LAUSR

In many parts of the world, forestry objectives are now shifting from focusing on maximum production to a wider perspective that includes biodiversity preservation and ecosystem functioning. To achieve these targets, managers are increasingly designing cutting regimes that imitate natural disturbances, with the aim of generating a more naturally structured forest. In many old-growth forests, tree fall gaps caused by tree death constitute the dominant type of disturbance. The application of gap dynamics theory appears to be a promising option for tropical forest management and conservation. In the present study of a tree species-rich and old-growth Ecuadorian Neotropical forest, we assessed the spatial distribution of gaps and gap size in relation to: i) tree number at the gap edge, ii) number of tree species at the gap edge, iii) number of tree species per stem at the gap edge, iv) species similarity, v) species evenness at the gap edge, vi) size differentiation at the gap edge, vii) gap isolation and viii) species mingling at the gap edge. Understanding natural gap processes in these forests is crucial for establishing new forestry practices in these forests to mimic natural processes of disturbance. Our results demonstrated that the spatial distribution of gaps was random. Various gap attributes were strongly associated with gap size. The number of tree species per stem at the gap edge was negatively correlated to the gap size. Gap sizes up to 50 m 2 were mostly sufficient to generate tree species-rich forest stands. Assuming that our results were representative for an old-growth neotropical rainforest in Ecuador, our study remarks the following management recommendations: 1) Rainforests have a very complex spatial and diversity structure and logging activities should preferably be omitted because of adverse effects. 2) If logging is inevitable, this should mimic a random choice of trees and tree species, to prevent special selection of tree dimension and species; and a random distribution of trees to be logged, to produce gaps smaller than 50 m 2 and never larger than 400 m 2 . Additionally, we suggest cutting not more than 5 % of the tree biomass per 10-20 years period, to preclude stronger alterations of ecosystem processes, and the reduction of existing dead wood from the ecosystem.