An ecological landscape can be modeled by a graph in which vertices represent the habitat areas and arcs represent the ecological corridors between habitat areas. Arcs lengths represent the difficulty for an individual to travel along the corresponding corridor. The landscape connectivity is then computed as a function of the all pair shortest paths of this graph. Among the corridors of the landscape, some may have their length reduced by paying a cost. We have proposed a MIP formulation for finding a subset of corridors to improve in order to maximize the landscape connectivity under a budget constraint. We will present a preprocessing algorithm that aims at reducing the size of the MIP formulation by identifying arcs of the graph that are (resp. are not) on a shortest path to a given vertex regardless of the improved corridors.