Hydrological alteration is one of the major threats to the integrity and functioning of riverine ecosystems. We propose to analyze such alterations by means of three approaches: the well‐established range of variability approach (RVA), wavelet transform analysis (WT), and a combination of the two. This allows us to separate the most relevant scales of variability and to detect their alteration independently; such information cannot be provided by purely statistical methods like RVA. We show the advantages of this multiple approach through the analysis of two streamflow data sets of the Adige River at Trento, northeastern Italy: a daily streamflow record spanning from 1923 to 2006, and a record of instantaneous measurements with time step of 15 min from 2002 to 2006. Results suggest that starting from the 1960s, after the construction of the reservoirs currently in operation, a progressive flattening occurred in the hydrograph, which attenuated the amplitude of seasonal variations. This negative tendency worsened in the last few years, possibly as an effect of a reduction in winter snowfall. Wavelet analysis shows that hydroelectric production caused the high‐frequency components of the signal to increase in amplitude and number, with fairly regular weekly and daily oscillations and sharp transitions. The whole spectrum of low flows is significantly altered. This alteration of the natural flow has adverse consequences on the ecological integrity of the ecosystem. Use of the range of variability approach, the wavelet transform analysis, and a new methodology which integrates both allows a better separation between natural and human‐induced effects at ecologically relevant scales of variability, thereby identifying changes and trends due to hydropower management.