The introduction of the turkey (Meleagris gallopavo) in early modern Sweden – historical and zooarchaeological evidence of husbandry and consumption

SUMMARY In this paper we describe how and why turkeys were introduced to Sweden during the 16th century, and how the bird spread to different social groups in the 17th century. We present data from unpublished financial records and provide a compilation of all archaeological findings of turkeys from the geographical area of present-day Sweden. The results show that turkeys, first imported by Duke Karl of Sweden in the 1580s, had spread to the Swedish nobility by the 1610s. During the first decades of turkey husbandry in Sweden, turkeys were items of conspicuous consumption, used to show off during elite dinners and as gifts to peers and subsequently also to subordinates. During the 17th century, the bird was adopted by the urban upper middle class. Early modern Swedish turkeys were small, and likely less affected by selective breeding when compared to modern heritage-breed turkeys.


INTRODUCTION
Turkeys (Meleagris gallopavo) were domesticated in Mexico and were brought to Europe in the early 16th century, soon after Spanish arrival in the Western hemisphere. 1 The subsequent introduction and spread of turkeys across the European continent has attracted growing attention in historical and archaeological research in recent years, with the publication of studies on their role and function in present day Hungary, the Czech Republic, Italy, Germany, France and Great Britain. 2 While some minor contributions and comments have been made in the historical and ethnographic literature, an in-depth analysis of the first introduction and spread of turkeys in Sweden that combines historical and archaeological material is lacking. 3 The earliest references to turkeys in Swedish written sources date back to the 1550s. Fragmentary examples of the consumption of turkeys in higher social circles can be found in letters and other sources from the first decades of the 17th century, with a much higher prevalence from the second half of the 17th century. 4 Thus, a glaring gap exists in our understanding of how these turkeys entered Sweden during the late 16th and early 17th century. In the archaeological material, turkeys have been identified in urban settings from 1612 onwards, where they have been considered to have been introduced by immigrant merchants or artisans, an interpretation first made in connection to a zooarchaeological analysis of food wasteincluding two turkey bonesthat were found in contexts associated with German immigrants in 17th century Norrk€ oping. 5 This interpretation has subsequently been repeated to a point of becoming an accepted truth. Early archaeological finds of turkeys in Sweden have thereafter habitually been understood in relation to the ethnicity of their assumed consumers. 6 In the following, turkeys will be interpreted from two other perspectives. On one hand they were an exotic new addition to the ducal and royal farms and dinner tablesan element in lavish banquets or ceremonies, suitable to be served to emissaries and continental visitors, or to be given as gifts. 7 Given their rarity, novelty, and association with the societal elite, turkeys became a good for conspicuous consumption. This term, coined by Thorstein Veblen, refers to the consumption of wastefully luxurious goods for the sake of demonstrating status, a kind of consumption that became increasingly common during the early modern era. 8 On the other hand, turkeys were also a new type of fowl, living creatures that looked, behaved, and tasted differently from the familiar birds. They had to be handled and raised in new ways. Moreover, they could be prepared and served in other ways than for example chickens (Gallus gallus domesticus) or geese (Anser anser domesticus). The raising of turkeys posed other challenges when compared to the rearing of other fowl. As will be seen, this affected the spread of turkeys geographically across Sweden and to different social environments.

TURKEY HUSBANDRY IN SWEDEN
An oft-repeated opinion in older Swedish literature on turkey farming is that turkeys were more difficult to rear when compared to other domestic birds. This has to do with the fact that turkey poults were notoriously hard to raise, with a high mortality rate. In addition, turkeys need more food when compared to chickens, and were therefore expensive to keep if they were not able to roam free and forage for food. They need more space than other birds, grow sick and lose weight if they are kept in cramped spaces, and should be kept separate from other poultry. 9 In England, the birds were also known for their troublesome behaviour, rendering them harder to manage. 10 The motivation for raising turkeys, despite their somewhat bad reputation, is their meat, which has been repeatedly praised over the centuries. 11 The sensitive nature of turkey poults has given rise to many and detailed recommendations concerning their care. Advice was apparently neededthe mortality rate for young turkeys was often very high at the Swedish crown estates investigated as part of this study. Consistent long-term growth of existing flocks was often hard to achieve.
Turkeys mate in early spring, with hens laying up to 30 eggs per clutch in March or April, and they subsequently brood for 4 weeks. The first poults hatch by the end of April. The poults must be protected from rain, cold and damp grass, and they need special fodder to survive and grow. 12 The recommended diet for newly hatched turkeys remained the same for centuries, including hard boiled eggs, dairy products, chives, onions and nettles. 13 Turkeys are born ready to leave the nest and forage. They are imprinted on their mother soon after birth. If the mother is not present, the poult will imprint on the nearest carer, be it a chicken hen or a person, and treat the foster parent as if it was a turkey. 14 This trait can be used to establish new flocks, where a farmer can allow chickens to incubate turkey eggswhich can be kept for about 14 days before they spoiland later raise the poults. 15 King Christian IV (r. 1588-1648) of Denmark had specifically designed ovens built at Frederiksborg Castle to hatch eggs from turkeys, mallards and swans. Eggs were placed on top of a pillow in a basket, carefully heated in the oven and regularly turned by designated handlers. 16 The most rational way to keep turkeys, which also agrees best with the behaviour and needs of wild turkeys, is to keep the birds on pasture. Wild turkeys range over large areas, and even turkey poults can move more than 1,000 meters on a daily basis. 17 In their natural habitat, wild turkeys form flocks with a dominant male, younger males, hens and poults with a clear pecking order. In a domesticated setting, a flock with one cock and at least four to six hens, would be considered viable and could yield more than 80 poults per year. When the poults reach sexual maturity, it is recommended to separate males from females or cull some of the young males to reduce conflict within the flock. 18 Modern heritage-breed turkeys take about seven months to grow ready for slaughter, typically achieving their maximum weight in mid-November. 19 In Sweden the slaughter historically started soon thereafter, peaked in December and continued until early January. Turkey poults were also slaughtered and sold in spring or early summer after reaching a size comparable to a small chicken hen. 20

REGIONAL FINANCIAL ACCOUNTS
A new understanding of the role of turkeys in Sweden around the turn of the 16th century can be gained from information gathered from regional financial accounts (landskapshandlingar). Before the advent of a modern administration in the 1630s based on collegial responsibilities and new methods of assembling and collecting information, the Swedish financial administration was built on a system where local stewards reported directly to the king. The structure was implemented in the 1530s by King Gustav I (r. 1523-1560), the first Swedish monarch of the Vasa dynasty. It remained in place relatively unchanged until the new system was introduced some 100 years later. 21 The recordstoday contained in more than 8,000 volumesconsist mainly of accounts of local tax collection, as reported by bailiffs with responsibilities for mainly rural administrative districts. Most important in the present context, they also include information of the various activities at crown estates and castles, including detailed information on agricultural production and animal husbandry, as well as distribution and consumption records. 22 While far from perfectly preserved, the records are still a crucial source that can cast light on the function and financial underpinnings of the early modern Swedish state.

CUSTOMS ACCOUNTS
In 1622, a new domestic customs system (lilla tullen) was implemented in Sweden, which introduced tariffs on most goods, including food and livestock, that were brought into towns from the countryside. In order to enforce the new regulation, fences were constructed around all towns, and the traffic was led through town gates. 23 The information generated by the new system, as recorded in customs ledgers, provide detailed records of most items reported by local farmers, fishermen and others, for the purpose of sale. They also contain information about individuals who resided in the towns and imported goods for their own consumption. For the present purpose the records have been searched for information about the import of turkeys, in order to understand the extent of the spread of the fowl in Swedish society and in an urban context.
Overall, the customs ledgers have rarely survived intact. In Stockholm, records have only been preserved for a few years during the first fifteen years after the implementation of the new system, and then, only from a couple of the town's six customs stations. 24 One of the best-preserved ledgers comes from the year 1624 for the customs station at Blockhusudden, which channelled all traffic that arrived to Stockholm by boat from the east, including all northern Swedish harbours. Another well preserved ledger comes from 1635 for Kapellporten, which handled traffic that arrived from the southern hinterland.

ZOOARCHAEOLOGICAL MATERIAL AND METHODS
Post-medieval archaeology is a growing area of research in Sweden. To a large extent this is due to legislation, which protects all archaeological remains pre-dating 1850. 25 While rural remains are included in the legal protection, a majority of the contract archaeological excavations affecting remains from the last 500 years are carried out in towns. Thus, while the number of identified bones from turkeys is continuously growing, the findings are most likely unevenly dispersed when compared to the original distribution. This becomes obvious when compiling all findings of turkey bones, including both previously published and non-published findings (Table 1). 26 The material is dominated by three towns in central Sweden: Norrk€ oping and Kalmar in the east, and Gothenburg situated on the west coast. Largescale excavations of post-medieval remains have been carried out in all of these towns. All but two of the sites discussed in this article are urban, with the two remaining sites consisting of a vicarage and a manor.
The identification of turkey bones is rather straightforward in a Swedish context. The main possible source of confusion is the capercaillie (Tetrao urogallus), but turkeys and capercaillies can be separated with the aid of modern reference specimens and the key provided by Tomek & Boche nski. 27 Turkeys, like several other species of Galliformes, exhibit large sexual dimorphism, and the male turkey has spurs that grow with age. 28 Thus, both the size of the bones and the presence or absence of spurs can be used as sexing criteria. In addition, just before and during the period of lay, hens produce medullary bone that can be used to sex the bones found on archaeological sites, but also to indicate the season of slaughter. 29 In general, the bird bones in the zooarchaeological reports have been labelled as either adult or subadult, only differentiating between fully mature and immature bones. In two of the assemblages, the bones have also been categorised as juvenile and pullus, however without any account of how the categorisations have been made. In some cases, no age estimations have been given for the bones.
In order to compare bone measurements from different sites and skeletal elements, the measurements have been transformed into logarithm size index (LSI) values, using a natural logarithm with the base e and the formula LSI ¼ ln(x/m), where x is the archaeological measurement and m the corresponding measurement of a standard individual. 30 In this case, the standard used for the calculation was a small female domestic turkey. 31 Length and breadth measurements were separated, since the former corresponds to body height and the latter is more influenced by body mass. 32 Length measurements (greatest length, GL) from Swedish sites (including several measurements from one female and from one male) were compared to those from early modern sites in present day Poland, the Czech Republic and Hungary as well as from a male wild turkey. 33 Only bones from adult birds were included.     . While we again lack information about when and from where the turkeys arrived, the records show that both birds were transferred to the ducal seat at Nyk€ oping Castle in S€ odermanland later that same year. Importantly, they were sent to the castle garden and not to the castle farms where the normal livestock was kept. The distinction is important. In the garden they were more akin to exotic objects, meant to be viewed and admired, rather than eaten. 35 The two turkeys that had been sent to the castle garden in 1572 were not alive when four new turkeys arrived in Nyk€ oping in 1583. The new introduction was to become the most significant, as it led up to the first recorded instances of successful breeding of turkeys in Sweden. 36 The new turkeys, arrived in Nyk€ oping as a result of a continental trip, on which Duke Karl embarked in the autumn of 1582. The destination was Heidelberg, the seat of his father-in-law, the Elector Palatine Louis VI, where the duke and his wife stayed for some six months, before returning to Sweden the following autumn. 37 The financial accounts from Nyk€ oping Castle show that a range of fowl were introduced that same year, explicitly denoted as having arrived from Germany. The new additions consisted of German ducks and German and Spanish chickens, but in the present context most importantly, the four abovementioned turkeys. It is possible that the birds were acquired in Heidelberg, but given the distances involved, it is more likely that they were purchased in northern Germany on the return voyage. 38 Although clearly a new source of protein, both in terms of meat and eggs, the turkeys were also the first in a range of more visually spectacular, exotic animals that arrived in Nyk€ oping between 1583 and 1587. These included turtle doves (Streptopelia turtur), grey partridges (Perdix perdix) and peacocks (Pavo cristatus). A particular unusual addition, and a true rarity in Europe, were guinea pigs (Cavia porcellus), two of which arrived in 1585, and were soon bred successfully. 39 Most of the new animals did not survive for long. By the 1590s, only the turkeys and a small number of peacocks remained. The guinea pigs, turtle doves and partridges all died (or escaped) in 1588. While all of the German ducks were lost only a year after their arrival, the German chickens and Spanish ducks and chickens that arrived with the turkeys bred successfully. The turkeys themselves survived and bred sufficiently between 1584 and 1587, with much more substantial growth of the flock from 1588 onwards (Fig. 1). The new growth would lead to one of the most important aspects of the turkey breeding operation in Nyk€ oping, namely allowing its spread across much of central Sweden. 40

THE DISPERSAL OF THE FLOCK
With a rapidly increasing turkey population in Nyk€ oping, Duke Karl in 1589 gave orders to send turkeys to other parts of his duchy. Twenty-six turkeys in total were ordered to be distributed from Nyk€ oping that year, with 23 assigned to € Orebro Castle and a further three to Eskilstuna Castle in northwestern S€ odermanland (Fig. 2). 41 The available records show that six turkeys reached € Orebro and four reached Eskilstuna. Five of the remaining sixteen birds ended up at H€ ojentorp in central V€ asterg€ otland in 1589, and another two at Vibyholm the following year. 42 A distribution to the parts of the duchy closest to Nyk€ oping Castle happened in 1591 and 1592, when turkeys were sent northwest to a range of crown estates and castles in S€ odermanland. The records show that three turkeys were assigned to Gripsholm Castle, five to R€ avsn€ as, five to Tynnels€ o and six to Julita. However, the intention and the actual result of the distribution diverged. The incoming records from the respective sites reveal that six turkeys arrived in Julita and ten in Tynnels€ o, but no birds arrived in either R€ avsn€ as or Gripsholm. The three final turkeys out of the nineteen originally distributed from Nyk€ oping instead ended up at either Sken€ as or Brån€ as, two crown estates in € Osterg€ otland, south of Nyk€ oping. 43 The establishment of several successful breeding flocks across the duchy eventually resulted in surplus turkeys being sent back to Nyk€ oping for consumption at the court. Forty turkeys arrived in Nyk€ oping in this manner between 1590 and 1594, from Tynnels€ o, Vibyholm, € Orebro, H€ ojentorp and Sken€ as. 44 The growing regional flocks also led to a redistribution of turkeys between the various castles and estates that did not involve the original flock in Nyk€ oping. € Orebro most notably provided the five turkeys which, in 1595, created a new flock in Segersj€ o, the other key crown estate in N€ arke, two of which were sent back to € Orebro the following year, after The various estates in S€ odermanland that received turkeys during the dispersal phase in the 1590s retained small but stable flocks throughout the first decade of the 17th century, with a number of flocks surviving also into the reign of King Gustav Adolf (r. 1611-1632) (Fig. 3).
Few of the more peripheral flocks survived very long, however. This was most notably the case for the flocks in V€ asterg€ otland and Småland. V€ asterg€ otland had been incorporated into the turkey distribution area as early as 1589, when turkeys were sent from Nyk€ oping to H€ ojentorp. The flock reproduced well, and in the following year, it grew to eighteen turkeys. This did not last. The flock disappeared sometime before the turn of the century. A new flock was established in 1603, when a number of new turkeys were sent from Nyk€ oping. 47 Successful breeding ensued again, and no less than 100 poults were added between 1604 and 1607. The surplus was used to create a new flock at nearby Skaraborg Castle. The H€ ojentorp flock and the new flock in Skaraborg were both gone by 1609, however, after the death of all turkeys at both sites within little more than a year. 48 The distribution to Småland appears to have been made by way of V€ asterg€ otland in 1603, when both J€ onk€ oping and Kalmar received turkeys. 49 The following year, six birds from Kalmar were sent by boat to Ottenby on the island of € Oland. The Kalmar flock disappeared in 1608, and the two final turkeys in J€ onk€ oping were pillaged by Danish troops in 1611. Only the Ottenby flock survived into the 1610s. Despite being reared on an island that essentially functioned as a large-scale game reserve, the flock never developed beyond ten animals (Fig. 4). 50 While most of the flocks in V€ asterg€ otland and Småland were gone by the 1610s, it is still probable that the earlier dispersion facilitated the spread of wider turkey farming in the regions. This was almost certainly so in V€ asterg€ otland, which had a denser population, more agricultural land, and, like S€ odermanland, a higher concentration of wealthy aristocratic landowners, many of whom developed an interest in turkey farming. It is less likely that turkey farming was established at many manors in Småland, a more forested region which lay on the outskirts of the Swedish Realm. 51 A fairly uncertain issue concerns the keeping of turkeys in Stockholm before 1620. Turkeys were definitely sent to and consumed at the Royal Palace in Stockholm. Twenty-four turkeys were sent from Nyk€ oping in 1590 and 1591, and according to a note from Stockholm, turkeys were recorded in the slaughterhouse at the palace in 1591. 52 The lack of records for the manors supplying Stockholm Castle for this period makes it impossible to determine if this was for consumption at the royal court or to set up a breeding population. Fourteen new turkeys were sent from Nyk€ oping in 1604 and another four in 1606, explicitly intended for court consumption. No fewer than 46 turkeys were sent to the kitchen at the Royal Palace from Kungs€ or in 1614. 53 While this shows that turkeys were eaten in Stockholm, no records are available to confirm that the home farms supplying Stockholm with fresh meat actually housed turkey flocks before 1620. 54

PREDATORIAL ATTACKS AND MORTALITY RATE
The records do not consistently mention the sexes of the turkeys, which makes it hard to say anything definite about the strategies behind the choice of animals distributed from Nyk€ oping to other crown estates. Lack of space, which might have affected the decision to distribute the flock initially, was no major concern over time, after the home farm system in Nyk€ oping was expanded, from two to three to eventually five home farms by 1610. 55 A definite reason for the dispersal was greater security, where different flocks made the system less vulnerable to both external and internal threats or risks. In short, with a more spread-out operation, unforeseen events could not threaten the entire population. The records make it very clear that various incidents affected the flocks. This included predatorial attacks. Turkeys were killed by hunting dogs at Vibyholm in 1592, at Kungs€ or in 1618, and at Nyk€ oping in 1619 and 1620. Foxes killed turkeys in Nyk€ oping in 1604 and 1619, when ravens also killed four young turkeys. Numerous examples of attacks on other livestock and fowl can be found in the sources, involving foxes, wolves and bears. More surprising issues also arose. No fewer than seven turkeys were stolen by foreign sailors at Liljestad in € Osterg€ otland in 1591, whereas the two remaining turkeys at J€ onk€ oping Castle in Småland were taken in 1611 by occupying Danish troops during the Kalmar War. 56 Yet more important were natural mortality ratesin particular among young turkeysthat affected the various populations during the period. With the exception of Nyk€ oping (Fig. 1), high poult and young turkey mortality was the leading reason for the depletion of the flocks over time (Fig. 5).

PATTERNS OF CONSUMPTION
The high mortality rate impaired the intention to create local turkey populations that could be made sufficiently stable to provide surplus animals for food, to be served when the duke stayed at the various estates or castles. Since medieval times, the crown estates served as residences where the king could stay when travelling around his domains. When Duke Karl took control of the estates and castles in his duchy in 1568, this function did not change. Rather, it intensified with his very frequent travels across his lands. 57 The sites served as residences not only for the duke and his retinue, but also for foreign and domestic dignitaries who travelled through Sweden on official business. This also meant that exclusive food was required on a regular basis. One of the more interesting examples of how turkeys were offered to prestigious visitors happened in the autumn of 1607, in connection to an overland voyage that a Persian ambassador made from Stockholm to € Alvsborg, located on the Swedish west coast. During the trip, the party stayed at Kungs€ or in the middle of October; at € Orebro Castle toward the end of the month; and at H€ ojentorp, V€ asterg€ otland, in early November, before arriving in € Alvsborg later that same month. The majority of the food offered to the members of the embassy at Kungs€ or, € Orebro, and H€ ojentorp was not markedly exotic, but rather of similar character, with good quality white bread, cheese and butter, finer quality beer, and small quantities of fish. Fresh lamb and chicken were served at all of the castles, and moreover, one turkey was served at € Orebro and two turkeys at H€ ojentorp. While turkeys undoubtedly were exotic additions, both € Orebro and H€ ojentorp then held sizable flocks, which meant that the offering made a relatively limited impact on the integrity of the breeding populations. 58 The situation changed markedly when the embassy reached € Alvsborg. During a stay which extended until April 1608, the embassy retinue was treated to much more luxury than before. This included vast quantities of fresh meat, including beef, veal, chicken, goose, sheep and lamb, as well as hare, wild fowl and roe deer. Most important for our discussion, five turkeys were served for the guests' stay. Furthermore, a much wider selection of wine and beer and large quantities of distilled spirits were made available, and the seaside location meant that 26 lobsters and good quantities of fresh fish could be obtained. To this could be added a long list of spices from the Far East, as well as various products sourced from the Mediterranean Basin. 59 The luxurious fare can be explained in part by an order sent from the king in Stockholm to the governor at € Alvsborg Castle, with orders to ensure that the embassy would be treated to the best possible cuisine that could be found locally, with as much wine, spirits and food as they would require. 60 That this in the end would include turkeys was the logical result of the fact that € Alvsborg Castle had harboured a flock of turkeys since the second half of the 1590s. Successful breeding saw the flock rise from three turkeys in 1600 to nineteen in 1607, despite relatively regular consumption by high profile individuals who passed through the area. € Alvsborg and the area around it were geographically unique in the Swedish early modern state, as the only part of Sweden with access to the North Sea. This made it an important place for commodity trade, but also a logical first or last point of contact for prominent people travelling between Western Europe and Sweden. In the years before the Persian embassy, turkeys had already been offered to French, Russian, and Danish diplomats and officials who visited or passed through the area. 61 Numerous other prominent guests who passed through the ducal and royal lands were also offered turkey for consumption. This not least included other members of the royal family, but also various foreign dignitaries. 62 One of the major consumers was Karl's nephew Duke Johan, the son of King Johan III. He ate turkeys in Nyk€ oping in 1605, 1606, 1610 and 1611; Gripsholm in 1605; Eskilstuna in 1606; € Orebro in 1605 and 1607; Ottenby in 1607; and Kungs€ or in 1614. Duke Johan's motherand Karl's sister-inlaw, the Queen Dowager Gunilla Bielkekept turkey flocks at her estate in € Osterg€ otland, already in the 1590s, as did Karl's elder sister Sofia at her estate in Uppland. 63 While they occasionally consumed birds from their flocks, turkeys were also served at more spectacular occasions. This practice was pioneered by Duke Karl, during his second wedding in 1592, when turkeys were served at Nyk€ oping. For his coronation in 1606, fourteen turkeys were sent from Nyk€ oping to Uppsala. The tradition was kept up by King Gustav Adolf, who had 40 turkeys sent from Kungs€ or to Uppsala in 1617, for his coronation ceremony. 64 Gustav Adolf also served turkeys on other occasions, such as at a banquet in 1613, and at a feast in Stockholm in 1623. 65 Turkey flocks at the crown estates were also a source of exotic, exclusive gifts, offered not only to high profile visitors or ducal or royal consumers, but also to aristocrats, and other officials. Prominent examples of gift-giving date to as early as 1589, when Duke Karl offered two turkeys to Jost Kurzel, one of his most trusted officials, who served as governor of N€ arke, and resided at an estate southeast of € Orebro. 66 In 1597, one turkey hen and one cock were gifted to Hans von Massenbach, who served in a leading role at Karl's court in Nyk€ oping, often residing at an estate northwest of Nyk€ oping. 67 In 1614, the Chancellor of the Realm, Axel Oxenstierna, requested two turkeys from Kungs€ or, for his estate at Fiholm. By this point, the effect of turkey breeding was already established in his close family circle, with evidence of turkeys having been offered at the weddings of his brothers and sisters in 1611 and 1613. 68 Usually, when turkeys were consumed at the ducal table, the main consumers were the duke and duchess themselves. Yet, the dinners also involved high-ranking members of his retinue. This type of broader consumption was also allowed when the duke or other royal guests were absent. Information from € Orebro Castle shows that, in 1596, four turkeys were served at the lord's table (essentially the level  below the ducal table), suggesting the governor and high-ranking guests as the primary consumers. By 1602, turkeys were served at a lower level still, when four turkeys were consumed by bailiffs and more general staff. The tendency continued, with six turkeys included in mid-level official and staff general consumption in 1607. At € Alvsborg Castle in 1606 and 1607, nine turkeys were consumed at the lord's table, again most probably by the governor and his retinue. 69 TOWARD A BROADER DEMAND AND SUPPLY OF TURKEYS An analysis of the customs accounts from the Stockholm town gate and customs point at Blockhusudden shows that at least 21 turkeys were imported in 1624. 70 As the gate was only one of six similar points of entry in Stockholm, the total number of imported turkeys was most likely considerably higher. 71 Five of the turkeys were imported by individuals who formed part of the resident urban middle class, and the remaining sixteen turkeys were imported by the Dutch captain and artillery expert Jakob Trello. 72 In 1635, the customs accounts from a different Stockholm town gate -Kapellportenregistered 45 turkeys imported from the southern hinterland. Two seasonal peaks in the imports can be identified, one in late spring/early summer and a second in December (Fig. 6). 73 Although a higher number of turkeys were imported in 1635 than in 1624, it was still a relatively limited amount, when compared to other birds, both wild and domestic, that passed through the southern gate. A number of turkeys brought through the Kapellporten customs point in April appear mainly to have been imported by common farmers. Of the remaining importers a majority were very high-ranking and influential individuals, including three aristocratic Councillors of the Realm. 74 Neither of the ledgers reveal the further fate of the turkeys. It is not known if the imported turkeys were then sold at the town market or consumed by the importers. Aristocrats were principally allowed to import goods for household consumption free from customs, which suggests that the registered turkeys were meant for some type of extraordinary consumption. Such imports probably expanded during the 1630s, after it was stipulated that the state administration was to be permanently based in Stockholma requirement that created a need for state servants to have residences in Stockholm to be able to partake in government work. The development led to a phase of intensive palace building, which meant that elite social settings and practices became transformed in the capital, with likely new demand for high-end foodstuff. 75 It can be assumed that the three resident burghers who imported turkeys in 1624 also did so for household consumption or for special occasions. This was however not the case for Jakob Trello. Although Trello had a Dutch background, he did not import the turkeys from abroad. It is also not likely that he brought sixteen turkeys into town for his own consumption. Most likely he instead brought the turkeys into town in order to sell them. It is even possible that he did so from his own farms. Unlike the other three importers, Trello did in fact not reside in Stockholm. He lived on a small estate, located some way north of Stockholm, that had been gifted to him by the crown five years earlier. 76 The actual sale of turkeys in the towns was a well-regulated process. A regulation from 1623 stipulated that farmers who arrived in town with goods to sell, after declaring the goods and paying the tax, had to sell the goods in a designated house to the town mongers. The prices for different sorts of birds were set, with turkeys being the most expensive, at more than three times the price of a goose. 77 The customs account suggests how turkeys, if not consumed by all, still were available in large enough numbers to enable different groups to acquire them, both for personal consumption and for sale. The information from Stockholm indicates that breeding flocks were present in the areas around the Swedish capital by the 1620s. It is unlikely that all of the importers kept turkey flocks, however. The import of turkeys by people living in direct proximity to the town centre, in particular, raises the possibility that some of the importers were individuals who themselves had bought the turkeys to sell them for profit in town.
Further indirect evidence for this practice can be seen in the information from Svartsj€ o Castle, a crown estate located on one of the islands west of Stockholm. The establishment of a flock at the site is interesting in its own right. With a high degree of certainty, the Svartsj€ o flock's establishment can be connected to an importation of turkeys and other fowl that was undertaken from northern Germany in 1622, indicating how turkeys began to enter Sweden through new channels. 78 The first information about the flock from 1623, when 40 adults and 89 poults were registered, documents a flock far larger than the quantity of imported turkeys from 1622. Moreover, the mixed age structure reported for 1623 entails that additional birds must have been sourced from existing Swedish flocks. 79 The successful breeding and the subsequent delivery of surplus turkeys for courtly consumption shows that the birds were readily available to the Royal Court. What primarily makes the Svartsj€ o flock interesting in the present context, however, was not how it was formed, but rather, the subsequent record from 1626, indicating that no fewer than 61 surplus turkeys were sold off from its rapidly expanding numbers. 80 Thus, the historical records reveal that from the 1620s, turkeys were also sold and consumed in Swedish towns. From this point in time, the archaeological material sheds light on how the turkey spread to different social groups in urban settings.

ARCHAEOLOGICAL FINDINGS OF TURKEYS IN SWEDEN
The oldest archaeological find of turkey remains in present-day Sweden is a coracoid bone from an adult bird, occurring in the fill of a well in the town of Ny Varberg (Fig. 7). The well was most likely filled when Swedish troops lay siege to the Danish town in February 1612, after which Ny Varberg was burned and abandoned. Household waste and valuables alike were dumped in the well, reflecting the chaos of the siege. The finds show that at least one of the wealthy households in the town had access to turkeys in the early 1610s, and that the same household consumed shellfish, including lobsters, crabs, and mussels. 81 Shellfish were mainly eaten by the elite at this time, which is also suggested in the regional accounts from € Alvsborg Castle, north of Ny Varberg. 82 Here, mussels, lobsters, and oysters were reserved for members of the royal family and important guests during the first decade of the 17th century. This indicates that the Ny Varberg household both had the means and connections to partake in the latest culinary trends amongst the elite. Since Ny Varberg was part of Denmark, the turkey would not have reached the household via Swedish networks, but through Danish ones. Turkeys were imported into Denmark at about the same time as in Sweden. King Frederik II of Denmark (r. 1559-1588) kept turkeys at Frederiksborg Castle on Zealand, and there are historical records of turkeys being kept at Danish crown estates in the 1570s and 80s. 83 The archaeological evidence, however, suggests that turkeys were imported to Denmark even earlier than that. Dating at the latest to 1550, the backfill of the moat surrounding Koldinghus Castle, Jutland, included three turkey bones. 84 A turkey bone has also been found in a cesspit dated to the late 16th or early 17th century in the Danish town of Elsinore. 85 In this context it is noteworthy that no turkey bones were found during extensive excavations of the Swedish town of Nya L€ od€ ose (1473-1624), situated just upstream from € Alvsborg Castle. The trading town housed a multi-ethnic population, including many wealthy burghers with connections to the castle. 86 Turkeys were kept at the castle between 1599 and 1610, but there is no evidence that they spread to Nya L€ od€ ose. 87 Nor have turkeys been identified in any other contemporary Swedish town, suggesting that they still had not spread outside the uppermost elite by the 1610s in Sweden, in contrast to the situation in Denmark.
Archaeological findings of turkeys only turn up from Swedish towns with the first datings from the late 1620s. This is consistent with data from the historical records. Some of the oldest findings derive from the town of Norrk€ oping. The wealthy entrepreneur Louis de Geer started building a palace in Norrk€ oping in 1627. During archaeological excavations of the palace, two turkey bones were found in construction layers of a kitchen building. 88 These findings have previously been interpreted in terms of de Geer's Dutch origin, 89 but could as well be connected to the economic and social status of the household. Two bones dated to the mid-17th century have been found in another kind of social setting in Norrk€ oping, namely in a city block inhabited by German immigrants who worked at a nearby brass factory. 90 Although a heterogenous group, these workers were generally highly-skilled and well-educated, and the archaeological material shows that their social status was probably equivalent to that of artisans and merchants. 91 Another early find originates from a municipal slaughter house in J€ onk€ oping, dated to c. 1642-1662. 92 As we have seen above, there is historical evidence of turkeys being imported to Stockholm already in the 1620s. Only one archaeological excavation in the Swedish capital has so far rendered turkey remains, however. 93 During the late 17th and the early 18th century, turkeys became more and more common in urban settings. In Kalmar turkey bones have been found in features dated from the 1670s to the late 18th century. Here, the identities of the plot owners are known from the early 18th century onwards, and these were a mix of relatively unassuming crafters, low-level officials, widows and sailors, but also a couple of merchants. 94 This shows that, by the early 18th century, turkeys were consumed in households from the lower to middle spectrum of the urban social ladder. Analyses of the material culture and food remains from the same plots show that while the overall consumption in the city parcels was not sumptuous in any way, individuals took the opportunity to buy and consume highranking foods and objects when available. The residents were active participants in the new consumption habits developing during the period, connected to ideals of modernity. 95 Excavations in Norrk€ oping and Kalmar provide a hint of the relative importance of turkeys, in comparison to other domestic birds, such as chickens and geese. Although turkeys appeared in different social settings during the 18th century in these towns, the relative importance of turkeys was insignificant. Bones from turkeys constitute only three percent of the identified bird bones from domestic birds found in one town plot in Kalmar. 96 In the city parcel Gubben in Norrk€ oping, turkeys constituted five percent of the bird bones during the years 1715 to 1768. 97 Backfill in a well in Gubben, dated to the 1760s and likely belonging to the household of Baron Johan Adelsw€ ard, contained 20 chicken hens, one chicken cock and one turkey hen. The birds were not butchered, and probably instead died in a fire which in turn suggests that they constituted the total number of birds kept in the household at the time. 98 Overall, the zooarchaeological material from Norrk€ oping and Kalmar is dominated by adult, or near-adult turkeys (Table 1). It is likely that turkeys were not raised inside these towns, but rather bought at the town market in late autumn or December, when the birds had reached their maximum slaughter weight. In Gothenburg younger birds dominate; a full 31% of the bones come from subadults, juveniles or (in one case) a hatchling. The large proportion of young birds in Gothenburg, indicates that turkeys might have been raised by the town residents. 99 Nevertheless, the presence of juveniles may not necessarily indicate turkey husbandry. The domestic customs accounts from Stockholm show that birds also were imported to the town in late spring. These birds were probably poults culled from the flock, sold for steaks to be consumed by wealthy residents. 100 All but two of the sites with identified turkey bones are urban. The two remaining sites consist of one vicarage and one manor. One of the findings originates from Vapn€ o manor, western Sweden. During the 17th century the manor was owned by various Danish andlater -Swedish wealthy men and women, including several members of the aristocratic Stenbock family. The main building was destroyed in a fire in 1733 and later rebuilt. 101 Six bones from one turkey were found in the moat backfill, which also contained demolition debris from the burnt building. The ceramics date to the 17th/18th century. From this archaeological context it is most likely that the turkey bones predate the fire in 1733. 102

THE SIZE OF THE TURKEYS
In some rare cases, the regional accounts mention carcass weights for animals consumed at the crown estates, including turkeys (Table 2). 103 The weights were not actual weights, but represent a mean approximated carcass weight, adopted as a standard for that particular place and in that particular year, according to the account keeper. 104 The fact that these weights vary considerably between different farms and years confirm that new estimates were made every accounting year. Table 2 shows that the turkeys kept at the crown estates in the early 17th century were remarkably small. 105 The approximate eviscerated percentage loss in a modern market turkey is about 20 per cent. 106 The carcass weights converted to live weights using the same weight loss results in live weights from c. 0.8-2.1 kilograms. As a comparison, the Italian agronomist Agostino Gallo, writing in the late 16th century, claimed that his turkey cocks grew to almost 11 kilograms. The live weight for modern wild turkeys is about 4-4.5 kilograms for hens and up to 9-10 kilograms for cocks. 107 Apart from the low weight, it is also interesting to note the large variation in size between the different crown estates. While domestic chickens and geese were about the same size at all the estates, turkeys kept at € Alvsborg Castle were less than half the size of turkeys kept at Nyk€ oping Castle in 1603.
The peak in weight in Nyk€ oping 1603 coincided with an unusually large addition of new poults and a low mortality rate (Fig. 1). The following year saw the opposite conditionsa high mortality rate and a relatively small number of new poults. During the following two years, the number of new poults continued to decline, and it was not until 1607 that the surviving poults increased in numbers again. The birds, however, were still a lot smaller compared to 1603. The chickens and geese also decreased in size between 1603 and 1607. While it is impossible to know the reason why the turkey flock suddenly decreased in number, with all the fowl reduced in weight, the change took place the year after a plague hit the country in 1603, and one of four of the employees hired to care for the animals in Nyk€ oping died. 108 This must have entailed a loss of competence and experience amongst the staff, perhaps affecting the overall success of the fowl rearing.
Measurements obtained from the archaeological finds provide a somewhat different picture of the overall size of the birds. Length measurements of the Swedish turkeys were compared to previously published data of archaeological turkey findings from present day Poland, the Czech Republic and Hungary, as well as a modern wild turkey cock  (Fig. 8). The Swedish measurements all derive from bones dated to the late 17th and the 18th century, with only one exceptionthe bone from Ny Varberg, also included in the graph, which is dated to 1612. While the sample is too small to draw any comprehensive conclusions from, a few patterns emerge. First, the measurements from Sweden cluster in two discrete groups, likely corresponding to females and males. Secondly, the Swedish male turkeys were about the same size as turkeys found in continental Europe. On the other hand, turkey hens from Sweden were smaller. It is, however, not clear whether the measurements from Poland and the Czech Republic include both females and males, and the difference might therefore be caused by a lack of females in the continental samples. Thirdly, the proportion between upper and lower limb size in Swedish early modern turkeys and the modern standard animal differs. Bones from the wing were larger compared to the standard animal, while bones from the lower limb were closer to the standard in size.

DISCUSSION
The turkey was the only exotic domestic animal to be successfully introduced in 16th century Sweden, mostly as a result of the efforts of Duke Karl of Sweden. 109 Even though the duke imported many different species to his ducal seat at Nyk€ oping Castle, only turkeys bred well enough to allow for new flocks to be established. By the mid-16th century, turkeys had spread to many parts of Europe, including Britain, France and Germany. A range of Italian towns also introduced the bird soon after its arrival in Europe. By the 1570s large flocks were held in areas in present day western Germany. During the second half of the century turkeys were also introduced to Denmark, to the areas which now constitute the Czech Republic, and as shown, to Sweden. 110 Turkey breeding was first mainly associated with the uppermost elite, and many of Duke Karl's European peers kept turkeys at their manors and castles. 111 The establishment of courtly menageries with exotic or unusual animals was a powerful statement, that not only showcased wealth and luxury, but also served to influence artists and the visual representation of the unusual. 112 The introduction of exotic species, but also the subsequent establishment of breeding flocks by Duke Karl fitted well with his attempt to develop his duchy into something more akin to a self-sufficient princely state modelled on European prototypes, but also his keen interest in most things new. During the same period when the exotic species were introduced to Nyk€ oping, he set out to develop international trade, which later would include the first Swedish attempts to send ships to the Western Hemisphere. While nothing came out of the latter endeavour, he corresponded with individuals who supplied him with information on global trade and supported the introduction of new products and goods to Sweden. 113 Turkeys in this context served as visual and symbolic centrepieces that underscored Duke Karl's ambition to function as a European prince with global ambitions.
The turkeys kept at the Swedish crown estates must have been an unfamiliar sight, when compared to all other domestic and wild birds, with an exotic body form and plumage. 114 Not only did the turkeys look different, but they behaved and sounded different, with their distinctive gobbling. The first turkeys to arrive in England, Italy and France were kept as exotic curiosities or pets, acting as symbols of their owners' wealth and connections. 115 Soon the turkeys were also served at the dinner tables of the rich. The fast integration of the turkey in European cookery, especially in comparison with other New World foods, has been attributed to the similarity turkeys shared with existing fowl. 116 By the 1580s, turkeys were deliberately bred to provide meat for the dinners of Duke Karl and his high-ranking guests. During the 1590s and 1600s, they were served at royal weddings and coronations. Throughout the first decades of Swedish breeding, turkeys were never eaten in large quantities. Instead, one or a small number of birds were enough to transform the meal into something special. 117 This was part of a more radical change in the way in which elites demonstrated their status through the consumption of food. While quantity had been the primary means to show off during the Middle Ages, quality, refinement, and elaborate dishes became the focus of elite food consumption in the early modern era (Fig.  9). 118 Already during the first half of the 16th century, it became increasingly important to mark social boundaries in food consumption. This was expressed through different menus, and tables, assigned to different socioeconomical groups employed at the Swedish crown estates. This trend was even more prominent at the ducal and royal courts in Sweden where the conspicuous-display aspect increased during the 16th century under the influence of continental Renaissance food culture. Imported luxurious food items gained in importance, e.g. expensive wines, sugar, spices and dried fruit. 119

Undersized and expensive to raise
Earlier archaeological research has suggested that turkeys were easy to rear, grew fast, and provided a relatively large quantity of meat. 120 As described above, turkeys were not considered easy to rear, however, but were on the contrary considered demanding and expensive to raise, with high mortality rates among young poults. Moreover, the claim that early modern Swedish turkeys yielded a lot of meat has been found to be incorrect. The turkeys kept at the crown estates often grew to a smaller size than ordinary geese. 121 The size varied considerably between different crown estates, and it is tempting to interpret this difference as the result of genetic variation, perhaps suggesting the existence of several breeds and/or populations of turkeys. However, as far as can be told all turkeys kept at the crown estates before the 1620s descended from the same original four birds that had been imported to Nyk€ oping in 1583. The variation must therefore be explained by other factors. One possible explanation is inbreeding, which might result in smaller birds. The problem of inbreeding must have been larger at estates situated on the outskirts of the kingdom, such as € Alvsborg, far from the large flock in Nyk€ oping. Another explanation is that the variation was connected to the presence or absence of expertise at different locations. Turkeys were a new sort of fowl, previously little known in Sweden. In order for the breeding of new imported animals to be successful, the import had to be coupled with proper carepreferably by experienced staff. Janken Myrdal has shown how this was manifested in the home farms connected to Nyk€ oping Castle, where Duke Karl kept many types of foreign farm animals. When the imported German cows were cared for by a Dutch and a German woman, the German cows milked up to 50% more than the Swedish ones. When the foreign staff left, the yield dropped to the level of the domestic cows. 122 The staff in Nyk€ oping was likely more experienced with turkey rearing, and this might have resulted in larger birds.
In geese, the lower limbs tend to increase in size with domestication and deliberate breeding towards higher slaughter weights. Simultaneously, the bones in the wing tend to decrease in size. 123 Brooklynne Fothergill has found that turkeys in early modern Britain had larger lower limbs than modern domestic turkeys. In line with the changes found in geese, she suggests that early modern British turkeys were increasingly husbanded for large size and rapid weight gain. 124 Bone measurements of Swedish early modern turkeys show that the wings were proportionally larger when compared to modern heritage breed turkeysthe opposite relationship compared to the British observations. This, in turn, may suggest that the turkeys identified in Sweden were less affected by selective breeding, and that they represented a less improved form, similar to their wild ancestors.

The turkey as an item of conspicuous consumption
When turkeys were introduced to Sweden in the late 16th century, they were items of conspicuous consumption in the Veblenian sense. As noted above, the termreferring to the consumption of luxurious goods for the sake of demonstrating statusis in many ways apt for describing turkeys. Veblen's definition of the term emphasises a dimension of wastefulness. In order for an item to be truly conspicuous, it must also be ostentatiously wasteful. 125 In line with this, the 'Veblen effect' describes how the demand for a product increases because it bears a higher price. 126 An example to illustrate this can be drawn from the above-mentioned oven at Frederiksborg Castle, designed for hatching eggs from rare birds as well as chickens. The latter were raised on a diet of sour milk and were described as weaker and considerably smaller when compared to chickens raised in a traditional way. Despite this, their price was about twelve times higher. Visiting high-ranking guests were offered these birds for dinner, with the main purpose to serve as a topic of conversation. 127 Similarly, the considerable effort and cost in raising turkeys was a matter of showing off an abundance of resources to peers and higher-ranking guests. The ostentatious waste was not in the consumption of the food itself, but in the way that care for the birds involved great cost, yielding only smaller, more precious morsels.
According to the theory formulated by Veblen, the leisure class defines a certain standard of lifestyle and consumption. All while the people within the leisure class continue trying to outdo one another in wasteful consumption, practices and preferences also spread to the lower classes, whose members strive to emulate the consumption of their social superiors. As a consequence, consumer goods first only available to elites, eventually become cheaper and available to more people, if supply can begin to catch up with demand. 128 This process also takes place with food culture, as the lower classes emulate the ingredients, recipes and cooking methods of the elites. 129 Our investigation has shown how turkeys, soon after their arrival in Sweden, began to spread beyond the dinner tables of the elites. The development led to a situation where the turkey evolved from an unobtainable luxury that previously had only been seen and heard in the areas around crown estates, into a much more tangible animal, and how, in the process, it acquired a meaning as a coveted foodstuff for a new range of consumers.

The emulation of turkeys
Our interpretation of how and why turkeys were introduced in early modern Sweden goes against the widespread, prevailing archaeological explanation, which attributes the introduction and spread to Dutch and German immigrants. It should be noted, that no evidence of turkeys has been found in the archaeological material in the first decades after their introduction to Sweden. This is likely explained by the fact that the material is heavily biased. Modern excavations of sites such as rural estates and castles where turkeys were kept during this period, are in fact largely lacking. The available archaeological material as a result cannot inform us about the first introduction of the bird.
The historical records, instead show that turkeys were introduced to Sweden through the efforts of Duke Karl and that the knowledge of turkey rearing, spread into the wider society from his crown estates. Already in the 1590s, turkey flocks had been established on estates controlled by other members of the royal family, and during the first decades of the 17th century turkey breeding and consumption appears to have spread to the aristocracy's rural estates. The development follows a similar pattern to the situation in the areas of the present-day Czech Republic, where turkeys first spread from the royal elites to the rural nobility. 130 The spread from the Swedish crown estates to the manors of the nobility often took the form of gift-giving. This parallels the development in the rest of Europe, where gift-giving was one of the main driving forces behind the spread of turkeys. 131 The aristocratic interest in turkeys, which took off during the 1610s, and the incorporation of turkeys in ceremonies and feasts, added new nodes to the existing knowledge system, further helping in the creation of a familiarity with the new bird. From the 1620s at the latest, turkeys were imported to Stockholm by the aristocracy and affluent urban burghers. Over time, turkey consumption spread to other strata in society, as can be exemplified for the early 18th century, when turkey bones were found in the food waste connected to lower middle-class urban households in Kalmar. The presence of bones from juvenile birds indicates that turkeys were most certainly raised inside Gothenburg, while the absence of juvenile birds in Kalmar and Norrk€ oping suggests that turkeys were brought in from the surrounding countryside to be sold to the burghers in these towns. These local differences in turkey rearing might be connected to differences in town layouts, e.g. the size of backyards or access to open areas that could be used to keep the birds. The local differences in turkey husbandry are worth further future investigation.
Turkeys were never common amongst the Swedish and Danish peasantry, and were instead primarily kept in smaller numbers at rural manors and vicarages. 132 In Sweden this meant that they never became emulated by all social groups or spread to all parts of the country. This also meant that turkey meat failed to play a significant role in Swedish cuisine, something which is noticeable even today. While there are several possible explanations for their limited spread, part of the explanation should be sought in the specific requirements that had to be fulfilled to rear the birds successfully. Overall, the rearing of turkeys was more expensive and demanding when compared to other poultry, and required specialized knowledge and substantial practical experience. 133 In order to be profitable, the birds were best kept on pasture, which limited their spread to urban environments or to households or farms without the means to buy the extra food that was needed to keep the birds in pens. Access to pasture was not the only prerequisite, however. Free-ranging birds were prone to attacks from predators. A Swedish turkey rearing manual from 1775 recommended using children or older women to herd the turkeys while on pasture. 134 On the other hand, the fact that turkeys could be raised by other domestic birds or humans, must have facilitated their spread. Eggs could be sent to new locations, where they could be raised by chicken hens, or even in special ovens built for hatching. While this hardly was an option for the peasantry, this might have contributed to the spread among the rural nobility, who had the financial and spatial means to raise the birds.

CONCLUSION
The results show that turkeys, first imported by Duke Karl of Sweden in the 1580s, had spread to the nobility by the 1610s. During the first decades of turkey husbandry in Sweden, turkeys were items of conspicuous consumption, used to show off during elite dinners and as gifts to peers and subsequently also to subordinates. During the 17th century, the bird was adopted by the urban upper middle class. The spread was both facilitated and hampered by the traits and particularities of the fowl. Early modern Swedish turkeys were small, and likely less affected by selective breeding when compared to modern heritage-breed turkeys.
The results also underscore the importance of combining archaeological interpretations with historical research. While the historical records provide a detailed picture of how and why the turkeys were introduced, the archaeological remains reveal how the birds spread to different social groups and their physical properties. It is only by studying both the archaeological remains and the historical records that we can get a full understanding of the earliest history of turkeys in Sweden.