well as new techniques for its analysis, will undoubtedly provide additional avenues to unlock its secrets.

However, the tried-and-true methods of dating, species identification, and isotopic analyses still stand as a formidable foundation in charcoal studies. Applying these methods in novel ways or to understudied regions is imperative in the continued growth of anthracological research. In this study, I will investigate the viability of using 87Sr/86Sr analyses as a method to provenance archaeological charcoal from iron smelting debris deposited from ca. 400

BCE – 400 CE at Meroe, Sudan.

124 Scott & Damblon 2010 125 Iles 2016 126 Caricchi et al 2014

Study Context & Research Overview

Regional Environment & History

The region of southern Egypt and Sudan

bracketed by the first and sixth Nile

cataracts (Figure 2) was called “Nubia” or

“Kush” in antiquity.127 The geographic area

is quite large, laying across the Sahara and

Sahel climate zones.128 This diversity of

landscape granted people both ancient and

modern access to a range of microclimates

and varied resources. Persistent ties between

ancient Kushite life and the river Nile

ensures a geographic continuity of site

formation restricted primarily to the

riverbanks and seasonal washes following

the African Humid Period (ca. 16,000 –

6,000 years ago) to the present.129 However,

variability and availability of natural

resources from north to south impacted the

Figure 2. Map of Nubia, with several key locations. types of sites formed, the technologies used, Map by Mark Dingemanse (www.vormdicht.nl), used with permission. and their archaeological remains. Most

127 Shinnie 1996 128 Williams and Talbot 2009 129 Hennekam et al 2015

29 important in the understanding of natural resource availability and environment in this region are the Nile alluvial system and local geology.

The Nile itself lies in a basin (Figure 3) extending from Lake Victoria in south-Central Africa to the Mediterranean Sea, including much of the Eastern African region.130 The Nile system is a complex one, beginning with the White Nile’s Headwaters in Uganda and the Blue Nile’s source in the Ethiopian highlands. After their confluence near modern Khartoum, Sudan, today the river is joined by the Atbara near the fifth cataract. While today regulated by dams at Aswan, Egypt and Merowe, Sudan, seasonal rains upstream in Ethiopia resulted in fluctuations in drainage in ancient times.131,132 According to geoarchaeological survey combined with aerial imagery, the Nile’s course has been revealed as highly mobile over the majority of the floodplain over the last five thousand years, focusing on more dramatic changes during one-thousand-year periods.133 Recent pedological study and 87Sr/86Sr analyses concur that the contribution of different branches of the Nile has varied during this time as well, with only 30% of the river’s discharge being attributed to the White Nile, though this percentage was likely greater in antiquity.134,135

130 Butzer and Hansen 1968 131 Williams and Talbot 2009 132 Wolf 2015 133 Bunbury and Lutley 2008 134 Sulieman et al 2016 135 Woodward et al 2015

Figure 3. Map of the Nile Basin in Eastern Africa. From the Nile Basin Initiative.

The Nile’s state of near constant erosion has resulted in the carving of terraces (Figure 4) across the landscape.136 However, very little geological research has been conducted in Sudan in recent years, likely as a result of political turmoil. In northwestern Sudan and farther into the Egyptian Sahara, much of the exposed landforms are comprised of bedrock strata due to the hyper-arid nature of the desert.137 Primarily, this consists of Mesozoic Nubian Series sandstone and the Precambrian basement complex (granite, gneiss, and schist) dotted with

136 Sampsell 2003 137 Butzer and Hansen 1968

31 outcrops of tertiary lava and Hubi chert.138,139 Bedrock also forms the solitary jebels (Arabic meaning, in this case, mountain) on the landscape; Jebel Barkal (Figure 5) near modern

Karima is perhaps the most well-known example.140 The Nubian Series itself is primarily composed of one facies of pale brown, medium-grained sandstone, though surface weathering can often darken the conglomerate to a deeper hue.141 In various watersheds, deposits of ferruginous “black” sandstone (Figure 6) occur, leading to the development of ferromanganese rinds or veins.142,143

Figure 4. Alluvial terrace on the bank of the Nile at Aswan, Egypt. Pharaonic Period tombs were cut into the terrace walls. Photo by Catie Witt (University of Chicago), used with permission.

138 Butzer and Hansen 1968 139 Worrall 1957 140 Ibid 141 Butzer and Hansen 1968 142 Rehren 2001 143 Butzer and Hansen 1968

Figure 5. Jebel Barkal near Karima, Sudan.

Figure 6. Ferruginous sandstone outcrops near al-Ghazali, Sudan.

Sedimentary deposits in the region consist primarily of aeolian silica sands and Nile alluvium.

The ratio of these materials has changed noticeably from the last glacial maximum, shifting in favor of greater proportions of aeolian deposits of silica sand due to drier conditions.144

Seasonal runoff channels, or wadis, demonstrate the depositional stratigraphy of the area.

Typically, the well-defined bedding consists of various marls, silts, and clays.145 Alluvium from as far upriver as the Ethiopian highlands exists in various regions along the riverbank, and clay is deposited in the floodplain inversely proportional to the distance from the river itself.146

Although the climate of Sudan ranges from arid deserts to semi-arid lowlands receiving annual precipitation, much of the region’s flora grows within the Nile floodplain as seen today (Figure 5).147 From antiquity, much of the indigenous population’s vegetal diet has consisted of cereals and grains such as emmer wheat and barley.148 Evidence for this grain- based subsistence can be seen in the botanical remains of settlements ranging in age from ca.

3000 – 1000 BP at archaeological sites located at Qasr Ibrim (now in modern Egypt), Nauri,

Amara West, and Kawa, among others.149,150,151 The substantial animal husbandry (and especially of sheep, goats, and cattle) historically practiced in this region likely necessitated a great deal of plant fodder to feed livestock and larger animals such as elephants, which were utilized in regional warfare.152

144 Woodward et al 2015 145 Butzer and Hansen 1968 146 Sulieman et al 2016 147 Buursink 1971 148 Bianchi 2004 149 Fuller and Edwards 2001 150 Ryan et al 2012 151 Fuller 2004 152 Ikram 2012

Some crops, such as foxtail millet and watermelon, may have been scavenged from the surrounding landscape rather than cultivated.153 More labor-intensive plants such as fresh fruit and vegetables suited to riverine environments supplemented the starchy diet. Prior to the modern industrial age, these crops were cultivated through agricultural techniques similar to those found in Egypt.154 Remains of advanced water management techniques such as hafirs

(dugout collection basins for wadi runoff), shaduf (or counterpoise lift) irrigation, and saqia

(donkey-drawn waterwheel) irrigation dating as early as 3000 BP can be observed near archaeological sites around Sudan.155,156 The addition of these techniques in the farming repertoire may have allowed imported crops such as grapes to prosper on Sudanese soil.157

Plants for purposes other than food also grow readily on the landscape, and include saffron, flax, and cotton.158

Hardy, shrub-like desert trees such as those in the Acacia and Tamarix genii and two species of fruit-bearing palm, Hyphaene thebaica (doum palm) and Phoenix dactylifera (date palm) grow primarily in well-irrigated areas, though some shrubs and trees persist out in the desert margins.159 Several species in the Acacia genus thrive in many extremes of the Sudanese climate—these include A. senegal, used to produce gum Arabic, A. tortillis, which survives both in the desert and near seasonal runoff areas, and Vechellia (formerly Acacia) nilotica

(Figure 7) within the moist soils of the Nile floodplain, though it too can survive in the drier

153 Fuller and Edwards 2004 154 Ryan et al 2012 155 Ikram 2012 156 Fuller 2004 157 Fuller and Edwards 2001 158 Fuller and Edwards 2001 159 Gorashi 2001

35 regions.160 In general, the semi-arid climate of the region does not support the growth of trees suited for robust building material such as oak, though Vechellia nilotica is sometimes used for this purpose.161 However, this is not to say that the area around Meroe was barren.

In fact, the arable soil provided in the Nile floodplain and its relatively consistent flow would have made its banks able to sustain growth of sizeable plant populations. Historical evidence from the historian Strabo (Geographika 1.2.25, 17.2.1-3)162 suggests that the southern reaches of Sudan (near the archaeological site Meroe) may even have received sufficient precipitation to sustain large forests of trees, including stands of ebony.163,164 However, paleoclimate studies with the potential to inform this gap in knowledge have yet to be undertaken.

Figure 7. Vachellia nilotica growing in the Wadi Abu Dom at al-Ghazali, Sudan.

160 Gorashi 2001 161 Gorashi 2001 162 Eide et al 1998 163 Eide et al 1998 164 Bianchi 2004

Site & Study Period

This study focuses on the ancient environment at the site of Meroe, and its use by local iron smelters. Meroe is located at 36Q 577348 1872323 in the Nahr al Nil state, Sudan (Figure 2).

The site lies on the Nile, and would have been subject to its seasonal fluctuations and seasonal wadi (dry riverbed) runoff prior to the construction of dams in the 20th century

CE.165 Today Meroe is characterized by the growth of acacia trees, fed by Nile runoff and marginal rains resultant of the Ethiopian monsoon system.166,167 The river system demonstrated large shifts over 1000-year periods during the Holocene, changing the distribution of its floodplain.168 Aerial photography suggests that the Nile is currently shifting west near Meroe, indicating the ancient city was situated closer to its banks in antiquity.169

The University College London – Qatar Research in Sudan Project has excavated, recorded, and analyzed archaeological material found at Meroe, Sudan since 2012. Meroe’s significance in Sudanese archaeology stems from its tenure as the capital of the Kingdom of Kush from ca. 400 BCE – 400 CE, and as a center of prolific iron smelting.170 The UCL Qatar project focuses primarily on better understanding the iron smelting activities and technology at

Meroe within the framework of the site’s culture. Included in the archaeological remains of iron smelting contexts is a great deal of charcoal.

165 Shinnie 1996 166 Williams and Talbot 2009 167 Wolf 2015 168 Bunbury & Lutley 2008 169 Wolf 2015 170 Humphris and Rehren 2014

Based on botanical analysis of the charcoal, Meroitic smelters preferentially selected a single acacia species, Vachellia nilotica (formerly Acacia nilotica), to fuel their metallurgical furnaces.171,172 Interestingly, V. nilotica does not have a particularly high heat capacity, making it perhaps an atypical fuel source for smelting in comparison to trees in the Tamarix genus— species of which were selected for kiln fuel at contemporaneous Karanis.173 The equal suitability of Tamarix spp. for iron smelting fuel, combined with its presence in metallurgical debris at Meroe and the ability of tamarisk trees to flourish in riverine environments suggests that, despite its availability, Tamarix species were not favored for smelting fuel at Meroe.174

Given that at least 50kg of charcoal was necessary for each smelt, iron production at Meroe would have necessitated vast amounts of V. nilotica.175

Current research indicates that Meroe possessed at least one iron ore mining site about 9km away, and technical ceramic176 materials necessary for their craft on site.177 Based on the pattern of obtaining valuable ore somewhat close to Meroe and more pedestrian ceramic material near the city, the use of exclusively local, common fuel is not unreasonable. It is possible that wood within easy reach was deemed “good enough” for the task at hand, with no need to harvest wood further afield. Potentially, changes in aridity in this semi-arid region may have affected human ability to use the wood in the local landscape due to slow regrowth, necessitating changes in pyrotechnological behavior.

171 Humphris and Rehren 2014 172 Humphris 2014 173 Marston et al 2017 174 Humphris and Eichhorn, forthcoming 175 Charlton and Humphris 2017 176 “Technical ceramics” is the name given to ceramic materials integral to the infrastructure of iron smelting, such as furnaces, tuyères (blowpipes), and brick linings. 177 Charlton and Humphris 2017

However, published Common Era climate studies specifically targeting Sudan prior to the

1960s are exceedingly rare. Instead, sedimentary records from the lower Nile floodplain and near its source lakes, Tana and Victoria, may be comparable proxies. This gap in paleoclimatic information makes Meroe a prime candidate to explore new techniques in environmental archaeology through study of metallurgical charcoal. Investigating the biogeochemistry of charcoal lends new information about the resilience of plant life, especially those in stressed environments. In the present study, the use of 87Sr/86Sr ratios in acacia charcoal from metallurgical contexts will be examined. Finding the geographic origin of these wood charcoal samples during two contrasting periods—during the Meroitic apex

(400-200 BCE) and during the city’s decline (200-400 CE)—would be incredibly beneficial in understanding both regional control and local climatic conditions during this time.178

Research Overview This research will incorporate archaeological data with methods from geochemistry to investigate the following questions:

1. Does the 87Sr/86Sr ratio (if present in charcoal) remain stable or change in a

predictable manner during carbonization of wood at temperatures consistent with

domestic hearths (400°C) and metallurgy (900°C)?

2. Does the carbonization process in Vachellia nilotica favor either 13C or 12C when

heated to the above temperatures?

3. Can 87Sr/86Sr data supported by carbon isotope (δ13C) analysis be applied to

archaeological material deposited 400-200 BCE and 200-400 CE from semi-arid

Meroe to determine the charcoal’s provenance(s) and possible temporal change?

178 Welsby 2002

4. How can this information be combined with other historic and paleoclimatic records

to form a better understanding of the local past environment?

Methods

Field Collection Field data was collected from two sites in Sudan: Meroe, and al-Ghazali.

Meroe Charcoal samples from Meroe were excavated on-site by the University College London

Research in Sudan archaeological team. The principal investigator, Dr. Jane Humphris, selected and mailed subsample assemblages from two loci, MIS 3 and MIS 6, to the author at the University of Arizona.

Al-Ghazali Modern wood for experimental charcoal production with was not able to be collected at

Meroe due to the restrictive nature of travel between Sudanese states. Instead, modern wood was collected by the author at al-Ghazali, Northern State, Sudan in January 2018. Al-Ghazali is located at 36Q 387153 2039414, near the ancient city Napata.179 This site was selected as an appropriate parallel for the Meroe archaeological material due to its accessibility and characteristic location in a desert oasis, which was predicted to present with an 87Sr/86Sr ratio unique from that of the Meroe area.

The site is situated far from the Nile, lying roughly 15 kilometers in the desert margins within the Wadi Abu Dom.180 Though the site today lies in vast desert, vestiges of its past local oasis brought from seasonal runoff are still seen in the remains of a paleochannel

(Figure 8). Al-Ghazali was constructed near substantial outcrops of ferruginous sandstone

179 Gabra and Takla 2013 180 Obluski et al 2015

(Figure 6), providing a likely source of ore for iron smelting activity, the remains of which can be found on the margins of the site.181 In addition, a small quantity of acacia trees dot the area, representing a small amount of what the vegetation which may have been present in wetter times (Figure 9). As al-Ghazali’s remote location (its only neighbors being the site keeper and his family) has rendered the landscape unaffected by modern industrialization, it was identified as a suitable site from which to collect modern material against which to compare the Meroe charcoal. Though input of aeolian sands from farther afield may have altered the 87Sr/86Sr values over time thus making the isotopic values incomparable, the strontium isotope ratios found in the trees should experience little (if any) external contamination.182 Even the dispersed nature of the acacia trees provides a possible parallel of the kinds of fuel Meroitic smelters may have procured on the desert margins. Further, the close locality of both iron ore and volcanic outcrops (on top of which the church was built) give the site a unique geological characteristic that can be paralleled at Meroe.183

Permissions to sample trees from an archaeological site and export the samples for analysis were granted by the General Director of the National Corporation for Antiquities and

Museums, Dr. Al-Hassan Ahmed Mohamed. Two Acacia sp. trees (GZL 1 and GZL 2) were selected based on visual species identification via the leaves and flowers, based on characteristics of Vachellia nilotica (formerly Acacia nilotica). Branches were selected based on accessibility from the ground as well as having a minimum diameter of 2 cm. Branches were cut off and trimmed of excess material using a small hand saw.

181 Obluski et al 2015 182 Woodward et al 2015 183 Gabra and Takla 2013

Figure 8. Paleochannel at al-Ghazali, Sudan. Possible channel sills bordered with red for clarity.

Figure 9. Acacia at al-Ghazali. Photographed in the most vegetation-dense area on site.

Laboratory Methods In order to determine the feasibility of applying 87Sr/86Sr analysis to archaeological charcoal, several experimental methods were tested to find a suitable protocol for creating wood charcoal in a controlled environment. Wood samples were carbonized in two groups, approximating conditions of iron smelting fuel, and domestic hearths. Following this, strontium isotope and carbon isotope (δ13C) analyses were then conducted on untreated

GZL samples, both experimental charcoal varieties, and archaeological charcoal samples.

Experimental Charcoal Production To approximate charcoal used as fuel, subsamples from both GZL 1 and GZL 2 were heated to 900°C, to approximate temperatures inside smelting furnaces. 184 Using the facilities of the University of Arizona Accelerated Mass Spectrometer laboratory, a subsample of two trees from al-Ghazali (GZL 1 and GZL 2) were carbonized in a Barnstead

Thermolyte 47900 muffler furnace. Each sample was carbonized one at a time. Fragments totaling approximately 600mg in mass were placed in a 9mm quartz tube (Figure 10). Each tube was placed in the top vent of the furnace, which was insulated with quartz wool. The tubes were adjusted to a depth such that the wood material inside was in the approximate center of the furnace, without the tube touching the furnace bottom. The tubes were capped with a compression valve under vacuum such that pyrolytic material and other volatiles could escape the tube without allowing oxygen in to prevent combustion and ensure carbonization (Figure 10). The compression valve’s tube was fed into the adjacent fume

184Charlton & Humphris 2017

44 hood’s vent to further these measures and ensure that no pyrolytic material escaped into the lab. Samples were heated in the furnace for 3-4 hours.

a b

Figure 10. a) Acacia fragments in tube for charcoal production, b) setup for heating in the furnace

As one aim of this study was to test how 87Sr/86Sr ratios may be affected by different types of fire (iron production vs domestic use) burning at different temperatures, I designed a series of experiments to determine the best approximation of a domestic hearth. The chosen method required tightly wrapping GZL wood fragments several times in Reynolds Wrap

Heavy Duty household aluminum foil (made from 99.85% non-Sr elemental metals with a thickness of 0.94 mil, approaching gas impermeability185), and placed inside a clean 15.25oz can capped with an 18.5oz can such that the charcoal was shielded on all sides. This apparatus was nestled within a pile of Kingsford charcoal within a backyard bowl-style fire basin, which was lit with a household lighter (Figure 11). This fire was monitored for a total

185 As reported on Reynolds’ website, and confirmed via personal communication with a customer service representative

45 of two hours. Though the temperature was not kept constant due to natural fluctuations, the temperature of the can was assessed every half hour using a handheld infra-red thermometer

(Etekcity Lasergrip 800) and recorded (Table 1). As the thermometer was only able to record the surface temperature of the outer can, the recorded values do not represent the actual fire temperatures, but rather their minimum values. After two hours, no additional kindling was added, and the fire was left to cool. After one hour of cooling, the can’s temperature was measured, the can was removed from the heating area, and the remaining coals were drenched in water to prevent further combustion. The can was left in the pit for 13 hours

(overnight) to cool completely and prevent combustion upon its opening. After complete cooling, the material inside the foil was recovered.

Figure 11. Charcoal production in the "hearth" simulation. Wood is placed within two cans, and heated in the bowl-style fire pit.

Charcoal Production Temperature (°C)

450 C) ° 400

350

300 0 30 60 90 120

Temperature ( Temperature Time Elapsed (minutes)

Table 1. Hearth temperature during charcoal production.

The resulting charcoal fragments were observed under a Meiji EMZ-8TR light microscope at various magnifications for confirmation of complete carbonization. The fragments were also photographed under magnification for future assistance in identification of unknown acacia species (Figure 12).

a b c

Figure 12. Experimental materials: a) untreated V. nilotica; b) V. nilotica carbonized in the hearth simulation; and c) V. nilotica carbonized under laboratory conditions. For the untreated wood used in the laboratory, see Figure 10.

87Sr/86Sr Analysis Archaeological samples from Meroe were selected for Sr analysis based on two criteria: excavation contexts (e.g. stratified in area, in furnace dumps, in iron slag, etc.) and condition of the sample. To incorporate a temporal component to this study, 10 samples were chosen from metallurgical contexts during two contrasting periods—MIS 3, used during Meroe’s apex (400-200 BCE) and MIS 6, used during the city’s decline (200-400 CE). Additionally, samples were preferentially chosen for Sr analysis from samples in too poor a condition to be assessed under a light microscope. Analyzing additional samples to form a larger dataset was not possible due to the expensive nature of 87Sr/86Sr analyses.

100mg was separated from each selected sample and sent to the University of Arizona’s

Arizona Laboratory for Emerging Contaminants for analysis. The samples were processed on a VG Sector 54 multicollector thermal ionization mass spectrometer (TIMS) in dynamic collection mode. Sample preparation from Benson et al (2010) was used prior to 87Sr/86Sr analysis to eliminate trace metal contaminants that may have accrued in the charcoal during the carbonization process. The samples were run in 4 batches over the course of one year, though using the same laboratory equipment. Remaining fragments large enough to be assessed under light microscope were compared visually, especially for the purposes of detecting any physical change to the wood structure during carbonization.

Carbon Isotope Analysis Carbon isotope (δ13C) analyses were conducted on both experimental and archaeological charcoal in the same manner as the samples processed for 87Sr/86Sr analysis. When possible,

48 the same samples were used for both analyses. The necessity of analyzing the experimental charcoal arises from the known variability of δ13C within species, and to assess possible effects of carbonization on carbon isotopes.186 Experimental charcoal fragments from each sample were chose based on mass value, totaling 1mg. Like the strontium isotope analysis, archaeological samples for δ13C analysis were also selected based on archaeological context.

In addition to the metallurgical material from Meroe, non-metallurgical charcoal filling the gap (ca. 300 – 100 BCE) was included in the carbon isotope analyses to create a more continuous chronological sequence. Fragments of uncontaminated areas measuring 100mg were manually removed each sample. Roughly 1mg portions was removed from each of these fragments and analyzed in the Environmental Isotopes Lab at the University of

Arizona. Carbon content, as well as δ13C, were measured on a continuous-flow gas-ratio mass spectrometer (Finnigan Delta PlusXL). Samples were combusted using an elemental analyzer (Costech) coupled to the mass spectrometer. Standardization is based on acetanalide for elemental concentration, NBS-22 and USGS-24 for δ13C. Precision is better than ± 0.06 for δ13C, based on repeated internal standards.

186 Leavitt 2010

Results

Experimental Charcoal Analysis The GZL 1 group produced 87Sr/86Sr ratios from 0.71715 – 0.71740 (Figure 13). The GZL 2 group produced 87Sr/86Sr ratios from 0.71678 – 0.71692. 87Sr/86Sr analysis of the GZL 1 and

GZL 2 groups determined that the untreated wood and samples heated in the lab at 900°C produced ratios within the ±0.0002 instrumental error range (

Table 2). Samples heated in a hearth environment produced 87Sr/86Sr values increased by

0.25 in the GZL 1 group, and 0.15 in the GZL 2 group.

87Sr/86Sr variation during charcoal formation 0.7175 0.7174 0.7173

0.7172 Sr Sr

86 0.7171 Sr/

87 0.717 0.7169 0.7168 0.7167 Wood 400°C 900°C GZL1 GZL2

Figure 13. Results from Sr isotope analysis on untreated V. nilotica, charcoal produced near 400 degrees Celsius, and charcoal produced under laboratory conditions at high heat. Material from two trees (GZL 1 and GZL 2) are represented here. Error bars are representative of instrumental error. (n=6)

Sample ID Context 87Sr/86Sr GZL 1 Wood 0.71719 GZL 1H ~400°C (hearth) 0.71740 GZL 1L 900°C (lab) 0.71715 GZL 2 Wood 0.71678 GZL 2H ~400°C (hearth) 0.71692 GZL 2L 900°C (lab) 0.71678

Table 2. Experimental charcoal strontium isotope results. Samples with suffix “H” represent charcoal created in a simulated hearth. Samples with suffix “L” represent charcoal produced in a laboratory setting.

Carbon isotope (δ13C) analyses on the GZL 1 group produced values from -28.2±0.06؉

) (VPDB) to -28.4±0.06؉ (VPDB)

- Table 3). Analyses of the GZL 2 group produced δ13C values from -26.4±0.06؉ (VPDB) to

VPDB)؉. The GZL 2 group present with differences outside of the ±0.06) 27.9±0.06 instrumental error range. Samples heated in a hearth environment produced lower δ13C values than the untreated wood and the laboratory-created charcoal (Figure 14. Results from carbon isotope (δ13C) analysis on untreated V. nilotica, charcoal produced near 400 degrees Celsius, and charcoal produced under laboratory conditions at high heat. Material from two trees (GZL 1 and GZL 2) are represented here.

Error bars represent instrumental error.).

δ13C variability during charcoal formation -25 -25.5 -26 -26.5

VPDB) -27 ؉

C ( C -27.5 13

δ -28 -28.5 -29 Wood 400°C 900°C

GZL1 GZL2

Figure 14. Results from carbon isotope (δ13C) analysis on untreated V. nilotica, charcoal produced near 400 degrees Celsius, and charcoal produced under laboratory conditions at high heat. Material from two trees (GZL 1 and GZL 2) are represented here. Error bars represent instrumental error. (n=6)

(Sample ID Context δ13C (؉ VPDB GZL 1 Wood -28.2 GZL 1H ~400°C (hearth) -28.4 GZL 1L 900°C (lab) -28.2 GZL 2 Wood -26.4 GZL 2H ~400°C (hearth) -27.9 GZL 2L 900°C (lab) -27.0

Table 3. Carbon isotope (δ13C ) results for experimental charcoal. Samples with suffix “H” represent charcoal created in a simulated hearth. Samples with suffix “L” represent charcoal produced in a laboratory setting.

Archaeometallurgical Charcoal:87Sr/86Sr 87Sr/86Sr ratios for the entire data set range from 0.70724 – 0.70785, with a mean of 0.70738, where σ = 0.00019 (

Table 4). 87Sr/86Sr ratios for the early metallurgy (MIS 3) group range from 0.70724 – 0.70737, with a mean of 0.70731, where σ = 0.00004. Comparatively, the late metallurgy (MIS 6) have a smaller range, from 0.70724 – 0.70742, with a mean of 0.70728, where σ = 0.0001. A one- way analysis of variance (ANOVA) test where p ≤ 0.05 determined that the early metallurgy and late metallurgy 87Sr/86Sr values were not statistically significant, with a p-value of 0.8513.

(

Figure 15).

Meroe Metallurgical Charcoal: 87Sr/86Sr

0.7067 0.7069 0.7071 0.7073 0.7075 0.7077 87Sr/86Sr

Early Metallurgy Late Metallurgy c. 400-200 BCE c. 200-400 CE

Figure 15. Strontium isotope analysis on early-period and late-period iron smelting contexts at Meroe, Sudan. Error bars are representative of instrumental error. (n=10)

Sample ID Context 87Sr/86Sr MIS 3-11-16 Superpositioned rubble 0.70724 MIS 3-3-16 Metallurgy: OS 0.70735 MIS 3-5-16 Metallurgy: OS017 0.70731 MIS 3-8-16 Metallurgy: OS004 0.70727 MIS 3-10-16 Metallurgy: OS021 0.70737

MIS 6-3-14 Slag heap: OS6276 0.70742 MIS 6-3-14 Slag heap: OS6180B 0.70724 MIS 6-1-14 Workshop FW224 0.70724 MIS 6-1-14 Workshop FW240 0.70725 MIS 6-1-14 Workshop FW233 0.70725

Table 4. Strontium isotope results for archaeological material from Meroe, Sudan. MIS 3 (400 - 200 BCE) and MIS 6 (200 - 400 CE) represent different iron smelting periods.

Archaeological Charcoal: δ13C The δ13C (VPDB) values for all archaeological charcoal fragments ranges from -26.9±0.06؉

- to -24.6±0.06؉ (Table 5). The early metallurgical material ranges from -26.9±0.06؉ to

-24.9±0.06؉, with a mean of -25.98±0.06؉ where σ=1.0035 (Figure 16). The intermediate non

metallurgical ranges from -26.3±0.06؉ to -25.5±0.06؉ with a mean of -25.9667±0.06؉ where

σ=0.4163. The late metallurgical material ranges from -26.9±0.06؉ to -24.6±0.06؉, with an

average of -25.72±0.06؉ where σ=0.9257. These values are typical, as δ13C values in woody

tissue are usually around -30؉ (VPDB).187 A one-way ANOVA test (p < 0.05) was performed on these data, indicating that the difference between the three groups is statistically insignificant, with a p-value of 0.880908.

Archaeological Charcoal from Meroe: δ13C

-30 -29 -28 -27 -26 -25 (δ13C ؉ (VDPB Early metallurgy Building timber Late metallurgy 400 - 200 BCE 300 BCE - 100 CE 200 - 400 CE

Figure 16. δ13C (VPDB) values from three contexts at Meroe: early metallurgy, building timber, and late metallurgy. See Table 4 for exact values. Error bars are representative of instrumental error. (n=13)

187 McCarroll & Loader, 2004

Sample ID Context δ13C ‰ MIS 3-9-16 Metallurgy: OS009B -26.8 MIS 3-4-16 Metallurgy: OS015B -24.9 MIS 3-7-16 Metallurgy: OS019 -24.9 MIS 3-3-16 Metallurgy: OS033B -26.9 MIS 3-3-16 Metallurgy: OS034 -26.4 MIS 3-2-14 Temple: pit fill, OS030 -25.5 MIS 3-2-14 Temple: pit fill, OS031 -26.3 MIS 3-2-14 Temple: pit fill, OS032 -26.1

MIS 6-1-14 Workshop FW224B -24.6

MIS 6-3-14 Slag heap: OS6276 -25.5

MIS 6-1-14 Workshop FW233A -25.2 MIS 6-1-14 Workshop FW233B -26.9 MIS 6-1-14 Workshop FW240B -26.4

Table 5. δ13C (VPDB) values for archaeological material from Meroe. The materials include archaeometallurgical material from MIS 3 and MIS 6, as well as possible building material from MIS 3 post-dating iron production.

Discussion Analyses on both experimentally-produced and archaeological charcoal gave rise to a series of interesting implications. Methodological, environmental, and archaeological hypotheses are discussed below.

87Sr/86Sr Stability during Acacia Charcoal Production The 87Sr/86Sr ratios derived from experimental charcoal remain relatively stable between untreated wood, experimental hearth charcoal, and laboratory-created charcoal. This is especially apparent when comparing the untreated wood to the laboratory-created material.

For example, the difference in ratios between GZL 1 and GZL 1L is 0.0004, which is within the margin of error for the TIMS analysis. Further, the analogous GZL 2 and GZL 2L presented identical 87Sr/86Sr ratios. Despite the small sample size (n = 6), these preliminary results suggest that temperatures achievable by ancient peoples may not impact 87Sr/86Sr values.

However, the samples created in the simulated hearth environment, GZL 1H and GZL 2H, produced 87Sr/86Sr ratios greater than their untreated wood counterparts by 0.0021 and

0.0014, respectively. These differences fall outside of the range prescribed for instrumental error, suggesting that these increased values are genuine, rather than artifacts of instrumental analysis. The increased 87Sr/86Sr ratios for GZL 1H and GZL 2H suggest that the environment in which the wood was heated impacted the 87Sr/86Sr ratios present in the wood before carbonization. This realization introduces the possibility that a range of burning environments, such as those found in ceramic kilns, domestic fires, iron production

57 contexts, or destruction sequences may artificially increase or decrease 87Sr/86Sr ratios for the woody material found therein.

Future study may have the potential to further hone the technique and better understand

87Sr/86Sr variation based on burning environment. Nonetheless, the confirmed stability of

87Sr/86Sr ratios derived from the experimental charcoal material creates a powerful, new method of analysis for woody material. Based on the 87Sr/86Sr ratios obtained from this study, it is clear that pyrolysis and carbonization of woody material alone:

• Do not eliminate Sr from carbonized material

• Do not preferentially remove one strontium isotope from the plant’s cellular

structure, enabling the 87Sr/86Sr ratios in carbonized wood to be treated as valid.

Carbon Isotope (δ13C) Stability during Acacia Charcoal Production Much like the Sr isotope data, the (δ13C) values show little variation during the charcoal formation process. GZL 1 group charcoal especially shows a linear consistency in the δ13C values across each treatment, with δ13C values from GZL 1 and GZL 1L both being -

28.2±0.06؉, and GZL 1H decreasing by only 0.02؉. GZL 2 group charcoal displays more variance, though with a similar pattern. As with the 87Sr/86Sr analysis, results from GZL 1H and GZL 2H deviate from the untreated wood and laboratory-produced charcoal (this time decreasing), suggesting again impacts from an uncontrolled burn environment.

The small sample size of this study prohibits the results from being applicable across other species or regions. Carbon isotope (δ13C) variation during carbonization is still poorly-

58 understood, and requires additional study across species.188,189 Further, as δ13C values may vary between growth rings (in species that produce annual rings) or growth regions as described in the literature review, bulk sampling as conducted here may present biases depending on the exact charcoal fragment analyzed. Despite this, the range of values for the

entire data set is -26.3؉ (VPDB) to -28.5؉ (VPDB), with a difference of only 2.2؉. Given

that the within-tree variability of carbon isotope values may be as much as 1 – 3؉, the experimental data suggests that the archaeological data may be treated as acceptable for this study on Vachellia nilotica.190 Based on the analyses conducted for this study, the stability of carbon isotope (δ13C) values across V. nilotica charcoal samples suggests that

• δ13C values do not seem significantly altered during charcoal formation.

Archaeological Charcoal: 87Sr/86Sr Analysis Based on the statistical insignificance (σ = 0.00019) between archaeometallurgical charcoal from MIS 3 (early metallurgy group) and MIS 6 (late metallurgy group), woody material from both groups likely came from the same location, or a location where widely similar climate patterns and geologies prevailed. Lack of variation in carbon isotope (δ13C) data (see below) supports this hypothesis, suggesting no major difference in aridity between 400 BCE – 400

CE. Determining the shared, exact origin of the fragments from both groups is not possible, due to a lack of extant 87Sr/86Sr data from the Meroe region, as travel to the site to collect both modern wood and sediment samples for compatible was not accomplishable (see

“Methods”).

188 Poole et al 2002 189 Jones et al 1993 190 Leavitt 2010

However, published 87Sr/86Sr data on the Nile system as a whole can help narrow the Meroe charcoal’s origin to a regional level. The frequency of strontium isotope analyses on Nile sediments is increasing, creating a larger dataset for future studies. Though 87Sr/86Sr data at several points along the Upper Nile offer a preliminary examination into the location of the charcoal from Meroe,191 Woodward et al. (2015) provides the most appropriate comparison for the Meroe dataset, as it offers contemporaneous material—this is critical, as aelion sand input into Nile sediment has increased over time since the end of the African Humid

Period.192

Upper Nile Valley 87Sr/86Sr

0.703 0.705 0.707 0.709 0.711 87Sr/86Sr Egyptian Nile Dongola Reach Atbara Meroe Blue Nile White Nile

Figure 17. Upper Nile Valley 87Sr/86Sr values at various locations, including: White Nile, Blue Nile, Atbara, and Egyptian Nile (Padoan et al 2011); Dongola (Woodward et al 2015); and Meroe (this study). Error bars for all material are representative of this study’s instrumental error for efficient comparison. Note that only the Dongola Reach material is contemporaneous with the Meroe data. Meroe data has an n-value of 10; n-values for other sets may be found in their respective publications.

191 Padoan et al 2011 192 Woodward et al 2015

Meroe lies on the stretch of the Nile north the confluence of the Blue and White Niles near

Khartoum, before the Atbara tributary joins the river (Figure 3). When the Meroe 87Sr/86Sr dataset is compared against 87Sr/86Sr data from various regions in the Upper Nile, the prevalent hypothesis in past research193,194 suggesting that the smelting fuel originated locally can be supported. 87Sr/86Sr from Meroe fall between those of the Blue Nile (0.7050 – 0.7054) and White Nile (0.7103 – 0.7113), somewhat equidistant between the 87Sr/86Sr ratios for the

Atbara (0.7040 – 0.7046) and the White Nile.195 (

Figure 17) These relationships suggest that the Meroe material grew in a watershed with a combined input of the White Nile and the Blue Nile. Further, the 87Sr/86Sr analysis on the

Meroe material supports a contribution of White Nile discharge larger than measured today.196

It is possible that the Meroe material may have grown downstream of the adjoining of the

Atbara’s waters with the Nile, in territory perhaps under Meroe’s control. However, more comprehensive study of the strontium geochemistry along this stretch of river must be completed to fully understand the Meroe material. Contemporaneous sample sites would also benefit this effort, ensuring that modern aeolian sediment increase would not skew the

87Sr/86Sr values. Using only the data measured in this study and available published records,

87Sr/86Sr analysis of all Meroitic charcoal described indicates that:

193 Wolf 2015 194 Humphris & Eichhorn, forthcoming 195 Padoan et al 2011 196 Woodward et al 2015

• Metallurgical contexts deposited at Meroe from 400 – 200 BCE and 200 – 400 CE

contain charcoal that likely came from the same region.

• The origin of metallurgical material may have extended north of the Atbara

confluence, but most probably no further north than the confluence of the Blue and

White Niles.

Archaeological Charcoal: Carbon Isotope (δ13C) Analysis Carbon isotope (δ13C) data can reveal variations in local aridity and amount of carbon

197 13 dioxide (CO2) in the atmosphere. In the case of plant matter, δ C values can broaden understanding of aridity in a given region. With favorable moisture conditions, photosynthesis preferentially uptakes the lighter isotope of carbon.198 However, during a particularly dry season, plant stomata open less frequently and the δ13C values in the tissue increase (Leavitt, 2010). Archaeologists studying past peoples’ relationship with their environment increasingly apply δ13C analysis (along with that of δ18O) to better understand past climates in their region of interest.199 This method can be applied to archaeometallurgical charcoal at Meroe.

The results of the one-way ANOVA test indicate that the variability in δ13C values between the three temporal groups is not statistically significant. A change in aridity is not supported during these three moments in time—change between the two dates is impossible to see from these temporally discontinuous data. The well-documented Ethiopian coast monsoon

197 Sharp 2006 198 Leavitt 2010 199 Pollard et al 2007

62 complicates this issue. Growth of this acacia species occurs in tandem with periods of heavy rainfall.200 However, the fragmentary nature of the sample set leaves the growth seasons represented unknown. Given the tight range of δ13C values, it is possible that this cyclic seasonality of the Ethiopian monsoon was not appropriately captured. Broadening the data set to include a longer period of time and a larger number of samples would be beneficial in building an environmental record more representative of the full possible range of δ13C values.

Interestingly, the δ13C variability within the samples from the intermediate non-metallurgical context are very low. Variability of δ13C values depends primarily on the species, but on

average ranges from 1-3؉(VPDB) within a single tree.201 Given that the range of variability

for these samples is at most 2؉, that the samples were excavated in very close proximity, and what can be seen of their anatomical features, it seems very likely that these charcoal timber fragments may have come from the same tree. Further investigation into maximizing this kind of “tree-matching” as an application of carbon isotope analysis may have future promise in archaeological contexts. Other samples in the dataset also have very similar (or in some cases, identical) δ13C values, but these samples were excavated in separate, discrete areas, making a common source tree less likely.

As there are not yet equivalent carbon isotope data or paleoclimate records for the study period and region, data from this study can only be interpreted in relationship to itself.

200 Khan 1970 201 Leavitt 2010

Further, methodological concerns regarding bulk sampling (as detailed above) may have biased the results depending on the part of the tree represented in the charcoal. However, the δ13C values derived from the metallurgical charcoal at Meroe supports the conclusions that:

• δ13C values between metallurgical charcoal deposited from 400 – 200 BCE, non-

metallurgical material deposited from 300 – 100 BCE, and metallurgical charcoal

deposited from 200 – 400 CE do not vary significantly.

• Aridity during the above periods at Meroe did not significantly change.

Implications for Ancient Meroe

Wood Fuel Procurement The lack of climate variability at Meroe from 400 BCE – 200 CE as suggested by the δ13C values, combined with the likely uniform region of origin (likely local to Meroe or its northern environs), suggests that the quantity of acacia being felled for charcoal was coming from one area whose microenvironmental conditions were relatively homogenous. A modern study in ecology determined that Vechellia nilotica best thrive in dry forests that are disturbed by human activity.202 However, their adaptive capacity is great: V. nilotica can thrive on annual rainfall values ranging from < 230 mm/year to > 1500 mm/year.203 Ecological study of V. nilotica trees reveals that young saplings may take up to 12 years to reach maturation under stressed conditions, while the mature tree may live to 25-60 years of age.204

These specific values vary depending on the growth environment of the specific tree—for

202 Khurana & Singh 2004 203 Kriticos et al 2003 204 Kriticos et al 1999

64 example, trees growing on the river edge would likely reach maturity more quickly, in as little as 5 – 7 years.205,206

Whether the people of ancient Meroe were employing forest management techniques for their acacia stands is also unknown. Coppicing, the act of trimming a tree down to a live stump, results in the growth of several quickly-growing saplings emanating from the same root system.207 It is possible that the acacia from ancient Meroe were products of coppicing to produce a greater yield, as in other acacia forest systems.208 However, as the charcoal from

Meroe has not yet been assessed for discernible presence of growth rings and V. nilotica life cycle data in undisturbed ecologies are relatively unknown, determining the validity of this hypothesis is difficult at present.

A Reduction or Cessation of Iron Production

Moving beyond the forest and towards the furnaces, human behaviors and their impacts for the city of Meroe have been revealed by dating of charcoal on site with a high stratigraphic resolution. Radiocarbon dating of the slag heaps at Meroe reveal a gap or reduction in iron smelting activity for approximately 200 years. Humphris and Scheibner209 were able to reconstruct the chronology of iron smelting at Meroe and a nearby iron smelting site (Figure

18). Interestingly, there are two distinct groups of smelting episodes, with a gap of no

205 Khan 1970 206 Kriticos et al 1999 207 Levesque et al 2011 208 Iles 2016 209 Humphris and Scheibner 2017

65 additional slag deposition for approximately 200 years.210 Following this pause in production, iron smelting deposition events at Meroe continue only for roughly an additional 200 years at MIS 6 and another substantial slag heap on the north side of the site (Jane Humphris, personal communication). Humphris and Scheibner posit that a period of lessened military exploits may have been a factor in this cessation or reduction of iron smelting, especially given the wealthy nature of Meroe during this time.211 While the absence of data should not be construed as its nonexistence, it will be worth noting in future research whether this gap remains.

210 Earlier radiocarbon dates (Shinnie and Bradley 1980) suggest that there was smelting/smithing activity occurring in the first few centuries CE. However, the sampling strategy for these dates is somewhat unreliable (Jane Humphris, personal communication) and not considered here. 211 Humphris and Scheibner 2017

Figure 18. Radiocarbon dates from Meroe and Hamadab, from Humphris & Schreiber 2017. Both MIS 3 and MIS 6 are notated on the top, horizontal axis.

Study of the technical ceramics212 used at Meroe may support the existence of this pause in intensive iron production. The study notes that the standardization of raw materials for bricks, furnace materials, and tuyères remained relatively constant during Napatan and early

Meroitic times—more or less contemporaneous to the features at MIS 3.213 However, late

212 Functional ceramic forms used in the production of iron, such as furnace components, tuyères (blowpipes), and bricks. 213 Ting & Humphris 2017

67 and Post-Meroitic technical ceramic production efforts are much less organized, with a greater degree in variability of raw materials. The deviation following the gap in the radiocarbon record indicates some manner of change at Meroe; Ting & Humphris posit change in administrative control over the iron smelting activities. It is also possible that this shift in organization came about as an artifact of a cease or reduction in iron production of over 200 years.

If there truly was a halt or reduction in smelting at Meroe, as discussed previously, exhaustion of the easily-accessible acacia populations along the Nile may have been a contributing factor. Prior to the pause, evidence for iron production at Meroe has been dated from 800 BCE – 200 BCE. Six hundred years of consistent resource pillaging (with an additional level of intensity from approximately 400 BCE – 200 CE) would have necessitated a substantial period for sufficient regrowth to occur. Given the intentional selectivity214 of V. nilotica as the prime fuel for iron smelting and the long wait for tree maturity, a downscaling of wood collection may have been necessary to allow sufficient regrowth of the acacia stands such that their harvest would be ample enough to yield sizable fuel amounts for smelting.

Environmental Considerations If anthropogenic exhaustion of the woody resources around Meroe did not occur, environmental instability may have caused sufficient disruption in the environment to slow the regrowth of in V. nilotica in the areas immediately surrounding the city, without depleting key agricultural crops serviced by irrigation. Several studies seem to support this possible paleoclimatic shift in regions in the Nile valley, as well as the Mediterranean and Near East.

214 Humprhis & Eichhorn, forthcoming

Though the shifts vary in direction and duration throughout geographic space, there seems to be an agreement that the period from 200 BCE – 200 CE was a time of quickly changing or oscillating climate.215

Because of current gaps in data, confident interpretations of climate disruption around

Meroe are difficult to obtain. Despite this dearth of information, climate data from different hydrological zones of the Nile may offer a punctuated image of ancient environmental conditions at Meroe. A sediment core taken at the White Nile’s source, Lake Victoria, reveals an increase in δ18O values from 3000 – 2000 cal BP, suggesting evaporation events and lowered lake levels concurrent with the start of Meroitic iron smelting around 800 BCE until the first century CE.216 Terrestrial leaf wax quantities in the same study reveal a decrease in plant matter entering the lake during the same time period, indicating less terrestrial growth.

Further, study of strontium isotope variation in the Mediterranean Nile fan suggests that, between 3000 – 2000 cal BP, discharge from the White Nile decreased at a rapid rate, while

Blue Nile discharge slowly increased.217 This change in river flux, especially if the Blue Nile did not increase its discharge to meet the disparity brought by White Nile reduction, may have decreased the Nile’s water levels near Meroe. In turn, this may have resulted in a smaller crop of quickly-growing, river-fed acacia.

215 Marston 2017 216 Cockerton et al 2015 217 Box et al 2011

Terrestrial records also offer supporting paleoclimate data. Speleothem records from the

Mediterranean confirm a sustained drier period from ca. 600 – 1 BCE, with an increase of moisture occurring beginning around 100 CE.218 Paleoclimate records based in Ethiopia also suggest that the land’s stability faltered from ca. 1500 – 500 BCE in the region, with tree-to- shrub ratios decreasing severely from ca. 1000 – 300 BCE, and only beginning to increase again near 100 CE—a very similar story to that found in the Mediterranean.219 However, other factors beyond Nile discharge may instead have been a cause for reduced acacia growth. For example, V. nilotica individuals whose limiting growth factor is moisture availability are less likely to flower and fruit during the spring.220 Minor fluctuations in local temperature, if present, could have instigated a decreased growth rate in the nearby acacia stands, limiting wood harvest.

These studies may support a drier or warmer climate at ancient Meroe between 1000 – 1

BCE. While impacts of a possible warm period have not yet been assessed through an archaeological lens, the amount and high quality of material culture during this time in the

Meroitic kingdom does not suggest unease in the elite members of this society.221 Perhaps the effects of a warm/dry period were not sever enough to impact faster-growing crops with available water from the Nile floodplain, but rather only the plants dependent on seasonal wadi flow or higher-than-normal floods. If this were the case, such a scenario would address the relative sociopolitical wealth of the city at this time, while also explaining the decrease or

218 Psomiadis et al 2018 219 Lanckriet et al 2017 220 Khan 1970 221 Shinnie and Bradley 1980

70 halt of acacia harvesting for the purpose of fueling iron furnaces. Further study on the paleoclimate surrounding Meroe is imperative if this story is to unfold.

Conclusions

Given the ubiquity of charcoal found at archaeological sites around the world, applying novel analytical techniques to anthracological study only increases the benefit of this powerful environmental, temporal, and anthropological multiproxy. Exploring 87Sr/86Sr analysis using charcoal creates a new opportunity for one such technique to be added to the anthracologist’s toolkit. Further, utilizing 87Sr/86Sr as an investigative method has the potential to unlock information on interactions between ancient peoples and their landscapes.

Experimental charcoal production using modern V. nilotica wood collected at al-Ghazali,

Sudan determined that 87Sr/86Sr values do not increase or decrease outside of the mechanical error of ±0.0002 for charcoal made in a laboratory environment under 900°C heat. This study is the first to explore 87Sr/86Sr applicability to charcoal, and its results are significant across disciplines: heat alone does not alter 87Sr/86Sr ratios found in wood. The preliminary success of this method opens the door to future development of strontium isotope studies using charcoal, a material apparent in a variety of analytical contexts.

Charcoal made in an experimental, hearth setting yielded 87Sr/86Sr values slightly higher than that of the original wood by 0.00014 – 0.00021. These data demonstrate that, though slightly biased due to external contamination, charcoal produced in uncontrolled environments can deliver approximate 87Sr/86Sr information. However, their similar increase indicates the

72 capability of direct burning conditions to alter 87Sr/86Sr found in wood—a hurdle that must be overcome to harness this technique to the fullest.

Using this 87Sr/86Sr geoprovenance technique, archaeometallurgical charcoal of two discrete periods from Meroe, Sudan can be potentially assigned to a growth region near Meroe no further south than the confluence of the Blue Nile and White Nile, perhaps reaching as far north as the Atbara-Nile confluence. Carbon isotope (δ13C) analysis suggests that the local growing conditions of trees made into these charcoal samples were not significantly different. Radiocarbon chronologies from the site suggest there was a reduction or cessation of iron smelting activity near Meroe from ca. 200 BCE – 200 CE. Interestingly, regional environmental records suggest that the Nile Valley may have been warmer or drier during the same time. Climatic conditions may have limited the growth of V. nilotica at the center of

Meroe’s acacia stands, resulting in the apparent gap in smelting activity.

Future Work A benefit to paleoenvironmental studies in Sudan must begin by widening the chronology of information towards the present day, to include development of cultures and increased human population sizes after the African Humid Period. Further development of a long, cross-regional paleoclimate record for Sudan is a key target for future investigations into long-term human ecology throughout the region. Increased sampling for 87Sr/86Sr geoprovenance analysis with the intent to form a more comprehensive comparative dataset for this Upper Nile Valley region will prove to be a useful tool. On a more practical scale, collecting wood samples from Meroe and its environs to create a larger, site-specific

73 spatially-resolved 87Sr/86Sr ratio network will improve the quality of this study. Further, assessment of 87Sr/86Sr ratio change at Meroe across time through sediment cores will help to facilitate a better understanding of local fluctuations in strontium isotope data.

Additionally, study of the physical properties of the Meroe charcoal may also address some of the gaps in the environmental record. Metrics typically used in reconstructing paleoforests may help shed light on this question. For example, the branch and stem diameters of the material could be used to assess patterns in method of wood fuel harvesting. Ethnographic evidence from central Africa indicates that branches are selected first under conditions of sufficient forest growth, while larger-diameter stumps are preferred during times of forest depletion.222 Searching for other markers of tree age at felling, such as sapwood, may also add to this study. The fact remains that there is still very much to learn about the process of iron production at ancient Meroe—information critical in understanding the human intersection with the environment. Continued study by the UCL Qatar team (as well as a continued pursuance of this project) may bring some of this knowledge to light.

Beyond site-specific data, the confirmation of 87Sr/86Sr geoprovenance as a valid technique to utilize during charcoal analysis may have a significant impact across fields. Most related to this work is the possibility of assessing fuel procurement strategies or import connections with archaeological materials. However, paleoclimatologists, ecologists, or dendrochronologists may find niche uses of this methodology to assign their charcoal to a geological space. Limitations to 87Sr/86Sr isotope studies are significant, encompassing the

222 Gelabert et al 2011

74 need for appropriate comparative material and the generally-high cost of analysis. As more study on the impact of different burn environments as trans-disciplinary collaborations increase, charcoal may begin to reach its highest potential as a lens for many questions—

87Sr/86Sr provenance now among them.

References

Abaker, W.E., Berninger, F. and Starr, M., 2017. Changes in soil hydraulic properties, soil moisture and water balance in Acacia senegal plantations of varying age in Sudan. Journal of Arid Environments.

Åberg, G., Jacks, G., Wickman, T. and Hamilton, P.J., 1990. Strontium isotopes in trees as an indicator for calcium availability. Catena, 17(1), pp.1-11.

Antal, M.J. and Grønli, M., 2003. The art, science, and technology of charcoal production. Industrial & Engineering Chemistry Research, 42(8), pp.1619-1640.

Babst, F., Poulter, B., Bodesheim, P., Mahecha, M.D. and Frank, D.C., 2017. Improved tree- ring archives will support earth-system science. Nat. Ecol. Evol, 1(0008).

Barakat, H.N., 1995. Middle Holocene vegetation and human impact in central Sudan: charcoal from the Neolithic site at Kadero. Vegetation History and Archaeobotany, 4(2), pp.101- 108.

Baillie, M.G., Hillam, J., Briffa, K.R. and Brown, D.M., 1985. Re-dating the English art- historical tree-ring chronologies. Nature, 315(6017), p.317.

Baton, F., Tu, T.T.N., Derenne, S., Delorme, A., Delarue, F. and Dufraisse, A., 2017. Tree- ring δ13C of archeological charcoals as indicator of past climatic seasonality. A case study from the Neolithic settlements of Lake Chalain (Jura, France). Quaternary International, 457, pp.50-59.

Benson, L.V., Taylor, H.E., Plowman, T.I., Roth, D.A. and Antweiler, R.C., 2010. The cleaning of burned and contaminated archaeological maize prior to 87Sr/86Sr analysis. Journal of Archaeological Science, 37(1), pp.84-91.

Bernabei, M. and Bontadi, J., 2012. Dendrochronological analysis of the timber structure of the Church of the Nativity in Bethlehem. Journal of Cultural Heritage, 13(4), pp.e54-e60.

Bert, D., Leavitt, S.W. and Dupouey, J.L., 1997. Variations of wood δ13C and water‐use efficiency of Abies alba during the last century. Ecology, 78(5), pp.1588-1596.

Beyin, A., Prendergast, M.E., Grillo, K.M. and Wang, H., 2017. New radiocarbon dates for terminal Pleistocene and early Holocene settlements in West Turkana, northern Kenya. Quaternary Science Reviews, 168, pp.208-215.

Bianchi, R.S., 2004. Daily life of the Nubians. Greenwood Publishing Group.

Billotte, N., Marseillac, N., Brottier, P., Noyer, J.L., Jacquemoud‐Collet, J.P., Moreau, C., Couvreur, T., Chevallier, M.H., Pintaud, J.C. and Risterucci, A.M., 2004. Nuclear microsatellite markers for the date palm (Phoenix dactylifera L.): characterization and utility across the genus Phoenix and in other palm genera. Molecular Ecology Resources, 4(2), pp.256- 258. Blanc-Jolivet, C., Kersten, B., Daïnou, K., Hardy, O., Guichoux, E., Delcamp, A. and Degen, B., 2017. Development of nuclear SNP markers for genetic tracking of Iroko, Milicia excelsa and Milicia regia. Conservation Genetics Resources, 9(4), pp.531-533.

Blondel, F., Cabanis, M., Girardclos, O. and Grenouillet-Paradis, S., 2018. Impact of carbonization on growth rings: dating by dendrochronology experiments on oak charcoals collected from archaeological sites. Quaternary International, 463, pp.268-281.

Box, M.R., Krom, M.D., Cliff, R.A., Bar-Matthews, M., Almogi-Labin, A., Ayalon, A. and Paterne, M., 2011. Response of the Nile and its catchment to millennial-scale climatic change since the LGM from Sr isotopes and major elements of East Mediterranean sediments. Quaternary Science Reviews, 30(3-4), pp.431-442.

Braadbaart, F. and Poole, I., 2008. Morphological, chemical and physical changes during charcoalification of wood and its relevance to archaeological contexts. Journal of archaeological science, 35(9), pp.2434-2445.

Brewer, P.W., 2016. Tellervo: A Guide for Users and Developers, http://dx.doi.org/10.5281/zenodo.51329, www.tellervo.org

Brooks, J.K., Bashirelahi, N. and Reynolds, M.A., 2017. Charcoal and charcoal-based dentifrices: A literature review. The Journal of the American Dental Association, 148(9), pp.661- 670.

Bunbury, J.M. and Lutley, K., 2008. The Nile on the move. Egyptian Archaeology, 32, pp. 3-5.

Butzer, K.W., 1968. Desert and river in Nubia. University of Wisconsin Press.

Buursink, J., 1971. Soils of central Sudan (Doctoral dissertation, Rijksuniversiteit te Utrecht).

Buzon, M.R. and Simonetti, A., 2013. Strontium isotope (87Sr/86Sr) variability in the Nile Valley: identifying residential mobility during ancient Egyptian and Nubian sociopolitical changes in the New Kingdom and Napatan periods. American journal of physical anthropology, 151(1), pp.1-9.

Caricchi, C., Vona, A., Corrado, S., Giordano, G. and Romano, C., 2014. 79 AD Vesuvius PDC deposits' temperatures inferred from optical analysis on woods charred in-situ in the

Villa dei Papiri at Herculaneum (Italy). Journal of Volcanology and Geothermal Research, 289, pp.14-25.

Chami, L., Hartl, D., Leboulleux, S., Baudin, E., Lumbroso, J., Schlumberger, M. and Travagli, J.P., 2015. Preoperative localization of neck recurrences from thyroid cancer: charcoal tattooing under ultrasound guidance. Thyroid, 25(3), pp.341-346.

Charlton, M. and Humphris, J., 2017. Exploring ironmaking practices at Meroe, Sudan—a comparative analysis of archaeological and experimental data. Archaeological and Anthropological Sciences, pp.1-18.

Cockerton, H.E., Street-Perrott, F.A., Barker, P.A., Leng, M.J., Sloane, H.J. and Ficken, K.J., 2015. Orbital forcing of glacial/interglacial variations in chemical weathering and silicon cycling within the upper White Nile basin, East Africa: Stable-isotope and biomarker evidence from Lakes Victoria and Edward. Quaternary science reviews, 130, pp.57-71.

Dai, Z., Webster, T.M., Enders, A., Hanley, K.L., Xu, J., Thies, J.E. and Lehmann, J., 2017. DNA extraction efficiency from soil as affected by pyrolysis temperature and extractable organic carbon of high-ash biochar. Soil Biology and Biochemistry, 115, pp.129-136.

Deguilloux, M.F., Pemonge, M.H. and Petit, R.J., 2004. DNA-based control of oak wood geographic origin in the context of the cooperage industry. Annals of Forest Science, 61(1), pp.97-104.

Deguilloux, M.F., Pemonge, M.H., Bertel, L., Kremer, A. and Petit, R.J., 2003. Checking the geographical origin of oak wood: molecular and statistical tools. Molecular Ecology, 12(6), pp.1629-1636.

Douglass, A.E., 1929. Secret of the Southwest Solved by Talkative Tree Rings. National Geographic Magazine.

Drake, B.L., Hanson, D.T. and Boone, J.L., 2012. The use of radiocarbon-derived Δ13C as a paleoclimate indicator: applications in the Lower Alentejo of Portugal. Journal of Archaeological Science, 39(9), pp.2888-2896.

Drake, B.L., Wills, W.H., Hamilton, M.I. and Dorshow, W., 2014. Strontium isotopes and the reconstruction of the Chaco regional system: evaluating uncertainty with Bayesian mixing models. PLoS One, 9(5), p.e95580.

Dufraisse, A., Coubray, S., Girardclos, O., Dupin, A. and Lemoine, M., 2017. Contribution of tyloses quantification in earlywood oak vessels to archaeological charcoal analyses: Estimation of a minimum age and influences of physiological and environmental factors. Quaternary International, 463, pp.250-257.

Dumolin‐Lapègue, S., Pemonge, M.H., Gielly, L., Taberlet, P. and Petit, R.J., 1999. Amplification of oak DNA from ancient and modern wood. Molecular Ecology, 8(12), pp.2137-2140.

Eide, T., Hägg, T., Pierce, R. H., and Török, L (eds), 1998. Fontes Historiae Nuborium: Textual Sources for the History of the Middle Nile Region Between the Eight Century BC and the Sixth Century AD (Vol. III. From the First to the Sixth Century AD). Universitetei i Bergen.

Enache, M.D. and Cumming, B.F., 2009. Extreme fires under warmer and drier conditions inferred from sedimentary charcoal morphotypes from Opatcho Lake, central British Columbia, Canada. The Holocene, 19(6), pp.835-846.

English, N.B., J.L. Betancourt, J.S. Dean, & J. Quade. 2001. Strontium isotopes reveal distant sources of architectural timber in Chaco Canyon, New Mexico. Proceedings of the National Academy of Science, 98, 21. pp. 11891-11896.

Fahmi, M., 2017. Climate, trees and agricultural practices: Implications for food security in the semi-arid zone of Sudan. Tropical Forestry Reports.

Francis, J.E. and Poole, I., 2002. Cretaceous and early Tertiary climates of Antarctica: evidence from fossil wood. Palaeogeography, Palaeoclimatology, Palaeoecology, 182(1-2), pp.47-64.

Fritts, H.C., 1976. Tree Rings and Climate. Academic Press.

Fuller, D.Q., 2004. Early Kushite Agriculture: Archaeobotanical Evidence from Kawa. Sudan & Nubia Bulletin, 8, pp.70-74.

Fuller, D.Q. and Edwards, D., 2001. Medieval plant economy in Middle Nubia: preliminary archaeobotanical evidence from Nauri. Sudan & Nubia Bulletin, 5, pp.97-103.

Gabra, G. and Takla, H.N. eds., 2013. Christianity and Monasticism in Aswan and Nubia (Vol. 5). Oxford University Press.

Gelabert, L.P., Asouti, E. and Martí, E.A., 2011. The ethnoarchaeology of firewood management in the Fang villages of Equatorial Guinea, central Africa: implications for the interpretation of wood fuel remains from archaeological sites. Journal of Anthropological Archaeology, 30(3), pp.375-384.

Gorashi, A.R., 2001. State of Forest Genetic Resources. Forest Genetic Resources Working Papers. Rome.

Gordon, R.B. and Killick, D.J., 1993. Adaptation of technology to culture and environment: bloomery iron smelting in America and Africa. Technology and Culture, 34(2), pp.243-270.

Graustein, W.C., 1989. 87 Sr/86 Sr ratios measure the sources and flow of strontium in terrestrial ecosystems. In Stable isotopes in ecological research (pp. 491-512). Springer.

Graustein, W.C. and Armstrong, R.L., 1983. The use of strontium-87/strontium-86 ratios to measure atmospheric transport into forested watersheds. Science, 219(4582), pp.289-292.

Guiterman, C.H., Swetnam, T.W. and Dean, J.S., 2016. Eleventh-century shift in timber procurement areas for the great houses of Chaco Canyon. Proceedings of the National Academy of Sciences, 113(5), pp.1186-1190.

Haneca, K., Wazny, T., Van Acker, J. and Beeckman, H., 2005. Provenancing Baltic timber from art historical objects: success and limitations. Journal of Archaeological Science, 32(2), pp.261-271.

Hayashi, K. and Kimoto, H., 2016. Microbial contamination of ice machines is mediated by activated charcoal filtration systems in a city hospital. Journal of environmental health, 78(10), p.32.

Hennekam, R., Donders, T.H., Zwiep, K. and de Lange, G.J., 2015. Integral view of Holocene precipitation and vegetation changes in the Nile catchment area as inferred from its delta sediments. Quaternary Science Reviews, 130, pp.189-199.

Hoffman, K.M., Gavin, D.G., Lertzman, K.P., Smith, D.J. and Starzomski, B.M., 2016. 13,000 years of fire history derived from soil charcoal in a British Columbia coastal temperate rain forest. Ecosphere, 7(7).

Hughes, J.D. and Thirgood, J.V., 1982. Deforestation, erosion, and forest management in ancient Greece and Rome. Journal of Forest History, 26(2), pp.60-75.

Humphris, J., 2014. Post-Meroitic Iron Production: initial results and interpretations. Sudan and Nubia, 18, pp.121-129.

Humphris, J., and Eichhorn, B. Forthcoming. Fuel choice during long-term ancient iron production in Sudan.

Humphris, J. and Rehren, T., 2014. Iron production and the Kingdom of Kush: An introduction to UCL Qatar’s research in Sudan. Ein Forscherleben zwischen den Welten, pp.177- 190.

Humphris, J. and Scheibner, T., 2017. A new radiocarbon chronology for ancient iron production in the Meroe region of Sudan. African Archaeological Review, 34(3), pp.377-413.

Ikram, S., 2012. From food to furniture: animals in ancient Nubia. In Fisher, M., Lacovara. (eds), Ancient Nubia. American University in Cairo Press.

Iles, L., 2016. The Role of Metallurgy in Transforming Global Forests. Journal of Archaeological Method and Theory, 23(4), pp.1219-1241.

Innes, J.B. and Blackford, J.J., 2003. The ecology of late Mesolithic woodland disturbances: model testing with fungal spore assemblage data. Journal of Archaeological Science, 30(2), pp.185- 194.

Inomata, T., Triadan, D., MacLellan, J., Burham, M., Aoyama, K., Palomo, J.M., Yonenobu, H., Pinzón, F. and Nasu, H., 2017. High-precision radiocarbon dating of political collapse and dynastic origins at the Maya site of Ceibal, Guatemala. Proceedings of the National Academy of Sciences, 114(6), pp.1293-1298.

Jones, T.P., Scott, A.C. and Mattey, D.P., 1993. Investigations of “fusain transition fossils” from the Lower Carboniferous: comparisons with modern partially charred wood. International Journal of Coal Geology, 22(1), pp.37-59.

Kagawa, A., Kuroda, K., Abe, H., Fujii, T. and Itoh, Y., 2007. Stable isotopes and inorganic elements as potential indicators of geographic origin of Southeast Asian timber. In International Symposium on Development of Improved Methods to Identify Shorea Species Wood and its Origin (p. 39).

Kagawa, A. and Leavitt, S.W., 2010. Stable carbon isotopes of tree rings as a tool to pinpoint the geographic origin of timber. Journal of wood science, 56(3), pp.175-183.

Khan, M.W., 1970. Phenology of Acacia nilotica and Eucalyptus microtheca at Wad Medani (Sudan). Indian Forester, 96(3), pp.226-48.

Khurana, E. and Singh, J.S., 2004. Germination and seedling growth of five tree species from tropical dry forest in relation to water stress: impact of seed size. Journal of Tropical Ecology, 20(4), pp.385-396.

Kriticos, D.J., Sutherst, R.W., Brown, J.R., Adkins, S.W. and Maywald, G.F., 2003. Climate change and the potential distribution of an invasive alien plant: Acacia nilotica ssp. indica in Australia. Journal of Applied Ecology, 40(1), pp.111-124.

Kriticos, D., Brown, J., Radford, I. and Nicholas, M., 1999. Plant population ecology and biological control: Acacia nilotica as a case study. Biological Control, 16(2), pp.230-239.

Kulkarni, S.V., Gupta, A.K. and Bhawsar, S., 2017. Formulation and Evaluation of activated charcoal peel off mask. International Journal of Pharmacy & Life Sciences, 8(5).

Kung, H.C., 1972. A mathematical model of wood pyrolysis. Combustion and flame, 18(2), pp.185-195.

Lanckriet, S., Rangan, H., Nyssen, J. and Frankl, A., 2017. Late Quaternary changes in climate and land cover in the Northern Horn of Africa and adjacent areas. Palaeogeography, Palaeoclimatology, Palaeoecology, 482, pp.103-113.

Leavitt, S.W., 2010. Tree-ring C–H–O isotope variability and sampling. Science of the Total Environment, 408(22), pp.5244-5253.

Leavitt, S.W., Long, A. and Dean, J.S., 1985. Tree-ring dating through pattern-matching of stable-carbon isotope time series. Tree-Ring Bulletin.

Leney, L. and Casteel, R.W., 1975. Simplified procedure for examining charcoal specimens for identification. Journal of archaeological science, 2(2), pp.153-159.

Lévesque, M., McLaren, K.P. and McDonald, M.A., 2011. Coppice shoot dynamics in a tropical dry forest after human disturbance. Journal of Tropical Ecology, 27(3), pp.259-268.

Lowe, A.J. and Cross, H.B., 2011. The Application of DNA methods to Timber Tracking and Origin Verification. IAWA Journal, 32(2), pp.251-262.

Ludemann, T. and Nelle, O., 2017a. Anthracology: Local to Global Significance of Charcoal Science - Part I, Quaternary International, 457, pp.1-240.

Ludemann, T. and Nelle, O., 2017b. Anthracology: Local to Global Significance of Charcoal Science - Part II, Quaternary International, 458, pp.1-232.

Ludemann, T. and Nelle, O., 2018. Anthracology: Local to Global Significance of Charcoal Science - Part III, Quaternary International, 463, Part B, pp. 219-424.

Marguerie, D., 2011. Short tree ring series: the study materials of the dendro-anthracologist. SAGVNTVM Extra, 11, pp.15-16.

Marguerie, D. and Hunot, J.Y., 2007. Charcoal analysis and dendrology: data from archaeological sites in north-western France. Journal of archaeological science, 34(9), pp.1417- 1433.

Marston, J.M., 2009. Modeling wood acquisition strategies from archaeological charcoal remains. Journal of Archaeological Science, 36(10), pp.2192-2200.

Marston, J.M., 2017. Agricultural Sustainability and Environmental Change at Ancient Gordion: Gordion Special Studies 8 (Vol. 8). University of Pennsylvania Press.

Marston, J.M., Holdaway, S.J. and Wendrich, W., 2017. Early-and middle-Holocene wood exploitation in the Fayum basin, Egypt. The Holocene, 27(12), pp.1812-1824.

McCarroll, D. and Loader, N.J., 2004. Stable isotopes in tree rings. Quaternary Science Reviews, 23(7-8), pp.771-801.

McNeill, J. and Turland, N.J., 2011. Major changes to the code of nomenclature— Melbourne, July 2011. Taxon, 60(5), pp.1495-1497.

Nabil, M. and Coudret, A., 1995. Effects of sodium chloride on growth, tissue elasticity and solute adjustment in two Acacia nilotica subspecies. Physiologia plantarum, 93(2), pp.217-224.

Namaalwa, J., Hofstad, O. and Sankhayan, P.L., 2009. Achieving sustainable charcoal supply from woodlands to urban consumers in Kampala, Uganda. International Forestry Review, 11(1), pp.64-78.

Obłuski, A., Ochala, G., Bogacki, M., Wieslaw, M., Maslak, S., and Mahmoud, Z., 2015. Ghazali 2012: preliminary report. Polish Archaeology in the Mediterranean, 24(1), pp.431-442.

Okuno, M., Nagaoka, S., Saito-Kokubu, Y., Nakamura, T. and Kobayashi, T., 2017. AMS Radiocarbon Dates of Pyroclastic-Flow Deposits on the Southern Slope of the Kuju Volcanic Group, Kyushu, Japan. Radiocarbon, 59(2), pp.483-488.

Padoan, M., Garzanti, E., Harlavan, Y. and Villa, I.M., 2011. Tracing Nile sediment sources by Sr and Nd isotope signatures (Uganda, Ethiopia, Sudan). Geochimica et Cosmochimica Acta, 75(12), pp.3627-3644.

Paradis-Grenouillet, S. and Dufraisse, A., 2018. Deciduous oak/chestnut: Differential shrinkage of wood during charcoalification? Preliminary experimental results and implications for wood diameter study in anthracology. Quaternary International, 463, pp.258- 267.

Pessenda, L.C., Valencia, E.P.E., Aravena, R., Telles, E.C.C. and Boulet, R., 1998. Paleoclimate studies in Brazil using carbon isotopes in soils. In Environmental geochemistry in the tropics (pp. 7-16). Springer, Berlin, Heidelberg.

Pingree, M.R.A., 2017. Fire, Charcoal, and the Biogeochemistry of Carbon and Nitrogen in Pacific Northwest Forest Soils (Doctoral dissertation).

Pollard, A.M., 2007. Analytical chemistry in archaeology. Cambridge University Press.

Poole, I., F. Braadbaart, J.J. Boon, P.F. van Bergen. 2002. Carbon isotope changes during artificial charring of propagules. Organic Geochemistry, 33, pp. 1675-1681.

Psomiadis, D., Dotsika, E., Albanakis, K., Ghaleb, B. and Hillaire-Marcel, C., 2018. Speleothem record of climatic changes in the northern Aegean region (Greece) from the

Bronze Age to the collapse of the Roman Empire. Palaeogeography, Palaeoclimatology, Palaeoecology, 489, pp.272-283.

Rees, G., 2015. Verifying the declared origin of timber using stable isotope ratio and multi-element analyses (Doctoral dissertation, University of York).

Rehren, T., 2001. Meroe, iron and Africa. Antike Sudan, 12, pp.102–109.

Reynolds, A.C., Betancourt, J.L., Quade, J., Patchett, P.J., Dean, J.S. and Stein, J., 2005. 87Sr/86Sr sourcing of ponderosa pine used in Anasazi great house construction at Chaco Canyon, New Mexico. Journal of Archaeological Science, 32(7), pp.1061-1075.

Rich, S., 2013. Ship Timber as Symbol? Dendro-Provenancing and Contextualizing Ancient Cedar Ship Remains from the Eastern Mediterranean/Near East (Doctoral dissertation, KU Leuven).

Rich, S., Manning, S.W., Degryse, P., Vanhaecke, F. and Van Lerberghe, K., 2012. Strontium isotopic and tree-ring signatures of Cedrus brevifolia in Cyprus. Journal of Analytical Atomic Spectrometry, 27(5), pp.796-806.

Robin, V., Bork, H.R., Nadeau, M.J. and Nelle, O., 2014. Fire and forest history of central European low mountain forest sites based on soil charcoal analysis: The case of the eastern Harz. The Holocene, 24(1), pp.35-47.

Rösler, H., Bönisch, E., Schopper, F., Raab, T. and Raab, A., 2012. Pre-industrial charcoal production in southern Brandenburg and its impact on the environment. Landscape Archaeology between Art and Science, pp.167-178.

Rozanski, K., Araguás‐Araguás, L. and Gonfiantini, R., 1993. Isotopic patterns in modern global precipitation. Climate change in continental isotopic records, pp.1-36.

Ruelle, J., 2014. Morphology, anatomy and ultrastructure of reaction wood. In The biology of reaction wood (pp. 13-35). Springer, Berlin, Heidelberg.

Ryan, P., Cartwright, C.R. and Spencer, N., 2012, July. Charred macroremains (seeds, fruits) and phytoliths from villa E12. 10 at Amara West, a pharaonic town in northern Sudan. In Proceedings from the 7th International Workshop for African Archaeobotany Vienna (pp. 2-5). Barkhuis Publishing.

Sampsell, B.M., 2003. The Geology of Egypt: A Traveler's Handbook. Amer Univ in Cairo Press.

Schiffer, M.B., 1986. Radiocarbon dating and the “old wood” problem: the case of the Hohokam chronology. Journal of Archaeological Science, 13(1), pp.13-30.

Scott, A.C. and Damblon, F., 2010. Charcoal: Taphonomy and significance in geology, botany and archaeology. Palaeogeography, Palaeoclimatology, Palaeoecology, 291(1-2), pp.1-10.

Scott, A.C. and Glasspool, I.J., 2007. Observations and experiments on the origin and formation of inertinite group macerals. International Journal of Coal Geology, 70(1-3), pp.53-66.

Sharp, Z.D. 2006. Principles of Stable Isotope Geochemistry. University of New Mexico.

Shinnie, P.L., 1996, Ancient Nubia. Routledge.

Shinnie, P.L. and Chittick, H.N., 1961. Ghazali-A Monastery in the Northern Sudan. Sudan Antiquities Service. Occasional Papers, 5.

Shinnie, P.L. and Bradley, R.J., 1980. The capital of Kush (Vol. 1). Otto Harrassowitz Verlag.

Speer, J.H., 2010. Fundamentals of tree-ring research. University of Arizona Press.

Speirs, A.K., McConnachie, G. and Lowe, A., 2009. Chloroplast DNA from 16th century waterlogged oak in a marine environment: initial steps in sourcing the Mary Rose timbers. Archaeological science under a microscope: studies in residue and ancient DNA analysis in honour of Thomas H. Loy. Terra Australis, 30, pp.175-189.

Springuel, I. and Mekki, A.M., 1994. Economic value of desert plants: Acacia trees in the Wadi Allaqi Biosphere Reserve. Environmental Conservation, 21(1), pp.41-48.

Stähli, M., Finsinger, W., Tinner, W. and Allgöwer, B., 2006. Wildfire history and fire ecology of the Swiss National Park (Central Alps): new evidence from charcoal, pollen and plant macrofossils. The Holocene, 16(6), pp.805-817.

Sulieman, H.M., 2018. Exploring Drivers of Forest Degradation and Fragmentation in Sudan: The Case of Erawashda Forest and its Surrounding Community. Science of The Total Environment, 621, pp.895-904.

Ting, C. and Humphris, J., 2017. The technology and craft organisation of Kushite technical ceramic production at Meroe and Hamadab, Sudan. Journal of Archaeological Science: Reports, 16, pp.34-43.

Venclová, N. and Dreslerová, D., 2013. Iron production, settlement, and environment: a regional approach. In: Rausz, S., A. Colin. K. Gruel, I. Ralston, T. Dechezleprê, eds, L'âge du Fer en Europe. Mélanges offerts a Olivier Buchsenschutz, Bordeaux, Ausonius Éditions Mémoires 32, pp. 291 - 304.

Welsby, D.A., 2002. The medieval kingdoms of Nubia: pagans, Christians and Muslims along the Middle Nile. British Museum Pubns Ltd.

Wickens, G.E., 1995. Role of Acacia species in the rural economy of dry Africa and the Near East (No. 27). Food & Agriculture Organization.

Wilk, M., Magdziarz, A., Kalemba, I. and Gara, P., 2016. Carbonisation of wood residue into charcoal during low temperature process. Renewable Energy, 85, pp.507-513.

Williams, M.A. and Talbot, M.R., 2009. Late Quaternary environments in the Nile basin. In The Nile (pp. 61-72). Springer, Dordrecht.

Witt, G.B., English, N.B., Balanzategui, D., Hua, Q., Gadd, P., Heijnis, H. and Bird, M.I., 2017. The climate reconstruction potential of Acacia cambagei (gidgee) for semi-arid regions of Australia using stable isotopes and elemental abundances. Journal of Arid Environments, 136, pp.19-27.

Wolf, P., 2015. The Qatar-Sudan Archaeological Project–the Meroitic town of Hamadab and the palaeo-environment of the Meroe region. Sudan & Nubia, 19, pp.115-131.

Woodward, J., M. Macklin, L. Fielding, I. Millar, N. Spencer, D. Welsby, M. Williams. 2015. Shifting sediment sources in the world’s longest river: A strontium isotope record for the Holocene Nile. Quaternary Science Reviews, 130, pp. 124-140.

Worrall, G.A., 1957. A Simple Introduction to the Geology of the Sudan. Sudan Notes and Records, 38, pp.2-9.

Xiao, X., Haberle, S.G., Li, Y., Liu, E., Shen, J., Zhang, E., Yin, J. and Wang, S., 2017. Evidence of Holocene climatic change and human impact in northwestern Yunnan Province: High-resolution pollen and charcoal records from Chenghai Lake, southwestern China. The Holocene, 28(1), pp. 127-139.

Xu, J., Pan, B., Liu, T., Hajdas, I., Zhao, B., Yu, H., Liu, R. and Zhao, P., 2013. Climatic impact of the Millennium eruption of Changbaishan volcano in China: New insights from high‐precision radiocarbon wiggle‐match dating. Geophysical Research Letters, 40(1), pp.54-59.

Xu, D., Ding, T., Li, Y., Zhang, Y., Zhou, D. and Wang, S., 2017. Transition characteristics of a carbonized wood cell wall investigated by scanning thermal microscopy (SThM). Wood Science and Technology, 51(4), pp.831-843.

Yost, C.L., Jackson, L.J., Stone, J.R. and Cohen, A.S., 2018. Subdecadal phytolith and charcoal records from Lake Malawi, East Africa imply minimal effects on human evolution from the∼ 74 ka Toba supereruption. Journal of human evolution, 116, pp.75-94.