ArticlePDF Available

Abstract and Figures

This paper presents the outlines of a new EU-funded research program for the long-term history, present-day management and further development of the European landscapes, including their natural and cultural heritage: HERCULES. One of the subprojects of this program (Work Package 2) links archaeological, historical and historical ecological data to the analysis of geo-information in order to develop models of long-term landscape change in three carefully chosen study regions in the Netherlands, Sweden and Estonia. This is framed theoretically by integrating insights from landscape biography, historical ecology and complex systems theory. The linking and analysis of data will be done using a Spatial Data Infrastructure and by means of dynamic modelling.
Content may be subject to copyright.
HERCULES: Studying long-term changes in Europe’s landscapes
Jan Kolen, Carole Crumley, Gert Jan Burgers, Kim von Hackwitz, Peter Howard,
Krista Karro, Maurice de Kleijn, Daniel Löwenborg, Niels van Manen, Hannes Palang, Tobias Plieninger, Anu
Printsmann, Hans Renes, Henk Scholten, Paul Sinclair, Martti Veldi and Philip Verhagen.
This paper presents the outlines of a new EU-funded
research program for the long-term history, present-day
management and further development of the European
landscapes, including their natural and cultural heritage:
HERCULES. One of the subprojects of this program
(Work Package 2) links archaeological, historical and
historical ecological data to the analysis of geo-information
in order to develop models of long-term landscape change
in three carefully chosen study regions in the Netherlands,
Sweden and Estonia. This is framed theoretically by
integrating insights from landscape biography, historical
ecology and complex systems theory. The linking and
analysis of data will be done using a Spatial Data
Infrastructure and by means of dynamic modelling.
1 IntroductIon
In December 2013 a new large-scale program was launched
for the research, protection and management of the
European cultural landscapes within EU Seventh
Framework Programme: HERCULES (Sustainable Futures
for Europe’s Heritage in Cultural Landscapes). The major
aim of HERCULES is “to empower public and private
actors to protect and sustainably manage cultural landscapes
that possess signicant cultural, socio-economic, historical,
natural and archaeological value, at a local, national and
Pan-European level” ( This
will be achieved through academic and applied research by
a network of 13 partners from 11 countries, including
European universities and research institutes, small and
medium-sized enterprises and non-governmental
organisations. Together they will develop an integrative
approach that incorporates diverse stakeholder perspectives
to appropriately address landscapes changes. The University
of Copenhagen and the Humboldt- Universität zu Berlin act
as the coordinators of the program.
HERCULES includes no less than 10 work packages
e.g. for case studies of short-term developments in the
European landscapes, landscape typologies, broad and
ne-scale modelling (scenarios) of future landscapes, the
re-coupling of social and ecological landscape components,
building a Knowledge Hub for landscape research and best
practices in management, etc. The research will focus on
a selection of nine case study landscapes in the UK, Sweden,
Finland, Estonia, The Netherlands, France, Switzerland,
Spain and Greece.
In this paper we present an overview of Work Package 2,
as it brings together archaeological, historical and ecological
data and insights in order to reconstruct the changes that
have taken place in the European landscapes from earlier
prehistory (the Neolithic) up to the present. In many ways,
these changes are still operating as landscape forming
processes in the European landscapes we live and work in
today. This is the case, for example, where water
management and land use had long-term or delayed effects,
such as in the extensive water management systems in the
Netherlands, in Mediterranean landscapes that suffer from
erosion, ‘desertication’ and loss of biodiversity, in urban
areas with long histories of ‘path-dependent’ trajectories of
urbanization, etc. The research in Work Package 2 is
coordinated and conducted by an interdisciplinary team of
archaeologists, geographers, experts in geo-information
science and land use managers from the Netherlands (Spatial
Information Laboratory, VU University; Faculty of
Archaeology, Leiden University), Sweden (Dept. of
Archaeology and Ancient History, University of Uppsala),
Estonia (Estonian Institute of Humanities, Tallinn University)
and the UK (Forest Communication Network Ltd.).
2 StudyIng long-term changeS In europeS
cultural landScapeS
Within HERCULES, Work Package 2 (from now on ‘WP2’)
specically addresses the long-term changes in the European
landscapes. Its objectives are:
1. To dene an innovative theoretical framework for
understanding the long-term development and transformation
of landscapes, drawing on recent insights from geography,
landscape archaeology, (historical) ecology, anthropology
and information science.
2. To develop and test an infrastructural facility for
retrieving and linking archaeological, historical and
ecological data and geo-information (SDI) to support the
interdisciplinary study of landscape change.
3. To develop dynamic and interpretive models for
analysing long-term trends in landscape history in three
98163_APL 45_15_Kolen.indd 209 8/05/15 08:45
3 the theoretIcal framework: IntegratIng
landScape bIography and hIStorIcal ecology
In 1996, The Danish-American geographer Kenneth Olwig
tried to synthesize ecological/environmentalist and cultural/
social science approaches in landscape research, based on a
thorough investigation of the origins of the landscape
concept, by re-introducing what he calls the ‘substantive’
nature of landscape (Olwig 2002; 1996). Although such a
synthesis of paradigms, if possible at all (as many
researchers oppose the idea of paradigms itself), is not an
explicit goal of the integrated long-term approach to be
developed in HERCULES, there is certainly a link between
WP2 and Olwig’s endeavour. WP2 too aims at developing an
integrated approach to the study of landscapes, combining
the long-term perspective of archaeology and history with
recent insights from cultural ecology, anthropology and
geography. To achieve this goal, elements of recent
‘landscape biography’ and ‘historical ecology’ will be used
as theoretical building blocks. Landscape biography and
historical ecology are among the prominent emerging
approaches to the study of long- term landscape history,
and are now being combined and integrated with a complex
systems approach to human-land and human-nature
Landscape biography studies the long-term transformations
in landscapes, preferably from prehistory to the present,
viewing landscape at each point in time as a complex
interplay between social and economic developments,
culturally specic perceptions of the environment, the history
of institutions and political formations, and ecological
dynamics (Kolen 2005; Roymans et al. 2009; Kolen et al.
2015). As a historical research strategy, it expresses a strong
sense of the multi-layered nature of landscapes. It
acknowledges the non-linear and path-dependent character of
landscapes and the active role they play in the life histories
and social memory of people (cf. Ingold 2000). This means
that landscapes are not only seen as the (interim) outcomes
of drivers, but in themselves are considered drivers for
social, economic and climate change as well.
Historical ecology, which emerged in the US within the
Boasian paradigm, developed a practical framework of
concepts and methods for studying the past and future of the
relationship between people and their environment (Balée
1998; Balée and Erickson 2006; Crumley 1994; 2012;
Hornborg and Crumley 2007; Meyer and Crumley 2011).
While historical ecology may be applied to spatial and
temporal frames at any resolution, it nds particularly rich
sources of data at the ‘landscape’ scale, where human
activity and cognition interact with biophysical systems, and
where archaeological, historical, ethnographic, environmen-
tal, and other records are plentiful. The term historical
ecology draws attention to a denition of ecology that
case study regions: the river landscape of the central
Netherlands, Uppland in the eastern part of central Sweden,
and Kodavere/Vooremaa area in the south-eastern part
of Estonia.
In the process, WP2 tries to tackle a number of key issues
in current landscape research and applied landscape
assessment. Thus a major issue is how to practically envision
future landscapes on the basis of a more sound understanding
of landscape changes that already started many centuries or
even millennia ago, such as increased human intervention in
river and lake systems and shoreline dynamics. This ts in
with the overall aim of HERCULES to contribute to future
management aspects by enhancing the role of long-term
thinking and analysis in Geo-design, urban planning,
landscape design, and stakeholder involvement.
Current historic landscape assessments are often poorly
matched with the needs of planners, policy makers and
public interest groups. Such knowledge communities actively
contribute to the further development of landscapes and
regions, including their heritage (Fairclough and Grau Moller
2008; Janssen et al. 2014; Kolen et al. 2014). WP2 wishes to
introduce these important knowledge communities to new
possibilities for understanding complex interactive processes
over the long term by clarifying the powerful roles of
narrative, social memory, and practical experience (from the
past and the present) in collaboration and design.
In WP2, landscape will be conceived of as being both a
socially constituted and meaningful whole and a dynamic
system that enables people to adapt to changing climatic and
environmental conditions. Starting from this broad denition,
the work package explores the methodological potential of
historical ecological and biographical approaches to
landscape (Crumley 1994; 2015; Kolen et al. 2015; Meyer
and Crumley 2011; Roymans et al. 2009). In order to assess
the long-term interactions between social and natural drivers
more accurately and systematically, the concepts of historical
ecology and landscape biography will be integrated with a
complex systems approach (see 3. Theoretical framework).
This has resulted already in a protocol for studying long-term
landscape dynamics (Crumley 2014), which will now be
tested and applied in some of the HERCULES study
landscapes, more particularly in the river delta landscape
(including the Holocene peat area) of the central and western
Netherlands, the Uppland area in Sweden and Kodavere/
Vooremaa in south-eastern Estonia. The technical aspects of
the research, which involves the building of dynamic models
and a Spatial Data Infrastructure (SDI), will be done in
collaboration with the Knowledge Hub of the program and a
core team of WP3, which explores the possibilities of
dynamic modelling for understanding short-term and current
changes in the landscape.
98163_APL 45_15_Kolen.indd 210 8/05/15 08:45
periods, and other temporal divisions as ripe for research and
not rewalls that protect temporal specialties. Such a
trans-temporal approach has several advantages. Particularly
important for planning and heritage, the coupled human/
environment system can be analysed with regard to effective
management strategies under specic (local, regional) social
and environmental conditions and the results used to
formulate future scenarios.
In order to tackle these challenges, the protocol for
understanding long-term landscape dynamics produced by
the WP2 team (Crumley 2014) starts from 15 premises.
Here we present only a selection thereof:
Much of what we know about landscape changes in the
past cannot be based on extrapolation from present
conditions. Yet initial conditions of landscape systems
are a strong predictor of later states. Past decisions
shape and constrain subsequent ones and small
differences are disproportionately the cause of later
circumstances. This is called path dependence. Thus
physical infrastructure, social practices, and other
conditions can impede necessary system-wide change.
Knowing a landscape’s history can be seen as using
completed experiments undertaken in the laboratory of
the past. Most (pre-)historic forms of land use have
proven not to be sustainable, but it is true that their
persistence is, at least in part, witness to their utility.
Landscapes have their own temporalities and rhythms,
in relation to but distinctive from individual and
community life cycles. The past is also always present
in the landscape of ‘today’. All landscapes incorporate
“the powerful fact that life must be lived amidst that
which was made before” (Meinig 1979, 44). Thus
landscape biography and historical ecology view
landscapes as palimpsests that are transforming
continuously, both through conscious interaction by
people with the material past in the environment and
through less conscious forms of agency. This again
illustrates that landscapes cannot simply be seen as the
outcomes of drivers, but that landscapes themselves are
also drivers of social, political and economic change.
Historical ecology and landscape biography both study
long-term transformations in landscapes from prehistory
to the present (Crumley 1994, 2015; Kolen et al. (eds.)
2015; Meyer and Crumley 2011; Roymans et al. 2009).
It is important to realize that the disciplines contributing
to this exercise, like landscape archaeology, historical
geography, historical anthropology and palaeo-ecology,
explore quite different datasets covering different
time-intervals and aspects of landscape change. These
datasets and the methods used to analyse and interpret
them must be related and integrated in systematic ways
in order to synthesize long-term changes.
includes humans as a component of all ecosystems and to a
denition of history that goes beyond the written record to
encompass both the history of the Earth system and the
social and physical past of our species. It provides tools to
construct an evidence-validated, open-ended narrative of the
evolution and transformation of specic landscapes, based on
records of human activity and changing environments at
many scales. Historical ecology offers insights, models, and
ideas for a sustainable future of contemporary landscapes
based upon this comprehensive understanding of their past.
Being complex systems, landscapes are self-organizing
and exhibit what are known as ‘emergent properties’, which
cannot be deduced from the individual natural or cultural
components of the system. Agent-based perspectives on
complex systems (cf. Van der Leeuw and McGlade 2013;
Bentley and Maschner 2003) combine the principles of
complex systems theory with the concept of interacting
agents. Dynamic modelling allows to study whether
developments inevitably lead in a certain direction (path
dependence), and whether different scenarios will produce
similar outcomes (equinality). By this, it is very suitable for
exploring long-term developments in cultural landscapes,
allowing to test different hypotheses of the development of
the cultural heritage embedded in these landscapes. As
dynamic modelling may also be explored to create insights
on how micro-scale processes give rise to macro-scale
phenomena, it is of great interest to landscape archaeology,
where we can usually only observe the macro-scale results of
micro-scale actions in the past. Several archaeological studies
have used dynamic modelling for this purpose (Kohler et al.
2007; Wilkinson et al. 2007).
Together, these frameworks encompass the range of
variation that is currently found in international landscape
studies. While the frameworks largely overlap, landscape
biographic approaches focus on the regional scale of analysis
and are more explicitly phenomenological and aimed at
heritage studies, while historical ecological approaches are
multi-scalar and are more comprehensive and explicitly
empirical. Both frameworks embrace the stakeholders,
planners, and managers of landscapes. For the rst time, the
HERCULES program aims at integrating the so far separate
concepts of landscape biographies and historical ecology
with a complex systems-based perspective on cultural
landscapes. With spatial dynamic models specically
designed for the needs of interdisciplinary study of landscape
change, HERCULES intends to provide landscape
researchers with new tools to understand long- term
developments in European landscapes by more effectively
linking archaeological, historical, ecological and social data.
By integrating landscape biography, historical ecology and
complex systems theory, WP2 wishes to realize a
trans-temporal approach to landscape, treating epochs,
98163_APL 45_15_Kolen.indd 211 8/05/15 08:45
Data and information that (can) inform landscape
researchers and stakeholders about the political
(territorial) and religious aspects of past landscapes,
like archaeological databases (burial sites, ritual
depositions), archival sources (monasteries, parishes,
manorial estates, etc.), cartographical databases
(historical maps), databases for specic monuments and
religious architecture (like churches), etc.;
Data and information that (can) inform landscape
researchers and stakeholders about past experiences and
meanings of landscape, like databases for eld and
place names, oral history databases, cartographical
databases (historical maps), visual databases for
landscape painting and historical photography, etc.
It has been acknowledged recently that in order to explore
and combine multidisciplinary datasets optimally these kinds
of data could best be organized by means of a so-called
Spatial Data Infrastructure or SDI (De Kleijn et al. 2014).
The core function of an SDI is to enable users to share
geospatial information beyond the level of a single institute
or organization. This need is generally found in landscape
research. In understanding what an SDI encompasses we
make a distinction between the user objectives, technological
components, Geospatial Information (GI)-literacy, content
and governance (De Kleijn et al. 2014). The combination of
the GI-literacy and the objectives determine the extent to
which the technological components need to be developed
and the content to which access is needed.
At the core of an SDI lies the technical infrastructure of
services, varying from data viewing services to download
and more complex processing services. On top of these
services applications can be built with which users, with
different objectives and GI-literacy levels, can perform their
tasks. The applications through which the services can be
accessed vary from web viewers for users to view and
validate data, to dedicated GIS software with which
modelling experts can perform complex analyses (e.g.
ArcMap, Quantum GIS, GeoDMS, MapINFO, etc.).
For governance of an SDI three aspects can be distinguished.
First, a party coordinative institute has to take the on leadership
and ensure long-term viability and educate where necessary
users how to use the tools. Second, the users’ requirements for
functionality and content have to be closely looked at in order
to ensure that their needs are translated to (technical)
requirements for the SDI and the applications that are
implemented on top of it. Third, considering the content, the
management of who can use what information for which
purposes is also a fundamental part of SDI governance.
Although the current trend is to publish data in the public
domain as open data, to which the HERCULES project also
strives to, some data cannot be put in the public domain due to
privacy issues and restrictions of the data providers.
Together, historical ecology and landscape biography
can link social memories to the long term, connecting
the micro-histories of places to broad-scale develop-
ments, and integrating experience and process. One of
the routes to this end is by the study of how, in different
mnemonic, religious and social systems, memories,
values and ideas concretely interact with the material
world of which landscapes are an integral part
(e.g. Küchler 2002).
Taking the above set of premises as its methodological
starting point, HERCULES will not produce a single
paradigm but rather offers a toolbox of concepts and
competencies (cf. De Kleijn et al., 2014; Kolen et al. 2015;
Meyer and Crumley 2011; Crumley 2015). At the same time,
the premises can be chosen as explicit theoretical guidelines
for research projects that tackle long-term changes in cultural
4 a SpatIal data InfraStructure (SdI) for
landScape reSearch
The availability of digital tools and data to study long term
changes in the landscape has, over the last decade, grown
tremendously. Landscape scholars and landscape practitioners
are more and more digitally skilled and the use of Geospatial
technologies has grown signicantly. Landscape research is
nowadays unthinkable without the use of Geographic
Information Systems (GIS) software to analyse, and Spatial
Data Infrastructures to systematically store and share digital
spatial information. The theoretical framework proposed in
WP2 promotes the integration of data and perspectives
produced by academics and Land Management practitioners.
Given their explicit multidisciplinary aim, studies of the
long-term history of landscapes and ecosystems make use of
datasets from various sources. Optimally, these will include a
combination of the following data:
Data and information that (can) inform landscape
researchers and stakeholders (spatial planners,
landowners, heritage managers, local interest groups
etc.) about the natural characteristics and physical
properties of the landscape, both past and present, like
geological and soil data, soil maps, digital elevation
maps, palaeo-geographical maps, botanical data, data on
climate and climate change, etc.;
Data and information that (can) inform landscape
researchers and stakeholders about social economic land
use and land use systems, both past and present, like
archival sources, cartographical databases, archaeologi-
cal databases (e.g. large-scale vectorised excavation
plans and survey databases), specic soil data and
botanical data, databases for historical landscape
features, monuments and historical urban structures,
98163_APL 45_15_Kolen.indd 212 8/05/15 08:45
visualization and exchange of spatial data sets. Current
approaches to mapping and visualizing landscapes of the past
can be described as predominantly static. Typically, they only
offer snapshots of archaeological or historical periods,
showing the places where people settled and used the
landscape, as well as the environmental setting itself (notably
in the form of palaeo-geographical, or better said
palaeo-geomorphological reconstructions). Predictive models
(which are essentially statistical extrapolations) are then
applied to predict the possible distribution of archaeological
remains (see e.g. Kamermans et al. 2009; Verhagen and
Whitley 2012). In this approach, time periods are lumped,
and the dynamics of landscape transformation are obscured,
tacitly assuming that in each period people were experienc-
ing the landscape as a kind of tabula rasa.
Approaches to overcome this static approach are now
emerging in geographical and archaeological landscape
research. Notably, techniques like Agent-Based Modelling
(e.g. Kohler et al. 2007; Wilkinson et al. 2007) and
Dynamical Systems Modelling have great potential for better
understanding the observed patterns of settlement and land
use, and the processes that led to these patterns (cf. Bentley
and Maschner 2003; McGlade and van der Leeuw 2013).
Such models are designed as heuristic tools, for example to
build scenarios of population development or to develop
more sophisticated theories of human (spatial) behaviour and
practices, but the outcomes should not primarily be seen and
used as accurate predictions of for example past land use.
It can therefore be doubted whether these models, in their
current state of application, would be of much use to heritage
professionals. From their point of view, information needs to
be accurate and usable, rather than multi-interpretable
For these reasons, WP2 will not adopt ABM as a core
modelling technique, despite its high potential for academic
analysis. The outcomes would not, at this stage, be able to
inform heritage policy, and hence not serve the overall aims
of HERCULES. WP2 will therefore adopt and further
enhance models that are able to close the gap between static
mapping aimed at heritage professionals and dynamical
modelling designed for academic research. This will be done
by developing models that link path dependencies in (pre)
historic land use to predictions and mappings that will make
sense in a heritage context.
Within WP2 the concrete possibilities of dynamic
modelling will be explored by adopting a case study
approach. This will be done for three different case study
landscapes, each being representative for more widespread
environmental and climatic conditions within Europe:
Atlantic (the river landscape of the Central Netherlands),
Boreal (the Uppland area, Sweden) and continental European
conditions (Kodavere/Vooremaa, Estonia). A Mediterranean
The study of long-term landscape change would benet
considerably from improved availability of data about the
history and heritage of the landscape and functionalities with
which the data can be processed and shared through an SDI.
In the process of studying long-term landscape change, ve
areas in which an SDI has the potency to play an important
role can be distinguished.
First, an SDI offers functionalities to integrate digital
spatial data (also from different repositories e.g. different
universities, governmental institutes etc.) in structured way
enabling users to work systematically.
Second, an SDI offers functionalities to communicate
historical and heritage spatial data to various stakeholders
ranging from landscape historians and heritage experts to the
people of the place for purposes of validation.
Third, an SDI offers functionalities to process and/or
download data into specialist software with which complex
long-term landscape change models can be developed and
Fourth, an SDI offers functionalities to share the models
and the outcomes of landscape research dynamically,
allowing changes to the data to automatically update the
Fifth, an SDI offers functionalities to disseminate the
research results as services that can be part of the Knowledge
Hub developed in HERCULES or existing local heritage
management data infrastructures.
The areas/themes identied above, except for the fth, are
not to be seen as separate phases, which follow up on each
other, but are to be approached as stages through which long
term landscape modelling goes through in several iterations.
A schematic overview of the SDI is shown in gure 1.
It shows how the data servers are related to the clients and
shows how the SDI aids landscape change modelling. It also
shows how the SDI components are related and how the SDI
interacts and can be integrated with HERCULES’ Knowledge
Hub developed in WP 7. The spatial modellers work
foremost with professional GIS software, whereas the
past-oriented and future-oriented landscape professionals and
local stakeholders provide feedback through workshops using
decision support tools. These last-mentioned groups are also
providing feedback, making use of custom-made data
viewers. Finally the project outcomes and results will be
transferred to the HERCULES knowledge hub, providing
insights in best practices for future landscape research.
5 modellIng landScape change
As stated above, the SDI is essential for giving stakeholders
access to relevant spatial information and validate outcomes
of spatial models about the past and current landscape.
However, in itself, it will provide nothing more than a
framework to accommodate the management and
98163_APL 45_15_Kolen.indd 213 8/05/15 08:45
wetlands and more open cultivated areas. In the Roman
Period, the study region formed the north-western part of the
Roman frontier on the continent. By then, land use had been
intensied considerably, creating a more open landscape with
an increased human impact on the water system.
In about 1000 AD, the inhabitants of the river villages in
the study region began building embankments along major
rivers like the Rhine and Meuse (Van de Ven (ed.), 1993;
Harten 2000). Along with the villages themselves, elds and
gardens occupied the highest parts of the banks, while the
slopes down to the ood basins behind the banks were used
as communal meadows and pastureland. In the period from
800 to 1250 AD, towns in the Dutch river area expanded
signicantly and there was growing demand for agricultural
products. To satisfy this demand, the agricultural land area
had to be extended to the low-lying peat areas and river
basins (Harten 2000; Renes 2005). But before these areas
could be drained and reclaimed, embankments had to be built
along the river courses and any obstructing ones had to be
dammed. Several centuries later, the still remaining open
area (Puglia, Italy) will be used for comparison. All case
studies, most notably the Dutch and Swedish ones, will start
from the premises dened in the protocol (see 3. Theoretical
framework). For each of the three study regions a somewhat
different modelling approach will be adopted to achieve this
aim. Therefore, we will briey introduce each of the case
study landscapes below.
6 caSe Study landScapeS
6.1. The Dutch river area
Parts of the Dutch river landscape were occupied already
during the Mesolithic and Neolithic. Initially, land use will
have been limited to the stream ridges of the rivers and the
adjacent parts of back swamps, as well as on Pleistocene
river dunes and their surroundings. In the Middle and Late
Bronze Age signicant sections of the stream ridges were
transformed into true rural landscapes, with scattered (and
roaming) farmyards with associated burial mounds, gardens,
eld systems and roads (Arnoldussen 2008). This rural
landscape was part of a mosaic environment with forest,
ClientsUsers Servers
Servers with Geodata
Future users People of the Place Spatial modeller(s)
History and Heritage Experts/
Landscape managers
Heritage agencies,
GIS modelling software;
ArcMap, QGIS, MapINFO,
GeoDMS etc.
(including direct scripting
on databases)
Decision support tools
In workshops (WP8)
(e.g. touch tables, smart
phones, tablets etc.)
Geoplaza viewer
Data portal
HUB Data Server
Geoplaza (VU)
municipalities, universities,
landscape organisations etc.
Data and feedback
Data and feedback
Modelling long term
landscape change
(high GI-literacy)
-Validating data
and models
-Using outcomes
long term modelling
for planning purposes
(medium/low GI-literacy)
-Validating data
and models
-Participating in
future landscape planning
for planning purposes
(medium/low GI-literacy)
Data and models
Figure 1 Schematic overview of a Spatial Data Infrastructure (SDI)
98163_APL 45_15_Kolen.indd 214 8/05/15 08:45
connect communities into smaller regional groups, so-called
river-based communities (Jordan 2003). Another such natural
connective element is a water catchment area. A water
catchment area constitutes the area from which all run-off
water comes together in a point or in a stream. A watershed
is the boundary between two such areas. Typically, a
watershed is a height where the rain falls on two different
sides forming two different water catchment areas. This
method has been used frequently in studies concerning
physical geography and to some extent in the study of
prehistoric regions where it has been argued that water
catchments have inuenced the development of territories
and administrative regions (Löwenborg 2007; Von Hackwitz
2012; Wijkander 1983).
For these reasons, the Uppland case focuses on the
modelling of shoreline dynamics in relation to isostatic land
rise. The model used here is based on a shoreline method
focused on archaeological sites combined with an isolation
method built on analyses to determine when lakes were
isolated from the sea. A regression equation is used for
making a shoreline reconstruction in order to consider both
the isostatic uplift and the eustatic variations. This means
that the reconstructions will be more accurately calculated,
especially for larger areas, as the uplift is uneven between
different land areas. Further, the shoreline can be modelled
from any given BP value which means that a site can be put
in its specic time context in terms of shoreline displacement
as long as there is a valid BP value (Sund 2010). The
regression equation used here was originally developed by
Risberg et al. (2007) and further developed by Sund for the
area of Eastern Central Sweden (Sund 2010). The accuracy
of Sund’s model is comparable with that of Risberg (Risberg
et al. 2007), but with the advantage of generating a
contemporary shoreline over a larger area (Sund 2010, 27).
The applied model is generic and well suited to create a
model for the Uppland region as a whole. Local deviations
might occur as topographic thresholds could later have been
eroded, making it difcult to accurately model the shoreline
in detail at every point, but overall the model would be fairly
accurate and relevant for the analyses proposed here.
Additionally, water catchments can be calculated from a
digital elevation model (DEM) using a set of hydrological
functions in a GIS (Geographical Information System). Pour
points, the points on the service at which water ows out of
an area, are determined. Relevant pour points are selected,
considering the modelled shoreline, and from the pour points
drainage basins can be calculated to identify the upland area
that is hydrologically joint at the pour point. The pour point
would also act as an important social node in the landscape,
connecting everyone using the upstream watercourses, and
thus forming a ‘natural’ region that would be easily
recognized. If there is a pronounced isostatic land rise in the
spaces between the village embankments were closed off and
long, uninterrupted dikes were built. This process was
completed in most parts of the Dutch delta by about 1300
AD. Inside the dikes, where in winter especially the river
water could dam up to a signicant extent, river forelands
were created.
Thus over the course of ve centuries, from 1000 to 1500
AD, the Dutch delta changed dramatically (Van de Ven 1993;
Renes 2005). It was transformed from an open delta where
the rivers had free reign and where large areas were taken up
by fens and marshes to a tightly ordered agricultural territory
under human control. With their far-reaching interventions
such as dike building, the inhabitants of the Dutch river
landscapes unconsciously reset the environmental agenda for
themselves. In the long run, their reshaping of wetlands and
stream valleys had unexpected repercussions, like dike
breaches and large-scale oods.
In the Dutch river delta, the so-called Land Use Scanner
(LUS) will be deployed to model land use changes (Hilferink
and Rietveld 1998; Koomen et al. 2011). This modelling
framework was originally designed for predicting land use
development in the near future, based on information about
the current situation and the hypothesized development of
future land use demand. The allocation methods applied in
LUS are the logit-based model to determine probabilities and
a discrete allocation method to generate an allocation that is
optimal given the suitability of different plots within the
region. A major advantage of this approach is that it shifts
the pervading focus in archaeology from local settlement
sites to various landscape scales as the object of interest.
It also looks at the landscape from the angle of its use, rather
than from the dominant, geomorphologically based point of
view aiming at predicting the landscape’s suitability for
settlement – which is of course only one aspect of what
people did in the past. The LUS is particularly suitable for
the Dutch river delta, given the very extensive data available
on settlement distribution and palaeo-geographic
reconstructions for the area over long periods of time.
6.2. Uppland, Sweden (gure 2)
Since the regressive shoreline displacement in Uppland,
like elsewhere in Sweden, changed the landscape
considerably over millennia and offered new land for
occupation, it is necessary to examine how the changing
topography inuenced the choice of locations for settlement
and land use.
A strategy to reconstruct this aspect of regional landscape
change is to take the physical characteristics of the landscape
into account, such as the presence of fresh water. Proximity
to fresh water is important in the choice of settlement
locations. Streams, lakes and rivers not only provide food
and water but are also means of communication. Rivers can
98163_APL 45_15_Kolen.indd 215 8/05/15 08:45
part of this case study relies on principles that are similar to
the Uppland case study. Water bodies in this area form an
extensive communication network, which was and still is
being used for food procurement, transport, and trade.
Occupation and land use focused on the western shore of
Lake Peipsi, which is on the one hand a lake, but on the
other hand has many similarities with a sea (i.e. the Baltic
Sea) (Karro 2012). Moreover, the lake is part of a larger lake
system including Peipsi and Pskov that has been documented
as part of a trade route system that covered an enormous area
stretching from Scandinavia to Byzantine (Mägi in press).
The method deployed for analysing landscape changes
throughout the Iron Age (500 BC – 1200 AD) correlates
geological information about the lakeshore with archaeologi-
cal data, historical maps and cultural historical information
(like historical narratives about particular places). The
dynamics of shorelines is reected in geographical shifts of
human settlement as well. Water levels in the lakes went up
gradually, but the land still rises faster in the southern part of
the study area. This implies that the present-day situation in
area this will affect both shorelines, thus making different
pour points relevant at different periods. In areas with level
terrain this might also cause the inland boundaries of the
watersheds to shift as the land surface is tilting. It is
therefore necessary to use a DEM that has been modied to
the relevant time period using a method like the one
described above. In the Uppland study area, a high resolution
DEM has been produced by the Swedish Cadastral Agency
(Lanmäteriet) using LiDAR technology.
Using the shoreline and watershed modelling together with
the digital database from the National Heritage Board
(Fornsök) the aim is to model the development of social
relations through regions and regionalism in the area, in
order to better understand the long term land use of the area
starting with the early Neolithic (ca 4000 BC) and ending
with the Viking Age (c. 1050 AD).
6.3. Kodavere and Vooremaa, Estonia (gure 3)
Kodavere and Vooremaa are neighbouring regions in the
south-eastern Estonian lakeside landscape. The theoretical
Figure 2 Landscape with royal burial mounds on a natural ridge along the lake, at the Medieval political, religious and economic centre of Old
Uppsala (Gamla Uppsala; photo Kim von Hackwitz)
98163_APL 45_15_Kolen.indd 216 8/05/15 08:45
maps and archaeological information are a good source for
detecting suitable arable land cultivated before the
mechanization of agriculture. The model may also help to
understand why some sites developed into centres of power
and –subsequently- into important places of local identity
‘surrounded’ by local narratives and beliefs. Historical
land-use data combined with contemporary soil maps and
adequate digital elevation models in the historical GIS also
provides a good platform for elementary predictive
modelling, which could be effectively used in preventive
heritage management.
7 epIlogue
In the next two years (2015-2016) the HERCULES WP2
team will study landscape changes in the three regions from
a comparative perspective, exploring the conceptual
framework and the techniques for linking and analysing
archaeological, historical and environmental data outlined
above. This exercise will focus on the question to what
extent landscape changes either converge or diverge, given
different or similar social and environmental conditions. It is
important, furthermore, to learn from the effects of
‘experiments’ that people have conducted in and with
landscapes in the past. More often than not these
experimental interventions caused unexpected and delayed
the northern part may roughly be indicative of the situation
in the past in the southern part. Lake Peipsi has seven visible
shoreline terraces dating from different periods since its
formation after the last Ice Age, of which the 38 m. shoreline
most strongly correlates with historical and present-day
settlement patterns. Folklore and historical photographs give
some hints about the traditional land cover (19th and the
beginning of the 20th century) that was drastically changed
during the Soviet period.
For the Vooremaa region a historical GIS is applied.
Detailed historical maps are geo-referenced and vectorized
into layers based on land use information (e.g. arable,
pasture, meadow, forest, swamp, fallow) and settlement
features (e.g. farms, manors, mills, taverns, churches, roads,
administrative boundaries). Additionally, a database with
relevant attribute data (information about land use) is used.
The layers based on historical maps can then be integrated
with known archaeological sites, which will result in a series
of detailed maps representing pre-industrial landscape
This provides us with a coherent sequence of about 350
years of land-use, indicating land use functions in terms of
arable land, pastures and meadows, forests, swamps and
bogs. Much of the wetland in Estonia was not drained until
the beginning of the 20th century, which is why historical
Figure 3 Viking Age hillfort at Peatskivi, Vooremaa region, Estonia (photo: Martti Veldi)
98163_APL 45_15_Kolen.indd 217 8/05/15 08:45
use in urban and rural areas. Research paper retrieved fromnaltext.pdf
Hornborg, A. and C.L. Crumley 2007. The World System and
the Earth System: global socioenvironmental change and
sustainability since the Neolithic. Walnut Creek CA.
Ingold, T. 2000. Perception of the Environment: Essays on
Livelihood, Dwelling and Skill. Oxford.
Janssen, J., E. Luiten, H. Renes and J. Rouwendal 2014.
Heritage planning and spatial development in the
Netherlands: changing policies and perspectives.
International Journal of Heritage Studies, 20(1), 1–21.
Jordan, P. 2003. Material culture and sacred landscape: the
anthropology of the Siberian Khanty. Walnut Creek CA.
Kamermans, H., M. van Leusen and P. Verhagen (eds.) 2009.
Archaeological Prediction and Risk Assessment: Alternatives
to Current Practice. ASLU 17. Leiden.
Karro, K. 2012. Lakeside and maritime landscapes: Cases of
Kodavere settlement district (Eastern Estonia) and Pöide
settlement district (Saaremaa). Lietuvos Archeologija 38,
Kohler, T. A., C.D. Johnson, M. Varien, S. Ortman, R.
Reynolds, Z. Kobti, J. Cowan, K. Kolm, S. Smith and L. Yap
2007. In T. Kohler and S. van der Leeuw (eds) Settlement
ecodynamics in the prehispanic central Mesa Verde region.
The Model-Based Archaeology of Socionatural Systems Santa
Fe, 61–104.
Kolen, J.C.A. 2005. De biograe van het landschap. Drie
essays over landschap, geschiedenis en erfgoed. Amsterdam
(VU University Amsterdam, PhD Dissertation).
Kolen, J., R. Hermans and H. Renes (eds) 2015. Landscape
biographies: geographical, historical and archaeological
perspectives on the production and transmission of
landscapes. Amsterdam.
Koomen, E., M. Hilferink and J. Borsboom-van Beurden
2011. Introducing land use scanner, in Koomen, E.,
Borsboom-van Beurden, J. (eds.), Land-use modelling in
planning practice. GeoJournal Library Vol. 101, 3-21.
Küchler, S. 2002. Malanggan: art, memory and sacrice.
Löwenborg, D. 2007. Watersheds as a Method for
Reconstructing Regions and Territories in GIS. In Digital
Discovery: Exploring new frontiers in human heritage.
Proceedings of the 34th International Conference on
Computer Applications and Quantitative Methods in
Archaeology (CAA), pp. 143-149. Fargo/Budapest.
effects for later generations, who had to ‘invent’ land use
systems anew in order to cope with these consequences. An
understanding of such iterative landscape histories can be of
value for stakeholders, landowners and planners who face the
challenge of environmental decision making today.
Arnoldussen, S. 2008. A Living Landscape: Bronze Age
settlements in the Dutch river area (c. 2000-800 BC). Leiden.
Balée, W. L. 1998. Advances in historical ecology. New York.
Balée, W. L. and C.L. Erickson 2006. Time and complexity
in historical ecology: studies in the neotropical lowlands.
New York.
Bentley, R. A. and H.D. Maschner 2003. Complex systems
and archaeology: Empirical and theoretical applications.
Salt Lake City.
Crumley, C. L. (ed.) 1994. Historical ecology: cultural
knowledge and changing landscapes. Santa Fe.
Crumley, C. L. 2012. A heterarchy of knowledges: tools for
the study of landscape histories and futures. In: T. Plieninger
and C. Bieling (eds), Resilience and the Cultural Landscape:
Understanding and Managing Change in Human-Shaped
Environments. Cambridge, 303-314.
Crumley, C. 2014. Deliverable 2.1 Innovative interdisciplin-
ary protocol for understanding long-term landscape
dynamics, based on the perspectives of historical ecology,
landscape biography, and complex systems theory.
HERCULES, FP7, GA no. 603447.
Crumley, C. L. 2015. New Paths into the Anthropocene.
In: C. Isendahl and D. Stump (eds) Oxford Handbook of
Historical Ecology and Applied Archaeology,.). Oxford.
De Kleijn, M., N. van Manen, J. Kolen and H. Scholten
2014. Towards a User-centric SDI Framework for Historical
and Heritage European Landscape Research. International
Journal of Spatial Data Infrastructures Research 9, 1–35.
Fairclough, G. and P. Grau Moller, P. (eds) 2008. Landscape
as Heritage–The Management and Protection of Landscape
in Europe, a summary by the Action COST A27 LAND-
MARKS. Geographica Bernensia: Berne, 79.
Harten, J. D. H. 2000. Rivierkleilandschap. In: S. Barends
(ed.), Het Nederlandse landschap. Een historisch-geogra-
sche benadering . Utrecht, 92–103.
Hilferink, M., and P. Rietveld 1998. Land Use Scanner. An
integrated GIS based model for long-term projections of land
98163_APL 45_15_Kolen.indd 218 8/05/15 08:45
Risberg, J., G. Alm, N. Björck and M. Guinard 2007.
Synkrona paleokustlinjer 7000-4000 kal. BP i mellersta och
norra Uppland. Uppsala.
Roymans, N., F. Gerritsen, C. van der Heijden, K. Bosma
and J.Kolen 2009. Landscape biography as research strategy:
The case of the South Netherlands project. Landscape
Research 34, 337–359.
Sund, C. 2010. Palegraska förändringar i östra Svealand
de senaste 7000 åren. NKA 16.
Van de Ven, G.P. (ed.) 1993. Man-made lowlands. History of
water management and land reclamation in The Netherlands.
Utrecht: Matrijs.
Verhagen, P. and T.G. Whitley 2012. Integrating archaeologi-
cal theory and predictive modeling: a live report from the
scene. Journal of Archaeological Method and Theory 19(1),
Von Hackwitz, K. 2012. The Creation of Regions: An
Alternative Approach to Swedish Middle Neolithic
Boundaries and Cultures. Norwegian Archaeological Review
45(1), 52–75.
Wijkander, K. 1983. Kungshögar och sockenbildning: studier
i Södermanlands administrativa indelning under vikingatid
och tidig medeltid. Nyköping.
Wilkinson, T. J., J.H. Christiansen, J. Ur, M. Widell and
M. Altaweel 2007. Urbanism within a Dynamic Environ-
ment: Modelling Bronze Age communities in Upper
Mesopotamia. American Anthropologist 109(1), 52–68.
Prof. dr. J.C.A. Kolen, Faculty of Archeology, World
Heritage, Universiteit Leiden, Van Steenis gebouw,
Einsteinweg 2, 2333 CC Leiden, the Netherlands,
Prof. dr. C.L. Crumley, Department of Archaeology and
Ancient History, Uppsala University, Engelska Parken,
Thunbergsvägen 3H, Uppsala, Sweden,
McGlade, J. and S.E. van der Leeuw 2013. Time, process
and structured transformation in archaeology. Oxford.
Mägi, M. in press. Bound for the eastern Baltic: trade and
centres 800-1200. In: J. Barrett and S. Gibbon (eds),
Maritime societies of the Viking and Medieval world.
Cambridge: MacDonald Monograph Series.
Meinig, D. W. 1979. The beholding eye: Ten versions of the
same scene. The Interpretation of Ordinary Landscapes:
Geographical Essays, Oxford, 33–48.
Meyer, W. J. and C.L. Crumley 2011. Historical ecology:
using what works to cross the divide. In Tom Moore and
Lois Armada (eds), Atlantic Europe in the First Millennium
BC: Crossing the Divide. Oxford, 109-134.
Olwig, K. R. 1996. Recovering the substantive nature of
landscape. Annals of the Association of American
Geographers 86(4), 630–653.
Olwig, K. 2002. Landscape, nature, and the body politic:
from Britain’s renaissance to America’s new world. Madison.
Renes, J. 2005. Water management and cultural landscapes in
The Netherlands. In: H. S. Danner,
J. Renes, B. Toussaint, G. P. Van de Ven and F. D. Zeiler
(eds), Polder pioneers. The inuence of Dutch engineers on
water management in Europe. Utrecht: Netherlands
Geographical Studies 338, 13-32.
98163_APL 45_15_Kolen.indd 219 8/05/15 08:45
98163_APL 45_15_Kolen.indd 220 8/05/15 08:45
... The selection of the study area was conditioned by two considerations: (1) testing the suggested methodology on heterogeneous land cover (from urban structures to forests, agricultural land and wetlands as land cover types, used as a subject of different disciplines) and (2) examining the colour harmony rates within protected and nonprotected areas to reveal the potential of further nature conservation. The Vooremaa protected area is one of the study landscapes of the HERCULES project (Kolen et al. 2015). Therefore, this study contributes to its objectives, providing a targeted case study on the colour harmony as a landscape value of typical Estonian land cover. ...
Full-text available
The concept of colour harmony, being rarely used in geography, landscape and environmental studies, has been significantly developed in psychology, art and computer science within the different approaches: colour wheel geometry and, more recently, numerical models applied to colour combinations. Using the main numerical principles of colour harmony, borrowed from the psychological literature, this study aims to investigate the ways of mapping the extent of the colour harmony of land cover, based on satellite Earth observations and explain the spatial distribution of colour harmony scores. The naturalness of environment, as well as heat and moisture balance, are confirmed to be the main drivers of the colour harmony of land cover. Crowdsourced photographs, collected from Mapillary service, were used to link satellite and ground-based estimations of the colour harmony of land cover as “proof of concept”. They have a limited applicability for ground-based assessment of scenic colour harmony. Therefore, remote sensing data provide a significant support for nature conservation and sustainable management, being used for mapping of the colour harmony of land cover as an indicator of the visual quality of the perceived environment. The read-only version of this paper is publicly available via the following link:
The intricate character of the landscape is one of the main difficulties when reaching an agreement on its values. This information is, however, essential to manage the landscape, a process which relies on methodologies that recognise those values and/or identifies. In this paper, the analytical methodology for an integrated plan of the territory is reviewed, and a method is presented to design cultural routes as a strategy for connecting the archaeological sites to their landscapes by restoring the dynamics of landscape formation in their immediate environments. Using the area surrounding the archaeological site of the Roman city of Italica in Andalusia (Spain) as a case study, actions and processes are identified that can enable projects based on ‘cultural routes’ to restore the dynamics of landscape formation, highlighting those processes that allow us to recognise the landscape values and to extract some of the landscape’s characteristic features.
Full-text available
Current advances in spatial simulation bring unprecedented possibilities for spatio-temporal modeling. In this paper, we focus on modeling the impact of settlement on land use in the Roman period in the Dutch river area, on the northern frontier of the Roman Empire. During this period, the area witnessed a strong population increase that put more demands on the available land to produce food, not only for the local population, but also for the soldiers stationed on the frontier and the citizens of the newly founded towns. We compare an agent-based model (ABM) of agricultural production in the region (ROMFARMS), and a model using the Past Land Use Scanner (PLUS. Both were used to estimate the effects of increased agricultural demand through simulations of food production, taking into account the available workforce and the productivity and availability of suitable land. However, how should we evaluate the model outcomes? What are the advantages and limitations of each? We discuss issues of scale, temporal resolution and model inputs, together with questions of technical implementation and validation. In this way, we aim to point the way to future researchers to implement these approaches effectively in other contexts.
Full-text available
Across Europe, landscape is recognised as a frame through which societal values are defined and embedded. The European Landscape convention and wider research has drawn attention to the need for integrating a diverse range of stakeholders to ensure landscape sustainability. Archaeology is increasingly recognised as having an important place in integrated landscape management but often remains relatively peripheral. This paper examines the place of archaeology in specific European regions and the potential ways of integrating archaeological heritage in landscape management. Emerging from a project funded by the Joint Programme Initiative on Cultural Heritage (Resituating Europe’s FIrst Towns (REFIT): A case study in enhancing knowledge transfer and developing sustainable management of cultural landscapes), we explore the place of a set of common European heritage assets, Iron Age oppida, in the management of the landscape they are a part of and how they might be used better to engage and connect stakeholders. Using four case studies, we review the present integration of archaeology within landscape management and how this operates at a local level. From this we explore what challenges these case-studies present and outline ways in which the REFIT project has sought to develop strategies to respond to these in order to enhance and promote co-productive management of these landscapes.
Full-text available
Learning about Burgundy In the mid-1970s, my colleague Bill Marquardt and I, with our students, began an archaeological project to excavate an Iron Age (Celtic) hill fort in Burgundy (France) called Mont Dardon. I had been there before, at least vicariously, in my years as a Latin student; the astonishing view of the landscape from Dardon’s summit is the setting for dramatic passages in Caesar’s Gallic Wars. The site was possibly an early capital of a polity called the Aedui, who were first Caesar’s allies and then his foes. Dardon was inhabited a thousand years before, during, and for fourteen hundred years after the Roman conquest. We realised that the site and its surrounding landscape provide the perfect setting for a tale only archaeology can tell – not in words carefully chosen by an ambitious general but in the ways life changed for those who dwelt in that landscape (Crumley & Marquardt, 1987). Archaeology, although only one of the ways our project learns about the past, lies at the heart of our work in Burgundy. Most early team members were archaeologists trained within the discipline of anthropology, the broad-in-space and deep-in-time study of human behaviour. We wished to pose anthropological questions: How did Aeduan politics and economy change after the Conquest? What spiritual and social adjustments to Roman rule were made? How did the region fare when the Western Empire crumbled and groups from eastern and northern Europe arrived?
Landscape, Nature, and the Body Politic explores the origins and lasting influences of two contesting but intertwined discourses that persist today when we use the words landscape, country, scenery, nature, national. In the first sense, the land is a physical and bounded body of terrain upon which the nation state is constructed (e.g., the purple mountain majesties above the fruited plain, from sea to shining sea). In the second, the country is constituted through its people and established through time and precedence (e.g., land where our fathers died, land of the Pilgrims' pride). Kenneth Olwig's extended exploration of these discourses is a masterful work of scholarship both broad and deep, which opens up new avenues of thinking in the areas of geography, literature, theater, history, political science, law, and environmental studies. Olwig tracks these ideas though Anglo-American history, starting with seventeenth-century conflicts between the Stuart kings and the English Parliament, and the Stuart dream of uniting Scotland with England and Wales into one nation on the island of Britain. He uses a royal production of a Ben Jonson masque, with stage sets by architect Inigo Jones, as a touchstone for exploring how the notion of "landscape" expands from artful stage scenery to a geopolitical ideal. Olwig pursues these contested concepts of the body politic from Europe to America and to global politics, illuminating a host of topics, from national parks and environmental planning to theories of polity and virulent nationalistic movements.