No abstract

The objective of this study was to test the use of repeat flight, airborne laser scanning data (lidar) for estimating changes associated with low-impact selective logging (approx. 10-15 m 3 ha −1 = 5-7% of total standing volume harvested) in natural tropical forests in the Western Brazilian Amazon. Specifically, we investigated change in area impacted by selective logging, change in tall canopy (30 m+) area, change in lidar canopy structure metrics, and change in above ground biomass (AGB) using a model-based statistical framework. Ground plot measurements were only available from the time of the 2010 lidar acquisition. A simple differencing of the 2010 and 2011 lidar canopy height models identified areas where canopy over 30 m tall had been removed. Area of tall canopy dropped from 22.8% in 2010 to 18.7% in 2011, a reduction of 4.1%. Using a relative density model (RDM) technique the increase in area of roads, skidtrails, landings, and felled tree gaps was estimated to be 17.1%. A lidar-based regression model for estimating AGB was developed using lidar metrics from the 2010 lidar acquisition and corresponding AGB ground plot measurements. The estimator was then used to compute AGB estimates for the site in 2010 and 2011 using the 2010 and 2011 lidar acquisition data, respectively. A model-based statistical approach was then used to estimate the uncertainty of the changes in AGB between the acquisitions. Change in RDMs between lidar acquisitions was used to classify each 50 m cell in the study area as impacted or non-impacted by logging. The change in mean AGB for the entire study area was − 9.1 Mg ha −1 ± 1.9 (mean ± SD) (P-value b 0.0001). The change in mean AGB for areas newly impacted in 2011 was − 17.9 ± 3.1 Mg ha − 1 (P-value b 0.0001) while the change in mean AGB for non-impacted areas was significantly less at − 2.6 ± 1.1 Mg ha −1 (P-value = 0.009). These results provide corroborating evidence of the spatial extent and magnitude of change due to low-intensity logging in tropical forests with heavy residual canopy cover.

Around 30 Mm 3 of sawlogs are extracted annually by selective logging of natural production forests in Amazonia, Earth's most extensive tropical forest. Decisions concerning the management of these production forests will be of major importance for Amazonian forests' fate. To date, no regional assessment of selective logging sustainability supports decision-making. Based on data from 3500 ha of forest inventory plots, our modelling results show that the average periodic harvests of 20 m 3 ha −1 will not recover by the end of a standard 30 year cutting cycle. Timber recovery within a cutting cycle is enhanced by commercial acceptance of more species and with the adoption of longer cutting cycles and lower logging intensities. Recovery rates are faster in Western Amazonia than on the Guiana Shield. Our simulations suggest that regardless of cutting cycle duration and logging intensities, selectively logged forests are unlikely to meet timber demands over the long term as timber stocks are predicted to steadily decline. There is thus an urgent need to develop an integrated forest resource management policy that combines active management of production forests with the restoration of degraded and secondary forests for timber production. Without better management, reduced timber harvests and continued timber production declines are unavoidable.

2017-08-25T16:27:58

No abstract