Assessing freshwater fish sensitivity to different sources of perturbation in a Mediterranean basin.

The accuracy of bioassessment programmes is highly limited by the precision of the systems used to derive sensitivity–tolerance values for the organisms used as indicators. We provide quantitative support to the objective evaluation of freshwater fish species sensitivity to different sources of disturbance, accounting for co-variation issues not only between perturbations and natural gradients (especially river size), but also between different perturbations. With this aim, we performed two different principal component analyses: (i) on a general environmental matrix to obtain a perturbation gradient independent of river size effects and (ii) on human impairment-related variables to extract independent synthetic perturbation gradients. Then, we checked each species responses to those gradients to assess their sensitivity–tolerance through an available-used chi-squared analysis in the first approach and through a t-test/ancova analysis in the second one. In this way, we obtained sensitivity–tolerance which could be included in future bioassessment tools, enabling effective evaluations. [ABSTRACT FROM AUTHOR]