Unraveling the performance of the benthic index AMBI in a subtropical bay: The effects of data transformations and exclusion of low-reliability sites.

We investigated the relationship between the AMBI index and different contamination proxies in an urbanized bay in South America (SE Brazil), and the effect of (a) abundance data transformation and (b) exclusion of low-reliability sites (high SD; low N) on the index' performance. Poor ecological quality and opportunistic species were related to an increase in contaminants concentrations and mud content. Good ecological status and sensitive species (EG I) were mainly related to increased hydrodynamics. Data transformation caused minimal changes to the overall relationships, but exclusion of low-reliability sites improved the relationship between ecological groups and contamination proxies. Our results show that AMBI is robust in detecting effects of different contaminants in the area and reinforce the importance of the index as a tool for coastal management, but local joint efforts are needed to improve and adjust local species classification in ecological groups to improve the index' performance.