4 Quantifying and mapping

Dependent on data availability and spatial and temporal scales of assessments, different methods are available for quantifying and mapping landscape functions/ services. For assessments at global level as well as for rapid assessments, landscape functions and services can be determined directly by land cover or ecosystems using general assumptions from literature reviews. These methods are often applied when the economic value of the area is interesting (e.g. Naidoo and Ricketts, 2006*; Troy and Wilson, 2006*). However, a proper presentation of landscape functions/services would require also additional data beyond land cover observations. For example, the recreational function of a landscape is not only defined by the land cover of a specific location (e.g. natural area) but depends also on accessibility properties (e.g. distance to roads) and characteristics of the surrounding landscape (de Groot et al., 2010). But in many cases this is only achievable at local or at least regional levels, because of data availability.

Kienast et al. (2009*) present a framework for a spatially explicit landscape functions assessment at European scale, linking land characteristics with a high number of landscape functions. However, the assessments are often primarily based on area measurements and only marginally on measurements of quality (e.g. land use diversity, forest structure).

At regional or local scale a more data-driven method can be used. Function and service data are originated mainly from field observations, including census data, spatial policy documents and biophysical data. Willemen et al. (2008*) present a methodological framework to quantify landscape functions and to make their spatial variability explicit. They distinguish three different methods depending on the measurable function: (1) linking landscape functions to land cover or spatial policy data, (2) empirical predictions using spatial indicators and (3) decision rules based on literature reviews (Willemen et al., 2008*). Whereas for some functions the exact location can be directly observed from the land-cover (e.g. wood for timber production), other functions such as recreation cannot be directly observed or only partially delineated and thus have to be empirically assessed based on landscape indicator analyses. If there does not exist any direct referenced information on the function’s location (e.g. leisure cycling), we have to rely on landscape data based on expert knowledge, literature reviews or process models.

A lot of studies dealt with these challenges aiming at providing spatial datasets to map landscape functions (e.g. Chan et al., 2006*; Haines-Young et al., 2006*; Gimona and Van der Horst, 2007*; Egoh et al., 2008*; Meyer and Grabaum, 2008*). However, by doing the analysis major problems encountered. Finding appropriate indicators related to the specific service providing unit and exploring how functions and services are correlated with different landscape scenarios are still unresolved questions. To investigate the capacity of landscapes to provide services, landscape complexity and configuration analysis have to be addressed. Aspects such as size, form, and the border length between neighbouring land use types as well as the spatial connectivity of landscape units have to be taken into account. However, current landscape service indicators are still limited by insufficient data and an overall low ability to convey information (Layke, 2009).

Some indicators available are inadequate in characterizing the diversity and complexity of the services provided by landscape functions, especially concerning regulation as well as cultural services, which occur at various spatial scales. Ecosystems are complex, interrelated systems, in which processes take place over a range of spatial and temporal scales (Tansley, 1935) varying from competition between individual plants at plot level, via meso-scale processes such as fire and insect outbreaks, to climatic and geomorphologic processes at largest spatial and temporal scales (Clark et al., 1979; Holling et al., 2002). As service supply is dependent on ecosystem processes and functions, it may occur at different scales. Some services are even relevant at more than one scale. For instance regulation services can occur both at global scale (climate regulation) and plot-scale (biological nitrogen fixation) (de Groot, 1992). Also pressures on ecosystem services can have effects at different scales. In general physical processes on small scales are often driven by the impact on long period phenomena at large scales (climate patterns, hurricanes, fires) (Limburg et al., 2002*). However, large scale processes are also strongly influenced by smaller scale occurrences, for example, microbes respire enough CO₂ to keep many lakes and rivers supersaturated (Levin, 1992; del Giorgio et al., 1997). Hence, for the analyses of the dynamics of ecosystem service supply it is very important to consider the drivers and processes at scales relevant for ecosystem service generation.

In addition, relevant to the time frame, ecosystems can act as service provider or suppressor (Martin and Blossey, 2009*). For example, wetlands dominated by Phragmites australis can act as source and sink for greenhouse gases, depending on time scale (Brix et al., 2001). The species assimilates atmospheric carbon dioxide through photosynthesis and through sequestration of organic matter produced in wetland soils. But it also emits methane into the atmosphere in a two stage process (Beckett et al., 2001). Therefore, before an ecosystem can be seen as a service supplier, a time frame has to be defined for evaluation.