The ratio of lentic to lotic habitat features strongly affects macroinvertebrate metrics used in southern Europe for ecological status classification

Biological quality in rivers based on benthic macroinvertebrates is typically assessed by comparison with ex-pected reference conditions, which represent relatively undisturbed situations. Commonly, reference conditions are set in agreement with river typologies to handle major ecological differences and limit biological variability. Although natural hydrological variation can be highly influential, site-specific tuning of reference conditions is rare in Mediterranean countries. River flow and local hydraulics change continuously over time, shaping the occurrence of lentic and lotic habitat features. Thus, biological reference conditions might require site-specific adjustment based on the ratio of lentic to lotic habitats assessed at the time of sampling. This would help reducing systematic bias in ecological assessments, interpreting benthic invertebrate responses to pressures, and diminishing the amount of unexplained biological variability. In this study, the response to the lentic-lotic character of river reaches was assessed for nineteen macroinvertebrate metrics and indices commonly used for the classification of ecological status in South European rivers. The study sites, with a prevalent temporary character, were located in Sardinia, southwestern Italy. Most metrics were significantly related to the lentic-lotic habitat conditions, both in pool and riffle mesohabitats, and their response curves were either parabolic or linearly decreasing at increasing lentic conditions. Taxonomic richness, score-based metrics, ovoviviparous taxa and multi-metric indices related well to the lentic-lotic conditions, while abundance metrics correlated less. The potential impact on ecological status classification was tested for the method formally used in Italy, which had a major role in comparing and inter-calibrating European assessment methods for the Water Framework Directive. After adjusting for bias due to the ratio of lentic to lotic habitat features, quality classification shifted towards better ecological status for approximate to 23% samples. This highlighted the impact of ignoring lentic-lotic information when defining reference conditions for assessing ecological status, varying from difficulties in understanding the biological response to pressures, to largely biased ecological status classification. The observed response of macroinvertebrate metrics to lentic-lotic conditions should be a key consideration for realistic ecological status assessment and could further be a valuable input for evaluating the effects of human-induced hydrological alteration and for assessing environmental flows.