Planning for reserve adequacy in dynamic landscapes; maximizing future representation of vegetation communities under flood disturbance in the Pantanal wetland

Aim Using a probabilistic modelling framework, we aimed to incorporate landscape spatiotemporal dynamics into reserve design. We employed a spatially explicit stochastic model, which integrates both hydrological and biological processes, to simulate the wetland's biological succession. Location Pantanal wetland (with 140,000 km²) between Brazil, Bolivia and Paraguay. Methods We used the reserve design software Marxan to optimize the current and future representation (up to 50 years) of 20% of five plant communities with maximum reliability (i.e. smallest uncertainty). The Kappa statistic was used to compare selection frequencies of individual sites through a set of planning timeframes (5, 17, 25 and 50 years) and the likely pattern of biological succession over these periods. Results Solutions based on static vegetation distributions were significantly dissimilar from solutions based on the expected modelled changes resulting from the flood disturbance and succession dynamics. Increasing the required reliability of biodiversity outcomes resulted in more expensive reserve solutions. We demonstrated the flexibility of probabilistic decision-making methods to illuminate the trade-offs between reliability and efficiency of site selection. Main conclusions Considering the importance of habitat heterogeneity to the principles and practice of systematic conservation planning, it is notable that landscape dynamics have not been a central theme in conservation planning. In the case of the Pantanal hydrosere, acknowledging and planning for temporal dynamics required an ability to model succession and define acceptable levels of outcome reliability, but ultimately improved the long-term Adequacy of resulting reserve networks.