Enhancing giant salvinia (Salvinia molesta) biocontrol with real-time monitoring using Google Earth engine

Salvinia molesta is an invasive aquatic weed native to Brazil that presents significant economic and socio-environmental challenges across Louisiana. The salvinia weevil (Cyrtobagous salviniae) has proven to be an effective biological control agent in tropical and subtropical regions. However, the widespread distribution of S. molesta in inland waterbodies makes it difficult to track, limiting the optimization of biocontrol strategies and early detection efforts. This study aims to 1) develop a near-real-time satellite-based monitoring tool for detecting S. molesta infestations, and (2 evaluate salvinia health and biocontrol impact using drone-based spectral data. To achieve these objectives, an open-access web application was created in Google Earth Engine using Sentinel-2 satellite imagery to visualize aquatic vegetation coverage. This web application allows users to visualize changes on aquatic vegetation using satellite imagery and generate time series coverage and health over time using NDVI values. For example, in Palm Lake (Slidell, LA), the tool detected 20% coverage (3.64 ha) of invasive aquatic vegetation for May 2025, with and NDVI value 0.652. Complementarily, controlled mesocosm experiments using UAV-based RGB and multispectral imagery were conducted to quantify the spectral response of salvinia under varying densities of C. salviniae herbivory. Results from these controlled trials are being used to establish spectral thresholds that link plant reflectance to known weevil activity, enabling future validation of biocontrol performance in the field. By integrating remote sensing through satellite-driven and drone data, this framework represents a step forward in aquatic weed management through remote sensing and early-detection systems.