Testing species abundance distribution models in tropical forest successions: Implications for fine-scale passive restoration

Tropical forests are recognized worldwide for providing ecosystem services that contribute greatly to biodiversity and ecosystem functioning. In this study, we hypothesize that tree species abundance distribution is determined by the commonness-rarity balance of species during tropical forest recovery through natural regeneration. We sampled 45 20 m x 50 m plots across three sites in secondary forests with different stand ages (5, 10, 15, and 20 years) after shifting cultivation, and an old-growth forest (> 100 years old) located in the Northern Amazon forest, Venezuela. We explored the variations in woody species abundance distribution and compositional similarity using a chronosequence approach. There was a clear pattern of increasing species abundance in forest successions, from second-growth to old-growth forests. Species abundance in the old-growth forest was much higher than in the second-growth forest. The niche-preemption model showed a high goodnessof-fit to the 5-year old second-growth forest, while 10- and 15-year old second-growth forests and the old-growth forest showed a significant fit to the lognormal model. Along the species-rank axis, the most common species accounted for 54% of the total proportional abundance at the earliest successional stage (5-year old), whereas the relative proportion of the most common species remained relatively stable at 22%-25% in the following successional stages (i.e., old-growth forest). Species dominance was lower in old-growth forest than those in second-growth forest, where there were seven species with the highest importance values within the community. There were obvious differences in relation to the old-growth forest, where rare species were present. The results we present here are crucial for designing and implementing proper, goal-oriented forest restoration strategies, and should aid future research.