The Victims at Sandby Borg

(1)

Author: Kerstin Calleberg Master thesis at the Archaeological Research

Laboratory, Stockholm University 2019

Supervisors:

Kerstin Lidén & Gunilla Eriksson

The Victims at Sandby Borg

- Tracing mobility and diet using

strontium analyses

(2)

Abstract

Sandby borg, an Iron Age ringfort on Öland, Sweden has been and is still at the center of attention in media and archaeological research. The massacre uncovered at the site during recent years opens many doors for analyses on the Migration Period (c. 400-550 AD) Iron Age skeletal remains. Eighteen teeth (molars) from 12 individuals and three rodent teeth were chosen for strontium (⁸⁷Sr/⁸⁶Sr) analyses. This was done to establish whether these individuals were locals or non-locals to Öland. The analyses displayed a, for the most part, local ⁸⁷Sr/⁸⁶Sr ratio. Two non-locals were identified, as well as a pattern of higher ⁸⁷Sr/⁸⁶Sr peaks on

numerous of the individuals during a certain age span, which could indicate a local weaning process with a special food.

Keywords: ⁸⁷^Sr/⁸⁶Sr, strontium, migration, Sandby borg, Öland, massacre, teeth, Migration Period, Iron Age, Laser Ablation.

Acknowledgment

Big thanks to my supervisors Kerstin Lidén and Gunilla Eriksson for interesting discussions and for sharing your knowledge on the subject. I would also like to thank Tammy Jamerson for reviewing the English in this thesis. Many thanks should also be given to my classmates for their wonderful dark humor, for backing each other, and for the nice hangs. Finally, all my love to my spouse Jimmy for your amazing support and patience. Also, thanks to my cat Iris for the lovely cuddles.

Cover illustration: Photographs by Kalmar county Museum and Kerstin Calleberg.

Illustration and cover assembly by Kerstin Calleberg

(3)

1. Introduction

1.1. Preface

During the European Migration Period (c. 400-550 AD) there was economic, political, and social instability among people and societies. The turbulence affected the Scandinavian realms and people’s movement across the land. The Swedes (Swedish: svear) and the Geats (Swedish: götar) discordance during this time is depicted in Beowulf and Old Norse sagas (Stjerna 1905). The dissonance and instability might be why numerous people were brutally slaughtered and left at the place of death inside ringfort Sandby borg on Öland in Sweden.

This has been interpreted as a massacre. Usually, people of the Migration Period buried the dead at cemeteries. Artifacts found inside the ringfort tell us that the people had long distance contacts to the Roman Empire and other parts of Europe during that time. This is therefore a very significant place and leaves many questions for researchers to answer. Many of the individuals show trauma to their head and body and no clear defensive wounds are present on the skeletons. DNA- and osteological analysis show that the majority of the adult individuals were men. Children of different ages and an infant tell us that there must have been women present on this site at the time of the massacre (Gunnarsson et al. 2015; Alfsdotter 2018). Of interest, to understand what happened on the site, is to identify if these people were locals (born and living on Öland) or non-locals (not born or living on Öland). To do this, teeth of some individuals were analyzed. Teeth grow in a special pattern and by analyzing the teeth’s strontium (⁸⁷Sr/⁸⁶Sr) values that grow during different times of the individual’s lifetime, a pattern of migration or diet could be detected.

1.2. Aims and research questions

The aim of this thesis is to establish whether these people were locals or non-locals to Öland by using strontium analysis on teeth with laser ablation.

- Are the individuals locals or non-locals?

- Can movement or dietary changes over time of the individuals be established?

- Is it possible to trace seasonal food intake?

- What can this tell us about the function of the ringfort?

- What can the analyses tell us about the events on the site?

(5)

2

2. Background

2.1. General background

The European Migration Period (c. 400-550 AD) started as a result of the weakened Roman Empire. What is seen as the springboard (375 AD) of the Migration Period was when a Mongolian nomadic people, the Huns, started to raid throughout Europe. During a hundred years they raided and plundered large parts of Central Europe. This was the starting point of the movement of people in Europe. The defeat of the eastern Goths was the start of the western expansion. At almost the same time the western Goths defeated the Romans and entered the Balkans in 378 AD and Italy the year 401. In 395 AD the Roman emperor Theodosius I divided the empire into two parts, the Western Roman Empire and the Eastern Roman Empire (Byzantium). The Goths occupied Rome in 410, which forced the Romans to leave Britannia. Around 400 AD the Franks and many other tribes entered the now open Gaul. These ravages have been described as barbaric and lively. Attila, who became the leader of the Huns in 446 AD, was defeated in 453, which led the Huns out of Europe. The Germanic people had in 492 taken the leading role of Rome and at the same time the kingdom of Francia was created (Kinder & Hilgemann 1981:113-117). Scandinavia was to a very high degree affected and involved in the warlike happenings on the continent. Evidence of this are for example, hillforts, ringforts, and buried treasures. Disturbance in society was probably not the only cause of building the forts and burying treasures. The period can be seen as an internal cultural growth and a regional distinctiveness, which were influenced by the continent (Burenhult 1999:287p). The import of solidi to Scandinavia has been of great importance. Most of the coins were used as jewelry pendants or were melted and reshaped (Burenhult 1990:293p). The solidi were most likely brought to Scandinavia together with returning soldiers from the Roman army. The coins were used as payment for serving the army (Fischer et al. 2011:192). The Iron Age societies were based on agrarian economy and animal production. The nuclear family worked as a central production unit. This would have included other relatives and allies with good coherence. Together they would have formed smaller realms where trade and needs of iron were satisfied (Burenhult 1990:298). In Denmark, during the same time, the society was based upon realms, political institutions, which controlled the social reproduction of craftsmanship. The basis of power for the realms was wars. The battles were between smaller realms and were also conflicts regarding trade- and plundering expeditions. Connections between the permanent elite and the travelling men during this time were also of great importance (Carlsson 2015:120).

Stora Alvaret, located in the southern parts of Öland, an island off the east coast of Sweden, started to be cultivated already during the Stone Age. Although, never to a large extent.

During the Iron Age the land was relentlessly used by the people of that time. This was probably when the Stora Alvaret got the look that is seen up to this day. No trees, flat, and open. The land was abandoned between the older and the younger parts of the Iron Age. It was once more cultivated during the Viking Age, but ultimately left as pasture land used by the local farmers (Blomkvist et al. 1990:75).

(6)

3

Fig. 1. The ringforts on Öland and the surrounding settlements dating from 200-700 AD. The red circles are the forts described in the text. Map based on Andrén 2014. Basemap: Esri, HERE, Garmin, ©OpenStreetMap contributors, and the GIS user community.

2.2. The ringforts on Öland

Öland is well known for the ringforts that are distributed around the island (fig. 1).

On mainland Sweden most of the forts are located near the larger lakes, smaller lakes, and on hills in the middle parts of Sweden.

These are called “hillforts”. In Uppland there are 60 of them, Västmanland has about the same amount, and Östergötland has up to 75 hillforts. In the northern part of Sweden there are only 15. The west coast also has a few. The island Gotland has 18 forts (Stenberger 1933:213) and Öland holds around 20 possible ringforts (Holmring 2014:4).

The ringforts on Öland are incredibly different compared to the forts on mainland Sweden, where mountains and hills, and other locations that were by far more defendable were used. Since Öland has a flat landscape, the inhabitants had to use other building techniques to build their forts. Hillforts from the mainland were built with irregular stone blocks, while the ringforts from Öland were built with limestone, which was modified for the purpose (Wegraeus 1974:33). Ismantorp, the best preserved ringfort still has visible house foundations in limestone. Eketorp, another ringfort is very well known because it was fully excavated between 1964-1974. Many of the ringforts show traces of settlement from the late 12^th to the early 13^th century. Fifteen forts (still visible) are known from historical records and old maps. Fifteen are still visible today. The forts are all of different sizes and shapes. The sizes

stretch from 60 m in diameter to 210 m in diameter. Most of them have a round or oval shape.

Two, Bårby and Treby are built in semi-circles and has three smaller circles in a row (Viberg et al. 2014:413p). The fort’s round form most likely derive from east European prototypes (Borg et al. 1976). Most commonly the forts have three gates, except for Ismantorp, which has nine gates. Not many of the ringforts have been excavated completely. Many of them are known from historical records, trial excavations, aerial photos and other sources. Other settlements from the same period (200-700 AD) are well known around the island. Around 2000 house foundations are known or preserved (fig. 1) (Viberg et al. 2014:413p).

Ismantorp, the best preserved ringfort, has nine gates through the wall and has a small open space in the middle. Early excavations have barely yielded any information about the ringfort.

The function of the fort has been interpreted by Stenberger in the early 20^th century as a

(7)

4

Fig. 2. Eketorp I, II & III. Based on Borg, Näsman &

Wegraeus, 1974.

Eketorp I EketorpII Eketorp III

shelter where people could take refuge in at times of instability during the Migration Period.

Stenberger also claimed that the ringfort was used as a cultic central place, because of the nine gates. Permanent inhabitation was never actual at the site, due to the absence of cultural layers found. The fort has a total of 95 houses, with foundations still visible today. These houses are radially joined. Radiocarbon dating and fieldwork during 1997-2001, lead by Anders Andrén, showed that Ismantorp’s past extends to a period over 700 years. Primarily, the fort was used during the years 300-600 AD. The function of the different houses has been interpreted as dwellings and workshops or byres, stables, barns, and stores (Andrén

2014:69pp).

Eketorp, as mentioned above, was fully excavated between 1964-1974. The excavations uncovered three phases of settlement (Borg et al. 1974:9) (fig. 2).

Eketorp I had an approximate area of 55-60 m in diameter. The wall-construction was about 3.4 m wide with a paved gate located in the south-west. The housing in this phase consists of conjoint long walls with two rows of post holes, which would have supported the roof. The number of houses was about 20-21. Water would have been supplied by the water- filled depression outside of the ringfort during this period (c. 300-400 AD). Cultural layers from this phase were thin. It has been suggested that the fort was not inhabited permanently, since no fireplaces were detected. The site might have been a meeting place for affairs, a cultic lay, or for fairs (Näsman 1974a:49p).

Eketorp II was larger than Eketorp I. Eketorp II (c 400- 700 AD) had a diameter of about 80 m. The wall was c 5 m broad and there were 3 gates, a wider one in the north and narrower ones in the south-west and in the east. The eastern one head right towards the water filled depression. Housing inside the fort forms a circular street-system. This is called a “Ring-street”

communication system. The houses in this phase are similar to the houses in Eketorp I, although the houses in Eketorp II have gables, which were built at the same time as the long walls. This phase, however, shows a great extent of inhabitance, compared to Eketorp I.

The cultural layers were rich in findings and

contained construction material, animal bones, charcoal, and artifacts. Many of the layers were also burnt, which indicate residential living. Houses in Eketorp II have been interpreted as storehouses, dwelling houses, byres, and workshops. The expansion of the fort in this phase can be seen as an enhanced need for shelter for the inhabitants and their livestock.

Eketorp II and the finds indicate that this was a highly functioning farming village. The finds from this phase show that the abandonment of the fort was peaceful, and perhaps caused by a changing surrounding environment. After the abandonment around 700 AD, the fort was uninhabited for 300-400 years. This abandonment layer contained scattered human bones, many of them fragmented, which could indicate some continuation of a sheltered place at unstable times (Näsman 1974a:51-55).

(8)

5

During the Viking Period, Eketorp was once again occupied. This phase is called Eketorp III.

The houses were different from those before but did to some extent follow the past housing plan. The main gate was the south-east entrance. The northern gate was closed. An outer wall was also built, which embraced the inner wall and the workshops. The middle part of the ringfort was used as a central space. The houses and the streets had poor constructions and were rebuilt as least two or three times. Evidence from the excavations show that the site had been permanently inhabited at some times during this phase. The site was most likely

occupied by high standard living people, because of the finds, such as, weapons and horse trappings. Connections to the Baltic were also drawn using some of the finds. It was not an ordinary farming community, but instead a special place with peculiar characters. There is no evidence of a ruthless abandonment; it was most likely peaceful and made due to changes in trading and shipping (Eketorp’s location does not support a harbor) (Näsman 1974a:56-59).

Bårby borg, another ringfort on Öland (see fig. 1) was built in a semicircle facing a 20 m high precipice, which act as a natural defense. There are two gates to enter Bårby. The main gate was built up with mortar some time during the Middle Ages. The mortar and other finds inside the fort further support the reuse of the fort during the Middle Ages. The fort was probably built during the 5^th century and finds show that Bårby was used during the 6^th- , 7^th, and the 8^th century (Stenberger 1933:222p).

Mossberga borg (see fig. 1), has an oval shape and the walls were probably about five meters wide during its standing days. The fort has two gates. One in the east, and one in the west.

The housing inside the fort was documented during the 1600s, as radially joined houses. In the 1900^th a solidus from the 5^th century was found. Other finds from the 5^th century are a bronze figurine of a sitting man, a bronze needle, bronze tweezers, lance heads, iron arrowheads, and bronze relief brooches (Stenberger 1933:243pp).

2.3. Sandby Borg

Sandby borg is placed right next to the eastern shore of Öland. The distance to the Baltic sea is about 42 m. The site has formerly been cultivated. Sandby borg has an oval shape and the inner measurements are c. 66-92 m. The wall is demolished, but at some places a cavity wall structure can be observed. The thickness of the wall is estimated to have been 4 m and the fort has had at least two gates: one in the south-east and one in the north. West of the wall is an amplification made up with larger blocks of granite (an outer wall) (Stenberger 1933:225p).

Sandby borg is known to be of a middle-sized fort compared to the other forts on Öland. The ringfort is well known from the 19^th century and was well documented during 1933 and thereafter in 1974. Aerial photos from 1974 revealed visible housing inside Sandby borg.

Here, it was proven that Sandby borg also had radially built houses (c. 53 houses in total) with a central block, just as Eketorp II had (Näsman 1974b:77). The houses inside the fort was noticed early, but not precisely documented until 2010, when geophysical prospections made it possible to see the housing more clearly (fig. 3) (Viberg 2012:8). In 2010 probable

plundering pits were noticed in Sandby borg. This was the start of the archaeological

excavations on the site, which are still active today. The finds, which they found when doing a search with a metal detector were prestigious jewelry depots containing, for example, rare gold-plated relief buckles (fig. 4) and beads buried near the limestone wall structures (Leivas

& Victor 2011:10pp). Through the finds, Sandby borg, has been typologically dated to the Migration Period (c. 400-550 AD). ¹⁴C-dating also confirms the dating of Sandby borg’s use (Gunnarsson et al. 2015:13). Of the c. 53 houses, only two of them have been fully excavated (house 40 & 4) (Gunnarsson et al. 2015; Papmehl-Dufay 2019).

(9)

6

Fig.3. Interpretations of geophysical surveys. The numbers in the picture are Viberg’s own interpretations of the houses.

The black and green lines represent anomalies recorded by the instrumentation. The blue circles are possible post holes.

G1-G3 marks the possible entrances (gates) (Viberg 2012:8).

Excavations of the site started in 2011 by Kalmar County Museum (Leivas & Victor

2011:10pp). This was when the first skeletons started to appear under the ground surface. As more individuals appeared inside the fort it was clear that Sandby borg was not reused after these people were killed, compared to other ringforts on Öland (Alfsdotter 2018:422pp).

Under the topsoil in house 40, there was a layer of rubble from when the houses collapsed.

Under the rubble the finds started to appear. This layer has been interpreted as representing the time when the houses were abandoned. The layer included charcoal and animal bones. At the bottom of this layer the skeleton emerged, which further supports the abandonment phase.

Underneath the skeletons the surface which would have used as a floor (the phase of

inhabitation) was revealed. This layer testifies to the time of the massacre and the moments right before the incident. For example, finds connected to weaving gives further insight on the use of house 40 (Gunnarsson et al. 2015:25pp).

One of the skeletons, an elderly man, had goat teeth stuck in to the mouth and had fallen over a heard sometime during or after his death (Papmehl-Dufay & Alfsdotter 2015:41-49). The brutally treatment of these people, both children and adults, indicates that this action could have been some kind of invasion. The cause and course of the invasion and slaughtering can be extensively discussed. After the massacre, which the event has been interpreted as, no survivors seem to have wanted, or could enter the site. The massacre must therefore have had

Fig. 4. Relief buckle found in 2010 inside the walls of house 52. Photo: Jan-Henrik Fallgren.

(10)

7

a great impact on the surrounding settlements and its residents. The event has been interpreted as a revenge act, as well as a part of a civil war on Öland during that time (Alfsdotter 2018:

441pp).

During the late 5^th century’s Öland, Sweden, and all of Europe were faced with an

apprehensive time. According to Helena Victor the massacre at Sandby borg could have been a launch of a civil war on Öland. Somebody was probably trying to attain the power, and might have succeeded, due to the scene found at Sandby. As mentioned above, it was a humiliation to not be buried after death and this act might have been some sort of a power manifestation. One of the reasons for thinking that the massacre was executed by other people from Öland is the fact that surviving relatives could not bury their dead. The enemy must have been guarding the fort, and therefore must have been living close to it. Barely any

weapons have been found at Sandby borg, yet, which also could indicate that the killing of the people was no siege but might have been an inside job (Wallén-Widung 2016).

The evidence of this trauma was also incorporated into the now living elderly generations on Öland. As children, they had been told not to enter Sandby borg. The reason for it was long forgotten (Wallén-Widung 2016).

2.4. Strontium isotope analysis

By measuring strontium isotopes in teeth enamel, information about migration and diet can be determined. There are four isotopes which exist naturally. These are: ⁸⁸Sr, ⁸⁷Sr, ⁸⁶Sr, and ⁸⁴Sr.

β-decay of ⁸⁷Rb (Rubidium, half-life 4, 88 x 10¹⁰yrs) is the process in which ⁸⁷Sr forms. The variation in ⁸⁷Sr/⁸⁶Sr depends on the rock age and the primary Rb/Sr ratio of the rock. The variation stretches from 0.702 to 0.750. Sea water today has an unvarying ⁸⁷Sr/⁸⁶Sr -ratio (0.7092). The Baltic Sea, which has brackish water, is much affected by the weathering of exposed older bedrock. The bedrock is transported from the rivers of the mainland to the Baltic. This gives these parts of the Baltic Sea a ⁸⁷Sr/⁸⁶Sr -ratio of c. 0.718 to 0.745. In the south, the bedrock of the mainland is younger, which gives river draining areas a ratio of c.

0.710. Today, the Baltic Sea’s ⁸⁷Sr/⁸⁶Sr ratio ranges between c. 0.7092 to 0.7097, though, the values increase in the northern parts. Soil and groundwater embody the bedrock weathering into the food chain. The local bedrock’s ⁸⁷Sr/⁸⁶Sr -ratio in a region where plants and animals thrive will be reflected in their embodied composition, thus, traceable for researchers

(Fornander et al. 2015:183, Bentley 2006). Strontium analysis requires a variation in the strontium isotope ratio in the surrounding area studied. If the ratio is not variated the use of strontium analysis will limit the results (Ericson 1989:254).

2.5. Other Sandby borg analyses

Stable isotope analysis on grains (barley) from Sandby borg shows high values of nitrogen.

This could point towards fertilization of the farm grounds (Eklund 2019). Dietary studies were conducted by Gunilla Eriksson. The results show a for the most part homogenous protein dietary intake with some mixture of terrestrial and marine protein (Eriksson 2019).

(11)

8

Fig. 5. Bedrock in Öland and the nearby mainland Sweden.

Data: © SGU. Basemap: Esri, HERE, Garmin,

©OpenStreetMap contributors, and the GIS user community.

Fig. 6. Soil distribution in Öland and the nearby mainland Sweden. Data: © SGU. Basemap: Esri, HERE, Garmin,

©OpenStreetMap contributors, and the GIS user community.

2.6. Öland geology

Most of Scandinavia has heterogeneous bedrock with old rocks, for example gneiss and granite. Öland, Gotland, and the southern areas (Denmark etc.) have younger bedrock, which is more calcareous. Öland and the coastal areas next to Öland are dominated by young bedrock units aged 850-34 million years. Other parts of the mainland are dominated by much older bedrock aged 1880-1740 million years (fig. 5). The northern parts of Öland consist of eroded land with a thin soil layer (fig. 6). The southern parts are framed with two long ridges, which were shaped by the fluctuations of the Baltic sea. The northern area between the ridges consists of fertile soil. The lower parts, Stora Alvaret, consist of eroded limestone with a thin soil layer. The soil distribution on Öland is, however, not as unvarying as the bedrock. During the last Ice Age, the melting of the glaciers wound has transported older rocks onto the fertile middle parts of Öland. This could give these parts a higher Strontium (Sr) (⁸⁷Sr/⁸⁶Sr) value.

Most parts of Öland is covered in till and glacial sand and silt. The northern parts are, however, dominated by postglacial sand and silt. The combinate heterogeneous ⁸⁷Sr/⁸⁶Sr - variation for the whole island would therefore be quite wide (Wilhelmson & Ahlström, 2015:33-34). Faunal samples from Öland do, to some extent, support a heterogenous ⁸⁷Sr/⁸⁶Sr -mixture in a homogenous outline on the island. Human samples, however, have not showed any indicators of correlation with the different soil dispersal (Wilhelmson & Ahlström 2015:37p).

(12)

9

3. Material

The material consists of human teeth (18 molars) from twelve individuals uncovered inside Sandby borg in 2014, 2015, and 2016 (fig. 7). The samples were chosen in collaboration with coexistent other analyses. Three samples from rodents from the fort were also included in this analysis. The teeth analyzed are listed in appendix no. 6.

Fig. 7. The approximate location of the individuals inside Sandby borg. Background photo: Kalmar county museum. House interpretations based on Viberg 2012.

Individual 3 (6948) from house 40

The remains of this individual (fig. 8) consist of only a few bone elements; a crushed cranium, a part of an upper arm, and a part of a forearm. The skeletal remains were located together in the middle part of the house. It is, however, not certain that the bones from the arm are from the same person as the cranium, but it is very likely. The cranium was found on top of a posthole and in between two ledges. Sooty spots also surrounded the individual. Finds found near the individual consisted of organic material, metal, ceramics, glass, and disarticulated human bone elements. The individual was an adult. The wisdom teeth from the maxilla were fully erupted. The teeth were affected by fire and the preservation of the teeth was therefore poor. The individual’s skull shows traces of sharp force trauma and the fire most likely burnt while the individual still had soft tissue present on the body (Gunnarsson et al. 2015).

(13)

10

Fig. 8. Individual 3 (6948). Photo: Kalmar county museum.

The individual (fig. 9) laid on the stomach facing west. The individual was found in the middle part of the house, near a fire stone and in between hollows in the ground. Finds found near the individual consisted of metal, burnt clay, and ceramics. Different finds near the individual were a mounting to a sword, a millefiori bead, and glass beads. The shanks from the individual were absent, but the femurs were still intact. Rubble from the housing caused the preservation of the skeletal to be poor. The calvarium was affected by fire. It is possible that the calvarium exploded during fire. The fire must have taken place when soft tissue was still present. It was hard to determine the sex of the skeleton, although some features and DNA-analysis suggest masculine. The teeth were affected by tartar and caries. No clear wounds have been detected, but the position of the individual suggests a sudden death (Gunnarsson et al. 2015).

(14)

11 Individual 5 (6356) from house 40

The preservation of this individual (fig. 10) was very poor. Most of the upper part of the body is missing. Cranial parts, the left arm, and some parts of the torso rested on a stone slab. The skeleton was disrupted and damaged by rubble from the house. Some parts of the skeleton were also affected by fire. The skeleton was found in the middle part of the house facing south. The head rested on a probable fire stone. Finds near the skeleton consisted of metal and ceramics. Different finds were fragments of iron, glass beads, and a gold-plated silver press plate mounting. In the temporal bone three small holes are shown. The teeth from the individual are fragmented. The molar tooth show signs of hard wear and some tooth roots were fire affected. It was hard to determine the sex of the individual due to poor preservation of important bone elements. The length of the individual, teeth wear, the cranial heaviness, and DNA-analysis suggest the skeleton to be masculine. Individual 6 partly laid over parts of individual 5, which suggest that individual 5 fell before individual 6 (Gunnarsson et al. 2015).

Individual 6 (fig. 11) is one of the two best preserved skeletons. The skeleton is almost complete. Hand bones from individual 5 were found under the fibula from individual 6. The cranium and the pelvic girdle laid on rocks and the cranium was crushed. Finds of metal was found near the skull. Near the skull was a gold-plated relief buckle. Close to the individual’s lumbar vertebrae a small femur was found. It belongs to an infant. The mandible was partly ragged by a rodent. Teeth eruption and ossification suggest that the individual was young, around 12-15 years old. According to DNA-analysis the individual was masculine. The teeth were affected by tartar and enamel hypoplasia, which vaguely could suggest that the

individual was malnourished. Blunt force trauma and sharp force trauma are suggested from reconstruction of the skull (Gunnarsson et al. 2015).

(15)

12

Individual 7 (fig. 12) is one of the two best preserved skeletons. The skeleton was articulated and laid in hocker position with the skull and torso towards west. Some parts of the skeleton, including the skull were badly preserved due to rubble from the house. Finds near the skeleton consisted of metal, glass, and amber. Other finds were glass beads and two loom weights. The skeletal remains showed no signs of being affected by fire. During excavation and preparation of the cranium a big crack was noticed. No sharp trauma was detected, but the individual was probably subjected to blunt force trauma. The injuries are located on the right-hand side of the skull, which is the side that was facing the ground. This further supports trauma to the head.

The hocker position also supports the head trauma and that the individual took this position some time before death. The individual was not fully grown, which makes it hard to

determine the sex of the individual. DNA-analysis suggests that the individual is masculine.

Skeletal elements from this individual showed inflammatory changes. The changes are suggested to be medial tibial stress syndrome. The individual also had a neglected back health. Deep Schmorl’s nodes were found on four of eleven of the vertebras. This could indicate hard physical work. The health of the teeth is good. Tartar was present, which is common (Gunnarsson et al. 2015).

(16)

This individual (fig. 13) was mainly articulated and had fallen over a hearth. Inside house 52 several jewelry depots have been uncovered. The finds include gold-plated relief buckles, gold and bronze rings, silver-beads, silver-bells, glass beads, arrow-heads, and bronze-beads.

Scattered human bones were also found inside house 52. Most of the cranial elements were absent, except for the maxilla. All the teeth from the maxilla were found, even though parts of the maxilla were missing. The disarticulation of some parts of the body is most likely due to taphonomic events after the individual’s death. The individual is most likely a man, that probably was probably at least 45 years old, and around 171 cm tall. The individual had substandard dental hygiene and tartar was present. When it comes to the maxilla, the bone was vigorously reduced. This was probably due to inflammatory diseases. The skeletal changes are not especially remarkable for a person of such age for that time. None of the joints show traces of stress changes, such as arthritis. No obvious trauma was detected on the individual, although some sort of trauma should have caused the position at death. The position of the skeleton suggests some sort of trauma to the head, based on the position and the absence of cranial parts. Fire affected many parts of the individual sometime during or after his death when soft tissue was still present on the body. Since he covered the fire, it must have burnt out after a short period of time. During excavation of the man other bone elements were found, two teeth from a goat or a sheep were found between the jaw halves. These teeth must have been inside the maxilla during time of death (Papmehl-Dufay et al. 2014:21-22, 41-47). This has been interpreted as some sort of desecration (Papmehl-Dufay et al. 2014:60).

(17)

This was the most articulated skeleton that was uncovered during 2016. The preservation was moderate, although the cranium was very fragmented. The skeleton was found inside house 4, by the entrance. Many of the taphonomic losses are probably due to rubble from the housing.

Dental eruption gives the individual an age of about 6-8 years. Gender determination was not possible by osteological analysis. The bone health was good and the teeth were in good condition (Alfsdotter, manus Humanosteologi 2018).

This young individual was found inside house 4, by the entrance. The skeleton was not articulated and was found just south east of individual 12 and was mixed with bone elements from individual 14. Bone elements found scattered inside house 4 might have belonged to individual 13. This individual was probably in his/hers early teens (10-15 years). The individual was decapitated, probably on the lower part of the neck. Disruption in the enamel is probably due to malnutrition during parts of the individual’s lifetime. Enamel hypoplasia was found on the molars and premolars (Alfsdotter, manus Humanosteologi 2018).

This individual was an elderly man, about 176 cm tall. Most of the joints had signs of wear and the right femur and a foot bone showed degeneration. The lumbar vertebra, the second- and the first cervical vertebra showed ossification and macroporosity. The pathologies of the joints strengthen the age determination (55+ yrs.). Enamel hypoplasia was present on the molars in the maxilla. This shows that the individual suffered from malnutrition during his childhood. The teeth bone has been sealed after permanent teeth have been lost. This also indicate a high age at death. The lower part of the skeleton was articulated, and the cranium was found closer to the lower parts than the upper part of the body. The man could have been decapitated (Alfsdotter, manus Humanosteologi 2018).

Individual 18 (8873) from the street

On the street near house 4, bones from individual 18 were found. Parts of the skeleton were also found further out on the street. The cranial fragments were found near the house wall.

The skull seams (lambdoid sutur) have almost closed and the wisdom teeth have erupted. The sacrum was not fused, which indicated that the individual was at most 25 years old. The dental wear indicates an age of about 25-30 years, although the dental development together with the cranial development indicates an age between 15-20 years. The frontal bone on the cranium, the glabella looks feminine. It is problematic to determine the gender because the pelvic bone from this individual was not found. Ongoing DNA-analysis will hopefully determine if this individual is female or male. The molars, premolars, and the incisors have enamel hypoplasia, which indicate that this individual was malnourished several times during the childhood. Tartar is also present on many of the teeth found. A sharp fracture can be seen on the shoulder blade. The surface of the bone is poorly preserved, which made it harder to determine the injury. The trauma is sharp and could have been caused by a stab to the shoulder with a sharp weapon (Alfsdotter, manus Humanosteologi 2018).

(18)

15 Individual 20 (8458) from the street

This individual was disarticulated. The bone fragments consisted of cranial fragments, 16 teeth, fragments from the maxilla and the mandible, and a metatarsal bone from a child aged c. 5 years. The skeletal remains were porous, except for the teeth. The individual was

disarticulated and found in a turbid area of the street, mixed with bones from animals and other individuals. Deciduous teeth (incisors) were stuck in the jaw bones. The permanent teeth are seen in the maxilla. They have not erupted but can be observed. Other teeth, which were in the development stage were loose, for example the M2, which consists of the crown with enamel that were not fully developed (Alfsdotter, manus Humanosteologi 2018).

This individual was found on the street in front of the gables of house 5 and 4. The finds consisted of a cranium, teeth and an upper extremity. According to the teeth eruption, the individual died at an age of about 12-15 years +- 36 months. One tooth was affected by caries.

Teeth from the maxilla were found articulated against the wall of house 4. This indicated the original location of the skull. Teeth from the maxilla were found in front of the gable of house 5, which most likely belongs to the same individual. Some of the cranial bone elements and teeth were affected by fire (Alfsdotter, manus Humanosteologi 2018).

Individual 23 was found on the street in front of house 4. The skeleton was porous. The dental wear gives the individual an age of about 25-35 years and the wisdom teeth have erupted. The ongoing fusion of the pelvic bone suggests that the individual was an older teenager/younger adult, which gives the individual an age of about 20-25 years. Fragments from the pelvic bone indicate that the skeleton belonged to a man. The front part of the mandible shows traces of fire. The left pelvic bone has an injury that was most likely caused by a weapon with an edge.

An injury like this would most likely have killed the individual; however, it is difficult to determine if other factors played a part in the death (Alfsdotter, manus Humanosteologi 2018).

Reference fauna

The reference fauna used in this study is based on published data that were collected from a passage grave in Resmo on Öland dating from c. 3500 BC to 1000 BC. Materials were also collected from the Neolithic site of Köpingsvik on Öland. The species included were mountain hare (Lepus timidus), wild boar (Sus scrofa), and roe deer (Capreolus capreolus).

Modern material, from the same study, was also included, which consisted of four shells from land-dwelling white-lipped snails (Cepaea hortensis). The marine samples (prehistoric marine mollusk shells (Mytilus edulis, Cerastoderma & Macomabaltica)) included were collected from the Baltic Proper and were dated to the Neolithic and the Bronze Age. Included in the analysis were also three bones from harp seal (Phoca groenlandica) and ringed seal (Pusa hispida) (Fornander et al. 2015:184pp).

Based on the faunal material, the local strontium isotopic range on Öland is estimated to a range between 0.7102 and 0.7158 (Fornander et al. 2015:187). To verify the local range, three rodent teeth from Sandby borg were analyzed in this study. The rodents with find number 6136 and 7178 were found inside house 40.

(19)

16

3.1. Teeth

Teeth are formed in the jaws and erupt through the gum. After the teeth are formed, no alteration is possible, unless it is made by teeth wear, demineralization, and breakage. Teeth carry most of the information about the human that possessed it. They can give information about sex, age, diet, health, and evolution. The environment an individual sojourn in will firstly interact with the teeth through consumption of food. Adult humans have 8 incisors.

These are formed as spatulas and are located in the front of the lower and upper jaws (the mandible and the maxilla). If the incisors have no wear they have sharp edges. The canines, which are next to the incisors function as an extension of the incisors and have a conical shape. The premolars come in pairs of fours in the upper and lower jaw. Molars are the largest teeth. They have a large chewing surface and emphasize grinding and crushing. Usually adults possess 12 molars. There are two sets of three molars in both the upper and the lower jaw. Deciduous (milk or primary) teeth erupt and function in the first years of an individual’s life. These teeth are systematically replaced by the permanent adult teeth (White & Folkens 2005:127p). The teeth are divided into two major parts, which are called crown and root. The core of the teeth is called dentine. Enamel coats the crown, while cement enfolds the roots of the teeth. The part where the crown and the root meet is called the cervix (neck). Inside the tooth there is a pulp chamber. This part is soft tissue and opens into canals, which connects to the roots (a tooth can have several roots) (Hillson 1996:8).

The permanent teeth start to develop long before they erupt. The crown’s enamel of the upper permanent first molar starts to develop at the age of birth and is complete at the age of c. 3-3.5 years. The crown of the upper permanent second molar start to develop at the age of c. 2-2.5 years and is fully developed at the age of c. 8-8.5 years. At the age of c. 8-8.5 years the crown of the third molar starts to form. The formation of the M3’s crown is finished at the

approximate age of c. 14-14.5 years. The estimated eruption ages of the molars are as followed: for the M1: 9-9.5, M2: 15.5, and M3: 21-21.5 years of age (Beaumont &

Montgomery 2015:409).

4. Ethical aspects

The study and handling of human remains can entail questions about ethics. Different

countries, cultures, and continents have diverse views on how dead people should be handled.

The beliefs and thoughts on this matter ultimately align with the sole individual (White et al.

2011: 358; Sellevold 2012:141). The main differences dealing with the remains in this thesis, concerns the scientists and the general public. Human remains are a main source of

information about past civilizations and cultures. Archaeologists and scientists have an ethical obligation to take part in the development of knowledge. This is one of more reasons why human remains should be available for research. Today, the indigenous people are those who express concern about the ethical treatment of human remains. Generally, people who have deceased in close connection to the today living people originate ethical concerns (Sellevold 2012:141pp). The remains uncovered at Sandby borg derive from the 5^th century and should therefore not give rise to major ethical concerns regarding time. Usually, the ethical

discussion involving human skeletal remains only include unburned remains. The Guiding Principles, proposed by the SAA – the Society for American Archaeologists, argue that remains that do not have a clear context or association in the archaeological senses are scientifically less appreciated (Sellevold 2012:152). This might be why cremated human bones acquire less ethical concerns. The intermediation of the remains from Sandby borg are

(20)

17

the prime ethical concern expressed by researchers on the project (Kalmar läns museum 2019;

Östra Småland Nyheterna 2018).

5. Methods

5.1. LA-MC-ICP-MS

The primary method that was used was Laser ablation multi-collector inductively coupled plasma mass spectrometry (LA-MC-ICP-MS). It is possible to gain information about lifecycles and migration patterns by measuring Strontium isotopes sequentially on the tooth enamel and thus obtaining time resolved variations of the isotope signature. The method only leaves small sampling lines that are barely noticable with the naked eye (Glykou et al.

2018:12515pp). The strontium ratios (⁸⁷Sr/⁸⁶Sr) are determined by the local geology and other external processes and vary regionally. When a human or an animal consumes local food, the skeletal tissue that is formed during that time will reflect the local strontium ratio (Willmes et al. 2016:102; Ericson 1989:252p; Fowler 2007; Bentley 2006).

The ⁸⁷Sr/⁸⁶Sr ratios in the tooth enamel were measured at Vegacenter at the Swedish Museum of Natural History, Stockholm using a NWR193 excimer ablation system, coupled to a Nu Plasma II multi-collector ICP mass spectrometer. Argon and helium gas was used to transport the extracted material from the laser ablation system to the mass spectrometer. After the teeth had been cleaned using an ultrasonic bath and gentle wiping of the enamel with ethanol they were mounted into the sample chamber (a TwoVol2 ablation cell) with the enamel parallel to the cell surface. The enamel surface was sampled with multiple tracks/lines (parallel to growth zones) spaced aproximately 200µm from each other to cover the exposed tooth enamel. The number of lines and the length of the lines was determined by the size and curvature of the tooth. In order to remove any possible contamination, prior to the actual measurements, a pre-ablation was performed. The instrument was tuned between each sample to adjust variations in gas flows and to maintain a high beam intensity. For quality control, tooth reference materials were ablated prior to, during, and after every run (Glykou et al. 2018:1218p; Vegacenter, Instrumentation 2017). For this, a tooth from a swedish hare and a tooth from a rodent from South Africa (otomys) were used. Those materials have a known homogenous isotope signature, as determined by solution TIMS (Thermal Ionization Mass Spectrometry) measurements. Laser ablation has gained more attention in archeology in recent years. Some of the reasons are that the sample preparation and measurement time is shorter than for other methods (e.g. solution TIMS) and it can provide excellent spatial resolution, however with the caveat of lower precision on individual isotope ratios. The method has also been more popular among museums and researchers because of the minimal destruction on the samples (Golitko 2016:23p).

5.2. Enamel formation age

To calculate the formation age per line on each individual the length of the enamel was estimated by adding extra sample lines on the enamel on some teeth to cover the whole enamel (fig. 14). The top part of the crown on the teeth from ind. 9 and 15 (that would have represented the earliest stages of the tooth’s formation) were well-worn. The avarage length for the teeth’s enamel that were in better condition (the M1, M2, & M3) were added to the worn teeth to hypothetically visualize the original length of the enamel (fig. 15).Then the spotsize and line spacing could be put together to get an approximate length of the enamel surface. To get the part of year for formation, the spotsize and the line spacing was divided

(21)

18

Fig. 15. Ind. 6 M1 & Ind. 15 M1. The crown surface of the tooth from ind. 15 is well-worn compared to the tooth from ind. 6.

with the enemel length. By multiplying the enamel length in years (Beaumont & Montgomery 2015) with the part of formation for the spotsize and the line spacing and by multiplying that with 365 the formation time in days will be procured. To get the start age (of formation) (Beaumont & Montgomery 2015) in days, the start age in years is multiplied with 365.

Ultimately the formation age per line in years is calculated by adding together the start age in days with the line number of interest multiplied with the formation time in days for the line.

By adding the line number minus one, multiplying with the formation time per days for the line spacing, and dividing it with 365 the formation age per line are projected.

Formation age equations:

(Spotsize mm × Lines amount) + (Line spacing mm × Spacing amount) = Enamel length mm Part of year for formation spot = Spotsize ÷ Enamel length

Part of year for formation space = Line spacing ÷ Enamel length

Formation time days spot = (Enamel length yrs × Part of formation spot) × 365 Formation time days space = (Enamel length yrs × Part of formation space) × 365 Start age days = Start age yrs × 365

Formation age per line in yrs = (Start age days + (Line nr × Formation time days spot) + ((Line nr – 1) × Formation time days space)) ÷ 365

Fig. 14. The arrows represent the teeth’s enamel growing direction from the tip of the crown to the root. The number on the arrows represent the amount of sample lines that were analyzed. The added numbers represent lines (not analyzed), which were added to cover the whole enamel. These need to be a part of the equation to calculate the formation age for each sample line.

(22)

19

5.3. Theory and definitions

The theoretical framework in this thesis is based upon scientific theory and is verified by means of scientific methods. Conclusions and theories which are presented are based on science and hermeneutics, which is about interpretations and understanding. The

interpretations are not presented as the truth between cause and effect, rather as new and rewarding ways of understanding people’s feelings, motives, patterns of thoughts, and other activities linked to the scene uncovered inside Sandby borg.

Migration, in this thesis, is used when trying to understand the movement of the people found inside Sandby borg. Movement, points to the Sandby borg people’s movement inside and outside of the island Öland during the childhood. Locals to Öland means that the people are born and raised on Öland and show a minimum to no movement outside of Öland. Non-locals are the people who display, for the most part, a non-local ⁸⁷Sr/⁸⁶Sr-ratio (not from Öland).

(23)

20

7. Results

All teeth sampled yielded useful results. The ⁸⁷Sr/⁸⁶Sr values are plotted against the age of enamel formation. By presenting the results like this, varieties may well be distinguished. All data sets are presented in the appendix.

Fig. 16. Age of enamel formation for individual 3.The local range is marked with light blue striped lines. This individual is represented with one tooth (M3). The age in years of enamel formation is plotted against the ⁸⁷Sr/⁸⁶Sr-values.

Individual no. 3 is outside of the local ⁸⁷Sr/⁸⁶Sr-range on Öland (fig. 16). The crown started to develop at the age of c. 8 years. The ⁸⁷Sr/⁸⁶Sr -ratio was higher at the age of c. 9.4 years. By the time the crown was complete, at the age of c. 14.8 years the individual had an even higher

87Sr/⁸⁶Sr -value (0.726). The individual was an adult male (?) and was represented in its context by a crushed cranium and an upper fore arm. The Total Sr-Beam (V) (concentration) ranges from 1.67 to 2.99.

(24)

21

Fig. 17. Age of enamel formation for individual 4. The local range is marked with light blue striped lines. This individual is represented with one tooth (M1). The age in years of enamel formation is plotted against the ⁸⁷Sr/⁸⁶Sr-values.

Enamel from line 13 and 14 on individual no. 4, were developed outside of the local range.

This would be at the age of c. 3-4 years. This skeleton was in poor shape; the cranium was scattered, and many parts of skeletal element were missing. The individual was most likely a middle-aged male adult. The Total Sr-Beam (V) (concentration) ranges from 0.44 to 0.83.

Line two to six on individual 5 could reflect the mother’s local ⁸⁷Sr/⁸⁶Sr -values (fig. 18).

During the rest of the formation stages the individual’s ⁸⁷Sr/⁸⁶Sr -values are for the most part within the local range. The condition of this individual was poor. Most skeletal elements from the upper body were missing. The individual was an adult and was probably a male. The Total Sr-Beam (V) (concentration) ranges from 1.15 to 2.03.

(25)

22

Line number three to six, which are outside of the local range, represent individual 6’s second year in life (fig. 19). The individual was at the age of c. 2.7 years inside the local range on Öland. Individual 6 is one of the best-preserved skeletons and is almost complete. The individual was a teenager, aged 12-15 years. DNA-analysis suggests the skeleton to be male.

The Total Sr-Beam (V) (concentration) ranges from 0.47 to 1.39.

Fig. 20. Age of enamel formation for individual 7. The local range is marked with light blue striped lines. This individual is represented with two teeth (M1 & M2). The age in years of enamel formation is plotted against the ⁸⁷Sr/⁸⁶Sr-values.

The ⁸⁷Sr/⁸⁶Sr-values for individual 7 are mostly local, but stands out at the age of c. 1.4 to 2.9 years (fig. 20). Individual 7 is one of the best-preserved skeletons. The individual was not fully grown. The age is estimated to 12-15 years. The individual had neglected back health, which could suggest hard physical work. According to DNA-analyses the skeleton is masculine. The Total Sr-Beam (V) (concentration) ranges from 0.62 to 2.18.

(26)

23

Fig. 21. Age of enamel formation for individual 9. The local range is marked with light blue striped lines. This individual is represented with three teeth (M1, M2, & M3). The age in years of enamel formation is plotted against the ⁸⁷Sr/⁸⁶Sr-values.

Individual 9 was represented by three teeth; a year span from a couple of months after birth to the age of c. 14.5 years. The ⁸⁷Sr/⁸⁶Sr -values are for the most part inside the local range, except for from the age of c. 6 years to the age of c. 7 years (fig. 21). Individual 9 was for the most part articulated. The individual was at least 45 years old and was a male. The condition of the skeleton before death was good, compared to his age. The Total Sr-Beam (V)

(concentration) ranges from 0.99 to 1.36 on the M1, 0.47 to 0.99 on the M2, and 1.45 to 2.96 on the M3.

From birth to the age of c. 1.7 years the ⁸⁷Sr/⁸⁶Sr -values for individual no. 12 are outside of local range (fig. 22). The preservation of this individual was moderate with a fragmented skull. This individual was young, aged c. 6-8 years at time of death. The Total Sr-Beam (V) (concentration) ranges from 0.38 to 0.70.

(27)

24

The ⁸⁷Sr/⁸⁶Sr -values for individual 13 are higher at the age of c. 1.9 to 3.3 (fig. 23). This individual was scattered, and bone elements were mixed with other individuals. The individual was decapitated on the lower part of the neck. Teeth wear, and other remarks suggests the individual to be of an age of c. 10-15 years. The Total Sr-Beam (V)

(concentration) ranges from 0.54 to 1.17 on the M1 and 1.33 to 1.75 on the M2.

Generally, the M1 on individual 15 has a high local values (fig. 24). The M2’s values at c.

age: c. 3.5 to 7.5 are not local. This individual was an elderly man, at least 55 years old. The skeleton showed traces of different pathologies, which most likely was present due to the individuals age. There is a possibility that this man was decapitated. The Total Sr-Beam (V) (concentration) ranges from 0.72 to 1.54 on the M1 and 0.29 to 0.53 on the M2.

(28)

25

Most of the sample lines on individual 18 are outside of the local range (fig. 25). The age span represented extends from c. 2 to 8.2. The ⁸⁷Sr/⁸⁶Sr peaks at the age of c. 5.8. This individual was not articulated and the teeth wear and other indicators suggest the individual was at most 25 years. The individual could have been a female. The Total Sr-Beam (V) (concentration) ranges from 0.69 to 0.87.

Fig. 26. Age of enamel formation for individual 20. The local range is marked with light blue striped lines. This individual is represented with one tooth (deciduous). The age in years of enamel formation is plotted against the ⁸⁷Sr/⁸⁶Sr-values.

The age span represented for individual 20 extends from c. 0.2 to 3.9 years (fig. 26). This individual was disarticulated and the skeleton, except for the teeth were porous. The teeth eruption, and the metatarsal bone suggest that the individual was c. 5 years old. This tooth was a deciduous tooth. The Total Sr-Beam (V) (concentration) ranges from 0.80 to 1.09.

(29)

26

Individual 21 has, compared to the other individuals much lower and constant local values (fig. 27). Individual 21 consisted of a skull, teeth, and an upper extermity. The individual was about 12-15 years old. The Total Sr-Beam (V) (concentration) ranges from 1.43 to 2.00.

The age span represented for individual 23 extends from c. 0 to 5 years. The ⁸⁷Sr/⁸⁶Sr -values are higher before the age of one year (fig. 28). Individual no. 23’s skeleton was porous and dental wear gives the individual an age of about 25-35 years. The pelvic bone would,

however, suggest that the individual was c. 20-25 years old. The gender of this individual was probably a male. The Total Sr-Beam (V) (concentration) ranges from 0.57 to 1.34.

Two of the individuals (20 & 21) are inside the local range. Individual no. 3 is far outside the local range. The strontium values greatly exceed the local ratio of Öland. Individual 18 does for the most part also exceed the local strontium ratio. The rest of the individuals have values that pendulate between the local ratio and the non-local ratio (Table 1 & fig. 29).

(30)

27

Fig. 29. Column chart showing the quantity of individuals from Sandby borg who are non-locals and locals at different ages.

Ind. Nr. Non-locals Locals Change of residence Age (at death) Sex

3 x Midde aged/adult M?

4 x Midde aged M?

5 x Adult M

6 x 12-15 yrs M

7 x 12-15 yrs M

9 x 45 + yrs M

12 x 6-8 yrs ?

13 x 10-15 yrs ?

15 x 55 + yrs M

18 x 20-25 yrs F?

20 x C 5 yrs ?

21 x 12-15 yrs ?

23 x 20-25 yrs M?

Table 1. Local or non-local determination and general information.

8. Discussion

Enamel is, compared to other skeletal parts much denser and thus less pliable to chemical, physical, or biological changes caused by external alteration (Budd et al. 2000:688).

Contamination disturbs the ⁸⁷Sr/⁸⁶Sr ratio of a local geology. The biological available

strontium can, however, without problems with contamination, be extracted from local fauna (Fornander et al. 2015:188). The faunal references used in this dissertation as comparable data is quite modest. Snail shells could give ratios affected by rain- and sea water, although it has, in this case, probably not affected the ratio due to Öland’s low yearly precipitation (Fornander et al. 2015:188).

After reviving the laser ablation analyses using a microscope, certain observation could be made regarding the precision. More lines toward both the root and the crown could have been fitted on quite a few of the teeth. These lines where added for the calculation of the formation age for each line (see fig. 14). Other observations are listed in appendix no. 5.

(31)

28

The diet of these individuals is hard to determine. The nitrogen levels are quite high according to the stable isotope data (Eriksson 2019). According to data collected from cereals from Sandby borg, the nitrogen levels are high (Eklund 2019). The high levels of nitrogen could be a result of cereal fertilization or marine food intake. Studies in Great Britain show that

fertilization with seaweed may affect the ⁸⁷Sr/⁸⁶Sr ratio. The use of seaweed as a fertilizer is well documented around Europe. Fertilizing with seaweed could lead to increased δ¹³C-, δ¹⁵N-, δ³⁴S values, and elevated strontium concentrations (Blanz et al. 2019:35; Evans et al.

2012:1pp), thus affecting the dietary and mobility research. Since the diet of these individuals seems to be a mixture, there is a possibility that a combination of two different sources of food with high (e.g. plants & marine) and low concentration (e.g. terrestrial- & marine animals with higher trophic levels) were consumed. The food with the highest concentration will affect the ⁸⁷Sr/⁸⁶Sr values more than the food with lower concentration (Eriksson 2019).

The Total Sr-Beam (V) (see appendix no. 3) is not the concentration of strontium, but the beam intensity. However, the intensity is correlated with the strontium concentration and is therefore useful as a corresponding unite to the concentration levels. Unfortunately, no direct parallels between different samples can be drawn due to different tuning conditions between each sample (Schmitt 2019). If the concentration was measured in this study, more parallels to seasonal food intake could have been made.

It is difficult to discuss sources of error when most of the individuals are represented with just one tooth. In those cases, the overlap of tooth growth cannot be studied. It is also hard to determine whether the individuals were local during later or earlier stages in their lives, because many of the individuals are represented with

only one tooth. Six of the twelve individuals analyzed were children (under the age of c. 15). With pretty good certainty, at least two individuals (ind. 12 &13), who were children, were local residents during time of death. The rest of the individual’s residence at time of death cannot be surely determined, because of the lack of data from later stages of their lives before death.

The growth of the enamel is in general individual. It is not certain that the enamel growth is linear for the twelve individuals. The model (Beumount &

Montgomery 2015) used for determining the age of enamel formation is just one of many other models; for

example, Schour & Massler (1940), Smith (1991), Fanning & Brown (1971), Anderson et al.

(1976), Haavikko (1970), and Lunt & Law (1974), which are discussed in Hillson (1996). The model used for this thesis was chosen because it was easier to visualize and compare the enamel surface formation to the samples. Many of the individuals analyzed have teeth that are affected by enamel hypoplasia, which disrupt the enamel growth and the surface (Hillson 1996:165pp). The sample lines were not always placed linear to the enamel surface and growing direction, for example, on individual 4 (fig. 30), 21, and 15 (M2). This might have caused the lines to overlap each other in formation in years. To be able to work out the age of formation in years, non-existing sample lines were added to some of the teeth. These lines were added to the pictures taken of the existing sample lines and are most likely not the exact same size as the existing lines. They were added to the calculation, but not used in the results.

The lines may, however, have caused errors in the calculation of formation in years per line.

The line spacing can also vary. It depends on the curvature and surface features of the tooth.

Fig. 30. Laser ablation from individual 4’s tooth.

The Victims at Sandby Borg

The Victims at Sandby Borg

- Tracing mobility and diet using

strontium analyses

Table of contents

1. Introduction

2. Background

3. Material

4. Ethical aspects

5. Methods

7. Results

8. Discussion