15 strategy and corporate social responsibility, C.P. 8888, succ. Centre-ville, Montréal, QC, Canada, 16 H3C 3P8. 17 18 2 Highlights 19 This paper is the first to develop global and empirically based characterization factors of the 20 impact of hydroelectricity production on aquatic ecosystems biodiversity, to be used in LCA; 21 The impact of hydroelectricity production on fish species richness was significant in the 22 tropics, of smaller amplitude in temperate and minimal in boreal biome; 23 The impact of hydroelectricity production on fish richness was consistent across scales -24 same directionality and statistical significance across sampling stations, reservoirs and 25 biomes; 26 The impact of hydroelectricity production on fish richness was sensitive to the duration of 27 the study, highlighting the need for a clear understanding of transient situations before 28 reaching steady states in LCA. 29 3 Abstract 30 Hydroelectricity is often presented as a clean, reliable, and renewable energy source, but is also 31 recognized for its potential impacts on aquatic ecosystem biodiversity. We used empirical data 32 on change in fish species richness following impoundment to develop Characterisation Factors 33 (CF) and Impact Scores (IS) for hydroelectricity production for use in Life Cycle Assessment 34 (LCA). We used data collected on 89 sampling stations (63 upstream and 26 downstream of a 35 dam) belonging to 27 reservoirs from three biomes (boreal, temperate and tropical). Overall, the 36 impact of hydroelectricity production on fish species richness was significant in the tropics, of 37 smaller amplitude in temperate and minimal in boreal biome, stressing for the need of 38 regionalisation. The impact of hydroelectricity production was also quite consistent across scales 39 (i.e., same directionality and statistical significance across sampling stations, reservoirs and 40 biomes) but was sensitive to the duration of the study (i.e., the period over which data have been 41 collected after impoundment), highlighting the need for a clear understanding of transient 42 situations before reaching steady states. Our CFs and ISs contribute to fill a gap to assist decision 43 makers using LCA to evaluate alternative technologies, such as hydropower, to decarbonize the 44 worldwide economy. 45 46