In the context of global change and current loss of biodiversity, there is an urgent need to improve spatial approaches for assessing and monitoring biodiversity in order to develop a conceptual framework based on biological indicators that can be used in conservation planning. The problem still remains in terms of decisions and knowledge on where to set appropriate conservation targets. Hence, we need detailed and reliable information about forest landscape structure and composition and methods for estimating this information over the whole spatial domain. This work is based on an approach involving conservation of a patrimonial species in a mountain landscape. The Regional Natural Park in the French sub-Alps, used for this study encounters multiple conflicts (i.e. ecological, economic, tourism pressures). The stakeholders are then very concern on "where to manage" in other words, they need tools to define the perimeter of good quality habitat to optimize management at an optimal scale. Landscape metrics are widely used to explore the spatial heterogeneity of landscapes that influence several species distribution. Few empirical researches have comparatively examined their indicator value for certain species occurrence at different scales. Within the framework of this study we test hypothesis concerning the best scale at which Hazel Grouse respond to the landscape structure. Our aim was to evaluate significant relations between different landscape level indicators and species occurrence and to assess the effects of the extent of the considered landscape on their performance. Results showed that spatial scale affected the performance of the metrics, since Hazel Grouse occurrence was usually better predicted at relatively smaller extents of surrounding landscape. In all, outputs of this work constitute support for forest management, as they provide a methodological framework and indicators for prioritizing areas in terms of habitat quality where species conservation planning can be optimized.
