Effects of sampling on benthic macroinvertebrate assemblages in assessment of biological quality of running water

In running waters, a picture of the macroinvertebrate community of a station is generally obtained by adding individuals sampled in different mesohabitats. Mesohabitats can be defined as combinations of a substrate type and a range of current speed. Simulations allowed us to study the variation in the community picture of a station when 8 mesohabitats were taken into account although eleven were sampled. Results were replaced in the framework of the use of the Indice Biologique Global Normalise (IBGN). Taxonomic structure described with indices varied greatly with the combination of mesohabitats sampled. Taxonomic richness was the more stable index considering the potential range of this index. In contrast, abundance was the more variable index. Between these extremes, diversity or dominance indices had an intermediate range of variability. The variability of a structure index seemed to increase when this index integrates taxa abundances, which depend on mesohabitats sampled. Taxonomic richness is integrated in IBGN in the form of classes; this expression reduces the effect of richness variability. Even if the sampling protocol was respected, the uncertainty on the IBGN can reach 3 units. Solutions have been considered to optimize the assessment of running water biological quality.