Since a few decades, underwater substrata can be characterized by methods using acoustic technology. Compared to traditional in situ grabbing techniques (bottom grab sampler, video camera, . . .), these methods allow a better spatial resolution. Until today, these acoustical techniques required specialized signal processing materials. This document presents works about lake-bed classification using single beam echo sounder and a post processing software (Sonar5) dedicated to the estimation of fish biomass. Indeed, an underwater substrate analysing module based on the theory developed by Orlowski (Oceanologia 19 (1984), pp. 6178) has just been implemented in the software. An optimum configuration have been selected and permitted to classify four different natures of substrata: silt, sand, sand & stones, and rock. A substrata nature map of an area known as an arctic charr (emblematic salmonid of the alpine big lakes) reproduction site has been done and shows the system's performances.