Monitoring and Hotspots Identification of Invasive Salvinia molesta in Mwanza Gulf Using Remote Sensing Technique

Abstract

Invasive plants, such as aquatic macrophytes, pose environmental challenges that require immediate attention, as they can cause considerable economic and social-ecological damage if left to spread. The invasion of Salvinia molesta (S.molesta) along the Gulf of Mwanza, Lake Victoria, was recently reported. Thus, this study focused on detecting and monitoring its presence in the Gulf of Mwanza waters, quantifying the extent of infestation, monitoring its spatiotemporal distribution from 2020 to 2025, and identifying the hotspots. The study applied a remote sensing technique using multispectral imagery from Sentinel-2, a Support Vector Machine, coupled with freely available high-resolution imagery delivered by Google Earth Pro. Three classes, namely water, S.molesta, and built-up areas, were employed in the analysis, which yielded satisfactory findings with overall accuracy metrics consistently exceeding 98.5% and Kappa coefficients greater than 0.985, confirming the reliability of the methodology for the operational detection of S.molesta invasion along Lake Victoria. The results revealed an alarming increase in S.molesta infestation along the Gulf of Mwanza, with approximately 7.72 km² coverage portraying a 60.3% increase over the past five years. Moreover, the edges of the Kigongo-Busisi Bridge were identified as highly significant hotspot locations for S.molesta infestation. These hotspots were aligned with field observations, suggesting an existing link to potential nutrient-rich runoff originating from adjacent agricultural and residential areas. This study has generated vital information for the responsible authorities to institute management strategies geared towards combating the escalation of S. S.molesta invasion in Lake Victoria.