Recent theoretical and field studies conducted in both terrestrial and marine ecosystems have suggested that diversity could be a more complex and multi-component concept than previously thought. However, it is still poorly understood to what extent the information provided by the various indices is complementary with regard to diversity, and to what extent this complementarity is reproducible in different situations or at different scales of observation. In the present work, we have analysed the reproducibility of the empirical relationships between 11 diversity indices related to four major aspects (components) of species diversity (species richness, rarity, evenness and species taxonomy) in order to determine whether there is a general pattern of redundancy or complementarity in diversity measures that holds across years, areas and spatial scales. We have applied this approach to groundfish diversity through the analysis of 2404 trawl hauls collected during the first large-scale annual surveys carried out using a single sampling design throughout the northern Mediterranean Sea (ranging from 36.3 to 45.7 degrees N, and 5.3 degrees W to 28 degrees E). Whatever the years, areas and scales studied, we found a strong reproducibility in the number and the nature of the main complementary components of diversity. The whole set of diversity indices considered may be roughly split into 6 complementary groups of descriptors. This result highlights that, in contrast to what is still the usual practice, monitoring diversity should not be based on only one or two of the most widely used components, even at large scale. Finally, in a context where developing tools for monitoring diversity is considered as a high priority worldwide, the strong reproducibility of the relationships between diversity indices we observed offers a basis for discussion of the feasibility and relevance of proposing a general shortlist of indices for monitoring diversity at different management scales.