Assessment of the claslite forest monitoring system in detecting disturbance from selective logging in the Selva Maya, Mexico

Abstract

Detecting and monitoring forest disturbance from selective logging is necessary to develop effective strategies and polices that conserve tropical forests and mitigate climate change. We assessed the potential of using the remote sensing tool, CLASlite forest monitoring system, to detect disturbance from timber harvesting in four community forests (ejidos) of the Selva Maya on the Yucatan Peninsula, Mexico. Selective logging impacts (e.g. felling gaps, skid trails, logging roads and log landings) were mapped using GPS in the 2014 annual cutting areas (ACAs) of each ejido. We processed and analyzed two pre-harvest Landsat images (2001 and 2013) and one post-harvest image (November 2014) with the CLASlite system, producing maps of degraded, deforested and unlogged areas in each ACA. Based on reference points of disturbed (felling and skidding), deforested (log landings and roads) and unlogged areas in each ACA, we applied accuracy assessments which showed very low overall accuracies (<19.1%). Selective logging impacts, mainly from log landings and new logging road construction, were detected in only one ejido which had the highest logging intensity (7 m3 ha–1).