Adoption of the Ecosystem Services concept in EU policies

38 The concept of ecosystem services has gained a strong political profile during the last 15 39 years. However, there is no specific EU policy devoted to the governing of ecosystem 40 services. This article shows that the ecosystem services concept is already embedded in 41 recent EU (environmentally-related) policies, such as the Biodiversity Strategy 2020 and 42 the Invasive Alien Species Regulation. A review of 12 policies shows that, overall, the 43 coherence between existing policies and the ecosystem services concept is moderate. 44 Policies showing very high coherence are confined to the policy arenas that address 45 natural ecosystems, forestry, or agriculture. Given the sectoral nature of most EU policies 46 and the limited options for revision in the near future, opportunities for improving 47 coherence are mostly expected in improving the integration of the ecosystem services 48 concept in the implementation of existing EU policies at national and regional levels. 49

 To assess the incorporation of the ES concept, its use in twelve EU policies was 28 reviewed 29  Overall, the incorporation of the ES concept in the reviewed policies is moderate, 30 yet it is gradually gaining a presence. as an important factor instigating policy change (Hall, 1993). Why some ideas become 53 policy relevant, and others not, and what triggers their adoption into policies, programs, 54 and philosophies, has been a subject of study in political science (Schmidt, 2008). At a 55 time when streamlining regulation and deregulation have been called for, it is important 56 to understand how new concepts fit in with existing policies (Taylor et al., 2012). The 57 way new concepts are operationalized to become a target and a means of steering, and 58 the ways in which they fit the existing policies, is a matter of policy coherence. Policy 59 coherence usually refers to the extent to which policies complement or are in line with 60 one another or form a meaningful ensemble (Nilsson et al., 2012b). Ensuring policy 61 coherence is particularly important in cases where the policies feature a mode of steering 62 that is detailed or complex. Among these, two major categories of instruments at the EU level can be identified: 1) 95 binding legislative instruments, such as Directives, Regulations, and Decisions, and 2) 96 non-binding program instruments, such as Strategies, Recommendations, and 97 Communications. In this article, 'policies' refers to both binding legislation and non-98 binding programs. 99 Although some specific environmental policy areas derived from particular ideas or 100 concepts can be distinguished, such as pollution prevention or biodiversity conservation, 101 concepts are not always framed as distinct policy areas. Indeed, there is no specific EU 102 policy framework addressing ecosystem services, despite the fast increasing use of the 103 concept. Instead, the ES concept might -and in fact is already to some extent implicitly 104 To get an overview of the EU policies, in which the ES concept is already addressed -121 either explicitly, i.e. actually using ES-related terminology, or implicitly, i.e. by referring 122 to particular ES or with terms referring to ecosystems as complex systems or ecosystem 123 functions -we carried out a policy scanning in three steps. First, a literature and 124 document review resulted in an initial list of 53 EU policies; second, the policies were 125 prioritised based on their relevance for a set of case studies analysing the 126 operationalisation of ecosystem services (OpenNESS, 2012); and, third, EU policy makers 127 identified key policies at a focus group workshop in Brussels in January 2014 (Schleyer et 128 al., 2015). 129 The eleven EU policies selected through this process included both binding and non-130 binding instruments, covering the policy fields of biodiversity, forest, climate, water, and 131 rural and urban areas, as well as a mobility and infrastructure-related policy (see Table 1  132 for policy fields and reviewed policies) (Schleyer et al., 2015). Finally, to account for 133 recent developments in the field of environmental policies, we supplemented the original 134 selection of policies with the Invasive Alien Species Regulation, which was adopted in 135 October 2014. 136 As we were interested in how the ES concept is used in EU policies, we did not apply one 137 particular standard definition or delineation. We reviewed the documents for the term 138 '(ecosystem) services' and whether particular services were mentioned. For analytical 139 and illustration purposes, however, we assigned specific ES mentioned in the EU policies 140 into the three ES categories of provisioning, regulating, and cultural ecosystem services. 141 The review focused on the main policy documents (Table 1), and did not review all 142 supporting documents (guidance manuals, impact assessments, plans and programs) 143 As we were also interested in how the policies were funded, we reviewed some of the 144 main funding instruments associated with the selected policies and the extent to which 145 they referred to the ES concept. The following funding mechanisms were reviewed: 146 Cohesion Fund; EU Agricultural Fund for Rural Development (EAFRD); EU Regional 147 Development Fund (ERDF): and LIFE/ LIFE+ Program. 148 incorporate the ES concept to 'produce' a meaningful and integrated policy at each of the 159 levels. We analysed internal (or vertical) coherence to understand the link between 160 goals, objectives, instruments, and the implementation processes within a particular 161 policy field as well as the coherence between the ES concept and the EU policy. External 162 (or horizontal) coherence was used to analyse the overlap or alignment of definitions, 163 objectives, instruments, or implementation processes across different policy fields. In 164 other words, the analysis sought to answer the question: what is the level of internal 165 coherence between the ES concept and the various dimensions of a particular EU policy. 166 To account for the gradual integration of the ES concept, we noted the date the policy 167 came into force and possible revisions made since then. 168 As EU policy is considered to be regulatory in nature ( The policies were assessed using a scoring system that was adapted from a 213 characterisation of impact assessments (based on (Helming et al., 2013); Table 2). 214

221
The analysis shows that, overall, the coherence between existing EU policies and the ES 222 concept is increasing gradually, but still confined to nature and natural resource policies. 223 Altogether, six policies had very high to high coherence and six others were assessed to 224 be lower. Six EU policies referred to ES explicitly, and only four reflected the ES concept 225 in the design of measures (Table 3). 226 for example the TEN-T or RED. Other communications, such as the Urban Strategy and 244 the Adaptation Strategy also focussing on environmental issues and the linkages between 245 ecosystems and humans, did not mention ES or the ES concept explicitly, though they 246 covered rather non-traditional environmental policy fields. 247 showed that regulating services were mentioned most frequently and also in greatest 262 detail, followed by provisioning ES and cultural ES (Figure 2, Table 4). Those policies, 263 however, which only mentioned ES indirectly tended to refer to just a small selection of 264 regulating ES, such as carbon sequestration or water quality (Table 4) Table 4).

285
Coherence at different levels of implementation 286 When analysing the mode of steering of the different policies, we found that all policies 287 that explicitly addressed ES and that had the ES concept fully embedded, featured an 288 advisory (even symbolic) mode of steering (Table 5) The ES concept was apparently integrated in a policy type for which the specific design of 299 measures and related ways of monitoring and -if applicable -sanctioning had been 300 delegated to MS or sub-national levels. A prominent example is the Biodiversity Strategy 301 where mapping, assessment, and valuation are to be undertaken by the MS, albeit with 302 support from the EU (Maes et al., 2016). This way, it is not necessary to organise a 303 formal consensus at the EU level about the selection of ES targeted or the methods for 304 mapping and assessment, and valuation of these services. This means that a detailed 305 operationalisation is not undertaken at the EU level. 306

309
The analysis showed that very few policies required MS to report on the stock/flow of a 310 particular ES. In general, mandatory reporting on environmental impacts appeared to be 311 rather specific to the policy area (e.g. WFD), rather than across different types of 312 impacts. Only the IAS (2014) requires that a surveillance system should monitor the 313 effectiveness of the policy on biodiversity, ES and if applicable human health and the 314 economy. This relative absence of pronounced and mandatory systems of measuring 315 environmental impacts could be due to the feasibility of establishing (environmental) 316 monitoring systems and the administrative burden of carrying out systematic monitoring. 317 In some cases, for example in the CAP/RDR, environmental impacts are not directly 318 measured, but assessments are made based on proxies that signal the change in a 319 (negative) driver, such as observed reduction in the use of pesticides. 320 For most policies, there were various EU funds available to finance measures. In some 321 cases, these EU funds were complemented by national and regional financing 322 mechanisms also incorporating private sector funding. There was, however, a small set of 323 EU funds that featured prominently across many of the analysed EU policies, including 324 the EU Cohesion Fund, the Life/Life+ Programme, the EU Regional Development Fund 325 (ERDF), and the EU Agricultural Fund for the Rural Development (EAFRD) (   Directives may be revised only once every 15-20 years, the average revision period of 385 the reviewed twelve policies is six to ten years (see Table 7 challenge to policy coherence, and will require adaptability, so that the complexity of a 414 policy does not undermine its feasibility and legitimacy (Ayres and Braithwaite, 1994;415 Fiorino, 1999;Rammel and van den Bergh, 2003). 416 Coherence and integration would also likely require political advocacy. Environmental 417 policies are often considered to oppose economic policies (Taylor et al., 2012). and where there are political windows of opportunity for new bridging concepts (Kingdon 442 and Thurber, 1984) and changing actors coalitions (Sabatier, 1998). categories, sometimes even exclusively on those specific ES, which are traditionally in 450 the domain of a particular policy. EU policy making follows a sectoral approach (Jordan 451 and Adelle, 2012), and the sectoral role-division maintains itself with path-dependent 452 processes and structures (Howlett and Rayner, 2007). This shows that there is a gap in 453 addressing the system-interdependence idea, which is at the heart of the ES concept 454 (MEA, 2005) and also underlies the justification for improving policy coherence (Howlett 455 and Rayner, 2007 If we assume that the ES concept will be gradually incorporated into EU policies, it is 487 important to look at the windows of opportunity for when this can be done. Table 7  488 shows when elements of the reviewed twelve EU policies might be revised. Overall, the 489 next opportunity for an overall revision of Communications (Strategies) and related 490 funding schemes is the year 2020. This means that the ES concept must be further 491 operationalised by then to match the sector's needs. 492 2018 (mid-term evaluation)

496
The article shows that the ES concept has not so far been coherently established in EU 497 policy making but that it is emerging in several of the strategy-setting policies. Only six 498 of the analysed twelve EU policies refer to ES explicitly, namely the Biodiversity Strategy, 499 GI Strategy, Forest Strategy, CAP/RDR, MSFD, and IAS. This shows that those policies 500 that address the natural environment and ecosystems are forerunners in integrating the 501 ES concept. Yet, even in these areas, the operationalisation is only partial. Only the 502 Biodiversity Strategy, GI Strategy, Forest Strategy, and IAS reflect the ES concept in the 503 design of measures. Some more recent policies (e.g., the Adaptation Strategy), however, 504 only take up the ES concept in supporting documents. The coherence of a particular 505 policy document with the ES concept might depend on the type of document, the 506 periodicity, and/or the stage of the implementation of a policy as well as the opposition 507 or support from involved actors. All four policies in which the ES concept is fully 508 embedded are either strategies featuring an advisory mode of steering, or, like the IAS, 509 leave details of operationalisation related to ecosystem services to the MS. This may be 510 due to the novelty of the ES approach, as well as because of the reluctance of MS to sign 511 up to more regulatory modes of steering or of involved actors to consider other interests. 512 Furthermore, very few policies require MS to report on the stock or flow of a particular 513

ES. 514
In conclusion, there is considerable scope to improve the mainstreaming of the ES 515 concept through, for example, common methods for monitoring and evaluation of ES, 516 developing dedicated financing mechanisms, and developing better tools to help policy 517 makers exploit cross-sectoral synergies and manage trade-offs between ES. 518 519 Acknowledgements 520 This research was undertaken in the framework of the EU project Operationalisation of 521 natural capital and ecosystem services. The research was funded by the EU FP7 grant 522 agreement No. 308428. Co-financing was provided by KBIV (KB-14) "Sustainable spatial 523 development of ecosystems, landscapes, seas and regions" which is funded by the Dutch 524 Ministry of Economic Affairs. 525 526