Recovery of the submersed vegetation is a target for the management of soft-water shallow lakes if they are to meet water quality and biodiversity standards. Knowledge of patterns of macrophyte space occupation and time to recovery is poor and mostly restricted to free floating species or riparian vegetation. Here we use pre- and post-impact monitoring data over 20 years showing the evolution of submersed aquatic vegetation of lake Baciver (Pyrenees), and develop models to infer space occupation and time to recovery. We use pre-impact macrophyte distribution in relation to bathymetry-derived data to fit logistic models to further simulate lake equilibrium scenarios. Depth and slope were found to be the best predictors, and models suggested that an assemblage dominated by Sparganium angustifolium was, at time of this study, over 95% of its potential distribution area. A dense, newly grown monospecific Isoetes lacustris population occupied <10% of its potential area and model projections suggest that it will take decades to recover. An I. lacustris residual population remains below the estimated depth threshold for survival and is bound to disappear. The lake appears to evolve towards a new steady-state where the current lake hypsography promotes the expansion of algae (Nitella sp.) over angiosperms.