Assessment and Spatial Planning for Peatland Conservation and Restoration: Europe’s Trans-Border Neman River Basin as a Case Study

: Peatlands are the “kidneys” of river basins. However, intensiﬁcation of agriculture and forestry in Europe has resulted in the degradation of peatlands and their biodiversity (i.e., species, habitats and processes in ecosystems), thus impairing water retention, nutrient ﬁltration, and carbon capture. Restoration of peatlands requires assessment of patterns and processes, and spatial planning. To support strategic planning of protection, management, and restoration of peatlands, we assessed the conservation status of three peatland types within the trans-border Neman River basin. First, we compiled a spatial peatland database for the two EU and two non-EU countries involved. Second, we performed quantitative and qualitative gap analyses of fens, transitional mires, and raised bogs at national and sub-basin levels. Third, we identiﬁed priority areas for local peatland restoration using a local hotspot analysis. Nationally, the gap analysis showed that the protection of peatlands meets the Convention of Biological Diversity’s quantitative target of 17%. However, qualitative targets like representation and peatland qualities were not met in some regional sub-basins. This stresses that restoration of peatlands, especially fens, is required. This study provides an assessment methodology to support sub-basin-level spatial conservation planning that considers both quantitative and qualitative peatland properties. Finally, we highlight the need for developing and validating evidence-based performance targets for peatland patterns and processes and call for peatland restoration guided by social-ecological research and inter-sectoral collaborative governance.


Study Area
Our case study area is the trans-border Neman River basin (97,928 km 2 ), which is the 14th largest river basin in Europe and the fourth largest in the Baltic Sea basin. It is located across the EU eastern border (56 • 15 -52 • 45 N and 22 • 40 -28 • 10 E), and is divided between 4 countries: the two EU members Lithuania and Poland, and the two non-EU countries Belarus and Russia (the Kaliningrad region) (Figure 1). The total length of the river is 937 km. Spring accounts for the highest seasonal river runoff (38%) and is followed by winter (26%), autumn (20%), and summer (16%) [44]. The mean slope of the riverbed varies from 0.16 m per km in the head waters to 0.23 m per km in the middle reaches, and 0.10 m per km in the downstream reaches (below the Neman-Neris River junction) [45]. The average water discharge of the river at Smalininkai is recorded at 535 m 3 /s [46]. The dominant land cover of the Neman River basin is agricultural land (57%) followed by forest (39%). The natural landscape development of the Neman River basin took place during two periods of the Pleistocene. The vast majority of Lithuania's territory was shaped during the last deglaciation of the Vistulian (Late Nemunas) ice sheet [47], resulting in numerous scattered depressions that have formed many small-sized peatlands. However, most of the peatlands of the Neman River basin in Belarus are of the Saalian age, with a monotonous morainic landscape. The dominant landforms are bottom moraines, fluvioglacial plains, and lowlands [48], which favours the formation of more extensive wetlands. The Lithuanian Neman River delta is home to the Aukštumala Land 2021, 10, 174 4 of 27 raised bog, the study area of the first comprehensive scientific study on the vegetation and development of raised bogs in the world and has thus made a unique contribution to mire science, including protection and restoration [49].
Land 2021, 10, x FOR PEER REVIEW 4 of 27 territory was shaped during the last deglaciation of the Vistulian (Late Nemunas) ice sheet [47], resulting in numerous scattered depressions that have formed many small-sized peatlands. However, most of the peatlands of the Neman River basin in Belarus are of the Saalian age, with a monotonous morainic landscape. The dominant landforms are bottom moraines, fluvioglacial plains, and lowlands [48], which favours the formation of more extensive wetlands. The Lithuanian Neman River delta is home to the Aukštumala raised bog, the study area of the first comprehensive scientific study on the vegetation and development of raised bogs in the world and has thus made a unique contribution to mire science, including protection and restoration [49].

Analytic Approach
An overview of the components of this study is presented in Figure 2. First, we compiled a spatial database of peatlands covering the Neman River basin in two EU (Lithuania and Poland) and two non-EU countries (Belarus and Russia) ( Figure 2, step 2). Second, to support strategic conservation planning, we made gap analyses to assess peatland quantity and quality relevant for (i) the highest levels of policy and governance (e.g., International/EU reporting of the national level situation), and (ii) regional sub-basins ( Figure 2, step 3a). Third, to support tactical planning for peatland restoration, we identified and quantified priority areas ( Figure 2, step 3b).

Analytic Approach
An overview of the components of this study is presented in Figure 2. First, we compiled a spatial database of peatlands covering the Neman River basin in two EU (Lithuania and Poland) and two non-EU countries (Belarus and Russia) (Figure 2, step 2). Second, to support strategic conservation planning, we made gap analyses to assess peatland quantity and quality relevant for (i) the highest levels of policy and governance (e.g., International/EU reporting of the national level situation), and (ii) regional sub-basins ( Figure 2, step 3a). Third, to support tactical planning for peatland restoration, we identified and quantified priority areas ( Figure 2, step 3b).
Land 2021, 10, x FOR PEER REVIEW 5 of 27 Figure 2. The analytic approach in four steps applied in this study. The strategic planning approach refers to the organizational process of defining high-level directions and making decisions on allocating resources to achieve an overall objective [50]. The cluster analysis aims at serving the subsequent tactical planning in local sub-basins.

Database Creation
Peatland areas are not always homogenous and can contain a large number of types Figure 2. The analytic approach in four steps applied in this study. The strategic planning approach refers to the organizational process of defining high-level directions and making decisions on allocating resources to achieve an overall objective [50]. The cluster analysis aims at serving the subsequent tactical planning in local sub-basins.  [39,51,52]. We focus on three peatland types, viz. (1) fens, i.e., minerotrophic peatlands, fed by mineral-rich groundwater or run-off water and atmospheric precipitation; (2) raised bogs, i.e., oligotrophic peatlands where the centre is higher than the edge, fed only by atmospheric precipitation and wind (aerosols), and are poor in minerals and nutrients; and (3) transitional mires, i.e., peatlands with features typical of both raised bogs and fens (as defined by Mitsch and Gosselink [53]). The detailed identification of peatland sub-types was not possible due to the diversity of methodological approaches used to acquire the peatland data in the different countries.
We created a spatial database of current peatlands for the Neman River basin ( Figure 2, step 2). This was created by compiling existing peatland data from Belarus (http://peatlands.by) and Lithuania [54]. However, as the Neman River basin peatland data for Poland was outdated, and not available for the Russian part of the case study area, we mapped their peatlands by combining remote sensing and field verification. Data compilation for peatland polygons included the three peatland types, area, protection status, drainage impact, and landcover (see www.neman-peatlands.eu). As the Russian and Belarussian data did not contain information on protection status and land cover types, we used supplementary data from protectedplanet.net and Broxton et al. [55]. The drainage status of peatlands was included within the Belarusian and Lithuanian peatland data, but for both Poland and Russia, they were captured using remote sensing and field verification. The data was compiled, harmonised, and analysed using GIS software. In addition, we corrected all topology errors. The minimum mapping unit of each peatland was 1 ha.

Amounts, Regional Distribution, and Characteristics of Peatland Types
To understand the spatial distribution of peatlands, we analysed their patch size distribution using the peatland database that we created. Consistent with percolation theory [56], the fragmentation and reduced size of peatland patches can have negative effects on water retention, nutrient filtration, carbon capture, and biodiversity. Peatland patch properties and species are closely linked. Using the umbrella species approach, namely that the presence of certain species can indicate that habitat patterns are satisfied for other less demanding species [57,58], the patch size requirements of wetland birds can be used to assess if benchmark conditions for habitat area and proportions in the local landscape of the catchment are satisfied or not. The black-tailed godwit (Limosa limosa), curlew (Numenius arquata), aquatic warbler (Acrocephalus paludicola), and black grouse (Tetrao tetrix), once common in peatlands of the Neman River basin, are relevant examples.
The approximate minimum area requirements to support local occurrence of these focal species ranges from 50-100 ha [37,59]. This patch size is consistent with the observation that peatland patches of >100 ha also support other species that require peatlands. To assess how the total area of peatland is distributed among different patch size intervals, we applied a geometric patch distribution of 0-50, 50-100, 100-200, 200-400, 400-800, 800-1600, and >1600 ha.
2.4. National and Regional Gap Analyses 2.4.1. Procedure Gap analysis [60] is a tool used to provide policy-makers with an assessment of the occurrence of potential gaps in the amount of different representative vegetation types for the maintenance of biodiversity and ecosystem services [61,62]. It can provide a quick summary at the policy level and for planners about the conservation status of different habitat types and ecosystems in terms of their extent, distribution, and representativeness. Using evidence-based conservation targets as a norm, gaps in terms of habitat types that are not sufficiently represented in the protected area networks can be identified. This forms the base for planning and actions to establish new conservation areas, changes in land Land 2021, 10, 174 6 of 27 management practices, and restoration of peatlands [60]. A gap analysis is an exercise based on a spatial comparison of a particular landcover (such as peatlands in this case) with existing protected areas that require detailed multiple data gathering, mapping, and analyses [63]. The gap analyses we applied to assess peatland protection included the following components: A = The area of peatlands. B1 = Current area of peatlands under protection (quantitative criterion). B2 = Current area of peatlands under protection not impacted by drainage (quantitative and qualitative criteria). C = Evidence-based or negotiated performance target. D = A × C-Long-term protection target. E = B1−(A × C)-Gap or surplus in protection. F = B2−(A × C)-Gap or surplus in protection not impacted by drainage.
Using this basic procedure, we performed multi-level quantitative and qualitative gap analyses that increased in complexity to assess the current protection status at different spatial scales. We thus focused on (i) the overall protection of peatlands within the Neman River basin at the national level (only quantitative), and (ii) regional level by peatland type and sub-basins (quantitative), and (iii) regional level by peatland type, and sub-basins excluding protected peatlands that are impacted by drainage (i.e., both quantitative and qualitative) (see Figure 2, step 3a).

Tipping Points for Patterns and Processes in Ecosystems
Performance targets for biodiversity conservation regarding the amount of land covers representing different ecosystems (i.e., "C" in the previous section) provide a good starting point to assess the status of current protected area networks able to sustain peatlands. As a negotiated performance target value reflecting evidence-based knowledge [28,29], as a proxy, we used the internationally agreed and ratified Convention on Biological Diversity's [21] Aichi Biodiversity Target #11, which states "By 2020, at least 17% of terrestrial and inland water, and 10% of coastal and marine areas, especially areas of particular importance for biodiversity and ecosystem services, are conserved through . . . protected areas and other effective area-based conservation measures." This Aichi target considers both pattern and process, and addresses both quantitative and qualitative criteria. Thus, we included the additional criteria of drained versus undrained peatlands. The logic for the analyses of peatland quality was determined by the fact that drainage affects the functionality of peatlands in terms of providing ecosystem services [64]. Thus, peatlands with drainage do not fulfil the requirements of CBD's Target 14 [21]: "By 2020, ecosystems that provide essential services, including services related to water, and contributed to health, livelihoods and well-being, are restored and safeguarded" and Target 15: "By 2020, ecosystem resilience and the contribution of biodiversity to carbon stocks have been enhanced, through conservation and restoration, including restoration of at least 15 percent of degraded ecosystems, thereby contributing to climate change mitigation and adaptation and to combating desertification". In addition, the Sustainable Development Goal 15 of the 2030 Agenda for Sustainable Development reiterates the importance of implementing the Strategic Plan for Biodiversity 2011-2020 and achieving the Aichi Biodiversity Targets. Moreover, the EU biodiversity strategy 2030 aims towards reducing the losses of nutrients from fertilisers by at least 50% [20]. Thus, the condition of peatland quality is even more important.

Cluster Analysis to Identify Peatland Hotspots and Coldspots
Although a gap analysis provides quantitative and qualitative results on the area amount required to satisfy a particular performance target, it does not provide precise spatial information on where priority areas for restoration are located (see Figure 2, step 3b). Therefore, strategic spatial planning of peatland protection, management, and restoration to identify peatland hotspots and coldspots is required [3,39]. Given that combined patches of different peatland types (e.g., fen, raised bog and transitional mire complexes) can be considered as a functional landscape element [32], we identified key peatland complexes by applying a 1 km 2 hexagon fishnet covering the entire Neman River basin. For each hexagon, we calculated the total peatland area proportions using ArcGIS. The rationale for selecting 1 km 2 hexagon units is that this is the approximate minimum home range area required to support local occurrence of wetland focal bird species that can indicate ecosystem health [37,59,65]. Indeed, the use of birds as a focal species has been shown to be a good indicator of wetland ecosystem health and functionality [66,67]. The aquatic warbler is a good example of a focal species that is dependent on fen management and restoration [40,68] with dominant open sedge fens or wet meadow habitats that are rich in invertebrates.
Subsequently, we used the Getis-Ord Gi * statistic cluster analysis tool [69] in ArcGIS to identify key peatland complexes. The cluster analysis evaluates the peatland area proportions for each hexagon and its neighbours. We applied a neighbourhood distance of 5 km to represent a local peatland landscape with a sufficient proportion of sufficiently large peatland patches [70]. The statistical variable Gi* is assigned to each of the hexagons and forms the z-score. For example, a high statistically significant positive z-score indicates more intense clustering of high-value peatlands and is thus a hotspot, whereas the opposite is a coldspot. Based on the cluster analysis outputs of the z-score, p-value, and reliability level (Gi_Bin), we created a hotspot map to identify key peatland complexes in the study area. The Gi_Bin field was defined at the statistical significance of hot spots ±2 bins, which represents a confidence level of 95%.

Priority Areas for Conservation and Restoration
Based on the results of the cluster analysis as well as the constraints of the available attributed data (e.g., protection status and drainage status), we prioritized the conservation and restoration potential of the peatland area within the Neman River basin for the identified hotspots (Table 1). We classified restoration as the re-wetting of drained peatlands through activities to remove and/or block the current drainage systems. We understand this is only a tactical step of the restoration process and that further operational restoration actions are required to be developed and formulated within ongoing management plans for peatlands [40,71], but this is beyond the scope of this paper. Table 1. Criteria used to determine priority actions for peatland re-wetting. We define this selection as a priority to reduce the peatland protection gaps in terms of both quality and quantity and thus their function to provide a greater range of ecosystem services within the hotspots of the cluster analysis.

Not Drained Drained
Protected

Secured Restoration needed Not protected
Conservation needed Conservation and Restoration

Distribution of Peatlands among Peatland Types and Sub-Basins
The Neman River basin peatland database consists of 1,006,802 ha of peatlands, with Belarus having the largest share (52%), followed closely by Lithuania (45%), while both the Polish and Russian parts of the Neman River basin contain only relatively small proportions of peatlands 2% and <1%, respectively ( Figure 3). Dividing the peatlands by type showed that fens made up 76%, transitional mire accounted for 12%, and raised bogs accounted for 12%. Overall, 44% of the Neman River basin's peatlands have been drained, with Poland recording the highest proportion of drainage 69% followed by Lithuania 66%, Russia 50%, and Belarus with only 23%, respectively. The allocation of peatlands by country and sub-river basin showed large variations in both Belarus and Lithuania. Given the small area sizes of the Polish and Russian parts of the Neman River basin, the sub-basins only contributed a small area amount. bogs accounted for 12%. Overall, 44% of the Neman River basin's peatlands h drained, with Poland recording the highest proportion of drainage 69% follo Lithuania 66%, Russia 50%, and Belarus with only 23%, respectively. The allo peatlands by country and sub-river basin showed large variations in both Bel Lithuania. Given the small area sizes of the Polish and Russian parts of the Nem basin, the sub-basins only contributed a small area amount. Spatial analyses of the Neman River basin showed the mean peatland p distribution varied by country and peatland type, with Belarus having the larg Spatial analyses of the Neman River basin showed the mean peatland patch size distribution varied by country and peatland type, with Belarus having the largest mean patches size for both fens (62 ha) and transitional mires (153 ha), and Russia for raised bogs (657 ha). The smallest peatland patch size for fens and transitional mires was recorded by Lithuania with 5 and 7 ha, respectively. Poland recorded the smallest mean raised bog patch size with 9 ha (Table 2, Figure 4). patches size for both fens (62 ha) and transitional mires (153 ha), and Russia for raised bogs (657 ha). The smallest peatland patch size for fens and transitional mires was recorded by Lithuania with 5 and 7 ha, respectively. Poland recorded the smallest mean raised bog patch size with 9 ha (Table 2, Figure 4).

Overall Protection of Peatlands within the Neman River Basin at the National Level
In total, 26% of the Neman River basin's peatlands are protected, with Poland having protected 90% of its peatlands. In Belarus, Lithuania, and Russia, the total area proportion of protected peatlands was much lower at 22%, 28%, and 26%, respectively. Thus, all four countries meet the CBD's Aichi Biodiversity Target No 11 of 17% in terms of overall area (quantity) protection of peatlands within the Neman River basin.

Regional Level by Peatland Type and Sub-Basin
At the national level, results showed surpluses in protection for all countries, with Poland leading the way at 70% followed by Belarus 21%, Lithuania 4%, and Russia 2%. However, at the sub-basin level, results showed that 9 out of the 23 sub-basins did not meet the CBD's 17% protection targets for fens ( Figure 5, Supplementary Material 1- Table S1). The Russian part of the Sesupe River sub-basin had the largest protection gap at 100%, whereas the Polish part of the Swislocz River sub-basin had the largest protection surplus, 72%.
Secondly, the analysis of transitional mires at the Neman River basin level showed an overall surplus in protection of 35% compared to the CBD's [21] nominated target of 17%. At the country level, Poland recorded the largest protection surpluses with 79% followed by Belarus 39% and Lithuania 25%. However, Russia had a 100% gap in transitional mire protection. The results at the sub-basin level show that only 3 sub-basins had protection gaps and did not meet the CBD's 17% target (Belarus-Merkys 100%, Russia-Sesupe 100%, and Lithuania-Jura 7% protection gaps) ( Figure 5, Supplementary Material 1- Table S2).
Thirdly, the analysis of raised bogs at the Neman River basin level showed an 18% surplus protection compared to the CBD's [21] nominated target of 17%. At a country level, Poland, Lithuania, and Russia recorded surpluses in protection, with an 83%, 46%, and 7%, respectively ( Figure 5, Supplementary Material 1- Table S3). However, Belarus showed a 1% gap in protection of raised bogs. The results at the sub-basin level show vast differences in protection gaps and surpluses, with 3 sub-basins recording protection gaps of 100% (Belarus-Czarna Hancza and Merkys sub-basins, and Russia-Sesupe sub-basin). Only 5 of the 23 sub-basins by country did not meet the CBD's 17% protection goals at this level of analysis ( Figure 5).

Impacts of Drainage
Our results show large variation at the finest level of the gap analysis for undrained protected peatlands. Firstly, the gap analysis for undrained protected fens showed a large decrease in protection, with an overall 3% gap in protection. At the sub-basin level, out of the 4 countries' 23 sub-basins, 17 did not meet CBD's 17% protection target ( Figure  6, Supplementary Material 1- Table S1). Lithuania had the biggest overall protection gap (8%), with 9 out of 10 sub-basins not meeting the proxy target of 17%.
Secondly, the gap analysis for undrained protected transitional mires showed an overall surplus in overall protection of 18%. Indeed, three of the countries still meet the international target applied in this assessment, with a surplus in protection of 37% for Belarus, 10% for Poland, and 5% for Lithuania. However, at the sub-basin level, 9 out of 23 sub-basins did not meet CBD's 17% protection goals ( Figure 6, Supplementary Material 1- Table S2).
Thirdly, the gap analysis for protected undrained raised bogs showed an overall 12% surplus in protection at the Neman River basin level showed compared to the CBD's (2010) nominated target of 17%. At this level of analysis, both Russia and Belarus recorded an overall protection gap of 1% and 14%, respectively. Out of the 4 countries' 23 sub-basins, 6 Figure 5. Regional distribution of peatland protection gaps and surpluses for each country by the Neman Rivers' sub-basins. In addition, the overall protection gap/surpluses are contained in the Neman River basin Total. Values equalling more than zero indicate the protection surplus in area proportion, values below zero indicate a gap in the protection of peatlands base on the Convention of Biological Diversity target #11 (17% protection).

Impacts of Drainage
Our results show large variation at the finest level of the gap analysis for undrained protected peatlands. Firstly, the gap analysis for undrained protected fens showed a large decrease in protection, with an overall 3% gap in protection. At the sub-basin level, out of the 4 countries' 23 sub-basins, 17 did not meet CBD's 17% protection target ( Figure 6, Supplementary Material 1- Table S1). Lithuania had the biggest overall protection gap (8%), with 9 out of 10 sub-basins not meeting the proxy target of 17%.  Table S3). Both Lithuania and Poland recorded surpluses in raised bog protection for all sub-basins.

Priority Areas for Restoration
The cluster analysis identified that 747,830 ha (74%) of peatlands were within hotspots, that 35,068 ha were within significant coldspots, and that 223,904 ha were identified as neither a hotspot nor a coldspot (Figure 7). Results show Belarus had the most peatland hotspots (456,049 ha), followed by Lithuania (274,053 ha), Poland (9510 ha), and Russia (8217 ha), respectively. Secondly, the gap analysis for undrained protected transitional mires showed an overall surplus in overall protection of 18%. Indeed, three of the countries still meet the international target applied in this assessment, with a surplus in protection of 37% for Belarus, 10% for Poland, and 5% for Lithuania. However, at the sub-basin level, 9 out of 23 sub-basins did not meet CBD's 17% protection goals ( Figure 6, Supplementary Material 1- Table S2).
Thirdly, the gap analysis for protected undrained raised bogs showed an overall 12% surplus in protection at the Neman River basin level showed compared to the CBD's (2010) nominated target of 17%. At this level of analysis, both Russia and Belarus recorded an overall protection gap of 1% and 14%, respectively. Out of the 4 countries' 23 sub-basins, 6 did not meet CBD's 17% protection goals ( Figure 6, Supplementary Material 1- Table S3). Both Lithuania and Poland recorded surpluses in raised bog protection for all sub-basins.

Priority Areas for Restoration
The cluster analysis identified that 747,830 ha (74%) of peatlands were within hotspots, that 35,068 ha were within significant coldspots, and that 223,904 ha were identified as neither a hotspot nor a coldspot (Figure 7). Results show Belarus had the most peatland hotspots (456,049 ha), followed by Lithuania (274,053 ha), Poland (9510 ha), and Russia (8217 ha), respectively. The priority areas analysis of peatlands within the hotspots of Neman River basin showed firstly that while 27% of all raised bogs have been secured by protection, only 12% of both fens and transitional mires, respectively, have been secured (i.e., protected and are not impacted by drainage) (Table 3). Secondly, the results show that in total, 20% of transitional mires and 5% of both fens and raised bogs are available for restoration (i.e., they are protected and impacted by drainage) (Table 3). Thirdly, we show that availability of peatland for conservation (i.e., peatland that is not protected and is not impacted by drainage) consists of 34%, 30%, and 15% for raised bogs, fens, and transitional mires, respectively. Finally, results show that 24%, 27%, and 29% of fens, raised bogs, and transitional mire, respectively, are within the category of needing both conservation and restoration (i.e., not protected and impacted by drainage) ( Table 3). The priority areas analysis of peatlands within the hotspots of Neman River basin showed firstly that while 27% of all raised bogs have been secured by protection, only 12% of both fens and transitional mires, respectively, have been secured (i.e., protected and are not impacted by drainage) (Table 3). Secondly, the results show that in total, 20% of transitional mires and 5% of both fens and raised bogs are available for restoration (i.e., they are protected and impacted by drainage) (Table 3). Thirdly, we show that availability of peatland for conservation (i.e., peatland that is not protected and is not impacted by drainage) consists of 34%, 30%, and 15% for raised bogs, fens, and transitional mires, respectively. Finally, results show that 24%, 27%, and 29% of fens, raised bogs, and transitional mire, respectively, are within the category of needing both conservation and restoration (i.e., not protected and impacted by drainage) ( Table 3). Table 3. Areas of secured peatlands (i.e., protected and are not impacted by drainage) and opportunities for peatland restoration (i.e., protected but impacted by drainage), peatland conservation (i.e., not protected and are not impacted by drainage), and peatlands that need both conservation and restoration (i.e., not protected and impacted by drainage) within the hotspots of the key peatland cluster analysis (Figure 7) of the Neman River basin for the three peatland types by country and sub-basin.

Gap Analysis Is an Assessment Tool Supporting Planning
Supporting the need for strategic and tactical spatial conservation planning, this case study of the trans-border Neman River basin demonstrates a methodology to assess the opportunities for conservation, management, and restoration of different peatland types at multiple scales. The trans-border context offered several challenges for the establishment of a harmonised spatial peatland database, which could be overcome by international collaboration made possible through EU InterReg funding. The peatland size distribution showed that the proportion of peatland patches exceeding 100 ha (i.e., the approximate minimum area requirement for peatland umbrella bird species) can be used as a criterion to estimate peatland functionality [37]. This minimum area size is assumed to also ensure key ecosystem processes, and that subsequent ecosystem service benefits are secured.
Nationally, our gap analysis showed that the protection of peatlands meets the international conservation target for the Neman River basin. However, assessment of different peatland types, and the exclusion of peatlands that are negatively impacted by drainage resulted in large gaps in the quantity and quality of peatlands for some sub-catchments. Thus, restoration of peatlands is required to improve their quality. The results also showed that raised bogs were better conserved than the other peatland types ( Figure 6). It is likely that this results from the larger size of raised bogs, and that they have been traditionally protected for years due to their inaccessibility and the larger costs to drain. This can be attributed to the perception that they are more ecologically valuable, being predominantly located on near-natural forest land. In contrast, fens and transitional mires, which are both relatively smaller in size and relatively uniformly distributed throughout the Neman River basin, have been subject to increased exploitation due to their greater economic value for agricultural use. Our results show that restoration is particularly needed for fens in agricultural landscapes, as they are both the most extensive (76 %, Figure 3) and also the most degraded peatland type, but the least protected type in the Neman River basin ( Figure 6). Additionally, using a cluster analysis, we could identify the most important peatlands and the area amounts available for conservation and restoration for each Neman River sub-basin. These priority peatland areas should be the focal points for conservation, management, and restoration ( Figure 7, Table 3).

Methodological Considerations That Underestimate Gaps
The peatland patch size distribution showed that both Belarus and Russia host larger peatland patches compared to Lithuania and Poland. This is due to both natural and anthropogenic factors. However, there are some caveats. For instance, in Lithuania, the peatland GIS layer was created using detailed spatial data [54], whereas in Poland and Russia, each peatland data set was created using remote sensing imagery validated in the field. However, for the peatlands of the Belarusian part of the Neman River basin, it can be argued that the peatland data was not of the same quality. In Lithuania, Poland, and Russia, we had peatland specialists on the ground, who could verify the data, whereas in Belarus, mapping such a large area is difficult, because data verification was lacking, and the high level of confidentiality of the spatial data. Thus, the results are less confident.
Moreover, considering the complex private landowner patterns in Lithuania and Poland, and subsequent difficulty for spatial planning of peatlands, would further decrease effective peatland patch sizes and increase fragmentation [72]. This applies in particular to fens. In Belarus, agricultural land is not subject to private ownership, thus management intensity is undertaken at an industrial scale [73]. In Russia, farming and agriculture is also dominated by industrial-scale operations [74], but there are also small landholders, which are in decline [75]. The collapse of the Soviet bloc in 1991 triggered large-scale land abandonment in Russia and Lithuania but not in Belarus and Poland [76]. Raised bogs, on the other hand, are usually embedded in forest landscapes, which are dominated by state ownership in Poland, Belarus, and Russia, and also in Lithuania, where only approximately 40% of forest land is privately owned [77].
Protected areas can be broadly divided into formal and voluntary [78], and management objectives and actions can vary enormously, from strict protection with no intervention to protected areas with management interventions. The four countries in the Neman River basin have different categories of protected areas, which are extremely complex and not harmonised. In the local context, the assignment of categories according to the IUCN World Commission on Protected Areas is thus extremely difficult [79,80]. In this light, we adopted a binary approach to peatland protection for this analysis. In summary, further data analysis on both land ownership and protection status is needed. We predict that such analysis would also show larger gaps in protection. Thus, we argue that the results in this paper are likely to be an underestimation of the protection gaps.
An important aspect of the gap analysis approach is the selection of the performance target that should be compared with indicators of ecosystem patterns and processes. In this study, we applied, as a negotiated and ratified performance target guided by evidencebased knowledge, the CDB's Aichi Target No. 11 of 17% as a proxy for sustainable ecosystems. However, evidence-based targets rather suggest that 30-40% is a critical threshold interval and natural tipping point for sustainable ecosystems [28,29]. Additionally, qualitative targets need to be met. The recent EU Biodiversity Strategy for 2030 [20] has thus set a re-negotiated target of 30% to be protected by 2030 for Lithuania and Poland. Belarus and Russia have both committed to the United Nations; Convention to Combat Desertification. Belarus has set targets of at least 60% of degraded land (natural meadows, forest land, woodlands and forest plantations, bogs and land of water bodies) to be stabilized and 60,000 ha of peatlands to be rehabilitated by 2030 [81] and Russia is still defining its targets [82]. Thus, applying either original evidence-based qualitative and quantitative targets, or revised negotiated targets, would reduce any surpluses and increase the gaps in protected peatland areas.
In summary, the methodology applied in this study is a promising avenue towards supporting assessments of ecosystem patterns and processes as foundations for strategic and tactical conservation planning. Concerning performance targets, a comprehensive research agenda is needed to define and validate evidenced-based knowledge on the tipping points and thresholds for variables supporting the conservation of peatland patterns and processes, which affect water quality, water retention, nutrient filtration, carbon storage, and the maintenance of biodiversity.
A gap analysis can vary from simple exercises based on a spatial comparison of a particular landcover with existing protected areas in terms of quantity to complex studies that can also include quality (such as the drainage of peatlands in this study). Moreover, gap analyses can be further developed to assess multiple landcover representing both potential natural vegetation types and cultural landscapes [37,62].

A Call for Adaptive Maintenance Actions of Fens
Our results show that restoration is particularly needed for fens in agricultural landscapes. This is partly determined by the fact that fens were drained and converted into agricultural land very intensively during the 20th century throughout the Neman River basin. Sustainable management of fens and the implementation of paludiculture approaches could stop further degradation and significantly improve the water quality. Therefore, focusing on fens as the most degraded peatland type, we discuss protection, management and restoration alternatives, and effective ways to mitigate the negative effects of intensive agriculture.
The evolution of peatland ecosystems is controlled by regional climate, landscape topography, water supply, nutrient, natural (e.g., fire, flooding) and anthropogenic (grazing, mowing, draining) disturbances, and autogenic processes associated with aging of individual peatlands [53,[83][84][85]. The aim of peatland restoration is to re-establish the desired vegetation and initiate self-regulatory mechanisms [86]. Hydrology is the critical element through the restoration of "pulse stability" [87], which maintains ecosystems' specific structures and functions. Although establishing the relevant hydrology can be achieved relatively quickly, restoration of fully functioning peat-accumulating mire ecosystems is a long-term process involving both active and passive measures.
Peatland vegetation may be re-established through natural succession [88] or be assisted by active measures using seeding diaspore (mosses and vascular plants) transferred from donor sites, planting potted young plants grown from seeds or from rhizomes, and other actions [71,[89][90][91][92]. This is less common in European restoration projects, where the regulation of hydrology and water chemistry dominates [86,93]. However, amelioration of internal eutrophication by sod cutting or topsoil removal, together with plant seeding, may be applied [71,89]. Passive restoration through natural re-vegetation is likely to succeed relatively quickly in fens [89]. However, the peat re-establishment action in mined peatbogs progresses very slowly and is even unlikely to succeed [94], thus requiring additional measures [90,95,96].
There are several trajectories of historic development of different peatlands types [84]. Fens are usually composed of large homogeneous patches with well-marked zonation from the waterlogged areas of riverbeds and oxbow lakes to the elevated edge-on parts of the valley that are not subject to inundation [38]. The typical vegetation of fens is relatively uniform and composed of ubiquitous and often expansive species. Habitats abundantly supplied with water are mainly occupied by Phragmites australis and Carex species, while Phalaris arundinacea, Calamagrostis canescens, Alopecurus pratensis, and Deschampsia caespitosa dominate the botanical composition of drier fen variants. According to Kołos and Banaszuk [97,98], the historic transformation of fens has resulted in five dominant vegetation types (Table 4). Most open wetlands ecosystems were developed in Eastern Europe to support animal husbandry through the removal of black alder (Alnus glutinosa) wet forests on floodplains. For centuries, they were transformed into wet meadows and pastures [99]. Regular mowing, grazing, and occasional fires maintained species-rich wet meadows and fen vegetation, and protected them from encroachment and overgrowing of shrubs and trees. Subsequently, intensive agriculture and widespread drainage of peatlands has led to a drastic decline in the area of species-rich wet grassland, meadows, and fens [37,100], and transformed the species composition [68,[101][102][103][104]. Finally, numerous peatlands were drained for peat mining. Peat extraction has affected 4.2% of the raised bogs in the Baltic States [105], and in Belarus, peat extraction covers 11.7% of the total peatland area [106]. Table 4. Five dominant types of vegetation currently found based on historically transformed fens according to Kołos and Banaszuk [97,98].

Vegetation Type Description Characteristic Plant Species
Permanent grasslands (hay meadows) Rarely flooded habitats, managed extensively every year and not well-fertilized with two variants: drier with low grasses and moist with low herbs and grasses To conclude, re-colonization of desirable peatland species to form the "ideal" historically natural biotopes and habitats of focal species is difficult, costly, and time demanding, and often not possible. There is concern that restoration will not be sustainable or successful under the unknown condition of future environmental conditions [107]. Restoration may also create a novel ecosystem, with no past analogue that are far away from an "ideally reconstructed" ecosystem by referring to its historical predecessors [86,108]. However, they may nevertheless provide ecosystem services comparable to natural mires. In addition, the benchmark for landscape restoration depends on the timeframe used as a reference point [37]. Thus, understanding the history of peatland development, past trajectories, and current trends and states is of key importance.
Changes in vegetation are difficult to predict due to synergistic interactions and the stochastic nature of these processes [109]. Fens are the most important mire type supporting the "kidney" and biodiversity conservation functions in sub-catchments. However, open fens require ongoing management to maintain their ecosystem functions [68]. This includes grazing of fens throughout late spring, summer, and autumn with low densities of large herbivores [110] or traditional mowing of fens to hinder the regeneration of trees (mostly Alnus glutinosa and Salix spp.) and the encroachment of shrubs [98]. Mowing should be performed at the beginning of August when plant species have finished flowering, so that seeds have had a chance to germinate in exposed areas, and wetland birds have finished nesting.
In addition to direct peatland restoration efforts through re-wetting, the establishment of wetland buffer zones surrounding peatlands can significantly improve water quality by filtering agricultural pollutants (mainly N and P) from the outflowing water by 43% for N (at a load of >500 kg N/ha/yr) and 21% for P (at a load of 20 kg P/ha/yr) [111]. Hence, a landscape perspective is needed, both in terms of spatial extent, and by considering landscapes as social-ecological systems [112].

Framing Peatland Restoration
To advocate restoration of peatlands, there are several relevant concepts that aim towards both balancing and maintaining landscapes' goods, services, and values, to mitigate global change and thus securing human well-being [113]. Firstly, the ecosystem services concept was presented in the 1980s, within the context of biodiversity conservation [114], and refers to "the direct and indirect contributions of ecosystems to human well-being" [14]. Ecosystem services emphasize societies' dependence on nature. However, this concept has been criticized, as it fails to include the complexity of both natural systems [115] and social-ecological systems [116,117].
Secondly, to support the vision of sustainable social-ecological systems, the landscape service concept was proposed to endorse participatory landscape planning [118]. The use of this concept is attractive to stakeholders from social and business disciplines [119] and can help facilitate inter-and trans-disciplinary research involving both researchers and practitioners [120]. The differences between ecosystem services and landscape services have arisen from the difference between an ecosystem viewed as a natural science phenomenon, and landscape as one integrating biophysical, anthropogenic, and perceived dimensions of social-ecological systems [121]. Moreover, landscape services have been deemed to address the spatial heterogeneity of landscapes more adequately [122].
Finally, "Nature's Contributions to People", which is used in the assessment by IPBES [123], acknowledges the central role that culture plays in defining all links between people and nature, and focuses on the role of indigenous and local knowledge [124]. While some believe that there is no fundamental difference between Nature's Contributions to People and ecosystem services [6], others claim that the ecosystem services concept already covers social sciences and other topics [113].
Irrespective of the framework chosen for analyses and valuation of peatlands, as a foundation for comprehensive spatial planning, portfolios of value items need to be identified, and the extent to which they are rival needs to be assessed. Gap analysis is such a tool. Adding analyses of spatial relations between particular complexes of peatlands in the Neman River basin, on top of indicating the data-supported needs for conservation and restoration, is an important foundation for the planning of peatland protection, management, and restoration. This applies both to ecosystem functions and conservation of habitat patterns for focal species.
Knowing the physical features of peatlands as well as their status may help in prioritising restoration oriented at systematic provision of ecosystem services. Referring to possible gains from increased water retention in rewetted mires to artificial retention in the catchment may indicate the relevance of restoration for mitigation low flows, acting simultaneously as a nature-based solution for reducing flood risk throughout the catchment [125]. Interrelation between the sites to be re-wetted and preserved may optimize a large-scale facilitation of nutrient retention in wetland buffer zones [111]. One should also consider that the costs of restoring wetland buffer zones are expected to be lower than the values of gains expressed as ecosystem services provided by the restored sites [84]. For instance, Valasiuk et al. [68] showed that citizens in Belarus are willing to pay a substantial amount of money for peatland habitat conservation, restoration, and maintenance for wetland birds, such as the aquatic warbler. This would support other key peatland functions, such as water retention, nutrient filtration, carbon capture, and support wetland biodiversity. Thus, restoration and integrated management of peatlands to combat land-degradation, through the provisions of water retention, nutrient filtration, carbon capture, and biodiversity maintenance, can have multiple positive societal impacts [126,127]. Concerted action for the protection and wise use of peatlands should therefore be a global priority linking planning and restoration activities at global, regional, and local levels.

Including Peatlands in River Basin Management Plans and Agricultural Strategic Plans
Our results emphasise the need to include peatland conservation in River Basin Management Plans and Agricultural Strategic Plans. For example, peatland re-wetting combined with paludiculture can provide win-win-options for various aspects of society, including social (additional employment in rural areas), economy (alternative incomes in agriculture), and environment (ecosystem services, substitution of fossil resources). Peatland conservation and restoration cuts across most United Nations Sustainable Development Goals and should be an instrumental part of the European Green Deal [16].
Regarding water policies, such as the EU Water Framework Directive, peatlands are still not adequately considered in the Neman River basin management plans (RMMPs), neither in terms of water retention, nutrient filtration and carbon capture [128], nor biodiversity conservation. This is in spite of positive affects at the entire sub-basin level. Therefore, the European Commission recommends the integration of wetlands including peatlands into the RBMP of the Water Framework Directive in its guidance for implementation [129]. This guidance should be adequately followed in drafting the update for the RBMPs of the Neman River catchment in the EU Member States Poland and Lithuania.
Besides water policies, agricultural policies, such as the EU Common Agricultural Policy (CAP), are the main drivers for management of drained organic soils including extensive drainage activities. Peatlands require a specific management approach due to their unique soil conditions. To maintain the carbon and nutrient stocks and reduce the release of large emissions, the raising of water levels up to or close to the soil surface is required. As a guiding principle, no landowner or user in the EU should be economically or socially disadvantaged by maintaining wetlands or developing re-wetted peatland management. This should be addressed by coherent standards for agricultural practices on peatlands and focused agri-environmental and climate schemes (AECSs) incentivising climate-smart water management, paludiculture, and implementation of wetland buffer zones. In the new CAP, which is currently under negotiation and will likely start in 2023, standards will set as conditionality with specific 'Good Agricultural and Environmental Conditions' (GAECs) [130]. For peatland management and water quality, two proposed GEACs are of special importance: GAEC 2-Preservation of carbon rich soils such as peatlands and wetlands and GAEC 4-Establishment of buffer strips along watercourses [130]. The detailed definition of the conditionality standards will be part of the National CAP strategic plans, which needs to be ambitious to fulfil other policy target-namely climate change mitigation and water quality.
Agri-environmental and climate schemes are programmed within the second pillar of the CAP, but the direct payments are contained within the first pillar. So far, the payments mostly serve biodiversity conservation purposes in the EU Member States. However, payments for the re-wetting and raising of water levels, which are instrumental towards mitigating climate change, are not included [131,132]. Thus, the payment schemes should be changed to support fit-for-purpose interventions described in the CAP strategic plans. For an overview of the different policy options for peatlands in the CAP, see Tanneberger et al. [16]. However, beneficiaries within the EU are individual farmers that operate as business enterprises. This complicates co-ordination among neighbouring landowners and often results in short-term commitments to managing individual landcover patches with many landowners. This could be solved by measures like AECS designed for environmental cooperatives of farmers [131] or with special programs for consolidation of land parcels. The complexity of land ownership in both Poland and Lithuania requires further analysis. Additionally, more harmonized information about possibilities of climatesmart management of wet organic soils in the Neman River basin including both EU and non-EU countries is needed.

Learning from Top-Down vs. Bottom-Up Legacies
Cross-border governance of peatlands is complicated by biophysical, historical, cultural, economic, and natural dimensions and the social-ecological system. All four countries that contain the Neman River basin were either part of the Soviet Union or part of the Soviet eastern bloc states with social systems characterised by state-centric top-down management control. Thus, there are several kinds of transitions affecting the approaches to planning and governance.
During the Soviet period (1922 to 1991), centralised planning ensured that peat extraction was concentrated in regions with significant peat resources. The gaps in peatland conservation between fens and raised bogs can be explained by traditional nature conservation where most protected areas were designated on forest land. In contrast, agricultural lands were managed for production and economic output and were not considered for nature conservation. Thus, many of these constraints were related to institutional, sociocultural, biophysical, and economic legacies of the Soviet and post-Soviet periods. The collapse of the Soviet bloc changed this; currently, peat extraction locations are determined by market-economic demand and agricultural lands are becoming important areas for nature conservation. However, EU bureaucracy and complex stakeholder portfolios also offer numerous governance challenges for Lithuania and Poland, which calls for actions at multiple scales [133].
The European Union's eastern border can be viewed as a fault line regarding the past level of modification of ecosystems with better conservation status in the East than the West [43,134]. Across Europe, peatland exploitation, protection and restoration have started to develop during different time periods, and at different basic levels of past transformation and rates of change [37]. While in the East, a significant proportion of natural mires have been retained, most other countries in the West have suffered severe losses [17]. The Central European trans-border regions, and regions where topography or other features hamper economic development, therefore often host valuable natural and cultural heritage [43,135]. This has led to improved retention of biodiversity, including species, habitat networks, and natural processes, compared to Western Europe [136], and cultural values [137]. However, regions located along the eastern border of the EU currently stand at a crossroad between increased production for economic benefits and the need for nature conservation [138,139]. Although Belarus still has a strong state-centric management control, they have been able to develop flexible nature conservation legislation, which has translated into success stories for peatland protection, management, and restoration [40]. This includes broader public awareness on nature values and ecosystem services in Belarus [68].
Therefore, central and eastern Europe's trans-border regions and landscapes are of particular importance for knowledge production and learning towards sustaining a wide variety of different ecosystems, ranging from those remaining with high levels of naturalness (i.e., raised bogs) to those built on traditional low-intensity farming including animal husbandry (i.e., fens). While the former requires protected area networks that allow natural disturbances, the latter requires maintenance of traditional multifunctional agricultural systems. This means that both historic permanent loss of peatlands as potential natural vegetation, and current transition trajectories in both ecological and social systems need to be understood [140,141]. However, trans-boundary collaboration both in terms of planning and management practices is not coherent because legislation and spatial planning are not effectively linked among countries [138].

Conclusions
This case study and the resulting discussion on the maintenance of peatlands through conservation, management, and restoration within the trans-border Neman River basin shows that the setting and interpretation of quantitative evidence-based performance targets need to be complemented with qualitative targets that mirror both ecosystem patterns and processes. At a national scale, all four countries meet the quantitative area protection targets for peatlands within the Neman River basin. However, factoring in additional qualitative aspects, including peatland type and history of drainage, shows that there are large protection gaps in some sub-basins. Fens were the dominant peatland type but also the most degraded and least protected. Our systematic regional gap analyses show that peatland restoration with sustained actions for maintenance is required, and the cluster analysis identified priority peatland hotspots for these actions. Thus, this study emphasises the need to include peatland conservation, management, and restoration into river basin management plans and agricultural strategic plans, and that planning should be adapted to meet the needs of different social-ecological systems. Comparative studies of trans-border regions can encourage knowledge production and learning about the past and current states and trends of both natural and anthropogenic peatlands. The governance and management of different green infrastructures, like peatlands, for human well-being is a concrete and suitable topic for place-based development cooperation among EU and non-EU countries. This requires research that integrates policy makers, planners, and stakeholders, as well as disciplines that mirror social-ecological systems, including landscape ecology, conservation biology, sustainability science, environmental policy, governance, assessment, and planning.  Data Availability Statement: Data used in this study can be found and downloaded at www.nemanpeatlands.eu.