The monitoring of biodiversity in space and time is the only way to show the impact of some pressures (e.g. in long-term ecosystem research sites LTER) and to assess the efficiency of conservation and restoration measures (e.g. in reserves, Natura 2000). Sampling methods must therefore be adapted to the studied taxa, be efficient towards the objective of the monitoring and be standardised to enable comparisons in space and time within the study and, hopefully, among studies. Moreover, sampling methods have to result from a compromise between level of exhaustiveness, comparability and feasibility. This goal of standardising, simplifying and harmonizing sampling methods is particularly relevant in the context of large scale monitoring in LTER sites, at the European level. Several national or European initiatives were devoted to these questions but they were restricted to a few biogeographical regions or taxonomic groups. Sampling methodologies need therefore to be assessed and optimised using some common methodology. Their efficiency and repeatability should be checked in a wider context -from the local and site scale to the global scale as ecological processes are likely to be expressed on different spatial and temporal scales. The purpose of this work would be to improve or even develop new sampling designs and to propose some integrated biodiversity monitoring for LTER sites. More precisely, the project will raise questions about (a) factors affecting elementary samples (bias, accuracy), (b) aggregation of samples in space and time and its links with the underlying mechanisms, (c) alternative allocation of effort between various spatial and temporal scales, always bearing the cost-effectiveness of the proposed methods in mind.