Freshwater ecosystems have experienced over the last three decades larger declines in biodiversity than terrestrial and marine ecosystems. In France, this degradation is represented by a decline in the quality and quantity of water, and by changes in the distribution and structure of aquatic biota for some rivers. These pressures on freshwater ecosystems are mainly human-induced and driven by land use changes. The objective of this paper is to evaluate the effects of different land use classes (agriculture, pasture, forest and urban) on a selected indicator of the ecological status of surface water, namely a Fish-Based Index (FBI) measured for various French rivers at the level of hydrographic sectors observed between 2001 and 2013. To this end, we couple a statistical spatial panel FBI model with an econometric land use share model. The FBI model explains the spatial and temporal distribution of the score of FBI by land use and pedo-climatic variables, while the land use share model describes how land use is affected by economic, physical and demographic factors. Our estimation results indicate that rivers in areas with more agricultural and urban land relative to forest area are associated with lower freshwater biodiversity. We use our estimations to simulate the impact of two fertilizer tax rates (+50% and +100%) on land use and their indirect impacts on FBI under two climate change (CC) scenarios (A2 and B1). Simulation results show that the fertilizer tax induces a decrease in agricultural land areas in favor of forest areas, which in turn improves the FBI. These results show also that a fertilizer tax helps cope partially with the negative impacts of CC on freshwater biodiversity.