Populations of the Middle Nile: Using Bioarchaeological and Paleogenetic Analyses to

Understand Nubian Ancestry

by

Abagail M. Breidenstein

A dissertation submitted in partial fulfillment

of the requirements for the degree of
Doctor of Philosophy
(Anthropology)
in The University of Michigan
2019

Doctoral Committee:

Associate Professor Abigail W. Bigham, Chair

Associate Professor Maureen Devlin
Associate Research Scientist Geoff Emberling
Associate Professor John Kingston
Professor Janet Richards
 Abagail M. Breidenstein

[email protected]

ORCID iD 0000-0002-1675-8399

© Abagail M. Breidenstein 2019

 Dedication

For my parents

ii
 Acknowledgements

I would like to thank the various funding sources which supported this dissertation. This
includes the Rackham Gradate Program through fellowships, conference support, and research
funds, various UM institutes (International Institute, African Studies Center, Mary Sue Coleman
Scholarship Fund, Weiser Center for Emerging Democracies), the Anthropology Department
through teaching fellowships, departmental funding, and research support for field and lab
research, and the UROP program at UM. Lastly, thank you to the National Geographic Society
for their continuing support of my project.
This project spanned five countries, four labs, and countless collaborations that I am
incredibly grateful for their invaluable contributions. To begin, a very special thanks to the
villagers of El-Kurru and officials at the National Corporation for the Antiquities and Museums in
Khartoum, Sudan, including Fakhri Hassan Abdallah Hassan and Dr. Abdelrahman Ali for their
permission and help exporting skeletal remains as well as your continuing trust with such
important objects of their heritage. Additionally, I would like to thank Mohamed Saad for opening
the NCAM collection to my project and more notably, helping me so much when I lived in
Khartoum for two weeks. I am grateful to my International Kurru Archaeological Project Team
members, especially Tim, Martin, Jalina, Jack, and Nacho, who helped in the field and packed
up remains when time was running short. Thank you to my collaborators abroad: Robert Stark
(McMaster University), Robert Mahler, Magna Srienc, and Dr. Mahmoud El-Tayeb (PCMA) for
taking a chance on a piloted paleogenomics project. I am so grateful for the donation of samples
and look forward to our work in the future.
To the three different labs that I have worked in throughout my paleogenomics training, I am
so grateful for your generosity and time. Firstly, I am especially grateful to Verena
Schuenemann and Frank Rühli for supporting my project, being immensely open to
collaborations, and our future work together. Thank you to the Archaeogenetics Group at the
University of Tübingen and Ella Reiter for teaching me enrichment techniques. And a big thank
you to Bastiaan Starr at the University of Oslo for introducing me to the Bone Crusher, then to
Agata Gondek and Giada Ferrari for our marathon lab sessions in Blindern. Then, many thanks
to Milford Wolpoff for letting me take over his Paleoanthropology lab at UM for a year to analyze

iii
the Nubian remains. And thank you to the students over than years that helped at various
stages of this project: Renee, Julian, Serena, and especially Zenta and Abby for their hard work
and enthusiasm for my project through UROP.
A special thank you to my mentor Geoff. We met so many years ago when I was a docent at
the Kelsey Museum, before I even started graduate school, and I never imagined I would be
able to do archaeological field work as part of my dissertation project. I am so honored you
brought me on your team to El-Kurru, supported me through two field seasons, showed me the
ropes, and laid the foundation for work beyond El-Kurru. Your research implementing
community engagement is inspiring and I hope to draw on your example for the Nuri Project. I
sincerely look forward to our work in the future.
Thank you to the Ancient Biomolecules Group at UZH, both times I was part of this team.
Round one including Giada, Claudia, Sabrina, and Abi for their training, guidance, side-jobs at
cheese shops, and being so open to have a new member of their team. You truly made Zürich
feel like home. Then round two: Gül, Judith, Enrique, and Christian. I would not have made it to
the finish line without your invaluable help, from finishing my sequencing lanes to drilling petrous
bones to running bioinformatic pipelines. I am excited we have the opportunity to continue
working together in Zürich.
A sincere thanks to my dissertation committee members for their advice, scientific or
otherwise, on navigating the graduate program and most importantly, becoming a better
instructor and scientist. A special thank you to my chair advisor, Abby, for years of support
before and during graduate school. I am grateful for the environment you cultivated to provide
me the freedom to explore my own project and grow as a scientist, gaining a multitude of skills
and independence along the way, and weathering the troubles with a level-head I still attempt to
emulate.
To my Michigan friends, Ann Arbor has been my home for well over a decade. The friends I
have made are more than friends, you are family. My sincere and heartfelt thanks to Amanda
and family, Jessie, Celia, Sylvia, Jack, Julian, Danielle, Emma & Henry, Liz, Vince, Elspeth,
Bree, Máire, Andrew, Obed, Greg, Ainash, Jordan, and Rachna. I would not have made it here
without you; I cherish our memories from Ann Arbor, Detroit, the lake house, and conferences,
and I look forward to new adventures.
Finally, to my family – Mom, Dad, Kurt, Jillian, Harrison, Lulu and Fiona. Thank you for your
unwavering support and immense patience. Your enthusiasm for my work has been the
backbone of my success. You never achieve anything alone and I am so happy to have you on
my team. Thank you.

iv
 TABLE OF CONTENTS

Dedication ..................................................................................................................................................... ii

Acknowledgements .......................................................................................................................................iii

List of Figures ............................................................................................................................................... vi

List of Tables ................................................................................................................................................ ix

List of Appendices ........................................................................................................................................ xi

Abstract ........................................................................................................................................................xii

CHAPTER 1: Introduction ............................................................................................................................. 1

CHAPTER 2: Life and Death at the End of Christian Nubia: Bioarchaeological Analyses of the Medieval
Population from El-Kurru, Sudan ................................................................................................................ 22

CHAPTER 3: Method Optimization of aDNA Extraction from Ancient Nubian Archaeological Materials ... 50

CHAPTER 4: Ancient Mitochondrial Genomes from Middle Nile Region of Sudan: A New Perspective of
Nubian Ancestry .......................................................................................................................................... 91

CHAPTER 5: Conclusion .......................................................................................................................... 129

Appendices ............................................................................................................................................... 132

References ................................................................................................................................................ 134

v
 List of Figures

Figure 1.1. Map of Ancient Nile River Valley ................................................................................................ 3

Figure 1.2. Modern Political Map of Nile River Valley with ethnic groups referenced in cited studies ......... 8

Figure 1.3. Mitochondrion in a human somatic cell .................................................................................... 11

Figure 1.4. Human mitochondrial DNA ....................................................................................................... 12

Figure 1.5. Lineage of Major Material Haplogroups for Mitochondrial DNA ............................................... 13

Figure 1.6. General workflow for NGS methodology (for archaeological samples) .................................... 15

Figure 1.7. European bias of paleogenomic work around the world, as of 2017........................................ 17

Figure 1.8. Published African mitochondrial and nuclear genomes, through August 2019 ........................ 18

Figure 2.1. Modern Map of Africa and ancient Nile Valley with close up of 4th Cataract region. ................ 23

Figure 2.2. El-Kurru Royal Cemetery plateau with Napatan Pyramid, temple in foreground. .................... 23

Figure 2.3. Excavation map of El-Kurru, including royal cemetery, medieval wall, and Napatan structures.
.................................................................................................................................................................... 24

Figure 2.4. Satellite view of El-Kurru excavations in relation to modern day agricultural and domestic land.
.................................................................................................................................................................... 25

Figure 2.5. Excavation of Medieval Christian structures at El-Kurru. ......................................................... 26

Figure 2.6. Excavations centered on Medieval Christian settlement and cemetery. .................................. 26

Figure 2.7. Mortality Distribution of 27 Christian Individuals from El-Kurru cemetery. ............................... 32

Figure 2.8. Artifacts uncovered within grave fill for El-Kurru burials. .......................................................... 34

Figure 2.9. El-Kurru Ind. 203 showing extended and supine burial. ........................................................... 35

Figure 2.10. Burial treatment and body treatment of El-Kurru individuals .................................................. 36

Figure 2.11. Cribra orbitalia severity and stages of healing observed in El-Kurru population. ................... 38

Figure 2.12. Stages of severity observed in cribra orbitalia lesions in El-Kurru population. ....................... 39

Figure 2.13. Healed lesions observed in El-Kurru population..................................................................... 39

Figure 2.14. Porotic Hyperostosis in El-Kurru population. .......................................................................... 40

Figure 2.15. Linear Enamel Hypoplasias found in El-Kurru population. ..................................................... 41

vi
Figure 2.16. Two instances of endocranial lesions observed in El-Kurru population. ................................ 42

Figure 2.17. Cribra femoralis found in the El-Kurru population ................................................................... 43

Figure 2.18. Cribra humeralis found in El-Kurru population. ....................................................................... 43

Figure 2.19. Four instances of trauma found in El-Kurru population. ......................................................... 44

Figure 3.1. Geographic locations of Nubian populations included in this study. ........................................ 52

Figure 3.2. Excavations centered on Medieval Christian settlement and cemetery. .................................. 55

Figure 3.3. Map of Ghazali monastery complex, including surrounding cemeteries and settlement area. 57

Figure 3.4. Topographical Map of Berber Meroitic Cemetery with dating (Bashir & David 2015). ............. 61

Figure 3.5. Laboratory facilities used for ancient DNA work at the University of Zürich. ............................ 65

Figure 3.6. Examples of teeth with dental calculus build up requiring sampling. ....................................... 66

Figure 3.7. Processing of tooth samples in clean lab. ................................................................................ 67

Figure 3.8. Set up for processing teeth within sampling hood in clean lab ................................................. 68

Figure 3.9. Crushing of root dentine using steel mortar and pestle. ........................................................... 69

Figure 3.10. Anatomical parts of temporal bone. ........................................................................................ 70

Figure 3.11. Visual of how the petrous portion of temporal bone is sectioned for processing. .................. 71

Figure 3.12. Preparation of bone samples from petrous portion using stainless steel mortar and pestle
with sleeve................................................................................................................................................... 72

Figure 3.13. Flowchart for all trialed extraction methodologies. ................................................................. 73

Figure 3.14. Library preparation of ancient DNA extractions and subsequent sequencing. ...................... 75

Figure 3.15. Sequencing and post-sequencing analysis. ........................................................................... 77

Figure 3.16. Mapped ancient reads listed by method. ................................................................................ 79

Figure 3.17. Impact of bleach on contaminating bacteria in extracts.......................................................... 81

Figure 3.18. Endogenous Content (%) comparison by method. ................................................................. 82

Figure 3.19. Cluster Factor value comparison by method. ......................................................................... 83

Figure 3.20. Comparison of tissue type to successfully obtain aDNA ........................................................ 84

Figure 3.21. Comparison of burial context to successfully obtain aDNA. ................................................... 85

Figure 4.1. Map of Berber Meroitic Cemetery. ............................................................................................ 94

Figure 4.2. Burial context of Individual BMC 6a and tooth sample. ............................................................ 94

Figure 4.3. Ghazali complex located near 4th Cataract, Sudan. ................................................................. 95

vii
Figure 4.4. Burial context of Ghazali-4-008.2/GHZ 2. ................................................................................. 96

Figure 4.5. Burial context and sampling of GHZ-1-010/GHZ 5. .................................................................. 97

Figure 4.6. Aerial view of the site of Kwieka near the 5th Cataract region. ................................................. 97

Figure 4.7. Burial context of KWC T2, associated finds, and sampling material. ....................................... 98

Figure 4.8. Burial context of KWC T6 and sampling material. .................................................................... 99

Figure 4.9. Right temporal bone from TARP T9. ........................................................................................ 99

Figure 4.10. Coverage of the Mitochondrial Genome of Nubian Individuals. ........................................... 107

Figure 4.11. Example of Coverage Plot of KWC 2 on the circular human mitogenome ........................... 108

Figure 4.12. Map Damage Profiles of Mapped Mitochondrial Reads from merged library samples. ....... 109

Figure 4.13. Read Fragmentation Plots of Mapped MT reads from merged library samples ................... 110

Figure 4.14. Haplotype Diversity of Northeast Africa region coordinated with geographic locations ....... 115

Figure 4.15. Haplotype Frequencies of modern Northeast African populations and Sudanese
subpopulations .......................................................................................................................................... 116

Figure 4.16. Principle Component Analysis based on haplotype frequencies, whole MT genomes ........ 118

Figure 4.17. Principle Component Analysis based on haplotype frequencies, ancient groups only ........ 119

Figure 4.18. Multi-Dimensional Scaling Plot mapping FST pairwise differences among Ancient and Modern
African and Eurasian populations ............................................................................................................. 122

Figure S1. Sequence counts histogram of Negative Blank libraries with more than zero reads. ............. 133

viii
 List of Tables

Table 2.1. Results from AMS Radiocarbon dating performed at ETH Zürich. ............................................ 31

Table 2.2. List of El-Kurru Individuals and demographic data. ................................................................... 37

Table 2.3. Summary Table of Health Status observations. ........................................................................ 45

Table 3.1. Archaeological information and sampling of Nubian populations included in this study. .......... 53

Table 3.2. List of El-Kurru individuals, including demographic data, archaeological context, and samples
collected. ..................................................................................................................................................... 55

Table 3.3. List of Ghazali Individuals, including demographic data, archaeological context, and samples
collected. ..................................................................................................................................................... 57

Table 3.4. List of Kwieka Individuals, including archaeological context and samples collected. ................ 58

Table 3.5. List of TARP Individuals, including demographic data, archaeological context, and samples
collected. ..................................................................................................................................................... 59

Table 3.6. List of BMC individuals, including archaeological context, contextual dating, and samples
collected. ..................................................................................................................................................... 61

Table 3.7. List of El-Zuma individuals, including demographic data, archaeological context, and samples
collected. ..................................................................................................................................................... 62

Table 3.8. List of el-Detti individual, including demographic data and samples collected. ......................... 63

Table 3.9. List of Tanqasi individuals, including demographic data and samples collected. ...................... 64

Table 3.10. Four samples selected for bleach pretreatment and results. ................................................... 80

Table 4.1. Sources and group counts for haplotype frequencies data for dataset assembly. .................. 102

Table 4.2. Libraries enriched and analyzed with Schmutzi. ...................................................................... 105

Table 4.3. Sequence Quality Analysis Summary - from EAGER analysis of enriched reads .................. 107

Table 4.4. Demographic and Archaeological information of Nubian individuals (ordered by date) .......... 111

Table 4.5. Results of Molecular Sexing .................................................................................................... 112

Table 4.6. Summary of consensus mitochondrial sequence information and haplogroup assignments .. 112

Table 4.7. Haplogroup Frequencies Grouped by Population or Subpopulations (in percentages) .......... 113

ix
Table 4.8. FST Values for comparisons to Ancient Nubians ...................................................................... 120

x
 List of Appendices

Appendix A - Supplementary information relating to methods: ................................................................. 132

Appendix B - Analysis / MultiFastQC of negative control samples ........................................................... 133

xi
 Abstract

This work investigated a new means to explore Nubian ancestry using ancient DNA
retrieved from archaeological skeletal material and marrying this perspective with
bioarchaeological and other data sources. The focus of this project is centered on sites within
the Middle Nile region of Sudan dated to before the Islamic conquest, where no paleogenomic
work has recently taken place, but there is a wealth of bioarchaeological research. To begin, a
sample population of 27 individuals was excavated from a cemetery, El-Kurru, northern Sudan,
(near the 4th Cataract). Burial traditions and body treatment identify these individuals as
Christian and their remains were bioarchaeologically analyzed for demographics (estimations of
sex and age) and physiological markers of stress and disease (orbital lesions and cranial
lesions, dental hypoplasias, non-specific lesions, trauma). The cemetery was contextually dated
to post-14th century CE, meaning these people were living on the brink of a cultural transition,
from Christian to Islamic dominance in the Upper Nubian region. Their remains showed high
frequencies of dental hypoplasias and non-specific lesions, especially in subadults, indicating a
heavy disease load and other stressors before adulthood; however, many adults lived to an
older age with little signs of trauma. The frequencies of stress signals were interpreted within
the context of Nubian contemporaneous populations and found to be consistent with other
reported ranges. Christian burial treatments and “normal” stress markers of this population
indicated the cultural transition had likely not arrived or did not have adverse consequences on
this Medieval community along the Nile.
To explore the genetic structure of Nubian ancestry in the Middle Nile region, a new
dataset needed to be built from archaeological materials and new protocols were trialed to
establish the most successful method to extract ancient DNA from human bone and tooth
samples. Trials included the use of bleach and/or a predigestion treatment prior to extraction to
increase the endogenous genetic content, leading to higher numbers of ancient sequence reads
without compromising the uniqueness of reads obtained. Next-generation sequencing
techniques were used to sequence these samples with a 40 percent success rate. Various
burial contexts, states of preservation, and tissue types used were compared across the skeletal
sample collection. These trials established that aDNA can be retrieved from material from the

xii
Middle Nile region of Sudan with extensive pretreatments of the tissue prior to extraction of the
DNA.
Mitochondrial genomes of six ancient Nubians were constructed using a hybridization
enrichment method, which was vital to obtain sufficient 10X coverage of the genome. As a
uniparental marker, mitochondrial DNA can reconstruct maternal heritage of individuals through
common ancestors with genetic similarities. Two individuals showed African lineage, while the
other four showed non-African ancestry. While limited, these results showed Nubians had a
strong African component with evidence of gene flow from Eurasia dating back to at least
Meroitic (4th century BCE) through Christian times. Through two dimension-reduction analyses,
ancient Nubians show most genetic affinity with modern Egyptians, Middle Easterners, and East
Africans, while less so to modern Sudanese. Although these individuals encompass varying
archaeological contexts and span well over one thousand years, these initial results hint at a
complexity of the genetic makeup and begin to reconstruct the impact of human migrations from
outside Africa. This would be expected given the dynamic history of this powerful kingdom
centered in the Middle Nile Valley.

xiii
 CHAPTER 1: Introduction

A distinguishing trait of our species is our demonstrated adaptability and flexibility to

occupy almost every niche of this planet. As our ancestors explored and inhabited new
landscapes, their genomes kept a record of past demographic events, environmental changes,
and cultural shifts. While archaeological, anthropological, and historic data can provide a
temporal aspect and context of external factors – that is political, social, environmental, health-
related forces, human genetic data serves as a significant source of information for
understanding the patterns of human variation and reconstruction of the past. A considerable
amount of work seeks to understand the movement of humans out of Africa and modern
discoveries rewrite the history of our species every year. In fact, recent findings lend
unexpected evidence suggesting H. sapiens migrations occurred much earlier (ca. 200,000 ya)
than previously thought (Wade, et al. 2019, Harvati, et al. 2019).
We are also beginning to understand human population movements within Africa. These
movements have shaped the incredible amount of biological diversity seen on this continent.
One particular place of interest is the Nile River Valley in northeast Africa. Positioned at the
corridor between Sub-Saharan Africa and Eurasia, this region has a long history of human
occupation and has a served as an important migration corridor, especially the Middle Nile
Basin or Ancient Nubia (modern-day southern Egypt and northern Sudan). Centuries of
archaeological expeditions in this area have uncovered ancient cities, temples, and settlements
as well as the skeletal remains of their inhabitants.
Investigating human history in the Nile River Valley is a complex topic which requires a
multidisciplinary approach integrating genetic and bioarchaeological components. Genetic data
allows us to explore an individual’s traits and ancestry, then on a population level, adaptation
and migration can be explored. Skeletal data provide snapshots of past people’s lives. When
used in conjunction with genetic data, they are a valuable means to study population history.
The usefulness of this mixed methods strategy has been demonstrated in modern publications
(i.e. Prendergast, et al. 2019, Aronsen, et al. 2019, Amorim, et al. 2018) and is well-exemplified
in work elucidating the Neolithic Transition in Europe (or the change from foraging to farming
subsistence strategies). The driving force of this transition was debated as either the spread of

1
people (demic diffusion) or the spread of ideas (cultural diffusion) or the process occurring
regionally (Childe 1968, Clark 1965, Ammerman & Cavalli-Sforza 1971, Menozzi, et al. 1978).
Archaeological data alone were unable to discern the driving force behind this major event in
population prehistory. With the inclusion of genetic data, the demic diffusion hypothesis was
supported (Sokal, et al. 1991). This multifaceted approach provides a template for analysis and
demonstrates the value of such a strategy.
Modeling this method, this dissertation research uses a two-pronged approach
combining bioarchaeological and paleogenomic analyses (both fields outlined below) to address
questions about the population history of Nubians in context of external forces shaping these
individuals. Such strategies have been implemented in the Nile Valley in the past. For example
the combination of archaeological, paleopathological, and paleodemographical data to study
people during cultural transitions and demographic shifts in Nubian history (Armelagos 1969,
Van Gerven, et al. 1995). However, these works did not utilize genetic evidence because it as
not yet available. Therefore, this is the first study in ancient Nubia to integrate these lines of
evidence to produce a more holistic view of these individuals and better address the complexity
of population dynamics in the Middle Nile Valley.

Ancient Nubia and Nubian Archaeology

For more than three centuries, explorers and archaeologists have been traversing the
Nile Valley uncovering a complex ancient history. Most of this endeavor has overlooked the
Nubian kingdoms in the south, in modern Sudan. Still, considerable progress has been made
during the past half-century, shedding light on this important chapter of African history. Known
as “the land between Cataracts”, Nubia occupies the region of the middle Nile basin, positioned
at an important corridor for human interactions along the Nile River (Figure 1.1). The successive
kingdoms of Nubia were influential in the development of humanity in Africa – they amassed
power and wealth while being masters of building, trade, and war.

2
Figure 1.1. Map of Ancient Nile River Valley.

LOWER
NUBIA

Kwieka
UPPER
NUBIA
Berber, Tinga

El Zuma, Ghazali
El Detti,
Tanqasi

Lower Nubia spanned from the 1st to the 3rd Cataract. Upper Nubia spanned from 3rd Cataract to Khartoum. Major
archaeological sites in Nubia and Egypt marked with black dots. Archaeological sites were samples were obtained for
this project marked with blue dots. El-Zuma marks the sites of El-Detti and Tanqasi, which are all in close proximity of
El-Kurru. Map from Nubia: Ancient Kingdoms of Africa (Emberling, 2011, p.3, Institute for the Study of the Ancient
World, New York University).

The Nile River Valley was occupied before the modern Holocene (ca. 10,000 BCE) as
evidenced by the presence of lithics and human burials (e.g. Wadi Kubbaniya) (Wendorf, et al.
1976, Leplongeon 2017, Wendorf, et al. 1988). Climatic and environmental changes greatly
impacted the development of human societies in this region. Specifically, the appearance of

3
more hospitable conditions around 8,000 years ago was pivotal (Edwards 2004). During the
early Holocene, inhabitants most likely subsisted by hunting, fishing, and gathering as
seasonally or semi-nomadic groups, centered around stable water sources outside the Nile
Valley (Edwards 2004). Evidence of the first permanent settlements dates to around 7,000 BCE
and the advent of pottery in this region appeared between the 10th and 9th millenniums BP (ca.
8,000-7,000 years ago) (Edwards 2004, Close 1995, Jesse 2010). Aridification of the Sahara
around 5,000 BCE changed the environment drastically, forcing settlements to closely hug the
Nile (deMenocal, et al. 2000, Claussen, et al. 1999). Around this time, people shifted away from
the hunter-gatherer lifestyle first with the domestication of animals then closely followed by crop
domestication (Edwards 2004). Shortly before 3,000 BCE, evidence of trade and more
advanced social structures were present (e.g. presence of foreign materials from the Red Sea
associated with burials) (Edwards 2004). The oldest distinguishable kingdoms of Nubia formed
in the region of Lower Nubia by 3,000 BCE, known as the Terminal A-Group (Edwards 2004).
Elite and royal burials with fine goods associated with this group demonstrated social
stratification, labor specialization, and wealth. Raids from neighboring cultures disturbed and
likely dispersed the A-Group and the area was not resettled for more than a half century by a
new population known as the C-Group (so-called “B-Group” culture was reclassified as early A-
Group due to work by Smith (1966)) (Edwards 2004, Gatto 2006). The origins of these
individuals are still debated, but they are characterized by distinct so-called “pan grave” burials,
and may have been related to eastern desert nomads, the Medjay (Bietak 1987, Emberling
2011). Others argue similarities to the later Kushites suggest a migration of people from the
south (Edwards 2004).
Around 2,500 BCE, the Kerma culture emerges among other political units in the 3rd
Cataract area (Figure 1.1) and develops into a sprawling urban center with palatial structures,
massive temples, and royal tombs, which still stand today (Bonnet 1986, Bonnet 1990). Here,
the “Kingdom of Kush” began. These powerful kings built an imposing military force, primarily
known for their superior skills in archery, who often undertook raids to the north in Egypt. The
Kermans appear to have maintained independence, although regularly engaging with the
ancient Egyptian civilization to the north, until 1,550 BC when Egyptian pharaohs of the New
Kingdom established direct control over Nubia (W. Adams 1984). When incorporating Nubia
within the Egyptian empire, this was a time of intensive intermingling of these populations.
Following the fall of the Egyptian New Kingdom (1550–1069 BCE), the Nile River Valley saw the
rise of the second Kingdom of Kushite, now centered in Napata near the 4th Cataract (Figure
1.1). This period was a revival of both Kushite and Egyptian beliefs, under the dominion of the

4
Napatan kings. During the 8th century BCE, the Napatan kings ruled all of Egypt and Nubia as
pharaoh for approximately a century. After losing control of Egypt, Napatan Kush flourished in
Nubia for nearly four more centuries (ca. 700–300 BCE).
The second Kingdom of Kush underwent a still poorly understood transition, when the
capitol was moved upriver to Meroe, ushering in the Meroitic period (ca. 300 BCE – 300 CE).
The kingdom continued to prosper during this time with far reaching control of the Middle Nile
and expansive trade networks. The fall of the Kingdom of Kush at the end of the Meroitic period
is also poorly documented but may be attributed to encroachment of the Kingdom of Aksum
from Ethiopia or the Nuba people from the Western Desert (Hintze 1967). For the next few
centuries, Persians pushed into North Africa and occupied both Egypt and Nubia. At this time,
Lower Nubia was occupied by smaller agrarian communities of the X-Group or the Ballana
culture, which later made up a new independent state known as Nobadia. The state of Makuria
rose to power in Upper Nubia at this time. By the 6th century, the region had largely converted to
Christianity. The Islamic conquest began in the north in the 7th century, itself taking centuries to
complete (Welsby 2002) and having a major impact the genetic structure of the entire Northeast
region as Arabic influence spread through the Nile Valley (Hollfelder, et al. 2017, Dobon, et al.
2015). During Medieval times, the Middle Nile was controlled by three Christian kingdoms with
regional capitols along the Nile River. Generally existing autonomously, these polities prospered
for eight centuries and were set amongst many other powerful African kingdoms, for example in
Mali (Canós-Donnay 2019), Chad, Darfur, Ghana, and Ethiopia (Edwards 2004). The Christian
kingdoms pushed back multiple attempts by Islamic invaders from the north to control the
Nubian heartland while also expanding the economy, keeping peace in Lower Nubia, and
maintaining robust trade networks moving through the Nile corridor. By the 14th century the
medieval Nubian Kingdoms came under Arab control and new polities arose, including the Funj
Sultanate of Sinnar (south of Khartoum) and the Sultanate of Darfur (in the west).
In order to understand ancient Nubian history from the site of El-Kurru and across time,
we implemented a paleogenomic approach that focused on mitochondrial DNA – an important
marker in understanding prehistory. Background information on mitochondrial DNA and its
usefulness in anthropological research are detailed below.

Bioarchaeological Research in Sudan and Nubia

The informative value of human remains is extensive, despite the small archaeological
footprint. Skeletal material serves as a direct way to study human populations and reconstruct
the lives of past individuals and is the primary material for bioarchaeological research. The Nile

5
River Valley has a long tradition of exploratory work with human remains tightly paired with
centuries of archaeological research in the Valley and was pivotal to the birth and advancement
of such fields as paleopathology and bioarchaeology (Baker & Judd 2012). Arid conditions and
cultural burial practices preserved substantial skeletal collections from this region. This is
particularly the case for Nubia, whose skeletal collections make up some of the world’s largest
and most expansive in time (Binder 2019). Like the collections from which the DNA samples
derive, the remains and burial context of these individuals provides an important means to
understand lived experiences and stressors, diet and nutrition, social identity, health and
disease, and demographic data to reconstruct population dynamics, social and biological
identity, and disease in antiquity. Understanding who these individuals were, is of paramount
importance to explore.
While the prehistory of Sudan has been the subject of exploration from outside scholars
since the early 19th century, Nubia was perhaps best characterized during large archaeological
campaigns in the 20th century (Baker 2016). The First Archaeological Survey of Nubia, from
1907-1911, explored hundreds of sites and unearthed thousands of individuals from all time
periods of Sudan’s history (Adams 1977). Excavators and medical professionals recognized the
value of studying the people themselves, who built the towns they excavated or crafted the
goods they were shipping home to museums (Smith & Jones 1910). This survey mostly focused
on demographic typologies, but soon grew to include the vast number of pathologies observed
(Waldron 2000). Many of the more curious or diagnostic examples were shipped out of Sudan to
contribute to or begin medical collections around the world (Baker & Judd 2012). The Second
Archeological Survey of Nubia, conducted from 1929 to 1934, was launched as a salvage
campaign with the raising of the Aswan Dam. Again, thousands of bodies were excavated and
the focus of studying these remains was mostly descriptive of racial affinities and included
observations of pathologies (Admas 1977, Waldron 2000).
The next surge of excavations came in the 1960s with another salvage campaign when
the Aswan Dam was raised again, putting over a thousand sites underwater as far as the 2nd
Cataract (Adams 1977). This time, biological anthropologists were employed alongside
archaeologists to excavate burials, which gave rise to a new paradigm in the field of
anthropology – the biocultural approach, as named by Borthwell (1967), then most notably
implemented by Armelagos (1969). Such an approach contextualized variation in human biology
observed in the skeletal remains within a cultural and environmental framework discerned from
the accompanying archaeological data (Baker 2016). This shift from descriptive to biocultural
studies gave rise to the “bridging” field of bioarchaeology, a broad reaching discipline which

6
describes methods for extracting biological and cultural data from archaeological skeletal
materials (Larsen 2015, Baker 2016).
As the field shifted, so did the location of work moving further south into Sudan to
include Dongola and Meroe as Lower Nubia was flooded (Binder 2019). Then in in the 1990s,
another dam project threatened the region near the 4th Cataract at Meroe, prompting thousands
of salvage excavations and providing more skeletal remains to be studied (Emberling 2012).
Undoubtedly supported by such an abundance of material, further expansion of archaeological
projects for the next few decades, and most importantly, the ability to export these materials for
more extensive investigations beyond the field (collections summarized in Binder 2019),
bioarchaeology in Nubia developed as a discipline unto itself and continues to be a staple of
investigation to elucidate the history and inhabitants of Nubia.

Modern Genetic Landscape of Sudan and Northeast Africa

Studies using genetic data, even from ancient sources, are not a new strategy to
understand Nubian history. Located in the Middle Nile Basin, ancient Nubians were positioned
at a corridor connecting Sub-Saharan Africa to North Africa that ultimately lead to Europe and
Asia. The Nile River served as a means to carry natural resources and other riches from the
interior of Africa to Egypt and beyond, turning Nubia into a strategic economic power (Adams
1977). This also meant the flow of people and ideas passed through this region, influencing the
population seen today as a unique blend of African and Mediterranean (Adams 1967). To verify
this observation, ancient DNA from Meroitic Lower Nubia and modern genetic data has been
used to show the Nile Valley was a genetic corridor for the bi-directional migration of people
since ancient times (Lalueza Fox 1997, Krings, et al. 1999). These works set the stage to further
understand the genetics of Nubia, this ancient group at a strategic place along the Nile River, as
well as demonstrate the viability of paleogenetic work with archaeological materials from this
region.
The past decade or more has seen a continued interest in the genetics of the Nile River
Valley, specifically Sudan and South Sudan, while also including Egypt and Ethiopia. Modern
studies utilize uniparental markers from the maternal lineage including mitochondrial (mtDNA)
(outlined below) and the paternal lineage including the non-recombining region of the Y
Chromosome (NRY), microsatellites or short tandem repeats (STRs), as well as autosomal DNA
including single nucleotide polymorphism (SNP) arrays and whole genome sequencing data.
Such studies follow a similar research design to genetically survey ethnic groups across region.
These include Nubians in the north, central groups of Arab descent further up the Nile River, the

7
eastern Beja near the Red Sea, the pastoral Nilotes in South Sudan, the Nuba near the
mountains, the Darfurian groups in the western desert, and some nomadic groups, like the
Fulani and the Meseria, with more expansive occupation areas across the country (Figure 1.2).

Figure 1.2. Modern Political Map of Nile River Valley with ethnic groups referenced in cited
studies.

Modern map of the Nile River Valley and surrounding counties. Orange labels represent ethnic groups included in
various genetic studies and their general locales; pastoral groups not included (Doban, et al. 2015, Babiker, et al.
2011, Hollfelder, et al. 2017, Hassan, et al. 2008, Hassan 2009). CAR – Central African Republic, DRC – Democratic
Republic of Congo. Map obtained from d-maps.com and modified by author.

While some of these datasets are more robust than others, what is published converges
on the uniqueness of Nubians among other ethnic groups and shows interesting signals of
complexity that were influenced by past demographic events, namely the Arab expansion from
the North. However, a new dataset from ancient individuals is required to fill in gaps in
knowledge to fully understand this complexity while also addressing new questions. For
example, what ancient groups contributed to the high diversity found in modern populations?
What is the genetic profile of Nubians before the Arabic expansion? And how can these data

8
augment anthropological, historical, and archaeological hypotheses and notions of the Nubian
past?
Nubians are unique compared to other ethnic groups in the Northeast region of Africa
and other Sudanese groups. They are characterized by higher intrapopulation variation,
meaning the most genetic diversity exists among individuals of this group, rather than diversity
between Nubians and another outside group. This had been observed with mtDNA (Alfonso, et
al. 2008, Hassan 2009) and nuclear SNP data (Hollfelder, et al. 2017). Autosomal (nuclear)
markers show Nubians are highly admixed through various measures of genetic diversity (e.g.
high allelic richness, an abundance of private alleles and shared private alleles, and short runs
of homozygosity) (Hollfelder, et al. 2017). The origin of this diversity has been the focus of
several recent studies using autosomal data that aim to understand the underlying genetic
structure in the Sudanese region. Such genetic variation is shaped by demographic forces,
including migration and admixture, as well as biological factors like genetic drift and natural
selection (Li, et al. 2008). While not mutually exclusive, the diversity present in Nubians is likely
to be the outcome of demographic events – namely gene flow (or the movement of genetic
material from one population to another) – which introduced material from Eurasia and East
Africa. Both uniparental, single locus markers add to this same narrative: haplogroup
frequencies indicate past gene flow or migration events, especially from outside Africa (Hassan,
et al. 2008, Hassan 2009, Lalueza Fox 1997, Krings, et al. 1999). Nubians have a significant
Eurasian component (roughly half) as part of their genetic makeup, with the remaining part
quantified as a “Nilo-Saharan” component that is distinct from Sub-Saharan or North African
signatures (Hollfelder, et al. 2017). Other ethnic groups in the surrounding region, the Beja,
central Arabic groups, and even modern Ethiopians, have this same two-part signature. The
Nilo-Saharan component is unique to East Africa and is hypothesized to characterize the
ancestral genetic state of this region (Hollfelder, et al. 2017). Archaeological samples can
directly test this hypothesis. South Sudanese Nilotes have this makeup almost exclusively and it
closely resembles the 4,500-year-old individual from Ethiopia with no Eurasian admixture
(Gallego Llorente, et al. 2015).
However, it remains unclear from which group these gene flow signals originate Genetic
data from extant Nubians show a particularly close connection with Egyptian populations
(Krings, et al. 1999, Babiker, et al. 2011, Lalueza Fox 1997). This would be expected from their
geographic proximity to each other and from archaeological and historical records that
document interactions between these two groups for millennia, especially during Egypt’s
colonial period in Sudan and subsequent “Egyptianization” and also during the rise of the

9
Napatan state during the 25th Dynasty. Both uniparental markers also attest to this narrative as
amounts of variance between the populations are low (Hassan, et al. 2008, Alfonso, et al.
2008). This suggests there was no sex bias in the intermingling of these two populations, at
least the recent past. Paleogenomic data would test if this closeness was the result of these
known events of their affinity being more basal.
Several genetic surveys of Africa support the idea that the genetic landscape of this
continent is profoundly shaped by geography (Scheinfeldt, et al. 2010, Tishkoff, et al. 2009) and
the Sudanese region keeps to this pattern. Based on autosomal data, Sudan and South Sudan
can be divided geographically into groups of the northeast (including Nubians, the Beja, Copts,
and central Arabs) and the southwest (including the Nuba, western desert groups, and Nilotes)
(Hollfelder, et al. 2017, Dobon, et al. 2015, Babiker, et al. 2011) (Figure 1.2). These two regional
groups were differentiated as a result of the presence of Eurasian and/or Middle Eastern genetic
signals in the northeast and general absence in the southwest. This shared Eurasian
component is the signature of the Arab expansion, which began in the 7th century in Egypt, but
did not reach the Middle Nile region until much later in the 14th century. Reconstructive analyses
showed successive admixture events moving southward along the Nile. However, the Nubians
are again distinct, even among the northeast ethnic groups, when concerning linguistic
affiliations. Nubians may be genetically close to those groups in geographic proximity, but they
speak Eastern-Sudanic languages while all others speak Afro-Asiatic ones (Hollfelder, et al.
2017, Dobon, et al. 2015). Eastern-Sudanic languages belong to the Nilo-Saharan language
family that are more commonly spoken among the southern Nilotes and other ethnic groups in
the southwest region. This would suggest population integration on a genetic level (i.e.
exchange of genetic material between people) but not all cultural transitions occurred as
happened with Arabic groups which adopted Afro-Asiatic languages (Hollfelder, et al. 2017).
Outlined here, the complexity of Nubian genetics is likely defined by East African
ancestry and was significantly impacted by recent migratory events. This begs the question of
what was the genetic landscape of Nubians predating the Arab expansion and was this the
source of diversity that is unique to Nubians in this region. DNA from the ancient from ancient
and historic populations in Sudan can refine the timing of admixture events and directly survey
the genetic landscape before the Arab expansion and characterize the basal state for
Northeastern Africans. Ancient DNA would also lend more data to help elucidate the linguistic
connection between Nubians and their southern neighbors.

10
Mitochondrial DNA

Mitogenomic data are an important means of reconstructing and detecting signatures of

human migrations with cross-disciplinary reaches. Human mitochondrial DNA (mtDNA) has
been pivotal to understanding the origins and evolution of our species (Cann, et al. 1987,
Krings, et al. 1997), human migrations and peopling events (e.g. Mulligan & Szathmáry 2017),
and the reconstruction of historical events, especially when biological sexes are differently
affected (Bamshad et al 1998).
Mitochondrial DNA is circular, double-stranded extranuclear genetic material that is
found in the energy-producing organelles of all somatic cells (save red blood cells) (Figure 1.3).
The 16,569 base pair genome is organized into 37 genes encoding proteins, rRNAs, or tRNAs
involved with cellular respiration or energy production (Stoneking 2017, p.111) (Figure 1.4).
While these molecules mutate at a rate 5-10x faster than nuclear DNA, they are inherited
uniparentally from an individual’s mother and thus do not undergo recombination with each
generation. Without recombination, mutation is the only source of variation. This makes mtDNA
ideal for tracing ancestral lineages, which can be used for constructing phylogenetic trees of
relatedness, even for closely related populations (Kaestle 2010). Especially advantageous for
forensic and paleogenomic applications, mtDNA is found at a high copy number per cell (i.e. up
to few thousand mitochondria per cell, 5-10 genomes per mitochondrion), in comparison to
nuclear DNA (Giles, et al. 1980, Pääbo and Wilson 1988) (Figure 1.3).

Figure 1.3. Mitochondrion in a human somatic cell.

(Created with BioRender.com by author)

11
 Mitochondrial (MT) genetic variation is a valuable tool to observe the impact of historical
events on genetic makeup and can reconstruct demographic history through the lens of
maternal ancestors (Cann, et al. 1987, Vigilant, et al. 1991). Analyses track mutations to create
unique sequence profiles that differ from the human reference sequence, including the
Cambridge Reference Sequence (CRS), first proposed in 1981 by Anderson, et al., the Revised
Cambridge Reference Sequence (rCRS) (Anderson, et al. 1999), or the most recent iteration,
the Reconstructed Sapiens Reference Sequence (RSRS) (Behar, et al. 2012). Specifically, most
mutations accumulate at two hypervariable regions (HVRI and HVRII), therefore sequence data
from these regions are the most helpful to infer genetic relationships among individuals and
populations (Figure 1.5). Individuals with genetic variants that share a common ancestor will
cluster together in clades as a result of shared common mutations; these clusters are defined as
haplogroups (groups of related sequences defined with nomenclature) (Stoneking 2018). Main
groups are named with letters (e.g. L, M, N) and subhaplogroups with numbers and letters (e.g.
L0a1a). L lineages originated in Africa, while M and N (and all those that derive from these two
lineages) – are found mostly outside of Africa (Stoneking 2018, Cerezo, et al. 2012). Typically,
research designs target HVRs for haplogroup assignment, rather than the whole mitogenome.
However, dropping costs and the implementation of massively parallel sequencing techniques
(next generation methods) have made whole mitogenomes more accessible and common for
projects.

Figure 1.4. Human mitochondrial DNA.

12
Genetic map of human mitochondrial DNA, spans 16,569 base pairs; inset shows the control region where the
Hypervariable Regions I and II are located (numbered, in red). (Figure 9.2 from Stoneking, et al. 2016, An
Introduction to Molecular Anthropology, p. 113)

Figure 1.5. Lineage of Major Material Haplogroups for Mitochondrial DNA.

Phylogenetic tree illustrating the relationships of the major mtDNA haplogroups. Macrohaplogroups L, M, N, and R
are indicated. Note that branch lengths are not proportional to mutational difference. (Figure 9.4 from: Stoneking, et
al. 2017, An Introduction to Molecular Anthropology, p. 118)

Despite the utility of mtDNA in population genetic analyses, there are some
disadvantages that should be considered here. First, it is a single genetic locus that tracks the
maternal lineage. Therefore, no information on paternal lineage history, sex-biased behaviors,
or cultural practices (e.g. marriage traditions or those with a sex-bias) can be obtained from this
locus. Illustrating this, MT haplogroup assignment has been found to provide limited information
concerning an individual’s continental region of origin or ancestry (Emery et al. 2015). This
indicates that identifying the mtDNA haplogroup for an ancient individual, then reconstructing
ancestry or even country of origin may not be as specific as it would seem. Furthermore, mtDNA
may be less informative of overall population history if it experienced locus specific genetic drift
or selection (Stoneking 2018). Second, the MT genome, only 16K base pairs, is a small fraction
of the human genome and is much less discriminatory than 2.9 base pairs found in nuclear
DNA. Third, new data show that heteroplasmy is possible, meaning the inherent assumption
that this marker is indicative of maternal heritage is currently under debate (Luo et al. 2018).
Despite these drawbacks, mtDNA is and continues to be an informative workhorse in
evolutionary studies (DeSalle et al. 2017).

13
 As is related to this dissertation research, this locus, among others, has been utilized to
characterize the genetic landscape of the modern Sudanese people as well as some ancient
populations. However, a largescale paleogenomics project has yet to be conducted in ancient
Nubia.

Paleogenomics and Ancient DNA Methodologies

More than 30 years ago, the field of paleogenomics (or the analysis of genetic
information of extinct living things) was developed. Since then, the scope of the field and the
number of projects undertaken have increased due to advances in sampling methods and the
implementation of NGS techniques (Hagelberg, et al. 2015). As research designs have moved
past the use of PCR products to NGS techniques, the fundamentals have remained the same
(Cooper & Poinar 2000, Fulton & Shapiro 2019). Today, ancient DNA studies are a useful tool in
various disciplines, including conservation genomics of endangered wildlife, evolutionary
genetics, and archaeological analysis (Leonard 2008, Green, et al. 2006, Amorim, et al. 2018).
In human population genetics, aDNA contributes to the exploration of population dynamics over
time and the timing of demographic shifts. Importantly, this genomic view into the past has
demonstrated that present-day population structure is the product of migrations and dynamic
admixture events, most ancestors likely originated from somewhere else, and unfortunately, this
complexity may not be reflected in the genetics of extant populations (Pickrell & Reich 2014).
Most commonly, aDNA is retrieved from bones or teeth, however other non-traditional
mediums are currently in use, including soft tissues, dental calculus, coprolites, and soil.
Regardless of the source material, successful extraction of aDNA faces the same hurdles.
ADNA is short, heavily damaged, found in low quantities, and contaminated. To address these
limitations, NGS methods are particularly well suited to handle the short, degraded molecules
(Figure 1.6). Briefly, DNA is extracted from a tissue sample, most commonly a powder. An
enzymatic treatment pulls the DNA into solution, and it is then purified and concentrated. These
short fragments of DNA are built into “libraries” by attaching adaptors and unique barcodes,
which are then massively sequenced in parallel on a sequencing platform (e.g. an Illumina)
(Dabney at el. 2013, Meyer and Kircher 2010, Kircher, et al. 2012) (Figure 1.6). Sequencing
products are processed bioinformatically, including assessment of damage, contamination
estimation, and consensus assembly.

14
Figure 1.6. General workflow for NGS methodology (for archaeological samples).

Figure created with BioRender by author.

Post-mortem damage is also a limitation to working with aDNA. While unavoidable, the
typical, chemically damaged profile is instead used as an authentication criterion to differentiate
ancient from modern sequences. Molecular degradation begins immediately after cellular death
when DNA repair mechanisms cease to function. In a burial context, decaying body tissues
damage genetic material as do environmental factors, including bacterial and chemical changes
in the surrounding soil (Allentoft, et al. 2012). This damage only increases with time, resulting in
fragmented (less than 100bp, usually 35-55bp) and chemically modified genetic material
(Briggs, et al. 2007, Poinar et al 2006). Hydrolysis is also a major problem causing strand
breakage and miscoding lesions, where the deamination of cytosine to uracil later causes C to T
transitions within the genetic code (Lindahl 1993). These C to T transitions are the most
common signature of damage and are used for authentication analysis by distinguishing
contaminating (those without extensive damage) sequences from ancient ones (Skoglund, et al.
2014, Fulton & Shapiro 2019).
Contamination from modern sources has plagued the field since its inception (e.g. Wall
& Kim 2007). From the start, the sample (bone, tooth, etc.) itself is at risk of contamination due

15
to the porousness of biological tissues. Environmental conditions and microorganisms may
further compromise the integrity of the tissue, allowing exogenous DNA to be absorbed.
Handling of the sample by archaeologists or paleogeneticists also may introduce contamination.
Both methodological and sampling strategies (i.e. petrous bone) have been implemented to
combat this problem. Sample processing and aDNA extraction occurs in clean-room facilities
with strict contamination-reducing protocols, including UV radiation, bleach, personal protection
equipment, and filtered air systems (Fulton & Shapiro 2019). Negative controls are implemented
during extraction, library construction, and PCRs to monitor any contamination. Additionally,
during the library-building step, unique tags are attached to sequences extracted in the clean
room to identify likely ancient material, while contaminating sequences acquired later will be
untagged (Henneberger, et al. 2019, Dabney, et al. 2013). Untagged modern DNA sequences
can be filtered out during computational processing steps later. Furthermore, there are several
bioinformatic tools or pipelines to quantify and remove contaminating reads (e.g. Schmutzi,
ContamMix).
Sampling is also a means to mitigate contamination. The success of petrous portion
sampling of the temporal bone in the skull, first described by Gamba, et al. (2014), was
revolutionary for the field. Previously, teeth were (and still are) used for tissue sampling. Teeth
allow for easy implementation of sampling strategies, they typically exhibit better preservation
than the rest of the skeleton as a result of higher mineralization, and DNA binds to these
minerals leading to higher endogenous contents, or the amount of human genetic material
compared to the amount of exogenous DNA (i.e. bacterial, fungal, environmental, unclassified
material, etc.) in a sample (Damgaard, et al. 2015, Hansen, et al. 2017). However, because the
petrous portion encases the cochlea and inner ear bones, it is extremely dense and fortified.
This protection helps to fend off contamination, leading to higher endogenous contents; thus,
often outperforming teeth as a useful sampling material. While destructive, obtaining bone
powder from this part of the skull is now the gold standard and continues to be optimized to be
less invasive and more efficient at obtaining the highest amount of aDNA (Sirak, et al. 2017,
Hansen, et al. 2017, Pinhasi, et al. 2015, Pinhasi, et al. 2019). For ancient samples, this amount
is typically less than 1% due to the conditions described above (Der Sarkissian, et al. 2015). It is
therefore crucial for all projects to assess if this small percentage can be utilized from samples
before proceeding with further testing and/or destruction of sample material.
Sample preservation can be assessed on a case-by-case basis and is highly dependent
on the depositional environment (Hoss, et al. 1996). Ideal conditions for DNA preservation
include low, stable temperatures and low humidity, like those in caves (e.g. Noonan, et al.

16
2005). Material from temperate and arctic regions have produced better yields for aDNA
compared to material from more hot or humid locations (Pinhasi, et al. 2015). This preservation
bias in addition to the abundance of preserved archaeological material in Europe has led to a
European bias in paleogenomics (Figure 1.7). Armed with new methodologies, the field of
paleogenomics has expanded its scope to include Africa which was previously inaccessible
(Pääbo 1985, Pääbo & Wilson 1988). Work in this continent was hindered for decades due to
extensive thermal degradation of genetic material found to be common with the archaeological
samples from extremely arid environments of Africa (e.g. the Nile River Valley). Early work with
African samples showed significantly lower DNA preservation with which to start, let alone
retrieve. However, the combination of NGS and sampling of the petrous portion has been a
successful combination to bring the genomic revolution to Africa. In 2014, the first ancient
African genome of a male individual from Mota Cave in Ethiopia was published by Gallego
Llorente, et al. (2014). In the five short years since this first success, the genomes of 230
African individuals have been fully sequenced for mitochondrial (230) or nuclear genomes (101)
(or both) encompassing more than 15,000 years of genetic history from the Canary Islands to
South Africa (Figure 1.8). Within the Nile Valley, the first project exploring the genetics of
Ancient Egyptian individuals was published by Schuenemann, et al. (2017). However, projects
using Ancient Nubian remains are scarce or forthcoming (e.g. Jugert, et al. 2018, Sirak, K.,
personal communication).

Figure 1.7. European bias of paleogenomic work around the world, as of 2017.

Figure obtained from Callaway (2017), Nature.

17
Figure 1.8. Published African mitochondrial and nuclear genomes, through August 2019.

Geolocations of published genomes in Africa with estimated dating of individuals. Grouped by archaeological site, or
publication, archaeological name from publications or (modern) country. Legend at left, circles denote mitochondrial
sequence data only, diamonds with dots represent whole genome and mitochondrial sequence data. WG - Whole
genome, Mt - mitochondrial, NE - Near Eastern, N - number of individuals, ya - years ago, *** designates non-NGS
protocols used. Some individuals from Lazaridis et al. 2016 publication mapped because were included in analyses.

Chapter Outline

For this dissertation research, I excavated and collected a Nubian sample population
from El-Kurru, Sudan (Figure 1.1), conducted bioarchaeological analysis of the remains, and
performed paleogenomic analyses. The goal was to join all lines of evidence for an
interdisciplinary approach, marrying archaeological, bioarchaeological, and genetic data, to
reconstruct the identity of Nubians near the 4th Cataract region (MacEachern 2013). Chapters of

18
this dissertation research are organized temporally and reflect a logical progression of the
project (i.e. collection of samples, protocol optimization, implementation). First, Nubian remains
were analyzed using a bioarchaeological approach (Chapter 2) to collect data on demographics
and indicators of health and trauma. In addition, the mortuary and archaeological context of this
population was investigated to understand cultural affiliations (e.g. Christian or Muslim) and the
use of the space (e.g. domestic, administrative, palatial). These bioarchaeological analyses
shed light onto our understanding of life along the Nile in this Late Christian town during a time
of transition to Islam as a state religion, following the Arab conquest.
In Chapter 3, ancient DNA (aDNA) extraction methods were trialed using various
standard and non-standard methodologies in order to maximize aDNA yields. Some
methodologies were new developments in the field (i.e. extraction from petrous bone), while
others have been implemented for a longer time (i.e. bleach treatment). Unfortunately,
paleogenomic analyses using the El-Kurru samples analyzed in Chapter 2 were unsuccessful
(i.e. no aDNA was obtained). However, these skeletal materials provided a valuable means to
pilot paleogenomic work, trial new methodologies, and ultimately led to the successful
extractions of other samples from the Middle Nile region (Chapters 3 and 4). ADNA extraction
from tooth and bone samples (mostly teeth) from the El-Kurru collection was attempted using a
PCR-based procedure. The results of this pilot work showed high contamination and poor
preservation of the tissue, which ultimately pushed the research design to move onto more
advanced extraction methods. These included a new sampling technique using bone tissue
instead of teeth, employing only NGS techniques, and including an enrichment step.
Furthermore, since this was one of the only paleogenomic projects in the ancient Nile River
Valley to date, there was a lack of comparative ancient sequence data. Only three studies have
been conducted on material from the Middle Nile region. However, none of the data from these
studies were able to be incorporated into this analysis as the data was generated using
outdated methodologies or the dataset was not comparable (i.e. different genetic markers
including Y-chromosomal or nuclear polymorphisms) (Lalueza Fox 1997, Jugert, et al. 2018,
and Hassan 2009). Due to this dearth of comparable datasets, an expanded sample set was
collected through various collaborations in Sudan and abroad. In total, seven additional
populations of at least four individuals each were obtained and sampled for comparison to the
El-Kurru group and each other (Figure 1.1). These populations spanned multiple archaeological
sites from the 3rd to past the 5th Cataract. This extensive geographic sampling is the first to
include remains from regions never previously analyzed using NGS techniques.

19
 The focus of this project was to obtain full mitogenomes of Nubian individuals from
varying archaeological contexts, creating a time transect, to understand the genetic landscape
of Nubia before the Arab conquest and trace population continuity through time (Buzon 2008,
Buzon, et al. 2016, Irish & Joel 2005). The goal was to address ways paleogenomic data can
help understand the demography of this region in context of archaeological, historical, and
anthropological records. In Chapter 4, the mitochondrial genomes were reconstructed for six
individuals. Two individuals had African ancestry, while four showed ties to the Near East.
These genomes were compared to publicly available ancient and modern mitochondrial
genomes from Africa, the Middle East, and Europe. We found that Ancient Nubians have close
affinities to modern Egyptians, Middle Easterners, and East Africans, and less affinity to modern
Sudanese and Ancient Egyptians.

Significance

There are few bioarchaeological studies from the Upper Nubia region. Those that do
exist are concentrated at the same archaeological sites providing a narrow perspective of the
region. The Christian population from El-Kurru will serve as a new dataset for comparison with
other contemporaneous groups, like other Christian groups in the Upper Nubia region or those
living further north in Lower Nubia. Similarly, it can be used to understand temporal changes in
the region by comparison to groups that vary in time. The El-Kurru population is as an example
of a group of individuals in transition. This collection may be a useful example for other
bioarchaeological investigations of groups or communities facing cultural transitions around the
globe. Although a small number of people, the skeletal remains help to build a narrative of life
along the Nile during an important cultural transition in this region.
Second, paleogenomic research using samples from arid environments is an important
advancement in the field but possesses new hurdles to overcome. Compared to Europe, where
most of the paleogenomic research was developed and is currently taking place, studies
analyzing skeletal remains from Africa are rare. Furthermore, the paleogenomics of the Nile
River Valley of Africa is particularly understudied. There is no standard protocol for aDNA
methodology using material from this region and no study to date has trialed different methods
to optimize aDNA extraction techniques. With more enthusiasm for African paleogenomic work,
it is critical to understand which methods should be employed to obtain ancient genetic material.
In addition, this will better inform sampling strategies for future work.
Third, to date, no aDNA study using NGS techniques has been conducted on
populations of the Middle Nile Basin. Paleogenomics can be used to reconstruct Nubian

20
demographic history to further understand the dynamics of this region. While drawing upon the
growing knowledge from archaeological work, archaeological samples provide glimpses of
various eras to reconstruct population dynamics (e.g. admixture, integration,
displacement/replacement events) and understand the mobility and interactions of these groups
across the landscape. ADNA can contribute a new line of evidence to complement those from
other disciplines, which can in turn address larger questions requiring a more multi-dimensional
approach.

21
 CHAPTER 2: Life and Death at the End of Christian Nubia: Bioarchaeological
Analyses of the Medieval Population from El-Kurru, Sudan

INTRODUCTION

The 4th cataract region of Sudan, the S-bend of the Nile (Figure 1.1), has been the
subject of recent salvage archaeological investigations due to the construction of a dam at
Merowe (Emberling 2012). This area is characterized by fast moving water, dangerous rapids,
and land not as suitable for agricultural practices. Long thought to be less ideal for settling, the
salvage projects from 1991-2008 revealed the opposite conclusion. Several thousand sites are
estimated to dot the Nile region here, dating from the Paleolithic through the Islamic era.
Furthermore, several significant monumental sites are located here, for example the royal
pyramids at Nuri, temples and palaces of Gebel Barkal, and the putative location of Napata’s
capital, all of which have provided a wealth of information to deepen our knowledge of Nubian
occupation and history (Figure 1.1). In this region, the first kingdom of Sub-Saharan Africa arose
at Kerma around 2500 BCE (see Chapter 1, Figure 1.1), then further upstream the Napatan (ca.
800 BCE) and Meroitic kingdoms through the 4th century CE. To inform the period following,
Scores of Post-Meroitic burials and a few settlements have been investigated downriver of the
Cataract (e.g. El-Zuma). During Medieval times, churches and fortified towns were built along
the Nile and were part of the Makurian kingdom. At its center was the capital of Old Dongola
where excavations have revealed extensive settlements, citadels, churches, tombs as well as
evidence of large-scale industry during this time of stability (Godlewski 2014).
Another of these 4th Cataract sites has multiple cultural horizons and the focus of this
chapter. Deep in the ancient Napatan heartland, the archaeological site of El-Kurru (18.406569,
31.773974) is located in northern Sudan (Figure 2.1A) on the west side of the Nile River, near
the 4th cataract (Figure 2.1B). This site is well known as the burial place for Napatan (25th
dynasty BCE) kings and queens beneath pyramids and the presumed first capital of the
powerful kingdom of Kush (Figure 2.2, 2.3) (Welsby 1996). Deeper excavation at this site has
revealed another cultural horizon beyond the Kushite one. Facing the Nile River, a fortified town
was uncovered dating to Medieval times, close to 2,000 years after the pyramids were built
nearby. Within the defensive perimeter, a complex of housing and other domestic structures

22
was built of mudbrick (Kendall 1992, Dann & Emberling, et al. 2016). Ceramic and midden
evidence indicated multiple phases of use from the 9th through the 14th century (Dann &
Emberling, et al. 2016). Currently, the modern town of Alkuro lies northwest of the defensive
wall and surrounds the plateau where the royal cemetery was built. These villagers trace their
ancestry to the Arabic kingdoms of the 14th century but may also have ties with the Medieval
Makurian kingdom which preceded it.

Figure 2.1. Modern Map of Africa and ancient Nile Valley with close up of 4th Cataract region.

A) Geographic location of Nile River Valley with ancient regions spanning modern Egypt and Sudan. B) Location of
archaeological field site El-Kurru (yellow star); map insert of 4th Cataract and surrounding Upper Nubia region.

Figure 2.2. El-Kurru Royal Cemetery plateau with Napatan Pyramid, temple in foreground.

El-Kurru was first excavated by George Reisner of Harvard University early in the 20th
century to explore the royal pyramids and burial tombs of the Kushite kings who ruled over
Egypt during the 25th Dynasty (747-646 BCE) (Figure 2.3, left). This royal cemetery offers a

23
unique opportunity to observe the development of the mortuary landscape of the Kingdom of
Kush (ca. 750-300 BCE). Here, a sequence of tombs become sequentially more elaborate while
appropriating themes from elite Egyptian culture: Nubian tumuli (large burial mounds),
rectangular structures with chapels, and steep-sided pyramids with underground tombs, which
was a uniquely Nubian feature (while Egyptians tombs were constructed inside pyramids).
Reisner excavated most of these burial structures including the central pyramid, a large rock-cut
funerary temple to the east (Figure 2.2), and exported the artifacts back to Boston. On the
periphery, he and his team also explored to some extent the areas surrounding the rocky
plateau. There, he uncovered what he postulated (incorrectly) to be an expansive Napatan
palace, which is known to be a fortified Medieval settlement (Figure 2.3, right lower).

Figure 2.3. Excavation map of El-Kurru, including royal cemetery, medieval wall, and Napatan
structures.

Left: Excavation plan of El-Kurru with pop-out (right) showing a close up of the defensive wall (in orange) and modern
features in this area (i.e. irrigation channels, etc.). Kurru cemetery map from Emberling & Dann, et al. (2013) with
modifications by author. Top: Radio carbon dating of samples from along the excavated city wall. Approximate date
range: ca. 600-1000 CE (Emberling & Dann, et al. 2013).

Very little work was conducted on the pyramids or the features in the periphery after the
exploratory excavations by Reisner, until the International Kurru Archaeological Project (IKAP)
was launched in 2013. Co-directed by Dr. Geoff Emberling (University of Michigan), Dr. Rachel
Dann (University of Copenhagen), and Dr. Abbas Sidahmed Mohammed Ali (University of
Dongola, Karima, Sudan), IKAP began new research initiatives at this site. This team of
international scholars has conducted archaeological excavation of the pyramid, its burial

24
chambers, and the settlement closer to the Nile River (Figure 2.3, Figure 2.4) as well as projects
on various topics including geophysical survey, cultural heritage, restoration and conservation of
art and architectural structures on site. This project sought to understand the multi-phase use of
this space and the significance of El-Kurru during the Kushite Kingdom, as well as engage the
local community who have become the stewards of their ancestral past.

Figure 2.4. Satellite view of El-Kurru excavations in relation to modern day agricultural and
domestic land.

Google map (2019) showing satellite view of the modern landscape, shaded blue area shows Medieval excavation
site in relation to agricultural land and a modern village. To the north, Royal pyramids on plateau labeled with a map
marker. Map obtained from Google (2019).

Recent work identified and completed excavation of structures from Reisner’s field
notes, including the stone block wall near the edge of the modern village (Figure 2.4). During
excavations, associated finds dated this settlement as Medieval Christian (ca. 600-1400 CE)
and it was later dated via AMS to be ca. 600-1000 CE (Figure 2.3 right upper), much later than
the Kushite pyramids. More than two meters thick, the wall ran parallel with the Nile in a
defensive manner facing the river and had multiple bastions along its 138-meter length,
including at the corners (Figure 2.3, Figure 2.5 left). The entrance was marked with a large
petrified log from the forest several kilometers away (Figure 2.5 left). The wall was built around
the 9th-10th century CE, followed by mudbrick domestic structures shortly after. These housing
structures within the wall had at least two phases of occupation and the presence of storage
jars, cooking pots, and grinding tools provide more evidence of a domestic use (Figure 2.5
right). Middens (trash/refuse piles) and ceramics from multiple Christian time periods suggested

25
an extended occupation (Emberling & Dann, et al. 2016). Following the second phase of the
domestic structures, the space within the wall was used for refuse dumping during the 12th-14th
centuries (Emberling & Dann, et al. 2016). The last phase of this space was the cemetery,
where at least 27 individuals were interred in approximate rows, stratigraphically on top of the
domestic layers, while avoiding the stone block wall (Figure 2.6). Given these graves cut into
this refuse area in this space, this cemetery is stratigraphically dated to post-14th century.

Figure 2.5. Excavation of Medieval Christian structures at El-Kurru.

Left: Entrance to defensive wall; threshold is a log of petrified wood (between the parallel walls). Right: Mudbrick
domestic spaces near wall entrance. Images from Dann & Emberling, et al. 2016.

Figure 2.6. Excavations centered on Medieval Christian settlement and cemetery.

26
Left: Map of settlement excavations, wall in orange (Emberling & Dann, et al. 2013). Blue box zooms in on cemetery,
wall entrance (east), and modern irrigation channel (west). Right: Expanded view of cemetery, graves outlined within
excavation squares. Courtesy of M. Uildriks.

The focus of this project concerns the skeletal remains excavated from the cemetery
within the town wall (Figure 2.6). As the probable inhabitants of this settlement, the skeletal
remains from this cemetery offered an unexplored perspective to understand more about those
who lived at El-Kurru. Anthropological analyses of these remains provide data about their
demography (i.e. age and biological sex), cultural practices and identities (i.e. burial treatments),
and life experiences (i.e. indicators of health status). Given the likely contextual date of post-14th
century and that Islam was officially instated in the southern most Christian Kingdom of Soba by
1493 CE with the establishment of the Funj Sultanate, this skeletal population may offer
important insights in to the increasing Arab influence on Nubian Kingdoms in the Middle Nile
region and impact of the transition to Islam (Hillelson 1930).

MATERIALS & METHODS

Excavation of El-Kurru Sample Population

Archaeological excavation of the Medieval cemetery found near the town wall and
anthropological evaluations of the human remains began in 2014 and continued through 2016
by various personnel associated with IKAP. Skeletal remains were excavated by Mohamad
Saad (National Corporation of Antiquities and Museums, Khartoum, Sudan), Tim Skuldbøl
(University of Copenhagen), Gretchen Emma Zoeller (Indiana University Purdue University,
Indianapolis), and Abagail Breidenstein (University of Michigan). The burials were uncovered
during the course of normal excavation of settlement architecture and domestic spaces. Within
the burials excavated, skeletal remains were exposed, photo-planned, and lifted. To limit the
risk of looting during excavation, each individual was excavated in one day. Outside the field,
remains were cleaned, sorted, and cataloged. Bone and teeth samples were collected for
ancient DNA analysis and exported from Khartoum, Sudan, to the University of Zürich,
Switzerland. All other remains were packed and exported from Khartoum to the University of
Michigan for detailed anthropological evaluation.

Age & Sex Estimation and Health Status Indicators

Age-at-death was estimated using as many features as possible, including degree of

epiphyseal fusion (Cunningham, et al. 2016), tooth formation (Thoma & Goldman 1960, Smith
1991) and dental eruption sequences (Buikstra & Ubelaker 1994, AlQahtani 2008) as well as (to

27
a lesser degree) occlusal wear or dental attrition (Gilmore & Grote 2012), the auricular surface
(Lovejoy 1984, Brooks & Suchey 1990), and the pubic symphyseal face (Buikstra & Ubelaker
1994). Dental attrition was observed or noted but not utilized to age an individual since
standards for this method are not calibrated for comparison with African populations in desert
environments. Biological sex was estimated with reference to non-metric cranial traits (Buikstra
& Ubelaker 1994), the morphology of the pubis (Phenice 1969), and the morphology of the
greater sciatic notch (Buikstra & Ubelaker 1994). Lastly, all pathological or traumatic features
were recorded and scored using standard reference texts and their methods as prompted on the
data collection sheets (Ortner 2003, Waldron 2009, Aufderheide & Rodriguez-Martin 1998,
Buikstra & Ubelaker 1994) or specialized methodologies specified below. These estimations
and observations were recorded following field data sheets “Human Remains Recording Sheet”
by Antione (2018).

Health Status Evaluation

To understand the life experiences of the El-Kurru individuals, the skeletal remains offer
a glimpse into these lives, especially those experiences which affected their health. In general,
the skeletal remains – cranial, dentition, and post-cranial – were examined to determine the
health status of every individual. Using a bioarchaeological approach, multiple tests are
conducted with this sampling to understand the stress response on a population level. These
tests record porotic hyperostosis, cribra orbitalia, linear enamel hypoplasias, and non-specific
lesions or periostitis (e.g. endocranial lesions, joint porosity/cribra). Skeletal tissue and enamel
formation are sensitive to insults that may increase or arrest these processes, such as
nutritional deficiencies, disease, or other hardships (Larsen 2015). The tests include evaluating
the presence or absence of health signals and usually a score for severity and/or status of
active or healing. Additionally, the presence of trauma was noted for all individuals with enough
preservation.
Cribra orbitalia (CO) (pitting lesions in the upper eye sockets) and porotic hyperostosis
(PH) (porous lesions on ectocranium) are both commonly assessed in skeletal remains from
archaeological contexts (Walker, et al. 2009, Ortner 2003). Currently, the exact etiology of these
lesions is hotly debated with possible causes including genetic anemias (any of the
thalassemias, G6PD deficiency, sickle-cell etc.), deficiency-related anemias including
malnutrition (rickets, scurvy, iron, B12 etc.), or those related to infectious disease (e.g. malaria,
hookworm infections) (Moseley 1974, Steinbock 1976, Walker, et al. 2009). The presence of CO
and/or PH is likely multifactorial, meaning these lesions are non-diagnostic and should be

28
considered with other stress indicators (Brickley 2018, McIlvaine 2013). Both conditions are
evaluated as per Rinaldo, et al. (2019). Briefly, individuals were included in the analysis if at
least one eye orbit was intact for inspection. One or both orbits were evaluated for the presence
or absence of lesions and the stage of severity and/or healing were both judged according to
the published chart and accompanying description of the lesions (Rinaldo, et al. 2019).
Next, when available, all dentition was checked for the presence of linear enamel
hypoplasias (Buikstra & Ubelaker 1994). Individuals were included in this test if at least one
anterior tooth (incisor and/or canine) was present for observations. Hypoplasias were noted as
present or absent and graded in severity and structure (i.e. pits or line) as per Buikstra and
Ubelaker (1994). LEHs present as lines, grooves, or a grouping of pits which encircle the crown
and can be macroscopically detected with a fingernail test (i.e. deep enough to catch when
running a fingernail over the surface of the tooth crown) (Buzon 2006). These defects are a
response to interruptions in the deposition of enamel during the formation of the tooth crown,
which varies from before birth to roughly 19 years old (Hillson 1996, White & Folkens 2005).
Disruptions in enamel formation can be caused by bouts of illness, malnutrition or diet
deficiencies, or local trauma (Goodman & Rose 1991, Goodman, et al. 1980). But differentiating
between causes is difficult, therefore the presence of these hypoplasias is seen as a general
marker for physiological stress (Larsen 2015, Goodman & Rose 1991).
Lastly, the presence of non-specific lesions is an important indicator of variable health.
Lesions are deposits of woven bone tissue as a response to inflammation (Larsen 2015,
Buikstra 2019, Waldron 2008). These lesions are judged as “non-specific” as most have a long
list of etiologies from infectious disease, genetics, environmental hazards, trauma, behaviors,
nutrition, etc., but are informative of an unhealthy environment and whether adaptation is
required to survive this hardship (Buzon 2006, Goodman & Armelagos 1988).

Dating of Skeletal Remains

Two individuals with good preservation from different depths and excavation squares
were chosen for AMS carbon dating. Bone and tooth samples from Ind. 206 (from area B249)
and Ind. 214 (from area B451) were sent to ETH Zürich (Department of Physics, Zürich,
Switzerland) for collagen extraction and c14 dating. When results were unsuccessful (See
Results), additional samples were sent to the University of California, Irvine (UCI), to the ESS
KECK/AMS Facility for troubleshooting with knowledge of the low collagen levels. Samples from
Ind. 212 (cranial fragments), Ind. 214 (mandible fragment), and Ind. 206 (mandible fragment)
were trialed for radiocarbon dating, but were again unsuccessful, likely due to poor preservation.

29
RESULTS

Excavation and preservation

The skeletal remains from the El-Kurru settlement were excavated over three
consecutive field seasons: two in 2014, ten in 2015, and fifteen in 2016. Over the course of
three excavation seasons, squares B350, B450, B249, B349, and B449 were excavated to
bedrock. Squares B351 (half opened) and B451 were not exposed to bedrock and are expected
to contain more individuals (Figure 2.6). In situ burials in the northwest balk of B350 confirms
that the cemetery extends further in the southwest and northwest directions and contains other
individuals. Due to time constraints and the proximity of the modern irrigation channel, these
squares remain unexcavated. The Reisner trench (within B348 and B448) may have contained
burials and still does, as indicated by the in situ burials remaining within the northeast corner of
B448 and another burial in B447. In general, the bodies were not buried in the vicinity of other
obvious structures, except the fortification wall. However, two individuals were excavated in
close proximity to mudbrick structures (i.e. Ind. 216 was between a mudbrick wall and door
jamb, Ind. 210 underneath a partial wall collapse), but these instances did not invalidate the
dating of the cemetery as ca. post-14th century.
In general, the skeletal remains found in the El-Kurru settlement were in various states
of preservation with most skeletal remains weathered, brittle, and mostly fragmented. Remains
from 2014 and 2015 were in fair to poor preservation, while those from 2016 were better
preserved. In contrast, the teeth were in better condition than the skeletal remains for most
individuals, allowing for a thorough evaluation of dental health and disease as well as sampling
for molecular analyses, (e.g. stable isotopes and ancient DNA). Preservation of the remains was
compromised by the cemetery's location near several water hazards – namely a large wadi
(seasonally active water drainage route) to the west, active agricultural fields to the east, and a
modern irrigation channel northwest of the cemetery (Figures 2.3 & 2.5). There was evidence of
several other hazards impacting the taphonomy of these remains, including but not limited to a
palm grove to the south (lateral roots systems damaged the skeletal remains), animal
scavenging, and heavy weathering from the arid climate. Lastly, there was no evidence of
looting or disturbance of burials.

Dating of Skeletal Remains

Despite being tested at two highly-reputable facilities and using the (perceivably) best
preserved biotissues, AMS dating was unsuccessful. Results from the testing at ETH are listed

30
in Table 2.1. Specifically, C/N ratios should be lower and collagen extraction by weight was not
enough. Results from UCI were reported as unsuccessful due to poor preservation which led to
insufficient collagen extraction. Therefore, the individuals interred in the cemetery will remain to
be contextually dated to ca. post-14th century.

Table 2.1. Results from AMS Radiocarbon dating performed at ETH Zürich.
Nitrogen Carbon Content Total Collagen
Sample Tissue Type C/N Ratio
Content (%) (%) Weight
IND. 206 (B249) Bone (Mandible) 0.56 2.33 4.1572 241.4 mg
Tooth 0.38 0.79 2.1007 --
IND. 214 (B451) Bone (Mandible) 0.39 2.43 6.2510 184.5mg
Tooth 0.41 1.45 3.5251 --

Anthropological Profiles of El-Kurru individuals

The El-Kurru skeletal sample is composed of 11 subadults (40.7%) (aged less than 20
years of age) and 16 adults (59.3%) (aged 20+ years), as summarized in Figure 2.7, with an
approximate mean age-at-death of 25 years. For biological sex, the adult group consists of nine
males (or probable males), five females (or probable females), and two of ambiguous or
indeterminate sex (Figure 2.7). Subadults were divided into five categories: Infants (0-1 year),
Early Child (2-5 years), Late Child (6-10), Adolescent (11-15 years), and Pubescent (16-19
years). Adults were divided into three age ranges: young adults (20-35 years old), middle adults
(35-50 years old), and older adults (50 years or older). No fetal or prenatal remains were found.
During excavation, it was found that preliminary assessment of biological sex was easier to
determine with in situ remains than after lifting the bones, especially with the pelvis. For an
unknown reason, the remains of male individuals were better preserved overall compared the
females and may present a bias for later analyses. Sub-adults were not assessed for biological
sex since puberty has not been reached. For adults, cranial features seemed to be misleading
while the pelvis seemed to be much more reliable. From experience of the author, the
population tended to have more gracile features, especially of the skull, and less robusticity
overall; therefore, features of the pelvis were weighted more than cranial ones for final
estimations of biological sex.

31
Figure 2.7. Mortality Distribution of 27 Christian Individuals from El-Kurru cemetery.

Age and sex distribution for El-Kurru sample population, N = 27. The y-axis is the number of individuals classified in
each age category. Colored bars represent estimated sex of each individual: blue – subadults, not assessed, orange
– male or probable male, yellow – female or probable female, grey – ambiguous sex, could not be determined due to
preservation or intermediate scores). X-axis lists age range (in years): Infant, Early Child, Late Child, Adolescent,
Pubescent, Young Adult, Middle Adult, Older Adult (respectively).

Within this sample of 27 individuals from El-Kurru, all age classes were present as well
as both sexes within the adult categories. The demographic data from this population showed a
typical “U-shaped” mortality curve, with some notable exceptions. In a “normal” graph, those
experiencing the most risk are the young and old (represented as larger groups for age-at-
death, or taller columns, at the far left and right, respectively), while the robusticity of the older
subadults and young adults are less at risk, thus creating a U-shaped graph when counts are
tallied from a skeletal population (Jackes 2011). Exceptions to this pattern seen in Figure 2.7
were a lack of infants or neonates (0-1 years) and middle adult males (36-50 years). The
missing infants from this sample may suggest these individuals are buried elsewhere (e.g. in the
home or a separate cemetery) (Pinhasi & Bourou 2007). The absence of males in the middle
adult category suggested either better survivorship that could be evidenced with the abundance
of older male adults (>50 years), or these individuals were dying elsewhere, which could be
expected given their proper age for military contribution. More broadly speaking of the 16 adults,

32
males and females were not equally dispersed (Figure 2.7), with a sex ratio of 180 (expressed
as the number of males per females). Additionally, given this mortality distribution of this
sample, more than 40% of the population did not reach adulthood. Eleven individuals were
categorized as subadults ranging from less than one year of age to 17.5 years old. The first
three categories (N = 8) suggested a mortality of nearly 30% for individuals under the age of ten
for the subpopulation represented at this cemetery. This is well within the rage of premodern
populations and this death curve is typical of ancient populations: on average, in the past two
millennia youth (less than 15 years of age) mortality is 46.2% and infant mortality is 26.9% (Volk
& Atkinson 2013). Figure 2.7 shows an abundance of individuals dying young – half of which are
under the age of five – then a typical drop off in the middle phases, speaking to the robusticity
and survivorship of the juveniles making it to adulthood. The timing of this life stage relates to
surviving weaning after early childhood, known to be a challenging stage for infants and young
children (Lewis 2007). The risk of death then steadily increases as individuals enter phases of
reproduction and more adult activity. For this population, however, the distribution of those dying
in adulthood is atypical, with more dying younger than those in the middle adult range. This is
possibly due to unequal or higher risk of death for females. However, it is more likely the result
of a small sample size. Lastly, the category with the most adults is the older adults, representing
individuals that died after the estimated age of 50. For the adults, the sex distribution is not
equal after the young adult category – suggesting a high risk of death for women dying younger
and males surviving longer to over 50 years. This could indicate females had a hard life in this
Medieval community. Overall, mortality peaks for older adults may actually indicate a well-
adapted population for this region. These two peaks – high mortality in the first few years
(approx. 19%) and later in adulthood (approx. 30%) – are typical; however, it seems more
atypical that adults were living to an advanced age in an ancient population. While it was not
unheard of, it is rare for people in ancient times to live past their early sixties. For example,
Roman working-class commoners had a mean age at death of 30 (Caldarini et al. 2015), while
those from various Nubian time periods is between 26 and 36 years (Armelagos 1969).

Mortuary Archaeology of El-Kurru Cemetery

A total of 27 individuals were excavated from the cemetery. Unmarked graves were
roughly organized into three or more rows running northeast to southwest, which were parallel
to the fortified wall and the bank of the Nile (Figure 2.6). Fully extended bodies were arranged
perpendicular to the town wall, head to the northwest and feet to the southeast. Graves were
dug as simple pit inhumations with a thin oblong shape, which were adapted to the size of the

33
individual with no evidence of coffins, boxes, or beds. Burials were typically single occupation,
with the exception of one double burial with an adult and a child (Ind. 106 and Ind. 113). Bodies
were buried about 1.5 feet (half a meter) from the modern surface up until a depth of 4 feet (1.2
meters) at the bedrock. Adult individuals were not buried with grave goods (e.g. vessels,
jewelry, etc.), but some children or infants had small tokens. Small beads of ostrich egg shell or
faience were found near the neck or wrists for two subadults suggesting they were likely buried
with simple jewelry. The burial style and lack of grave goods is consistent with typical Christian
burial treatments.
Some grave markers were thought to be associated with burial pits, although these
stones could also have been deposited serendipitously with a body. For example, a carved
sandstone block was in context with burial 206, indicating that it could possibly be a headstone.
During excavation, the stratigraphy of the bodies (ie. overlapping, varying depth) indicated
multiple phases of use. For example, burials 206 and 209 overlap in such a way that Ind. 209’s
grave cuts Ind. 206’s, which was excavated first. This could be explained by those maintaining
the cemetery forgetting where the first body was buried without the use of grave markers.
Organics from the wall date the fortification to ca. 600-1000 CE (Figure 2.3). The burials
were located at a higher elevation than the threshold of the fortified wall. Furthermore,
stratigraphy and the arrangement of the burials to the wall suggested that the burials post-date
the wall and were interred later than the 14th century. Given that the treatment of the bodies
exhibited typical Christian characteristics, it is estimated the group of individuals date to after the
upper limit of the wall’s contextual dating. Other finds associated with the graves included
decorated pottery sherds in the grave fill that are not considered grave goods, but likely refuse
mixed with the dirt used to fill graves. Sherds were dated to Classic Christian (ca. 800-1100 CE)
or Late Christian (ca. 1100 – 1450 CE) periods based on painted motifs (e.g. “twisted rope
motif”) and fine ware symbols (Figure 2.8A & C) (Klimaszewska-Drabot 2008, Phillips 2003).

Figure 2.8. Artifacts uncovered within grave fill for El-Kurru burials.

34
Examples of contextual finds from the grave fill used for dating. A) Example of pottery sherds in context with skeletal
remains used to date burials; white fabric with “twisted rope motif” are Classic Christian (ca. 800-1100 CE) and used
a finer fabric than red or black course ware. B) River “prayer” pebbles. C) White fine ware with cross symbol.

Another observation of note for those buried at El-Kurru was the ubiquitous presence of
colorful river pebbles found in the top layers of the grave fill (Figure 2.8B). Found on the banks
of the Nile, river pebbles were prayers or tokens of affection for the dead. This tradition has a
deep history in this region of Sudan as far back are the Meroitic period (Cavendish 1966). It is
common during the Christian period (e.g. El-Ar (early Christian 6th cent.) and Nuri (ca. early
Christian)) and has continued into the present (e.g. Manasir) (Cavendish 1966, Zurawski 2007).

Burial Treatment

Most graves did not have a super structure. A few individuals had a rough mud cap over
the grave, shallowly constructed 0.5m above the body (e.g. Ind. 201). There was no evidence of
coffins or box graves within the pit (structure made of flat rocks outlining the body). Although it
was expected to find traces of a burial shroud, as is typical of Christian style treatments (Welsby
2016), no organics or fabrics were preserved. However, for those buried supine, features of the
articulated body suggested that burial shrouds or some materials were possibly used for
binding. For example, the clavicles had a characteristic verticalization arrangement (Harris &
Tayles 2012, Nilsson Stutz, 2003) (Figure 9, blue arrow) in situ caused by the arms being drawn
tightly to the body in a close bind and for some individuals, hands were very closely clasped
together (Figure 2.10C). This suggests that burial shrouds could have been in use and likely
decomposed as a result of the inundation of the area since antiquity.

Figure 2.9. El-Kurru Ind. 203 showing extended and supine burial.

Close-up photograph of Ind. 203 within the grave before lifting. Blues arrows point to clavicles, in situ.

35
 Each individual shared three characteristics of how the body was arranged during burial:
clasped hands, fully extended, and supine (body facing upwards) or side-facing placement
(Figure 2.10A-C). The bodies of all individuals were fully extended, never flexed. Sometimes,
bodies were turned to fit into tight spaces for burials. Most bodies were facing to either side but
not in a uniform way, specifically 15 to the right (65%) and eight to the left (35%), while four
were supine. Body position may have shifted during decomposition as indicated by twisted, but
still articulated legs (Figure 2.10B). This suggested that the body started in a supine position
then may have turned after burial (e.g. Ind 207). All burials had clasped hands positioned over
the pelvis, regardless of the body arrangement. This also indicated the use of binds at the time
of burial and during the decomposition or skeletonization of the body (Welsby 2016). All bodies
were oriented with the head to the northwest (west when aligned to Nile) and feet to the
southeast (east when aligned to the Nile) (Figure 2.5). All individuals were found in anatomical
position with little to no disruption of the skeletal remains post-deposition. The only exception
was Ind. 209, where the left arm and hand was found cutting the grave cut of Ind. 206;
additionally, a few epiphyses of 209’s hand bones were found inside the mouth of 206.
Little variability was identified in the El-Kurru cemetery population. All individuals appear
to be uniformly treated. There was no apparent relationship between body positioning or burial
orientation with relation to sex or age of the individuals (Table 2.2). The treatment of the body
(i.e. extended, bound) was uniform with no variation based on the age or sex. Lastly, interment
position within the whole cemetery does not show a pattern based on age or sex; for example,
children and infants were buried among adults, instead of being buried elsewhere.

Figure 2.10. Burial treatment and body treatment of El-Kurru individuals.

36
Examples of body treatment of individuals buried at El-Kurru. A) Fully extended body positioning, e.g. Ind. 104. B)
Body placed in a supine or side-facing position, e.g. Ind. 207. C) Hands probably bound by cloth or a shroud, without
preservation of textiles or organics, e.g. close up of Ind. 210.

Table 2.2. List of El-Kurru Individuals and demographic data.

Age at
Ind. No. Season Death Sex β Body Position Burial Orientation
(in years)α
004 2014 3-5 SA Extended, right side W-E, facing S
005 2014 20-35 M Extended, supine W-E, facing up
104 2015 20-35 M Extended, right side W-E, facing S
105 2015 20-35 F Extended, left side W-E, facing N
106 2015 36-50 I Extended, left side W-E, facing N
107 2015 16 F Extended, right side W-E, facing S
108 2015 50+ M Extended, right side W-E, facing S
109 2015 7-9 SA Extended, left side W-E, facing N
110 2015 36-50 F Extended, left side W-E, facing N
111 2015 36-50 F Extended, supine W-E, facing S
113 2015 12.5-15.5 SA Extended, right side W-E, facing S
115 2015 <1 SA Extended, right side W-E, facing S
201 2016 50+ M Extended, right side W-E, facing down
202 2016 7.5 SA Extended, left side W-E, facing N
203 2016 20-35 M Extended, supine W-E, facing S
204 2016 50+ I Extended, right side W-E, facing S
205 2016 50+ F Extended, left side W-E, facing down
206 2016 20-35 F Extended, supine W-E, facing up
207 2016 50+ M Extended, right side W-E, facing S
209 2016 17.5 SA Extended, left side W-E, facing N
210 2016 50+ M Extended, right side W-E, facing S
211 2016 4.5 SA Extended, right side W-E, facing up
212 2016 3.5 SA Extended, right side W-E, facing S
213 2016 5.5 SA Extended, right side W-E, facing S
214 2016 5.5-6.5 SA Extended, left side W-E, facing N
215 2016 50+ M Extended, right side W-E, facing S
216 2016 50+ M Extended, right side W-E, facing S
α Cunningham et al 2016, Thoma & Goldman 1960, Smith 1991, Buikstra & Ubelaker 1994, AlQahtani 2008, Gilmore
& Grote 2012, Lovejoy 1984, Brooks & Suchey 1990
β Buikstra & Ubelaker 1994, Phenice 1969

+ = more than
< = less than
F = Female or probable female
M = Male or probable male
I = Indeterminate sex or could not be determined as male or female

37
SA = Subadult, no sex estimate
W-E = skull at west end, feet at east end; directions not cardinal, oriented toward the river

Health Status Evaluation

Cranial remains were evaluated for cribra orbitalia (CO) as per Rinaldo, et al. (2019).
Fifteen of the 27 individuals (ten adults and five subadults) were able to be evaluated and those
not included had poor preservation. Of the ten adults, two females and two males were affected
or 40% of the sample. These results show no bias for sex, although this is a small sample. For
the five subadults evaluated, four or 80% exhibit CO. Age estimates range from 3.5 – 17 years
old. The lesions judged were most commonly at stage 3, or “presence of holes that join within
inner bone tissue” (Rinaldo, et al. 2019) (Figure 2.11, Figure 2.12). In general, all lesions were
in some stage of healing, none were judged as active. The most common stage observed was
stage 2 (Figure 2.11, Figure 2.13). Taken together, 40% of the population was estimated to
have moderate signs of CO, but the lesions were in remission. This would suggest these
individuals were vulnerable to stress, but adapted enough to live through the hardship.

Figure 2.11. Cribra orbitalia severity and stages of healing observed in El-Kurru population.

Left: Counts for stages of severity observed in those individuals with orbit lesions or cribra orbitalia (dark blue). Lower
stages refer to small area affected and small pores, while higher stages have a large affected area with deep, joining
pores. Right: Counts for stages of healing observed in those individuals with orbit lesions or cribra orbitalia (light
blue). Lower scores describe no signs of healing, while higher stages are for healing or healed lesions. Counts above
bars. Y-axis= count of individuals.

38
Figure 2.12. Stages of severity observed in cribra orbitalia lesions in El-Kurru population.

Individuals used as examples: Stage 1 = Ind. 205 (left orbit), Stage 2 = Ind. 212 (right orbit), Stage 3 = Ind. 203 (right
orbit).

Figure 2.13. Healed lesions observed in El-Kurru population.

Individuals used as examples. Stage 2 healed = Ind. 105 (right orbit), Stage 3 healed = Ind. 214 (left orbit). Of note
are the rounded margins of the pores within the lesions.

Porotic hyperostosis (PH) could be evaluated in 16 of the 27 individuals (i.e. more than
50% cranial surface observable). Those not included were due to poor preservation of the
crania. The severity and healing stages were scored as per Rinaldo, et al. (2019) and Stuart‐
Macadam (1989). Of the 19 evaluated, only three were affected: Ind. 203, a young male adult,
Ind. 205, an older probable female adult, and Ind. 216, an older male adult. Thus, PH is
observed in roughly 19% of the population. All individuals were at stage 2 severity, and were
either stage 2 or stage 4 healing (Figure 2.14).

39
Figure 2.14. Porotic Hyperostosis in El-Kurru population.

Examples of Porotic Hyperostosis (PH) found in El-Kurru population. Top: Ind. 203 right parietal and frontal bones.
Bottom: Ind. 203 frontal bone and Ind. 216 parietal bones.

For the evaluation of linear enamel hypoplasias (LEH), 22 individuals from El-Kurru
could be included in this test (Figure 2.15). Five were not included due to poor preservation of
the teeth or attrition (tooth wear) was so severe that enamel could not be viewed. Of these 22
individuals, 16 (73%) had at least one tooth with an LEH and 6 (27%) had none. Those affected
included all ages and both sexes: 7 subadults ranging in age from 3 – 17 years and 9 adults,
including 2 females, 2 of indeterminate sex, and 5 males. This sex bias may indicate males
were more susceptible to hardships, but the sample size is very small. Despite heavy attrition,
LEHs were very common among adults and subadults.

40
Figure 2.15. Linear Enamel Hypoplasias found in El-Kurru population.

Blue arrows point to linear grooves or lines where enamel production was arrested. Ind. 209: right mandible. Ind. 109:
mandibular incisors (and some calculus build up). Ind. 107 left: unerupted third molar. Ind. 107 right: right maxillary
teeth. Ind. = individual

Endocranial lesions were observed in 2 of 18 individuals able to be evaluated, (i.e.

preservation of more than 50% of the cranium). The other individuals had crania too
fragmentary to observe the inner table of the skull or for which reconstruction was impossible.
Ind. 004 (child 4.5 years old) had extensive, active (i.e. “hair-on-end”) lesions within the
endocranium, while those found within the skull of Ind. 211 (child 4.5 years old) were already
well healed (Lewis 2017) (Figure 2.16). The etiology of such lesions, especially found in
children, is still debated; from non-pathological to vitamin deficiency to infectious disease (Lewis
2004, Wolbach & Bessey 1941, Mitchell 2006).

41
Figure 2.16. Two instances of endocranial lesions observed in El-Kurru population.

Both examples found on occipital bones of skull. Blue arrows show areas of where the lesions are located. Ind. =
individual

Porous lesions found on the humerus (cribra humeralis) and of the femur (cribra
femoralis) were quite common, but only among the younger individuals. Eighteen individuals
had adequate preservation of the post-cranial skeleton to observe the joints where these lesions
appear. Seven out of 18 (39%) showed porotic lesions at the proximal metaphysis of the femur
and/or the distal metaphysis of the humerus (Figure 2.17, Figure 2.18). These lesions are often
associated with each other (Smith- Guzmán 2015b). All instances were bilateral, occurring in
both the arms and leg bones, and characterized by deep pores penetrating the cortical bone.
One individual (Ind. 005) was a young adult (20-35 years old) while the six others were aged
between approx. 4 - 17 years old. The etiology of these lesions is poorly understood and not
clinically linked with anemia or a deficiency in the way CO or PH is (Lewis 2017). For those
individuals that could be evaluated for CO (i.e. preservation of at least one orbital roof), all (N=4)
also had CO. Together, this is known as the “cribrous syndrome” and is the result of red bone
marrow expansion during red blood cell proliferation, the subsequent thinning of the cortical
bone superficial to marrow, and the substitution of the trabecular bone underneath (Miquel-
Feucht, et al. 1999a,b, Djuric, et al. 2008). Furthermore, the association of these types of
lesions with malaria has also been examined by Smith-Guzmán (2015b). Using the lesion
presence or absence algorithm, two individuals can be diagnosed with malaria: Ind. 206 and
Ind. 212. The lesions are in variable states of activity; those present in Ind. 209 seemed quite
sharp or active, while those present in Ind. 212 have more rounded edges and have signs of

42
healing (Mensforth, et al. 1978) (Figure 2.17). Erosion and taphonomic damage made it difficult
to judge these lesions consistently.

Figure 2.17. Cribra femoralis found in the El-Kurru population.

All proximal ends of femora. Blue arrows point to lesions on cortical bone. Ind. = Individual

Figure 2.18. Cribra humeralis found in El-Kurru population.

All distal ends of humeri. Blue arrows point to lesions on cortical bone. Ind. = Individual

43
 Four instances of trauma were observed in the entire El-Kurru population (15%) (Figure
2.19). All were well-healed fractures as evidenced by the boney callus and subsequent
resorption of lamellar bone. All fractures occurred in quite mundane locations of the skeleton,
specifically a right second toe (Ind. 216), a left upper arm (Ind. 205), and a right clavicle (Ind.
210), and all to older (50+ years or older) adults of both sexes. All of these injuries are
extremely common, especially the right clavicle that is the most commonly broken bone in the
human body (Arizona Science Center). The healed fracture in Ind. 204 (indeterminate sex) is
classified as a “parry fracture,” or an injury associated with a defensive wound from a frontal
(likely right-handed) attack (Judd 2000). This was a left ulna (forearm bone) and the fracture
only involved this bone, not the adjacent radius, and was located at the distal (closer to hand)
end (Figure 2.19, Ind. 204). These criteria fit the definition of a “parry fracture” exactly and the
fracture was well healed indicating the individual likely sustained the injury early in life.

Figure 2.19. Four instances of trauma found in El-Kurru population.

Ind. 204 “parry fracture” of left ulna. Ind. 205 healed left humerus. Ind. 210 (poorly) healed right clavicle. Ind. 216 right
second proximal metatarsal. Blue arrows show the boney callus, all healed fractures. Ind. = Individual

44
Table 2.3. Summary Table of Health Status observations.
Test No. evaluated No. Afflicted Percentage
Porotic Hyperostosis 16 3 19 %
Cribra Orbitalia 15 8 40 %
Linear Enamel Hypoplasias 22 16 73 %
Endocranial Lesions 18 2 11 %
Joint Cribra 18 7 39 %
No. – number of individuals

DISCUSSION

Archaeological records of late Medieval Nubia in the 4th cataract region in Upper Nubia
are sparse (Edwards 2013). The excavation of the cemetery at El-Kurru and its subsequent
bioarchaeological analyses offers a perspective of the climate at the end of the Christian time
period in the Middle Nile Region. Political constructs began to weaken after the first Muslim king
was put in place up river at Dongola in the early 13th century (Welsby 2002). There is also a
possibility that plague, which spread through Egypt and Ethiopia, played a part in the instability
at this time (Borsch 2005). Contextually dated to post-14th century, the cemetery at El-Kurru
contained at least 27 individuals culturally associated to Christians as per associated pottery
sherds, grave constructions, and body treatment (i.e. extended position, arms at the pelvis,
supine or side-facing). Using bioarchaeological evaluations, the presence of stress signals were
detected in this skeletal population that can identify if these individuals were living in unhealthy
environments. Since the etiologies of these disruptions are not easy to discern and health itself
is influenced by a multitude of factors, stress markers were taken together to portray a more
general view on stressors in the population, which can originate from biological, environmental,
social-political, etc. causes (Goodman & Martin 2002, Larsen 2015). In the El-Kurru population,
high frequencies of LEHs indicate stressors during developmental years (< 20 years) and the
presence of distinctive lesions suggest parasitic infections were common among the subadult
population with two probable cases of malaria.
In general, the percentages of stress signals (CO, PH, LEH, other lesions) detected in
the El-Kurru population (Table 2.3) were consistent with those reported from this region and
time. The appearance of porotic hyperostosis (19%) was well within the context of other data
reported on the presence of these lesions. From Lower Nubia, PH was present in 4% - 68% of
Christian individuals, while for subadults the frequency range was 40 - 80 % (Van Gerven, et al.
1981, Carlson, et al. 1974, Hurst 2014, Soler 2012). The presence of PH begins to build an
overall narrative of a stress response and can be an indicator of anemias, metabolic disorders
(e.g. scurvy), chronic infectious disease, or malnutrition (Buikstra 2019, Brickley 2018). Next,
the presence of CO was also within the range of reported populations in the Northeast Africa.

45
CO was recorded in 40% of the individuals at El-Kurru, with 80% of those being subadults.
Among past Nile River Valley populations, instances of CO range ca. 30 – 60 % and malaria
has been suggested as the causative agent being present at endemic levels from prehistoric to
Christian time periods (summarized by Smith-Guzmán 2015a). Specifically, rates from Upper
Nubia were on the lower side at some sites, e.g. 11% all individuals, 43% subadults from
Tombos, 21% all individuals, 32% subadults from Wadi Halfa (Carlson, et al. 1974, Buzon
2006), but higher in others, e.g. over 60% at Amarna West or 45% from Kulubnarti (Mittler &
Van Gerven 1994, Binder 2014, Hummert & Van Gerven 1994). The cause of these lesions was
speculated to be from nutritional deficiencies, namely iron, from a millet and wheat-centered diet
(Carleson, et al. 1974). However, a heavy disease burden should not be discounted, as malaria,
hookworm, and schistosomiasis are also endemic in the region presently and likely present in
the past (Hibbs, et al. 2011, Larsen 2015). In Upper Nubia, Christians from Mis Island show
49% prevalence, with more than 80% of those affected being subadults (Soler 2012, Hurst
2014). The prevalence of CO from El-Kurru and the higher rates in subadults is consistent with
other populations close in space and time. Given the lesions are all in healing states (non-
active), this suggests stressors likely occurred in the early years of life and individuals lived
through the stressful conditions, for example iron deficiency anemia triggered by poor nutrition,
weaning, or bouts of infectious disease.
The high frequency of LEHs (73%) present in the El-Kurru population is consistent with
other prevalence data reported from Christian Nubia. At Kulubnarti, (300-1400 CE) in Lower
Nubia, every individual examined had at least one hypoplasia event recorded, while at Mis
Island roughly 50% were affected (Van Gerven, et al. 1990, Soler 2012, Hurst 2014). While
methodologies may be responsible for this difference, sampling of Nubian groups deeper in time
showed similar variation in values. Two surveys conducted during the New Kingdom period and
after (ca. 1400-800 BCE) show that 21% of individuals at Tombos displayed LEHs, while 69% of
individuals from Amarna West showed these lesions (Buzon 2006, Binder 2014). While each of
these frequencies were interpreted within the context of these sites, especially varying political
or cultural situations, the etiology of these lesions can be attributed to high chronic stress during
childhood when the teeth are forming (Goodman & Rose 1990, Hillson 2008). These stressors
may be a consequence of an iron-deficient diet, a heavy disease load, or weaning. Thus, the
high frequency found in the Late Christians from El-Kurru would indicate a majority of those
living at this fortified town experienced hardships when growing up, but given the abundance of
older individuals, these Nubians were surviving to old age. Lastly, other lesions present (joint
cribra and endocranial lesions) in the El-Kurru population have not been evaluated in other

46
Nubian populations for comparison. However, another dataset from Medieval Serbia attributed
high frequencies of the “cribrous syndrome” in subadults to parasitic infections, causing
hemolytic anemia and/or diarrhea (Djuric, et al. 2008). Given that these non-specific lesions
occurred in the subadult population from El-Kurru, their occurrence speaks to a heavy disease
burden on the youngest individuals of the population.
Four instances of healed, non-lethal trauma (14%) implies the El-Kurru population was
one not likely in conflict (Figure 2.19). The fractures observed in this selection of individuals are
mundane injuries or accidents. The clavicle is the most broken bone in the human body, and
this fracture in Ind. 210 (Figure 2.19) is the right side no less; it is evidence of a fall, where the
man stuck out his right hand to brace the fall, shattering his collarbone as the force radiated up
his arm. It was poorly set and the lack of woven bone indicated the callus has been smoothed
away by many years, even decades, of healing. Another example, Ind. 216’s right foot with a
broken second metatarsal. One can imagine this man dropping something on the top of his foot,
breaking only that one bone. Again, this break has a smooth callus speaking to years of healing.
A few of these foot bone injuries also have been recorded in other New Kingdom individuals
(Binder 2014). The other two instances, the broken humerus of the older woman and the parry
fracture for Ind. 204 (indeterminate sex) possibly suggest interpersonal violence. The arm
fracture was well healed and there are myriad scenarios, with and without aggression, to
reconstruct this injury. However, the “parry fracture” observed on Ind. 204’s left ulna is a more
convincing signal of violence (Judd 2000). However, no other signs of interpersonal violence
were recorded, for example head lesions. From Lower Nubia, 14% of individuals display head
lesions from Meroitic through Christian time periods (Armelagos 1969). On the other hand, post-
cranial lesions of violence are low (ca. 4%) in the earlier periods and increase to 11% in
Christian times, indicating a possible uptick in aggression (Armelagos 1969). These are both
much lower than occurrences recorded in the Kerma period (ca. 2500-1500 BCE) which ranged
from 40-80% with at least one injury (Judd 2002). Taken together, the frequency of trauma at El-
Kurru was low, suggesting that this sample of the population was not experiencing intra-group
or inter-group violence.
The limitations of this study include the sample size, the preservation of remains, and
the inability to radiocarbon date these remains. The cemetery was not completely excavated
and therefore more individuals could have been included. However, even with only 27
individuals, demographically all categories were covered with two notable exceptions (infants
and middle adults), implying that this was approximately a representative sample (Jackes 2011).
Moreover, the preservation of these skeletal remains impacted the post-field analyses and

47
molecular analyses. The heavy weathering and severe water damage made these bones
chalky, brittle, and diagenetically compromised. Several individuals were excluded from post-
cranial analyses due to poor preservation. This impacted the bioarchaeological recordings (e.g.
difficult to judge the severity of lesions due to erosion, etc.), the molecular extraction of DNA,
and both attempts at radiocarbon dating. These previous two failures likely indicate collagen
levels were damaged to the point of no recovery, likely from the heat, but perhaps more so from
the constant or seasonal presence of water that leached the protein out of the bone tissue
matrix.
Future directions of this work would include fitting these individuals within the context of
more Medieval Nubian sample populations in the Middle Nile region when these data become
publicly available. This could include more bioarcheological studies or those which explore the
identity of these Nubians expressed through their mortuary archaeology. Another perspective
would be to pilot stable isotopic analysis (i.e. carbon and nitrogen) of the El-Kurru individuals to
further contextualize signals of the childhood stress and the presence of lesions. From previous
work, Nubians generally had a mixed diet between C3 and C4 plants (i.e. wheat and barley, fruits
and vegetables; sorghum and millet, respectively) and had little contribution of protein from land
animals (not aquatic ones) (Basha, et al. 2016, White & Schwarcz 1994). Modern villagers in the
region, including those living in Alkuro, subsist as sedentary agriculturalists and it is likely that
little has changed since Christian times. Specifically, a stable isotopic study could shed further
light on their means of subsistence and test if malnutrition was a challenge. Lastly, since there is
so little knowledge, the plague hypothesis could be tested using pathogen screening of the
remains (reviewed in Spyrou, et al. 2019). This would be limited in scope (i.e. small number of
individuals) and likely not viable given the little success with other molecular analyses due to
poor preservation. However, any vascularized tissue may be used to screen for the Yersinia
pestis genome (Spyrou, et al. 2019). Plague would explain the many deaths of younger
individuals but there are no indications of special body treatment (e.g. encased in lye), which
would lend credence to this hypothesis.

CONCLUSION

Recently, a small Christian cemetery at El-Kurru, Sudan, was excavated in context of a

fortified, Medieval town close to the Nile River. Mortuary treatment of the bodies was consistent
with Christian traditions, as were the pit graves and a general lack of burial goods. Demographic
analysis revealed both sexes and all age categories were buried here, from infants to older
adults. A slight sex bias was observed where younger females were dying sooner than older

48
males and subadult mortality was normal, less than 20%. Bioarchaeological analysis of these
individuals recorded markers of stress, including high levels of childhood hardship or
malnutrition and likely exposure to parasitic infections, which affected the subadult population
the most. The lesions found in two subadults were likely from malaria. Additionally, only one
case of trauma which can be attributed to interpersonal violence suggesting this population was
not exposed to interpersonal violence or those who did were buried elsewhere.
This sample population from El-Kurru represents a group of individuals living at a time of
transition, the end of Christianity and the rise of Islam in the Middle Nile region. There does not
seem to be an over-abundance of hardships or deaths, as all the frequencies of stress markers
are mostly congruent with those from the\is time period. This fact ultimately hints that transition
to Arab rule and the Islamitization at this location did not impact their well-being, but certainly
warrants further investigation.

49
 CHAPTER 3: Method Optimization of aDNA Extraction from Ancient Nubian
Archaeological Materials

INTRODUCTION

When ancient DNA sampling is performed using skeletal material of our past ancestors,
it is destructive of some tissue which cannot be used for future work. Therefore, increasing the
endogenous human DNA to start with is paramount when utilizing precious archaeological
materials. In addition to advancements in sampling (outlined in Chapter 1), other strategies that
were previously developed may also be used on samples with very low endogenous levels,
including those from arid regions of Africa. These strategies include additional treatments during
the sample processing step to boost the endogenous content prior to DNA extraction. A bleach
pre-treatment, first introduced by Kemp and Smith (2005), has been used to decontaminate
ancient samples and is standard procedure of those samples with high contamination (Barta, et
al. 2013, Korlević, et al. 2018, Korlević & Meyer 2019). This mild pre-treatment uses low
percentage sodium hypochlorite (NaClO or bleach) to decrease the amount of exogenous DNA
that has been absorbed by the tissue. This step has been shown to increase the amount of
target DNA by almost five-fold and decreases the amount of contaminating DNA. However, the
impact of bleach on library complexity, or the uniqueness of DNA reads, is debated (Korlević &
Meyer 2019, Boessenkool, et al. 2017). In addition to a bleach pretreatment, the use of a mild
enzymatic digest before the DNA extraction has also shown to be useful in removing
contamination and subsequently increase the endogenous content (Schroeder, et al. 2019).
Also known as a “double digest,” the sample is subjected to an enzymatic solution for a short
time (15 min to 1 hour), this solution is removed, and the sample is ready for an overnight, full
digestion followed by extraction the next day. Like the bleach pre-treatment, this method
removes exogenous DNA from the sample powder to increase the endogenous content prior to
DNA extraction (Damgaard, et al. 2015). Both of these pre-treatments can be used in
combination (i.e. Boessenkool, et al. 2017) or separately (e.g. Damgaard, et al. 2015)
depending on the extent of contamination or sample fragility.
Samples from Africa are particularly difficult to work with due to thermal degradation and
low DNA preservation, thus alternative methods must be considered to retrieve any DNA that

50
remains in the tissue. For these reasons, and the fact that only one protocol has been recently
published outlining successful DNA extraction using material from Sudan and none from the
Middle Nile region, for this study it was necessary to optimize a new protocol. In the present
work, multiple combinations of the steps described above were piloted to assess the extent of
DNA viability in Ancient Nubian samples from the Middle Nile region. Furthermore, it was
investigated which pretreatments were most helpful to retrieve aDNA, quantify differences
between material used (bone versus tooth dentine and if certain environments or burial contexts
impact the survival of DNA. These trials and the optimized procedure may be used for current
paleogenomic work with archaeological materials from Sudan and will better inform sampling
strategies in the future.

MATERIALS

Samples were collected from eight archaeological collections of the Middle Nile Basin
representing nearly 3,000 years of history in this region (Figure 3.1, Table 3.1). Field sites span
nearly 1000 kilometers (km) along the Nile River, from the 5th Cataract at Berber to past the 3rd
near Abri, Sudan. With well over 200 tooth or bone samples, 89 individuals represent five
cultural horizons in Nubian history, from the Napatan period of the Kushite Kingdom to Medieval
Nubian times at the end of the Christian Kingdoms (Table 3.1). The archaeological and
anthropological contexts of all sites and samples are briefly described here to present the
variables introduced from different burial treatments and environments. Samples listed in Table
3.1 were shipped from collaborators or collected by the author for aDNA extraction at the
Institute of Evolutionary Medicine, University of Zürich, Zürich, Switzerland. Any remaining
tissue after processing, including enamel, tissue powder, or bone fragments, was returned to
the collaborators.

51
Figure 3.1. Geographic locations of Nubian populations included in this study.

Left: Satellite map of modern Nile River Valley with labels of approximate regions in the ancient world. Kwieka is
located in Lower Nubia, near the between the 2nd and 3rd Cataract. Blue box zooms in on Upper Nubia. Right:
Expanded view of Upper Nubian sites near the 4th and 5th Cataracts. Maps obtained from Google (2019), edited by
author. Note: ez-Zuma is another version of El-Zuma, both are used in the literature.

52
Table 3.1. Archaeological information and sampling of Nubian populations included in this study.

Nubian Group Location/Site Date(s) N Sample Material Excavated by (year(s)):

Napatan, 800 – 350 BCE, NCAM, Tinga Archaeological Rescue Project

Enepis 10 27 bone and tooth samples
Meroitic 350 BCE – 350 CE (2017)

NCAM, Kwieka Archaeological Rescue Project

Meroitic Kwieka 350 BCE – 350 CE 4 8 bone and tooth samples
(2018)

Meroitic Berber 350 BCE – 350 CE 12 28 bone and tooth samples NCAM, Berber Meroitic Cemetery (2009 - 18)

450-550 CE (2nd phase Early Polish Centre for Mediterranean Archaeology,

Post-Meroitic El-Zuma 16 26 bone and tooth samples
Makurian period) Early Makuria Project (2015-17)

450-550 CE (2nd phase Early Polish Centre for Mediterranean Archaeology,

Post-Meroitic El-Detti 4 10 bone and tooth samples
Makurian period) Early Makuria Project (2015-17)

Polish Centre for Mediterranean Archaeology,

Post-Meroitic Tanqasi 350-600 CE 4 9 bone or tooth samples
Early Makuria Project (2015-17)

Early Polish Centre for Mediterranean Archaeology

Al-Ghazali 650-1000 CE 12 12 tooth samples
Christians (2016)

Medieval 40+ bone samples, 80+ tooth

El-Kurru 900-1400 CE 27 International Kurru Archaeological Project (2016)
Christians samples, post-cranial remains

N = Number of individuals
BCE = Before Common Era
CE = Current Era
NCAM = National Committee for Antiquities and Museums, Khartoum, Sudan

53
El-Kurru

Known for its royal pyramids, tombs, and mortuary temple dating to the Napatan time
period, El-Kurru was an important site in Upper Nubia since the Kingdoms of Kush. It is located
on the east bank of the Nile, near the 4th Cataract, and has been the focus of the International
Archaeological Kurru Project since 2013 (Figure 3.1, 3.2). Recent archaeological work has
concentrated on further exploration of the pyramids and burial tombs on the plateau as well as
exploration of the unfinished excavations by George Reisner (from the 1910s) closer to the Nile
(Emberling & Dann, et al. 2015). This work has uncovered a fortified Medieval settlement and
an associated cemetery located within the town walls. This cemetery contained at least 27
individuals and was excavated over three seasons (2014-2016) (Figure 3.2). Anthropological
analysis of these remains was conducted, including sex and age estimations in addition to
evaluations of stress markers and health status. All age groups were represented by at least
one individual, ranging from infant to older adult (50+ years) (mean of 26 years), and there was
a slight abundance of males from this group. Mortality distributions peak during first years of life
and at the end, demonstrating survival conditions typical of ancient populations. Material
remains, including ceramics and cookware, typical of Middle to Classic Christian periods were
found near burials and were associated with the surrounding domestic spaces. Radiocarbon
dating of organics from the wall date the fortification to 600-1000 CE; the Christian burials are
likely dated after this phase to around the arrival of Islam in Nubia, ca. 1450 CE (Skuldbøl, et al.
2016, Welsby 2016). Preservation of the remains was generally fair (likely due to proximity to
the Nile, the natural hydrology of the site, and modern irrigation strategies (Figure 2.4)), but still
provided many opportunities for tissue samples; specifically, the temporal bones seemed in
better condition, while tooth dentine was chalky and demineralized. A summary of individuals,
including excavation season, demographics, burial observations, and sampling materials for
ancient DNA extraction, are summarized in the table below. More than 200 teeth and temporal
bones were exported to Zürich for piloting methods for extracting DNA (Table 3.2).

54
Figure 3.2. Excavations centered on Medieval Christian settlement and cemetery.

Left: Map of settlement excavations, wall in orange (Emberling & Dann, et al. 2013). Blue box zooms in on cemetery,
wall entrance (east), and modern irrigation channel (west). Right: Expanded view of cemetery, graves outlined within
excavation squares. Courtesy of M. Uildriks.

Table 3.2. List of El-Kurru individuals, including demographic data, archaeological context, and
samples collected.
Age at Death
Ind No. Season Sexβ Body Position Burial Orientation Tooth Bone
(in years)α
004 2014 3-5 SA Extended, right side W-E, facing S 

005 2014 20-35 M Extended, supine W-E, facing up 

104 2015 20-35 M Extended, right side W-E, facing S  
105 2015 20-35 F Extended, left side W-E, facing N 
106 2015 36-50 I Extended, left side W-E, facing N 
107 2015 16 F Extended, right side W-E, facing S  
108 2015 50 + M Extended, right side W-E, facing S
109 2015 7-9 SA Extended, left side W-E, facing N  
110 2015 36-50 F Extended, left side W-E, facing N  
111 2015 36-50 F Extended, supine W-E, facing S 
113 2015 12.5-15.5 SA Extended, right side W-E, facing S  
115 2015 <1 SA Extended, right side W-E, facing S 

55
 201 2016 50 + M Extended, right side W-E, facing down  
202 2016 7.5 SA Extended, left side W-E, facing N  
203 2016 20-35 M Extended, supine W-E, facing S  
204 2016 50 + I Extended, right side W-E, facing S  
205 2016 50 + F Extended, left side W-E, facing down 
206 2016 20-35 F Extended, supine W-E, facing up  
207 2016 50 + M Extended, right side W-E, facing S  
209 2016 17.5 SA Extended, left side W-E, facing N  
210 2016 50 + M Extended, right side W-E, facing S  
211 2016 4.5 SA Extended, right side W-E, facing up  
212 2016 3.5 SA Extended, right side W-E, facing S  
213 2016 5.5 SA Extended, right side W-E, facing S  
214 2016 5.5-6.5 SA Extended, left side W-E, facing N  
215 2016 50 + M Extended, right side W-E, facing S  
216 2016 50 + M Extended, right side W-E, facing S  
α Cunningham 2016, Thoma & Goldman 1960, Smith 1991, Buikstra & Ubelaker 1994, AlQahtani 2008, Lovejoy, et al.
1985, Brooks & Suchey 1990
β Buikstra & Ubelaker 1994, Phenice 1969

+ = more than
< = less than
F = Female or probable female
M = Male or probable male
I = Indeterminate sex or could not be determined as male or female
SA = Subadult, no sex estimate
W-E = skull at west end, feet at east end; directions not cardinal, oriented toward the river
 = at least one available

Ghazali Monastery Complex

Ghazali is a medieval monastery complex dated to the 7th-13th century CE and is

located in the 4th Cataract region, northwest of the Bayuda desert (Figure 3.1, Figure 3.3). This
site has been excavated since 2012 by the Polish-Sudanese Ghazali Archaeological Site
Preservation Project (Obłuski, et al. 2015). In addition to the monastery, four cemeteries have
been excavated with variable mortuary architecture– for example, simple rock cut pits with stone
cairns or mud-brick mastabas with double vaulted chamber tombs (Obłuski 2014). Cemetery 2
contains the graves of monastic monks; here, all the interred are male, burials were located in
close vicinity to the church, and associated goods indicated the individuals’ religious status.
While cemeteries 1-3 are likely associated with the religious complex, Cemetery 4 is more
isolated and had an unknown function (Figure 3.3, Cemetery 4 not pictured). Here, 15 box
graves were excavated demonstrating typical Christian burial style, but atypical body treatments
(Welsby 2016). Due to the great preservation of the skeletal remains, samples were requested

56
for ancient DNA extraction. Currently, human remains are stored at the team field house in
Karima or at McMaster University (Ontario, Canada). Intact tooth samples from 12 individuals
were selected by the project anthropologist (Robert Stark) for ancient DNA analysis. At least
one individual from each of the four cemeteries was selected for comparison of genetic makeup,
but also to evaluate tissue preservation at this site for future analyses. Demographic data of
these 12 individuals were evaluated by Stark and are summarized in the table below (Table
3.3).

Figure 3.3. Map of Ghazali monastery complex, including surrounding cemeteries and
settlement area.

Obtained from Stark & Ciesielska (2019).

Table 3.3. List of Ghazali Individuals, including demographic data, archaeological context, and
samples collected.
Ind No. Site Cemetery Sampling No. Age-at-death (in years) α Sex β Tooth Sample
4-010 Ghazali 4 (none) ~ 40 M 3rd M (?)
4-008.2 Ghazali 4 (none) ~ 35 F RM2
3-002 Ghazali 3 (none) 50 + F RM1
3-009 Ghazali 3 (none) ~ 40 M RM1

57
 1-010 Ghazali 1 (none) ~ 45 M LM2
1-004.1 Ghazali 1 4.22.1 ~ 25 M RM2
2-009 Ghazali 2 3.03.1 ~ 45 M LM3
2-010 Ghazali 2 3.30.2 ~ 45 M RM1
2-011 Ghazali 2 3.12.1 50 + M RM3
2-004.2 Ghazali 2 4.23.1 ~ 50 M RM2
2-004.1 Ghazali 2 3.06.1 ~ 40 M RM3
2-013.2 Ghazali 2 3.14.1 ~ 40 M LM3
α Buikstraand Ubelaker 1994
βSchaeler 2009, AlQahtani, et al. 2010
~ = approximately
+ = more than
M = Male
F = Female
R or L = Right or Left, M = Molar, (superscript) = Maxillary (upper), (subscript) = Mandibular (lower), numbers
reference which molar (i.e. 1 = first, etc.)
(?) = unknown assignment, likely third molar based on morphology of crown and roots
Sampling No. = refers to internal numbering for the Ghazali excavation project

Kwieka Cemetery

The archaeological site of Kwieka is located on the east bank of the Nile, near the
modern town of Abri and Sai Island (Figure 3.1). More than 200 mound graves were identified in
the area between the highway and settlement areas in proximity to the Nile River. To date, 17
graves have been excavated over two rescue seasons in 2016 and 2018 by National
Corporation for Antiquities and Museums (NCAM, Khartoum, Sudan) officials. Mortuary
excavations suggest multiple phases of occupation, spanning from Late Meroitic to Christian
periods. Various types of graves and burial architecture were categorized by a suite of features,
including super- and subterranean structures, which were useful for contextual dating when
compared to other 4th Cataract burials (Abdallah 2018). Most burials showed evidence of
looting in antiquity, but still contained fine grave goods (decorated ceramics, beads, metal
objects), indicating more elite social status of these individuals (Abdallah 2018). Samples
representing four individuals were collected from storage at NCAM and transported to Zürich for
processing. Of the four collected, only one (T5) was contextually (i.e. from artifacts) dated to the
late Meroitic period (3rd - mid 4th century CE). For each individual, at least one sample was
collected (mostly two per individual, Table 3.4), including teeth and the petrous portion of the
temporal bone. A total of eight samples were exported to Zürich (Table 3.4).

Table 3.4. List of Kwieka Individuals, including archaeological context and samples collected.
Sample
Ind No. Site Burial Type (Abdallah 2018) Samples
Details
Type B - small mound, rectangular pit,
Tomb 1 KWC 2 teeth RPM1, M2
oval rock-cut burial chamber

58
 Type F - rounded rock superstructure, 2 bone, 1 L&R
Tomb 2 KWC
rectangular pit, oval burial chamber tooth Petrous, Molar
Type C - circular superstructure,
Tomb 5 KWC 1 bone L Petrous
trapezoid descendary, oval burial chamber
Type D - rounded super structure,
Tomb 6 KWC 2 bone L & R Petrous
rectangular shaft, rectangular burial chamber
R or L = Right or Left
PM = Premolar
M = Molar
(subscript) = Mandibular
No sub- or superscript = Unknown assignment
Numbers reference which molar (i.e. 1 = first, etc.)

Enepis Cemetery - Tinga Archaeological Rescue Project (TARP)

The Tinga Archaeological Rescue Project (named for the modern canal in the vicinity)
began in July 2017 by NCAM officials as a salvage project to excavate 34 graves (containing 32
individuals) discovered during the expansion of modern agriculture projects. The TARP
cemetery site is located in the Berber region, south of the modern town of Enepis and on the
east side of the Nile River (Figure 3.1). The site has a large cemetery of grouped tumuli. Tombs
were dug into the alluvium layer, sometimes down to the bedrock layer, with 6 types of
superstructures: 1) circular mound of large pebbles with stone slab substructure, 2) rectangular
grave with EW orientation with variable shapes, 3) oval-shaped grave with EW orientation with
variable skeleton positioning, 4) rounded grave pit with a diameter of nearly one meter, 5)
cylindrical (long and rounded) with EW orientation, 6) Meroitic shape extending EW with burial
niche to the west (Bushara et al 2017). Various body positions and orientations were observed:
flexed, contracted, crouched, extended – which are possibly indicative of various cultural
occupations at this site. Grave goods were discovered, with preservation of ochre, metal goods,
and leather items in some cases. Overall good preservation of the skeletal remains allowed for
preliminary analyses of demographics, paleopathology, and stature. Twenty-six samples from
ten individuals were sampled from those excavated in 2017 for genetic analysis, including two
from a double burial. Samples include temporal bones and/or single rooted permanent or
deciduous teeth (Table 3.5). Of those individuals collected, only G18 was dated to the Medieval
period.

Table 3.5. List of TARP Individuals, including demographic data, archaeological context, and
samples collected.
Grave Ind. Age Sex Grave Type &
Season Samples Sample Details
No. No. Estimation α β Body Position
T6
G06 2017 Young Adult F Rectangular, flexed 1 bone, 2 teeth L Petrous, RC1, RI2
D3
G07 C3 2017 Mature Adult M Circular, crouched 2 bone, 2 teeth L & R Petrous, 2 PM
G09 C4 2017 Child SA Rectangular, flexed 1 bone R Petrous

59
 G17 - 2017 Mature Adult F Rectangular, flexed 2 teeth LC1, RC1
G18 D4 2017 Infant SA Oval, extended 1 tooth LC1, RC1
G21 - 2017 Adolescent F Oval, crouched 2 bone, 1 tooth L & R Petrous, RC1
G22 T3 2017 Child SA Circular, crouched 2 bone L & R Petrous
G23 E3 2017 Child SA Circular, contracted 2 bone, 1 tooth L & R Petrous, I2
G27 T4 2017 Mature Adult F Oval, extended 3 teeth RI1, RI2, RC1
L & R Petrous,
G28 T5 2017 Child SA Circular, crouched 2 bone, 2 teeth
deciduous teeth
α Buikstra & Ubelaker 1994, Lovejoy, et al. 1985, Brooks & Suchey 1990, Walker, et al. 1991
β Buikstra & Ubelaker 1994
F = Female or probable female
M = Male or probable male
SA = Subadult, no sex estimation
Samples: R or L = Right or Left, I = Incisor, C = Canine, PM = Premolar, M = Molar, (superscript) = Maxillary (upper),
(subscript) = Mandibular (lower), numbers reference which tooth in series (i.e. 1 = first, etc.)

Berber Meroitic Cemetery

In 2009, a cemetery was discovered on the East bank of the Nile south of the 5th
Cataract while construction crews were digging foundations for the construction of a factory
(Figure 3.1). A rescue project was launched by NCAM to excavate burial structures, human
remains, and associated funerary objects at this archaeological site - The Berber Meroitic
Cemetery (BMC). Contextual finds date this cemetery to the Meroitic Kingdom (350 BC – 350
CE) and the necropolis contained more than 57 individuals of good preservation, despite the
disturbance from foundation trenches and ancient or modern looting (Figure 3.4). Burials yielded
large caches of wooden objects, decorated ceramics, and finer items associated with the
skeletons as evidence of good preservation. The burial formations include those typical of a
Meroitic tradition common in central Sudan (i.e. slope to burial niche, blocking wall) and likely
represent an elite community, especially with the fine grave goods (Bashir 2010, 2013).
Variation in burial practices, representative ceramics, and selective c14 dating of organic
material have constructed a finer chronology of the burials on site (Figure 3.4) (Bashir & David
2015).
Skeletal remains from the six seasons of excavation at BMC are currently housed at
NCAM in Khartoum. After assessing individuals with good preservation, samples from twelve
individuals were selected for ancient DNA analysis. Samples from these remains were
catalogued, photographed, and packaged for exportation in 2018 and transported to Zürich for
processing. For each individual, at least one sample was collected (mostly two per individual),
including teeth and the petrous portion of the temporal bone. At least one individual was
selected from each occupation phase of the site (Bashir & David 2015), totaling 22 samples
from twelve individuals. The table below summarizes these samples, including tissue type and

60
excavation data of the graves and is also color-coded to match the internal chronology of the
publication (Table 3.6, Figure 3.4) (Bashir & David 2015, Bashir 2010).

Figure 3.4. Topographical Map of Berber Meroitic Cemetery with dating (Bashir & David 2015).

Table 3.6. List of BMC individuals, including archaeological context, contextual dating, and
samples collected.
Tomb Ind. Demographic &/or
Season Sample Type Sample Details
No. No. Archaeological Notes
T34 - 2012 Extended, wooden bed 2 teeth LC1, LPM (Lower)
T26 SK1 2012 Semi flexed; double burial 1 bone, 1 tooth R Petrous, LPM (Lower)
T3 SKA 2009 Double burial; fine grave goods 1 bone, 1 tooth R Petrous, PM
SKB
T3 2009 Double burial 2 bone L & R Petrous
A-45
T33A - 2012 Flexed; fine grave goods 1 bone L & R Petrous
Double burial, semi-flexed,
T16 SKB 2012 1 bone L& R Petrous
male and female adults
Multiple burials (adult, two
T32 - 2012 1 bone L & R Petrous
children)
63-
T6 2009 Meroitic script on finds 3 teeth RPM1, lower Incisor, RC1
6(A)
T57 - 2016 Bone lesions 1 bone, 2 teeth R Petrous, PM2 (L?), C1 (R?)
T53 SKB 2016 Double burial 2 teeth 2xC
T53 SKA 2016 Double burial 3 teeth LI2, LC1, RPM1
T49 SKB 2016 (none) 2 teeth RC1, LC1
Green – before mid-1st century CE
Blue – second half of 1st – beginning of 2nd century CE
Red – 2nd – beginning of 3rd century CE
Yellow – 3rd century CE
Grey – no date available, recent excavation
SK - skeleton

61
Samples: R or L = Right or Left, I = Incisor, C = Canine, PM = Premolar, M = Molar, (superscript) = Maxillary (upper),
(subscript) = Mandibular (lower), numbers reference which tooth in series (i.e. 1 = first, etc.), (?) = questionable
assignment

El-Zuma

Excavations of the Early Makuria Research Project at El-Zuma are directed by Dr.
Mahmoud El-Tayeb of the Polish Centre of Mediterranean Archaeology (PCMA) (University of
Warsaw). Explorations of this site began in 2005 and excavation seasons take place every year
(El-Tayeb 2007). El-Zuma is located in the northern providence between the 3rd and 4th
Cataract on the right side of the bank (Figure 3.1). Finds date the site to the second phase of
the Makuria period (mid-4th to mid-6th century CE). More than 30 tumuli (some up to 50m in
diameter and 10m in height) are the subject of intense research and varying burial practices
suggest different social status of those interred, all from the Meroitic period. The tumuli have
been classified into three type by the archaeologists: Type 1 – large superstructure made of
sand and stone with rock cut tunnel and burial chambers with additional rooms, presence of
pillars; Type 2 – flat top superstructure, smaller in diameter than Type 1, shaft that leads to a
main burial chamber with additional rooms but no tunnels, presence of pillars, closed off with
stones or bricks, dating to Meroitic period; Type 3 – have the smallest superstructures with a
stone ring around the base, one rock-cut shaft with a bench and one burial chamber located on
the West, indicating Meroitic traditions (El-Tayeb & Czyżewska 2011, El-Tayeb, et al. 2016a).
Most of these burials were looted in antiquity, suggesting the large amount of wealth buried with
these individuals.
After remains were assessed for preservation, i.e. some were glued or bone tissue was
excessively moldy or chalky, 16 individuals were selected for sampling. For each individual, at
least one sample was collected (mostly two per individual), including intact teeth with no glue or
cracks, and temporal bones with complete petrous portions or areas which were able to still be
sampled, totaling 26 samples (Table 3.7). Summary table constructed from El-Tayeb, et al.
(2014) and personal communication with anthropologist Robert Mahler (PCMA).

Table 3.7. List of El-Zuma individuals, including demographic data, archaeological context, and
samples collected.
Age-at-death
Ind. No. Sex Burial Type Sample Type Sample Details
(in years)
T.8 35-45 F I 1 bone R Petrous
T.10 30-40 M III 2 bone L & R Petrous
T.11 16-18 SA II 2 bone L & R Petrous
T.14 20-30 F II 1 bone R Petrous
T.15 21-24 F II 1 bone, 2 teeth L Petrous, LPM2, LC1

62
 T.16 16-24 M II 1 bone R Petrous
T.17 30-40 M III 1 bone L Petrous
T.18 50+ F III 2 bone L & R Petrous
T.19 35-45 M III 1 bone R Petrous
T.20 50+ M III 2 bone L & R Petrous
T.22 35-45 M III 1 bone L Petrous
T.24 15-18 F II 1 bone R Petrous
T.25 24-35 F II 2 bones, 2 teeth L & R Petrous, RPM1, RPM2
T.26 45-55+ F II 1 bone L Petrous
T.27 35-45 M III 1 tooth LPM2, RPM2
T.28 40-55 F III 1 bone R Petrous
+ = more than
F = Female or probable female
M = Male or probable male
SA = Subadult, no sex estimation
Samples: R or L = Right or Left, I = Incisor, C = Canine, PM = Premolar, M = Molar, (superscript) = Maxillary (upper),
(subscript) = Mandibular (lower), numbers reference which tooth in series (i.e. 1 = first, etc.), (?) = questionable
assignment

El-Detti

The archaeological site of El-Detti is located on the right bank of the Nile River between
the 3rd and 4th cataracts (Figure 3.1). Excavation work has been on-going since 2014 and is
most notable for over 50 tumuli for comparison with other burial architecture in the region –
namely Tanqasi and El-Zuma (El-Tayeb, et al. 2016b). Tumuli from this site are categorized as
type II (multi-chambered) and type III (single-chambered), much like those found in El-Zuma.
Skeletal remains from the tumuli have been recovered by the Early Makuria Research Project.
This site is dated to the Meroitic Period, specifically Early Vistur period, phase II or mid-4th to
the end of 6th century CE (El-Tayeb, et al. 2016b). The remains from four individuals were
selected for sampling for ancient DNA analysis. For each, at least one sample was collected (up
to four for one individual), including teeth and/or the petrous portions of the temporal bone. A
total of 10 samples were brought to Zürich for aDNA extraction. Samples for these four
individuals are summarized in Table 3.8, including demographics (personal communication,
Magda Srienc).

Table 3.8. List of el-Detti individual, including demographic data and samples collected.
Age-at-death
Ind. No. Sex Sample Type Sample Details
(in years)
T.2 45-55 M 2 bone, 2 teeth L & R Petrous, LPM2, LI1
T.4 45-55 M 1 bone, 2 teeth L Petrous, RI1, LPM2
T.5 40-50 M 1 bone R Petrous
T.7 40-50 M 2 bone L & R Petrous
M = Male or probable male

63
Samples: R or L = Right or Left, I = Incisor, PM = Premolar, M = Molar, (superscript) = Maxillary (upper), (subscript) =
Mandibular (lower), numbers reference which tooth in series (i.e. 1 = first, etc.), (?) = questionable assignment

Tanqasi

The Tanqasi archaeological site is located in the vicinity of El-Zuma and El-Detti, near
the 4th Cataract (Figure 3.1). This site has been part of the Early Makuria Research Project
since 2005, directed by the PCMA. Skeletal remains have been recovered from inside tumuli
and likely represent a cultural group from the Post-Meroitic time period. Due to preservation
issues, only four individuals were selected for sampling. For each, at least one sample was
collected (up to five per individual), including teeth and the petrous portion of the temporal bone.
A total of nine samples were brought to Zürich, as summarized in Table 3.9, including
demographics (personal communication, Magda Srienc).

Table 3.9. List of Tanqasi individuals, including demographic data and samples collected.
Ind. No. Age-at-death Sex Sample Type Sample
(in years)
T.16 30-35 F bone R Petrous
T.23 25-35 M 2 bone, 3 teeth L & R Petrous, LI1, RPM1, LC1
T.46 35-45 M tooth LPM1
T.179 45-55 F tooth RI1, RPM2
F = Female or probable female
M = Male or probable male
Samples: R or L = Right or Left, I = Incisor, PM = Premolar, M = Molar, (superscript) = Maxillary (upper), (subscript) =
Mandibular (lower), numbers reference which tooth in series (i.e. 1 = first, etc.), (?) = questionable assignment

METHODS

Clean Lab facilities

All DNA extractions were performed at the Ancient DNA Laboratory at the Institute for
Evolutionary Medicine (IEM) at the University of Zürich, Switzerland (Figure 3.5). This facility is
dedicated solely to aDNA research and has four self-contained rooms with independent HEPA
air filtration systems where researchers adhere to strict contamination control protocols
including uni-directional workflows, regular decontamination of work surfaces with UV
irradiation, DNase treatment, and/or bleach solution. The ancient lab is located in a separate
building from the modern genomics labs to avoid cross-contamination of stable PCR products.
All researchers wear full Tyvek suits, filtered masks, goggles, shoe covers, sleeve covers, hair
nets, and double disposable gloves during lab work. All reagents, solutions, tools, and
consumables were decontaminated before use with UV radiation (i.e. at least 30 minutes in UV
crosslinker) or DNase treatment, whenever possible. Additionally, all tools, surfaces, hoods, and
some elements of PPE (i.e. gloves and sleeves) are bleached, UV-sterilized, treated with

64
DNase, and/or changed between individuals to prevent cross contamination (Appendix A). The
lab is equipped with separate air-filtered laminar fume hoods for dedicated sample extraction
and PCR preparation. For these experiments, no positive controls were used, as these could be
a possible source of contamination. Instead, negative controls were processed in parallel with
all samples, continuously monitored for contamination, and carried through final sequencing
steps.

Figure 3.5. Laboratory facilities used for ancient DNA work at the University of Zürich.

Left) View from entry room to Ancient Laboratory at the Institute of Evolutionary Medicine, University of Zürich,
Zürich, Switzerland. Right) Bone cellar accessible to IEM members; bone and tooth samples not stored in the clean
room are stored here, which is in the same building as the clean lab, but not modern PCR labs on campus to prevent
contamination of stable PCR products.

General sample preparation and clean lab protocols

Sampling materials from the archaeological collections were intact rooted teeth or
temporal bones – the petrous is known to contain more authentic DNA, even in hot climates - or
both, as specified in previous Materials section (Gamba, et al. 2014, Pinhasi, et al. 2015,
Gallego Llorente,, et al. 2015). Prior to processing tissue samples, some samples were cleaned
in a non-PCR wet lab at the IEM to remove external sand and dirt. Digital photographs were
taken of teeth or temporal bones before processing. All data required for sex or age estimations
or notable/pathological characteristics were recorded and photographed (e.g. dental eruption
patterns, the mastoid process, dental disease, presence of bone lesions). In the clean lab, tooth
and bone samples were UV-decontaminated for at least 10 minutes on each side before

65
processing with additional treatments. However, for tooth specimens with dental calculus build-
up, this calculus was sampled prior to this step (see below for steps followed) (Figure 3.6).

Figure 3.6. Examples of teeth with dental calculus build up requiring sampling.

A) Tanqasi T.179, right second maxillary premolar. B) Berber Meroitic Cemetery T6 63-6a, right first mandibular
premolar. C) Ghazali-3-002, right first maxillary molar.

Processing of teeth samples for Ancient DNA Extraction

Tooth calculus sampling technique follows Warinner “Dental Calculus, Sampling

Protocol v.4, 2016” (personal communication). Working in the sampling laminar hood, calculus
build up was removed/scraped off from outer enamel or dentine surfaces with a sterile dental
scaler, collected, and weighed in a UV-treated tube. Samples were stored in the clean lab at
4°C and protected from decontamination UV light, i.e. in carton sample boxes or covered with
aluminum foil. Sterile calculus samples were taken whenever possible for proteomic or
additional genetic analyses in the future (Figure 3.6A-C).
Sterile samples of dentine tissue were processed via drilling with a dental drill (Nakanishi
(Hoffman Estates, IL) Emax EVOlution electric micro grinder (Model No. EV410-120, Cat No.
8233)) as per standard lab protocols outlined in Damgaard, et al. (2015) or crushed with a
stainless-steel mortar and pestle as per Gondek, et al. (2018). Each sample was assessed for
the most optimal method for dentine tissue extraction – including friability, taphonomic damage,
rooted or unrooted tooth, and size of root or crown. Generally, samples that showed signs of
weathering or were quite friable or chalky were known from experience to be difficult to drill or
obtain a sterile sample. Such samples were chosen for crushing after extensive UV treatment,
ablation of outer surfaces along dentine tissue, including the cementum when necessary, and
mechanical removal of enamel, as per personal communication with researchers who
developed techniques using the steel mortar and pestle (Gondek, et al. 2018). Removal of these
layers is important to increase the endogenous content of the samples. This method has proved
beneficial for teeth that were especially friable, would not withstand drilling, or were broken
during handling within the clean lab. With this method, sterile tissue was isolated and crushed,
where drilling would have been impossible. Tissue samples of at least 200 mg were extracted

66
for processing; unused powder was stored at 4°C. For subadults with small roots, dentine
powder was pooled from multiple teeth to reach 100mg. Unused enamel was stored at room
temperature in the clean lab for future analyses or returned to collaborators.
Dentine powder can be collected from within the crown using a dental drill. First, using a
dental drill and a sterile spherical cutter bit drill bit, a generous layer of the outer surface was
removed and discarded (Figure 3.7B). This outer layer typically consisted of remaining calculus,
residual dirt, and degraded cementum. Following this cleaning step, the root tissue then looked
polished (Figure 3.7C). Dust was wiped away using a damp lab tissue and the sample was UV
treated for 10 min per side. Next, the vise and jaws were lined with wax paper and the tooth was
firmly loaded into the vise (crown clasped horizontally within the jaws, roots pointing to the side)
within the sample hood (Figure 3.8A, B). The roots were removed from the crown with a coping
saw equipped with a sterilized fine-tooth blade (Figure 3.8C-E). Roots were retained for
crushing, as detailed below. For each tooth, crown dentine was drilled with a new spherical
cutter drill bit and collected into a tube using a sterilized glass funnel (6cm diameter) placed
below the crown (Figure 3.8F). Powder was weighed and stored at 4°C until extraction.

Figure 3.7. Processing of tooth samples in clean lab.

Cleaning of external surface, before drilling of crown dentine or crushing root dentine, to decrease contamination
during extraction step. Ghazali individual 4, GHZ-3-009, right first maxillary molar. A) Tooth sample before calculus
sampling or cleaning in the clean lab. B) Removal of outer layer using electric dental drill with a spherical cutter drill
bit. C) Tooth sample after cleaning showing a polished surface, inferior to the cementum-enamel junction (CEJ).

67
Figure 3.8. Set up for processing teeth within sampling hood in clean lab.

A) Cleaned tooth is loaded into a vise. B) Roots are positioned as such to leave crown in the vise to later drill the
dentine from beneath the enamel; sterile padding is used to keep the crown from cracking. C) Roots are cut from the
crown using a coping saw with new blade per sample. D) Cut is using made just inferior to the cementum-enamel
junction (CEJ). E) Root dentine is collected for crushing later; crown is left in the vise for drilling. E) Crown dentine
ready for drilling.

Following collection of crown dentine, the roots were wiped of dust then UV treated
again for 10 minutes per side before crushing in the stainless-steel mortar and pestle as per
Gondek, et al. (2018). Dentine tissue was loaded in the mortar with sleeve and crushed using
the pestle and a rubber mallet (Figure 3.9A). Several rounds of crushing (Figure 3.9B-E) were
necessary depending on how mineralized the tissue was and degree of preservation. Between
rounds, a sterilized spatula was used to move larger chunks around or scape powder back into
the well of the mortar. Powder should be a consistency finer than cornmeal (Figure 3.9E).
Powder was collected into a tube, weighed, and stored at 4°C until extraction.

68
Figure 3.9. Crushing of root dentine using steel mortar and pestle.

A) Set up of stainless-steel mortar and pestle with sleeve (inside dish to contain loose powder). B) Root dentine after
first round of crushing using mallet. C) Root dentine after second round of crushing. D) Root dentine after third round
of crushing. E) Root dentine after fourth round of crushing. F) Weighing of final amount of powder produced.

Processing temporal bone samples for Ancient DNA Extraction

Samples of petrosal bone tissue were obtained by opening the pyramidal/petrous portion
of the temporal bone to access sterile tissue for drilling with an electric dental drill (Nakanishi
(Hoffman Estates, IL) Emax EVOlution electric micro grinder (Model No. EV410-120, Cat No.
8233)) as per published protocols (Gamba, et al. 2014, Pinhasi, et al. 2015) or the petrous
portion was removed, cleaned, and crushed with a stainless-steel mortar and pestle as per
Gondek, et al. (2018) (Figure 3.10A-D). Samples were individually assessed for the most
optimal method for bone tissue processing – including degree of friability, taphonomic damage,
and size or morphology of petrous portion – to isolate the inner ear canals and surrounding
dense bone tissue for sterile sampling (Figure 3.11). Generally, samples with heavy weathering
or those that were friable or chalky were known from experience to be difficult to drill (i.e. tissue
does not stand up well to the mechanical force) to obtain a sterile sample (Hansen, et al. 2017).
Such samples were chosen for crushing after extensive UV treatment, manual ablation of outer
surfaces, and mechanical removal of less dense tissue surrounding the densest portion of the
pyramidal part. In general, outer tissue is heavily contaminated by environmental and modern
DNA. Removal of these layers is important to increase the endogenous content of the samples.
This method has proved beneficial for petrous bones with weathering (Gondek, et al. 2018).
Samples that were selected for drilling were processed according to published protocols

69
(Gamba, et al. 2014, Pinhasi, et al. 2015, Hansen, et al. 2017, Gallego Llorente, et al. 2015).
Tissue samples of at least 200 mg were extracted for processing; powder is stored at 4°C until
extraction step.

Figure 3.10. Anatomical parts of temporal bone.

Example of temporal bone from current archaeological collection, right temporal bone of TARP 2017 G9. A) External
view of temporal bone, showing external auditory meatus, zygomatic process, mastoid process, and squamous
portion. B) Internal view of temporal bone, showing squamous and petrous or pyramidal portion. C) Inferior view of
same bone. D) Anatomical location of temporal bone (in green), showing lateral view and superior view of skull.

The procedure to drill petrous portions of temporal bone was adapted from various
published protocols utilizing especially degraded samples (i.e. hot, dry) and the researcher’s
experience with skeletal remains (Pinhasi, et al. 2015, Gamba, et al. 2014, Hansen, et al. 2017,
Gallego Llorente, et al. 2015). Following cleaning steps and UV decontamination, a diamond
dust-encrusted circular blade attached to a dental drill was used to remove the petrous portion
of the temporal bone from the squamous portion to expose the inner ear labyrinth (Figure
3.10A-C, Figure 3.11A-C). Another layer from the freshly cut surface and surrounding area was
then removed with a new spherical cutter bit. The sample was wiped with a lab tissue and UV
irradiated for 10 minutes per side. Using a new spherical cutter bit, the dense bone of the
cochlea and surrounding tissue was drilled and collected using a 6cm funnel inside a 2mL tube.

70
Care was taken to not overheat the tissue by using short scraping movements with the drill bit
and using the lowest speed possible with the dental drill. Powder was weighed and stored at
4°C until extraction. Remaining portions of the petrous were wrapped in sterile aluminum foil
and stored in the clean lab for additional sampling, when necessary.

Figure 3.11. Visual of how the petrous portion of temporal bone is sectioned for processing.

A) Coronal (anterior/posterior) cross section of petrous portion to reveal cochlea and dense bone surrounding the
labyrinth; orange dotted line shows where a cut is made to separate the petrous portion from the squamous portion;
shaded area (yellow) is targeted for drilling to collect the most sterile sample for extraction; yellow arrow shows the
orientation of the drill bit during drilling. Specimen from Anatomy Collection, University of Zürich, Zürich, Switzerland.
B) Example of right petrous portion TARP 2017 G28 T5; orange dotted line shows approximately where the cut is
made, lateral to the flat ridge, to expose the target area for drilling.

To prepare a temporal sample for crushing, the petrous portion was removed from the
squamous portion of the temporal bone as specified above (Figure 3.11). Next, the petrous
portion was manually trimmed (including trabecular and/or cortical bone) to be small enough to
fit into the steel sleeve (Figure 3.12A). Using an electric dental drill with a sterile spherical or
conical cutter drill bit (New Technology Instruments, Kahla, DE), the outer surface of the sample
was removed and discarded. Bone dust was wiped away with damp lab tissue (KimWipe with
purified Milli-Q Water). The sample was UV treated again for 10 min each side before being
crushed. As per Gondek,, et al. (2018) and personal communications, samples were loaded into
the mortar with the sleeve in place and crushed lightly with one or two blows with a mallet and
pestle to separate the less dense tissue from denser portions being targeted. Less dense bone
was flakey, darker in color, while denser pieces were irregular and chunky after crushing (Figure
3.12B). These dense chunks were manually separated with sterile tweezers for another round of
crushing, while the less dense pieces were removed from the mortar (and stored/labeled
separately.) Shattered bone was swept away first using a dry lab tissue, taking care to
especially remove powder from the mortar, pestle, and sleeve. The dense pieces were reloaded

71
into the mortar with sleeve for mallet crushing. Between uses of the mallet, a sterile metal
spatula was occasionally used to move around uncrushed or overly crushed parts within the
mortar. The goal was to create an evenly mixed powder with both small and large chunks of
tissue, rather than perfectly homogenous (Figure 3.12C). Sample powder was collected by
placing a 50mL Falcon tube over the mortar well and carefully flipping the powder into the tube.
Lastly, powder was weighed and stored in the clean room until extraction.

Figure 3.12. Preparation of bone samples from petrous portion using stainless steel mortar and
pestle with sleeve.

A) Set up with sleeve loaded in the mortar (left), pestle at the side. B) Petrous portion after first few strikes to
separate less dense bone from more dense bone of the inner ear channels and surrounding tissue. C) Final result is
a fine powder, which will be weighed and divided into 100mg or 200mg aliquots for DNA extraction.

Extraction of ancient DNA from prepared dentine or bone powder

To release DNA from the powder generated during the processing steps, bone or
dentine samples were subjected to three different extraction methods to optimize and boost
endogenous DNA content: 1) single digest method, as per Schuenemann and Peltzer, et al.
(2017) with modification from Dabney at el. (2013), 2) double digest method, as per Damgaard,
et al. (2015) and “Orlando Lab June 2015 protocol, with Viking lab modifications,” (used in
Ancient DNA Lab in Blindern, Oslo) and 3) double digest with bleach pretreatment, as per Kemp
& Smith (2005) and Bossenkool, et al. (2017) (Figure 3.13). Additionally, the amount of powder
used for these extraction methods varies between 50mg and 200mg. Cryo-milling, where whole
chunks of teeth or bone are powdered while being chilled by liquid nitrogen (Figure 3.13). This
technique is no longer used because other methods (i.e. drilling) are more sterile.

72
Figure 3.13. Flowchart for all trialed extraction methodologies.

Methodologies are depicted in this flowchart; a total of ten different methods were trialed. Solid grey lines show
methods piloted; dotted lines have not been trialed. For more details, see Chapter 3. (Image created with
BioRender.com by author)

For the single digest method, for each individual, approximately 50-100mg of powder
(dentine or bone tissue) was digested with a lysis solution of 0.45M EDTA and 0.25mg/mL
proteinase K overnight with rotation at 37°C (Figure 3.13). Following centrifugation, DNA was
isolated from the supernatant using binding buffer and a QIAGEN (Hilden, Germany) silica
column with centrifugation (as per Gamba, et al. 2015 or Dabney & Meyer 2019), followed by
purification steps with in-house modifications (binding centrifugation speed 1500 rpm, washing
14,000 rpm, and eluted in 50uL of TET buffer twice). Final volume for DNA was 100uL in TET
for library preparation.
For the double digest method (Figure 3.13), also referred to as mild digestion
pretreatment, approximately 200mg of powder (dentine or bone tissue) for each individual was
digested in duplicate (pooled later) with a lysis solution of 0.45M EDTA, 0.25mg/mL proteinase
K, and 0.5% N-lauryl-Sarcosyl for 1 hour at 37°C with nutation. Following centrifugation, the first
digest supernatant was removed, stored in a labeled tube, and pellet was resuspended in fresh

73
lysis solution for overnight incubation at 37°C with nutation. Following centrifugation, DNA was
isolated from supernatant using a binding buffer and a QIAGEN silica column with a Manifold
vacuum system, followed by purification steps using MinElute PCR Purification reagents and
protocols (QIAGEN, Hilden, Germany). DNA was eluted in 65uL of elution buffer for library
preparation.
Some samples were selected for bleach pretreatment, namely those with visible mold
and heavy weathering. For these samples, approximately 200mg of powder (dentine or bone
tissue) was incubated with 0.5% bleach solution for up to 15min at RT with nutation (Figure
3.13). Following centrifugation, the pellet was washed three times with purified Milli-Q water.
Then, the protocol proceeds with a double digestion, as outlined in the section above.

Library Preparation

From this step forward, all samples were treated the same. In the clean lab, double-
stranded Illumina libraries were prepared with 20μL of sample extract (or extraction blanks) as
outlined in Meyer and Kircher (2010) (Figure 3.14). Briefly, libraries of sample DNA were blunt-
end repaired, adaptors ligated to the fragments, and gaps filled in with silica-column
purifications (QIAGEN) between each step to remove interfering molecules (i.e. hairpin
fragments with identical adaptors or dimers) (Meyer & Kircher, 2010). Before indexing, the
quantity of molecules was measured by qPCR on a LightCycler (Roche Life Science, Basel,
Switzerland) for quality control and to monitor the process. Next, libraries were dual indexed (i.e.
indexing or ‘barcoding’ both adaptors) as described in Kircher, et al. (2012) to decrease the
chances of downstream misidentification. The quantity was measured again via qPCR and
libraries were reamplified with Herculase II Fusion Polymerase (Agilent, Santa Clara, CA, USA)
according to the qPCR measurement to reach a copy number of 1013; concentrations and
fragment length of libraries were quantified on an Agilent 2200 TapeStation (Agilent, Santa
Clara, CA, USA). Libraries were then ready for shotgun sequencing for screening or a targeted
enrichment step, followed by sequencing (Figure 3.14).

74
Figure 3.14. Library preparation of ancient DNA extractions and subsequent sequencing.

Genetic material from original extraction is depicted with various colors: green – human mitochondrial DNA, blue –
human nuclear DNA, yellow – modern contaminating DNA, orange – environmental DNA. Purification steps are
depicted as the tube with silica column. Library material is sequenced for shotgun sequencing or is enriched for
mitochondrial DNA with in-solution capture step with magnetic beads, then sequenced. (Image created with
BioRender.com by author)

Shotgun Sequencing Screening and Processing

Pooled DNA libraries were screened for human DNA through shotgun sequencing on the
Illumina HiSeq 4000 platform (San Diego, CA, USA), with 2x150bp read lengths, at the
Functional Genomics Center Zürich (UZH) or the Genomics Core Facilities at the Oslo
University Hospital. Sequencing data were analyzed with the pipeline EAGER v1.92 (Efficient
Ancient Genome Reconstruction) as it is tailored for ancient DNA obtained via NGS (Peltzer, et
al. 2106). This tool uses FastQ files generated by the Illumina platform Reporter that are
demultiplexed and sent as raw files for analysis. Raw reads were analyzed by EAGER to map to
the human reference sequence and the mtDNA revised Cambridge Reference Sequence
(rCRS; NC012920), as well as assess endogenous content, average coverage, damage, and
average fragment length (Fu, et al. 2013). Data authenticity was established by examining the
molecular damage patterns and characteristic fragment distribution using mapDamage2.0

75
(Ginolhac, et al. 2011, Jόnsson, et al. 2013). Samples not demonstrating typical molecular
behaviors were excluded from the following enrichment step. Those libraries with authentically
ancient human reads (nuclear or mtDNA) were selected for an mtDNA capture step for
enrichment of human mitochondrial DNA. For comparison among processing methodologies,
libraries were also analyzed with MALT (MEGAN Alignment Tool) to categorize the
contaminating sequencing DNA not mapping to the human genome (Herbig, et al. 2016).

mtDNA Capture and Sequencing

Those libraries selected in the previous step were enriched for human mtDNA using
bead capture hybridization as outlined in Maricic, et al. 2010 (Figure 3.14). Subsequent steps
for enriched library amplification follow Schuenemann & Peltzer, et al. 2017, with sequencing
performed on an Illumina MiSeq or NexSeq (Mid-output) (San Diego, CA, USA) at the
Functional Genomics Center Zürich (UZH). Results relating to the enrichment step are detailed
in Chapter 4.

mtDNA Processing, Alignment, and Analysis

For read processing, EAGER again was used on sequenced libraries (Peltzer, et al.
2106). As outlined in Schuenemann & Peltzer, et al. (2017), options like CircleMapper, DeDup,
and other included features of the EAGER pipeline were utilized as they are optimized for NGS
datasets from mitochondrial aDNA. To measure contamination from modern human mtDNA,
Variant Called Files (VCF) were analyzed using the statistical program Schmutzi (Renaud, et al.
2015) with strict criteria settings to identify contamination and to estimate the percentage of
endogenous DNA for each sample and its quality. Samples with more than 5% contamination
were excluded from further analysis. For those under this threshold, consensus sequence files
for each library were generated by this tool (Figure 3.15) and libraries were then merged if they
were from the same individual (e.g. KWC_2 single digest library + KWC_2 double digest
library). Samples with a mapping quality score of over 30, high coverage of the mitochondrial
genome (roughly 10X or more), and low contamination were retained for haplogroup analysis.
Negative controls were processed through this step and analyzed to identify DNA introduced
during lab work. More detailed analysis of the results from the enrichment part of this research
design are detailed in Chapter 4.

76
Figure 3.15. Sequencing and post-sequencing analysis.

Following amplification, libraries are demultiplexed, removing adaptors and barcodes, using various bioinformatic
pipelines for the assembly of a consensus sequence for individuals (using either mitochondrial (green) or nuclear
DNA (blue) (not performed)). (Image created with BioRender.com by author)

RESULTS

Results presented here are comparisons with shotgun sequencing data, not results from
enrichment (see Chapter 4). Read counts refer to mapped reads to the human genome, not
necessarily the human mitochondrial genome. Shotgun sequencing is a screening step
performed to assess the viability of ancient genetic material and can be used for mitochondrial
or autosomal screening. Mitochondrial reads are expected to be higher given the high quantity
of mitochondria in each cell (few thousand mitochondrial genomes compared to one nuclear
genome per cell (Giles, et al. 1980)). Three metrics were compiled to measure the effectiveness
of the methodologies trialed: number of ancient mapped reads (human genome as reference),
percent of endogenous content (the amount of aDNA in a sample), and library complexity
measured by “cluster factor” (uniqueness of reads).
In total, 43 individuals were trialed, most with two or more extracts processed in three
different ways including bleach pretreatment with double digestion, double digestion, or single
digestion (Figure 3.13). Given the low endogenous content typical of Nubian samples recovered

77
from harsh environments, the goal of these pretreatments (i.e. bleach or double digestion or
both) was to increase the endogenous content prior to extraction to obtain more ancient reads
while not compromising library complexity. Seventy-five NGS libraries from 43 individuals were
built using these extraction methods. Twenty-five libraries (33% success rate), from 17 unique
individuals were successful in obtaining ancient human DNA, a 40% success rate.
To compare methodologies, pairs of libraries – where a single library pair consisted of
two libraries constructed from different extracts from the same individual – were compared with
the metrics outlined above. Twenty-seven bone or tooth powder samples were trialed in Zürich
with a double digest and were compared to 27 parallel extracts using the single digestion
method. These trials included samples from all eight archaeological contexts (Table 3.1). Eight
libraries from the double digest method were successful in obtaining ancient reads and eight
were successful for the single digest method. From these successful trials, seven library pairs
for both methods were compared for analysis.
Mild bleach treatment can increase an extract’s endogenous content while also
removing contaminating genetic material introduced as the result of taphonomic factors
(Damgaard, et al. 2015). Therefore, in addition to comparing samples processed with a double
digest to the single digest, samples prepared with a mild bleach treatment followed by a double
digest were compared to samples processed with only a double digest. For this analysis, four
bone samples were trialed at the Ancient DNA Laboratory (Centre for Ecological and
Evolutionary Synthesis) in the Department of Biosciences at the University of Oslo. Samples of
different preservation were selected – two graded as good (TARP 9, KWC 6) and two graded as
poor (ZUM 18, KU 210) – to test if macroscopic tissue integrity impacted our ability to obtain
aDNA. Two aliquots were extracted at the same time during tissue processing; one aliquot first
was treated with bleach, then both were treated with a double digest, totaling eight trials. Four
extracts were successful in obtaining aDNA, while four were not. Of the successful trials, only
one library pair was generated for method comparison (KWC 6). Furthermore, samples
extracted at Oslo were shotgun sequenced twice, at two facilities (i.e. Oslo and Zürich),
resulting in a duplicate library pair from KWC 6 for analysis. For the unsuccessful library pairs,
one of the two extracts failed to produce aDNA, and were thus removed from downstream
analysis.

Ancient Mapped Reads

Counts of mapped ancient reads showed an increase with both pretreatments compared
to the single digest method (Figure 3.16A, B). First, the double digest method increased the

78
number of ancient reads by 278% on average when compared to the single digest method. A
paired t-test was performed to determine significance of the reads increase. For all trials, the
double digest method increased the number of ancient reads retrieved over the single digest
method, but this increase was not significant (t = 1.3014, df = 12, p-value = 0.2176).
For the samples extracted at Oslo, aDNA was obtained for three of the four samples:
TARP 9B, KWC 6B, KU 210B (Table 3.10, Figure 3.16B). For KWC 6, we obtained aDNA for
both aliquots, with and without bleach. However, the bleach treatment resulted in a 56%
increase in reads. For TARP 9B and KU 210B, the double digestion method retrieved no ancient
reads while the bleach pretreatment facilitated aDNA extraction. These samples had 2024 and
1442 reads mapping to the human genome, respectively. We failed to obtain aDNA from ZUM
18. For this sample, 8,181 reads were reported, however damage pattern analyses indicated
the reads were not ancient. Overall, the mild bleach pretreatment increased the aDNA yield of
the ‘good’ tissue samples. However, it is unclear if bleach pretreatment would aid in aDNA
recovery for ‘poor’ samples as it was successful for one poor sample, but unsuccessful the
other.

Figure 3.16. Mapped ancient reads listed by method.

A) Single vs. double digest comparison B) Double digest vs. bleach pretreatment. Counts of ancient mapped reads
are listed above columns for each library. TARP G9 and KU 210 did not have ancient human mapped reads for the
double digest method and were listed as zero without bars. Ancient mapped reads were not significantly different for
either method comparison based on student’s t-test. Oslo and Zurich below the sample IDs in (B) designate the lab
where the sequencing was performed. BMC = Berber Meroitic Cemetery, GHZ = Ghazali, KWC = Kwieka Cemetery,
TARP = Tinga Archaeological Rescue Project.

79
Table 3.10. Four samples selected for bleach pretreatment and results.
Sample ID & Read Count Read Count
Sample Picture Bone Type Preservation Notes No Bleach With Bleach Conclusion

El-Kurru 2016 Poor preservation;

78 reads 1442 reads
Ind. 210 covered in silty dirt,
(KU 210) moldy throughout Ancient DNA
NO
cortical bone, obtained
ANCIENT ANCIENT
Right Temporal extensive
DNA DNA
bone weathering

Kwieka 2016
3115 reads 4845 reads Ancient DNA
Tomb 6
Good preservation, obtained
(KWC 6)
subadult, good bone
integrity ANCIENT ANCIENT 55.5%
Left Temporal
DNA DNA increase ↑
bone

TARP 2017
456 reads 2024 reads
Area G, Tomb 9 Good preservation,
(TARP 9) high bone integrity, Ancient DNA
NO
hard compact bone, obtained
ANCIENT ANCIENT
Right Temporal subadult
DNA DNA
bone

El-Zuma 106 reads 8181 reads

Moldy, chalky, soft
Tomb 18
bone tissue; much of
(ZUM 18)
external surface Failed
NO NO
removed for
Right Temporal ANCIENT ANCIENT
sampling
bone DNA DNA

Read counts include all mapped reads, both ancient and those determined to be from modern contamination.

To examine the impact of bleach on these tissues, bacterial profiles for each extract
were constructed and examined with MALT (Herbig, et al. 2016). This tool classifies and counts
bacterial reads from shotgun sequence data. Data from the bleach trials were selected to
organize the bacterial reads by phylum (Figure 3.17). The most prominent phyla across the
extracts were: Actinobacteria phylum (approx.10-95%) (highly diverse terrestrial and aquatic
bacteria), Proteobacteria (approx. 4-82%) (nitrogen fixing bacteria in soil, some are pathogenic),
Firmicutes (approx. 2-72%) (abundant cell wall bacteria or fermenters, some are pathogenic),
and Nitrospirae (up to 4%) (common aquatic or nonaquatic bacteria involved in nitrogen cycle)
(Figure 3.17).

80
Figure 3.17. Impact of bleach on contaminating bacteria in extracts.

Graphs generated by MALT, legend of bacterial phyla detected at far right. Both graphs organized per individual,
listed as without bleach treatment then with bleach treatment where B following the sample ID indicates bleach,
followed by negative controls from the Oslo Ancient Lab. A) Bacterial profiles, scaled to 100 percent (y-axis – number
of reads %). B) Reads of bacterial DNA from each trail sample (y-axis – number of reads).

Bacterial read counts increased with bleach treatment for three of the four trials, (KWC 6
decreased), but were not statistically different (t=1.18, df=3, p-value=0.32). This suggested that
the bleach treatment does not have a decontaminating effect. However, upon visual inspection,
the bacterial profiles did change for three of the four individuals. Again, KWC 6 differed from the
pattern of the other four samples, mostly remaining the same. Following the bleach treatment,
KU 210 showed an increase in Proteobacteria and a decrease in Actinobacteria. The TARP 9
profile was dominated by Actinobacteria, not Firmicutes, and the ZUM 18 sample lost all of the
Actinobacteria signal. The negative controls (i.e. water) contained Proteobacteria only. This
indicated that these Proteobacteria were likely contaminating bacteria introduced during the
laboratory processing of the extracts. Previously, these bacteria phyla have been tagged as lab
contaminants found in the reagents used during library construction that should be ignored
(Salter, et al. 2014). Proteobacteria also dominated the ZUM 18 sample that was treated with
bleach, likely showing a two-step process where pretreatment removed most of the bacteria,
then lab processes introduced contaminant bacterial reads since the profile is identical to
negative controls. This comparison lends more support that this phylum of bacteria was
introduced by lab processes.

81
Endogenous Content

Endogenous content showed the same positive results as mapped reads where its
proportion increased with pretreatments, either bleach treatment or double digest (Figure 3.18A,
B). First, when comparing single to double digest methods, the endogenous content increased
an average of 89%. Only GHZ 5 showed a 28% decrease with a double digestion. If we remove
GHZ 5 from the comparison, endogenous content increased 109%. When the values for all
library pairs were compared via a paired t-test, the difference between the endogenous content
obtained from the single digest method versus the double digest was not significant (t = 2.01, df
= 6, p-value = 0.091). Bleach treatment increased the endogenous content for all library pairs
(percent increase 117% to over 800%). However, the difference was not statistically significant
(t = 2.6885, df = 4, p-value = 0.05).

Figure 3.18. Endogenous Content (%) comparison by method.

A) Single vs. double digest comparison B) Double digest vs. bleach pretreatment. Percentages are listed above
columns for each library. TARP G9 had zero endogenous content for the double digest method and were listed as
zero without bars y-axis is Endogenous content (%). Ancient mapped reads were not significantly different for either
method comparison based on student’s t-test. Oslo and Zurich below the sample IDs in (B) designate the lab where
the sequencing was performed. BMC - Berber Meroitic Cemetery, GHZ - Ghazali, KWC - Kwieka Cemetery, TARP -
Tinga Archaeological Rescue Project.

Library Complexity

Library complexity is an important metric that judges the impact of pretreatments on the
uniqueness of reads which are built into libraries. Overall, libraries with low complexity are less
informative for genetic projects as they provide less distinct reads and more duplicates.
Therefore, to observe if pretreatments decrease the number of unique reads, we compared
library complexity across extraction methodologies. The EAGER pipeline produces a metric
called a “cluster factor” as a measure of complexity; it is defined as the number of duplicates
divided by the number of mapped reads. Duplicates are removed during bioinformatic

82
processing, are not helpful for constructing genomes, and do not contribute to higher coverage.
Thus, a higher cluster factor indicates more duplicates and therefore less complexity. A cluster
factor of 1 is an ideal complexity. Cluster results of all pretreatment trials (bleach and/or double
digest) were relatively the same when compared to the single digest method (Figure 3.19A, B).
When these values were compared via a paired t-test, the results were not significant,
supporting the idea that digestion method did not impact library complexity (t = 2.306, df = 8, p-
value = 0.070). This analysis established that pretreatments do not impact read uniqueness and
therefore do not lead to poor quality libraries.

Figure 3.19. Cluster Factor value comparison by method.

A) Single vs. double digest comparison B) Double digest vs. bleach pretreatment. Cluster factors are listed above
columns for each library. BMC = Berber Meroitic Cemetery, GHZ = Ghazali, KWC = Kwieka Cemetery, TARP = Tinga
Archaeological Rescue Project.

Tissue Comparison

Two tissue types were collected for sampling which allowed us to compare the
performance of tooth dentine and petrous bone tissue when obtaining aDNA. For this analysis,
processing was not taken into account, therefore tooth tissue was drilled or crushed dentine and
bone tissue was drilled or crushed petrous portions of the temporal bone. Extraction method
also was not considered as a cofactor. Thus, a trial was successful if aDNA was obtained and
not successful if no aDNA was obtained regardless of extraction methodology. The total number
of libraries was 75. Nine out of 49 libraries were successful using bone tissue, while 16 out of 26
trials libraries were successful using teeth (Figure 3.20). Extractions from teeth yielded
significantly more successful extractions compared to bone extractions from the petrous portion
(Fisher’s Exact test, p-value = 0.0003).

83
Figure 3.20. Comparison of tissue type to successfully obtain aDNA.

Trial outcomes are listed in the 2x2 contingency table below columns. The Fisher’s Exact test statistic value is
0.0003. The result is significant at p < 0.05. X-axis is tissue type. Y-axis is number of aDNA libraries obtained.

Burial Context Comparison

Varying archaeological contexts were represented in the eight populations included in

the full collection of skeletal samples (listed in Table 3.1). The burial conditions in which the
skeletons were interred were compared to understand if grave type had an impact on the ability
to retrieve aDNA from the sample material. Four material contexts were designated: box grave
(skeleton encased within stone-slab structure, e.g. Ghazali individuals), niche burial (skeleton
placed within an undercut or underneath a blocking wall, e.g. BMC individuals), pit grave
(skeleton placed at the bottom of a shaft, e.g. El-Kurru individuals), and tumulus (large, multi-
chambered, open tomb, e.g. El-Zuma individuals). Tissue type and extraction method were not
designated for this analysis. Successful trials produced libraries with aDNA and unsuccessful
trials did not; again, 75 trials were counted. Successful trials included 10 of 12 box graves
burials, 10 of 19 niche burials, 3 of 21 pit inhumations, and 2 of 23 tumuli burials (Figure 3.21).
A Chi-squared test for independence was applied to these data and results showed burial
context was significantly associated successful aDNA recovery (χ2 = 26.40, p-value = 0.00001,
significant at the 0.05 level). Samples from tumuli or pit graves did not perform well, while those

84
from niche burials and box graves did better to retrieve aDNA. These results are illustrated in
Figure 3.21.

Figure 3.21. Comparison of burial context to successfully obtain aDNA.

Trial outcomes are listed in the 2x4 contingency table below columns of four burial types. The Chi-squared statistic is
significant at the 0.05 level. X-axis is grave type. Y-axis is number of aDNA libraries obtained.

DISCUSSION
Paleogenomic research in Africa lags behind equivalent research in other regions of the
world, particularly Europe. High heat and variable preservation have been barriers to aDNA
recovery from samples excavated from this region of the world (Pinhasi, et al. 2015).
Methodologies that were successful with samples from cold, temperate regions with adequate
endogenous DNA preservation (i.e. Europe) have not been suitable for samples from arid or
humid environments. These environmental conditions expose tissues to destructive processes,
resulting in samples with very low endogenous contents. To utilize these small amounts of
aDNA and successfully assemble ancient African genomes, we trialed new method
combinations for optimal aDNA extraction that produced high quality libraries for NGS
techniques. This included a digest with an enzymatic solution prior to an overnight digestion (i.e.
double digest) or bone or tooth powder pretreated with a bleach solution before performing the
double digest. Here, we tested whether tissue pretreatment increased mapped ancient reads
and endogenous content and impacted library complexity. These results serve as a “best

85
practice” suggestion for sampling of ancient Nubian materials and may be extended to other
human remains preserved in similar conditions.
At this time, it is critical to optimize methods to prevent over- or unnecessary sampling of
tissues as paleogenomic work extends to additional African archaeological sites and more
samples are collected from museum assemblages. In the future, sensitive techniques for difficult
samples may be developed. Strategic sampling schemes should strive to preserve tissue
samples for future developments while also maximizing aDNA output from the tissues
consumed now.
We compared the performance of three extraction methods including the standard single
digestion, a bleach pretreatment, and double digestion on ancient Nubian tooth or bone
samples. We found that both bleach pretreatment and double digestion increased the
endogenous content and number of ancient mapped reads over the single digest method.
Furthermore, these two pretreatment methods maintained library complexity, an important
feature for downstream NGS applications. Similar results were obtained in other methodological
evaluations including Boessenkool, et al. (2017) and Damgaard, et al. (2015). However, t-tests
comparing the extraction techniques did not yield significant differences. This is likely explained
by our limited sample size, and we would expect our findings to reach significance thresholds
with increased sampling. In particular, we found that the bleach pretreatment was particularly
robust to rescue samples that would have otherwise had unsuccessful extractions of aDNA.
Differences in read counts of bleached vs. non-bleached samples with double digest was
approaching significance (p=0.05). Nevertheless, we suggest that these pretreatments should
be implemented in all future work.
Successful paleogenomic studies in Africa have utilized multiple types of skeletal tissue.
For example, Skoglund, et al. (2017) used teeth, petrosal bones, and long bone samples from
South and East Africa. In this trial, the type of tissue used for sampling was found to be
significantly associated with the ability to obtain aDNA libraries. In fact, teeth outperformed bone
tissue by producing more aDNA libraries. This result may have been driven by the many
successful trials from Ghazali (6 of 7 trialed). However, the biological composition of dentine
tissue also may explain better preservation of DNA. While bone and teeth are both composed of
two materials, specifically inorganic minerals (mostly hydroxyapatite) and organic protein
(collagen), the degree of mineralization differs. Tooth dentine has a mineral content of roughly
70%, while bone tissue is approximately 60% (Berkovitz, et al. 2017). This differentiation speaks
to the tissue’s biological functions within the human body: bone tissue is more flexible, dynamic,
and adaptable to stressors, while teeth are mostly unchanging (dentine may remodel, but not

86
enamel) and fortified for heavy load-bearing activities including mastication. A commonly held
theory is that DNA is preserved within the organic portion of the tissue; however, the
biochemical particulars are currently being explored. Previous studies have demonstrated that
the mineral component is as important, or more, to binding and preserving DNA than collagen
and this component is enriched with small-length genetic molecules as a result of tissue
degradation and cellular death (Campos, et al. 2012, Schwarz, et al. 2009). These findings may
explain why teeth outperformed bone samples when retrieving aDNA from these tissues.
Furthermore, when devising a sampling strategy, using teeth samples is less invasive given
there are up to 32 potential tooth samples compared to two temporal bones per individual.
Therefore, our findings encourage sampling of teeth rather than first harvesting petrous bones.
This is a key finding given that the currently recommended method for African materials is the
sampling of petrous portions, which is far more invasive than sampling from teeth.
The viability of genetic material also may be impacted by the burial context. We found
that interment method was significantly associated with successful extraction of aDNA
(p=0.00001). Skeletal remains from tumuli and pit inhumations performed poorly, indicating the
interment conditions likely destroyed DNA beyond recovery. The populations that performed the
worst were those from El-Detti (individuals T.2, T.4, T.5, T.7) and El-Zuma (individuals T.10,
T.11, T.15, T.17, T.18, T.28); none of these individuals were successful. These individuals were
interred in large rock-cut burial chambers or underneath tumuli. The humidity here seems to be
the likely problem and many samples were externally moldy, overly porous, and had a chalky
texture. Thus, we suggest that skeletal remains in these contexts may be trialed (e.g. teeth may
work better due to higher mineralization of tissue), but should not be intensively sampled if they
do not first yield any endogenous content.
Additionally, the El-Kurru population samples did not perform well or were highly
contaminated. These skeletal remains were inundated for hundreds of years and were in close
proximity to agricultural land. Furthermore, these remains were interred directly into the ground
with no physical barriers, like a box-grave constructed from slabs of rocks (i.e. Ghazali monastic
site), to protect the remains. This preservation difference has been corroborated with another
aDNA study conducted in the 4th Cataract region at Wadi Abu Dom. While this study was PCR-
based, the preservation of aDNA was poor with tumuli and best with box graves (Jugert, et al.
2018). Additionally, the successes from Sirak, et al. (2016) are notable, however the remains
from Kulubnarti (near modern Sudan’s north border and Aswan) had impeccable preservation
due to more extreme dryness in northern Sudan or Lower Nubia. In many cases, flesh and
organics were perfectly preserved by natural mummification (Armelagos & Van Gerven 2017,

87
Van Gerven, et al. 1995). In the Middle Nile region, the area is (relatively) wetter and
archaeological sites closer to the Nile River are more at risk. Therefore, the sample processing
should be specific to those remains from the Middle Nile basin. Lastly, in general, the
populations that performed the best were individuals from Ghazali (70% success rate) and
Kwieka (100%). The preservation of the tissues may have been better due to the site locations
in the desert, further from the Nile River bank, and/or not in proximity to wadis or irrigation that
would cause water level fluctuations. Including BMC and TARP, all field sites which produced
successful samples were located on the east bank of the Nile with no apparent (modern) wadis
or other water sources in the vicinity (with the exception of Ghazali which is somewhat near
seasonal wadis). The exception to this rule is El-Kurru, which was less than 400m from the west
bank and between two artificial water sources described above. While distance from the Nile
likely has an influence, no pattern was apparent. For example, the Ghazali, BMC, TARP,
Tanqasi are located far from the banks, around ca. 14km, 2.5km, 5.5km, and 4km, respectively,
while Kwieka and El-Kurru were less than a kilometer from the bank and all sites had at least
one successful ancient library produced. Those sites with no recovery of aDNA, El-Zuma and
El-Detti, range from 2.5km to 800m from the Nile and are in close proximity to other modern
sources of water, including irrigation channels or wadis, which may be responsible for water
fluctuations. Thus, these eight sites suggest distance from the river may not be a driving factor
for preservation; instead, seasonal or flooding water may affect the preservation of skeletal
tissue. In conclusion, the burial context and the resulting preservation of biological tissues of the
skeletal remains as well as the geographic location with proximity to water sources (that are not
the Nile River) should be considered when implementing an invasive sampling strategy for
future work in this region.
One of the major limitations of this study was the small sample size of library pairs that
were evaluated for number of mapped reads, endogenous content, and library complexity.
Forty-three individuals were trialed with at least two different methods. With a 40% success rate,
the final number of individuals compared was small. Additional trials may increase this success
rate and produce more pairs of libraries to compare. Next, we did not trial a single digestion with
bleach treatment. In the context of the results from the other trials, this method would assess
the necessity for a double digestion following the bleach pretreatment. Given that the single
digestion method produced viable libraries with aDNA with appropriate cluster factors (Figure
3.19), it would be worthwhile to trial this extraction strategy as it would lead to decreased
reagent consumption as well as subject the sample to less treatment and thus may reduce
contamination.

88
 Future directions of this work will include the implementation of both pretreatments for all
processing of samples from the Middle Nile region to verify these results. In addition, future
studies whose aim is to extract aDNA from the Middle Nile region should employ a strict
sampling strategy wherein sampling of remains from burial contexts that have been shown to be
problematic are excluded. Here, the impact of burial context and taphonomy has been shown to
impact the success of paleogenomic work and these factors should be considered when
sampling in the field. The impact of external factors on tissue preservation, such as aridity and
water fluctuation, requires further investigation and the geography of each archaeological site
should be observed to discern a pattern in the future. Considerations of burial context and
geography will encourage a more conservative strategy for collecting samples that will leave
material for future paleogenomic work (or other fields like paleoproteomics). Furthermore, the
comparison of sampling tissues also should be pursued to a larger extent. While less invasive
techniques have been developed (e.g. Sirak, et al. 2017), harvesting bone powder from the
petrous portion of the temporal bone remains destructive. If teeth are available and they
produce high-quality libraries, these tissues should be used in place of temporal bones.
Moreover, the use of dental calculus should also be piloted for human aDNA retrieval because it
has been successful with other Nile Valley samples (Neukamm, in prep) and is the least
invasive sampling technique (e.g. Ziesemer, et al. 2018). Again, this will conserve material for
future work and is typically all that is allowed for sampling of skeletal remains, especially where
museum collections are concerned.
As the field of paleogenomics expands across the globe, it is important to utilize the
materials we choose to sample in the most impactful way. As the first NGS study using skeletal
samples from the Middle Nile, this study demonstrated the viability of ancient DNA research in
region of Sudan and trialed a combination of methods to be implemented for further work.
These methodologies are recommended to other researchers performing aDNA extracts from
bone or teeth samples from Sudan.

CONCLUSION

Skeletal remains from archaeological excavations in the Middle Nile basin of Sudan was
used to extract ancient DNA from Nubian individuals. Forty percent of samples were successful
for NGS. Multiple methodologies were trialed to compare the amount of human DNA able to be
obtained. Two pretreatments, bleach and a double digestion, of the bone or tooth powder were
piloted separately and in combination before a standard overnight digestion. Except for one trial,
both pretreatments resulted in the increase of endogenous contents and amount of mapped

89
ancient human DNA reads. Additionally, these treatments did not impact the complexity of the
libraries built from these extracts. Tooth samples outperformed bone samples for obtaining
aDNA. Additionally, bone or tooth samples from protected burials (e.g. box grave construction)
performed the best, while those recovered from tumuli contexts did not. Moving forward, both
pretreatments should be used for future work involving skeletal materials from the Middle Nile
region. In order to include paleogenomics in research projects in Africa, especially those from
the Nile River Valley, viable access to these molecules using precious materials was a critical
hurdle to overcome. Now, DNA analysis from archaeological remains may be added to current
or future projects in this region as a valuable line of evidence, such as the research designed
employed here to join bioarchaeological data with paleogenomic while in context with the
archaeological settings.

90
CHAPTER 4: Ancient Mitochondrial Genomes from Middle Nile Region of Sudan:
A New Perspective of Nubian Ancestry

INTRODUCTION

Decades of archaeological, ethnographic, and linguistic work attest to the vast diversity
found in North East (NE) Africa, the region that encompasses the Nile River Valley including
Egypt, Sudan, South Sudan, and Ethiopia (Figure 1.2). The occupational history of this region is
one of the deepest known to humanity – from the origin of our species through developed
civilizations that flourished here for millennia. Furthermore, modern genomic studies show that
this region houses some of the greatest diversity on the African continent, which itself has the
most genetic variation on the planet (Tishkoff, et al. 2009).
Within the Valley, the Sudanese region (Sudan and South Sudan) is of particular interest
given its own extensive diversity and the presence of a historically underrepresented group of
inhabitants, the Nubians. Ancient Nubia spanned from modern-day Aswan in southern Egypt to
Khartoum in Northern Sudan (Figure 1.1). Past populations in this region were mobile with
broad trade contacts since prehistoric times, through the empire of Kush, the Medieval Christian
kingdoms, and the arrival of Islam in the 14th century CE. Nubians had a dynamic history with
their neighbors to the north – robust trade, conquest, colonialism – all of which may have
contributed to their genetic ancestry (as outlined in Chapter 1). Despite this entangled
relationship, the modern profile of Nubians is unique and has been the subject of ongoing
genetic projects during the past three decades, summarized in Chapter 1 (Alfonso, et al. 2008,
Krings, et al. 1999, Lalueza Fox 1997, Hassan 2009, Hassan, et al. 2008, Dobon, et al. 2015,
Hollfelder, et al. 2017). However, the lens of modern population data has only described the
genetic landscape up until the Islamic conquest, while ancient demographic trends before this
large migratory event remain speculative or unknown. Sampling from archaeological
populations provides an opportunity address these unknowns by directly characterizing genomic
structure in the past. Such data will contribute to our understanding of human prehistory in the
region and reconstruct the multifaceted history of Nubians.

91
 When the field of paleogenomics appeared, samples from the Nile Valley were some of
the first to be trialed (Pääbo 1985). Unfortunately, serious complications with contamination
sidelined the exploration of this resource for ancient molecular studies (Hofreiter, et al. 2014,
Pääbo & Wilson 1985). Paleogenomic work was simply unsuccessful when using archaeological
material from such a hot, dry environment which degrades ancient DNA to the point of no
recovery. However, recent advances in biotechnology and optimization of methodologies have
opened an exciting opportunity to scientific work. This includes the use of parallel sequencing
platforms or next-generation sequencing techniques, specialized extraction treatments (e.g.
bleach pre-treatment), and use of petrous portions of the temporal bone for sampling, which is
known to contain more authentic DNA, even from hot, dry climates (Hansen, et al. 2017,
Gamba, et al. 2014, Pinhasi, et al. 2015, Gallego Llorente, et al. 2015, Dabney, et al. 2013,
Kircher, et al. 2011). These advances now open the Middle Nile, and likely much more of Africa,
to the genomic revolution and will continue to expand our knowledge about this continent’s
populations and human migrations (Hofreiter, et al. 2015, e.g. Schlebusch, et al. 2017,
Skoglund, et al. 2017).
While most knowledge about this region of the world is informed by historical,
anthropological, and archaeological data, a paleogenetic study using complete mtDNA has yet
to be published, making the present study the first. Furthermore, it includes an unprecedented
time-series of samples from underrepresented African groups. The samples elucidate the
demographic history of this region dating to the time of Kushite power along the entire Nile
(around 8th century BCE), the later invasion of foreign rulers, followed by restructuring of the
Kingdom of Kush at Meroe (3rd century BCE), the rise of Christian Kingdoms (7th century CE),
and ending with the Arab expansion (14th century CE). Such events (e.g. invasions, migrations,
entanglement) affect the genetic ancestry of the individuals, which in turn can be analyzed to
reconstruct past demographic events (Krings, et al. 1999, Hollfelder, et al. 2017). This sample
set, which varies in time and space (Table 3.1), was uniquely powered to observe the dynamics
of population structure in real-time while simultaneously contributing to the growing interest
centered on the genetic past of indigenous African populations.
The primary aim was to genetically characterize the inhabitants of Nubia by extracting
DNA from archaeologically excavated human remains and using these data for population
genetic analyses. Sequence data from these ancient individuals allowed us to describe variation
within the Nubian populations and begin to reconstruct ancestry of each individual to understand
the diversity that exists between and within the ancient groups. Due to the lack of ancient
comparative data, extant Sudanese are used for comparison. Additionally, metapopulations of

92
modern Sudanese are grouped by region from previous datasets (e.g. Hassan 2009, Babiker, et
al. 2011). Specifically of interest was the genetic makeup of Nubians before the arrival of Islam,
which can be used to more accurately time the major influx of Eurasian component in the Middle
Nile basin. These data stand to test hypotheses of population continuity as well as how ancient
Nubians have contributed to the genetic diversity of modern inhabitants of the Nile Valley by
comparing them to extant Sudanese and regional neighbors, given geography plays such an
influencing role. Lastly, these data may characterize a distinct genetic signature present in the
Middle Nile region which may be differentiated from other areas, i.e. Lower Nubia, Egyptians, or
other ancient populations.

MATERIALS

Of the eight sample populations and 43 unique individuals, six individuals from four
sample populations had aDNA successfully extracted from skeletal materials. Outlined below
are the specific archaeological contexts for each individual in addition to a few details on the
sites and the sampling tissue.

Berber Meroitic Cemetery Individual 6 / BMC 6a

Near the 5th Cataract, the Berber Meroitic Cemetery has been excavated by NCAM for
several seasons since 2009 after building foundation trenches disturbed these burials (Figure
1.1). This site was occupied during the Meroitic period ca. 50-350 CE, has 38 tombs, and three
mudbrick pyramids (Bashir & David 2015) (Figure 4.1). Body positioning was related to the
burial chambers found at this site. Oval graves had contracted or semi-flexed skeletons, while
east-west burials had extended body positioning within a box grave.

93
Figure 4.1. Map of Berber Meroitic Cemetery.

Dating estimations based on radiocarbon dating, pottery types, and decorations found in context with the burials
(Bashir & David 2015). Tomb 6 is dated to the 3rd Century CE (arrow).

Located in the southern part of the cemetery, the burial niche for BMC 6 was disturbed
by a foundation trench, but excavations revealed it contained intact pottery, kohl pots of ebony
and ivory, and preserved wood (Figure 4.1) (Bashir & David 2015). Regardless of robber
disturbances, a sandstone offering table with cursive Meroitic writing naming the deceased,
Sobt, was found in the grave (Figure 4.2A). This artifact dates the grave to mid-3rd century CE
(ca. 250 BCE) (Bashir & David 2015). Skeletal data is unavailable at this time. The right
mandibular 1st premolar was extracted from the mandible for aDNA sampling (Figure 4.2B).

Figure 4.2. Burial context of Individual BMC 6a and tooth sample.

A B

A) Sand stone offering table artifact with cursive Meroitic writing discovered in context with BMC 6a. B) Right
mandibular 1st premolar for Individual BMC 6a.

94
Ghazali Individual 4-008.2 / GHZ 2

Ghazali 4-008.2/GHZ 2 Individual was interred in Cemetery 4 within the Early Christian
monastery complex at Ghazali in the 4th Cataract region (Figure 4.3). The main church has been
radiocarbon dated to 670-770 CE. This cemetery was somewhat isolated on site and those
buried here have inconsistent burial treatments, as opposed to those in the other larger
cemeteries at Ghazali, and remains show evidence of interpersonal violence (Stark & Ciesielska
2019). Project archaeologists suggest this cemetery has a different, but unknown function
compared to the other three in the vicinity (Stark & Ciesielska 2019).

Figure 4.3. Ghazali complex located near 4th Cataract, Sudan.

Line map drawing (left) of cemeteries, settlement area and monastery complex at Ghazali. Aerial kite photograph
(right) of the monastery. Cemetery 1 visible to the west of the monastery where GHZ 5 was interred, Cemetery 4
(where GHZ 2 is buried) is not mapped. Map obtained from Stark & Ciesielska (2019).

Excavated in 2016, GHZ 2 was buried beneath a superstructure consisting of vertically

arranged rocks from the local landscape that outlined the grave (Figure 4.4A). The body
treatment shows typical Christian burial traditions, including a fully extended body, arms
extended at the side, head is encased with flat rocks (two framing the skull and one over the
face). Also, some traces of a burial shroud still remain in situ (Figure 4.4B). Anthropologists
estimated GHZ 2 as a female, aged 30-40 years old. The right 2nd maxillary molar was obtained
for aDNA extraction (Figure 4.4C).

95
Figure 4.4. Burial context of Ghazali-4-008.2/GHZ 2.

Ghazali individual 4-008.2 from cemetery 4. A) Superstructure covering the grave of this individual. B) Skeletal
remains of GHZ 2 in situ. C) Right 2nd maxillary molar, before cleaning.

Ghazali Individual 1-010 / GHZ 5

GHZ 5 was buried in Cemetery 1, west of the Early Christian monastery complex at
Ghazali. The main church has been radiocarbon dated to the Early Christian period (670-770
CE) (Figure 4.3). This cemetery was reserved for ad sanctos burials for people in the
surrounding areas wishing to be interred near a holy place, i.e. the monastery. The skeletal
remains of GHZ 5 were contained within a stone box-grave, oriented Northwest-Southeast, with
a large stone superstructure (Figure 4.5A). The grave was constructed in typical Nubian
Christian traditions with regards to the superstructure (like GHZ 2). The body was fully extended
and supine with the arms extended at the sides and hands clasped over the pelvis (Figure
4.5B). Evidence of a burial shroud were discovered. The left upper 2nd molar was sampled for
aDNA extraction (Figure 4.5C).

96
Figure 4.5. Burial context and sampling of GHZ-1-010/GHZ 5.

A) Exposed rocky super structure of burial for GHZ-1-010 / GHZ 5. B) In situ skeletal remains of GHZ 5 at the bottom
of the burial pit. C) Left Maxillary 2nd molar, before cleaning.

Kwieka T2 Individual / KWC 2

The cemetery at Kwieka was likely occupied during Meroitic (350 BCE – 350 CE), Post-
Meroitic (350 – 550 CE), and Christian periods (550 – 1450 CE) as per the burial traditions
observed during excavations (Abdallah 2018). This site was excavated in 2016 by NCAM
officials and its archaeological context is described by Abdallah (2018) (Figure 4.6).
Figure 4.6. Aerial view of the site of Kwieka near the 5th Cataract region.

97
 Kwieka T2 individual was interred near the eastern edge of the excavated area of the
cemetery, dated contextually to the Meroitic period based on grave construction, approximately
350 BCE-350 CE (Figure 4.6, Figure 4.7A). The burial had a rounded superstructure of quartz
stones (Type F), a rectangular tomb shaft, and an oval chamber cut into the sandstone (Figure
4.7A, B) (Abdallah 2018). The skeletal remains were found within this grave cut, along with a
clay storage bin (or quseiba), and a broken beer cup (Figure 4.7A). NCAM archaeologists
determined the skeleton was completely disturbed by robbers, thus its original positional cannot
be determined. Kwieka T2 was estimated to be an adult and the intact right temporal bone was
used for aDNA sampling (Figure 4.7C).

Figure 4.7. Burial context of KWC T2, associated finds, and sampling material.

A) Circular superstructure of KWC T2 (upper) and rectangular substructure (lower), with associated finds (inset)
(Saad 2018). B) Adult skeletal remains of KWC 2 with evidence of disturbance. C) Right temporal bone for sampling.

Kwieka T6 Individual / KWC 6

The grave of individual KWC T6 was discovered by local villagers when mining gravel
used for local building projects. The grave had a rounded, flat-topped gravel superstructure, up
to 7m in diameter and half a meter high (Classified as a Type D tumulus) (Abdallah 2018). The
underground substructure was a 1.3m-deep rectangular shaft with a sub-rectangular burial
chamber and mud-brick blockage wall that showed signs of robber activity (Figure 4.8A). Within
the burial chamber, the skeletal remains were found to be disturbed and some ceramic sherds
were found within the fill (Figure 4.8A). The skeleton was that of a child and thus, no skeletal
sex estimation was conducted. The intact left temporal bone was used for sampling to extract
aDNA (Figure 4.8B).

98
Figure 4.8. Burial context of KWC T6 and sampling material.

A) Disturbed skeletal remains of KWC T6, immature long bones seen without distal epiphyses. B) Left temporal bone
used for aDNA sampling.

TARP G9 individual / TARP 9

The Tinga Archaeological Rescue Project took place near the 5th Cataract and in close
geographic proximity to the Berber Meroitic Cemetery. Excavations were conducted in 2017 by
NCAM officials. The archaeological context of burials exhumed were described by Bushara, et
al. (2018). TARP individual G9 was interred in a rectangular grave, oriented East-West, with a
traditional wooden bed. This grave type is similar to Napatan style burials found down river in
Meroe and at Berber, another 5th Cataract site north of TARP. The skeleton was found in a
typical flexed position on a wooden bed, head towards the west end of the grave and facing
south. The remains belong to a child and were not assessed for biological sex. The right
temporal bone was used for aDNA sampling (Figure 4.9).

Figure 4.9. Right temporal bone from TARP T9.

TARP = Tinga Archaeological Rescue Project, T = tomb number

99
METHODS

Ancient DNA Extraction, Library Assembly, and Sequence Data Processing

A full description of the processing, pretreatment, and extraction of DNA is outlined in

the Chapter 3. All samples are approved for destructive analysis as per agreements with the
directors of the archaeological excavations and/or the National Committee for Antiquities and
Museums from Sudan. In brief, all work was performed in a designated clean lab at the Institute
of Evolutionary Medicine at the University of Zürich or the Ancient Laboratories at the University
of Oslo in Blindern, Norway. Bone and teeth samples were externally decontaminated with UV
irritation treatments in a crosslinker. Sterile powder was collected from the petrous portion of the
temporal bone or the inner dentine of intact teeth using methods optimized for the tissue type
and integrity (i.e. crushing or drilling). Powder (100-200mg) was subjected to a pretreatment of
bleach or a mild pre-digestion step or both as per Korlevic & Meyer (2019) and Schroeder, et al.
(2019). DNA was extracted following Gamba, et al. (2014) with modifications from Dabney &
Mayer (2019). DNA extracts were converted into paired-end Illumina libraries as outlined in
Meyer & Kircher (2010), indexed for sequencing as per Kircher, et al. (2012). All libraries were
shotgun sequenced on an Illumina HiSeq4000 to screen for libraries containing DNA displaying
typical molecular characteristics of degraded genetic material and estimate endogenous
content. Those libraries selected for further analysis (i.e. low contamination, endogenous DNA
present, typical damage profiles) were enriched for human mitochondrial DNA using in-solution
hybridization capture (Maricic, et al. 2010). These libraries were sequenced at a high
percentage on an Illumina NexSeq sequencing platform and negative control samples on an
Illumina MiSeq Micro.
A more in-depth description of the processing of sequence data, is found in the previous
chapter. Generally, raw reads from the ancient Nubians and negative controls were processed
with the bioinformatic pipeline EAGER (v1.92) to assess mapped reads to the human
mitochondrial genome and again estimate criteria for authenticity (Peltzer, et al. 2016). Those
libraries demonstrating typical molecular behavior of genetic material consistent with age and
degradation were analyzed with the tool Schmutzi to estimate modern contamination and build
endogenous consensus sequences (Renaud, et al. 2015). Libraries with low contamination (<
5%) were merged when necessary for each individual and consensus sequences were used for
haplogroup assignment. Coverage plots were generated for each individual used in the
haplogroup analysis with an in-house R script.

100
 Samples later included in the haplogroup analysis were molecularly sexed by utilizing a
bioinformatic pipeline and shotgun read data from the first screening of viable genetic material
(Mittnik, et al. 2016). This pipeline used the reads which mapped to the sex chromosomes and
measured the ratio of the X and Y chromosomes to estimate a male or female karyotype
(Mittnik, et al. 2016). Values classified individuals as being “assigned” with a certain karyotype
or “consistent” with one karyotype but not the other. These data were compared to those
estimations of sex (and age) done by archaeologists (listed above or Chapter 3, Materials) using
standard bioarchaeological methods, characterizing discrete traits of the skull and post-crania to
estimate age-at-death and degree of sexual dimorphism (Buikstra & Ubelaker 1994). Sex
estimation was not performed on subadult individuals.

Haplogroup Assignment

Using the consensus sequence files generated from the previous step, haplogroups
were assigned using HaploGrep 2 (https://haplogrep.uibk.ac.at/) (Weissensteiner et al 2016).
Consensus sequences used for this analysis was chosen by balancing the highest rank
achieved and highest quality data (i.e. lower quality (Q20) may have less ‘no calls’ or N’s, but
rank or confidence of haplotype assignment may not increase with more nucleotide calls;
therefore, a higher quality sequence would be selected for assignment).

Comparative Dataset Assembly and Alignment

Two comparative datasets – haplogroup frequencies and a whole MT genome dataset –

were assembled from publicly available sources and databases: HmtDB (Clima, et al. 2016),
National Center for Biotechnology Information Nucleotide Database (Benson et al 2018),
European Molecular Biology Laboratory (EMBL) Nucleotide Sequence Database (Baker, et al.
2000), other collaborations, and other published manuscripts outlined below.
For haplogroup frequencies, there is a dearth of mitochondrial genomic data for the
country of Sudan; however, there are two data sources with limited availability and coverage of
the mitogenome that are useful for context of Sudan. One is Krings, et al. (1999) sequenced the
HVRI of 68 Egyptians, 80 Nubians, and 76 Southern Nilotes. The other is Hassan (2009, Table
4.1) that provides haplotype frequencies for six geographic groups that include 15 ethnic groups
for comparison. These frequencies were derived from sequence data from the HVRI region, but
these sequences are no longer accessible. These frequencies and others from the Northeast
region of Africa provided by Abu-Amero, et al. (2008) (Nile River Valley countries) and
Fadhlaoui-Zid, et al. (2011) (Libya) were merged into one dataset (Table 4.1).

101
Table 4.1. Sources and group counts for haplotype frequencies data for dataset assembly.
Population or Subpopulation N Published source
Libyans 268 Fadhlaoui-Zid, et al. 2008
Egyptians 68 Krings, et al. 1999
58 Stevanovitch, et al. 2003
118 Rowold, et al. 2007
Nubians (Northern Sudan) 79 Krings, et al. 1999
29 Hassan 2009
Beja (East) 48 Hassan 2009
Central Sudanese (Arab) – Arakein, Copts, Gaalien, Hausa 103 Hassan 2009
Western Sudanese – Borgu, Masalit, Fur 87 Hassan 2009
Southern Nilotes – Dinka, Shilluk, Nuer, Nuba 94 Hassan 2009
Pastoralists – Fulani, Meseria 43 Hassan 2009
Ethiopians 74 Thomas, et al. 2002
270 Kivisild et al 2004
Kenyans 99 Brandstätter, et al. 2004
42 Rowold, et al. 2007
Geographic groups in bold typeface and the subpopulations included in macrogroup listed underneath.
N = number of individuals, count.

For whole MT genome analyses, both modern and ancient whole mitogenomes were
obtained for 2,006 individuals from all of Africa, northern, western, and central Europe, and the
Middle East, dating as far back as 15,000 years ago to as recent as the past few centuries.
While a large part of this dataset is from extant human populations (i.e. 1,740 individuals), a
significant amount are ancient genomes. Two hundred sixty-six ancient individuals cover North
Africa, from the Canary Islands to Egypt (Schuenemann, et al. 2017, van de Loosdrecht, et al.
2018, Vai, et al. 2019, Fregel, et al. 2019, Fregel, et al. 2018), East Africa, including Kenya and
Ethiopia (Prendergast, et al. 2019, Gallego Llorente, et al. 2015), then continuing southward to
Tanzania, Malawi, & South Africa (Skoglund, et al. 2017, Schlebusch, et al. 2017, Morris, et al.
2014) (Figure 1.8). Those outside the African continent include 22 individuals from the Levant
area – namely ancient Natufians, Anatolians, and other Near Eastern farmers (Lazardis, et al.
2016) (Figure 1.8) – and five ancient individuals from Europe (Fu, et al. 2013, not pictured in
Figure 1.8).
For those ancient individuals with no publicly accessible mitogenomes, raw files were
obtained from databases mentioned above, mapped to the human mitochondrial reference
sequence, and VCF files were created to be uploaded in Haplogrep 2 to generate a FASTA file

102
containing the mitogenome for each individual (Weissensteiner et al 2016). For those
publications with published polymorphism files for ancient individuals, these data were formatted
and uploaded into Haplogrep 2 to generate FASTA files containing the mitogenome for each
individual (Weissensteiner et al 2016). Haplogroup assignments were checked for each
generated FASTA file with those in the published works. Additionally, those with low overall
ranks (< 75%) from Haplogrep 2 were not included.
All MT sequence data, including those of the six ancient Nubians, were aligned using
AliView v1.26 (Lasson 2014). For whole MT analyses, populations groups were collapsed into
“macrogroups” for haplogroup frequency PCA analysis and pairwise difference analyses (e.g.
PCA, FST, MDS). Table 4.2 outlines the countries of origin which characterized these collapsed
groups. MitoBench v1.1 beta (Neukamm & Peltzer 2019) was used to compile and generate
files to be used in Arlequin (Excoffier & Lischer 2010) for FST calculations.

Table 4.2. Macrogroups used for analyses and which populations are included within each one.

Macrogroup N Countries or Population Groups Included in Macrogroup

AncientEuropean 5 Czech Republic, Germany, France, Luxemburg
AncientNearEastFarmers 43 Turkey, Armenia, Jordan, Israel, Iran
AncientNorthAfrican 67 Libya, Morocco, Canary Islands
AncientEgyptiansPPP 60 Northern Egypt (Abusir el-Meleq)
AncientEgyptiansPtoP 16 Northern Egypt (Abusir el-Meleq)
AncientEgyptiansRoman 14 Northern Egypt (Abusir el-Meleq)
AncientNubians 6 Sudan (Upper Nubia)
AncientEastAfrican 50 Kenya, Malawi, Tanzania, Ethiopia
AncientSouthAfricans 11 South Africa
WesternEuropeModern 81 Spain, Portugal, France
NorthernEuropeModern 17 United Kingdom, Ireland, England, Scotland, Finland
CentralEuropeModern 21 Netherlands, Poland, Germany, Czech Republic
SouthernEuropeModern 90 Italy, Sicily, Bulgaria, Romania, Greece
SardiniaModern 6 Sardinia
Saudi Arabia, Kuwait, Oman, Yemen, United Arab Emirates, Iraq,
MiddleEastModern 301 Syria, Lebanon, Cyprus, Israel, Jordan, Palestine, Georgia, Iran,
Pakistan, Armenia, Turkey, Azerbaijan
NorthAfricaModern 165 Algeria, Canary Islands, Libya, Morocco, Tunisia
EgyptModern 70 Egypt
SudanModern 68 Sudan
EthiopiaModern 56 Ethiopia
EastAfricaModern 147 Eritrea, Ethiopia, Kenya, Rwanda, Somalia, Tanzania, Uganda
Chad, Equatorial Guinea, Cameroon, Central African Republic,
CentralAfricaModern 74 Democratic Republic of Congo, Sao Tome and Principe, Lake
Chad
SouthernAfricaModern 592 Angola, Botswana, Mozambique, Namibia, South Africa, Zambia

103
 Guinea-Bissau, Gambia, Ghana, Ivory Coast, Mali, Mauritania,
WestAfricaModern 52
Nigeria, Senegal, Niger, Burkina Faso
PPP, Pre-Ptolemaic Period; PtoP, Ptolemaic Period
“Ancient” populations refer to where samples were excavated within modern political boundaries.
N = number of individuals, count

Haplogroup Profile Construction

To visualize the varying genetic components that describe the ancestry of a defined
population, haplogroup profiles were compiled from published data or calculated from
haplogroup frequency counts found in published data. Since mitochondrial DNA is a single
marker, each individual contributes one data point as a haplogroup assignment. Therefore, a
large number of individuals, or a large sample number, was required to create an accurate
profile of the population. Raw counts of individuals were tallied and frequencies of haplogroups
were calculated as percentages, then subgroups were collapsed into macro-haplogroups to
simply the profile (e.g. L0a1a grouped as L0). Profiles were constructed based on these
percentages and were visualized as staked columns or pie charts (both equal to 100%), then
were color-coded to correspond to informational groupings of these haplogroups (e.g. all African
L lineages are warm colors).

Population Comparison Analyses

Using the Haplogroup frequencies dataset, a Principle Component Analysis plot was
constructed in MitoBench v1.1 beta (Neukamm & Peltzer 2019). MitoBench was also used to
construct Arelquin files for FST values calculations to estimate between population differences.
FST values were generated in Arelquin 3.5 (Excoffier & Lischer 2010). The MDS plot to visualize
the genetic distances was created with an in-house R script and implemented in R Studio.

RESULTS

Contamination Estimation Analysis

Bone and teeth samples from 43 Nubian individuals were piloted using various
extraction methods to obtain ancient genetic material. With the first screening step of shotgun
sequencing, these methods were successful in yielding authentically ancient DNA – either
mitochondrial or nuclear – from 17 of these individuals, a 40% recovery rate. Given low reads
mapping to the human mitochondrial genome, a capture step was performed using baits and in-
solution beads to enrich for the mitogenome. Twenty-two libraries (and all negative controls)
were enriched for human mitochondrial DNA (Table 4.2). (Some libraries with ancient reads
were not included due to over lapping indexes and time constraints). Negative controls

104
established no cross contamination of ancient material and a negligible amount of modern
contamination throughout the process; these results are detailed in Appendix B. Of these
enriched libraries, seven failed, five were successful, but were estimated to be heavily
contaminated, and 10 were successful enrichments with low contamination (Table 4.2). For
those libraries from the same individual, raw files were merged and run through Schmutzi
(“_merged”, listed at the end of table, green). All merged files had less contamination and the
consensus sequence built from the merged reads was used for all analyses following this. For
sample BMC 6a, merging the reads decreased the overall contamination rate, despite being
constructed from the double digest library having a moderate contamination estimation.

Table 4.2. Libraries enriched and analyzed with Schmutzi.

Final
Archaeological
Sample Name contamination Low (%) High (%) Comment
Site
estimation (%)
Berber Meroitic
BMC 6a_DD 0.12 0.11 0.13 Switching
Cemetery
BMC 6a_SD 0.01 0.0 0.02 Stable
BMC 6a_merged 0.02 0.01 0.03 Stable
GHZ 2_merged 0.01 0.0 0.02 Stable
GHZ 5_merged 0.01 0.0 0.02 Stable
KWC 2_merged 0.01 0.0 0.02 Stable
KWC 6_merged 0.01 0.0 0.02 Stable
El-Detti DET 7_oslo NR - - -
Ghazali GHZ 1_DD NR - - -
GHZ 2_DD 0.01 0.0 0.02 Switching
GHZ 2_SD 0.01 0.0 0.02 Stable
GHZ 3_SD 0.99 0.98 0.99 Stable
GHZ 5_DD 0.01 0.0 0.02 Stable
GHZ 5_SD 0.01 0.0 0.02 Stable
GHZ 6_SD 0.99 0.98 0.99 Stable
El-Kurru KU 210B_oslo NR - - -
KU 212_oslo 0.99 0.98 0.99 Stable
Kwieka Cemetery KWC 2_DD 0.02 0.01 0.03 Stable
KWC 2_SD 0.01 0.0 0.02 Stable
KWC 6_oslo 0.01 0.0 0.02 Stable
KWC 6B_oslo 0.01 0.0 0.02 Stable
TAQ 179_oslo 0.0 0.0 0.005 Stable
Tinga
Archaeological TAR 22_DD NR - - -
Rescue Project
TAR 9B_oslo 0.01 0 0.02 Stable
El-Zuma ZUM 18B_oslo NR - - -

105
 ZUM 28_DD NR - - -
ZUM 28_SD NR - - -
Reads were run through the Schmutzi pipeline to build a consensus mitochondrial sequence using endogenous (i.e.
ancient) reads. The contamination estimation (%) of modern human contamination is reported (“Final”) as is the
average between the “High” and “Low” values reported from the multiple iterations of the pipeline. Comments about
the contamination estimation are noted in that last column; “stable” refers to consensus, calling, and contamination
estimations run iteratively that reached a stable estimation value, “switching” indicates inconsistent values were
reached during iterative runs, suggesting that the results should not be necessarily trusted. Samples with “NR” as the
final estimation have too few reads for this tool.
BMC - Berber Meroitic Cemetery, DET - El-Detti, GHZ – Ghazali, KU - El-Kurru, KWC - Kwieka Cemetery, TAQ –
Tanqasi, TAR - Tinga Archaeological Rescue Project, ZUM - El-Zuma, DD - double digestion method, SD - single
digestion, oslo – Oslo method used (i.e. double digestion) and prepped at the University of Oslo, B - bleach
pretreatment, NR - Not Reported.
Green – enrichment successful, not contaminated
Yellow – enrichment successful, contaminated
White – enrichment failed

Sequence Quality Analysis

Six individuals had a viable number of endogenous reads for consensus sequence
reconstruction of the entire mitogenome (with contamination estimations of 2% or less): BMC
T6a (BMC 6a_SD/BMC 6a_DD/BMC 6_merged), Ghazali 4-008.2 (GHZ 2/GHZ 2_SD/GHZ
2_DD), Ghazali 1-010 (GHZ 5/GHZ_SD/GHZ_DD), KWC T2 (KWC 2/KWC 2_SD/KWC
2_DD/KWC_merged), KWC T6 (KWC 6B_oslo/KWC 6_oslo/KWC 6_merged), and TARP G9
(TAR9B_oslo). TAQ179_oslo was not contaminated, but read count and endogenous content
were low. These six individuals had reads mapping to human mitochondrial reference genome
ranging from 3,356 to 146,987 (Table 4.3). Samples had a mean coverage ranging from 3X to
450X with some individuals (i.e. GHZ 5 and KWC 2) up to 100% of bases covered at least
fivefold or more (Table 4.3, Figure 4.10). Two individuals, GHZ 5 and KWC 2, had immensely
high coverage resulting in very confident reconstructions of their respective full mitogenomes
(an example, KWC 2 depicted in Figure 4.11). Some individuals had two libraries built from their
sample extracts. For these individuals, raw data was merged for analysis. For example, libraries
for individual KWC T6 (KWC6) were obtained from two separate bone powder samples,
“KWC6_oslo” and “KWC6B_oslo", obtained using different methods (see Chapter 3 for more
details), and were merged into the file “KWC6_merged” for analysis. Sequence quality (read
count, damage, etc.) was determined for all libraries, merged or unmerged.

106
Figure 4.10. Coverage of the Mitochondrial Genome of Nubian Individuals.

Graph of six Nubian individuals with full genome coverage. The x-axis depicts the extent of coverage as a function of
the entire mitochondrial genome, shown on the y-axis up to 100%. KWC - Kwieka Cemetery, GHZ – Ghazali, BMC -
Berber Meroitic Cemetery, TARP - Tinga Archaeological Rescue Project. Coverage graph was generated using
Qualimap BamQC.

Table 4.3. Sequence Quality Analysis Summary - from EAGER analysis of enriched reads.
Damage Damage Average
Number Mean
Endogenous Coverage(%) at 1st at 1st Fragment
Sample Name of Mito MtDNA
DNA (%) 1X 5X Base 5’ Base 3’ Length
Reads Coverage
(%) (%) (bp)
BMC 6a_SD 2,112 0.84 5.9 95.74 59.63 0.199 0.246 46.39
BMC 6a_DD 1,071 1.34 3.4 42.81 29.11 0.201 0.160 53.00
BMC 6a_merged* 3,641 1.23 10.6 96.98 79.95 0.202 0.218 48.04
GHZ 2_SD 2,756 0.80 7.8 98.71 79.09 0.216 0.236 47.06
GHZ 2_DD 673 1.01 2.12 38.27 18.31 0.210 0.310 52.37
GHZ 2_merged* 4,443 0.91 13.0 99.04 90.73 0.218 0.239 48.60
GHZ 5_SD 29,647 11.96 116.2 100 100 0.247 0.256 64.96
GHZ 5_DD 8,927 16.05 27.4 99.72 97.54 0.224 0.222 50.80
GHZ 5_merged* 146,987 12.84 450.6 100 100 0.258 0.269 50.82
KWC 2_SD 4,296 0.10 11.5 99.77 88.29 0.456 0.542 44.53
KWC 2_DD 14,171 0.50 38.2 100 99.93 0.489 0.546 44.65
KWC 2_merged* 27,465 0.44 71.3 100 99.99 0.486 0.526 43.03
KWC 6_oslo 3,609 0.07 10.3 98.42 79.35 0.497 0.494 47.16
KWC 6B_oslo 2,808 0.11 7.2 97.08 65.84 0.475 0.493 42.69
KWC 6_merged* 6,356 0.09 17.4 99.32 93.18 0.490 0.498 45.27
TARP 9B_oslo* 3,691 1.36 11.1 99.83 91.22 0.244 0.250 49.87

107
Results obtained from EAGER pipeline (Peltzer, et al. 2016). Output includes number of reads mapped to
mitochondrial reference sequence. Endogenous content was calculated by the number of reads obtained/reads
mapped to the human reference. Coverage was calculated from the number of unique reads that included a given
nucleotide in a reconstructed sequence. Percentages reported for coverage at 1X to 5X coverage. Percent damage
(misincorporations/lesions) recorded for both ends of reads. Naming example: “KWC 6_merged” combined reads
from “KWC 6_oslo” and “KWC 6B_oslo.”
* Libraries used for haplogroup analysis
GHZ - Ghazali, KWC - Kwieka Cemetery, TARP - Tinga Archaeological Rescue Project, BMC - Berber Meroitic
Cemetery; bp - base pair

Figure 4.11. Example of Coverage Plot of KWC 2 on the circular human mitogenome.

Logarithmic scale for coverage amount (X). KWC 2 (Ind. Kwieka Cemetery 2) averaged 71X.
To authenticate the mapped reads as ancient, nucleotide misincorporation patterns were
observed to be consistent with typical ancient genetic material. Percent damage at the first base
on either the 5′ or 3′ end ranged from 20% to 53% (Table 4.3, Figure 4.12). Further
authenticating these reads, the average fragment length of the six individuals varied from 43 to
51 base pairs (Table 4.3, Figure 4.13).

108
Figure 4.12. Map Damage Profiles of Mapped Mitochondrial Reads from merged library
samples.

Misincorporation patterns observed for mapped sequences to the human mitochondrial genome following enrichment
steps. Reads used totals: A) GHZ 2 (4,443 reads) B) GHZ 5 (146,987 reads) C) KWC T2 (27,465 reads) D) KWC T6
(6,356 reads) E) TARP T9 (3,691 reads) F) BMC 6a (3,641 reads). X-axis maps the position from the end of the
sequence. Y-axis is the frequency of misincorporations of the mapped reads, in percentage. The cytosine to thymine
(C>T) deamination rate at the 5’ end is labeled in red (left graph). Guanine to adenosine (G>A) deamination rate at
the 3’ end is labeled in blue (right graph). Fuchsia indicates insertions, green indicates deletions, grey indicates other
misincorporations. Scales are not identical for each individual. Map Damage Profiles, generated by mapDamage 2.0

109
(Jonsson, et al. 2013). GHZ - Ghazali, KWC - Kwieka Cemetery, TARP - Tinga Archaeological Rescue Project, BMC
- Berber Meroitic Cemetery.

Figure 4.13. Read Fragmentation Plots of Mapped MT reads from merged library samples.

Read length distributions for samples mapping to human mitochondrial reference, generated by mapDamage2.0
(Jonsson, et al. 2013). Reads used totals: A) GHZ 2 (4,443 reads) B) GHZ 5 (146,987 reads) C) KWC T2 (27,465
reads) D) KWC T6 (6,356 reads) E) TARP T9 (3,691 reads) F) BMC 6a (3,641 reads). X-Axis displays Read Length,
as measured in base pairs; ranges are not equal for all graphs. Y-axis measures all reads mapped by abundance
including positive and negative strands, ranges different for all graphs. EAGER pipeline removes all sequences below

110
30bp that are likely contamination or too damaged and are typically not included in analysis. This parameter can be
changed/extended, however sequences less than 30bp were not observed to be mapped for these samples. Scales
are not identical for each individual. Map Damage Profiles, generated by mapDamage 2.0 (Jonsson, et al. 2013).
GHZ - Ghazali, KWC - Kwieka Cemetery, TARP - Tinga Archaeological Rescue Project, BMC - Berber Meroitic
Cemetery.

Molecular Sexing, Archaeological Context, and Haplotype Assignment

These six individuals span a wide breadth in time (Napatan (ca. 800-300 BCE) to
Christian periods (ca. 600-1450 CE)) and space (3rd to 5th Cataracts). Additionally, various
methodologies were utilized to obtain full coverage of the mitogenomes (Table 4.4). The
archaeological context, sample material, and specifics of molecular sexing are individually
outlined below in Table 4.4 and Table 4.5. Results of haplotype assignments with confidence
percentages and specifics of the consensus sequence used (including contamination rates) are
outlined in Table 4.6.
Two individuals were assigned to L lineage haplogroups, GHZ 2 – L2a1 and TARP 9B –
L0a1a. Both of these haplogroups are common in modern Sudanese, 7.7% for L2a1 and 10.2%
for L0a. However, haplogroup L0a is only common today within Southern Nilotes (Nuba, Dinka)
(Hassan 2009). The other four Ancient Nubians were assigned non-Africa haplogroups (Figure
1.5). H2a was assigned to two individuals, GHZ 5 and KWC 2. Lineage N was assigned to KWC
6 (N1a1a3). Lastly, the Berber individual (BMC 6a) from the 5th Cataract was assigned to
haplogroup T1. Taken together, all haplogroup assignments were superficially plausible and
very broadly speak to a blend of African and Near Eastern descent (Table 4.6). In later
haplogroup frequency surveys, all haplogroups were present in modern groups of Northeast
Africa in various proportions, with the exception of Near Eastern haplogroups (N, T, H) in the
southern Nilotic groups, Western Sudanese groups, and Pastoralists (Figures 4.14 and 4.15).

Table 4.4. Demographic and Archaeological information of Nubian individuals (ordered by date).

Sample Excavation Sex Sample Extraction Molecular

Age Est.α Dating
Name ID Est.α Material Method Sexingµ

30-40 Tooth SD, DD Cemetery 4

GHZ 2 GHZ-4-008.2 Female Female
years RM2 (UZH) 670-770 CE

Tooth SD, DD Likely Cemetery 1

GHZ 5 GHZ-1-010 ~45 years Male
LM2 (UZH) Male 670-770 CE

BMC 2009 Tooth SD, DD Meroitic

BMC 6a n/a NA Female
63-6a (T6) RPM1 (UZH) ca. 3rd Century CE

Bone SD, DD Meroitic

KWC 2 KWC T2 Adult None Male
R Petrous (UZH) 350 BCE-350 CE

Bone DD, B+DD Meroitic

KWC 6 KWC T6 Child NA Female
L Petrous (Oslo) 350 BCE-350 CE

111
 Bone B+DD Likely Napatan
TARP 9 TARP G9 Child NA
R Petrous (Oslo) Female 800-350 BCE
αMethods compiled by Buikstra & Ubelaker 1994, performed by archaeologists and/or anthropologists
µMittnik,et al. 2016
SD - Singe Digest, DD - Double Digest, B+DD - Bleach pretreatment + Double Digest;
UZH - University of Zürich; Oslo - University of Oslo
CE - Current Era, BCE - Before Common Era;
R/L - Right, Left; M - Molar; PM - Premolar;
GHZ – Ghazali, BMC - Berber Meroitic Cemetery, KWC - Kwieka Cemetery, TARP - Tinga Archaeological Rescue
Project; T - tomb; NA - data not available because individual was a subadult, not performed, or data not available; Est
- Estimation.

Table 4.5. Results of Molecular Sexing.

Sample R-Squared Value, confidence
Pipeline output/assessment Outcome
Name intervals (CI - 95%)

Sample should be assigned as

GHZ 2 Rx = 0.970, 95% CI: 0.904, 1.036 Female
Female

Sample consistent with an XY

GHZ 5 Rx = 0.615, 95% CI: 0.564, 0.667 Likely Male
karyotype, but not XX

Sample should be assigned as

BMC 6a Rx = 0.910, 95% CI: 0.872, 0.9488 Female
Female

KWC 2 Rx = 0.509, 95% CI: 0.470, 0.549 Sample should be assigned as Male Male

Sample should be assigned as

KWC 6 Rx = 1.035, 95% CI: 0.953, 1.117 Female
Female

Sample is consistent with an XX

TARP 9 Rx = 0.822, 95% CI: 0.766, 0.877 Likely Female
karyotype, but not XY

Table 4.6. Summary of consensus mitochondrial sequence information and haplogroup

assignments.
Number % No Overall Rank Mean
Sample Name Quality Haplogroup Contamination
of N's Call (Quality) Coverage
GHZ 2_merged Q20 462 2.8 L2a1+143 91.81% 13X 1.0%
GHZ 5_merged Q80 0 0.0 H2a 83.51% 450X 1.0%
BMC 6a_merged Q30 1813 10.9 T1 76.83 11X 2.0%
KWC 2_merged Q30 3 0.0 H2a 88.34% 71X 1.0%
KWC_6 merged Q30 519 3.1 N1a1a3 89.62% 17X 1.0%
TARP9B_oslo Q20 176 1.1 L0a1a 94.38% 11X 1.0%
Results of haplogroup assignments from Haplogrep 2 (Weissensteiner, et al. 2016), quality of consensus sequence
used, number of missing nucleotides (counts and as a percentage of whole human mitogenome). Overall quality rank
output from Haplogrep 2 as a confidence measure of the assignment, contamination percentages from Schmutzi
(Renaud, et al. 2015).
N indicates “no call” or “missing data” at nucleotide position

Haplogroup Frequencies Dataset Analysis

To put the mitogenomes of Ancient Nubians within context of the Nile Valley and the
wider Northeast region, the haplogroup frequencies dataset was assembled from published

112
sources outlined in Methods (Table 4.7). These data were visualized via two different means
(Figure 4.14, 4.15).

Table 4.7. Haplogroup Frequencies Grouped by Population or Subpopulations (in percentages).

Haplogroup Libyans Egyptians Nubians Beja Central Western S. Nilotes Pastoralists Ethiopians Kenyans

L0 1.5 2.9 15.7 6.3 1.9 20.7 19.2 4.7 7.3 19.9
L1 3.4 1.6 5.6 8.3 7.8 14.9 11.7 23.3 2.9 4.2
L5 0.0 2.4 4.6 2.1 2.9 2.3 18.1 0.0 2.9 2.8
L2 8.6 6.1 11.1 16.7 20.4 25.3 26.6 23.3 14.2 9.2
L4 1.1 0.0 2.8 2.1 0.0 8.1 7.5 0.0 0.0 0.0
L3 12.7 5.2 17.6 35.4 34.0 28.7 17.0 39.5 8.1 30.4
L/N/M* 0.4 7.4 7.4 0.0 0.0 0.0 0.0 0.0 17.7 23.4
M 3.4 7.3 8.3 4.2 3.9 0.0 0.0 0.0 15.4 4.2
D 0.0 0.0 0.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0
N 1.9 4.9 0.9 0.0 1.0 0.0 0.0 0.0 2.3 0.7
I 0.0 1.6 0.0 0.0 0.0 0.0 0.0 0.0 0.6 0.0
W 0.0 1.2 0.0 0.0 0.0 0.0 0.0 0.0 0.9 0.7
X 2.2 1.2 0.9 0.0 0.0 0.0 0.0 0.0 0.6 0.0
R0 6.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
HV 8.6 6.5 9.3 0.0 6.8 0.0 0.0 0.0 12.5 2.1
H 17.2 9.9 5.6 8.3 0.0 0.0 0.0 0.0 2.0 0.0
J 9.7 9.0 2.8 0.0 7.8 0.0 0.0 9.3 2.1 0.7
T 6.0 13.1 2.8 2.1 4.9 0.0 0.0 0.0 3.8 0.0
U 11.9 8.8 2.8 2.1 8.7 0.0 0.0 0.0 5.3 0.7
K 5.2 3.7 0.9 12.5 0.0 0.0 0.0 0.0 1.5 0.7
Other 0.0 6.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Total N 268 244 108 48 103 87 94 43 344 141
Haplogroup frequencies expressed as percentages for populations and subpopulations in the Northeast Africa region,
see Table 6 for ethnic groups included in geographic groups. Major Haplogroups (column 1) are collapsed for
simplicity. Individual totals are listed at the end of the table. L/N/M* refers to undifferentiated individuals with not
enough diagnostic variation sites to classify further. Grouping breakdown and references listed in Table 4.2. S;
Southern.

Analysis of Mitochondrial Genomes: Haplotype Profiles

Haplotype profiles are a useful way to visualize and quantify genetic differentiation and
homogeneity among populations (Brandt, et al. 2013, Schuenemann, et al. 2017). From these
haplogroup profiles, we observed highly similar signatures for southwestern ethnic groups of
Sudan that showed less diversity than those in the north-east part of the country. Western,

113
Pastoralists, and Nilotes shared similar compositions of the Africa-specific L lineages, in slightly
different proportions. Other subpopulations within Sudan showed much more haplotypic
variation, including lineages outside of the L macrogroup. In general, the L3 lineage was the
greatest proportion for a majority of populations or subpopulations of the Sudanese region,
followed by L2 (Table 4.7, Figure 4.14).
Haplotype frequencies were visualized as stacked columns to observe homogeneity
between populations in this region (Figure 4.15). Arranged roughly geographically, the African
signature increased in a step-wise fashion, with the other part of the profile being a Eurasia-
originating component. The African ancestry of Libyans and Egyptians was less than 30%.
Northern Sudanese, including the Nubian, Beja, and Central Arab ethnic groups have roughly
70% African ancestry. To be expected, the southern and south western Sudanese groups had
none or very little Eurasian gene flow. Ethiopians also had a large non-African component like
Egyptians and Libyans (ca. 70%), but this component was more intermediate (ca. 50%) when
compared to the northern Sudanese groups (ca. 30%). Kenyans had an approximately 10%
non-African component; however, 20% of this profile was not diagnostic to discern more specific
haplogroups (i.e. classified as L/N/M*). While the six Nubians did not constitute a population and
cannot be directly compared, these profiles placed the Ancient Nubians in context of the
northeast region.

114
Figure 4.14. Haplotype Diversity of Northeast Africa region coordinated with geographic
locations.

MtDNA haplogroup frequencies for modern populations of the Nile Valley and Eastern Africa, including Libya for
comparison. A total of 1480 individuals are represented in these charts. Sample sizes for the populations are as
follows: Libyans (n=268), Egyptians (n=244), Nubians (n=108), Beja (n=48), Central Sudanese (n=103), Western
Sudanese (n=87), South Nilotes (n=94), Sudanese Pastoralists (n=43), Ethiopians (n=344), Kenyans (n=141).
Haplogroup profiles for populations and subpopulations are depicted as pie charts coordinated with geographic
locations; pastoralist groups do not have a set geographic location and instead occupy a larger expanse due to
nomadic lifestyles. Sudanese subpopulations total 483 Individuals that is encompassed in the simplified pit chart to
the left to emphasize the abundance of macrohaplogroups (legend right).

115
Figure 4.15. Haplotype Frequencies of modern Northeast African populations and Sudanese
subpopulations.

MtDNA haplogroup frequencies for modern populations of the Nile Valley and Eastern Africa, including Libya, for
comparison (N=1480). Sample sizes for the populations are as follows: Libyans (N=268), Egyptians (N=244),
Nubians (N=108), Beja (N=48), Central Sudanese (N=103), Western Sudanese (N=87), South Nilotes (N=94),
Sudanese Pastoralists (N=43), Ethiopians (N=344), Kenyans (N=141). All populations combined: individuals
representing the North East African region around the Nile River Valley. Legend at right. Warm colors (red, orange,
yellow) describe Sub-Saharan content of populations and subpopulations. Cool colors describe other lineages: back
to Africa (M, N) and Eurasian lineages (H, etc.). N - number of individuals, counts.

Analysis of Mitochondrial Genomes: Haplotype PCA and Pairwise Differences

To further test genetic affinities and shared ancestry with extant African and Eurasian
populations, a Principle Component Analysis (PCA) based on haplogroup frequencies and
Multi-dimensional Scaling (MDS) of pairwise genetic distances using the FST statistic were
performed. This dataset includes the six whole mitogenomes of the Ancient Nubians captured
here, eight ancient groups, and 14 modern populations spanning Africa, the Middle East, and
Europe (Table 4.1). For PCA analysis, mitogenomes were transformed into macrohaplotype
frequencies and collapsed into large haplogroups based on geography given the role of
geography in genetic structure (Hassan 2009, Babiker, et al. 2011, Hollfelder, et al. 2017,
Dobon, et al. 2015). For FST or pairwise distance calculations, groups are specified in Table 4.1.
PCA is a standard tool used in genetic studies. It is a means to detect genetic structure,
or any pattern in the genetic makeup of individuals or subpopulations within a population

116
(Chakraborty 1993). For this analysis, haplotype lineages, derived from full MT sequence data,
were compared across 23 population groups. PC 1 characterized 42.39% of the variation, PC 2
characterized 17.46%, and PC 3 characterized 14.12%. Differences were observed among
populations based on geography as dispersed by the first component in a south to north cline
(left to right in Figure 4.16). The groups with the largest variation have the most influence on PC
1, namely South Africans and Europeans, which also were separated by the most physical
distance. Populations with the highest genetic affinity cluster together. This was evident with
Ancient and Modern Europeans and especially Near Easterners and Ancient Egyptians, which
was expected as per Schuenemann, et al. (2017). North Africans of the Maghreb (or
Northwestern Africa) also showed a close genetic signal to those populations geographically to
the north. The Ancient Nubians clustered with Modern Egyptians and Middle Easterners. The
Ancient Nubians and Modern Egyptians were not separated by PC 1, only by PC 2. Modern
Sudanese and East Africans (modern and ancient) were separate from Ancient Nubians
indicating genetic differences among these groups.
Generally, almost all African groups are on the left side of the plot, while Eurasian
populations are the right (Figure 4.16). The six Nubians have more affinity to African
populations. The only exception to this divide is the Ancient Egyptians. This result is consistent
with previously reported data: these Ancient Egyptian populations from the northern Fayum and
temporally from three different time periods showed high affinities to modern populations in the
Levant and Near East, as well as showed population continuity (Schuenemann, et al. 2017). In
general, PC 2 does not separate the groups, except for the South Africans (both ancient and
modern).
When only highlighting the ancient populations (Figure 4.17), again the role of
geography was prominent and the most important driving factor for variation among these
groups. While it was expected to observe the clustering of Ancient Egyptians with Near Eastern
farmers and Europeans, Ancient Nubians fell between East African populations and groups
centered around the Mediterranean. With considering modern populations only, Ancient
Nubians showed the most affinity to Modern Egyptians and Middle Easterners, and somewhat
to Southern Europeans and North Africans; in general, Mediterranean populations, rather than
modern African populations. Overall the Ancient Nubians show affinity toward both African and
European groups and much more so than Ancient Egyptians.

117
Figure 4.16. Principle Component Analysis based on haplotype frequencies, whole MT genomes.
(17.46% variance)

(42.39% variance)
PCA constructed in MitoBench. Legend on the right, with corresponding colors of groups listed in Table 4.1.

118
Figure 4.17. Principle Component Analysis based on haplotype frequencies, ancient groups
only.

PCA constructed in MitoBench. Modern populations in grey to highlight only ancient populations. Legend on the right,
with corresponding colors of groups listed in Table 4.1. Ancient Egyptians colored the same for simplicity. PC 1
characterizes 42.39% of the variance, PC 2 characterizes 17.46% of the variance.

Pairwise FST is a measure of population differentiation (Weir & Cockerham 1984,

Holsinger & Weir 2009). Much like a percentage, values of this statistic range from 0 to 1, with
zero indicating no genetic difference and one indicating complete genetic divergence. Whole MT
genome data was used to examine pairwise differences of all populations listed in Table 4.2. In
general, populations that were geographically close had lower pairwise FST. Values ranged from
-0.015 to 0.363 (Table 4.8). (Negative values were the result of the transformation of FST values
by Arlequin and likely the smaller population sizes given that the estimator requires large
population sizes (Weir & Cockerham 1984)). Pairwise FST was the greatest between South
Africans and modern Northern Europeans (FST = 0.305). The negative FST values are effective
not different from zero, we can therefore conclude that the Ancient Nubians had the closest
affinity (or lowest FST) with modern Egyptians (FST = -0.015), with modern Ethiopians (FST = -
0.001), and Modern East Africans (FST = -0.011). Following these were close affiliations with
modern Central Africans (FST = 0.0041), modern Middle Easterners (FST = 0.012), and modern
Western Africans (FST = 0.013). Slightly less similarity was shown to modern Sudanese (FST =
0.037) and North Africans (FST = 0.049) as well as ancient East Africans (FST = 0.029). Ancient
Nubians were the least similar to Northern Europeans (FST = 0.305) and modern Southern

119
Africans (FST = 0.363). Overall, the values obtained from the FST analysis support the clustering
shown in the PCA and show the differentiation of South African groups.

Table 4.8. FST Values for comparisons to Ancient Nubians.

Population FST
EgyptModern -0.015*
EastAfricaModern -0.011*
EthiopiaModern -0.001*
CentralAfricaModern 0.004
MiddleEastModern 0.012
WestAfricaModern 0.013
AncientEastAfrican 0.029
SudanModern 0.037
NorthAfricaModern 0.049
AncientEgyptiansRoman 0.065
AncientEuropean 0.072
AncientEgyptiansPPP 0.087
AncientEgyptiansPtoP 0.100
WesternEuropeModern 0.102
CentralEuropeModern 0.105
AncientNorthAfrican 0.113
AncientNearEastFarmers 0.120
AncientSouthAfricans 0.127
SardiniaModern 0.151
SouthernEuropeModern 0.163
NorthernEuropeModern 0.305
SouthernAfricaModern 0.363
* Negative values were the result of the transformation of FST values by Arlequin and likely the smaller population
sizes given that the estimator requires large population sizes (Weir & Cockerham 1984). Negative values were
changed to “zero” for the heat map and ‘0.0001’ for the Multi-Dimensional Scaling plot.

FST values were visualized using an MDS plot (Figure 4.18). Like the PCA analysis, this
dimension-reduction is a means to graphically represent the similarities (clustering) and
differences (distances) captured in the FST values. Distance calculations derived from genetic
variation across the entire MT genome were used for analysis. PC 1 explained 62.91% of the
differences, while PC 2 characterized the next 19.35% variation. Ignoring the clear outliers,
Northern Europeans and Southern Africans, two clusters were visible (Figure 4.18). One
included all (other) Europeans, North Africans, Middle East or Near Easterners, and all Ancient
Egyptians. The second cluster included Modern Egyptians and Africans south of North Africa
(East, Central, Western) (Figure 4.18). Ancient Nubians were again between East Africans and
the Egyptians. Focusing on the right cluster, these groups (modern Egyptians, Ethiopians,

120
Sudanese, Central, West, and East Africans, Ancient Nubians, and East Africans) vary amongst
one another with Dimension 2 more so than the cluster with Eurasian populations, North
Africans, and Ancient Egyptians. Within this cluster, Nubians (and Modern Egyptians) showed
the most differences between Ancient East Africans and Modern Sudanese. Like the PCA plot,
PC 1 served as a means to classify groups by geography, where this dimension separated
African populations groups from Eurasian groups, with the exception of Modern and Ancient
North Africans and Ancient Egyptians. Overall, Ancient Nubians clustered most closely with
Modern Egyptians, Modern East Africans, and Modern Ethiopians.

121
Figure 4.18. Multi-Dimensional Scaling Plot mapping FST pairwise differences among Ancient and Modern African and Eurasian
populations.

Population labels simplified to geographic region. Colors of light and dark shades correspond with these simplified regions for better visualization.

122
DISCUSSION

In the ancient past, northern Sudan was home to multiple powerful and dynamic Nubian
kingdoms (i.e. Kermites, Kushites, Christians). We know much of this time from various fields of
study including archaeological research, historical and ethnographic records, and
anthropological research, but the population genomics of this time has yet to be investigated.
Modern genetic data of Nubian Sudan tells two stories: possible links exist to an ancestral past
with East Africa and the effects of the Arab expansion dominate the genetic profiles of modern
individuals. What was the genetic landscape of Nubians before this major event?
Archaeological, historical, and anthropological work has demonstrated that Nubians had a
dynamic history in the Nile River Valley with notable demographic shifts – colonialism,
entanglement, conquest, expansion – that likely had a quantifiable impact on their genetic
structure. To explore this and begin to join these lines of evidence, paleogenomic methods were
used to sample mtDNA from ancient Nubians, providing a new line of data to understand
Sudan’s ancient past and begin to define their biological ancestry.
ADNA was extracted from archaeological samples from Sudan dating (c14 or
contextually) to before the Arab expansion. Sequence data from targeted MT genome capture
was used to characterize the haplogroups of the ancient Nubians. The full mitogenomes
obtained here were the first ancient Nubian genomes to date. Only one other study has
published genomes of Nile River Valley individuals (Schuenemann, et al. 2017). The analyses
performed here were the first to contextualize Ancient Nubians within the African and Eurasian
landscape. In general, Nubians are more closely affiliated with African populations than Ancient
Egyptians.
Two individuals were assigned to have African ancestry (L haplogroups) and four were
assigned to non-African lineages, namely those originating in the Near East (H, N. T
haplogroups). These individuals derive from varying archaeological contexts and therefore
different cultural horizons in Nubian history. The individual deepest in time, the Napatan period,
is likely the oldest Nubian to be sequenced yet and helps address our questions about Nubian
genetic structure in the past. Individual TARP 9B originated from deep within the Nubian lands
and was assigned the haplogroup L0a, which demonstrates the presence of this haplogroup
during the Napatan (ca. 800-300 BCE) time period. The macrogroup L0 is the most ancient
lineage of mtDNA and its subgroup L0a1 dates to roughly 25,000 years ago with a probable
origin in Southeast Africa (Rito, et al. 2013, Soares, et al. 2009) (Figure 1.5). The other
individual with African ancestry is GHZ 2, who was dated much later in the early Christian period

123
and was assigned L2. This group is a very common macrohaplogroup within Sudan, being the
second most abundant following L3 (Hassan 2009). Kulubnarti individuals from ancient Lower
Nubia have also been classified with this haplogroup (Sirak, et al. 2016). Both of African
lineages were within the range of Sub-Saharan African variability and are extremely rare outside
Africa (Salas, et al. 2004). Beyond Sudan, the macrogroup L2a likely originated in central or
western regions of Africa (Silva, et al. 2015) and its subgroup L2a1, which dates to roughly
20,000 years ago, is common all over the continent (Salas, et al. 2002, Soares, et al. 2009).
Generally, within the ancient Nile Valley, these two L assignments of African lineages have not
been detected in Egypt (Schuenemann, et al. 2017), but have been in ancient Tanzania and
Malawi (Skoglund, et al. 2017, Prendergast et al. 2019). These over overall rare, and more
sampling of ancient individuals inside Africa will offer more context. In the meantime, these data
build our narrative of the African component of Nubian genetic background and specifically
characterize the unique background of inhabitants in this region.
The four other individuals which we obtained full mitogenomes for were assigned to
lineages which have origins outside Africa (H, N, T) and demonstrate the presence of these
haplogroups within the ancient Nubian genetic landscape. Individuals GHZ 5 (Christian era
monk) and KWC 2 (Meroitic individual from 5th Cataract) were both assigned H2a haplogroups.
The H macrogroup is the most common one in Europe and Eurasia but has Near Eastern
origins with a modern distribution around the Middle East, particularly in the Caucasus region
and the Arabian Peninsula (Pereira, et al. 2005, Achilli, et al. 2004, Roostalu, et al. 2007). While
the H or HV lineages are common in Ancient Egypt, none have the H2a assignment. The
appearance of this haplotype deep in Sudan suggests migrations from the Near East or Levant
regions and it is especially plausible that one individual, GHZ 5, was buried at a monastery
where people may have traveled for pilgrimages.
The other individual from Kwieka, KWC 6, was assigned to the haplogroup N1a1a3. The
macrogroup N1a originated in the Near East and is common amongst Neolithic Europeans
(Haak, et al. 2005). N1a is distributed throughout Eurasia and northeastern Africa, especially
with Afro-Asiatic speakers (Hassan 2009). This haplogroup also originated in the Near East and
is a subgroup of the R and JT lineages (Figure 1.5). This specific haplogroup was present in
one Ancient Egyptian dating to around the same period and another dating to dynastic eras
(Schuenemann, et al. 2017, Neukamm in prep). Within modern genetic context of Sudan, this
haplogroup is prevalent among modern Afro-Asiatic speakers (i.e. those groups of Arab
descent), while being virtually absent in Nilo-Saharan speakers (Bekanda, et al. 2015). This is
also the case for the Berber 6a individual (BMC 6a) who was assigned as belonging to

124
haplogroup T1. This group has origins in the Near East, but more so in West Eurasia
(Fernandez, et al. 2015). Although these are only two individuals, this hints at dating the
presence of a Eurasian component well before the Arab expansion and reaffirms the idea that
the Nile Valley was a corridor of movement in antiquity (Krings, et al. 1999, Lalueza Fox 1997).
Additionally, this challenges the hypothesis that the genetic structure of Ancient Nubians
resembled Nilotic groups, which have no Eurasian signatures within autosomal or mitochondrial
genomes (Hollfelder, et al. 2017). However, this is but one individual cannot speak for a
population and should certainly be carefully interrupted since multiple markers are in use when
investigating the genetic structure of Sudan, which have their own biases and advantages. More
sampling and collecting autosomal data are of the utmost importance moving forward.
All haplotypes that were classified in the Ancient Nubians were found in modern Nubians
(Table 4.7), so population continuity cannot be ruled out. Specifically, the Berber 6a individual
was assigned as T1; this lineage was found in all groups except those groups that represent
ancestral East Africans. The TARP 9 and GHZ 2 individuals were assigned African L lineages,
shared by all populations in this region, while all others were non-African. Furthermore, these
two individuals were the separated by the most time, likely more than one thousand years. The
haplogroup H2a was assigned to two individuals from two sites separated by both time and
space. This haplotype was present in Egyptians, Libyans, Ethiopians, northern Sudanese
groups, but not the Central Arab group. The N lineage haplogroup assigned to the Lower
Nubian KWC 6 individual was much rarer within this context but was present in extant Nubians.
This lineage was most common in the Egyptians. Moving forward, more individuals are needed
to formally test continuity between past and extant populations.
When grouping these individuals as a very approximate population, FST calculations and
PCA showed that Nubians were genetically close with modern Egyptians, followed by modern
Middle Easterners and modern East African groups. This close affinity with Egyptians also has
been demonstrated with modern Northern Sudanese genetic data and has been hypothesized
to extend back in time (Babiker, et al. 2011, Krings, et al. 1999). Results presented here do not
indicate population continuity, but also do not rule it out. Further testing of this hypothesis is
warranted. The PCA also showed Ancient Nubians were not as closely affiliated with their
descendants, but showed clear signs of gene flow during antiquity as indicated by Lalueza Fox
(1997), Dobon, et al. (2015), and Babiker, et al. (2011). Overall, Nubians were characterized by
a significant African component within their profile (more so than Ancient Egyptians) with
enough gene flow from the Middle East and Europe to make them distinct from other African
populations. These signals suggest more genetic structure than hypothesized by Hollfelder, et

125
al. (2017) when surveying modern Nubians in northern Sudan. The haplotype assignments,
even only six, show vast differentiation from the Southern Nilotes, which were genetically
isolated and hypothesized as being ancestral East African, as per Hollfelder, et al. (2017). The
variety of haplogroups suggests more admixture due to the presence of non-African lineages
while Southern Nilotes have none. These data are consistent with the notion Nubians have
African and Mediterranean or Eurasian ancestry, which is reflected in anthropological,
archaeological, and ethnographic observations and confirms this dimension of Nubian character
(Armelagos & Van Gerven 2017).
The analyses of the mitogenomes reinforce the influential role geography plays in
shaping the genetic landscape of Africa. The most variance captured in PC 1 differentiated
African groups in a south to north cline in both dimension-reducing analyses. This suggests that
geography is a sound indicator of mitochondrial structure. This finding reinforces the idea that
population geneticists should consider geography when investigating genetic diversity across
wide areas (Peter, et al. 2018).
Relating to the methodological specifics of this study, we performed an enrichment step
in order to retrieve full ancient mitogenomes. For many individuals, initial shotgun sequencing of
reads mapping to the mitochondrion were as low as zero but increased to several thousand
post-enrichment. Merged consensus sequences produced deeper coverage of the mitogenome
and provided subsequent haplogroup assignments with more confidence. For example, merged
reads for KWC 6 assigned the haplogroup N1a1a3 with 89.62% confidence, which was higher
than the separate reads, and with 2 times less N’s (or no calls/data) for diagnostic loci. Based
on these findings, we advocate that enrichment be implemented for all future work with Nubian
samples. In addition, the use of two separate, parallel extracts with enrichment was required for
full reconstruction of five out of the six genomes obtained. Each extract contributed unique
sequence reads that increased the coverage of the MT genome. In other words, each extract
exhibited high complexity and little to no overlap. For example, the mean coverage of the
merged reads for KWC 6 was 10X and 7X for the non-bleached and bleached powder,
respectively, and following merger, the coverage increased to 17X, demonstrating the
uniqueness of the libraries from each extract (Table 4.3). However, it should be noted that two
extracts require more starting tissue and therefore multiple extractions from the same individual
should only be performed after careful consideration of the benefits. Libraries should be
immortalized, especially those not enriched, to extend the use of these extracts to the fullest
capacity (i.e. to be used for whole genome or SNP-enrichment assays).

126
 Several limitations narrowed the scope of this study. First, our final sample size was
small. Out of 43 individuals trialed, six individuals had full genomes for analysis (14% success
rate). Furthermore, this small number of individuals does not create a homogeneous
“population” since they were excavated from four different field sites with unique archaeological
contexts and span up to 1,000 years in time (Figure 1.5, Table 4.4). These analyses would
greatly benefit from a larger dataset with each archaeological context considered independently
(e.g. Brandt, et al. 2013). Second, our analysis lacked resolution of the Sudanese mitochondrial
landscape. Despite robust research projects focused on Sudanese population genetics,
mitochondrial sequence data has not generated. Only one seminal dataset, Krings, et al. (1999),
is available. However, it only characterizes two regions of the country (i.e. Northern Nubians
and Southern Nilotes) and is limited to HVR1 sequence data. In the NGS age, trends toward
larger datasets mean full mitogenomes will likely be available in the future. Fortunately, Dr.
Hisham Yousif Hassan (from Hassan (2009), Bahrain Defense Force Hospital) has full
mitogenome sequences for 15 known ethnic groups (used in Dobon, et al. 2011) within the
country. These data will be published in the next year and will bring meaningful resolution to the
mitochondrial landscape of modern Sudan for future work. Lastly, mtDNA analyses only
represent the maternal landscape. Additional loci would be informative for understanding the
history of paternal lineages, sex-biased gene flow, and genetic adaptation. Nevertheless,
mtDNA can address cross-disciplinary questions from archaeologists and anthropologists about
ancestry and migrations and should be included whenever possible.
Future directions of this work include further sampling of individuals across the Nubian
landscape, possibly expanding into Lower Nubia, as well as augmenting the sample size of the
populations that were successful and previously demonstrated very good preservation of
sample tissue, i.e. Ghazali and Kwieka individuals. Data from the nuclear genome would be
very informative for understanding Nubian paleodemography beyond the maternally inherited
mtDNA. Shotgun sequencing results of the samples extracted here showed preservation of
ancient nuclear DNA, even in the absence of mitochondrial DNA. This suggests that nuclear
enrichment would be plausible to explore the Nubian autosomal genome. Further
anthropological context of the successful individuals and their archaeological sites should also
be collected. Prendergast, et al. (2019) serves as a notable publication worth emulating as it
provides equal footing of archaeological context with genetic data within a manuscript.
This work demonstrated the viability of paleogenomic work using archaeological material
from Middle Nile region. Even with mitogenomes from only six individuals, the dynamic ancestry
of Nubia is reflected in their genetic past and begins to build the narrative centered on Nubian

127
genetic structure. Given the long tradition of the research in this region of the Nile Valley by
other disciplines, aDNA will supplement and advance our knowledge of Nubian history. Overall,
paleogenomic work offers a unique perspective to join with archaeological and historical data
and can be especially helpful when these sources are sparse, contested, or unavailable.

CONCLUSION

Recent advances ancient DNA methodologies has opened the continent of Africa to the
paleogenomic revolution. Mitochondrial DNA from ancient Nubian sources showed that these
individuals have gene flow from the Eurasia, but have close genetic ties with Africans. This is
consistent with the population interactions and demographic shifts attested to in archaeological
and anthropological studies as well as historical texts. These data are the first ancient
mitochondrial genomes to be reported and offer a new facet to understanding Nubian ancestry,
specifically the ties to Near Eastern haplogroups but also the ancestral African ones. This
blending can be further explored with more sampling and better comparative datasets built with
mitochondrial DNA.

128
 CHAPTER 5: Conclusion

This dissertation synthesized Nubian paleogenomic data with skeletal bioarchaeological

evidence to understand the population history of the El-Kurru population, dating to the end of
the Christian era and beginning of Arab expansion in Upper Nubia. This multidisciplinary
approach included paleogenomic analyses of the ancient individuals set within an
archaeological and bioarchaeological framework. Since modern genetic data does not shed light
on the genetic landscape prior to the Arab expansion in Sudan and there is a lack of ancient
Nubian genetic datasets for comparison, seven additional archaeological sites were sampled to
define Nubian ancestry in this region. Additionally, new paleogenomic methodologies were
optimized using newly or previously developed techniques because African samples are difficult
to work with on account of poor DNA preservation and thermal degradation. While the ancient
DNA extraction was not successful for the El-Kurru individuals, it was with individuals from other
archaeological sites. These six individuals help build what we know about Nubian genetics in
the ancient past and establish the viability for continued work with material from the Middle Nile
region of Sudan.
With a viable method to extract aDNA, the two part approach can be implemented in
various other regions of the Nile Valley, namely pairing paleogenomic data with
bioarchaeological analyses of individuals and populations. Mortuary and bioarchaeological data
can add a meaningful viewpoint on the daily lives and environments in which these individuals
lived. As was demonstrated with the Neolithic Transition example, the genetic perspective can
be very useful when in context with multiple layers of evidence. For the population from El-
Kurru, a group likely at the brink of transition to Islam as it expanded to the south, their Christian
religious affiliation was expressed in death with typical body treatment and grave construction.
Their hardships with malnutrition, disease, and/or other stressors was recorded in their remains
and told a story that despite a heavy disease burden on the younger individuals, adults were still
living to an old age. Their burial practices suggested that this time of transition to Islam had not
arrived, or even if it had and the inhabitants kept to their Christian identities, the health profiles
as a community speak to little to no negative impact of this transition. More populations from this

129
area and time period would supplement this narrative and give greater insight into life during this
transitional time period, in addition to successful paleogenomic anlayses.
This work defined a workable protocol that should be shared with other colleagues
working with African materials or those from similar arid climates. There is also room for further
improvements. To build full mitochondrial genomes at 10X coverage or more, either or both
pretreatments of sample powder were enough to prepare the samples for enrichment, but two
separate extracts were necessary. This is an important area to improve upon to be less
consuming of archaeological material. For those samples from tumuli burials which performed
incredibly poorly, more trials could be conducted to obtain aDNA, as these individuals represent
an important transitional time period (the Post-Meroitic, ca. 350-650 CE) in Nubian history and
the skeletal remains are accompanied by extensive bioarchaeological data. This would
represent another area to implement the two-part research design. However, we would
generally caution restraint when sampling material for a second time as it is destructive and it is
important to leave some tissue unsampled for future developments within the field. Given the
blistering pace at which paleogenomics progresses, it may not take much time for improved
techniques to yield robust aDNA yields from samples that yield minimal to no aDNA with current
methods.
With an established methodology, an expansion of this project is achievable.
Mitochondrial DNA was the first survey step and initial successes open more options for further
research in this field, including autosomal DNA retrieval for more impactful comparisons with
available datasets from modern Sudanese and other ancient Africans (Figure 1.8). Expansion is
also possible during the laboratory phase; for example, the inclusion of automated steps during
the enrichment part would make the process more high-throughput since multiple extracts are
required to build the mitogenomes. Despite the small number of successful genomes obtained,
the initial results, where two Nubians have African ancestry and four have non-African ancestry,
supported the role of the Nile River Valley as a corridor. Compared to modern mitochondrial
DNA, these Nubians are most genetically close to modern Egyptians, Middle Easterners, and
East Africans. They are less affiliated to modern Sudanese, Ancient Egyptians, and Ancient
East Africans. This lack of genetic closeness to the modern Sudanese may be a product of
sampling, given the high diversity within Sudan which is heavily driven by geography. More
informative comparative populations with provenience are required to better contextualize the
results and also those concerning the closeness to modern Egyptians. Lastly, additional Nile
Valley populations from Egypt and Ethiopia would increase this resolution

130
 To address questions concerning ancestry and characterizing of the genetic make-up of
Nubians before the Arab expansion, development of more homogeneous populations for each
site is critical for further analyses. This would involve obtaining genetic sequences from more
individuals at each site, all of which represent distinct time periods and cultural contexts. Given
the vast number of collections previously excavated and those yet to be discovered, building a
larger sampling base is possible and will further advance the foundational research described
here. It will be especially imperative to augment those individuals who are temporally, spatially,
or contextually linked (e.g. individuals from El-Zuma, El-Detti, and Tanqasi which are close in
proximity and date to the same Post-Meroitic period). In general, deeper sampling will create
valuable resolution for work in this area.
As we expand our knowledge of the genetics of Upper Nubia, this narrative would
specifically benefit from a contribution of sample populations from Lower Nubia, or the region
between modern Aswan and the 2nd Cataract. This region was significant as the buffer-zone
between Upper Nubia, where the capitals of Kush, Napata and later Meroe, are located, and
southern ancient Egyptians. During dynastic time periods, this region fluctuated between
peaceful and forceful Egyptian occupation, often to protect the flow of trade goods (Flammini
2008). During the Roman period, this area had an ever-shifting frontier zone, fortified or not, but
continued to be an interface for populations (Török 2008). Later, this region played a crucial role
in keeping the Arab forces at bay beginning in the 6th century CE through to when the Arab
forces took control of northern Sudan. Given this long history of interaction, the contribution of
samples from this region is of great importance when mapping the genetic landscape of Nubia
and widens the scope to include the greater Nile River Valley.

131
 Appendices

Appendix A - Supplementary information relating to methods:

1. Stainless steel mortar and pestle (or “Bone Crusher”) cleaning – after Oslo Lab protocols
- After each use, clean each piece (mortar, pestle, sleeve) with dish detergent (crystal
form for more abrasiveness) and a tooth brush to clear away bone dust
- Rise all pieces with tap water until no bubbles remain, usually 3x
- Place all pieces, laying flat, in a small tray or bin. It is helpful to reassemble the
mortar and sleeve for bleach bath in next step
- Cover with 2X bleach solution, making sure all used surfaces are completely covered
2. Sodium Hypochlorite solution is typically available at 14X concentration and should be
diluted to 2X, i.e. 100mL 14X bleach diluted with 700mL Milli-Q water (UV irradiated)
- Incubate for at least 15 minutes to disinfect
- Rinse thoroughly with Milli-Q water (UV irradiated)
- Rinse with 20% Ethanol solution and dry with lab tissues
- UV treat for at least 10 minutes before next use

132
Appendix B - Analysis / MultiFastQC of negative control samples
As outlined in the methodology, negative blanks were carried through all parts of the
protocol including mitochondrial enrichment and in both laboratories. In total, there were 15
blanks from all enrichment trials: three from Oslo, three from extractions via single digest
method, three from extractions via double digest method, six blanks from the library preparation
step. Most libraries were completely empty, while some had few reads that mapped the human
reference sequence. Those with read counts ranged from 3 to 259 (Figure S1). The only sample
with enough reads to observe typical damage patterns of ancient material was Library Prep
Blank 6 that had an average MT coverage of 0.0057. These reads showed exactly zero percent
damage suggesting no cross-contamination of the ancient samples during lab work.

Figure S1. Sequence counts histogram of Negative Blank libraries with more than zero reads.

MultiQC Analysis (Ewels, et al. 2016).

133
 References

Abdallah, M. S. (2018). Preliminary report on the excavations in Kwieka cemetery (KWC) – seasons 2016
and 2018. Sudan & Nubia, 22, 142-149.
Abu-Amero, K. K., Larruga, J. M., Cabrera, V. M., & González, A. M. (2008). Mitochondrial DNA structure
in the Arabian Peninsula. BMC Evolutionary Biology, 8(1), 45.
Achilli, A., Rengo, C., Magri, C., Battaglia, V., Olivieri, A., Scozzari, R., Cruciani, F., Zeviani, M., Briem,
E., Carelli, P., Dugoujon, J., Roostalu, U., Loogväli, E., Kivislid, T., Brandelt, H., Richards, M., Villems.,
R., Santachiara-Benerecetti, A.S., Semino, O., & Torroni, A., (2004). The molecular dissection of mtDNA
haplogroup H confirms that the Franco-Cantabrian glacial refuge was a major source for the European
gene pool. The American Journal of Human Genetics, 75(5), 910-918.
Adams, W. Y. (1967). Continuity and change in Nubian cultural history. Sudan Notes and Records, 48, 1-
32.
Adams, W. Y. (1977). Nubia: corridor to Africa. 1st edition. London: Allen Lane.
Adams, W. Y. (1984). The first colonial empire: Egypt in Nubia, 3200–1200 BC. Comparative Studies in
Society and History, 26(1), 36-71.
Afonso, C., Alshamali, F., Pereira, J. B., Fernandes, V., Costa, M., & Pereira, L. (2008). MtDNA diversity
in Sudan (East Africa). Forensic Science International: Genetics Supplement Series, 1(1), 257-258.
Allentoft, M. E., Collins, M., Harker, D., Haile, J., Oskam, C. L., Hale, M. L., Campos, P.F., Samaniego,
J.A., Gilbert, M.T., Willerslev, E., Zhang, G., Scofield, R.P., Holdaway, R.N., & Bruce, M. (2012). The half-
life of DNA in bone: measuring decay kinetics in 158 dated fossils. Proceedings of the Royal Society B:
Biological Sciences, 279(1748), 4724-4733.
AlQahtani S. J. (2008). Atlas of tooth development and eruption. Barts and the London School of
Medicine and Dentistry. London, Queen Mary University of London. MClinDent.
Ammerman, A. J., & Cavalli-Sforza, L. L. (1971). Measuring the rate of spread of early farming in Europe.
Man, 674-688.
Amorim, C. E. G., Vai, S., Posth, C., Modi, A., Koncz, I., Hakenbeck, S., La Rocca, M.C., Mende, B.,
Pohl, W., Baricco, L. P., Bedini, E., Francalacci, P., Giostra, C., Vida, T., Winger, D., von Freeden, U.,
Ghirotto, S., Lari, M., Barbujani, G., Krause, J., Caramelli, D., Geary, P., & Veeramah, K.R. (2018).
Understanding 6th-century barbarian social organization and migration through paleogenomics. Nature
Communications, 9(1), 3547.
Andrews, R. M., Kubacka, I., Chinnery, P. F., Lightowlers, R. N., Turnbull, D. M., & Howell, N. (1999).
Reanalysis and revision of the Cambridge reference sequence for human mitochondrial DNA. Nature
Genetics, 23(2), 147.
Antoine, D. (2018). Human Remains Recording Sheet. Institute for Bioarchaeology, British Museum,
London.
Armelagos, G. J. (1969). Disease in ancient Nubia. Science, 163(3864), 255-259.
Armelagos, G. J., & Van Gerven, D. P. (2017). Life and Death on the Nile: A Bioethnography of Three
Ancient Nubian Communities. Gainesville: University Press of Florida.
Aronsen, G. P., Fehren-Schmitz, L., Krigbaum, J., Kamenov, G. D., Conlogue, G. J., Warinner, C., Ozga,
A.T., Sankaranarayanan, K., Griego, A., DeLuca, D.W., Eckels, H.T., Byczkiewicz, R.K., Grgurich, T.,

134
Pelletier, N.A., Brownlee, S.A., Marichal, A., Williamson, K., Tonoike, Y., & Bellatoni, N.F. (2019). "The
dead shall be raised": Multidisciplinary analysis of human skeletons reveals complexity in 19th century
immigrant socioeconomic history and identity in New Haven, Connecticut. PloS one, 14(9), e0219279.
Aufderheide, A. C., Rodríguez-Martín, C., & Langsjoen, O. (1998). The Cambridge encyclopedia of
human paleopathology (Vol. 478). Cambridge: Cambridge University Press.
Babiker, H. M., Schlebusch, C. M., Hassan, H. Y., & Jakobsson, M. (2011). Genetic variation and
population structure of Sudanese populations as indicated by 15 Identifier sequence-tagged repeat (STR)
loci. Investigative Genetics, 2(1), 12.
Baker, B. J. (2016). Biocultural investigations of ancient Nubia. In Zuckerman, M.K. & Martin, D.L. (Eds.)
New Directions in Biocultural Anthropology, (p. 181-199). Hoboken: Wiley Blackwell.

Baker, B. J., & Judd, M. (2012). Development of paleopathology in the Nile valley. In Buikstra, J &
Roberts, C. (Eds.) The Global History of Paleopathology, (p. 209-34). Oxford: Oxford University Press.

Baker, W., van den Broek, A., Camon, E., Hingamp, P., Sterk, P., Stoesser, G., & Tuli, M. A. (2000). The
EMBL nucleotide sequence database. Nucleic Acids Research, 28(1), 19-23.
Bamshad, M. J., Watkins, W. S., Dixon, M. E., Jorde, L. B., Rao, B. B., Naidu, J. M., Prasad, B.V.,
Rasanayagam, A. & Hammer, M. F. (1998). Female gene flow stratifies Hindu castes. Nature, 395(6703),
651-52.
Barta, J. L., Monroe, C., & Kemp, B. M. (2013). Further evaluation of the efficacy of contamination
removal from bone surfaces. Forensic science international, 231(1-3), 340-348.
Basha, W. A., Chamberlain, A. T., Zaki, M. E., Kandeel, W. A., & Fares, N. H. (2016). Diet reconstruction
through stable isotope analysis of ancient mummified soft tissues from Kulubnarti (Sudanese Nubia).
Journal of Archaeological Science: Reports, 5, 71-79.
Bashir, M. (2010). A recently discovered Meroitic cemetery at Berber, River Nile State, Sudan.
Preliminary report. Sudan & Nubia, 14, 69-74.
Bashir, M. (2013). A Third Season of Rescue Excavations in the Meroitic Cemetery at Berber, October
2012: Preliminary Report. Sudan & Nubia, 17, 90-100.
Bashir, M. and R. David. G. (2015). The Meroitic Cemetery at Berber. Recent Fieldwork and Discussion
on Internal Chronology. Sudan & Nubia, 19, 97-105.
Behar, D. M., Van Oven, M., Rosset, S., Metspalu, M., Loogväli, E. L., Silva, N. M., Kivisild, T., Torroni,
A., & Villems, R. (2012). A “Copernican” reassessment of the human mitochondrial DNA tree from its
root. The American Journal of Human Genetics, 90(4), 675-684.
Bekada, A., Arauna, L. R., Deba, T., Calafell, F., Benhamamouch, S., & Comas, D. (2015). Genetic
heterogeneity in Algerian human populations. PloS one, 10(9), e0138453.
Benson, D. A., Cavanaugh, M., Clark, K., Karsch-Mizrachi, I., Ostell, J., Pruitt, K. D., & Sayers, E. W.
(2018). GenBank. Nucleic Acids Research, 46(D1), D41-D47.
Berkovitz, B. K., Holland, G. R., & Moxham, B. J. (2017). Oral Anatomy, Histology and Embryology E-
Book. Elsevier Health Sciences.
Bietak, M. (1987). “The C-Group and Pan-Grave culture in Nubia.” In T Hägg (Ed.) Nubian Culture Past
and Present. Stockholm: Almqvist and Wiksell.
Binder, M. (2014). Health and Diet in Upper Nubia through Climate and Political Change-A
bioarchaeological investigation of health and living conditions at ancient Amara West between 1300 and
800BC (Doctoral dissertation, Durham University).
Binder, M. (2019). The Role of Physical Anthropology in Nubian Archaeology. In Raue, D. (Ed.)
Handbook of Ancient Nubia (p. 103-127). Berlin: de Gruyter.

135
Boessenkool, S., Hanghøj, K., Nistelberger, H. M., Der Sarkissian, C., Gondek, A. T., Orlando, L., Barrett,
J.H., & Star, B. (2017). Combining bleach and mild predigestion improves ancient DNA recovery from
bones. Molecular Ecology Resources, 17(4), 742-751.
Bonnet, C. (1986). Kerma, Territoire et Metropole, Cairo:IFAO.
Bonnet, C. (ed.) (1990). Kerma, Royaume de Nubie, Geneva: Musee d’Art et d’Histoire.
Borsch, S. (2005). The Black Death in Egypt and England. University of Texas Press. Retrieved from
http://www.jstor.org/stable/10.7560/706170
Brandstätter, A., Peterson, C.T., Irwin, J.A., Mpoke, S., Koech, D.K., Parson, W., & Parsons, T.J. (2004).
Mitochondrial DNA control region sequences from Nairobi (Kenya): inferring phylogenetic parameters for
the establishment of a forensic database. International Journal of Legal Medicine, 118:294-306.
Brandt, G., Haak, W., Adler, C. J., Roth, C., Szécsényi-Nagy, A., Karimnia, S., Möller-Rieker, S., Meller,
H., Gansleier, R., Friedrich, S., Dresely, V., Nicklisch, N., Pickrell, J., Sirocko, F., Reich, D., Cooper, A.,
Alt, K.W., & the Genenographic Consortium. (2013). Ancient DNA reveals key stages in the formation of
central European mitochondrial genetic diversity. Science, 342(6155), 257.
Brickley, M. B. (2018). Cribra orbitalia and porotic hyperostosis: A biological approach to diagnosis.
American Journal of Physical Anthropology, 167(4), 896-902.
Briggs, A. W., Stenzel, U., Johnson, P. L., Green, R. E., Kelso, J., Prüfer, K., Meyer, M., Krause, J.,
Ronan, M.T., Lachmann, M., & Pääbo, S. (2007). Patterns of damage in genomic DNA sequences from a
Neandertal. Proceedings of the National Academy of Sciences, 104(37), 14616-14621.
Brooks, S., & Suchey, J. M. (1990). Skeletal age determination based on the os pubis: a comparison of
the Acsádi-Nemeskéri and Suchey-Brooks methods. Human evolution, 5(3), 227-238.
Brothwell, D. (1967). The bio-cultural background to disease. In Brothwell, D & Sandison, A. (Eds.)
Diseases in Antiquity (p. 45-68). Springfield: Charles C. Thomas.

Buikstra, J. E. & Ubelaker, D.H. (1994). Standards for Data Collection from Human Skeletal
Remains. Arkansas Archaeological Survey, Research Seminar Series 44. Arkansas Archaeological
Survey: Fayetteville, AR.
Buikstra, J. E. (Ed.). (2019). Ortner's Identification of Pathological Conditions in Human Skeletal Remains.
Academic Press.
Bushara, M., Abdalah, M.S., and Bashir, M.S. (2017). Between Napata and Meroe: a newly discovered
cemetery at Enapis (TARP) in the Middle Nile Region. Sudan & Nubia, 21, 128-133.
Buzon, M. R. (2006). Health of the non‐elites at Tombos: Nutritional and disease stress in New Kingdom
Nubia. American Journal of Physical Anthropology: The Official Publication of the American Association
of Physical Anthropologists, 130(1), 26-37.
Buzon, M. R. (2008). A bioarchaeological perspective on Egyptian colonialism in Nubia during the New
Kingdom. The Journal of Egyptian Archaeology, 94(1), 165-182.
Buzon, M. R., Smith, S. T., & Simonetti, A. (2016). Entanglement and the Formation of the Ancient
Nubian Napatan State. American Anthropologist, 118(2), 284-300.
Caldarini, C., Catalano, P., Gazzaniga, V., Marinozzi, S., & Zavaroni, F. (2015). The study of ancient
bone remains. In Bones (pp. 3-38). Springer, Cham.
Callaway, Ewen. (2017). African’s genomes expose Africa’s past. Nature, 13 July 2017, 149.
Campos, P. F., Craig, O. E., Turner-Walker, G., Peacock, E., Willerslev, E., & Gilbert, M. T. P. (2012).
DNA in ancient bone–where is it located and how should we extract it?. Annals of Anatomy-Anatomischer
Anzeiger, 194(1), 7-16.
Cann, R. L., Stoneking, M., & Wilson, A. C. (1987). Mitochondrial DNA and human
evolution. Nature, 325(6099), 31-36.

136
Canós-Donnay, S. (2019, February 25). The Empire of Mali. Oxford Research Encyclopedia of African
History. Retrieved 16 Sep. 2019, from
https://oxfordre.com/africanhistory/view/10.1093/acrefore/9780190277734.001.0001/acrefore-
9780190277734-e-266
Carlson, D. S., Armelagos, G. J., & Van Gerven, D. P. (1974). Factors influencing the etiology of cribra
orbitalia in prehistoric Nubia. Journal of Human Evolution, 3(5), 405-410.
Cavendish, M. 1966. The Custom of Placing Pebbles on Nubian Graves’, Sudan Notes and Records 47,
151-156.
Cerezo, M., Achilli, A., Olivieri, A., Perego, U. A., Gómez-Carballa, A., Brisighelli, F., Lancioni, H.,
Woodward, S.R., Lopez-Soto, M., Carracedo, A., Capelli, C., Torroni, A., & Salas, A. (2012).
Reconstructing ancient mitochondrial DNA links between Africa and Europe. Genome research, 22(5),
821-826.
Chakraborty R. (1993) Analysis of Genetic Structure of Populations: Meaning, Methods, and Implications.
In Majumder P.P. (Ed.) Human Population Genetics. Boston: Springer.
Chiang, C. W., Marcus, J. H., Sidore, C., Biddanda, A., Al-Asadi, H., Zoledziewska, M., Pitzalis, M.,
Busonero, F., Maschio, A., Pistis, G., Steri, M., Angius, A., Lohmueller, K, E., Abecasis, G, R.,
Schlessinger, D., Cucca, F., & Novembre, J. (2018). Genomic history of the Sardinian population. Nature
Genetics, 50(10), 1426.
Childe, V. G. (1968). The Dawn of European civilization. London: Routledge.
Clark, J. G. D. (1965). Radiocarbon dating and the spread of farming economy. Antiquity, 39(153), 45-48.
Claussen, M., Kubatzki, C., Brovkin, V., Ganopolski, A., Hoelzmann, P., & Pachur, H. J. (1999).
Simulation of an abrupt change in Saharan vegetation in the mid‐Holocene. Geophysical Research
Letters, 26(14), 2037-2040.
Clima, R., Preste, R., Calabrese, C., Diroma, M. A., Santorsola, M., Scioscia, G., Simone, D., Shen, S.,
Gasparre, G., & Attimonelli, M. (2016). HmtDB 2016: data update, a better performing query system and
human mitochondrial DNA haplogroup predictor. Nucleic Acids Research, 45(D1), D698-D706.
Close, A.E. (1995). Few and far between: Early ceramics in North Africa. In W.K. Barnett and J.W. Hoops
(Eds.) The Emergence of Pottery, Washington, D.C.: Smithsonian.
Cooper, A., & Poinar, H. N. (2000). Ancient DNA: do it right or not at all. Science, 289(5482), 1139-1139.
Cunningham, C., Scheuer, L., & Black, S. (2016). Developmental Juvenile Osteology. Academic Press.
Dabney, J., & Meyer, M. (2019). Extraction of Highly Degraded DNA from Ancient Bones and Teeth. In
Shapiro, B & Holfreiter, M. (Eds.) Ancient DNA (pp. 25-29). New York: Humana Press.
Dabney, J., Knapp, M., Glocke, I., Gansauge, M. T., Weihmann, A., Nickel, B., Valdiosera, C., García, N.,
Pääbo, S., Arsuaga, J.L., & Meyer, M. (2013). Complete mitochondrial genome sequence of a Middle
Pleistocene cave bear reconstructed from ultrashort DNA fragments. Proceedings of the National
Academy of Sciences, 110(39), 15758-15763.
Dabney, J., Knapp, M., Glocke, I., Gansauge, M. T., Weihmann, A., Nickel, B., Valiosera, C., Garcia, I.,
Pääbo, S., Arsuaga, J., & Meyer, M. (2013). Complete mitochondrial genome sequence of a Middle
Pleistocene cave bear reconstructed from ultrashort DNA fragments. Proceedings of the National
Academy of Sciences, 110(39), 15758-15763.
Damgaard, P. B., Margaryan, A., Schroeder, H., Orlando, L., Willerslev, E., & Allentoft, M. E. (2015).
Improving access to endogenous DNA in ancient bones and teeth. Scientific Reports, 5, 11184.
Dann, R. J., Emberling, G., et al. 2016. El Kurru 2015-16. Preliminary Report. Sudan & Nubia, 20, 35-49.
deMenocal, P., Ortiz, J., Guilderson, T., Adkins, J., Sarnthein, M., Baker, L., & Yarusinsky, M. (2000).
Abrupt onset and termination of the African Humid Period: rapid climate responses to gradual insolation
forcing. Quaternary Science Reviews, 19(1-5), 347-361.

137
Der Sarkissian, C., Allentoft, M. E., Ávila-Arcos, M. C., Barnett, R., Campos, P. F., Cappellini, E., Ermini,
L., Fernández, R., da Fonseca, R., Ginolhac, A., Hansen, A.J., Jónsson, H., Korneliussen, T., Margaryan,
A., Martin, M.D., Moreno-Mayar, J.V., Raghavan, M., Rasmussen, M., Velasco, M.S., Schroeder, H.,
Schubert, M., Seguin-Orlando, A., Wales, N., Gilbert, M.T., Willerslev, E., & Orlando, L. (2015). Ancient
genomics. Philosophical Transactions of the Royal Society B: Biological Sciences, 370(1660), 20130387.
DeSalle, R., Schierwater, B., & Hadrys, H. (2017). MtDNA: The small workhorse of evolutionary studies.
Frontiers in Bioscience, 22, 873-887.
Djuric, M., Milovanovic, P., Janovic, A., Draskovic, M., Djukic, K., & Milenkovic, P. (2008). Porotic lesions
in immature skeletons from Stara Torina, late medieval Serbia. International Journal of
Osteoarchaeology, 18(5), 458-475.
Dobon, B., Hassan, H. Y., Laayouni, H., Luisi, P., Ricaño-Ponce, I., Zhernakova, A., Wijmenga, C., Tahir,
H., Comas, D., Netea, M.G. & Bertranpetit, J. (2015). The genetics of East African populations: a Nilo-
Saharan component in the African genetic landscape. Scientific Reports, 5, 9996.
Edwards, D. N. (2004). The Nubian Past: An Archaeology of the Sudan. New York: Routledge.
Edwards, D. N. (2013). Medieval and Post-Medieval States of the Nile Valley. In Mitchell, P. & Lane, P.J.
(Eds.) The Oxford Handbook of African Archaeology.
El Tayeb, M., Juszczyk-Futkowska, K. and Czyżewska, E. (2014). El-Zuma 2011: the fourth season of
excavations on the site. Preliminary report. Polish Archaeology in the Mediterranean 23/1, 357–374.
El-Tayeb, M. (2007). Early Makuria Research Project. el-Zuma excavations preliminary report on the
second season, 2007. Polish archaeology in the Mediterranean, 19, 467-479.
El-Tayeb, M., & Czyżewska, E. (2011). Early Makuria Research Project. Excavations at ez-Zuma. The
third season, Jan.–Feb. 2009. Sudan & Nubia, 15, 108-118.
El-Tayeb, M., Czyżewska-Zalewska, E., Kowarska, Z., & Lenarczyk, S. (2016a). Early Makuria Research
Project: interim report on the excavation at el-Detti in 2014 and 2015. In PAM (Vol. 25, pp. 403-430).
El-Tayeb, M., Skowrońska, E., Czyżewska, E. (2016). Early Makuria Research Project. The Results of
Three Seasons of Excavation at El-Zuma Cemetery, 2013, 2014 and 2015. Sudan & Nubia 20, 110–126.
Emberling, G. (2012). Archaeological salvage in the Fourth Cataract, Northern Sudan (1991-2008).
Ancient Nubia: African Kingdoms on the Nile, 71-77.
Emberling, G. (2012). Archaeological salvage in the Fourth Cataract, Northern Sudan (1991-2008).
Ancient Nubia: African Kingdoms on the Nile, 71-77.
Emberling, G., Dann, R. J., et al. 2013. New Excavations at El Kurru: Beyond the Napatan Royal
Cemetery. Sudan & Nubia, 17, 42-60.
Emberling, G., Dann, R. J., et al. 2015. In a Royal Cemetery of Kush: Archaeological Investigations at El
Kurru, Northern Sudan, 2014-15. Sudan & Nubia, 19, 54-70.
Emberling, Geoff. (2011). “Nubia: Ancient Kingdoms of Africa.” The Institute for the Study of the Ancient
World, New York University.
Emery, L. S., Magnaye, K. M., Bigham, A. W., Akey, J. M., & Bamshad, M. J. (2015). Estimates of
continental ancestry vary widely among individuals with the same mtDNA haplogroup. The American
Journal of Human Genetics, 96(2), 183-193.
Ewels, P., Magnusson, M., Lundin, S., & Käller, M. (2016). MultiQC: summarize analysis results for
multiple tools and samples in a single report. Bioinformatics, 32(19), 3047-3048.
Excoffier, L., & Lischer, H. E. (2010). Arlequin suite ver 3.5: a new series of programs to perform
population genetics analyses under Linux and Windows. Molecular Ecology Resources, 10(3), 564-567.
Fadhlaoui‐Zid, K., Rodríguez‐Botigué, L., Naoui, N., Benammar‐Elgaaied, A., Calafell, F., & Comas, D.
(2011). Mitochondrial DNA structure in North Africa reveals a genetic discontinuity in the Nile Valley.
American Journal of Physical Anthropology, 145(1), 107-117.

138
Fernandes, V., Triska, P., Pereira, J. B., Alshamali, F., Rito, T., Machado, A., Fajkosova, Z., Cavadas, B.,
Cerny, V., Soares, P., & Richards, M. B. (2015). Genetic stratigraphy of key demographic events in
Arabia. PloS one, 10(3), e0118625.
Flammini, R. (2008). Ancient Core-Periphery Interactions: Lower Nubia during Middle Kingdom Egypt (ca.
2050-1640 BC). Journal of World-Systems Research, 14(1), 50-74.
Fregel, R., Méndez, F. L., Bokbot, Y., Martin-Socas, D., Camalich-Massieu, M. D., Santana, J., Morales,
J., Avila-Arcos, A.E., Underhill, P.A., Shapiro, B., Wojcik, G., Rasmussen, M., Soares, A.E.R., Kapp, J.,
Sockell, A., Rodriquez-Santos, J., Mikdad, A., Trujillo-Medros, A., Bustamonte, C. (2018). Ancient
genomes from North Africa evidence prehistoric migrations to the Maghreb from both the Levant and
Europe. Proceedings of the National Academy of Sciences, 115(26), 6774-6779.
Fregel, R., Ordóñez, A. C., Santana-Cabrera, J., Cabrera, V. M., Velasco-Vázquez, J., Alberto, V.,
Moreno-Benítez, M.A., Delgado-Darias, T., Rodríguez-Rodríguez, A., Hernández, J.C., Pais, J.,
González-Montelongo, R, Lorenzo-Salazar, J.M., Flores, C., Cruz-de-Mercadal, M.C., Álvarez-Rodríguez,
N., Shapiro, B., Arnay, M., & Bustamante, C.D. (2019). Mitogenomes illuminate the origin and migration
patterns of the indigenous people of the Canary Islands. PloS one, 14(3), e0209125.
Fu, Q., Mittnik, A., Johnson, P. L., Bos, K., Lari, M., Bollongino, R., Sun, C., Giemsch, L., Schmitz, R.,
Burger, J., Ronchitelli, A.M., Martini, F., Cremonesi, R. G., Svoboda, J., Bauer, P., Caramelli, D.,
Castellano, S., Reich, D., & Krause, J. (2013). A revised timescale for human evolution based on ancient
mitochondrial genomes. Current Biology, 23(7), 553-559.
Fulton, T. L., & Shapiro, B. (2019). Setting up an ancient DNA laboratory. In Shapiro, B & Holfreiter, M.
(Eds.) Ancient DNA (pp. 1-13). New York: Humana Press.
Gallego Llorente, M., Jones, E. R., Eriksson, A., Siska, V., Arthur, K. W., Arthur, J. W., Curtis, M.C.,
Stock, J.T., Coltorti, M., Pieruccini, P., Stretton, S., Brock, F., Higham, T., Park, Y., Hofreiter, M., Bradley,
D.G., Bhak, J., Pinhasi, R., & Manica, A. (2015). Ancient Ethiopian genome reveals extensive Eurasian
admixture throughout the African continent. Science, 350(6262), 820-822.
Gamba, C., Jones, E. R., Teasdale, M. D., McLaughlin, R. L., Gonzalez-Fortes, G., Mattiangeli, V.,
Domboróczki, L., Kővári, I., Pap, I., Anders, A., Whittle, A., Dani, J., Raczky, P., Higham, T.F., Hofreiter,
M., Bradley, D.G., & Pinhasi, R. (2014). Genome flux and stasis in a five millennium transect of European
prehistory. Nature Communications, 5, 5257.
Gatto, M. C. (2006). The Nubian A-group: A reassessment. Archéo-Nil, 16, 61-76.
Giles, R. E., Blanc, H., Cann, H. M., & Wallace, D. C. (1980). Maternal inheritance of human
mitochondrial DNA. Proceedings of the National Academy of Sciences, 77(11), 6715-6719.
Ginolhac, A., Rasmussen, M., Gilbert, M. T. P., Willerslev, E., & Orlando, L. (2011). mapDamage: testing
for damage patterns in ancient DNA sequences. Bioinformatics, 27(15), 2153-2155.
Godlewski, W. (2014). Dongola Capital of early Makuria: Citadel–Rock Tombs–First Churches. Der Antike
Sudan. Mitteilungen der Sudanarchäologischen Gesellschaft zu Berlin e.V. Sonderheft. Berlin.
Gondek, A. T., Boessenkool, S., & Star, B. (2018). A stainless-steel mortar, pestle and sleeve design for
the efficient fragmentation of ancient bone. BioTechniques, 64(6), 266-269.
Goodman, A. H., & Martin, D. L. (2002). Reconstructing health profiles from skeletal remains. In Steckel,
R. H. & Rose, J. C. The Backbone of history: Health and Nutrition in the Western hemisphere (pp. 11-60).
Cambridge: Cambridge University Press.
Goodman, A. H., & Rose, J. C. (1991). Dental enamel hypoplasias as indicators of nutritional status.
Advances in dental anthropology, 5, 225-240.
Goodman, A. H., Armelagos, G. J., & Rose, J. C. (1980). Enamel hypoplasias as indicators of stress in
three prehistoric populations from Illinois. Human Biology, 52(3), 515-528.
Goodman, A. H., Brooke Thomas, R., Swedlund, A. C., & Armelagos, G. J. (1988). Biocultural
perspectives on stress in prehistoric, historical, and contemporary population research. American Journal
of Physical Anthropology, 31(S9), 169-202.

139
Green, R. E., Krause, J., Ptak, S. E., Briggs, A. W., Ronan, M. T., Simons, J. F., Du, L., Egholm, M.,
Rothber, J.M., Paunovic, M. & Pääbo, S. (2006). Analysis of one million base pairs of Neanderthal DNA.
Nature, 444(7117), 330.
Haak, W., Forster, P., Bramanti, B., Matsumura, S., Brandt, G., Tänzer, M., Villems, R., Renfrew, C.,
Gronenborn, D, Alt, K.W., & Burger, J. (2005). Ancient DNA from the first European farmers in 7500-year-
old Neolithic sites. Science, 310(5750), 1016-1018.
Hagelberg, E., Hofreiter, M., & Keyser, C. (2015). Ancient DNA: the first three decades.
Hansen, H. B., Damgaard, P. B., Margaryan, A., Stenderup, J., Lynnerup, N., Willerslev, E., & Allentoft,
M. E. (2017). Comparing ancient DNA preservation in petrous bone and tooth cementum. PloS one,
12(1), e0170940.
Harris, N. J., & Tayles, N. (2012). Burial containers–A hidden aspect of mortuary practices:
Archaeothanatology at Ban Non Wat, Thailand. Journal of Anthropological Archaeology, 31(2), 227-239.
Hassan Mohamed, H.Y. (2009). Genetic Patterns of Y-chromosome and Mitochondrial DNA Variation,
with Implications to the Peopling of the Sudan. (Doctoral Thesis, University of Khartoum.) Retrieved from
https://www.docdroid.net/8GAIp0X/genetic-patterns-of-y-chromosome-and-mitochondrial-hassan-
2009.pdf
Hassan, H. Y., Underhill, P. A., Cavalli‐Sforza, L. L., & Ibrahim, M. E. (2008). Y‐chromosome variation
among Sudanese: restricted gene flow, concordance with language, geography, and history. American
Journal of Physical Anthropology, 137(3), 316-323.
Henneberger, K., Barlow, A., & Paijmans, J. L. (2019). Double-Stranded Library Preparation for Ancient
and Other Degraded Samples. In Shapiro, B. & Hofreiter, M. (Eds.) Ancient DNA (pp. 65-73). New York:
Humana Press.
Herbig, A., Maixner, F., Bos, K. I., Zink, A., Krause, J., & Huson, D. H. (2016). MALT: Fast alignment and
analysis of metagenomic DNA sequence data applied to the Tyrolean Iceman. BioRxiv, 050559.
Hibbs, A. C., Secor, W. E., Van Gerven, D., & Armelagos, G. (2011). Irrigation and infection: the
immunoepidemiology of schistosomiasis in ancient Nubia. American Journal of Physical Anthropology,
145(2), 290-298.
Hillelson, S. (1930). Nubian Origins. Sudan Notes and Records, 13(1), 137-148.
Hillson, S. (1996). Dental Anthropology. Cambridge: Cambridge University Press.
Hintze, F. (1967). "Meroe and the Noba", Der Antike Sudan. Mitteilungen der Sudanarchäologischen
Gesellschaft zu Berlin 10:49-55.
Hofreiter, M., Paijmans, J. L., Goodchild, H., Speller, C. F., Barlow, A., Fortes, G. G., Thomas, J. Ludwig,
A., & Collins, M. J. (2014). The future of ancient DNA: Technical advances and conceptual shifts.
BioEssays, 37(3), 284-293.
Hollfelder, N., Schlebusch, C. M., Günther, T., Babiker, H., Hassan, H. Y., & Jakobsson, M. (2017).
Northeast African genomic variation shaped by the continuity of indigenous groups and Eurasian
migrations. PLoS Genetics, 13(8), e1006976.
Holsinger, K. E., & Weir, B. S. (2009). Genetics in geographically structured populations: defining,
estimating and interpreting F ST. Nature Reviews Genetics, 10(9), 639.
Höss, M., Jaruga, P., Zastawny, T. H., Dizdaroglu, M., & Paabo, S. (1996). DNA damage and DNA
sequence retrieval from ancient tissues. Nucleic Acids Research, 24(7), 1304-1307.
Hummert, J. R., & Van Gerven, D. P. (1983). Skeletal growth in a medieval population from Sudanese
Nubia. American Journal of Physical Anthropology, 60(4), 471-478.
Hurst, C. V. (2013). Growing up in medieval Nubia: Health, disease, and death of a medieval juvenile
sample from Mis Island. Ph.D. thesis. Michigan State University, Lansing.

140
Irish, D., & Joel, D. (2005). Population continuity vs. discontinuity revisited: Dental affinities among late
Paleolithic through Christian‐era Nubians. American Journal of Physical Anthropology, 128(3), 520-535.
Jackes, M. (2011). Representativeness and bias in archaeological skeletal samples. In Agarwal, S.C. &
Glencross, B.A. (Eds.) Social bioarchaeology, (pp. 107-146). Hoboken: Blackwell Publishing Inc.
Jesse, F. (2010). Early pottery in northern Africa–an overview. Journal of African Archaeology, 8(2), 219-
238.
Jónsson, H., Ginolhac, A., Schubert, M., Johnson, P. L., & Orlando, L. (2013). mapDamage2. 0: fast
approximate Bayesian estimates of ancient DNA damage parameters. Bioinformatics, 29(13), 1682-1684.
Judd, M. (2002). Ancient injury recidivism: an example from the Kerma period of ancient Nubia.
International Journal of Osteoarchaeology, 12(2), 89-106.
Judd, M.A. (2000). Trauma and interpersonal violence in ancient Nubia during the Kerma period (ca.
2500-1500 BC). Ph.D. thesis, University of Alberta, Edmonton.
Jugert, F., Klatt, T. W., Grosskopf, B., and Hummel, S. (2018). Anthropologic and Paleogenetic Studies
on Human Skeletal Remains from the Wadi Abu Dom, Sudan. In Lohwasser, A., Kerberg, T., and
Auenmüller, J. (Eds.), Bayuda Studies: Proceedings of the First International Conference on the
Archaeology of the Bayuda Desert in Sudan (pp. 231-246). Wiesbaden, Germany: Harrassowitz Verlag.
Kaestle, F. A. (2010). Paleogenetics: Ancient DNA in Anthropology. In Larsen, L (Ed.) A Companion to
Biological Anthropology, 427.
Kemp, B. M., & Smith, D. G. (2005). Use of bleach to eliminate contaminating DNA from the surface of
bones and teeth. Forensic Science International, 154(1), 53-61.
Kendall, T. (1999). The origin of the Napatan State: El Kurru and the evidence for the royal ancestors. In
S. Wenig, ed., Studien zum antiken Sudan: Akten der 7. Internationalen Tagung für meroitische
Forschungen: 2–117. Wiesbaden
Kircher, M., Sawyer, S., & Meyer, M. (2011). Double indexing overcomes inaccuracies in multiplex
sequencing on the Illumina platform. Nucleic acids research, 40(1), e3-e3.
Kivisild T, Reidla M, Metspalu E, Rosa A, Brehm A, Pennarun E, Parik J, Geberhiwot T, Usanga E,
Villems R. (2004). Ethiopian Mitochondrial DNA Heritage: Tracking Gene Flow Across and Around the
Gate of Tears. American Journal of Human Genetetics, 75, 752-770.
Klimaszewska-Drabot, E. (2008). Early Makuria Research Project: The Pottery. Polish Archaeology in the
Mediterranean, (XVIII), 477-491.
Korlević, P., & Meyer, M. (2019). Pretreatment: Removing DNA Contamination from Ancient Bones and
Teeth Using Sodium Hypochlorite and Phosphate. In Shaprio, B. & Hofreiter, M. (Eds.) Ancient DNA (pp.
15-19). New York: Humana Press.
Korlević, P., Gerber, T., Gansauge, M. T., Hajdinjak, M., Nagel, S., Aximu-Petri, A., & Meyer, M. (2015).
Reducing microbial and human contamination in DNA extractions from ancient bones and teeth.
Biotechniques, 59(2), 87-93.
Krings, M., Halim Salem, A., Bauer, K., Geisert, H., Malek, A.K., Chaix, L., Simon, C., Welsby, D., Di
Rienzo, A., Utermann, G., Sajantila, A., Pääbo, S., & Stoneking, M. (1999). mtDNA analysis of Nile valley
populations: A genetic corridor or a barrier to migration? American Journal of Human Genetics,
64(4),1166-1176.
Krings, M., Stone, A., Schmitz, R. W., Krainitzki, H., Stoneking, M., & Pääbo, S. (1997). Neandertal DNA
sequences and the origin of modern humans. Cell, 90(1), 19-30.
Lalueza Fox, C. (1997). mtDNA analysis in ancient Nubians supports the existence of gene flow between
sub-Sahara and North Africa in the Nile Valley. Annals of Human Biology, 24(3), 217-227.
Larsen, C. S. (2015). Bioarchaeology: Interpreting Behavior from the Human Skeleton. Cambridge:
Cambridge University Press.

141
Larsson, A. (2014). AliView: a fast and lightweight alignment viewer and editor for large data sets.
Bioinformatics 30(22): 3276-3278. http://dx.doi.org/10.1093/bioinformatics/btu531
Lazaridis, I., Nadel, D., Rollefson, G., Merrett, D. C., Rohland, N., Mallick, S., Fernandes, D., Novak, M.,
Gamarra, B., Sirak, K., Connell, S., Stewardson, K., Harney, E., Fu, Q., Gonzalez-Fortes, G., Jones, E.R.,
Roodenberg, S.A., Lengyel, G., Bocquentin, F., Gasparian, B., Monge, J.M., Gregg, M., Eshed, V.,
Mizrahi, A.S., Meiklejohn, C., Gerritsen, F., Bejenaru, L., Blüher, M., Campbell, A., Cavalleri, G., Comas,
D., Froguel, P., Gilbert, E., Kerr, S.M., Kovacs, P., Krause, J., McGettigan, D., Merrigan, M., Merriwether,
D.A., O'Reilly, S., Richards, M.B., Semino, O., Shamoon-Pour, M., Stefanescu, G., Stumvoll, M., Tönjes,
A., Torroni, A., Wilson, J.F., Yengo, L., Hovhannisyan, N.A., Patterson, N., Pinhasi, R., & Reich D. (2016).
Genomic insights into the origin of farming in the ancient Near East. Nature, 536(7617), 419.
Leonard, J. A. (2008). Ancient DNA applications for wildlife conservation. Molecular Ecology, 17(19),
4186-4196.
Leplongeon, A. (2017). Technological variability in the Late Palaeolithic lithic industries of the Egyptian
Nile Valley: The case of the Silsilian and Afian industries. PloS one, 12(12), e0188824.
Lewis, M. E. (2004). Endocranial Lesions in Non - Adult Skeletons: Understanding Their Aetiology.
International Journal of Osteoarchaeology 14:87–97.
Lewis, M. E. (2007). The Bioarchaeology of Children: Perspectives from Biological and Forensic
Anthropology. Cambridge: Cambridge University Press.
Lindahl, T. (1993). Instability and decay of the primary structure of DNA. Nature, 362(6422), 709-715.
Lovejoy, C. O., Meindl, R. S., Mensforth, R. P., & Barton, T. J. (1985). Multifactorial determination of
skeletal age at death: a method and blind tests of its accuracy. American Journal of Physical
Anthropology, 68(1), 1-14.
Luo, S., Valencia, C. A., Zhang, J., Lee, N. C., Slone, J., Gui, B., Wang, X., Li, Z., Dell, S., Brown, J.,
Chen, S. M., Chien, Y.H., Hwu, W.L., Fan, P.C., Wong, L.J., Atwal, P.S., & Huang, T. (2018). Biparental
inheritance of mitochondrial DNA in humans. Proceedings of the National Academy of Sciences, 115(51),
13039-13044.
MacEachern, S. (2013). Genetics and Archaeology. In The Oxford handbook of African Archaeology.
Oxford: Oxford University Press.
Maricic, T., Whitten, M., & Pääbo, S. (2010). Multiplexed DNA sequence capture of mitochondrial
genomes using PCR products. PloS one, 5(11), e14004.
McIlvaine, B. K. (2015). Implications of reappraising the iron‐deficiency anemia hypothesis. International
Journal of Osteoarchaeology, 25(6), 997-1000.
Menozzi, P., Piazza, A., & Cavalli-Sforza, L. (1978). Synthetic maps of human gene frequencies in
Europeans. Science, 201(4358), 786-792.
Mensforth, R. P., Lovejoy, C. O., Lallo, J. W., & Armelagos, G. J. (1978). Part two: the role of
constitutional factors, diet, and infectious disease in the etiology of porotic hyperostosis and periosteal
reactions in prehistoric infants and children. Medical Anthropology, 2(1), 1-59.
Meyer, M., & Kircher, M. (2010). Illumina sequencing library preparation for highly multiplexed target
capture and sequencing. Cold Spring Harbor Protocols, 2010(6), pdb-prot5448.
Miquel-Feucht, M. J., Polo-Cerda, M., & Villalaín-Blanco, J. D. (1999). Anthropological and
paleopathological studies of a mass execution during the war of independence in Valencia, Spain (1808-
1812). Journal of Paleopathology, 11(3), 15-24.
Mitchell, P. D. (2006). Child health in the crusader period inhabitants of Tel Jezreel, Israel. Levant, 38(1),
37-44.
Mittler, D. M., & Van Gerven, D. P. (1994). Developmental, diachronic, and demographic analysis of
cribra orbitalia in the medieval Christian populations of Kulubnarti. American Journal of Physical
Anthropology, 93(3), 287-297.

142
Mittnik, A., Wang, C. C., Svoboda, J., & Krause, J. (2016). A molecular approach to the sexing of the
triple burial at the Upper Paleolithic Site of Dolní Věstonice. PloS one, 11(10), e0163019.
Morris, A. G., Heinze, A., Chan, E. K., Smith, A. B., & Hayes, V. M. (2014). First ancient mitochondrial
human genome from a prepastoralist southern African. Genome Biology and Evolution, 6(10), 2647-2653.
Moseley J.E. (1974). Skeletal changes in the anemias. Seminars in Roentgenology 9(3), 169-184.
Mulligan, C. J., & Szathmáry, E. J. (2017). The peopling of the Americas and the origin of the Beringian
occupation model. American Journal of Physical Anthropology, 162(3), 403-408.
Neukamm, J., & Peltzer, A. (2009, March 9), mitobench/MitoBench: MitoBench version 1.1 bets(Version
v1.1-beta). Zenodo. http://doi.org/10.5281/zenodo.2588120
Nilsson Stutz, L. (2003). Embodied rituals and ritualized bodies: tracing ritual practices in Late Mesolithic
burials. Acta Archaeologica Lundensia. Series in 8, 46.
Noonan, J. P., Hofreiter, M., Smith, D., Priest, J. R., Rohland, N., Rabeder, G., Krause, J., Detter, J.C.,
Pääbo, S., & Rubin, E.M. (2005). Genomic sequencing of Pleistocene cave bears. Science, 309(5734),
597-599.
Obłuski, A. (2014). Ghazali Site Presentation Project 2012 – 2014: Preliminary Results. Der Antike
Sudan, 25, 197-204.
Obłuski, A., Ochała, G., Bogacki, M., Małkowski, W., Maślak,S., and Zaki ed-Din, M. (2015). Ghazali
2012: preliminary report. Polish Archaeology in the Mediterranean, 24(1), 431-442.
Ortner, D. J. (2003). Identification of pathological conditions in human skeletal remains. Academic Press.
Pääbo, S. (1985). Molecular cloning of ancient Egyptian mummy DNA. Nature, 314(6012), 644.
Pääbo, S., & Wilson, A. C. (1988). Polymerase chain reaction reveals cloning artefacts.
Nature, 334(6181), 387.
Palanichamy, M. G., Zhang, C. L., Mitra, B., Malyarchuk, B., Derenko, M., Chaudhuri, T. K., & Zhang, Y.
P. (2010). Mitochondrial haplogroup N1a phylogeography, with implication to the origin of European
farmers. BMC evolutionary biology, 10(1), 304.
Peltzer, A., Jäger, G., Herbig, A., Seitz, A., Kniep, C., Krause, J., & Nieselt, K. (2016). EAGER: efficient
ancient genome reconstruction. Genome Biology, 17(1), 60. https://github.com/apeltzer/EAGER-GUI
Pereira, L., Richards, M., Goios, A., Alonso, A., Albarrán, C., Garcia, O., Behar, D.M., Gölge, M., Hatina,
J., Al-Gazali, L., Bradley, D.G., Macauley, V., & Amorim, A. (2005). High-resolution mtDNA evidence for
the late-glacial resettlement of Europe from an Iberian refugium. Genome research, 15(1), 19-24.
Peter, B. M., Petkova, D., & Novembre, J. (2018). Genetic landscapes reveal how human genetic
diversity aligns with geography. BioRxiv, 233486.
Phenice, T. W. (1969). A newly developed visual method of sexing the os pubis. American Journal of
Physical Anthropology, 30(2), 297-301.
Phillips, J. (2003): An Overview of the Ceramics. In: Żurawski, Bogdan T. (Ed.) Survey and Excavations
between Old Dongola and Ez-Zuma. Southern Dongola Reach of the Nile from Prehistory to 1820 AD
based on the Fieldwork conducted in 1997-2003 by the Polish Archaeological Joint Expedition to the
Middle Nile. Southern Dongola Reach Survey 1. (pp. 387-437) Warsaw: Nubia II.
Pickrell, J. K., & Reich, D. (2014). Toward a new history and geography of human genes informed by
ancient DNA. Trends in Genetics, 30(9), 377-389.
Pinhasi, R., & Bourbou, C. (2007). How representative are human skeletal assemblages for population
analysis?. In Pinhasi, R. & Simon, M. (Eds.) Advances in Human Palaeopathology (p. 31-44). Hoboken:
John Wiley & Sons, Ltd.
Pinhasi, R., Fernandes, D. M., Sirak, K., & Cheronet, O. (2019). Isolating the human cochlea to generate
bone powder for ancient DNA analysis. Nature Protocols, 14(4), 1194.

143
Pinhasi, R., Fernandes, D., Sirak, K., Novak, M., Connell, S., Alpaslan-Roodenberg, S., Anders, A.,
Pietrusewsky, M., Rollefson, G., Jovanovic, M., Trinhhoang, H., Bar-Oz, G., Oxenham, M., Matsumura,
H., & Hofreiter, M. (2015). Optimal ancient DNA yields from the inner ear part of the human petrous bone.
PloS one, 10(6), e0129102.
Poinar, H. N., Schwarz, C., Qi, J., Shapiro, B., MacPhee, R. D., Buigues, B., Tikhonov, A., Huson, D.H.,
Tomsho, L.P., Auch, A., Rampp, M., Miller, W., & Schuster, S.C. (2006). Metagenomics to
paleogenomics: large-scale sequencing of mammoth DNA. Science, 311(5759), 392-394.
Prendergast, M. E., Lipson, M., Sawchuk, E. A., Olalde, I., Ogola, C. A., Rohland, N., Sirak, K.A.,
Adamski, N., Bernardos, R., Broomandkhoshbacht, N., Callan, K., Culleton, B.J., Eccles, L., Harper, T.K.,
Lawson, A.M., Mah, M., Oppenheimer, J., Stewardson, K., Zalzala, F., Ambrose, S.H., Ayodo, G., Gates,
H.L. Jr., Gidna, A.O., Katongo, M., Kwekason, A., Mabulla, A.Z.P., Mudenda, G.S., Ndiema, E.K., Nelson,
C., Robertshaw, P., Kennett, D.J., Manthi, F.K., & Reich, D. (2019). Ancient DNA reveals a multistep
spread of the first herders into sub-Saharan Africa. Science, eaaw6275.
Renaud, G., Slon, V., Duggan, A. T., & Kelso, J. (2015). Schmutzi: estimation of contamination and
endogenous mitochondrial consensus calling for ancient DNA. Genome Biology, 16(1), 224.
Rinaldo, N., Zedda, N., Bramanti, B., Rosa, I., & Gualdi-Russo, E. (2019). How reliable is the assessment
of Porotic Hyperostosis and Cribra Orbitalia in skeletal human remains? A methodological approach for
quantitative verification by means of a new evaluation form. Archaeological and Anthropological
Sciences, 1-11.
Rito, T., Richards, M. B., Fernandes, V., Alshamali, F., Cerny, V., Pereira, L., & Soares, P. (2013). The
first modern human dispersals across Africa. PloS one, 8(11), e80031.
Roostalu, U., Kutuev, I., Loogväli, E. L., Metspalu, E., Tambets, K., Reidla, M., Khusnutdinova, E.K.,
Usanga, E., Kivisild, T. & Villems, R. (2006). Origin and expansion of haplogroup H, the dominant human
mitochondrial DNA lineage in West Eurasia: The Near Eastern and Caucasian perspective. Molecular
Biology and Evolution, 24(2), 436-448.
Rowold, D. J., Luis, J. R., Terreros, M. C., & Herrera, R. J. (2007). Mitochondrial DNA geneflow indicates
preferred usage of the Levant Corridor over the Horn of Africa passageway. Journal of Human
Genetics, 52(5), 436.
Salas, A., Richards, M., De la Fe, T., Lareu, M. V., Sobrino, B., Sánchez-Diz, P., Macauley, V., &
Carracedo, Á. (2002). The making of the African mtDNA landscape. The American Journal of Human
Genetics, 71(5), 1082-1111.
Salas, A., Richards, M., Lareu, M. V., Scozzari, R., Coppa, A., Torroni, A., Macauley, V., & Carracedo, A.
(2004). The African diaspora: mitochondrial DNA and the Atlantic slave trade. The American Journal of
Human Genetics, 74(3), 454-465.
Salter, S.J., Cox, M.J., Turek, E.M., Calus, T.S., Cookson, W.O., Moffatt, M.F., Turner, P., Parkhill, J.,
Loman, N.J. & Walker, A.W. (2014). Reagent and Laboratory contamination can critically impact
sequence-based microbiome analyses. BMC Biology 12(1), 87.
Schaefer, M., Black, S. M., Schaefer, M. C., & Scheuer, L. (2009). Juvenile osteology. London: Academic
Press.
Scheinfeldt, L. B., Soi, S., & Tishkoff, S. A. (2010). Working toward a synthesis of archaeological,
linguistic, and genetic data for inferring African population history. Proceedings of the National Academy
of Sciences, 107(Supplement 2), 8931-8938.
Schlebusch, C. M., Malmström, H., Günther, T., Sjödin, P., Coutinho, A., Edlund, H., Munters, A.R.,
Visente, M., Steyn, M., Soodyall, H., Lombard, M., & Jakobsson, M. (2017). Southern African ancient
genomes estimate modern human divergence to 350,000 to 260,000 years ago. Science, 358(6363),
652-655.
Schroeder, H., de Barros Damgaard, P., & Allentoft, M. E. (2019). Pretreatment: Improving Endogenous
Ancient DNA Yields Using a Simple Enzymatic Predigestion Step. In Shapiro, B. & Hofreiter, M. (Eds.)
Ancient DNA (pp. 21-24). New York: Humana Press.

144
Schuenemann, V. J., Peltzer, A., Welte, B., van Pelt, W. P., Molak, M., Wang, C. C., Furtwaegler, A.,
Urban, C., Reiter, E., Nieselt, K., Teßmann, B. Francken, M., Harvati, K., Haak, W., Schiffels, S. &
Krause, J. (2017). Ancient Egyptian mummy genomes suggest an increase of Sub-Saharan African
ancestry in post-Roman periods. Nature Communications, 8: 15694.
Schwarz, C., Debruyne, R., Kuch, M., McNally, E., Schwarcz, H., Aubrey, A. D., Bada, J., & Poinar, H.
(2009). New insights from old bones: DNA preservation and degradation in permafrost preserved
mammoth remains. Nucleic Acids Research, 37(10), 3215-3229.
Silva, M., Alshamali, F., Silva, P., Carrilho, C., Mandlate, F., Trovoada, M. J., Černý, V. & Soares, P.
(2015). 60,000 years of interactions between Central and Eastern Africa documented by major African
mitochondrial haplogroup L2. Scientific Reports, 5, 12526.
Sirak, K. A., Fernandes, D. M., Cheronet, O., Novak, M., Gamarra, B., Balassa, T., Bernert, Z6, Cséki A7,
Dani J8, Gallina JZ9, Kocsis-Buruzs G10, Kővári I11, László O12, Pap I6, Patay R13, Petkes Z14,
Szenthe G15, Szeniczey T16, Hajdu T16, Pinhasi R2,17 & Kocsis-Buruzs, G. (2017). A minimally-
invasive method for sampling human petrous bones from the cranial base for ancient DNA analysis.
BioTechniques, 62(6), 283-289.
Sirak, K., Frenandes, D., Novak, M., Van Gerven, D., & Pinhasi, R. (2016, January). A community
divided? Revealing the community genome (s) of Medieval Kulubnarti using next-generation sequencing.
In IUAES Inter-Congress 2016.
Skoglund, P., Northoff, B. H., Shunkov, M. V., Derevianko, A. P., Pääbo, S., Krause, J., & Jakobsson, M.
(2014). Separating endogenous ancient DNA from modern day contamination in a Siberian Neanderthal.
Proceedings of the National Academy of Sciences, 111(6), 2229-2234.
Skoglund, P., Thompson, J. C., Prendergast, M. E., Mittnik, A., Sirak, K., Hajdinjak, M., Salie, T.,
Rohland, N., Mallick, S., Peltzer, A., Heinze, A., Olalde, I., Ferry, M., Harney, E., Michel, M., Stewardson,
K., Cerezo-Román, J.I., Chiumia, C., Crowther, A., Gomani-Chindebvu, E., Gidna, A.O., Grillo, K.M.,
Helenius, I.T., Hellenthal, G., Helm, R., Horton, M., López, S., Mabulla, A.Z.P., Parkington, J., Shipton, C.,
Thomas, M.G., Tibesasa, R., Welling, M., Hayes, V.M., Kennett, D., Ramesar, R., Meyer, M., Pääbo, S.,
Patterson, N., Morris, A.G., Boivin, N., Pinhasi, R., Krause, J., & Reich, D. (2017). Reconstructing
prehistoric African population structure. Cell, 171(1), 59-71.
Skuldbøl, T., Uildriks, M., Rose, K., Philips, J., & Breidenstein, A. (2016). “The Medieval Fortification,
Settlement, and Cemetery” in El-Kurru 2015-16: Preliminary Report. Sudan & Nubia, 20, 40-46.
Smith, B. H. (1991). Standards of Human Tooth Formation and Dental Age Assessment. Wiley-Liss Inc.
Smith, G.E., & Jones, F.W. (1910). The Archaeological Survey of Nubia: Report for 1907–1908, volume
II, Report on Human Remains. Cairo: National Printing Department.
Smith-Guzmán, N. E. (2015a). Cribra orbitalia in the ancient Nile Valley and its connection to malaria.
International Journal of Paleopathology, 10, 1-12.
Smith‐Guzmán, N. E. (2015b). The skeletal manifestation of malaria: An epidemiological approach using
documented skeletal collections. American journal of physical anthropology, 158(4), 624-635.
Soares, P., Ermini, L., Thomson, N., Mormina, M., Rito, T., Röhl, A., Salas, A., Oppenheimer, S.,
Macauley, V., & Richards, M. B. (2009). Correcting for purifying selection: an improved human
mitochondrial molecular clock. The American Journal of Human Genetics, 84(6), 740-759.
Sokal, R. R., Oden, N. L., & Wilson, C. (1991). Genetic evidence for the spread of agriculture in Europe
by demic diffusion. Nature, 351(6322), 143.
Soler, A. (2012). Life and death in a medieval Nubian farming community: the experience at Mis Island.
Ph.D. thesis. Michigan State University. Lansing.
Spyrou, M. A., Bos, K. I., Herbig, A., & Krause, J. (2019). Ancient pathogen genomics as an emerging
tool for infectious disease research. Nature Reviews Genetics, 20, 323-340.

145
Stark, R. J., & Ciesielska, J. (2019). Vertebral infection in a male individual buried in the monastic
cemetery (Cemetery 2) at Ghazali (ca. 670–1270 CE), Northern Sudan. International Journal of
Paleopathology, 24, 34-40.
Steinbock, R.T. (1976). Paleopathological Diagnosis and Interpretation: Bone Diseases in Ancient Human
Populations, Charles C. Thomas, Illinois.
Stevanovitch A, Gilles A, Bouzaid E, Kefi R, Paris F, Gayraud RP, Spadoni JL, El-Chenawi. (2003).
Mitochondrial DNA Sequence diversity in a sedentary population from Egypt. Annals of Human Genetics,
68, 23-39.
Stoneking, M. (2018). Mitochondrial DNA. In Trevathan, W. (Ed.) The International Encyclopedia of
Biological Anthropology. John Wiley & Sons, Inc.
Stuart‐Macadam, P. (1989). Porotic hyperostosis: relationship between orbital and vault lesions.
American Journal of Physical Anthropology, 80(2), 187-193.
The Arizona Science Center. (2011, May 13). Broken Bones. ASU - Ask A Biologist. Retrieved August 8,
2019 from https://askabiologist.asu.edu/how-bone-breaks
Thoma, K. H., & Goldman, H. M. (1960). Oral pathology (p. 1192). St. Louis: Mosby.
Thomas, M.G., Weale, M.E., Jones, A.L., Richards, M., Smith, A., Redhead, N., Torroni, A., Scozzari, R.,
Gratrix, F., Tarekegn, A., Wilson, J.F., Capelli, C., Bradman, N., & Goldstein, D.B. (2002). Founding
mothers of Jewish communities: geographically separated Jewish groups were independently founded by
very few female ancestors. Am J Hum Genet 70:1411-1420.
Tishkoff, S. A., Reed, F. A., Friedlaender, F. R., Ehret, C., Ranciaro, A., Froment, A., Hirbo, J.B,
Awomoyi, A.A., Bodo, J.M., Doumbo, O., Ibrahim, M., Juma, A.J., Kotze, M.J., Lema, G., Moore, J.H.,
Mortensen, H., Nyambo, T.B., Omar, S.A., Powell, K., Pretorius, G.S., Smith, M.W., Thera, M.A.,
Wambebe, C., Weber, J.L., & Williams, S.M. (2009). The genetic structure and history of Africans and
African Americans. Science, 324(5930), 1035-1044.
Török, L. (2008). Between two worlds: the frontier region between ancient Nubia and Egypt 3700 BC-500
AD. Leiden: Brill.
Vai, S., Sarno, S., Lari, M., Luiselli, D., Manzi, G., Gallinaro, M., Mataich, S., Hübner, A., Modi, A., Pilli, E.,
Tafuri, M. & Tafuri, M. A. (2019). Ancestral mitochondrial N lineage from the Neolithic
‘green’Sahara. Scientific Reports, 9(1), 3530.
van de Loosdrecht, M., Bouzouggar, A., Humphrey, L., Posth, C., Barton, N., Aximu-Petri, A., Nickel, B.,
Nagel, S., Talbi, E.H., El Hajraoui, M.A., Amzazi, S., Hublin, J.J., Pääbo, S., Schiffels, S., Meyer, M.,
Haak, W., Jeong, C., & Krause, J. (2018). Pleistocene North African genomes link near Eastern and sub-
saharan African human populations. Science, 360(6388), 548-552.
Van Gerven, D. P., Beck, R., & Hummert, J. R. (1990). Patterns of enamel hypoplasia in two medieval
populations from Nubia's Batn el Hajar. American Journal of Physical Anthropology, 82(4), 413-420.
Van Gerven, D. P., Sandford, M. K., & Hummert, J. R. (1981). Mortality and culture change in Nubia's
Batn el Hajar. Journal of Human Evolution, 10(5), 395-408.
Van Gerven, D. P., Sheridan, S. G., & Adams, W. Y. (1995). The health and nutrition of a medieval
Nubian population: the impact of political and economic change. American Anthropologist, 97(3), 468-
480.
Vigilant, L., Stoneking, M., Harpending, H., Hawkes, K., & Wilson, A. C. (1991). African populations and
the evolution of human mitochondrial DNA. Science, 253(5027), 1503-1507.
Volk, A. A., & Atkinson, J. A. (2013). Infant and child death in the human environment of evolutionary
adaptation. Evolution and Human Behavior, 34(3), 182-192.
Waldron, H. A. (2000). The study of the human remains from Nubia: the contribution of Grafton Elliot
Smith and his colleagues to palaeopathology. Medical history, 44(3), 363-388.

146
Waldron, T. (2008). Palaeopathology. Cambridge: Cambridge University Press.
Walker, P. L., Bathurst, R. R., Richman, R., Gjerdrum, T., & Andrushko, V. A. (2009). The causes of
porotic hyperostosis and cribra orbitalia: A reappraisal of the iron‐deficiency‐anemia hypothesis. American
Journal of Physical Anthropology: The Official Publication of the American Association of Physical
Anthropologists, 139(2), 109-125.
Walker, P. L., Dean, G., & Shapiro, P. (1991). Estimating age from tooth wear in Archaeological
Populations. Advances in Dental Anthropology, 169-178.
Wall, J. D., & Kim, S. K. (2007). Inconsistencies in Neanderthal genomic DNA sequences. PLoS
Genetics, 3(10), e175.
Weir, B. S., & Cockerham, C. C. (1984). Estimating F‐statistics for the analysis of population structure.
Evolution, 38(6), 1358-1370.
Weissensteiner, H., Pacher, D., Kloss-Brandstätter, A., Forer, L., Specht, G., Bandelt, H. J., Kronenber,
F., Salas, A. & Schönherr, S. (2016). HaploGrep 2: mitochondrial haplogroup classification in the era of
high-throughput sequencing. Nucleic Acids Research, 44(W1), W58-W63.
Welsby, D. 2016. Observations on the graves of the Medieval period in the SARS concession at the
Fourth Cataract’, in A. Łajtar, A. Obłuski and I. Zych (eds), Aegyptus et Nubia Christiana. The
Włodzimierz Godlewski Jubilee Volume on the Occasion of his 70th Birthday, 613-628.
Welsby, D. A. (1996). The Kingdom of Kush: The Napatan and Meroitic Empires. London: British Museum
Press.
Welsby, D. A. (2002). The Medieval Kingdoms of Nubia: Pagans, Christians and Muslims along the
Middle Nile. London: British Museum Press.
Wendorf, F., Schild, R., & Schild, R. (1976). Prehistory of the Nile Valley (Studies in Archeology). New
York: Academic.
Wendorf, F., Schild, R., Close, A. E., Hillman, G. C., Gautier, A., Van Neer, W., Donahue, D.J., Jull, A.J.T,
& Linick, T. W. (1988). New radiocarbon dates and late Palaeolithic diet at Wadi Kubbaniya,
Egypt. Antiquity, 62(235), 279-283.
White, C. D., & Schwarcz, H. P. (1994). Temporal trends in stable isotopes for Nubian mummy tissues.
American Journal of Physical Anthropology, 93(2), 165-187.
White, T. D., & Folkens, P. A. (2005). The Human Bone Manual. New York: Elsevier.
Wolbach, S. B., & Bessey, O. A. (1941). Vitamin A deficiency and the nervous system. Archives of
Pathology, 32, 689-722.
Ziesemer, K. A., Ramos‐Madrigal, J., Mann, A. E., Brandt, B. W., Sankaranarayanan, K., Ozga, A. T.,
Hoogland, M., Hofman, C.A., Salazar-Garcia, D.C., Frohlich, B., Milner, G.R., Stone, A.C., Aldenderfer,
M., Lewis, C.M., Hofman, C.L., Warinner, C., & Schroeder. (2019). The efficacy of whole human genome
capture on ancient dental calculus and dentin. American Journal of Physical Anthropology, 168(3), 496-
509.
Zurawski, B. (2007). Where the water is crying. Survey and excavations in Shemkhiya, Dar el-Arab (Suegi
el-Gharb) and Saffi Island carried out by the Polish Expedition to the Fourth Cataract in the winter of
2004/2005. Preliminary report. In Näser, C., & Lange, M. (Eds.). Proceedings of the Second International
Conference on the Archaeology of the Fourth Nile Cataract, Berlin, August 4th-6th, 2005 (Vol. 23). Otto
Harrassowitz Verlag.

147