Selective logging effects on abundance, diversity, and composition of tropical understory herbs

Abstract

Little is known about the impacts of selective logging on herbaceous plants of tropical forests. If logging impacts on nontimber species are low and little time is required for recovery, logged forests could be compatible with conservation of some elements of the understory flora. As many nontimber species may have commercial value, knowledge about the impacts of different logging intensities is necessary to determine levels of timber extraction compatible with multiple use of forests. In this study, we examined the effects of logging intensity, time after logging, and skidder tracks, on the composition and diversity of the ground-level herb community of a terra-firme forest in central Amazonia. Logging was carried out at varying intensities in eight 4-ha experimental plots in 1987, three plots in 1993, and five plots were controls. Herb community composition, analyzed with hybrid multidimensional scaling (HMDS), was not significantly affected by logging intensity, but was significantly related to the area covered by logging gaps and skidder tracks. There was no significant difference in community composition and richness between treatments with different times after logging. Species richness was also not related to the logging intensity. Some species (18.2%) were found only in logged plots and were mainly from habitats outside the rain forest. There were significant differences in herb composition between controls and skidder tracks, but recent and old skidder tracks were mixed in the ordination space, suggesting that recovery in skidder tracks is slower than in other areas affected by logging. Overall, our results indicate that the ground herb community is not severely affected by selective logging at the intensities used in this experiment. The alterations in composition are mainly local, and restricted to the most disturbed patches. We conclude that forests managed with logging intensities similar to those used in this experiment could be compatible with conservation of understory herbs. Skidder tracks should be minimized, as they have the greatest long-term impact and allow species from other habitats to invade the forest.