The improvement of scientific knowledge to be used to define relevant retention practices of dead wood in favour of saproxylic diversity, is really at stake in the Western European Nemoral forest context. The RESINE project (2006-2009) aims at assessing, in a pluri-disciplinary framework: (i) the social representation of dead wood by different forest users, and (ii) the relationship between dead wood parameters (volume, quality, diversity) and saproxylic fungus and beetle diversity, in order to define sustainable forest practices and biodiversity indicators based on dead wood. The ecological study was lead in two forest regions: (i) the private pine forest in South-Western Landes, (ii) the state lowland oak forest of Rambouillet in Northern France, with a multi-scale approach from micro-habitat to stand and landscape. This talk will focus on preliminary results at piece and stand scales in Northern oakwoods. The in situ monitoring of oak dead wood pieces by sampling fungi and emerging beetles (using emergence traps) was meant to assess the relative influence of wood properties on species preference and assemblage composition. The effect of log diameter (from 2 to 70 cm), strata (stumps, canopy, standing and lying dead wood) and decay stage was measured. The beetle species richness was higher in snags and stumps than in lying dead wood, whereas the opposite trend was observed for fungi. The poorly known dead wood in living canopies showed the lowest species richness, but a distinct species composition. The highest number of beetle species was observed in the largest (>30cm) and mid-decayed pieces. The fungus assemblages were richer in the largest (all classes > 10cm) and late decayed classes. At the stand scale, the relative influence of dead wood parameters (volume, diversity, profile) on saproxylic local diversity was investigated. In 60 plots along a range of dead wood volume between 3 and 98 m3/ha, dead wood stocks were described, beetle assemblages were sampled using window-flight traps, and moss and fungus communities were assessed. In contrast with French pine and boreal stands, beetle and moss species richness did not seem to be correlated with the volume of dead wood at a 1ha-scale. Nonetheless, the number of fungus species increased with the volume of dead wood, and large dead wood volume in particular. In parallel with the development of this scientific background, the RESINE project included the planning of a network of dead wood adaptive management plots in French state lowland forests in collaboration with the National Forest Office. Contrasting forest management strategies were defined, including slash removal, target volumes of dead wood or number of dead or veteran trees per ha, % of overmature stands.