Heterogeneity in macroinvertebrate sampling strategy introduces variability in community characterization and stream trait-based biomonitoring: Influence of sampling effort and habitat selection criteria

摘要

Benthic macroinvertebrate assemblages have been of major importance in the biomonitoring of rivers for many decades. To more closely fit with new regulatory purposes, standardized methods of macroinvertebrate field sampling have changed over time both in terms of sampling effort and criteria of habitat type sampling priority (e.g. according to a decreasing gradient of "bottom surface coverage" versus "substratum favorability"). Such methodological heterogeneity has rendered the exploitation of biomonitoring data far more complex because they are indeed not fully comparable over time and may induce bias in long term series studies. In this context, we assessed the influence of variation in field sampling strategy on (i) the taxonomical characterization of benthic macroinvertebrate assemblages (297 taxon abundances, taxonomic richness and diversity, EPT-related metrics), (ii) the response of sensitivity descriptors (ASPT, SPEAR), (iii) the bio-ecological trait description of assemblages and (iv) the resulting ecological diagnostic of river reaches, based on 10,074 sampling events across France. Six sampling strategies were compared, corresponding to three levels of sampling effort (4, 8 and 12 individual samples) and different criteria for substratum selection; e.g. marginal (= 5% coverage at reach scale) vs. dominant; hosting capacity (= substratum favorability) vs. bottom coverage. A strong influence of the sampling strategy was observed on the characterization of the macroinvertebrate assemblages. Marginal substrata revealed a unique picture of the community at the reach scale and presented a higher taxonomic diversity and richness than dominant ones, as well as a higher number of sensitive EPT and SPEAR taxa. Both strategies based on dominant substratum sampling provided similar pictures of the taxonomical and bio-ecological characteristics of macroinvertebrate assemblages. The high hosting capacity of the most frequent marginal substrata has been identified as an important driver of the observed responses. When using a tool calculating the probability that sampled macroinvertebrate assemblages were impaired by several pressures related to water quality degradation, hydromorphological alteration or land-use, differences in sampling strategy affected the probability estimation. As a main trend, the results were especially reliable when they were obtained from an 8-sample strategy, even more if they included the information provided by the marginal substrata. If sampling effort definitely matters for environmental diagnostic, substratum selection criteria play also a crucial role on (i) the description of local assemblages and (ii) the reliability of calculated impairment risks, due to the strong relationship between stenotopic taxa and specific habitat conditions.