LAUSR

Abstract Multimetric indices are considered a low-cost and rapid means of assessing ecological integrity in streams. The present study aimed to develop and validate multimetric indices that permit evaluation of reduced impact logging (RIL) effects on the ecological integrity of streams in the Eastern Amazon. We sampled 47 stream sites, 13 of them in reference areas and 11 in conventional logging (CL) areas, which were used to develop indices. The remaining 23 stream sites were sampled in RIL areas and used for validation of the developed indices. In each stream site we measured metrics relating to habitat structure and ecological features of fish assemblages, which were used to develop a habitat integrity index (HII), biotic integrity index (BII), and a general multimeric index (GMI). The values of the indices obtained for the RIL areas were compared and tested both in relation to the values observed in the reference and CL areas, and in relation to temporal gradients (i.e. time since logging) in RIL areas. All developed indices differed between RIL and CL, however only the HII differed between RIL and reference areas. We also detected a decrease in the GMI in recently logged RIL areas compared with streams in older RIL areas. The lack of distinction in GMI and BII indices between reference and RIL areas demonstrates the benefits of RIL practices for fish assemblages compared to conventional logging methods. However, the intermediate values obtained from these indices in RIL areas for all developed indices, the distinction of HII between reference and RIL, and the temporal variation of GMI in RIL areas can be considered as an important warning regards the possible impacts of RIL practices on ecological integrity in Amazonian streams. In summary, the multimetric indices were shown to be good indicators of ecological integrity in Amazonian streams subject to RIL. Even where the deleterious effects of logging are controlled by technical guidelines, indices presented a high sensitivity to anthropogenic disturbance, reinforcing their value as tools to evaluate and monitor ecological integrity of streams in human-modified tropical landscapes.