Year: 2016
Type: dissertation
Abstract: Background: Tropical rain forests are highly biodiverse and provide globally important ecosystem services. In the face of increasing dominance of human-modified landscapes in tropical regions, sustainably managing these forests may contribute to maintain ecosystem integrity while fulfilling society’s demand for forest goods. Sustainable forest management is based on the assumption that forest ecosystems are resilient to silvicultural interventions and that their ecosystem functions recover across felling cycles. So far, our understanding of tropical forest responses to silvicultural interventions is rather limited, especially regarding the medium to long-term post-logging recovery. Aims: The aim of this thesis was to analyse the effects of silvicultural intervention intensity and type (harvesting and thinning) on recovery of important forest attributes and functions related to biodiversity conservation, carbon sequestration and timber provision. In addition to the disturbance regime, vegetation attributes of the remaining forest stand may affect recovery. In less diverse ecosystems, it has been demonstrated that diversity in species or traits may increase resilience to disturbances because species complement each other for resource use (niche-complementarity hypothesis), or dominant species with their traits may drive ecosystem processes (mass-ratio hypothesis). However, such effects have not been investigated yet in relation to the resilience of tropical forests. Here, the effects of niche-complementarity and biomass-ratio hypotheses on biomass resilience were investigated. Research questions: 1. What are the dynamics of tree communities in the process of recovering diversity and composition relative to intervention intensity and type? 2. How is the aboveground biomass recovery affected by disturbance intensity and attributes of the remaining tree community? 3. How is the development of growing stock of commercial timber species affected by silvicultural intervention intensity and type? Study site and experiment: Data analysed in this thesis come from a unique long-term study on management effects on forest ecosystem recovery. The experiment with an area of 180 ha was established in 1981 in the Tapajós National Forest, Pará, Brazil by the forest management group of Embrapa Amazônia Oriental. Trees (DBH ≥ 5 cm) were measured and permanently labelled in 35 permanent sample plots of 0.25 ha each. Eight repeated inventories took place from 1981 to 2012. Additionally, six control plots (0.25 ha) have been monitored in unlogged forest as of 1983. Intervention intensities comprised logging (on average 61 m³ ha−¹) and associated damage to remaining trees (1982) and thinning (1993/1994). These interventions together resulted in a gradient of reduction in basal-area ranging on average from 19 to 53% relative to pre-logging stocks. Analytical approach: Relative recovery of diversity, composition, biomass and timber stocks (in relation to pre-logging conditions) and other response variables (e.g., diversity and recruitment) were compared among intervention intensities, types and to the unlogged control. Furthermore, temporal changes in response variables were investigated in some cases (e.g., composition and growth rates). Effects of post-disturbance tree-community attributes on biomass resilience were further investigated using structural equation modelling. Results and discussion: Tree diversity was not impaired by silvicultural interventions but high intervention intensities (logging followed by medium and high thinning intensities) substantially affected the recovery of tree species composition (DBH ≥ 10 cm). This change in composition was underlined by shifts in the trait-mean of wood density towards lower values overtime, indicating a transition towards pioneer dominated communities. Such change may affect ecosystem functioning, for instance, species interactions due to lower availability of fleshy fruits. Moreover, the reduction of basal area through harvesting damage and thinning had a stronger effect on species composition than logging of mature trees itself. These findings demonstrate that intensive thinning should be avoided and they point to the importance of reduced impact logging for easing damage to retained trees and hence lessen possible unintended changes to tree species composition. These measures may enhance species composition recovery and maintenance of biodiversity at other trophic levels. Stand-level aboveground biomass recovery rates (DBH ≥ 10 cm) were largely determined by basal area remaining and less by post-disturbance composition. Below a post-logging basal area of ca. 23 m² ha−¹ and following intensive thinning interventions, recovery rates were impaired. The weak support for both the niche complementarity and the biomass-ratio hypotheses was possibly related to high functional redundancy in these species-rich forests, or because some species modulate post-intervention biomass recovery rather than overall diversity or community-mean trait values. These results suggest that heavy interventions may impair biomass resilience. Moreover, these forests retained high biomass stocks and fast recovery rates (on average 6.8 Mg ha−¹ year−¹), indicating the potential of managed tropical forests to contribute to climate change mitigation. Recruitment of small stems (5 cm ≤ DBH < 15 cm) and volume increment of future crop trees (15 cm ≤ DBH < 50 cm) temporarily increased following silvicultural interventions. High recruitment and growth rates were mainly driven by long-lived pioneer species so that future harvests will have to deal with a different pool of commercial species. When only the 22 commercial tree species that were harvested in 1982 were considered, the recovery in harvestable volume (DBH ≥ 50 cm) over 30 years ranged from 19 to 57%. However, if 30 additional tree species were considered that have been recognized as commercial timber species since, the recovery in the volume of the harvestable growing stock ranged from 27 to 143%. Light to medium thinning intensities promoted recovery, whereas intensive selection harvesting and follow-up thinning impeded recovery of harvestable growing stocks. These findings point to the importance of silvicultural interventions oriented towards future crop trees of target species to enhance recovery and reduce changes in composition of the commercial growing stock. Conclusions: A common finding was that the recovery in different important ecosystem attributes (tree species composition, biomass and timber stocks) within 30 years following initial logging was impaired when basal area was reduced more than 20% relative to pre-logging stocks. Furthermore, the results indicate that heavy harvesting damage and thinning may lead to unintended ecosystem changes that may affect the provision of ecosystem services. Results further suggest that current maximum harvesting intensity allowed by legislation (i.e., 30 m³ ha−¹ for a 30-year felling cycle) for the Brazilian Amazon is not likely to impair tree species composition and biomass recovery. Lastly, this thesis constitutes one of the few studies to date that addresses a comprehensive view of post-logging recovery over such a long-time period. The knowledge presented here contributes to improving our understanding of the resilience of tropical rain forests to silvicultural interventions.
Author: de Avila, Angela Luciana
Institution: Wageningen University & Research
Language: English
Subfield: Forestry
Domain: Life Sciences
Keywords: tropical rainforest, post-logging regeneration, forest management practices, logging, thinning, tree species diversity, tree community composition, commercial timber species, carbon sequestration, forest biodiversity conservation, Tapajós National Forest, long-term forest monitoring, disturbance intensity
SDG: Life on land
Open Access status: green
OpenAlex ID: W7033435940
