Full text of DOI:10.1080/00291950119811

This article was downloaded by: [Florida Atlantic University]
On: 18 November 2014, At: 09:51
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Norsk Geografisk Tidsskrift - Norwegian
Journal of Geography
Publication details, including instructions for authors and subscription
information:
http://www.tandfonline.com/loi/sgeo20
Heterogeneity as a measure of spatial
pattern for monitoring agricultural
landscapes
W. J. Fjellstad , W. E. Dramstad , G.-H. Strand & G. L. A. Fry
Published online: 05 Nov 2010.
To cite this article: W. J. Fjellstad , W. E. Dramstad , G.-H. Strand & G. L. A. Fry (2001) Heterogeneity as a
measure of spatial pattern for monitoring agricultural landscapes, Norsk Geografisk Tidsskrift - Norwegian
Journal of Geography, 55:2, 71-76, DOI: 10.1080/00291950119811
To link to this article: http://dx.doi.org/10.1080/00291950119811
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our
licensors make no representations or warranties whatsoever as to the accuracy, completeness, or
suitability for any purpose of the Content. Any opinions and views expressed in this publication
are the opinions and views of the authors, and are not the views of or endorsed by Taylor &
Francis. The accuracy of the Content should not be relied upon and should be independently
verified with primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities
whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial
or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or
distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use
can be found at http://www.tandfonline.com/page/terms-and-conditions

Norsk geogr. Tidsskr. Vol. 55, 71–76. Oslo. ISSN 0029-1951
Heterogeneity as a measure of spatial pattern for monitoring agricultural
landscapes
W. J. FJELLSTAD, W. E. DRAMSTAD, G.-H. STRAND & G. L. A. FRY
Fjellstad, W. J., Dramstad, W. E., Strand, G.-H. & Fry, G. L. A. 2001. Heterogeneity as a measure of spatial pattern for monitoring
agricultural landscapes. Norsk GeograŽ sk Tidsskrift–Norwegian Journal of Geography Vol. 55, 71–76. Oslo. ISSN 0029-1951.
The Norwegian monitoring programme for agricultural landscapes is a sample-based programme that aims to document the current
state and patterns of change in the agricultural landscape for the entire country. For this broad geographical coverage an indicator
approach is the most cost-effective methodology . Landscape metrics are used as indicators in the monitoring programme based on
the assumed importance of landscape composition and spatial pattern for a number of countryside interests. However, although
indicators have been used successfully over a long period of time in various disciplines, their use in landscape monitoring is rela-
tively new and there is little empirical data with which to assess their validity. This paper provides information on the performance
of landscape diversity and landscape heterogeneity indices as estimated by speciŽ c biodiversity measures. Preliminary results
show increasing numbers of breeding bird species and numbers of vascular plant species with increasing landscape diversity
and heterogeneity, two indices that capture information on landscape content and structure. Overall, the heterogeneity index pro-
vided a better surrogate for species richness than the diversity index.
Keywords: Diversity, heterogeneity, indicators, landscape monitoring
Ê
W. J. Fjellstad, W. E. Dramstad, G.-H. Strand, G. L. A. Fry, Norwegian Institute of Land Inventory, P.O. Box 115, N-1431 As,
Norway. E-mail: wef@nijos.no, wed@nijos.no, ghs@nijos.no, gary.fry@ninaosl.ninaniku.n o
Introduction
information to indicate trends of change in various aspects of
Norwegian farmland (Dramstad et al. in press). Four themes
The concept and deŽ nition of a ‘landscape ecology’ was
introduced by the German geographer Carl Troll while
working on the interpretation of aerial photographs (Troll
of interest have been deŽ ned for the monitoring programme:
spatial structure, biodiversity, cultural heritage and accessi-
bility. The monitoring programme involves an explicit need
to compare, simplify and communicate Ž ndings on complex
issues, which is exactly what indicators are intended to be
used for (Smeets & Weterings 1999). The selected indicators
are related to landscape content and spatial pattern. Whilst
metrics of spatial structure can be extracted directly from
aerial photographs, indicators for the other themes are
indirect measures, building particularly on a theoretical basis
from the Ž eld of landscape ecology (Forman & Godron 1986,
Forman 1995). Yet the links between these themes and
landscape spatial structure are far from proven.
In the case of indicators of cultural heritage, accessibility
and landscape experience, the relationship with landscape
structure is highly uncertain and there is a lack of empirical
data with which to test hypotheses regarding possible links.
In the context of biodiversity indicators, landscape content
obviously in uences the distribution of species in as much as
their occurrence is restricted to areas of suitable habitat. In
addition, aspects of landscape structure have been shown to
in uence the occurrence of several species (Merriam 1988,
Opdam 1990, Rolstad 1991, Forman 1995, Hanski 1997).
However, in order to use landscape information as a
surrogate for biodiversity information, there are a number
of important questions to be answered. For example, can
‘habitat’ be accurately mapped, what are the relative roles of
habitat quality, habitat quantity and spatial structure (Fahrig
1997), and what level of biodiversity is realistic (and useful)
to quantify (Tucker 1999)?
1968). In 1986, the growing discipline of landscape ecology
was deŽ ned by Forman & Godron as a branch of ecology
focusing on three characteristics of the landscape: structure,
function and change. There are now a number of deŽ nitions
of landscape ecology (see, e.g., Farina 1998, O’Neill et al.
1988, Forman 1995) and most include aspects of the
temporal dynamics of landscapes.
Landscape ecology, with its roots in geography, links to
remote sensing and focus on change, was a natural choice to
provide a theoretical framework for a national programme
for monitoring landscape change. With a rapidly developing
theoretical basis and its applied perspective, landscape
ecology also provides practical tools that may be of use in
monitoring. Landscape indicators are one such tool that has
become increasingly popular in recent years (McGarigal &
Marks 1995, Farina 1998, Brouwer & Crabtree 1999).
Indicators are simple measures that summarise and convey
complex information. They have been used successfully over
a long period in various disciplines, familiar examples being
the various indices from the Ž nancial world, such as the Dow
Jones index. Although the use of landscape indicators is still
an issue of some scientiŽ c and policy debate (Cale & Hobbs
1994, Dramstad et al. 1998), there are currently few other
options available for monitoring and reporting over large
areas (Bastian & Bernhardt 1993). Within the OECD, for
example, landscape indicators are currently being developed
as tools for international reporting on agri-environmental
state and change (OECD 1994, 1997, 1999).
For national monitoring, indicators are required that
summarise conditions for biodiversity in general, for other
themes such as cultural heritage, and for landscapes across
The Norwegian monitoring programme for agricultural
landscapes (the ‘3Q-programme’) aims to use landscape

72
W. J. Fjellstad et al.
NORSK GEOGRAFISK TIDSSKRIFT 55 (2001)
the entire country. However, empirical studies investigating
the relationship between landscape structure and, for
example, species occurrence tend to focus on a single
species or taxon. In addition, due to the costs of landscape
scale studies and the need to limit the number of variables
involved, these investigations seldom cover more than a few
Table 1. Classification system for land types, showing number of subclasses
at levels 2 and 3 in the hierarchy.
Number of sub-classes
Land type
Letter code Level 2
Level 3
Agricultural land
Bare ground
Semi-natural grassland
Other semi-natural open vegetation
Open wetland vegetation
Forest
A
B
C
F
M
S
3
2
2
3
2
22
8
8
13
6
3
‘
landscapes’ (Wiens & Milne 1989, Ims 1990). In spite of
these limitations, these studies have provided important
empirical data illustrating links with landscape structure for a
range of species groups (Rolstad 1991, Wegge et al. 1992,
Forman 1995).
3
Built-up areas
Water
Total
U
V
8
6
3
24
38
6
104
At a national level, economy and logistics preclude
investigations at very Ž ne resolution. Nevertheless, it is
essential that coarse-resolution indicators adequately re ect
the processes of importance at the species and habitat level.
Obtaining evidence to determine the nature of the relation-
ships between indicators and the ‘real world’ will require
years of study, yet landscape change is more rapid and wide-
reaching than ever before (Meeus 1993, McInerney 1994)
and there is an urgent need to monitor its effects. Research to
clarify the links between landscape and biodiversity must
therefore be conducted in parallel to coarse-resolution
monitoring. Detailed research studies and coarse-resolution
monitoring can thus be mutually beneŽ cial – research can
make use of the landscape data resulting from monitoring
programmes (Burt 1994) and monitoring programmes can be
augmented based on new research results. Ideally, intensive
monitoring (species inventories) should be located on sites
used for extensive monitoring (mapping of land types) so that
the relationships between data gathered at these two different
levels of resolution can be identiŽ ed.
In this paper, we focus on indices of landscape diversity
and landscape heterogeneity as indicators of biodiversity
since these two indicators best re ect landscape content and
spatial pattern. We use Ž eld data from south-eastern Norway
to test whether these indicators are of relevance to the species
richness of breeding birds and vascular plants. We compare
the performance of the two indicators and we discuss
methodological issues, both regarding the nature of Ž eld
data required to test the indicators and how methods of
calculation in uence indicator values.
major land classes (agricultural land, natural bare ground,
semi-natural grasslands, other semi-natural open vegetation,
natural wetland vegetation, forest, built-up areas and water/
snow/ice). Each major land class was further sub-divided,
with a total of 24 land types at level two in the classiŽ cation
system and 104 land types at level three (Table 1).
Shannon’s diversity index was used as a measure of the
diversity of land types (Magurran 1988). The index was
calculated at level two in the classiŽ cation system.
X
0
Shannon’s diversity index:
H ˆ ¡
piln pi
where p is the proportional occurrence of a particular land
i
type, n, in relation to the total occurrence of all land types, N,
and thus n /N.
i
The heterogeneity index (Hix) was developed for the 3Q
programme. It was designed to distinguish between ‘large-
scaled landscapes’ with few elements per unit area, and
‘
small-scaled landscapes’ with many elements per unit area.
Hix can therefore be seen as a measure of grain-size (sensu
Wiens 1989).
Hix is calculated from land type recorded on a lattice of
points spread out across the study area. The study area in the
2
3
Q programme is the 1 km sampling square and the points
are organised as a 10 £ 10 lattice. Each point in the lattice
has a number of neighbours, usually 8 but fewer along the
edges of the study area. The total number of neighbour
relationships in a 10 £ 10 lattice with no missing data is thus
684. The heterogeneity index Hix is calculated as
Methods
n
n
P
P
wijcij
The landscape data reported in this paper were collected
through the Norwegian 3Q-programme. The programme was
initiated in 1998 as a sample-based programme for monitor-
ing agricultural landscapes (Mathiesen et al. 1999, Dramstad
et al. in press). A total of 1475 squares of 1 km £ 1 km will
be mapped over a Ž ve-year period (1998–2002), after which
landscape change will be recorded through a second Ž ve-
year cycle. The data presented in this paper are from 1998
and 1999. For further details of the background and
methodology of the monitoring programme see Dramstad
et al. (in press).
iˆ1
jˆ1
i
6ˆ j
Hix ˆ 1 ¡
Pn Pn
wij
iˆ1i
6
where w is a binary weight set to 1 if i and j are neighbours,
ij
otherwise 0; and c is the binary similarity index set to 1 if i
and j are identical, otherwise 0.
ij
Hix will equal 1.0 in an extremely heterogeneous area
where no two neighbouring points have identical land type
and 0.0 in a completely homogeneous area where all points
have the same land type. Hix represents the probability of
Ž nding identical land types when two points separated by a
certain distance represented by the mesh of the lattice are
compared. It is thus a very simple measurement of
autocorrelation at this particular lag distance.
Indicators of spatial structure were derived from landscape
maps constructed from the interpretation of true colour aerial
photographs (scale 1:17,000). Land types were classiŽ ed
according to a hierarchical classiŽ cation system, with eight

NORSK GEOGRAFISK TIDSSKRIFT 55 (2001)
Heterogeneity in agricultural landscapes 73
with equal number of quadrats in each land type. The
location of quadrats was marked onto maps before going into
the Ž eld. The quadrats were distributed such that they fell in
different Ž elds (map polygons) within the square, in order to
capture variation within the sample squares. Quadrats were
placed randomly within each Ž eld. All vascular plant species
within each quadrat were recorded.
The biological data were expressed as species richness
(
number of species) per square. This is a simple and
uncontroversial measure of biodiversity that is easily under-
stood and interpreted. Relationships between the biological
data and the landscape indices were analysed using simple
linear regression and multiple regression. These tests can be
easily in uenced by outliers and they assume that the data are
normally distributed, so we also used non-parametric
correlation analysis (Kendall’s t). Since the conclusions
were the same for both types of test we have chosen to report
only the regression statistics.
Fig. 1. Number of breeding pairs of skylarks (Alauda arvensis) and robins
(Erithacus rubecula) in relation to the area of cereal Ž elds and meadows in 15
sample squares (1 £ 1 km) in South-eastern Norway.
Field data were collected for breeding birds and vascular
plants. Representatives from these taxa have been negatively
affected by modern changes in agricultural land use, the
groups are generally amenable to Ž eld study, and they are
currently recorded in monitoring programmes in a number of
countries (Haila 1985, Ekstam & Forshed 1992, Wascher
Results
Results from individual bird species show different relation-
ships with the various landscape indicators recorded from the
3Q-squares. To illustrate how landscape composition in u-
ences different species differently, Figure 1 shows the
number of breeding pairs of skylarks (Alauda arvensis),
which increases with increasing area of cereal Ž elds and
meadows, whilst the number of robins (Erithacus rubecula),
a woodland species, declines.
Analyses of total number of bird species per square
showed signiŽ cant positive relationships between number of
bird species and both landscape diversity and heterogeneity
values (Fig. 2 and Table 2). The relationship was clearest for
the heterogeneity values.
1996).
Birds were recorded as number of breeding pairs within a
0 m radius from 9 points, evenly spaced throughout the
km £ 1 km sample squares. Recording was carried out
5
1
from May to June, using the same basic methodology as is
used in the Norwegian Breeding Birds Survey run by the
Norwegian Ornithological Society. All species seen or heard
within a 10-minute period at each point were recorded.
Information on the vegetation cover at each point and
weather conditions were also recorded, and there was no
recording on days with strong wind or heavy precipitation.
Plant species lists and percent cover were recorded from
Total number of plant species per square was not
signiŽ cantly correlated with landscape diversity but was
correlated with heterogeneity values (Fig. 3 and Table 2).
Multiple regression analysis showed that the Shannon
diversity index did not contribute signiŽ cantly to the
2
0 quadrats (1 m £ 1 m) per 3Q-square, from the land type
categories ‘agricultural land’ and ‘semi-natural grasslands’,
Fig. 2. Relationship between number of bird species and (a) Shannon’s diversity index values and (b) heterogeneity index values per square in 15 sample squares
1 £ 1 km) in South-eastern Norway.
(

74
W. J. Fjellstad et al.
NORSK GEOGRAFISK TIDSSKRIFT 55 (2001)
Table 2. Simple linear regressions of numbers of breeding birds and numbers
of vascular plant species per square against heterogeneity index values (Hix)
and Shannon’s diversity index values per square. Only regressions involving
Hix are clearly significant at the 95% level. Regression involving Shannon is
not significant for plants and barely significant for birds.
measurement, limited to a Ž xed lag distance. An autocorrela-
tion measurement that spans several lags could be an
improvement. However, we believe that a statistic should
only be used if the logic of the measurement is fully
understood. Hix is thus preferred due to its simplicity and
mathematical transparency.
2
Dependent (y) Independent (x) Regression
R
p
The nature of Ž eld data on biodiversity will also be highly
in uenced by sampling effects. For birds, which were
sampled at 9 points evenly spaced throughout the square
and independent of land type, the number of species recorded
in the different squares would seem to be a good relative
measure that captures the relative species richness at the
required resolution. Measures of landscape diversity and
Plants
Plants
Birds
Birds
Hix
Shannon
Hix
y = 70.74x ‡ 43.22
y = 17.27x ‡ 50.23
y = 31.43x ‡ 2.03
y = 14.35x ¡ 7.35
0.52
0.09
0.63
0.27
<0.01
0.21
<0.01
0.05
Shannon
explanatory power of the regression for either vascular plant
species or breeding bird species (Table 3).
2
landscape heterogeneity at the km level are thus relevant.
For plants, a larger number of quadrats would have been
desirable to give a better re ection of variation at the level of
the entire square. This was not possible due to the economic
constraints of the project. Clearly, uncommon species are
unlikely to be recorded in the limited number of quadrats
used. Nevertheless, the standardised method employed
provides a relative measure that is suitable for comparisons
between squares and that provides data at a much Ž ner
resolution than the data extracted from aerial photographs.
The correlation between the map-based heterogeneity index
and the simple measure of vascular plant species richness
provides support for the use of Hix as a coarse-resolution
indicator. Further work should focus on more detailed
sampling of the vegetation, to determine whether the
relationship also holds true at still Ž ner levels of resolution.
For both breeding birds and vascular plants, we cannot
assume that the results from south-eastern Norway can be
extrapolated to different parts of the country. Although the
monitoring programme focuses speciŽ cally on agricultural
landscapes, the nature of and causes of land type hetero-
geneity in the landscapes of south-eastern Norway will be
very different than those in, for example, the far north of
Norway. In addition, the species present in the landscape are
clearly different in different regions. This must be taken into
account when interpreting national indicator values.
Discussion
Methodological considerations
Diversity and heterogeneity indices are dependent on the
land classiŽ cation system used and other mapping rules,
including the minimum size of mapping units and the number
of land type categories that are distinguished. Such indices
are, therefore, most appropriate for comparative use, where
relative values are compared for maps created using identical
mapping rules.
The Hix index is sensitive to the spacing of sample points.
This spacing should, therefore, be related to the scale of
landscape phenomena of interest. In studies of pattern
analysis a rule of thumb is that sampling should be at a
greater frequency than half the length of the smallest pattern
of interest (Greig-Smith 1983). The choice of 100 m spacing
between points was made partly on the basis of average Ž eld
sizes in Norwegian agricultural landscapes. However, further
testing is needed to deŽ ne the optimal spacing of points for
measuring landscape heterogeneity in agricultural land-
scapes for various interests.
The heterogeneity index Hix is a simple autocorrelation
Fig. 3. Relationship between number of vascular plant species and (a) Shannon’s diversity index values and (b) heterogeneity index values per square in 20
sample squares (1 £ 1 km) in South-eastern Norway.

NORSK GEOGRAFISK TIDSSKRIFT 55 (2001)
Heterogeneity in agricultural landscapes 75
Table 3. Coefficients for the multiple regression of numbers of breeding
birds against heterogeneity index values (Hix) and Shannon’s diversity index
values per square; and numbers of vascular plant species against both index
values per square: y = a (Hix) ‡ b (Shannon) ‡ c
Indicator relevance
Interpreting landscape indices is often complicated by the
fact that a change in one landscape element within a given
area will affect the values possible for other indicators. For
example, the greater the proportion of forest in a 3Q-sample
square, the lower the proportion of open land. Since the main
land types in the agricultural landscapes of south-eastern
Norway are agricultural land and forest, increasing hetero-
geneity values in the squares almost exclusively re ect an
increasing amount of forest. Lower heterogeneity values tend
to represent large areas of open farmland rather than large
areas of forest. To disentangle the relative effects of different
landscape content and conŽ guration clearly requires a more
thorough analysis. Nevertheless, the heterogeneity index
explains more of the variation in the species data than does
area of forest. This suggests that simple measurements of
forest area alone are insufŽ cient to explain patterns of species
diversity and that spatial pattern can provide a better
indicator in the agricultural landscape of south-eastern
Norway. Spatial heterogeneity is thus a valuable addition
to indicators based merely on landscape composition, such as
areas of given land types or diversity of land types
Dependent variable (y)
Plants
Value
Birds
SigniŽ cance
Value
SigniŽ cance
a
b
c
84.65
¡13.77
61.93
<0.01
0.27
<0.01
<0.01
34.44
¡2.87
5.76
<0.01
0.69
0.57
2
R
0.55
0.63
<0.01
many inter-related variables is a challenging task that
requires data from a large and wide-ranging sample of
landscapes.
Indicators should be ‘efŽ cient’ in two senses. They should
efŽ ciently measure the phenomenon we wish to study (for
example, biodiversity) and they should be economically
efŽ cient such that they can be used in a large survey.
Economic efŽ ciency requires a high degree of simpliŽ cation
and the challenge is to Ž nd the intersection between what can
realistically be measured within available budget constraints
and what is an acceptable degree of simpliŽ cation.
(
Shannon’s index). Whether this relationship will hold true
in different types of landscape remains to be seen.
The difference between landscape diversity as measured
by the Shannon’s index and landscape heterogeneity as
measured by Hix is that the former includes information only
on landscape composition – the types of land present and
their abundance – whilst the latter provides information on
spatial variability in the landscape. However, the hetero-
geneity index describes only whether adjacent blocks of land
are different from one another and therefore only indirectly
re ects content (number of land types) (Fig. 4).
It would thus appear that Shannon’s diversity index and
the heterogeneity index should be seen as complementary to
one another. However, adding the diversity measures did not
contribute signiŽ cantly to the explanatory power of regres-
sion models for our test squares (Table 3). We interpret this
result as a re ection of the landscape structure, where high
heterogeneity is usually a result of a number of different,
interspersed land types. This illustrates the difŽ culties of
interpreting landscape data. Distinguishing the effects of
Great caution is needed in interpreting indicator results
because even when a relationship is shown to exist, the
application of this information can be difŽ cult. For example,
although we have shown that species numbers in the
agricultural landscape are higher in more heterogeneous
squares, the indicator provides no information about which
species are present. If many species are common, generalist
species, then achieving a high species number may not be a
relevant goal for the agricultural sector. Rather, there may be
other more appropriate goals, for example, related to the
conservation of species that are strongly associated with
traditional agricultural practices.
Similarly, indicators and goals associated with the
management of agricultural landscapes may not be appro-
priate for other types of landscape. For example, to maintain
national biological diversity requires diversity of landscapes
at a range of scales. If all landscapes were heterogeneous at
the 1 km £ 1 km scale, there would be no habitat remaining
for species requiring large core areas of habitat. Similarly, a
homogeneous, species-poor habitat such as heathland may
support species of high conservation interest. Thus hetero-
geneity may be positive in some situations but negative in
others and must be related to the wider landscape context.
As a general rule, the interpretation of landscape indices
requires an understanding both of what the indices describe
and how this information relates to speciŽ c environmental
targets. The indices cannot tell us how best to manage the
landscape, only how well we are achieving pre-deŽ ned
targets.
Fig. 4. Two examples of landscape structure. Both landscapes comprise the
same number of land types (2) and the same distribution of land between
types (50%). Both landscapes will thus have the same Shannon’s index value
but different Hix values. Hix will be low in landscape 1 (many identical
neighbouring points) but high in landscape 2 (few identical neighbouring
points).
Conclusions
The results presented in this paper support the hypothesis that
spatial patterns in the landscape in uence patterns of species

76
W. J. Fjellstad et al.
NORSK GEOGRAFISK TIDSSKRIFT 55 (2001)
occurrence. Interestingly, the heterogeneity index (Hix) was
a better indicator of species richness of birds and plants then
the Shannon’s diversity index, which has previously been
suggested for use in national landscape monitoring.
Methodological considerations must be thoroughly eval-
uated to ensure that sampling and mapping methods enable
calculation of indicators that are of relevance to the topic of
study. Evidence so far suggests that a single indicator will not
be relevant for all landscape functions, or even all aspects
within one function. Rather, a suite of indicators will be
necessary.
The development of new landscape indices is proceeding
at great speed and with today’s technology these can often be
quickly applied to existing map data. However, when using
landscape metrics as indicators it is important to ensure that
the limitations of the indicator approach are understood –
both in terms of what we can expect from indicator values
and what we can use them for. For example, not all of the
attributes that are measurable are of relevance for monitoring
and those that are may be relevant only at a given scale. We
will follow up this study with an investigation of how the Hix
indicator is in uenced by sampling intensity and by the size
and scale of the landscape where it is applied.
Ekstam, U. & Forshed, N. 1992. Om H a¨ vden Upph o¨ r: K a¨ rlv a¨ xter som
Indikatorarter i Angs- och Hagmarker. AB F a¨ lths Tryckeri, V a¨ rnamo.
¨
Fahrig, L. 1997. Relative effects of habitat loss and fragmentation on
population extinction. Journal of Wildlife Management 61, 603–610.
Farina, A. 1998. Principles and Methods in Landscape Ecology. Chapman &
Hall, London.
Forman, R. T. T. 1995. Land Mosaics: the Ecology of Landscape and
Regions. Cambridge University Press, Cambridge.
Forman, R. T. T. & Godron, M. 1986. Landscape Ecology. John Wiley &
Sons, New York.
Greig-Smith, P. 1983. Quantitative Plant Ecology. Blackwell ScientiŽ c
Publications, Oxford.
Haila, Y. 1985. Birds as a tool in reserve planning. Ornis Fennica 62, 96–100.
Hanski, I. 1997. Metapopulatio n dynamics: from concepts and observations
to predictive models. Hanski, I. & Gilpin, M. E. (eds.) Metapopulation
Biology: Ecology, Genetics and Evolution, 69–91. Academic Press,
London.
Ims, R. A. 1990. Hva er landskapsøkologi? Problem og metoder. Fauna 43,
1
51–171.
Magurran, A. E. 1988. Ecological Diversity and Its Measurement. Croom
Helm, London.
Mathiesen, H. F., Fjellstad, W. J. & Stokland, J. N. 1999. Tilstandsoverv aÊ k-
ing og resultatkontroll i jordbrukets kulturlandskap ved hjelp av utvalgs-
kartlegging. Landbruksøkonomisk Forum 3/99, 49–59.
McGarigal, K. & Marks, B. 1995. FRAGSTATS: Spatial Analysis Program
for Quantifying Landscape Structure. General Technical Report PNW-
GTR-351. Portland, Oregon: US Department of Agriculture, Forest
Service, PaciŽ c Northwest Experiment Station. 122 pp.
McInerney, J. P. 1994. Duality
–
the future face of farming. Journal of the
Although we urge caution, we feel that the indicator
approach has great potential in large-scale monitoring. The
heterogeneity index Hix has so far been shown to re ect
landscape spatial pattern and show clear links with aspects of
biodiversity in the agricultural landscape of south-eastern
Norway.
Royal Agricultural Society of England 155, 25–33.
Meeus, J. H. A. 1993. The transformation of agricultural landscapes in
Western Europe. The Science of the Total Environment 129, 171–190.
Merriam, G. 1988. Landscape dynamics in farmland. Trends in Ecology and
Evolution 3, 16–20.
OECD 1994. Environmental Indicators: OECD Core Set. Organisation for
Economic Co-operation and Development, Paris.
OECD 1997. Environmental Indicators for Agriculture: OECD Progress
Report. Organisation for Economic Co-operation and Development, Paris.
OECD 1999. Environmental Indicators for Agriculture. Volume 2: Issues and
Design – the York Workshop. Organisation for Economic Co-operation and
Development, Paris.
O’Neill, R. V., Krummel, J. R., Gardner, R. H., Sugihara, G., Jackson, B.,
DeAngelis, D. L., Milne, B. T., Turner, M. G., Zygmunt, B., Christensen,
S. W., Dale, V. H. & Graham, R. L. 1988. Indices of landscape pattern.
Landscape Ecology 1, 153–162.
Acknowledgements.–This project was co-Ž nanced by the Norwegian Re-
search Council, the Norwegian Institute of Land Inventory (NIJOS) and the
Norwegian Institute for Nature Research (NINA). 3Q data were made
available through the 3Q monitoring programme, jointly funded by the
Ministry of Agriculture and the Ministry of the Environment. Thanks to Mary
Edwards and two anonymous reviewers for helpful comments on the
manuscript.
Manuscript submitted 30 November 2000; accepted 7 March 2001
Opdam, P. 1990. Dispersal in fragmented populations: the key to survival.
Bunce, R. G. H. & Howard, D. C. (eds.) Species Dispersal in Agricultural
Habitats, 3–17. Belhaven Press, London.
Rolstad, J. 1991. Consequences of forest fragmentation for the dynamics of
bird populations : conceptual issues and the evidence. Biological Journal of
the Linnean Society 42, 149–163.
References
Bastian, O. & Bernhardt, A. 1993. Anthropogenic landscape changes in
Central Europe and the role of bioindication. Landscape Ecology 8, 139–
Smeets, E. & Weterings, R. 1999. Environmental Indicators: Typology and
Overview. Technical Report No. 25. European Environment Agency,
Copenhagen.
1
51.
Brouwer, F. & Crabtree, B. (eds.) 1999. Environmental Indicators and
Agricultural Policy. CAB International, Wallingford, UK.
Burt, T. P. 1994. Long-term study of the natural environment – perceptive
Troll, C. 1968. Landschaft o¨ kologie. Tuxen, R. (ed.) P anzensoziologi e und
Landschafts o¨ kologie, 1–21. Verlag Dr. W Junk N. V., The Hague.
Tucker, G. 1999. Measuring the impacts of agriculture on biodiversity.
science or mindless monitoring. Progress in Physical Geography 18, 475–
Brouwer, F.
& Crabtree, B. (eds.) Environmental Indicators and
4
96.
Agricultural Policy, 89–104. CAB International, Wallingford, UK.
Wascher, D. M. 1996. Methodology for Assessment of Europe’s Biodiversity.
European Centre for Nature Conservation, Tilburg.
Wegge, P., Rolstad, J. & Gjerde, I. 1992. Effects of boreal forest
fragmentation on capercaillie grouse: empirical evidence and management
implications. McCullough, D. R. & Barret, R. H. (eds.) Wildlife 2001:
Populations, 738–749. Elsevier Science Publications, New York.
Wiens, J. A. 1989. Spatial scaling in ecology. Functional Ecology 3, 385–
398.
Cale, P. G. & Hobbs, R. J. 1994. Landscape heterogeneit y indices: problems
of scale and applicability, with particular reference to animal habitat
description. PaciŽc Conservation Biology 1, 183–193.
Dramstad, W. E., Fjellstad, W. J. & Fry, G. L. A. 1998. Landscape indices –
useful tools or misleading numbers? Dover, J. W. & Bunce, R. G. H. (eds.)
Key Concepts in Landscape Ecology, 63–68. International Association of
Landscape Ecology – IALE (UK), Myerscough.
Dramstad, W. E., Fjellstad, W. J., Strand, G.-H., Mathiesen, H. F., Engan, G.
&
Stokland, J. N. Development and implementation of the Norwegian
Wiens, J. A. & Milne, B. T. 1989. Scaling of ‘landscapes’ in landscape
ecology, or, landscape ecology from a beetle’s perspective. Landscape
Ecology 3, 87–96.
monitoring programme for agricultural landscapes. Journal of Environ-
mental Management. In press.