Biodiversity indicators are tools that are more and more used to formalize and concentrate the information we wish to share on biodiversity. We here wish to highlight some properties of indicators that arise from their place at the boundary between the academic and management worlds. We illustrate these properties by recent research specifically refering to Driver, Pressure and Response indicators within the DPSIR framework applied to forest biodiversity. First, especially from a management point of view, we think we should distinguish indicators according to the magnitude of their relationship with biodiversity - and not only according to the statistical significance of the relationship. Second, due to logical and ecological reasons, we need to take into account the fact that the relationship between the indicator and biodiversity might vary in space, time or according to ecological conditions. This in particular relates to the notion of domain of validity of the indicator. Third, we need to include the possibility that the relationship be non-linear rather than linear, with especially the possibility to estimate thresholds in the relationship. Fourth and finally, we need to better apprehend which parts of biodiversity are or are not related to the indicator. These different pieces of advice would help documenting the "identity card" of indicators in their relatonship to biodiversity, by better documenting the parts of biodiversity related to the indicator under different conditions, and those that are unrelated to the indicator. This should help enhance indicators as a boundary object between research and management.