R

Hr

OD

a

ARRAA

KELMMSS

C

h1

Ecological Indicators 73 (2017) 492–504

Contents lists available at ScienceDirect

Ecological Indicators

jo ur nal ho me page: www.elsev ier .com/ locate / ecol ind

eview

ow to analyse ecosystem services in landscapes—A systematiceview

skar Englund ∗, Göran Berndes, Christel Cederbergiv. of Physical Resource Theory, Chalmers University of Technology, Gothenburg, Sweden

r t i c l e i n f o

rticle history:eceived 29 July 2016eceived in revised form 6 October 2016ccepted 9 October 2016vailable online 24 October 2016

eywords:cosystem servicesandscape

a b s t r a c t

Ecosystem services (ES) is a significant research topic with diverse modelling and mapping approaches.However, the variety of approaches—along with an inconsistent terminology—cause uncertainties con-cerning the choice of methods. This paper identifies and qualitatively assesses methods for mapping ESin terrestrial landscapes, based on a systematic review of the scientific literature. It further aims to clarifythe associated terminology, in particular the concept of landscape and landscape scale. In total, 347 casesof ES mapping were identified in the reviewed papers. Regulating and maintenance services were mostcommonly mapped (165), followed by cultural (85), and provisioning services (73). For individual ES, alarge variation in number of mapping cases was found. This variation may either reflect the perceived

appingethods

patial analysisystematic review

importance of the ES, or that different ES can be more or less easily mapped. Overall, Logical models andEmpirical models were most commonly used, followed by Extrapolation, Simulation/Process models, Dataintegration, and Direct mapping. Only twelve percent of all ES mapping cases were validated with empir-ical data. The review revealed highly diverging views on the spatial extent of landscapes in studies of ES,and that the term landscape is sometimes used rather arbitrarily.

© 2016 Elsevier Ltd. All rights reserved.

ontents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4932. Material and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493

2.1. Meta review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4932.2. Systematic review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493

2.2.1. Assessment framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4933. Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494

3.1. Mapping of ecosystem services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4943.2. Typology and terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4963.3. The concept of landscape and landscape scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496

3.3.1. Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4963.3.2. Landscape services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498

4. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4994.1. “Landscapes” in ES literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4994.2. Methods for mapping ecosystem services at a landscape scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499

4.2.1. Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5014.3. Further reflections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502

5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author.E-mail address: oskar.englund@chalmers.se (O. Englund).

ttp://dx.doi.org/10.1016/j.ecolind.2016.10.009470-160X/© 2016 Elsevier Ltd. All rights reserved.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503

al Indi

1

mwtttwfaitMgnms

is2Epeuuatcmmaafcsb

motst

2

2

taawStT

2

efta

O. Englund et al. / Ecologic

. Introduction

Ecosystems provide various goods (e.g. food and constructionaterial) and services (e.g. regulation of water flows) to society,hich contribute to our survival and well-being. Such “ecosys-

em services” (ES) (Daily, 1997; MEA, 2005) have been evidento humans throughout history, but not explicitly considered untilhe late 1960s and 1970s (Hermann et al., 2011; Portman 2013),hen scientists began to address the societal value of nature’s

unctions (King 1966; Helliwell 1969; Dee et al., 1973; Bormannnd Likens 1979). The term “ecosystem services” was introducedn 1981 (Ehrlich and Ehrlich, 1981) and, following important con-ributions by, e.g., Daily (1997) and Costanza et al. (1997), the

illennium Ecosystem Assessment (MEA) (MEA, 2003) broughtlobal attention to its importance. Today, the concept ES is recog-ised in policy and it is a significant research topic with diverseodelling and mapping approaches supporting studies at different

patial and temporal scales (Burkhard et al., 2013).Mapping—the organization of spatially explicit quantitative

nformation—is essential for many assessments of ES since bothupply and demand can be spatially explicit (Troy and Wilson,006). Mapping can allow full assessment and quantification ofS (Crossman et al., 2013), including the spatial distance betweenroviding areas and benefiting areas (Fisher et al., 2009; Bastiant al., 2012). Crossman et al. (2013) argue the need for betternderstanding of where ES are supplied so that “stocks of nat-ral capital and the flow of ES can be monitored and managedcross spatial and temporal scales”. They also point out that spa-ially explicit understanding of conditions of, and threats to, naturalapital, will facilitate that resources are allocated to where they areost needed. The usefulness of maps (i.e., spatial products fromapping) to support governance and management of ecosystems

nd their services is noted by Hauck et al. (2013). Many methodsnd tools exist to map and quantify ES, applicable for highly dif-ering scales. This, along with inconsistencies in the terminology,reates uncertainties concerning the choice of methods. The incon-istent terminology can even cause uncertainty in what is actuallyeing mapped (Crossman et al., 2013).

This paper identifies and qualitatively assesses methods forapping ES in terrestrial landscapes, based on a systematic review

f the scientific literature. It further aims to clarify the associatederminology, in particular the concept of landscape and landscapecale, based on a meta-review of recent literature and outcomes ofhe systematic review.

. Material and methods

.1. Meta review

In order to clarify the terminology used in studies of ES, in par-icular the concept of landscape and landscape scale, and to developn assessment framework for the systematic review of methods,

meta-review of recent literature was performed. Review papersere identified from keyword searches in the Scopus and Web of

cience databases. Additional papers were identified by examininghe bibliographies in the review papers and papers that cite these.he outcome of this review is presented in Section 3 (Theory).

.2. Systematic review

Papers included in two previous review publications (Crossman

t al., 2013; Andrew et al., 2015) were reviewed on methods usedor mapping ES at a landscape scale. These two review publica-ions also cover papers previously reviewed by Egoh et al. (2012)nd Martinez-Harms and Balvanera (2012). An additional literature

cators 73 (2017) 492–504 493

search was carried out to identify relevant papers published after2012. The full literature selection process is described in Table 1.The outcome of this review is presented in Section 4 (Results andDiscussion).

The 1112 papers identified in the literature search werescreened to determine if they met two relevance criteria: (1) spa-tially explicit results (i.e., maps) presented for at least one ES; and(2) study stated to be made at a landscape scale, for the purpose oflandscape planning, or referring to a study area as a landscape or ascontaining landscapes. A total of 170 papers fulfilled these criteria.

2.2.1. Assessment framework2.2.1.1. General information. The 170 papers were reviewed on:(a) their targeted scale: global, continental, international, national,or sub-national; (b) the country/countries in which the study wasperformed; and (c) the year the paper was published.

2.2.1.2. References to landscapes. The papers were then reviewedon whether or not specific areas (the study area or any other area)were referred to as landscapes. The size of such areas was noted tofacilitate a discussion of the spatial extent of landscapes.

2.2.1.3. Limitations in resolution and method. The papers were thenreviewed on the resolution at which the spatial results were pre-sented. Papers using a resolution of approximately 1 km2 or coarserwere not further reviewed. In addition, papers mapping only themonetary value of ES using value transfer, i.e. assigning monetaryvalues to areas without prior quantification of biophysical or otherestimates to support the monetization, were also excluded fromfurther review. A total of 49 papers were eliminated in this step.

2.2.1.4. Ecosystem services studied. The remaining 121 papers werethen reviewed on which ES that were mapped. A modified versionof CICES v4.3 classification system was used (see Fig. 4, cf. Table 2).In case several ES that fall under the same ES category were studiedseparately (e.g., biomass for food and energy, respectively) usingthe same or similar methods, they were counted as one ES. A totalof 347 cases of ES mapping was identified.

2.2.1.5. Mapping methods used. [•]• The papers were then reviewed on the type of method that was

used to map each ES. A categorisation system similar to Andrewet al. (2015) was used, as follows:

• Direct mapping refers to methods where survey and censusapproaches provide complete spatial information of the distri-bution of an ES.

• Empirical models refer to models based on point-based mea-surements of ES. Values are then explained and consequentlyestimated elsewhere using, e.g., regression analysis.

• Simulation and process models attempt to simulate or modelecosystem processes to identify [changes in] ES values [givenchanges in ecosystem properties]. Such models require no mea-surements of ES except possibly for calibration and validation.

• Logical models map ES based on a set of indicators using decisionrules.

• Extrapolation methods parameterize ecosystem properties (oftenland-cover classes) for their level of ES supply, based on aspatialsummary values.

• Data integration methods synthesize pre-existing spatial prod-ucts to generate ES maps, often with rule-based approaches.The first four types roughly constitute ecological production

function methods, i.e., estimating the level of ES provisioning at aparticular location given the biotic and abiotic characteristics ofthat site. The latter two roughly constitute benefit transfer methods,i.e., estimating the value of ES provisioning in one context by adapt-

494 O. Englund et al. / Ecological Indicators 73 (2017) 492–504

Table 1Literature selection process for systematic review.

Source Number of papers Cumulative number of papers

Papers from Crossman et al. (2013) 113a 113Papers from Andrew et al. (2015) 144 257Additional search in Scopusb 757 1014Additional search in Web of Sciencec 687 1701Removing duplicates −589 1112Title, abstract and full text screening −942 170Included in review 170

a Total number of references in the appended list of reviewed papers.b TITLE-ABS-KEY(“ecosystem service*OR ‘landscape service*OR “ecosystem function*OR “ecosystem process*) AND TITLE-ABS-KEY(“Geospatial” OR ‘Geographic informa-

t ndscaposyst

O 4 OR

ic

s

••

t

2nw

3

3

mSoaasvteis

italCbrbFswptft

oda

ion system’ OR “GIS’ OR ‘map’ OR “spatial” OR “indicator*) AND TITLE-ABS-KEY(‘lac TS = (“ecosystem service*OR ‘landscape service*OR “ecosystem function*OR “ecR ‘map’ OR “spatial” OR “indicator*) AND TS = (‘landscape’) AND PY = (2013 OR 201

ng an estimate of the value of the same ES provisioning in anotherontext (Andrew et al., 2015).

Two additional method types were added to the categorisationystem:

[•] Combination, in case different types of methods were used. Unknown, in case the method was not sufficiently well describedto determine the type.In addition to noting the method type, each method was quali-

atively summarised.

.2.1.6. Validated results. The papers were reviewed on whether orot they attempted to validated their results, how it in that caseas done, and with what outcome.

. Theory

.1. Mapping of ecosystem services

Several recent reviews have been published on methods forapping and quantifying ES. In 2015, a special issue in Ecosystem

ervices (vol. 13) was dedicated to the subject, including reviewsn ES mapping across scales and continents (Malinga et al., 2015)nd empirical (public) participatory GIS mapping of ES (Brownnd Fagerholm, 2015). In the editorial, Willemen et al. (2015)ummarised “best practices for mapping ecosystem services”. Pre-iously, Martinez-Harms and Balvanera (2012) reviewed 70 papershat mapped the supply of ES based on social-ecological data. Egoht al. (2012) reviewed 67 papers that mapped and/or modelled ESn order to identify spatial indicators used and their available dataources.

Crossman et al. (2013) revisited the papers in the two preced-ng reviews and reviewed additional papers not included in thesewo reviews. They found that (following the typology used by theuthors) regulating services have been most often mapped, fol-owed by provisioning, cultural, and supporting/habitat services.limate regulation was the most commonly mapped ES, followedy recreation and tourism, food provision, provision of water, andegulation of water flows. These results are in line with resultsy Martinez-Harms and Balvanera (2012) and Egoh et al. (2012).urthermore, most assessments were carried out on a regional, orub-national, scale, although provisioning and regulating servicesere more commonly mapped at larger scales compared with sup-orting and cultural services. Secondary data were used more oftenhan primary data, especially for regulating services, and the mostrequently applied indicators were land use/-cover, soils, vegeta-ion, and indicators related to nutrients.

Andrew et al. (2015) reviewed nearly 150 studies in order toutline socioecological drivers of ES and identify representativeatasets; categorize models used to map ES; and discuss the efficacynd credibility of ES assessments. They suggest that there are three

e’) AND PUBYEAR > 2012.em process*) AND TS = (“Geospatial” OR ‘Geographic information system’ OR “GIS’2015).

primary paradigms for mapping the provisioning of ES—ecologicalproduction functions, ES providers, and functional traits—and thatthese provide a strong scientific framework. They however expressconcerns of the widespread use of non-validated proxy-basedmethods.

Nemec and Raudsepp-Hearne (2013) reviewed the differentways that GIS has been used to estimate ES and their values,including: (1) static estimates, i.e., data-derived spatial estimates ofpresent or past ES distributions; (2) models that can be used to anal-yse how changes in landscapes impact ES; and (3) approaches thatemphasize social preferences and priority-setting for ES manage-ment. They claim that it is difficult to generalize about which toolsprovide the most credible results and that the choice of tool shoulddepend on context. They also stress that caution should be appliedto the use and interpretation of models developed for broad use,and that practitioners with advanced GIS skills may benefit fromcreating their own models rather than using an existing platform.

Andrew et al. (2014) highlight the ways that remote sensingcan contribute to information needs in ES assessments and presentnovel methods for assessing ecosystem processes and servicesusing remote sensing. In particular, they state that remote sens-ing can provide spatially nuanced depictions of plant functionaltraits and soil properties, and that it can monitor aspects of criticalbiogeochemical processes. Finally, “urban ES”, i.e., ES in an urbanor peri-urban environment, have been reviewed by, e.g., Mao et al.(2015).

A cause for concern, and an indication of that many methodsmay be unsuitable for landscape scale studies, is that a majorityof ES assessment studies so far use proxy methods (Egoh et al.,2012), such as benefits transfer (e.g. Costanza et al. (1997)). Suchmethods are much less complex than for example direct map-ping with survey and census approaches or empirical productionfunction models, and may thus be an appealing approach for ESassessments. However, there are several disadvantages with proxybased methods, such as the risk for generalization error, whichmakes them prone to error (Eigenbrod et al., 2010a; Stephens et al.,2015). Since landscapes are typically not mere combinations ofecosystems, but shaped by the interactions between ecosystemstructures/processes and humans (Council of Europe, 2000), theuse of proxies at the landscape level is particularly sensitive tolocal conditions. Careful calibration and validation is thus necessary(Stephens et al., 2015), but this has typically not been done (Seppeltet al., 2011; Martinez-Harms and Balvanera, 2012). Proxies may besuitable for identifying broad-scale trends in ES, or for global leveland rapid assessments, but they are likely to be unsuitable for iden-tifying hotspots or priority areas for multiple ES (Eigenbrod et al.,2010b; Hermann et al., 2014). Additional data beyond land cover

observation are therefore often necessary for a proper assessmentof ecosystem functions or services, especially at the landscape scale(Hermann et al., 2011; Nemec and Raudsepp-Hearne, 2013).

O. Englund et al. / Ecological Indicators 73 (2017) 492–504 495

Table 2Comparison of CICES v4.3, MEA, TEEB, and Costanza et al., 1997 classification systems: Provisioning and Regulating and maintenance services. The MEA Supporting ServicesPhotosynthesis and Primary production are not included in the table.

CICES v4.3 (EEA 2013) TEEB (2010a,b) MEA (2003) Costanza et al.(1997)

Provisioning Nutrition Biomass Cultivated crops • Food • Food • Food productionReared animals and their outputsWild plants, algae and theiroutputsWild animals and their outputsPlants and algae from in-situaquacultureAnimals from in-situ aquaculture

Water Surface water for drinking • Fresh water • Fresh water • Water supplyGround water for drinking

Materials Biomass Fibres and other materials fromplants, algae and animals for directuse or processing

• Raw materials• Medicinal

resources

• Fibre• Biochemicals,

naturalmedicines, andpharmaceuticals

• Ornamentalresources

• Raw materials

Materials from plants, algae andanimals for agricultural useGenetic materials from all biota • Maintenance of

genetic diversity• Genetic resources • Genetic resources

Water Surface water for non-drinkingpurposes

• Fresh water • Fresh water • Water supply

Ground water for non-drinkingpurposes

Energy Biomass-basedenergy sources

Plant-based resources • Raw materials • Fuel • Raw materials

Animal-based resourcesMechanical energy Animal-based energy – – –

Regulating andmaintenance

Mediation ofwaste, toxics andother nuisances

Mediation by biota Bio-remediation bymicro-organisms, algae, plants, andanimals

• Waste-watertreatment

• Local climateand air quality

• Waterpurification andwaste treatment

• Air qualityregulation

• Waste treatment• Gas regulation

Filtration/sequestration/storage/accumulationby micro-organisms, algae, plants,and animals

Mediation byecosystems

Filtration/sequestration/storage/accumulationby ecosystemsDilution by atmosphere,freshwater and marine ecosystemsMediation of smell/noise/visualimpacts

– – –

Mediation of flows Mass flows Mass stabilization and control oferosion rates

• Erosionprevention

• Erosionregulation

• Erosion controland sedimentretention

Buffering and attenuation of massflows

–

Liquid flows Hydrological cycle and water flowmaintenance

• Fresh water • Water regulation• Water cycling

• Water regulation

Flood protection • Moderation ofextreme events

• Natural hazardregulation

• Disturbanceregulation

Gaseous/air flows Storm protectionVentilation and transpiration – – –

Maintenance ofphysical, chemical,biologicalconditions

Lifecyclemaintenance,habitat and genepool protection

Pollination and seed dispersal • Pollination • Pollination • Pollination

Maintaining nursery populationsand habitats

• Habitats forspecies

– • Refugia

Pest and diseasecontrol

Pest control • Biological control • Pest regulation • Biological control

Disease control • Diseaseregulation

–

Soil formation andcomposition

Weathering processes • Maintenance ofsoil fertility

• Soil formation• Nutrient cycling

• Soil formation• Nutrient cycling

496 O. Englund et al. / Ecological Indicators 73 (2017) 492–504

Table 2 (Continued)

CICES v4.3 (EEA 2013) TEEB (2010a,b) MEA (2003) Costanza et al.(1997)

Decomposition and fixingprocesses

Water conditions Chemical condition of freshwaters • Habitats forspecies

– –

Chemical condition of salt waters – –Atmosphericcomposition andclimate regulation

Global climate regulation byreduction of greenhouse gasconcentrations

• Carbonsequestration andstorage

• Climateregulation

• Climateregulation

• Gas regulation

Micro and regional climateregulation

• Local climate andair quality

• Air qualityregulation

t2Clpaidp

3

edTfgapmdces“isEi2cEEpacP(T

cseiiT2

Many have argued in favour of a standardized approach to quan-ify and map ES (TEEB, 2010a; Eppink et al., 2012; Maes et al.,012; Martinez-Harms and Balvanera, 2012; Seppelt et al., 2012;rossman et al., 2013). This has however been difficult to estab-

ish (Nemec and Raudsepp-Hearne, 2013). Crossman et al. (2013)roposed instead a blueprint for documenting studies, to provide

template and checklist concerning information needed for carry-ng out a modelling and mapping study, and to provide a growingatabase of completed “blueprints” to the benefit of researchers,ractitioners and policymakers.

.2. Typology and terminology

Several ES classification systems have been proposed (Costanzat al., 1997; Daily, 1997, 1999; De Groot et al., 2002; MEA, 2003;e Groot 2006; Boyd and Banzhaf, 2007; Fisher and Turner, 2008;EEB, 2010a). Costanza (2008) states that there are many use-ul ways to classify ecosystem goods and services, and that theoal should not be to have a single, consistent system, but rather

pluralism of typologies that can be useful for different pur-oses. A drawback is that the use of multiple classification systemsakes comparisons and integration of assessments with other

ata difficult (Haines-Young and Potschin, 2011). Among the moreommonly used classification systems, Costanza et al. (1997) cat-gorized ES into 17 major groups while MEA (2003) categorized aimilar set of ES as “provisioning”, “regulating”, “supporting” andcultural”. The Economics of Ecosystems and Biodiversity (TEEB)nitiative (2010a) adopted a classification based on MEA, but con-idered “supporting services” to be ecological processes rather thanS, and emphasized the importance of “habitat services” by assign-ng it a separate category. Initially, the TEEB classification contained2 ES (TEEB, 2010a) but was reduced to 17 ES (TEEB, 2010b). A newlassification system, the Common International Classification ofcosystem Services (CICES), is being developed by the Europeannvironment Agency (www.cices.eu). The aim of CICES is to pro-ose a universal classification of ES that is both consistent withccepted categorizations and allows easy translation of statisti-al information between different applications (Haines-Young andotschin, 2011). A comparison of CICES v. 4.3 (EEA 2013) with TEEB2010a), MEA (2003), and Costanza et al. (1997), is provided inable 2 and 3.

In addition to—possibly to some extent due to—inconsistentlassification, the terminology in ES research has remained incon-istent. For example, studies that use the MEA typology (Andrewt al., 2015) include supporting services. The same “services” are

n other studies (Crossman et al., 2013) considered to be ecolog-cal (or ecosystem) processes, following, e.g., the TEEB typology.hese are also sometimes referred to as intermediate ES (Boyd007; Crossman et al., 2013). Furthermore, some (Wallace 2007;

• Climateregulation

Andrew et al., 2015) consider ecosystem functions to be synony-mous with ecosystem processes, while others (Crossman et al.,2013; Potschin et al., 2014) do not. It should be noted that,although many use terms highly arbitrarily, inconsistency in thiscase reflects an ongoing scientific discourse. It has been arguedthat definitions of ES are purpose-dependent and should be judgedon their usefulness for a particular purpose (Zhang et al., 2007;Lamarque et al., 2011). However—as noted also for classificationsystems above—co-existence of different terminologies and defini-tions could impede on-the-ground use of the concept (Lamarqueet al., 2011). Diversity is important for advancing science andknowledge, but can create difficulties in situations where gover-nance agreements are to be made—particularly where multiplegoals need to be considered. At present, work is in progress to estab-lish working definitions of commonly used terms (Potschin et al.,2014). This may, along with the advancement of the CICES classifi-cation, help to harmonize the terminology and make studies moreconsistent and comparable. Definitions of commonly used terms,which are also used in this paper, are presented in Table 4.

3.3. The concept of landscape and landscape scale

Academic landscape research has primarily been conductedwithin the fields of ecology/landscape ecology, geography, and spa-tial planning (Conrad et al., 2011). Landscape ecology was for along time focused on the relationships between spatial patternsand ecological processes (Wiens and Moss, 2005), not acknowledg-ing people as part of the landscape, but rather a cause of landscapechange (Termorshuizen and Opdam, 2009). In the field of spatialplanning, however, a central notion was that people are a partof the landscape, and that landscape change should benefit peo-ple (Termorshuizen and Opdam, 2009). Research in the fields oflandscape ecology and spatial planning have therefore been basedon different views on what a landscape is, including its value tohumans. As stated by Higgins et al. (2012): “as a theoretical tool, theconcept is steeped in ambiguity and complexity, with innumerablenuances accompanying its multiple interpretations”.

3.3.1. DefinitionsSimilar to the views in the field of spatial planning, the Euro-

pean Landscape Convention (ELC) defined landscape as ‘an area, asperceived by people, whose character is the result of the action andinteraction of natural and/or human factors’ (Council of Europe,2000). As a result of the ELC, as well as the Convention for theSafeguarding of the Intangible Cultural Heritage (UNESCO, 2003)

and the Framework Convention on the Value of Cultural Heritagefor Society (Council of Europe, 2005), landscape received a higherstatus in spatial planning (Herlin, 2007) since it became formallyrecognised and highlighted as central to matters of sustainability

O. Englund et al. / Ecological Indicators 73 (2017) 492–504 497

Table 3Comparison of CICES v4.3, MEA, TEEB, and Costanza et al., 1997 classification systems: Cultural services.

CICES v4.3 (EEA 2013) TEEB (2010a,b) MEA (2003) Costanza et al.(1997)

Cultural Physical andintellectualinteractions withbiota, ecosystems,andland-/seascapes[environmentalsettings]

Physical andexperientialinteractions

Experiential use ofplants, animals andland-/seascapes indifferentenvironmentalsettings

• Recreation andmental andphysical health

• Tourism

• Recreation andecotourism

• Recreation

Physical use ofland-/seascapes indifferentenvironmentalsettings

Intellectual andrepresentativeinteractions

Scientific • Aestheticappreciation andinspiration forculture, art anddesign

• Knowledgesystems,

• Educationalvalues

• Culturaldiversity

• Cultural

EducationalHeritage, cultural • Cultural

diversity• Social relations• Cultural heritage

values

EntertainmentAesthetic • Aesthetic values

Spiritual, symbolicand otherinteractions withbiota, ecosystems,andland-/seascapes[environmentalsettings]

Spiritual and/oremblematic

Symbolic • Spiritualexperience andsense of place

• Spiritual andreligious values

• Sense of place

Sacred and/orreligious

ce

t

asbmsitgemospm

liegtgoct(

2014). With this view, landscape scale is referred to as havinga landscape—which in general should be larger than a field andsmaller than a physiographic region—as a study area. In addition,

Other culturaloutputs

Existen

Beques

nd the management of public spaces. This led to further discus-ions on the meaning of ‘landscape’ − not only what a landscape is,ut also what it does − in legislation, policy, planning, and manage-ent. Nassauer (2012) proposed two laws of landscape function, to

how how landscape can be used for synthesis science and ecolog-cal design: (1) ‘Different environmental processes operate in andhrough the same landscape, and each landscape inherently inte-rates these processes’; and (2) ‘Landscapes are visible in everydayxperience and can be made visible in spatial representations. Thisakes it possible for different people to have the same experience

f visible characteristics of a given landscape’. The latter empha-izes that because landscapes can be seen, different people canoint to characteristics that they notice and discuss the differenteanings those characteristics convey.Humans are today considered to be an integral part of the

andscape, and landscapes are thus often regarded a result ofnteractions between natural and human processes (Vallés-Planellst al., 2014). The definition of landscape in the recently proposed ESlossary by Potschin et al. (2014) places slightly less emphasis onhe interactions between natural and human processes: ‘A hetero-eneous mosaic of land cover, habitat patches, physical conditions orther spatially variable elements viewed at scales relevant to ecologi-

al, cultural-historical, social or economic considerations’. Accordingo this definition, a landscape can be an area of (widely) varying sizesee Fig. 1) that for various reasons is relevant to consider a coher-

ent unit. Besides recognizing that a landscape can have ecological,cultural, social, and economic importance, this indicates that land-scapes can have varying spatial extent. For example, an ecologicalperspective considers, e.g., topological features while social or cul-tural perspectives consider, e.g., social connections, administrativeunits, or tradition. It can be argued that landscapes need not beheterogeneous but can consist of largely homogeneous,1 land cover(Soto and Palomares, 2015), such as cropland or forest (DeVrieset al., 2015), provided that it is relevant from ecological, cultural-historical, social or economic point of view. Therefore, it may bemore appropriate to refer to landscapes as: ‘An area viewed at a scaledetermined by ecological, cultural-historical, social and/or economicconsiderations’.

Landscape scale has been defined as an intermediate integra-tion level between the field and the physiographic region (Turner,1989; Burel and Baudry, 2003), but with an extent depending onthe spatial range of the biophysical and anthropogenic processesdriving the processes (or services) under study (Lacoste et al.,

1 Note also that homogeneity, at a high level of abstraction, can be defined asgeneralized information of internal heterogeneity (Bastian et al., 2013).

498 O. Englund et al. / Ecological Indicators 73 (2017) 492–504

Table 4Definitions of commonly used terms, which are also used in this paper. Adapted from Potschin et al. (2014); Hermann et al. (2011); Andrew et al. (2015); Mastrangelo et al.(2014) and Bastian et al. (2014).

Term Definition

Ecosystem structure Static ecosystem characteristics: spatial and aspatial structure, composition and distributionof biophysical elementsExample: land use, standing crop, leaf area, % ground cover, species composition

Ecosystem processes Dynamic ecosystem characteristics: Complex interactions among biotic and abiotic elementsof ecosystems causing physical, chemical, or biological changes or reactions.Examples: decomposition, photosynthesis, nutrient cycling and energy fluxes.

Ecosystem functions The subset of processes and structures that, if benefiting to human well-being, provide ES. Canbe defined as the capacity of ecosystems to provide ES.Example: carbon sequestration

Ecosystem properties Refers collectively to ecosystem structure and processes.Ecosystem services Direct and indirect contributions of ecosystem functions to human well-being.

Example: climate regulationIntermediate ecosystemservice

Ecosystem functions that do not directly benefit to human well-being, but that support otherfunctions that do. Synonymous with ‘supporting services’

Ecosystem service providers The ecosystems, component populations, communities, functional groups, etc. as well asabiotic components such as habitat type, that are the main contributors to specific ES.Example: Forest tree communities are ES providers for global climate regulation.

Human well-being A state that is intrinsically or instrumentally valuable for a person or society.Example: The MEA (2005) classifies components (or drivers) of human well-being into: basicmaterial for a good life, freedom and choice, health and bodily wellbeing, good social relations,security, peace of mind, and spiritual experience.

Ecosystem service supply ES provisioned by a particular area over a given time period.Ecosystem service demand ES demanded in a particular area over a given time period.Ecosystem service providingunits/areas

Spatial units that are the source of ES. Commensurate with ecosystem service supply.

Ecosystem service benefitingareas

The complement to ES providing areas. ES benefiting areas may be far distant from respectiveproviding areas. Commensurate with ES demand.

Landscape An area viewed at a scale determined by ecological, cultural-historical, social and/or economicconsiderations

Landscape services The contributions of landscapes and landscape elements to human well-beingLandscape multifunctionality The capacity of a landscape to simultaneously support multiple benefits to society

F is spec e largo

BadmTov

ig. 1. Size of the 94 areas referred to as “landscape” in the reviewed papers. Sizeountries of an approximately equivalent size for the areas at the far right. Due to thutline. Hence, they appear similar in size.

astian et al. (2013) states that “landscape units can be aggregatedt various levels of abstraction, resulting in a number of sub-imensions within the so-called chorological dimension: nano-,

icro-, meso- and macro-chores (choros: Greek ‘space’, “land”)”.

his means it may even be possible to build a hierarchical systemf different landscape levels. Landscapes can therefore have highlyarying character and size, and two studies that use very differ-

cified using absolute numbers for the areas at the far left of the figure, and usinge differences, the smallest 15 areas would not be visible in this figure without their

ent kinds of study areas may still both claim to be performed at alandscape scale.

3.3.2. Landscape servicesTermorshuizen and Opdam (2009) claimed that landscapes can-

not be synonyms for ecosystems (e.g. de Groot (2006)), but ratheras spatial human-ecological systems that deliver functions valued

al Indicators 73 (2017) 492–504 499

btausssaiaes(

i2(vEtoei

2pwbmtcttbtstiswTtl

shima

4

4

i(Sst2iTt

0

5

10

15

20

25

30

35

40

45

1997 2000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Numberofpapers

individual ES or not covered by the CICES classification system,were mapped. This includes examples of “landscape services” (seeSection 3.3.2), where landscapes or specific landscape elements,

O. Englund et al. / Ecologic

y humans for economic, socio-cultural, and ecological reasons. Forhat reason, they claimed that landscape services would be a moreppropriate term, as it would enable the consideration of both nat-ral and cultural aspects, spatial patterns, and the involvement oftakeholders. The necessity of this concept can be disputed since (1)everal cultural values of landscapes are already regarded as ES; (2)patial patterns are essential to assessing, e.g., aesthetic and recre-tional values as well as the spatial configuration of habitats thatnvolve lateral flows of material or organisms (Andrew et al., 2014);nd (3) the involvement of stakeholders is already widely acknowl-dged as important for meaningful ES assessments on a national, ormaller, level—see, e.g., Koschke et al. (2012) and Fagerholm et al.2012).

Nevertheless, the concept has been used in several recent stud-es (Hermann et al., 2014; Ungaro et al., 2014; Vallés-Planells et al.,014; Hainz-Renetzeder et al., 2015), and Vallés-Planells et al.2014) even suggest a new classification system for landscape ser-ices that includes “services” not previously included in the majorS classification schemes. One advantage may be that the referenceo “landscapes” instead of “ecosystems” may attract the interestf other scientific disciplines than those usually concerned withcosystems and ecology, and thus contribute to interdisciplinaryn ES assessments (Hermann et al., 2011).

Besides using landscape services as a synonym (Willemen et al.,012) or an alternative (Vallés-Planells et al., 2014) to ES, it may beossible to use it as a complement. Bastian et al. (2014) examinedhether a stronger focus on landscape services would be useful

y examining case studies in Germany, and found no strong argu-ents for replacing the concept of ES with landscape services at

he landscape scale. They did, however, find that both conceptsan be useful depending on the situation. The decisive difference ishe object they refer to: ES to ecosystems, and landscape serviceso landscapes or landscape elements (which need not necessarilye ecosystems). They distinguish between three cases to illustratehis: (1) where the service is supplied by ecosystems and land-cape issues play no role, the term ES should be used; (2) wherehe service is supplied by ecosystems and also landscape issues aremportant, both terms may be used; and (3) where the service is notupplied by an ecosystem but by landscape elements, the term ESould be obsolete and instead landscape services should be used.

hey suggest a new definition of landscape services to illustratehis: “Landscape services are the contributions of landscapes andandscape elements to human well-being”.

Analogous to landscape services, the term landscape functions isometimes used to describe the capacity of a landscape to fulfiluman needs (de Groot, 2006) (cf. ecosystem functions), although

ts original definition (Krovakova et al., 2015) “the flows of energy,aterials and species among the component ecosystem” (Forman

nd Godron, 1986) is more biocentric than anthropocentric.

. Results and discussion

.1. “Landscapes” in ES literature

A total of 94 areas referred to as “landscape” were identifiedn the systematic review (Fig. 1). Their sizes range from 24 ha (ha)roughly 34 football pitches) to 122 million ha (roughly the size ofouth Africa). Lacoste et al. (2014) suggest that the extent of a land-cape can be considered to range from 100 to 10,000 ha. However, ofhe 94 areas referred to as landscapes in the reviewed papers, only

3 fall within that range. It is thus obvious that there are diverg-

ng views on the spatial extent of landscapes in the ES literature.he review revealed that the term is sometimes used rather arbi-rarily. To avoid this, it can be advisable to describe areas referred

Pub lica�on year

Fig. 2. Chronological distribution of reviewed papers (n = 170).

to as landscapes in a way that explains why they are consideredlandscapes (as in, e.g., Vihervaara et al. (2010)).

Given the diverging views on the spatial extent of a landscape,there are also diverging views on the meaning of landscape scale.The view that landscape scale is referred to as having a landscapeas a study area (see Section 3.3.1) is common in the ES litera-ture, although while some attempt to map ES across the landscape,others aggregate the ES under study to one value for the entirelandscape. A study area can also be described as containing several“landscapes”, for which ES values are aggregated. In such cases,some also refer to the entire study area as a landscape. Two studiesmay thus focus on the same area, refer to it as a landscape, but havewidely varying views on what is meant by landscape scale.

4.2. Methods for mapping ecosystem services at a landscape scale

A total of 170 papers were identified as having mapped ES ata landscape scale2 and 121 of these mapped ES at a relatively fineresolution across landscapes. The remaining papers mapped ES ata coarser resolution (approximately 1 km or higher), in monetaryterms only, or aggregated for administrative units. Almost half ofthe papers were published in 2015 and 2016, while only 14% ofthe papers were published before 2010 (Fig. 2). This is in line withobservations by, e.g, Andrew et al. (2015) and Martinez-Harms andBalvanera (2012) and confirms that ES research—also at the land-scape scale—is a relatively recent and rapidly growing area.

Most studies were carried out in Europe (87), followed by NorthAmerica (31), Asia (15), Africa and Australia (12 each), and SouthAmerica (11) (Fig. 3).3 This is similar to the geographical distribu-tion of similar studies irrespective of scale (Crossman et al., 2013).At a country level, most studies were carried out in the USA (26),followed by Germany (15), Australia (12), United Kingdom (11), theNetherlands (11), and Spain (10).

In total, 347 cases of ES mapping were identified in the reviewedpapers (Fig. 3), with an average of 2.8 ES mapped per paper (min = 1,max = 14). All cases of ES mapping are summarised and qualita-tively described in the supplementary online material. Regulatingand maintenance services were most commonly mapped (165),followed by cultural (85), and provisioning services (73). An addi-tional 24 “services”, that were either a combination (bundle) of

2 That is, studies stated to be made at a landscape scale, for the purpose oflandscape planning, or referring to a study area as a landscape or as containinglandscapes. See Section 2 (Material and Methods).

3 The two remaining papers were not focused on any particular country.

500 O. Englund et al. / Ecological Indicators 73 (2017) 492–504

Fig. 3. Geographical distribution of reviewed studies (n = 170). The number of studies performed in each country ranges from 1 (light grey) to 26 (black). White = zero.

0 10 20 30 40 50 60

Biom ass

Sur face water

Groundwater

Mechanical energy

Media�on of waste, toxics and other nuisances

Media�on of mass flows

Media�on of water flows

Media�on of gaseous/air flows

Lifecycle maintenance

Pest and disease control

Soil forma�on and com posi�on

Water cond i�ons

Atmosph eric compo si�on and cl imate regula�on

Physical and exp erien�al interac�ons with natur e

Intellectu al and representa�ve interac�ons with nature

Spiritual and/or emblema�c interac�ons with nature

Existence and bequest

Other or combina�on

Provision

ing

Regula�n

gandmaintenance

Cultural

Other

Number of mapping a�empts

Direct mapp ing Empirical mod el Simula�on or process mod el Logical mod el

F d at a

r(wBvs

Extrapola�on Data integra�on

ig. 4. Number of times different (groups of) ecosystem services have been mappe

ather than ecosystems, provides benefits to human well-beingsee Section 4 in Supplementary online material). A comparisonith previous reviews (Egoh et al., 2012; Martinez-Harms and

alvanera, 2012; Crossman et al., 2013) indicates that cultural ser-ices are more often mapped at the landscape scale than at largercales.

Com bina�on Unknown

landscape scale based on the review, divided into different method types used.

Atmospheric composition and climate regulation, usually relatingto global climate regulation, and Biomass, were the most commonlymapped ES (51 and 50 times, respectively), followed by Physical and

experiential interactions with nature (42 times), and Lifecycle main-tenance (40 times). The least commonly mapped ES include Pest anddisease control (4 times), Soil formation and composition (5 times),Existence and bequest, and Spiritual interactions with nature (both

O. Englund et al. / Ecological Indicators 73 (2017) 492–504 501

Fig. 5. Number of cases where mapping results were validated (dark grey) and not v

Fv

6oTeeghapqmsm

t

ig. 6. Number of cases where mapping results were validated (dark grey) and notalidated (light grey) with empirical data, for the different method types.

times), and Groundwater (7 times). No study included mappingf Mechanical energy, Mediation of air flows, or Water conditions.4

his large variation may reflect the perceived importance of differ-nt ES, but it may also reflect that different ES can be more or lessasily mapped. For example, the two most frequently mapped ES,lobal climate regulation and biomass production, are indisputablyigh priority in society and also easily mapped with adequateccuracy using proxies and statistics. Other ES that are also highriority, e.g., surface water and flow mediation, are much less fre-uently mapped. This may be explained by the more complicatedethods—such as process models or simulation—required to map

uch ES with adequate accuracy. Furthermore, the supply of ES isuch more commonly mapped than the demand, and few studies

4 Of course, a large number of scientific studies exist that do not refer explicitlyo ES, but have a landscape scope and address ES such as soil formation.

alidated (light grey) with empirical data, for the different ecosystem services.

attempt to analyse or discuss spatial links between providing andbenefiting areas.

Logical models and Empirical models were most commonly used(86 and 84 times, respectively), followed by Extrapolation (66times), Simulation/Process models (51 times), Data integration (24times), and Direct mapping (17 times). In ten cases, a combina-tion of several method types were used, and in 9 cases the typeof method was not possible to determine. Proxy based methodsare thus widely used also at the landscape scale. Methods were inmany cases difficult to assess and categorize due to very brief orotherwise insufficient method description; in nine cases we wereunsuccessful in determining which type of method had been used.This should serve as a reminder that method descriptions in sci-entific literature should not only facilitate understanding, but alsoreproduction. Several of the reviewed papers failed to facilitate thelatter. The method class definitions by Andrew et al. (2015), thatwere used to categorize methods, proved to be useful and it wasmostly straight-forward to classify methods accordingly. We agreehowever with the authors that there are some ambiguities in theseclass definitions, and that hybrid methods are possible. The methodclass definitions by Martínez-Harms and Balvanera (2012) may beuseful as a basis for further improvement of the typology.

4.2.1. ValidationOnly twelve percent of all ES mapping cases were validated with

empirical data. This is exclusive of the cases using direct mapping,since this does not require validation. No difference was foundbetween recent and older articles in this regard. In absolute terms,validation was most common for biomass, lifecycle maintenance,and physical and experiential interactions with nature (6 timeseach), followed by mediation of waste (5 times), and mediation of

mass flows (4 times). For all mapped ES at least one case includedvalidation (Figs. 4 and 5).

Validation was almost exclusively applied in studies based onempirical models, simulation and process models, or logical mod-

5 al Indi

erfsv(

eveeHbpshe

4

aeceae

ficsCarrii

aaHmr2tcv

cwnooaiiarivFrs

ib

02 O. Englund et al. / Ecologic

ls; constituting 42%, 35%, and 19% of all validated mapping cases,espectively (Fig. 6). Relative to the total number of mapping casesor each method type, validation was most common in studies usingimulation and process models (29% of all such mapping cases werealidated), followed by empirical models (21%) and logical models11%) (Fig. 6).

The common lack of validation is noteworthy. Collection ofmpirical data is time consuming and this probably explains whyalidation is most commonly made in studies that map ES usingmpirical models, or simulation and process models (fed withmpirical data), where empirical data must be collected anyway.owever, invalidated results can be difficult to evaluate and thuse of limited use for both academia and society in, e.g., landscapelanning. Validation should therefore be prioritised in ES mappingtudies. The widespread use of invalidated proxy-based methodsave previously been raised as a reason for concern by, e.g., Andrewt al. (2015).

.3. Further reflections

In order to use ES assessment as a basis for spatial planningnd decision-making, a high level of detail and accuracy is nec-ssary at varying spatial and temporal scales. Since landscapesommonly are heterogeneous—i.e., having patterns of differentcosystems—the ES supply is unequally distributed across spacend should therefore be assessed in spatially explicit ways (Nelsont al., 2009; Willemen et al., 2010, 2012).

Similar to Nemec and Raudsepp-Hearne (2013), we find it dif-cult to generalize about which methods that provide the mostredible results. Methods are context dependent and should be cho-en on the basis of competence, data availability, time frame, etc.arefully calibrated empirical or process based models, validatedgainst empirical data, can provide accurate and easily evaluatedesults, but they might not be relevant for certain ES, study areas, oresearch groups. The use of simple proxies in landscape level stud-es may generate misleading results, as has been discussed earliern the paper.

Regarding the use of models developed for broad use, Nemecnd Raudsepp-Hearne (2013) suggested that practitioners withdvanced GIS skills may benefit from creating their own models.owever, some models, e.g., the InVEST model, have been appliedany times, in several cases with validated and acceptably accu-

ate results (Hoyer and Chang 2014; Hamel et al., 2015; Meyer et al.,015). When using third-party models, it is however imperativehat they are properly evaluated on their suitability for the spe-ific project beforehand, and that they are carefully calibrated andalidated using empirical data.

The reviewed papers used different classification systems, butonversion of the reviewed ES into the CICES classification systemas in most cases fairly straight-forward. Most of the ES that couldot be fitted into CICES were either bundles of ES mapped togetherr examples of ecosystem processes rather than ES. Further devel-pment of CICES should consider whether to only include final ESnd thus exclude ecosystem processes and functions. For example,t can be argued that soil formation and composition by weather-ng, decomposition and fixing processes is not a final ES, but rathern intermediate ES, or an ecosystem function. The final ES shouldather be associated with what benefits to humans the soils facil-tate; e.g., cultivation of crops, or—indirectly, since soils facilitateegetation growth —mediation of water flows, and climate control.urthermore, “water conditions” was found to be redundant, as itefers to ensuring favourable living conditions for biota, which is

imilar to “lifecycle maintenance”.

Possible additions to CICES could be mediation of UV radiation,.e., shade, which is an ES commonly used by humans and animalsut very seldom described in the literature (no example was found

cators 73 (2017) 492–504

in the review). Finally, the comprehensiveness and use of moretechnical terms in CICES make it less suited for communication withstakeholders (including academics) that lack in-depth understand-ing of ES. Given the importance of stakeholder involvement in ESassessments, this is a clear disadvantage. It may therefore be ben-eficial to review the wording with this in mind, or to complementthe typology with alternative, less technical, descriptions.

5. Conclusions

Most of the 347 cases of ES mapping concerned regulating andmaintenance services followed by cultural and provisioning services.The results indicate that cultural services are more often mapped atthe landscape scale than at larger scales. The supply of ES is muchmore commonly mapped than the demand. Few studies attempt toanalyse or discuss spatial links between providing and benefitingareas.

While some ES—mediation of air/gaseous flows, mechanicalenergy, and water conditions—were not mapped in any of thereviewed studies, others—most notably atmospheric compositionand climate regulation, and biomass production—were commonlymapped. The large variation may reflect the perceived importanceof different ES, but it may also reflect that different ES can be moreor less easily mapped.

Overall, logical models and empirical models are most com-monly used, followed by extrapolation, simulation/process models,data integration, and direct mapping. Proxy based methods are thuswidely used also at the landscape scale. This review can serve as aninformation source concerning methods for ES mapping, hopefullyuseful for the design of new studies.

Methods were in many cases difficult to assess and categorizedue to very brief or otherwise insufficient method descriptions,which in some cases fail to facilitate reproduction.

Only twelve percent of all ES mapping cases were validatedwith empirical data. Invalidated results can be difficult to evalu-ate and thus be of limited use for both academia and society in, e.g.,landscape planning. Validation should be prioritised in ES mappingstudies.

There are highly diverging views on the meaning of landscapeand landscape scale, and the terms are sometimes used rather arbi-trarily. A total of 94 areas referred to as “landscape” were identified,with sizes ranging from 24 hectares to 122 million hectares.

Given the importance of high resolution and need for morecomplex methods and validation, most ES assessments with a land-scape scope will need to limit the number of ES included in thestudy. To ensure that the most relevant ES are included, it is essen-tial to involve stakeholders in the selection process. Furthermore,the capacity of the research group and available resources for theproject may determine which ES that can be included. ES thatcannot be studied in other ways than with simple proxies, or besufficiently validated, may preferably be omitted.

Acknowledgements

The authors would like to thank two anonymous reviewers forvaluable comments on the manuscript. Funding from E.ON., theSwedish Energy Agency (STEM), IEA Bioenergy, and the SwedishResearch Council for Environment, Agricultural Sciences and Spa-tial Planning (FORMAS) is gratefully acknowledged.

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.ecolind.2016.10.009.

al Indi

R

A

A

B

B

B

B

B

B

B

BB

C

C

C

C

C

C

D

D

d

D

D

D

E

E

E

E

E

E

F

O. Englund et al. / Ecologic

eferences

ndrew, M.E., Wulder, M.A., Nelson, T.A., 2014. Potential contributions of remotesensing to ecosystem service assessments. Prog. Phys. Geogr. 38 (3), 328–353.

ndrew, M.E., Wulder, M.A., Nelson, T.A., Coops, N.C., 2015. Spatial data, analysisapproaches, and information needs for spatial ecosystem service assessments:a review. GI Sci. Remote Sens. 52 (3), 344–373.

astian, O., Grunewald, K., Syrbe, R.-U., 2012. Space and time aspects of ecosystemservices, using the example of the EU Water Framework Directive InternationalJournal of Biodiversity Science. Ecosyst. Serv. Manage. 8 (1–2), 5–16.

astian, O., Lupp, G., Syrbe, R.-U., Steinhäußer, R., 2013. Ecosystem services andenergy crops −spatial differentiation of risks. Ekológia (Bratislava) 32 (1),13–29.

astian, O., Grunewald, K., Syrbe, R.U., Walz, U., Wende, W., 2014. Landscapeservices: the concept and its practical relevance. Landscape Ecol. 29 (9),1463–1479.

ormann, F.H., Likens, G.E., 1979. Catastrophic disturbance and the steady state inNorthern Hardwood Forests: a new look at the role of disturbance in thedevelopment of forest ecosystems suggests important implications forland-use policies. Am. Sci. 67 (6), 660–669.

oyd, J., Banzhaf, S., 2007. What are ecosystem services? The need for standardizedenvironmental accounting units. Ecol. Econ. 63 (2–3), 616–626.

oyd, J., 2007. Nonmarket benefits of nature: what should be counted in greenGDP? Ecol. Econ. 61 (4), 716–723.

rown, G., Fagerholm, N., 2015. Empirical PPGIS/PGIS mapping of ecosystemservices: a review and evaluation. Ecosyst. Serv. 13, 119–133.

urel, F., Baudry, J., 2003. Landscape Ecology. Science Publishers.urkhard, B., Crossman, N., Nedkov, S., Petz, K., Alkemade, R., 2013. Mapping and

modelling ecosystem services for science policy and practice. Ecosyst. Serv. 4,1–3.

onrad, E., Christie, M., Fazey, I., 2011. Is research keeping up with changes inlandscape policy? A review of the literature. J. Environ. Manage. 92 (9),2097–2108.

ostanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K.,Naeem, S., O’Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., van den Belt, M.,1997. The value of the world’s ecosystem services and natural capital. Nature387, 253–260, Available from: http://www.esd.ornl.gov/benefits conference/nature paper.pdf.

ostanza, R., 2008. Ecosystem services: multiple classification systems are needed.Biol. Conserv. 141 (2), 350–352.

ouncil of Europe, 2000. European Landscape Convention. Council of Europe,Strasbourg.

ouncil of Europe, 2005. Council of Europe Framework Convention on the Value ofCultural Heritage for Society. Council of Europe, Faro.

rossman, N.D., Burkhard, B., Nedkov, S., Willemen, L., Petz, K., Palomo, I., Drakou,E.G., Martín-López, B., McPhearson, T., Boyanova, K., Alkemade, R., Egoh, B.,Dunbar, M.B., Maes, J., 2013. A blueprint for mapping and modelling ecosystemservices. Ecosyst. Serv. 4, 4–14.

aily, G., 1997. Nature’s Services: Societal Dependence on Natural Ecosystems.Island Press, Washington DC.

aily, G.C., 1999. Developing a scientific basis for managing Earth’s life supportsystems. Ecol. Soc. 3 (2), 14.

e Groot, R., 2006. Function-analysis and valuation as a tool to assess land useconflicts in planning for sustainable, multi-functional landscapes. LandscapeUrban Plann. 75 (3–4), 175–186.

e Groot, R.S., Wilson, M.A., Boumans, R., 2002. A typology for the classification,description and valuation of ecosystem functions, goods and services. Ecol.Econ. 41 (3), 393–408.

eVries, B., Decuyper, M., Verbesselt, J., Zeileis, A., Herold, M., Joseph, S., 2015.Tracking disturbance-regrowth dynamics in tropical forests using structuralchange detection and Landsat time series. Remote Sens. Environ. 169, 320–334.

ee, N., Baker, J., Drobny, N., Duke, K., Whitman, I., Fahringer, D., 1973. Anenvironmental evaluation system for water resource planning. Water Resour.Res. 9, 523–535.

EA, 2013. Common International Classification of Ecosystem Services (CICES) v.4.3 (January 2013). European Environment Agency (EEA). Available from:http://cices.eu/content/uploads/sites/8/2015/09/CICES-V4-3- -17-01-13a.xlsx.

goh, B., Drakou, E.G., Dunbar, M.B., Maes, J., Willemen, L., 2012. Indicators forMapping Ecosystem Services: a Review. Publications Office of the EuropeanUnion, Luxembourg.

hrlich, P.R., Ehrlich, A.H., 1981. Extinction: the Causes and Consequences of theDisappearance of Species. Random House, New York.

igenbrod, F., Armsworth, P.R., Anderson, B.J., Heinemeyer, A., Gillings, S., Roy, D.B.,Thomas, C.D., Gaston, K.J., 2010a. The impact of proxy-based methods onmapping the distribution of ecosystem services. J. Appl. Ecol. 47 (2), 377–385.

igenbrod, F., Armsworth, P.R., Anderson, B.J., Heinemeyer, A., Gillings, S., Roy, D.B.,Thomas, C.D., Gaston, K.J., 2010b. Error propagation associated with benefitstransfer-based mapping of ecosystem services. Biol. Conserv. 143 (11),2487–2493.

ppink, F.V., Werntze, A., Mäs, S., Popp, A., Seppelt, R., 2012. Land Management andEcosystem Services. How Collaborative Research Programmes Can Support

Better Policies. Gaia: Okologische Perspektiven in Natur-, Geistes- undWirtschaftswissenschaften, 21 (1), 55–63.

agerholm, N., Kayhko, N., Ndumbaro, F., Khamis, M., 2012. Communitystakeholders’ knowledge in landscape assessments − Mapping indicators forlandscape services. Ecol. Indic. 18, 421–433.

cators 73 (2017) 492–504 503

Fisher, B., Turner, R.K., 2008. Ecosystem services: classification for valuation. Biol.Conserv. 141 (5), 1167–1169.

Fisher, B., Turner, R.K., Morling, P., 2009. Defining and classifying ecosystemservices for decision making. Ecol. Econ. 68 (3), 643–653.

Forman, R.T.T., Godron, M., 1986. Landscape Ecology. John Wiley & SonsIncorporated.

Haines-Young, R., Potschin, M., 2011. Common International Classification ofEcosystem Services (CICES): 2011 Update. European Environment Agency.

Hainz-Renetzeder, C., Schneidergruber, A., Kuttner, M., Wrbka, T., 2015. Assessingthe potential supply of landscape services to support ecological restoration ofdegraded landscapes: a case study in the Austrian-Hungarian trans-boundaryregion of Lake Neusiedl. Ecol. Modell. 295, 196–206.

Hamel, P., Chaplin-Kramer, R., Sim, S., Mueller, C., 2015. A new approach tomodeling the sediment retention service (InVEST 3.0): Case study of the CapeFear catchment North Carolina, USA. Sci. Total Environ. 524–525, 166–177.

Hauck, J., Görg, C., Varjopuro, R., Ratamäki, O., Maes, J., Wittmer, H., Jax, K., 2013.‘Maps have an air of authority’: Potential benefits and challenges of ecosystemservice maps at different levels of decision making. Ecosyst. Serv. 4, 25–32.

Helliwell, D.R., 1969. Valuation of wildlife resources. Reg. Stud. 3 (1), 41–47.Herlin, I.S., 2007. New challenges in the field of spatial planning: landscapes.

Landscape Res. 29 (4), 399–411.Hermann, A., Schleifer, S., Wrbka, T., 2011. The concept of ecosystem services

regarding landscape research: a review. Liv. Rev. Landsc. Res. 5 (1), 1–37.Hermann, Á., Kuttner, M., Hainz-Renetzeder, C., Konkoly-Gyuró, É., Tirászi, Á.,

Brandenburg, C., Allex, B., Ziener, K., Wrbka, T., 2014. Assessment frameworkfor landscape services in European cultural landscapes: an Austrian Hungariancase study. Ecol. Indic. 37, 229–240, PART A.

Higgins, S., Mahon, M., McDonagh, J., 2012. Interdisciplinary interpretations andapplications of the concept of scale in landscape research. J. Environ. Manage.113, 137–145.

Hoyer, R., Chang, H., 2014. Assessment of freshwater ecosystem services in thetualatin and Yamhill basins under climate change and urbanization. Appl.Geogr. 53, 402–416.

King, R.T., 1966. Wildlife and man. N. Y. Conserv. 20 (6), 8–11.Koschke, L., Fuerst, C., Frank, S., Makeschin, F., 2012. A multi-criteria approach for

an integrated land-cover-based assessment of ecosystem services provision tosupport landscape planning. Ecol. Indic. 21, 54–66.

Krovakova, K., Semeradova, S., Mudrochova, M., Skalos, J., 2015. Landscapefunctions and their change—a review on methodological approaches. Ecol. Eng.75, 378–383.

Lacoste, M., Minasny, B., McBratney, A., Michot, D., Viaud, V., Walter, C., 2014. Highresolution 3D mapping of soil organic carbon in a heterogeneous agriculturallandscape. Geoderma 213, 296–311.

Lamarque, P., Quetier, F., Lavorel, S., 2011. The diversity of the ecosystem servicesconcept and its implications for their assessment and management. Comp.Rendus Biol. 334 (5–6), 441–449.

MEA, 2005. Ecosystems and Human Well-being: Synthesis. Millennium EcosystemAssessment. Island Press, Washington DC.

MEA, 2003. Ecosystems and Human Well-Being: A Framework for Assessment.Island Press, Washington DC.

Maes, J., Egoh, B., Willemen, L., Liquete, C., Vihervaara, P., Schägner, J.P., Grizzetti,B., Drakou, E.G., Notte, A.L., Zulian, G., Bouraoui, F., Luisa Paracchini, M., Braat,L., Bidoglio, G., 2012. Mapping ecosystem services for policy support anddecision making in the European Union. Ecosyst. Serv. 1 (1), 31–39.

Malinga, R., Gordon, L.J., Jewitt, G., Lindborg, R., 2015. Mapping ecosystem servicesacross scales and continents—a review. Ecosyst. Serv. 13, 57–63.

Mao, Q.Z., Huang, G.L., Wu, J.G., 2015. Urban ecosystem services: a review. Chin. J.Appl. Ecol. 26 (4), 1023–1033.

Martinez-Harms, M.J., Balvanera, P., 2012. Methods for mapping ecosystem servicesupply: a review. Int. J. Biodiver. Sci. Ecosyst. Serv. Manage. 8 (1–2), 17–25.

Mastrangelo, M.E., Weyland, F., Villarino, S.H., Barral, M.P., Nahuelhual, L., Laterra,P., 2014. Concepts and methods for landscape multifunctionality and aunifying framework based on ecosystem services. Landscape Ecology 29 (2),345–358, http://dx.doi.org/10.1007/s10980-013-9959-9.

Meyer, M.A., Chand, T., Priess, J.A., 2015. Comparing bioenergy production sites inthe southeastern US regarding ecosystem service supply and demand. PLoSOne 10 (3), e0116336.

Nassauer, J.I., 2012. Landscape as medium and method for synthesis in urbanecological design. Landscape Urban Plann. 106 (3), 221–229.

Nelson, E., Mendoza, G., Regetz, J., Polasky, S., Tallis, H., Cameron, D.R., Chan,K.M.A., Daily, G.C., Goldstein, J., Kareiva, P.M., Lonsdorf, E., Naidoo, R., Ricketts,T.H., Shaw, M.R., 2009. Modeling multiple ecosystem services, biodiversityconservation, commodity production, and tradeoffs at landscape scales. Front.Ecol. Environ. 7 (1), 4–11.

Nemec, K.T., Raudsepp-Hearne, C., 2013. The use of geographic informationsystems to map and assess ecosystem services. Biodiver. Conserv. 22 (1), 1–15.

Portman, M.E., 2013. Ecosystem services in practice: challenges to real worldimplementation of ecosystem services across multiple landscapes—a criticalreview. Appl. Geogr. 45, 185–192.

Potschin, M., Haines-Young, R., Heink, U., Jax, K., 2014. OpenNESS Glossary (V2.0).[online]. Available from: http://www.openness-project.eu/library/reference-

book.

Seppelt, R., Dormann, C.F., Eppink, F.V., Lautenbach, S., Schmidt, S., 2011. Aquantitative review of ecosystem service studies: approaches, shortcomingsand the road ahead. J. Appl. Ecol. 48 (3), 630–636.

5 al Indi

S

S

S

T

T

T

T

T

U

Manage. 100, 86–95.Willemen, L., Burkhart, B., Crossman, N., Drakou, E.G., Palomo, I., 2015. Editorial:

best practices for mapping ecosystem services. Ecosyst. Serv. 13, 1–5.Zhang, W., Ricketts, T.H., Kremen, C., Carney, K., Swinton, S.M., 2007. Ecosystem

04 O. Englund et al. / Ecologic

eppelt, R., Fath, B., Burkhard, B., Fisher, J.L., Grêt-Regamey, A., Lautenbach, S., Pert,P., Hotes, S., Spangenberg, J., Verburg, P.H., van Oudenhoven, A.P.E., 2012. Formfollows function?: proposing a blueprint for ecosystem service assessmentsbased on reviews and case studies. Ecol. Indic. 21, 145–154.

oto, C., Palomares, F., 2015. Coexistence of sympatric carnivores in relativelyhomogeneous Mediterranean landscapes: functional importance of habitatsegregation at the fine-scale level. Oecologia.

tephens, P.A., Pettorelli, N., Barlow, J., Whittingham, M.J., Cadotte, M.W., 2015.Management by proxy? The use of indices in applied ecology. J. Appl. Ecol. 52(1), 1–6.

EEB, 2010a. The Economics of Ecosystems and Biodiversity: Ecological andEconomic Foundations. Earthscan, London.

EEB, 2010. The Economics of Ecosystems and Biodiversity: Mainstreaming theEconomics of Nature: A synthesis of the approach, conclusions andrecommendations of TEEB. doc.teebweb.org.

ermorshuizen, J.W., Opdam, P., 2009. Landscape services as a bridge betweenlandscape ecology and sustainable development. Landscape Ecol. 24 (8),1037–1052.

roy, A., Wilson, M.A., 2006. Mapping ecosystem services: practical challenges andopportunities in linking GIS and value transfer. Ecol. Econ. 60 (2), 435–449.

urner, M.G., 1989. Landscape ecology—the effect of pattern on process. Annu. Rev.Ecol. Syst. 20 (1), 171–197.

NESCO, 2003. Convention for the Safeguarding of the Intangible CulturalHeritage. unesdoc.unesco.org. Paris: The General Conference of the UnitedNations Educational, Scientific and Cultural. Organization (UNESCO).

cators 73 (2017) 492–504

Ungaro, F., Zasada, I., Piorr, A., 2014. Mapping landscape services: spatial synergiesand trade-offs. A case study using variogram models and geostatisticalsimulations in an agrarian landscape in North-East Germany. Ecol. Indic. 46,367–378.

Vallés-Planells, M., Galiana, F., Van Eetvelde, V., 2014. A classification of landscapeservices to support local landscape planning. Ecol. Soc. 19 (1) (art44).

Vihervaara, P., Kumpula, T., Tanskanen, A., Burkhard, B., 2010. Ecosystemservices–A tool for sustainable management of human–environment systems.Case study Finnish Forest Lapland. Ecol. Complexity 7 (3), 410–420.

Wallace, K.J., 2007. Classification of ecosystem services: problems and solutions.Biol. Conserv. 139 (3–4), 235–246.

Willemen, L., Hein, L., Verburg, P.H., 2010. Evaluating the impact of regionaldevelopment policies on future landscape services. Ecol. Econ. 69 (11),2244–2254.

Willemen, L., Veldkamp, A., Verburg, P.H., Hein, L., Leemans, R., 2012. A multi-scalemodelling approach for analysing landscape service dynamics. J. Environ.

services and dis-services to agriculture. Ecol. Econ. 64 (2), 253–260.