This thesis aimed to develop methodologies for identifying biodiversity conservation priorities in fragmented landscapes while considering meta-population dynamics and species’ habitat connectivity. The research encompassed several key aspects of biodiversity conservation in degraded landscape as mapping biotope and species corridors connecting them.
To begin, the study focused on delineating landscape elements using the concept of ecotopes, which are the smallest homogeneous units within a cartographic system. A novel approach utilizing spectral information and topography enhanced the ecological homogeneity of ecotope delineation. These ecotopes proved suitable for modeling biotope distribution, which is crucial for understanding landscape conservation priorities and guiding field inventories.
Subsequently, a biotope modeling framework was developed to accurately predict the presence of multiple biotopes within a region by incorporating vegetation dynamics. The concept of Potential Natural Vegetation (PNV) was utilized, assigning biotopes to PNV categories. This approach resulted in a robust presence/absence dataset for calibration and exhibited significant improvements in model sensitivity compared to traditional methods. The importance of PNV modeling in capturing the historical ecological envelope of biotopes was underscored.
Moving forward, the research explored the mapping of species corridors, using the case of the wildcat as an example. Various approaches were compared, including expert knowledge and modeling methods based on species observations. A data-driven approach utilizing presence-only data outperformed others in terms of efficiency, while all approaches identified the same critical corridors, emphasizing the importance of maintaining connectivity. Graph analysis revealed different central patches crucial for landscape connectivity, suggesting the data-driven approach when accurate data are available and the knowledge-driven approach when understanding of species habitat is well established.
The study then extended to modeling multiple species habitat networks to ensure landscape connectivity for species with different connectivity needs. A knowledge-driven approach was employed by considering fragmentation-sensitive focal species and their associated life history traits. Cluster analysis grouped species based on sensitivity to fragmentation, facilitating species choice to perform graph-based analyses for prioritizing connectivity stakes.
In conclusion, this thesis developed methodologies for identifying biodiversity conservation priorities in fragmented landscapes. The findings contribute to understanding and promoting landscape connectivity and emphasize the importance of incorporating such ecological considerations in landscape planning and management.