RESEARCH ARTICLE
264 years of change and persistence in an agrarian
landscape: a case study from the Swiss lowlands
Matthias Bu
¨
rgi
•
Daniel Salzmann
•
Urs Gimmi
Received: 12 May 2014 / Accepted: 16 March 2015 / Published online: 29 March 2015
Springer Science+Business Media Dordrecht 2015
Abstract
Context Century-long interaction between societies
and their environment has shaped the cultural land-
scapes across Europe. Specific farming systems have
persisted for long period s, and their abandonment has
led to far-reaching and rapid changes in patterns of
land use and land cover.
Objective Taking the example of the municipality of
Limpach, located in the Swiss lowlands, we studied
the dynamics of cultural landscape and related driving
forces in a long-term perspective.
Methods Using various data types, we reconstruct
264 year s of change and persistence and evaluated the
processes involved. The sources permit not only the
examination of land use and land cover change, but
also linear elements (paths, waterways) and the study
of changes in parcel structure.
Results For centuries, efforts to increase the inten-
sity and productivity of agriculture, such as the so-
called ‘‘melioration projects’’, which included exten-
sive drainage of wetlands and the establishment of an
efficient network of roads and paths, greatly influ-
enced landscape changes. In the last few decades,
driving forces not directly aiming at an intensification
of agricultural production, but rather promoting eco-
logically valuable landscape elements, have left traces
in the open landscape of Limpach.
Conclusions Changes are an inherent characteristic
of cultural landscapes. However, the rates of changes
are highly v ariable over time. While comparatively
stable periods might be used as a reference/baseline
and the region itself may be referred to as a traditional
cultural landscape, such long periods should not be
misinterpreted as being stable from a long term
historical perspective.
Keywords Landscape history Rates of change
Driving forces Historical maps Agricultural history
Switzerland
Introduction
Originally an activity of historical geographers, the
study of landscape changes has become an important
topic in landscape ecology and mor e recently an
important approach used in landscape ecology (Rhem-
tulla and Mladenoff 2007) and land change science
(Turner et al. 2007). Four major groups of studies can
be identified: (a) descriptive approaches (Bender et al.
2005; Hamre et al. 2007), (b) studies focussing on
specific trends, such as urbanization (Antrop 2004;
Seto and Fragki as 2005), loss of wetlands (Gimmi
et al. 2011) or agricultural abandonment (Gellrich and
M. Bu
¨
rgi (&) D. Salzmann U. Gimmi
Research Unit Landscape Dynamics, WSL Swiss Federal
Institute for Forest, Snow and Landscape Research,
Zu
¨
rcherstrasse 111, 8903 Birmensdorf, Switzerland
e-mail: matthias.buergi@wsl.ch
123
Landscape Ecol (2015) 30:1321–1333
DOI 10.1007/s10980-015-0189-1
Zimmermann 200 7 ), (c) studies examining the conse-
quences of landscape changes on, for example,
biodiversity (O lsson et al. 2000; Coppedge et al.
2001; Stouffer et al. 2006) or ecosystem services, such
as regional climate (Schneider and Eugster 2007), and
(d) analytical approaches attempting to understand the
forces driving landscape changes (Bu
¨
rgi et al. 2004;
Long et al. 2007; Bieling et al. 201 3 ). Landscapes are
the sphere in which human acts and natural factors and
processes are intertwined. Landscapes can therefore
change due to changes in the natural system and/or in
the social system (Bic
ˇ
ı
´
k et al. 2001), and these changes
in turn have effects on both systems alike. Thus, many
of the frameworks sugges ted for the study of such
social–ecological systems (Binder et al. 2013) can be
applied to the study of landscape changes.
Current research focussing on the driving forces of
landscape change is not often motivated by quest ions
such as ‘‘Can landscape change be directed into a
societally more desirable direction?’’ or ‘‘What
factors are accelerating or slowing down the rate of
landscape change?’’ (Schneeberger et al. 2007) and
we believe that these aspects have thus far not yet
received the attention they deserve. By focusing
primarily on changes, the analysis of persistence and
impeding and stabilizing factors in landscapes are
largely neglected, despite the fact they may be highly
relevant for planning and management (Bu
¨
rgi et al.
2004).
Long-term studies of landscape change can only be
conducted based on suitable historical sources, most
often a series of historical maps which are necessarily
consistent and comparable in quality and map content
(Gimmi et al. 2011). The comparable map content
determines, for example, whether the focus is on land
cover changes (open land–forests–settlements) or
whether land use changes (e.g., pasture vs. meadow)
can be analysed. Limited source availability is
relevant for a descriptive approach, and further
accentuated in more analytica l appro aches since this
limits the quantity of available information on poten-
tial forces driving the change observed.
Historical sources are not randomly available, but
certain historical circumstances can lead to favorable
source availability. Producing maps, which necessi-
tates surveying the land, requires significant resources.
Therefore historical maps were often initiated by
government activities that required spatial informa-
tion, such as tax collection that required cadastral
maps, i.e. a detailed cartographic record of land
ownership showing the location and structure of
parcels (Cousins 2001; Hamre et al. 2007), or major
infrastructural projects with an impact on the land.
In this paper we present a study on landscape
changes in the agrarian municipality of Limpach,
located in the Swiss lowlands, taking advantage of two
such circumstances: cadastral maps avai lable from the
eighteenth century used for tax collection purposes
and planning documents on an encompassing melio-
ration project from the early twentieth century, that
included river corrections, drainage and reparcelling,
i.e. the re-arrangement of parcels into larger entities to
enhance efficient farming. Both map series permit the
analysis of landscape elements, often not included in
studies of landscape change, such as different uses of
agricultural land, land parcelling and linear elements
(paths, waterways).
The municipality of Limpach is located in the
Limpach valley, which as a whole has been the subject
of a recent study on the driving forces in landscape
change (Bu
¨
rgi et al. 2010). In this study, landscape
change is assessed with a focus on peat mining, which
was prevalent in some parts of the Limpach valley. In
the present paper, we zoom in on the municipality of
Limpach, characterized by the favourable source
situation outlined above. For this location, we
specifically address the following research questions:
How did the landscape change over the past
264 year s? What aspects of the landscape are highly
dynamic? What aspects show high persistence? What
are the specific processes that transform agricultural
landscapes?
With this study we therefore want to contribute to
the analytical approaches attempting to understand the
forces driving landscape changes (i.e. the fourth group
of landscape change studies as listed at the beginning
of the introduction), but also explicitly studying
driving forces of landscape persistence.
Study area
The municipality of Limpach is located in the
Limpach Valley, in the central Lowlands of Switzer-
land (Fig. 1). The municipality covers an area of
445 ha and is located at about 480 masl. The yearly
average precipitation is 1000–1100 mm and the
average temperature is 8.5 C (Luterbacher 1992).
1322 Landscape Ecol (2015) 30:1321–1333
123
The study area consists of organic, partly peaty soils
on the valley floor and mineral soils on the slopes and
hills. In the 1940s, the average depth of the peat layer
varied from 100 to 300 cm (Sta
¨
hli 1941). Today, due
to peat mining and drainage with subsequent miner-
alization, this depth is generally reduced.
Historically, due to its position in the lowlands,
Limpach was part of the so-called ‘‘Grainland’’
(‘‘Kornland’’—Mathieu 1992), which encompassed
most rural areas of the Swiss lowlands. Consequently,
the main developments observed in the municipality
of Limpach can be taken as an example of the trends
occurring in large parts of the Swiss lowlands.
Characteristic for the Grainland, grain production
was a core land use and the arable land was cultivated
according to the three-field system. In this land use
system, which was common practice throughout large
parts of Europe since about the thirteenth century
Fig. 1 The municipality of
Limpach is located in the
lowlands of Switzerland
(data land cover: pixmaps
2013 swisstopo (5704 000
000), DEM: dhm25
2013
swisstopo (5704 000 000))
Landscape Ecol (2015) 30:1321–1333 1323
123
(Pfister 1983; Renes 2014), about one-third of private
land was sown in turn with winter grain, a second third
with summer grain (oats, barley, beans), the last third
remained fallow.
From the mid-eighteenth century onwards, an
agricultural reform movement, the so-called agricul-
tural revolution, spread from England to large parts of
Europe, and resulted in Switzerland in the abandon-
ment of this three-field system (Pfister 1983; Lowood
1987). The agrarian revolution also included the
partitioning of common lands, the introduction of
potatoes and legumes (e.g., clover, alfalfa), improve-
ments in nutrient management through the introduc-
tion of summer stable feeding and dung collection,
and—for Switzerland—the start of cheesemaking in
the lowlands.
Today, Limpach is a largely rural municipality, and
employment for some of the 341 inhabitants (http://
www.fin.be.ch/) is found in nearby larger towns and
cities, such as Berne, the capital of Switzerland, which
is only about 20 km away.
Materials and methods
Landscape change in Limpach was reco nstructed
based on historical map s, supplemented by historical
documents and secondary literature, most of which
were found at local, state, and federal libraries and
archives. Despite the fact that not all maps show
exactly the same land cover categories, we were able
to build a remar kably thorough and homogenous
documentation on landscape changes based on a series
of historical maps with sufficient thematic overlap.
The first map included in the study shows the situation
before the agrarian revolution, as it dates from 1749.
The original map consists of a series of single maps
which were drawn by Jean David and Abraham
Vissaula, who were famous for outstanding surveying
in their day (Grosjean 1996). These 1749 maps, drawn
at the approximate scale of 1:1450, show parcels,
buildings, paths, waterways, fences and hedgerows.
Information on ownership and land cover type (gar-
dens, arab le land, meadows, pastureland and forest
areas) is listed in a separate volume (‘‘Urbar’’—details
in Grosjean 1960). The aim of this survey was to have
an overview of land ownership as a basis for tax
collection. The second point in time considered is
based on the first modern cadas ter map of 1882, which
was drawn on 15 sheets at a scale 1:1000. This maps
shows appro ximately the same signatures as the 1749
map. Additional information on agricultural land use
was taken from the contemporary cadastral register,
however for some parcels, no information was avail-
able (white patches in Fig. 2). An encompassing
drainage and melioration project conducted in the
1930s and 1940s was the reason for two more map
series from 1939 (situation before the start of the
project) and 1943 (situation after project com pletion).
Both maps also include information on arable land,
and artificial and natural meadows. The 1943 land use
map does not cover the whole municipality of
Limpach (white patch in the north-eastern corner in
Fig. 2—1943). This gap only exists with reference to
land cover information, whe reas information on parcel
boundaries, paths and waterways is available. The last
step in time included in our study is the situation in
around 2013, based on modern digital maps at a scale
1: 25,000 [Vektor25—swisstopo (DV033594)] com-
bined with spatial and statistica l information on the
area under crop rotation obtained from the cantonal
office for geoinformation (Amt fu
¨
r Geoinformation,
Fruchtfolgefla
¨
chen des Kantons Bern). Since the crop
rotation system leads to period changes between
grassland and arable land, it does not make sense to
attempt locating these land use types for a specific
year. We introduced, for the last time step, a new land
cover type called ‘‘crop rotation’’ and depicted only
ecological compensation areas as permanent grass-
lands, as they are not part of the crop rotation system.
However, based on the statistical data, we have
information on the extent of grassland and arable land
today, which makes it possible to calculate the total
area and the proportion of the two land use types
(Table 1).
All historical maps were scanned, the single map
sheets georeferenced, mosaicked, and, together with
the modern map, integrated into a geodatabase using
ArcGIS 10.1 (ESRI). For the extraction of land-cover
information, the modern map was taken as a reference
layer on which the olde r maps were vectorized usin g a
back-dating approach. This approach allows to re-
draw boundaries only for real changes, avoiding sliver
polygons resulting from changes due to inaccuracies
of older maps. From the GIS database, we calculated
net changes in all analyzed signatures. For changes in
settlements, forests, paths and waterways, we deter-
mined increases and decreases in area or length
1324 Landscape Ecol (2015) 30:1321–1333
123
separately for all periods. For all other land cover
categories, this comparison was only possible from
1749 to 1943, as the modern crop rotation system does
not permit such analysis for the last time step. We
separately give the ratio of grassland to arable land,
which reflects a core characteristic of the agricultural
system. To account for large differences in length of
periods between the maps, we also calculated annual
relative changes.
Based on this map analysis, we determined the
main trends of land cover change for 1749, 1882,
1939, 1943 and 2013. Th e changes and related driving
forces were interpreted using information from the
written documents and secondary literature. The
oldest written document evaluated is a description of
Limpach conducted by the local minister Mu
¨
ller
(1804), containing information on land use at the
beginning of the nineteenth century, a relatively short
period before the beginning of the agricultural
revolution in Limpach. Two reports (Sta
¨
hli 1941,
1953) contain detailed information on the drainage
and melioration project. Information on development
that occurred in the twentieth century was taken from
Straub (2008; see also Bu
¨
rgi et al. 2010).
Results
Land cover and land use
In the first period from 1749 to 1882, a slight
expansion in settlements from 2 to 3 % of the whole
municipality, and a decline in forest area from 28 to
22 % can be discerned (Fig. 2; Table 1). Agricultural
land (i.e., the remaining 70–75 %) was about equally
divided between grassland and arable land at both
points in time (i.e. ratios of 0.95 and 0.98, respective-
ly). During the next 57 years, the major land cover
classes remained stable, except for a significant
expansion in grassland at the expens e of arable land,
Fig. 2 Land cover in Limpach from 1749 to 2013. A large melioration and drainage project conducted between 1939 and 1943 enabled
the conversion of formerly wet grassland in the northern part to arable land (sources discussed in text)
Landscape Ecol (2015) 30:1321–1333 1325
123