Background: Do tree species that have wood of high commercial value which are being extracted in the Saracá-Taquera National Forest have diameter and spatial distribution patterns that enable continuous production of timber to meet the criteria of ecological conservation and sustainable management of their populations? To answer this question, we evaluated diameter and spatial distributions of 15 commercial tree species of ecological and economic importance, in a concession area in the Saracá-Taquera National Forest, Eastern Amazon. Methods: The data were obtained from a forest inventory carried out in 24 50 m × 50 m plots, considering a minimum diameter of 30 cm. The 15 species were selected based on three criteria: importance value index; volume stock; and commercial value of the wood. The trees were distributed into diameter classes with an interval of 10 cm between them to analyse the diameter distribution. The spatial distribution patterns of the species were analysed using the univariate Ripley’s K function. Results: The results showed that the presence of large trees and the aggregated spatial distribution in most of the tree species studied may determine the feasibility of continuous timber production to meet conservation and sustainable management criteria. The spatial distributions of stocks should guide planning of forest management activities, including pre-logging, logging and post-logging activities. However, the low population density of most species suggests that logging must be well planned to avoid population declines. Although forest management on a sustainable basis has advanced substantially in recent years in the Amazon, the factors that promote different diameter and spatial distributions of species are still considered complex and little studied. Conclusions: The results point to a need to expand research on this topic in managed areas, including the same technological approach as the work described here. This type of analysis can constitute an important tool for defining population management strategies for each species, considering their ecological characteristics and environmental adaptation at a local level.
Selective logging alters forest structure and demographic processes in tropical forests, yet long-term post-harvest recovery remains poorly understood. We evaluated growth dynamics, mortality, recruitment, biomass, and canopy structure in a managed forest in the southwestern Amazon six and ten years after reduced-impact logging. The study combined permanent sample plots monitored between 2013 and 2023 with airborne and UAV-based LiDAR surveys conducted in 2015, 2023–2024, and 2025. Temporal patterns in stand dynamics were assessed using generalized linear mixed models, while LiDAR metrics were used to estimate aboveground biomass (AGB), canopy height, and gap dynamics. Tree density declined consistently over time across all functional groups, with time since logging emerging as the dominant driver, whereas logging year showed no significant effect. Recruitment remained low and temporally stable, particularly for commercial species, indicating limited demographic compensation. Mortality increased in the most recent census interval, contributing to persistent net biomass losses despite moderate diameter growth. Field inventories and LiDAR-based estimates indicated sustained reductions in AGB. Canopy trajectories differed between logging cohorts. Areas logged earlier showed partial recovery of canopy height, while more recently logged areas remained structurally open. Gap analyses indicated increasing gap frequency and size over time in logged and fire-affected areas, reflecting ongoing canopy instability rather than progressive closure. These results indicate that forest biomass recovery following reduced-impact logging extends beyond a decade and may exceed the standard 25-year cutting cycle period.
The Southwestern Amazon (SWA) comprises more than 20 million hectares of tropical forest threatened by illegal logging and deforestation. Accurate regional estimates of aboveground biomass (AGB) are essential to support conservation policies. LiDAR data are widely used to estimate biomass, but the generalization of LiDAR-derived AGB models remains constrained by environmental heterogeneity. Our objective was to develop a regional LiDAR-based modeling framework for both airborne (ALS) and UAV-LiDAR platforms capable of generating reliable AGB estimates across SWA forests without requiring additional local ground calibration. We developed and compared multi-site LiDAR linear regression models calibrated using permanent sample plot (PSP) data, evaluating models based exclusively on LiDAR metrics against models integrating geographic covariates (X, Y, and Z). Independent validation was performed using datasets from Jamari National Forest and Iracema II Farm. Models calibrated solely with LiDAR structural metrics overestimated AGB when applied to the independent validation sites. In contrast, integrating elevation (Z) or longitude (X) significantly reduced prediction bias and improved model generalization by capturing environmental variation not represented by canopy structure alone. Tested against independent datasets, these integrated models demonstrated high predictive accuracy and low error, yielding AGB estimates statistically equivalent to field observations. These results demonstrate that simple geographic covariates, can substantially improve the regional generalization of LiDAR-based biomass models. This parsimonious framework enables accurate AGB predictions at independent sites, showing strong potential for reducing inventory costs and expanding regional biomass monitoring. Furthermore, leveraging regional UAV models lowers costs and adds versatility to LiDAR-based forest management planning and monitoring.
21 July 2025 in Southern Forests: a Journal of Forest Science
Monitoring reduced-impact logging (RIL) forest management activities in large and complex areas such as the Brazilian Amazon is an expensive and imprecise activity. Airborne Laser scanning (LiDAR) allows for estimating structural parameters for large areas of the forest quickly and accurately. Herein we propose and evaluate the applicability of five RIL indicators based on LiDAR for areas under a forest concession regime. Five Annual Production Units (APUs) were studied in the Forest Management Unit (FMU) III of the Jamari National Forest located the Southwest of the Brazilian Amazon. These locations were flown over by LiDAR in 2014 and 2015. Digital Terrain (DTM), Surface (DSM), and Canopy Height (CHM) Digital Models were made for each APU. The proposed indicators were: i. dominant and codominant trees; ii. the presence of gaps; iii. impacts on the understory; iv. Canopy vertical profile; and v. Permanent Preservation Areas (PPA) and restrictive areas. The main results were: 3.95% reduction in the occurrence of higher canopy crowns after logging; greater occurrence of small gaps (λ > 1), in which λ values were higher in APU 1 (2.34), with an older exploration history; Gini coefficient values in all locations were lower than 0.5, revealing the low intensity of disturbances in the forest canopy; γ and β parameters of the understory and canopy were not significantly correlated with factors of selective logging; impacts of exploitation on PPAs and restricted areas were less than 2%. The results demonstrate that the proposed LiDAR-derived indicators showed potential for monitoring managed areas in the Amazon.