Science of the Total Environment 521–522 (2015) 108–122

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Methodology to assess and map the potential development of forest ecosystems exposed to climate change and atmospheric nitrogen deposition: A pilot study in Germany Winfried Schröder a,⁎, Stefan Nickel a, Martin Jenssen b, Jan Riediger a a b

University of Vechta, Chair of Landscape Ecology, P.O.B. 1553, 49377 Vechta, Germany Waldkunde-Institut Eberswalde GmbH, Hohensaatener Dorfstraße 27, 16259 Bad Freienwalde, Germany

H I G H L I G H T S • • • • •

A spatial explicit methodology for evaluating integrity of forests was developed. Data on vegetation, soil condition, climate change and atmospheric were used. Forest types were classified based on data from 21,600 stands. Integrity was investigated by comparing current, future and reference states. Potential future conditions of forests were proved to be positive and negative.

a r t i c l e

i n f o

Article history: Received 10 November 2014 Received in revised form 11 March 2015 Accepted 12 March 2015 Available online xxxx Editor: J. P. Bennett Keywords: Ecological functions Ecosystem classification Soil modelling Climate modelling Mapping

a b s t r a c t A methodology for mapping ecosystems and their potential development under climate change and atmospheric nitrogen deposition was developed using examples from Germany. The methodology integrated data on vegetation, soil, climate change and atmospheric nitrogen deposition. These data were used to classify ecosystem types regarding six ecological functions and interrelated structures. Respective data covering 1961–1990 were used for reference. The assessment of functional and structural integrity relies on comparing a current or future state with an ecosystem type-specific reference. While current functions and structures of ecosystems were quantified by measurements, potential future developments were projected by geochemical soil modelling and data from a regional climate change model. The ecosystem types referenced the potential natural vegetation and were mapped using data on current tree species coverage and land use. In this manner, current ecosystem types were derived, which were related to data on elevation, soil texture, and climate for the years 1961–1990. These relations were quantified by Classification and Regression Trees, which were used to map the spatial patterns of ecosystem type clusters for 1961– 1990. The climate data for these years were subsequently replaced by the results of a regional climate model for 1991–2010, 2011–2040, and 2041–2070. For each of these periods, one map of ecosystem type clusters was produced and evaluated with regard to the development of areal coverage of ecosystem type clusters over time. This evaluation of the structural aspects of ecological integrity at the national level was added by projecting potential future values of indicators for ecological functions at the site level by using the Very Simple Dynamic soil modelling technique based on climate data and two scenarios of nitrogen deposition as input. The results were compared to the reference and enabled an evaluation of site-specific ecosystem changes over time which proved to be both, positive and negative. © 2015 Elsevier B.V. All rights reserved.

1. Background and goal Approximately 30% and 40% of the German and European land surface, respectively, are covered by forests (EU, 2014). Their structures and functions and, subsequently, their service to human societies may ⁎ Corresponding author. E-mail address: [email protected] (W. Schröder).

http://dx.doi.org/10.1016/j.scitotenv.2015.03.048 0048-9697/© 2015 Elsevier B.V. All rights reserved.

be influenced by climate change and air pollution as structures and functions are reconstructed and other ecosystem types evolve (Allen, 2009; Allen et al., 2010; Bobbink et al., 2010; de Vries et al., 2014; de Vries and Posch, 2011; EEA, 2012; EUROSTAT, 2012; FAO, 2009, 2010; FOREST EUROPE, UNECE, FAO, 2011; Gobiet et al., 2014; Luedeling et al., 2013; Manion, 1991; Maroschek et al., 2009; Nagajyoti et al., 2010; Paoletti et al., 2010; Richardson et al., 2013; Stankovic et al., 2014; Tanino et al., 2010; Vitasse et al., 2009a,b, 2011; Vose et al.,

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

2012; Wulff et al., 2012). Simpson et al. (2014) were able to show that climate and emission changes impact nitrogen (N) deposition. Changes in ecosystems often are noted in references to ecosystem quality, ecosystem state, ecosystem condition, ecosystem health, and ecosystem integrity (Holyoak and Hochberg, 2013) and linked to various diversity/ stability and scaling concepts (Fränzle, 1994; Chase and Knight, 2013; Chave, 2013; Ives and Carpenter, 2007; Loreau and de Mazancourt, 2013; Saint-Béat et al., 2015; Thibaut and Connolly, 2013). Certain functions and structures of ecosystems are valued by people because they serve to regulate ecosystem conditions and the related aspects mentioned above, provide material products and contribute nonmaterial human benefits. To safeguard the regulating, provisioning and cultural services of ecosystems, their structures and functions need to be monitored and protected. This is the aim of environmental directives and conventions such as the EU Biodiversity Strategy to 2020 and the UNECE Convention on Long-range Transboundary Air Pollution (Air Convention, UNECE, 2013). The framework of the Air Convention, and the activities therein that refer to climate change and ecological structures and functions, comprises the Cooperative Programme for Monitoring and Evaluation of the Long-range Transmission of Air Pollutants in Europe (EMEP) and the Working Group on Effects, which manage six International Cooperative Programmes (ICP). The ICP Forests, ICP Vegetation and ICP Integrated Monitoring are of special relevance for assessing the impacts of climate change and air pollution on ecological structures and functions. Theoretical models and empirical studies have identified linkages between changes in biodiversity and ecological functions (Balvanera et al., 2006; Doherty et al., 2000; Hooper et al., 2005). Even minor losses in the number of species may reduce ecosystem functions and stability aspects such as resilience when faced with environmental change. This is especially true for those species that each uniquely contributes to the functioning of the ecosystem. The possibility of reduced ecosystem function is increased as more species are lost due to reductions in substitutability. Thus, biodiversity and ecosystem functions are codependent, and therefore, biodiversity is vital to maintaining functioning ecosystems and vice versa (Maynard et al., 2010, 2011; Midgley, 2012; Petter et al., 2013). Under action 5 of the EU Biodiversity Strategy to 2020, the condition of ecosystems and their services should be mapped and assessed across Europe. To this end, information about drivers and pressures, such as, for instance, air pollution and climate change, as well as their “impacts on structure and function of each ecosystem type,” should be assessed by using available data (EU, 2014:20). Thus, a strategy for the Mapping and Assessment of Ecosystems and their Services (EU, 2014) was developed. This Europe-wide approach addresses ecological structures and functions and encompasses ecosystem type and ecosystem condition mapping. Mapping ecosystem condition is used to deliver information about the services each ecosystem type can provide while taking into account climate, geology and other natural factors, as well as the drivers and pressures to which the ecosystem types are exposed. Changes in ecosystem condition due to environmental changes such as land use, air pollution or climate change provide further information about the ecosystem's capacity to deliver services over time. Mapping ecosystems provides information about the spatial extension and distribution of the main ecosystem types and is regarded as the starting point for assessing the condition of each ecosystem type. The ecosystem typology differentiates at three levels and takes into account mapping feasibility at the European scale while aiming at compatibility with national mapping approaches. Additionally, national and subnational data sources should be used in pilot studies to detail the ecosystem coverage and condition across Europe (EU, 2014). In Germany, an integrative approach that can cope with potential modifications in ecological structures and functions due to climate change and atmospheric N deposition is still lacking. Therefore, the objective of the study at hand was to develop a comprehensive and spatially explicit methodology for generating and verifying hypotheses on the integrity of forest ecosystems using available data. The methodology

109

should enable an evaluation of ecosystem integrity (ESi) both at the site level and across Germany. 2. Materials and methods Focusing on forest ecosystems, the methodology, which was quantitatively developed, achieves the following six objectives: 1. defining ESi; 2. classifying (forest) ecosystem types and establishing an indicator-based reference system; 3. mapping ecosystem types (EsT); 4. generating spatial hypotheses (predictive maps) on potential patterns of EsT regions across time (1961–2070); 5. generating hypotheses (projections) on potential developments of site-specific ESi indicators for the years 2011–2070 by numeric modelling; and 6. evaluating potential developments in ESi. 2.1. Defining ecosystem integrity (ESi) Ecosystems are dynamic open systems encompassing plant, animal, and microorganism communities and components such as air, water, soil, and bedrock that interact through fluxes of energy, information and matter that serve as functional units (Fränzle, 1994; Hassan et al., 2005; White et al., 1992). From the interactions among the above-mentioned compartment-specific biophysical structures – the ‘architecture’ of ecosystems in terms of their horizontal and vertical setting, geographical location, and topographical features – functions (processes) emerge. For scientific understanding environmental policies affecting the physical, chemical and biological conditions (state) of an ecosystem across time need to be monitored and compared to reference systems. These reference systems could be agreed-upon targets in environmental directives such as, at the European level, the Habitats Directive, the Water Framework Directive or the Marine Strategy Framework Directive (European Commission, 2014), or historical or projected potential future conditions. Because many ecological phenomena such as biodiversity or theoretical concepts such as ecosystem integrity are too complex for direct measurements, they are subdivided into measurable units, which serve, when each is taken as a part of the whole, as indicators (Pesch and Schröder, 2006). We are aware that a multitude of definitions of ecosystem integrity exists, each regarding specific aspects of ecosystem structure

Table 1 Ecosystem functions, corresponding indicators and respective data for quantification of ecosystem integrity. Ecosystem function

Indicators

(1) Habitat function

Deviations of compositions of species to reference condition, e.g., Kullback-distance (Kullback, 1951; Jenssen, 2010) (2) Net primary Deviation of NPP to reference production (NPP) condition (3) Carbon Deviation of biomass and sequestration humus condition to reference conditions

(4) Nutrient flow

(5) Water flow

(6) Resilience

Deviation of C/N ratio, soil acidity, base saturation and nutrients concentration in needles and leaves to reference conditions Deviation of soil moisture index to reference conditions Deviation of tree species composition from reference conditions

Data Vegetation databases ICP Level 2

NPP models Forest management Forest management ICP Level 2 Forest soil data Long-term soil observation Atmospheric deposition Indicator value models Forest soil data Long-term soil observation Atmospheric deposition Indicator value models Forest soil data Long-term soil observation ICP Level 2 Potential natural vegetation Forest management Agriculture management

110

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

and functioning and reflecting the subjective perspectives of humans on the value, importance, and role of the respective ecosystem features (Carignan and Villard, 2002; de Leo and Levin, 1997; Manuel-Navarete et al., 2004); nevertheless, according to our understanding, ecosystem integrity (ESi) refers to a consistency in ecosystem structures and ecosystem functions that correspond to the particular local or regional natural potential or other reference, such as historical or normative ecosystem conditions (Andreasen et al., 2001; Karr and Dudley, 1981; Haines-Young and Potschin, 2010; Stoll et al., 2015; Tierney et al., 2009). Ecosystem functions are understood as the relations among the compartments that constitute ecosystem structures: atmosphere, hydrosphere, pedosphere, and biosphere. The biosphere comprises communities, populations

and species of organisms interacting with each other and with abiotic features of their environments in terms of production, consumption and decomposition. Here, fluxes of energy, substances, and information, which exhibit characteristic patterns on a range of scales of time and space, as well as in structural and functional complexity, are processed (Jenssen et al., 2013 in accordance with: Angermeier and Karr, 1994; Cairns, 1977; Limburg et al., 1986; Lindenmayer and Franklin, 2002; Parrish et al., 2003). For a spatially explicit operationalization of this ESi definition based on collected data, an ecosystem classification is needed that takes into consideration ecosystem structures and functions. The goal of an ecological integrity assessment is to provide information on the condition of ecosystem types (Faber-Langendoen et al., 2012a).

Fig. 1. Geographical location of modelling sites within Germany. LII UNECE International Cooperative Programme Forests Level II. W.I.E. Waldkunde-Institut Eberswalde.

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

111

Fig. 2. Basic scheme for stepwise evaluation of ecological integrity for single indicators, for ecological functions and for ecosystems.

2.2. Ecosystem classification and indicator-based reference system From the above, it is clear that indicators of ecological integrity should be based on an understanding of the structure and functions of ecosystems. Ecological classifications can help categorize the variability within and among ecosystem types so that differences among several grades of ecosystem integrity can be more clearly recognized (Faber-Langendoen et al., 2012a,b). According to Hofmann (1997), ecosystem types should be classified as entities characterized by certain homogeneity of significant features of their structures and functions. Accordingly, ecosystem types were categorized using data collected for 21,600 forest sites across Germany during the years 1961–1990. The data quantify site factors (soil type, topography, and climate) as well as ecological functions and related structures in terms of habitat function, net primary production, carbon sequestration, nutrient and water flow, in addition to adaptability to climate change and atmospheric N deposition (resilience). With respect to these functions, structures were selected according to their ecological significance (Hofmann, 1997; Jenssen et al., 2013) and the availability of data for the quantification of corresponding indicators (Section 2.5). These data should not only allow for classifying ecosystem types but also enable a comparison of the state of these ecosystem types over the years from 1961–1990, referred to herein as reference state. The comparison will involve i) the current ecosystem state as defined in this investigation by the years 1991–2010, and ii) the potential future ecosystem conditions (2011– 2040, 2041–2070), calculated with climate change projections, and with decision tree-based nationwide mapping (Sections 2.3, 2.4) and site-specific numerical modelling (Section 2.6) being applied to both current and future conditions. The data representing the reference state ecosystem types (EsT) were derived from and referred to the potential natural vegetation (Bohn et al., 2000/2003, 2005; Suck et al., 2010, 2013) and the European Habitats Directive Annex I habitat types (EEC, 1992).

According to the respective parameter value interval of the particular ecological function, and by application of a vegetation similarity measure developed by Hofmann and Passarge (1964), each ecosystem type was assigned a code that comprehensively addressed the geographic region, the soil water budget and the biogeochemical budget as indicated by the humus condition. In this way, 135 semi-natural forest ecosystem types, 16 semi-natural open land ecosystem types, 45 cultivated forest ecosystem types, and 18 cultivated grassland types were classified and detailed quantitatively by Jenssen et al. (2013). The classification system includes not only ecosystem type-specific reference conditions but also potential succession states. This is of significance inasmuch, at present, future states for only three of the six functional indicators can be estimated by numerical modelling (Sections 2.6, 2.7).

2.3. Mapping ecosystem types Linking the potential natural vegetation (pnV) map (Suck et al., 2010) with the dominating ecosystem type that is spatially included in the pnV complexes within a geographic system (GIS) and applying a vegetation similarity measure according to Jenssen (2010) enabled mapping of the potential natural ecosystem types (pnEsT) across Germany. Then, the pnEsT GIS map was connected with GIS maps of recent tree species coverage1 and actual land use,2 and current ecosystem 1 http://www.eea.europa.eu/data-and-maps/data/external/change-of-habitatsuitability-coverage The spatial distribution of 30 forest tree species was modelled by application of classification and regression tree analysis and Random Forest algorithms and mapped on a 1 km by 1 km grid. 2 http://www.eea.europa.eu/data-and-maps/data/corine-land-cover-2000-clc2000seamless-vector-database Corine Land Cover 2000. European Environmental Agency (EEA). The Corine Landcover 2005-map was not available nationwide for Germany.

112

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

types (cEsT) were identified by application of the following two conditional statements to the aforementioned geodata and mapped: 1. IF dominating land use categories are consistent with pnEsT AND tree species coincide with pnEsT, THEN pnEsT are mapped as current seminatural ecosystem type (csnEsT); 2. ELSE IF dominating tree species correspond to a cultivated ecosystem type that takes the place of the pnEsT on the particular site, THEN this cultivated ecosystem type is mapped as csnEsT; 3. ELSE the ecosystem is mapped as current non-natural ecosystem type (cnnEsT). 2.4. Mapping EsT-regions across time For identifying potential areal shifts of EsT over time due to climate change, the map of ecosystem types (reference state 1961–1990), as derived by the methodology explained in Section 2.3, was related to geodata on elevation a.s.l., soil texture, and climate (average monthly minimum, maximum and mean of air temperature, monthly means of relative air humidity, evapotranspiration and precipitation) collected within the reference period 1961–1990. The multiple statistical relations were modelled by Classification and Regression Trees (CART; Breiman et al., 1984). The resulting if–then–else rules identified for the reference period 1961–1990 were then applied to the abovementioned geodata on EsT, elevation a.s.l., soil texture and, iteratively, climate in terms of the aforementioned meteorological phenomena for the periods 1991–2010, 2011–2040, and 2041–2070. These data were computed by the Potsdam Institute for Climate Impact Research for two climate change scenarios with a regional climate model (Section 2.6). The application of the CART rules describing the statistical relation between the ecosystem types (reference state 1961–1990), climate (1961–1990, 1991–2010, 2011–2040, and 2041–2070), elevation above sea level (a.s.l.) and soil texture to the geodata enabled predictive mapping of regions for EsT. Each of these spatial clusters contained EsT with similar relations to elevation a.s.l., soil texture, and climate. 2.5. Indicating ecological functions To define and make measurable ecosystem integrity (Section 2.1), six ecological functions were determined according to their relevance and data availability (Table 1) and then specified quantitatively by the following indicators: habitat function, net primary production, carbon sequestration, and water and nutrient budget, as well as adaptability. The present (1991–2010) or potential future (2011–2040 and 2041– 2070) features of these functions were determined for seven sites that correspond to seven different ecosystem types. Then, they were compared to the respective reference state data (1961–1990), which are available for 33 ecosystem types at present. For instance, the habitat function indicated by composition and spatial density of vegetation coverage was assessed by calculating similarity measures developed by Kullback (1951) and Jenssen (2010). The carbon sequestration was quantified by the contents of organic carbon within the humus layer and the upper 80 cm of the soil column. The carbon to nitrogen (C/N) ratio was used as a bulk parameter for the nutrient budget of the ecosystem. Further indicators, such as pH value and base saturation, were derived by vegetation-based indicator models (Jenssen, 2010; Jenssen et al., 2013). The water budget was measured by scores based on indicator values for moisture according to Hofmann (2002). The adaptability to changing environmental boundary conditions was quantified by measuring the similarity between current and natural tree species percentages (Jenssen and Hofmann, 2003). Due to the ecological relevance and the functions of the modelling technique applied to estimate the potential future development of ecosystem state, three out of the six ecological functions used to assess ecosystem integrity were selected: carbon sequestration, nutrient budget, and water budget (Section 2.6).

2.6. Projecting site-specific ESi indicators (soil) The sustainability of forest ecosystem functions depends upon biogeochemical cycles over the long term. Therefore, estimating the integrity of ecosystems exposed to climate change and atmospheric N deposition (de Vries et al., 2014) is, if indicated by projecting potential future developments, one way to draft and decide upon precautionary measures (Matyssek et al., 2012). For estimating potential future ecosystem states, the Very Simple Dynamic (VSD + 3.6.1.2) model was used. VSD + is an extension of the soil acidification VSD model (Bonten et al., 2009; Posch and Reinds, 2009) and includes organic C and N dynamics. VSD + allows the estimation of C and N pools in litterfall, root turnover, and N fixation and allows the calculation of the C/N ratio. Litterfall and root turnover are distributed over two fresh litter pools (i.e., easily degradable and recalcitrant fresh litter), depending on the C/N ratio of litterfall and root turnover. The use of VSD+ facilitates the calculation of the accumulation of N and C in humic mat− ter and the calculation of the release of NH+ 4 and NO3 through mineralization and nitrification. VSD+ is well approved (Posch et al., 2003) and was used in several scientific studies — in particular, in the work of the International Cooperative Programme on Modelling and Mapping of

Table 2 Abstract from the complete list of Ecosystem Types (EsT) according to Jenssen et al. (2013) (two left columns) and related potential natural vegetation (PnV) (Suck et al., 2010) and Habitats Directive Annex I habitat types (HabT). 1. Natural ecosystem types 1.3. Lowland ecosystems, Atlantic, sub-Atlantic and central European 1.3.1 Beech tree ecosystems, lowland PnV Eb–7n–C2 Hygrophilous moder beech forests Lb2f Eb–7n–E2 Hygrophilous lime-mull ash and beech Nc6a forests Eb–5r–E2 Calciphilous lime-mull beech forests Nc4b on slopes Eb–4n–C2 Sandy moder sessile oak and beech Lb2c forests Eb–5n–C2 Moder beech forests on bunter Lc3a 1.8. Coniferous woods, subalpine and higher montane level 1.8.1. Spruce tree ecosystems, higher montane level Cg–7ü1–T4 Alluvial spruce forests of the montane S3 level C4–8o–T3 Acidophilous spruce bog forests of the S2 altimontane level C4–7m–Ta1N Spruce block forests of the altimontane S5F, T3 level C4–6d–B1 Raw humus spruce forests of the S1a altimontane level C4–6d–Ta1N Calciphilous spruce forests of the T2, T1 altimontane level 2. Cultivated ecosystem types Coniferous forests Pine tree forests, semi-dry Ec–2n–b1 Thermophilous raw humus pine forests Eb, Ec–3n–c1 Thermophilous raw humus-moder pine forests Thermophilous moder pine forests Ec–3n–c2

HabT 9110 9130 9130 9110 9110

9410 *91D0, 9410 9410 9410 9410

Ec–3n–B2 Eb,Ec–3n–C1

–

Ec–3n–D1

–

Ecosystem-Code: Regional hypsometric and horizontal allocation: Cg = Montane level — generally; C4 = Montane level — (boreal); Eb = Sub-Atlantic lowland, Ec = Middle European lowland; soil moisture: 2n = extremely dry; 3n = dry; 4n = moderately dry; 5n = fresh; 6d = constantly fresh; 7n = moist; 7ü1 = constantly moist-flooded; 7m = constantly moist-mineral; 8o = constantly wet-organic; humus condition: T4 = acidicalkaline-peat; T3 = acidic peat; Ta1N = Tangel/lime; B1/b1 = raw humus; B2 = moder– raw humus; C1/c1 = raw humus–moder; C2/c2 = moder; D1 = brown mull; E2 = mull. Potential natural vegetation (Suck et al., 2010): S1a = Calamagrostis villosa—Piceetum; S2 = Calamagrostis villosa–Piceetum sphagnetosum; S3 = Equisetum sylvatica—Abietetum; S5F = Betula carpatica–Picea abies; Lb2c = Majanthemo—Fagetum sylvaticae; Lb2f = Frangula—Fagetum sylvaticae; Lc3a = Luzulo fagetum; Nc4b = Hordelymo–Fagetum; Nc6a = Fraxino–Fagetea; T1 = Homogyne–Piceetum; T2 = Adenostyles glabrae– Piceetum. Habitats Directive Annex I habitat types: 9110 = Luzulo–Fagetum (beech forests); 9130 = Asperulo–Fagetum beech forests *91D0 = Bog woodland; 9410 = Acidophilous Picea forests of the montane to alpine levels (Vaccinio–Piceetea).

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

Critical Levels and Loads and Air Pollution Effects, Risks and Trends of the UNECE Air Convention (Slootweg et al., 2015). The modelling of soil parameters such as pH, base saturation, concentrations of anions and cations in the soil solution, C and N content and C/N ratio with VSD+ was performed for seven forest sites and for the time periods from the first year with available data until 2010, for 2011–2040 and for 2041–2070 (Fig. 1). The respective results were compared with reference values (1961–1990) and used to assess ecosystem integrity, as detailed in Section 2.7. Initial information on soils and forest stands were derived from the particular ecosystem type profile (Section 2.2) and from the databases of the UNECE International Cooperative Programme Forests Level II and the Waldkunde-Institut Eberswalde. The model runs were calibrated with multiple measurements from different years for input parameters such as pH and concentrations of anions and cations in the soil solution, C and N content, and C/N ratio. The modelling covered two scenarios of atmospheric N deposition: one according to the European average deposition in 1900 (5 kg N/ha a in all years after 2010) and one representing the average European deposition occurring in 1980 (15 kg N/ha a in all years after 2010) (de Vries and Posch, 2011; Torseth et al., 2012). The air temperature and precipitation data needed for projecting potential soil conditions using the indicators for the three soil functions mentioned above were produced by regional climate modelling. Regional climate models bridge the gap between the general circulation models that operate on a global scale at rather coarse horizontal resolution and the climate impact models that focus on specific ecosystem functions operating at a regional or local scale (Reyer et al., 2013). To this end, impact models such as VSD + use the results of regional climate models as input data.

113

For this investigation, the modelling of potential future soil conditions with VSD+ used climate data calculated with the Statistical Analogue Resampling Scheme (STARS) for the RCP 2.6 and 8.5 scenario (Feldhoff et al., 2014; Lutz and Gerstengarbe, 2014; Orlowsky et al., 2008). STARS stochastically resamples meteorological data according to an observed trend of a meteorological variable that is known as robust in the context of global warming, such as air temperature (Feldhoff et al., 2014). The simulation output is physically consistent. However, contrary to dynamic models, STARS is not able to create new extreme values that are greater than the observation data (Feldhoff et al., 2014). Nevertheless, projected droughts, for example, can be longer than in the reference period. Observation climate data covering the years 1951–2010 and derived from the German weather service were checked, homogenized and interpolated on a 10 km×10 km grid (Österle et al., 2006). Data on air temperature, sunshine duration and precipitation for the RCP 2.6 and 8.5 scenarios covering the time period 2011–2070 and modelled daily were generated by STARS and provided by the Potsdam Institute for Climate Impact Research on and interpolated on the same grid (Orlowsky et al., 2008). For each grid cell, the STARS projections consist of 100 different model runs, each driven by different temperature gradients. In the modelling, the model run that represents the 95% quartile and a high temperature increase for the RCP 2.6 and 8.5 scenarios was considered. This model run is based on results from the ACCESS1.0 model provided by the Commonwealth Scientific and Industrial Research Organisation, Melbourne, Australia and is explained by Bi et al. (2013). The RCP 2.6 and 8.5 scenarios are driven by an increase in the atmospheric CO2 concentration by approximately 40 and 210% of the 1850 global average up to the year 2090 (Arora et al., 2011). This leads to a temperature increase of approximately 2.3 °C in RCP 2.6 and of

Fig. 3. Potential natural Ecosystem Types (pnEsT) in the Kellerwald National Park (Germany).

114

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

approximately 5 °C in RCP 8.5 in the time period 1850–2090 on a global average. The RCP 8.5 scenario is the fiercest emission scenario regarded in climate change modelling (IPCC, 2013; Moss et al., 2010; van Vuuren et al., 2011). Nevertheless, the RCP 8.5 scenario is surpassed by actual observed emissions (Peters et al., 2013) and is, up to the year 2060, the most similar emission scenario so far to the widely used SRES A1B scenario (IPCC, 2013).

other three indicators addressing the development of the vegetation and related functions (habitat, net primary production, and resilience) were estimated based on the ecosystem classification. This system included not only ecosystem type-specific reference states but also potential succession states (Section 2.2).

2.7. Evaluating site-specific ESi

3.1. Ecosystem classification

The assessment of ecosystem integrity basically relies on a comparison of measured current or estimated future ecosystem states with the reference conditions of the particular ecosystem type. The impact on ecosystem integrity was graded in accordance with increasing differences between the current or future ecosystem conditions and then referenced as very low, low, medium, high, and very high. Then, in the case of ecological functions being implemented by more than one indicator, as holds true for the habitat function (2 indicators) and the nutrient flow (3 indicators), the central tendency of the multiple scores was indicated by the most frequent grade, i.e., the mode, or by the median of the grades. The same procedure was used for aggregating stepwise the impact grades from ecological indicators of the six ecological functions to one grade, thereby indicating the integrity of an ecosystem as a whole (Fig. 2). The comparison of recent ecosystem states with the respective reference can include all indicators by quantifying them by the use of measurement data. Up to now, the assessment of potential future ecosystem states only included three of the six ecological indicators for ecosystem integrity as modelled by VSD +: carbon sequestration, hygric regime, and trophic regime. Based on respective results, the

Table 2 contains an abstract from the complete list of all ecosystem types (EsT) classified (135 semi-natural forest ecosystem types, 16 semi-natural open land ecosystem types, 45 cultivated forest ecosystem types, and 18 cultivated grassland types). Each EsT was described by a specific profile noting the parameter values of the respective ecological functions that indicate the reference state. To this end, in addition to the geographic position, macroclimate and soil type information on ecological functions (Table 1) were quantified by measured and modelled values (Section 3.4). Regarding the geographic region, the soil water budget and the biogeochemical budget the EsT (Table 2) are more homogeneous than those developed by Riecken et al. (2006) and Ssymank et al. (1998).

3. Results

3.2. Map of current EsT Linking the map on potential natural vegetation (Suck et al., 2010) with the dominating EsT included in the pnV complexes enabled mapping of the potential natural ecosystem types (pnEsT) in Germany, with an example shown in Fig. 3 for the Kellerwald National Park. This

Fig. 4. Map of current semi-natural ecosystem types (csnEsT) in the Kellerwald National Park.

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

pnEsT map differentiates the potential natural vegetation according to ecosystem functions as indicated in the ecosystem classification (Section 2.2). From the pnEsT map, the map on current semi-natural ecosystem types (csnEsT) was derived and, again, is depicted in an example of the Kellerwald National Park (Fig. 4).

115

3.3. Map of EsT regions over time The CART model describing the statistical relations between the spatial patterns of current semi-natural ecosystem types (Fig. 4) with maps on elevation a.s.l., climate and soil texture (Section 2.4) yielded 44

Fig. 5. EsT-regions 1961–1990, 1991–2010, 2011–2040, and 2041–2070. Explanation: The map shows the spatial distribution of the EsT regions (=spatial EsT classes) with a grid size of 500 by 500 m for the periods 1961–1990, 1991–2010, 2011–2040 and 2041–2070. The underlying CART model consists of 44 leaf nodes (=EsT classes) with a characteristic variance and a predominant csnEsT: 9, 15 = Ea–5n–c2 Atlantic moder pine forest; 18 = Ed–3n–b1 Subcontinental raw-humus pine forest; 19, 28 = Eb–7n–D1 Hygrophilous brown mull beech forest; 20 = Eb–5n–D1 Brown mull beech forest; 24 = Eg–5n–c3 Calciphilous spruce forest; 29, 32, 48, 65, 78 = Ebc–4n–c2 Moder pine forest; 34 = Dg–5n–c2 Moder spruce forest of the mountain level; 35 = C3–6d–B2 Raw-humus spruce, fir and beech forest of the altimontane level; 36 = D2–7s–B2 Hygrophilous raw-humus fir forest of the montane level; 38 = C4–6d–Ta1N Calciphilous spruce forest of the altimontane level; 46, 63, 74 = Eb–5n–D1 Brown mull beech forest; 47 = Ebc–3n–c2 Thermophilous moder pine forest; 49, 69, 72 = Eg–5n–b1 Rawhumus spruce forest; 51, 85 = C3–7n–C2 Hygrophilous moder spruce, fir and beech forest of the altimontane level; 52, 81 = C2–6d–Ta1L Calciphilous spruce, fir and beech forest of the altimontane level; 56, 71, 73 = Eb–5n–C2 Moder beech forest on bunter; 57, 67 = D1–6d–C2 Moder beech forest of the montane level; 58, 68, 79 = Dg–5n–c2 Moder spruce forest of the montane level; 59, 60 = D2–6d–C1 Moder Douglas-fir forest of the montane level; 61 = Eg–5n–c3 Calciphilous spruce forest; 62 = Eb–5r–E2 Calciphilous mull beech forests on slopes; 76 = Eg–5n–c2 Moder spruce forest; 80, 83 = Ebc–4n–c1 Raw humus-moder pine forest; 84 = Ea–5n–c2 — Sandy moder oak and beech forest; 86 = D2–6d–C2 Moder fir and beech forest of the montane level.

116

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

Table 3 Climate change-induced spatial shifting of EsT grouped by their hypsometric and horizontal allocation. Ecoclimatic classification

1961–1990

1991–2010

2011–2040

2041–2070

B — Subalpine knee timber C — Higher montane level D — Montane level E — Lowland to lower montane level En — Northern European Ea — Atlantic Eb — Sub-Atlantic Ec — Middle European Ed — Subcontinental Ee — Sub-Mediterranean Eg — Generally

0.10% 5.37% 14.82% 79.71% 0.04% 7.02% 25.61% 16.97% 4.81% 0.05% 25.23%

0.10% 4.98% 13.93% 80.99% 0.00% 14.17% 29.65% 8.34% 4.11% 0.10% 24.63%

0.08% 4.02% 12.40% 83.50% 0.00% 10.58% 27.94% 10.34% 8.12% 0.07% 26.46%

0.05% 3.27% 10.23% 86.45% 0.00% 17.60% 34.77% 8.02% 0.39% 0.15% 25.53%

Explanation: = increase; Kendall b 0.1.

= significant increase with p-value according to Mann Kendall b 0.1;

Table 4 Climate change-induced spatial shifting of EsT grouped according to assigned Habitat Directive Annex I habitat types. Habitat type

1961–1990

1991–2010

2011–2040

2041–2070

*4070 9110 9130 *91G0 9410

0.10% 12.76% 27.01% 0.16% 1.40%

0.10% 14.54% 28.96% 0.09% 1.57%

0.08% 13.88% 29.38% 0.15% 1.18%

0.05% 16.03% 35.11% 0.02% 0.95%

Trend

Explanation: *4070 = Bushes with Pinus mugo and Rhododendron hirsutum (Mugo — Rhododendretum hirsutii); 9110 = Luzulo–Fagetum (beech forests); 9130 = Asperulo– Fagetum beech forests; 9410 = Acidophilous Picea forests of the montane to alpine levels (Vaccinio–Piceetea); *91G0 = Pannonic Woods with Quercus petraea and Carpinus betulus = significant increase with p-value according to (Tilio–Carpinetum); = increase; = significant decrease with p-value according Mann Kendall b 0.1; = decrease; to Mann–Kendall b 0.1.

spatial clusters detailed by Jenssen et al. (2013). Each of these regions contains one dominant ecosystem type and several other ecosystems with lower percentages. Nevertheless, the areal percentages of the ecosystem types joined to one region, and all cluster members feature the same relations to elevation, climate and soil texture. The application of the CART rules describing the relations between ecosystem types, elevation and soil texture to the climate projections for RCP 8.5 resulted in the maps depicting the shift of spatial patterns of ecosystem type regions due to climate change (Fig. 5). Taking regions 9 and 32 as examples, their areas changed from 9.2% and 1.1% of the German territory (1961–1990) to 28.5% and 0% (2041–2070), respectively.

= decrease;

Trend

= significant decrease with p-value according to Mann

Grouping the ecosystem types by their hypsometric and horizontal allocation (first item of the ecosystem type code, Table 2) comprehensively corroborates spatial trends (Table 3). Accordingly, the areal percentages of subalpine knee timber (B) and (high) mountain forests (C, D) was estimated to decrease from approximately 20% to 14%, while the coverage of the forest ecosystem types in the low mountains and in the lowlands is expected to increase from ca. 80% to 87% (1961–2070). Furthermore, northern European (En), subcontinental (Ed) and central European (Ec) ecosystem types potentially could decrease contrary to the (sub)Atlantic (Ea, Eb) and sub-Mediterranean (Ee) types. Most of these temporal trends were significant with p b 0.1 according to Mann–Kendall. Up to now, the weakness of this statistical analysis is that ecosystem types not yet existing in Germany, but, for instance, existing in countries with sub-Mediterranean climates (e.g., central France), should be included in further computations because they could potentially develop with ongoing climate change. Considering nature protection measures, the ecosystem types were referred to habitat types as defined by the European Habitats Directive (Table 4), enabling an analysis of the areal shifts due to climate change. Compared to the habitat types, the ecosystem types are characterized by a relatively high degree of differentiation and internal homogeneity with respect to specific aspects of ecosystem structure and functioning. For example, the ecosystem type predominantly represents a specific characteristic of a corresponding habitat type. For the habitat types listed in Table 4, the current geographical location potential for future developments of areal extent were mapped (Fig. 6). Decreases of N 80% between 1961–90 and 2041–70 were

Fig. 6. Current geographical location and potential future development of the areal extent of European Habitats Directive Annex I habitat types. Explanation: Dark grey = main occurrence, light grey = low occurrence (Ssymank et al., 1998); black = potential future areal decrease: N80% = high (LRT *4070 and 9130), N40–80% = medium (LRT 9410). The spatial allocation of the habitat type 9130 is limited to the current semi-natural ecosystem type with potential occurrence of the red listed association Galio rotundifolii—Abietetum.

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

117

Table 5 An example of ecosystem condition for one ecosystem type: Reference (left column data), measured data (2006–2009) and modelled data (2041–2070) (underlined in columns 2 ff ).

Based on measured data 2006–2009 Site:

LII–1605

Potential natural ecosystem type: Raw humus spruce forests of the altimontane level (C4–6d–B1)

Ecosystem type: Raw humus spruce forests of the altimontane level

Ecosystem type code: C4–6d–B1

Habitat type BfN: 44.03.02.01

Impact onecosystem integrity Very low

Low

Medium

High

Very high

High

Very high

Habitat function Very low

Low

Medium Kullback distance

0–0.52

0.53–1.39

1.40–2.26

2.27–3.13

3.13–4.0

Similarity of mean cover ratio to reference conditions (plant species) 65–100

49–64

33–48

16–32

0–15

Net primary production Very low

Low

Medium

High

Very high

Mean net primary production of tree wood in t TS/ha ≥2.2

1.6–1.1

1.7–2.1

0.6–1.0

0–0.5

Carbon sequestration Very low

Low

Medium

High

Very high

Corg concentration in the humus / surface soil layer and the uppermost 80 cm of the soil ≥80

60–79

40–59

20–39

0–19

Nutrient flow Very low

Low

Medium

High

Very high

pH value in the uppermost 5 cm of the soil (including litter) 2.65–2.87

2.49–3.40

2.33–3.94

2.17–4.47

2.00–5.00

Base saturation in the uppermost 5 cm of the soil (including litter) 12.9–19.9

12.2–24.9

11.5–30.0

10.7–35.0

10–40

C/N ratio in the uppermost 5 cm of the soil (including litter) 29.2–26.2

30.7–23.2

1.32–1.36

1.24–1.57

32.1–20.1

33.6–17.1

35–14

1.08–1.99

1.0–2.2

0.08–0.29

0.06–0.30

0.29–1.27

0.2–1.4

N concentration in leaves and needles (1–year–old shoots) 1.16–1.78 P concentration in leaves and needles (1–year–old shoots) 0.14–0.24

0.12–0.26

0.10–0.27 K concentration in leaves and needles (1–year–old shoots)

0.54–0.88

0.46–1.01

0.37–1.14

Ca concentration in leaves and needles (1–year–old shoots) 0.62–0.72

0.52–0.89

0.13–0.19

0.11–0.22

0.41–1.06

0.31–1.23

0.2–1.4

Mg concentration in leaves and needles (1–year–old shoots) 0.10–0.25

0.08–0.27

0.06–0.30

Water flow Very low

Low

Medium

High

Very high

Soil moisture index 5.1–6.5

4.7–6.9

4.3–7.3

3.9 –7.6

3.5–8.0

Resilience Very low

Low

Medium

High

Very high

118

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

Table 5 (continued)

Similarity of tree species composition to potential natural vegetation 60–100

45–59

30–44

15–29

0–14

Based on modelled data 2041–2070 Nutrient flow Very low

Low

Medium

High

Very high

pH value in the uppermost 5 cm of the soil (including litter) 2.65–2.87

2.49–3.40

2.33–3.94

2.17–4.47

2.00–5.00

Base saturation in the uppermost 5 cm of the soil (including litter) 12.9–19.9

12.2–24.9

11.5–30.0

10.7–35.0

10–40

C/N ratio in the uppermost 5 cm of the soil (including litter) 29.2–26.2

30.7–23.2

32.1–20.1

classified as critical, and decreases of N 40 to 80%, as medium. From this analysis, it becomes clear that by the year 2070, the areal coverage of priority habitat types *4070 “Bushes with Pinus mugo and Rhododendron hirsutum (Mugo — Rhododendretum hirsutii)” (p N 0.1), *91G0 = Pannonic Woods with Quercus petraea and Carpinus betulus (TilioCarpinetum) (p N 0.1), and 9410 “Acidophilous Picea forests of the montane to alpine levels (Vaccinio-Piceetea)” (p b 0.1) are expected to decrease. By contrast, habitat type 9110 “Luzulo-Fagetum (beech forests)” and 9130 “Asperulo-Fagetum (beech forests)” (p N 0.1) presumably will increase until 2070. For the association “Galio rotundifolii—Abietetum” as a special characteristic of 9130 and part of the German Red List of Plant Communities (Rennwald, 2000), a significant decrease could be determined (p b 0.1). All these trends corresponded clearly with a loss of areas in the montane and subalpine zone and a respective increase in ecosystem types in the low mountains and lowlands (Table 3). At first glance, the decline of the priority habitat type *91G0 “Pannonian woods with Quercus petraea and Carpinus betulus” (p b 0.1) might be surprising considering the respective typical environmental conditions (Ssymank et al., 1998). The major reason could be found in the increase in the mean annual precipitation from 549–576 mm/a (1961–91) to 779 mm/a (2041–70), thus accounting for at least 203 mm/a. This corresponds to fundamental climate change from subcontinental to sub-Atlantic conditions in certain parts of Germany (Table 4). 3.4. Evaluation of site-specific ESi For one of the seven forest ecosystem sites investigated (Fig. 1), ecosystem type C4–6d–B1 (Raw humus spruce forests of the altimontane level), some results derived from a site-specific evaluation of ecosystem integrity are given as an example in Table 5. This example evaluation was quantified with data collected between 2006 and 2009 and with VSD+ modelling results from the period 2041–2070 (scenario RCP 8.5, 15 kg N/ha a). The value intervals in the left column indicate the reference state of the respective indicator. In the case that observed and modelled values, which are indicated underlined, coincided with the reference interval, the particular ecological function was not impacted or was only slightly impacted. According to the similarity between monitored values and values modelled from the reference, grades for impacts on ecosystem integrity were assigned as very low, low, medium, high or very high. When comparing the data recorded with the potential future values modelled, it becomes clear that potential future modifications of ecosystem functions/ structures could affect the ecosystem integrity both positively (base saturation, C/N ratio) and negatively (pH value).

33.6–17.1

35–14

A comprehensive overview derived from the seven site-specific assessments of ecosystem integrity is given in Table 6. At three of the seven sites evaluated, as shown in columns 2 and 3, the current ecosystem type corresponds to the potential natural ecosystem types. Concerning potential future ecosystem types, the changes in structures and functions were classified as either negative (13 out of 14 cases) or positive (1 out of 14 cases), as considerably (4 out of 13 negative cases) or moderately (2 out of 13 negative cases) significant, or as not significant (7 out of 13 negative cases). The estimations of future ecosystem types relied on scenarios RCP 2.6 (5 and 15 kg N/ha a) and RCP 8.5 (5 and 15 kg N/ha a), except in the case of site LII-1405. Here, significant changes in the herb layer were only found for the scenario RCP 8.5 (15 kg N/ha a), whereas the other scenarios did not indicate any changes. The potential future conditions of the W.I.E. sites Kahlenberg 75 (current ecosystem type: Raw humus-moder pine forests), Biesenthal 1534 a (current ecosystem type: Sandy moder sessile oak and beech forests) and Peitz 150 (current ecosystem type: Subcontinental raw humus pine forests) are not expected to change significantly negatively (Table 6). The proportion of broadleaved trees should be increased on all sites to strengthen their adaptability to climate change and to reduce forest fire risk, which is expected to rise in Eastern parts of Germany (Lindner et al., 2008; Linke et al., 2010). Impacts affecting the current EsT Raw humus pine forests at LII site 1405 could be moderate but significantly negative due to the high atmospheric N deposition leading to a disharmonious nutritional condition for the occurring Pinus sylvestris. An increased proportion of broadleaved trees would counteract this. On LII site 1602 (current EsT: Moder spruce, fir and beech forests of the altimontane level) a decline or even loss of plant species is possible if the atmospheric N deposition stays at a high level. LII site 1605 (current ecosystem type: Raw humus spruce forests of the altimontane level) will undergo significantly negative functional/structural changes. Increasing acidification of the top soil of the current moder spruce forests (LII site 1609) is expected due to the humid climate and the planting of non-indigenous spruce. A higher proportion of broadleaved trees could counteract the acidification. 4. Discussion 4.1. General appraisal Major aims of ecology are to quantitatively describe and explain the condition of ecosystems and try to project potential future developments of the respective ecosystems. To this end, data are collected to

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

119

Table 6 Potential natural, current and future EsT at 7 sites and an evaluation of the changes. Site

Potential natural EsT

Current EsT (1991–2010)

LII-1405

Sandy moder sessile oak and beech forests (Eb–4n–C2)

Raw humus pine forests (Eb–4n–b1)

LII-1602

Moder spruce, fir and beech forests of the altimontane level (C3–6d–C2) Raw humus spruce forests of the altimontane level (C4–6d–B1)

LII-1605 LII-1609

W.I.E.-Kahlenberg 75 W.I.E.-Biesenthal 1534 a W.I.E.-Peitz 150

Moder fir and beech forests of Moder spruce forests of the the montane level montane level (Dg–5n–c2) (D2–6d–C2) Moder beech forests on bunter Raw humus-moder pine (Eb–5n–C2) forests (Ebc–4n–c1) Sandy moder sessile oak and beech forests (Eb–4n–C2) Moder pine and sessile oak forests (Ed–3n–C2)

Subcontinental raw humus pine forests (Ed–3n–b1)

Future EsT (2070) Scenario RCP 2.6

Scenario RCP 8.5

Raw humus pine forests (Eb–4n–b1)a–c

Raw humus pine forests (Eb–4n–b1)a,c (Eb–4n–b1)b,d Moder beech forests on bunter (Eb–5n–C2)a,b,e Raw humus spruce forests of the montane level (Dg–5n/6d–b1)a,b,e Moder spruce forests (Eg–5n–c2)a,b,d

Moder fir and beech forests of the montane level (D2–6d–C2)a,b,e Raw humus spruce forests of the montane level (Dg–5n/6d–b1)a,b,e Moder spruce forests of the montane level (Dg–5n–c2)a–c Raw humus-moder pine forests (Ebc–4n–c1)a–c Sandy moder sessile oak and beech forests (Eb–4n–C2)a–c Raw humus pine forests (E–3n–b1)a–c

Subcontinental raw humus-moder pine forests (Ed–3/4n–c1)a–c Moder oak-hornbeam forests (Ec–4n–C2)a,b,f Subcontinental raw humus-moder pine forests (Ed–3n–c1)a–c

Evaluation of changes in ecological functions/structures. a Scenario 5 kg/ha a N deposition after 2010. b Scenario 15 kg/ha a N deposition after 2010. c Negative (not significant). d Negative, moderately significant. e Negative, considerably significant. f Positive (significant).

verify hypotheses as well as methods for collecting and evaluating data, for verification of hypotheses and for projecting potential future developments. Dynamic and statistical models can support these objectives. Models prescind from the real system which they represent. Among others, the degree of abstraction depends on the knowledge about the modelled system, the respective data that are available and the modelling technique (Fränzle et al., 2008; Jørgensen and Bendoricchio, 2001; White et al., 1992). As a matter of course, this holds fully true for this investigation and generally defines the validity and restrictions of the methodology presented in this paper. A general problem is also given by the fact that environmental monitoring programmes often are not adapted sufficiently to the models developed and vice versa (Schröder et al., 1996). Nevertheless, even if all these boundary conditions were to be optimal, it does not seem realistic that environmental sciences one day would be able to quantify the complexity of single ecosystems holistically or the spatial coverage of ecosystems and their future development. Therefore, as an empirical science, ecology relies on a stepwise approach that reflects structural and functional representativeness. This can be reached by focusing on selected ecosystems and representative ecosystem types and quantifying their essential structures and functions. To this end, indicators that are meaningful representatives for regarded structures and functions of the ecosystem(s) need to be investigated (Egoh et al., 2012; Fränzle et al., 2008; Schröder et al., 1996, 2004). Finally, the verification of inductive reasoning, as realized with hypothesis testing or spatial and temporal estimations in terms of interpolation and, subsequently, with predictions/projections, is disputed in the philosophy of science (Popper and Miller, 1983). The modelling approaches presented in Sections 2.4/3.3 and 2.6/3.4 make use of climate projections. In contrast to meteorological predictions (short- to midterm weather forecasts), which attempt to predict what actually will happen on the basis of current information, climate projections such as those calculated with STARS (Section 2.6) describe what will happen under a set of assumptions on changes in boundary conditions (scenarios), such as an increase in greenhouse gases, which might influence the future climate. Thus, projections are conditional expectations about the likelihood that something will happen several decades to centuries in the future if certain influential conditions develop (IPCC, 2013). The same applies for any modelling that makes use of climate projections and thus for the predictive mapping of potential patterns of ecosystem

type regions across time and the generating of potential developments of site-specific ecosystem integrity indicators. 4.2. Specific appraisal Beyond this basic appraisal of the approach presented, specific methodological considerations are discussed. First, it should be noted that the most comprehensive dataset for mapping ecosystems at the EU level is Corine Land Cover. This should be complemented by the best available data from sub-national and national data sources at appropriate scales (EU, 2014). Maps are spatial models of reality, and thus, from the above considerations, it follows that mapping of ecosystems needs abstracts from ecological reality, which are gained by classifying ecosystems and quantifying their structures and functions by ecological indicators and related biological, chemical and physical quantities (Figs. 1, 2; Table 1). The ecosystem classification developed and presented in this article is, as far as the authors are aware, unique regarding the size of the dataset used. Based on this assessment, the diversity of ecosystem coverage of Germany was ‘reduced’ to types described quantitatively by well-distinguished intervals of quantities indicating basic ecological functions. Each combination of indicator features characterized a specific ecosystem type, ensuring a sufficient homogeneity with respect to ecosystem structures and functions (Section 3.1). The databases used allow the quantifying of intervals of indicator values as reference for site-specific evaluations of ecosystem integrity. Considering other investigations that aim to analyse ecosystem integrity (Andreasen et al., 2001; Balvanera et al., 2006; Cairns, 1977; Carignan and Villard, 2002; EEA, 2012; Egoh et al., 2012; Faber-Langendoen et al., 2012a,b; EU, 2014; Petter et al., 2013; Stoll et al., 2015; Tierney et al., 2009; Vose et al., 2012; Wulff et al., 2012), the mapping and analysis of ecosystem types across space and time in combination with ecosystem type-/site-specific reference intervals for functional and structural indicators broadens and deepens ecosystem integrity assessments. Climate change and N deposition are significant driving forces shifting ecosystems toward modified or even completely changed states. As shown in this study, depending on the intensity, the duration, and the combination of climate and deposition impacts, ecosystem states may be shifted within the type-specific parameter intervals. But ecosystem conditions may also cross the interval boundaries of one or several indicators, giving rise to other ecosystem types that have

120

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

been observed at other sites or even never before in the region (Sections 3.3, 3.4). Linking site-specific assessments of ecosystem integrity via ecosystem types with dynamic modelling of quantities indicating functions of forest ecosystems at specific sites could prove to be a useful tool for estimating potential future ecosystem development and for deriving critical limits of driving forces. Dynamic modelling enables the projection of potential future ecosystem condition considering different climate projections and atmospheric N deposition scenarios. This is – together with the predictive mapping and the estimation of ecosystem integrity – important for addressing, in a precautionary manner, nature and environmental protection goals (Jenssen et al., 2013; Kriebel et al., 2001). By using the modelling results, whether a specific ecosystem type will be impacted or can be expected to be impacted by climate change and N deposition can be estimated. If impacts are predicted, countermeasures can be taken at an early stage and specified for the respective functional and structural indicators. Even if the modelling results of this investigation are not understood as strict forecasts but as projections, they reveal a loss of vertical zoning and an increase in the Atlantic and sub-Mediterranean zone and their climatic characteristics. The latter is mainly due to the increase in intensity and duration of dry phases in the summer, while decreasing continentality is indicated by increases in precipitation and temperature in the winter. Comparisons of the spatial distribution of the EsT regions between 1961–90 and 2041–70 enable mapping of climate change risks for European Habitats Directive Annex I habitat types. For instance, this can be used for selecting sample sites to observe the conservation status according to Article 11 of the Habitats Directive. The spatial analyses presented revealed that a gradual increase in temperature will alter the growing environment of certain tree species, reducing the growth of some species (especially in dry forests) and increasing the growth of others (especially in high-elevation forests). Mortality may increase in older forests stressed by low soil moisture, and regeneration may decrease for species affected by low soil moisture and competition with other species during the seedling stage. Species habitats will move upward in elevation and northward in latitude and will be reduced in current habitats at lower elevations and lower latitudes. Nevertheless, due to the high genetic diversity of most tree species, which enables a tolerance of a broad range of environmental conditions such as temperature, tree growth and regeneration may be affected more by extreme weather events and climatic conditions than by gradual changes. Such a biophysical disturbance might change the structure and function of ecosystems across large areas over a short period of time. Increased atmospheric N deposition will potentially alter the physiological function and productivity of forest ecosystems (Vose et al., 2012). Therefore, further modelling should include additional structural and functional indicators and boundary conditions such as extreme weather events. To give an example, in Germany during the last several decades, 20–60 tornados were verified yearly, with a maximum frequency in July. Tornado-related windthrow may change the eco-climate in impacted forest stands. Due to this factor, decomposition and soil chemistry may be affected and change the integrity of respective forest ecosystems (Seidl et al., 2011). The dynamics and magnitude of climate change could exceed the stability of forest ecosystems, and novel ecosystems without historical analogues could develop (Vose et al., 2012). Acknowledgements The investigation was funded by the Federal Environment Agency (UFOPLAN FKZ 3710 83 214) (Umweltbundesamt) of Germany. References Allen, C.D., 2009. Climate induced forest dieback: an escalating global phenomenon? Unasylva 60 (1–2), 43–49 (231/232).

Allen, C.D., Macalady, A.K., Chenchouni, H., Bachelet, D., McDowell, N., Vennetier, M., Kitzberger, T., Rigling, A., Breshears, D.D., Hogg, E.H., Gonzalez, P., Fensham, R., Zhang, Z., Castro, J., Demidova, N., Lim, J.-H., Allard, G., Running, S.W., Semerci, A., Cobb, N., 2010. Drought-induced forest mortality: a global overview reveals emerging climate change risks. For. Ecol. Manag. 259 (4), 660–684. Andreasen, J.K., O'Neill, R.V., Noss, R., Slosser, N.C., 2001. Considerations for the development of a terrestrial index of ecological integrity. Ecol. Indic. 1, 21–35. Angermeier, P.L., Karr, J.R., 1994. Biological integrity versus biological diversity as policy directives. Protecting biotic resources. Bioscience 44 (10), 690–696. Arora, V.K., Scinocca, J.F., Boer, G.J., Christian, J.R., Denman, K.L., Flato, G.M., Kharin, V.V., Lee, W.G., Merryfield, W.J., 2011. Carbon emission limits required to satisfy future representative concentration pathways of greenhouse gases. Geophys. Res. Lett. 38, 1–6. Balvanera, P., Pfisterer, A.B., Buchmann, N., He, J.S., Nakashizuka, T., Raffaelli, D., Schmid, B., 2006. Quantifying the evidence for biodiversity effects on ecosystem functioning and services. Ecol. Lett. 9, 1146–1156. Bi, D., Dix, M., Marsland, S.J., O'Farrell, S., Rashid, H.A., Uotila, P., Hirst, A.C., Kowalczyk, E., Golebiewski, M., Sullivan, A., Yan, H., Hannah, N., Franklin, C., Sun, Z., Vohralik, P., Watterson, I., Zhou, X., Fiedler, R., Collier, M., Ma, Y., Noonan, J., Stevens, L., Uhe, P., Zhu, H., Griffies, S.M., Hill, R., Harris, C., Puri, K., 2013. The ACCESS coupled model: description, control climate and evaluation. Aust. Meteorol. Oceanogr. J. 63, 9–32. Bobbink, R., Hicks, K., Galloway, J., Spranger, T., Alkemade, R., Ashmore, M., Bustamante, M., Cinderby, S., Davidson, E., Dentener, F., Emmett, B., Erisman, J.W., Fenn, M., Gilliam, F., Nordin, A., Pardo, L., de Vries, W., 2010. Global assessment of nitrogen deposition effects on terrestrial plant diversity: a synthesis. Ecol. Appl. 20 (1), 30–59. Bohn, U., Neuhäusl, R., Gollub, G., Hettwer, C., Neuhäuslová, Z., Schlüter, H., Weber, H., 2000/2003. Map of the Natural Vegetation of Europe. Scale 1:2.500.000. Part 1: Explanatory Text: 1-655. Part 2: Legend:1-153. Part 3: Maps (9 sheets, Legend sheet, General Map 1:10 million). Landwirtschaftsverlag, Münster. Bohn, U., Hettwer, C., Gollub, G. (Eds.), 2005. Application and analysis of the map of the natural vegetation of Europe. Bundesamt für Naturschutz, Bonn. BfN-Skripten 156, pp. 1–452. Bonten, L., Posch, M., Reinds, G.J., 2009. The VSD+ soil acidification model. Model Description and User Manual. Version 0.11pp. 1–19. Breiman, L., Friedman, J., Olshen, R., Stone, C., 1984. Classification and Regression Trees. Wadsworth, Belmont, CA. Cairns, J., 1977. Quantification of biological integrity. In: Ballentine, R.K., Guarraia, L.J. (Eds.), The Integrity of Water. U.S. Environmental Protection Agency, Office of Water and Hazardous Materials, Washington, D.C., USA, pp. 171–187. Carignan, V., Villard, M.-A., 2002. Selecting indicator species to monitor ecological integrity: a review. Environ. Monit. Assess. 78, 45–61. Chase, J.M., Knight, T.M., 2013. Scale-dependent effect sizes of ecological drivers on biodiversity: why standardised sampling is not enough. Ecol. Lett. 16, 17–26. Chave, J., 2013. The problem of pattern and scale in ecology: what have we learned in 20 years? Ecol. Lett. 16, 4–16. De Leo, G.A., Levin, S., 1997. The multifaced aspects of ecosystem integrity. Conserv. Biol. Online 1 (1), 3. de Vries, W., Posch, M., 2011. Modelling the impact of nitrogen deposition, climate change and nutrient limitations on tree carbon sequestration in Europe for the period 1900– 2050. Environ. Pollut. 159, 2289–2299. de Vries, W., Dobbertin, M.H., Solberg, S., van Dobben, H.F., Schaub, M., 2014. Impacts of acid deposition, ozone exposure and weather conditions on forest ecosystems in Europe: an overview. Plant Soil 380, 1–45. Doherty, M., Kearns, A., Barnett, G., Sarre, A., Hochuli, D., Gibb, H., Dickman, C., 2000. The Interaction Between Habitat Conditions, Ecosystem Processes and Terrestrial Biodiversity — A Review. State of the Environment, Second Technical Paper Series (Biodiversity), Department of the Environment and Heritage, Canberra, Australia, pp. 1–114. EEA (European Environment Agency), 2012. Climate change, impacts and vulnerability in Europe 2012. An indicator-based report. EEA Report No 12/2012, pp. 1–300. EEC (European Economic Community), 1992. Council directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora. Off. J. L 206, 7–50. Egoh, E., Drakou, E.G., Dunbar, M.B., Maes, J., Willemen, L., 2012. Indicators for Mapping Ecosystem Services: A Review. Joint Research Centre – Institute for Environment and Sustainability. EUR – Scientific and Technical Research Series. Publications Office of the European Union, Luxembourg, pp. 1–111. EU (European Commission), 2014. Mapping and assessment of ecosystems and their services indicators for ecosystem assessments under Action 5 of the EU Biodiversity Strategy to 2020. 2nd Report — Final, February 2014pp. 1–80. EUROSTAT, 2012. Energy, transport and environment indicators — EUROSTAT Pocketbooks. EUROSTAT, European Commission. Publications Office of the European Union, Luxembourg http://dx.doi.org/10.2785/19616. Faber-Langendoen, D., Hedge, C., Kost, M., Thomas, S., Smart, L., Smyth, R., Drake, J., Menard, S., 2012a. Assessment of wetland ecosystem condition across landscape regions: a multi-metric approach. Part A. Ecological Integrity Assessment Overview and Field Study in Michigan and Indiana. EPA/600/R-12/021a. U.S. Environmental Protection Agency Office of Research and Development, Washington, DC. Faber-Langendoen, D., Rocchio, J., Thomas, S., Kost, M., Hedge, C., Nichols, B., Walz, K., Kittel, G., Menard, S., Drake, J., Muldavin, E., 2012b. Assessment of wetland ecosystem condition across landscape regions: a multi-metric approach. Part B. Ecological Integrity Assessment Protocols for Rapid Field Methods (L2). EPA/600/R-12/021b. U.S. Environmental Protection Agency Office of Research and Development, Washington, DC. FAO, 2010. Global Forest Resources Assessment 2010 — Main Report. Food and Agriculture Organisation of the United Nations (FAO), Rome, pp. 1–340. FAO (Food and Agriculture Organization of the United Nations), 2009. Adapting to climate change. Unasylva 60 (231/232), 1–93.

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122 Feldhoff, J.H., Lange, S., Volkholz, J., Donges, J.F., Kurths, J., Gerstengarbe, F.-W., 2014. Complex networks for climate model evaluation with application to statistical versus dynamical modeling of South American climate. Clim. Dyn. http://dx.doi.org/10. 1007/s00382-014-2182-9. FOREST EUROPE, UNECE, FAO, 2011. State of Europe's Forests 2011 — Status and Trends in Sustainable Forest Management in Europe. FOREST EUROPE, UNECE and FAO, Oslo, pp. 1–337. Fränzle, O., 1994. Thermodynamic aspects of species diversity in tropical and ectropical plant communities. Ecol. Model. 75 (76), 63–70. Fränzle, O., Kappen, L., Blume, H.P., Dierßen, K. (Eds.), 2008. Ecosystem organization of a complex landscape. Long-term research in the Bornhöved Lake District, Germany. Ecological Studies 202. Springer, Heidelberg, pp. 1–392. Gobiet, A., Kotlarski, S., Beniston, M., Heinrich, G., Rajczak, J., Stoffel, M., 2014. 21st century climate change in the European Alps — a review. Sci. Total Environ. 493, 1138–1151. Haines-Young, R.H., Potschin, M.P., 2010. The links between biodiversity, ecosystem services and human well-being. In: Raffaelli, D., Frid, C. (Eds.), Ecosystem Ecology: New Synthesis. BES Ecological Reviews Series. CUP, Cambridge, UK, pp. 110–139. Hassan, R., Scholes, R., Ash, N. (Eds.), 2005. Ecosystems and human wellbeing: current state and trends. Findings of the Condition and Trends Working Group of the Millennium Ecosystem Assessment volume 1. Island Press, Washington D.C. Hofmann, G., 1997. Mitteleuropäische Wald- und Forstökosystemtypen in Wort und Bild. 2 ed. AFZ - Der Wald. Deutscher Landwirtschaftsverlag. 2. erweiterte Auflage, Münchenpp. 1–85. Hofmann, G., 2002. Entwicklung der Waldvegetation des nordostdeutschen Tieflandes unter den Bedingungen steigender Stickstoffeinträge in Verbindung mit Niederschlagsarmut. Mit Anlagen. In: Anders, S., et al. (Eds.), Ökologie und Vegetation der Wälder Nordostdeutschlands 24–41. Dr. Kessel, Oberwinter, pp. 201–283 (www. forstbuch.de). Hofmann, G., Passarge, H., 1964. Über Homogenität und Affinität in der Vegetationskunde. Arch. Forstwes. 13, 1119–1138. Holyoak, M., Hochberg, M., 2013. Ecological effects of environmental change. Ecol. Lett. 16, 1–153 (Special Issue). Hooper, D.U., Chapin, F.S., Ewel, J.J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J.H., Lodge, D.M., Loreau, M., Naeem, S., Schmid, B., Setala, H., Symstad, A.J., Vandermeer, J., Wardle, D.A., 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecol. Monogr. 75, 3–35. IPCC (Intergovernmental Panel on Climate Change), 2013. Climate change 2013. The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK and New York, NY, USA, pp. 1–1535. Ives, A.R., Carpenter, S.R., 2007. Stability and diversity of ecosystems. Science 317, 58–62. Jenssen, M., 2010. Modellierung und Kartierung räumlich differenzierter Wirkungen von Stickstoffeinträgen in Ökosysteme im Rahmen der UNECE-Luftreinhaltekonvention. Teilbericht IIII: Modellierung der Wirkung der Stickstoff-Deposition auf die biologische Vielfalt der Pflanzengesellschaften von Wäldern der gemäßigten Breiten. UBA-Texte 09/2010. Dessau-Roßlaupp. 1–50. Jenssen, M., Hofmann, G., 2003. Die Quantifizierung ökologischer Potentiale der Phytodiversität und Selbstorganisation der Wälder. Beitr. Forstwirtsch. Landschaftsökol. 39 (3), 132–141. Jenssen, M., Hofmann, G., Nickel, S., Pesch, R., Riediger, J., Schröder, W., 2013. Bewertungskonzept für die Gefährdung der Ökosystemintegrität durch die Wirkungen des Klimawandels in Kombination mit Stoffeinträgen unter Beachtung von Ökosystemfunktionen und -dienstleistungen. UBA-Texte 87/2013. Dessau, Textband + 9 Anhängepp. 1–381. Jørgensen, S.E., Bendoricchio, G., 2001. Fundamentals of Ecological Modelling. 3rd ed. Elsevier, Oxford, UK, pp. 1–430. Karr, J.R., Dudley, D.R., 1981. Ecological perspective on water quality goals. Environ. Manag. 5 (1), 55–68. Kriebel, D., Tickner, J., Epstein, P., Lemons, J., Levins, R., Loechler, E.L., Quinn, M., Rudel, R., Schettler, T., Stoto, M., 2001. The precautionary principle in environmental science. Environ. Health Perspect. 109, 871–876. Kullback, S., 1951. Information Theory and Statistics. Wiley, New York. Limburg, K.E., Levin, S.A., Harwell, C.C., 1986. Ecology and estuarine impact assessment: lesson learned from the Hudson River (USA) and other estuarine experiences. J. Environ. Manag. 22, 255–280. Lindenmayer, D.B., Franklin, J.F., 2002. Conserving Forest Biodiversity: A Comprehensive Multiscaled Approach. Island Press, Washington, DC, pp. 1–351. Lindner, M., Garcia-Gonzalo, J., Kolström, M., Green, T., Reguera, R., Maroschek, M., Seidl, R., Lexer, M.J., Netherer, S., Schopf, A., Kremer, A., Delzon, S., Barbati, A., Marchetti, M., Corona, P., 2008. Impacts of climate change on European forests and options for adaptation. AGRI-2007-G4-06, Report to the European Commission DirectorateGeneral for Agriculture and Rural Development. Brusselspp. 1–173. Linke, C., Grimmert, S., Hartmann, I., Reinhardt, K., 2010. Auswertung regionaler Klimamodelle für das Land Brandenburg – Darstellung klimatologischer Parameter mit Hilfe vier regionaler Klimamodelle (CLM, REMO10, WettReg, STAR2) für das 21. Jahrhundert. Fachbeiträge des Landesumweltamtes des Landes Brandenburg 113, Potsdampp. 1–305. Loreau, M., de Mazancourt, C., 2013. Biodiversity and ecosystem stability: a synthesis of underlying mechanisms. Ecol. Lett. 16, 106–115. Luedeling, E., Guo, L., Dai, J., Leslie, C., Blanke, M.M., 2013. Differential responses of trees to temperature variation during the chilling and forcing phases. Agric. For. Meteorol. 181, 33–42. Lutz, J., Gerstengarbe, F.-W., 2014. Improving seasonal matching in the STARS model by adaptation of the resampling technique. Theor. Appl. Climatol. http://dx.doi.org/10. 1007/s00704-014-1205-0. Manion, P.D., 1991. Tree Disease Concepts. Prentice-Hall, Englewood Cliffs.

121

Manuel-Navarete, D., Kay, J.J., Dolderman, D., 2004. Ecological integrity discourses: linking ecology with cultural transformation. Hum. Ecol. Rev. 11 (3), 215–229. Maroschek, M., Seidl, R., Netherer, S., Lexer, M.J., 2009. Climate change impacts on goods and services of European mountain forests. Unasylva 69 (1–2), 76–80 (231/232). Matyssek, R., Wieser, G., Calfapietra, C., de Vries, W., Dizengremel, P., Ernst, D., Jolivet, Y., Mikkelsen, T.N., Mohren, G.M.J., Le Thiecj, D., Tuovinen, J.P., Weatherall, A., Paoletti, E., 2012. Forests under climate change and air pollution: gaps in understanding and future directions for research. Environ. Pollut. 169, 57–65. Maynard, S., James, D., Davidson, A., 2010. The development of an ecosystem services framework for South East Queensland. Environ. Manag. 45 (5), 881–895. Maynard, S., James, D., Davidson, A., 2011. An adaptive participatory approach for developing an ecosystem services framework for South East Queensland, Australia. Int. J. Biodivers. Sci. Ecosyst. Serv. Manag. 7 (3), 182–189. Midgley, G.F., 2012. Biodiversity and ecosystem function. Science 335 (6065), 174–175. Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S., 2010. The next generation of scenarios for climate change research and assessment. Nature 463, 747–756. Nagajyoti, P.C., Lee, K.D., Sreekanth, T.V.M., 2010. Heavy metals, occurrence and toxicity for plants: a review. Environ. Chem. 8, 199–216. Orlowsky, B., Gerstengarbe, F.W., Werner, P.C., 2008. A resampling scheme for regional climate simulations and its performance compared to a dynamical RCM. Theor. Appl. Climatol. 92 (3–4), 209–223. Österle, H., Gerstengarbe, F.-W., Werner, P.C., 2006. Ein neuer meteorologischer Datensatz für Deutschland, 1951–2003, Proceedings der 7. Deutschen Klimatagung 2006. Klimatrends: Vergangenheit und Zukunft. Meteorologisches Institut der LudwigMaximilians-Universität, München, pp. 1–3. Paoletti, E., Schaub, M., Matyssek, R., Wieser, G., Augustaitis, A., Bastrup-Birk, A.M., Bytnerowicz, A., Günthardt-Goerg, M.S., Müller-Starck, G., Serengil, Y., 2010. Advances of air pollution science: from forest decline to multiple-stress effects on forest ecosystem services. Environ. Pollut. 158, 1986–1989. Parrish, J.D., Braun, D.P., Unnasch, R.S., 2003. Are we conserving what we say we are? Measuring ecological integity within protected areas. Bioscience 53, 851–860. Pesch, R., Schröder, W., 2006. Mosses as bioindicators for metal accumulation: statistical aggregation of measurement data to exposure indices. Ecol. Indic. 6, 137–152. Peters, G.P., Andrew, R.M., Boden, T., Josep Canadell, G., Ciais, P., Le Quéré, C., Marland, G., Raupach, M.R., Wilson, C., 2013. The challenge to keep global warming below 2°C. Nat. Clim. Chang. 3, 4–6. Petter, M., Mooney, S., Maynard, S., Davidson, A., Cox, M., Horosak, I., 2013. A methodology to map ecosystem functions to support ecosystem services assessments. Ecol. Soc. 18 (1), 31. Popper, K.R., Miller, D.W., 1983. A proof of the impossibility of inductive probability. Nature 302, 687–688. Posch, M., Reinds, G.J., 2009. A very simple dynamic soil acidification model for scenario analyses and target load calculations. Environ. Model. Softw. 24, 329–340. Posch, M., Hettelingh, J.P., Slootweg, J., 2003. Manual for dynamic modelling of soil response to atmospheric deposition. Coordination Center for Effects. RIVM Report 259101012, Bilthoven, The Netherlands, pp. 1–69. Rennwald, E., 2000. Verzeichnis und Rote Liste der Pflanzengesellschaften Deutschlands. Schriftenr. Vegetationsk. 35, 1–800. Reyer, C., Lasch-Born, P., Suckow, F., Gutsch, M., Murawski, A., Pilz, T., 2013. Projections of regional changes in forest net primary productivity for different tree species in Europe driven by climate change and carbon dioxide. Ann. For. Sci. http://dx.doi. org/10.1007/s13595-013-0306-8. Richardson, A.D., Keenana, T.F., Migliavacca, M., Ryu, Y., Sonnentag, O., Toomey, M., 2013. Climate change, phenology, and phenological control of vegetation feedbacks to the climate system. Agric. For. Meteorol. 169, 156–173. Riecken, U., Finck, P., Raths, U., Schröder, E., Ssymank, A., 2006. Rote Liste der gefährdeten Biotoptypen Deutschlands. Naturschutz Biol. Vielfalt 34, 1–318. Saint-Béat, B., Baird, D., Asmus, H., Asmus, R., Bacher, C., Pacella, S.R., Johnson, G.A., David, V., Vézina, A.F., Niquil, N., 2015. Trophic networks: how do theories link ecosystem structure and functioning to stability properties? A review. Ecol. Indic. 52, 458–471. Schröder, W., Fränzle, O., Keune, H., Mandry, P. (Eds.), 1996. Global Monitoring of Terrestrial Ecosystems. Ernst und Sohn, Berlin, pp. 1–281. Schröder, W., Pesch, R., Schmidt, G., 2004. Soil monitoring in Germany. Spatial representativity and methodical comparability. J. Soils Sediments 4, 49–58. Seidl, R., Fernandes, P.M., Fonseca, T.F., Gillete, F., Jönsson, A.M., Merganičováh, K., Netherer, S., Arpaci, A., Bontemps, J.-D., Bugmann, H., González-Olabarria, J.R., Lasch, P., Meredieu, C., Moreira, F., Schelhaas, M.-J., Mohren, F., 2011. Modelling natural disturbances in forest ecosystems: a review. Ecol. Model. 222 (4), 903–924. Simpson, D., Andersson, C., Christensen, J.H., Engardt, M., Geels, C., Nyiri, A., Posch, M., Soares, J., Sofiev, M., Wind, P., Langner, J., 2014. Impacts of climate and emission changes on nitrogen deposition in Europe: a multi-model study. Atmos. Chem. Phys. 14, 6995–7017. Slootweg, J., Posch, M., Hettelingh, J.P., Mathijssen, L., 2015. Modelling and mapping the impacts of atmospheric deposition on plant species diversity in Europe. Status Report 2014. Coordination Center for Effects. RIVM Report 2014-075, Bilthoven, Netherlands, pp. 1–160. Ssymank, A., Hauke, U., Rückriem, C., Schröder, E., 1998. Das europäische Schutzgebietssystem Natura 2000. BfN-Handbuch zur Umsetzung der Fauna-FloraHabitat-Richtlinie und der Vogelschutz-Richtlinie. Schriftenr. Landschaftspflege Naturschutz (53), 1–560. Stankovic, S., Kalaba, P., Stankovic, A.R., 2014. Biota as toxic metal indicators. Environ. Chem. Lett. 12, 63–84. Stoll, S., Frenzel, M., Burkhard, B., Adamescu, M., Augustaitis, A., Baeßler, C., Boneth, F.J., Carranzai, M.L., Cazacu, C., Cosor, G.L., Díaz-Delgado, R., Grandin, U., Haase, P., Hämäläinen, H., Loke RMüller, J., Stanisci, A., Staszewski, T., Müller, F., 2015. Assessment of ecosystem integrity and service gradients across Europe using the LTER Europe network. Ecol. Model. 295, 75–82.

122

W. Schröder et al. / Science of the Total Environment 521–522 (2015) 108–122

Suck, R., Bushart, M., Hofmann, G., Schröder, L., Bohn, U., 2010. Karte der Potentiellen Natürlichen Vegetation Deutschlands: Band 1. Maßstab 1: 500.000; Karten + Legende. BfN-Schriftenvertrieb im Landwirtschaftsverlag Münster. Kartenteil: 7 Karten; Legende: 1–24. Suck, R., Bushart, M., Hofmann, G., Schröder, L., 2013. Karte der Potentiellen Natürlichen Vegetation Deutschlands. Band 1: Kartierungseinheiten. BfN-Skripten 349, Bonnpp. 1–305. Tanino, K.K., Kalcsits, L., Silim, S., Kendall, E., Gray, G., 2010. Temperature-driven plasticity in growth cessation and dormancy development in deciduous woody plants: a working hypothesis suggesting how molecular and cellular function is affected by temperature during dormancy induction. Plant Mol. Biol. 73, 49–65. Thibaut, L.M., Connolly, S.R., 2013. Understanding diversity–stability relationships: towards a unified model of portfolio effects. Ecol. Lett. 16, 150–160. Tierney, G.L., Faber-Langendoen, D., Mitchell, B.R., Shriver, W.G., Gibbs, J.P., 2009. Monitoring and evaluating the ecological integrity of forest ecosystems. Front. Ecol. Environ. 7, 308–316. Torseth, K., Aas, W., Breivik, K., Faeraa, A.M., Fiebig, M., Hjellbrekke, A.G., Myhre, C.L., Solberg, S., Yttri, K.E., 2012. Introduction to the European Monitoring and Evaluation Programme (EMEP) and observed atmospheric composition change during 1972– 2009. Atmos. Chem. Phys. 12, 5447–5481. UNECE (United Nations Economic Commission for Europe), 2013. Strategies and policies for air pollution abatement. 2010 Review Prepared Under the Convention on Longrange Transboundary Air Pollution. ECE/EB.AIR/123. New York, Geneva, pp. 1–57.

van Vuuren, D.P., Edmonds, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K., Hurtt, G.C., Kram, T., Krey, V., Lamarque, J.-F., Masui, T., Meinshausen, M., Nakicenovic, N., Smith, S.J., Rose, S.K., 2011. The representative concentration pathways: an overview. Climate Change 109, 5–31. Vitasse, Y., Porté, A., Kremer, A., Michalet, R., Delzon, S., 2009a. Responses of canopy duration to temperature changes in four temperate tree species: relative contributions of spring and autumn leaf phenology. Oecologia 161, 187–198. Vitasse, Y., Delzon, S., Dufrêne, E., Pontaillerc, J.-Y., Louvet, J.-M., Kremer, A., Michalet, R., 2009b. Leaf phenology sensitivity to temperature in European trees: do within species populations exhibit similar responses? Agric. For. Meteorol. 149 (5), 735–744. Vitasse, Y., François, C., Delpierre, A., Dufrên, E., Kremer, A., Chuine, I., Delzon, S., 2011. Assessing the effects of climate change on the phenology of European temperate trees. Agric. For. Meteorol. 151 (7), 969–980. Vose, J.M., Peterson, D.L., Patel-Weynand, T. (Eds.), 2012. Effects of climatic variability and change on forest ecosystems: a comprehensive science synthesis for the U.S. forest sector. Gen. Tech. Rep. PNW-GTR-870. U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Portland, OR, pp. 1–265. White, I.D., Mottershead, D.N., Harrison, S.J., 1992. Environmental Systems. Chapman and Hall, London, pp. 1–623. Wulff, S., Lindelöw, Å., Lundin, L., Hansson, P., Axelsson, A.-L., Barklund, P., Wijk, S., Ståhl, G., 2012. Adapting forest health assessments to changing perspectives on threats—a case example from Sweden. Environ. Monit. Assess. 184, 2453–2464.