International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132

Contents lists available at ScienceDirect

International Journal of Applied Earth Observation and

Geoinformation
journal homepage: www.elsevier.com/locate/jag

Satellite Earth observation data to identify anthropogenic pressures in

selected protected areas
Harini Nagendra a,∗ , Paola Mairota b , Carmela Marangi c , Richard Lucas d ,
Panayotis Dimopoulos e , João Pradinho Honrado f , Madhura Niphadkar a ,
Caspar A. Mücher g , Valeria Tomaselli h , Maria Panitsa e , Cristina Tarantino j ,
Ioannis Manakos i , Palma Blonda j
a
Ashoka Trust for Research in Ecology and the Environment (ATREE), Royal Enclave, Srirampura, Jakkur Post, Bangalore 560064, Karnataka, India
b
Dipartimento di Scienze Agro-Ambientali e Territoriali, Università degli Studi di Bari “Aldo Moro”, Via Orabona, 4–70126 Bari, Italy
c
National Research Council- Istituto per le Applicazioni del Calcolo, Via Amendola 122/D, 70126, Bari, Italy
d
Centre for Ecosystem Science, School of Biological, Earth and Environmental Sciences, the University of New South Wales, High Street, Kensington, NSW
2052, Australia
e
Department of Environmental and Natural Resources Management, University of Patras, Seferi 2, 30100 Agrinio, Greece
f
CIBIO-Centro de Investigação em Biodiversidade e Recursos Genéticos & Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre, s/n,
4169-007 Porto, Portugal
g
Alterra, Wageningen UR, Droevendaalsesteeg 3, 6708 PB, Wageningen, The Netherlands
h
National Research Council – Institute of Biosciences and BioResources (CNR-IBBR), Via G. Amendola 165/A, 70126 Bari, Italy
i
Information Technologies Institute (ITI), Centre for Research & Technology Hellas, 6th km Harilaou – Thermi, 57001 Thessaloniki, Greece
j
National Research Council – Institute of Intelligent Systems for Automation (CNR-ISSIA), Via G. Amendola 122, 70126 Bari, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Protected areas are experiencing increased levels of human pressure. To enable appropriate conserva-
Available online 22 November 2014 tion action, it is critical to map and monitor changes in the type and extent of land cover/use and habitat
classes, which can be related to human pressures over time. Satellite Earth observation (EO) data and
Keywords: techniques offer the opportunity to detect such changes. Yet association with field information and expert
Biodiversity conservation interpretation by ecologists is required to interpret, qualify and link these changes to human pressure.
Earth observation
There is thus an urgent need to harmonize the technical background of experts in the field of EO data
Changes in state
analysis with the terminology of ecologists, protected area management authorities and policy makers in
Protected areas
order to provide meaningful, context-specific value-added EO products. This paper builds on the DPSIR
framework, providing a terminology to relate the concepts of state, pressures, and drivers with the appli-
cation of EO analysis. The type of pressure can be inferred through the detection of changes in state (i.e.
changes in land cover and/or habitat type and/or condition). Four broad categories of changes in state
are identified, i.e. land cover/habitat conversion, land cover/habitat modification, habitat fragmentation
and changes in landscape connectivity, and changes in plant community structure. These categories of
change in state can be mapped through EO analyses, with the goal of using expert judgement to relate
changes in state to causal direct anthropogenic pressures. Drawing on expert knowledge, a set of pro-
tected areas located in diverse socio-ecological contexts and subject to a variety of pressures are analysed
to (a) link the four categories of changes in state of land cover/habitats to the drivers (anthropogenic pres-
sure), as relevant to specific target land cover and habitat classes; (b) identify (for pressure mapping) the
most appropriate spatial and temporal EO data sources as well as interpretations from ecologists and
field data useful in connection with EO data analysis. We provide detailed examples for two protected
areas, demonstrating the use of EO data for detection of land cover/habitat change, coupled with expert
interpretation to relate such change to specific anthropogenic pressures. We conclude with a discussion
of the limitations and feasibility of using EO data and techniques to identify anthropogenic pressures,
suggesting additional research efforts required in this direction.
© 2014 Elsevier B.V. All rights reserved.

∗ Corresponding author. Tel.: +91 80 23635555; fax: +91 80 23530070.

E-mail addresses: nagendra@atree.org (H. Nagendra), paola.mairota@uniba.it (P. Mairota), c.marangi@ba.iac.cnr.it (C. Marangi), richard.lucas@unsw.edu.au (R. Lucas),
pdimopoulos@upatras.gr (P. Dimopoulos), jhonrado@fc.up.pt (J.P. Honrado), madhura.niphadkar@atree.org (M. Niphadkar), sander.mucher@wur.nl (C.A. Mücher),
valeria.tomaselli@ibbr.cnr.it (V. Tomaselli), mpanitsa@upatras.gr (M. Panitsa), imanakos@iti.gr (I. Manakos), blonda@ba.issia.cnr.it (P. Blonda).

http://dx.doi.org/10.1016/j.jag.2014.10.010
0303-2434/© 2014 Elsevier B.V. All rights reserved.
 H. Nagendra et al. / International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132
Introduction
Local scale anthropogenic pressures on the Earth have led to
accelerated declines in biodiversity (Hooper et al., 2012), affect-
ing the provisioning of ecosystem services essential for human
wellbeing (Mace et al., 2012). Thus, international bodies such as
the recently established Intergovernmental Panel on Biodiversity
and Ecosystem Services (IPBES) have stressed the need to assess
human pressures on biodiversity and ecosystem services across all
scales, including at local scales relevant to management interven-
tion (Honrado et al., 2013).
Although protected areas constitute a cornerstone of local and
international conservation approaches (Jenkins and Joppa, 2009),
they continue to experience anthropogenic pressure (Nagendra,
2008; Vicente et al., 2013). For effective management response,
spatial knowledge of the type and location of pressure is required.
EO and airborne data from passive optical and active Synthetic
Aperture Radar (SAR) and Light Detection and Ranging (LiDAR) are
available at different spectral and spatial resolutions, and can be
very useful for detecting land cover/use and/or habitat changes over
time (Nagendra et al., 2013).
EO and associated techniques, coupled with landscape pattern
analysis and habitat modelling of biodiversity distributions, can
provide information on changes in the state/condition of biodiver-
sity and ecosystems. These changes can be used to infer evidence
of pressures. However, there is an urgent need to harmonize the
technical terminology of experts in the field of EO data analysis
and of ecologists, with the applied scientific terminologies of policy
makers and management authorities. Policy makers and protected
area management authorities use terms such as impact, pressures,
threats, drivers, state, stress–response (Salafsky et al., 2008). It is
difficult to directly relate these terms to corresponding indicators
derived from EO data analysis, which mainly provides information
on the extent and location of changes in land cover/use (Tarantino
et al., 2007).
The aim of this paper is to harmonize the discourse between
field experts, ecologists and protected area managers on the one
hand, and EO experts on the other hand, by proposing a unified
approach to facilitate the provision of value-added products from
EO sources that can be useful for biodiversity conservation pur-
poses. This is illustrated for selected protected areas, located in
diverse socio-ecological contexts and representing diverse habitat
types including wetlands, forests, grasslands and bogs subject to
a variety of pressures, with two detailed examples provided. The
paper concludes with a brief discussion of the advantages and lim-
itations of EO data techniques for pressure assessment in protected
areas.
This study was developed and tested within the European
Union’s Seventh Framework Programme (EU-FP7) project Biodi-
versity Multi-SOurce Monitoring System: From Space To Species
(BIO SOS), that aimed to develop tools and models for consistent
multi-annual monitoring of protected areas and their surroundings
by the integrated use of EO and in-field data.
Methods: definition and categorization of changes in state
We propose to build on the Driving forces, Pressure, State,
Impact and Response (DPSIR) framework (EEA, 1995), defining four
broad categories of changes in state which can be mapped and mon-
itored through EO analyses. Changes in state are likely to be site-
and habitat-specific (Nagendra et al., 2012). These are described,
drawing on a systematic analysis based on literature review and
expert knowledge (judgement), for a number of Natura 2000 sites
in Europe and a protected area in southern India. The sites are fur-
ther described in Table 1. European sites are located in countries
125
belonging to the Mediterranean (Italy, Greece and Portugal) and
Atlantic (Portugal, The Netherlands and Wales) biogeographical
regions (Habitats Directive–92/43/EEC). For each site, interpreta-
tion from ecologists, drawing on field data, is used to link specific
changes in state with the direct pressures considered as drivers in
the protected area. Appropriate EO data are then used to analyse
specific changes in state. Detailed examples are provided for two
protected area sites, illustrating the use of EO data for detecting
changes in state related to pressures.
Taxonomy for pressure assessment
A number of different terms have been used by researchers and
conservation agencies to define pressures and consequent changes
in state of habitats in protected area assessments. Perhaps the
most widely used by policy makers and managers is the DPSIR
framework, although there are a number of other frameworks
and approaches such as strengths, weaknesses, opportunities and
threats (SWOT) analysis (widely used in other contexts), the
WCPA/IUCN framework for assessment and monitoring of pro-
tected areas (Hockings et al., 2006), and the work of Salafsky et al.
(2008), reviewed further in Nagendra et al. (2012).
According to the DPSIR framework, Drivers of environmental
change (e.g., urbanization) result in Pressures on the environment
(e.g., discharge of industrial and domestic waste), which create
changes in the State of the environment (e.g., water quality), cre-
ating Impacts (e.g., modifications of ecological functions and in
ecosystem services provision) on ecology and society. This leads
to Responses (e.g., policy/management responses) – which in turn
feedback on Drivers, thus reinstating the overall cycle. Thus the
DPSIR framework, through state and impact monitoring, allows
for the assessment of response effectiveness and appropriateness
within an adaptive management perspective.
Pressures in the DPSIR framework refer to processes of disturb-
ance that influence changes in biodiversity and ecosystem state.
It is difficult to directly detect pressures, or to assess impacts of
such pressures on society and ecology through EO. EO can however
determine changes in land cover/use or habitat e.g. in forest struc-
ture (state of the forest) as a consequence of pressure (Nagendra
and Rocchini, 2008). Through expert knowledge, changes in state
can be used to infer both the type of disturbance (e.g., wildfire, wind
storm, pest infestation, silviculture, overharvesting) as a driving
pressure, and the type of impacts likely to ensue, so that appropriate
responses can be devised (e.g. Nagendra et al., 2010).
Broad categorization of changes in state
Based on literature review and the knowledge of ecologists,
four broad categories of changes in state are identified, which can
be mapped through EO data analyses, and related to causal direct
anthropogenic pressures based on expert ecological interpretation.
These are
1 Land cover/habitat conversion;
2 Land cover/habitat modification (e.g., in vegetation structure, as
indicated by changes in density or cover, height);
3 Habitat fragmentation and changes in landscape connectivity;
4 Changes in plant community structure.
In some instances land cover/use classes (e.g. deciduous forest)
coincide with habitat categories, but in general the definition of a
habitat (sensu Bunce et al., 2013) is linked to that of an ecosystem,
and includes additional ecological concepts and components (e.g.,
lithology, slope, water salinity).
Expert knowledge is crucial in the EO analytical phase. Ecologists
and other field experts rely on processes of human discrimination
 126
Table 1
Description of study sites.

Natura 2000 site name Natura Habitats of conservation concern (*) General description
2000 Code

Rios Sabor and Maçãs PTZPE0037 Intermittently flowing Mediterranean rivers of the Paspalo-Agrostidion
and (3290)
PTCON0021 Stable xerothermophilous formations with Buxus sempervirens on rock
slopes (Berberidion p.p.) (5110)
Siliceous rocky slopes with chasmophytic vegetation (8220)
Thermophilous Fraxinus angustifolia woods (91B0)
Alluvial forests with Alnus glutinosa and Fraxinus excelsior (Alno-Padion,
Alnion incanae, Salicion albae) (91E0*)
Salix alba and Populus alba galleries (92A0)
Quercus suber forests (9330)
Quercus ilex and Quercus rotundifolia forests (9340)
Endemic forests with Juniperus spp. (9560*)
Peneda-Gerês PTZPE0002 Temperate Atlantic wet heaths with Erica ciliaris and Erica tetralix (4020*)
and European dry heaths (4030)
PTCON0001 Arborescent matorral with Laurus nobilis (5230*)
Oro-Iberian Festuca indigesta Grasslands (6160)
Species-rich Nardus grasslands, on siliceous substrates in mountain areas
(and submountain areas, in Continental Europe) (6230*)
Lowland hay meadows (Alopecurus pratensis, Sanguisorba officinalis) (6510)
Siliceous rocky slopes with chasmophytic vegetation (8220)
Siliceous rock with pioneer vegetation of the Sedo-Scleranthion or of the Sedo
albi-Veronicion dillenii (8230)
Alluvial forests with Alnus glutinosa and Fraxinus excelsior (Alno-Padion,
Alnion incanae, Salicion albae) (91E0*)
Galicio-Portuguese oak woods with Quercus robur and Quercus pyrenaica
(9230)
Cors Fochno UK0014791 Active raised bogs (7110)
Degraded raised bog (7120)
Wekeromse Zand and NL9801023 Inland dunes with open Corynephorus and Agrostis grasslands (2330)
Ederheide & Ginkelse Dry sand heaths with Calluna and Genista (2310), Eur. dry heaths (4030)
heide Species-rich Nardus grasslands, on silicious substrates (6230)
Atlantic acidophilous beech forests with Ilex (9120), Old acidophil. oak
woods with Q. robur on sandy plains (9190)
Murgia Alta IT9120007 Semi-natural dry grasslands and scrubland facies on calcareous substrates
(Festuco-Brometalia) (important orchid sites) (6210) Pseudo-steppe with
grasses and annuals of the Thero-Brachypodietea (6220*)
Eastern sub-Mediterranean dry grasslands (Scorzoneratalia villosae) (62A0)
Calcareous rocky slopes with chasmophytic vegetation (8210)
Eastern white oak woods (91AA)
Quercus trojana woods (9250)
H. Nagendra et al. / International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132
Portuguese river valley system with Mediterranean type of climate and landscape mosaic
dominated by open sclerophyllous woodland and scrub of Quercus suber, Quercus ilex and Juniperus
oxycedrus, in mosaic with dry grasslands. In uplands and plateau areas oak forests, old chestnut
groves and dry scrub predominate in the landscape. Cliffs and steep slopes are common in the
main valley as well as several types of riparian vegetation.
Jackson, A.L.R., 2011. Renewable energy vs. biodiversity: policy conflicts and the future of nature
conservation. Glob. Environ. Change 21, 1195–1208.
Portuguese upland mosaic of mountain farmland and grazing lands, including native deciduous
oak forests, hay meadows, acidophyllous grasslands and heathlands. Lowland valleys are
dominated by pine plantations, oak woodlands and dry heath. Riparian forests and laurel thickets
are common along river valleys. Saxicolous habitats are common at high elevations and host
endemic plant species and vegetation types.
Cerqueira, Y., Araújo, C., Vicente, J., Pereira, H.M., Honrado, J., 2010. Ecological and cultural
consequences of agricultural abandonment in the Peneda-Gerês National Park (Portugal). Chapter
2 in: Evelpidou, N., Figueiredo, T., Mauro, F., Tecim, V., Vassilopoulos, A. (Eds.) “Natural Heritage
from East to West”. 1st Edition, 2010, XIII, 363 p., Springer.
One of the largest surviving lowland bogs in the UK with both an active raised bog and degraded
raised bog within an estuarine setting.
http://jncc.defra.gov.uk/ProtectedSites/SACselection/sac.asp?EUCode=UK0014791
A part of Veluwe National Park, the largest end moraine in the Netherlands that was created during
penultimate glacial period. The final landscape of alternating sand dune areas, heathlands and dry
forests were created by a long history of intensive land use (starting from Celtic fields in Iron Age).
Mücher, C.A., Kooistra, L., Vermeulen, M., Vanden Borre, J., Haest, B., Haveman, R., 2013.
Quantifying structure and function of Natura 2000 heathland habitats using spectral mixture
analysis and segmentation techniques on hyperspectral imagery. Ecol. Indic. 33, 71–81.
Italian semi-natural grasslands and other semi-natural vegetation (garrigues, scrubs, Quercus
pubescens, Q. trojana and Q. coccifera coppices) and conifer plantations cast in an agricultural
landscape matrix dominated by cereal crops and, to a lesser extent, comprised of olive groves,
almond orchards and vineyards.
Mairota, P., Leronnii, V., Xi, W., Mladenoff, D.J., Nagendra, H., 2014. Using spatial simulations of
habitat modification for adaptive management of protected areas: Mediterranean grassland
modification by woody plant encroachment. Environ. Conserv. 41, 144–156.
 Le Cesine IT9150032 Coastal lagoons (1150) Italian complex of coastal lagoons with intricate coenological pattern. Fine grained mosaic of
and Annual vegetation of drift lines (1210) helophytic stands, stripes of halophilous and dry therophytic vegetation, water pools, marshes.
IT9150014 Salicornia and other annuals colonizing mud and sand (1310) Sand dunes banks host remnants of Juniperus macrocarpa subsp. macrocarpa communities. Inland
Mediterranean salt meadows (Juncetalia maritimi) (1410) this complex is surrounded by a forest landscape mosaic composed of Mediterranean maquis and
Mediterranean and thermo-Atlantic halophilous scrubs (Sarcocornetea garrigues, Quercus ilex woods and Pinus halepensis plantations.
fruticosi) (1420) Tomaselli, V., Sciandrello, S., Di Pietro, R., 2011. Plant communities structure and composition in
Embryonic shifting dunes (2110) three coastal wetlands in southern Apulia (Italy). Biologia 66, 1027–1043.
Shifting dunes along the shoreline with Ammophila arenaria (‘white dunes’) Tomaselli, V., Tenerelli, P., Sciandrello, S., 2012. Mapping and quantifying habitat fragmentation in

H. Nagendra et al. / International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132
(2120) small coastal areas: a case study of three protected wetlands in Apulia (Italy). Environ. Monit.
Malcolmietalia dune grasslands (2230) Assess. 184, 693–713.
Coastal dunes with Juniperus spp. (2250) Tomaselli, V., Dimopoulos, P., Marangi, C., Kallimanis, A.S., Adamo, M., Tarantino, C., Panitsa, M.,
Mediterranean temporary ponds (3170) Terzi, M., Veronico, G., Lovergine, F., Nagendra, H., Lucas, R.M., Mairota, P., Mücher, C.A., Blonda, P.,
Calcareous fens with Cladium mariscus and species of the Caricion davallianae 2013. Translating land cover/land use classifications to habitat taxonomies for landscape
(7210) monitoring: a Mediterranean assessment. Landsc. Ecol. 28, 905–930.
Kalamas River delta GR2120001 Coastal lagoons (1150) Landscape mosaic in Greece, comprised of riparian forest stands and a coastal lagoon as well as
Salix alba and Populus alba galleries (92AO) and salt meadows, halophilous scrubs, reed and tamarisk thickets, estuaries, coastal lagoons, and
Southern riparian galleries and thickets (Nerio-Tamaricetea and Securinegion coastal dune systems.
tinctoriae) (92DO) Goedecke, F., 2009. Diversität und Zustandsbewertung der Auwälder- und Gebüsche im
Sarcopoterium spinosum phryganas (5420) FFH-Gebiet Kalamas, NW-Griechenland. In: Biologischen Fakultät – Albert-von-Haller Instituts für
Mediterranean salt meadows (Juncetalia maritimi) (1410) Pflanzenwissenschaften, Göttingen: Georg-August-Universität Göttingen, pp. 143.
Mediterranean and thermo-Atlantic halophilous scrubs (Sarcocornetea Tomaselli, V., Dimopoulos, P., Marangi, C., Kallimanis, A.S., Adamo, M., Tarantino, C., Panitsa, M,
fruticosi) (1420) Terzi, M, Veronico, G, Lovergine, F, Nagendra, H, Lucas, RM, Mairota, P, Mücher, C.A., Blonda, P.,
2013. Translating land cover/land use classifications to habitat taxonomies for landscape
monitoring: a Mediterranean assessment. Landsc. Ecol. 28, 905–930.
Kalodiki Fen GR2120002 Brackish marshes (Cladium mariscus) (7210) Wetland in Greece consisting of two lakes embedded in a topogenous mire, and Kalodiki Fen, an
Sarcopoterium spinosum phryganas (5420) inland, highly biodiverse freshwater wetland.
Hellenic forests with Quercus coccifera (934A) Dimopoulos, P., Sýkora, K., Gilissen, C., Wiecherink, D., 2005. Vegetation ecology of Kalodiki Fen
(NW Greece). Biologia (Bratislava) 60, 69–82.
Sarika-Hatzinikolaou, P., Koumpli-Sovantzi, M.L., Yannitsaros, A., 1996. The vascular flora of Lake
Kalodiki (Ipiros, NW Greece). Webbia 50, 223–236.
Sarika-Hatzinikolaou, P., Koumpli-Sovantzi, M.L., Yannitsaros, A., 1996. The vascular flora of Lake
Kalodiki (Ipiros, NW Greece). Webbia 50, 223–236.
Tomaselli, V., Dimopoulos, P., Marangi, C., Kallimanis, A.S., Adamo, M., Tarantino, C., Panitsa, M.,
Terzi, M., Veronico, G., Lovergine, F., Nagendra, H., Lucas, R.M., Mairota, P., Mücher, C.A., Blonda, P.,
2013. Translating land cover/land use classifications to habitat taxonomies for landscape
monitoring: a Mediterranean assessment. Landsc. Ecol. 28, 905–930.
Kalamas River Gorge GR2120004 Riparian forests with Platanus orientalis (92C0) Riparian forest landscape in Greece dominated by deciduous woodlands and riparian forest.
Salix alba and Populus alba galleries (92A0) Goedecke, F., 2009. Diversität und Zustandsbewertung der Auwälder- und Gebüsche im
Sarcopoterium spinosum phryganas (5420) FFH-Gebiet Kalamas, NW-Griechenland. In: Biologischen Fakultät – Albert-von-Haller Instituts für
Pflanzenwissenschaften. Göttingen: Georg-August-Universität Göttingen, pp. 143.
Tomaselli, V., Dimopoulos, P., Marangi, C., Kallimanis, A.S., Adamo, M., Tarantino, C., Panitsa, M.,
Terzi, M., Veronico, G., Lovergine, F., Nagendra, H., Lucas, R.M., Mairota, P., Mücher, C.A., Blonda, P.,
2013. Translating land cover/land use classifications to habitat taxonomies for landscape
monitoring: a Mediterranean assessment. Landsc. Ecol. 28, 905–930.
Biligiri Rangaswamy N/A Evergreen forests Highly biodiverse landscape located between the Eastern and Western Ghats region of India. The
Temple Tiger Semi-evergreen forests area surrounding the park is densely populated, with agriculture providing the main occupation
Reserve† Shrub Savanna (grasslands) and high dependence on natural vegetation for fuelwood and fodder.
Riparian forests Ramesh, B.R., 1989. Flora of Biligiri Rangan Hills. Ph.D. thesis, Madras University.
Shrub woodland to thickets with bamboo
Tree Savanna

(*) For European Union sites habitats types of Annex I of the European Habitats Directive (Council Directive 92/43/EEC) are indicated according to the European Commission, 2013. Interpretation Manual of European
Habitats–EUR28. Published by the European Commission, DG Environment, Nature and Biodiversity.
(†) Not a European Union site, thus Natura 2000 system and Annex I habitat type definitions do not apply.

127
128 H. Nagendra et al. / International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132
for classifications of habitat types and habitat conditions. These
processes, often fuzzy, scale dependent and site specific, need to
be translated into explicit rules and algorithms that can be sequen-
tially applied across locations, to convert EO data into analytical
products. The importance of expert interpretation is highlighted by
Shao and Wu (2008). The sections below provide a brief overview
of the approaches used by analysts for conversion of EO data into
information on various categories of changes in state.
Land cover/habitat conversion
The detection of land cover/habitat change frequently relies
on supervised processes of change detection (Tarantino et al.,
2007), requiring ground reference information, which can be pro-
vided from field data, by photointerpretation or from prior expert
knowledge of the site, for training the classifiers and subsequently
validating the results. Classification is typically conducted using
per-pixel classifiers, but object based change detection techniques
often provide better performance for very high spatial resolution
(VHR) imagery (Chen et al., 2012).
Land cover/habitat conversions, once detected, can be related
with information on changes in species diversity (Nagendra, 2001;
Buchanan et al., 2008). These can also be related to habitat fragmen-
tation and changes in landscape connectivity, through the coupling
of GIS and ecological modelling with EO.
Land cover/habitat modification
It is challenging to map land cover and habitat modification, for
instance when these involve changes taking place below the top
canopy in forest habitats (Joseph et al., 2011). Land cover mapping
through VHR EO images and subsequent translations to habitats,
based on the Food and Agricultural Organization (FAO) Land Cover
Classification System (LCCS) taxonomy (Di Gregorio and Jansen,
2005), have been demonstrated to be helpful for assessing land
cover and habitat modification (Lucas et al., this issue; Tomaselli
et al., 2013). As an example, habitat modification in native vegeta-
tion can be detected by relating changes in the LCCS classifier cover,
which ranges from closed (up to 65%) to open (65–15%) and sparse
(15–1%). VHR images are very useful to assess habitat modification
due to pressures such as pest infestation (Fuller, 2007) and grazing
(Allard, 2003) in diverse terrestrial habitats.
Direct associations have been established between spectral
bands or indices, and plant diversity (Joseph et al., 2011). Textu-
ral, spectral and structural features developed from EO (Koch and
Ivits, 2004; Costanza et al., 2011) have also been related to species
richness of plants, birds and insects. Changes in within-patch spa-
tial patterns have provided indicators of change in habitat quality
using indicators such as texture (St-Louis et al., 2006), and spectral
mixture analysis (Mücher et al., 2013). Active remote sensing such
as LiDAR, with its ability to penetrate through the top vegetation
canopy, has been used to map changes in the three-dimensional
structure of vegetation (Vierling et al., 2008), relating these changes
to silvicultural practices (Pirotti, 2011), and to habitat modifications
(Graf et al., 2009), sometimes in combination with VHR imagery
(Adamo et al., 2014).
Habitat fragmentation and changes in landscape connectivity
Indicators of habitat fragmentation and changes in landscape
spatial pattern have been used to monitor changes in state due to
human and natural pressure (Pôças et al., 2011). Changes in spa-
tial pattern can be related to habitat suitability for specific species
(Torres et al., 2010). EO derived land use/land cover (LULC) maps
can be very important to derive baseline characterization of spa-
tial pattern for monitoring (Mairota et al., 2013), and can assist in
modelling the ecological implications of fragmentation on habitat
quality (Mairota et al., 2014a,b). Changes in spatial pattern observed
through EO can also provide useful inputs to improve modelling of
species and community distribution (Padoa-Schioppa et al., 2012),
of landscape functional connectivity (Mairota et al., 2011), and of
future vegetation dynamics (Mairota et al., 2014a,b).
Changes in plant community structure
VHR imagery has been utilized for the delineation of top canopy
vegetation such as invasive plants (He et al., 2011), trees (Agarwal
et al., 2013), reeds (Everitt et al., 2005), and grasses (Hall et al.,
2011). Pressures that cause disruptions in plant community struc-
ture are more difficult to assess using EO. The date of image
collection is critical, and should ideally be timed to coincide with
specific phenological stages that make species of interest easy to
identify (Turner et al., 2003). For species mapping, there are impor-
tant trade-offs between spatial and temporal resolutions of satellite
imagery, which are difficult to resolve (Nagendra and Rocchini,
2008).
Results: applications of the approach
Utilizing the approach described above, and drawing on infor-
mation from EO analysts familiar with the protected areas studied
in BIO SOS, a site and habitat specific categorization of changes in
state and pressures is described (Supplement 1). Based on expert
information from collaborations between ecologists and EO data
analysts familiar with each site, we found that the majority of the
changes in state in the sites seem to be distinguishable through
EO. However certain types of changes in state, such as soil dis-
turbance due to sod cutting in The Netherlands, or changes in
habitat quality due to point pollution sources in various sites, are
unlikely to be detected using EO. Expert understanding and cross-
talk between ecologists and EO analysts provides important inputs
into the process of target identification, EO data procurement, and
identification of the most appropriate methods of analysis. Across
sites belonging to different biogeographic regions characterized by
a range of habitat types and socio-ecological conditions, the spatial
and temporal resolution of EO datasets should match the temporal
and spatial scales of the pressures observed in each site (Supple-
ment 1).
Two case studies reported as an extended demonstration of this
approach illustrate how expert knowledge can be used to relate
changes in state with specific ongoing pressures in each protected
site.
Case study – Le Cesine
Le Cesine (Natura 2000 SCI code IT9150032, SPA code
IT9150014) is a coastal area, mainly composed by a complex of
coastal lagoons and characterized by a system of ponds, marshes
and wet meadows. The site covers an area of about 21.5 km2
(2148 ha). The wetland is one of the most important sites in south-
ern Italy in terms of biodiversity. Reed and sedge communities
are prone to fires spreading from the adjacent agricultural areas
and from arson. Outside the boundary of the protected area, a
golf course was created in an area formerly covered by a mosaic
of Mediterranean grassland (Annex I 6220* “Pseudo-steppe with
grasses and annuals of the Thero-Brachypodietea” of the 92/43 EEC
Directive–EUNIS E1.313 for the annual grasslands) and scrub habi-
tats (EUNIS F6.2C). Part of the same area was subsequently modified
to create artificial water bodies (tanks) to supply the golf green. EO
data was used to assess land cover conversion from grassland to
artificial water bodies. Such an assessment requires the post clas-
sification comparison of two land cover maps from different years
(Bruzzone and Bovolo, 2013). Multi-seasonal information (in this
location, for spring and winter images) would have been useful
for classification purposes given the low number of spectral bands
(four) in QuickBird images. However, due to the limited archive
 H. Nagendra et al. / International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132 129

Fig. 1. (a) Quickbird image at time T1; (b) Quickbird image at time T2; (c) NDVI feature at time T1; (d) NDVI feature at time T2; (e) NDVI image difference between time T1
and T2; (f) thresholded NDVI image difference between time T1 and T2.

availability for specific years of interest, a procedure of unsuper-
vised direct image comparison (Bruzzone and Bovolo, 2013) was
adopted. The analysis utilized two VHR QuickBird images, at 2.4 m
spatial resolution, acquired on July 2005 (Fig. 1a), and on June 2009
(Fig. 1b). Image-differencing was applied to the normalized dif-
ference vegetation index (NDVI) computed from the two images,
after relative radiometric calibration of each of the images, aimed
to reduce false changes. A threshold was applied to the differ-
ence image to detect areas of change. Visual interpretation of VHR
images by expert ecologists working in combination with EO ana-
lysts was essential to relate changes observed in the images with
land cover class conversion. Thus, the pressure active in this region
was related to the presence of a golf course. Besides the direct loss
of important sections of the grassland habitat at the margins of
the protected area, such pressures can ultimately result in indi-
rect impacts on aquatic habitat types located in the protected area,
through changes in water-table salinity induced by surface water
extractions from the tanks.
130 H. Nagendra et al. / International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132

Fig. 2. (a) Worldview image at time T1; (b) worldview image at time T2; (c) red/blue ratio at time T1; (d) red/blue ratio at time T2; (e) red/blue image difference between
time T1 and T2; (f) thresholded red/blue image difference between time T1 and T2.

Case study – Ederheide and Ginkelse heide
Ederheide and Ginkelse heide are located within the National
Park “Veluwe” in the Province of Gelderland (SCI code NL9801023,
SPA code NL3009017). The Veluwe Natura 2000 site in The
Netherlands (Fig. 2) consists of sand dune areas alternating with
heathlands and dry forests, and is shaped by a long history of
intensive land use. Ederheide and Ginkelse heide cover an area
of approximately 1000 ha, where heathland is the dominant land
cover. Urbanization, tourism and agriculture are the main pressures
affecting the site.
Inland sand dunes in this region have been impacted from nitro-
gen deposition caused by intensive agriculture outside the park,
leading to encroachment of moss, grass, shrubs and trees. The goal
of the analysis was to detect land cover changes associated with cut-
ting of forest plantation and sod, implemented by park managers,
to restore sand dune dynamics that were impacted due to excessive
nitrogen deposition. The analysis utilized two Worldview II images,
 H. Nagendra et al. / International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132
at 2 m spatial resolution, acquired on June 2011 (Fig. 2a) and March
2013 (Fig. 2b) respectively (Mücher et al., this issue). The approach
taken was similar to the previous study. Following relative radio-
metric calibration, a red/blue band image-differencing algorithm
was applied to the two images to identify locations of change.
Expert knowledge was used to relate these changes to cutting of for-
est plantation and sod covered areas, and to associate these changes
with direct interventions by protected area managers.
Discussion
The usefulness of EO data to inform biodiversity conservation
and environmental management has been highlighted by remote
sensing specialists and ecologists in recent years (Nagendra et al.,
2013; Pettorelli et al., 2014). The progress of such integration has,
it seems, been slower than anticipated (Pettorelli et al., 2014), and
requires greater coordination between these two research commu-
nities to develop a common shared understanding, terminology,
and problem definition. Such synchronization is key to improving
the potential for EO data to effectively support environmental man-
agement decisions by providing adequate value added products.
The paper builds on collaborations between ecologists and EO
analysts familiar with site-specific habitat and land cover related
changes in state in specific protected areas, to elicit the linkage
between the ecological concept of pressures and the EO detec-
tion of change in state. While high resolution (HR) Landsat data
series are available for free and can be used to extract specific
class trends through post-classification comparison of LULC maps
(Tarantino et al., 2007), change detection is frequently requested
at fine resolution for use in local/regional decision making. The
lack of multi-seasonal very high resolution (VHR) archive data on
the same area can limit the production of accurate LULC maps to
be compared for the detection of from-to class transitions, with
these closely related to pressures on the environment. In such
cases, automatable, unsupervised techniques can be combined
with expert knowledge and visual interpretation of images to infer
class transitions, as demonstrated in the two case studies presented
within this paper. Some of the applied EO data analysis routines
are well established. The importance lies in the close collabora-
tion between ecologists and EO analysis, both of whom are familiar
with the site specific context of pressures and changes in state in the
areas of study. This enables the selection of appropriate site-specific
spatial and temporal scales of EO data for identifying pressures and
changes in state, and of identification of best EO analytical tech-
niques.
VHR data are only tasked on specific areas and requirements
but not collected globally, as HR Landsat data is for example. Use
of different VHR sensors for comparison of land cover change is
problematic due to leads to variation in critical factors such as
classification accuracy, spatial resolution, data processing, mean
accuracy, and seasonality (Shao and Wu, 2008). However, with the
advent of new imagery such as that provided by the launch of ESA’s
SENTINEL, open VHR data will become available with improved
temporal resolution, providing new opportunities for ecological
insights to be incorporated into EO analysis and monitoring.
In order to address the widespread demand of detecting VHR
changes in biodiversity state within and outside protected areas,
agreements need to be established between space agencies and
private companies (such as DigitalGlobeTM , which provides the
largest sub-metre constellation of satellites, including QuickBird,
WorldView-1, WorldView-2, IKONOS and GeoEye-1) to enable the
regular acquisition of multi-seasonal images. Such agreements, as
produced in the past within the European Framework Programme
(e.g. DataWarehouse) etc., can greatly advance the provision of
ecologically annotated EO monitoring data on changes in state to
131
provide support to national or regional conservation authorities
and research bodies.
Selection of site-specific appropriate EO images and processing
techniques by EO analysts should be based on prior, expert
site-specific knowledge from ecologists. This should include infor-
mation on plant physiology and phenology, vegetation structure,
spatial relationships and temporal dynamics, and habitat and
landscape level disturbance regimes that are of use for discrim-
inating land cover/habitat classes with similar signatures. This
is the main recommendation of a joint white paper by FP-7
SPACE projects BIO SOS (http://www.biosos.eu/) and MS.MONINA
(http://www.ms-monina.eu/).
Based on expert judgement, the approach suggested and docu-
mented in this paper (elaborated in the case studies) is applicable
to selected protected areas in diverse biogeographic regions and
with a diversity of land cover/habitat types. This approach con-
nects to the growing data sets from remote sources and pattern
recognition techniques for their automatic interpretation. We urge
it to be tested and applied over a wide range of pressures and
social–ecological settings.
Acknowledgments
This research draws on the European Union’s Seventh Frame-
work Programme (EU-FP7) project titled Biodiversity Multi-SOurce
Monitoring System: From Space To Species (BIO SOS, GA 263435)
coordinated by CNR ISSIA, Bari, Italy (http://www.biosos.eu).
We thank the anonymous reviewers of earlier versions of this
manuscript for their very useful feedback.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jag.2014.10.010.
References
Adamo, M., Tarantino, C., Tomaselli, V., Kosmidou, V., Petrou, Z., Manakos, I., Lucas,
R.M., Mücher, C.A., Veronico, G., Marangi, C., De Pasquale, V., Blonda, P., 2014.
Expert knowledge for translating land cover/use maps to general habitat cate-
gories (GHC). Landsc. Ecol., http://dx.doi.org/10.1007/s10980-014-0028-9.
Agarwal, S., Vailshery, L.S., Jaganmohan, M., Nagendra, H., 2013. Mapping urban tree
species using very high resolution satellite imagery: comparing pixel-based and
object-based approaches. ISPRS Int. J. Geo-Inf. 2, 220–236.
Allard, A., 2003. Detection of vegetation degradation on Swedish mountainous
heaths at an early stage by image interpretation. Ambio 32, 510–519.
Buchanan, G.M., Butchart, S.H.M., Dutson, G., Pilgrim, J.D., Steininger, M.K., Bishop,
K.D., Mayaux, P., 2008. Using remote sensing to inform conservation status
assessment: estimates of recent deforestation rates on New Britain and the
impacts upon endemic birds. Biol. Conserv. 141, 56–66.
Bruzzone, L., Bovolo, F., 2013. A novel framework for the design of change-detection
systems for very-high-resolution remote sensing images. Proc. IEEE 101 (3),
609–630.
Bunce, R.G.H., Bogers, M.M.B., Evans, D., Halada, L., Jongman, R.H.G., Mücher, C.A.,
Bauch, B., de Blust, G., Parr, T.W., Olsvig-Whittaker, L., 2013. The significance of
habitats as indicators of biodiversity and their links to species. Ecol. Indic. 33,
19–25.
Chen, G., Hay, G.J., Carvalho, L.M.T., Wulder, M.A., 2012. Object-based change detec-
tion. Int. J. Remote Sens. 33, 4434–4457.
Costanza, J.K., Moody, A., Peet, R.K., 2011. Multi-scale environmental heterogeneity
as a predictor of plant species richness. Landsc. Ecol. 26, 851–864.
Di Gregorio, A., Jansen, L.J.M., 2005. Land Cover Classification System (LCCS): Clas-
sification Concepts and User Manual. Food and Agriculture Organization of the
United Nations, Rome.
EEA, 1995. Europe’s Environment: The Dobris Assessment. European Environment
Agency, Copenhagen.
Everitt, J.H., Yang, C., Deloach Jr., C.J., 2005. Remote sensing of giant reed with Quick-
Bird satellite imagery. J. Aquat. Plant Manage. 43, 81–85.
Fuller, D.O., 2007. Tropical forest monitoring and remote sensing: a new era of
transparency in forest governance? Singap. J. Trop. Geogr. 27, 15–29.
Graf, R.F., Mathys, L., Bollmann, K., 2009. Habitat assessment for forest dwelling
species using LiDAR remote sensing: capercaillie in the Alps. For. Ecol. Manage.
257, 160–167.
132 H. Nagendra et al. / International Journal of Applied Earth Observation and Geoinformation 37 (2015) 124–132
Hall, K., Reitalu, T., Sykes, M.T., Prentice, H.C., 2011. Spatial heterogeneity of Quick-
Bird satellite data is related to fine-scale plant species spatial turnover in
semi-natural grasslands. Appl. Veg. Sci. 15, 145–157.
He, K.S., Rocchini, D., Neteler, M., Nagendra, H., 2011. Benefits of hyper-
spectral remote sensing for tracking plant invasions. Divers. Distrib. 17,
381–392.
Hockings, M., Stolton, S., Leverington, F., Dudley, N., Courrau, J., 2006. Evaluating
Effectiveness. A Framework for Assessing Management Effectiveness of Pro-
tected Areas, second ed. IUCN, Gland, Switzerland/Cambridge, UK.
Honrado, J.P., Viera, C., Soares, C., Monteiro, M.B., Marcos, B., Pereira, H.M., Partidário,
M.R., 2013. Can we infer about ecosystem services from EIA and SEA practice?
A framework for analysis and examples from Portugal. Environ. Impact Assess.
40, 14–24.
Hooper, D.U., Adair, E.C., Cardinale, B.J., Byrnes, J.E.K., Hungate, B.A., Matulich, K.L.,
Gonzalez, A., Duffy, J.E., Gamfeldt, L., O’Connor, M.I., 2012. A global synthesis
reveals biodiversity loss as a major driver of ecosystem change. Nature 486,
105–108.
Jenkins, C.N., Joppa, L., 2009. Expansion of the global terrestrial protected area sys-
tem. Biol. Conserv. 142, 2166–2174.
Joseph, S., Murthy, M.S.R., Thomas, A.P., 2011. The progress on remote sensing tech-
nology in identifying tropical forest degradation: a synthesis of the present
knowledge and future perspectives. Environ. Earth Sci. 64, 731–741.
Koch, B., Ivits, E., 2004. Results from the Project BIOASSESS – relation between
remote sensing and terrestrial derived biodiversity indicators. In: Monitoring
and Indicators of Forest Biodiversity in Europe–From Ideas to Operationality EFI
Proceedings No. 51, pp. 315–332.
Mace, G., Norris, K., Fitter, A.H., 2012. Biodiversity and ecosystem services: a multi-
layered relationship. Trends Ecol. Evol. 27, 19–26.
Mairota, P., Leronni, V., Cafarelli, B., Baveco, J.M., 2011. Connectivity in a real frag-
mented landscape: distance vs. movement model based approaches. In: Spatial
Data Methods for Environmental and Ecological Processes, second ed. CDP
Service edizioni, Foggia, pp. 145–148, ISBN: 9788896025123.
Mairota, P., Cafarelli, B., Boccaccio, L., Leronni, V., Labadessa, R., Kosmidou, V., Nagen-
dra, H., 2013. Using landscape structure to develop quantitative baselines for
protected area monitoring. Ecol. Indic. 33, 82–95.
Mairota, P., Cafarelli, B., Labadessa, R., Lovergine, F.P., Tarantino, C., Nagendra, H.,
Didham, R.K., 2014a. Very high resolution Earth observation features for testing
the direct and indirect effects of landscape structure on local habitat quality. Int.
J. Appl. Earth Observ. Geoinf., http://dx.doi.org/10.1016/j.jag.2014.07.003
Mairota, P., Leronni, V., Xi, W., Mladenoff, D., Nagendra, H., 2014b. Using spatial sim-
ulations of habitat modification for adaptive management of protected areas:
Mediterranean grassland modification by woody plant encroachment. Environ.
Conserv. 41, 144–156.
Mücher, C.A., Kooistra, L., Vermeulen, M., Vanden Borre, J., Haest, B., Haveman, R.,
2013. Quantifying structure and function of Natura 2000 heathland habitats
using spectral mixture analysis and segmentation techniques on hyperspectral
imagery. Ecol. Indic. 33, 71–81.
Nagendra, H., 2001. Using remote sensing to assess biodiversity. Int. J. Remote Sens.
22, 2377–2400.
Nagendra, H., Rocchini, D., 2008. High resolution satellite imagery for tropi-
cal biodiversity studies: the devil is in the detail. Biodivers. Conserv. 17,
3431–3442.
Nagendra, H., 2008. Do parks work? Impact of protected areas on land cover clearing.
Ambio 37, 330–337.
Nagendra, H., Rocchini, D., Ghate, R., 2010. Beyond parks as monoliths: spatially
differentiating park-people relationships in the Tadoba Andhari Tiger Reserve
in India. Biol. Conserv. 143, 2900–2908.
Nagendra, H., Lucas, R., Honrado, J.P., Jongman, R.H.G., Tarantino, C., Adamo, M.,
Mairota, P., 2013. Remote sensing for conservation monitoring: assessing pro-
tected areas, habitat extent, habitat condition, species diversity and threats. Ecol.
Indic. 33, 45–59.
Nagendra, H., Mairota, P., Blonda, P., Marangi, C., Torri, D., Lucas, R., Dimopolous,
P., Honrado, J.P., Niphadkar, M., Mücher, S., Tomaselli, V., 2012. Methodology
to Identify and Quantify Local Pressures. Biodiversity Multisource Monitoring
System: From Space TO Species (BIO SOS) Deliverable D6.8.
Padoa-Schioppa, E., Ficetola, F.G., Bonardi, A., Mücher, C.A., Nipadkar, M., Nagendra,
H., Leronni, V., Mairota, P., Tarantino, C., Adamo, M., 2012. Report on Ecological
Niche Modeling Biodiversity Multisource Monitoring System: From Space TO
Species (BIO SOS) Deliverable D6.7.
Pettorelli, N., Laurance, W.F., O’Brien, T., Wegmann, M., Nagendra, H., Turner, W.,
2014. Satellite remote sensing for applied ecologists: opportunities and chal-
lenges. J. Appl. Ecol. 51, 839–848.
Pirotti, F., 2011. Analysis of full-waveform LiDAR data for forestry applications: a
review of investigations and methods. iForest Biogeosci. For. 4, 100–106.
Pôças, I., Cunha, M., Pereira, L.S., 2011. Remote sensing based indicators of changes
in a mountain rural landscape of Northeast Portugal. Appl. Geogr. 31, 871–880.
Salafsky, N., Salzer, D., Stattersfield, A.J., Hilton-Taylor, C., Neugarten, R., Butchart,
S.H.M., Collen, B., Cox, N., Master, L.L., O’Connor, S., Wilkie, D., 2008. A standard
lexicon for biodiversity: unified classifications of threats and actions. Conserv.
Biol. 22, 897–911.
Shao, G., Wu, J., 2008. On the accuracy of landscape pattern analysis using remote
sensing data. Landsc. Ecol. 23, 505–511.
St-Louis, V., Pidgeon, A.M., Radeloff, V.C., Hawbaker, T.J., Clayton, M.K., 2006. High-
resolution image texture as a predictor of bird species richness. Remote Sens.
Environ. 105, 299–312.
Tarantino, C., Blonda, P., Pasquariello, G., 2007. Remote sensed data for automatic
detection of land-use changes due to human activity in support to landslide
studies. Nat. Hazards 41, 245–267.
Tomaselli, V., Dimopoulos, P., Marangi, C., Kallimanis, A.S., Adamo, M., Tarantino, C.,
Panitsa, M., Terzi, M., Veronico, G., Lovergine, F., Nagendra, H., Lucas, R., Mairota,
P., Mücher, S., Blonda, P., 2013. Translating land cover/land use classifications
to habitat taxonomies for landscape monitoring: a Mediterranean assessment.
Landsc. Ecol. 28, 905–930.
Torres, J., Brito, J.C., Vasconcelos, M.J., Catarino, L., Gonçalves, J., Honrado, J., 2010.
Ensemble models of habitat suitability relate chimpanzee (Pan troglodytes) con-
servation to forest and landscape dynamics in Western Africa. Biol. Conserv. 143,
416–425.
Turner, W., Spector, S., Gardiner, N., Fladeland, M., Sterling, E., Steininger, M., 2003.
Remote sensing for biodiversity science and conservation. Trends Ecol. Evol. 18,
306–314.
Vicente, J.R., Fernandes, R.F., Randin, C.F., Broennimann, O., Gonçalves, J., Marcos,
B., Pôças, I., Alves, P., Guisan, A., Honrado, J.P., 2013. Will climate change drive
alien invasive plants into areas of high conservation value? An improved model-
based regional assessment to prioritize the management of invasions. J. Environ.
Manage. 131, 185–195.
Vierling, K.T., Vierling, L.A., Gould, W.A., Martinuzzi, S., Clawges, R.M., 2008. Lidar:
shedding new light on habitat characterization and modeling. Front. Ecol. Envi-
ron. 6, 90–98.