Spatial patterns of logging-related disturbance events: a multi-scale analysis on forest management units located in the Brazilian Amazon

• Published in: Landscape ecology Vol. 35; no. 9; pp. 2083 - 2100
• Main Authors: Lima, Thaís Almeida, Beuchle, René, Griess, Verena C., Verhegghen, Astrid, Vogt, Peter
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.09.2020; Springer Nature B.V
• Subjects: Biomedical and Life Sciences; Disturbance; Ecology; Environmental Management; Forest management; Harvesting; Landscape Ecology; Landscape/Regional and Urban Planning; Life Sciences; Logging; Mapping; Multiscale analysis; Nature Conservation; Pattern analysis; Research Article; Small mammals; Spatial analysis; Sustainable Development; Tropical forests; Brazil; Disturbance profiles; Composition; Multi-scale analysis; Configuration; Selective logging
• ISSN: 0921-2973; 1572-9761
• DOI: 10.1007/s10980-020-01080-y

Context Selective logging has been commonly mapped using binary maps, representing logged and unlogged forests. However, binary maps may fall short regarding the optimum representation of this type of disturbance, as tree harvest in tropical forests can be highly heterogeneous. Objectives The objective of this study is to map forest disturbance intensities in areas of selective logging located in the Brazilian Amazon. Methods Selective logging activities were mapped in ten forest management units using Sentinel-2 data at 10 m resolution. A spatial pattern analysis was applied to the logging map, using a moving window approach with different window sizes. Two landscape metrics were used to derive a forest disturbance intensity map. This map was then compared with actual disturbances using field data and a post-harvest forest recovery analysis. Results Disturbed areas were grouped into five distinct disturbance intensity classes, from very low to very high. Classes high and very high were found to be related to log landings and large felling gaps, while very low intensities were mainly related to isolated disturbance types. The post-harvest forest recovery analysis showed that the five classes can be clearly distinguished from one another, with the clearest differences in the year of logging and one year after it. Conclusions The approach described represents an important step towards a better mapping of selectively logged areas, when compared to the use of binary maps. The disturbance intensity classes could be used as indicators for forest monitoring as well as for further evaluation of areas under forest management.