Spatial prioritization of Natura 2000 coastal ecosystems: single-site or network approach?

One of the main challenges in conservation biology is to identify and preserve the most representative areas of natural landscapes, optimizing at the same time the costs associated with their conservation (Margules & Pressey 2000). Although the Natura 2000 network has been formally established across all European countries, more analysis is needed of its efficiency and effectiveness. To contribute to this, we aimed to identify the irreplaceable areas for the conservation of biodiversity features (habitats and species) within three coastal dune Natura 2000 sites, considering them individually and as a network.

The chosen sites are representative of the Adriatic seashore, hosting 13 EC habitat types and 63 species of fauna relevant for conservation (sensu 2009/147/EC and 92/43/EEC Directives). 

In a GIS environment, each study area was divided into a 50 x 50 m cell grid (PU, planning units). Three different values were assigned to each PU: 1) a monetary cost, based on the market values pertinent to the land cover types included (Ban & Klein 2009); 2) the shared boundary with adjacent PU; 3) the extents of every EC habitat type and of suitable habitat for every EC species existing in each PU. Specific conservation targets (i.e. the minimum percentage of EC habitat type and of suitable habitat for every EC species to be preserved) were fixed for each biodiversity feature. Targets for habitat types were set according to their conservation status in Europe (priority versus non-priority habitats sensu 92/43/EEC Directive), while targets for species conformed to national or global Red List assessments.

A multi-criteria Simulated Annealing algorithm (software MARXAN, Ball & Possingham 2000) was used to assess how many and which PU are necessary to achieve the conservation targets and to identify their best location in the landscape. Such analysis was performed individually for each site and globally for the entire Natura 2000 network, with a 100-time reiteration for each approach.

Results indicate that in the proposed context a network approach is more economically efficient than one that considers single sites individually (4106 selected PU vs 4871). Although a network approach could have lower costs for preserving and managing fewer PU of high conservation value, it showed a more spatially fragmented pattern (many isolated PU) than the single site approach, with a consequent low ecological effectiveness for the disconnected patches. As the economical efficiency of the network approach seems to be related to the presence of a very similar set of coastal dune habitats and species in the three selected sites, analogous results would be expected in ecological networks where the biodiversity features to be preserved are equally distributed in the different core zones.