The effects of selective logging on tree diversity, changes in tree species composition and plant functional types were studied with the use of seven permanent plots in virgin and in logged forest. All plots were located in a lowland dipterocarp rain forest in East Kalimantan on the island of Borneo. Just after logging and during the following 20 yr tree diversity measured as Fishers' α was not significantly affected in logged forest plots. Temporal shifts in tree species composition were analysed with Principal Component Analysis (PCA). Logged forest plots had much larger changes over time than virgin forest plots. In the smallest diameter class, some logged forest plots showed a distinct trajectory in PCA space compared to virgin forest plots, while in larger diameter classes movement of logged plots in PCA space was random. This suggests that there is no pre-determined community to which logged forest plots tend to shift when recovering from logging. We found a significant negative correlation between diameter increments and the species-specific wood densities of tree species. Species-specific wood density and potential tree height were used to assign species to five PFTs. As expected, logging increased the fraction of softwood stems in small diameter classes. In the largest diameter classes (≥ 50 cm DBH) a strong decrease of softwood emergent stems was found in logged forest plots. After more than 20 yr no recruitment was found of softwood emergent stems in selectively logged forest.

Nomenclature: Sidiyasa et al. (1999); Anon. (1955–1994).

Abbreviations:GRH = Gap-phase Regeneration Hypothesis; IDH = Intermediate Disturbance Hypothesis; LLOG = Low-intensity logging; MLOG = Moderate-intensity logging.