Mapping aboveground biomass of Moso bamboo (Phyllostachys pubescens) forests under Pantana phyllostachysae Chao-induced stress using Sentinel-2 imagery

Moso bamboo (Phyllostachys pubescens) stands as a pivotal economic bamboo species globally, holding substantial potential for carbon sequestration. Accurate estimation of aboveground biomass (AGB) in Moso bamboo forests is crucial due to its close ties with the ecosystem's carbon cycle. Despite the maturation of monitoring techniques for Pantana phyllostachysae Chao, a significant pest of Moso bamboo, its interplay with AGB in these forests remains enigmatic. This study addressed this gap by categorizing P. phyllostachysae's impact on Moso bamboo forests into four levels: healthy, mild damage, moderate damage, and severe damage. By scrutinizing field data, we delved into the shifts in Moso bamboo leaf biomass under P. phyllostachysae stress. Leveraging Sentinel-2A/B imagery, we extracted diverse correlation factors, including original wave bands, vegetation indices, texture attributes, and vegetation's physical and chemical parameters. Subsequently, machine learning algorithms-namely, random forest (RF), support vector machine (SVM), and extreme gradient boosting (XGB) were employed to achieve remote sensing inversion of AGB in Moso bamboo forests, accounting for the presence of insect pests. We analyzed the response of Moso bamboo biomass sensitive factors and to further clarify the changes of AGB of Moso bamboo forests under insect pest stress at the remote sensing level ultimately. The results showed that (1) the degree of Moso bamboo leaf biomass damage was positively related to the damage level, which gradually increased from 15.15 % to 59.42 %; (2) the RF algorithm excelled in estimating Moso bamboo forest AGB, particularly in May, and inclusion of insect pest considerations enhanced AGB estimation accuracy; (3) among the four factor types, Band information and vegetation indices emerged as most impactful, and Band5, Band11, Band12, NDVI68a and MSAVI were selected the most often; (4) at the remote sensing level, AGB in Moso bamboo forests significantly varies under P. phyllostachysae stress. Healthy areas demonstrate an AGB of 66.9037 Mg ha−1, while heavily affected regions drop to 52.6591 Mg ha−1. It can be seen that combining pest factors for Moso bamboo biomass estimation solves the problem of rough biomass estimation, and this study provides a more promising method for forest growth monitoring.