Using the River Ecosystem Service Index to evaluate “ Free Moving Rivers ” restoration measures: A case study on the Ammer river (Bavaria)

Restoring natural fluvial dynamics is fundamental for sustaining biodiversity and functional integrity of river and floodplain ecosystems. In Central Europe, however, pervasive river regulation and bank protection have greatly impaired ecosystem functioning and many water bodies fail to achieve a good ecological status within the European Water Framework Directive. The “ Free Moving Rivers ” approach seeks to restore the ecological integrity of rivers and floodplains by creating appropriate conditions for natural fluvial dynamics. Principal goals of the approach include removing artificial constraints on river processes and expanding the river corridor to restore natural river habitats and structures. Lacking, however, are complementary tools that evaluate and predict changes to ecosystem services (ESSs) after implementation. Here, we describe a case study of the Ammer river in Bavaria, Germany, to (i) calculate the extent of the “ Free Moving Rivers ” corridor, and (ii) assess changes to ESSs of a proposed river restoration measure under two alternative land ‐ use scenarios. To do this, we apply the River Ecosystem Service Index (RESI), whereby individual ESSs are assessed in a spatially explicit way. We show how a proposed implementation of the “ Free Moving Rivers ” approach enhances three investigated ESSs: flood retention, sediment balance and habitat provision. We conclude that RESI is a potentially useful tool with wide applicability for restoration planning that synthesises floodplain complexity in such a way that facilitates decision making. Due to their diversity, varied hydrological high rates of sediment and biomass


| INTRODUCTION
Natural rivers and floodplains are valuable landscape components that provide various benefits for human well-being and ecological functioning (Fischer et al., 2019;Robinson et al., 2002;Schneider et al., 2018;Scholz et al., 2012;Ward et al., 1999). Due to their geomorphic diversity, their varied hydrological conditions, high rates of sediment and biomass turnover, and high productivity, they provide a multitude of ecological functions. These result in numerous ecosystem services (ESSs), such as the provision of food, clean water, fuel, the regulation of flood and drought, the reduction of excess nutrients, the provision of habitats with a high level of biodiversity, cultural identification and cultural assets, like, recreational hiking, fishing and natural experience (Costanza et al., 1997;Feld et al., 2011;Millennium Ecosystem Assessment (MA), 2005;Tockner & Stanford, 2002). Despite their great ecological value and the ESSs they provide, the functional integrity of many European river landscapes is impaired (Schinegger et al., 2012), with 62% having moderate to poor ecological status (Grizzetti et al., 2017). In Germany, about two-thirds of the former river floodplain areas are separated from their river channels and only 9% are classified as near-natural (Bundesministerium für Umwelt Naturschutz und nukleare Sicherheit [BMU] & Bundesamt für Naturschutz [BfN], 2021). A principal reason for the poor ecological status of European river landscapes is the reduction of fluvial dynamics caused by hydraulic-engineering measures, such as channelisation, bank protection, damming, straightening of river courses, the building of levees and intensive land-use adjacent to rivers (Brunotte et al., 2009;Feld et al., 2011;Schneider et al., 2018;Tockner et al., 2010).
Because of a strong correspondence between fluvial dynamics and the provision of ESSs, modern concepts of sustainable floodplain restoration are often based on providing space for re-establishing natural fluvial processes (Biron et al., 2014;Buffin-Bélanger et al., 2015;Piégay et al., 2005;Ward et al., 2001;Wohl et al., 2005). Enlarged floodplain corridors, for example, not only provide natural flood retention areas addressing increased flood risk in particular concerning climate change (Lobanova et al., 2018), but also sustain a natural mosaic of habitats for floodplain-specific biota (Scholz et al., 2012;Ward et al., 1999). Reestablishing fluvial dynamics therefore directly addresses ongoing biodiversity and climate crises, as well as regional governance objectives (e.g., Natura 2000) and sustainable water usage and management (e.g., Water Framework Directive [WFD]; Biron et al., 2014;Piégay et al., 2005).
The French "Espace de Liberté" approach (Freedom Space for Rivers) was the first in Europe to use characteristics of historical river mobility as a basis for recovering fluvial dynamics (Malavoi et al., 1998). Within newly expanded corridors, morphodynamic river processes are allowed to proceed with minimal constraints (Charrier, 2012;Sauvade et al., 2015). The German "Free Moving Rivers" approach ("Freier Pendelraum für Fließgewässer") is also based on this idea, seeking to provide a minimum space for ecologically functioning rivers and floodplains to re-establish important ESSs with positive effects for humans and biodiversity alike . Implementation requires an appropriately defined river corridor without settlements close to the river and the absence or removal of artificial structures that constrain lateral river movement. A natural or largely unaltered flow regime allowing natural erosive processes is also essential . Bank protection measures remain only near essential infrastructure, such as bridge pillars. Restorations of formerly altered river sections are likely to increase native biodiversity (Kail et al., 2015;Meli et al., 2014;Schindler et al., 2016;Tomscha et al., 2021), add to the multifunctionality of river systems, and thus increase ESS supply (Funk et al., 2021;Meli et al., 2014;Schindler et al., 2014). The environmental and biological complexity of river and floodplain ecosystems and their multifunctionality thus require multidisciplinary assessments that not only quantify positive and negative effects of restoration measures, but also synthesise these in such a way that facilitates decision making (Funk et al., 2021). The River Ecosystem Service Index (RESI) provides a spatially explicit tool for this multiperspective analysis (Pusch, 2016;Podschun, Thiele, et al., 2018). The index evaluates restoration measures on rivers and floodplains using existing data or, alternatively, is based on assumptions of future conditions (Stammel et al., 2021). Thus, it can be useful in establishing benchmarks for evaluating project goals and successes, as well as for comparing changes to ecosystems services under different hypothetical scenarios (Fischer et al., 2019). Furthermore, effects on ESSs can be illustrated in an easily understandable and spatially explicit format that enhances communication with public authorities and local stakeholders (Podschun, Thiele, et al., 2018).
Here, we apply the RESI approach to a proposed restoration We assessed the space requirements for restoration measures and resulting changes to ESSs on the lower Ammer river by answering the following questions: (i) What is the minimum space required to implement the "Free Moving Rivers" approach in the study area on the lower Ammer river?
(ii) How will ESS provision change after implementing the proposed "Free Moving Rivers" restoration measures? 2 | METHODS

| Study area
The Ammer river originates in the Northern limestone Alps in Bavaria close to the Austrian border and flows into the Ammersee lake (Ringler et al., 2000). The study area is located upstream of the Annual precipitation at the study site totals 956 mm and the mean temperature is 8.5°C (Climate-Data, 2020).

| Calculation of the "Free Moving Rivers" corridor
The "Free Moving Rivers" approach  determines an appropriate width of a river corridor sufficient for the development of habitat diversity and geomorphological features. This corridor is based on geomorphological and hydrological variables of the given river type. River type is initially classified according to the scheme of Pottgießer and Sommerhäuser (2008) which takes into account the geographic location of the river, hydromorphological, biological and physicochemical parameters. Each river type is associated with a set of factors which, when multiplied by the current river width, estimate a potential natural river bed width (current river width multiplied by 3), a minimum river corridor (potential natural river bed multiplied by 3) and maximum river corridor (potential natural river bed multiplied by 10). These are meant to be added as lateral buffers to each side of the river (Dahm et al., 2014). The corridor is later modified to exclude man-made structures which must not be flooded, such as cities or buildings. In our study area (Figure 1  Flood retention judges the natural capability of floodplains to reduce peak flow discharge and is a function of retention volume and the systematic mapping of river and floodplain structure . Retention volume is the ratio of ac- Sediment balance refers to the capacity of rivers to transport sediment and suspended material, and is linked to erosion and sedimentation processes . The sediment transport capacity of rivers is reduced by structures that impede sediment flow such as reservoirs, dams, weirs, sills or ramps as well as by structures that inhibit erosive processes, such as bank protections. The calculation is based on the river bed assessment of the Bavarian river structure data (Environment Agency of Bavaria [LfU], 2017).
These are manually altered according to assumed changes of restoration measures using expert-defined improvements due to the removed structures. An evaluation of the effect of these regulations for each river segment is converted into five classes ranging between class 5, with a mostly unimpeded sediment balance, and class 1, with a highly disturbed sediment balance .
Habitat provision assesses structural and functional habitat quality for floodplain-specific biota (Fischer et al., 2019). In contrast to flood retention and sediment balance where assessments are performed on the whole morphological floodplain, habitat provision is calculated only in the proposed restoration area allowing for the representation of single habitats that would otherwise not be represented by the area-averaged assessment. Current habitat types were classified according to the biotope assessment of Bavaria (Environment Agency of Bavaria [LfU], 2020) and the land-use model of the Federal Agency for Cartography and F I G U R E 2 Schematic figure of morphological floodplain for calculating RESI at river sections (modified after Lotti, 2018). RESI, River Ecosystem Service Index Geodesy (LBM-DE2018, BKG) which were revised by ground-truthing during field investigations. The study area is mostly covered by grassland (about 14 ha) and riparian forests (11 ha). Several water bodies such as abandoned meanders and lateral tributaries lie in the area (almost 1 ha in total). Biotope types are assigned an expert-defined habitat value (1-5) defined in the RESI manual and based on land-use and floodplain uniqueness. This value is reduced if the habitat is affected by artificially caused impounded water or is located in the inactive floodplain and increased by good conservation status of Natura 2000 habitat types . This results in an ordinal scaled assessment between class 5 which describes areas with a very high importance for habitat provision and class 1 with very low importance .

| Development of scenarios
We assessed the current conditions (status quo) in the study area and

| Dimension of the "Free Moving Rivers" corridor
The lower Ammer river is classified as a small river of the upper moraine in the Alpine foreland (river type 3.2 according to Pottgiesser & Sommerhäuser, 2008). The current modified river bed width of 25 m results in a potential natural river bed width of 75 m, a "Free Moving River" minimum corridor of 225 m and a maximum corridor of 750 m (according to Dahm et al., 2014).
F I G U R E 3 Minimum and maximum "Free Moving Rivers" corridor according to the approach of the German Federal Environment Agency (Dahm et al., 2014) for the area of the proposed restoration measure BECKER ET AL.

| 5
Constraints that narrow the river corridor in the study area are the cities of Peißenberg and Weilheim and to a smaller extent roads and settlements (Figure 3 inset). The proposed restoration measure is located between the minimum and maximum river corridor (Figure 3).

| RESI assessment
Results of RESI evaluation indicate weak to strong enhancements for the three ESSs assessed through the restoration measure.

| Flood retention
Using the ratio between active and inactive floodplain, the RESI flood retention index of the status quo reveals settled areas with high floodplain losses (low RESI values) but also areas in the centre with preserved inundation capacity and therefore higher RESI values ( Figure S1 left, see Supporting Information). In the proposed restoration area one segment improves from moderate to low loss ( Figure S1 right, see Supporting Information).

| Sediment balance
In the status quo, anthropogenic impacts on the sediment balance by technical structures are mainly low to moderate with two very low segments at the eastern end of the section (Figure 4 left). The proposed removal of the bank fixations enhances the parameter "bank protection" in the assessment data of river structure (Environment Agency of Bavaria [LfU], 2017), thus leading to an upgrade of many segments in the restoration area mainly by one RESI-class (Figure 4 right).

| Habitat provision
In the status quo, areas close to the river are covered by riparian forest whereas areas distant to the river are dominated by grassland

| DISCUSSION
In the study area on the lower Ammer river, the calculated "Free  (Müller, 2017) and on a straightened section close to the mouth into the Ammersee lake . However, fundamental changes in site conditions of rivers always have to be considered when planning to implement "Free Moving Rivers" corridors. Human impacts can lead to profound changes that may alter entire river types (Gregory, 2006). As shown for the Ammer downstream of the proposed restoration measure, shortening of the river course and regulation measures lead to incision of the river bed turning a former multichannel river into a single thread river type (Guzelj et al., 2020).
In this case, providing space for the river is not likely to re-establish its historical river type. Consequently, restoration objectives have to be adapted accordingly and changed river types have to be considered in calculations (Shields et al., 2003) making feasibility studies generally necessary before implementation of river revitalisation projects (Stammel et al., 2021).
Possible limitations of restoration measures are shown by a metaanalysis of studies on different riparian ecosystems which revealed lower biodiversity and ESSs in restored than in pristine sites (Meli et al., 2014; for different ecosystems see also Benayas et al., 2009).
Nevertheless, removing structures that reduce natural river processes in formerly altered areas enhances rivers' ability to provide ESSs (Dehnhardt, 2002;Scholz et al., 2012). Restoration measures often increase the multifunctionality of sites by enhancing the provision of various ESSs. For example, Funk et al. (2021) found the removal of embankments and connection of side channels on a Danube river section resulting in an increase in habitat availability and nutrient retention.
Dyke relocations lead to an increase in many wetland ESSs, such as recreational opportunities, water availability and food supply or nutrient and water quality regulation (Hornung et al., 2019;Schindler et al., 2014). Together with reduced maintenance costs for bank protection this can lead to economic benefits, even if decreases in agricultural or forest productivity are included (Buffin-Bélanger et al., 2015;Grossmann et al., 2010). For example, Buffin-Bélanger et al. (2015) quantified ESSs and costs of implemented river corridors for three rivers in Quebec over a 50-year period, finding that monetary and societal benefits increased 1-5-fold over costs.
In this paper, the RESI approach was used on a limited set of ESSs considered important for the Ammer case study. It could have been expanded by adding other ESSs. Customising the set of evaluated ESSs is indeed one of the assets of the RESI method (Stammel et al., 2021). Cultural ESSs could have been added, with other studies revealing a decrease by reduced length of roads resulting in smaller areas with recreational functions (Vallecillo et al., 2019;Woolsey et al., 2007) which might also be the case by the restoration measure in the study area. However, reduced impact of tourists may also reduce physical disturbance of sites or nutrient input (Hornung et al., 2019).
The RESI approach is based on assumed changes in the future development of the site, often relying on expert knowledge that F I G U R E 8 Calculations of three ESSs with RESI in status quo and two scenarios for the proposed restoration measure cannot consider all external factors influencing ESSs. This results in some uncertainty concerning the future development of restoration measures. Data quality might sometimes be a limitation to the method, which largely relies on external data. Nevertheless, calculation of the RESI is simple and comprehensible, and the required data are mostly available across Germany. Different restoration measure options can thus be compared and a holistic evaluation of advantages and impacts can be compared, which facilitates communication with stakeholders and authorities to support decision making (Stammel et al., 2021). The RESI can therefore be considered a useful tool to demonstrate changes in ESSs.

| CONCLUSION
The dynamic and often unpredictable nature of rivers, as well as a lack of unaltered rivers that serve as natural benchmarks for setting restoration goals makes successful restoration planning a difficult task (Ward et al., 2001). Reinstating more natural fluvial processes of the river is therefore considered a very promising pathway. The "Free Moving Rivers" approach offers a fast, simple and transferable framework for the estimation of the required space of river ecosystems for ecological and regulative processes . However, for detailed planning of such measures, especially in highly urbanised areas, more comprehensive studies and hydraulic models are needed.
Assessing changes in river ecosystems can be performed by the RESI approach, which offers a standardised assessment method for a larger number of ESSs beyond those tested here. Thus RESI provides a comprehensive tool for the comparison of different measurement options Podschun, Thiele, et al., 2018).