Application of the continuum suitability index (csi) model to display the permeability of the alpine-dinaric landscape and to define intervention priorities for ecological linkages

The Alps and the Dinaric Mountains are European biodiversity hubs and still show one of the lowest levels of landscape fragmentation in Europe, although some parts of the two mountain ranges are subjected to heavy human-related pressures. The aim of the present study was to analyse the current state of landscape connectivity and identify the most important geographical areas for preservation of unimpeded movement of species and the flow of natural processes from northern Italy to northern Greece, using the Continuum Suitability Index (CSI) in combination with a least-cost path (LCP) model. Results show 21.2% of the of the area displays high level of permeability with functioning ecological linkages (named Strategic Connectivity Areas, SACA1) of which 82% are already under environmental protection. However, results indicate a noticeable difference in the proportion of SACA1 area under environmental protection between EU and non-EU countries. The proportion of the national territory assigned to SACA1 varies considerably among countries from 7.2% in Bosnia and Herzegovina to 31.3% in Albania. The identified macro-regional LCP covers 105,669 km2 interlinking 52% of the SACA1 surface. Almost half of the SACA1 are transboundary, indicating high importance of synchronised cross-border conservation efforts and spatial planning. Results of this structural connectivity model are useful for policy maker on the macro-regional scale, however future efforts should focus on functional connectivity on the local and regional scales.


Introduction
Ecological connectivity (EC) represents the backbone of biodiversity conservation (Beier 1995;Bennett 1998;Taylor et al. 2006;Favilli et al. 2015) by ensuring the flow of organisms and ecological processes across landscapes (Taylor et al. 1993) and counteracting the negative effects of habitat fragmentation and climate change (Heller and Zavaleta 2009). The Alps and the Dinaric Mountains are exceptionally rich in biodiversity and, although they still show one of the lowest levels of landscape fragmentation in Europe and feature the largest Natura2000 sites as measured by area (EEA 2011;Lawrence et al. 2021), they are threatened by climate change and habitat loss (Laner et al. 2020). Several projects implemented in the Alpine area have clearly shown that the Alps are a sort of island embedded in a matrix of human presence and urban areas, underlying the need for the enhancement of the ecological network (i.e. Plassmann et al. 2019).
The main reason to enhance EC, especially between different mountain ranges, is that this should facilitate the ability of species to move into new regions as the climate changes (Krosby et al. 2010), thereby decreasing the probability of extinction. A network of connected areas has the potential to allow populations to persist following different spatial dynamics (e.g. classic metapopulations, source-sink dynamics, or patchy populations, as in Harrison 1991) even in the absence of vast intact natural areas. This is possible through long-distance animal movements such as juvenile dispersal, seasonal migrations, and geographical range shifts in response to habitat turnover (Van Teeffelen et al. 2012), environmental stochasticity, and climate change (Hannah 2008;Heller and Zavaleta 2009;Verboom et al. 2010), if the distance or landscape resistance to movement between habitat fragments is not excessive.
DINALPCONNECT (https://dinalpconnect.adrioninterreg.eu/) is the first EU project attempting to analyse the current state of EC in the Dinaric Mountains and the Alps-Dinaric connection, and to identify priority areas, barriers, and the most important linkages related to biodiversity conservation and EC (as already stated by Gazoulis et al. 2022). The aim of the present paper is to analyse and display the general permeability of the Alpine-Dinaric landscape, highlighting the Strategic Connectivity Areas (SACAs), and the LCPs connecting them.
The Geographic Information System (GIS) habitat suitability model applied in this study was adapted from the ALPBIONET2030 project (Plassmann et al. 2019), which did not concentrate on specific species, but analysed the overall Alpine connectivity and barriers using the structural approach of the Continuum Suitability Index (CSI). The CSI aims at detecting the general permeability of habitats and landscapes based on the physical features and arrangements of habitat patches, as well as the presence or absence of human pressure (Swiss National Park 2019). This structural connectivity model prioritises areas with a low degree of human disturbance, which are assumed to be permeable for terrestrial species (Hilty et al. 2020). LCP modelling is one of the methods used in landscape ecology to measure ecological connectivity by representing the landscape as an energy-cost surface. LCPs represent the route of maximum efficiency between two locations as a function of the distance travelled and the costs expended (Douglas 1994, Adriaensen et al., 2003, Affolter et al. 2011, Etherington and Holland 2013, Favilli et al. 2015. This methodology enabled the identification of the main SACAs, demonstrating that the CSI is suitable for a large-scale analysis, although it requires the adaptation to major disturbing and connecting elements (https://www.jecami.eu/; Swiss National Park 2019).
The present paper will also attempt to highlight this methodology in order to help local decision-makers identify the most valuable areas to be protected to enhance EC and develop the best strategies to promote local human-wildlife coexistence.

Study area
The total study area included in the analysis was 274,981 km 2 , corresponding to the DINALP-CONNECT project area, which extends along approximately 1,300 km from Greece in the south to Italy in the north. This comprises eight countries, of which the entire state territories of Albania, Bosnia and Herzegovina, Croatia, Montenegro, and Slovenia were included. The project area also includes the north-eastern part of Italy (the NUTS 2 regions Friuli-Venezia Giulia, Trentino-Alto Adige, and Veneto), the southern part of Austria (the NUTS 2 regions Carinthia and Styria, and the NUTS 3 Eastern Tyrol) and the central and western mainland of Greece (the NUTS 2 regions Western Macedonia, Epirus, Central Greece, and the NUTS 3 region Aetolia-Acarnania) (Fig. 1).

The continuum suitability index (CSI) modelling approach
The CSI is a spatial suitability analysis of landscape permeability, using a structural connectivity approach, focussing on three main ecological objectives in accordance with Ferretti and Pomarico (2013): the minimization of human pressure, the current biotic factors, and the physical environment. The CSI approach was developed and improved in several EU projects on Alpine ecological connectivity between 2008 and 2020 (see Affolter et al. 2011;Haller 2016;and Plassmann 2019).
The CSI is composed of five indicators that cover all of the above-mentioned issues: • Land use/land cover (LAN): defines landscape permeability based on land use, with landscape permeability values assigned to different land-use classes by experts. Data source: Corine Land Cover 2019, 100 × 100 m raster. • Population (POP): describes the pressure of the human population in and around settlements, with higher population reducing landscape permeability. Data source: GEOSTAT Population Grid 2011, updated with municipal data for 2018, 1 × 1 km grid. • Environment (ENV): highlights the level of environmental protection of a specific site, representing a positive contribution to ecological integrity. It considers protection at different levels, from regional and national to the Natura 2000 EU network. For areas with several levels of environmental protection, the highest level of protection was noted. Data sources: World Database on Protected Areas, national repositories, shapefiles. • Fragmentation (FRA): describes the rate of landscape fragmentation by human infrastructure, which reduces landscape permeability. Data sources: Roads and railways from the European Global Map 2019 and national repositories for non-EU countries. The effective mesh density was calculated for a 1 × 1 km grid. • Topography (TOP): combines the absolute altitude of the terrain and the slope as landscape permeability factors. The assumption is that general movement decreases with increasing altitude and slope. Data source: EU-DEM v.1.1 (2017), 25 × 25 m raster.
For each indicator, an indicator value from 0 (lowest permeability) to 10 (highest permeability) is assigned according to the landscape characteristic. The values for each indicator were defined using the AlpBioNet2030 values (Affolter et al. 2011;Swiss National Park 2019;Lüthi et al. 2018a, b, c, d, e), which were revised by DINALPCONNECT's experts in each project country (Laner and Favilli 2022a). These values were transposed onto a 100 × 100 m raster file of the project area and the overall CSI value for each cell was calculated as a weighted mean of all five indicators: CSI = (2*LAN + 2*POP + ENV + FRA + TOP)/7. For the CSI model details and calculations, please see: Affolter et al., 2011;Swiss National Park 2019;Laner and Favilli 2022a. The aim of the CSI model is to detect the most important areas for sustaining the macro-regional ecological network, named Strategic Connectivity Areas (SACAs) (Swiss National Park 2019; Laner and Favilli 2022a, b

Definition of ecological linkages and the identification of best paths
The identification of ecological linkages was conducted on two levels: firstly, to investigate the best connections of SACA1 areas to define the south-north ecological corridor from Greece to Italy. Secondly, to identify the most important regional linkages within SACA2 areas. The identification of the linkages was based on the LCP approach. It was conducted with the Linkage Mapper Toolset and its Linkage Pathways Tool (LinkageMapper.org). The LCP represents the route along which the least resistance -imposed by the landscape matrix on the general wildlife movements -occurs (Favilli et al. 2015). The LCP identifies the route of least cumulative resistance for a species moving between two SACA1 areas (McRae and Kavanagh 2012; Jones 2015; Beier 2018; Cayton 2019). To build a resistance raster for the corridor design, the values of the CSI were transposed to resistance values ranging from 0 to 100 by a data transformation ( Table 1). Linear resistances were selected for the CSI values 5-10, but higher resistance values for the CSI values 1-5 ( For the macro-regional linkage, several test runs were carried out using the southernmost and most important SACA1 areas in Greece and the northernmost ones in Italy and Austria. For the regional linkages, it would not have been possible to consider each single SACA1 area due to the limited time and computing capacity, as well as the excessive number of resulting ecological linkages. Therefore, a selection of the most important SACA1 areas was conducted based on an expert evaluation, considering size, the importance for potential connections to other countries and the south-north direction, and the general biological diversity. This approach allowed a reduction in the number of SACA1 areas from 2,208 to 159. The calculation of regional linkages was done with "bounding circles" of 60 km, since this limits the cost-weighted distance calculations from each source core to including only the portion of the landscape likely to be relevant to the connectivity between the sources and target cores, thus reducing processing time (McRae and Kavanagh 2012). The width of the corridors in the LCP approach is the deviation from the LCP and is determined by the resistance of the surrounding cells; therefore, the corridor is narrow if the resistance of the surrounding landscape is high. Instead of defining the maximum width of the corridors at any point of the corridor, we defined the maximum deviation of the corridor from the LCP along the entire length of the corridor network. The maximum deviation was truncated at 40,000 m for regional linkages and at 400,000 m for the macro-regional corridor. To build a resistance raster for the corridor design, the values of the continuum suitability index were transferred to resistance values ranging from 0 to 100 by a data transformation to obtain linear resistances (Res lin ): to build a resistance raster for the corridor design, the values of the continuum suitability index were transferred to resistance values ranging from 0 to 100 by a data transformation to get linear resistances: Res lin =(10 − CSI) ×10.
To remove linkages produced as artifacts along the border of the project area and the coast of the Adriatic Sea, all the regional linkages with a Euclidean distance between two core areas longer than 140 km were removed. In addition, 38 linkages were deleted by hand as they were long-distance connections between core areas which had alternative shortdistance connections via a series of intermediate core areas.

Implementation of the continuum suitability index (CSI) model in the DINALPCONNECT project area
The land cover indicator showed low values due to intensive agricultural land use mostly in lowlands (Fig. 5 -supplementary material). The project area is relatively sparsely populated except for the Po River valley in Italy and the surroundings of larger cities in all countries; therefore, the human population pressure in most of the area was low (Fig. 6 -supplementary material). The environmental protection indicator showed considerable differences among EU and non-EU countries, with a larger overall area under protection in EU countries, mostly due to the Natura 2000 network (Fig. 2). The fragmentation indicator showed higher connectivity in Albania, Bosnia and Herzegovina, and Montenegro, and at higher altitudes in the Alps, reflecting less developed traffic infrastructure (Fig. 7  -supplementary material5). The topography indicator showed that connectivity in the area is hindered by high altitude and steep slopes only in the Alps (Fig. 8 -supplementary material). Although the average indicator values do not differ greatly between countries, they give some indication of the main causes of low or high ecological permeability in different countries ( Table 2).

SACA areas in the DINALPCONNECT project area
The identification of SACA areas using the CSI approach in the DINALPCONNECT project area revealed that the structure of the landscape is as follows (Fig. 3): • 21.2% of the total project area is classified as SACA1 areas. They represent the main ecological conservation and core areas of the overall ecological network of the investigated area. 82.0% of these areas are part of an already protected site. • 75.4% of the area falls inside the SACA2 areas. This result shows that there exists a high potential for establishing measures to enhance the connectivity between Ecological Conservation Areas (SACA1). • The remaining 3.4% of the total area is classified as SACA3 areas, which represent the most visible barriers, such as major cities and agglomerations, as well as important linear transport infrastructure. Although it is a small percentage, the barrier effect of these areas can have a big impact on connectivity. SACA3 areas have the lowest fragmentation, land-use, and population index values.
A total of 2,208 SACA1 areas covering 58,235 km 2 with an average size of 26.4 km 2 were identified in our analysis. The largest cumulative SACA1 area was detected in Croatia and Greece (13,958 km 2 and 12,547 km 2 , respectively), which correlates to the large area of these countries included in our analysis ( Table 2). The SACA1 areas in these two countries are the largest in size, with an average area of 38.2 km 2 in Croatia and 37.2 km 2 in Greece ( Table 2). However, the proportion of the national territory in the SACA1 category among different countries varied considerably, with 31.3% of Albania's territory assigned to  Table 2). Bosnia and Herzegovina had the lowest percentage of territory designated as SACA1 areas (7.2%).
The results show that a highly permeable landscape can also be achieved without protection. There was a wide gap between EU and non-EU countries, with Bosnia and Herzegovina, Montenegro, and Albania having the lowest proportions of SACA1 areas within protected areas and the highest outside protected areas ( Table 2). This difference is clearly visible at the national borders. For example, on the border between Croatia and Bosnia and Herzegovina there are SACA1 areas on the Croatian side, with a sharp cut-off on the Bosnian side (Fig. 3). On the contrary, in EU countries the proportion of SACA1 areas under protection was high, between 85.4 and 99.5% ( Table 2). A high proportion of nationally designated protected areas which fall outside SACA1 can be found in Slovenia (52.9%), Croatia (47.0%), and Greece (36.1%). Those countries also have the most fragmented landscapes, with low average fragmentation indicator values ( Table 2).  Approximately half of all SACA1 areas had a transboundary characteristic (they extend on both sides of a national border), or they reached a national border. A comparison between national and transboundary SACA1 areas showed a similar total area in both categories, with a considerable difference in the average size, at 14.3 km 2 for national and 184.6 km 2 for transboundary SACA1 areas (Table 3).

Macro-regional connection
The CSI model test was applied for the identification of the LCP between northern and southern SACA1 areas and produced the same results from Slovenia to Albania regardless of the SACA1 area starting and ending points in Italy and Greece.
The total area covered by the south-north connection was 105,669 km 2 . It interlinked 108 of the total of 2,208 SACA1 areas; however, these 108 SACA1 covered 30,572 km 2 , which represents 52% of the total SACA1 surface. The 11 biggest SACA1 areas were larger than 1,000 km², and among them only one was not connected to the identified LCP of the southnorth connection. These results show that the LCP tends to select larger SACA1 areas and those with a higher environmental protection index value ( Fig. 9 -supplementary material).
The largest width of the macro-regional connection was approximately 100 km near the border of Montenegro and Bosnia and Herzegovina. The narrowest width of the south-north connection was 9 km and was in Albania between Lake Prespa and Lake Ohrid. Beside this bottleneck, two other bottlenecks are visible by a cartographic interpretation: in Slovenia between Triglav National Park and the more southern protected areas, and in the south of Croatia near the city of Sinj ( Fig. 9 -supplementary material).
The south-north connection covered the largest area in Greece (25,210 km 2 ), followed by Italy (16,650 km 2 ) and Albania (15,990 km 2 ). The smallest area of the south-north connection was in Slovenia and Austria (Table 4). The highest proportion of the identified southnorth connection is under environmental protection in Slovenia (79.4%), Croatia (63.7%), and Greece (43.5%). The relatively low number of protected areas in Bosnia and Herzegovina, Montenegro, and Albania was reflected again when considering only the south-north connection. In these countries protection was also lacking on the identified south-north connection (Table 4).
Considering some of the potential threats to ecological connectivity, we found that 8.1% of the south-north connection was used for intensive agriculture (Table 4), with the highest proportion in Greece (15.5%) and the lowest in Austria (2.4%) and Montenegro (2.2%). The most important agricultural barriers were the Adige Valley and the Non Valley in Italy, the surroundings of Postojna in Slovenia, the surroundings of the Croatian cities of Sinj and Ploče, and the surroundings of the city of Lezha in Albania (Fig. 4). Furthermore, we found 11 intersections between the south-north connection and motorways in total, of which five in Greece, four in Croatia, one in Slovenia, and one in Italy, with no apparent connectivity mitigation measures.

Indicator
Total

Discussion and conclusions
The importance of maintaining EC The aim of the DINALPCONNECT project is to identify actions on the ground in order to enhance the macro-regional ecological network, to respond to the political and economic circumstances that have created transboundary barriers between associated countries, which thus pose a major challenge for the conservation of EC and biodiversity (Gazoulis et al. 2022). In addition thereto, the area between Italy, Slovenia, and Austria is part of the European Green Belt, as is the area between Greece, Montenegro, and Albania -making the project area of great importance for the overall Trans-European Nature Network (TEN-N). According to that, non-EU countries should be better supported by EU neighbouring countries in order to enhance and maintain their national ecological network, by improving the presence and extension of protected areas, and the measures to connect them. A substantial part of the Balkan peninsula in the DINALPCONNECT project area has been affected by turbulent political conflicts in the recent past, which have weakened transboundary cooperation in biodiversity conservation. However, the importance of transboundary connectivity in this area has been advocated for the conservation of flagship species, such as the Balkan lynx (Schwaderer 2012). Furthermore, a study on the long-distance dispersal of grey wolf from the Dinaric-Balkan population in Slovenia into the Alps, and the consequent establishment of a mixed Alpine and Dinaric-Balkan pack in Italy, highlighted the importance of maintaining the transboundary connectivity between the Alps and the Dinaric Mountains (Ražen et al. 2016). Landscape corridors are important features for wildlife, and ensure the connectivity needed to counteract habitat fragmentation. Connectivity between SACA1 areas should be maintained and enhanced to ensure unimpeded movement of species and the flow of natu- ral processes. Therefore, it is essential to move to the pilot region level, after the macroreginal analysis and include the behavioural ecology and landscape structure of the target species when evaluating the connection type ("functional connectivity", as in Favilli et al. 2015). The reason for this is that the species characteristics are considered to be the most important factors in the evaluation of landscape resistance. Resistance maps should be carefully explored in order to create corridors as well as to implement biological conservation measures. The presented results have identified corridors between the core areas and major barriers within these corridors. Although the SACA1 areas were mapped based on structural landscape permeability, rather than on the needs of species, SACA1 areas are expected to meet the needs of most species in each region. Nevertheless, the actual adaptability of the SACA1 network to local and transnational wildlife movements is uncertain and could vary greatly among species and locations.

Fig. 4
Identified barriers on the macro-regional south-north connection

The impact of the DINALPCONNECT project
To enhance knowledge and identify major issues for ecological connectivity in the project area, the DINALPCONNECT project consortium proposed a strategy for coordinated and effective interventions at all levels of governance, structured in eight themes: Forestry, Agriculture, Wildlife Management, Water management, Tourism, Role of protected areas, Spatial planning and Infrastructure. The strategy encourages joint interventions based on an interdisciplinary, transboundary and horizontal approach. The project was able to overcome the political differences between the countries in the project area, and to re-establish collaboration for joining biodiversity conservation and local development. The fact that some of the countries involved in the project are not members of the European Union could pose barriers between partners and territories but was actually seen as support for the non-EU countries to better align with EU conservation laws and directives. The GIS analysis of EC highlighted the lack of adequate protected areas in non-EU countries (as in Table 2). The transboundary approach for enhancing EC, together with the other activities on the ground (i.e., SWOT), is a priority for biodiversity conservation and highlights the problems of weak governance, which result in habitat and ecosystems fragmentation due to artificial geopolitical country borders and linear infrastructures (as in Liu et al. 2020). In addition, the Balkan peninsula is one of the three major European areas which has acted as glacial refugia for almost all European species of Mediterranean origin, from where these species re-colonised Central and Northern Europe (Schmitt 2007). In view of the current and future climatic changes, connectivity in this area will be crucial for enabling range shifts of species to the north, following the spatial distribution of well-preserved natural areas with lower human pressures, as identified by the CSI macro-regional model based on a structural connectivity approach (Hilty et al. 2020). Structural connectivity is easier to measure and visualize compared to functional connectivity (Crooks and Sanjayan, 2006) and it was therefore useful in a large study area spanning eight countries with very different availability and accessibility of spatial data. The assessment of the human pressures on landscape naturalness and permeability was based on expert opinions, which were harmonised among countries, taking into account that land-use management can differ considerably within the same land-use category. Similarly, expert opinion-based harmonization was necessary in order to understand the levels of environmental protection in the different countries, which necessarily introduced a certain degree of subjectivity. Although the expert opinions introduced some bias, they were successfully used in biodiversity conservation studies where sufficient data was not available (Broome et al. 2019).

SACAs and protected areas
The conservation of EC is a major challenge for the whole of Europe (Perrin et al. 2022), which aims at developing and connecting a larger EU-wide network of protected areas on land and at sea, legally protecting a minimum of 30% of the EU's land area, and "integrating ecological corridors as part of a true Trans-European Nature Network" (European Commission 2021). Our results confirm the importance of transnational cooperation in biodiversity conservation, considering that approximately half of the SACA1 areas in the study area span or traverse national borders. There were considerable differences among the countries in the proportion of the national territories assigned to SACA1, ranging between 7.2% in Bosnia and Herzegovina and 31.3% in Albania. A higher proportion of SACA1 areas was included in established protected areas in the EU compared to non-EU countries. This partly reflects the high percentage of land under environmental protection in Slovenia, Croatia, and Greece compared to the global average (15.3% of land and freshwater environments, see Maxwell et al. 2020), but also the EU-27 average (26.4% of land; EC, 2021). In EU countries, where the level of protection is high, the width of the connection is influenced by protected areas. The non-EU countries have in general less land designated as protected areas mostly due to the lack of Natura 2000 sites, although the CSI demonstrated that environmental protection is not always decisive for the identification of SACA1 areas, as seen for Bosnia and Herzegovina, Montenegro, and Albania, where approximately half of the SACA1 areas fall outside of protected areas. These results highlight the opportunities of non-EU countries to protect the identified SACA1 areas, as part of the European TEN-N.With 21.2% of the surface assigned to SACA1 category, the DINALPCONNECT study area shows a higher degree of naturalness compared to the EUSALP macro-region (8%, as in Plassmann et al. 2019). This difference can be attributed to relatively lower population pressure, lower fragmentation by transport infrastructure, and lower terrain elevation, as well as to the presence of large areas of well-preserved natural ecosystems in the Dinaric-Alpine area. However, caution must be taken when considering average low population pressure on a large scale, because regional or local ecological linkages can be threatened by urbanization on a local scale. SACA1 areas represent the backbone of biodiversity conservation in the area and the main management goal should be to preserve and promote connectivity between them. Although the LCP method tended to favour the largest SACA1 areas with the highest environmental protection index value, the macro-regional south-north connection linked approximately half of the area of all SACA1. Along the south-north connection, several bottlenecks were identified, which are in need of special attention in order to designate the area as an ecological corridor and manage it over the long term to maintain or restore effective EC (Hilty et al. 2020). The narrowest point, between Lake Prespa and Lake Ohrid in Albania, should be interpreted with caution as it is very close to the edge of the study area (the border between Albania and North Macedonia), and the CSI model lacks information on the North Macedonia side. In eleven locations, the south-north connection was traversed by a motorway without any mitigation measures for EC. In addition to these threats to connectivity due to linear infrastructure, the model showed that approximately 8% of the southnorth connection is currently used for intensive agriculture. In the 2020, State of Nature in the EU report, agriculture was the most frequently reported pressure on habitats and species (EEA, 2020). Agricultural intensification between 1990 and 2007 was more pronounced in the Western and Central European countries (Levers et al., 2016), but now this trend is spreading to Eastern European countries.

Spatial analysis and involvement of stakeholders
For the above-mentioned reasons, it is therefore important to develop a methodology that allows both the identification and prioritization of the most important areas enhancing EC, and of the main barriers hindering it. The involvement of regional and transnational stakeholders is considered a fundamental step in the discussion on the results of the spatial analysis and the development of scenarios that take into consideration the needs of the local people and the effects of the current climate crisis (i.e. Sahraoui et al. 2021). Due to the large extent of the Dinaric Mountain range and to the national differences, it is impossible to give the derived results the same priority in all participating countries; the same topic/problem is tackled in different ways at different locations due to different reasons: the socio-economic environment, the national/local laws, the conformity of the landscape, the species present and their interactions with human society, and the personal relation of the people with their natural capital. In developing a regional connectivity analysis, it is important to involve end-users early in the design process in order to jointly determine what types of areas they want to connect, which areas need connectivity, and which areas merit the highest priority. The entire process should be transparent and replicable in order to build trust and allow the updating of new or better data as it becomes available. Keeley et al. (2018) highlighted that, for a successful connectivity implementation, it is fundamental to establish partnerships between stakeholders at local level supporting on-ground action. For macro-regional strategy development, the authors recommend a macro-regional connectivity analysis to identify SACA1 areas and LCPs. An assessment of the ecological value of linkages, coupled with an analysis of the threats and opportunities, is fundamental before moving to the regional scale, which will involve local decision-makers and stakeholders.

The CSI model for policymakers
The CSI model results provide information for policymakers and managers on the macroregional level and can be explored online at https://maps.eurac.edu/maps/1140/view. However, at the macro-regional level, a lot of valuable detailed information is not available or can be misinterpreted. Local data on the current and future presence of infrastructure are needed to detect potential barriers to wildlife dispersal and to define the best mitigation measures. In many cases, depending on the investigated region, this data can be obsolete and incomplete. Therefore, a GIS analysis of the identified linkages must be confirmed by field verification and validation. The CSI GIS model applied in the DINALPCONNECT project is a powerful tool that requires publicly accessible data to create reliable maps of the macro-regional EC. To perform a valuable analysis, input from local experts regarding the indicator values is needed. If the input data, factors, values, and weights used are consistent with the actual situation, the model can visualize structural connectivity of the studied areas, and detect the paths that may provide safe and alternative routes for wildlife in view of infrastructure development and climate change.

Future work
Future work should focus on assessing the functionality of the Alpine-Dinaric network for diverse wildlife species at the regional and local levels and refining the Ecological Connectivity Map and the recommendations based on the results.
Recent studies advocate an alternative framework wherein habitat quality, connectivity, or other conservation metrics are derived from attributes of both the biophysical and social landscapes (Carter et al. 2020). The incorporation of social factors into spatial analysis was initiated during the ALPBIONET2030 project (Plassmann et al. 2019), highlighting locations where EC-related measures and coexistence strategies were both biologically critical and socially feasible. Moreover, a broader incorporation of the social factors that inhibit or promote conservation may help identify a more diverse suite of targeted interventions to achieve desired conservation outcomes (Carter et al. 2020;Fortin et al., 2020).
The authors hope that this study will be regarded as a guide for researchers and planners to identify ecological network features, and for policymakers to adopt policies helping biological conservation and eco-regional landscape planning.