Assessing the recovery in species, size and location diversities of a lowland tropical rainforest after shifting cultivation by multiple indices at stand and neighborhood scales.

Tropical forests are rapidly being converted for agricultural use, but abandoned agricultural lands can recover naturally through secondary succession. However, comprehensive knowledges of how species composition, size structure and spatial patterning (represented by species, size and location diversities) change during recovery at multiple scales are still lacking. Our aim was to explore these change patterns to understand the underlying mechanisms of forest recovery and propose corresponding solutions for restoring regrowing secondary forests. Here, twelve 1ha forest dynamics plots (4 plots each in young-secondary forests (YS), old-secondary forests (OS) and old-growth forests (OG) from a chronosequence of tropical lowland rainforest after shifting cultivation) were used to assessed the recovery in species, size and location diversity of trees at stand (plot) and neighborhood (focal tree and its neighbors) scale by using 8 indices. The relative recoveries of YS and OS were quantified by dividing each of the indices in YS and OS to those in OG. Results showed that species and size diversity increased while location diversity decreased with the recovery process. The relative recovery of location diversity was higher than those of species and size diversity in both YS and OS, while species diversity was only higher than size diversity in YS. The relative recovery of species diversity at neighborhood scale was higher than that at stand scale in OS, while there were no differences between scales in size and location diversity. Additionally, using only the Shannon index and Gini coefficient at two scales can provide consistent insights into the recovery patterns of diversity as indicated by the 8 indices. Our study demonstrated that recovery rates of secondary forests relative to old-growth counterparts could be comprehensively quantified using multiple diversity indices in three types at two scales. This quantitative assessment on the relative recovery of disturbed forests could be helpful in applying appropriate management activities and selecting rational approaches to speed up restoration process of degraded forest ecosystems.