Change Detection of Invasive Swietenia macrophylla (Mahogany) in Bohol Man-Made Forest from 2013-2023 Through Vegetation and Soil Spectral Indices Using Landsat 8 in Google Earth Engine

Abstract. Swietenia macrophylla (Mahogany), an invasive species introduced in Bohol over half a century ago to mitigate deforestation, was primarily used to reforest the now ecologically vulnerable Bohol Man-Made Forest. Allelopathic effects of IAPS on Philippine ecosystems were previously unknown, but recent studies show Mahogany inhibits nearby biodiversity. This study employed remote sensing to delineate and classify the Mahogany population in Bohol Man-Made Forest, detect land classification differences over ten years, and determine its effects on local vegetation and soil using various indices. Google Earth Engine was utilized to select Landsat 8 Level 2 images, apply cloud masks, and create a 70:30 generalization for Random Forest classification. The timeline and mapping units for using Landsat in training data relied heavily on a 2018 reference map. With this, a key limitation was the assumption that Mahogany points identified in 2018 remained unchanged in 2013 and 2023, based on its growth development. Results show 90.77% and 91.36% overall accuracy for classification in 2013 and 2023. Mahogany class had the highest positive change in total area at 5.334 sq. km. Non-Mahogany vegetation showed the highest change with a decrease of 6.039 sq. km. Bivariate maps of NDVI, MSI, and BSI showed Mahogany’s effects on non-Mahogany vegetation, degrading health and moisture content, and consistently preventing small vegetation growth, allowing high intensity of bare soil. This study recommends planting the right trees in the right places and enforcing strict reforestation policies, especially in forests and protected areas, to preserve natural biodiversity.