To the Caribbean and Beyond:
Complete Mitogenomes of Ancient Guinea Pigs
(Cavia porcellus) as a proxy for Interaction in
the Late Ceramic Age.
Edana Lord
A thesis submitted in fulfillment of the requirements for the Degree of Master of Science in Anatomy at the University of Otago, Dunedin, New Zealand.
22nd March 2017 Abstract
The late ceramic age in the Caribbean occurred from AD500 and was associated with increased interaction between the islands and mainland South America. The domestic guinea pig (Cavia porcellus) was introduced to the Caribbean post-AD500 through human transportation, along with the peccary, agouti, armadillo and opossum. The main goal of this thesis was to use guinea pigs as a commensal model to identify likely human migration routes and interaction spheres within the wider Caribbean region, using complete mitogenomes of ancient guinea pigs. Furthermore, the thesis aimed to identify the number of guinea pig introductions to the Caribbean and locate the origins of early historic European and North American guinea pigs.
A total of 23 ancient and two modern complete mitogenomes were obtained. The identified haplogroups indicate that two introductions of guinea pigs to the Caribbean occurred, and that ancient Caribbean guinea pigs were most closely related to those from Peru. The first introduction occurred through previously established trade networks from coastal Colombia to Puerto Rico. A second introduction occurred post-
AD1000 to the southern Lesser Antilles, likely as a result of coastal migrations through Peru and the northern coasts of Colombia and Venezuela into the Caribbean.
In addition, a potential origin for European domestic guinea pigs was found to be in the Andean region encompassing southern Peru and northern Bolivia. A historic period North American guinea pig was found to have come from the Caribbean.
Furthermore, it was identified that there may have been two domestication events on the mainland of South America, which contributed to the current population of domestic guinea pigs. This study is the first to use next-generation sequencing to obtain complete mitogenomes of a commensal animal to investigate prehistoric
ii interaction in the pan-Caribbean region, and results are in agreement with current archaeological and genetic evidence for human mobility into the Caribbean.
iii Acknowledgements
First and foremost, I would like to thank my supervisor, Lisa Matisoo-Smith, for the supervision of this project and valuable advice and inspiration throughout. Secondly, to Susan deFrance, Michelle LeFebvre, Katherine Moore, Martha Zierden, and
Fabienne Pigiere for kindly providing the samples of guinea pig and archaeological information, on behalf of the archaeologists and to Sian Halcrow for providing the modern guinea pig. I would also like to thank my advisory committee, Stephen Bunn and Michael Knapp, for their valuable suggestions during the year.
I wish to thank Catherine Collins, for assistance in all aspects of the project from laboratory work to phylogenetic analyses and editing. Additionally, I wish to acknowledge Anna Gosling, Olga Kardailsky for the teaching of, and assistance with laboratory procedures. To Sophia Cameron-Christie for developing the bioinformatic pipeline and answering my many questions regarding bioinformatic analyses! For assistance in proof reading and editing, I wish to thank Eric Lord and Patrick Lepine.
To Denise, Natalie, Hanna, Kate, Susie and the rest of Room 333, thank you for your valuable support, laughs and ice cream during this project. Finally, I would like to thank my family and Jay for your unconditional support throughout this project.
iv Table of Contents
Abstract ...... ii
Acknowledgements ...... iv
Table of Contents ...... v
List of Tables ...... viii
List of Figures ...... ix
List of Abbreviations ...... xi
Chapter 1: Introduction ...... 1
Chapter 2: Caribbean Prehistory ...... 5
2.1 Overview ...... 5
2.2 Archaic Period ...... 9
2.3 Early Ceramic Period ...... 10
2.4 Late Ceramic Period ...... 14
2.5 Regionalisation and Interaction ...... 14
2.6 Introduction of Animals ...... 19
2.7 Strontium Isotopic Analyses of Human and Faunal Remains ...... 21
Chapter 3: Genetic Variation ...... 24
3.1 Overview ...... 24
3.2 Development of Genetic Variation Research ...... 25
3.3 Genetic Studies in the Caribbean ...... 27
3.3.1 Modern DNA Studies ...... 27
3.3.2 Ancient DNA Studies ...... 29
3.4 Commensal Approach ...... 32
3.4.1 Guinea Pigs as a Commensal Model for Understanding Prehistoric
Human Interaction in the Caribbean ...... 34
v 3.5 Research Aims ...... 39
Chapter 4: Methodology ...... 40
4.1 Ancient DNA: Uses and Limitations ...... 40
4.1.1 Contamination and Sequence Authenticity ...... 40
4.1.2 The Application of Next Generation Sequencing for Ancient DNA ...... 42
4.1.3 New Methodologies and Limitations ...... 43
4.2 Modern DNA Processing ...... 44
4.2.1 Sample Collection and DNA Extraction ...... 44
4.2.2 Long Range PCR Amplification ...... 44
4.2.3 Library Preparation and Sequencing of a Modern Tissue Sample ...... 45
4.2.4 Bait Preparation ...... 46
4.3 Ancient DNA Processing ...... 47
4.3.1 Sample Collection ...... 47
4.3.2 Sample Preparation ...... 48
4.3.3 DNA Extraction ...... 54
4.3.4 Double Stranded Library Preparation ...... 55
4.3.5 Hybridisation Capture ...... 57
4.3.6 Ancient DNA Sequencing ...... 58
4.4 Bioinformatics ...... 59
4.5 Phylogenetic Analyses ...... 60
Chapter 5: Results ...... 63
5.1 DNA Preservation and Authenticity ...... 63
5.1.1 Sequence Coverage ...... 63
5.1.2 Contamination Analysis ...... 70
5.1.3 Damage Patterns and Read Lengths ...... 70
vi 5.2 Nucleotide Diversity ...... 72
5.3 Phylogenetic Analyses ...... 75
5.3.1 Median Joining Network of Complete Mitogenomes ...... 75
5.3.2 Bayesian and Maximum Likelihood Tree of Complete Mitogenomes .... 77
5.3.3 Cytochrome B Maximum Likelihood Tree ...... 77
5.3.4 Cytochrome B Temporal Network ...... 80
Chapter 6: Discussion ...... 82
6.1 Preservation of Ancient DNA ...... 82
6.2 Possible Misidentification of Guinea Pig Samples ...... 83
6.3 Domestication of Guinea Pigs ...... 90
6.4 Complete Mitogenomes ...... 91
6.5 Guinea Pig Origins and Implications for Late Ceramic Age Interaction ...... 93
6.5.1 Origins of Caribbean Guinea Pigs ...... 93
6.5.2 Historic Movements of Guinea Pigs ...... 99
6.6 Guinea Pigs as a Caribbean Commensal Model ...... 101
Chapter 7: Conclusions and Future Directions ...... 103
7.1 Conclusions ...... 103
7.2 Future Directions ...... 105
Appendix ...... 108
Reference List ...... 116
vii List of Tables
Table 2.1: Summary of ceramic styles present in the Caribbean from 500 BC...... 15
Table 4.1: Primers created for Long Range PCR amplification of the complete Cavia
porcellus mitochondrial genome...... 45
Table 4.2: Archaeological specimens analysed in the study...... 50
Table 5.1: Summary statistics for PCR amplification and sequencing of ancient
(n=45) and modern (n=2) samples...... 64
Table 5.2: High Coverage ancient and modern samples used in phylogenetic analyses
(n=25)...... 66
Table 5.3: Additional archaeological samples used in analyses of the Cytochrome B
region (n=8)...... 68
Table 5.4: Nucleotide diversity statistics for each geographic group...... 74
Table 8.1: Genbank accession numbers of samples used in CytB analyses...... 108
Table 8.2: Post-sequencing data for complete mitogenomes from archaeological and
modern samples...... 111
Table 8.3: Nucleotide diversity (π) across the complete mitochondrial genome.
Maximum nucleotide diversity is highlighted in red...... 114
Table 8.4: Haplogroups in Temporal Network of Cytochrome B sequences of modern
and ancient domestic guinea pig...... 115
viii List of Figures
Figure 2.1: Map of islands in the pan-Caribbean area mentioned in the text...... 6
Figure 2.2: Culture areas discussed in the text...... 7
Figure 2.3: Movement of raw and finished materials in the Early Ceramic Age
(500BC-AD500)...... 12
Figure 2.4: Interaction in the Late Ceramic Age (AD500-1492)...... 18
Figure 3.1: From LeFebvre and DeFrance (2014): Distribution of archaeological sites
in the Caribbean containing domestic guinea pig remains...... 35
Figure 4.1: Map of archaeological sites included in the analysis...... 49
Figure 4.2: Specimen MS10653 (Llusco Site, Bolivia), right hand side portion of a
mandible...... 54
Figure 5.1: Coverage plots of the complete mitochondrial genome...... 69
Figure 5.2: Nucleotide frequency at the 5' and 3' ends of reads for sample MS10625,
shown as an example...... 71
Figure 5.3: Read length distribution for MS10625...... 72
Figure 5.4: Nucleotide diversity (π) plotted on coverage distribution for one of the
Colombian samples (MS10639)...... 74
Figure 5.5: Median Joining network of ancient, historic and modern complete Cavia
porcellus mitogenomes...... 76
Figure 5.6: Bayesian Tree with Maximum Likelihood bootstrap support for ancient,
historic and modern complete mitogenomes...... 78
Figure 5.7: Maximum Likelihood Tree (with Bayesian posterior probabilities)
demonstrating the relationship between modern (n=105) and ancient (n=29)
domesticated guinea pig cytochrome B regions...... 79
ix Figure 5.8: Temporal Network of Cytochrome B sequences from modern (n=68) and
ancient (n=29) domestic guinea pigs...... 81
Figure 6.1: Possible late ceramic age translocation routes and timings of the guinea pig into the Greater and Lesser Antilles...... 95
x List of Abbreviations
°C Degrees in celsius π Nucleotide diversity 3' Three prime 5' Five prime A Adenine AD Anno domini aDNA Ancient DNA BC Before Christ BIC Bayesian Information Criteria bp Base pairs BWT Bind and Wash Buffer with Tween 20 BWA Burrows-Wheeler Aligner C Cytosine Cyt B Cytochrome B DNA Deoxyribonucleic acid dNTP Deoxyribonucleotide triphosphates EDTA Ethylenediaminetetraacetic acid FASTA File format for DNA sequences g Gram g Gravitational acceleration G Guanine GTR Nucleotide substitution model GuSCN Guaninadine Thiocyanate HKY Nucleotide substitution model HVR Hypervariable region HWT Hot Wash buffer with Tween 20 k Average number of nucleotide differences between two samples kb kilobases M Molar mg Milligram
MgCl2 Magnesium Chloride
xi mM Millimolar MJ Median-joining mL Millilitre ML Maxium Likelihood mtDNA Mitochondrial DNA NaCl Sodium Chloride NADH1 Nicotinamide adenine dinucleotide subunit 1 NADH5 Nicotinamide adenine dinucleotide subunit 5 ng Nanogram NGS Next Generation Sequencing NRY Non-recombining Y chromosome PCR Polymerase chain reaction qPCR Quantitative PCR RFLP Restricted fragment length polymorphism RNA Ribonucleic acid rRNA Ribosomal RNA SNP Single nucleotide polymorphism T Thymine TE Tris-EDTA TET Tris-EDTA with Tween TIM2 Nucleotide substitution model
TM Melting temperature tRNA Transfer RNA U Unit USA United States of America VCF Variant Call File µL Microlitre µM Micromolar
xii Chapter 1: Introduction
Studying patterns of mobility and interaction in past populations is important for understanding how cultures moved through time and space. Traditional methods for studying past migration of humans include analyses of material culture such as stone tools, ceramics, and architecture, as well as through the spread of language families
(Gray et al., 2009; Hofman et al., 2010). Recently, DNA from modern and ancient human populations has been used to support or challenge theories of migration and interaction established using traditional archaeological and linguistic methodologies
(Cann et al., 1987; Haak et al., 2010; Heupink et al., 2016; Westaway and Lambert,
2014). By using non-recombining, single-locus elements of the genome such as the mitochondrial genome (maternally inherited) or the Y-chromosome (paternally inherited) and very recently, full nuclear genomes, the migration and interaction of humans throughout the past has been investigated (Duggan et al., 2014; Llamas et al.,
2017; Sankararaman et al., 2016).
As an alternative to and to support archaeological and human genetic data, the commensal approach has been developed. The commensal approach can be defined as the “Examination of biological variation in species that have been transported by or with humans” (Storey et al., 2013, pg 38), where the genetic relationships of animals, plants, or microorganisms that were transported by humans, intentionally or otherwise, can be used as a proxy for determining human interaction and migration (Matisoo-
Smith, 1994; Storey et al., 2013). The commensal model, when applied to archaeological remains, can be used to address a range of questions such as: the timings of migration, the presence of multiple introductions, and identifying possible
1 interaction spheres in prehistory. This approach is particularly useful in island environments, as the water barrier between islands generally prevents the natural distribution of land mammals from reaching islands without human involvement, although the commensal approach has also been used in continental contexts (Thaw et al., 2004).
The Pacific region has been the focus for a number of commensal approach studies using both modern and ancient variation, and has provided researchers with a more complete view of the complex patterns of mobility and interaction of humans that led to the transportation and establishment of non-native fauna and flora throughout the region. Commensal species studied in this region include the Pacific rat (Rattus exulans), dog (Canis familiaris), pig (Sus scrofa), chicken (Gallus gallus), and various plant species (Greig et al., 2015; Larson et al., 2010; Matisoo-Smith, 1994; Matisoo-
Smith and Robins, 2004; Savolainen et al., 2004; Seelenfreund et al., 2011; Storey et al.,
2007). Some of these species were clearly transported intentionally as part of a transported landscape, and others may have been introduced unintentionally (Storey et al., 2013).
In the Pacific, commensal studies have used either modern or ancient DNA, or a combination of both to infer patterns of migrations and interaction. A limitation with using modern DNA from commensal species is that modern populations may not be direct descendants of ancient populations due to more recent introductions and admixture of the same species, and therefore do not represent the migrations undertaken in prehistory (Matisoo-Smith, 2009). Thus, the use of ancient DNA from archaeological remains of commensal species allows for a better understanding of
2 prehistoric migration and can determine if there is continuity between the ancient and modern populations of a commensal species. Until recently, the study of ancient DNA from archaeological samples in tropical island environments, like the Pacific, has been difficult due to the degraded nature of the DNA in these contexts. Studies have been limited to analyses of very short fragments of DNA, typically from the mitochondrial genome, which has provided limited information regarding variation within a limited geographic region, such as East Polynesia (Matisoo-Smith and Barnes, 2006).
With the advancement of DNA sequencing technologies such as next-generation sequencing (NGS), the use of ancient DNA to explore past migrations and interactions has become more feasible (Millar et al., 2008). The application of NGS has allowed smaller fragments to be sequenced, which is useful for the short, damaged fragments typical of ancient DNA (Willerslev et al., 2007). Combined with the use of capture protocols (Marcic et al., 2010), the use of NGS technology has allowed for the sequencing of complete mitogenomes, or even in some cases complete nuclear genomes from the small, degraded fragments of ancient DNA (Reich et al., 2010). Due to the read depth attainable through NGS, it has also allowed for the characterisation of damage patterns associated with ancient DNA, as well as deciphering contamination in the sequences from exogenous sources. Furthermore, the cost of sequencing has decreased, allowing these technologies to become more affordable.
Within the Caribbean region, archaeological and modern human genetic data have recently been used to unravel the complex prehistory of this region (Hofman et al.,
2014a; Martinez-Cruzado, 2013). Modern human genetic data, however, is confounded by historical and recent admixture between Native Americans, Europeans and Africans
3 (Vilar et al., 2014), and few ancient human DNA studies have been undertaken in the
Caribbean. With the application of the commensal model and NGS, further insight into human mobility and interaction in this region can be uncovered as animals and plants native to South America were transported to the Caribbean in prehistory (Giovas et al.,
2012; Jimenez, 2007; LeFebvre and deFrance, 2014). To date, NGS methodology has not yet been applied to questions of migration and interaction of in the Caribbean region.
The present study aims to analyse the relationships between archaeological specimens of domestic guinea pig (Cavia porcellus) from the Caribbean, South America, North
America and Europe, using complete mitogenomes. As guinea pigs were introduced to the Caribbean via human migration, these relationships will be used to infer mobility and interaction in pan-Caribbean prehistory. This is the first study that aims to use NGS to obtain complete mitogenomes of a commensal animal in the Caribbean region to infer prehistoric patterns of mobility and interaction.
The literature review presented in Chapter 2 will discuss prehistoric migration and interaction networks in the broader Caribbean region inferred from existing studies of archaeology and isotopes. The use of human genetic variation to infer migration and mobility in the Caribbean will be discussed, along with the merits of using archaeological remains of the guinea pig to undertake a commensal study in the
Caribbean in Chapter 3. Finally, the benefits and limitations of using ancient DNA and the revolution of NGS for interpreting the past will be discussed in Chapter 4.
4 Chapter 2: Caribbean Prehistory
2.1 Overview
Understanding the prehistory of the Caribbean and surrounding regions is important for identifying spheres of interaction and potential origins of migration for the pre-
Columbian (prior to the arrival of Columbus) people of the Caribbean. The Caribbean region is split into three major areas: the Greater Antilles, and the Northern and
Southern Lesser Antilles, though the Caribbean Sea also reaches to Central America and the northern coast of South America (Figure 2.1). The Greater Antilles encompasses the islands from Cuba to Puerto Rico (Fitzpatrick, 2006). The Northern
Lesser Antilles includes the region from the Virgin Islands to Guadeloupe, with the
Southern Lesser Antilles comprising the area from Dominica to Trinidad. Most islands within the island chain are intervisible, meaning that each island can be seen from neighbouring islands (Hofman et al., 2014a).
The area including the Caribbean Islands and the coasts of Central and northern South
America has recently been referred to as the “pan-Caribbean” region, in order to aptly describe the level of interaction between these areas during pre-Columbian times
(Hofman and Bright, 2010; Hofman et al., 2014a; Rodriguez-Ramos et al., 2010).
Additionally, the area encompassing Central America from Guatemala to Colombia has been referred to as the Isthmo-Colombian area, which has distinct cultural ties to the
Caribbean in prehistory (Figure 2.2). The continental regions bordering the Caribbean
Sea were instrumental in pre-Columbian times in providing access and trade for materials not present in the Caribbean Islands (Rodriguez-Ramos et al., 2010), thus broadening the geographical scope of the Caribbean area allows for better
5 THE BAHAMAS!
CUBA! TURKS & CAICOS ISLANDS!
DOMINICAN HAITI! VIRGIN CAYMAN ISLANDS! REPUBLIC! ISLANDS! ANGUILLA! (HISPANIOLA)! JAMAICA! PUERTO RICO! ST MARTIN! VIEQUES!
Greater Antilles! ST KITTS & NEVIS! ANTIGUA! MONTSERRAT! Northern GUADELOUPE! Lesser Antilles!
DOMINICA ! HONDURAS! CARIBBEAN SEA! MARTINIQUE! Southern Lesser Antilles! ST LUCIA! ST VINCENT! BARBADOS! NICARAGUA! ARUBA! BONAIRE! CARRIACOU!
CURACAO! GRENADA!
TOBAGO!
TRINIDAD! COSTA RICA! COLOMBIA! VENEZUELA!
Figure 2.1: Map of islands in the pan-Caribbean area mentioned in the text. All Base maps in this thesis are from ArcGIS® software by ESRI.
6 Caribbean&
CARIBBEAN SEA!
Amazonian& Isthmo- Colombian&
Andean&
Figure 2.2: Culture areas discussed in the text. Red: Isthmo Colombian region, Yellow: Caribbean region, Green: Amazonian region, Blue: Andean region.
7 interpretations of the movements of people in the past.
The prehistory of the Caribbean region has been divided into four phases: the pre-
Arawak archaic, the Arawak archaic, and the Early and Late Ceramic phases
(Fitzpatrick, 2006; Rouse, 1989b; Rouse, 1989a; Petersen et al., 2004). The Late
Ceramic phase culminated in the arrival of Columbus in AD1492, ending the prehistoric period in the Caribbean. The four prehistoric periods have been defined according to the presence or absence of particular ceramic styles, raw and processed materials such as various forms of jade, and other archaeological remains of the indigenous Caribbean cultures.
A notable scholar in Caribbean prehistory, Irving Rouse, was the first to outline
Caribbean prehistory and establish the initial ceramic sequence, with the main outcome being to classify cultures in space and time (Rouse, 1989b). Rouse (1989b) outlined four major cultures; the first (I) corresponds to the pre-ceramic cultures, which existed in the pan-Caribbean region around 5000-1000/300BC. The next three of Rouse’s periods (II, III, IV) defined different ceramic cultures present in the Caribbean over the time period from 1000/300BC until European contact. However, a major flaw in his original findings was the lack of a concept that two cultures could co-exist in the same geographic area, which led to some misclassifications of the timings and definition of the proposed ceramic cultures (Petersen et al., 2004). The timings and cultural definitions of these periods have since been refined with the discovery and analysis of more material (Fitzpatrick et al., 2008; Garcia-Casco et al., 2013; Hofman et al., 2008;
Keegan, 2000; Knippenberg, 2007; Rouse, 1993) and the use of radiocarbon dating
8 (Fitzpatrick, 2006; Fitzpatrick and Giovas, 2011), which have led to a better understanding of Caribbean prehistory.
2.2 Archaic Period
The first movement of people into the Caribbean region began around 4000BC in Cuba and Hispaniola (the island encompassing Haiti and the Dominican Republic, Figure
2.1) and was due to the expansion of Central American lithic societies (Fitzpatrick,
2006; Hofman et al., 2014a; Petersen et al., 2004). Sites from this period are relatively rare compared to later periods. Artefacts include flaked-stone tools (Keegan, 1994).
These societies were hunter-gatherer societies and lacked ceramic production.
The first movement into Puerto Rico and the Lesser Antilles occurred approximately
2500-2000BC and was the result of expanding South American hunter-gatherer societies known as the ‘Arawak’, thus this period is referred to as the Arawak Archaic.
Evidence of these groups is seen earliest in northern South America and there is evidence of their arrival in Trinidad around 2500BC, before moving northwards into the Greater and Lesser Antilles. Puerto Rico and the Virgin Islands, in the Greater
Antilles region of the Caribbean, were the first islands settled by these hunter-gatherer groups (Fitzpatrick, 2006). Evidence of this migration in the Southern Lesser Antilles is almost absent except for two sites on Martinique (Allaire and Mattioni, 1983) and possibly one on St Vincent (Callaghan, 2003).
The material culture and language associated with the Arawak is thought to have its origins in the Amazonian region, and spread northwards through South America to the
Caribbean (Keegan, 1994; Rouse, 1989b; Walker and Ribeiro, 2011). These
9 populations lacked ceramics, though had developed ground-shell and stone tools and had personal adornments in the form of serpentinite beads (Rodriguez-Ramos et al.,
2010). These Arawak populations relied on the marine economy for subsistence as demonstrated by the abundance of mollusk remains in middens (archaeological waste sites) (Keegan, 1994). There is also evidence that manioc, maize and sweet potatoes were transported from mainland South America to the Caribbean and cultivated, likely as a supplement to the marine-based economy (Jiménez, 2007).
2.3 Early Ceramic Period
Ceramic cultures, thought to have developed in northern South America, spread into the
Caribbean from Venezuela approximately 500BC (Fitzpatrick, 2006; Petersen et al.,
2004), 500 years later than previously suggested by Rouse (1989b). The early ceramic cultures are defined by two ceramic styles, the first is the Saladoid (500BC-AD500) and towards the end of the early ceramic period, we see the intrusion of the Barrancoid style (AD350-400) (Hofman, 2013; Keegan, 2000). Ceramics from this period are found throughout the Antillean island chain, from Puerto Rico to Trinidad. Based on this widespread ceramic distribution, previous studies had suggested that the early ceramic colonisers followed a south-to-north stepping stone model from northern South
America through the Antillean island chain (Keegan, 2000; Rouse, 1989b). However, a recent compilation of radiocarbon dates and a chronometric hygiene approach applied by Fitzpatrick (2006) has shown that the first islands to be settled in this period were the most northern islands: Puerto Rico, Vieques, St Martin and Monsterrat. Computer simulations of voyaging routes have further supported that the most plausible route was to voyage to the northern islands (Callaghan, 2003; Callaghan and Bray, 2007).
10 Lithic artefacts abundant in this period were flint tools, greenstone (“indigenous gray- green, partly recrystallised mudstone” Hofman et al., 2007, pg 248) axes, and jadeitite and serperntinite pendants (Knippenberg, 2007; Rodriguez-Ramos et al., 2010). Flint from Long Island, Antigua was distributed from Nevis to Guadeloupe as nodules, and finished products from Puerto Rico to Martinique (Knippenberg, 1999) (Figure 2.3).
Greenstone axes were predominantly produced from a source in St Martin
(Knippenberg, 2007), and were distributed throughout the northern Lesser Antilles and
Puerto Rico. Jadeitite axes from Guatemala or Cuba are also present in Puerto Rico, as well as personal adornments, though no evidence of production of this material is seen in the Greater Antilles, suggesting importation of finished products from the Isthmo-
Colombian region (García-Casco et al., 2009; Rodriguez-Ramos et al., 2010).
Further evidence of contact with the Isthmo-Colombian region is seen through the distribution of personal adornments and guanin (a gold-copper alloy). Personal adornments such as serpentinite pendants from the Greater Antilles show similarities in manufacture and iconography to jadeitite pendants found throughout the Isthmo-
Colombian region dating to the same period (Hofman et al., 2010; Rodriguez-Ramos et al., 2010). The presence of guanin also suggests interaction with the Isthmo-
Colombian region during the early ceramic period as guanin metallurgy was developed in Colombia and spread northwards through the Isthmo-Colombian region and into the
Caribbean from AD100 (Cooper et al., 2008; Siegel and Severin, 1993). No evidence for guanin is seen outside of the Isthmo-Colombian region and the Greater Antilles. In addition, interaction between the Lesser Antilles and Puerto Rico, and mainland South
America is seen through the movement of serpentinite. Serpentinite from sources in
11 TURKS & CUBA! CAICOS ISLANDS!
ANGUILLA! HISPANIOLA! PUERTO RICO! JAMAICA! ST MARTIN! Greater Antilles! ST KITTS ANTIGUA! Northern & NEVIS! Lesser Antilles! GUADELOUPE! GUATEMALA! HONDURAS! MARTINIQUE!
Southern ST LUCIA! Lesser Antilles! NICARAGUA! CARIBBEAN SEA! GRENADA! TOBAGO!
COSTA RICA! TRINIDAD!
VENEZUELA! COLOMBIA! PANAMA! Key:% Guatemalan%Jadeitite% St%Martin%Greenstone% Serpentinite% Long%Island%Flint% Guanin%
Figure 2.3: Movement of raw and finished materials in the Early Ceramic Age (500BC-AD500).
12 Puerto Rico and northern South America was used for manufacture of pendants and beads from Puerto Rico to Grenada, though evidence of their manufacture does not occur in sites in this area thus it appears that they were distributed as finished products.
From AD350-400, the development of interaction zones within the Caribbean and between the Caribbean and northern South America becomes evident (Hofman et al.,
2010; Rodriguez-Ramos et al., 2010). Firstly, Barrancoid ceramics were introduced to the Lesser Antilles from the South American mainland, through Trinidad and Tobago.
Stylistic similarities in ceramics are seen between Grenada and Guadeloupe, Antigua and Saba, and the Virgin Islands and Puerto Rico, suggesting regionalisation within each of these areas (Hofman and van Gijn, 2008). The distribution areas of St Martin limestone and Long Island flint increased southward during this period, though distribution to Puerto Rico subsides (Rodriguez-Ramos et al., 2010). Finally, three- pointed artefacts, which are objects made from shell or stone (calcirudite), likely used for a variety of ritual purposes, first appeared in Puerto Rico (Siegel, 1989). These are thought to have been an indigenous development within Puerto Rico, which later spread to the northern Lesser Antilles (Breukel, 2013). Due to the regionalisation of ceramic styles and the development of three-pointers in the Greater Antilles, along with the cessation of Lesser Antillean raw materials to this region, it suggests a change in culture. This was possibly due to interaction with expanding complex societies from mainland South America (Stanish, 2001). The early ceramic phase ended with the development of new ceramic cultures, increased population size, and further development of and interaction between cultures in the pan-Caribbean region.
13 2.4 Late Ceramic Period
The late ceramic period occurred from AD500-1500 (Fitzpatrick, 2006; Hofman, 2013;
Keegan, 2000; Wilson, 2007). This period saw the development of new pottery styles, an increase in regionalisation of raw materials, and the introduction of non-native animals demonstrating interaction to the western Greater Antilles and between the
Lesser Antilles and South America. Post-AD1000 complex, hierarchal societies developed within the Greater Antilles as a result of an increase in population size and interaction (Keegan, 2000). Population size increased in the Lesser Antilles as well, although the development of complex societies is not seen in this region (Hofman,
2013), implying that differing settlement processes and interaction may have occurred between the Greater and Lesser Antilles. In AD1492, a voyage from Spain led by
Christopher Columbus made landfall on Puerto Rico, signaling the end of the prehistoric period. During the years AD1492-1503, Columbus made repeated voyages to Puerto Rico and Hispaniola. Numerous arrivals from Spain and the Netherlands soon followed, having a large impact on the cultures and peoples of the Caribbean and South
American regions (Deagan, 1988; Keegan, 1996).
2.5 Regionalisation and Interaction
During the late ceramic period, regionalisation of ceramic styles within and between the Greater and Lesser Antilles occurred (Fitzpatrick, 2006; Fitzpatrick et al., 2008;
Hofman et al., 2010). Within the Greater Antilles, the Saladoid was gradually replaced by the Ostionoid ceramic culture (Table 2.1). Elanan Ostionoid, the most dominant of the Ostionoid series, is found in Puerto Rico and the Virgin Islands, indicating a interaction sphere encompassing this area. Post-AD1200 saw the development of the
Chican Ostionoid, thought to represent the “Taino” culture of the Greater Antilles. The
14 Table 2.1: Summary of ceramic styles present in the Caribbean from 500 BC.
Ceramic Style Date Location Northern Colombia, northern Venezuela, Saladoid 500BC-AD500 Puerto Rico, Virgin Islands, Lesser Antilles Trinidad, Tobago, Barrancoid AD350-400 southern Lesser Antilles Guiana, Venezuela, Araquinoid AD500-600 Trinidad Eastern Puerto Rico, Elanan AD600-1200 Virgin Islands Jamaica, Cuba, Ostionan AD600-1200 Hispaniola, Western Puerto Rico Ostionoid Jamaica, Cuba, Mellican AD900-post-1492 Hispaniola Bahamas, Cuba, Chican AD1200-1400 Hispaniola, Puerto Rico, Virgin Islands, Anguilla Troumassan AD500/600-1000 Southern Lesser Antilles Mamoran AD500/600-1500 Northern Lesser Antilles Troumassoid Suazan Guadeloupe to Tobago AD1000-1500 (Caliviny) (St Vincent to Tobago) AD1250-early Guiana, southern Lesser Cayo colonial Antilles
Taino culture developed in Puerto Rico and is characterised by large complex settlements and ceremonial courts (Lalueza-Fox et al., 2001). Artefacts from the Taino culture are found as far south as Trinidad, suggesting widespread interaction in late prehistory (Hofman et al., 2008; Petersen, 1997).
15 In the Lesser Antilles, a new ceramic style, the Troumassoid, developed out of the
Saladoid, and underwent regional diversification with influences from the South
American mainland. In the Southern Lesser Antilles, the Troumassan series developed around AD500-600 and persisted until AD1000. This style was greatly influenced by the Araquinoid style of Trinidad, Venezuela and the Guianas (Hofman et al., 2007).
From AD1000 the Troumassan was replaced by the Suazan style, and this coincides with expansion into previously unoccupied locations in the islands (Boomert, 2000). A notable style within the Suazan phase is the Caliviny, which originates in northern
Venezuela and has a distribution from Tobago to St Vincent (Boomert and Kameneff,
2005).
In the Northern Lesser Antilles, the Saladoid was replaced by the Mamoran
Troumassoid series around AD800-1200, coinciding with an increase in population density. Some more northern areas of the Southern Lesser Antilles show similarities to the Northern Lesser Antillean Mamoran styles, suggesting interaction and fusion of ceramic styles in this region (Hofman and Hoogland, 2004). In addition, ceramics from
Anguilla and Saba show elements of non-local production, though a source was unable to be identified (Crock and Petersen, 2004). Towards the end of the late ceramic phase
(post AD1250), an introduction of the Cayo style is seen in the southern Lesser Antilles.
This style is thought to originate in the Guiana region and represents the migration of the ‘Island Carib’ people into the Caribbean region (Hofman et al., 2014a; Hofman et al., 2010; Petersen, 1997).
The distribution of St Martin greenstone axes and Long Island flint tools became concentrated within the Anguilla-Guadeloupe region, although St Martin greenstone is
16 found at low frequency outside of this area (Figure 2.4) (Hofman et al., 2007;
Knippenberg, 2007). The reduction in distribution of these raw materials demonstrates further regionalisation within the Northern Lesser Antilles (Hofman et al., 2010;
Knippenberg, 2007; Rodriguez-Ramos et al., 2010). In addition, communities on
Anguilla and Saba became involved in the production and distribution of greenstone tools. Post-AD500 saw the introduction of jadeitite artefacts into Hispaniola, and the
Lesser Antilles (to St Lucia) (Rodriguez-Ramos et al., 2010). The introduction of
Cuban and Hispaniolan sources of jadeitite into the exchange networks occurs from
AD1000, coinciding with the introduction of jadeitite artefacts to the Bahamas and
Jamaica (Allsworth-Jones, 2008; Carlson, 1995). This likely reflects wider interaction networks that occurred with the development of the Taino society in the Greater
Antilles.
The movement of guanin objects into the Greater Antilles continued in the late ceramic age, again reflecting the importance of interaction with the Isthmo-Colombian region
(Cooper et al., 2008). An increase in three-pointed artefacts is seen in the Northern
Lesser Antilles from AD600-1200, and then in the Greater Antilles from AD800
(Hofman et al., 2008). Manufacture of three-pointed artefacts occurred in St Martin and
Anguilla and they were distributed throughout the Northern Lesser Antilles and Puerto
Rico, reflecting the influence of the Greater Antilles societies on the Northern Lesser
Antilles (Breukel, 2013). Interaction networks to and from the northeastern South
American mainland and the Isthmo-Colombian region were likely instrumental in the development of complex societies that is exhibited in the Antilles in the late ceramic period (Hofman et al., 2010).
17 CUBA! TURKS & CAICOS ISLANDS!
HISPANIOLA! ANGUILLA! PUERTO RICO! ST MARTIN! JAMAICA! ST KITTS ANTIGUA! Greater Antilles! & NEVIS! Northern GUADELOUPE! GUATEMALA! Lesser HONDURAS! Antilles! MARTINIQUE!
Southern ST LUCIA! BARBADOS! NICARAGUA! Lesser CARIBBEAN SEA! Antilles! GRENADA! TOBAGO!
COSTA RICA! TRINIDAD!
VENEZUELA! PANAMA! COLOMBIA!
Key:% Guatemalan%Jadeitite% St%Martin%Greenstone% Cuban%Jadeitite% Long%Island%Flint% Interaction%spheres% Guanin% Caliviny%Ceramics%
Figure 2.4: Interaction in the Late Ceramic Age (AD500-1492).
18 2.6 Introduction of Animals
In addition to the movement of materials and goods, several studies (Giovas et al.,
2012; Laffoon et al., 2015; LeFebvre and deFrance, 2014) have revealed and identified animal species that exist in archaeological sites that are non-native to the islands. These animals must have been transported to the islands via humans as they lack the ability to cross water barriers themselves. These species include: dogs (Canis familiaris), opossum (Didelphis marsupialis), agouti (Dasyprocta sp), pecarry (Tayassu/Pecari sp.), armadillo (Dasypus sp.) and guinea pig (Cavia porcellus). All species apart from the dog are native to South America.
The domesticated dog was brought into South America with the first movement of people into this area (Clutton-Brock, 1988). Dogs were subsequently introduced to the
Caribbean when people expanded into the Caribbean. The dog was likely not used as a food resource due to their appearance in human burials and not in middens (Newsom and Wing, 2004). The agouti and opossum underwent captive management and, along with the guinea pig, pecarrry and armadillo, were used as a food resource (Wing and
Wing, 1995). Due to the low abundance of remains compared to marine taxa, it is suggested that these commensal animals were used to supplement the marine and agricultural economy established in the earlier archaic and ceramic periods (Giovas et al., 2012).
The domestic guinea pig has the widest distribution of the commensal animals, existing in archaeological sites from Jamaica to Curacao (LeFebvre and deFrance, 2014). The greatest concentration of guinea pig remains are seen in Puerto Rico and Vieques, which led LeFebvre and deFrance (2014) to suggest that guinea pigs were introdued to
19 the Greater Antilles and then underwent subsequent dispersal to the Lesser Antilles and the western Greater Antilles. Agouti are seen in high frequency in the Lesser Antilles during the late ceramic age (Giovas et al., 2012). There is some suggestion that they may have been introduced during the early ceramic age (Giovas et al., 2012; Stahl,
2009), however no secure dating of these early remains has been performed. There is a lower frequency of agouti in sites in the northern Lesser Antilles (Newsom and Wing,
2004) and no agouti remains are present in the Greater Antilles, suggesting dispersal from the northern South America. The oppossum has been described in four islands:
Grenada, Carriacou, St Lucia and Guadeloupe, all in the Lesser Antilles. Carriacou has the highest frequency of opossum remains, with seven individuals represented by 28 bones (Giovas et al., 2012). All other sites only report a few bones of opossum
(Newsom and Wing, 2004).
Only a few remains of pecarry and armadillo occur in the Antilles and these have not been as well described as guinea pig, agoutiand opossum. Modified specimens of pecarry found in Vieques date prior to the ceramic age (AD100-400) however, these likely represent the movement of pecarry teeth or bones that had beend used for pendants, not live animals. The transport of live pecarry occurred later in the ceramic age and is predominantly seen in the southern Lesser Antilles, though one occurance is also seen on Puerto Rico (Giovas et al., 2012). The remains of the armadillo are found in the Lesser Antilles in contexts dating post AD500, though only a few remains have been described (Giovas et al., 2012).
These fauna have been described in several sites from mainland South America, particularly Colombia (Correal Urrego, 1990; Garcia, 2012; Rivas, 1999), which
20 suggests a link between this region and the Caribbean. Aside from recent isotopic and genetic research (Giovas et al., 2016; Kimura et al., 2016), little research has been undertaken to determine the exact origins of each of these commensal species.
Determing the origins of these fauna would add to knowledge of interaction and mobility of the ceramic societies of the Caribbean. Several species of these non-native fauna (armadillo (D. novemcinctus), agouti (D. leporina, D. puncuata and D. mexicana), guinea pig, and opossum) are present in the Caribbean today. However, it is unclear if modern populations present in the Caribbean today are descendants of prehistoric populations or more recent introductions during the historic period (Giovas et al., 2012), an issue, that along with determing the origins of non-native fauna in the Caribbean, may be addressed using ancient DNA analyses.
2.7 Strontium Isotopic Analyses of Human and Faunal Remains
Recently, biosphere strontium ratios have been studied throughout the Caribbean region in order to determine whether skeletal remains from sites within this region can be characterised as local or non-local, thus locating possible areas of origin for the ceramic societies (Giovas et al., 2016; Hoogland and Hofman, 2010; Laffoon et al., 2015).
Strontium isotopes are useful in studies of migration as strontium is absorbed by the enamel or bone during the formation of the tissue. Thus differences in strontium ratios between individuals’ teeth and bone indicates differences in location during the formation of their enamel (birth) and bone (death), as enamel is formed in early life and bone is constantly remodelling throughout life (Hodell et al., 2004). The use of strontium isotope ratios has been applied to human (Hoogland and Hofman, 2010) and dog remains from two sites in Guadeloupe dating to AD500-1400 (Laffoon et al., 2015), as well as the agouti and opossum from Carriacou (Giovas, 2016).
21
Strontium ratios suggest a mix of local and non-local human and dog remains on
Gaudeloupe (Hoogland and Hofman, 2010; Laffoon et al., 2015), with non-local remains likely to have originated elsewhere in the Lesser Antilles. Hoogland and
Hofman (2010) state that approximately a quarter of the human samples analysed in the study were classed as non-local, with individuals from this group having multiple different origins. Dogs were also shown to have differing strontium ratios within the population suggesting non-local origins, however, Laffoon et al. (2015) ruled out
Venezuela and Guiana as potential origins for the non-local dogs based on differing strontium ratios to those on the mainland, suggesting the movement of these dogs was limited to within the Caribbean islands. Giovas et al. (2016) concluded that, based on strontium ratios, opossum and agouti were established as living populations on
Carriacou in the Lesser Antilles, and may have been subject to inter-island mobility.
They also attempted to use lead isotopic ratios, but these results were found to be contaminated with modern lead found in the environment, which can alter the isotopic value, making them not representative of the prehistoric values (Giovas et al., 2016).
Caution needs to be applied to isotopic studies as they may be affected by several factors, such as contamination (as demonstrated from modern anthropogenic lead in
Carriacou) and similarity in isotopic profiles across broad regions. Isotopic studies may have limited value in the Caribbean as many islands have similar or identical strontium isotopic profiles. The islands including Les Saintes, Matinique, St Lucia and Tobago in the Southern Lesser Antilles all have similar isotopic profiles, along with St Kitts and
St Martin in the Northern Lesser Antilles as well as Puerto Rico in the Greater Antilles
(Laffoon et al., 2015; Laffoon et al., 2014). However, strontium ratios can be useful for
22 identifying mainland versus island origins, as these strontium profiles differ significantly enough.
To summarise, the picture emerging from recent reviews of the archaeological record in the pan-Caribbean region, is one of complex interaction networks, connecting various groups across the region. Most early archaeological studies focused on a Venezuelan homeland for the ceramic Caribbean cultures based on analyses of ceramics. Through the movement of jadeitite, sperpentinite and guanin, more recent studies have begun to recognise the interactions that may have occurred within the Amazonian and Isthmo-
Colombian regions prior to movement into the Caribbean that also contributed to the cultures that appeared in the Caribbean. Isoptopic methodologies have also recently contributed to further understanding the origins of the Caribbean culture, although were unable to locate origins for human or animals on mainland South America. To supplement knowledge based on archaeological and isotopic methodologies, an additional approach to assess origins of populations moving into the Caribbean is through exploring the genetic variation of humans, both modern and ancient, and their transported animals, as will be discussed in the following chapter.
23 Chapter 3: Genetic Variation
3.1 Overview
Genetic variation, particularly mitochondrial genome variation, is a useful tool for studying population movements and interaction worldwide (Behar et al., 2008; Cann et al., 1987; Kivisild et al., 2005; Macaulay and Richards, 2013; Van Oven and Kayser,
2009; Vigilant et al., 1991). Much of the early research into population movements focused on using modern genetic variation to reconstruct prehistoric migrations and the genetic histories of past populations (Stoneking et al., 1991; Vigilant et al., 1991).
However, this has its limitations due to historic and recent admixture between populations (Xu et al., 2012). Thus studies using ancient DNA of humans, and the plants and animals associated with them (Storey et al., 2013) are becoming more common for providing evidence of prehistoric populations origins and interaction.
Only since the advent of polymerase chain reaction (PCR) in the mid-1980s have studies of ancient DNA (aDNA), i.e. “DNA recovered from archaeological or palaeontological remains” (Hofreiter et al., 2001b), been able to be conducted, due to the ability to amplify short DNA fragments. Ancient DNA studies provide a better picture of migrations and population variation in the past, though analyses are often limited by small sample sizes and highly degraded material. With the recent innovation of powerful Next Generation DNA sequencing technologies, more information can now be analysed from small molecules and degraded sequences of DNA, as well as improvements in being able to classify damage patterns and contamination in ancient
DNA studies (discussed further in Chapter 4). This can allow for more in depth studies
24 of variation and interaction in modern and past populations (Linderholm, 2016; Llamas et al., 2017; Xu et al., 2012).
3.2 Development of Genetic Variation Research
Early studies into molecular variation in humans focused on proteins and ABO blood groups (Hill et al., 1987), and later used restriction enzyme analysis to determine genetic relationships between populations, based on modern data (Wallace et al., 1985).
The advent of DNA sequencing in the late 1970s (Sanger et al., 1977) allowed for variation to be characterised at the DNA sequence level through differences identified between two sequences. Until recently, most studies have focused on short fragments of non-recombining, single-locus elements of genomes. These include the mitochondrial genome (maternally inherited) or the non-recombining portion of the Y chromosome
(NRY; paternally inherited) in order to explore variation (Cann et al., 1987; Underhill et al., 2000).
Regions of the mitochondrial genome with higher mutation rates such as the hypervariable region (HVR) have been used most commonly in phylogenetic studies to determine relationships within a species or population (Stoneking et al., 1991; Kivisild et al., 2005; Vigilant et al., 1991). However, recent studies have demonstrated that complete mitochondrial genome sequences provide better resolution of variation in populations than short regions such as the HVR (Benton et al., 2012; Duggan et al.,
2014; Fu et al., 2013; Greig et al., 2015; Knapp et al., 2012b). Benton et al. (2012),
Duggan et al. (2014), and Knapp et al. (2012b) identified new mitochondrial haplotypes
(sequences that contain a specific set of mutations or variants) in Maori and Polynesian populations by complete mitochondrial genome sequencing that were unable to be
25 identified using only the HVR. Based on this research, it can be concluded that variation exists outside of the HVR that is essential to fully understand the relationships between haplotypes in modern and ancient human populations. The Cytochrome B (Cyt
B) region of the mitochondrial genome has been used to explore variation. Due to the conservation of this region, it is more applicable to determining variation between species and has recently been used for species barcoding analyses (Clare et al., 2007;
Hebert et al., 2003; Yacoub et al., 2015).
Modern and ancient DNA from humans and their associated animals and plants has recently been used to decipher prehistoric migrations and interactions (Heupink et al.,
2016; Kimura et al., 2016; Larson et al., 2010; Matisoo-Smith and Robins, 2004;
Matisoo-Smith et al., 2016). However, due to historical and recent admixture, modern populations may not contain the same variations that were present in past populations
(Xu et al., 2012). Thus analysis of DNA from archaeological remains of past populations can provide a more accurate picture of prehistoric movements and interaction (Benton et al., 2012; Llamas et al., 2017; Knapp et al., 2012b; Reich et al.,
2010). Typically, mitochondrial DNA is used to explore variation between past populations, however recent studies have begun to explore variation and admixture using ancient nuclear genomes (Reich et al., 2010; Sankararaman et al., 2016). Though ancient nuclear genome studies are increasing, mitochondrial genome studies are still the most common tool to assess population origins in ancient DNA studies. Commensal studies using archaeological remains of plants and animals associated with human movement can also aid in creating a broader view of mobility and interaction in the past.
The use of modern and ancient human DNA, along with DNA from commensal animals to explore interaction will be discussed for the Caribbean region.
26 3.3 Genetic Studies in the Caribbean
3.3.1 Modern DNA Studies
Several studies using genetic tools such as those described above, have attempted to determine the origins of populations that settled within the Caribbean Islands and possible interaction spheres in late prehistory (Kimura et al., 2016; Lalueza-Fox, 1996;
Martínez-Cruzado, 2010; Martinez-Cruzado et al., 2001; Martinez-Cruzado et al., 2005;
Mendizabal et al., 2008; Toro-Labrador et al., 2003; Vilar et al., 2014; Wallace and
Torroni, 2009). As with the early archaeological studies, most of these studies have focused on uni-linear patterns of migration from one source population rather than multi-linear migrations from around the pan-Caribbean region. Also, much of the genetic understanding of the Caribbean comes from studies focused in Puerto Rico and the Greater Antilles alone, which does not allow for a complete picture of prehistoric interactions based on modern genetic data as based on archaeological evidence, the settlement histories and interaction patterns of prehistoric cultures of the Greater and
Lesser Antilles differ. More genetic data is required from throughout the Greater and
Lesser Antilles in order to form a more cohesive picture of interaction in the Caribbean region.
The first genetic studies of modern populations in the Caribbean and South America determined that there were four of the five (A, B, C, D, X2a) Native American mitochondrial lineages present in the Caribbean populations. Haplotypes from the major lineages A, B, C and D were first described by Wallace et al. (1985) using restriction fragment length polymorphism (RFLP) and HVR mutation data. These have since been refined by a number of studies focusing on mutations in the HVR to determine haplotypes within the major haplogroups (Martinez-Cruzado et al., 2001;
27 Martinez-Cruzado et al., 2005; Toro-Labrador et al., 2003). X2a is the only major
Native American haplogroup that does not appear to be present in the modern Native
American population of the Caribbean.
There are some drawbacks to using modern population data in order to reconstruct past population variation, mainly due to the historical decimation of the Native American populations associated with the arrival of Europeans and admixture associated with the
African slave trade (Salas et al., 2004). Recent studies by Martínez-Cruzado (2010) and
Vilar et al. (2014) have highlighted the lower levels of indigenous Native American haplotypes in some populations of the Caribbean compared to others. Studies indicate that Puerto Rico has the highest level of Native American mitochondrial haplotypes, found in 61% of the population, followed by Cuba with 33%, Dominica with 28%, and
5-15% in Hispaniola (Benn-Torres et al., 2008; Gaieski et al., 2011; Martinez-Cruzado et al., 2001; Martinez-Cruzado et al., 2005; Mendizabal et al., 2008; Simms et al., 2010;
Vilar et al., 2014; Wilson et al., 2012). The effect of colonial impacts and historical admixture is even more extreme on NRY haplotypes, with less than 2% of the male populations in Hispaniola, Cuba and Puerto Rico exhibiting Native American Y- chromosome haplotypes (Mendizabal et al., 2008).
Despite the low levels of indigenous mitochondrial DNA in most modern Caribbean populations, Martínez-Cruzado (2010) suggested there were three major migrations into the Caribbean, based on analyses of HVR haplotypes. The first is likely to have come from Central America and the latter two from South America. These interactions have since been refined by Vilar et al. (2014) using haplotypes determined from complete mitochondrial genome and HVR sequencing. They described haplotype links with
28 Venezuela and Brazil through the presence of haplotypes C1b4 and C1b2, respectively, in the Puerto Rican population. The researchers have also suggested that the presence of the haplotype A2k1 in Puerto Rico populations may reflect the interaction and movement of people within Puerto Rico as a result of the development of the Taino culture. Only from the sequencing of complete mitochondrial genomes was the refinement of haplotypes C1b and A2k possible.
The use of DNA sequences from extant populations provides a picture of the current variation within the Caribbean populations analysed and studies have attempted to make some links to interactions and migrations in the past (Martínez-Cruzado, 2013).
However, due to the low levels of indigenous DNA in the current populations, they may not accurately represent the level of variation that was present in the prehistoric populations in the Caribbean. In addition, most studies have focused on populations within the Greater Antilles, which may have different migration and interaction histories to those of the Lesser Antilles. To get a better picture of the variation and origins of the pre-historic populations, aDNA analysis can be carried out on archaeological remains. Three studies have reported the results from aDNA analyses on ancient populations from Cuba, Hispaniola and the Lesser Antilles (Lalueza-Fox et al.,
2001; Lalueza-Fox et al., 2003; Mendisco et al., 2015), and these are described below.
3.3.2 Ancient DNA Studies
A short segment (354bp) of the HVR was amplified from ancient remains of the
Ciboney population of Cuba, dating to 2770BC-AD380 (Lalueza-Fox et al., 2003). The
Ciboney were a hunter-gather culture believed to have descended from the Pre-Arawak and Arawak migrations into the Greater Antilles. The same short fragment has also
29 been amplified from remains of the Taino peoples of Hispaniola, dating to AD600-
1600 (Lalueza-Fox et al., 2001). In contrast to the Ciboney population, the Tainos descended from populations associated with the ceramic age migrations in to the
Caribbean, thus the two groups may exhibit differing genetic histories. The haplogroups represented by both of these groups are A2, C and D, which are also found in Native Americans in North, Central and South America. When compared with modern Native American populations, the Ciboney haplotypes fell within Native populations from both Central and South America, showing a mix of ancestry from the pre-Arawak migration from Central America, and the Arawak migration from South
America. In the Taino population, however, haplotypes from haplogroup C were most closely related to modern haplotypes from South America. In addition, some haplotypes from the Ciboney population were identical to those of the Taino, suggesting an ancestral link between the two groups, which is unsurprising as the two cultures occupied the same regions and likely interacted during Taino expansion into the western Greater Antilles.
Mendisco et al. (2015) focused on mitochondrial HVR and single nucleotide polymorphism (SNP) analysis (13 individuals) as well as NRY-SNP analysis (1 individual) from prehistoric sites in the Guadeloupe archipelago, Lesser Antilles. This is the only ancient genetic study thus far to focus on this region of the Caribbean. The human remains analysed by Mendisco et al. (2015) date to AD1200-1600, which is later in the prehistoric sequence than the emergence of the Ciboney or Taino populations and during the time of significant population growth, interaction and arrival of Europeans in the region. The haplogroups determined from HVR and SNP analyses from these samples are A2, C1 (C1b, C1d, C1c) and D1, which are most common in
30 modern populations from Northern South America. Haplotypes found in this ancient population most closely matched sequences from modern Native American tribes in
Colombia and the Amazonian area, which the authors suggest may be due to continued interaction with these populations following settlement of the Caribbean.
A common thread in the three studies described above is that all found little resemblance between the ancient and extant populations within the Caribbean, with haplotypes of the ancient populations most closely related to those of Central America or northern South America. Furthermore, a lack of similarity between ancient and modern populations was also demonstrated in a recent study by Llamas et al. (2016) using complete ancient mitochondrial genomes from South America. None of the 84 haplotypes determined from 92 prehistoric individuals in this study were identical to modern Native American haplotypes from the same regions. However, the level of variation between the haplotypes was shown to fall within the current levels of variation. Llamas et al. (2016) attribute these results to a combination of geographical separation after arrival in South America and a genetic bottleneck due to decimation of indigenous populations after the arrival of Europeans in the early contact period. This further illustrates why variation from extant populations may not fully resemble the variation in prehistoric populations.
An additional issue with the currently completed ancient DNA studies in the Caribbean is that aside from the comparison of haplotypes between the Ciboney and Taino populations from the Greater Antilles, no study has compared ancient Caribbean data to ancient populations in the pan-Caribbean region such as the northern coasts of
Venezuela and Colombia or the Isthmo-Colombian region, or vice versa. Several
31 ancient DNA studies from the Caribbean coasts of Venezuela, Colombia and Central
America, have been published in the literature (Lalueza-Fox, 1996; Monsalve et al.,
1994; Monsalve, 1997; Monsalve et al., 1996). These studies describe the presence of haplogroups A, B, C and D in Central and South America based on characterisation through restriction enzyme and HVR analysis, and how the distributions of the haplogroups differ to each other in these regions. Comparisons using data from ancient populations in both the Caribbean and the mainland areas that border the Caribbean Sea may help elucidate theories of migration and interaction established using archaeology.
Llamas et al. (2016) study described above was the first to use complete mitochondrial genome sequencing on ancient human remains in the South American region and demonstrated a better resolution of haplotypes than previous South American and
Caribbean aDNA studies. Future studies in the pan-Caribbean region should aim to undertake complete mitochondrial DNA analyses of human remains, if possible, to generate comparative data that will provide a broader picture of interaction during the prehistoric period. With the current limitations on extant and ancient human data to determine migration and interaction in the past, other approaches using genetics, such as the commensal approach (Matisoo-Smith, 1994; Matisoo-Smith and Robins, 2004), can provide valuable details relating to the mobility and interaction in prehistory in the
Caribbean region.
3.4 Commensal Approach
As an alternative to and also to support existing data from archaeological and modern or ancient human DNA, the commensal approach, has been developed to better understand population interaction and mobility in the past. Commensal species can be
32 defined as plants, animals or microbes that have been intentionally or unintentionally transported by humans beyond their natural range (Giovas, 2016; Matisoo-Smith 1994;
Storey et al., 2013). This model is particularly useful when human remains have not been recovered, or are not available to be analysed, or can supplement already established human DNA research. Commensal studies in the Pacific region have highlighted the use of the commensal model in an island settting to determine patterns of migration and interaction, as commensal species are typically restricted to islands unless transported by humans (Greig et al., 2015; Larson et al., 2010; Matisoo-Smith,
1994; Matisoo-Smith and Robins, 2004; Savolainen et al., 2004; Seelenfreund et al.,
2011; Storey et al., 2007). This approach may also be applied to other island settings, such as the Caribbean where a number of small mammals are found to have been transported from mainland South America during the late ceramic age.
As mentioned previously, several species of animals (dogs, armadillos, peccaries, opossums, agoutis and guinea pigs) were introduced to the Caribbean islands via human transportation during the Late Ceramic period. Therefore, these animals may be useful for a commensal study in this region to address the likely origins of the ceramic populations who transported these animals, and whether they support the current migration models outlined in Chapter 2. Limited analyses have attempted to determine the origins of Caribbean agoutis and dogs through isotopic methods (Giovas, 2016;
Laffoon et al., 2015), but only the guinea pig has been studied thus far using a genetic approach (Kimura et al., 2016). Recent archaeological and molecular work has highlighted the value of domestic guinea pigs as a commensal model for the migrations to the Caribbean due to their wide distribution throughout the islands (Kimura et al.,
2016; LeFebvre and deFrance, 2014). The following section will focus on the merits for
33 using DNA studies of the guinea pig as a commensal model for human migration and interaction in South America and the Caribbean.
3.4.1 Guinea Pigs as a Commensal Model for Understanding Prehistoric Human
Interaction in the Caribbean
Guinea pigs originated in South America and were thought to have been domesticated in the Andean highlands of Peru by 2500BC (Morales, 1994; Sandweiss and Wing,
2013; Wing, 1986). Molecular data suggest that the domestic species (C. porcellus) descended from the wild Cavia tschudii (Dunnum and Salazar-Bravo, 2010; Spotorno et al., 2007; Spotorno et al., 2006). Guinea pigs were commonly used as a food source, sacrificial ritual object, and for medicinal diagnoses in prehistoric South America
(Moseley et al., 2005; Rofes, 2004; Rofes and Wheeler, 2003; Sandweiss and Wing,
2013). Several species of guinea pig currently inhabit Colombia, Peru, Brazil, Bolivia,
Chile, and Puerto Rico where they are primarily used as a food source (Spotorno et al.,
2006).
Archaeological remains of domesticated guinea pigs are found across the Caribbean archipelago post-AD500, from Jamaica and Hispaniola to Curacao (Figure 3.1), though they appear to have been introduced first to Puerto Ricoas the earliest archaeological dates are seen in this location (LeFebvre and deFrance, 2014). Remains are found in low numbers across all the sites compared to marine taxa, suggesting guinea pigs along with other small animals such as hutia, agouti and opposum were a supplementary food source to the marine economy. There is little to no evidence of guinea pigs being used in ritual or religious contexts in the Caribbean as they were in the Andean communities
(LeFebvre and deFrance, 2014).
34
Figure 3.1: From LeFebvre and DeFrance (2014, Figure 1): Distribution of archaeological sites in the Caribbean containing domestic guinea pig remains.
The largest number of archaeological guinea pig remains in the Caribbean are found in
Puerto Rico which, as suggested by LeFebvre and deFrance (2014), may signal Puerto
Rico as a prehistoric distribution centre for the animal. Guinea pig remains are found across six different archaeological sites in Puerto Rico; Finca Valencia, Jacana, Tibes,
Hacienda Grande, Rio Tanama and Paso Del Indio (Carlson, 2008; deFrance et al.,
2010; DuChemin et al., 2010; Quitmyer and Kozuch, 1996; Singleton, 2012; Wing et al., 1990). They date from AD600-AD1500, although the majority of remains found date from post AD1000, suggesting an expansion in population size during this time.
Other islands with recorded presence of pre- Columbian guinea pig remains are:
Jamaica (Green Castle), Hispaniola, Vieques (Lujan), St John (Cinnamon Bay), Nevis
(Coconut Walk), Antigua (Mill Reef, Indian Creek, Coconut Hall), St Lucia (Giraudy),
Carriacou (Grand Bay), and Curacao (Santa Barbara), though these occurrences are limited to only a few bones at each site (Allgood, 2002; Allsworth-Jones and Wesler,
35 2001; Giovas et al., 2012; Healy et al., 2003; Phulgence, 2007; Quitmyer and Wing,
2001; Solís Magaña and Rodríguez, 2000).
Following European arrival in the Caribbean and South America from AD1492, guinea pigs were translocated to Europe as pets (Wagner and Manning, 1976). The first archaeological evidence of guinea pigs in Europe dates to AD1550-1640, from Mons,
Brussels in Belgium (Ansieau and Denis, 2009; Pigière et al., 2012). Eight bones were recovered, all likely representing a single individual. Early evidence of guinea pigs can also be found at Hill Hall Manor in Essex, England (Hamilton-Dyer, 2009) where a partial skeleton was found, dating to AD1574-5. The only other archaeological remains of guinea pigs in Europe discovered thus far, are a maxilla and skull fragment dating to the 19th Century, at the Royal London Hospital, England (Morris, 2014). During the
18th and 19th Centuries guinea pigs became a useful species for early medical and genetic studies and research undertaken using guinea pigs contributed to many Nobel prizes (Endersby, 2012).
The earliest known evidence of guinea pigs in North America dates to the early 19th
Century. The remains of a single guinea pig have been found at Heyward-Washington
House in Charleston, South Carolina (Zierden and Reitz, 2009). No evidence of butchery or charring due to cooking is found on the remains suggesting that in this case the guinea pig was likely a pet. Archaeological remains of other exotic animals such as several species of parrot were also found in conjunction with the archaeological remains of the guinea pig. It is unclear where guinea pigs from North America originate from, as they may have been brought over from Europe during colonial expansion or
36 acquired as a ‘curiosity’ through trade with populations in South America (Zierden and
Reitz, 2009).
In the present day, the majority of guinea pigs outside of South America are kept as house pets or are bred specifically for laboratory studies as model organisms for investigations on a range of diseases (Baldwin et al., 1998; McMurray, 1994). This differs in South America where guinea pigs are still regularly consumed and meat markets have encouraged breeding for high meat quality and quantity. Over time, this may have an impact on the genetics of meat-producing populations (Spotorno et al.,
2006).
The majority of molecular studies using guinea pigs in South America have focused on inter-species phylogenetic relationships or the origins of the domestic species (Burgos-
Paz et al., 2011; Dunnum and Salazar-Bravo, 2010; Spotorno et al., 2007; Spotorno et al., 2006; Spotorno et al., 2004). These studies have compared cytochrome B sequences from wild and domestic guinea pigs from throughout South America to determine that the wild species, Cavia tschudii tschudii, is the most likely ancestor to the domestic species. To date, molecular evidence from modern guinea pigs has not yet identified a founder lineage for the European guinea pigs (Spotorno et al., 2006), however this has the potential to be resolved with genetic analysis of archaeological specimens from
South America and Europe.
Recently, Kimura et al. (2016) has used genetic variation from archaeological remains of guinea pigs in the Caribbean to attempt to determine a likely origin. Based on analysis of a 288bp region of the cytochrome B gene as well as a very short fragment of
37 the HVR (79bp), Kimura et al. (2016) suggested that the likely origin of domestic guinea pigs in the Caribbean was Colombia. Phylogenetic relationships were based on analysis of the ancient samples with modern samples from Colombia, Peru, Ecuador and Chile as no ancient data was available at the time from these regions. There are, however, limitations with this study. Firstly, Kimura et al. (2016) only used a short fragment of the mitochondrial genome, which is unlikely to contain the full level of variation present in the genome (Greig et al., 2015; Knapp et al., 2012b). Secondly, the sample size was small (n=5 ancient samples), and all but one of the samples was from the same island (Puerto Rico). Thirdly, the ancient samples were compared to modern data, primarily from Colombia which may have biased the relationships. In order to really test the hypothesis put forward by Kimura et al. (2016), a larger sample size, inclusion of ancient samples from mainland South America, and sequencing of complete mitogenomes is required.
38 3.5 Research Aims This study aimed to use complete mitochondrial genomes from ancient guinea pigs,
Cavia porcellus, as a proxy for studying human interaction and exchange networks in
South America and the Caribbean, and to better understand the trade of guinea pigs both prehistorically and historically. Archaeological remains of guinea pigs from the
Caribbean, Colombia, Peru, and Bolivia as well as historic remains from Belgium and
South Carolina, were analysed in this study. Based on previous archaeological and genetic research detailing patterns of trade and migration into the Caribbean described above, I hypothesise that ancient Caribbean guinea pigs may be most closely related to ancient Colombian guinea pigs. An alternative hypothesis may be that migration occurred from elsewhere in South America, possibly coastal Venezuela, in which case
Caribbean guinea pigs may be genetically distinct from guinea pigs in western South
America.
Specific objectives of the research were:
1. To assess the value of complete mitogenomes for understanding prehistoric
population interaction in the Caribbean region.
2. To investigate the likely origin for the historic translocation of guinea pigs to
Europe and the United States of America.
3. To determine if complete mitogenomes provide greater haplotype resolution
than the Cytochrome B or Hypervariable sequences.
4. To demonstrate the value of Next Generation Sequencing for ancient DNA
studies in the Caribbean and South America.
39 Chapter 4: Methodology
4.1 Ancient DNA: Uses and Limitations
Ancient DNA (aDNA) analysis is a useful tool for understanding the past at the molecular level. The first successful aDNA study was undertaken by Higuchi et al.
(1984), where DNA was extracted from a 150-year-old Quagga (Equus quagga). Since then, the advent of PCR and Next Generation Sequencing (NGS) technologies have allowed for a multitude of aDNA research (Linderholm, 2016). Most aDNA studies until very recently have focused on the use of mitochondrial DNA (mtDNA) as it exists in multiple copies per cell, thus is more likely to be retrieved than nuclear DNA in ancient remains (Pääbo, 1989). With advancements in technology, more recent studies have begun to reconstruct whole genomes from archaeological material (Shapiro and
Hofreiter, 2014). Limitations of using aDNA include: risk of contamination by exogenous DNA, determining the authenticity of the sequence, and achieving sufficient amplification of highly degraded sequences. These will be discussed below as well as the impact of NGS and other recent methodological advances that have contributed to the better retrieval of DNA from archaeological materials.
4.1.1 Contamination and Sequence Authenticity
DNA in archaeological samples can become contaminated with a variety of exogenous sources of DNA throughout post-mortem degradation, excavation, and laboratory processing (Höss et al., 1996; Skoglund et al., 2014). As the DNA becomes fragmented, the infiltration of bacterial and fungal DNA occurs, followed by further contamination from human and other environmental DNA (Höss et al., 1996). PCR products from modern DNA sequencing laboratories are also a large source of contamination for
40 aDNA studies (Hofreiter et al., 2001b; Cano et al., 1993). Due to potentially high levels of contamination affecting research, stringent procedures have now been put in place for the handling of archaeological materials for aDNA analysis. These include the use of clean facilities specific for aDNA (i.e. no PCR), regular bleaching or UV radiation of surfaces and materials used in the processing of aDNA, as well as the bleaching, UV radiation or removal of the outer surface of archaeological samples prior extraction to eliminate surface contamination (Green et al., 2009; Knapp et al., 2012a; Linderholm,
2016).
Several studies have identified the way DNA is degraded post mortem (Gilbert et al.,
2003; Lindahl, 1993; Pääbo, 1989; Wandeler et al., 2003). There are two types of damage patterns seen in ancient DNA sequences: shortening lesions and miscoding lesions. Shortening lesions occur through the activity of nucleases that cause strand breaks, creating fragments typically less than 150bp (Pääbo et al., 2004). Miscoding lesions occur through the deamination of cytosine (C) and guanine (G) to produce uracil and xanthine (Hofreiter et al., 2001a; Millar et al., 2008). This leads to the incorporation of thymine (T) and adenine (A) respectively. Several studies have shown that C to T transitions are more likely to occur than G to A transitions (Briggs et al.,
2007; Gilbert et al., 2003).
The extent of damage in an ancient DNA sample cannot be determined by age alone.
The environment heavily influences the rate of degradation (Smith et al., 2003;
Willerslev et al., 2007). Cold, high salinity or extremely dry climates such as the permafrost have provided the best preservation for ancient DNA sequences (Willerslev et al., 2007). Under these conditions, the nucleases responsible for degrading DNA
41 become inactive (Smith et al., 2003). Other mechanisms still lead to the degradation of
DNA, but at a much slower rate. In contrast, tropical environments such as the Pacific have typically produced poor quality, highly degraded DNA sequences (Knapp et al.,
2012b).
4.1.2 The Application of Next Generation Sequencing1 for Ancient DNA
The development of NGS in the early 2000s revolutionised the field of aDNA (Millar et al., 2008). NGS technologies sequence all available DNA in a sample through short reads, and are capable of sequencing to high read depth (number of times each base is covered), unlike traditional cloning and Sanger sequencing methods. The ability to sequence short fragments is practical for ancient DNA studies due to the typically fragmented DNA in archaeological and museum samples (Pääbo, 1989). The ability of being able to sequence short fragments has allowed for the reconstruction of complete mitochondrial genomes, and recently nuclear genomes from degraded DNA of archaeological samples (Green et al., 2008; Krause et al., 2010; Shapiro and Hofreiter,
2014). Due to the read depth available through NGS, it has become easier to visualise and characterise DNA damage patterns, thus authenticating aDNA sequences. This also makes it easier to identify modern contamination in a sequence, as the DNA will not display damage patterns like those of aDNA. Typical DNA damage patterns such as those described above have now been incorporated in bioinformatic programs, such as
MapDamage (Jónsson et al., 2013), that determine between damage and true sequence.
In addition, with the development of NGS, the feasibility and cost of sequencing has decreased, thus the amount produced has rapidly increased. This powerful technology
1 Next generation sequencing here refers to second generation sequencing techniques, though NGS is the term most often used to describe these techniques in the literature and will be applied here. 42 will likely continue to revolutionise aDNA studies in the future by allowing for higher amounts and quality of data to be produced.
4.1.3 New Methodologies and Limitations
With the continued development of aDNA extraction and library preparation techniques, as well as sequencing technologies, it is now possible to recover more information from degraded DNA. Extraction methodologies developed by Rohland and Hofreiter (2007) and Dabney et al. (2013) have increased the level of DNA extracted compared to previous methods. The method by Dabney et al. (2013) is well suited to very short fragments (i.e. highly degraded DNA). Enrichment of the ancient DNA library by hybridisation capture helps to preserve as much endogenous DNA as possible by hybridising DNA to beads using bait sequences and removing any unbound
(exogenous) DNA (Maricic et al., 2010). Furthermore, with the innovation of NGS, new bioinformatics approaches have been developed to process increasing amounts of sequence data.
There are, however, limitations to aDNA analyses. Firstly, it is a destructive process, which may mean that rare or important specimens cannot be further studied, if at all.
Secondly, analyses are restricted to material that is recovered from the archaeological record, meaning that samples sizes are often small. Thirdly, no matter how powerful the techniques, if the DNA in a sample is too degraded, it won’t be able to be analysed as it will be too fragmented and damaged for adequate sequence to be reconstructed. This can affect already small sample sizes if a number of samples have DNA that is too degraded to be retrieved. Despite limitations, the field of aDNA has already produced
43 valuable results addressing a range of biological and anthropological questions, and will continue to do so with the innovation of more high-powered technologies.
4.2 Modern DNA Processing
4.2.1 Sample Collection and DNA Extraction
Muscle tissue from a single modern, recently deceased domestic pet guinea pig (“Mrs
Chloe”) was obtained for analysis. DNA was extracted from the muscle tissue using the
MagJet Genomic DNA kit (Thermo Fisher Scientific, Inc), as per the manufacturers instructions, adding 150μL of elution buffer. The purified DNA sample was quantified using a NanoDrop 2000 (Thermo Fisher Scientific, Inc) and stored at 4°C until further use.
4.2.2 Long Range PCR Amplification
Primers for long-range PCR amplification were created using PrimerBLAST from the
Cavia porcellus complete mitogenome sequence available on Genbank (Accession number: NC_000884.1, D’Erchia et al, 1997). Three overlapping primer sets covering the whole mitogenome (total size: 16.8kb) were created (Table 4.1), one of which spans the HVR (primer set 3).
Long-range PCR was performed using the KAPA LongRange HotStart PCR Kit
(KAPA Biosystems) with the following concentration of reagents: 1x KAPA
2+ LongRange Buffer (no Mg ); 1.75mM MgCl2; 0.3mM dNTPs; 0.5µM of each forward and reverse primer; 0.75U KAPA LongRange HotStart DNA polymerase; and 2µL of
DNA template (20ng). Optimal primer efficiency was found using the following
44 Table 4.1: Primers created for Long Range PCR amplification of the complete Cavia porcellus mitochondrial genome.
Fragment Primer Set: Name Sequence TM (°C) length (kb) 1: CP_Mit1311F CCCGAAACCAAACGAGCTAC 56.2 6.8 1: CP_Mit8190R AGGCCTAACAGGTTGGTGGA 58.5
2: CP_Mit7406F CCGGGGAACTTCGACTACTT 56.5 7.1 2: CP_Mit14553R CCCTATGAATGCGGTAGCCA 57.0
3: CP_Mit14041F GCAGATAGGCACACAAAGCC 56.7 4.7 3: CP-Mit1970R GCCGTTGAACGAATGTCACT 55.9
protocol: 94°C for 3 minutes, followed by 35 cycles of: 94°C for 25 seconds, 55°C for
15 seconds, 68°C for 8 minutes; one cycle of 72°C for 12 minutes followed by a hold at10°C. PCR products were visualised by electrophoresis on a 1% agarose gel stained with GelRed (Biotium). A KAPA Universal DNA Ladder (KAPA Biosystems) was used for estimation of product sizes. PCR products that successfully amplified were purified using Mag-Bind RXNPure Plus beads (OMEGA Biotek), following the manufacturers’ instructions. Purified PCR products were quantified using a NanoDrop
2000 (Thermo Fisher Scientific, Inc) and stored at 4°C until further use.
4.2.3 Library Preparation and Sequencing of a Modern Tissue Sample
A double-stranded barcoded library was created from the purified PCR products of modern DNA obtained from ‘Mrs Chloe’ following Kircher et al., (2012) for sequencing on an Illumina MiSeq platform (San Diego, CA, USA). Long-range PCR products were pooled at equimolar ratios and sonicated using a Picorupter (Diagenode) for 9 cycles of 15 seconds on, 45 seconds off to shear the DNA into smaller fragments
45 (approximately 500bp). Purification using Mag-Bind RXNPure Plus beads (OMEGA
Biotek) was carried out between each step. Steps followed were: blunt end repair and phosphorylation by T4 DNA Polymerase and T4 Polynucleotide Kinase (both New
England BioLabs Inc), A-tailing with the enzyme Klenow (New England BioLabs Inc), and ligation of Illumina compatible double stranded adaptors (Adapter_P5 and
Adapter_P7) (Knapp et al., 2012b). A high fidelity PCR was carried out using KAPA
HiFi PCR Kit (KAPA Biosystems). Reagents in the 30µL reaction included: 1x KAPA
HiFi Buffer (with Mg2+); 0.3mM dNTPs; 0.3µM each of the P5_extension and
P7_extension primers; 0.6U KAPA HiFi DNA Polymerase; and 19µL of PCR product.
The amplification protocol carried out was: 94°C for 3 minutes; 20 cycles of 94°C for
25 seconds, 60°C for 15 seconds, 72°C for 15 seconds; one cycle of 72°C for 15 minutes. The sample was then purified once more with Mag-Bind RXNPure Plus beads
(OMEGA Biotek) and quantified using a Qubit 2.0 Fluorometer (Invitrogen). The DNA library was sequenced using a 2 x 250 base paired-end run on an Illumina MiSeq platform at the Otago Genomics & Bioinformatics Facility at the University of Otago,
Dunedin, New Zealand.
4.2.4 Bait Preparation
Hybridisation capture requires bait DNA to ligate to beads, which was produced using
DNA extracted from the modern domestic guinea pig (Section 4.2.1). To create the bait, the mitochondrial genome was amplified in the same three fragments as described above, however amplification was carried out using a KAPA Long-Range PCR protocol in 50 L reactions with the following reagent concentrations: 1x KAPA
2+ Buffer (with no Mg ); 1.75mM MgCl2; 0.3mM dNTPs; 0.5µM of each primer; 1.25U
KAPA DNA Polymerase; with 1µL of DNA template (10ng). The same PCR protocol
46 was used as described in Section 4.2.2. PCR products were analysed by electrophoresis on a 1% Agarose gel stained with GelRed (Biotium). Purification of successfully amplified products was carried out using MinElute PCR Purification Kit (QIAGEN) following the manufacturer’s instructions, with an additional ethanol wash. Purified
PCR products for each fragment were pooled and quantified on a NanoDrop 2000
(Thermo Fisher Scientific, Inc).
Bait preparation was carried out following Maricic et al. (2010). The three long-range fragments were pooled in equimolar ratios and sonicated to shear the DNA into smaller fragments (~250bp) for 9 cycles of 15 seconds on, 45 seconds off, using a Picoruptor
(Diagenode). Sonicated DNA was analysed on a 1.5% Agarose gel, stained with
GelRed (Biotium) and compared with unsonicated DNA to ensure sonication had been successful, resulting in fragments sizes of 150-450bp. The sonicated bait DNA was then re-quantified using a NanoDrop 2000 (Thermo Fisher Scientific, Inc). Sonicated
DNA then underwent blunt end repair (described above) followed by purification with a MinElute PCR Purification Kit (QIAGEN). Blunt-ended samples were ligated with biotinylated (Bio-T/B) adapters and again purified with a MinElute PCR Purification
Kit (QIAGEN). Bio-T/B adapter-ligated bait was quantified using a NanoDrop 2000
(Thermo Fisher Scientific) and stored at -20°C until further use.
4.3 Ancient DNA Processing
4.3.1 Sample Collection
A total of 43 ancient (pre AD1500) and three historic (post AD1500) domestic guinea pig (C. porcellus) samples from the Caribbean, Colombia, Peru, Bolivia, Belgium and
47 South Carolina (see Figure 4.1 for a map of archaeological sites, Table 4.2 for a list of samples by region) were used for complete mitochondrial genome extraction and sequencing. The samples were kindly provided for use in this project by Dr. Susan deFrance and Dr. Michelle LeFebvre (Florida Museum of Natural History, Gainsville,
Florida), Dr. Katherine Moore (University of Pennsylvania Museum of Archaeology and Anthropology, Philadelphia, Pennsylvania), Dr. Fabienne Pigiere (Royal Belgian
Institute of Natural Sciences, Brussels, Belgium), and Martha Zierden (Charleston
Museum, Charleston, South Carolina).
All samples from the Caribbean and South America were excavated from midden areas in archaeological sites. The samples from Belgium and Charleston were excavated from historic house sites. The majority of the archaeological samples were mandible fragments or teeth, as these elements are more prominent in archaeological sites due to size and robusticity.
4.3.2 Sample Preparation
All ancient and historic samples used for DNA analyses were processed in a purpose- built ancient DNA facility, located in the Richardson Building at the University of
Otago, prior to PCR amplification steps (Knapp et al., 2012a). As ancient DNA extraction is a destructive process, all samples were photographed before starting the extraction process (see Figure 4.2 for an example). Samples were immersed in 5% bleach, rinsed in ultrapure water and dried overnight under UV lighting to ensure all surface contamination was removed.
48 Key:" " 18" 1. Tibes"A"&"B,"Puerto"Rico" 2. Coconut"Walk,"Nevis" 1" 2" 3" 3. Coconut"Hall,"Antigua" 4. Giraudy,"St"Lucia" 4" 5. Grand"Bay,"Carriacou" 5" 6. El"Venado,"Colombia" 7. Checua"II,"Colombia" 8. Madrid"Site,"Colombia" 6" 9. Aguazuque"I,"Colombia" 7" 10. Kuelap,"Peru" 8" 9" 11. Lo"Demas,"Peru" 12. Torata"Alta,"Peru" 13. Moqi,"Peru" 14. Kala"Uyuni,"Bolivia" 15. Llusco,"Bolivia" 16. HeywardZWashington"House," South"Carolina" 10" 17. Mons,"Belgium" 18. San"Sebastian,"Puerto"Rico" 17"
11" 16" 12" 15" 13"14"
A" B"
Figure 4.1: Map of archaeological sites included in the analysis. Boxed area in Figure B indicates the area covered in Figure A.
49 Table 4.2: Archaeological specimens analysed in the study.
Lab ID Site Location Region Date Element (side) Reference
MS10625 Tibes A Puerto Rico Caribbean AD600-900 Mandible (R) 1
MS10626 Tibes B Puerto Rico Caribbean AD900-1200 Mandible (L) 1
MS10627 Grand Bay Carriacou Caribbean AD985-1030 Maxilla (R) 2
MS10630 Coconut Hall Antigua Caribbean AD1035, 1045 Mandible (L) 3
MS10631 Giraudy St Lucia Caribbean AD1200-1400 Atlas, Lumbar Vertebrae 4
MS10632 Coconut Walk Nevis Caribbean AD760-1440 Femur Epiphysis (R) 5
MS10633 Coconut Walk Nevis Caribbean AD760-1440 Femur Epiphysis (L) 5
MS10634 Coconut Walk Nevis Caribbean AD760-1440 Femur Epiphysis (L) 5
MS10635 San Sebastian Puerto Rico Caribbean Modern Complete Mandible 6
MS10636 El Venado Valle de Samaca Colombia AD1400-1540 Mandible (R) 7
MS10637 El Venado Valle de Samaca Colombia AD1400-1540 Cranial Fragments 7
MS10638 El Venado Valle de Samaca Colombia AD1400-1540 Mandible (L) 7
50 Lab ID Site Location Region Date Element (side) Reference
MS10639 El Venado Valle de Samaca Colombia AD1400-1540 Mandible (R) 7
MS10640 El Venado Valle de Samaca Colombia AD1400-1540 Mandible (L) 7
MS10641 El Venado Valle de Samaca Colombia AD1400-1540 Mandible (R) 7
MS10672 Madrid Site Cundinamarca Colombia 200-100 BC Mandible (R) 8
MS10673 Aguazuque I Cundinamarca Colombia 1885-1815 BC Incisors and molar 9
MS10674 Aguazuque I Cundinamarca Colombia 1885-1815 BC Molar 9
MS10675 Aguazuque I Cundinamarca Colombia 1885-1815 BC Incisors 9
MS10676 Checua II Sabana de Bogota Colombia 6200-2000BC Premaxilla Fragment and 10
Incisor
MS10677 Checua II Sabana de Bogota Colombia 6200-2000BC Maxilla Fragment with Molar 10
MS10678 Checua II Sabana de Bogota Colombia 6200-2000BC Molars 10
MS10679 Checua II Sabana de Bogota Colombia 6200-2000BC Maxilla and Molars 10
MS10680 Checua II Sabana de Bogota Colombia 6200-2000BC Incisors 10
51 Lab ID Site Location Region Date Element (side) Reference
MS10628 Moqi Locumba Valley Peru Post AD1400 Mandible (L) 11
MS10629 Moqi Locumba Valley Peru Post AD1400 Mandible (L) 11
MS10645 Moqi Locumba Valley Peru Post AD1400 Mandible (L) 11
MS10646 Moqi Locumba Valley Peru Post AD1400 Mandible (L) 11
MS10647 Moqi Locumba Valley Peru Post AD1400 Mandible (L) 11
MS10648 Moqi Locumba Valley Peru Post AD1400 Mandible (L) 11
MS10649 Moqi Locumba Valley Peru Post AD1400 Innominate (L) 11
MS10642 Lo Demas Chincha Valley Peru AD1480-1540 Femur (L) 12
MS10643 Lo Demas Chincha Valley Peru AD1480-1540 Mandible (L) 12
MS10644 Lo Demas Chincha Valley Peru AD1480-1540 Mandible (L) 12
MS10650 Torata Alta Torata Valley Peru AD1550-1600 Mandible (L) 11
MS10681 Kuelap Chachapoyas Peru AD500-1400 Femur Fragment 13
MS10682 Kuelap Chachapoyas Peru AD500-1400 Mandible Fragment 13
52 Lab ID Site Location Region Date Element (side) Reference
MS10683 Kuelap Chachapoyas Peru AD500-1400 Humerus Fragment 13
MS10684 Kuelap Chachapoyas Peru AD500-1400 Femur Fragment 13
MS10685 Kuelap Chachapoyas Peru AD500-1400 Femur Fragment 13
MS10651 Kala Uyuni Titicaca Basin Bolivia 200BC-AD500 Mandible (R) 14
MS10652 Kala Uyuni Titicaca Basin Bolivia 200BC-AD500 Mandible (R) 14
MS10653 Llusco Titicaca Basin Bolivia 800-200BC Mandible (R) 15
MS10654 Llusco Titicaca Basin Bolivia 800-200BC Mandible (L) 15
MS10655 Mons Brussels Belgium AD1550-1640 Tibial Shaft 16
MS10671 Heyward- Charleston South Carolina 1820s Femur Fragment 17
Washington House
1: deFrance (2010), 2: Giovas (2016), 3: Healy et al. (2003), 4: Phulgence (2007), 5: Kaye et al. (2010), 6: Kimura et al. (2016), 7:Rivas (1999), 8: Rodríguez and Toro (2005), 9:Correal Urrego (1990), 10:de Mahecha (1992), 11: deFrance (1993), 12: Sandweiss and Wing (2013), 13: Doig (2000), 14: Bandy and Hastorf (2007), 15: Whitehead (2007), 16: Pigière et al. (2012), 17: Zierden and Reitz (2009).
53 MS10653:"Mandible"(R)"
Figure 4.2: Specimen MS10653 (Llusco Site, Bolivia), right hand side portion of a mandible.
4.3.3 DNA Extraction
DNA extraction using a silica-binding method was carried out following Rohland and
Hofreiter (2007). For each sample, approximately 150-250mg of bone or teeth was ground to a fine powder using a sterile mortar and pestle and digested overnight in an extraction buffer comprised of 0.5M EDTA, 0.10mg/ml Proteinase K. An extraction blank was processed in parallel with each set of five samples. Extraction blanks were treated as samples, but with no added DNA, thus are negative controls to ensure no contamination was present in lab reagents or due to processing. Following the digestion overnight, the samples (and blank) were centrifuged at 1200 g and the supernatant was transferred to a 15mL falcon tube containing 2.5mL of binding buffer (comprised of:
5M GuSCN, 3M Sodium Acetate) and 100µL of silica suspension. Samples were rotated overnight at room temperature (RT) to bind the DNA to the silica. Following
54 binding, samples were centrifuged at 1200 g for two minutes and the supernatant was discarded. The pellet was washed twice with binding buffer to ensure all DNA was bound and then washed three times with a wash buffer (100% ethanol, 5M NaCl, 1M
Tris pH 8.0, 0.5M EDTA pH 8.0), to remove PCR-inhibiting agents. DNA was eluted in 70µL 1x TE and samples were stored at -20°C until further use.
4.3.4 Double Stranded Library Preparation
Double-stranded barcoded libraries were prepared from the extracted DNA following
(Greig et al., 2015) for hybridisation capture and paired-end sequencing on an Illumina platform. Blunt–end repair was conducted using T4 DNA Polymerase and phosphorylated with T4 Polynucleotide Kinase (New England Labs Inc), followed by a purification with MinElute PCR Purification Kit (QIAGEN) with the addition of a second PE wash and eluted in 20µL 0.1xTE + 0.05% Tween 20. Adapter ligation followed, where Illumina compatible adapters, sol_adap_P5 and sol_adap_P7-BIO
(biotinylated at the 5’ end), were ligated onto the DNA fragments. Adapter-ligated
DNA was purified with a MinElute PCR Purification Kit (QIAGEN) with the following modifications: the addition of a second PE wash and elution of DNA in 25µL 1xTE.
The biotinylated adapter-ligated DNA was immobilised onto streptavidin-coated beads and then washed thoroughly to remove any unincorporated adapters. Adapter-ligated
DNA libraries were denatured from the beads and eluted in 20µL 0.1xTE.
DNA libraries were then prepared for quantitative PCR (qPCR) to determine the number of cycles required for amplification of each library in the subsequent PCR step.
Each reaction for qPCR amplification contained: 1x SYBR Green Master Mix (Thermo
Fisher Scientific, Inc); 0.25µM of each primer (sol_quant_P5 and sol_quant_P7); 1µL
55 of DNA template; and ultrapure water to make up a total volume of 25µL. The following protocol was used for qPCR: 95°C for 10 minutes; 40 cycles of 94°C for 30 seconds, 58°C for 30 seconds, 72°C for 60 seconds; and a final extension of 72°C for
10 minutes on a MxPro 3000P (Stratagene). The number of cycles at which amplification plateaued was recorded for each sample, including blanks. Post-qPCR libraries were electrophoresed on a 2% agarose gel, stained with GelRed (Biotium) with a KAPA Universal Ladder (KAPA Biosystems). For any samples that had not successfully amplified or showed adapter dimers, a new library was prepared from the original DNA extract, or from a new extract. If no amplification was successful, this sample was omitted from the study.
Preparation for the indexing of the ancient DNA libraries was carried out in the ancient
DNA laboratory and libraries were then transferred to a modern DNA laboratory
(located in the Department of Anatomy, University of Otago) for amplification by PCR.
Each reaction for indexing PCR contained: 1x Taq Buffer, 0.5mM MgCl2, 1mM dNTPs,
0.2µM of each primer (sol_prim_ext_P5 and sol_prim_ext_P7 to individually barcode each sample), 3.75U of Amplitaq Gold polymerase (Thermo Fisher Scientific, Inc),
19µL of DNA library and ultrapure water added to a total volume of 50µL. The amplification protocol for PCR was: 95°C for 12 minutes, 14-26 cycles (number of cycles taken to reach plateau in qPCR, minus four) of: 94°C for 30 seconds, 58°C for
30 seconds, 72°C for 60 seconds; and a final extension of 72°C for 10 minutes.
Following amplification, libraries were purified using a MinElute PCR Purification Kit
(QIAGEN), with the modification of two PE washing steps and libraries eluted in 20µL
0.1xTE with 0.05% Tween 20.
56 Prior to hybridisation capture, libraries were further amplified by high fidelity PCR to obtain 2µg of each library. Each PCR reaction contained: 1x KAPA HiFi Buffer,
0.3mM dNTPs, 0.3µM of each primer (sol_amp_P5 and sol_amp_P7), 1U of KAPA
Hifi DNA Polymerase, 1µL of indexed library. The following PCR protocol was carried out: 94°C for 5 minutes, 10 cycles of 94°C for 30 seconds, 58°C for 30 seconds,
72°C for 60 seconds, with a final extension of 72°C for 10 minutes and a hold at 10°C.
PCR reactions were purified using a MinElute PCR Purification Kit (QIAGEN) with the modification of two PE washing steps and amplified libraries were eluted in 20µL
0.1xTE. Libraries were quantified using a Qubit 2.0 Fluorometer (Invitrogen) to ensure adequate library concentration for hybridisation capture.
4.3.5 Hybridisation Capture
Hybridisation of ancient libraries to modern bait DNA was conducted following
Maricic et al. (2010) with modifications. Five hundred nanograms of biontinylated bait
DNA (modern C. porcellus DNA ligated with Bio-T/B adapter) was aliquoted with an equal volume of BWT (2M NaCl, 10mM Tris pH 8.0, 1mM EDTA pH 8.0, 0.1%
Tween-20) and denatured at 98°C for 1 minute. Streptavidin beads were purified and bait DNA was added to the beads and rotated at 25°C for 20 minutes to allow bait to bind to the beads. Beads were then purified to remove any un-incorporated bait and re- suspended in 50µL 1xTE with 0.05% Tween 20 and stored at 4°C until the ancient libraries were prepared for capture.
A hybridisation mixture comprising of: 0.9x Agilent hybridisation buffer, 0.9x Agilent blocking agent, 1.8µM of each of the eight blocking oligonucleotides (designed to block adapters used in the preparation of libraries), and 1.8µg of indexed library was
57 prepared for each sample. The hybridisation mixture and library underwent two temperature steps: 95°C for 3 minutes, then 37°C for 30 minutes. The hybridisation mixture was added to the baited beads and samples were rotated at 12rpm for 36-48 hours in a 65°C hybridisation oven to hybridise the libraries to the bait. Following hybridisation, the baited beads, with libraries attached were purified via three washes with BWT, two washes with HWT (2.5mM MgCl, 50mM KCl, 15mM Tris-HCl pH 8.0,
0.1% Tween-20) and a final wash with 1xTET, to remove unbound DNA. Libraries were then the re-suspended in 15µL of 0.1xTE. The solution was heated to 95°C for 3 minutes to denature the captured libraries from the beads. Captured libraries were eluted into a siliconised tube and stored at -20°C.
4.3.6 Ancient DNA Sequencing
A re-amplification using a KAPA HiFi PCR Kit (KAPA Biosystems) was required post-capture to ensure adequate concentration of captured library for DNA sequencing.
Each PCR reaction contained: 1x KAPA Hifi Buffer (with Mg2+), 0.3mM dNTPs,
0.3µM of each of the amplification primers (sol_amp_p5 and sol_amp_p7), 1U of
KAPA Hifi DNA polymerase and 15µL of captured library. PCR conditions were as follows: 94°C for 5 minutes, 20 cycles of 94°C for 20 seconds, 55°C for 55 seconds,
72°C for 15 seconds, with a final extension of 72°C for 5 minutes. Amplified libraries were purified using a MinElute PCR Purification Kit (QIAGEN) with the modification of two PE washing steps and libraries were eluted in 20µL 0.1xTE. Captured libraries were quantified on using a Qubit 2.0 Fluorometer. Libraries were also visualised and quantified on an Agilent 2100 Bioanalyser, to ensure that the size distribution looked as expected for aDNA libraries and to confirm the absence of adapter-dimers. All libraries were pooled in equimolar concentrations and sequenced using a 2 x 75bp paired-end
58 run on either an Illumina MiSeq or HiSeq Rapid at the Otago Genomics &
Bioinformatics Facility at the University of Otago, New Zealand. Libraries constructed from extracted blanks were sequenced separately.
4.4 Bioinformatics
Both modern and ancient samples were processed through an in-house bioinformatics pipeline adapted from Boocock (2014)
(https://github.com/theboocock/ancient_dna_pipeline). For the modern sample, reads were mapped to a reference genome of C. porcellus (Genbank: NC_000884.1) using
Burrows-Wheeler Aligner (BWA v0.7.15) (Li and Durbin, 2009). Duplicate reads were removed via Picard’s mark duplicates tool (http://broadinstitute.github.io/picard/) and variants were called using GATK (v3.5) (McKenna et al., 2010). A FASTA file was created from the variant call file (VCF) produced by GATK, using an in-house script, for use in phylogenetic analyses.
For the ancient samples, raw reads were processed by AdapterRemoval (v2) (Lindgreen,
2012) to remove adapters and bases with low quality scores. Reads were also mapped to a reference genome of C. porcellus (Genbank: NC_000884.1) using BWA (v0.7.15)
(Li and Durbin, 2009). In order to determine if any contamination had occurred during laboratory protocols, reads were also mapped to a composite genome of Homo sapiens
(NC_012920.1), Sus scrofa (NC_00845.1), Gallus gallus (NC_001323.1), Canis lupus
(NC_002008.4), Bos taurus (NC_006853.1) and Rattus exulans (NC_012389.1). PCR duplicates were removed using Picard’s mark duplicates and mapDamage (v 2.0) was used to assess damage patterns typical of ancient DNA (Jónsson et al., 2013). A VCF was produced using GATK (v3.5) (McKenna et al., 2010) and was converted to a
59 FASTA file for subsequent analyses. Sequence coverage was determined and coverage plots created using R (Ihaka and Gentleman, 1996; Team, 2000). Only samples above
90% coverage were retained for downstream analyses.
4.5 Phylogenetic Analyses
Two modern, two historic and 22 ancient complete mitochondrial genome sequences, as well as the reference genome of C. porcellus (Genbank: NC_000884.1) were aligned using the MUSCLE alignment option in Geneious (v9.1.6) (Kearse et al., 2012). The alignment was exported in the Nexus format and a Median Joining network was created using PopArt (Leigh and Bryant, 2015) with a built in traits block created using an in- house script.
Nucleotide diversity (π) across the complete ancient and modern mitochondrial genomes, the Cyt B region and the HVR was calculated using DnaSP (v.5.10.1)
(Librado and Rozas, 2009). Nucleotide diversity measures the average number of variations per site between two samples. This was performed to determine the level of variation in the complete mitochondrial genome compared to the Cyt B region and
HVR alone. It was also conducted to determine whether missing data in samples occurred at variable positions or not. Nucleotide diversity was also assessed within each geographic group of samples: the Caribbean, Colombia, Peru, Bolivia and
Europe/Historic.
In order to create Bayesian and Maximum Likelihood phylogenies for the complete mitochondrial sequences, the best-fit model for nucleotide substitution was determined in jModelTest (v2.1.10) (Darriba et al. (2012), accessible from:
60 https://github.com/ddarriba/jmodeltest2). Models were determined from 11 substitution schemes, using the Bayesian information criteria (BIC). For the complete mitochondrial genome alignment, the best-fit model was determined to be TIM2+I (I = proportion of invariable sites).
A Bayesian analysis of the complete mitochondrial genome phylogeny was undertaken using BEAST (v1.8.3) (Drummond and Rambaut, 2007). The nucleotide model and chain parameters were set using BEAUTi (v1.8.3) (Drummond et al., 2012). The nucleotide substitution model was set to GTR+I, with parameters set equal to the ac parameter to get to the TIM2 model. The analysis was run for 10,000,000 interations, with sampling at every 1000 chains. Burn-in of 10% was discarded using
TreeAnnotater (v1.8.3) (Rambaut and Drummond, 2013). The consensus phylogeny was visualised in FigTree (v1.4.2) (Rambaut, 2007), with posterior probability displayed on the branches. A Maximum Likelihood phylogeny was also created, using
IQ-Tree (v1.5.3) (Nguyen et al., 2015). The TIM2+I model was used as above.
Bootstrap (1000 replicates) values were visualised on the phylogeny, using FigTree
(v1.4.2).
An alignment of modern Cyt B sequences (1140bp) from domestic (C. porcellus) and wild (C. tschudii, C. aperea, C. aperea guianae, C. anolaimae, and C. magna) using
105 sequences available from Genbank (Appendix Table 8.1) and 29 ancient domestic guinea pigs from this study was also undertaken using MUSCLE in Geneious. The best-fit nucleotide substitution model was determined to be HKY+G (G = gamma distributed rate variation among sites) using jModelTest (v2.1.10) (Darriba et al., 2012).
Bayesian and Maximum Likelihood trees were created, as described above but with
61 parameters set for the HKY+G model and visualised in FigTree (v1.4.2) with posterior probability and bootstrap (1000 replicates) values, respectively. This was to determine whether all ancient samples fell were most closely related to modern domestic samples and to conclude if any ancient sequences were ancestral to modern guinea pigs. To further explore the relationships between ancient (n=29) and modern domestic guinea pigs (n=68) (using the Cyt B region), a temporal network was created using TempNet
(Prost and Anderson, 2011) in R. This allowed for identification of haplogroups shared between ancient and modern samples.
62 Chapter 5: Results
5.1 DNA Preservation and Authenticity
5.1.1 Sequence Coverage
Out of a total of 47 samples, 43 DNA libraries prepared from archaeological remains and one DNA library prepared from muscle tissue were sent for DNA sequencing, giving an overall library amplification success rate of 93% (Table 5.1). Three ancient samples failed to produce a DNA library after two attempts, which was most likely due to inadequate endogenous DNA as shown by the presence of adapter-dimers after DNA library preparation, therefore were removed from the study. Post sequencing, a total of
23 ancient and two modern samples (56% of amplified libraries) had adequate coverage
(over 90%) and read depth (>2 reads per base) when mapped to a complete C. porcellus mitochondrial reference genome (Genbank accession ID: NC_000884.1) for inclusion in further analyses (Table 5.2). The samples represent four of six sites in the Caribbean, and all sites analysed from Colombia, Peru, Bolivia, Belgium and South Carolina. An additional eight samples had adequate coverage (over 98%) of the cytochrome B (Cyt
B), thus were able to be used in analyses of these regions (Table 5.3).
Total and average coverage of the complete mitochondrial genome for each sample was determined and plotted using R, an example of which is shown in Figure 5.1A. Total coverage ranged between 21.1% (MS10631) and 99.5% (MS10650, MS10681,
MS10683) (Appendix Table 8.2). Average coverage of the mitochondrial genome ranged between 0.2 reads per base (MS10631) to 3297.4 reads per base (MS10683). All samples had zero coverage for a repeat region that occurs from position 16259-16562,
63 Table 5.1: Summary statistics for PCR amplification and sequencing of ancient (n=45) and modern (n=2) samples.
Samples Amplified Sequences over 90% Period Location Site Success Rate (%) (n) libraries (n) coverage (n) Puerto Rico Tibes (A and B) 2 2 1 50% Carriacou Grand Bay 1 1 1 100% Antigua Coconut Hall 1 1 1 100% St Lucia Giraudy 1 1 0 0% Nevis Coconut Walk 3 3 0 0% El Venado 6 6 1 17% Madrid Site 1 1 1 100% Colombia Prehistoric Aguazuque I 3 1 1 33% Checua II 4 4 1 25% Lo Demas 3 3 1 33% Moqi 7 7 3 43% Peru Torata Alta 1 1 1 Kuelap 5 5 5 100% Kala Uyuni 2 2 2 Bolivia Llusco 2 2 2
64 Samples Amplified Sequences over 90% Period Location Site Success Rate (%) (n) libraries (n) coverage (n) Belgium Mons 1 1 1 Historic South Charleston 1 1 1 Carolina 100% Puerto Rico San Sebastian 1 1 1 Modern New Zealand Dunedin 1 1 1 Total (% of total) 47 44 (94%) 25 (53%)
65 Table 5.2: High Coverage ancient and modern samples used in phylogenetic analyses (n=25).
Average Coverage Location Site Sample ID Date Total Coverage (%) (reads per base) Puerto Rico Tibes A MS10625 AD600-900 99.2 418.4 Carriacou Grand Bay MS10627 AD985-1030 99 26.8 Antigua Coconut Hall MS10630 AD1035, 1045 99.2 143.3 El Venado MS10639 AD1400-1540 93.5 228.4 Madrid Site MS10672 200-100BC 94.4 65.1 Colombia Aguazuque I MS10675 1885-1815 BC 94.5 42.1 Checua II MS10677 6200-2000BC 99.1 445.6 Moqi MS10628 Post AD1400 99.2 687.4 Moqi MS10645 Post AD1400 99.2 398.8 Moqi MS10649 Post AD1400 98.6 23.6 Peru Lo Demas MS10643 AD1480-1540 98.9 37.3 Torata Alta MS10650 AD1550-1600 99.5 1238.6 Kuelap MS10681 AD500-1400 99.5 1831.9 Kuelap MS10682 AD500-1400 99.4 2902
66 Average Coverage Location Site Sample ID Date Total Coverage (%) (reads per base) Kuelap MS10683 AD500-1400 99.5 3297.4 Peru Kuelap MS10684 AD500-1400 99.4 2406.4 Kuelap MS10685 AD500-1400 99.3 2961.1 Kala Uyuni MS10651 200BC-AD500 99.1 143.4 Kala Uyuni MS10652 200BC-AD500 99.1 442.4 Bolivia Llusco MS10653 800-200BC 99.1 478.5 Llusco MS10654 800-200BC 98.8 31.5 Belgium Mons MS10655 AD1550-1640 99.5 3253.1 South Carolina Charleston MS10671 AD1820 99.2 399.8 Puerto Rico San Sebastian MS10635 Modern 99.2 623.2 New Zealand Dunedin MrsChloe Modern 100 734
67 Table 5.3: Additional archaeological samples used in analyses of the Cytochrome B region (n=8).
Total Coverage of Location Site Date Sample ID Cytochrome B (%) Puerto Rico Tibes B AD900-1200 MS10626 99.8
El Venado AD1400-1540 MS10636 98.7
El Venado AD1400-1540 MS10637 98.3
El Venado AD1400-1540 MS10638 98.5 Colombia Checua II 6200-2000BC MS10678 100
Checua II 6200-2000BC MS10679 99.3
Checua II 6200-2000BC MS10680 99.6
Peru Moqi Post AD1400 MS10629 99.3
as current mapping software are not able to align reads through repeat regions.
For each of the Colombian samples, an unusual but consistent pattern was observed in the coverage plots (Figure 5.1B), with some regions having considerably higher coverage (up to 3000x coverage per base) than other regions (0-10x coverage per base).
Only one out of ten Colombian samples sequenced had over 99% coverage of the mitochondrial genome (MS10677), with three more (MS10639, MS10672 and
MS10675) having over 90% coverage, giving the Colombian samples a low post- sequencing success rate (33.3%) compared to other locations (Caribbean 37.5%, Peru
62.5%, Bolivia 100%, European/North America 100%). As all Colombian samples portrayed this pattern despite the differences in ages (6200BC-AD1540), this may be due to factors other than age.
68 A
B
Figure 5.1: Coverage plots of the complete mitochondrial genome. Total coverage (%), and average read depth (red line). A: MS10681 (Peru, an example of a normal coverage plot). B: MS10639, MS10672, MS10675, MS10677 (Colombia).
69 5.1.2 Contamination Analysis
To determine if any contamination had occurred from laboratory reagents or equipment, reads from all ancient and modern sequences were mapped to a composite reference genome of human, guinea pig, rat, sheep, dog, pig, and chicken (Genbank accession
IDs: NC_012920.1, NC_000884.1, NC_012389.1, NC_006853.1, NC_002008.4,
NC_00845.1, NC_001323.1). No samples contained any significant level of reads mapping to any of the above species other than guinea pig, thus the reagents and equipment used in the study were assumed to be free from contamination. In addition, no extraction blank that was processed in parallel with the ancient samples contained any contamination, indicating that no cross-contamination occurred during the process of generating and amplifying ancient DNA libraries.
5.1.3 Damage Patterns and Read Lengths
Damage patterns characteristic of ancient DNA were also analysed, along with fragment length (Jónsson et al., 2013). All ancient samples showed a higher proportion of thymine residues in the first 10 nucleotides following the 5’ strand break, and conversely a higher proportion of adenine residues within the last 10 nucleotides prior to the 3’ strand break (Figure 5.2A). In addition, a higher level of C to T transitions at the first nucleotide position of the 5’ strand is seen in the ancient samples (Figure 5.2B).
Average fragment length differs across the ancient samples, indicating various stages of degradation. For most samples, average fragment length is between 50-100bp (Figure
5.3). In summary, the short fragment lengths, DNA damage patterns, and a lack of contamination from other species indicate that the sequence data obtained represent endogenous mitochondrial DNA from ancient guinea pig specimens.
70 MS10625_collapsed_remdupA) A A C C 0.5 0.5
● 0.4 0.4 ● ● ●
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.3 ● ● 0.3
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2
Frequency ● ● ● ● ●
● 0.1 ● 0.1
0.0 0.0
G G T T 0.5 0.5
● 0.4 0.4 ● ● ●
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.3 ● ● ● ● ● 0.3 ●
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 Frequency ● ● ● ● ●
● 0.1 ● 0.1
0.0 0.0 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10 10 10 10 10 10 10 10 − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − 0.30 0.30 B) 0.25 0.25
0.20 0.20
0.15 0.15
Frequency 0.10 0.10
0.05 0.05
0.00 0.00 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 − − − − − − − − − − − − − − − − − − − − − − − − −
Figure 5.2: Nucleotide frequency at the 5' and 3' ends of reads for sample MS10625, shown as an example. A) The frequencies of the bases A, C, G and T at the first and last ten nucleotides following strand breaks (shown as grey brackets). B) Nucleotide misincorporations in the first and last 25 positions of T (red), C (grey), A (blue), G (purple).
71 MS10625_collapsed_remdup
Single−end read length distribution Single−end read length per strand + strand
800 − strand 400
600 300
400 200 Occurences Occurences subplus$Occurences
200 100
0 0 26 36 46 56 66 76 86 96 26 36 46 56 66 76 86 96 106 116 126 136 106 116 126 136 Read length Read length subplus$Length
Figure 5.3: Read length distribution for MS10625.C>T The occurrence of each fragment G>A length is plotted. The most common fragment length for this sample is 90bp.
1 1 + strand + strand 5.2 Nucleotide0.9 Diversity− strand − strand 0.9 0.8 0.8 Nucleotide diversity (π) of the 25 high coverage (>90%) complete mitogenomes and 0.7 0.7 the reference0.6 sequence (NC_000884.1) was determined using DnaSP (v5.10.1) 0.6
0.5 0.5 (Librado & Rozas 2009). Within the ancient, modern and historic samples, there are a 0.4 0.4 total of 20 mutations, 17 of which are informative. Eight of these informative mutations 0.3 0.3 occur within0.2 the HVR, one in the CytB region, and seven in the coding regions outside 0.2 Cumulative frequencies Cumulative frequencies Cumulative
0.1 0.1 of the HVR and CytB. Within the coding region, the mutations are present in sequences 0 0 encoding the 16S ribosomal subunit, NADH1, NADH5 and several transfer RNAs 0 0 10 20 30 40 50 60 70 10 20 30 40 50 60 70
mut]))) c(0, cumsum(subplus[, mut]/sum(subplus[, 72Read position mut]))) c(0, cumsum(subplus[, mut]/sum(subplus[, Read position Index Index (tRNAs). The peak nucleotide diversity is found in the HVR (π = 0.01034) (Appendix
Table 8.3). The average number of nucleotide differences between two samples in the population (k) is 5.194. The number of haplotypes with the data for complete mitogenomes, the Cyt B region and the HVR was determined using PopArt and DnaSP.
Fourteen haplotypes are present using the complete mitogenome, eight in the HVR and only one in the Cyt B region.
Nucleotide diversity is illustrated in Figure 5.4, using overlapping windows of 100bp
(step size: 25) that are centered at the midpoint. This is plotted against the coverage plot of one of the Colombian samples, MS10639. This analysis was undertaken to determine if the missing data seen in the coverage plots of Colombian samples coincided with variable positions. Sequences were also assessed in Geneious (Kearse et al., 2012) to determine which variable positions the Colombian sequences may have been lacking.
All Colombian samples had missing data in at least two variable positions, and had lower coverage in regions with variable sites between the sequences.
Nucleotide diversity was also assessed within each of the geographic groups represented in the study: Caribbean, Colombia, Peru, Bolivia and European/Modern
(including the Charleston and modern Caribbean samples) (Table 5.4). The highest nucleotide diversity (π) and number of differences between two samples (k) within a group is seen in the Colombian samples. There is more diversity within the Colombian samples than across the whole population, which may reflect the antiquity of the samples. The lowest diversity is seen within the Caribbean and Bolivian samples,
73 MS10639) 3000" 0.012" Coverage"
Pi"
2500" 0.01"
2000" 0.008" Nucelo'de)Diversity)(
1500" 0.006" Coverage)per)bp) π ))
1000" 0.004"
500" 0.002"
0" 0" 0" 1000" 2000" 3000" 4000" 5000" 6000" 7000" 8000" 9000" 10000" 11000" 12000" 13000" 14000" 15000" 16000" bp)
Figure 5.4: Nucleotide diversity (π) plotted on coverage distribution for one of the Colombian samples (MS10639). Coverage distribution is filled in grey, with nucleotide diversity plotted in blue. The majority of variable sites occur in low coverage or missing regions in Colombian samples.
Table 5.4: Nucleotide diversity statistics for each geographic group. Showing nucleotide diversity (π), the average number of differences between two sequences (k), the number of polymorphic sites (S) and the number of haplotypes (h) within each of the geographic groups: Caribbean, Colombia, Peru, Bolivia and European/Modern, as well as all samples combined.
Number of Group π k S h sequences
Caribbean 3 0.00008 1.333 2 2
Colombia 4 0.00069 7.833 13 4
Peru 10 0.0001 1.600 3 4
Bolivia 4 0.00007 1.167 2 3
European/Modern 5 0.00037 6.200 15 5
All 26 0.00047 5.194 20 14
74 possibly reflecting the bottleneck situation involved with translocation to the Caribbean and proximity of the Bolivian sites to each other.
Nucleotide diversity of the Cyt B region was assessed for the modern and ancient populations, using an additional 65 modern domestic guinea pig (C. porcellus) sequences from across South America (Cyt B sequences obtained from Genbank,
Appendix Table 8.1 for accession IDs). The additional ancient samples (Table 5.3) were also used in this analysis as they had adequate coverage of the Cyt B region
(>98%), therefore a total of 99 domestic guinea pig sequences were analysed. There were 75 mutations in the modern Cyt B population (n=70), and only one mutation within the ancient samples (n=29), suggesting recent expansion of the domestic population.
5.3 Phylogenetic Analyses
5.3.1 Median Joining Network of Complete Mitogenomes
A median joining (MJ) network was produced for the 26 complete mitogenomes using
PopArt (Leigh and Bryant, 2015). Four major haplogroups based on geographic area are represented by fourteen haplotypes, eleven of which contain ancient samples
(Figure 5.5). The first major haplogroup contains four haplotypes comprising samples from the Caribbean, Peru and Charleston. Northern and southern Caribbean samples are split into two haplotypes with northern and coastal/southern Peruvian samples respectively. The northern Peru and Caribbean haplotype is also shared by the historic
Charleston sample. The second major haplogroup represents all four Bolivian samples.
75 3
1
4
2
Key:( Northern(Caribbean( Northern(Peru( Modern(European( (Puerto(Rico,(An4gua)( Southern(Peru( Modern(Caribbean( Southern(Caribbean( Bolivia( Belgium( (Carriacou)( Colombia( Charleston(
Figure 5.5: Median Joining network of ancient, historic and modern complete Cavia porcellus mitogenomes. Showing four main haplogroups 1: Peru, Caribbean and Charleston, 2: Bolivia, 3: Modern and Historic European, 4: Colombia. Mutations between samples are denoted by hatch marks. Haplotypes are coloured based on geographic location. Size of circle corresponds to the number of sequences that share that haplotype. Black circles are indicative of intermediate haplotypes not represented by samples in the analysis.
76 The third haplogroup contains the modern and historic European samples, as well as the modern Caribbean sample. The fourth haplogroup consists of samples from Colombia.
5.3.2 Bayesian and Maximum Likelihood Tree of Complete Mitogenomes
PopArt does not include variable sites where there is missing data, such as positions missing from within the Colombian sequences. Consensus Bayesian and Maximum likelihood (ML) trees were constructed, which do not remove variable sites where there is missing data. Both the Bayesian and ML methods displayed similar trees, thus the
Bayesian tree is presented here, with either posterior probability or bootstrap support over 0.7/70 respectively (Figure 5.6). The tree also displayed geographically separated clades evident in the MJ network. The Colombian samples form the most distant clade, followed by the historic and modern European samples. The Bolivian and Peruvian,
Caribbean and Charleston clades are the most closely related.
5.3.3 Cytochrome B Maximum Likelihood Tree
Due to the diversity within the Colombian samples and their distinct pattern of coverage, a ML tree was created using sequences of the Cyt B region from domestic and wild guinea pigs to determine whether the Colombian samples were domestic or wild samples, with 105 additional samples acquired from Genbank (Appendix Table
8.1), and 29 ancient sequences (Figure 5.7). All ancient samples, including the
Colombians, fell within a clade of modern Colombian and Peruvian domestic guinea pig sequences. As all ancient Colombian samples fell within domesticated guinea pig sequences, it appears that they may be domestic guinea pig, C porcellus. However, since the Colombian samples underwent targeted-capture for the domestic rather than wild guinea, this may have hindered the analysis. Full mitogenome sequencing of wild
77 0.73/42' MS10672' 0.91/80' MS10639' 1" 1/64' MS10675'
MS10677'
0.55/79'0.55/79' MrsChloe' 0.94/97'0.94/97' MS10635' 1/92' 2" MS10655' NC_000884.1' MS10682' MS10625' MS10684' MS10683' MS10630' 0.59/80' MS10671' 0.98/76' MS10685' MS10681' 0.98/72' MS10645' 3"
0.82/59' MS10627' MS10643'
MS10649' 0.63/72' 0.77/62' MS10628' Key:' MS10650' Northern'Caribbean' Northern'Peru' Modern'European' MS10654' (Puerto'Rico,'AnEgua)' 0.94/57' Southern'Peru' Modern'Caribbean' 0.76/77' MS10651' Southern'Caribbean' Bolivia' Belgium' 4" (Carriacou)' MS10653' Colombia' Charleston' MS10652' 6.0EQ5'subsEtuEons/site'6.0E-5 Figure 5.6: Bayesian Tree with bootstrap support for ancient, historic and modern complete mitogenomes. Branch support where either posterior probability or bootstrap support is above 0.7/70 respectively is displayed. Four main geographic haplogroups are seen; 1: Colombia, 2: Modern and historic European, 3:Peru, Caribbean, Charleston, 4: Bolivia.
78 AY382790.1_ 98/1$ GU136753.1 91/0.99$ GU136742.1_ GU136752.1 100/1$ AY382791.1_ AY382792.1_ GU136748.1_ C.#aperea# 100/1$ GU136751.1_ 100/1$ GU136749.1_ GU136750.1_ GU136756.1 99/1$ 100/1$ GU136757.1_ GU136758.1_ GU136759.1_ 100/1$ GU136726.1_ GU136731.1_ 100/1$ AY245099.1_ GU136730.1_ DQ017048.1_ 94/1$ 99/1$ 93/0.99$ DQ017051.1_ C.#tschudii# DQ017050.1_ 84/0.83$ DQ017049.1_ 99/1$ DQ017052.1 DQ017053.1_ 88/0.42$ GU136729.1_ AY228361.1_ AY247008.1_ DQ017037.1_ HM447150.1_ HM447182.1_ HM447174.1 DQ017044.1_ DQ017045.1_ 95/0.99$ DQ017046.1_ 90/0.85$ DQ017047.1_ GU136732.1_ HM447146.1_ 85/1$ DQ017041.1_ AY245096.1_ 1% AY245097.1_ AY245098.1_ 83/0.39$ DQ017042.1_ DQ017043.1_ GU136727.1_ HM447163.1_ AY382793.1_ *$ DQ017039.1_ DQ017040.1_ 99/1$ *97/0.41$ MS10675_6 MS10636_6 MS10639_6 MS10672_6 MS10649_6 MS10680_6 MS10625_6 100/1$ HM447155.1_ MS10685_6 MS10652_6 MS10682_6 HM447152.1_ HM447173.1_ HM447147.1_ HM447157.1_ MS10654_6 MS10630_6 HM447161.1_ HM447187.1_ MS10653_6 HM447171.1_ C.# HM447148.1_ HM447177.1_ MS10681_6 porcellus# MS10651_6 MS10650_6 MS10643_6 HM447168.1_ MS10683_6 MS10628_6 HM447158.1_ GU136728.1_ HM447186.1_ HM447170.1_ HM447169.1_ 2% HM447156.1_ MS10645_6 HM447172.1_ MS10684_6 HM447167.1_ MS10671_6 MS10677_6 MS10627_6 gi_5835988_ref_NC_000884.1_ MS10678_6 HM447164.1_ MS10635_6 HM447159.1_ Key:% MS10655_6 MrsChloe MS10638_6 DQ017038.1_ MS10679_6 3% C.#aperea# MS10626_6 *$ AY228362.1_ *82$ AY245095.1_ AY245094.1_ HM447165.1_ C.#aperea#pamparum# HM447166.1_ MS10637_6 HM447179.1_ MS10629_6 C.#anolaimae/C.#aperea#guianae# HM447153.1_ HM447149.1_ HM447183.1_ HM447184.1_ HM447160.1_ C.#tschudii# GU136733.1_ HM447175.1_ HM447162.1_ HM447178.1_ C.#porcellus%# HM447185.1_ HM447180.1_ HM447181.1_ HM447151.1_ C.#magna# HM447154.1_ _AF490405.1_ HM447176.1_ 100/1$ GU136734.1_ GU136735.1_ GU136736.1_ C.#magna#
0.02$subs1tu1ons/site$0.02
Figure 5.7: Maximum Likelihood Tree (with Bayesian posterior probabilities) demonstrating the relationship between modern (n=105) and ancient (n=29) domesticated guinea pig cytochrome B regions. Branch support where either posterior probability or bootstrap support is above 0.7/70 respectively is displayed. Grey shading represents archaeological samples. Three distinct clades are seen within C. porcellus labelled: 1 (samples from Peru, Bolivia, Chile, Ecuador, Colombia), 2 (samples from Colombia, Peru, archaeological samples)-3 (modern laboratory and Peru market samples).
79 guinea pigs will ensure the identification of the Colombian samples, as will be discussed in the chapter 6.
Also identified in the Cyt B ML tree is two distinct clades within C. porcellus. The first contains modern sequences from Peru, Bolivia, Ecuador, Chile and Colombia (1,
Figure 5.7). The second contains all the ancient sequences, as well as modern sequences from predominantly Colombia, and Peruvian markets (2, Figure 5.7).
Additionally, the Peruvian market sequences fit in a subclade with the modern/European samples analysed in this thesis (3, Figure 5.7).
5.3.4 Cytochrome B Temporal Network
A temporal network was created using Cyt B sequences from modern (n=68) and ancient (n=29) domestic guinea pigs (Figure 5.8). A list of the geographic areas represented by each numbered haplogroup can be found in the Appendix (Table 8.4).
The temporal network revealed that there are two haplogroups in the ancient dataset, both of which are also found in the modern dataset. The first (Haplogroup 2, Figure
5.8) contains sequences ancient sequences from Colombia and Puerto Rico, and modern sequences from Colombia, and Europe. The second (Haplogroup 3, Figure 5.8) contains ancient sequences from Puerto Rico, Antigua, Carriacou, Colombia, Peru, and
Bolivia, as well as modern sequences from Colombia and Charleston. The temporal network shows a clear affinity between the ancient samples and those of modern-day
Colombia, suggesting an ancestral link. No ancestral haplogroups were found for modern day southern Peruvian, Chilean, Ecuadorian and Bolivian samples despite Peru and Bolivia being covered by ancient samples.
80
Figure 5.8: Temporal Network of Cytochrome B sequences from modern (n=68) and ancient (n=29) domestic guinea pigs. Pink circles represent ancient (pre- AD1500) sequences and the grey circles represent modern (post-AD1500) sequences. The black dots indicate mutations. The white dots represent haplogroups in the modern dataset that are not present in the ancient dataset. The blue square represents modern Colombian sequences. The black square indicates modern European or Peruvian market sequences. The other modern haplogroups represent sequences from Peru, Colombia, Chile, Bolivia and Ecuador.
81 Chapter 6: Discussion
This study is the first to use complete mitogenomes of a commensal animal to explore interaction in the late prehistoric Caribbean region. It has highlighted the use of ancient
DNA, complete mitogenome sequencing through next generation sequencing, and the application of the commensal model. The origins of Caribbean, European and North
American guinea pigs will be discussed based on phylogenetic relationships observed in complete mitogenome and Cyt B analyses. This study adds to the growing literature highlighting broader-scale interaction networks within the Caribbean in late prehistory.
6.1 Preservation of Ancient DNA
The preservation of ancient DNA differed greatly between the archaeological samples.
Environmental conditions during deposition, age, and post-excavation handling can affect the preservation and recovery of ancient DNA (Pruvost et al., 2007). In particular, environmental conditions heavily influence the preservation of DNA in archaeological specimens (Hofreiter et al., 2001b; Adler et al., 2011). Tropical environments, such as the Caribbean islands, increase the rate at which DNA is degraded by molecular processes, whereas cold and dry environments slow the degradation of DNA by preventing the activity of nucleases that fragment DNA (Smith et al., 2003; Collins et al., 2002).
In this study, complete mitogenomes were recovered from three of the eight archaeological specimens from the Caribbean Islands. This low success rate (37.5%) is likely due to poor conditions for preservation of DNA in the island environment. The
DNA in samples excavated from the Coconut Walk site on Nevis and Giraudy, St Lucia
82 would have been too fragmented to be recovered. Additionally, of three samples from the coastal site of Lo Demas in the Chincha Valley, Peru, sufficient DNA sequence was only recovered from one sample (success rate of 33.3%), despite these samples being mummified and some of the youngest prehistoric samples in this study. The preservation results from this study are consistent with previous Caribbean ancient
DNA studies (Lalueza-Fox et al., 2003; Mendisco et al., 2015). Mendisco et al. (2015) reported an overall success rate of 34% for the amplification of DNA from human burials in Guadeloupe, and for Cuba, Lalueza-Fox et al. (2003) found an overall success rate of 38.3%. Tropical island or coastal environments are not conducive for
DNA preservation, as shown in the Pacific (Knapp et al., 2012b), where archaeological remains between 200-700 years old had an average fragment size of 70bp, and the success rate of amplification of DNA from samples was also low.
In contrast to low DNA preservation in the Caribbean and coastal Peru, the highlands of Peru and Bolivia showed better preservation of DNA with complete mitogenomes recovered from 70% of samples from these regions. The Andean highlands have a lower average temperature (12-15°C) compared to the Pacific coast of Peru (25°C). The dry environment aids in the preservation of the archaeological specimens (Spotorno et al., 2007). DNA is more likely to survive under these conditions compared to the warmer environments of the Caribbean and coastal South America, as cooler and drier conditions inhibit the activity of the nucleases that fragment DNA (Smith et al., 2003).
6.2 Possible Misidentification of Guinea Pig Samples
The lowest success rate (26.6%) for any site in terms of recovering complete mitogenomes was from Colombia, despite more favourable site conditions for
83 preservation (high altitude, dry environment) than in other areas in this study. Three samples failed to produce DNA libraries altogether, as evidenced by the presence of adapter dimers and lack of a sufficient peak in the expected library region. However, it is likely that the low success rate of Colombian samples is not due to preservation issues related to age or environment, as will be discussed below.
All Colombian samples showed an identical, distinct pattern of coverage and no samples from other geographical areas in the study shared this pattern (refer to Figure
5.1B). As all Colombian samples were extracted in separate batches, captured at different times, and sequenced in different runs and on different platforms (HiSeq and
MiSeq), it can be ruled out that this patterning is due to issues during processing. The consistent gaps in coverage suggest that targeted-capture of the mitogenomes using domestic guinea pig DNA was not able to fully capture several regions, possibly due to the Colombian sequences being too different to the bait sequence at these regions.
Higher coverage was observed in regions that typically are more conserved between species or within mammals, such as tRNA coding regions (Figure 5.1). Such regions are more conserved because they are vital to basic organism function through their role in translating RNA sequence to amino acids (Ojala et al., 1981). In regions of the guinea pig mitochondrial genome demonstrated to be more variable, the Colombian samples have lower or no coverage (refer to Figure 5.4). Hypotheses for the lack of coverage at variable positions may be that these samples represent one of the wild species (or subspecies) of guinea pig that exist in South America, possibly C. tschudii or C. aperea, or that they represent a different species of rodent present in the archaeological sites. Each of these three explanations will be discussed further below.
84 Three of four Colombian samples are considerably older than most of the other samples used in the study, dating to the pre-ceramic period and overlapping proposed dates for when guinea pig domestication is thought to have occurred. Wing (1986) proposed that the domestication of guinea pigs might have occurred as early as 5000BC, but definitely by 2500BC. Colombian archaeological remains of guinea pigs from Checua
II (6000-2000BC) may predate the suggested domestication age of 2500BC and therefore could represent the wild predecessor of the domestic population (C. tschudii) or an early form of the domestic species. However, the same distinct coverage pattern is also observed in Colombian guinea pigs from El Venado, a site that dates to AD1400-
1540, which is well after domestication occurred. It may be possible that the prehistoric communities of the Colombian highlands continued to utilise a wild population of guinea pigs after the domestication event occurred.
Studies by Dunnum and Salazar-Bravo (2010), Spotorno et al. (2006), Spotorno et al.
(2007) proposed that the domestication of guinea pigs occurred in Peru, although
Colombia was also suggested as a possible domestication area (Spotorno et al., 2004).
However, the current distribution patterns of the recognised species and subspecies of the Cavia genus show that the wild C. tschudii is restricted to the Andean region of
Peru, Bolivia, and northern Chile and Argentina (Dunnum and Salazar-Bravo, 2010).
There are two other wild species of guinea pig (C. aperea guianae and C. anolaimae) that currently inhabit the northern South America region, including Colombia.
Although distribution patterns today may not reflect those in the past, other candidate species with a known distribution in Colombia may provide a better explanation for the missing data in the Colombian samples.
85 In Colombia there are two species of wild guinea pig (C. anolaimae and C. aperea guianae) along with the domestic C. porcellus, that occupy the region today (Dunnum and Salazar-Bravo, 2010). C. anolaimae is more prevalent in the highland regions, particularly Cundinamarca (which encompasses the region in which some of the samples in this thesis were recovered), and C. aperea guianae is found in the lowland areas of Colombia, Venezuela and Guiana. Both of the wild species are thought to be derived from, or closely related to C. aperea, which is found in Brazil and southern
Bolivia (Woods and Kilpatrick, 2005). In a phylogenetic analysis of the cytochrome B region undertaken by Dunnum and Salazar-Bravo (2010) both C. anolaimae and C. aperea guianae are shown to be part of the C. aperea clade, which forms a sister clade to C. tschudii (including C. porcellus). The analysis was not able to determine between
C. anolaimae and C. aperea guianae, though only one sequence of C. anolaimae was analysed in the study and the specimens were acquired from nearby areas. Thus further analyses will be required to determine the relationship between the two possible subspecies. The modern distribution of C. anolaimae occurs in the same region of
Colombia as the sites from which the Colombian archaeological specimens were excavated. This implies that the Colombian archaeological specimens may belong to C. anolaimae rather than C. porcellus.
In the literature, the term ‘C. anolaimae’ has previously been synonymous with the domestic form, C. porcellus, including C. anolaimae which has been described as an
“escaped” or feral form of the domestic guinea pig (Cabrera, 1961; Eisenberg and
Redford, 1992; Woods and Kilpatrick, 2005). Alternatively, C. anolaimae could be a domesticated form of C. aperea, which would likely carry several similar morphological characteristics as C. porcellus, due to the domestication process. These
86 characteristics inlcude: larger overall size, reduction in brain size, and coat colour polymorphisms (Spotorno et al., 2007). Dunnum and Salazar-Bravo (2010) described a general lack of morphological diversity within the Cavia genus, with the exception of
C. magna that evolved aquatic abilities and is genetically distant from all other Cavia species. Pictet and Ferrero (1951) and Rood (1972) described the result of fertile offspring from crosses between C. aperea and C. porcellus, suggesting that hybrids between the species may exist in the wild where species ranges overlap. This further supports a lack of distinction between the two species. Therefore, due to the current distribution of wild and domestic Cavia species and the similarity in morphology between C. anolaimae and C. porcellus, it is possible that the Colombian samples in the present study may be representatives of C. anolaimae (or C. aperea guianae if they are the same species), or a hybrid between C. anolaimae and C. porcellus.
It can be difficult to identify the correct species from fragmentary faunal remains in the archaeological record, therefore a third explanation is that the Colombian samples may represent another species of rodent entirely. Several other small mammals were also present in prehistoric sites in Colombia. At Checua II and Aguazuque I, archaeological remains of a variety of small mammals such as rat, agouti (Dasyprocta sp.) and pecarry
(Tayassu pecari) are found in conjunction with those identified as domestic guinea pig
(Correal Urrego, 1990; de Mahecha, 1992). The majority of these animals are typically larger than a domestic or wild guinea pig, though the size of juvenile individuals of the above species may overlap with adult guinea pigs.
Genetic analyses have previously corrected misidentifications of zooarchaeological material, both between species and also within wild and domestic forms of a species
87 (Horsburgh et al., 2016). Orton et al. (2013) determined that a maxilla from an archaeological site South Africa, thought to be early evidence of domestic cattle in
Africa was in fact gemsbok (Oryx gazella). Likewise, ancient genetic analyses showed that archaeological specimens thought to represent domestic sheep were actually representative of several species of springbok (Horsburgh and Moreno-Mayar, 2015). A study conducted by Yang et al. (2005) was able to expand the known ecological range for snowshoe hare (Lepus americanus) in Colorado, based on misidentification of
Lepus californicus. Interestingly, based on analysis of Cyt B sequences, two specimens of C. tschudii may have been morphologically incorrectly identified as they fall within the C. porcellus clades, separate to the other C. tschudii subspecies (Figure 5.7).
Dunnum and Salazar-Bravo (2010) suggested these two samples represent the wild progenitor to the domestic form, though they could also represent a hybrid between C. porcellus and C. tschudii that is phenotypically ‘wild’ and genetically ‘domestic’, or they could be the result of a misidentification of the wild species.
Misidentification is more likely to occur with very fragmentary remains or between two closely related species that are morphologically similar (Horsburgh et al., 2016). Many archaeological faunal remains are fragmentary, thus identification at the species or subspecies level can be a challenge. This is especially true when trying to categorise remains from morphologically similar species. Almost all archaeological specimens presented in this thesis were recovered from midden sites, where fragmentary remains are common as a result of processing for consumption. Many of these specimens were incomplete portions of elements such as a mandible, maxilla or tooth. Thus the fragmentary nature of the remains may have hindered identification at the species or subspecies level.
88
Another complication in identifying between species or subspecies based on morphological data is the range of variation exhibited within a species or population. In the case where species or subspecies have similar morphological characteristics, such as the guinea pigs, the range in morphological measurements of the two may overlap.
In this case, it would be impractical to identify the remains at the species or subspecies level for risk of misidentification. Many archaeological studies in South America and the Caribbean refer to guinea pig remains as ‘Cavia’, and make the assumption that they are of the domestic species, however in the Colombian highlands, this may not necessarily be the case. For sites that may overlap with the time during which domestication occurred throughout the Andean region or in sites in an area where two species may exist, identification based on morphology alone should be undertaken with caution.
It was determined through comparison of the Cyt B region between the archaeological remains, and modern wild and domestic Cavia species, that the ancient Colombian samples affiliate most closely with modern domestic Colombian or European samples
(Figures 5.7, 5.8). This observation may have been hindered by missing data in the
Colombian samples due to targeted-capture of the DNA libraries with a different species. In order to consolidate which species the Colombian samples belong, further analyses will be required, such as the generation of complete mitogenome sequencing of wild guinea pigs (C. tschudii, C. anolaimae, C. aperea guianae) and other rodents that may have occurred in the same archaeological sites, as these data are not currently available.
89 6.3 Domestication of Guinea Pigs
From the Cyt B ML Tree and the Temporal Network (Figures 5.7, 5.8), there appears to be two distinct clades within C. porcellus. One is comprised of all the ancient sequences presented in this thesis, and modern sequences from southeastern Colombia, coastal Peru, and Europe. The second clade consists of modern samples from the
Andean regions of Peru, Bolivia, Ecuador, and Chile. The appearance of two clades may be evidence of two domestication events occuring in the past. Spotorno et al.
(2004) previously suggested that Colombia may have been a centre of domestication, although later research pointed to a domestication centre in Peru (Dunnum and Salazar-
Bravo, 2010; Spotorno et al., 2007; Spotorno et al., 2006). Given the modern distribution pattern of the wild C. tschudii, it is unlikely that a domestication event occurred in Colombia, but that the modern Colombian samples probably represent an expansion of the guinea pig populations associated with prehistoric trade through this region.
All of the ancient samples from Peru and Bolivia are closely related, thus it is possible that they represent populations that were translocated throughout these regions. It is assumed that these populations moved through Colombia, thus the modern Colombian populations may be remnants of these networks. The European population appears to represent a subset of this domestication event. This domestication event may have occurred in the Andean region, possibly Bolivia. In contrast, the modern samples from inland Peru, Ecuador, Chile and Bolivia may represent a second or concurrent domestication event that occurred in the area encompassing the Andean region of southern Peru and northern Chile and is subsequently exhibited in modern populations throughout the Andean region.
90
The timings and locations of these two possible domestication events are unclear and further research will be required to resolve these questions. Complete mitogenomes of modern and additional ancient domestic guinea pigs will help to determine if the Cyt B results do represent two domestication events, and clarify the timings and nature of the the two possible events. Complete mitogenomes provide better resolution of the relationships within a species than the Cyt B region, as will be discussed below.
6.4 Complete Mitogenomes
Most previous phylogenetic studies of the guinea pig have been limited to analyses of the Cyt B or short fragments of the HVR of modern samples, focusing on determining the wild predecessor to the domestic species (Burgos-Paz et al., 2011; Dunnum and
Salazar-Bravo, 2010; Spotorno et al., 2007; Spotorno et al., 2006; Spotorno et al.,
2004). Kimura et al. (2016) used 288bp of the Cyt B region and 79bp of the HVR in order to determine phylogenetic relationships within prehistoric guinea pigs from the
Caribbean. The HVR has been used frequently in studies to determine variation within a population or species. This is due to the HVR having a higher rate of mutation than elsewhere in the mitochondrial genome (Stoneking et al., 1991). The Cyt B region has a lower mutation rate than the HVR, thus is most often used for ‘barcoding’ projects to identify species, rather than examine variation within a species (Clare et al., 2007;
Hebert et al., 2003; Yacoub et al., 2015). Recent studies have shown that short fragments of mitochondrial DNA such as the HVR and Cyt B do not contain the full level of variation within or between populations (Benton et al., 2012; Greig et al., 2015;
Knapp et al., 2012b). Instead, complete mitogenomes are now being utilised to further investigate the variation within populations.
91
Figure 5.4 illustrates the level of nucleotide diversity across the mitochondrial genome of archaeological guinea pigs analysed in this study. Unsurprisingly, the highest level of variation is seen in the HVR. As this region mutates faster in comparison to the coding regions of the mitochondrial genome, it is expected that it will accumulate more mutations and thus have a higher level of diversity (Stoneking et al., 1991). Low diversity is seen within the Cyt B region, with only one mutation separating the
European guinea pigs from the South American, Caribbean and Charleston samples. In comparison, increased diversity is seen within modern Cyt B sequences (Figure 5.7), suggesting recent expansion of populations for meat markets in South America has had an impact on the genetic diversification within the Cyt B region (Spotorno et al., 2006).
Analysis of phylogenetic relationships based on the HVR alone revealed eight haplotypes, compared to the fourteen present in analysis of the complete mitogenomes.
Expansion in the haplotypes occurred within the Peruvian and Caribbean clades and within the European and Bolivian sequences. Therefore, despite the HVR being the most variable region of the mitogenome, mutations outside of the HVR are also crucial in determining correct phylogenetic relationships. This has also been shown in complete mitogenome analyses of modern and ancient human studies undertaken in the
Pacific by Benton et al. (2012); Duggan et al. (2014); Knapp et al. (2012b), which expanded the B4a1a1 haplogroup. Llamas et al.’s (2016) analysis of complete mitogenomes of ancient human remains in South America has provided an updated view of the haplotypes present in the pre-Columbian mainland. The results of this thesis agree with those in the recent literature, highlighting the value of complete mitogenome sequencing for phylogenetic studies since important variation is found outside of short
92 fragments of the mitochondrial genome.
6.5 Guinea Pig Origins and Implications for Late Ceramic Age Interaction
6.5.1 Origins of Caribbean Guinea Pigs
The earliest evidence and the greatest number of guinea pig remains in the Caribbean are found in the Greater Antilles, in particular Puerto Rico and Vieques (LeFebvre and deFrance, 2014; Wing, 2000). Wing (2000) suggests that the first introductions of guinea pigs to the Caribbean occurred in the Greater Antilles, before being translocated later in the ceramic age to Hispaniola and Jamaica and south to the Lesser Antilles.
However, it was unknown where in South America the Caribbean guinea pigs may have been translocated from. The identification of a haplogroup in the Caribbean that is shared with the mainland (Colombia, Peru or Bolivia) may represent a potential origin of Caribbean guinea pigs, and therefore potential interaction spheres between the island archipelago and the mainland during this period. Based on recent archaeological and genetic evidence for migration into the Caribbean in this period, I hypothesised that the ancient Caribbean guinea pigs would be most closely related to ancient guinea pigs from Colombia.
From the MJ network and ML tree of complete mitogenomes of archaeological and modern guinea pigs (Figures 5.5, 5.6), it is clear that there are two separate haplogroups for archaeological guinea pigs present in the Caribbean. The first is seen in the Greater
Antilles and the Northern Lesser Antilles, at sites from Puerto Rico (Tibes A) and
Antigua (Coconut Hall) (Figure 6.1). This haplogroup is shared by samples from
Kuelap in northern Peru and the historic specimen from Charleston, USA. The second
93 haplogroup is present in Carriacou, in the Southern Lesser Antilles, and is shared with a sample from Lo Demas in the Chincha Valley, Peru. The presence of two separate haplogroups in the Caribbean is possibly indicative of at least two introductions of guinea pigs to the region, although it may also represent the genetic diversity present within a single introduction
Current archaeological evidence shows the transformation of Saladoid to Ostionoid ceramics in Puerto Rico from approximately AD500 (Fitzpatrick, 2006; Hofman et al.,
2014a; Keegan, 2000). Routes for the transportation of raw materials throughout the pan-Caribbean region during this time may have also played an important role in the translocation of guinea pigs to the Greater and Lesser Antilles. Interaction from Puerto
Rico to the western Greater Antilles occurred with the introduction of Ostionoid ceramics to the Dominican Republic, Haiti, Jamaica, Cuba, and the Bahamas and
Caicos (Keegan, 2000), where evidence of guinea pigs occur post-AD1000. Interaction is also seen south to the northern Lesser Antilles, as the Mamoran Troumassoid styles on Antigua have stylistic affiliations to the Ostionoid subseries in eastern Puerto Rico
(Figure 6.1).
The presence of a shared guinea pig haplogroup between the northern Caribbean
(Puerto Rico and Antigua) and northern Peru (Kuelap) samples may represent interaction between mainland communities prior to translocation to the Caribbean. This conflicts with my hypothesis based on current archaeological and genetic data that suggest a possible Colombian origin for Caribbean guinea pigs. However, it can be inferred that translocation may have occurred through an inland or a coastal route, through Colombia, prior to expansion into the Caribbean (Figure 6.1). To date
94 CUBA! TURKS & CAICOS ISLANDS!
AD1000% ANGUILLA! HISPANIOLA! AD1000% PUERTO RICO! ST MARTIN! JAMAICA! ST KITTS ANTIGUA! Greater Antilles! & NEVIS! Northern GUADELOUPE! GUATEMALA! CARIBBEAN Lesser Antilles! HONDURAS! SEA! MARTINIQUE!
Southern ST LUCIA! BARBADOS! Lesser NICARAGUA! Antilles! CARRIACOU!
AD1000% TOBAGO! COSTA RICA! TRINIDAD!
From% VENEZUELA! PANAMA! Peru% COLOMBIA! Key:% Guatemalan%Jadeitite% St%Martin%Greenstone% Interaction%spheres% Cuban%Jadeitite% Long%Island%Flint% Possible%guinea% Guanin% Caliviny%Ceramics% pig%translocations%
Figure 6.1: Possible late ceramic age translocation routes and timings of the guinea pig into the Greater and Lesser Antilles.
95 no ancient guinea pig remains have been recovered from coastal Colombia, so this cannot yet be confirmed through archaeological or genetic evidence of guinea pigs, though there is archaeological evidence for interaction in this location that may have contributed to the movement of guinea pigs through the area.
Interaction networks established in the early ceramic period, such as the introduction of metallurgy (guanin) to the Greater Antilles via Colombia (Cooke, 2005; Rodriguez-
Ramos et al., 2010), were likely used to transport guinea pig and other non-local fauna to Puerto Rico. Later interaction between the Ostionoid societies of Puerto Rico and the
Mamoran Troumassoid societies of Antigua allowed for the introduction of guinea pig to the northern Lesser Antilles (Figure 6.1). This likely occurred through the previously established jadeitite and serpentinite networks (Hofman and Bright, 2010; Rodriguez-
Ramos et al., 2010; Schertl et al., 2007). Extensive interaction networks with the
Isthmo-Colombian and mainland South American regions likely led to the rise of complex, hierarchal societies that developed during the late ceramic age in the
Caribbean (Hofman and Bright, 2010). No previous archaeological or genetic studies have described a link between the pre-Columbian societies of Peru and the Caribbean.
Pre-Columbian societies of South America were not restricted by today’s geographic boundaries, thus future studies should aim to use samples from a broader region including Venezuela, Colombia, Ecuador and Peru in studies of exchange and migration through the region.
The presence of a second haplogroup in the southern Lesser Antilles (Carriacou) was likely the result of a separate migration from mainland South America to the Lesser
Antilles, during a period (from AD1000) when an interaction encompassing these two
96 regions occurred archaeologically (Figure 6.1). The sample from Carricacou dates to
AD985-1030. A southern Lesser Antilles-mainland South America interactionsphere in the late ceramic period is well documented in the archaeological record with similar ceramic styles being transported between these regions (Hofman, 2013; Hofman and
Hoogland, 2011). The Troumassoid ceramic culture present in the southern Lesser
Antilles, developed from AD500, has stylistic links to the Araquanoid of Trinidad and the Orinoco Valley in Venezuela. The Suazan Troumassoid developed out of the
Troumassan around AD1000. In addition, the presence of a distinct style of ceramics, the Caliviny, appeared in St Vincent, St Lucia, Martinique, Grenadines, Barbados, and
Tobago, following the introduction of the Troumassoid ceramics (Keegan, 2000).
In contrast, it is possible that this second haplotype was introduced as part of a single migration to Puerto Rico, and then translocated down to the southern Lesser Antilles through trade networks within this region (such as St Martin greenstone) (Rodriguez-
Ramos et al., 2010). Given the small size of the ancient Caribbean dataset it may be difficult to determine whether the haplotypes seen in the Caribbean are indicative of a single or multiple migrations. However, societies in the Greater Antilles and southern
Lesser Antilles showed marked regionalisation during this period, in terms of ceramics
(Ostionoid in the north, Troumassoid in the south). Therefore, due to the affinity of this haplotype to a different region in Peru than the northern haplotype, and the increasing levels of trade and exchange between the southern Lesser Antilles and South America during the latter late ceramic age, it is likely that this is a result of a second introduction. The migration of people, and thus guinea pigs, probably occurred through a coastal route from Peru to Colombia and Venezuela, then to the southern Lesser
Antilles (Figure 6.1). This same route likely led to the introduction of the ‘Island Carib’
97 people, described by Mendiscio et al. (2015), later in prehistory. It may be expected that the Island Carib are genetically similar to ancient human populations in Guiana and northern Venezuela, though further genetic analyses will have to be undertaken to analyse this.
The affiliations of Caribbean guinea pigs to those of Peru are likely indicative of interaction within mainland communities prior to expansion into the Caribbean. Most research regarding migration into the Caribbean focuses on the origin of the ceramic societies in the Venezuelan region and the lithic items transported from the Isthmo-
Colombian region, and does not consider interaction in the mainland prior to expansion to the Caribbean (Hofman et al., 2008; Keegan, 2000; Petersen et al., 2004; Rodriguez-
Ramos et al., 2010; Rouse, 1989b). It would be expected that if interaction occurred through Colombia, the Caribbean guinea pigs would be genetically most similar to the
Colombian guinea pigs. However, as previously discussed in Section 6.2, Colombian guinea pigs in this analysis do not appear to be representatives of the domestic guinea pig that was translocated to the Caribbean. C. porcellus was probably introduced into coastal Colombia following increased interactions in the area, during the ceramic age, as this species is present there today. No archaeological specimens of the domestic guinea pig have been recorded for coastal Colombia or Venezuela (LeFebvre and deFrance, 2014), thus these regions were not able to be included in this study, although archaeological remains of C. aperea or C. guianae have been recovered from the
Venezuelan highlands (Garson, 1980). Despite this, it is suggested from the analysis of complete mitogenomes of ancient South American and Caribbean guinea pigs that domestic guinea pigs were first introduced into the Greater Antilles via previously established trade networks, likely from Colombia, and secondly to the Lesser Antilles
98 following later migrations of ceramic populations from Venezuela and Guiana.
6.5.2 Historic Movements of Guinea Pigs
Spotorno et al. (2006) attempted to determine the origins of European guinea pigs based on Cyt B sequences, and found that they most likely came from southern Peru or northern Chile. A definitive origin for European guinea pigs could not be established from the results of this thesis, as no haplotypes were shared between the European and ancient South American or Caribbean samples. However, it is most likely that they came from an area encompassing southern Peru and northern Bolivia as they are most closely related to ancient guinea pigs from these areas (Figure 5.5, 5.6). The present study did not include any ancient guinea pigs from Chile, thus it cannot be ruled out that guinea pigs translocated to Europe were acquired from this region. Kimura et al.
(2016) suggested a possible Caribbean origin for European guinea pigs, however this is unlikely given the genetic distance between the Caribbean and European mitogenomes
(Figure 5.5, 5.6). The lack of diversity within the historic and modern European guinea pigs (Table 5.4) can be attributed to a genetic bottleneck situation during translocation to Europe, where only a subset of the South American population was transported.
For populations of guinea pigs in Puerto Rico today, primarily bred for market purposes, it is unclear whether they are the descendants of an ancient population or from a historic re-introduction. One of the modern European samples shared a haplotype with the modern sample from the Caribbean (San Sebastian, Puerto Rico), suggesting a recent re-introduction, possibly from Europe during the historic period.
However, modern day Peru should not be excluded as a possible origin since breeding for market is common in this region. The diversity within the modern population can
99 be attributed to a rise in breeding for the meat markets of South America and the
Caribbean, leading to population expansion and breeding for meat quantity and quality
(Burgos-Paz et al., 2011; Spotorno et al., 2006).
A likely origin can be determined for the historic Charleston (South Carolina) sample, recovered from the historic Heyward-Washington House. This sample shared a haplotype with the northern Peruvian and Caribbean samples, suggesting that the specimen was acquired directly from the Caribbean or Peru, rather than introduced from Europe during colonial expansion to North America. This is unsurprising given the popularity of having exotic collections of flora, fauna and material culture during the 19th Century.
In summary, this study has explored possible origins for European and North America guinea pigs using complete mitogenomes of archaeological and modern domestic guinea pigs from the Caribbean, South America, Belgium and South Carolina. Results indicate a southern Peruvian or northern Bolivian origin for European guinea pigs, though archaeological specimens from northern Chile were not available for use in this project. Such samples, along with further samples from Peru and Europe, may help refine the origins for European domestic guinea pig. A European origin was established for a modern specimen of guinea pig on Puerto Rico, suggesting that modern populations on the islands do not resemble ancient populations that were established during the late ceramic age. Finally, a Caribbean or Peruvian origin was determined for a specimen from a historic site in Charleston, South Carolina, indicating that North
American guinea pigs may have been acquired directly from these areas, as opposed to from Europe during colonial expansion into North America.
100
6.6 Guinea Pigs as a Caribbean Commensal Model
Domestic guinea pigs were introduced to the Caribbean via human-mediated transport from AD500, along with several other small mammals (LeFebvre and deFrance, 2014;
Giovas et al., 2012). The use of archaeological guinea pigs as a commensal model to determine possible origins of or interactions in Caribbean late ceramic age societies has demonstrated that there may have been at least two introductions of guinea pig to the
Caribbean, representing multiple migrations into the Caribbean region. An initial introduction occurred to Puerto Rico and a later introduction to Carriacou, both following previously established trade networks. This compliments recent archaeological research on interaction in the late Ceramic period which shows that there were multiple interaction spheres which could have led to multiple migrations into the Caribbean (Hofman and Carlin, 2010; Hofman et al., 2014b; Rodriguez-Ramos et al., 2010).
Within the pan-Caribbean region, using guinea pigs as a commensal model has provided further insight into the prehistoric migration and interactions in this area. To date no study has found an archaeological or genetic link between prehistoric societies of Peru and those of the Caribbean. It is likely that this represents interaction within mainland South America, prior to expansion into the Caribbean. Therefore, using the commensal approach supplements previously established archaeological data, and has been able to broaden knowledge nature of mobility and interaction in the prehistoric
Caribbean.
The commensal model has also been used in other island environments, such as the
101 Pacific, as a proxy for human movement to determine the likely origins and timings of settlement (Matisoo-Smith and Robins, 2004; Seelenfreund et al., 2011; Storey et al.,
2007). In this region, the use of multiple commensal studies has developed the origins and timings of introductions and interactions in the Pacific, expanding on archaeological, linguistic and human genetic data. It can be concluded that the development of the commensal model has greatly aided research into prehistoric migration, particularly in island settings such as the Pacific and now the Caribbean.
102 Chapter 7: Conclusions and Future Directions
7.1 Conclusions
This thesis aimed to use complete mitogenomes of ancient domestic guinea pig from the Caribbean and South America in order to explore human migration into the
Caribbean during the Late Ceramic Age (AD500-1492), and to determine an origin for
European guinea pigs. Two possible introductions of domestic guinea pig into the
Caribbean were identified, one to Puerto Rico and Antigua in the Greater Antilles and northern Lesser Antilles respectively, and a second introduction to Carriacou in the southern Lesser Antilles. The first introduction likely occurred through interaction networks previously established in the archaeological record from the early ceramic age, for the dispersal of raw materials and artefacts from the South American mainland
(Peru and Colombia) to the Caribbean. It is suggested that these networks allowed for the rise of complex society and increase in interaction within the northern Caribbean as they provided a link to the complex societies developing in South America. The second introduction likely occurred through migrations from Peru and Venezuela, coinciding with the development of a regional interaction sphere between the southern Lesser
Antilles and north-eastern South America, approximately AD1000. The correlation between the Caribbean guinea pigs and those from Peru was not predicted based on current archaeological and genetic evidence, thus has potentially extended the known range for interaction in this period.
The likely origin of European guinea pigs was identified as southern Peru or northern
Bolivia, though northern Chile is also possible. The historic Charleston guinea pig, and thus possibly the US guinea pig stock appears to have originated in the Caribbean or
103 Peru, rather than from Europe through colonial establishment. Modern Caribbean guinea pigs, bred for market purposes, are likely to be derived from modern European or Peruvian guinea pigs, and are not remnants of an ancient population established in the late ceramic age.
Analyses also revealed the possibility of two guinea pig domestication events occurring in prehistory. Based on analysis of the Cyt B region of modern and ancient domestic guinea pigs there are two distinct clades within C. porcellus. It is suggested that this may be to due to two domestication events, one possibly in the southwestern Andean region of Peru and the other in the northern Chilean Andean region. Complete mitogenome sequencing of modern samples and further ancient samples from South
America will be able to refine this hypothesis and the timings and location(s) of guinea pig domestication.
The application of complete mitogenome sequencing has shown that more variation is present in the population compared to using short fragments such as the HVR or Cyt B region. With continuing improvements in methodologies for recovery of ancient DNA, next-generation sequencing, and targeting genomes through hybridisation-capture, retrieval of complete mitogenomes from archaeological samples is now more reliable and should have greater use in future ancient DNA studies in the pan-Caribbean region.
Furthermore, the use of complete mitogenomes allowed for the observation that the
Colombian samples are a species of wild guinea pig that may have been misidentified based on morphological similarity between guinea pig species.
In summary, complete mitogenomes of archaeological guinea pigs have added to the
104 complex picture of interaction in the late ceramic age Caribbean and South America, and allowed for inferences on the interaction networks present within the pan-
Caribbean region. Furthermore, it has allowed for the identification of possible origins for European and North American guinea pigs, and shown that modern populations of guinea pigs in the Caribbean are not descended from ancient populations established in the late ceramic age. Also identified in this study was the possibility of two guinea pig domestication events in South America, and the probable misidentification of the
Colombian samples. This study is the first to use complete mitogenomes of a commensal animal within the Caribbean region, and has highlighted the use of the commensal model and next generation sequencing within the Caribbean context.
7.2 Future Directions
Firstly, a review of the taxonomy of the Cavia genus is warranted. Genetic data, in combination with morphological data will help define phylogenetic relationships between the various species and subspecies. Currently there are six recognised species and several recognised subspecies. Determining a phylogeny from complete mitogenomes of wild and domestic guinea pigs would aid in assessing the taxonomic classificiation for those currently recognised as subspecies, such as C. anolaimae and
C. aperea guianae. This would also help to examine morphological characteristics that may be different between each species, and thus assist with faunal identifications in the future.
Future whole genome sequencing projects would also aid in addressing issues regarding domestication and selection. Whole genome studies have recently determined the phylogenetic relationships between wild and domestic form, as well as the timings
105 of domestication in a range of species of plants and animals (Skoglund et al., 2015;
Frantz et al., 2016; Wang et al., 2016). This methodology could be applied to modern and ancient guinea pigs to determine the number, timing and location of possible different domestication events. Full genome studies are also useful in determining hybridisation between species (Reich et al., 2010), which may help to infer the level of hybridisation, if any, between species and subspecies of Cavia.
For this project, further archaeological samples from the Caribbean would help to ascertain if one or two introductions occurred during the late Ceramic Age, and possible dispersal events within the Caribbean. Additional samples from Puerto Rico will elucidate if the haplotype present in the southern Lesser Antilles is present in the
Greater Antilles, or whether it does in fact represent later interaction with mainland
South America. In order to further pinpoint the origins of migrations into the
Caribbean, samples representative of C. porcellus from Colombia and Venezuela will help to determine the dispersal routes that potentially went through these areas prior to translocation to the Caribbean. Currently no archaeological specimens of domestic guinea pig have been excavated along coastal Colombia or Venezuela, though archaeological specimens of C. aperea have been identified in the Venezuelan highlands (Garson, 1980). Furthermore, samples from Chile as well as additional samples from Peru and Bolivia may assist in locating the origin of European domestic guinea pigs.
Phylogenetic studies could be carried out using archaeological remains of other commensal animals (agouti, pecarry, opossum, and armadillo) present in the Caribbean for comparison to the guinea pig. This would aid in determining if all commensals
106 originated from the same location, as well as if they all were part of the same interaction spheres that led to their establishment in the Caribbean. This would further pinpoint the origins of late Ceramic Age migrations into the Caribbean, and interactions that took place with the surrounding mainland areas during this period. The use of mulitple commensal studies in the Pacific has strengthened the evidence for relationships between each of the island societies (Larson et al., 2007; Matisoo-Smith,
1994; Matisoo-Smith and Robins, 2004; Storey et al., 2007), and a similar approach could be applied to the Caribbean using agouti, pecarry, opossum and armadillo.
Finally, re-analysis of existing ancient human remains, as well as future excavated material, using complete mitogenome methodology will also provide further insight into prehistoric migration and interaction in the pan-Caribbean region. The data generated in the study by Llamas et al. (2016) provides excellent comparative data for a future study using complete ancient human mitogenomes from the Caribbean. This would strengthen the hypotheses on mobility and interaction in the Caribbean and
South American areas that have been established through archaeology and now through a commensal study using complete mitogenomes.
107 Appendix
Table 8.1: Genbank accession numbers of samples used in CytB analyses. References: 1. Dunnum and Salazar-Bravo (2010) 2. Spotorno et al. (2004), 3. Burgos-Paz et al. (2011), 4. Spotorno et al. (2006).
Accession Species Location Reference Number GU136741.1 C. aperea Bolivia: Santa Cruz 1 GU136742.1 C. aperea Bolivia: Santa Cruz 1 GU136743.1 C. aperea Bolivia: Beni 1 GU136752.1 C. aperea Bolivia: Santa Cruz 1 GU136753.1 C. aperea Bolivia: Santa Cruz 1 GU136754.1 C. aperea Bolivia: Santa Cruz 1 AY382790.1 C. aperea Bolivia: Santa Cruz 2 AY382791.1 C. aperea Paraguay: Concepcion 2 GU136758.1 C. aperea anolaimae Colombia: Cundinamarca 1 GU136755.1 C. aperea guianae Colombia: Meta 1 GU136756.1 C. aperea guianae Colombia: Meta 1 GU136757.1 C. aperea guianae Colombia: Meta 1 GU136759.1 C. aperea guianae Suriname: Nickeri 1 GU136740.1 C. aperea hypoleuca Paraguay: Concepcion 1 GU136744.1 C. aperea pamparum Argentina: Buenos Aires 1 GU136745.1 C. aperea pamparum Argentina: Buenos Aires 1 GU136746.1 C. aperea pamparum Argentina: Santa Fe 1 GU136747.1 C. aperea pamparum Argentina: Santa Fe 1 GU136748.1 C. aperea pamparum Argentina: Santa Fe 1 GU136749.1 C. aperea pamparum Argentina: Buenos Aires 1 GU136750.1 C. aperea pamparum Argentina: Buenos Aires 1 GU136751.1 C. aperea pamparum Uruguay: Rio Negro 1 GU136734.1 C. magna Uruguay: Rocha 1 GU136735.1 C. magna Uruguay: Rocha 1 GU136736.1 C. magna Uruguay: Rocha 1 GU136732.1 C. porcellus Bolivia: La Paz 1 GU136733.1 C. porcellus Colombia: Tolima 1 HM447146.1 C. porcellus Colombia: Pupiales 3 HM447147.1 C. porcellus Colombia: Pupiales 3 HM447148.1 C. porcellus Colombia: Pupiales 3 HM447149.1 C. porcellus Colombia: Pupiales 3 HM447150.1 C. porcellus Colombia: Potosi 3 HM447151.1 C. porcellus Colombia: Potosi 3 HM447152.1 C. porcellus Colombia: Potosi 3 HM447153.1 C. porcellus Colombia: Potosi 3 HM447154.1 C. porcellus Colombia: Potosi 3 HM447155.1 C. porcellus Colombia: Obonuco 3
108 Accession Species Location Reference Number HM447156.1 C. porcellus Colombia: Obonuco 3 HM447157.1 C. porcellus Colombia: Obonuco 3 HM447158.1 C. porcellus Colombia: Obonuco 3 HM447159.1 C. porcellus Colombia: Udenar 3 HM447160.1 C. porcellus Colombia: Udenar 3 HM447161.1 C. porcellus Colombia: Udenar 3 HM447162.1 C. porcellus Colombia: Udenar 3 HM447163.1 C. porcellus Colombia: Udenar 3 HM447164.1 C. porcellus Colombia: Udenar 3 HM447165.1 C. porcellus Colombia: Udenar 3 HM447166.1 C. porcellus Colombia: Udenar 3 HM447167.1 C. porcellus Colombia: Botana 3 HM447168.1 C. porcellus Colombia: Botana 3 HM447169.1 C. porcellus Colombia: Botana 3 HM447170.1 C. porcellus Colombia: Botana 3 HM447171.1 C. porcellus Colombia: Botana 3 HM447172.1 C. porcellus Colombia: Botana 3 HM447173.1 C. porcellus Colombia: Botana 3 HM447174.1 C. porcellus Colombia: Botana 3 HM447175.1 C. porcellus Colombia: Botana 3 HM447176.1 C. porcellus Colombia: Botana 3 HM447177.1 C. porcellus Colombia: Botana 3 HM447178.1 C. porcellus Colombia: Botana 3 HM447179.1 C. porcellus Colombia: Jose M Hernandez 3 HM447180.1 C. porcellus Colombia: Jose M Hernandez 3 HM447181.1 C. porcellus Colombia: Jose M Hernandez 3 HM447182.1 C. porcellus Colombia: Jose M Hernandez 3 HM447183.1 C. porcellus Colombia: Pasto 3 HM447184.1 C. porcellus Colombia: Pasto 3 HM447185.1 C. porcellus Colombia: Pasto 3 HM447186.1 C. porcellus Colombia: Pupiales 3 HM447187.1 C. porcellus Colombia: Pupiales 3 DQ017037 C. porcellus Chile: Andina Arica 4 DQ017038 C. porcellus Chile: Pirbright 4 DQ017039 C. porcellus Chile: Andina Arica 4 DQ017040 C. porcellus Chile: Andina Arica 4 DQ017041 C. porcellus Peru: Creole Junin 4 DQ017042 C. porcellus Ecuador: Mejocuy Auqui 4 DQ017043 C. porcellus Ecuador: Mejocuy Auqui 4 DQ017044 C. porcellus Bolivia: Mejocuy Nativa 4 DQ017045 C. porcellus Bolivia: Mejocuy Nativa 4 DQ017046 C. porcellus Peru: Mejocuy Tamborada 4 DQ017047 C. porcellus Peru: Mejocuy Tamborada 4
109 Accession Species Location Reference Number AY382793.1 C. porcellus Chile: Arica Market 2 AY245094 .1 C. porcellus Peru: Piura Market 2 AY245095.1 C. porcellus Peru: Trujillo Market 2 AY245096.1 C. porcellus Peru: Cajamarca Market 2 AY245097.1 C. porcellus Peru: Cusco Market 2 AY245098.1 C. porcellus Peru: Puno Market 2 AY247008.1 C. porcellus Peru: Arequipa Market 2 AY228363.1 C. porcellus Peru: Tacna Market 2 AY228362.1 C. porcellus Argentina: Pet Shop 2 AY228361.1 C. porcellus Chile: San Pedro de Atacama, 2 pet AF490405.1 C. porcellus Colombia: Palmira Market 2 GU136727.1 C. tschudii Peru:Ica 1 GU136728.1 C. tschudii Peru:Ica 1 GU136729.1 C. tschudii Bolivia: Santa Cruz 1 GU136731.1 C. tschudii Peru: Cuzco 1 DQ017048 C. tschudii Chile: Wild Lluta 4 DQ017049 C. tschudii Chile: Wild Lluta 4 DQ017050 C. tschudii Chile: Wild Lluta 4 DQ017051 C. tschudii Chile: Wild Lluta 4 DQ017052 C. tschudii Chile: Wild Lluta 4 DQ017053 C. tschudii Peru: Wild Chiguata 4 AY382792.1 C. tschudii Peru: Cuzco Market 4 AY245099.1 C. tschudii Peru: Puno Market 4 GU136730.1 C. tschudii osgoodi Peru: Puno 1 GU136726.1 C. tschudii sodalis Argentina:juyjuy 1
110 Table 8.2: Post-sequencing data for complete mitogenomes from archaeological and modern samples. Total coverage (%) and average read depth are listed for each sample. '-' denotes inadequate read depth for mapping to the guinea pig reference genome. Green highlight denotes samples that were not sequenced due to inadequate DNA libraries.
Average Total Sample Region Site Date Coverage Coverage (%) MS10625 Caribbean Tibes A, Puerto Rico AD600-900 418.4 99.2 MS10626 Caribbean Tibes B, Puerto Rico AD900-1200 2.4 88.6 MS10627 Caribbean Grand Bay, Carriacou AD985-1030 26.8 99 MS10628 Peru Moqi Post AD1400 687.4 99.2 MS10629 Peru Moqi Post AD1400 2 82.3 MS10630 Caribbean Coconut Hall, Antigua AD1035, 1045 143.3 99.2 MS10631 Caribbean Giraudy, St Lucia AD1200-1400 0.2 21.1 MS10632 Caribbean Coconut Walk, Nevis AD760-1440 - - MS10633 Caribbean Coconut Walk, Nevis AD760-1440 - - MS10634 Caribbean Coconut Walk, Nevis AD760-1440 - - MS10635 Caribbean (Modern) Puerto Rico Modern 623.2 99.2 MS10636 Colombia El Venado AD1400-1540 61.2 82.9 MS10637 Colombia El Venado Date 47.3 81.4 MS10638 Colombia El Venado AD1400-1540 76.3 80.5 MS10639 Colombia El Venado AD1400-1540 228.4 93.5 MS10640 Colombia El Venado AD1400-1540 7.1 62.1 MS10641 Colombia El Venado AD1400-1540 14.4 76.2 MS10642 Peru Lo Demas AD1400-1540 - - MS10643 Peru Lo Demas AD1480-1540 37.3 98.9
111 Average Total Sample Region Site Date Coverage Coverage (%) MS10644 Peru Lo Demas AD1480-1540 2.3 68.1 MS10645 Peru Moqi Post AD1400 398.8 99.2 MS10646 Peru Moqi Post AD1400 0.4 29.9 MS10647 Peru Moqi Post AD1400 1.6 73.6 MS10648 Peru Moqi Post AD1400 0.8 52.6 MS10649 Peru Moqi Post AD1400 23.6 98.6 MS10650 Peru Torata Alta AD1550-1600 1238.6 99.5 MS10651 Bolivia Kala Uyuni 200BC-AD500 143.4 99.1 MS10652 Bolivia Kala Uyuni 200BC-AD500 442.4 99.1 MS10653 Bolivia Llusco 800-200BC 478.5 99.1 MS10654 Bolivia Llusco 800-200BC 31.5 98.8 MS10655 Belgium (Historic) Mons, Brussels AD1550-1640 3253.1 99.5 MS10671 Charleston (Historic) Heyward-Washington House 1820s 399.8 99.2 MS10672 Colombia Madrid Site 150 +/- 50 BC 65.1 94.4 MS10673 Colombia Aguazuque I 1885-1815 BC - - MS10674 Colombia Aguazuque I 1885-1815 BC - - MS10675 Colombia Aguazuque I 1885-1815 BC 42.1 94.5 MS10676 Colombia Checua II 6200-2000BC - - MS10677 Colombia Checua II 6200-2000BC 445.6 99.1 MS10678 Colombia Checua II 6200-2000BC 158.1 89.8 MS10679 Colombia Checua II 6200-2000BC 69.3 84.9 MS10680 Colombia Checua II 6200-2000BC 11 87.6
112 Average Total Sample Region Site Date Coverage Coverage (%) MS10681 Peru Kuelap AD500-1400 1831.9 99.5 MS10682 Peru Kuelap AD500-1400 2902 99.4 MS10683 Peru Kuelap AD500-1400 3297.4 99.5 MS10684 Peru Kuelap AD500-1400 2406.4 99.4 MS10685 Peru Kuelap AD500-1400 2961.1 99.3 MrsChloe NZ (Modern) Dunedin Modern 734 100
113 Table 8.3: Nucleotide diversity (π) across the complete mitochondrial genome. Maximum nucleotide diversity is highlighted in red. Points in the genome where nucleotide diversity is 0 are not displayed.
Window Midpoint π Window Midpoint π 1477-1576 1526 0.00312 12566-12665 12615 0.00312 1502-1601 1551 0.00312 12591-12690 12640 0.00312 1527-1626 1576 0.00312 13566-13665 13615 0.00312
1552-1651 1601 0.00312 13591-13690 13640 0.00312
3442-3542 3491 0.00312 13616-13715 13665 0.00312
3467-3582 3516 0.00312 13641-13740 13690 0.00312
3492-3607 3542 0.00312 14452-14551 14501 0.00312
3517-3632 3582 0.00312 14477-14576 14526 0.00312
3658-3777 3727 0.00485 14502-14601 14551 0.00312 3683-3802 3752 0.00485 14527-14626 14576 0.00312 3728-3827 3777 0.00485 15310-15410 15359 0.00312 3753-3852 3802 0.00485 15335-15435 15384 0.00312 5090-5191 5139 0.00173 15360-15460 15410 0.00312 5115-5216 5164 0.00173 15385-15485 15435 0.00312 5140-5241 5191 0.00173 15461-15639 15589 0.00485 5167-5267 5216 0.00173 15486-15664 15614 0.00485 9022-9121 9071 0.00091 15590-15703 15639 0.01004 9047-9146 9096 0.00091 15615-15728 15664 0.01004 9072-9172 9121 0.00091 15640-15753 15703 0.00519 9097-9199 9146 0.00091 15665-15778 15728 0.00519 12006-12105 12055 0.00312 16546-16645 16595 0.01039 12031-12130 12080 0.00312 16571-16670 16620 0.01039 12056-12155 12105 0.00312 16596-16695 16645 0.01039 12081-12180 12130 0.00312 16621-16720 16670 0.01039 12531-12640 12590 0.00312
114 Table 8.4: Haplogroups in Temporal Network of Cytochrome B sequences of modern and ancient domestic guinea pig.
Number of Haplogroup Location Sequences 1 1 Colombia 2 (ancient) 5 Puerto Rico, Colombia 2 (modern) 6 Europe, Puerto Rico, Peru Puerto Rico, Antigua, Carriacou, Colombia, Peru, 3 (ancient) 23 Bolivia Charleston, Colombia (Pupiales, Potosi, Obonuco, 3 (modern) 19 Udenar, Botana) 4 2 Chile, Peru (Arequipa) 5 1 Argentina Pet Shop 6 1 Peru Market (Piura) 7 1 Peru Market (Trujillo) 8 1 Peru Market (Cajamarca) 9 1 Peru Market (Cusco) 10 1 Peru Market (Puno) 11 1 Chile (Atacama, pet) 12 1 Chile (Andina Arica) 13 1 Chile (English pet breed) 14 1 Chile (Andina Arica) 15 1 Chile (Andina Arica) 16 1 Peru 17 1 Ecuador 18 1 Ecuador 19 1 Bolivia 20 1 Bolivia 21 1 Peru (Tamborada) 22 1 Peru (Tamborada) 23 1 Bolivia (La Paz) Colombia (Palmira, Udenar, Botana, Jose M 24 7 Hernandez, Pasto) 25 1 Colombia (Pupiales) 26 1 Colombia (Pupiales) 27 3 Colombia (Potosi, Botana, Jose M Hernandez) 28 1 Colombia (Potosi) 29 1 Colombia (Potosi) 30 1 Colombia (Udenar) 31 1 Colombia (Udenar) 32 1 Colombia (Udenar) 33 1 Colombia (Udenar) 34 1 Colombia (Botana) 35 1 Colombia (Jose M Hernandez) 36 1 Colombia (Pasto) 37 1 Colombia (Pasto)
115 Reference List
Adler CJ, Haak W, Donlon D, et al. (2011) Survival and recovery of DNA from ancient teeth and bones. Journal of Archaeological Science 38: 956-964. Allaire L and Mattioni M. (1983) Boutbois et le Godinot: deux gisements acéramiques de la Martinique. Comptes Rendus du IXe Congrès International d'Etudes des Civilisations Précolombiennes des Petites Antilles: 27-38. Allgood J. (2002) Faunal Analysis of the Green castle Estate Assemblage (STM25). unplublished report. Allsworth-Jones P. (2008) Pre-Columbian Jamaica, Alabama: University of Alabama Press. Allsworth-Jones P and Wesler K. (2001) Excavations at Green Castle, Jamaica, 1999- 2001. Proceedings of the XIXth International Congress for Caribbean Archaeology. 186-193. Ansieau C and Denis M. (2009) Fouilles preventives d'une parcelle sise a Mons, rue Jean Lescarts (Ht.). Archaeologia Mediaevalis 32: 81-84. Baldwin SL, D’Souza C, Roberts AD, et al. (1998) Evaluation of new vaccines in the mouse and guinea pig model of tuberculosis. Infection and immunity 66: 2951- 2959. Bandy M and Hastorf C. (2007) Kala Uyuni an Early Political Center in the Southern Lake Titicaca Basin: 2003 Excavation of the Taraco Archaeological project: Berkley: University of California Press 1-11. Barnes S, Matisoo-Smith E and Hunt TL. (2006) Ancient DNA of the Pacific rat (Rattus exulans) from Rapa Nui (Easter Island). Journal of Archaeological Science 33: 1536-1540. Behar DM, Villems R, Soodyall H, et al. (2008) The dawn of human matrilineal diversity. The American Journal of Human Genetics 82: 1130-1140.
Benn‐Torres J, Bonilla C, Robbins C, et al. (2008) Admixture and population stratification in African Caribbean populations. Annals of human genetics 72: 90-98. Benton M, Macartney-Coxson D, Eccles D, et al. (2012) Complete mitochondrial genome sequencing reveals novel haplotypes in a Polynesian population. PloS one 7: e35026.
116 Boocock J. (2014) Ancient DNA Pipeline. available at: https://github.com/theboocock/ancient_dna_pipeline Boomert A. (2000) Trinidad, Tobago, and the Lower Orinoco interaction sphere: an archaeological/ethnohistorical study. Leiden: Leiden University Press . Boomert A and Kameneff L. (2005) Preliminary report on archaeological investigations at Great Courland Bay, Tobago. Proceedings of the 20th Congress of the International Association for Caribbean Archaeology (2). 457-468. Breukel T. (2013) Threepointers on Trial: A biographical study of Amerindian ritual artefacts from the pre-Columbian Caribbean. Leiden: University of Leiden Press. Briggs AW, Stenzel U, Johnson PL, et al. (2007) Patterns of damage in genomic DNA sequences from a Neandertal. Proceedings of the National Academy of Sciences 104: 14616-14621. Burgos-Paz W, Ceron-Munoz M and Solarte-Portilla C. (2011) Genetic diversity and population structure of the Guinea pig (Cavia porcellus, Rodentia, Caviidae) in Colombia. Genet Mol Biol 34: 711-718. Cabrera A. (1961) Catálogo de los mamíferos de América del Sur: Coni. Callaghan RT. (2003) Comments on the mainland origins of the Preceramic cultures of the Greater Antilles. Latin American Antiquity 14: 323-338. Callaghan RT and Bray W. (2007) Simulating prehistoric sea contacts between Costa Rica and Colombia. The Journal of Island and Coastal Archaeology 2: 4-23. Cann RL, Stoneking M and Wilson AC. (1987) Mitochondrial DNA and human evolution. Nature 325: 31-36. Cano RJ, Poinar HN, Pieniazek NJ, et al. (1993) Amplification and sequencing of DNA from a 120-135-million-year-old weevil. Nature 363: 536. Carlson B. (1995) Strings of command: Manufacture and utilization of shell beads among the Taino. Proceedings of the Fifteenth International Congress for Caribbean Archaeology, Centro de Estudios Avanzados de Puerto Rico y el Caribe, San Juan. 97-110. Carlson L. (2008) A Multidisciplinary Approach to Site Testing and Data Recovery at Two Village Sites (AR-38 and AR-39) on the Lower Rıo Tanamá, Municipality of Arecibo, Puerto Rico. Prepared by Southeastern Archaeological Research. Inc.(SEARCH) for the for US Army Corps of Engineers, Jacksonville District. On file at the Puerto Rico State Historic Preservation Office, San Juan.
117 Clare EL, Lim BK, Engstrom MD, et al. (2007) DNA barcoding of Neotropical bats: species identification and discovery within Guyana. Molecular Ecology Notes 7: 184-190. Clutton-Brock J. (1988) Carnivore remains excavated from Fell's cave in 1970. In: Hyslop J (ed) Travels and Archaeology in South Chile by Junius B. Bird. Iowa City: University of Iowa Press, 188-195. Collins MJ, Nielsen-Marsh CM, Hiller J, et al. (2002) The survival of organic matter in bone: A review. Archaeometry 44: 383-394. Cooke R. (2005) Prehistory of Native Americans on the Central American land bridge: Colonization, dispersal, and divergence. Journal of Archaeological Research 13: 129-187. Cooper J, Martinón-Torres M and Rojas RV. (2008) American gold and European brass: Metal objects and indigenous values in the cemetery of el chorro de maíta'Cuba. In: Hofman, C and Gijn, A (eds) Crossing The Borders: New Methods and Techniques in The Study of Archaeology Materials from The Caribbean. Alabama: University of Alabama Press,34-42. Correal Urrego G. (1990) Aguazuque: evidencias de cazadores, recolectores y plantadores en la altiplanicie de la Cordillera Oriental. Fundacion de Investigaciones Arqueologicas Nacionales: Publicaciones. Crock JG and Petersen JB. (2004) Inter-island exchange, settlement hierarchy, and a Taino-related chiefdom on the Anguilla Bank, Northern Lesser Antilles. In: Delpuech A. and Hofman CL. Late Ceramic Age Societies in the Eastern Caribbean: BAR International Series Paris Monographs on American Archaeology, Oxford: Archaeopress, 139-158. Dabney J, Knapp M, Glocke I, et al. (2013) Complete mitochondrial genome sequence of a Middle Pleistocene cave bear reconstructed from ultrashort DNA fragments. Proceedings of the National Academy of Sciences 110: 15758-15763. Darriba D, Taboada GL, Doallo R, et al. (2012) jModelTest 2: more models, new heuristics and parallel computing. Nature methods 9: 772-772. de Mahecha AMG. (1992) Checua: una secuencia cultural entre 8500 y 3000 años antes del presente. Banco de la República: Fundación de Investigaciones Arqueológicas Nacionales.
118 Deagan K. (1988) The Archaeology of the Spanish Contact Period in the Caribbean. Journal of World Prehistory 2: 187-233. deFrance S. (2010) Chiefly fare or who’s feeding the cacique? Equality in animal use at the Tibes ceremonial center, Puerto Rico. In: Campana D, Crabtree P, deFrance S, et al. (eds) Anthropological Approaches to Zooarchaeology: Colonialism, Complexity and Animal Transformations. Oxford: Oxbow Books, 76-89. deFrance S, Hadden CS, LeFebvre MJ, et al. (2010) Faunal use at the Tibes ceremonial site. In: Curet LA and Stringer LM (eds) Tibes: people, power, and ritual at the center of the cosmos. University of Alabama Press. deFrance S. (1993) Ecological imperialism in the South-Central Andes: Faunal data from Spanish colonial settlements in the Moquegua and Torata valleys. University of Florida. Doig FK. (2000) Primera expedicion arquelogica a los mausoleos Chachaopya(s) de la laguna de las Momias (Dpto. de Amazonas, Perú). Chungara: Chungara Revista de Antropología Chilena, 49-54. Drummond AJ and Rambaut A. (2007) BEAST: Bayesian evolutionary analysis by sampling trees. BMC evolutionary biology 7: 214. Drummond AJ, Suchard MA, Xie D, et al. (2012) Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular biology and evolution 29: 1969-1973. DuChemin GR, Cannarozzi N and deFrance S. (2010) The Cultural Landscape of Jácana: Archaeological Investigations of Site PO–29, Municipio de Ponce, Puerto Rico. New South Associates, Inc., Stone Mountain, Georgia. Submitted to the Jacksonville District, US Army Corps of Engineers, Jacksonville, Florida. On file at the Puerto Rico State Historic Preservation Office, San Juan. Duggan AT, Evans B, Friedlaender FR, et al. (2014) Maternal history of Oceania from complete mtDNA genomes: contrasting ancient diversity with recent homogenization due to the Austronesian expansion. The American Journal of Human Genetics 94: 721-733. Dunnum JL and Salazar-Bravo J. (2010) Molecular systematics, taxonomy and biogeography of the genus Cavia (Rodentia: Caviidae). Journal of Zoological Systematics and Evolutionary Research 48: 376-388.
119 Eisenberg JF and Redford KH. (1992) Mammals of the Neotropics, Volume 2: The Southern Cone: Chile, Argentina, Uruguay, Paraguay, Chicago: University of Chicago Press. Endersby J. (2012) A guinea pig's history of biology: the plants and animals who taught us the facts of life, London: Random House. Fitzpatrick SM. (2006) A critical approach to C-14 dating in the Caribbean: Using chronometric hygiene to evaluate chronological control and prehistoric settlement. Latin American Antiquity 17: 389-418. Fitzpatrick SM, Carstensen JA, Marsaglia KM, et al. (2008) Preliminary petrographic and chemical analyses of prehistoric ceramics from Carriacou, West Indies. Journal of Caribbean Archaeology 8: 59. Fitzpatrick SM and Giovas CM. (2011) New Radiocarbon Dates for the Grenadine Islands (West Indies). Radiocarbon 53: 451-460. Frantz LA, Mullin VE, Pionnier-Capitan M, et al. (2016) Genomic and archaeological evidence suggest a dual origin of domestic dogs. Science 352: 1228-1231. Fu Q, Mittnik A, Johnson PL, et al. (2013) A revised timescale for human evolution based on ancient mitochondrial genomes. Current Biology 23: 553-559. Gaieski JB, Owings AC, Vilar MG, et al. (2011) Genetic ancestry and indigenous heritage in a Native American descendant community in Bermuda. American Journal of Physical Anthropology 146: 392-405. Garcia JL. (2012) The Foods and Crops of the Muisca: A Dietary Reconstruction of the Intermediate Chiefdoms of Bogotá (Bacatá) and Tunja (Hunza), Florida: University of Central Florida. Garcia-Casco A, Knippenberg S, Ramos RR, et al. (2013) Pre-Columbian jadeitite artifacts from the Golden Rock Site, St. Eustatius, Lesser Antilles, with special reference to jadeitite artifacts from Elliot's, Antigua: implications for potential source regions and long-distance exchange networks in the Greater Caribbean. Journal of Archaeological Science 40: 3153-3169. García-Casco A, Vega AR, Párraga JC, et al. (2009) A new jadeitite jade locality (Sierra del Convento, Cuba): first report and some petrological and archeological implications. Contributions to Mineralogy and Petrology 158: 1- 16.
120 Garson A. (1980) Prehistory, Settlement and Food Production in the Savanna Region of La Calzada de Paez, Venezuela. Department of Anthropology. New Haven: Yale University. Gilbert MTP, Hansen AJ, Willerslev E, et al. (2003) Characterization of genetic miscoding lesions caused by postmortem damage. The American Journal of Human Genetics 72: 48-61. Giovas C, Kamenov G, Fitzpatrick S, et al. (2016) Sr and Pb isotope investigation of paleomobility in Pre-Columbian introduced mammals of the Lesser Antilles. Journal of Archaeological Science 69: 39-53. Giovas CM. (2016) Pre-Columbian Amerindian Lifeways at the Sabazan Site, Carriacou, West Indies. The Journal of Island and Coastal Archaeology: 1-30. Giovas CM, LeFebvre MJ and Fitzpatrick SM. (2012) New records for prehistoric introduction of Neotropical mammals to the West Indies: evidence from Carriacou, Lesser Antilles. Journal of Biogeography 39: 476-487. Gray RD, Drummond AJ and Greenhill SJ. (2009) Language phylogenies reveal expansion pulses and pauses in Pacific settlement. Science 323: 479-483. Green RE, Briggs AW, Krause J, et al. (2009) The Neandertal genome and ancient DNA authenticity. The EMBO Journal 28: 2494-2502. Green RE, Malaspinas A-S, Krause J, et al. (2008) A complete Neandertal mitochondrial genome sequence determined by high-throughput sequencing. Cell 134: 416-426. Greig K, Boocock J, Prost S, et al. (2015) Complete mitochondrial genomes of New Zealand’s first dogs. PloS one 10: e0138536. Haak W, Balanovsky O, Sanchez JJ, et al. (2010) Ancient DNA from European Early Neolithic Farmers Reveals Their Near Eastern Affinities. PLoS Biology. 8: e1000536. Hamilton-Dyer S. (2009) Animal Bones. In: Dury P and Simpson R (eds) Hill Hall: A Singular House Devised by a Tudor Intellectual. London: Society of Antiquities/EH Monograph, 345-351. Healy P, Thornton E, Fuess M, et al. (2003) The Post Saladoid Faunal Assemblage from The Coconut Hall Site, Antigua, West Indies. In Proceedings from the XX International Congress for Caribbean Archaeology. Museo del Hombre Dominicano and Fundación García Arévalo, Santo Domingo. 189-200.
121 Hebert PD, Cywinska A and Ball SL. (2003) Biological identifications through DNA barcodes. Proceedings of the Royal Society of London B: Biological Sciences 270: 313-321. Heupink TH, Subramanian S, Wright JL, et al. (2016) Ancient mtDNA sequences from the First Australians revisited. Proceedings of the National Academy of Sciences 113: 6892-6897. Higuchi R, Bowman B, Freiberger M, et al. (1984) DNA sequences from the quagga, an extinct member of the horse family. Nature 312: 282-284. Hill A, Gentile B, Bonnardot J, et al. (1987) Polynesian origins and affinities: globin gene variants in eastern Polynesia. American Journal of Human Genetics 40: 453. Hodell DA, Quinn RL, Brenner M, et al. (2004) Spatial variation of strontium isotopes (87 Sr/86 Sr) in the Maya region: a tool for tracking ancient human migration. Journal of Archaeological Science 31: 585-601. Hofman C and Bright A. (2010) Mobility and Exchange from a Pan-Caribbean Perspective.. Journal of Caribbean Archaeology, Special Publication 3. Hofman C, Mol A, Hoogland M, et al. (2014a) Stage of encounters: migration, mobility and interaction in the pre-colonial and early colonial Caribbean. World Archaeology 46: 590-609. Hofman CL. (2013) The Post-Saladoid in the Lesser Antilles (AD 600/800-1492). In: Keegan, W, Hofman, CL and Rodriquez-Ramos, The Oxford Handbook of Caribbean Archaeology, Oxford: Oxford University Press, 205-220. Hofman CL, Bright AJ, Boomert A, et al. (2007) Island rhythms: The web of social relationships and interaction networks in the Lesser Antillean archipelago between 400 BC and AD 1492. Latin American Antiquity 18: 243-268. Hofman CL, Bright AJ, Hoogland ML, et al. (2008) Attractive ideas, desirable goods: Examining the Late Ceramic Age relationships between Greater and Lesser Antillean societies. The Journal of Island and Coastal Archaeology 3: 17-34. Hofman CL, Bright AJ, Ramos RR, et al. (2010) Crossing the Caribbean Sea: towards a holistic view of pre-colonial mobility and exchange. Journal of Caribbean Archaeology 10: 1-18. Hofman CL and Carlin EB. (2010) The ever dynamic Caribbean: Exploring new approaches to unraveling social networks in the pre-colonial and early colonial
122 periods. Linguistics and archaeology in the Americas: The historization of language and society 2: 107. Hofman CL and Hoogland ML. (2004) Social dynamics and change in the Northern Lesser Antilles. In: Delpuech A. and Hofman CL. Late Ceramic Age Societies in the Eastern Caribbean: BAR International Series Paris Monographs on American Archaeology, Oxford: Archaeopress, 47-58. Hofman CL and Hoogland ML. (2011) Unravelling the multi-scale networks of mobility and exchange in the pre-colonial circum-Caribbean. Communities in Contact: Essays in Archaeology, Ethnohistory and Ethnography of the Amerindian Circum-Caribbean: 14-44. Hofman CL, Mol A, Ramos RR, et al. (2014b) Networks Set in Stone: Archaic- Ceramic interaction in the early pre-Colonial northeastern Caribbean. Archéologie Caraibe. Leiden: Sidestone Press, 119-132. Hofman CL and van Gijn AL. (2008) Crossing the Borders: New Methods and Techniques in the Study of Archaeology Materials from the Caribbean: Alabama: University of Alabama Press. Hofreiter M, Jaenicke V, Serre D, et al. (2001a) DNA sequences from multiple amplifications reveal artifacts induced by cytosine deamination in ancient DNA. Nucleic acids research 29: 4793-4799. Hofreiter M, Serre D, Poinar HN, et al. (2001b) Ancient DNA. Nature Reviews Genetics 2: 353-359. Hoogland M and Hofman C. (2010) Island Dynamics. In: Fitzpatrick SM and Ross AH (eds) Island Shores, Distant Pasts: Archaeological and Biological Approaches to the Pre-Columbian Settlement of the Caribbean. Gainsville: University of Florida Press, 148-162. Horsburgh KA and Moreno-Mayar JV. (2015) Molecular identification of sheep at Blydefontein Rock Shelter, South Africa. Southern African Humanities 27: 65- 80. Horsburgh KA, Orton J and Klein RG. (2016) Beware the Springbok in Sheep’s Clothing: How Secure Are the Faunal Identifications upon Which We Build Our Models? African Archaeological Review 33: 353-361. Höss M, Jaruga P, Zastawny TH, et al. (1996) DNA damage and DNA sequence retrieval from ancient tissues. Nucleic acids research 24: 1304-1307.
123 Ihaka R and Gentleman R. (1996) R: a language for data analysis and graphics. Journal of computational and graphical statistics 5: 299-314. Jiménez JRP. (2007) De antiguos pueblos y culturas botánicas en el Puerto Rico indígena: el archipiélago borincano y la llegada de los primeros pobladores agroceramistas: British Archaeological Reports Limited. Jónsson H, Ginolhac A, Schubert M, et al. (2013) mapDamage2. 0: fast approximate Bayesian estimates of ancient DNA damage parameters. Bioinformatics: 29:1682-4. Kaye Q, Fitzpatrick S, Kappers M, et al. (2010) Beyond Time Team: Archaeological Investigations at Coconut Walk, Nevis, West Indies, 1st July–4th August 2010. Papers from the Institute of Archaeology 20. Kearse M, Moir R, Wilson A, et al. (2012) Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28: 1647-1649. Keegan WF. (1994) West Indian archaeology. 1. Overview and foragers. Journal of Archaeological Research 2: 255-284. Keegan WF. (1996) West Indian Archaeology. 2. After Columbus. Journal of Archaeological Research 4: 265-294. Keegan WF. (2000) West Indian Archaeology. 3. Ceramic Age. Journal of Archaeological Research 8: 135-167. Kimura BK, LeFebvre MJ, deFrance SD, et al. (2016) Origin of pre-Columbian guinea pigs from Caribbean archeological sites revealed through genetic analysis. Journal of Archaeological Science: Reports 5: 442-452. Kircher M, Sawyer S and Meyer M. (2012) Double indexing overcomes inaccuracies in multiplex sequencing on the Illumina platform. Nucleic acids research 40: e3. Kivisild T, Shen P, Wall DP, et al. (2005) The role of selection in the evolution of human mitochondrial genomes. Genetics 172:373-87. Knapp M, Clarke AC, Horsburgh KA, et al. (2012a) Setting the stage–building and working in an ancient DNA laboratory. Annals of Anatomy-Anatomischer Anzeiger 194: 3-6. Knapp M, Horsburgh KA, Prost S, et al. (2012b) Complete mitochondrial DNA genome sequences from the first New Zealanders. Proceedings of the National Academy of Sciences 109: 18350-18354.
124 Knippenberg S. (1999) Anse des Pères. Lithics. Archaeological investigations on St. Martin (Lesser Antilles). The sites of Norman Estate, Anse des Pères, and Hope Estate with a contribution to the ‘La Hueca problem’. Archaeological Studies Leiden University 4: 87-104. Knippenberg S. (2007) Stone artefact production and exchange among the Lesser Antilles. Amsterdam: Amsterdam University Press. Krause J, Fu Q, Good JM, et al. (2010) The complete mitochondrial DNA genome of an unknown hominin from southern Siberia. Nature 464: 894-897. Laffoon JE, Plomp E, Davies GR, et al. (2015) The Movement and Exchange of Dogs in the Prehistoric Caribbean: An Isotopic Investigation. International Journal of Osteoarchaeology 25: 454-465. Laffoon JE, Ramos RR, Baik LC, et al. (2014) Long-distance exchange in the precolonial Circum-Caribbean: A multi-isotope study of animal tooth pendants from Puerto Rico. Journal of Anthropological Archaeology 35: 220-233. Lalueza-Fox C. (1996) Mitochondrial DNA haplogroups in four tribes from Tierra del Fuego Patagonia: Inferences about the peopling of the Americas. Human biology 68: 855-871. Lalueza-Fox C, Calderon FL, Calafell F, et al. (2001) MtDNA from extinct Tainos and the peopling of the Caribbean. Annals of Human Genetics 65: 137-151. Lalueza-Fox C, Gilbert MT, Martinez-Fuentes AJ, et al. (2003) Mitochondrial DNA from pre-Columbian Ciboneys from Cuba and the prehistoric colonization of the Caribbean. American Journal of Physical Anthropology 121: 97-108. Larson G, Cucchi T, Fujita M, et al. (2007) Phylogeny and ancient DNA of Sus provides new evidence for human dispersal routes in Island Southeast Asia and the Pacific. Proceedings of the National Academy of Sciences USA 104: 4834- 4839. Larson G, Liu R, Zhao X, et al. (2010) Patterns of East Asian pig domestication, migration, and turnover revealed by modern and ancient DNA. Proceedings of the National Academy of Sciences 107: 7686-7691. LeFebvre MJ and deFrance SD. (2014) Guinea Pigs in the Pre-Columbian West Indies. Journal of Island & Coastal Archaeology 9: 16-44. Leigh JW and Bryant D. (2015) popart: full‐feature software for haplotype network construction. Methods in Ecology and Evolution 6: 1110-1116.
125 Li H and Durbin R. (2009) Fast and accurate short read alignment with Burrows– Wheeler transform. Bioinformatics 25: 1754-1760. Librado P and Rozas J. (2009) DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451-1452. Lindahl T. (1993) Instability and decay of the primary structure of DNA. Nature 362: 709-715. Linderholm A. (2016) Ancient DNA: the next generation–chapter and verse. Biological Journal of the Linnean Society 117: 150-160. Lindgreen S. (2012) AdapterRemoval: easy cleaning of next-generation sequencing reads. BMC research notes 5: 337. Llamas B, Fehren-Schmitz L, Valverde G, et al. (2016) Ancient mitochondrial DNA provides high-resolution time scale of the peopling of the Americas. Science advances 2: e1501385. Llamas B, Willerslev E and Orlando L. (2017) Human evolution: a tale from ancient genomes. Philosophical Transactions of the Royal Society B 372: 20150484. Macaulay V and Richards MB. (2013) Mitochondrial genome sequences and their phylogeographic interpretation. eLS. Maricic T, Whitten M and Pääbo S. (2010) Multiplexed DNA sequence capture of mitochondrial genomes using PCR products. PloS one 5: e14004. Martínez-Cruzado J. (2010) The History of Amerindian Mitochondrial DNA Lineages in Puerto Rico. In: Fitzpatrick SM and Ross AH (eds) Island Shore, Distant Pasts: Archaeological and Biological Approaches to the Pre-Columbian Settlement of the Caribbean. Gainsville: University of Florida Press. Martínez-Cruzado J. (2013) The DNA Evidence for the Human Colonization and Spread Across the Americas. In: Keegan, W, Hofman, CL and Rodriquez- Ramos, The Oxford Handbook of Caribbean Archaeology, Oxford: Oxford University Press, 470-488. Martinez-Cruzado JC, Toro-Labrador G, Ho-Fung V, et al. (2001) Mitochondrial DNA analysis reveals substantial Native American ancestry in Puerto Rico. Human biology 73: 491-511. Martinez-Cruzado JC, Toro-Labrador G, Viera-Vera J, et al. (2005) Reconstructing the population history of Puerto Rico by means of mtDNA phylogeographic analysis. American Journal of Physical Anthropology 128: 131-155.
126 Matisoo-Smith E. (1994) The human colonisation of Polynesia. A novel approach: genetic analyses of the Polynesian rat (Rattus exulans). The Journal of the Polynesian Society 103: 75-87. Matisoo-Smith E and Robins JH. (2004) Origins and dispersals of Pacific peoples: evidence from mtDNA phylogenies of the Pacific rat. Proceedings of the National Academy of Sciences of the United States of America 101: 9167-9172. Matisoo-Smith EA, Gosling AL, Boocock J, et al. (2016) A European Mitochondrial Haplotype Identified in Ancient Phoenician Remains from Carthage, North Africa. PloS one 11: e0155046. McKenna A, Hanna M, Banks E, et al. (2010) The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome research 20: 1297-1303. McMurray DN. (1994) Guinea pig model of tuberculosis. Tuberculosis. American Society of Microbiology, 135-147. Mendisco F, Pemonge MH, Leblay E, et al. (2015) Where are the Caribs? Ancient DNA from ceramic period human remains in the Lesser Antilles. Philosophical Transactions of the Royal Society of London B 370: 20130388. Mendizabal I, Sandoval K, Berniell-Lee G, et al. (2008) Genetic origin, admixture, and asymmetry in maternal and paternal human lineages in Cuba. BMC Evolutionary Biology 8: 213. Millar CD, Huynen L, Subramanian S, et al. (2008) New developments in ancient genomics. Trends in Ecology & Evolution 23: 386-393. Monsalve M, Restrepo H, Espinel A, et al. (1994) Evidence of mitochondrial DNA diversity in South American aboriginals. Annals of human genetics 58: 265-273. Monsalve MV. (1997) Mitochondrial DNA in ancient Amerindians. American Journal of Physical Anthropology 103: 423-425. Monsalve MV, Fernando C, GUHL F, et al. (1996) Phylogenetic analysis of mtDNA lineages in South American mummies. Annals of human genetics 60: 293-303. Morales E. (1994) The Guinea-Pig in the Andean Economy - from Household Animal to Market Commodity. Latin American Research Review 29: 129-142. Morris J. (2014) Explorations in anatomy: the remains from Royal London Hospital. Anthropozoologica 49: 109-120.
127 Moseley ME, Nash DJ, Williams PR, et al. (2005) Burning down the brewery: Establishing and evacuating an ancient imperial colony at Cerro Baúl, Peru. Proceedings of the National Academy of Sciences of the United States of America 102: 17264-17271. Newsom LA and Wing ES. (2004) On land and sea: Native American uses of biological resources in the West Indies, Alabama: University of Alabama Press. Nguyen L-T, Schmidt HA, von Haeseler A, et al. (2015) IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Molecular biology and evolution 32: 268-274. Ojala D, Montoya J and Attardi G. (1981) tRNA punctuation model of RNA processing in human mitochondria. Nature, 290: 470-4 Orton J, Mitchell P, Klein R, et al. (2013) An early date for cattle from Namaqualand, South Africa: implications for the origins of herding in southern Africa. Antiquity 87: 108-120. Pääbo S. (1989) Ancient DNA: extraction, characterization, molecular cloning, and enzymatic amplification. Proceedings of the National Academy of Sciences 86: 1939-1943. Pääbo S, Poinar H, Serre D, et al. (2004) Genetic analyses from ancient DNA. Annual Review Genetics 38: 645-679. Petersen JB. (1997) Taino, Island Carib, and prehistoric Amerindian economies in the West Indies: tropical forest adaptations to island environments. In: Wilson, S. Indigenous Peoples of the Caribbean. Gainsville: University Press of Florida, 118-130. Petersen JB, Hofman CL and Curet LA. (2004) Time and culture: chronology and taxonomy in the Eastern Caribbean and the Guianas. In: Delpuech A. and Hofman CL. Late Ceramic Age Societies in the Eastern Caribbean: BAR International Series Paris Monographs on American Archaeology, Oxford: Archaeopress 17-32. Phulgence W. (2007) Zooarchaeological Report on Pit 8 Level 1, Giraudy, St. Lucia. Manuscript on file with Florida Museum of Natural History, Gainesville. Pictet A and Ferrero MA. (1951) La descendance d'un croisement interspécifique de cobayes (Cavia aperea d'az.× Cavia Cobaya marcg.) analysée durant 25 années. Genetica 25: 357-515.
128 Pigière F, Van Neer W, Ansieau C, et al. (2012) New archaeozoological evidence for the introduction of the guinea pig to Europe. Journal of Archaeological Science 39: 1020-1024. Prost S and Anderson CN. (2011) TempNet: a method to display statistical parsimony networks for heterochronous DNA sequence data. Methods in Ecology and Evolution 2: 663-667. Pruvost M, Schwarz R, Correia VB, et al. (2007) Freshly excavated fossil bones are best for amplification of ancient DNA. Proceedings of the National Academy of Sciences of the United States of America 104: 739-744. Quitmyer I and Kozuch L. (1996) Phase II Zooarchaeology at Finca Valencia (NCS-1) and Site NCS-4. Northwest Puerto Rico with a Note on the Guinea Pig Remains, manuscript on file, Florida Museum of Natural History, University of Florida, Gainesville. Quitmyer I and Wing E. (2001) The Luján Site: A Record of Native American Animal Use on Vieques Island. Submitted to Virginia Rivera and on file at the Florida Museum of Natural History, Puerto Rico. Rambaut A. (2007) FigTree, a graphical viewer of phylogenetic trees. See http://tree.bio.ed.ac.uk/software/figtree. Rambaut A and Drummond A. (2013) TreeAnnotator v1. 7.0. Available as part of the BEAST package at http://beast.bio.ed.ac.uk. Reich D, Green RE, Kircher M, et al. (2010) Genetic history of an archaic hominin group from Denisova Cave in Siberia. nature 468: 1053-1060. Rivas AMB. (1999) Organización social y económica en la aldea muisca de El Venado (Valle de Samacá, Boyacá). Revista colombiana de Antropología: 5. Rodríguez JV and Toro AC. (2005) Un yacimiento formativo ritual en el entorno de la antigua laguna de La Herrera, Madrid, Cundinamarca. Maguaré. Rodriguez-Ramos R, de Utuado R and de Ciencias Sociales P. (2010) What is the Caribbean? An archaeological perspective. Journal of Caribbean Archaeology 3: 19-51. Rofes J. (2004) Prehispanic guinea pig sacrifices in southern Peru, the case of el Yaral. In: O'Day, S., van Neer, W., Ervynck, A. (Eds.), Behavior Behind the Bones: The Zooarchaeology of Ritual, Religion, Status and Identity. Oxford: Oxbow Books, 95–100.
129 Rofes J and Wheeler JC. (2003) Sacrificio de cuyes en los Andes: el caso de El Yaral y una revisión biológica, arqueológica y etnográfica de la especie Cavia porcellus. Archaeofauna 12: 29-45. Rohland N and Hofreiter M. (2007) Ancient DNA extraction from bones and teeth. Nature Protocols 2: 1756-1762. Rood J. (1972) Ecological and behavioural comparisons of three genera of Argentine cavies. Animal Behaviour Monographs 5: 83-84. Rouse I. (1989a) Peoples and cultures of the Saladoid frontier in the Greater Antilles. Early Ceramic population lifeways and adaptive strategies in the Caribbean, BAR, International Series 506: 383-403. Rouse I. (1989b) Peopling and repeopling of the West Indies. In: Woods, CA (ed) Biogeography of the West Indies: Past, present, and future, Gainsville: Sandhill Crane Press, 119-136. Rouse I. (1993) The Tainos: rise and decline of the people who greeted Columbus: Yale University Press. Salas A, Richards M, Lareu MV, et al. (2004) The African diaspora: mitochondrial DNA and the Atlantic slave trade. American Journal of Human Genetics 74: 454-465. Sandweiss DH and Wing ES. (2013) Ritual rodents: the guinea pigs of Chincha, Peru. Journal of Field Archaeology, 24: 47-58. Sanger F, Nicklen S and Coulson AR. (1977) DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences of the United States of America 74: 5463-5467. Sankararaman S, Mallick S, Patterson N, et al. (2016) The combined landscape of Denisovan and Neanderthal ancestry in present-day humans. Current Biology 26: 1241-1247. Savolainen P, Leitner T, Wilton AN, et al. (2004) A detailed picture of the origin of the Australian dingo, obtained from the study of mitochondrial DNA. Proceedings of the National Academy of Sciences of the United States of America 101: 12387-12390. Schertl H, Krebs M, Maresch W, et al. (2007) Jadeitite from Hispaniola: a link between Guatemala and Antigua? 20th Colloquium on Latin American Earth Sciences, Kiel, Germany, Abstract Volume.
130 Seelenfreund D, Clarke A, Oyanedel N, et al. (2011) Paper mulberry (Broussonetia papyrifera) as a commensal model for human mobility in Oceania: anthropological, botanical and genetic considerations. New Zealand Journal of Botany 48: 231-247. Shapiro B and Hofreiter M. (2014) A paleogenomic perspective on evolution and gene function: new insights from ancient DNA. Science 343: 1236573. Siegel PE. (1989) Site structure, demography, and social complexity in the early Ceramic Age of the Caribbean. Early Ceramic Population Lifeways and Adaptive Strategies in the Caribbean 506: 193-245. Siegel PE and Severin KP. (1993) The first documented prehistoric gold-copper alloy artefact from the West Indies. Journal of Archaeological Science 20: 67-79. Simms TM, Rodriguez CE, Rodriguez R, et al. (2010) The genetic structure of populations from Haiti and Jamaica reflect divergent demographic histories. American Journal of Physical Anthropology 142: 49-66. Singleton H. (2012) The Zooarchaeology of Paso del Indio: Investigations of Subsistence at a Prehistoric Village Site in Puerto Rico. Manuscript on file with author. Skoglund P, Ersmark E, Palkopoulou E, et al. (2015) Ancient wolf genome reveals an early divergence of domestic dog ancestors and admixture into high-latitude breeds. Current Biology 25: 1515-1519. Skoglund P, Northoff BH, Shunkov MV, et al. (2014) Separating endogenous ancient DNA from modern day contamination in a Siberian Neandertal. Proceedings of the National Academy of Sciences of the United States of America 111: 2229- 2234. Smith CI, Chamberlain AT, Riley MS, et al. (2003) The thermal history of human fossils and the likelihood of successful DNA amplification. Journal of Human Evolution 45: 203-217. Solís Magaña C and Rodríguez M. (2000) Final Phase II Archaeological Evaluation for the Finca Valencia Site North Coast Superaqueduct Project Municipality of Arecibo, Puerto Rico. Prepared by LAW Environmental-Caribe for Thames- Dick Superaqueduct Partners, Guaynabo, Puerto Rico. Spotorno AE, Manriquez G, Fernandez A, et al. (2007) Domestication of Guinea Pigs from a Southern Peru-northern Chile Wild Species and their Middle Pre-
131 Columbian Mummies. In Kelt, DA, Lessa EP, Salazar-Bravo, J and Patton, JL (eds) The Quintessential Naturalist: Honoring the Life and Legacy of Oliver P. Pearson Berkley: University of California Press, 134: 367. Spotorno AE, Marin JC, Manriquez G, et al. (2006) Ancient and modern steps during the domestication of guinea pigs (Cavia porcellus L.). Journal of Zoology 270: 57-62. Spotorno AE, Valladares JP, Marin JC, et al. (2004) Molecular diversity among domestic guinea-pigs (Cavia porcellus) and their close phylogenetic relationship with the Andean wild species Cavia tschudii. Revista Chilena de Historia Natural 77: 243-250. Stahl PW. (2009) Adventive vertebrates and historical ecology in the pre-Columbian Neotropics. Diversity 1: 151-165. Stanish C. (2001) The origin of state societies in South America. Annual Review of Anthropology 30: 41-64. Stoneking M, Hedgecock D, Higuchi RG, et al. (1991) Population variation of human mtDNA control region sequences detected by enzymatic amplification and sequence-specific oligonucleotide probes. American Journal of Human Genetics 48: 370. Storey AA, Clarke AC, Ladefoged T, et al. (2013) DNA and Pacific Commensal Models: Applications, Construction, Limitations, and Future Prospects. Journal of Island & Coastal Archaeology 8: 37-65. Storey AA, Ramirez JM, Quiroz D, et al. (2007) Radiocarbon and DNA evidence for a pre-Columbian introduction of Polynesian chickens to Chile. Preceddings of the National Academy of Sciences of the United States of America 104: 10335- 10339. Team RC. (2000) R language definition. Vienna, Austria: R foundation for statistical computing. Thaw, S., Jaarola, M., Searle, J.B., Dobney, K.M., 2004. The origin of the Orkney vole Microtus arvalis orcadensis: a proxy for reconstructing human movements, In: RA, H., G, C. (Eds.), Atlantic Connections and Adaptations: Economies, Environments and Subsistence in the North Atlantic Realm, Oxford: Oxbow Books, 114-119.
132 Toro-Labrador G, Wever OR and Martinez-Cruzado JC. (2003) Mitochondrial DNA analysis in Aruba: Strong maternal ancestry of closely related Amerindians and implications for the peopling of northwestern Venezuela. Caribbean Journal of Science 39: 11-22. Underhill PA, Shen P, Lin AA, et al. (2000) Y chromosome sequence variation and the history of human populations. Nature genetics 26: 358-361. Van Oven M and Kayser M. (2009) Updated comprehensive phylogenetic tree of global human mitochondrial DNA variation. Human mutation 30: E386-E394. Vigilant L, Stoneking M and Harpending H. (1991) Human mitochondrial DNA. Science 53: 1503-1507. Vilar MG, Melendez C, Sanders AB, et al. (2014) Genetic diversity in Puerto Rico and its implications for the peopling of the Island and the West Indies. American Journal of Physical Anthropology 155: 352-368. Walker RS and Ribeiro LA. (2011) Bayesian phylogeography of the Arawak expansion in lowland South America. Proceedings of the Royal Society of London B: Biological Sciences 278: 2562-2567. Wallace DC, Garrison K and Knowler WC. (1985) Dramatic founder effects in Amerindian mitochondrial DNAs. American Journal of Physical Anthropology 68: 149-155. Wallace DC and Torroni A. (2009) American Indian Prehistory as Written in the Mitochondrial DNA: A Review. Human biology 81: 509-521. Wandeler P, Smith S, Morin PA, et al. (2003) Patterns of nuclear DNA degeneration over time—a case study in historic teeth samples. Molecular Ecology 12: 1087- 1093. Wang G-D, Zhai W, Yang H-C, et al. (2016) Out of southern East Asia: the natural history of domestic dogs across the world. Cell research 26: 21-33. Westaway MC and Lambert D. (2014) First Australians: Origins. In: Smith, C (ed) Encyclopedia of Global Archaeology. Berlin: Springer, 2787-2800. Whitehead WT. (2007) Exploring the wild and domestic: Paleoethnobotany at Chiripa, a Formative site in Bolivia, Berkley: University of California Press. Willerslev E, Cappellini E, Boomsma W, et al. (2007) Ancient biomolecules from deep ice cores reveal a forested southern Greenland. Science 317: 111-114.
133 Wilson JL, Saint‐Louis V, Auguste JO, et al. (2012) Forensic analysis of mtDNA haplotypes from two rural communities in Haiti reflects their population history. Journal of forensic sciences 57: 1457-1466. Wilson SM. (2007) The archaeology of the Caribbean, Cambridge: Cambridge University Press. Wing E. (2000) Guinea pig ( Cavia porcellus) remains from Finca Valencia (NCS-1), Northwest Puerto Rico, contribution to Phase II Zooarchaeology at the Finca Valencia (NCS-1) and NCS-4 Site, Northwest Puerto Rico. Final Phase II Archaeological Evaluation for the Finca Valencia Site North Coast Superaqueduct Project Municipality of Arecibo, Puerto Rico. Prepared by LAW Environmental-Caribe for Thames-Dick Superaqueduct Partners, Guaynabo, Puerto Rico. Wing E, Rouse I and Alegria R. (1990) Animal remains from the Hacienda Grande site. In: Excavations at Maria de la Cruz Cave and Hacienda Grande Village Site, Loiza, Puerto Rico. Yale University Publications in Anthropology 80: 87-101. Wing ES. (1986) Domestication of Andean mammals. In: Vuilleumier, F (ed) High altitude tropical biogeography, Oxford: Oxford University Press 246-264. Wing ES and Wing SR. (1995) Prehistoric ceramic age adaptation to varying diversity of animal resources along the West Indian archipelago. Journal of Ethnobiology 15: 119-148. Woods C and Kilpatrick C. (2005) Infraorder Hystricognathi, Baltimore: John Hopkins University Press. Xu S, Pugach I, Stoneking M, et al. (2012) Genetic dating indicates that the Asian– Papuan admixture through Eastern Indonesia corresponds to the Austronesian expansion. Proceedings of the National Academy of Sciences of the United States of America 109: 4574-4579. Yacoub HA, Fathi MM and Sadek MA. (2015) Using cytochrome b gene of mtDNA as a DNA barcoding marker in chicken strains. Mitochondrial DNA 26: 217-223. Yang DY, Woiderski JR and Driver JC. (2005) DNA analysis of archaeological rabbit remains from the American Southwest. Journal of Archaeological Science 32: 567-578.
134 Zierden MA and Reitz EJ. (2009) Animal use and the urban landscape in colonial Charleston, South Carolina, USA. International Journal of Historical Archaeology 13: 327-365.
135