Acacia longifolia, like any other invasive species, poses a threat to the natural ecosystems in South Africa and beyond. In South Africa, the species is reported to cause significant loss to vegetation biodiversity and is associated with high water use. It remains a challenge to understand the extent of their invasion in complex landscapes, where the use of ground-based surveys is difficult. Considering the heterogeneity and complexity of hydrological and ecological processes at catchment scale, the implementation of remote sensing to detect the distribution of these species becomes critical for continuous monitoring of their extent and associated impacts. We, therefore, aim at species-specific characterization of the spatial distribution of Acacia longifolia invasive wood species at catchment scale within the Cape Floristic region located in the Western Cape Province, using Sentinel 2 remotely sensed data. Overall, the distribution of A. Longifolia was found to be patchy and heterogeneous, across the catchment. However, the main hot spots were detected within the fynbos vegetation, low-lying areas, as well as along the southern coastal belt. Although the catchment is predominantly under cultivation, A. longifolia was found to occupy almost 10% of its area. Classification using combined variables, using both vegetation indices and bands produced the highest overall accuracy of 88.66%, than when bands (66.19%) and indices (76.05%) were used as independent dataset. For example, the producer’s classification accuracy of A. longifolia increased by almost 15%, whereas the user’s increased by almost 10% from spectral bands accuracies. The classification of A. longifolia was primarily attributed to near-infrared, short wave-infrared (centered at 2190 nm), red edge (705 and 740 nm), as well as, chlorophyll vegetation index, the leaf water content index, and normalized difference vegetation index, derived using the near-infrared and red-edge band (at 740 nm) variables. The findings of this study underscore the relevance of satellite-based remote sensing in monitoring invasions in natural ecosystems and can be used as baseline information for invasive-species clearing endeavors in the area. |
ACCESS THE FULL ARTICLE
No SPIE Account? Create one
CITATIONS
Cited by 6 scholarly publications.
Vegetation
Remote sensing
Ecosystems
Short wave infrared radiation
Atmospheric modeling
Image classification
Satellites