Assessing the ecosystem service flood protection of a riparian forest by applying a cascade approach

Assessing the ecosystem service flood protection of a riparian forest by applying a cascade approach

Ecosystem Services 21 (2016) 39–52 Contents lists available at ScienceDirect Ecosystem Services journal homepage: www.elsevier.com/locate/ecoser As...

1MB Sizes 1 Downloads 63 Views

Ecosystem Services 21 (2016) 39–52

Contents lists available at ScienceDirect

Ecosystem Services journal homepage: www.elsevier.com/locate/ecoser

Assessing the ecosystem service flood protection of a riparian forest by applying a cascade approach Nina-Christin Barth, Petra Döll n Institute of Physical Geography, Goethe University Frankfurt, Altenhöferallee 1, 60438 Frankfurt am Main, Germany

art ic l e i nf o

a b s t r a c t

Article history: Received 10 September 2015 Received in revised form 6 July 2016 Accepted 19 July 2016

We present a method for assessing the ecosystem service (ES) flood protection of riparian wetlands and apply it to a riparian forest in Germany. The suggested workflow implements a cascade approach to ES characterization in which current provisioning is assessed in four steps: (1) qualitative description of biophysical processes and structures, (2) definition and quantification of main and additional ecosystem functions, (3) qualitative description of economic and social benefits and (4) valuation. Future provisioning is addressed by identifying pressures and analyzing potential enhancements. Using flood hazard and risk maps produced in response to the EU floods directive, quantification of the ecosystem function water retention as well as monetary valuation by the replacement cost and avoided damage cost methods were achieved without site-specific hydrological-hydraulic modeling. Technical structures with the same water retention volume as the investigated ecosystem in case of an extreme flood would cost 68 million EUR (equivalent ES value EUR 1900/ha/yr). In case of a 10-year flood, the riparian forest avoids damage costs of at least 26 million EUR (EUR 4300/ha/yr). We provide suggestions for standardizing the application of both monetary valuation methods and discuss their information content as well approaches for non-monetary valuation of the ES flood protection. & 2016 Published by Elsevier B.V.

Keywords: Flood protection Ecosystem service cascade Replacement costs Avoided damage costs Valuation Flood hazard map

1. Introduction Reported flood damages have increased from about USD 7 billion per year world-wide in the 1980s to about USD 24 billion per year in 2011 (adjusted for inflation; Kundzewicz et al., 2014). Since 1970, the annual number of flood-related deaths has been in the thousands, with more than 95% in developing countries (Handmer et al., 2012). The main reason for increased losses is greater exposure of people and assets (Handmer et al., 2012; Kundzewicz et al., 2014). As heavy rainfall events are very likely to become more intense and frequent due to climate change, except in areas with strongly reduced total rainfall, floods are also expected to increase in the future in many areas of the world (Döll et al., 2015). Whether flooding occurs after a heavy rainfall or snowmelt event strongly depends on the natural characteristics of the drainage basin and the floodplain. While technical measures such as dykes or man-made reservoirs may serve as additional flood protection, the focus is put increasingly on flood protection by the ecosystem itself, in particular the floodplain ecosystem that can store floodwater and decrease downstream peak discharges (Damm et al., 2011; Scholz et al., 2012). Thus, protection of existing natural n

Corresponding author. E-mail address: [email protected] (P. Döll).

http://dx.doi.org/10.1016/j.ecoser.2016.07.012 2212-0416/& 2016 Published by Elsevier B.V.

floodplain ecosystems or their restoration may be an appropriate strategy for flood protection and adaptation to climate change (e.g. BMU and BfN, 2009). Decision-making in favor of this type of flood protection is impacted by financially attractive alternative land uses including urban and agricultural uses. Therefore, assessment of the ecosystem service (ES) flood protection as provided by natural floodplains, e.g. riparian forests or riparian wetlands, is essential for making well-grounded decisions. Such an assessment should demonstrate the value of a particular ecosystem for flood protection in downstream areas, ideally in both monetary and non-monetary terms. Not all the benefits that humans derive from ecosystems can or should be expressed in terms of money. However, expression of at least a part of the total ES value in monetary terms allows internalizing so-called externalities in economic accounting procedures (TEEB, 2010a). Please note that assessment of ES does not imply that ecosystems only have value for the services they provide to people, i.e. an instrumental value, or that non-human living beings do not have intrinsic rights and intrinsic value (Hunter et al., 2014; but see Goulder and Kennedy (1997)). The anthropocentric concept of ecosystem services may be recognized as a pragmatic approach in support of sustainable development also by those who share a biocentric (or another non-anthropocentric) worldview. The challenge is that there is no blueprint or universal framework for ecosystem service assessments (ESA) (Haines-Young and

40

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

Potschin, 2009; Seppelt et al., 2012). As is the case for other regulating ES, the value of the ES flood protection cannot be derived based on market prices. A specific challenge of this ES is that it is very site-specific, such that its valuation can hardly rely on existing studies for other locations. Of the 1310 monetary ES values collected by van der Ploeg and de Groot (2010), only 13 referred to flood protection by inland wetlands. Applied monetary valuation methods for ES flood protection are the calculation of replacement costs and avoided damage costs (Leschine et al., 1997; van der Ploeg and de Groot, 2010; Brander et al., 2013). In former studies, usually only one of the two methods was applied. Grygoruk et al. (2013) and Leschine et al. (1997), for example, calculated replacement costs, Kousky and Walls (2014) avoided damage costs. Other studies did some type of non-monetary valuation (e.g. Nedkov and Burkhard, 2012) or only investigated the ecosystem function leading to the ES (Posthumus et al., 2010). There are a number of software tools for assessing ES (Peh et al., 2013). The Natural Capital Project provides the popular InVEST software which, however, does not include flood protection yet. An ESA software tool that covers the ES flood protection on a site scale is TESSA (Toolkit for Ecosystem Service Site-based Assessment). TESSA allows calculating avoided flood damage costs by wetlands but only if hydrological information on the impact of the wetland on inundated area during floods with a certain return period is available (Peh et al., 2013, 2014). This information needs to be derived by involved and costly hydrological-hydraulic modeling of climate and land cover-driven streamflow dynamics and inundation as affected by river and floodplain morphology. With our study on the ES flood protection, we contributed to a project that aimed at assessing all important ES of the riparian forest Bulau located directly upstream of the city of Hanau in the Federal State of Hesse, Germany. Through the Bulau and the city runs the river Kinzig which inundates almost the whole riparian forest and parts of the city during large flood events. Due to the water storage capacity of the riparian forest, peak discharges of floods and therefore inundation of Hanau are reduced as compared to other land uses. To assess the ES flood protection in a comprehensive manner, we developed a workflow that allows implementing the cascade approach of de Groot et al. (2010), Haines-Young and Potschin (2010) and TEEB (2010a). The assessment includes the consideration of the current state of the ES with its biophysical processes and structures, the definition and quantification of an ecosystem function, the detailed description of the resulting benefits and the valuation of the ES. Both replacement costs and avoided damage costs were computed on the basis of easily available data and without site-specific hydrological-hydraulic modeling. The future ES provisioning was addressed by discussing pressures on the ecosystem and potential enhancements of ES provisioning (Haines-Young and Potschin, 2009; MA, 2005a; Rounsevell et al., 2010; TEEB, 2010a). In this paper we present a structured and comprehensive approach for assessing the ES flood protection of a riparian wetland. To support the development of standards of valuing the ES flood protection, we clarify various aspects that affect monetary valuation and discuss how complementary non-monetary valuation may be approached. In addition, we wish to promote the comparability of ESAs, and therefore summarize our assessment using the Purpose, Scope, Analysis, Recommendations, and Monitoring (PSARM) Blueprint by Seppelt et al. (2012) (see Appendix).

2. Methods 2.1. Workflow for assessing ES Fig. 1 shows the workflow we applied for assessing the ES flood protection, based on a cascade approach. The workflow includes four steps for assessing the current provisioning of an ES and two regarding its future provisioning. Assessment of the current state of an ES starts with the qualitative description of the biophysical structures and processes of the ecosystem that are related to ES provisioning. Then a “subset of ecosystem processes and components that is directly involved in providing the service” is defined as an ecosystem function (TEEB, 2010a: 15, their Fig. 5). In this study, we quantified relevant indicators of the ecosystem functions (Fig. 1). The next step of the ESA is a detailed qualitative description of the economic and social benefits of the ES to the society. A benefit is defined as “the positive change in wellbeing from the fulfillment of needs and wants” (TEEB, 2014). In contrast to the TEEB cascade, we do not include “ecological (sustainability)” benefits (TEEB 2010a: 15, their Fig. 5) but rather add a second phase to the workflow in which the future provisioning of the ES is considered. Following the description of the benefits, i.e. the increase of human wellbeing due to the ecosystem, the value of the ES is characterized. This is done by assessing the importance of the ES for human wellbeing, expressing the importance either in monetary or in nonmonetary terms. In our case study, ES was valued monetarily by two alternative methods, the replacement cost and the avoided damage costs method. Non-monetary valuation was not done but approaches for its application for the ES flood protection are discussed. Finally, future provisioning of the ES is assessed. We identified (1) pressures that may negatively affect ES provisioning and (2) options for enhancing ES provisioning (Fig. 1 bottom) (Haines-Young and Potschin, 2009; MA, 2005a; Rounsevell et al., 2010; TEEB, 2010a). In other cases, modeling of the impact of future pressures and enhancement measures on the four elements of the workflow (Fig. 1 top) may be useful. The first four steps of the workflow (Fig. 1 top) are also elements of the cascade models of Haines-Young and Potschin (2010) and TEEB (2010a). But those cascades show “ecosystem services” as an additional element between “ecosystem functions” and “economic and social benefits”. Cascade models clarify that ecosystem functions become ES (only) if humans exist who may enjoy the benefits generated by ecosystem functions; putting “ecosystem services” between “ecosystem functions” and “benefits” illustrates that ES are derived from ecosystem functions and lead to benefits for humans. However, cascade models may not be interpreted as a workflow for assessing an ES. ES are defined as the benefits that humans derive from ecosystems (MA, 2005a). Therefore, when assessing ES one cannot, as the cascade models may suggest, first characterize the ES and then the benefits as these are per definition the same. Instead, as shown in Fig. 1 (top), it is the joint description of (1) biophysical structures and processes, (2) ecosystem functions, (3) the benefits derived from the ecosystem function and (4) the value of the benefits that make up the characterization of an ES. 2.2. Data Quantitative calculations carried out in this study were based on easily accessible data. Central to this study were flood hazard maps for Hanau including the riparian forest Bulau. The maps were produced in response to the EU Floods Directive (EU, 2007), a Europeanwide effort to improve flood protection. This regulatory framework prescribes the assessment of flood risks in all 27 member states of the EU (Müller, 2010). Flood hazard maps show flooded areas and inundation heights for statistical flood events with different return periods and thus magnitude. Additional flood risk maps indicate the

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

41

Fig. 1. Workflow for assessing current and future provisioning of an ES, implementing the cascade models of Haines-Young and Potschin (2009, 2010) and TEEB (2010a) in an ESA.

different land use categories within the floodplains. Based on the flood hazard maps, the storage capacity of the Bulau could be estimated, and also to what extent the riparian forest Bulau reduces inundation within the City of Hanau. In combination with damage functions and specific asset values provided by IKSR (2001b), the flood risk map enabled computation of flood damages. Flood hazard and flood risk maps for a 10-year flood (HQ10), a 100-year flood (HQ100) and an extreme flood (HQex) were provided as raster data with a spatial resolution of 2 m by 2 m by the Hessian State Agency for Environment and Geology (HLUG, 2014a).

3. Study area The riparian forest Bulau is located directly upstream of the city Hanau in the Federal State of Hesse, Germany, where a total of 93,000 people live (Stadt Hanau, 2015) (Fig. 2). The Bulau covers an area of 604 ha and is part of the floodplain of the river Kinzig. It is a protected area according to the Flora-Fauna-Habitat Directive of the EU (Buttler et al., 2003). The Kinzig runs about 80 km from its source at 400 m a.s.l. to its confluence with the river Main in Hanau at 100 m a.s.l. (Ditter, 1991). The Kinzig basin comprises 1058 km2 (HGN, 2004). The forest community in the Bulau is

Fig. 2. Location of the riparian forest Bulau (green within black outline) in Germany and within the Kinzig basin (left). The floodplain of the Kinzig is located directly upstream of the city of Hanau, where the river Kinzig flows into the river Main (right). Database: HLUG (2014a, 2014b, 2011). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

42

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

Fig. 3. Flooded areas with inundation heights [m] in case of a 10-year flood (HQ10), a 100-year flood (HQ100) and an even more extreme flood (HQex) within the Bulau (black outline) and downstream until the confluence of Kinzig and Main. Database: HLUG (2014a, 2014b).

strongly influenced by groundwater and is referred to as chickweed-oak-hornbeam forest (Stellario-Carpinetum) (Buttler et al., 2003). The dominant soil types are brown alluvial soil and (pseudo-) gleyish brown alluvial soil (HLUG, 2014b). Fig. 3 shows flooded areas and inundation heights for a 10-year flood (HQ10), a 100-year flood (HQ100) and an extreme flood (HQex) of the Kinzig in the Bulau (outlined in black) and in the downstream city of Hanau until the confluence of Kinzig and Main. In case of floods, the riparian forest Bulau acts as large flood water storage; during a 100-year flood and an extreme flood almost the whole riparian forest is inundated, while inundation of the downstream city is relatively small (Fig. 3).

4. Assessing the ES flood protection of a riparian forest 4.1. Current provisioning 4.1.1. Biophysical structures and processes A flood results from a high value of runoff generation within a short period of time, either driven by heavy rainfall or intense snowmelt, and concentration of discharge in the receiving stream. Flooding (inundation) occurs if discharge in the river is so high that the river water cannot be contained any more within the river channel but overflows into the floodplain. The fast runoff component direct runoff that is relevant for flood generation depends not only on rainfall or snowmelt; it also depends on structures within the river basin ecosystem that enhance evapotranspiration

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

(before the rainfall event), interception and infiltration. A forest such as the Bulau mostly leads to higher evapotranspiration before the rainfall event as compared to cropland, grassland or settlement areas. This results in a lower value of soil saturation and thus more pore space for infiltration of rainfall. Interception, i.e. retention of precipitation on leaves, is also generally higher in case of forests than in case of crops or grasslands. The lower amount of rainfall that reaches the ground in forests may also infiltrate better in forests due to larger pore space compared to areas with a low vegetation cover or artificial cover (e.g. concrete). Once direct runoff has reached the river, the floodplain ecosystem can reduce the flood peak significantly through the retention of excess water by water storage as affected by storage space and reduction of flow velocity (Scholz et al., 2012). Flat terrain and high roughness increase the retention of water (Dittrich and Worm, 2006). Water is retained above the land surface where topography is flat like in the case of the Bulau, and in the subsurface pore space if the subsurface is not already saturated (Russi et al., 2013). In addition, the roughness of the floodplain, which is high in case of a floodplain forest, slows down the flow rate in the inundated floodplain (Damm et al., 2011; Scholz et al., 2012). In general, reduction of direct runoff and the water retention in the floodplain are influenced by topography, ground cover or vegetation, soil and geology (Dittrich and Worm, 2006). 4.1.2. Ecosystem functions Among the biophysical structures and processes, two key processes, i.e. the ecosystem functions, were identified that may lead to provisioning of the ES flood protection by the riparian forest. The main ecosystem function of the ES flood protection of a riparian forest is retention and the additional ecosystem function is regulation of direct runoff. The additional ecosystem function has a significantly smaller influence on the reduction of a flood peak, as the Bulau covers less than 0.6% of the Kinzig basin area upstream of Hanau, such that its regulation function regarding the generation of direct runoff in the basin during a rainfall event is expected to be very small. Ecosystem functions were quantified with the help of state indicators that describe “[…] what ecosystem process or component is providing the service and how much” (de Groot et al., 2010: 262). To represent the current state of the ecosystem, the main ecosystem function retention was quantified by the state indicator surface and subsurface retention volume, and the additional ecosystem function regulation of direct runoff by a model estimate of mean annual direct runoff per unit area of the Bulau. 4.1.2.1. Quantification of main ecosystem function. Taking into account inundation heights and areas provided by the flood hazard maps, surface retention volumes of the Bulau were determined using the GIS software ArcGIS by computing water volumes in the inundated area within the Bulau during the statistical floods HQ10, HQ100 and HQex (Fig. 3). The volume of water bodies (the Kinzig itself and a small lake) was subtracted. The surface retention volumes range between 2.7 million and 5.9 million m3 (Table 1). Calculation of the subsurface retention volume was computed by multiplying a representative subsurface volume by a literaturebased estimate of maximum water retention capacity in percent of subsurface volume. Maximum water retention capacity was estimated as 25 vol% (difference between the porosity value of 53.5 vol% and the water content at the permanent wilting point of 28.5 vol%) (AG Boden, 2005; HLUG, 2014b). The representative subsurface volume of the Bulau was computed by simply multiplying the inundated area of the riparian forest (Fig. 3) with an average depth to the groundwater table of 1 m. The computed subsurface retention volumes range between 1.0 and 1.3 million m3 for the three types of flood events (Table 1) and may be an overestimation as water content in the soil is likely to be above the

43

Table 1 State indicators of the main ecosystem function retention and the additional ecosystem function regulation of direct runoff. The size of the riparian forest Bulau is 604 ha. Ecosystem function

State indicator

Main ecosystem function retention

Surface retention [m3]

Add. ecosystem function regulation of direct runoff

a

Time scale

HQ10 HQ100 HQex Subsurface retention [m3] HQ10 HQ100 HQex Direct runoff [mm/yr]a Annual average Regulation of direct runoff % on x% of total catchment area

Value

2.674.582 4.597.853 5.928.655 1.040.000 1.280.000 1.345.000 20 0.6

Nina Stiehr (personal communication, 2014).

water content at wilting point before the onset of a flood. 4.1.2.2. Quantification of the additional ecosystem function. Different from the water retention volumes, the computed direct runoff values are not those during a particular statistical flood as estimates of direct runoff during those floods were not available. Mean annual direct runoff, evapotranspiration and groundwater recharge rate in the Bulau were calculated by Nina Stiehr (personal communication, 2014) using a simple soil water balance model. The model applied the method of Meßer (2013), taking into account data on climate, soil, depth to water table and land use in the Bulau. Vertical diffuse groundwater recharge in the Bulau was calculated to be 115 mm/yr, based on 736 mm/yr of precipitation. Evapotranspiration amounts to 601 mm/yr. The average annual direct runoff is only 20 mm/yr. Influence of the Bulau on runoff generation in the Kinzig basin is very small, since the area of the Bulau is only 0.6% of total basin area upstream of Hanau (Fig. 2, Table 1). 4.1.3. Economic and social benefits Benefits for humans are “the positive change in wellbeing from the fulfillment of needs and wants” (TEEB, 2014). Due to its ecosystem functions, the Bulau can provide an ecosystem service that is in demand by the inhabitants of downstream Hanau flood protection, or more precisely reduction of inundated area and inundation height, from which a variety of economic and social benefits arise (Table 2). Economic benefits of the Bulau with respect to flood protection Table 2 Economic and social benefits resulting from the ES flood protection of the riparian forest Bulau. Benefits

1a 2a 3a 4a 5a 6

Economic Avoided damage to buildings including mobile assets and infrastructure Avoided crop failure of agricultural areas or forestry Avoided production stoppage and business interruption Avoided damages to vehicles Avoided costs of emergency management, fire brigade and emergency services Lower (no) expenses for man-made flood protection structure Social Protection of human life, health and income generation capacity Protection of identity including sites of cultural significance and personal belongings Sense of security

7 8 9 a

IKSR (2001a: 5).

44

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

include prevention of damages and financial losses in case of floods as well as lower expenses for technical flood protection structures like dykes. Social benefits focus on the protection of people living in areas at risk of flooding, including the protection of their lives, health and income generation capacity, but also the protection of their cultural and personal identity. The place they bond to will not be destroyed and they will not suffer from emotional stress due to the experienced flood event. The riparian forest provides a feeling of security to flood-prone residents. An additional benefit for some people might be the good feeling that nature itself and not expensive and possibly unreliable technical infrastructure protects people from floods. A clear description of the benefits of an ES, i.e. of the human welfare increase due to an ES, is necessary for understanding the ES, and an indispensable basis for determining the ES value. Even though benefits and values are closely related as it is the benefits that are valued, a separate (qualitative) description of benefits is an essential step of the workflow for assessing ES. One reason is that provisioning of a benefit may remain constant over time while its valuation changes because, for example, costs for replacing destroyed assets increase due to higher market prices (Haines-Young and Potschin, 2009). Another reason is that different valuation methods may value different benefits. 4.1.4. Values The value of a benefit describes how important a benefit is. The concept of “Total Economic Value” describes the value of ecosystems for humans as the sum of certain components such as use and non-use values (Christie et al., 2012). While there are a large number of valuation methods that are suitable for particular types of values, there are methodological difficulties with all of them (Christie et al., 2012). In our study, we considered the actual indirect use value that the ecosystem has for the downstream population of Hanau regarding flood protection. We applied two market-cost based approaches to value, in monetary terms, part of the benefits described in the previous section. The replacement cost method provides a proxy measure for the value of the benefit that due to the riparian forest there is no need for (additional) technical flood protection infrastructure, by estimating the potential cost of the infrastructure (benefit 6 in Table 2). This method was applied by Grygoruk et al. (2013), Leschine et al. (1997) and in most of the 38 wetlands studies compiled by Brander et al. (2013). The avoided damage cost method values the reduction or prevention of flood damages to buildings, mobile assets and infrastructure as well as to agricultural land and forests (benefits 1 and 2 in Table 2). This monetary valuation method was used by e.g. Dehnhardt and Meyerhoff (2012) and Kousky and Walls (2014) and a few of the studies of Brander et al. (2013). Neither stated preference methods nor non-monetary valuation were performed in the case study but are discussed in Section 5. Regulation of direct runoff by the Bulau was not taken into account in the ES valuation due to the small contribution of this

ecosystem function to the benefits that human derive from the Bulau. Only the main ecosystem function water retention was considered. For the ES valuation the current situation, where the Bulau stores water in case of floods, was compared to the following hypothetical alternative situation. The riparian forest Bulau does not exist anymore as the whole piece of land has been converted into a residential area and covered with buildings. To protect the area from flooding, a dyke has been constructed along the Bulau. In case of floods, the floodwater is therefore no longer stored by the Bulau but flows downstream where it leads to increased flooding of the built-up area of Hanau that has not been equipped with (higher) dykes. 4.1.4.1. Replacement costs. Replacement costs are the amount of money that would be required for replacing the ES by a technical flood protection structure that provides the same service (Brander et al., 2013). Technical substitutes include man-made reservoirs or smaller retention basins. Hypothetical investment and operational costs are used to express the value of the ES that would be replaced in monetary terms (MA, 2005b). The technical alternative considered should be the least-cost alternative (Shabman and Batie (1978) as cited in Grossmann (2012)). There are different approaches for determining replacement costs (e.g. Grygoruk et al., 2013; Leschine et al., 1997; Jiang et al., 2007). In this study, we used the investment costs of existing and planned flood protection structures together with information on their water retention volumes to compute the cost of storing 1 m3 of water by some technical substitutes of the Bulau. With this approach, it is unlikely that we overestimate the cost of the alternative flood protection infrastructure. It can be assumed that the existing flood protection infrastructure was located at a costefficient location, i.e. where the ratio of water storage over cost was high. It is likely that additional built infrastructure to replace the riparian wetland would have to be located at a location with a less suitable relief, for example, and therefore a lower water storage over cost ratio. We only took into account flood protection structures within the Kinzig basin as they represent the real commitment of the region to invest in such protection measures (Grygoruk et al., 2013; Leschine et al., 1997). The total costs for the existing Kinzig reservoir and the two planned retention basins (Table 3) refer to planning and construction of the reservoirs and do not include costs of reservoir maintenance or hydropower generation (Grygoruk et al., 2013). The arithmetic mean of the unit cost of water storage, 11.40 EUR/m3, was multiplied by one of the state indicators for the ecosystem function retention, the surface retention volume of the Bulau, for the three types of floods (Table 1). Computation for the case of an extreme flood when the Bulau retains 5.9 million m3 (Table 1) results in the maximum ES value of the riparian forest. If the surface retention volume of the riparian forest were to be replaced by technical flood protection measures like those listed in Table 3 in case of an extreme flood, the costs would amount to EUR 68 million (Table 4). Considering the area of the Bulau of 604 ha and the average useful life of the flood protection structures of 60

Table 3 Costs of the existing and planned flood protection structures in the catchment of the Kinzig (price basis 2014) (Verlag Parzeller GmbH and Co. KG, 2013). Flood protection structure

Retention volume [m3] Total planning and construction costs [Mio. EUR]

Unit cost of water storage [EUR/m3]

Details

Bad Soden-Salmünster, Salz (planned)

350,000

6.80

19.42

Birstein, Lahnemühle, Bracht (planned) 817,000

3.80

4.65

63.40b

10.39

Dam height: 6 m Dam width: 285 m Dam height: 6,5 m Dam width: 315 m Large dam

Kinzig reservoir (built in 1982) a

6,100,000a

HMUELV (2007: 27), Tab. 1. Estimate of the total costs of planning and construction: DM 73,744,209 (Hessischer Landtag, 1988), value corresponds to EUR 63,442,973 for the price basis 2014 (Hainke, 2014). b

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

Table 4 Replacement costs and avoided damage costs for the ES flood protection of the riparian forest Bulau. Replacement costs Total [EUR] [EUR/yr]

HQ10 30,730,947 HQ100 52,829,331 HQex 68,120,246

Avoided damage costs [EUR/ha/ yr]

512,183 848 880,489 1458 1,135,337 1880

Total [EUR] [EUR/yr]

25,843,492 n.c. n.c.

[EUR/ha/ yr]

2,584,349 4279 n.c. n.c. n.c. n.c.

n.c. not computed.

years (DWA, 2012), replacement costs for an extreme flood are 1.1 million EUR/yr or 1900 EUR/ha/yr. Replacement costs for less severe and more frequent floods are smaller (Table 4) due to the lower retention volume in the Bulau that would have to be replaced (Table 1). If the subsurface retention volume were included in the calculations, replacement costs for the case of an extreme flood would increase by 23%. Due to the uncertainty of the calculated subsurface retention volume, it was neglected here. 4.1.4.2. Avoided damage costs. We calculated both the potential damage costs without flood protection by the Bulau and the damage costs that occur even with the riparian forest in place, for each of the three statistical floods. The difference is the damage cost avoided due to the ecosystem, i.e. the ES value (Dehnhardt and Meyerhoff, 2012; MA, 2005b). Both damage scenarios were computed on the basis of the flood hazard and flood risk maps of HLUG (2014a) using ArcGIS. No hydrological-hydraulic modeling was performed for the study. We chose to only take into account direct, tangible and primary flood damages (benefits 1 and 2 in Table 2), like in most flood damage assessments (IKSR, 2001a, 2001b; Müller, 2010). The land use categories of the flood risk maps considered to be susceptible to damage were settlement, culture and service, industry, transport, agriculture and forestry. Damage likely to occur in case of flooding was calculated by combining specific asset values per m2 with relative damage functions that related the percentage of the asset value lost to inundation height (Müller, 2010; IKSR, 2001a). Each land use category within the flood zone requires its own damage function (Table 5). There is also a distinction between immobile (such as buildings) and mobile (e.g. building contents) values within one land use category (IKSR, 2001b). In this work we used the land use-specific relative damage functions and specific asset values for Hesse (Table 5) provided by the IKSR Rhine Atlas (IKSR, 2001b). Please note that the damage functions from the IKSR Rhine Atlas were developed for CORINE land use categories Table 5 Relative damage functions and specific asset values for Hesse to calculate damage costs (h ¼inundation height [m], d¼ percentage of asset value lost [%]). Land use category

Settlement Industry Transport

Agriculture Forestry a b

Immobile

Mobile

Relative daSpecific asmage functiona set valueb [EUR/m2]

Relative damage functiona

Specific asset valueb [EUR/m2]

d ¼2h2 þ 2h d ¼2h2 þ 2h d ¼10h for h o 1; d ¼10 for h 41 d ¼1 d ¼1

d ¼11.4hþ 12.625 d ¼7h þ 5 d ¼10h for h o 1;

51 80 3

IKSR (2001b): 38, Tab. 5.13. IKSR (2001b): 22, Tab. 5.5.

231 258 300

d ¼10 for h 41 1 1

d ¼1 d ¼1

7 1

45

while flood risk maps used here show aggregated ATKIS data. The category culture and service from the land use data was also considered as settlement area. The category green space was considered insensitive to damage and was not taken into account. First we considered a 10-year flood. In case of the alternative scenario without flood protection by the Bulau, the volume of water that would be stored by the Bulau during a 10-year flood was assumed to inundate the city of Hanau. From the flood hazard maps from HLUG, inundated areas and inundation heights for the three statistical floods HQ10, HQ100 and HQex were available (Fig. 3). In case of a 10-year flood, the surface retention volume of the Bulau in its current state is about 2.7 million m3 (Table 1) and the volume of the inundation water in the city of Hanau amounts to about 0.3 million m3. Thus, in the alternative scenario without flood protection by the Bulau, a total volume of about 3.0 million m3 can be assumed to inundate the city of Hanau in case of a 10year flood. Inundation of the city of Hanau for the case of dykes along the Bulau would be best simulated by hydrological-hydraulic modeling but such modeling results were not available as modeling exercises done for generating flood hazard maps do not include such specific model runs. Based on the flood hazard maps (Fig. 3) we calculated that the volume of inundation water downstream of the Bulau in its current state is approx. 0.8 million m3 in case of a 100-year flood and 1.8 million m3 in case of an extreme flood. These volumes are smaller than the volume of water that is stored by the Bulau during 10-year floods. This means that inundation and therefore flood damages that would occur on average every 10 years if the riparian forest would not be able to store water would be much more severe than for a very rare and extreme flooding event with the current flood protection by the Bulau (HQex, Fig. 3). Thus, evaluation of the flood hazard map for HQex allows deriving a lower bound for the damage that would occur without the Bulau in case of a 10-year flood. Data on inundation height and land use category for each 2 m by 2 m area inundated were combined to compute, for the two scenarios HQ10 and HQex, flood damage in the area downstream of the Bulau to the confluence of Kinzig and Main (Fig. 4), using damage functions and specific asset values shown in Table 5. While the financial damage for 10-year floods is estimated as EUR 1.5 million, it would increase to over EUR 24.0 million if the Bulau would no longer serve to retain the flood water (Table 6). Accordingly damage costs avoided by the ES flood protection of the Bulau are about EUR 22.5 million in case of a 10-year flood. As the applied specific asset values were for 2001, we adjusted the computed avoided economic damage value to 2014 based on the development of the Germany gross domestic product. This resulted in a value for the avoided damage costs in case of a 10-year flood of approximately EUR 26 million. As without the Bulau this (additional) damage would occur on average every 10 years, the annual value is EUR 2.6 million/yr or EUR 4300/ha/yr (Table 4). These avoided damage costs for the case of a 10-year flood are a very conservative estimate for various reasons. The computed value only refers to the damage that only about 60% of the flood water volume that would probably flow into the city of Hanau without the Bulau would cause. Apart from this, we neglected additional damage costs such as replacement, production and operational failures and cost of civil protection (IKSR, 2001a). Besides the specific asset values for the land use category settlement may have been underestimated (IKSR, 2001b). To obtain total annual avoided damage costs, flood events with different return periods should have been considered and their damages should have been summed up in proportion to the frequency (Kousky and Walls, 2014). However, as our data base was restricted to the three available flood hazard maps, it was only possible to estimate avoided damage costs for the 10-year flood.

46

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

Fig. 4. Damage costs [EUR/4 m2] for a HQ10 (top) and HQex (bottom) under the current situation. In the alternative scenario without the riparan forest Bulau, inundation and thus damage in case of a 10-year flood would be even larger than shown in the bottom panel. Database: HLUG (2014a, 2014b), IKSR (2001b).

4.2. Future provisioning

the basis for the assessment of the future provisioning of the ES.

An important aspect of the ES approach is sustainable use and safe-guarding of the future provisioning of the ES (MA, 2005a). Therefore, the second part of the assessment (Fig. 1 bottom) identified pressures on the ecosystem including the risk of an ecosystem collapse to analyze whether the future ES provision was ensured (Haines-Young and Potschin, 2009; TEEB, 2010a; Rounsevell et al., 2010). In addition, we considered possibilities to enhance the future provisioning of the ES (MA, 2005a; TEEB, 2010a). The already quantified indicators of the ecosystem functions form

4.2.1. Pressures The most important pressures on flood-reducing structures and processes in a riparian forest are expansion of urban areas and agricultural land as well as river modification e.g. for transport (BMU and BfN, 2009). This is currently no threat to the Bulau as it has been declared a protected area. The riparian forest is part of a landscape conservation area (Buttler et al., 2003). As a result, the regulations of the Federal Water Act apply, e.g. regarding land use (Federal Water Act, 2009, § 78). In addition, the Bulau is a Flora-

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

Table 6 Financial damage [EUR] in case of HQ10 and HQex in the area downstream of the Bulau ending up at the confluence of Kinzig and Main. The financial damage in case of HQex in the current situation (as shown here) is less than the damage in case of HQ10 if the Bulau is no longer able to store flood water (price basis 2001) (Database: HLUG (2014a), IKSR (2001b)). Land use category

Settlement Industry Transport Agriculture Forestry Total

Financial damage [EUR] HQ10

HQex

1,464,357 49,277 5288 13,063 716 1,532,701

13,939,741 7,400,698 2,639,749 24,111 1004 24,005,303

Fauna-Habitat area (area number: 5819-308) (Buttler et al., 2003). However, some less important pressures do exist. In the retention cadaster for the Kinzig catchment there is currently a proposal to gain additional retention volume in the area of the Bulau by increasing the water storage, e.g. by constructing small dams (HGN, 2004). This intervention could affect the total water balance and retention behavior of the riparian forest in a yet unknown way. 4.2.2. Enhancements We evaluated whether an alternative vegetation in the Bulau would lead to reduced direct runoff (an ecosystem function). Direct runoff in the current state (forest), with an average of 20 mm/ yr, was compared to the values computed for three different land use variants in the Bulau (Nina Stiehr, personal communication, 2014). These were grassland, intensive agriculture, and urban area with an average sealing of 40–50%. Grassland may lead to a direct runoff of 68 mm/yr, and intensive agriculture to 57 mm/yr. Conversion of the Bulau into urban area may result in an almost tenfold higher direct runoff of 212 mm/yr. Compared to these three land use variants the Bulau in its current state shows the best ecosystem function. Regarding water retention, the main ecosystem function, the alternative land use scenarios would result in reduced surface roughness and therefore a reduced retention. An improvement of the ecosystem functions and thus the ES is therefore not possible by alternative land uses. A small optimization potential still exists; the path network within the Bulau has a lower infiltration capacity and surface roughness than the surrounding forest, both of which could be increased to a certain extent. But even removal of the path network is likely to have only a small impact on water retention of the Bulau.

5. Discussion As pointed out by Cordier et al. (2014), it is the benefits to humans that should be valued; only benefits depend on (individual) human preferences and can therefore be valued. To avoid double counting, other elements of the cascade should not be valued (Cordier et al., 2014). 5.1. Monetary valuation There are good reasons for expressing the value of benefits to humans in monetary terms. According to Emerton and Bos (2004: 29), “[…] economic concerns remain a powerful determinant of how people behave, how decisions are made and how policies are formulated” (Emerton and Bos, 2004: 29). A monetary value is “[…] directly comparable with other sectors of the economy […]” (MA, 2005b: 61). De Groot et al. (2012: 51) stated that “[…] expressing the value of ES in monetary units is an important tool to raise

47

awareness and convey the (relative) importance of ecosystems and biodiversity to policy makers”. According to Leschine et al. (1997: 50) replacement costs show “[…] that there exists a strong economic rationale […] to protect wetlands today in order to avoid what are likely to be much higher costs of flood protection in the future”. However, it is important to recognize the limits of monetary valuation. Below we discuss methodical issues, and limits to the information content of monetized values. Methodical differences in non-standardized monetary valuation lead to different numerical results. Applications of the replacement cost method may vary with respect to the type of the replacement, the reference price or the choice of reference floods. Regarding calculation of avoided damage costs, results vary due to consideration of different flood events, e.g. with different return periods. In addition, studies on monetary valuation of the ES flood protection vary in terms of considered types of damage, damage sensitive land use categories, applied damage functions and whether the calculations are conducted in an area- or object-based way. To improve the comparability of monetary values of the ES flood protection that are computed by the same valuation method, application standards would be helpful. Based on the experience gained in this study, we provide some preliminary suggestions on how to best compute monetary values of the ES flood protection by replacement cost method (Table 7) and the avoided damage cost method (Table 8). Certainly, computed monetary ES values cannot be transferred easily to other situations. Both investment costs for technical substitutes and specific asset values strongly depend on the economic context (Haines-Young and Potschin, 2009; TEEB, 2010b). This is why the TEEB database of monetary ES values indicates the “Country Income Group” (van der Ploeg and de Groot, 2010). Average investment costs for flood protection in China, for example, are much lower than in the Kinzig basin in Germany (Jiang et al., 2007). Even within Europe, transfer among basins may be problematic given that replacement infrastructure needs to be suitable for the specific basin. Regarding comparability of the damage costs avoided by wetlands, this value depends strongly on site-specific hydrological conditions and on land use in the area downstream of the investigated wetland. The specific asset values for agricultural or forestry areas are less than 10 EUR/m2, but between 200 and 300 EUR/m2 (see Table 5) in urban areas. Besides methodical issues, there is the issue of the information content of monetary ES values. Flood damage costs avoided by the Bulau amount to much more than EUR 2.6 million/yr, only considering the statistical 10-year flood. This value is much larger than the annual costs for a technical substitute of the Bulau (manmade reservoirs) of EUR 1.1 million/yr that would retain all the water the Bulau can retain in case of an extreme flood. According to de Groot et al. (2012; 58) “[…] it should be noted that values derived from different valuation methods may not be measuring the same economic construct and therefore values from different methods may not be directly comparable”. The two different numbers for the value of the ES flood protection represent different perspectives, and their use for decision making should fit with the specific problem setting. When communicating the ES value derived as replacement cost, it should be considered whether capital and technical skills are available for such an investment and if there is an appropriate location in the catchment for building large reservoirs. If it is not possible to build a technical flood protection structure that may substitute the ecosystem, the ES value is much higher than the calculated replacement cost. Then, avoided damage costs are more informative. Another situation in which to use avoided damage costs is when a riparian forest may be converted into an urban area and avoided damage costs are equal to the opportunity costs of this land use. In case of a 10-year flood, the riparian forest Bulau avoids a damage of EUR 2.6 million/yr.

48

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

Table 7 Computation of the monetary value of the ES flood protection by the replacement cost method: Proposals for supporting the establishment of application standards. Aspect

Examples from literature

Proposal

Indicator

Subsurface retention volume (Jiang et al., 2007) Surface retention volume (Leschine et al., 1997) Reduction of flood peak (Leschine et al., 1997)

Reference for the indicator

Statistical flood (e.g. 100-year flood) (Leschine et al., 1997) Historic flood event (Leschine et al., 1997) Whole wetland area (no specific or statistical flood event) (Jiang et al., 2007)

Technical substitute

Man-made reservoirs, reconnection of different wetlands

Reference price

Single technical substitute (Leschine et al., 1997) Several technical substitutes (Leschine et al., 1997) Average investment in flood protection structures in the entire country within a defined period of time (Jiang et al., 2007)

The choice of the indicator for comparing the performance of the riparian wetland and the technical substitute may play a minor role and could be selected depending on the available data. It is easier to compute (sub-)surface retention volumes than to simulate the reduction of runoff during flood events due to the ecosystem. It is better to consider only the water retention volume that is actually needed during a flood event and not the total retention volume of the wetland. Consideration of rare floods (e.g. HQ100 or HQex) may represent maximum flood protection performance of the wetland. Consideration of statistical floods instead of a specific historic flood event allows better comparability with other studies, and a generalization of the ES with respect to future floods. The technical substitute should be realistic option for the study area. It is therefore best to use costs of (least-cost) technical substitutes within the catchment under consideration, as these reflect the actual willingness to pay for such protection measures (Leschine et al., 1997). We recommend to average costs of several flood protection structures since the price of a single substitute may vary greatly depending on type and location.

The investment into technical substitutes needed to provide the same service as the riparian forest amounts to 0.5 million EUR/yr, while the riparian forest provides the service for free. In general, it is informative to monetize the value of the ES flood protection by both methods. However, it is necessary to clearly distinguish the resulting numerical values and, for example, not to average over numerical values that have been determined using those two methods (as has been done in the meta-analysis by Brander et al. (2013)). The monetary ES values determined in this study are very conservative (low) estimates. Regarding replacement costs, subsurface retention volume was neglected. Regarding the avoided damage costs, inundation area and height during a 10-year flood, without water retention by the Bulau, were estimated to be larger more extensive than in case of an extreme flood event with the existing water retention by the Bulau. Besides, the damage costs could double if costs to replace the lost assets were considered, or could increase up to fourfold if production failure and business stoppage were included (IKSR, 2001a). The underlying assumption of applying the replacement cost and the avoided damage cost methods is that the benefits are at least as great as the costs involved in repairing, avoiding or compensating for damage (Brander et al., 2013). The two market-cost based methods cannot determine the total value of an ES but only represent a proxy (Christie et al., 2012). “[…] it is important to realize that economic and esp. monetary valuation will always capture only part of the “true” or total value (…) of an ecosystem or service” (de Groot et al., 2010: 262). Social benefits (see benefits 7–

9 in Table 2) such as avoiding psychological stress are not included by the two methods applied in this study. Stated preference methods including contingent valuation, however, are suitable for identifying the value of social benefits as well as the “Total Economic Value” (Christie et al., 2012). They reflect more clearly the demand for the ES than the two market-cost based methods. In contingent valuation studies potential beneficiaries are asked for their willingness-to-pay for flood protection. Contingent valuation studies are expensive and need to be designed very carefully in order to result in valid ES values (e.g. Carson et al., 2001). Stated preference methods have very rarely been applied to value the ES flood protection of riparian wetlands (but compare Zhai et al. (2006) for a contingent valuation study on willingness-to-pay for general flood risk reduction). Non-monetary valuation (Section 5.2) should complement any monetary valuation to obtain a more integrated characterization of the ES flood protection. Stakeholders have to be informed about relevance and limitations of the communicated ES values. It is necessary to clearly state the assumptions made in the ES assessment and to describe uncertainties at least qualitatively. Quantitative uncertainty estimates are preferred but often not achievable. For example, in our case no quantitative information on the uncertainty of the flood hazard maps was available. We compared the computed monetary values of the ES flood protection of the Bulau to values from other studies (Leschine et al., 1997; Jiang et al., 2007; van der Ploeg and de Groot, 2010; Brander et al., 2013). We found that the values derived in this study are within the very wide range of values from the literature.

Table 8 Computation of the monetary value of the ES flood protection by the avoided damage cost method: Proposals for supporting the establishment of application standards. Aspect

Examples from literature

Proposal

Assumptions for the alternative scenario

No wetland (e.g. dam around newly created residential area) (this study) Land use change (Kousky and Walls, 2014) Specific flood event (ACOE, 1976) Several statistical floods (Kousky and Walls, 2014) Single statistical flood (Gerrard, 2004)

To reflect the entire ES flood protection it seems reasonable to create a scenario in which the main wetland ecosystem function of the wetland is not performed anymore.

Reference flood

Quantifying the alternative scenario (performance of relevant indicator)

Modeling, usage of alternative data and simplified assumptions (this study)

Consideration of statistical floods allows a better comparability and generalization. Which statistical floods should be considered remains to be discussed; it is preferable to consider a large number of floods with different return periods and then aggregate their damage costs avoided (Kousky and Walls, 2014). It should be investigated to what extent costly hydrological-hydraulic modeling of alternative scenarios can be replaced (and with what accuracy) by often easily available data such as flood hazard maps.

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

For 26 wetlands mostly in the US and Europe, for example, Brander et al. (2013) determined a mean and median value of the ES flood protection of 6923 USD/ha/yr and 427 USD/ha/yr, respectively. 5.2. Non-monetary valuation Non-monetary valuation can determine the value of the benefits that cannot be described in terms of money. These benefits include protection of human health and human lives as well as of cultural and personal identity that may be negatively affected by flooding (Table 2). Assume the hypothetical example of two riparian forests A and B that provide the ES flood protection with the same monetary value (e.g. because they avoid the same amount of damage cost). The total value of the riparian forest A would be higher than that of riparian forest B if in case A more people were affected by floods, if society were less prepared for a flood event or if it had a lower capacity for coping with flood damages. Vulnerability of society to floods and demand for flood protection is positively correlated with the value of the ES flood protection (Nedkov and Burkhard, 2012). Vulnerability can be characterized qualitatively or by identifying suitable indicators (e.g. number of deaths avoided) and then quantifying them. If these indicators are scaled (e.g. 0–1), they allow a comparison between ES flood protection of different wetlands. Table 9 suggests questions that may help to determine exposure and vulnerability to floods and thus the value of the ES flood protection. We regard consideration of these questions as a necessary part of an assessment of the value of the ES flood protection as monetary values alone do not provide a complete characterization of the value of an ES. When, for example, comparing the ES of wetlands across country boundaries, the consideration of the monetary value only would be misleading even if it were corrected for the average income or purchasing power in the different countries. The total value of the wetland would be underestimated for the country where construction of a technical substitute would not be feasible (question 6 in Table 9). Consultative and deliberative methods that may be applied to value the ES in a non-monetary way are described in Christie et al. (2012).

6. Conclusions We conclude that the workflow we designed for implementing the cascade approach (Fig. 1) is suitable for assessing the ES flood protection of a wetland in a systematic, consistent and comprehensive way. While each of the six elements of the workflow provides important information on its own, it is the combination, with qualitative and quantitative characterizations, that provides a consistent and rather complete assessment of the ES. The division of the assessment into the two phases “assessment Table 9 Questions that help identifying the value of the ES flood protection in a certain area. Questions What is the population density in the floodplain? Do functioning evacuation plans and civil protection exist? Are there any special precautions against flood damages, e.g. on buildings? Were recent flood events existence- or life-threatening? Are there any high risk sites or important cultural heritages within the floodplain? 6 What possibility does society have to provide a technical substitute for this ES? 7 Are there other wetlands in the area that could provide the same protection performance? 1 2 3 4 5

49

of current ES” and “assessment of future ES” provides a clear assessment structure, as does the sequence of (1) qualitatively describing relevant biophysical structures and processes, (2) defining key biophysical structures and processes as main and an additional ecosystem functions, and quantifying both with state indicators by using easily available data, (3) qualitatively describing social and economic benefits that humans enjoy due to the ecosystem functions, and (4) quantitatively valuing the benefits. Following this sequential workflow, a characterization of the current provisioning of the ES can be achieved in the first phase of the assessment. The second phase on future provisioning of the ES could follow the same steps if it were important to analyze e.g. the impact of future climate change or socio-economic changes on ES provisioning. Otherwise, like in our study, the second phase can be restricted to considering pressures on and possible future enhancements of the ES. A major impediment for quantifying the ecosystem function leading to the ES flood protection is the need for costly site-specific hydrological-hydraulic modeling of water storage and inundation in case of flood events. Such a modeling study would ideally encompass simulations for a large number of statistical floods and for a number of alternative scenarios with respect to the state of the ecosystem under consideration (in our case with the wetland under current condition and with the wetland being disconnected from the river). We found that flood hazard maps that were generated throughout Europe due to implementation of the EU floods directive may be used to quantify the main ecosystem function water retention and to monetize the ES. Flood hazard maps show inundation areas and heights for a number of statistical floods both within the ecosystem that provides the ES and in the area that needs to be protected. They are a sufficient data base for monetizing ES with the replacement cost method that only requires information on the water retention volume of the wetland. As the avoided damage costs method requires information on inundation both with and without the wetland, and flood hazard maps only simulate current conditions, usability of flood hazard maps depends on local conditions. In our case study, we could utilize the existing flood hazard maps to estimate at least a lower bound of ES value for the smallest statistical flood. Already for this type of flood the additional amount of water that would inundate the city if not retained in the wetland exceeded the amount that inundates the city in case of an extreme flood (with the wetland being present). We recommend to do monetary valuation of the ES flood protection by the two well-established replacement cost and avoided damage cost methods. The necessarily different monetary values that are derived by using the two methods are relevant in different decision-making contexts. In addition, we suggest advancing the standardization of the two valuation methods (Tables 7 and 8). Then, the results obtained in different studies would be better comparable to studies in which the same valuation method was used. If the surface retention volume of the riparian forest Bulau during an extreme flood had to be replaced by technical flood protection structures, the costs would amount to at least EUR 68 million, i.e. EUR 1.1 million/yr or EUR 1900/ha/yr if a useful life of the flood protection structures of 60 years is assumed. These replacement costs are strongly exceeded by the avoided damage costs. Even if only a 10-year flood is considered, the Bulau avoids damage costs of at least EUR 26 million, equivalent to EUR 2.6 million/yr or EUR 4300/ha/yr. For a number of reasons, these monetary values are very conservative estimates, and the actual replacement and avoided damage costs are certainly higher. Monetary valuation delivers tangible economic reasons to protect the ES. However, monetary values derived from marketcost based methods underestimate the total value of an ES.

50

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

Benefits of the ES to human well-being such as protection of sense of security and human health should explored by additional valuation methods. Further research into non-monetary valuation of the ES flood protection by riparian wetlands is recommended. The method for calculating the ecosystem function water retention as well as inundation with and without the wetland from flood hazard maps, i.e. the basic information for applying the described monetary valuation methods, should be tested for other areas in Europe where flood hazard and risks maps as required by the EU flood directive exist. It should be investigated under what circumstances such maps can replace costly site-specific hydrological-hydraulic modeling.

Acknowledgements This paper is based on the master thesis of Nina Barth who was co-supervised by Nina Stiehr in the framework of her doctoral dissertation; we thank her for the support. We thank Tuck Fatt Siew for his insightful comments to the manuscript. In addition, we are grateful for the comments of the two anonymous reviewers that helped us to improve the manuscript.

Appendix

Table A1 Description of the assessment of the ES flood protection of the riparian forest Bulau according to the Purpose, Scope, Analysis, Recommendations, and Monitoring (PSARM) Blueprint for reporting ecosystem service studies (Seppelt et al., 2012). (1) Purpose and design

Scope of study: Assessment (including valuation) of the ecosystem service (ES) flood protection of a riparian forest in Germany by applying a cascade approach. Project goals: Design of a workflow for a comprehensive and consistent ES assessment that implements the cascade approach; assessment of the current state of the ES with (1) its biophysical processes and structures, (2) definition and quantification of main and additional ecosystem function, (3) qualitative description of economic and social benefits and (4) valuation of the ES (calculation of the replacement cost & damage costs avoided on the basis of easily available data, and without site-specific hydrological-hydraulic modeling, suggestions for non-monetary valuation); addressing the future ES provisioning by identifying pressures and analyzing potential enhancements. Main threats: The ES flood protection of unprotected riparian forests and other natural floodplain vegetation faces major pressures worldwide, e.g. from conversion to urban or agricultural land use. Targets: To provide a well-structured method for the comprehensive assessment of the ES flood protection. Examination of relevant questions of the ES approach, amongst others (de Groot et al., 2010): Which indicators are suitable to analyze the capacity of an ecosystem to provide a certain ES? Is there easily available data for the quantification? What methods are best suited for the assessment of the ES (which one delivers the main monetary value)? How can the comparability of the results of the same valuation method be improved? Team of scientist: Nina-Christin Barth, Prof. Dr. Petra Döll, Goethe University Frankfurt, Frankfurt am Main, Germany. Stakeholders: none

(2) Scope of problemscape and concept

System description, landscape and policy measures: The riparian forest Bulau (604 ha) acts as large flood water storage in case of a flood event in the river Kinzig and protects parts of the city of Hanau that is directly located downstream. The Bulau is part of the floodplain of the river Kinzig and a protected area according to the Flora-Fauna-Habitat Directive of the EU. Ecosystem service: The contribution of the ecosystem Bulau to the well-being of flood-prone residents is an ecosystem service (ES) (Haines-Young and Potschin, 2009) that we defined as flood protection. Storyline of possible futures: To calculate the avoided damage costs we created an alternative scenario without the ecosystem function water retention of the Bulau (e.g. with dykes to protect t a residential area from being flooded). We examined the change of the ecosystem function regulation of direct runoff for three land use scenarios: grassland, intensive agriculture and settlement area with an average sealing of 40–50%.

(3) Analysis, assessment

Assessment based on a modified cascade approach: (1) Current ES: Description of biophysical processes and structures; Definition of the main ecosystem function “water retention” and the additional ecosystem function “regulation of direct runoff” Ecosystem functions quantified by the Indicators: surface and subsurface retention volume [m3], based on hazard maps for three statistical floods as developed due to implementation of EU floods directive; direct runoff [mm/a] and regulation of direct runoff on x% of the total catchment area; Description of the economic and social benefits; Valuation & documentation of methods applied: Monetary - replacement costs and damage costs avoided (on the basis of easily available primary data including flood hazard and flood risk maps, without site-specific hydrological-hydraulic modeling). Non-monetary – presentation of questions about exposure and vulnerability that can form the data basis for recommended non-monetary valuation. Test with real world data: Monetary results are within the range of other studies (Leschine et al., 1997; Jiang et al., 2007; van der Ploeg and de Groot, 2010). (2) Future ES: Identification of pressures for safe-guarding the future provisioning of the ES; Scenario analysis to address potential enhancements of the provision of the ES (Quantification scenarios: Computation of land use-dependent direct runoff by simple water balance model.) Suggestions for standardizing application of monetary valuation methods replacement cost and avoided damage costs.

(4) Recommendations and results

Monetary results: The costs of a comparable technical flood protection structure with the same retention performance as the ecosystem would amount to 68 million EUR considering an extreme flood (1,880 EUR/ha/yr). In addition, the flood protection of the riparian forest avoids damage costs of at least 26 million EUR in case of a 10-year flood (4,279 EUR/ha/yr) (very conservative estimate for various reasons). Summary: The developed workflow implements the cascade approach in a structured and sequential way and enables a consistent and comprehensive assessment of the ES flood protection. To improve comparability of monetary ES values, we suggest the definition of standardized application guidelines of the replacement costs and avoided damage costs methods. Recommendations: The method for calculating the ecosystem function water retention as well as inundation with and without the wetland, i.e. the basic information for applying the described monetary valuation methods, should be tested, e.g. for other areas in Europe where flood hazard and risks maps as required by the EU flood directive exist. It should be investigated under what circumstances such maps can replace costly site-specific hydrological-hydraulic modeling. Further research should examine how to value social benefits that are not covered by monetary valuation methods, and how to combine the results of an overall assessment, since none of the values represent the true value of the ES. Indicators: Subsurface and surface water retention volume, regulation of direct runoff

(5) Monitoring

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

References ACOE - U.S. Army Corps of Engineers, 1976. Water Resources Development Plan, Charles River Watershed, Massachusetts, New England Division. AG Boden, 2005. Bodenkundliche Kartieranleitung, 5. Aufl. Bundesanstalt für Geowissenschaften und Rohstoffe und Niedersächsisches Landesamt für Bodenforschung. Schweizerbart, Stuttgart. BMU and BfN - Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit and Bundesamt für Naturschutz, 2009. Auenzustandsbericht: Flussauen in Deutschland. Brander, L., Brouwer, R., Wagtendonk, A., 2013. Economic valuation of regulating services provided by wetlands in agricultural landscapes: a meta-analyses. Ecol. Eng. 56, 89–96. Buttler, Dr. K.P., Hemm, K., Fehlow, M., 2003. Grunddatenerfassung für Monitoring und Management im FFH-Gebiet 5819-308 “Erlensee bei Erlensee und Bulau bei Hanau”. Regierungspräsidium Darmstadt. Carson, R.T., Flores, N.E., Meade, N.F., 2001. Contingent valuation: controversies and evidence. Environ. Resour. Econ. 19, 173–210. Christie, M., Fazey, I., Cooper, R., Hyde, T., Kenter, J.O., 2012. An evaluation of monetary and non-monetary techniques for assessing the importance of biodiversity and ecosystem services to people in countries with developing economies. Ecol. Econ. 83, 67–78. Cordier, M., Pérez Agúndez, J.A., Hecq, W., Hamaide, B., 2014. A guiding framework for ecosystem services monetization in ecological–economic modeling. Ecosyst. Serv. 8, 86–96. Damm, C., Dister, E., Fahlke, N., 2011. Auenschutz – Hochwasserschutz – Wasserkraftnutzung: Beispiele für eine ökologisch vorbildliche Praxis. In: Bundesamt für Naturschutz Bad Godesberg (Eds.), Naturschutz und Biologische Vielfalt Heft 112. De Groot, R.S., Alkemade, R., Braat, L., Hein, L., Willemen, L., 2010. Challenges in integrating the concept of ecosystem services and values in landscape planning, management and decision making. Ecol. Complex. 7, 260–272. De Groot, R.S., Brander, L., Van der Ploeg, S., Costanza, R., Bernard, F., Braat, L., Christie, M., Crossman, N., Ghermandi, A., Hein, L., Hussain, S., Kumar, P., Mc Vittie, A., Portela, R., Rodriguez, L.C., Ten Brink, P., Van Beukering, P., 2012. Global estimates of the value of ecosystems and their services in monetary units. Ecosyst. Serv. 1 (2012), 50–61. Dehnhardt, A., Meyerhoff, J., 2012. Ökonomische Bewertung der Ökosystemdienstleistungen von Auen. Methodische Ansätze und Ergebnisse des Fallbeispiels Elbe. BfN, Internationale Naturschutzakademie: Der Nutzen von Ökonomie und Ökosystemdienstleistungen für die Naturschutzpraxis. Workshop II: Auen, Moore und Gewässer, April 2012, Insel Vilm. URL: 〈http://www.bfn.de/fi leadmin/MDB/documents/ina/vortraege/2012/2012-OekonomiePraxis-IIMeyerhoff-Dehnhardt.pdf〉 (20.10.14). Ditter, G., 1991. Hydrographische und hydrologische Untersuchungen über das Hochwasserabflussverhalten des Kinziggebietes (Maingebiet). Institut für Physische Geographie, Johann Wolfgang Goethe-Universität Frankfurt am Main. Dittrich, S., Worm, W., 2006. Dezentraler Hochwasserschutz. In: Schriftenreihe der Sächsischen Landesanstalt für Landwirtschaft, Heft 11, 2006. Sächsische Landesanstalt für Landwirtschaft, Dresden. URL: 〈http://d-nb.info/996225080/34〉, (20.10.14). Döll, P., Jiménez-Cisneros, B., Oki, T., Arnell, N.W., Benito, G., Cogley, J.G., Jiang, T., Kundzewicz, Z.W., Mwakalila, S., Jiang Nishijima, A., 2015. Integrating risks of climate change into water management. Hydrol. Sci. J. 60 (1), 3–14. http://dx. doi.org/10.1080/02626667.2014.967250. DWA - Deutsche Vereinigung für Wasserwirtschaft, Abwasser und Abfall e. V., 2012. Leitlinien zur Durchführung dynamischer Kostenvergleichsrechnungen (KVRLeitlinien): 8. überarbeitete Auflage, Juli 2012. Herausgeber 1. – 7. Auflage: LAWA. Emerton, L., Bos, E., 2004. Value: Counting Ecosystems as an Economic Part of Water Infrastructure. IUCN, Gland, Switzerland und Cambridge, UK. EU - European Union, 2007. Directive 2007/60/EC of the European Parliament and of the Council of 23 October 2007 on the assessment and management of flood risks, Official Journal of the European Union L 288 of 06.11.2007. Federal Water Act, 2009. Wasserhaushaltsgesetz vom 31. Juli 2009 (BGBl. I S. 2585), das zuletzt durch Artikel 4 Absatz 76 des Gesetzes vom 7. August 2013 (BGBl. I S. 3154) geändert worden ist. § 78: Besondere Schutzvorschriften für festgesetzte Überschwemmungsgebiete. URL: 〈http://www.gesetze-im-internet.de/ bundesrecht/whg_2009/gesamt.pdf〉 (letzter Aufruf: 20.10.2014). Gerrard, P., 2004. Integrating Wetland Ecosystem Values into Urban Planning: The Case of That Luang Marsh, Vientiane, Lao PDR. IUCN  The World Conservation Union, Asia Regional Environmental Economics Programme and WWF Lao Country Office, Vientian. URL: 〈http://www.mekongwetlands.org/Common/ download/WANI_economics_ThatLuang%20Marsh.pdf〉 (20.10.14). Goulder, L.H., Kennedy, D., 1997. Valuing ecosystem services: philosophical bases and empirical methods. In: Daily, G.C. (Ed.), Nature's Services: Societal Dependence on Natural Ecosystems. Island Press, Washington, DC, pp. 23–47. Grossmann, M., 2012. Economic value of the nutrient retention function of restored floodplain wetlands in the Elbe River basin. Ecol. Econ. 83, 108–117. http://dx. doi.org/10.1016/j.ecolecon.2012.03.008. Grygoruk, M., Mirosław-Świątek, D., Chrzanowska, W., Ignar, S., 2013. How much for water? Economic assessment and mapping of floodplain water storage as a catchment-scale ecosystem service of Wetlands. Water 5, 1760–1779. Haines-Young, R.H., Potschin, M.B., 2009. Methodologies for defining and assessing ecosystem services. Final Report, JNCC, Project Code C08-0170-0062. URL:

51

〈http://www.nottingham.ac.uk/cem/pdf/JNCC_Review_Final_051109.pdf〉, (20.10.14). Haines-Young, R.H., Potschin, M.B., 2010. The links between biodiversity, ecosystem services and human well-being. In: Raffaelli, D.G, Frid, C.L.J. (Eds.), Ecosystem Ecology: A New Synthesis. Cambridge University Press, British Ecological Society, 110-139. URL: 〈http://www.nottingham.ac.uk/cem/pdf/Haines-Young andPotschin_2010.pdf〉, (20.10.14). Hainke, T., 2014. EUR-DM-Rechner. 〈http://www.altersvorsorge-und-inflation.de/ euro-rechner.php?dm_eur ¼ DM_EUR〉, (20.10.14). Handmer, J., Honda, Y., Kundzewicz, Z.W., Arnell, N., Benito, G,. Hatfield, J., Mohamed, I.F., Peduzzi, P., Wu, S., Sherstyukov, B., Takahashi, K., Yan, Z., 2012: Changes in impacts of climate extremes: Human systems and ecosystems. In: Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mastrandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, M. Tignor, and P.M. Midgley (eds.), Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK and New York, USA, 231-290. Hessischer Landtag, 1988. Antwort des Ministers für Umwelt und Reaktorsicherheit auf die kleine Anfrage der Ag. Dr. Müller (Gelnhausen) und Lenz (Hanau) (CDU) betreffend Stausee bei Bad Soden-Salmünster, Drucksache 12/1722 (25.08.1988). URL: 〈http://starweb.hessen.de/cache/DRS/12/2/02842.pdf〉 (20.10.14). HGN – Hydrogeologie GmbH Nordhausen, 2004. Retentionskataster Flussgebiet Kinzig,km 0þ000 bis 73 þ354. URL: 〈http://static.hlug.de/medien/wasser/rkh/ berichte/2478_Kinzig_km0_bis_km73.354.pdf〉, (01.03.14). HLUG - Hessisches Landesamt für Umwelt und Geologie (Eds.), 2011. Das JanuarHochwasser 2011 in Hessen. Hydrologie in Hessen Heft 6. HLUG – Hessisches Landesamt für Umwelt und Geologie, 2014a. Raster data sets with inundation heights and shapefiles containing flood zones and respective land use categories for the statistical flood HQ10, HQ100 and HQex. HLUG – Hessisches Landesamt für Umwelt und Geologie, 2014b. Digitale Bodenkarte von Hessen 1:25.000, Blatt 5819 Hanau (BK 25) and Themenkarten der Bodenflächendaten 1:50.000, L5918 Frankfurt Ost (BFD 50). HMUELV - Hessisches Ministerium für Umwelt, ländlichen Raum und Verbraucherschutz, 2007. Landesaktionsplan Hochwasserschutz Hessen. URL: 〈https://umweltministerium.hessen.de/sites/default/files/HMUELV/land esaktionsplanhochwasserschutzhessen.pdf〉 (20.10.14). Hunter, M.L., Redford, K., Lindenmayer, D.B., 2014. The complementary niches of anthropocentric and biocentric conservationists. Conserv. Biol. 28 (3), 641–645. IKSR – Internationale Kommission zum Schutz des Rheins, 2001a. Atlas der Überschwemmungsgefährdung und möglichen Schäden bei Extremhochwasser am Rhein. IKSR – Internationale Kommission zum Schutz des Rheins, 2001b. Übersichtskarten der Überschwemmungsgefährdung und der möglichen Vermögensschäden am Rhein: Abschlussbericht: Vorgehensweise zur Ermittlung der hochwassergefährdeten Flächen, Vorgehensweise zur Ermittlung der möglichen Vermögensschäden. Bearbeiter: Ruiz Rodriguez þZeisler (Wiesbaden), geomer GmbH (Heidelberg), PlanEVAL (München), Haskong (Nijmegen). Jiang, M., Lu, X., Xu, L., Chu, L., Tong, S., 2007. Flood mitigation benefit of wetland soil – a case study in Momoge National Nature Reserve in China. Ecol. Econ. 61, 217–223. Kousky, C., Walls, M., 2014. Floodplain conservation as a flood mitigation strategy: examining costs and benefits. Ecol. Econ. 104, 119–128. Kundzewicz, Z.W., Kanae, S., Seneviratne, S.L., Handmer, J., Nicholls, N., Peduzzi, P., Mechler, R., Bouwer, L.M., Arnell, N., Mach, K., Muir-Wood, R., Brakenridge, G.R., Kron, W., Benito, G., Honda, Y., Takahashi, K., Sherstyukov, B., 2014. Flood risk and climate change: Global and regional perspectives. Hydrol. Sci. J. 59 (1), 1–28. http://dx.doi.org/10.1080/02626667.2013.857411. Leschine, T.M., Wellman, K.F., Green, T.H., 1997. The Economic Value of Wetlands: Wetland's Role in Flood Protection in Western Washington. Ecology Publication No. 97-100. Washington State Department of Ecology. MA - Millennium Ecosystem Assessment, 2005a. Ecosystems and Human WellBeing Synthesis. Island Press, Washington, DC. MA - Millennium Ecosystem Assessment, 2005b. Ecosystems and Human WellBeing: Wetlands and Water Synthesis. World Resources Institute, Washington, DC. Meßer, J., 2013. Ein vereinfachtes Verfahren zur Berechnung der flächendifferenzierten Grundwasserneubildung in Mitteleuropa. Veröffentlichung der Emscher Wassertechnik GmbH, Lippe Wassertechnik GmbH. URL: 〈http://www. gwneu.de/downloads.html〉, (20.10.14). Müller, U., 2010. Hochwasserrisikomanagement: Theorie und Praxis: PRAXIS. Viewegþ Teubner Verlag, Springer Fachmedien Wiesbaden GmbH. Nedkov, S., Burkhard, B., 2012. Flood regulating ecosystem services — Mapping supply and demand, in the Etropole municipality, Bulgaria. Ecol. Indic. 21, 67–79. Peh, K.S.-H., Balmford, A.P., Bradbury, R.B., Brown, C., Butchart, S.H.M., Hughes, F.M. R., et al., 2013. TESSA: a toolkit for rapid assessment of ecosystem services at sites of biodiversity conservation importance. Ecosyst. Serv. 5, 51–57. Peh, K.S.-H., Balmford, A.P., Field, R.H., Lamb, A., Birch, Bradbury, R.B., et al., 2014. Benefits and costs of ecological restoration: rapid assessment of changing ecosystem service values at a U.K. Wetl. Ecol. Evol. 2014 20 (4), 3875–3886. http://dx.doi.org/10.1002/ece3.1248. Posthumus, H., Rouquette, J.R., Morris, J., Gowing, D.J.G., Hess, T.M., 2010. A framework for the assessment of ecosystem goods and services; a case study on lowland floodplains in England. Ecol. Econ. 69, 1510–1523.

52

N.-C. Barth, P. Döll / Ecosystem Services 21 (2016) 39–52

Rounsevell, M.D.A., Dawson, T.P., Harrison, P.A., 2010. A conceptual framework to assess the effects of environmental change on ecosystem services. Biodivers. Conserv. 19, 2823–2842. Russi, D., ten Brink, P., Farmer, A., Badura, T., Coates, D., Förster, J., Kumar, R., Davidson N., 2013. The Economics of Ecosystems and Biodiversity for Water and Wetlands. IEEP, London/Brussels; Ramsar Secretariat, Gland. URL: 〈http://www. teebweb.org/wp-content/uploads/2013/04/TEEB_WaterWetlands_Report_2013. pdf〉 (20.10.14). Seppelt, 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. Form follows function? Proposing a blueprint for ecosystem service assessments based on reviews and case studies. Ecol. Indic. 21, 145–154. Scholz, M., Mehl, D., Schulz-Zunkel, C., 2012. Ökosystemfunktionen von Flussauen: Analyse und Bewertung von Hochwasserretention, Nährstoffrückhalt, Kohlenstoffvorrat, Treibhausgasemissionen und Habitatfunktionen. In: Bundesamt für Naturschutz Bad Godesberg (Eds.), Naturschutz und Biologische. Shabman, L., Batie, S., 1978. Economic value of natural costal wetlands: a critique. Coast. Zone Manage. J. 4, 231–247. Stadt Hanau, 2015. Statistische Basisdaten. URL: 〈http://www.hanau.de/rathaus/ statistik/daten/005334/#anchor_10_65〉, (16.04.15). TEEB, 2010a. Integrating the ecological and economic dimensions in biodiversity

and ecosystem service valuation. In: Kumar, Pushpam (Ed.), The Economics of Ecosystems and Biodiversity: Ecological and Economic Foundations. Earthscan, London and Washington (14.10.14) http://www.teebweb.org/wp-content/up loads/2013/04/D0-Chapter-1-Integrating-the-ecological-and-economic-dimen sions-in-biodiversity-and-ecosystem-service-valuation.pdf. TEEB, 2010b. Socio-cultural context of ecosystem and biodiversity valuation. In: Kumar, Pushpam (Ed.), The Economics of Ecosystems and Biodiversity: Ecological and Economic Foundations. Earthscan, London and Washington (20.10.14) http://doc.teebweb.org/wp-content/uploads/2013/04/D0-Chapter-4-Socio-cul tural-context-of-ecosystem-and-biodiversity-valuation.pdf. TEEB, 2014. “Benefit” and “Valuation”. URL: 〈http://www.teebweb.org/resources/ glossary-of-terms/〉, (20.10.14). Van der Ploeg, S., de Groot, R.S., 2010. The TEEB Valuation Database – a searchable database of 1310 estimates of monetary values of ecosystem services. Found. Sustain. Dev., Wagening., Neth. Verlag Parzeller GmbH and Co. KG, 2013. Rückhaltebecken an der Salz zum Schutz vor Hochwasser (28.06.2013). URL: 〈http://www.fuldaerzeitung.de/artike lansicht/artikel/175685/hochwasserschutz-an-der-salz-661182〉, (20.10.14). Zhai, F., Sato, T., Fukuzono, T., Ikeda, S., Yoshida, K., 2006. Willingness to pay for flood risk reduction and its determinants in Japan. J. Am. Water Res. Assoc. 42 (4), 927–940. http://dx.doi.org/10.1111/j.1752-1688.2006.tb04505.x.