Quantitative analysis of landslide impact on vegetation: Insights from field surveys and UAV imagery

Abstract

Deep-seated landslides can have significant and long-lasting impacts on surrounding ecosystem by altering topography and destabilizing the ground. This study presents a quantitative investigation of the impacts of deep-seated landslides, which have actively supplied sediments for more than one hundred years, on damaged trees, incorporating various factors such as distance from landslide borders, slope gradient, factor of safety (FS), Wind Exposure Index (WEI), Convergence Index, and plan curvature in four landslides in Japan (Shichimenzan, Senmai, Sarugare, and Akakuzure). The study aims to enhance our understanding of the relationship between landslides and vegetation, providing valuable insights for landslide management and ecological restoration. Using a combination of field surveys and high-resolution UAV (Unmanned Aerial Vehicle) images from 2020 and 2023, a total of 2057 damaged trees were identified across all four study landslide areas. Statistical analysis (One-Way ANOVA) show there is significant differences among different damaged tree types in studied variables in 99 % of confidence level. Analysis of the distance from landslide borders revealed significant differences among damaged tree types. The average distance of Fallen and Green (FG) trees from the landslide borders was found to be 7.3 m, indicating their proximity to the affected area. Furthermore, the density of FG trees within the expanded area of the landslide was observed to be higher. Stand and Dead (SD) trees, in contrast, were situated at an average distance of 68.6 m from landslide border, suggesting their vulnerability to animal-related damages. Fallen and dead (FD) trees were associated with steep slope gradients, averaging 40.8 degrees, and exhibited low FS values (0.87), indicating their susceptibility to slope instability. Stand and Dead partially trees (SDP) and SD trees demonstrated higher FS values suggesting their presence in areas with superior slope stability. In addition, results show, FG, SD, and SDP trees were predominantly located in wind-exposed higher elevation areas. FD trees were primarily situated in areas with negative Convergence Index values (−0.7), indicating slope convergence. FG and SDP trees were found in areas with positive values, suggesting slope divergence. The findings enhance our understanding of the complex and long-lasting landslide impacts on forests, informing landslide management and ecological restoration strategies.