Urbanised Nature in the Past Site formation and environmental development in two Swedish towns, AD 1200-1800 Jens Heimdahl Avhandling i Kvartärgeologi Thesis in Quaternary Geology No. 5 Department of Physical Geography and Quaternary Geology Stockholm University 2005 1 Cover painting by Jens Heimdahl ISBN 91-7155-143-3 ISSN 1651-3940 Akademitryck AB Urbanised Nature in the Past Site formation and environmental development in two Swedish towns, AD 1200-1800 Jens Heimdahl This doctoral thesis has ﬁve appendant papers, listed below. They are referenced by their roman numerals in the text. All ﬁgures included in the thesis are original illustrations made by the author. Paper I Heimdahl, J., Menander, H., and Karlsson, P., in press: A New Method for Urban Geoarchaeological Excavation, Example from Norrköping, Sweden. Reprinted from Norwegian Archaeological Review, Vol.38:2 (www.tandf.no/nar) by permission of Taylor & Francis AS. Paper II Heimdahl, J.: Urban Sediments as Indicators of Changes in Land use and Building Tradition in two Swedish Towns, AD 1200-1800 in two Swedish Towns. Manuscript Paper III Heimdahl, J.: Botanical Evidence of Changes in Vegetation and Cultural Landscape in PostMedieval Karlstad, Sweden. Manuscript Paper IV Heimdahl, J.: Botanical Evidence of Changes in Environment and Tradition in Norrköping, Sweden, AD 1200-1660. Manuscript Paper V Heimdahl, J.: Archaeobotanical Evidence of Early Tobacco Cultivation in Norrköping, Sweden. Manuscript accepted by Vegetation History and Archaeobotany Content 1 Introduction 2 Interdisciplinarity Geology and archaeology Interdiciplinary urban research Occupational layers and archaeological stratigraphy Clashes in terminology Stratigraphic terms used in this thesis 3 Urban occupational deposits Urbanisation and towns Urban deposits Dark earths The stratigraphy of urban constructions Stratigraphy of waste Urban pedogene soil processes 4 Urban plant ecology Earlier studies of urban archaeobotany in Sweden and neighbouring countries 5 Taphonomy of urban macro remains 1. Biology 2. Depositional processed and ecology in urban environments A. The local ﬂora B. The regional ﬂora C. Plants from pastures and meadows D. Cultural plants 3. Human inﬂuence 4. Preservation of plant macrofossils 5. The reworking processes 6. Sampling preparation and registration 7. Personal biases 6 Study sites 7 Methods Norrköping Karlstad Fieldwork Macrofossil analysis Pollen analysis Diatom analysis Grain size distribution Organic carbon and loss on ignition 14 C-Dating Other dating methods Ecological grouping Biological nomenclature 8 Summary of papers Paper I Paper II Paper III Paper IV Paper V 9 Discussion Urban strata, a formation model The formation of dark earth Sediments, soil horizons and culture 10 Conclusions References Svensk sammanfattning Tack! 7 8 8 9 10 13 14 14 14 15 16 18 20 21 23 24 26 26 26 27 28 28 28 28 29 29 29 29 29 29 31 32 32 33 33 33 33 33 34 34 34 34 34 34 35 36 37 37 38 38 41 42 43 44 55 59 Abstract In order to explore site formations and reconstruct environmental development in Medieval and Post-Medieval towns, urban occupational strata in Norrköping and Karlstad were studied according to biostratigraphy, sedimentology and pedology. New ﬁeld procedures including continuous pilot sampling, parallel archaeological and geological stratigraphic interpretation, and on-site analysis of plant macrofossils were developed and applied at archaeological excavations in both towns. Representation of both disciplines in the ﬁeld during excavations greatly contributed to more complete ﬁeld interpretations. Stratigraphical analyses indicate that geological processes have been active in both towns, and reveal similarities in site formation. The earliest proto-urban phase is represented by the presence of dark earths, formed by the combination of alluvial processes and cattle tramping. Alluvial processes were common in Karlstad due to the ﬂooding of the river delta, and in Norrköping due to the sloping topography. Both situations were enhanced by human activity, which caused drainage problems. A signiﬁcant change in composition and origin of house foundation ﬁll was also noted. The oldest foundations contained ﬁne-grained material of local origin in contrast to younger foundations, which contained coarser material, sometimes of regional origin. This is interpreted as a professionalisation of the urban building tradition, which in Norrköping occurred during the beginning of the 17th century and in Karlstad during the 18th century. Site formations of urban strata are regulated by three major factors: deposition, post-depositional soil formation and erosion/truncation, which all may occur both culturally and naturally. Plant macrofossil analyses in Norrköping and Karlstad resulted in a fossil record with a total amount of 203 and 169 different types of plant species and taxa respectively. The records indicate that site formation processes seem to have been inhibited during wintertime. The results also conﬁrm the idea of the early Scandinavian towns as rural, also during the Post-Medieval time. The ﬁnds of cultural plants in Karlstad indicate 18th century cultivation of Fragaria moscata and 17th century import of Pimento ofﬁcinalis. In Norrköping remains of beer additives conﬁrm that the tradition of combining Humulus lupulus and Myrica gale disappeared after the 15th century, but also indicate a the use of Filipendula ulmaria as a beer addative. Finds of seeds from Nicotiana rustica suggests that tobacco cultivation occurred in Norrköping 1560-1640, which is some decades earlier than known so far in Sweden. Keywords: Urban stratigraphy, environmental reconstruction, site formation, geoarchaeology, dark earth, macrofossil analysis Urbanised Nature in the Past 1 Introduction more carefully discussed in the chapter “Urban occupational deposits”. The foundation of palaeoenvironmental studies of urban areas is the interpretation of the urban stratigraphy. It is important to understand the site formation processes of urban occupational deposits, since these deposits derive from both cultural and natural processes. Ideally, archaeologists and Quaternary geologists should perform the ﬁeld interpretation together. However, the archaeological method of recording and interpreting urban stratigraphy, developed during the last 25 years, has often proven to be difﬁcult and demanding to combine with the methods used by visiting geologists (cf., Lucas 2001). The present thesis is based on site studies and analyses of samples collected from two urban excavations performed within the frames of so called ‘rescue archaeology’. A rescue excavation is conducted as a controlled form of the destruction of material that is classiﬁed as cultural heritage when making room for new constructions. The task of archaeologists is to retrieve, document and interpret the archaeological remains, leaving a description and interpretation in its place. Ideally the description should leave possibilities for researchers in the future to make new interpretations. In Sweden rescue archaeology is ﬁnanced by those responsible for the exploitation of the ground where ancient remains are present. Economical frameworks deﬁned by the county administrative board limit the excavation time and the analytical work performed after the excavation. It affects also the choice of strata that are to be documented. Generally urban occupational deposits in Sweden that are older than the 18th century are considered to be part of the cultural heritage and are classiﬁed as ancient remains that should be preserved. Younger strata are excavated without documentation. Towns are often built on older remains called ‘urban occupational deposits’ that contain information about the history of the town. Geological and biological remains are mixed and interbedded with archaeological traces of houses, streets and artefacts (man-made objects). In other words, urban occupational layers do not only contain ancient cultural remains, but also environmental archives from which palaeoenvironmental records can be extracted. Urban nature occurs alongside and intermingled with the cultural life in a town, and adapts to and affects urban cultural expressions. In this way, the study of urban nature results in the knowledge about urban culture, and thus we may ﬁnd examples throughout history of the reactions and responses in the relation between culture and nature. Traditionally, the main focus of geological and biological studies have been placed on the ‘undisturbed’ nature – the nature untouched and unaffected by humans. In later decades, with growing concern about the anthropogenic effects on the environment, studies of ‘disturbed’ nature have become more common. Among culturally affected areas, the urban environment can be considered the most extreme. Geological and ecological studies of towns – urbanised nature – is however a young and growing discipline (Gilbert 1989, McDonnell & Pickett 1990, Botkin & Beveridge 1997, McDonnell 1997, Walbridge 1998, Nilon et al. 1999). In this ﬁeld, studies of the palaeoecological traces in urban occupational layers may contribute new perspectives (e.g., Latalova et al. 2003, Wittig 2004). The term ‘urban’, used in the title of this thesis, has been deﬁned in many different perspectives, for example according to law, history and demography. I use this term to characterize a speciﬁc type of environment. The deﬁnitions will be A 20 10 B Karlstad County of Värmland Sweden 60 County of Örebro County of Södermanland Vänern Norrköping Baltic Sea 55 200 km County of Västra götaland County of Östergötland 50 km Figure 1: (A-B) location of the investigated towns of Norrköping and Karlstad in southern Sweden. 7 J. Heimdahl The sites investigated within this study are situated in Norrköping and Karlstad, middle Sweden (Fig. 1). The main aims are: I) To develop methods for integrated ﬁeldwork between archaeologists and Quaternary geologists/palaeoecologists (Paper I). How can the members of the different disciplines adapt their working procedures to utilize each other’s capacity of interpretation? II) To reconstruct site formation and environmental development at two urban sites (Paper II-V). Reconstructions of site formation and environmental development are performed according to two main perspectives. Site formation is mainly studied by sedimentology and pedology, an approach that I refer to as ‘geoarchaeological’ (Paper II). Environmental development is principally studied from the perspective of biostratigraphy/archaeobotany (Paper III-V). However, the two perspectives of geoarchaeology and archaeobotany are sometimes integrated to conﬁrm or contradict speciﬁc results. This thesis has been written from the perspectives of Quaternary geology and palaeoecology/archaeobotany. The environmental development of Karlstad and Norrköping has been previously studied from the perspectives of archaeology and history, studies which I take into account, and the conclusions of which are compared with the results and conclusions in this thesis. Hence, it is my intention that archaeologists and historians, as well as Quaternary scientists and botanists, should be able to understand, use and scrutinize this research. Accordingly, the different disciplines require an introduction and my basic approach to this interdisciplinary work is presented below. 2 Interdisciplinarity The description and interpretation of objects and subjects may be facilitated if they are approached from different disciplinary perspectives. Geological deposits may contain biological remains and hence, a biological perspective can contribute to the understanding of the deposit. The objects of the archaeological study are the remains of material culture, which are also possible to study according to the methods of natural sciences, history and social sciences. These statements and examples may sound trite, but while they are easy to make, they are, however, difﬁcult to implement. Geology and archaeology Since the disciplines of geology and archaeology tra- 8 ditionally belong to different faculties, natural sciences and art, this may give an impression that they are fundamentally different. It has been stated that: “…archaeology […] straddles the gulf which separates the arts from the sciences.” (Hodder 1992). But I don’t believe that it is so. I don’t think there is a gulf. A popular opinion concerning those differences is that natural sciences are based on empirical principles and use quantitative methods while the humanities are based on hermeneutical principles and mainly use qualitative approaches. This is incorrect; the fundaments of both disciplines rest upon empirical principles and hermeneutics (sensu lato interpretation – not only of text) and use both quantitative and qualitative methods (Kieffer 2005). For example, geological interpretation of stratigraphical sequences is inherently hermeneutic and qualitative. Similarly, the botanical identiﬁcation of plant remain based on morphology, is qualitative. These studies can be quite subjective, perhaps more so than either discipline would like to admit. There is currently a debate among Scandinavian archaeologists concerning the question of what may be considered archaeology (Svestad 2004; Wienberg 2004; Cornell 2005; Hegardt 2005; Herschend 2005; Karlsson 2005; Kristiansen 2005; Notelid 2005; Rundqvist 2005). It relates to the criticism of post-processual archaeology, which emphasises the symbolic meaning of material culture, partly as a reaction to positivistic trends, the so-called ‘New Archaeology’ that inﬂuenced archaeology in the 60s and 70s and attempted to reform archaeology to become more like the natural sciences. The debate is ongoing and trends are developing in all scientiﬁc disciplines but I would like to emphasise that my studies of archaeological material from a perspective of natural science are not to be seen as a contribution to this debate. I will not debate the nature of archaeology, only some of the study objects of archaeology, although from a different perspective. Some of the differences between humanities and natural sciences fall back on the classic and controversial question of what is considered nature versus culture. One perspective is that humanity exists beside or on the top of nature and that human activity (culture) is considered separate from nature. The contrasting perspective is that humanity is part of nature and that culture should be considered as a part of nature as well. The controversial aspect of these questions partly derives from the deﬁnitions of the terms that apply to societal values. ‘Nature’ is often used as synonymously to positive terms such as ‘normal’, ‘ordinary’, Urbanised Nature in the Past Figure 2: Conceptual model of how an object – urban occupational strata and the processes that caused them (marked with grey) – may be studied by different disciplines. More disciplines may also be added, such as chemistry, physics and human geography. These disciplines may further be divided into sub-disciplines e.g., geoarchaeology which is at the boundary zone between archaeology and geology. ‘usual’, ‘regular’ or ‘original’, and to describe something that is ‘good’, ‘harmonious’ or according to a higher or ‘God-given’ purpose. On the other hand, the term ‘culture’ is more complex, referring to a way of life, customs, beliefs, art, identity etc. It may refer to something sophisticated but also something that opposes nature – something ‘unnatural’. From the perspective of natural science, humanity might also be seen as a part of nature because humans and human culture are observable in the physical world. This does not mean that all aspects of humanity and culture can or should be studied according to the methods of natural science. In fact, most cultural aspects are far too complex for these methods and instead, require methods from the humanities and social sciences. In the following, I will use the term ‘culture’ to describe human activity while ‘nature’ will be used for everything that is not cultural. It is merely a practical use that results from the need to separate human actions from non-human actions, not a moral statement. In this regard, it is also important to note that irregardless of whether we include culture as part of nature or not, nature and culture strongly affect each other and often in such a way that cause and effect are difﬁcult to distinguish. It is not always possible to separate nature and culture, and it is in such cases that interdisciplinarity is put to the test. Interdisciplinary urban research A multidisciplinary approach is to study an object from the perspective of many different disciplines parallel to each other. This multidisciplinary work becomes interdisciplinary in the dialogue between the different disciplines, and when they thereby af- fect and inﬂuence each other. In this study the objects consist of the physical remains of towns – urban occupational strata. These strata, and some of the processes that have caused them, may be studied from the perspectives of archaeology and natural sciences. However, some of the causes behind the processes may be of such nature that they only can be understood by the use of historical or sociological methods, even if the physical object itself cannot be explained according to those methods (Fig. 2). Since the end of the 19th century, urban occupational deposits have been considered part of our cultural heritage and the main responsibility to investigate them has fallen on archaeologists. European towns generally belong to the period deﬁned as ‘historical’ – the period that can be studied through written sources. When archaeologists began to study the material from this period they collaborated with historians. The archaeological branch that studies remains from the historical period is called ‘historical archaeology’, and it has to consider different types of sources of information. The speciﬁc study of town remains may be called ‘urban archaeology’. In Sweden urban archaeology includes ‘Medieval’ and ‘Post-Medieval archaeology’, two periods that are separated by the time of reformation of the church deﬁned at AD 1527 by king Gustavus Vasa. Urban archaeology has undertaken many questions that traditionally fall within social sciences and human geography, questions important to the urbanisation process. Many archaeologists remedy the need for multidiciplinary knowledge by taking supplementary training in Quaternary geology, and there are archaeologists that are also good geologists or geoarchaeologists. 9 J. Heimdahl This supplementary knowledge is however not enough to cover the need at all excavations. Archaeologists and geologists work with different scientiﬁc questions, not only because of the differences in the study objects, but also because of the differences in disciplinary background. Archaeologists need not only to determine what process or behaviour has formed a certain record, but must also ask questions about why this process or behaviour occurred and attempt to understand the context of social interaction behind it (cf., McLees et al. 1994, Larsson 2000). Geologists need to ask questions about the mechanisms behind processes but since those processes traditionally do not include human activity, social interactions are not normally considered as a factor. If a geologist works with occupational layers, the processes behind these layers may be caused by social interactions and should therefore be considered. However, since geologists generally lack the necessary knowledge concerning cultural explanations, the interpretation process should be performed in conjunction with archaeologists or/and historians (Fig. 2). Occupational layers and archaeological stratigraphy The sediments that compose the stratigraphy in occupational sites may be of natural or cultural origin. Natural deposits that can be formed within, or cover an occupation may be of eolian (wind transported), alluvial (water transported) or colluvial (gravity transported) origin but most common are depositions created by humans. Such deposits are called ‘occupational layers’, ‘occupational deposits’ or sometimes ‘cultural layers’. It is here worth mentioning that the geological terminology concerning occupational deposits is poor. On the maps of Quaternary deposits published by the Geological Survey of Sweden (SGU), urban occupational layers are marked as “Artiﬁcial ﬁll: the original ground surface is covered by an alien material” (Persson & Svantesson 1994). The word ‘alien’ may here also serve as an example of how humans sometimes are viewed from the perspective of geology. The archaeological terminology is hence developed independently from geology. A general difference between geological and archaeological stratigraphical methods is that those used by archaeologists, to a greater extent, are focused to describe the surfaces buried within the stratigraphy (contacts between layers called ‘interfaces’). Geological terms and methods are more focused around the layers and sediments themselves and their changes with depth. Different deﬁnitions of ‘occupational layer’ exist, partly because ﬁeld-terminology varies locally and 10 Figure 3: Many excavations during the ﬁrst half of the 20th century were performed according to the ‘Box and baulk system’. The excavations were performed according to a regular system of rectangular areas. The bottom of the pits was always horizontal and often cut through strata that were dipping. Stratigraphic control was poor and stratigraphic interpretation difﬁcult. also because of the methodological development within stratigraphy. Traditionally there has been a need for archaeologists to separate stratigraphical units formed intentionally, like ﬁll used as foundation for houses or other buildings, from units that are formed unintentionally, either by cultural or natural action. Strata formed intentionally may in fact be considered as artefacts (cf. Arrhenius 1996) since humans create them, and their form and placement may reveal information about the people that formed them. On the other hand, strata formed unintentionally, for example as a secondary effect of a cultural activity, may reveal information about this activity also by its content. Excavation techniques differ slightly between countries and regions. Here I will brieﬂy describe the development of excavation methods in Sweden. They have many international connections, especially to the development in Britain (cf., Roskhams 2001). For a long time excavations were made in rectangular, horizontal blocks, the so-called “box and baulk system” (Larsson 2000, Roskhams 2001), without concern for how the layers dipped (Fig. 3). The stratigraphy was carefully documented by the interpretation and drawing of sections when the excavation was completed, but the stratigraphical control of the ﬁnds was poor. During the 1970s, the stratigraphical awareness started to grow among urban archaeologists who had begun to confront the complex occupational deposits of towns. Järpe et al. (1979) suggested a drawing standard for documenting occupational layers. Also a narrow deﬁnition of what was to be called ‘occupational layers’ was suggested, according to the idea that the strata had “grown” through “unintentional cultural sedimen- Urbanised Nature in the Past tation”. Beside those slowly and continuously deposited sediments called ‘accumulations’ (Sw. avsatta lager), there were also intentionally created layers called ‘depositions’ (Sw. påfö f rda lager), which were fö not to be considered occupational layers. The idea was that the unintentionally accumulated occupational layers that successively had been growing during occupation (due to the habit of waste handling) would contain more information about the everyday life on the site than the intentionally formed deposits. Järpe et al. (1979) also called this type of occupational layers ‘cultural gyttja’ and presented a number of symbols, representing different types of layers, to be used use within section drawing (Fig. 4). The type of layers presented in the system were however not explained and deﬁned, and even though many of the words used (such as ‘peat’, ‘gyttja’, ‘alluvial layer’ etc.) were of geological origin, it is obvious that the deﬁnitions used by the authors were not geological. For a period the paper by Järpe et al. and the symbols they presented had a great inﬂuence on stratigraphical methods in Sweden (Larsson 2000, Tagesson 2000a) since it contributed a tool to handle stratigraphy in a more interpretative way. It also seems to have corresponded to former ideas about site formation process in Britain (Matthews 1993). During the 80s Andrén (1985, 1986, 1989a, b) introduced the terms ‘latent’ and ‘manifest’ which aimed to separate deposits formed due to unintentional action from those formed by intentional acts. Since the different deposits are formed according to different degrees of intent, they have to be separated and sometimes interpreted in closed contexts (cf. Andrén 1989a, b). The ideas of different degrees of intent in site formation processes have also been discussed by Florén and Dahlgren (1996) who suggested a separation between ‘acts’ (intentional acts) and ‘behaviour patterns’ (acts according to habits without speciﬁed intention). The need for better stratigraphic control within urban archaeology was emphasised by Beronius-Jörpeland (1992) and a broader deﬁnition of occupational deposits was suggested. She suggests that all deposits created by humans are to be called ‘occupational deposits’, and stressed the importance of separation between depositional and post-depositional processes. During the 80s and 90s there was a continuous development of excavation techniques, although they differed locally. The technique of excavating stratigraphically was developed. One stratum at a time was excavated and the extension of the stratum was followed horizontally. By this method it was possible to uncover and separate old ground surfaces, which made the interpretation of the site development easi- er (e.g., Anund 1995, Ros 1996, Larsson 2000). A considerable shift in excavation techniques, especially at urban sites, occurred in the beginning of the 21st century. The shift had its prologue in 1993 when the method of so-called Single Context Recording (hence referred to as SCR) was introduced in Sweden after experiences from Norway (Larsson 1993). SCR is a stratigraphic recording method adapted to archaeology that deals with complex stratigraphy. It was developed in Britain by Edward Figure 4: Stratigraphical drawing according to the system of Järpe et al. (1979). According to this system only the ‘accumulated’ units and the ‘cultural gyttja’ are considered as true occupational layer. The system presented a large number of symbols for the different types of material, but did not deﬁne the material itself. A comparison of different stratigraphical records obtained with this system, shows that the terms are not classiﬁed according to any geological or archaeological deﬁnition. 11 J. Heimdahl Fig. 5: Example of application of Harris matrix (Harris 1979). In the matrix to the left stones are marked with boxes, layers with rings, holes with pentagons, and hole inﬁllings with triangles. In this case, layer (4) is the settling sand for the cobblestones (3). If this cobbled ground were interpreted as a part of the construction that was held up with posts, e.g. a shed, posthole (2) would also be a part of the shed-context, marked within the dotted line in the matrix. The hole inﬁlling (1) is a later deposition and will not be a part of the context. Harris (Harris 1979 and Harris et al. 1991) and was built on the idea of classifying and uniting different stratigraphical units (which are deﬁned as consisting of ‘depositions’ and ‘interfaces’) and artefacts into time dependent contexts. The contexts are in turn connected to groups. The stratigraphical relation between those units, contexts and groups, are illustrated by the drawing of a so-called ‘Harris Matrix’ (Fig. 5). This type of matrix was used earlier, for example at Helgeandsholmen in Stockholm but the excavation only used parts of this method (Broberg & Douglas 1983). In this method both layers, interfaces and “negative” units, such as holes, are given the same treatment. The method has proven to be a powerful and effective tool in order to document and interpret stratigraphical information, and it forces the user to carefully consider all strata and contacts. SCR has dominated urban excavations in the latest years, for instance at excavations in Lund from 1993 (Larsson 1993, 2000); Norrköping 1998 and 2000 (Hållans et al. 1999, Tagesson 2000b Karlsson et al. in press); Sigtuna 1999-2000 (Fogelberg & Wickström 2003) and Karlstad 2001 (Bäck 2002). There has been criticism against the weaknesses within the SCR-system, for example the problem of catching the cultural activities within the time-gaps between the stratigraphical units described within the matrix (Fogelberg et al. 2004). Harris (1979) states that occupational layers consist of two components: depositions and interfaces. In the later years, the need for separating the strata into depositions and soil horizons (depositional and post-depositional remains) within the interpretation has been stressed (Heimdahl 2003, 2004). The depositions generally reﬂect short-term events, while soil horizons devel- 12 oped within the depositions are generated by the effect from a longer sequence of every-day activities. In later years, the development of digital tools and GIS-techniques have greatly improved urban stratigraphical documentation and new techniques are tested alongside the use of SCR. Extensions of layers and positions of ﬁnds and samples are measured, digitalised and registered in GIS-programs (for example Arc-View). Also 3-D visualization of complex urban stratigraphy has been tested, for example in Naantali, Finland (Lehtonen & Uotila 2004). The extensive redeposition that has occurred in urban occupational deposits sometimes makes it difﬁcult to interpret ﬁnds therein. This has led to a need to classify different deposits according to the risk that its content (of for example artefacts) is redeposited. In Britain, a system of classifying ﬁnds into ‘primary’ (original) and ‘secondary’ (redeposited) positions has been in use, but was criticised by Roskhams (1992) as being too coarse. He instead suggested a system of four classes, relating the ﬁnd to chronological, spatial and functional contexts. In Sweden, this system developed into a three-grade scale where the ﬁnds are classiﬁed as: • primary deposited – according to original time, space and function contexts, for instance household tools dropped on the ﬂoor in the house where it was used. • secondary deposited – contemporaneous but isolated from original context in function and space, for example waste deposits. • tertiary deposited – redeposited material, isolated in time, space and function (Larsson & JohanssonHevren 1998), for example material and artefacts in Urbanised Nature in the Past layers that are excavated and used as ﬁll in new constructions on another site. This system was later adapted to account for the depositions themselves (Tagesson 2000a). Tertiary deposits consisting of ﬁll that is redeposited may then have had a primary function, for instance as a foundation for a house. Clashes in terminology When discussing terminology, we must assume that terminological use within a discipline often differs between different groups of researchers, both locally and regionally. However in this discussion, I refer to geologists and archaeologists, although simpliﬁed, as two homogeneous groups and discuss some similarities and differences that appear to be widespread. Both geology and archaeology work with the physical material that holds information on past conditions, and both disciplines attempt to gain knowledge through the use of stratigraphic principles. Many of the stratigraphic terms used in archaeology derive from geology (c.f., Larsson 2000). It can therefore be presumed that archaeologists and geologists use the same stratigraphical terminology. This is partly true, especially when it comes to descriptive terms (e.g., ‘deposition’, ‘interface’) and technical terms (e.g., ‘gravel’ or ‘clay’). However, there are major differences between the uses of terminology in the two disciplines. Archaeologists usually do not apply many terms used in geology. Examples are: ‘autochthonous’/‘allochthonous’ and ‘concordance’/‘discordance’, which are descriptive terms that probably would be useful also to archaeologists. Although when these terms have been used by archaeologists (e.g., Roslund 1997), other archaeologists have found them complicated (Larson & Johansson-Hervén 1998). In other cases, archaeologists have developed their own terms to describe stratigraphy. Good examples are the terms referring to SCR and the use of Harris matrix. There are also examples of terms that are used by both disciplines but differ in deﬁnition. Some of those differences are necessary and depend on the differences in approach, objects of study and methods. The different use of terms has led to interdisciplinary misunderstandings and conﬂicts. Since geologists often have a clearer view of terms associated with sediments and soils, sometimes archaeologists just leave it up to the geologists to classify the material. This can create a false sense that the material is classiﬁed according to strict and correct principles. This is perhaps so, but sometimes geological termi- nology is insufﬁcient to describe the material that occurs in archaeological contexts, and the geologist may lack knowledge of the formation process of “archaeological” strata. Also, terms with common deﬁnitions used in archaeology may at times be more appropriate than geological or pedological terms when describing occupational deposits (cf., Heimdahl et al. 2003). One example is the Swedish word torv, which has two possible translations in English: ‘peat’ and ‘turf’. In Swedish geological terminology torv is deﬁned strictly as ‘peat’ – an autochthonous accumulation of plant remains, for instance vitmosstorv (Sphagnum peat) or starrtorv (Carex peat). However, an older meaning of the word, probably of agrarian origin, is the uppermost vegetative zone of soils in grasslands – the root horizon (easy to cut with a spade to sods) of turf. In Sweden as in many other European countries, turf has been used as a building material, for example in ramparts, earthwork and as a cover on house roofs. At some archaeological sites, ground vegetation and root horizons may be covered and become preserved as buried horizons within the stratigraphy, which also may be considered turf. Furthermore different types of peat have been used throughout history. For instance, Sphagnum was used as a packing material, as ‘toilet paper’, as moist buffering material in beds and as litter in stables (Krzywinski et al. 1983, Geraghty 1996, Hansson & Dicksson 1997). In some cases, when a site is located in a moist area (for instance an alluvial valley), peat may also have accumulated naturally in the urban environment. It is possible to ﬁnd both remains of peat and turf in urban occupational layers and of course, it is important to separate these two different types of material since they indicate different environmental origin and cultural use. This does not mean that the Swedish term torv should be applied only in one way but it is of great importance to be aware of the different deﬁnitions in order to clarify descriptions. Another example of confusing terminology is the Swedish term jordart, which may be translated to ‘regolith’ – an unconsolidated deposit. It may have both a genetic deﬁnition like ‘till’, ‘glacioﬂuvium’, ‘littoral sand’, or a descriptive function, for example ‘gravel’, ‘sand’ and ‘clay’. Occupational deposits may also be referred to as a type of jordart. In Scandinavia, deposits of this type are of Quaternary origin but the term is also used to describe weathered material that may be much older in other parts of the world. The geological term jordart should not be confused with the pedological term jordmån, which corresponds to ‘soil’ – the upper part of a regolith that has been affected by post-depositional proc- 13 J. Heimdahl esses associated with physical, chemical, biological or cultural factors. The problem is that the English term ‘soil’ is not only used in the strict pedological deﬁnition but also in a more general manner, even by scientists. It does not then refer to a speciﬁc type of soil but either to all soils covering the world’s land area (Brady 1990), or at times to the material composing a soil or terrestrial sediment. This use may correspond to the word jord, which lacks a geological deﬁnition, but may sometimes refer to what in a geological deﬁnition should be considered a regolith (jordart). It may also be added that the term ‘earth’ is not used as a substitute for soil. Many Swedish archaeologists are not aware of this complexity of deﬁnitions and do not notice when a British or American author refers to the pedological deﬁnition of soil or merely uses the word in a general, undeﬁned way. For instance, it is common that archaeologists deﬁne culturally affected soil horizons as ‘occupational layers’. Since ‘layer’ refers to a deposit, and the culturally affected soil to the main part consisting of the minerogenic sediment that has been there before the occupation, this often leads to misunderstandings. Torv and jord- (art/ art mån) are two examples of how art/ problems concerning geoarchaeological terminology may arise. These, and many more examples not mentioned here, reveal the importance of awareness of the differences and similarities between the terminologies used within the two disciplines. Clear and open communication between the disciplines is necessary, not only through an explanation of terms, but also through a sensitive and receptive dialogue. Stratigraphic terms used in this thesis The urban occupational layers I have worked with are composed of two main physical units: depositional layers and post-depositional soil horizons. Since it is of great importance to keep these apart, I will use the terms sensu stricto according to the working deﬁnitions presented in Table 1. 3 Urban occupational deposits Urbanisation and towns Historians, human geographers and archaeologists have mainly performed research on the history of towns during the last 25 years. Before the 1970s, historians dominated the arena of urban science and at that time very few urban archaeological excavations had been performed in Sweden. Exceptions were the towns of Birka, Lund and Gamla Lödöse, 14 where long traditions of urban archaeology existed. Large building projects during the 1970s and 1980s caused a huge expansion in urban archaeology. As a response to the need for information concerning ancient remains in towns, a project called ‘The Medieval Town’ (Sw: Medeltidsstaden) was initiated by the National Heritage Board (RAÄ). The aim of this project was to provide an overview of the urban medieval remains in Sweden, which concerned about 70 towns that were given privileges before the Reformation in 1527 (Forsström 1982, Andersson 2001, Lindeblad 2002). The history of towns differs according to the deﬁnition of ‘town’. In the juridical aspect, a town is an area that has received special privileges to regulate trade. In Scandinavia, this system was ﬁrst introduced in Denmark (11th century) and Norway (12th century). Therefore, the oldest Swedish towns, according to this deﬁnition, are located in areas that previously belonged to those countries. Lund and Helsingborg in the county of Skåne was founded in Denmark during the end of the 11th century. Kungahälla in the county of Bohuslän was founded 1130 in Norway. In Sweden, the system of town privileges was introduced during the early 13th century. Examples of early Swedish towns, according to this deﬁnition, are Visby, Söderköping and Sigtuna (Ahlberg & Redin 1994). Another type of deﬁnition of town is social and structural, for example “a dense settlement, where inhabitants form a social unit in economic or/and juridical aspect according to common interests in trade” (Schück 1926). According to this deﬁnition, Scandinavian towns are older. Ribe in Denmark seems to have been founded as a central trading place in the early 8th century. Birka in Lake Mälaren (Sweden) existed c. 760-1000 AD, and many of the later medieval towns seem to have been founded upon older occupations, which in many cases may be called towns, for instance Visby and Sigtuna. Within the juridical deﬁnition of towns, those types of settlements are sometimes referred to as ‘proto-towns’ (Holmqvist-Olausson 1993). Deﬁnitions related to the environmental aspects of the terms ‘town’ and ‘urban’ are also used among urban ecologists (cf., Mcintyre et al. 2000). In urban ecology, the environment is often compared with rural areas along so-called ‘urban – rural gradients’ (e.g., McDonnell & Pickett 1990, McDonnell et al. 1997, Simmons et al. 2002, Moffatt et al. 2004). ‘Urban’ and ‘rural’ are then commonly deﬁned according to population density in speciﬁc areas (e.g., McDonnell et al. 1997). Urban palaeoecology often confronts similar gradients, although Urbanised Nature in the Past Table 1: Explanation of stratigraphical terms used in this thesis Term Contact and Interface Dark earth Explanation The boundary between layers/beds/deposits. Archaeologists most commonly use ‘Interface’ while geologists use ‘contact’. A dark, organic rich stratum (generally over 2 dm thick) that commonly pre-dates the well stratiﬁed medieval and post-medieval urban occupational deposits in Scandinavia and Europe. Dark earths are more thoroughly discussed in Chapter 3. Deposit, layer and bed The terms are used synonymously for depositional units or sediments of natural or cultural origin. The material in layers and beds may have been deposited during a single event or by successive deposition, for example in water bodies where sedimentation of single beds may take some time. Layers, beds and deposits can be divided into subunits that are also called layers, beds and deposits. Facies A series of speciﬁc stratigraphical units developed during similar conditions (cf. Reading & Levell 1996). The stratigraphy at the sites studied here is divided into four different Facies. A Facies may further be divided into subfacies. Horizon or soil horizon A post-depositional feature developed within a deposit/layer/bed by pedogenic processes under the inﬂuence of climate, soil chemistry and/or biology. Horizons in urban occupational layers are often developed by tramping by humans and cattle. When horizons are formed through several strata, contacts between layers can be destroyed and older stratigraphical features disappear or become obscured. Phase A term used by archaeologists (based on e.g. different building styles or other cultural habits reﬂected by artefacts) to separate time dependent units in a stratigraphy Sediment Transported and deposited minerogenic and/or organic material consisting of one or several layers. Soil Not used as a single word in this thesis in order to avoid confusion. It is, however, used to describe postdepositional process or strata e.g., soil horizon, soil process or soil bioturbation. Stratum and unit Refers to an unspeciﬁed stratigraphical unit that has not yet been classiﬁed as a layer/bed/deposition or a soil horizon or a combination of both. Urban deposits occurring chronologically, but there is generally a lack of information concerning population density. The term ‘urban’ is instead used to name a certain type of environment with speciﬁc characteristics. Early Scandinavian towns generally had a strong rural character since farms were often situated within the towns. This means that the early urban environments had rural character that eventually decreased and disappeared. Because of this process, the terms ‘urban’ and ‘rural’ are not mutually exclusive (cf., Clark 1992). I will use the term ‘rural’ as referring to an environment dominated by activities occurring in a rural economy (as crop processing and live stock handling), n.b. the degree of self-sufﬁciency not deﬁned. The term ‘urban’ will refer to a densely and permanently populated area that is environmentally characterised by trade, craft and/or industry. A central issue within urban history is the processes that give rise to the towns – the urbanisation. The causes behind urbanisation are complex. They include geographical, social, historical, economical and physical factors. Ambitions of a central political power often seem to have initiated the process. In Scandinavia, urbanisation occurred irregularly from the late Iron Age (Viking Age) until modern times (Andersson 2001). The character of urban occupational strata in Scandinavia differs not only between but also within the towns themselves. Compared to geological deposits, an urban stratigraphical unit varies on a smaller scale due to its cultural origin. Of course there are also many urban sites that lack older occupational layers, both because of uneven deposition and as an effect of truncation. According to my own observations at urban excavations (in Norrköping 2000; Sigtuna 2001; Stockholm 1999, 2001; Lund 2003; Mariestad 2003; Karlstad 2003-2005 and Skänninge 2003-2005), as well as studies of records from older excavations (Beronius-Jörpeland 1992, Larsson 2000), some general features may be recognised. Urban sites are commonly located close to the water (for instance, rivers, lakes or in coastal areas) since waterways were the natural routes for trade and transport. As a result, many towns have been exposed to ﬂooding and alluvial deposits are often found interbedded within urban deposits. Waterlogging of deposits is also common, which leads to anaerobic conditions. This inhibits the decay of uncharred organic material and leads to the preservation of organic material, for example plant remains. The thickness of urban occupational deposits 15 J. Heimdahl Figure 6: Type proﬁle of occurrence of older urban occupational deposits. They tend to be concentrated to depressions in the natural substratum and as slope front ﬁlls. Older urban occupational layers are generally covered by modern ﬁll, and excavations occur where this modern ﬁll would replace them. seems to be dependent on local topographical features and therefore have a levelling effect on the urban topographical landscape, since they tend to ﬁll depressions and become thinner on the top of hills (Fig. 6). In a ﬂat landscape, the thickness of urban deposits is generally even. In Sweden, the thickness of urban deposits is up to 12 meters thick, excluding 19-21st century ﬁlls. In many towns there is a general trend in the degree of stratiﬁcation according to relative age. Younger units are generally thinner and more distinctly stratiﬁed than older units, which are often composed of thick (20-70 cm), dark, homogeneous strata rich in organic matter. Such old strata have proven difﬁcult to interpret through archaeological methods. Internationally they have been called ‘dark earths’ (more thoroughly discussed in the following section). The upper, well-stratiﬁed strata are generally easier to interpret since they mainly consist of two main units: remnants of constructions (e.g., ﬁlls, houses and streets) and waste. The material that constitutes urban deposits varies from minerogenic to more or less organic but is mostly a mixture of minerogenic and organic components. Charcoal, fragments of wood and faecal material from both humans and animals often dominate the organic component. Reworking processes were extensive, both due to deliberate actions such as digging and truncation, and unintentional effects of tramping by humans and animals. The material in the deposits is therefore often from different origins. Dark earths The term ‘dark earth’ (in some early texts referred to as ‘black earth’) originates from Britain and was coined to describe the thick, homogeneous, poorly stratiﬁed, dark soils that often separate the clearly stratiﬁed urban deposits of the Roman and Medieval Ages (Fig. 7). The term was traditionally associated with the abandonment and decline of urban environments during the early Middle Ages – the ‘Dark Ages’ c. AD 500-1000 (Norman & Reader 1912). Dark earths were formed as a result of cultural activities 16 that occurred at former urban and proto-urban sites (Macphail et al. 2003). (Note that both ‘dark earth’ and ‘black earth’ are terms also used in the classiﬁcation of natural Amazonian soils and humid step soils. Those natural soils have nothing to do with the dark earths found in the archaeological contexts discussed here.) In later decades archaeologists in, for example Italy, Belgium and France (terres noires) have adopted the term. The European use of the term gave rise to a broader deﬁnition, and ‘dark earths’ are no longer considered to have been formed during a speciﬁc time period, in a speciﬁc urban environment, or by a speciﬁc human activity. The term is now more generally used to describe dark, thick (often 20-70 cm), poorly stratiﬁed stratigraphic, organically rich units that occur at archaeological sites. The genesis of dark earths on different sites is probably diverse. When occurring at urban sites, dark earth is usually associated with some type of urban decline, and the local environment is often interpreted as changing into waste ground or midden (Macphail et al. 2003). Ideas about the processes that create dark earths have varied. Macphail (1994) showed that dark earth might be formed by natural soil formation due to total abandonment of the area. A ruderal ﬂora may begin to grow in the house debris, giving rise to entisols or inceptisols (developed by earthworm and root bioturbation) in occupational deposits. Cattle grazing (Gebhardt 1997), middening (Macphail 1994), cultivation (Gebhardt 1997), market activities (Reece 1980), ‘backyard activities’ (Gebhardt 1997) and natural ﬂood events (Corty et al. 1989) have all been used to explain the formation of speciﬁc dark earths. In the 1980s and 1990s, a discussion took place in Italy concerning the formation process of dark earth. The question was whether the formation process occurred during a few years or decades, or longer time periods of hundreds of years (Macphail et al. 2003). In later years, it has been obvious that dark earth often reveals a more complex stratigraphy, obscured by the bioturbation (e.g., Sidell 2000). In Sweden, the term ‘dark earth’ has so far not been Urbanised Nature in the Past adapted in this European aspect. The term svarta jorden (‘black earth’) has long been used to describe soils with occupational remnants rich in carbon dust, for example the soils in the area of the Viking Age town Birka on Björkö in Lake Mälaren. Use of this term is documented from 1683, probably as an adaptation to the terminology used locally (Haldorph 1683, Hyenstrand 1992). After this, and since the excavations during the 19th century (e.g., Stolpe 1871-1879), the term ‘black earth’ has been regularly used when referring to the complete stratigraphy of the occupational layers in Birka (c. 1-2 m thick). Also, well-stratiﬁed units preserved within the occupational layers at Birka are referred to as belonging to the ‘black earth’. The Swedish term svarta jorden is also used to describe the carbon-rich occupational layers of the 12th century town of Kungahälla (Andersson & Carlsson 2001). Engelmark (2002) used the term ‘dark earth’ for the urban occupational layers at Sigtuna, deposited between the 10th and 13th century. Those layers are, however, well-stratiﬁed units of urban occupational layers so in this case, the term is used in a different way. In Norway, the term ‘black earth’ has been used to describe the remain- Figure 7: Typical stratigraphical position and habitus of dark earths in Britain (left) and Scandinavia (right). ing deposits of the Viking Age marketplace Kaupang (Sørensen et al. 2002). The term ‘black/dark earth’ in Scandinavian contexts seems to refer to carbon/organic-rich urban occupational deposits in general. Thick, poorly stratiﬁed, organic-rich strata that predate well-stratiﬁed medieval urban units are also commonly found in Scandinavia (Fig. 7). Apparently, they correspond to the European dark earths but have not been referred to as such. In the following, I refer to these Scandinavian units as ‘dark earths’, in the broad, descriptive sense not connected to any speciﬁc formation process or human activity. The occurrences of dark earths have also been debated in Scandinavia, but since this area lacks the traces of Roman occupation as on the European continent and on the British Isles, the Scandinavian dark earths have generally been looked upon as indications of a new development – a development that preceded the advanced medieval urban phase, rather than as a decline of an earlier advanced phase. To explain the shift from dark earths to well-stratiﬁed urban units, the focus has mainly been placed upon the changes in the thickness of the strata. The Scandinavian discussion concerning the dark earths has generally been focused on their disappearance. Many Scandinavian authors note that this type of thick occupational strata was formed between the 10th and 14th century (e.g., Andrén 1986). BeroniusJörpeland (1992) suggested that the high ‘growthrate’ of deposits, resulting in thick occupational layers, might indicate that settlement expansion was at its height during this period. Runer (1999) noted the difference between urban and agrarian development during Medieval time. While the occurrence of thick urban strata decreases from the 14th century, they seem to remain in many agrarian environments, which are interpreted as an indication of a ‘social integration’ into new hierarchies that occurred within towns but not in rural settings. In Lund and many other towns, the thinning of occupational strata from the 14th to the 17th century has been described as a decrease in ‘growth-rate’ (Andrén 1986). The thick, early occupational strata in Uppsala have been explained as being caused by a ‘semi-agrarian’ economy with large cattle stocks within the town (Ersgård 1986). The shift to thinner strata in Visby is explained by the increase in waste control (e.g., Westholm 1990, Nydolf et al. 1992). Also Andrén (1986) refers to the changes in waste handling when explaining this general shift within urban occupational stratigraphy. The increasing use of manure on ﬁelds during the 15th century may have decreased the amount of cattle faeces deposited within the urban environment, and the increasing use of pavements 17 J. Heimdahl and permanent plots may have made it easier to keep the town clean. Andrén also suggests that the wastehandling shift may indicate a changed cultural view on waste. This view was supported by Solli (1989) who suggested that the shift in the thickness of urban strata in Norwegian towns was caused by social and political changes. Accordingly, the ‘growth of layers’ may be a measure of a town’s capacity for organisation. It is interesting to note the differences and similarities between the Scandinavian and European discussions concerning dark earths. One of the main differences is what is considered status quo in the environment where dark earth is formed. At many urban sites where Roman towns preceded Medieval towns, the interlayering dark earth is considered as a break in an urban activity continuum. Apparently pedogenic processes acting on the debris of Roman buildings formed dark earth at some sites. Others have been formed by rural activities (Macphail et al. 2003). In Scandinavia, dark earth seems to indicate new development at a site. Apparently they are formed during the initial phase of the riddling urbanisation process. Naturally, this phase has never been referred to as an ‘abandonment’ or ‘decline’ – rather the opposite – since the preceding environment commonly has been a non-occupied site, even though the later urban phase with its thinner and better-deﬁned stratigraphical units is considered to be formed during a more advanced phase. The similarities in these discussions are found in the ideas concerning speciﬁc activities and processes that may have caused the formation of dark earth. Cultivation, grazing/cattle handling, middening/waste ground use or less intensive occupational activities have been common interpretations in both the European and the Scandinavian debate. The methodological approach to dark earth and obscured stratigraphy varies between the traditions in the different countries. In England and France micromorphological studies have been used (e.g., Corty et al. 1989, Gebhardt 1997), for example in London (Corty et al. 1989), Worchester (Dalwood 1992) and Paris (Cammas 1997). In Sweden there has been no tradition of incorporating micromorphological studies in archaeology so far. The exception is the studies of soil proﬁles from cultivated terraces in the counties of Halland and Närke and the occupational strata of Birka (mentioned in Håkansson 1995 and Håkansson 1997). Unfortunately, the results of those studies have not been published thus far. Measurements of various biological, chemical and physical parameters are applied in geology in order to make stratigraphical interpretations. Some 18 of this multiproxy approach has also been applied to trace the possibly obscured stratigraphy of dark earths. One approach has been to determine vertical or horizontal compositional variations (Sidell 2000). Chemical analysis and measurements of magnetic susceptibility may also be useful when separating dark earths formed in urban and rural environments (Macphail et al. 2000). A clear indication of hidden strata are the remnants of constructions found in situ within the homogeneous stratigraphical unit, resting on a former surface that now invisible in sections. The stratigraphy of urban constructions From an archaeological point of view, the most important components in urban occupational layers are the physical remains of the town itself – its buildings, streets and yards. In the discussion concerning what part of the occupational deposits are intentionally or unintentionally created (Järpe et al. 1979; Andrén 1985, 1986), these constructions are considered intentionally created deposits (and may also be referred to as ‘manifest’). It is important to realize that medieval houses, and parts of houses, were often constructed during different time periods (e.g., Eriksdotter 1996) and that old structures rarely have been left untouched by later activities. Therefore, a building remnant may represent building phases that span over several decades. Parts of buildings preserved in urban strata generally consist of the lower units – the foundation, which in turn may consist of ﬁll, stones, sills, posts/ postholes and sometimes ﬂoor remains. Foundations or parts of ovens and kilns are also sometimes preserved. In archaeological studies, the focus is generally placed on documenting the structural proportions of house remains in order to interpret the type of building and its function, and the spatial distribution between different buildings in order to locate backyards, lanes and streets (Augustsson 1992). In order to interpret building techniques, there has been a focus on the parts that most clearly reveal the upper structures of the building, for example the arrangement and styles of sills and postholes (cf. Goodburn 1995). The foundation layers of houses, which generally are the remains that built up the occupational layers, have been studied in terms of their general morphology. Only few, if any, studies have until now focused on the material in those foundations, the ﬁll that composes them. Since they often contain reworked material, ﬁnds within the ﬁlls are difﬁcult to interpret and relate to time and place dependent contexts and are therefore considered as be- Urbanised Nature in the Past Figure 8: The collapse of a building controls the extension of debris layers deposited above the foundations. Examples of collapses are: A) Outward collapse, B) Inward collapse and C) Collapse through tipping. Samples taken from building debris (grey) may therefore be plotted as being outside the building, since constructions commonly are marked on excavation plans by the extension of their foundations (dotted lines). ing of limited archaeological value (e.g., Roskhams 1992, 2001, Tagesson 2000a). In some cases, the upper parts of buildings are preserved in the stratigraphy. This is especially evident when buildings have been burned but sometimes also deliberately demolished, generally in order to make space for other constructions. These types of remains are preserved as depositions of house debris consisting of fragments of the former houses and in some cases, their content in the form of charcoal, wood, fragmented bricks, chalk, mortar, glass, stones and pieces of turf. Sometimes it is possible to identify layers of different composition within the house debris, which may make it possible to understand which materials composed the different units of the building. When dealing with the stratigraphy of collapsed constructions, it is important to consider how the building came down (e.g., Keevill 1995, Robins 2003). Did the building collapse on its foundation, or did it tip over in the process? If the building tipped over, or if its walls fell outwards, the house debris would also be distributed outside its foundations (Fig. 8). The investigation of how a collapse of a building took place may also reveal evidence of the building techniques (Keevill 1995). When new buildings were constructed in medie- val towns, it was common to use debris from older buildings as ﬁll in the new foundations. Augustsson (1992) states that building materiasl in Swedish towns were limited, which may be reﬂected in the habit of recycling house debris as ﬁll. A special valuable type of stratigraphical remains of destructed buildings is the traces of large city ﬁres, which often were well documented in historical records (Fig. 9). Residues of such ﬁres may be preserved as extensive and continuous charcoal horizons that contain building remains and objects affected by ﬁre (melted glass, burned clay etc.). These layers are often possible to identify even if they have been truncated. If such identiﬁcation is possible (and supported by artefact typology) the dating of the strata can be pinpointed, sometimes down to hours. Also streets, lanes and yards may be preserved within the stratigraphy, both as deposits and as horizons. When cobble streets and yards are preserved in urban occupational layers, they generally consist of a settling material (commonly sand) into which the cobblestones are pressed. Sometimes the stones have been removed and the settling horizon may then be recognised by its surface morphology, with preserved imprints of the stones. At waterlogged sites, wooden boardwalks are sometimes found. The most common types of streets, lanes and yards in Scan- 19 J. Heimdahl Figure 9: Fire could be extremely destructive among densely built timber constructions and many Medieval and Post Medieval towns were burnt down at several occasions. Traces of large town ﬁres are sometimes preserved stratigraphically and may be used for dating. dinavian towns were probably naked, tramped soil without any paving at all. Remains of such surfaces may be recognised by the compacted topsoil or the tramping horizon. Old walking grounds may also be recognised by the orientation of artefacts and other objects that almost exclusively are deposited in horizontal positions. Finds of former tramping grounds are generally considered as being of high interpretative value. Objects found on those grounds may be considered as primary deposits, and can therefore be connected to a time and space dependent context (Tagesson 2000a). Another type of urban construction is the largescale ﬁll that was used to transform ground surfaces within the town. Examples of this are the rearrangements or damming of rivers, ﬁlling and foundations of banks or shores, and ﬁlling in depressions (Beronius-Jörpeland 1992). In order to transform landforms or prepare an area for building, negative forms such as pits, cavities, hollows and truncations were also created (Yule 1992, Clark 2000, Tagesson 2000a). In geological terms the contact between the truncated material and a later deposition is ‘erosive’ (indication of hiatus). In archaeological terms it is sometimes called a ‘truncation horizon’. The stratigraphy of waste ‘Waste’ would possibly be the most common answer to the question of what mainly composes urban occupational layers, and probably many archaeologists and geologists would agree on this point. There are many reasons for this. Waste may be considered a term for unwanted material. What is considered as ‘unwanted’ in a speciﬁc situation is, of course, dependent on social conventions. Things that are ‘waste’ in one situation or in 20 a speciﬁc social context, may be considered valuable and useful in another. In Scandinavian medieval towns, the things that were interpreted as waste can be classiﬁed in different groups: by-products, broken or used objects, and faeces – which also may be considered a type of by-product (Fig. 10). For instance, by-products (or waste products) may be wooden chips created during carpentry, bones during the cutting of meat, or ash and charcoal from hearths. This type of waste is the material that in one way or another was originally brought into the town from the surroundings. Another type of by-product has its origin within the town itself, for example natural (or cultural) sediments that may be considered as waste when being redistributed by digging. Besides by-products, broken, used or old products will often also be considered waste. It is the waste of past times, the small things forgotten, that in our times will consist the absolute majority of the ‘artefacts’ studied by archaeologists (cf., Deetz 1996). A third type of waste is faeces from humans and animals. Cattle, horses, pigs, sheep, goats, dogs, cats and fowls were kept on the streets during daytime. This is the reason that dung constitutes a major part of the organic compound in many urban occupational layers. Faecal material from animals is often found in urban refuse hips, which indicates that this material was removed from the streets. Human faeces are generally found concentrated in latrines and on dump piles where latrine buckets or chamber pots were emptied. There are many popular myths concerning waste handling in Medieval and Post-Medieval times and many have a mental picture of the medieval towns as extremely dirty. Numerous authors have described how household waste was dumped directly on the street, which has contributed to the idea of the medieval town as a constantly growing waste pile – an Urbanised Nature in the Past idea that in turn has been used in order to explain the presence of urban occupational deposits, since they have been interpreted as the actual waste pile. Archaeologists initially supported this picture since it was impossible to ﬁnd traces of cleaning and truncation, according to earlier excavation techniques (Larsson 2000). During the latest decade this picture has been adjusted. Partly because of the introduction of SCR, it has been possible to interpret patterns of refuse removal within urban environments. Speciﬁc dumping places and dumping pits are often identiﬁed, and an organised system of waste dispersal with wooden barrels seems to have been in use in many towns. The systematic cleaning in Medieval and Post-Medieval towns is also supported by written sources (Beronius-Jörpeland 2001). Urban production of waste during the Medieval and Post-Medieval times was limited compared to later times and the recycling of material was probably common (cf. Drangert & Hallström 2002). Organic by-products from cooking was used to feed pigs, wooden waste was often used as fuel and when things could not be recycled in other ways, they could serve as ﬁll in foundations for new constructions (Beronius-Jörpeland 2001). In many towns there are examples of how waste was used as ﬁll in the construction of extensive land forming, for instance the foundation of new waterfronts in Stockholm during the 14th century (Hansson 1970, Dahlbäck 1982). Urban pedogene soil processes Alongside the depositional processes, like construction and waste handling, the urban environment and its associated activities also led to the formation of strata accorded to post-depositional process and formation of horizons. The soil horizons formed in urban sites can be developed by natural pedogene processes that occur independently of the urban environment and also by processes directly linked to the urban environment. Tramping by humans and domesticated animals, grubbing of pigs and pecking of fowls cause culturally induced bioturbation in the urban topsoil. At the same time horizons may form by natural bioturbation, for instance caused by roots, earthworms and other soil organisms. A majority of the topsoil horizons in towns were probably truncated, but sometimes they were buried and thereby preserved. In the south Swedish lowland two orders of ‘natural’ pedogene soils dominate: Spodosoles (Sw: podsoler) and a type of Entisoles (sometimes also referred to as Inceptisoles), (Sw: brunjordar). The Entisols are developed in grasslands and broad leaf forests through earthworm bioturbation (Fig. 11). Organic litter on the ground is mixed with the upper part of the minerogenic material that constitutes the sub-soil. This leads to the formation of A and B-horizons above the unaffected C-horizon: The surface A-horizon is characterised by humiﬁed organic matter mixed with Figure 10: Deposition of waste occurs through different degrees of awareness. A) By-products, such as wood chip litter, e.g. created through carpentry. They are unintentionally accumulated as litter on the ﬂoor or ground and later mixed with the topsoil through post depositional tramping. B) Minerogenic waste created as a by-product from digging. They can be left as unintentionally created waste piles, be dumped intentionally in waste piles or be reused as ﬁll. C) Objects may be dropped accidentally and are then considered waste by mistake. If new protecting layers do not cover the surface, the objects dropped can be either tramped into the ground or cleaned away and deposited in waste pits or piles. D) Waste from e.g. households was commonly and intentionally thrown in waste pits or given to pigs. 21 J. Heimdahl Fig 11: Typical Entisol (Sw. brunjord), developed in Middle Swedish grassland. It consists of three horizons. The A-horizon, a surface horizon of humiﬁed organic matter mixed with minerogenic particles. The B-horizon, a subsurface minerogenic horizon enriched with organic matter. The C-horizon, the unaffected minerogenic mother material. minerogenic particles and the subsurface minerogenic B-horizon is characterised by the enrichment of organic matter. If an area, for instance, a vacant lot within a town is left untouched, it generally leads to the establishment of ruderal plants and earthworms begin to thrive. After some years this will develop natural A and B-horizons. Such horizons (sometimes partly truncated) are commonly found buried in archaeological sites and indicate old ground surfaces that have been open, and relatively undisturbed, for several years. Human interaction with soil forming processes is sometimes considered as ‘disturbance’ but soils formed due to anthropogenic activity may also be classiﬁed as its own order: Anthrosols (Gilbert 1989) that may include soils affected by ploughing, digging, tramping, or chemical pollution. Individual horizons within soils, affected by ploughing (often rich in phosphate due to fertilizing) may also be called ‘anthropic’ (e.g., Brady 1990). However, the classiﬁcation of Anthrosols is generally adapted to modern soils (Gilbert 1989) and the letter system of classifying different horizons within soils is not sufﬁcient when classifying horizons in urban archaeological contexts. Instead the individual horizons are referred to by the descriptions of the activity that created them such as: ‘tramping horizon’, ‘earthworm horizon’, ‘root horizon’ or in combination such as ‘tramping horizon, affected by earthworms’. 22 The two major horizon-types formed due to cultural activities found in urban occupational deposits are tramping horizons (developed by humans and animals) and cultivation horizons (developed through ploughing and digging). Cultivation horizons, or cropmark horizons, are sometimes difﬁcult to separate from horizons developed through cattle tramping. Their content and appearance in cross section may be similar. Both contain a mixture of cattle faeces and minerogenic material, and both may be developed to similar depths (c 10-20 cm). If studied in detail in the ﬁeld, they may be separated by different features in their contact with the unaffected substratum, for example in cultivation horizons the contact may be impressed by marks from ploughs or shovels (e.g., Petersson 1999, Hedwall et al. 2000, Karlenby 2002). Ploughing is a very rough mechanical force that generally destroys remnants of buildings and structures within the affected horizon, and archaeological interpretation in plough horizons is difﬁcult, although possible (Alexander & Armit 1993, Sarnäs 2004). Tramping, on the other hand, may be gentler, and the turbative force is not as large but still has an effect on objects within the affected horizon. Experiments have proven that tramping effects by humans, especially in sand, may cause dispersal effects on artefacts, not only on the horizontal plan but also in their vertical displacement (Stockton 1973, Villa & Courtin 1983). Another soil affecting process may be heating of the ground, for example below hearths or burned buildings (Corty et al. 1989). The effect of earthworms is even smaller than tramping but still these organisms can be capable of moving larger objects, such as bones and stones. The effects of earthworms on artefacts and archaeological sites have been investigated several times (e.g., Armour-Chelu & Andrews 1994; Grave & Kealhofer 1999; Lawson et al. 2000; Davidson 2002). Traces of earthworms may be found either by the structural analysis of the soil, by macrofossil ﬁnds or by worm tracks in the sediment. Earthworms’ faeces are composed of mineral particles and humic material that stick together in aggregates. These aggregates form a structure called ‘crumb microstructure’. Generally they are easy to identify with ﬁeld observations but sometimes micromorphological studies are required (Corty et al. 1989). A soil that contains crumb microstructures obtains speciﬁc properties, for instance it becomes much more resistant to erosion by rain and alluvial runoff (Brady 1990). The cocoons formed on the worms’ clitellum constitute the most common remains of earthworms (Fig 12). They are easy to ﬁnd and identify during macrofos- Urbanised Nature in the Past of water along the roots also leads to the transport of oxygen. The rate of decomposition can ﬂuctuate through time due to changes in climate and local environment, like nearby excavations. Excavations lead to the drainage of surrounding strata, which then increases decomposition rate. This in turn may lead to the settling of the ground and the destruction of buildings. The processes and effects of decomposition of occupational strata have been studied several times (e.g., Borg 1993, Oxley 1993, Gardelin 2002, Lagerlöf & Nord 2002, Wikström 2003). 4 Urban Plant Ecology Figure 12: A) Earthworm cocoon formed on the clitellum (jointed segments at the anterior third of the body of oligochaetes). B) Earthworm traces obscuring the contact between two strata. sil analysis, and may reveal the former presence of earthworms even if a crumb microstructure has not been identiﬁed. Tunnels burrowed by earthworms are identiﬁable by characteristic tracks. They are especially easy to identify when the worms have moved from one strata into another since the material from the old strata will occur as tunnel ﬁll in the new strata (Fig. 12). If earthworms occur in and between several strata separated by sharp contacts, these contacts will gradually become obscured. This makes it possible to exclude bioturbation by earthworms between strata whose contacts are sharp and undisturbed. Another important post-depositional effect is the decomposition of organic material in the depositions. The decomposition occurs through activities by microorganisms, mainly fungi and bacteria. This activity is largely controlled by the access to oxygen, which in turn is controlled by the amount of water in the material. Decomposition activities are therefore low in waterlogged material. The presence of water depends on the ground water table and the capillarity of the material. Richness in organic material increases the ﬁne-grained fraction through humus particles and the water is easily kept within the strata. Preservation in coarse-grained material, like gravel and stones, is generally poor if it is not situated below the groundwater surface. Preservation along roots is generally poorer because the transport From a perspective of botanical ecology, the urban environment is a mosaic of ecological niches, providing habitats for a multitude of communities characterised by species with different strategies. The urban environment has changed in many ways since the Medieval and Post-Medieval times. One of the most important changes is the urbanisation of the rural economy. The end of animal husbandry within towns and the disappearance of horse driven transport have decreased the content of nutrients in the systems. The transport of seeds and fruits from meadow and pasture plants by animal faeces and ﬁeld weeds from crops has almost ceased. Streets and yards are almost exclusively covered with asphalt or stones and few cultivation plots for kitchen plants remain. On the other hand, cultivation for esthetical or recreational purposes has increased since the mid 19th century when public parks were introduced (c.f., Nolin 1999). In those parks, plant communities of cultivated species and weeds co-exist. There are also examples of urban environments that remain relatively unchanged. Past and present processes of soil exposure are probably similar. Loosely packed and exposed soil provides a niche for r-strategists (cf., Hitchmough et al. 2001) – pioneer plants adapted to quick environmental ﬂuctuations by a rapid reproductive capacity and the capacity to form seed banks (Begon et al. 1996, Crawley 1997). Plants that are resistant to mechanical disturbances, such as tramping, ﬁnd their niches in both past and present towns on paths and between cobblestones on streets and yards. It is also common to ﬁnd refugias of natural environments in towns (Gilbert 1989, Botkin & Beveridge 1997). Along cliffs and water lines, it is also common to ﬁnd a ﬂora dominated by K-strategists – competitive plants with long life cycles. Present urban environments are often characterised by a large number of exotic plants introduced both intentionally and unintentionally (Gilbert 23 J. Heimdahl 1989). Gardens and parks are especially rich in foreign plants, and it happens that imported plants naturalise in vacant ecological niches or take over niches by interspecies competition. Unintentionally introduced plants commonly spread in harbours and places where cargo is unloaded, for instance on market places. The high introduction rate of neophytes provides a stress to the urban ecology in general (Gilbert 1989) and it has been observed that many alien species become a part of the urban seed bank (Kostel-Hughes et al. 1998). The introduction pattern of alien taxa has changed since the Medieval and Post-Medieval times. Both intentional introduction of cultural plants and unintentional introduction have increased in the last centuries (Lundqvist 2000). Proportionally, the intentional introduction has increased much more than the unintentional introduction because of the large quantities of cultural plants. Although the early towns were probably rich in neophytes and past urban environments should also be suitable to study from this perspective (cf., Preston et al. 2004, Wittig 2004). Urban fossil ﬂoras commonly consist of a mixture of plants derived from different environments (see Chapter 5). This mixture may be analysed through the ecological grouping of those species that can provide important information about both local and regional environments (Latalowa et al. 2003). Earlier studies of urban archaeobotany in Sweden and neighbouring countries The material in this study has been compared with the results from other investigations of urban Medieval sites in Sweden, and also from Finland, Norway and Denmark where archaeobotanical methods have been applied (Fig. 13). Some results are unpublished or exist as reports from local museums or university departments. In other nearby countries, such as Germany, Poland and Great Britain, urban biostratigraphical records have been more frequently studied and published than in Scandinavia. Examples are the investigations in Cracow (Wasylikowa 1978), Göttingen (Hellwig 1995), Kiel (Wiethold 1995), Lednicki (Polcyn 1995), York (Kenward & Hall 1995), Wroclaw (Kosina 1995), Dublin (Geraghty 1996), London (Giorgi 1997), and Elbag and Kołobrzeg (Latalova et al. 1998, 2003). Most archaeobotanical studies in Sweden have focused on prehistorical sites but there are some examples of the studies of plant remains from Medieval and Post-Medieval urban sites. Urban archaeobotanical work has been performed on plant material 24 from 11th – 13th century Lund, 12th-14th century Uppsala (Hjelmqvist 1963, 1991, Regnell in Carelli & Lenntorp 1994, Regnell in Carelli 1995, Påhlsson 1983, 1991a, 1991b) and 16th century Gävle (Elfwendahl & Påhlsson 1991). During the archaeological excavations at Helgeandsholmen in Stockholm, there was a close collaboration between the National Heritage Board (RAÄ) and the Geological Survey of Sweden (SGU). Palaeoecological reconstructions, based on diatom and pollen ﬂoras, were performed by Miller and Robertsson (1982) and the plant macrofossils were studied by Grifﬁn (1983) and Berggren (1984). In Stockholm, studies have also been carried out on the material from the Post-Medieval area of ‘Kvasten’ (Heimdahl 1999b) and from the park ‘Humlegården’ (Hansson & Dyhlén-Täckman 1999). Urban archaeobotanical studies in Göteborg have been performed at four Post-Medieval sites (Larsson 1985, 1986). Geoarchaeological studies in a wider sense have been carried out around and in the Viking Age town of Birka (Miller & Clarke 1997 and Risberg et. al 2002), including archaeobotanical analysis (Hansson 1995, 1996, Hansson & Dickson 1997, Heimdahl 1999a). Birka, however, differs from other Swedish urban contexts since the occupational deposits are not waterlogged and the preservation of organic remains is poor. Much of the study has therefore focused on the clay-gyttja deposited offshore, just outside the town. This is similar to the sediments of the former lake Fatburssjön, that represent depositions from the Bronze Age to the 19th century (Robertsson et al. 1995). In addition, earlier analyses of plant macrofossils of samples from the 16th -17th century have been performed in Norrköping (Ranheden 1999). Macrofossils were also sampled in the castle of the Medieval town of Vadstena (Ranheden in Hedvall et al. 2000). Macro remains from the Medieval town of Sigtuna (9th-13th century) were studied by Engelmark (2002). Urban palaeoenvironmental reconstructions in Finland have been undertaken in the Helsinki area using different biostratigraphical methods (Vuorela & Lempiäinen 1993). In Turku macrofossil analyses were carried out by Aalto (1994) and Lempiäinen (1985, 1988, 1994 &1995). During the reconstruction of the environmental history of Turku, a combination of palaeoecological methods (pollen, diatoms and plant-macrofossils) was used (Vuorela et al. 1996). The Finnish data is probably well suited for comparison with the Swedish sites since Finland was strongly inﬂuenced by Sweden during the Medieval and Post-Medieval times as it was under Swedish administration during AD1323-1809. Extensive geoarchaeological and palaeoecological Urbanised Nature in the Past 13 Sweden Norway 60 Finland 14 12 11 1 2 3 4 15 Estonia 16 5 6 7 Denmark 10 55 9 8 10 Baltic Sea 20 Figure 13: Location of urban sites, which has been archaeobotanically investigated, and are here compared with the results of this thesis. 5. Norrköping 1. Uppsala Grifﬁn et al. (1988) Ranheden (1999) Påhlsson (1983, 1991a, 1991b) Viklund (2000) 12. Bergen 6. Vadstena Krzywinski et al. (1983) Ranheden in Hedvall et al. (2000) 2. Sigtuna Krzywinski & Kaland (1983) 7. Göteborg Engelmark (2002) Hjelle (2002) Larsson (1985, 1986) 3. Birka 13. Trondheim 8. Lund Hansson (1995, 1996) Grifﬁn & Sandvik (1989) Hjelmqvist (1963, 1991) Miller & Clarke (1997) Sandvik (1992, 1995) Regnell in Carelli & Lenntorp (1994) Hansson & Dickson (1997) Regnell in Carelli (1995) Heimdahl (1999a) 14. Turku Risberg et al. (2002) Lempiäinen (1985, 1988, 1994, 1995) 9. Copenhagen Vuorela et al. (1996) Moltsen (2002) 4. Stockholm Aalto (1994) Miller & Robertsson (1982) 10. Svendborg Grifﬁn (1983) 15. Helsinki Jensen (1979) Vuorela & Lempiäinen (1993) Berggren (1984). Robertsson et al. (1995) 11. Oslo 16. Tartu Hansson & Dyhlén-Täckman (1999) Grifﬁn (1979) Heimdahl (1999b) Sillasoo (2002) work has been carried out in urban medieval contexts in Norway, for example in Bergen (Krzywinski et al. 1983, Krzywinski & Kaland 1983). Investigations speciﬁcally focusing on plant macrofossil anal- ysis have been made in Oslo (Grifﬁn 1979, Grifﬁn et al. 1988) and Trondheim (Grifﬁn & Sandvik 1989, Sandvik 1992, 1995). Denmark has a long tradition of archaeobotanical 25 J. Heimdahl Figure 14: Examples of the regional origin of plant diaspores recorded in an urban environment. investigations of Medieval contexts. However, during the Medieval Time Denmark was heavily inﬂuenced by northern continental Europe, and the Medieval Danish archaeobotanical material is more comparable to the German or Polish than to the Scandinavian material. Results from Svendborg (Jensen 1979), København (Moltsen 1999, 2002) and a summary of plant macrofossil data from the Medieval Danish sites (Karg and Robinson 2002) were used as northern European case studies in this investigation. 5 Taphonomy of urban macro remains The study in this thesis is largely built on the analysis of plant macrofossils. In order to understand a fossil record, both palaeoecological and taphonomic aspects must be considered. Taphonomy of urban material is especially complex, and it is therefore necessary to include it as a background for palaeoecological analyses and interpretation of results. Taphonomy includes all factors, both natural and cultural, that affect the ﬁnal setting of a fossil record, including post-depositional processes as well as the sampling and treatment of material in the laboratory. This account will mainly focus on the plant macrofossil material – seeds, fruits and other diaspores. A more extensive taphonomic account of pollen grains can be found in Vuorela (1999). I have chosen to divide the taphonomic factors into seven groups: biology; deposition; human inﬂuence; preservation; reworking; sampling, preparation and registration; and personal bias. 26 1. Biology Life strategies of plant species vary. K-strategists have a long and less effective reproductive cycle than r-strategists (e.g., Begon et al. 1996, Crawley 1997). K-strategists more commonly form wood tissue and have thicker roots. Differences in seed dispersal and seed production (Howe & Westley 1997) will inﬂuence the fossil record. Some plants are dioicic (separated sexes of the individuals) and only female, fruit setting plants will be registered through seed or fruit identiﬁcation. Plants of the same species may act differently from time to time depending on environmental factors. They may produce different amounts of seeds or set fruits in different periods, and thus give signals of varying magnitude in the fossil record. Since plant macrofossil analysis commonly deals with diaspores (as they are relative to other plant tissue that are both resistant and easier to determine to species), we will come across plants that have been going through seed setting. This also means that male individuals (in the case of dioic plants) will not be found in this fossil record. It is also important to take into account that seed and fruit production varies between different plant species in quantity, season and frequency. 2. Depositional processes and ecology in urban environments Depositional processes within the urban environment are complex (Fig. 14 and 15) (cf. Thomas 1997). In this discussion, the fossil plant record has been di- Urbanised Nature in the Past vided into four groups, based on taphonomic origin of the plants: A: the local ﬂora; B: the regional ﬂora; C: plants from pastures and meadows, and D: cultural plants. These groups can in turn be subdivided as discussed below. A: The local ﬂora The local ﬂora is considered to include plants, which have grown in the immediate vicinity of the investigated area. The Medieval and Post-Medieval urban environment was characterised by continued spatial rearrangement of buildings and structures, and by the large production and deposition of waste and nutrients. Soils were exposed and covered by new depositions rapidly, both by human activities and large-scale domestic animal bioturbation (e.g., pigs). In this type of environment, alluvial and colluvial processes lead to the exposure of soils. These environmental factors will disturb plant communities, and small-scale migration of plants will occur between patches in the area. The in situ urban ﬂora with ruderal plants and weeds has pioneer strategies with a high capacity of dispersal and low competition ability, according to trade-off effects (e.g., Crawley 1997). This means that these plants often have a rich production of seeds and can form seed-banks or seeds that are easily dispersed (Rees 1997, Howe & Westley 1997). The Medieval urban ﬂora generally developed different biotopes directly connected to the soil types within the site (Gilbert 1989). These biotopes can be divided into eight groups (Fig. 15): Figure 15: Local urban ﬂora environments (1-8) and transport of material and seeds within the town (A-I). 1. Tramping resistant ﬂora on streets and in yards 2. Loose soil ﬂora 3. Nitrophilous ﬂora 4. Turf roof ﬂora 5. Refugal ﬂora on cliffs, etc. 6. Harbour ﬂora, neophytes etc. 7. Weeds in local cultivation plots 8. Flora thriving along walls, etc. A. Seed dispersal through footwear and wheels B. Import of (sometimes exotic) seeds C. Storing and processing of seeds before marketing D. Export of seeds E. Cleaning of streets and yards, dumping of waste F. Cleaning of dung from streets and yards, used as fertilizer G. Dung from stables H. Stored and processed crops I. Processed cereals used within the town or exported 27 J. Heimdahl 1. Tramping resistant communities occurring on pathways, backyards and in streets. separated if their occurrence is connected to other species that are speciﬁc to certain local or regional environments. 2. Loose soil communities growing on loosely packed debris and soil stacks. C: Plants from pastures and meadows 3. Nitrophilous communities found near latrines, waste piles, dung heaps and along house walls and ditches, where moisture and nutrients are concentrated in the soil. 4. Turf-roof communities will occur if roofs are overgrown with e.g., grass or plants selected for ﬁre protection (Krzywinski et al. 1983). 5. Remains of the pre-urban natural ﬂora may occur as refugium biotopes, for example river and seashore ﬂora, or plants growing on cliffs or rocks. 6. Harbours where cargo is handled may act as a ﬁrst environment for neophytes. Floras in harbours and wharves often contain speciﬁc species. 7. Weeds and kitchen plants may grow on cultivation plots within the town. Cultivation within and just outside towns is documented in historical sources. 8. The ground adjacent to walls may act as a substrate for weeds. The environmental factors along walls vary according to geographical position, house type, street-type etc. B: The regional ﬂora In this study the regional ﬂora is deﬁned as all ruderal and natural plants occurring outside the excavated area, both outside and in the town. The regional ﬂora is mainly represented by plants along roads and in the nearby town blocks. These plants can be spread by wind or by attachment to animals, humans (Clifford 1956) or wheels (Millberg 1991). Seeds transported in this way are, however, difﬁcult to differentiate from seeds grown in situ, since they are often derived from identical biotopes. Sometimes they may also arrive from different environments, such as forests and wetlands. Many plant species thriving in urban areas can also grow as weeds in cultivated ﬁelds. Among the ﬁnds from Norrköping and Karlstad, there are also plants that are common in many different environments and thus are hard to correlate to any speciﬁc habitat. This is especially relevant for the Chenopodium album-type and also the species Cirsium arvense, Fumaria ofﬁcinalis, Galeopsis tetrahit/G. biﬁda, Lithospermum arvense, Solanum dulcamara and Stellaria media. These plants could have been growing in ﬁelds, along roads and pathways, in the town or locally within the investigated area. They may be 28 Grazing by cattle can lead to a concentration of plant remains from pastures and meadows in urban areas by deposition of faeces and hay. This taphonomic factor plays a major role in most Medieval and PostMedieval urban sites. Cattle sometimes also graze ruderal plants and weeds, and some of the ruderal plants are abundant in pastures as well. If isolated coprolites (in this case, faeces that are distinguished as coming from one individual) are found, it may be possible to determine the species of the animal that produced it by the morphology or species-speciﬁc parasites. Plant macrofossils in coprolites can also provide important information on the grazed area, or during which season the grazing or cutting of the hay occurred (Moe 1996, Moltsen 1999, Karlsson 2000). D: Cultural plants Cultural plants are referred here as plants that are known to have been regularly used by humans. They could have been collected from a wild or cultivated population and might have been brought into the town for trade, storage and/or processing. If a cultural plant is represented, both in the pollen ﬂora and by macrofossils, it may be an indication of regional cultivation. This is not always the case since pollen grains can also be brought into the town in an artiﬁcial way (cf., Faegri and Iversen 1989). Cultural plants are generally processed in speciﬁc ways which may affect the fossil record. For example, harvested cereal crops are processed through different methods of threshing, sieving and cleaning. Fossil records of crops may differ according to which steps have been taken in their handling. (cf., Hansson 1997, Viklund 1998, Hambro Mikkelsen 2003). 3. Human inﬂuence Humans can selectively have affected plant communities due to cultural habits that may be unknown to us. Some species may have been cleaned away or favoured for different reasons, for example due to unpleasant characteristics (like stinging or burning tissues), pleasant characteristics (like beauty), or due to mythical ideas (like luck or misfortune). Smallscale cultivation or the collecting of species, according to religion, tradition, medicine and decoration, Urbanised Nature in the Past must also be considered (cf., Johnson 1999). the biotopes for plant remains found in this type of deposit must be handled with care. 4. Preservation of plant macrofossils Macrofossils are preserved by water logging and occasionally by carbonisation through heating. Different preservation processes of the same original plant community will lead to different fossil records. Water logging generally leads to a wider and more complete spectra of preserved plant remains but there are exceptions. Cereal grains and wild grasses are more easily preserved by carbonisation, while fruits of e.g., the Apiaceae family are much better preserved by water logging (Gustafsson 2000). Another example is the differences between r- and K-strategists. Since r-strategists often have the capacity to form seed banks, the seeds tend to be bigger and more hardcoated than seeds from K-strategists. This may lead to that seeds from r-strategists are over represented in materials with bad preservation status. At Norrköping and Karlstad, about 95% of the counted macrofossils were preserved as un-carbonised specimen, since the embedding medium consisted of a dense, ﬁne-grained deposit with high capillarity and low permeability. Only 5% of the material consisted of carbonised diaspores, concentrated mostly on the old surfaces in connection with ﬁreplaces or debris from burnt houses. Samples with carbonised material were generally lacking un-carbonised material and vice versa but in some cases they occurred together. This may indicate different taphonomic origin of the plant material and the two groups were interpreted separately. 5. The reworking processes The reworking of deposits in urban environments is usually extensive. It may affect the preservation status of the fossil content by mechanical, biological and chemical deterioration. This can lead to the relative concentration of hard macrofossils in the redeposited material, (given that they also occur during the reworking process or are present in the mixed material). This also makes robust macrofossils less suitable to use for reconstructions since they can be reworked many times and still remain well preserved. For the same reason, we can put more trust in fragile material. Some samples from ‘Konstantinopel’ were characterised by high concentrations of plant macrofossils with the same size as the medium sand fraction (0.20.6 mm). This occurred in well-sorted, water deposited alluvial material, dominated by silt and is probably an effect of sorting by water. Conclusions about 6. Sampling, preparation and registration Additional taphonomic factors occur by sampling, storing, processing and ﬁnally by analysing the fossil content of the samples. For example, the ﬂotation process is a method that works differently with different sedimentological mediums, and it is a good idea not to handle all samples in the same way (Wright 2005). During wet sieving, the choice of sieve size will have an effect on the composition of the retrieved material, which in turn may affect ecological interpretation (e.g., Birks 2001; Zohar & Belmaker 2005). During analysis, it is also important to have a clear classiﬁcation and quantiﬁcation strategy for fragmented material, and for the comparison of the different anatomical features preserved as plant macrofossils. How do we compare leaves with seeds? 7. Personal biases The ﬁnal taphonomic factor is connected to the person performing the analysis. It is therefore important to ask the question: what are my personal biases? It appeared to me that I ﬁrst observed certain types of fossils. These were usually the most common types of moderate size (1-2 mm) like Chenopodiumtypes, Carex sp. and Urtica sp. This was probably due to their high abundance within the samples and that they were easily noted with the magniﬁcation used (c. ×8-15). But it also may have been a kind of “favouring the expected”. To avoid this, the samples were analysed on a number of occasions to ensure that the more unusual or discrete fossils (e.g., Poaceae-fruits) were identiﬁed as well as the more common remains. The trade-off of favouring the expected is to favour the unexpected. This seemed to happen when I made unusual or exotic ﬁndings. This then led to attempts to ﬁnd more fossils of the same type, which perhaps made me ignorant to the more common ﬁndings. Again, to go through the samples repeatedly was my strategy to avoid this bias. 6 Study sites Norrköping (Paper I, II, IV, V) Norrköping is situated along the river Motala Ström, between Lake Glan in the west and a bay of the Baltic Sea, Bråviken, in the east (Fig. 16). The excavated 29 J. Heimdahl Figure 16: A) Map over Norrköping. The investigated area is marked with black in the upper left corner of the marked block. The present river Motala ström is shown in dark grey, and the light-shaded area marks the former extension of the stream before the rearrangements after AD 1640. B) Simpliﬁed map of Quaternary deposits in the area (After Bergström and Kornfeldt 1973). Till and bedrock are marked with triangles, glacioﬂuvium with dots, ﬁne grained marine sediments with white, and light grey marks the alluvial deposits of the river (dark grey). site is located in the centre of the town (Fig. 16). The bedrock consists of migmatised granites, gneisses and mica schists (Bergström & Kornfält 1973) and is partly exposed in the highland area southwest of the town. In the east, thicker Quaternary deposits cover the bedrock. About one kilometre south of Norrköping, the deposits are generally dominated by different littoral sediments (Fig. 16). The eastern lowland is characterised by ﬂat areas with post-glacial clay and silt underlain by varved glacial clay and silt and southwest of the town, the landscape is hummocky and consists of till and bedrock outcrops. An esker runs straight through the town from northwest to southeast. The investigated site is situated where the river Motala Ström cuts the esker. The glacioﬂuvial deposits are surrounded by coarse silt and till. The slope within the excavated area is part of the river valley of Motala Ström, and the esker itself does not seem to affect the local topography (cf., Lindgren-Hertz 1999). Downstream, the riversides are dominated by alluvial deposits, from clays to coarse silt rich in organic material. The distribution of this alluvium is poorly known, especially the area along the southern riverside, where the investigated area is situated (Bergström & Kornfält 1973). The investigation area is situated close to a bay of the Baltic and thus the shore displacement must be considered. Present land uplift in the area is 3.0 mm/yr and according to a relevant shore displacement curve (Persson 1979), the rate has probably not changed signiﬁcantly during the last 3000 yrs. The altitude of the shoreline of the Baltic Sea during the 30 13th century can be estimated to be about 2.4 m a.s.l. The present surface of the excavation site is located about 9 m above present sea level and the substratum of the oldest occupation layers is situated about 7 m a.s.l. Since the Medieval times, the site has not been affected by the regressive sea level. The elevation of the earliest medieval ground surface in the excavated area was about 4.5 m a.s.l. The ﬂat landscape is suitable for cultivation. Humans have been present in the area since the land in the vicinity started to rise above sea level about 2000 years ago. The river Motala Ström, which in this part is characterised by rapids and turbulent water, made the place attractive for salmon ﬁshing and ideal for establishing mills and therefore, it became a place of economic interest (Ljung 1965). Norrköping was ﬁrst mentioned as a trade centre in 1283, but it was not given its town privileges until 1384. Written sources indicate that during the Medieval times, Norrköping was an important centre for trade and administration, where both the state and the church had interests (Lindeblad 1997). During the middle of the 16th century, as a result of the reformation under the regiment of Gustavus Vasa, Norrköping became the most important Swedish harbour for the export of iron. The town expanded rapidly and during the 17th century, due to industrialisation, Norrköping became the second largest town in Sweden, next to Stockholm. The Medieval constructions in Norrköping have so far remained unknown because of the major reconstructions of the city plan during the 16th and 17th centuries, including a major reconstruction and Urbanised Nature in the Past ﬁlling of the river Motala Ström, which was made narrower and deeper (Fig. 16). Therefore, the oldest remains normally found during archaeological excavations are from the 16th century or later (Parr 1987, Kjellén 1996, Hållans et al. 1999). In the excavated area the preserved occupation layers are sloping towards the west and the former river course, which according to the rectiﬁed map of 1640 was very close, perhaps only some meters away from the border of the present investigation area (Fig. 1). Due to the sloping substratum, the layers are thicker and better preserved towards the west. The occupation layers disappear about 50 meters upwards from the western edge of the excavated area (LindgrenHertz 1998). Karlstad (Paper II and III) Karlstad is situated in southwestern Sweden on the river Klarälven delta that ﬁlls out a bay on the northern shore of lake Vänern (Fig. 17). The lake was isolated from the sea due to land uplift c. 7000 BC (Fredén 1988). Karlstad is the only town in Scandinavia that is located on an active river delta. Hummocky terrain with exposed and sparsely till-covered bedrock surrounds and partly penetrates the delta plain. Between 7000 and c. 2500-2000 BC the shoreline of Lake Vänern retreated ca 20 km southwards due to land uplift (Sandegren 1939, Heijkenskiöld 1981, Sundborg & Heikenskiöld 1972 and Fredén 2000) and the delta sedimentation moved southwards to the present area. During the last 2000, years the shoreline of Lake Vänern has remained more or less stable and local changes are mainly due to a continuous built up of the delta. Coring in the delta revealed silty and sandy sediments with a depth between 10 and 20 m (Fredén 2000). The present water level of Lake Vänern is 44.0 m a.s.l. and the water level ﬂuctuates about ±1.3m between low and high water (Heikenskiöld 1981). Heikenskiöld (1981) estimates that the delta situated at the excavation area was built up over the sea level about 700 BC to AD 1. During the beginning of the Roman Iron age, a sandbank dividing the river into two branches was formed. Later the delta mainly grew in the eastern and western branches of the river. By AD 700 the central part of the delta had the same characteristic features as today. From the 13th century the expansion of the delta shifted to the south, and since then the central parts of the delta have remained stable, although certain morphological changes of the spit of Sandgrund was seen during the 19th century. The hills around the ﬂat delta plain display several Iron Age and Medieval settlements. There seems to have existed a Medieval trade centre located at a place called ‘Tingvalla’. In AD 1290 Tingvalla is mentioned as the centre where markets and juridical processes were held (Nygren 1934), but the exact location and nature of Tingvalla is not known, and no archaeological evidence has so far been found. However, single artefacts from the Middle Ages found within Karlstad indicate nearby activities during this period (Andersson & Schedin 2001). Karlstad was founded as a strategic iron-shipping town between the mining areas in middle Sweden Figure 17: A) Map of central Karlstad. The investigated area is marked with black on the left side of the marked block. The river Klarälven is shown in dark grey. The line that surrounds and lies southeast of the church marks the hill ‘Lagberget’. A part of the hill ‘Marieberg’ is marked in the southwestern corner. B) Simpliﬁed map of Quaternary deposits in the area (after Fredén 2000). The alluvial delta material is light grey, and the river Klarälven and Lake Vänern are dark grey. White areas are dominated by till and bedrock but also include ﬁne-grained sediments and peat accumulations. 31 J. Heimdahl (Bergslagen) and the Swedish west coast. Karlstad was given its town privileges in AD 1584 by King Karl IX and until now, no urban structures pre-dating that year has been found (Lundh et al. 1994, Ängeby 1995). The town had its main economical growth during the 18th and 19th centuries. The accumulation of sand seems to have been a problem in Karlstad for a long time and regular dredging of the river was necessary for the shipping. Flooding in the town centre also occurred regularly, except for the hilly areas of till-covered bedrock that are seen above the ﬂat delta plain (Andersson & Schedin 2001). The excavated area is located on the ﬂat delta area in the town centre between the till-covered hills of Lagberget in the northeast and Marieberg in the southwest (Fig. 17). The c. 1 m thick occupational layers were deposited directly on, and partly developed as horizons in the ﬂat delta surface and they are evenly distributed over the investigated area. 7 Methods Field methods and sampling strategies of this project are described in more detail in Paper I. Because of the nature of the urban stratigraphy, continuous sampling was not undertaken in all stratigraphical units. The main method used was plant macrofossil analysis, which was suitable to describe the local environment as well as human activities. The macrofossil content was diagnosed in all stratigraphical units excavated. In remains directly related to human activities, e.g., house debris, the method was limited to the search for indicators of past vegetation and cultural plant use. In these contexts, sampling in each stratigraphical unit became more extensive in order to adapt to the archaeological excavation strategy. Other biostratigraphical methods used were pollen and diatom analysis. These methods were only applied when more speciﬁc questions arose. Pollen analysis was used as a complement to plant macrofossil analysis on the oldest strata because of the lack of cereal grains in these layers. In order to trace the origin of the deposits, diatom analysis was used where the stratigraphy presumably was inﬂuenced by alluvial action. Some stratigraphical units were studied more carefully for sedimentological features and in these cases, samples were taken for texture analysis and the determination of organic carbon content. Fabric analysis was carried out when it was difﬁcult to distinguish between dumped ﬁlling masses and till. 32 Field work Since the geological ﬁeldwork was carried out during the archaeological excavations it had to be adapted to the archaeological framework (Menander & Karlsson 2002, Heimdahl et al. 2003; Paper I). At each site the archaeological excavation was performed according to SCR methodology and the stratigraphy was plotted in a Harris Matrix. The positions of all stratigraphical units, as well as artefacts and samples, were documented three-dimensionally in a database and visualised in the GIS-program Arc View. A sampling technique for macrofossil analysis was developed and adapted to the single context methodology used at the excavation (Paper I). Pilot sampling was performed in each layer except those that were unfavourable for the preservation of organic remains, for instance deposits including coarse gravel or stones. The samples aimed for macrofossil analyses were collected in cardboard boxes with a plastic inner ﬁlm. The samples were normally ﬂoatated and wet-sieved immediately after sampling, according to the method described by Wasylikowa (1986). The enriched remains were investigated under a dissecting microscope at the site and the plant macrofossil content was qualitatively diagnosed in order to evaluate the need of additional sampling before the layer was totally excavated. After this pilot analysis, the samples were stored in water ﬁlled plastic containers. Stratigraphical cross sections were described and drawn both by the archaeologists and the author. The colour of the strata was classiﬁed according to the Munsell Soil Colour Cart (Munsell Colour 2000). In some cases the drawings were dissimilar because of differences in the interpretation of the stratigraphy. In layers where sedimentological structures could be of interest for sampling, metal boxes were pressed into the wall and cut out so that sufﬁcient parts of the strata were preserved intact. The ﬁlled boxes were subsequently wrapped in plastic and stored in a cold room before sub-sampling. In order to minimise contamination by recent vegetation, a survey was made of the seed-producing plants in the present local urban ﬂora. The composition of the modern local ﬂora was also compared with the fossil urban ﬂora investigated. Investigations were carried out on ﬁve occasions during each excavation in an area up to 10 metres away from the excavation site. Plants with seeds spread by wind were documented with extra care, even if they were growing a long distance from the excavation site e.g., Betula spp. and Salix spp. The time of seed setting was noted and plants that were about to set fruits close to the site were, if possible, removed. Urbanised Nature in the Past Macrofossil analysis Pollen analysis In order to facilitate an ecological interpretation of the macrofossil content, between 300-500 diaspores were identiﬁed to species level in each sample investigated. Other plant remains which could be identiﬁed, such as needles, leaves and bark, were registered but not included in the percentage calculations. The macrofossil samples were, when necessary, wetsieved for a second time and cleaned in the laboratory according to methods described by Wasylikowa (1986). After the wet sieving the samples were examined as subsamples of c. 3-5 ml at a time. The microscope used in the laboratory had a magniﬁcation of ×8-100. The samples were ﬁrst described according to their general composition of tissues and biota. The groups used were wood, charcoal, bark, herbacea epigaeic (stems and leaves), herbacea hypogaeic (roots), insects, statoblasts (hibernate-capsules of bryozoans), puparia (pups, eggs, cocoons etc. of lower animals), Ichtyes (bones and scales from ﬁsh), Mammalia (bones and hair from mammals) and artefacts, e.g., slag, glass fragments and brick fragments were also included in this description. The amounts of the most frequent components: wood, charcoal and bark was roughly quantiﬁed as ‘1’ (single objects), ‘2’ (<5% of the visual ﬁeld) or ‘3’ (>5%). The diaspores were counted as single units. Fragmented material was noted as parts of units if the original parts could be reconstructed or the amounts were roughly quantiﬁed into separate units. Fragments of a seed taxon, represented by a single element from the original carpum (e.g., a hilum), could be counted as one unit but in contrast, fragments without these special elements were not considered for quantiﬁcation. The remains with characteristic cell patterns, such as Poaceae and Juncus fruits, were determined under a light-microscope with higher magniﬁcations (×1001000). For specimens where identiﬁcation was of extra importance, further analyses were undertaken with a SEM-microscope (Paper II). The identiﬁcation of the seeds was mainly carried out by consulting a reference collection and the works of Korsmo (1926); Bertsch (1941); Körber-Grohne (1964, 1991); Katz et al. (1965); Berggren (1969, 1981); Beijerink (1976); Tallantire (1976); Jacomet (1987); Schoch et al. (1988); Jacomet et al. (1989); Blidow and Krause (1990); Anderberg (1994) and Krause (1997). After identiﬁcation, the seeds were stored in distilled H2O in 5ml plastic containers. A mixture of glycerine, ethanol and thymol was used for storing very fragile material. Pollen analysis was undertaken to investigate the cereal pollen content in the dark earths of Norrköping. Samples designated for pollen analysis were sub-sampled from the sediment collected in the metal boxes. In order to concentrate pollen and spores from the sediment, a standard acetolysis method was used (Erdtman 1936; Berglund & Ralska-Jasiewiczowa 1989). A collection of reference slides and the work of Moore et al. (1991) were used for identiﬁcation. The cereal type pollen grains were measured and surface structures were studied according to Beug and Firbas (1961) and Andersen (1979). Diatom analysis Analysis of diatom composition were undertaken in order to investigate ecological signals of terrestrial or aquatic environments. Samples for diatom analysis were sub-sampled from the metal boxes. Siliceous microfossils were extracted according to the method described by Battarbee (1986). Identiﬁcation of the diatoms and the interpretation of the ecological conditions were made with the use of Kramer and LangeBertalot (1986; 1988; 1991a; 1991b) and Round et al. (1990). The diatoms were grouped according to their habitat, either as terrestrial or aquatic. Only one of the six samples contained signiﬁcant amounts of diatoms for a quantitative analysis. Grain size distribution Texture analysis was performed as a complementary stratigraphical tool in order to reveal obscured stratigraphical changes and in order to trace possible mother material for ﬁlls. When sediment is reworked (for example by digging), its structural features will disappear and changes but the texture will remain intact (cf., Paper I). Samples for grain size distribution were taken directly from the open transects and on surfaces where stratigraphical units had been exposed. Texture analyses were performed on gravel and sand particles by dry sieving using sieves with square-formed holes. The silt and clay fractions were determined by hydrometer analysis (Kompendium i jordartsanalys 1995). Organic carbon and loss on ignition The analysis of organic carbon content and the loss on ignition (LOI) was performed in order to reveal obscured stratigraphy of the dark earths. It was determined on the dark earths from both Norrköping and Karlstad. Continuous samples, each 1 cm thick 33 J. Heimdahl yielding c. 1-2 cm3, were extracted from the dark earth at 3-5 cm intervals. Organic carbon content was determined by a carbon sulphur detector (ELTRA CS 500). The LOI was calculated as percentages from the weight loss of dried samples, which were slowly heated to +550ºC and kept at that temperature for two hours (Kompendium i jordartsanalys 1995). C-Dating 14 Seven plant macrofossil samples (ﬁve from Norrköping and two from Karlstad) were AMS dated at the Ångström Laboratory, Uppsala University. The material used was sub-sampled from the macrofossil samples (Table 1), which were stored in distilled H2O in a cold room at +4ºC, between the summer of 2000 and the autumn of 2001. Fragile seeds from terrestrial plants found in a primary position (e.g., ﬂoors, tramping horizons and hearths) were chosen for dating to minimize the risk of using reworked material. The seeds were gently cleaned in distilled H2O and detritus was washed away with a small nylon brush under a dissecting microscope. This material was ﬁrst slowly heated at +70ºC for 2h and then dried at +105ºC for 8h and placed in small glass jars with plastic screw lids. Other dating methods Dendrochronology and the typology of artefacts and coins were also used for dating purposes. Thomas Bartholin at the National Museum in Copenhagen carried out the dendrochronological work. The archaeologists at the excavation undertook the typological datings. Monica Golabiewski Lannby at the Royal Coin Cabinet in Stockholm performed the identiﬁcation of coins. Ecological grouping The identiﬁed plants were grouped according to ecology using the systems described by Grime et al. (1988), Ellenberg et al. (1991) and Mossberg et al. (1992, 2003). The system of Ellenberg et al. is based on Central European data, and groups the plants according to their different requirements (e.g., light, moisture, nutrients and temperature) expressed in numbers. The system is hierarchic and a plant can only be described as thriving either in light or shadow. This compilation covers about 90% of the species found in Norrköping and Karlstad. Grime et al. (1988) describe the biotopes of plants in the British Isles according to where the plants are found and 34 evaluate their degree of abundance with a number (a scale from 1-5). The system is heterarcic, which means that a plant may occur with the same frequency in different habitats. In this way plants with generalist strategies can be separated from plants with special strategies according to their habitat. This system covers about 60% of the plant species found in the investigations. Mossberg et al. (1992, 2003) is a comprehensive ﬂora with descriptions of the ecology and habitats of plants found in the Scandinavian Peninsula. This was mainly used for plants that were not included in the systems of Grime et al. (1988) or Ellenberg et al. (1991) Mossberg et al. (1992, 2003) was also useful for comparing descriptions of plant ecology in order to check if there are differences between the regions. The following publications were also used for information on single plant species: Jensen (1985, 1998); Kroll (1995, 1996, 1997, 1998, 1999, 2000); Schultze-Motel (1992, 1993, 1994); Johnson (1999) and Mabberley (2000). Biological nomenclature The Latin names of plant species and taxas are according to the continuously updated species list published on the internet by the Museum of Natural History in Stockholm. The version used here was updated 2004-01-19 (Karlson 2004). The English names are according to Qualtrocchi (2000). 8 Summary of Papers Paper I Jens Heimdahl, Hanna Menander and Pär Karlsson, In press 2005: A New Method for Urban Geoarchaeological Excavation, Example from Norrköping, Sweden. Norwegian Archaeological Review. In press November 2005. During excavations of complex urban occupational deposits in Norrköping, new procedures were elaborated in order to combine archaeological, geological and archaeobotanical ﬁeldwork. This ﬁeldwork was also based on parallel work by members of the different disciplines throughout the whole excavation. The archaeological ﬁeldwork was based on SCR. In order to understand the archaeological ﬁeldwork procedures and the basis of interpretation, the geologist (Jens Heimdahl) also spent time excavating with the archaeologists. The elaborated procedures resulted in continuous pilot sampling and on site preparation and di- Urbanised Nature in the Past agnostics of samples. When excavation of a new stratigraphical unit started, it was sampled by the geologist who immediately enriched and analysed the material under a dissecting microscope at the site. Hence, a pilot survey of the macroscopic content of the stratigraphical unit was available before it had been fully excavated. The pilot surveys facilitated the ﬁeld interpretation of the strata. On several occasions the results revealed the functions of constructions, and provided environmental glimpses already in the ﬁeld. For example, it was possible to take additional pilot samples in order to investigate if the macrofossil content within the strata differed spatially, and a comparison to other strata could also quickly be performed. The method also made the correlation easier between strata cut by younger truncations. The pilot surveys were also used to trace former root-horizons. Such horizons were analysed to determine which type of plants that had grown on the site, information that in turn could give clues for how long time surfaces had been left open and untouched. All strata were geologically studied according to sedimentological structure and texture, parallel to the archaeological stratigraphical interpretations. These double interpretations resulted in constantly ongoing ﬁeld discussions. On some occasions the different interpretations conﬁrmed each other, in others they opposed each other. The discussions that followed proved important for the development of a mutual understanding and in some cases led to new, sometimes surprising conclusions. The geological interpretations revealed that the thick minerogenic strata, located below the dark earths, were ﬁlls consisting of reworked glacioﬂuvium and littoral sand. Without the geological expertise present, they would probably have been interpreted as the natural substratum. The geological ﬁeld interpretations also revealed new ways to critically treat stratigraphical units as information sources. This especially counted for material that traditionally is considered reworked, like ﬁll. By analysing the ﬁll texture and comparing it with the texture in other local material of natural origin, it was on some occasions possible to determine the source material for the ﬁll. Another way of dealing with the problem of reworking was to look at the degree of fragmentation of fragile organic macro compounds. Finds of large pieces of leafs or fragile seeds may be used to exclude reworking. Geological ﬁeld interpretation also resulted in the discovery that the occupational deposits to a large extent were of alluvial origin. Many of those alluvial deposits consisted of silt lenses that apparently had accumulated in water ponds. Some of the water ponds had acted as traps for material from surrounding ground surfaces. Surfaces are important for archaeological interpretation, and since many surfaces had disappeared though truncations or were located outside the excavation area, the macro content of water pond sediments could sometimes be used to reveal cultural activities and environment on those surfaces. Paper II Jens Heimdahl, 2005: Urban Sediments as Indicators of Changes in Land use and Building Tradition in two Swedish Towns, AD 1200-1800. (Manuscript) Results and conclusions from sedimentological studies of occupational deposits at the excavations in Norrköping and Karlstad were compared in order to reconstruct urban environmental development and site formation processes. The site in Norrköping is situated on a slope towards the river Motala ström, and the occupational layers were deposited as slope front ﬁll on a substratum of littoral and alluvial sand. The occupational deposits were 3-4 meters thick downslope and disappeared 10-15 m upslope from the excavation limit. The lowest 1m thick layers consisted of minerogenic sediments, probably of littoral and glacioﬂuvial origin, dumped partly in the ﬂooded stream during the 13th century. On top of these ﬁll, two dark earth units with obscured strata developed during the 13th to 14th century. Wooden boardwalks were built on top of the uppermost dark earth during the mid 14th century. Above the boardwalks the stratigraphy is characterised by c. 1 m well-stratiﬁed urban occupational deposits originating from 1400-1660. The occupational layers at the site in Karlstad are deposited on top of the active delta of river Klarälven. The substratum consists of the top set beds of the delta. The oldest culturally affected deposits at the site is the uppermost delta sand in which a 3040 cm thick dark earth with obscured strata have formed, probably during the 16th century, but it may also contain older units. This dark earth is overlain by 1 m well stratiﬁed urban occupational layers deposited between c. 1600-1800 and at some places, interlayered by delta beds from ﬂooding events. The dark earths at Norrköping and Karlstad seem to have developed in similar ways. They consist of 85-95% of minerogenic material with the identical texture as the unaffected substratum. The organic component seems to be dominated by cattle faeces. Sedimentological structures indicate that this dung 35 J. Heimdahl was tramped into the ground, possibly during wet conditions. There are also several structures revealing that the homogeneous dark earths contain obscured strata. The dark earths are believed to have formed during a combination of repeated alluvial sedimentation and a continuous tramping of cattle, thus successively deepening the horizon and homogenising the individual alluvial deposits. In Karlstad the alluvial deposits originate from regular ﬂood events on the active delta. In Norrköping they are formed by alluvial runoff on the slope. This also explains why the dark earths here are thicker downslope. The alluvial deposits in Norrköping continued to form during the formation of the well-stratiﬁed urban occupational deposits. Silty alluvium accumulated repeatedly in the depressions of the town, mainly in streets and lanes between the houses, and on some occasions, created negative casts of the urban landscape. When alluvial silt ﬁlled the streets, new constructions had to be elevated in order to avoid moisture. This habit is thought to have formed a cycle of human response to natural effects thereby inducing each other and resulting in the growth of urban occupational deposits. At both Norrköping and Karlstad ﬁlls used in the foundations for houses were investigated with grain size analyses. The results show that the ﬁll used in the early stages of the urban development was local material of a ﬁne fraction, not suitable for foundations. In the foundations built later, the material was coarser and taken from more distant deposits. This shift is thought to represent a professionalisation of building construction. In Norrköping it seems to have occurred during the beginning of the 17th century. In Karlstad the shift occurs later, during the 18th century. Paper III Jens Heimdahl, 2005: Botanical Evidence of Changes in Vegetation and Cultural Landscape in Post-Medieval Karlstad, Sweden. (Manuscript) Macrofossil evidences from Karlstad reveal glimpses of environmental development of the urban and regional area, cultivation, trade, and differences between grazed pastures and mowed meadows. The oldest 14C-dated samples taken in the delta sediments had calibrated ages of 215 (±40) BC and AD 165 (±40). A dense forest vegetation represented by birch and alder formed on the newly built up delta surface. Light demanding plants, for example Prunus spinosa, Rubus idaeus and Bidens tripartite, thrived along channel banks, and different Juncus 36 and Carex species were growing at the waterline. Submerged plants like Nitella opaca, Isoëtes echinospora and Potamogeton spp. probably grew in the delta channels. Cattle grazing and the creation of meadows deforested the delta during the time when the black earth was formed, sometime between AD 500 and 1500. An increase of ruderal plants in the uppermost part of the undisturbed delta sediments provides an indication of human occupation predating the formation of the dark earth. The macrofossil concentration in the dark earth is heterogeneous and varies both spatially and vertically. The composition of the ﬂora, however, is homogeneous, and is heavily dominated (up to 90%) by Juncus spp., mainly Juncus bufonius and other wet meadow plants like Carex spp. The taphonomy of wet meadow plants is interpreted as cattle faeces. Samples taken from fresh cattle faeces at a nearby wet meadow with ungrazed Juncus tufts reveal that the faecal material becomes heavily dominated by Juncus fruits, due to the grazing of fruits stuck to the surrounding grazed vegetation. The macrofossil record in the urban occupational deposits is dominated by a species composition that also characterises cattle faeces. This conﬁrms the historical notes from the 16-17th centuries that the town was largely composed of farms. Hay remains were found stuck to the inside of a compressed bucket that was situated on the ﬂoor in a burnt down small stable from the early 17th century. The macrofossil composition of the hay, dominated by Filipendula ulmaria and Poa palustris, differs completely from the composition in the cattle faeces, dominated by Juncus bufonius, found on the stable ﬂoor (that seems to have been covered by spruce branches). This show that the cattle kept in this stable had been grazing and was not fed with hay mowed from meadows. Since hay was likely used as winter fodder, this indicates that the stable was not used for winter stabling. An explanation to this may be that it was used speciﬁcally for cattle that were to be slaughtered. The urban cultural layers were also rich in the remains of cultivated, collected and imported plants. The ﬁnds of cereal grains were dominated by Avena sativa (oat). According to historical sources, this was the most frequently cultivated plant in the county of Värmland at this time, which indicates that locally cultivated crops dominated the urban use of cereals. Macrofossil ﬁnds also indicate the cultivation of Fragaria moschata (Hatbois strawberry) during 1718th century, Rubus subg. Rubus sect. Rubus (blackberries), Pisum sativum var. arvense (ﬁeld peas) and Fagopyrum esculentum (buckwheat). Further, there is botanical evidence of cultivation of Nicotiana rus- Urbanised Nature in the Past tica (wild tobacco). This cultivation is historically documented to have been introduced in Karlstad by Gustav Claren during the mid 18th century. Historical documentation and ﬁnd positions in the debris of a burnt house may contribute to a dating of the destruction of the building to June 8, 1752, when there was a large ﬁre in Karlstad. The discovery of a fruit from Pimenta ofﬁcinalis (allspice) in a stratum from the latter part of the 17th century indicates the importation of exotic species. Allspice has earlier only been archaeologically found twice, in England and Poland, and this is probably the oldest ﬁnd in Europe so far. There is also other evidence of long distance trade in the form of rare species speciﬁcally known to grow at docking places and in harbours, for example Erucastrum gallicum and Chenopodium murale/ C. vulvaria. Paper IV Jens Heimdahl, 2005: Botanical Evidence of Environmental and Traditional Changes in Norrköping, Sweden, AD 1200-1660 (Manuscript) During excavations in Medieval and Post-Medieval Norrköping, macrofossil and pollen evidences revealed local and regional ﬂoral changes, cultivation of plants, preparation of food and importation of exotic plants. The oldest macrofossils were found in alluvial deposits underlying the oldest occupational deposit – a ﬁll from c. AD 1200. The composition indicates that the slope towards the river Motala ström partly contained a wet meadow with tufts of Juncus sp. and Urtica dioica, and partly a drier sand bank vegetation with for example, Arenaria serpyllifolia, Potentilla argentaea and Arabidopsis thaliana. The urban phase at the site seems to have started with the dumping of ﬁlling masses on the slope towards the river. On these ﬁlling masses, a dark earth developed through cattle tramping between AD 1200-1400. Nitrophilous plants with pioneer strategies like Thlaspi arvense and Lamium album besides species thriving on faeces like Urtica dioica and Chenopodium glaucum/rubrum seem to have dominated the site together with plants resistant to tramping, for instance Polygonum aviculare and Plantago major. This combination of plants indicates a rural environment with dung heaps, tramped paths and plants growing along fences and buildings. Above the dark earths, complex strata of urban occupational layers were deposited between AD 1400 and 1660. The composition in the layers indicate a shift in the local plant ecology since earlier common plants like Plantago major and Lamium album disappear and are replaced by newcomers like Aethusa cynapium and an increase of Hyoscyamus niger. Centaurea cyanus, Rumex acetosella and Spergularia arvensis are examples of plants interpreted as regional ﬁeld weeds. Field weeds especially common in ﬁelds with autumn sawed rye, for example Bromus secalinus and Agrostemma githago, was found. The cattle that were brought into town were probably mainly grazing on wet meadows with different species of Carex spp. Secale cereale (rye) dominates among the cereals in the pollen record from the dark earths (1200-1400) and in the cereal macrofossil composition between 1400 and 1550. If this reﬂects local cultivation, it is partly in line with historical evidences since the area became part of the so called ‘rye belt’ during the 15th century. Of special interest is a burnt ﬁnd of what is interpreted as threshed remains from a harvested rye crop from 1400-1550. The cereal grains are here mixed with seeds from typical ﬁeld weeds like Centaurea cyanus and Rumex acetocella but the most common non-cereal plants are Poa palustris and Trifolium hybridum, which are typical meadow plants common in mowed hay. This may be either a mixture of a crop and hay, or possibly a crop from a ﬁeld that previously was used as a meadow. Among the most commonly found cultivated plants were hops (Humulus lupulus), well known to have been used for the ﬂavouring of beer. Beer making seems to have been common at the site between the 14th and 16th centuries. In some samples, hops were found together with Myrica gale, which was the most commonly used beer ﬂavour in Sweden before hops was introduced. The ﬁnds also indicate that a mixture of H. lupulus and M. gale was used for beer ﬂavouring in 1400-1550. Filipendula ulmaria was over-represented in the samples that contained other beer ﬂavours. This plant is earlier known to have been used as a ﬂavouring of mead during the Viking Age. The ﬁnd in Norrköping may indicate that this tradition survived into the late Middle Ages. Further, seeds from imported plants like grapes and ﬁgs were found. An early ﬁnd of grape, probably deriving from the 14th century, indicates the presence of a clerical, or upper class, community. Paper V Jens Heimdahl, 2005: Archaeobotanical Evidence of Early Tobacco Cultivation in Norrköping, Sweden (Manuscript) Botanical evidence of tobacco cultivation was found 37 J. Heimdahl during the investigations in Norrköping. It is the ﬁrst archaeological ﬁnd of tobacco in Sweden. Light microscope and SEM analysis resulted in the identiﬁcation of the seeds as Nicotiana rustica (wild tobacco) and Nicotiana spp. (tobacco unspec.) Since ﬁnds of fossil tobacco seeds are rare, the seeds are carefully described. No leaf tissue from tobacco was found in the samples containing the seeds. In total, 42 seeds were identiﬁed. Although a 14C-date indicates a younger age, artefacts and stratigraphical information suggests that cultivation occurred sometime during 1560-1660 – possibly the earliest tobacco cultivation in Sweden. The conclusion that the seeds indicate cultivation in Norrköping is based on several facts. Tobacco for smoking or snifﬁng was imported as leaves. In order to maximise the size of cultivated tobacco leafs, the ﬂower buds were cut off the plants. The possibility that the seeds derive from imported tobacco leaves is therefore minimal. So far the oldest known cultivation of tobacco in Sweden took place in the garden of Uppsala castle in 1632. During the 17th century, tobacco cultivation in Sweden was not aimed for large-scale production but for medical use. Tobacco was considered a panacea – a miracle medicine able to cure many types of diseases. Early tobacco cultivation for medical or scientiﬁc use is a probable explanation to the ﬁnd in Norrköping. From written sources, tobacco cultivation for industrial use is known to have occurred in this town between 1720 and 1930. 9 Discussion Urban strata, a formation model From a geological perspective, it is natural to deﬁne urban strata as two different types of phenomenon: sediments and soil horizons, meaning strata formed by depositional and post-depositional processes respectively. This classiﬁcation has earlier been stressed in archaeological contexts by both geologists (Limbrey 1975, Lundqvist 1986, Corty et al. 1989, Carter 1992), and archaeologists (Beronius-Jörpeland 1992). A third important factor that affects site formation are the negative effects from truncations, erosion (Tagesson 2000a) and decomposition. The processes that controlled the formation of the urban stratigraphy in Norrköping and Karlstad can be divided into three major groups: Depositional, Postdepositional and Truncation/erosion/decomposition processes. Deposition provides the material, which is redistributed and reworked during post-deposition, and the third process, truncation/erosion, removes the material. Truncation/erosion may be also followed by redeposition. Again, each of these groups may contain both natural and cultural elements. One of the major questions archaeologists ask about site formation is which of the units were formed by intentional acts, and which were formed unintentionally (Järpe et al. 1979, Andrén1986, Beronius-Jörpeland 1992, Larsson 2000). The answer inﬂuences the information potential of each stratum. Moving towards a classiﬁcation of possible site formation causes as being either intentional or unintentional, I would like to add natural processes as a third factor. Table 2 presents how the tree groups of processes, depositional, post-depositional and truncational/ erosional, combined with the tree groups of causes, intentional, unintentional and natural, may result in different types of stratigraphic features. A compilation of processes that affects the formation of urban occupational strata can be combined into a model (Fig. 18). This model combines the geological and archaeological approaches to the stratigraphy. The model is divided in three main steps: causes, cultural actions/natural events and stratigraphic Table 2: Classiﬁcation of urban occupational strata Natural Alluvial (water) Depositions Colluvial (gravity) Eolian (wind) Horizons Earthworm horizons Root horizons Erosion/truncation/de- Alluvial erosion composition Microbiological decomposition 38 Unintentional Cattle droppings Unintentional waste Tramping horizons (humans and cattle) Live stock turbation Tramping erosion Intentional Constructions Fills Waste and dump Plough horizons Cultivation horizons Truncations Urbanised Nature in the Past Cause Stratigraphic result Action/event Organic g Waste and dumps enin id d /m g in ts Fills and fundaments amen mp und f Du f o n tion atio Building debris/remains estruc Cre on/d i t c ru t s n Co Cattle faeces n tio c tru Alluvial es ng pi l Deposition may be fal Deposition ain post-depositionally r or affected (reworking on d Post-deposition small scale) oo Di gg in g /p l Depth scale in meters o ug hin g 1 0,001 0,01 0,1 Fl Culture Dr op D Strata formation Intentional Minerogenic Mixed Unintentional Tramping horizons stock Live stock turbation horizons orms of earthw ence s e r P plants Growing of oo M ic r do ob iol io r ra og infa ic a Root horizons n Truncations ll Erosion l ac tivity Preservation and disappearence of strata may affect natural and cultural conditions Decomposition Disappearence of strata Erosive contacts Preservation of strata Strata destruction at nc ru g/t gin Dig ng lki wa g. Fl Earth worms horizons Deposition and horizons may disappear or be reworked by truncation or erosion E. Natural Reworking g ng of live rking Nature H oldi Plough/crop horizons Rewo Tramp in Strata (Layers and horizons) Figure18: Model for the formation of occupational strata within a site. Disappearance of material from this model will result in depositions on other sites. The system is inﬂuenced by the opposite situation. results. The stratigraphic results are in turn divided into: strata formation (depositional and post-depositional) and strata destruction (including truncation/ erosion/decomposition) effects. The model illustrates how cultural and natural causes contribute to site formation in both the depositional and the postdepositional phase, and also how the stratigraphic results in turn can act as a trigger mechanism for cultural and natural responses (cf., the cyclic model in Paper II). All types of processes that bring material to the site are here referred to as depositional. From a geological perspective, they can be considered as ‘sedimentological’ processes. The depositional processes result in accumulations of different layers. The intentional deposits dominate the occupational layers in both Norrköping and Karlstad. The main volume of the deposits consists of minerogenic layers that were transported to the site with speciﬁc intentions, mostly to reshape the local landscape, for example to make a terrace on a slope, as in Norrköping, or to construct elevated foundations for buildings. For 39 J. Heimdahl these purposes mainly minerogenic material was used. Reworked house debris are also intentional deposits. The destruction of houses may have been unintentional but often the material of destructed houses was truncated and used as ﬁlls in new foundations, and the unintentionally created debris is thereby intentionally reworked (cf., Tagesson 2000a). A special type of intentionally deposited material was sand that was spread on icy grounds during wintertime. This sand was deposited with a purpose but when the snow melted in the spring, it became redistributed by alluvial processes and by cleaning. Waste piles, dung heaps and latrines are examples of intentionally created organic deposits but all traces of organic waste piles found in Norrköping and Karlstad contained 40-80% minerogenic material. The fact that cattle faeces were dumped within the town and not brought to the surrounding ﬁelds indicates an overproduction, which is also conﬁrmed by written sources (see Paper V). The second most common depositional factor in Norrköping and Karlstad was natural events (Paper II). The major natural process was alluvial and resulted in the sedimentation of silt and sand in water ponds formed in depressions. In Norrköping there is also an example of small mudﬂows. Regularly occurring exposure of sandy material in the slope results in erosion during rainfall and snow melt. The alluvial slope runoff in Norrköping probably worked as a trigger mechanism for a certain cultural responses. The concentration of alluvial and mudﬂow deposits to depressions results in drainage problems in these areas, which led to that buildings had to be elevated by their foundations in order to avoid the moisture, which in turn created new depressions in the streets. Hence, this includes a nature-culture cycle of reactions and responses (cf. Paper II). This is probably one of the reasons why occupational deposits are thicker in depressions. The cycle was inhibited when the foundation materials became coarser, with better permeability and more resistance to alluvial erosion. (In Fig. 18 the bottom arrow that leads from the strata and the erosive cintacts back to the cultural and natural causes represent these processes.) A fully natural process dominates the alluvial material in Karlstad – the regular ﬂooding of the river Klarälven and the formation of top set beds on the still active delta. Examples of unintentional deposits are dropped objects and animal faeces dropped on the ground in the streets and yards. Normally, material deposited on the ground was mixed into the topsoil by post-depositional processes such as tramping and bioturbation by earthworms. To ﬁnd these types of 40 deposits unaffected by post-depositional processes is extremely rare. Tramping by humans and animals was probably the most important of the post-depositional processes that occurred in the Medieval and Post-Medieval urban environment. Experiments have proved this process to be able to redistribute objects up to 17 cm in depth (Villa & Cortuin 1983) in sandy material during just one day of tramping. However, with the exception of dark earths, tramping horizons on the sites in this study are 3-10 cm deep. Cobblestones, wooden boardwalks and a compact humic ground inhibited the tramping effect. Natural soil formation processes occurred parallel to the tramping. There are several examples of strata in both Norrköping and Karlstad where earthworm traces were present. This is indicated both by horizons with crumb microstructure and macroscopic ﬁnds of earthworm cocoons. It was quite common to ﬁnd traces of root horizons together with traces of earthworms. Traces of smaller soil-penetrating organisms like nematodes and Oribateids were also found in many samples. Examples of intentionally created post-depositional processes are ploughing or digging in order to prepare a soil for cultivation. Soils of this type were not found at any of the sites although a 16th century plough-horizon had earlier been identiﬁed in Karlstad, c. 200 m east of the present site (Lund et al. 1994). The macrofossil record from Karlstad indicates that the development of soil horizons occurred during summer and autumn (cf. Paper III). Due to frozen ground no organic material was mixed into the soils during wintertime. Cattle faeces loaded with fodder plants remnants from mowed meadows (winter signal) are absent in the urban soils and must therefore have been deposited during the grazing season. Since deposition of winter droppings loaded with fodder plants most likely occurred, their absence implies that the droppings were cleaned from the streets before the thawing of the ground. It could have occurred during snow ploughing but since snow is not always present, their absence also implies that the frozen streets were cleaned from dung also when they were snow-free – probably due to a regular habit of cleaning the streets. This is in line with the idea that an organised cleaning occurred in the towns during the Medieval and Post-Medieval times (cf., BeroniusJörpeland 2001). Intentional truncation and unintentional/natural erosion created negative features found in Norrköping and Karlstad. These features are revealed by erosive contacts between layers. Truncations may be Urbanised Nature in the Past Figure 19: Formation of dark earth through continuous land use (in this case, cattle handling causing tramping horizon) and repeated alluvial events causing deposition of alluvial sand. The stratigraphy is repeatedly obscured by the tramping but may also be preserved, e.g. by root horizons from vegetation and by postholes from constructions. during snow ploughing. Natural erosion may occur, for example through alluvial processes. In this study erosion has been harder to recognise than truncation, as it occurs on a smaller scale. On the other hand, the material that is eroded and redeposited by alluvial process has been easier to identify. The formation of dark earth s rie es tu ad C en ec D Ye a rs s th on M ks The dark earths of Norrköping and Karlstad were formed by a combination of depositions by natural events through alluvial processes and the post-depositional unintentional formation of soil horizons by W ee s ay H D ou rs es ut in M Se co nd s recognised by their regular extension that excludes natural or unintentional processes. When a truncation has cut one or several strata, it is sometimes possible to identify and trace the eroded or truncated material that is redeposited as ﬁlls nearby. This is possible when the redeposited material contains a mixture that is possible to identify, for instance through recognisable speciﬁc pieces or lumps. It is possible to make these observations in the ﬁeld, although soil micromorphology provides a more powerful tool when dealing with these features (cf., Carter 1993). Erosion by unintentional acts may be, for example hollowed walking paths or streets that are eroded Faecal dropp Deposits Fills Alluvial deposits Waste/dump/middening piles Fire debris Accumulation of charred material in kilns and on cooking places Accumulation of litter on floors Plowing/cultivation horizons Tramping horizons Horizons Root horizons Earth worm horizons Fig. 20: Duration of urban strata formation. Note that deposits generally have a longer duration than horizons. 41 J. Heimdahl cattle tramping (Fig. 19). The post-depositional processes led to the formation of soil horizons in the exposed ground surfaces. When cattle were present at the site, new organic material in the form of dung was deposited on top of the sand. The mixture of sand and dung occurred by cattle tramping but probably also by other pedogene soil forming processes, such as bioturbation by earthworms and roots. In this way a tramping horizon mainly composed of a dung-and-sand mixture was formed. However, alluvial processes continued to affect the area, and additional sand and silt were deposited on top of the horizon. Since the cattle handling at the site continued, the deposition and mixing of the dung into the tramping horizon also continued. The horizon affected the ground all the way below the contact between the old ground surface and the newly deposited alluvial sand, and erased this contact. This process continued and the result was dark earths that contained several obscured beds of alluvial material and several overlapping obscured tramping horizons (Fig. 19). The obscured stratigraphy of the dark earths in Norrköping and Karlstad was revealed through the traces within them, such as preserved fragments of minerogenic units, traces of constructions that apparently were built on top of obscured surfaces and the varying organic compound and macrofossil content that differed through depth (Paper II, cf., Sidell 2000). The reason why the dark earths at the investigated sites ceased to form was probably due to the change in land use when the area was divided into plots in order to raise houses. Even if cattle were still kept within the towns, movements of the animals were restricted to streets and backyards that were partly paved, inhibiting alluvial effects. Tramping horizons of cattle do occur also in the strata above the dark earths but were not developed to the same depth or complexity. Sediments, soil horizons and culture The formation of deposits in the urban environment generally occurred during short-term intervals while post-depositional formation of soil horizons normally had a longer duration (Fig. 20). Fills in foundations and terraces were most likely formed during minutes, hours or days. Waste deposits may, as a whole unit, have a longer duration of months or years but parts of the waste heaps can be formed during seconds. Geologically speaking, the urban occupational deposits are extremely rich in hiatuses since no continuous accumulation takes place. 42 The effective depositional time of the urban layers in Norrköping and Karlstad may be a few days or weeks in total, even though the whole stratigraphy represents a development during several millennia. The pedogene post-depositional soil formation process had a longer duration. The tramping horizons developed in some of the deposits represent units that were formed through ‘the unintentional activities of the daily life’. If the total formation time of these soil horizons are added to the total formation time of the deposits, the result is more likely to represent decades instead of days or weeks but it is still important to remember that we are trying to reconstruct millennia from the fragments of those decades. Harris’ (1979) statement that a stratigraphic sequence comprises depositions and interfaces is, in my opinion, too generalized. Larsson (2000) may be sociological correct when he states that “the surface is the scene upon which the social interactions occurred” but it may also be stratigraphically misleading. The idea that the surfaces are the most interesting stratigraphical units to the archaeologist (Gardelin et al. 1997, Larsson 2000) may lead to some confusion when trying to understand how the ground was affected by activities. Surfaces are twodimensional and constantly changing on a substrate and if they are buried, they are extremely difﬁcult to trace through excavation. Surfaces themselves are rarely preserved in a stratigraphy because of truncation and erosion. In fact, most of the contacts/interfaces between the urban deposits of Norrköping and Karlstad are erosive. Instead the traces after those ‘activity surfaces’ have to be searched for in the soil horizons developed as three-dimensional strata in the deposits that constituted the ground at a speciﬁc period. The social interaction mainly took place during the post-depositional phase, and it is therefore the post-depositional features – the tramping horizons etc. – that must be studied in order to interpret this daily life activity, and thus the cultural and social complexes. Still, it has also been proven fruitful to trace social and cultural changes by studying the deposits. Alluvial beds may contain reworked objects from lost surfaces and objects found in minerogenic deposits with known origin may be considered as primary deposited (Paper I). In addition, textural studies of ﬁll have proven rewarding. The shift from locally used material in the ﬁlls to material with ‘exotic’ and coarser texture probably reﬂects a professionalisation of the building tradition (Paper II). Among urban historians such a change has long been known but it has not been known when it occurred, or if it developed at different times in different towns, Urbanised Nature in the Past and in different blocks within the town (Augustsson 1992). The investigations in Norrköping and Karlstad suggest that this shift occurred at different times in different towns, but it is not known if it occurred simultaneously in all areas within the these towns. This professionalisation may also have inhibited the further formation of well-stratiﬁed urban occupational strata. The use of coarser material led to better drainage and higher resistance to erosion by alluvial processes. This change also resulted in more advanced drainage systems and the paving of streets and backyards become more common. The paving inhibited soil formation by tramping, and the streets became easier to keep clean (cf., Beronius-Jörpeland 2001). It is probably not a coincidence that well stratiﬁed urban strata are not found after the reorganisation of building tradition was introduced. The younger urban occupational deposits are coarser, water logging is inhibited and the preservation status of organic material in this material is therefore poor. Larsson (2000) stated that we might expect a decrease of natural (geological) processes within the urban environment in comparison to a rural environment. The results in this thesis speak in favour of the opposite: The speciﬁc urban environmental conditions at Norrköping seem to have enforced geological processes. In Karlstad, the geological, alluvial processes were not affected at all until the control of the river in the beginning of the 20th century. It is also worth mentioning that urban occupational deposits generally are thicker than rural occupational deposits. The comparisons I have made between excavations indicate rather that naturally formed geological deposits are more commonly found interlayered within urban deposits than rural deposits. This also suggests that the urbanisation in Medieval and PostMedieval towns, with their constant ongoing soil exposure, reworking and land use shifts, increased the activity of local geological processes. 10 Conclusions • Stratigraphic interpretations are more complete and reliable when archaeologists and geologists carry out excavations together. Possibilities of continuous ﬁeld discussions enhance the prospects of mutual understanding between the disciplines. Methods connected to Single Context Recording are fully adaptable to a continuous geological ﬁeldwork during archaeological excavation. • Continuous pilot diagnoses of macro contents in samples during excavation greatly contribute to ﬁeld interpretations, concerning both interpretation of stratigraphy, functions of structures and past local environments. • The presence of a geologist during archaeological excavations contributes with knowledge concerning both natural and cultural factors connected to the site formation processes. Naturally formed sediments may be studied to obtain cultural information, for example alluvial sediments may have acted as traps for material on truncated surfaces or areas situated outside the excavation area. • There is evidence of small changes in the local urban ﬂora through time, probably as an effect of interspecies competition. Marked ﬂoral changes occurred according to changes in land use. In Karlstad there is evidence of a neophyte harbour ﬂora with rare species, nowadays limited to sporadic occurrences on the west coast. Many species are found which probably had a more northerly distribution during the 14th –19th centuries than today. • Botanical evidence combined with historical data suggests that the main consumption of cereals in the towns derived from local crops. Rye dominated in Norrköping and oat dominated in Karlstad. In Norrköping evidence of a transition between a grass-clover meadow and a rye ﬁeld was found. There are also indications of cultivation of berries, for example Fragaria moschata in Karlstad during the 17th and 18th centuries. In Karlstad and Norrköping evidence of tobacco cultivation was found. In Norrköping tobacco cultivation occurred between 1560-1660, and may be one of the earliest tobacco cultivations in Sweden. In Karlstad it occurred during the 18th century, and by historical sources it is possible to connect to Gustav Clareen’s cultivation of tobacco 1741-1772. • Comparison between the content of plant macrofossils in mowed hay and cattle dung in Karlstad strongly suggests the use of a speciﬁc stable as a slaughterhouse, not used for winter stabling. The ﬂoral composition of meadows and pastures differed and there existed several types of meadows that were managed differently. The grazed pastures in the area were characterised by ungrazed tufts of Juncus bufonius. Sampling of recent cattle droppings from a pasture with an ungrazed Juncus community shows that the cattle graze the Juncus fruits that fall on the surrounding vegetation. The composition of diaspores in cattle faeces may therefore also include ungrazed species. • The formations of dark earths in Sweden are not connected to any speciﬁc time period but probably to rural environments. The dark earths found in Norrköping and Karlstad were formed due to a combi- 43 J. Heimdahl nation of recurrent alluvial deposition in areas that were used for intense cattle handling. By the tramping of the cattle, humic soil horizons were formed that obscured the contacts between the alluvial sand and silt. In Norrköping the dark earth was formed during the 13th century, in Karlstad up to the 17th century. • Deposition of urban strata was generally more intense in depressions and at the base of slopes. The causes of this seem partly to have been from a cultural need of levelling the ground due to drainage, and partly because alluvial deposits from rain water, snow melting and ﬂooding were concentrated in depressions. Since alluvial deposits were concentrated to depressions in the town, they may sometimes have created casts of streets and lanes and backyards. • Development of urban soil horizons was concentrated during spring, summer and autumn, when the ground was not frozen. This is indicated by the macrofossil record originating from cattle dung in Karlstad that consists entirely of species that were grazed at meadows. Remains of mowed winter fodder have not been found within the preserved soils. • In the occupational deposits of Norrköping and Karlstad there is a signiﬁcant textural trend of coarsening upwards of the minerogenic material because the local sand and silt was used as ﬁll in the early phases, and more coarse sand and gravel was used in later constructions. (This trend is also observed during ﬁeld studies in Skänninge and Stockholm, and according to many archaeologists it seems to be common in the occupational deposits of most towns.) The shift was identiﬁed to have occurred during the 17th century in Norrköping and during the 18th century in Karlstad. It is probably a reﬂection of a professionalisation of the urban building procedures that apparently occurred at different times in different towns. • Waste material does not seem to have been used as a primary ﬁlling. Places where the dumping of minerogenic ﬁlling material occurred were probably secondarily used as places for waste dumping. The different taphonomy of the urban macrofossil record may be utilized to make interpretations of the composition of mixed and reworked depositions. Macrofossil remains may contribute to model the development of both local and regional environments. Changes in the local urban ﬂora may reﬂect both environmental changes and interspecies competition. When such changes occur, they can be locally used for indirect dating of strata. • Formation processes of urban strata are regulated by three major factors: Deposition, post-depo- 44 sitional soil formation and erosion/truncation. All these factors may occur as both natural and cultural. Culturally controlled factors dominate but natural factors are always present and even dominate on some occasions. Intentional and natural depositions seem to have contributed to the main part of the minerogenic material. The main part of the organic material was successively deposited unintentionally and unintentionally mixed into the topsoil through tramping. Past urban environments in Norrköping and Karlstad were geologically active and geological processes contributed to and greatly affected the site formation process. References Aalto, M., 1994: The vegetation of the Turku castle. Turun Maakuntamuseo Raportteja 16, 21-38. Ahlberg, N. & Redin, L., 1994: Städerna. In: Selinge, K.G. (ed.), Kulturminnen och kulturmiljövård, 110-117 Sveriges nationalatlas, SNA. Bra Böcker Förlag. Stockholm. Alexander, D. & Armit, I., 1993: Unstratiﬁed stratigraphy: methodologies for interpreting and presenting cropmark sites. In: Barber, J., (ed.): Interpreting Stratigraphy – 1992 Edinburgh, 37-41. AOC (Scotland) Ltd, Edinburgh. Anderberg, A.-L., 1994: Atlas of seeds. Part 4. Resedaceae-Umbelliferae, 281 pp. Naturhistoriska Riksmuseet Stockholm. Andersen, S. Th., 1979: Identiﬁcation of wild grass and cereal pollen. Danmarks Geologiske Undersøgelse. Årbog 1978, 69-92. Andersson, H., 2001: Perspektiv bakåt och framåt på medeltida stadsforskning. Bebyggelsehistorisk tidskrift 42: 7-16. Andersson, H. & Carlsson, K., 2001: Kungahällaprojektet – en bakgrundsteckning. In: Andersson, H., Carlsson, K., Vretemark, M., (eds.) Kungahälla. Problem och forskning kring stadens äldsta historia (Kungahälla. Problems and research on the early history of the town). Bohusläns museums förlag. Stockholm, Uddevalla. 9-28. Andersson, S. & Schedin, P., 2001: Den värmlä rml ndska urrmlä baniseringen. Urbaniseringsprocesser i V Västsverige. En utvärdering av uppdragsarkeologins möjligheter att belysa historiska processer [The Urbanization in Värmland. Processes of Urbanization in Western Sweden. An Evaluation of the Possibilities of Rescue Archaeology to elucidate Historical Processes], 144 pp. Värmlands museum, Karlstad. Urbanised Nature in the Past Andrén, A., 1985: Den urbana scenen. Städer och samhälle i det medeltida Danmark. Acta Archaeologica Lundensia. Series in 80 Nr. 13 . 282 pp. Berggren, G., 1981: Atlas of seeds. Part 3. SalicaceaeCruciferae. Swedish Natural Science Research Council. Stockholm. 261 pp. Andrén, A., 1986: I städernas undre värld. Medeltiden och arkeologin. Festskrift till Erich Cinthio. In: Andrén , A., (ed.) Lund Studies of Medieval Archaeology 1: 259269. Berggren, G., 1984: Utgrävningen på Helgeandsholmen. Fröanalysrapport. Riksantikvarieämbetet, Rapport UV 1984:41. 31 pp. Andrén, A., 1989a: Spåren förskräcker? META 1989: 2, 42-47. Andrén, A., 1989b: I Vidars fotspår. META 1989: 4, 4042. Anund, J., 1995: Stadsarkeologi på UV Uppsala. Stratigraﬁsk metodhistorik och lägesrapport. META 95:2, 3-18 Anund, J., 2001: The Curses and Possibilities of Wooden Architecture: Domestic Buildings in Medieval Uppsala. Lübeck Kolloquium zur Stadtarchäologie im Hanseraum III: Der Hausbau. Bereich Archäologie der Hansestadt Lübeck, 635-656. Verlag Schmidt-Römhild. Armour-Chelu, M. & Andrews, P., 1994: Some Effects of Bioturbation by Earthworms (Oligochaeta) on Archaeological Sites. Journal of Archaeological Science 21: 433443. Arrhenius, B., 1996: The Archaeological Cultural Layer as an Artifact. In: Robertsson, A.-M., Hackens, T., Hicks, S., Risberg, J. & Åkerlund, A. (eds.): Landscape and Life. Studies in Hounour of Urve Miller. Pact 50: 367-376. Berglund, B.E. & Ralska-Jasiewiczowa, M. 1989: Pollen analysis and pollen diagrams. In: Berglund, B.E., (ed.): Handbook of Holocene Palaeoecology and Palaeohydrology, 455-484. John Wiley, Chichester. Bergström, R. & Kornfält, K.-A., 1973: Description of the Geological Map Norrköping NV, Sveriges geologiska undersökning, Ae 14. 71 pp. Beronius-Jörpeland. L., 1992: The Formation of Occupation Layers as an Archaeological Source. In: Ersgård, L., Holmström, M., Lamm, K., (eds.) Rescue and Research. Reﬂections of Society in Sweden 700-1700 A.D. Riksantikvarieämbetet, Arkeologiska undersökningar No 2. Stockholm. 127-149. Beronius-Jörpeland, L., 2001: Den grå vardagen – städning och renhållning i den medeltida staden. Bebyggelsehistorisk tidskrift 42: 59-74. Bertsch, K., 1941: Früchte und Samen. Ein Bestimmungsbuch zur Pﬂanzenkunde der vorgeschichtlichen Zeit. In: Reinerth, H. (ed.): Handbücher der praktischen Vorgeschichtsforschung, Band 1, 246 pp. Verlag Ferdinand Enke, Stuttgart. Augustsson, J.-E., 1992: Medeltida husbyggande i Sverige. Bebyggelsehistorisk tidskrift 23: 55-85. Beug, H.-J., & Firbas, F., 1961: Leitfaden der Pollenbestimmung ffür Mitteleuropa und angrenzende Gebiete, 28-43. Gustav Fisher Verlag, Stuttgart. Bäck, M., 2002: a,e,l,m,n,r,s och t – eller hur man daterar arkeologiska lämningar kontextuellt. Vyer från det äldsta Karlstad. In: Lindgren-Hertz, L., (ed.): Blick för Bergslagen, 36-46. UV Bergslagen. Årsbok 2002. Riksantikvarieämbetet. Örebro. Birks, H., 2001: Plant macrofossils. In: Smol, J., Birks, J., Last, W., (eds.), Tracking Environmental Change Using Lake Sediments Volume 3: Terrestrial, Algal, and Siliceous Indicators. 49-74. Kluwer Academic Publishers. Dordrecht. Battarbee, R.W., 1986: Diatom analysis. In: Berglund, B.E., (ed.): Handbook of Holocene palaeoecology and palaeohydrology, 527-570. John Wiley, Chichester. Blidow, I., & Krause, W., 1990: Bestämningsnyckel för svenska kransalger. Svensk Botanisk Tidskrift 84: 119160. Begon, M., Harper, J. L. & Townsend, C. R., 1996: Ecology. Individuals, populations and communities. Blackwell Science. Oxford. 1068 pp. Borg, G. Ch., 1993: Hotet mot medeltida kulturlager ur naturvetenskaplig synpunkt. META 2: 44-57. Beijerinck, W., 1976: Zadenatlas der Nederlandsche Flora, 316 pp. Backhuys & Meesters, Amsterdam. Berggren, G., 1969: Atlas of seeds. Part 2. Cyperaceae. Swedish Natural Science Research Council. Stockholm. 64 pp. 39 Plates. Botkin, D. B., & Beveridge, C. E., 1997: Cities as environments. Urban Ecosystems, 1: 3-19. Brady, N. C., 1990: The Nature and properties of Soils. 621 pp. Macmillian Publishing Company. New York. Broberg, B. & Douglas, M., 1983: Det arkeologiska arbetet. In: Dalbäck, G., (ed.): Helgeandsholmen. 1000 45 J. Heimdahl år i Stockholms ström, 15-25. Monograﬁer utgivna av Stockholms kommun. Stockholm. Cammas, C., 1997: Etude micromorphologique des terres noires de College de France, 4 pp, In: Macphail, R., (ed.): Bullentin 1 of the Archaeological Soil Micromorphology Working Group. Webpage: http://www.gre.ac.uk/~at05/ micro/soilmain/papers/paper12.htm. (2005-09-09). Carelli, P. 1995: På kulturens bakgård. Arkeologisk undersökning inför Vita husets tillbyggnad Kv. Tegnér nr 1. Lund. Arkeologiska rapporter från Lund 18: 54-57. Carelli, P. & Lenntorp, K.-M., 1994: Kv S:t Mårten nr 2526 Lund, arkeologisk undersökning 1993. Arkeologiska rapporter från Lund 8: 223-227. Clark, P., 1992: Contrasts in the Recording and Interpretation of ‘Rural’ and ‘Urban’ Stratiﬁcation. In: Steane, K. (ed.): Interpreting Stratigraphy 1992 (Lincoln). 17-19. City of Lincoln Archaeology Unit. Lincoln. Clark, P., 2000: Negative Features and Interfaces. In: Roskams, S., (ed.): Intrerpreting Stratigraphy. Site evaluation, recording procedures and stratigraphic analysis. Papers presented to the Interpreting Stratigraphy Conferences 1993-1997, 103-105. BAR International Series 910 Carter, S., 1993: Soil thin sections as an aid to stratigraphic interpretation. In: Barber, J., (ed.): Interpreting Stratigraphy – 1992 Edinburgh. 63-64. AOC (Scotland) Ltd, Edinburgh. Christopherssen, A., 1997: Byer? Hvilke byer? Tanker om byutviklingen i Norge i middelalderen. META 4: 3-26. Clifford, H. T., 1956: Seed dispersal on footwear. Proceedings of the Botanical Society of the British Isles 2: 129131. Cornell, P., 2005: Vad bör göras? Om arkeologi, eklektoretorik och Svestads ordning. META 1: 21-27. Courty, M. A., Goldberg, P., & Macphail, P., 1989. Soils and Micromorphology in Archaeology, Cambridge manuals in Archaeology. 344 pp. Cambridge University Press, Cambridge. Crawley, M. J., (ed.), 1997: Plant Ecology. Second edition. 717 pp. Blackwell Science, Oxford. City of Lincoln Archaeology Unit. Lincoln. Davidson, D. A., 2002: Bioturbation in Old Arable Soils: Quantitative Evidence from Soil Micromorphology. Journal of Archaeological Science 29: 1247-1253. Deetz, J., 1996: In Small Things Forgotten – An Archaeology of Early American Life. 285 pp. Anchor Books Doubleday. New York. Drangert, J.-O. & Hallström, J., 2002: Den urbana renhållningen i Stockholm och Norrköping: – från svin till avfallskvarn? Bebyggelsehistorisk tidskrift 44: 7-24. Elfwendahl, M. & Påhlsson, I., 1991: Palaeoecological/Archaeological correlation – an example from Gävle, Sweden. Laborativ arkeologi 5: 55-59. Ellenberg, H., Weber, E. H., Düll, R., Wirth, V., Werner, W. & Paulissen, D. 1991: Indicator values of plants in Central Europe. Scripta Geobotanica 18. 166 pp. Engelmark, R. 2002: The Dark Earth of Sigtuna. In: Viklund, K. (ed.): Nordic Archaeobotany — NAG 2000 in Umeå, Archaeology and Environment 15: 49-57. Erdtman, G. 1936: New method in pollen analysis. Svensk Botanisk Tidskrift 30, 154-164. Eriksdotter, G., 1996: Byggnadernas stratigraﬁ – exempel på byggnadsarkeologisk metodik. META 1996: 2, 319. Ersgård, L., 1986: Expansion och förändring i Uppsalas medeltida bebyggelse. In: Cnattingius, N. & Nevéus, T., (eds.): Från Östra Aros till Uppsala. Uppsala stads historia VII. 78-100. Uppsala kommun. Uppsala. Faegri, K., Kaland, P. E., & Krzywinski, K., 1989: Archaeopalynology. Pollen analysis of the human environment. In: Faegri, K., & Iversen, J., (red.): Textbook of Pollen Analysis, 175-199. John Wiley & sons, New York. Florén, A., & Dahlgren, S., 1996: Fråga ga det fförﬂutna. En introduktion till modern historieforskning, 323 pp. Studentlitteratur, Lund. Fogelberg, K., Gardelin, G., & Menander, H., 2004: Ve och fasa. En diskussion kring bruket av faser. META 2004:2: 15-29. Dalbäck, G., (ed.), 1982: Helgeandsholmen. 1000 år i Stockholms ström, 504 pp. Monograﬁer utgivna av Stockholms kommun. Stockholm. Fogelberg, K. & Wickström, A., 2003: Single Context i Sigtuna. In: Bodilsen, A., Hjermind, J., Iversen, M. (eds.) 2003. Stratigraﬁens Mangfoldigheder. 4. Nordiske Stratigraﬁmøde. Viborg Middelalderseminar 2: 55-60. Dalwood, H., 1992: The Use of Soil Micromorphology for Investigating Site Formation Processes. In: Steane, K. (ed.), Interpreting Stratigraphy 1992 (Lincoln). 3-6. Forsström, M., 1982: Medeltidsstäderna under 1970-talet – antikvarisk verksamhet och forskning. Bebyggelsehistorisk tidskrift 3: 9-20. 46 Urbanised Nature in the Past Fredén, C., 1988: Marine life and deglaciation chronology of the Vänern basin, southwestern Sweden. Sveriges geologiska undersökning Ca 71, 80 pp. Fredén, C., 2000: Beskrivning till jordartskartorna 10D Karlstad NV och SV. Sveriges geologiska undersökning Ae 140 & 141: 21-64. Grifﬁn, K. & Sandvik, P. U. 1989: Frukter, frö og andre makrofossiler: funksjoner og aktiviteter belyst gjenom analyser av jordprover. Fortiden Trondheim Bygrunn Folkebibliotekstomten. Meddelelser 19, 1-108. Grime, J. P. Hodgson, J. G. & Hunt, R. 1988: Comparative Plant Ecology. 742 pp. Unwin Hyman. London. Gardelin, G., 2002: Nedbrytning av urbana kulturlager. En översiktlig sammanställning av erfarenheter och kunskapsl ge. 28 pp. Kulturen i Lund. Lund. kunskapslä Gustafsson, S. 2000: Carbonized Cereal Grains and Weed Seeds in Prehistoric Houses—an Experimental Perspective. Journal of Archaeological Science 27: 65-70. Gardelin, G., Goksör, S., Johansson Hevrén, C. & Larsson, S., 1997: Askallén, Lundagård. Arkeologisk förundersökning 1996. Arkeologiska rapporter från n Lund 19, 180 pp. Håkansson, T., 1995: Soil Micromorphology. Field strategies and sample preparation. Laborativ arkeologi 8: 1925. Gebhardt, A., 1997: Dark Earth: Some Results in Rescue Archaeological Context in France Machphail, R. (ed.): Bullentin 1 of the Archaeological soil micromorphology working group. http://www.gre.ac.uk/~at05/micro/soilmain/papers/paper13.htm. (21-06-2005). Geraghty, S., 1996: Viking Dublin: Botanical Evidence from Fishamble Street. 118 pp. Royal Irish Academy. Dublin. Gilbert, O. L., 1989: The Ecology of Urban Habitats, 369 pp. Chapman and Hall, London. Giorgi, J., 1997: Diet in late Medieval and early modern London: the archaeobotanical evidence. Society for Medieval Archaeology Monography 15, Oxbow Monography 98:197-213. Goodburn, D., 1995: Beyond the Poste-Hole: notes on stratigraphy and timber buildings from a London perspective. In: Shepherd, L., (ed.), Interpreting Stratigraphy 5 – 1994 Norwich. 43-52. Norwich Castle Museum. Norwich. Grave, P. & Kealhofer, L., 1999: Assessing Bioturbation in Archaeological Sediments using Soil Morphology and Phytolith Analysis. Journal of Archaeological Science 26: 1239-1248. Grifﬁn, K. 1979: Plant remains from Oslogate 7. In: Schia, E., (ed.): De arkeologiske utgravninger i Gamlebyen, Oslo 2, 124-133. Alvheim & Eide Akademisk forlag, Øvre Ervik. Grifﬁn, K. 1982: Utgrävningen på Helgeandsholmen. Analyser av jordprov från sju båtar. Riksantikvarieämbetet och Statens historiska museeer, Rapport 1982:2. 25 pp. Grifﬁn, K., Halvordssen Ökland, R., Jones, A. K. G., Kenwars, H., Lie, R.W. & Schia, E. 1988: De Arkeologiske utgravninger i Gamlebyn, Oslo 5:5, 200 pp. Alvheim & Eide Akademisk forlag, Øvre Ervik. Håkansson, T., 1997: Soil Micromorphology with Special Reference to ’Dark Earth’. In: Miller, U. & Clarke, H (eds.): Environment and Vikings. Scientiﬁc Methods and Techniques. Birka Studies 4// PACT 52: 151-154. Haldorph, J., 1687: Biärköa Rätten, 40 pp. Stockholm. Hållans, A.-M., Karlsson, P. & Tagesson, G., 1999: Kvarteret Dalkaren, Bebyggelse och industri i stormaktstidens Norrköping. Riksantikvarieämbetet. Rapport UV Öst 1999:1, 36 pp. 7 Plates. 1 Compact Disk. Hambro Mikkelsen, P.,2003: Agerbruget. In: Hambro Mikkelsen, P., & Nørbach, L., C.: Drengsted. Bebyggelse, jernproduktion og agerbrug i yngre romersk og ældre germansk jernalder, 115-225. Jysk Arkæologisk Selskab. Mosegård. Hansson, A.-M. 1995: Seeds and Fruits found in the Black Earth of Birka: Excavation Season 1990. Birka Studies 2: 117-122. Hansson, A.-M., 1997: On Plant Food in the Scandinavian Peninsula in Early Medieval Times. Archaeological Research Laboratory. Stockholm University. Theses and Papers in Archaeology B:5, 72 pp. Hansson, A.-M. & Dickson, J. H. 1997: Plant Remains in Sediment from the Björkö Strait Outside the Black Earth at the Viking Age Town of Birka, Eastern Central Sweden. In: Miller, U. and Clarke, H. (eds.): Environment and Vikings — Scientiﬁc Methods and Techniques. PACT 52: 205-216. Hansson, A.-M. & Dyhlén-Täckman, I. 1999: Humlegården in Stockholm, Sweden—glimpses from a garden´s history. Arkeologisk Museum Stavanger — Varia 12A: 89-98. Hansson, H., 1970: Kulturhistoriskt avskräde. Fataburen 1970. 187-196. Harris, E. C., 1979: Principles of Archaeological Stratig- 47 J. Heimdahl raphy, 136 pp. 1st Edition. Academic Press, London and San Diego. Harris, E. C., Brown III, M. R. & Brown, G. J. 1993: Practices of Archaeological Stratigraphy, 296 pp. Academic Press, London. Hedvall, R., Karlsson, P., Persson, B., Persson, K., Stibélius, M. & Tagesson, G. 2000: Stadsgårdar i den senmedeltida stadsdelen Sanden, Vadstena. Riksantikvarieämbetet. Rapport UV Öst 2000:26, 110-111. Hegardt, J., 2005: Adjö Asgeir Svestad. Välkommen, Jes Wienberg. META 2005, 1: 28-34. Heijkenskjöld, R., 1981: Naturlandskapets utveckling i Karlstadtrakten, 90 pp. Karlstad kommun. Karlstad. Heimdahl, J. 1999(a): Växtmakrofossil som indikatorer på vikingatida landskapsutnyttjande på Björkö. Department of Quaternary Research. Stockholm University. Quaternaria B. 17. 29 pp. Heimdahl, J. 1999(b): Kvasten 8, Analys av botaniska makrofossil och sedimentinnehåll. Unpublished report. Stockholm stadsmuseum, 7 pp. Heimdahl, J., 2003: Den urbana naturen. Synen på urbana kulturlager i mötet mellan arkeologi och kvartärgeologi. META 2003, 3: 3-19. Heimdahl, J., 2004: Ögonblick och kontinuitet. Horisontbegreppets användbarhet inom kulturlagerstratigraﬁ. META 2004, 2: 65-74. Heimdahl. J., Karlsson, P. & Menander, H., 2003: Arkeologiskt och geologiskt samarbete i fält — Erfarenheter från den arkeologiska undersökningen i kv. Konstantinopel, Norrköping. In: Bodilsen, A., Hjermind, J. & Iversen, M., (eds.): Viborgs Middelalderseminar 2. Stratigraﬁens mangfoldigheder. 4. Nordiske Stratigraﬁmøde. Viborg 3.-5. augusti 2001: 7-17. Hellwig, M., 1995: Paradieskörner Aframomum melegueta (Roscoe) K. Schum. – Ein Gewürz aus Westafrika im frühneuzeitlichen Göttingen. In: Kroll, H. & Pasternak, R., (eds.): Res archaeobotanicae. International Workgroup for Palaeoethnobotany. Proceedings of the nineth Symposium Kiel 1992: 39-48. Herschend, F., 2005: Argumentation ad hominem idag och för 160 år sedan. META 2005, 1: 53-59. Hitchmough, J., Kendle, T., Paraskevopoulou, A. T., 2001: Seedling emerge, survival and initial growth of forbs and grasses native to Britain and central/southern Europe in low productivity urban “waste” substrates. Urban Ecosystems 5: 285-308. Hjelle, K. L. 2002: Palaeobotanical investigations in Bergen, Norway — development of an urban settlement and the use of plants in the medieval town. In: Viklund, K. (ed.): Nordic Archaeobotany — NAG 2000 in Umeå, Archaeology and Environment 15: 85-94. Hjelmqvist, H. 1963: Frön och frukter från det äldsta Lund. In: Blomqvist, R. & Mårtensson, A. W., (eds.): Fynd från Ultima Thule, 262 pp. Kulturhistoriska museét, Lund. Hjelmqvist, H. 1991: Några trädgårdsväxter från Lunds medeltid. Svensk Botanisk Tidskrift 85: 225-248. Hodder, I., 1992: Theory and practice in archaeology, 285 pp. Routledge. London. Holmqvist-Olausson, L., 1993: Aspects on Birka : investigations and surveys 1976-1989. Stockholm University. Archaeological Research Laboratory. 161 pp. Howe, H. F. & Westley, L. C. 1997: Ecology of Pollination and Seed Dispersal. In: Crawley, M. J. (ed.): Plant Ecology, 262-283. Second edition. Blackwell Science, Oxford. Hyenstrand, E., 1992: Early discoveries in the Black Earth. In: Ambrosiani, B., & Clark, H., (eds.): Investigations in the Black Earth. Volume 1: 23-51. Jacomet, S. 1987: Prähistorische Getreidefunde, Eine Anleitung zur Bestimmung prähistorischer Gersten- und Weizen- Funde, 70 pp. Botanisches Institut der Universität Abteilung Pﬂanzensystematik und Geobotanik, Basel. Jacomet, S. Brombacher, C. & Dick, M. 1989: Archäobotanik am Zürichsee. Ackerbau, Sammelwirtschaft und Umwelt von neolitischen und bronzezeitlichen Seeufersiedlungen im Raum Zürich. Züricher Denkmalpﬂege, Monograﬁen 7, 330-347. Kommissionsverlag Orell Füssli Verlag, Zürich. Järpe, A., Redin, L. & Wahlöö, C., 1979: Ett standardiserat beteckningsschema för sektionsritning. Fornvännen 74: 32-44. Jensen, H. A., 1979: Seeds and other Diaspores in Medieval Layers from Svendborg. In: Jansen, M. H. (ed.): The Archaeology of Svendborg — Denmark, Volume 2, 101 pp. Odense University Press, Odense. Jensen, H. A., 1985: Catalogue of late- and post-glacial macrofossils of Spermatophyta from Denmark, Schleswig, Scania, Halland, and Blekinge dated 13,000 B.P. to 1536 A.D. Danmarks Geologiska Undersøgelse. A6, 95 pp. Jensen, H. A., 1998: Bibliography on Seed Morphology, 48 Urbanised Nature in the Past 310 pp. A.A Balkema, Rotterdam & Brookﬁeld. Johnson, T., 1999: CRC Ethnobotany Desk Reference, 1211 pp. CRC Press, London. Korsmo, E., 1926: Ogräs. Ogräsarternas liv och kampen mot dem i nutidens jordbruk, 410 pp. Albert Bonniers förlag, Stockholm. Karlenby, L., 2002: Blick för Bergslagen. Avdelningen fför arkeologiska undersökningar UV Bregslagen. Årsbok 2002, 72 pp. Kosina, R., 1995: Botanical synopsis of Medieval Wroclaw. In Kroll H. & Pasternak, R. (eds): Res archaeobotanicae – 9th symposium IWGP. 101-116. Kiel 1992. Oetger-Voges-Verlag, Kiel. Karg, S. & Robinson, D., 2002: Secondary food plants from medieval sites in Denmark: fruits, nuts, vegetables, herbs and spices. In: Viklund, K. (ed.): Nordic Archaeobotany — NAG 2000 in Umeå, Archaeology and Environment 15, 49-57. Kostel-Hughes, F., Young, T. P., & McDonnel, M. J., 1998: The soil seed bank and its relationship to the aboveground vegetation in deciduous forests in New York City. Urban Ecosystems, 2: 43-59. Karlsson, H., 2005: En retorisk bumerang. Asgeir Svestad och (u)arkeologin. META 1: 35-40. Karlsson, S., 2000: Studies of coprolites from the Viking Age Birka. In: Karlsson, S., 2000: Vegetation history as a key to living conditions for domestic animals and man in Viking Age Birka. Licentiate thesis, 34 pp. Department of Quaternary Research, Stockholm University. Unpublished. Karlsson, T., 2004: Cheklista över Nordens kärlväxter – verision 2004-01-19 Web page: http://www2.nrm.se/ fbo/chk/chk3.htm (08-09-2005). Karlsson, P., Menander, H. & Heimdahl, H. (in press): Kvarteret Konstantinopel, Norrköping. Arkeologisk rapport, Riksantikvarieämbetet, UV Öst. Katz, N.J., Katz, S.V. & Kipiani, M.G. 1965: Atlas of fruits and seeds occuring in Quaternary deposits of the USSR [in Russian], 356 pp. Nauka, Moskva. Keevill, G., 1995: Process of Collapse in Romano-British Buildings: a review of evidence. In: Shepherd, L., (ed.): Interpreting Stratigraphy 5 – 1994 Norwich. 26-37. Norwich Castle Museum. Norwich. Kenward, H.K., Hall, A.R., 1995: Biological Evidence from 16-22 Coppergate. In: P.V. Addyman (ed.): The Archaeology of York. Vol 14: The Past Environment of York. 435-797. Krammer, K., & Lange-Bertalot, H., 1986: Bacillariophyceae. 1. Teil: Naviculaceae. In: Ettl, H., Gerloff, J., Heynig, H. & Mollenhauer, D. (eds.): Süsswasserﬂora von Mitteleuropa 2/1, 876 pp. Gustav Fischer Verlag, Stuttgart. Krammer, K., & Lange-Bertalot, H., 1988: Bacillariophyceae. 2. Teil: Bacillariaceae, Epithemiaceae, Surirellaceae. In: Ettl, H., Gerloff, J., Heynig, H. and Mollenhauer, D. (eds.): Süsswasserﬂora von Mitteleuropa 2/2, 596 pp. Gustav Fischer Verlag, Stuttgart. Krammer, K., & Lange-Bertalot, H., 1991a: Bacillariophyceae. 3. Teil: Centrales, Fragilariaceae, Eunotiaceae. In: Ettl, H., Gerloff, J., Heynig, H. and Mollenhauer, D. (eds.): Süsswasserﬂora von Mitteleuropa 2/3, 576 pp. Gustav Fischer Verlag, Stuttgart. Krammer, K., & Lange-Bertalot, H., 1991b: Bacillariophyceae. 4. Teil: Achnanthaceae. Kritische Ergänzungen zu Navicula (Lineolate) und Gomphonema. In: Ettl, H., Gerloff, J., Heynig, H. and Mollenhauer, D. (eds.): Süsswasserﬂora von Mitteleuropa 2/4, 437 pp. Gustav Fischer Verlag, Stuttgart. Krause, W., 1997: Charales (Charophyceae). Süsswasserﬂora von Mitteleuropa 18, 202 pp. Gustav Fischer Verlag, Jena. Kristiansen, K., 2005: Hvorfor nysgerrigheden må komme først. Svar till Asgeir Svestad. META 2005, 1: 60-63. Kieffer, R., 2005: Hermeneutik. Nationalencyklopedin. Web page: http://www.ne.se/jsp/search/article.jsp?i_art_ id=202053. (22-06-2005). Kroll, H., 1995: Literature on archaeological remains of cultivated plants (1992/1993). Vegetation History and Archaeobotany 4, 51-66. Kjellén, U., 1996: 1500- och 1600-talslämningar i kvarteret Gripen, Norrköping. Riksantikvarieämbetet, UV Stockholm, Rapport 1996:61. 31 pp. Kroll, H., 1996: Literature on archaeological remains of cultivated plants (1994/1995). Vegetation History and Archaeobotany 5, 169-200. Kompendium i jordartsanalys – laboratorieanvisningar, 1995: Rapporter och meddelanden Nr. 1. 3rd revised edition. Stockholm University, Department of Quaternary science. Quaternaria B, 1: 17-45. Kroll, H., 1997: Literature on archaeological remains of cultivated plants (1995/1996). Vegetation History and Archaeobotany 6, 25-67. Kroll, H., 1998: Literature on archaeological remains of 49 J. Heimdahl cultivated plants (1996/1997). Vegetation History and Archaeobotany 7, 23-56. Kroll, H., 1999: Literature on archaeological remains of cultivated plants (1997/1998). Vegetation History and Archaeobotany 8, 129-163. Kroll, H., 2000: Literature on archaeological remains of cultivated plants (1998/1999). Vegetation History and Archaeobotany 9, 31-68. Krzywinski, K., Fjelldal, S. & Soltvedt, E.-C. 1983: Recent Palaeoethnobotanical Work at the Mediaeval Excavations at Bryggen, Bergen, Norway. In Proudfoot, B. (ed.): Site, Environment and Economy. Symposia of the association for Environmental Archaeology 3, 145-169. BAR International Series 173, Oxford. Krzywinski, K., & Kaland, P. E., 1983: Bergen — from Farm to Town. In Herteig, A. (ed.), The Bryggen Papers. Supplementary Series 1, 1-38. Körber-Grohne, U., 1964: Bestimmungsschlüssel für subfossile Juncus-Samen und Gramineen-Früchte. Probleme der Küstenforschung im südlichen Nordseegebiet. Schriftenreihe des Niedersächsischen Landesinstitutes f r Marschen- und Wurtenforschung. Band 7, 1-32. 21 fü plates. Körber-Grohne, U., 1991: Bestimmungsschlüssel für subfossile Gramineen-Früchte, Identiﬁcation key for subfossil Gramineae fruits. Probleme der Küstenforschung im südlichen Nordseegebiet. Schriftenreihe des Niedersächsischen Landesinstitutes ffür Marschen- und Wurtenforschung. Band 18, 169-234. 24 plates. Lagerlöf, A. G., & Nord A., 2002: Påverkan på arkeologiskt material i jord: redovisning av två forskningsprojekt, 32 pp. Riksantikvarieämbetet. Stockholm. Larsson, E.-L., 1985: Frukter och frön från det äldsta Göteborg — en botanisk makrofossilanalys av fyra 1600-1700-tals kvarter. Enskilt arbete i Biologi. Göteborg University, Department of Botany, section of Systematic botany. 62 pp. Larsson, E.-L., 1986: Särtryck ur Andersson, S., Jönsson Kihlberg, E., & Broo, B., (eds.): Livet i det gamla Göteborg, Archaeobotanical papers in ﬁve chapters, with summary in English. Göteborgs arkeologiska museum. Arkeologi i V Västsverige 2, 122-207. Larsson, S., 1993: Lunda-arkeologin i ljuset av erfarenheterna från de senaste utgrävningarna. META 1993 3-4: 39-65. Larsson, S. 2000: The Hidden Layers. Conditions and conceptual frameworks for urban archaeology in Lund ex- 50 pressed in its praxis for production of data 1890-1990. Archaeologica Lundensia, Investigationes de antiqvitatibus urbis Lundae IX, 378 pp. Kulturhistoriska museet, Lund. Larsson, S. & Johansson Hervén, C., 1998: Källmaterialsproduktion och förståelsehorisonter i stadsarkeologin. META 1998 2: 20-41. Latalova, M., Jarosinska, J., & Badura, M., 1998: Medieval Elbag in the light of present archaeobotanical material. Archeol Polski 43: 147-166. Latalowa, M., Badura, M. & Jarosi�ska, J., 2003: Archaeobotanical samples from non-speciﬁc urban contexts as a tool for reconstructing environmental conditions (examples from Elb�g and Kołobrzeg, northern Poland). Vegetation History and Archaeobotany 12: 93-104. Lawson, T., Hopkins, D. W., Chudek, J. A., Janaway, R. C., & Bell, M. G., 2000: The Experimental Earthwork at Wareham, Dorset after 33 Years: 3. Interaction of Soil Organisms with Buried Materials. Journal of Archaeological Science 27: 273-285. Lehtonen, H., Uotila, K., 2004: Digitalization of stratigraphy. Experiences from the excavations of the medieval town of Naantali, Finland. META 2004, 2: 57-63. Lempiäinen, T. 1985: Plant remains from the ancient lake Mätäjärvi in Turku, SW Finland. Finska Fornminnesf reningen. Iskos 5, 258-271. fö Lempiäinen, T. 1988: Pﬂanzliche Makroreste von dem Innenhof des Qwenselschen Anwesens in Turku, SW Finnland, vom 17.-19 Jahrhunderten. Annales Botanici Fennicae 25, 47-54. Lempiäinen, T. 1994: Plant remains from the castle Kuuisto, Museoviraston Rakennushistorian Osaston Rapportteja 8, 80-99. Lempiäinen, T. 1995: Macrofossil ﬁnds from the medieval Turku, SW Finnland. In: Kroll H. & Pasternak, R. (eds): Res archaeobotanicae – 9th symposium IWGP, 149-164. Kiel 1992. Oetger-Voges-Verlag, Kiel. Limbrey, S. 1975: Soil Science and Archaeology, 384 pp. Academic Press. London Lindeblad, K. 1997: The Town and the Three Farms. On the Organisation of the Landscape in and around a Medieval Town. In: Andersson, H., Carelli, P. & Ersgård, L. (eds.): Visions of the past. Trends and traditions in Swedish Medieval Archaeology. Riksantikvarieämbetet, arkeologiska undersökningar, skriftserie nr 24, 491512. Urbanised Nature in the Past Lindeblad, K., 2002: Urban Archaeology in Östergötland. In: Hedvall, R., (ed.): Urban Diversity – Archaeology in the Swedish Province of Östergötland. Riksantikvarieämbetet. Arkeologiska undersökningar 45: 7-15. Lindgren-Hertz, L., 1999: Mellan landsförsamlingens kyrka och Strömmen, Kv Konstantinopel. Riksantikvarieämbetet UV Linköping. Rapport 1999:4, 25 pp. Ljung, S., 1965: Tiden intill 1568. In: Helmfrid, B. & Kraft, S., (eds.): Norrköpings historia, vol. I, från Forntid till Vasatid, 58-73. Norrköpings Stads Historiekommitté, P. A. Nordstedt & Söner, Stockholm. Lundh, G., Lundqvist, L., Stibéus, M. & Ängeby, G., 1994: De arkeologiska undersökningarna i kvarteret Mercurius i Karlstad. Arkeologi i V Värmland 3, 60 pp. Riksantikvarieämbetet och Länstyrelsen i Värmlands län. Karlstad. Lucas, G., 2001: Critical Approaches to Fieldwork. Contemporary and historical archaeological practice, 246 pp. Routledge. London and New York. Lundqvist, J., 1986: Sediments and soils. Geology and Palaeoecology for Archaeologists. PACT 24: 15-22. Lundqvist, K., 2000: Bidrag till kännedom om begreppet trädgård och om trädgårdsväxternas historia i Sverige. Sveriges lantbruksuniversitet, Institutionen fför landskapsplanering, Alnarp. Rapport 01: 111-112. Mabberley, D. J., 2000: The Plant-Book, A portable dictionary of the vascular plants, 859 pp. 2:d ed. Cambridge University Press. Macphail, R. I., 1994: The reworking of urban stratigraphy by human and natural process. In: Hall, A. R. & Kenward, H. K., (eds.), Urban-Rural Connexions: Perspectives from Environmental Archaeology, 13-43. Oxbow. Oxford. Macphail, R. I., Cruise, G., Engelmark, R., & Linderholm, J., 2000: Integrating Soil Micromorphology and Rapid Chemical Survey Methods: new developments in recording past rural settlement and landscape organisation. In: Roskams, S. (ed.): Interpreting Stratigraphy. Site evaluation, recording procedures and stratigraphic analysis. Papers presented to the Interpreting Stratigraphy Conferences 1993-1997. BAR International Series 910: 7180. Macphail R. I., Galinié, H., & Verhaege, F., 2003: A future for Dark Earth? Antiquity 77: 349-358. Matthews, K., 1993: A futile occupation? Archaeological meanings and occupation deposits. In: Barber, J., (ed.), 1993: Interpreting Stratigraphy – 1992 Edinburgh. 55- 61. AOC (Scotland) Ltd, Edinburgh. McDonnell, M. J., 1997: A Paradigm Shift. Urban Ecosystems, 1: 85-86. McDonnell, M. J., & Pickett, S. T. A., 1990: Ecosystem structure and function along urban-rural gradients: an unexploited opportunity for ecology. Ecology 71: 12321237. McDonnell, M. J., Pickett, S. T. A., Groffman, P., Bohlen, P., Pouyat, R. V., Zipperer, W. C., Parmelee, R. W., Carreiro, M. M., & Medley, K., 1997: Ecosystems processes along an urban-to rural gradient. Urban Ecosystems 1: 21-36. Mcintyre, N. E., Knowles-Yánez, K., & Hope, D., 2000: Urban ecology as an interdiciplinary ﬁeld: differences in the use of “urban” between the social and natural sciences. Urban Ecosystems, 4: 5-24. McLees, C., Nordeide, S. W., Petersén, A., & Saunders, T., 1994: The production of archaeological knowledge: The theory and method of urban excavation. META 1994 2: 3-29. Menander, H., & Karlsson, P., 2002: Hypotheses, Soils and Seeds — Presentation of a Method. In: Hedvall, R. (ed.): Urban Diversity — Archaeology in the Swedish Province of Östergötland. Riksantikvarieämbetet Arkeologiska Undersökningar skrifter 45: 65-76. Millberg, P., 1991: Car tyres as seed dispersers. Fauna och Flora 86: 266-270. Miller, U., & Clarke, H., (eds.) 1997: Environment and Vikings, Scientiﬁc Methods and Techniques. PACT 52: 275 pp. Miller, U., & Robertsson, A.-M., 1982: The Helgeandsholmen excavation: An outline of biostratigraphical studies to document shore displacement and vegetational changes. Second Nordic Conference on the application of scientiﬁc methods in archaeology. PACT 7: 311-328. Moe, D., 1996: Coprolite Material, a Little Used Source for Local Environmental Studies, a Survey with Examples from Pollen-analytical Studies. In: Robertsson, A.M., Hackens, T., Hicks, S., Risberg, J. & Åkerlund, A. (eds.): Landscape and Life—Studies in Honour of Urve Miller, PACT 50: 451-461. Moffatt, S. F., McLachan, S. M. & Kenkel, N. C., 2004: Impacts of land use on riparian forest along an urban – rural gradient in southern Manitoba. Plant Ecology, 174:119-135. Moltsen, A., 1999: Rester af gammelt staldmiljø fra Ninna 51 J. Heimdahl Bangs Plads / Pilestræde i København (KBM 2030). Nationalmuseets Naturvidenskabelige Undersøgelser Rapport 11, 9 pp. Moltsen, A., 2002: Arkaeobotaniske undersøgelser fra middelalderbyen København — metodik og udvalgte eksempler. In Viklund, K. (ed.): Nordic Archaeobotany — NAG 2000 in Umeå, Archaeology and Environment 15: 49-57. Moore, P.D. Webb, J.A., & Collinson, M.E., 1991: Pollen analysis, 216 pp. Second edition. Blackwell Scientiﬁc Publication, Oxford. Mossberg, B., Stenberg, L. & Ericsson, S. 1992: Den nordiska ﬂoran, 696 pp. Wahlström & Widstrand, Stockholm. Mossberg B, Stenberg L & Karlsson T., 2003: Den nya nordiska ﬂoran. 928 pp. Wahlström & Widstrand, Stockholm. Munsell Color, 2000: Munsell Soil Color Charts, 10 pp. 9 Plates. Baltimore, Maryland, USA. Nilon, C. H., Berkowitz, A. R., & Hollweg, K. S., 1999: Editorial: Understanding urban ecosystems: A new frontier for science and education. Urban Ecosystems, 3: 34. Nolin, C., 1999: Till stadsbornas nytta och fförlustande. Den offentliga parken i Sverige under 1800-talet. 393 pp. Byggförlaget. Stockholm. Norman P., & Reader, F.W., 1912: Further discoveries relating to Roman London, 1906-12. Archaeologia LXIII: 257-344. Notelid, N., 2005: Varför begrundan måste komma före disciplinen. META 2005, 1: 41-45. Nydolf, N-G., Runeby, C., Swanström, E., & Zerpe, L., 1992: Occupation Layers as Historical Evidence for Visby – Problems and Interpretations. In: Kyhlberg, O., (ed.): Rescue and Research. Arkeologiska undersökningar. Riksantikvarieämbetet. Stockholm. Skrifter No 2: 98-123. Nygren, C. E., 1934: Karlstad stads historia, fförsta delen omfattande tiden 1584-1719, 427 pp. Nya Wermlandstidningens aktiebolag. Karlstad. Olsson, A., 1999: Spår av avfallshantering i ett medeltida urbant sammanhang. META 1999, 1: 49-65. Oxley, J., 1993: Leave it on the Ground! Preserving Archaeological Deposits in York. META 1993, 2: 58-68 Parr, M. 1987: Spår av det medeltida Norrköping i kvar- 52 teret Kronan. Arkeologisk undersökning 1981. Riksantikvarieämbetet. Rapport UV 1987:10. 28 pp. Persson, C. 1979: Shore displacement during Ancylus time in the Rejmyra area South Central Sweden. Sveriges geologiska undersökning, C755, 23 pp. Persson, C. & Svantesson, S.-I., 1994: Metodik och jordartsindelning tillämpad vid geologisk kartering i skala 1:50 000, SGU Ae. Sveriges geologiska undersökning. Uppsala, 19-20. Petersson, M., 1999: Arkeologiska utgrävningar av fossil åkermark - en metodstudie. In: Ericsson, A., (ed.): Odlingslandskap och uppdragsarkeologi. RÄA skrifter nr. 29. Riksantikvarieämbetet Polcyn, M., 1995: Plant remains in a Early Medieval pot from Ostrów Lednicki. In: Kroll, H. & Pasternak, R., (eds.): Res archaeobotanicae. International Workgroup for Palaeoethnobotany. Proceedings of the nineth Symposium Kiel 1992. 249-260. Preston, C. D., Pearman, D. A. & Hall, A. R., 2004: Archaeophytes in Britain. Botanical Journal of the Linnean Society, 145: 257-294. Påhlsson, I. 1983: Växtlämningar i medeltida avlagringar från kvarteret Svalan, Uppsala. In: Mogren, M. & Påhlsson, I. (eds.): 1983: Svalan, ett medeltida Uppsalakvarter. Riksantikvarieämbetet, Rapport UV 1983:1, 141 pp. Påhlsson, I. 1991a: Bryggaren – ett kvarter i centrum. En medeltidsarkeologisk undersökning i Uppsala. Riksantikvarieämbetet och Statens Historiska Museer. Rapport UV 1991.1, 199-206. Påhlsson, I. 1991b: Arkeologisk för-/slutundersökning. Uppland, Uppsala, Kvarteret Svava (Dragarbrunn 31:1), Fornlämning 88. Riksantikvarieämbetet, 4 pp. (unpublished report) Qualtrocchi, U., 2000: CRC World Dictonary of Plant Names. Common Names, Scientiﬁc Names, Eponyms, Synonyms, and Etymology, 2896 pp. CRC Press, London. Ranheden, H. 1999: Makrofossilanalys från kv. Dalkarlen 3:8 och 10 i Norrköping. In: Hållans, A.-M., Karlsson, P. & Tagesson, G. (eds.): Kvarteret Dalkarlen, Bebyggelse och industri i stormaktstidens Norrköping, Riksantikvarieämbetet. Rapport UV Öst 1999:1, 7 pp. Reading, H.G. & Levell, B.K. 1996: Controls on the sedimentary rock record. In: Reading, H.G. (ed.): Sedimentary Environments: Processes, Facies and Stratigraphy (third ed.), 5-36. Blackwell Science. Cambridge. Reece, R., 1980: Town and country: the end of Roman Urbanised Nature in the Past Britain. World Archaeology, 12: 77-92. Risberg, J., Karlsson, S., Hansson, A.-M., Hedenström, A., Heimdahl, J., Miller, U. & Tingvall, C. 2002: Enviromental changes and human impact as recorded in a sediment sequence offshore a Viking Age town Birka, southeastern Sweden. The Holocene 12: 445-458. Robertsson, A.-M., Karlsson, S. & Helm, K., 1995: Lake Fatburssjön – an Archive of 3000 Years of Environmental History in Stockholm, Sweden. In: Robertsson, A.-M., Hackens, T., Hicks, S., Risberg, J. & Åkerlund, A.(eds.): Landscape and Life—Studies in Honour of Urve Miller, PACT 50: 283-296. Robins, G., 2003: The stratigraphy of violence. In Bodilsen, A., Hjermind, J. & Iversen, M., (red.): Viborgs Middelalderseminar 2. Stratigraﬁens mangfoldigheder. 4. Nordiske Stratigraﬁmøde. Viborg 3.-5. augusti 2001, 79-85. Ros, J., 1996: Stratigraﬁ med erfarenheter från Sigtuna. META 1996, 4: 45-55. Roslund, M., 1997: På drift i tid och rum? Om informationspotentialen i komposita dokumentationsmaterial. META 97:3, 37-53. Roskhams, S., 1992: Finds Context and Deposit Status. In: Steane, K. (ed.), Interpreting Stratigraphy 1992 (Lincoln). 27-29. City of Lincoln Archaeology Unit. Lincoln. Roskhams, S., 2001: Excavation. Cambridge Manuals in Archaeology. 311 pp. Cambridge University Press. Cambridge. Round, F. E., Crawford, R. M. & Mann, D. G. 1990: The Diatoms. Biology & Morphology of the genera, 747 pp. Cambridge University Press, Cambridge. Runer, J., 1999: Kulturlager och verklighetsproduktion. Aktuell Arkeologi VII. Stockholm Archaeological Reports 36: 49-53. Rundqvist, M., 2005: Arkeologer tittar på gamla saker för att ta reda på hur det vara att leva för länge sedan. META 1995, 1: 46-52. Sandegren, R., 1939: Nedre Klarälvsdalens postglaciala utvecklingshistoria. Sveriges geologiska undersökning. Årsbok 33, no. 2. 38 pp. Sandvik, P. U. 1992: Utgravningarne i Erkebispegården 1991, Delfelt 1A og 1B: Botanisk analyse. Arkeologiske Undersøkelser i Trondheim 8, 55 pp. Sandvik, P. U. 1995: The Archbishop´s Palace, Trondheim, Norway – results from the archaeobotanical investigations in 1991. In: Kroll, H. & Pastenak, R. (eds.): Res archaeobotanice – 9 th symposium IWGP, 267-279. Kiel 1992, Oeter-Voges-Verlag, Kiel. Sarnäs, P., 2004: Att ha metoden som mål? Några reﬂektioner med utgångspunkt i matjordsarkeologi. META 2004 2: 75-80 Schoch, W. H., Pawlik, B. & Schweingruber F. H. 1988: Botanical macro-remains, 228 pp. Paul Haupt Publishers, Stuttgart. Schultze-Motel, J. 1992: Literature on archaeological remains of cultivated plants (1989/1990). Vegetation History and Archaeobotany 1, 53-62. Schultze-Motel, J. 1993: Literature on archaeological remains of cultivated plants (1990/1991). Vegetation History and Archaeobotany 2, 47-59. Schultze-Motel, J. 1994: Literature on archaeological remains of cultivated plants (1991/1992). Vegetation History and Archaeobotany 3, 33-61. Schück, A., 1926: Studier rörande det svenska stadsväsendets uppkomst och äldsta utveckling, 472 pp. Stockholms högskola. Uppsala. Sidell, E. J., 2000: Dark earth and obscured stratigraphy. In: Huntley, J. P. (ed.): Taphonomy and Interpretation, 35-42. Oxbow books, Oxford. Sillasoo, Ü. 2002: Gardens and Garden Products in Medieval Tartu, Estonia. In: Viklund, K. (ed.): Nordic Archaeobotany — NAG 2000 in Umeå, Archaeology and Environment 15, 181-192. Simmons, C. S., Sorrensen, C., Walker, R., 2002: Urban rural linkages and environmental change: Addressing the human dynamics of urban ecologies. Urban Ecosystems, 6: 5-8. Solli, B., 1989: Norske middelalderbyer og bydannelse. Et arkologisk utsyn. Viking LII. Oslo. 133-143. Sørensen, R., Bukholm, K., Piløe, L., Henningsmoen, K. E., & Høeg, H. I., 2002: [Abstract] Geology and soils at the ﬁrst Norwegian town – Kaupang, southeastern Norway. Environment and settling along the Baltic coasts through time, p 53. NorFA seminar, 3-6 October 2002 in Pärnu, Estonia. Stockton, E. D., 1973: Shaw’s Creek shelter: human dicplacement of artifacts and its signiﬁcance. Mankind 9: 112-117. Sundborg, Å., & Heijkenskjöld, R., 1972: Klarälvsdeltat i 53 J. Heimdahl Karlstad. Ymers årsbok 1972, 14-25. Svestad, A., 2004: Hvorfor oldsakene må komme først. Svar til Jes Wienberg. META 2004, 3: 38-47. Tagesson, G., 2000a: Bodde dom där nere? Om kulturlagerbildning och avfallshantering i stormaktstidens Norrköping. In: Eriksdotter, G., Larsson, S. & Löndahl, V., (eds.): Att tolka stratigraﬁ. Det tredje nordiska stratigraﬁmötet, Åland 1999. Meddelanden från Ålands högskola 11: 153-173. Tagesson, G., 2000b: När metoden kom till byn. Kontextuell grävningsmetod i Norrköping. In: Eriksdotter, G., Larsson, S. & Löndahl, V., (eds.): Att tolka stratigraﬁ. Det tredje nordiska stratigraﬁmötet, Åland 1999. Meddelanden från Ålands högskola 11: 130-132. Tallantire, P.A., 1976: Provisional key for the Identiﬁcation of Sub-Fossil seeds of Vaccinium spp. Folia Quaternaria 47, 39-40. Thomas, K. D. 1997: Environmental Archaeology and Early Towns. In: Miller, U. & Clarke, H. (eds.): Environment and Vikings — Scientiﬁc Methods and Techniques, PACT 52, 155-165. Walbridge, M. R., 1998: Growing Interest in Urban Ecosystems. Urban Ecosystems 2: 3. Wasylikowa, K. 1978: Plant Remains from the Early and Late Medieval Time found on the Wawel Hill in Cracow. Acta Palaeobotanica XIX: 2, 198 pp. Wasylikowa, K. 1986: Analysis of fossil fruits and seeds. In: Berglund, B. E., (ed.): Handbook of Holocene Palaeoecology and Palaeohydrology, 571-590. John Wiley & Sons Ltd, Chichester. Westholm, G., 1990: Gaturenhållning och hushållsavlopp – Modern service eller medeltida teknik? Om Visbys medeltida stadsmiljö utifrån det arkeologiska materialet. In: Bohm, R: Gotlandia Irredenta. Festschrift ffür Gunnar Svahnström zu seinem 75. Geburtstag, 303323. Sigmaringen. Thorbecke. Wienberg, J., 2004: Arkaeologi, ﬁlosoﬁ og retorik. META 2004, 3: 26-37. Wiethold, J., 1995: Plant remains from town-moats and cesspits of Medieval and Post-Medieval Kiel (SchleswigHolstein, Germany). In: Kroll, H. & Pasternak, R., (eds.): Res archaeobotanicae. International Workgroup for Palaeoethnobotany. Proceedings of the nineth Symposium Kiel 1992. 359-384. 54 Wikström, A. (red.), 2003: Rapport från grävningen av Stora Gatan 2002, 16 pp. Sigtuna Museums Undersökningsverksamhet (SMUV). Wittig, R., 2004: The origin and development of the urban ﬂora of Central Europe. Urban Ecosystems, 7: 323339. Wright, P. J., 2005: Flotation samples and some paleoethnobotanical implications. Journal of Archaeological Science 32: 19-26. Viklund, K., 1998: Cereals, Weeds and Crop Processing in Iron Age Sweden. Methodological and interpretative aspects of archaeobotanical. Archaeology and Environment. Vol 14. Umeå universitet. Umeå. 192 pp. Villa, P. & Courtin, J., 1983: The Interpretation of Stratiﬁed Sites: A View from Underground. Journal of Archaeological Science 10: 267-281. Vuorela, I. 1999: Taphonomy and Sources of Error in Palynological Research, In: Huurre, M., (ed.): Dig it all — papers dedicated to Ari Siiriäinen, 65-69. The Finnish Antiquarian Society, The Archaeological Society of Finland, Helsinki. Vuorela, I., Grönlund, T. & Lempiäinen, T. 1996: A reconstruction of the environment of Rettig in the city of Turku, Finland on the basis of diatom, pollen, plant macrofossil and phytolith analyses. Bulletin of the Geological Society of Finland 68 (2), 46-71. Vuorela, I. & Lempiäinen, T. 1993: Palynological and palaeobotanical investigations in the area of the postmedieval Helsinki old town. Vegetation History and Archaeobotany 2, 101-123. Yule, B., 1992: Truncation Horizons and Reworking in Urban Stratigraphy. In: Steane, K., (ed.), Interpreting Stratigraphy 1992 (Lincoln), 20-22. City of Lincoln Archaeology Unit. Lincoln. Zohar, I., Belmaker, M., 2005: Size does matter: methodological comments on sieve size and species richness in ﬁshbone assemblages. Journal of Archaeological Science 32: 635-641. Urbanised Nature in the Past Den urbaniserade naturen De äldre städernas spår Städer är ofta byggda på resterna av sina forna lämningar. Äldre fyllnadsmassor, husgrunder, gator, avfallshögar och enstaka glömda och borttappade föremål utgör s.k. stadskulturlager. Vid sidan av arkeologiska kulturlämningar vittnar geologiska och biologiska spår om den miljö som funnits i och omkring staden under olika perioder. Det är lätt att se ’staden’ som en motsats till ’naturen’, men miljön i städer utgör också en ekologisk nisch för växter och djur med särskilda krav. I staden skapas också förutsättningar för vissa geologiska processer. På samma sätt som man kan tala om en ”urbaniserad människa/kultur” kan man också tala om en ”urbaniserad natur”. Traditionen att arkeologiskt gräva ut stadskulturlager är ung och uppkom när lagren, och fynden i lagren började klassiﬁceras som fornminnen. Före 1970-talet hade bara en handfull utgrävningar av städer gjorts i Sverige. Stadskulturlager är ofta stratigraﬁskt komplexa och svårtolkade, och inte sällan är de ﬂera meter tjocka, och när arkeologin ställdes inför dem tog utvecklingen den arkeologiska stratigraﬁn fart. Man frågade sig vad alla dessa lager egentligen representerade, vad de bestod av och hur de bildats. En central fråga för tolkningen, och för att kunna förstå fyndmaterial och konstruktioner, blev om dessa lager var medvetet skapade, eller tillkommit omedvetet. I ett tidigt skede kallade man bara de omedvetet tillkomna lagren för kulturlager och man antog att de till stor utsträckning bildats genom att människor hällt ut avfall på gatan och tappat saker – alltså att ett visst omedvetet vardagslivsbeteende gav upphov till kulturlager. På senare år har denna syn förändrats radikalt. Alla lager som skapats av människan deﬁnierar man idag som kulturlager. Den gängse synen är, att den övervägande delen av stadskulturlagren är medvetet skapade. Man är idag kritisk mot synen på den medeltida människan som boende på en ständigt växande sophög. Ändå ﬁnns bilden av medeltidens miljö som genomsyrad av smuts kvar i mångas medvetande, delvis på grund av synen på stadskulturlagren. Den moderna synen på stadskulturlager hänger bl.a. samman med utvecklingen av den arkeologiska grävtekniken. De tidiga grävteknikerna omöjliggjorde upptäckandet av s.k. avröjningar, t.ex. spår efter städning. Modern grävteknik är bättre anpas- sad för att tillvarata dessa typer av information som ﬁnns bevarad i kulturlagren. Både i Sverige och internationellt har stadskulturlager främst varit studieobjekt för arkeologer, men de senaste decennierna har också naturvetare såsom. biologer (arkeobotaniker) och geologer (geoarkeologer), alltmer kommit att knytas till utgrävningarna. Vanligen är det fråga om tillfälliga, i bästa fall regelbundna, besök i fält, eller analys av jordprover som skickas in till laboratorier. Samtidigt har praxis inom exploateringsarkeologin (som står för nästan alla utgrävningar i Sverige) varit att fältarbetet utförs av arkeologer, vilket innebär att den viktiga fälttolkningen av kulturlagren, som är avgörande för slutresultaten, nästan uteslutande görs av arkeologer. Det kan också uttryckas som så att naturvetenskapliga studier av kulturlager ofta saknat en tillfredställande fältfas. En del av detta har berott på outtalad praxis, en annan på uppfattningen att modern arkeologisk grävteknik är knepig att kombinera med deltagande av naturvetare under fältarbetet. Mot denna bakgrund kan den här avhandlingens syfte beskrivas som: Att utveckla och tillämpa metoder för integrerat fältarbete mellan arkeologi och kvartärgeologi/arkeobotanik, och att från ett kvartärgeologiskt/paleoekologiskt perspektiv tolka och rekonstruera miljöutvecklingen och tillkomsten av kulturlager i två olika städer: Norrköping och Karlstad. Jag har fältarbetat på grävningen i kvarteret ”Konstantinopel” i Norrköping sommaren 2000, och under undersökningen av kvarteret ”Druvan” i Karlstad hösten och vintern 2003-2004. Utgrävningarna Norrköping ﬁck stadsprivilegier 1384 och är den äldsta av de två städerna. Namnet ﬁnns omnämnt redan 1283 som en viktig handelsplats. Staden är byggd runt forsarna i Motala ström, där man anlagt kvarnar och bedrivit laxﬁske. Under stormaktstiden växte staden i samband med att kronan anlade vapenfabriker där, och den blev en viktig exporthamn. På 1660-talet skedde stora ombyggnader, gatunätet ritades om och stora delar av den gamla bebyggelsen revs och schaktades bort. Samtidigt omformade man strömfåran så att den blev smalare än den varit tidigare. Ombyggnationen var så omfattande att spåren efter det medeltida Norrköping nästan försvann helt, och det var först 1999, under förundersökningen av kvarteret Konstantinopel, som man hittade spår efter den medeltida staden. Utgrävningsområdet ligger på sluttningen ner mot Motala ström, på en plats, som enligt äldre kartor ligger mycket nära strömmens gamla strandkant. 55 J. Heimdahl År 1589 grundade Hertig Karl (senare Karl IV) Karlstad, som en strategisk utskeppningsplats mot väster. Staden har en möjlig medeltida föregångare som kallas ”Tingvalla”. Den är omnämnd som marknadsplats i området, men har hittills inte lokaliserats arkeologiskt, även om man då och då hittar lösa fynd i området, som visar att människor varit här under medeltiden. Karlstad ligger på Vänerns norra strand, där Klarälven har sitt aktiva delta. Staden är byggd på själva deltaytan, vilket medför att översvämningar har varit regelbundet förekommande i Karlstad genom historien. Under utgrävningen i Norrköping utvecklades metoder i syfte att anpassa de kvartärgeologiska och arkeologiska fältarbetsmetoderna till varandra. Alla stratigraﬁska tolkningar i fält gjordes parallellt, både arkeologiskt och kvartärgeologiskt, vilket ledde till mycket fruktbara diskussioner. Det blev uppenbart att förutsättningarna för förståelsen av kulturlagren kraftigt ökade vid gemensamt fältarbete. Det blev också tydligt att varken arkeologisk eller kvartärgeologisk kompetens enskilt var tillräckligt för att tillfredsställande tolka och förstå stadsstratigraﬁn. Närvaron av de två disciplinerna bidrog till att ﬂera nya typer av fenomen kom att undersökas. Exempelvis lager, som bildats genom ”naturliga” geologiska processer såsom sediment bildade i vattenpölar och som nu framträdde som laminerade (randiga) linser i stratigraﬁn. Linserna användes för att spåra äldre markytor som annars försvunnit i avröjningar och de var vanliga i både Norrköping och Karlstad. Dessa linser visade sig vid ﬂera tillfällen innehålla makro- och mikroskopiska spår efter aktiviteter som ägt rum på nu helt försvunna markytor, och i områden som låg utanför själva undersökningsytorna. Det verkar till exempel som om ölbryggning ägde rum i området ovanför den sluttning, som grävdes ut i Norrköping. Den geologiska undersökningen i Norrköping visar att de äldsta kulturlagren i området bestod av metertjocka fyllnadsmassor av sand och grus, som använts för att någon gång på 1200-talet skapa en terrass på sluttningen, kanske som ett skydd mot översvämningar av Motala ström. Fyllnadsmassorna har delvis blivit dumpade i vatten, antagligen under en översvämning. Utan den geologiska undersökningen hade dessa tidiga fyllnadsmassor antagligen blivit tolkade som naturliga sediment, och inte kommit att ingå i den arkeologiska tolkningen av platsen. Eftersom analys av växtmakrofossil (växtrester synliga för blotta ögat, främst frukter och frön) tillämpades som huvudmetod utvecklades ett system för att kontinuerligt och preliminärt analysera alla stratigraﬁska enheter och arkeologiska lämningar 56 redan i fält. På så viss kunde provtagningen effektiviseras och arkeologerna kunde få svar på vilka växtrester som fanns i olika lager, innan de var helt utgrävda. Förutom att ge svar på hur den lokala ﬂoran i kvarteret såg ut, visade också analyserna t.ex. spår av latrinavfall, kodynga, odlingsväxter och åkerogräs, rester av ölkryddor, exotiska (importerade) växer, men också spår av hantverk som t.ex. smidesloppor. Växterna i Norrköping och Karlstad Växtrester som bevarats i stadskulturlager har ett komplext ursprung. För att kunna tolka sammansättningen av växtfossilen, måste man förstå vilka processer som lett till att de hamnat där. Staden har en lokal ﬂora med växter som trivs i själva stadsmiljön och som kan avslöja hur miljön varit på olika platser, t.ex. vilka områden som betrampats, var avfallsplatser har funnits, och vilka områden som legat i skugga eller sol. Förändringar i lokalﬂoran kan avslöja förändringar i stadsmiljön. En stor mängd frön och frukter kommer in i staden via djur som betat i hagar eller ätit vinterfoder slaget på ängar. Frön och frukter i dynga ger signaler om hur stadens omgivande hag- och ängsﬂora sett ut vid olika tider. En tredje viktig fyndgrupp är frön och frukter från odlade och insamlade växter, som förts in i staden för bearbetning, försäljning och konsumtion. Makrofossil som hittas i spiskonstruktioner eller i latrinavfall kan ge indikationer på vilken typ av kost som konsumerats, vilket i sin tur kan ge antydningar om den lokala ekonomiska situationen och klasstillhörighet hos de inneboende i kvarteret. I Norrköping fanns tydliga tecken på att lokalﬂoran skiftat karaktär genom tiden. Före stadsbebyggelsens tillkomst verkar platsen ha utgjort en del av en betad ﬂodstrandäng med kvarlämnade tuvor av tågväxter och brännässlor. Längre upp i sluttningen fanns ett naturligt strandhak utbildat i sand, och ovanpå detta hak återfanns torrmarksväxter som sandnarv och backtrav. Denna vegetation begravdes helt av de metertjocka fyllnadsmassor som dumpades i sluttningen för att forma en terrass någon gång på 1200talet. Byggnader restes på denna terrass och kreatur hölls i området, vilket ledde till att näringshalten i marken ökade kraftigt. Större delen av marken var förmodligen barmark till följd av intensiv kreaturstrampning, Näringskrävande växter som brännässla och blåmålla växte förmodligen längst husväggar och inhägnader. Likaså fanns en tramptålig ﬂora med t.ex. trampört på vägar och bakgårdar. Sammansättningen av den lokala ﬂora verkar i stor utsträckning ha varit oförändrad från 1300-talet och framåt, Urbanised Nature in the Past men vissa förändringar kan dock spåras, i synnerhet kring mitten av 1500-talet då djurhållningen får ge vika för en mer hantverkspräglad miljö. Vissa örter, t.ex. trampört och vitplister försvinner ur kvarteret, medan andra, t.ex. bolmört och vildpersilja ökar i betydelse. Även rester av bevuxna hustak hittades på platsen. Innehållet i de rikliga resterna av dynga i kulturlagren visar att boskapen betat på fuktiga strandängar dominerade av starr och tågväxter runt Motala ström, utanför själva staden. Det fanns rikligt med spår av kulturväxter i Norrköping. Bland sädesslagen dominerar rågen från 1300-talet och framåt, vilket överensstämmer med historiska uppgifter om Östergötland som en del av dåtidens ”rågbälte”. Det indikerar kanske också att den lokala konsumtionen präglades av lokalt odlade grödor. En annan intressant fyndgrupp utgörs av ölkryddor som hittas i lämningar från 1300-talet och framåt. Förutom den populära humlen verkar ölet fram till 1550-talet också ha kryddats med pors, men därefter verkar humlen ha blivit den helt dominerande ölkryddan. Ytterligare en intressant växt i sammanhanget är älggräs, som är överrespresenterat i prover, som också innehåller humle och pors. Älggräset användes som öl-/mjödkrydda under vikingatiden, men inget är hittills känt om eventuell användning under medeltid, eller i kombination med humle och pors. Från 1550-talet ökar fynden av exotiska och exklusiva frukter som vindruva och ﬁkon, vilket väl överensstämmer med kvarterets utveckling till borgarkvarter. Av särskilt intresse är fyndet av frön från bondtobak. Eftersom import av tobak innebär import av blad, och eftersom ingen bladvävnad hittades tillsammans med tobaksfröerna, indikerar dessa med stor sannolikhet att tobaksodling ägt rum i eller omkring Norrköping. Tobaksodling blev vanlig i Sverige under början av 1700-talet. Dessförinnan odlades växten i liten skala, mest som kuriosum, eller som medicinalväxt. Den tidigaste Svenska odlingen av bondtobak känd från historiska källor ﬁnns angiven 1632 från Uppsala slottsträdgård. Tobaksfynden från Norrköping dateras med hjälp av stratigraﬁn och artefakttypologi till mellan 1560 och 1660, men den tidigare delen av perioden, alltså ca 1560-1610 är troligast. Stämmer denna datering är detta spåren efter den tidigaste odlingen av tobak i Sverige. Spåren efter den tidigaste lokala ﬂoran i Karlstad hittades i deltasedimenten på en meters djup. De visar att deltat kring Kristi födelse varit skogsbevuxet med björk och al, och att snår av ljuskrävande växter som slån och hallon vuxit längst stränderna av de slingrande älvkanalerna. Någon gång, troligtvis under medeltiden röjs skogen och deltat omvandlades till betesmark, karaktäriserad av tågtuvor. Någon gång mellan 1000- och 1600-talet började man hålla boskap i någon form av fållor och vegetationen på den trampstörda marken ersattes lokalt av en ruderatﬂora dominerad av revormstörel och pilört. De omgivande fuktängarna dominerades alltjämt av tåg. När hus började byggas på platsen i slutet av 1500talet tillkom näringskrävande växter (t.ex. nässlor, blåmålla och tiggarranunkel) och tramptåliga växter (ex. groblad och trampört). I mitten av 1600-talet försvann den tidigare dominerande revormstöreln och mer ovanliga växter uppträdde, såsom kålsenap och gat- eller stinkmålla, vilka idag blott sporadiskt förekommer längst hamnar på västkusten. Förekomsten av dessa växter visar på Karlstads dåvarande betydelse som utskeppningshamn till västkusten. Likaså verkar många växter som idag har sitt utbredningsområde begränsat söder om Vänern ha förekommit i Karlstads lokalﬂora på 1600-1700-talet. I lämningarna av ett litet nedbrunnet stall i Karlstad hittades en sammanpressat laggkärl med rester av bevarat hö, som satt fastklibbat på kärlets insida. Jämförelser av innehållet i detta hö med innehållet i den kreatursdynga som fanns på golvet i stallet visar radikalt olika sammansättning, och ger en unik möjlighet att separat urskilja ängs- och betesmarksmiljöerna runt 1600-talets Karlstad. Resterna i spannen består med all sannolikhet av hö från ﬂera olika ängar, och visar att foder tagits från både hävdade och ohävdade ängar. Sannolikt motsvarar sammansättningen den man kan förvänta sig ﬁnna i det vinterfoder som använts lokalt. Sammansättningen av dyngan på stallgolvet är däremot en helt annan och domineras av tåg, och liknar all annan dynga som hittades i Karlstads kulturlager. Detta innebär förmodligen att sammansättningen av dyngan representerar betad hagmark och inte slaget foder. Det innebär att det lilla stallet sannolikt inte använts för vinterstallning av djur, utan att djuren som hållits där nyligen varit på sensommar eller höstbete, förmodligen inför slakt. Idén om funktionen av det lilla stallet som slakteri förstärks av att en grop i närheten innehöll koskallar, vilket därmed skulle knyta gropen till stallbyggnaden. Prover tagna från modern dynga på Mariebergs strandängar visar att frukter från tågväxter är mycket rikligt förekommande, trots att djuren undviker denna växt som lämnas kvar som tuvor. Djuren får i sig tågfrukterna genom att dessa sprids från tuvorna till den omgivande betade vegetationen. Också i Karlstad gjordes rikligt med fynd av rester efter odlade och insamlade växter. Liksom i Norrköping dominerar det lokalt odlade sädesslaget, i detta fall havre. Spår av insamlade frukter och bär, bl.a. björnbär, hallon, smultron och blåhallon, före- 57 J. Heimdahl kommer rikligt. I lager från 1700-talet förekommer också spår av parksmultron, en äldre odlingsväxt som sedermera kom att ersättas av jordgubben. Ett fynd av kryddpeppar från tidigt 1600-tal utgör ett intressant exempel på en mer exklusiv växt, som nyligen börjat öka i användning på kontinenten under denna tid. Endast två arkeologiska fynd av denna växt är tidigare kända från Europa. I resterna av ett nedbrunnet hus från 1700-talet hittades ett antal tobaksfrön. Tobaksodlingen i Karlstad kan knytas till en speciﬁk historisk person: Gustav Clareen, som odlade tobak i trakten mellan 1741 och 1772. Fyndet av tobaken i detta nedbrunna hus skulle därmed kunna knyta det till stadsbranden den 8 juni 1752. Lagrens och horisonternas tillkomst Det är tydligt att bildningen av kulturlager både i Norrköping och Karlstad starkt inﬂuerats av de rådande lokala topograﬁska och geologiska förutsättningarna. Den sluttande topograﬁn i Norrköping skapade förutsättningar för vattentranport av material utför sluttningen så att ﬁnsand och silt ansamlades i de lägre belägna delarna av kvarteret. Ansamlandet av vattentransporterat slam var så omfattande att decimetertjocka lager bildades och fyllde ut gränder och gator mellan husen och kom att skapa ett slags negativa avgjutningar av gatunätet och angränsande byggnader. Det blev därigenom möjligt att, genom studier av det vattenavsatta sedimentets form, ﬁnna avgjutningar av föremål och konstruktioner som fanns på platsen när slammet avsattes men som nu är försvunna. När slammet avsattes i gränderna och på gatorna i stadens lägre belägna delar, ledde de till dräneringsproblem. Detta löstes genom att nya byggnader restes på upphöjda grunder över gatuplanet, vilket skapade nya sänkor av gatunätet, och snart var problemet tillbaka när nytt slam täppte till gatorna, varpå husgrunderna ånyo ﬁck höjas. Följden blev ett cykliskt förlopp i en miljö där erosionskänsliga jordmassor exponerades. Erosion genom regnfall och slamavsättning ledde till en viss kulturell respons som innebar att nya jordmassor exponerades. Förekomsten av denna natur-kulturcykel skulle kunna förklara de tjocka kulturlagren och den rikliga förekomsten av vattenavsatta sediment i städernas svackor. Undersökningar av texturen (kornstorlekssammansättningen) i materialet som använts i husgrunder vid olika tidpunkter i stadens historia, visar liknande tendenser för Norrköping och Karlstad. I det tidiga skedet användes ﬁnkornigt material (silt och ﬁnsand) som hämtats lokalt (förmodligen inte mer än några meter bort). I senare tid skedde en 58 omsvängning, först med användning av en kombination av lokalt ﬁnkornigt material och grövre material, och sedermera till bruk av grovsand och grus, som i Karlstad måste ha hämtades från mer avlägsna täkter. Denna förändring är tolkad som en professionalisering av byggandet. Från att husbyggarna varit inﬂyttande bönder, går ansvaret över till ett byggarskrå med andra traditioner. Det ﬁnkorniga materialet i grunderna kan förklaras av att det på landsbygden, där husen mer fritt kunnat placeras på dränerade platser, funnits en tradition av att ha betraktat grundmaterialet främst som köldisolering, medan de professionella byggarna i städerna främst såg det som en stabil inﬁltrationsbädd. Professionaliseringen av byggandet sker vid olika tidpunkter i de två städerna, i Norrköping sker den omkring 1600 och i Karlstad omkring 1700. Förändringen leder till att de yngre kulturlagren är bättre dränerade vilket gör att organiskt material bevaras sämre än i de äldre lagren. Det grövre materialet står också emot erosion bättre, varför också förekomsten av vattentransporterade lager minskar. Vid studierna av stadsstratigraﬁ har jag funnit det nödvändigt att skilja mellan två huvudtyper av stratigraﬁska enheter (strata): lager och horisonter. Med lager, eller depositioner, menas här ett material som transporteras och deponerats ovanpå en yta och kommit att bilda en sammanhängande enhet. Med en horisont menas en del av markskiktet (oftast de översta centimetrarna) som påverkats av den miljö som existerat ovanför. Det kan t.ex. vara fråga om markskiktet som påverkats av växtrötter, eller att marken trampats. Man kan också tala om lagren som depositionella och horisonterna som postdepositionella stratigraﬁsk enheter. Depositionerna, eller lagren är i allmänhet bildade under kortare tidsperioder, från några sekunder upp till ett antal veckor, medan horisonterna ofta är bildade under betydligt längre tidsperioder. Många tramphorisonter har förmodligen utbildats under ﬂera års tid. I syfte att skapa en teoretisk modell över hur stadskulturlager bildas kan de två huvudtyperna av strata kombineras med den arkeologiska frågeställningen om vilka enheter som bildats under medvetna och omedvetna processer. För att alla lager i stadsstratigraﬁn skall täckas in krävs också att man frågar sig vilka enheter som bildats genom naturliga processer. Modellen, som framträder, är ett nätverk av medvetna, omedvetna och naturliga handlingar, aktiviteter och processer som bidrar till bildning av olika typer av lager och horisonter, men också till deras försvinnande. Modellen möjliggör också en direkt tolkning av de handlingar, händelser och förutsättningar som ligger bakom ﬂertalet stadsstratigraﬁska Urbanised Nature in the Past enheter. Analyserna av makrofossil ger också indikationer på hur kulturlagren och tramphorisonterna bildas. I Karlstad, där det varit möjligt att skilja mellan vinterfoder och sommar-/höstbete, är det uppenbart att dyngan bevarad i stadskulturlagren domineras av det senare. Detta tyder på att bildandet av tramphorisonter hämmades vintertid då marken var frusen. Frånvaron av spår av vinterfoder indikerar också att den dynga, som avsatts i staden vintertid, antagligen hunnit städats undan innan marken tinade och den trampades ner i marken. Detta antyder i sin tur på att gatorna i Karlstad städades regelbundet. Det sammantagna intrycket av bildningen av stadskulturlager är ett komplext system av kulturella och naturliga processer. De mänskliga aktiviteterna i de äldre städerna skapade förutsättningar för en mängd geologiska processer som i sin tur gav upphov till motåtgärder. Den geologiska miljön i de äldre städerna kan både betraktas som aktiverad och i viss mån kontrollerad av människan, samtidigt som människan i viss mån styrts av geologiska förutsättningar och påverkats av geologiska processer. Tack! Som den strikta men förvirrade forskare jag nu kommit att bli, har jag valt att dela in detta tack till olika personer som riktat till olika skaror. Dessa är fem till antalet och individerna inom dem har haft betydelse för avhandlingen på olika sätt. Jag vill inleda med att tacka den första skaran, nämligen mina handledare och mentorer som alltid funnits där när jag behövt dem, först min huvudhandledare Ann-Marie Robertsson. En gång för snart tio år sedan bad hon mig vänligt men bestämt återgå till övningspreparaten när jag försökte se hur mina anteckningar i blyerts såg ut genom mikroskopet. Sedan dess har jag alltid haft stor respekt för vad hon sagt. Hon har genom åren oförtröttligt stöttat mig med sin kunskap, sitt tålamod och sitt kulturella engagemang. Sedan kommer jag till mina biträdande handledare i bokstavsordning: Ann-Marie Hansson (Arkeologiska forskningslaboratoriet) ledde mig genom grinden till arkeobotanikens ängder. Förutom att hon alltid varit mycket generös med sitt kunnande, gav hon mig dessutom en frack som jag sedermera gifte mig i! Jan Lundqvist har handlett mig i de sedimentologiska frågorna och med outsägligt glatt humör hjälpt mig på många sätt. Hans öppenhet och förmåga att ständigt framhålla hur lite han vet (trotts att han vet mest av alla) har inspirerat mig mycket. Mats Regnell var den som först ledde mig över arkeobotanikens ängder. Hans osvikliga tro på mig har många gånger fått mig att må bättre än jag förtjänat. Slutligen vill jag i denna skara också tacka Claes Hättestrand, som inte varit handledare i formell bemärkelse, men som ”utomstående forskare” tog sig tid att läsa min kappa och som kommit med många värdefulla kommentarer. Den andra skaran jag vill tacka är de arkeologer som tagit hand om mig under fältarbetena. I Norrköping, där jag tog mina första stapplande steg bland kulturlagren, och först hade svårt att förstå vad det var jag såg, leddes utgrävningen av Hanna Mennander och Pär Karlsson (UV- öst). De gav mig sprit och värme och såg till att jag hade det bra. De blev sedermera mina medförfattare och genom dem har många dörrar till den arkeologiska världen öppnats för mig. Också de andra arkeologerna i Norrköpingsgropen Anna, Annelie, Clas, Jasmine och Kattis är jag skyldig ett stort tack för allt socialt utbyte av både och kunskapsmässigt utbyte. I Karlstad leddes utgrävningen av Martin Karlsson (Värmlands museum) och Kerstin Fogelberg (UV-bergslagen) som tillsammans med Ebba och Björn gjorde arbetet och vistelsen där mycket angenäm. Jag vill också tacka Kerstin och Ebba för att de delade sin kvart med mig, och Kerstin för att hon hjälpt mig med språket i en av artiklarna. Andra arkeologer som betytt mycket för mig i sammanhanget, är Ingrid Dyhlén-Täckman på Stockholm stadsmuseum, Mattias Bäck och Johan Anund på UV-mitt, Stefan Larsson på UV-syd, Ulrika Söderlind på Stockholms universitet, Rickard Hedvall, Karin Lindeblad, Maria Petersson, Magnus Stibéus och Göran Tagesson på UV-öst. De har alla på sitt sätt bidragit till att jag kunnat utveckla idéerna i denna avhandling. Till den tredje skaran räknar jag mina arbetskamrater. De har alla hjälpt och stöttat mig på olika sätt. Särskilt vill jag nämna Barbara Wohlfarth för att hon visat mig stort förtroende och stöttat mig sen den dag hon började leda avdelningen. Urve Miller för hennes brinnande tro på tvärvetenskap och hennes uppmuntran och stöd genom åren. Tack vare hennes har jag kunnat delta i många konferenser och symposier runt Östersjön. Lars Brunnberg har delat med sig av sin pedologiska kunskap och förgyllt åren genom att ösa ur sin kopiösa fatabur med anekdoter. Jan Risberg har hjälpt mig med diatomékunskaper och Sven Karlsson med pollenkunskaper. De har båda med sin erfarenhet av arkeologer också kunnat le överseende när jag stundom farit ut i frustration över de språkförbistringar som råder mellan kvartärgeologer och arkeologer. Helena Alexandersson höll mig i handen 59 J. Heimdahl när jag började fumla med layoutprogrammet. Jag vill också tacka dem som varit mina doktorandkolleger genom åren. Anna, Greger, Martina, Kristian, Tiit, Gun, Anders, Prem, Linda, Soﬁa, Joshi, Dannile, Jonas, Amélie och Tim som jag haft mycket roligt med. Särskilt vill jag tacka Anders Borgmark för alla spel han laddat ner till mig, all öl han trugat i mig, alla efterrätter han lagat och allt kul vi haft tillsammans, och Jonas Bergman för alla ickegeologiska diskussioner, all badminton och allt bus i källaren. Jag vill också tacka Danny, Siwan, Tina, Martin och Kerstin som hjälpt mig med språket i denna avhandling, och Anna-Lena Anderberg (Naturhistoriska riksmuseet) som hjälpt mig med identiﬁering av vissa fröer som inte ville ge sig tillkänna. Ekonomiskt stöd har erhållits från Gerhard De Geer’s stiftelse. Dateringar har bekostats av Swedish Match North Europé AB via Inga Junhem som också entusiastiskt hjälpte mig med referenser om äldre svensk tobaksodling. Den fjärde kategorin är mina nära vänner som bidragit till min lyckliga status även sedan jag lämnat det otroligt stimulerande dagsverket av att sortera tusentals millimeterstora frön i mycket små burkar. Främst vill jag tacka min vän och före detta sambo Erik som i sanning outat mitt humanist-jag och för- 60 gyllt min tillvaro med långa diskussioner om varia, ﬁlosoﬁ, arkitektur, historia, Victorian ways of Firing Automatically och kostymer. Sedan alla i den mytomspunna Mystiska klubben, däri, förutom Erik, också Danny, Axel, Helena, Caroline, Lotta och Jonas ingår. Tack vare dem har jag gjort många värdefulla internationella bekantskaper som t.ex. Walter Talcott, Patricia Herrington och Robert Tindale, vilka alla bidragit till utvecklingen av detta arbete. Det femte kategorin att tacka är min släkt för att de alltid stöttat mig och, helt okritiskt, tycker allt jag gör är bra. Mina föräldrar vill jag tacka för att de alltid uppmuntrat mitt petimeterintresse för naturen, och mina systrar för att de stått ut med (och fortfarande står ut med) att jag på ett självgott sätt försökt imponera på dem med dessa petimeterkunskaper. Sist, men kanske ändå främst, vill jag tacka min fru Jenny för att hon den där kulna höstkvällen för fem år sedan repade mod och frågade om hon ﬁck se mitt ”laboratorium”. Det ﬁck hon, och jag saknar ord att tacka henne för det.
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project