Abstract

Aim: Flooded forests are critically important for supplying water-related ecosystem services, but their restoration has been severely limited by low seedling performance due to flooding stress. We classified the flooding tolerance of 15 native tree species in a nursery-based method to guide species selection for restoring a flooded forest. Location: Atlantic Forest, São Paulo State, Brazil. Methods: We first exposed 15 native tree species typical of flooded forests to four flooding conditions (i.e., normoxia and three increasing hypoxia levels): no flooding, up to the middle of the root system, up to the root collar, and 2 cm above the root collar. After nine months, we evaluated seedling survival, growth (height and stem diameter), and the occurrence of hypertrophied lenticels and adventitious roots. Seedling survival and growth data were combined to create a flooding tolerance index. Further, we planted the same species in two degraded swampy areas under three soil preparation methods: planting seedlings in regular planting holes, on mounds, and in between furrows. We assessed seedling survival and stem diameter of these seedlings after one year. Results: Flooding tolerance index was significantly correlated to the occurrence of hypertrophied lenticels, but not of adventitious roots. Classifying flooding tolerance of tree species was effective in predicting species growth, but not seedling survival, in restoration plantations. Overall, seedling performance in restoration plantations was highly influenced by species identity, but not by soil preparation methods. Conclusions: Species selection to improve restoration performance of flooded forests can be guided by simple and fast screenings performed in nurseries before planting. Pioneer trees may facilitate the establishment of species less tolerant to flooding by reducing water saturation of superficial soil layers and enhancing micro-relief heterogeneity by litter deposition. Only a few native species should initially be used in the restoration of flooded forests with high water tables.