Recovery of species composition over 46 years in a logged Australian tropical forest following different intensity silvicultural treatments

Currently, more than 400 million hectares of tropical forests worldwide are now being managed for timber production. Understanding the long-term responses of tropical forests to management practices is critical for managing tropical forests sustainably. To investigate the responses of tropical forest dynamics to different silvicultural treatment intensities, permanent plots were established in 1967 in an Australian tropical forest with four treatments: selective logging only as a control and selective logging followed by three differing intensity silvicultural treatments in 1969. We investigated changes in the number of species, species dominance, and species composition of trees (DBH ≥ 10 cm) from 1967 to 2015. Before selective logging, the number of tree species did not differ between the treatments, ranging from 70 to 75. Selective logging alone had small immediate effects on the number of species and species abundance distributions. After silvicultural treatment, the number of species in the low-intensity treatment, medium-intensity treatment and high-intensity treatment were reduced to 48, 42, and 18 respectively. The number of species in the control, low-intensity, and medium-intensity treatments recovered to their pre-logging levels within 46 years, but recovery in the high-intensity treatment was incomplete due to much greater initial species loss through silvicultural treatment. Silvicultural treatments increased species dominance, with the differences being progressively more marked as the level of treatment intensity increased. Over 46 years, tree species abundance distribution in the silvicultural treatments became more even and largely returned to pre-logging conditions, with more rapid recovery after low and intermediate silvicultural treatments. Following silvicultural treatments, species composition in the logging only, low-intensity and medium-intensity silvicultural treatments did not change markedly, whereas species composition was substantially altered by high-intensity silvicultural treatment and was subsequently distinct from the other treatments. Within 46 years following treatments, species composition in the high intensity treatment showed a recovery trajectory towards pre-logging conditions, leading to increased species compositional similarity among the four treatments. Increasing the intensity of silviculture treatment led to greater time required for recovery of species diversity, composition and compositional similarity. We recommend that high-intensity silvicultural treatments should be avoided if rapid recovery of species diversity and composition is the desired management outcome in tropical rainforests similar to those in our study area.