< Back to previous page
Publication
Forest degradation in the Dry Chaco: A detection based on 3D canopy reconstruction from UAV-SfM techniques
Journal Contribution - Journal Article
Forest degradation is a gradual process through which the forest’s biomass declines, the species composition and vertical complexity change and the soil physicochemical properties degrade. Evaluating forest degradation is challenging, as it needs measurable indicators in order to compare different states of degraded forest
with forests in optimal conditions. The UAV-SfM technique, combining imagery taken from unpiloted aerial vehicles (UAV) and Structure from Motion (SfM) algorithms, offers the opportunity to model forests in three dimensions and to derive ecological indicators describing forest conditions. The research objective of this
paper is to evaluate to what extent information on forest degradation can be derived with UAV-SfM techniques.
The Dry Chaco ecosystem in Argentina offers us the possibility to study the effects of a long history of anthropogenic disturbances on subtropical broadleaf dry
forests. We surveyed a sample of 54 forest plots with different degradation histories with a UAV. Based on existing literature on forest ecology, two main degradation
mechanisms were defined: loss of vertical integrity and vertical complexity. Multivariate statistical techniques were applied on structural indicators extracted from
the 3D canopy model to classify the forest plots into groups or clusters with similar forest conditions. Next, the structural characteristics of the clusters were discussed
in the light of a conceptual framework on forest degradation in the Dry Chaco. The results show that objective criteria derived from UAV-SfM 3D canopy models can
be used to differentiate (1) forests in good condition from (2) forests that were subject to extensive processes of degradation, (3) forest with a dominance of shrubs
and low trees, (4) forests intensively disturbed and (5) low vegetation dominated by bare soils with shrubs.
Journal: Forest ecology and management
ISSN: 0378-1127
Volume: 526
Publication year:2022
Accessibility:Closed