The effect of spring flooding on management and distribution of cotton bollworm (Helicoverpa armigera) by flood mapping using SAR sentinel-1 and optical imagery landsat-8; a case study in golestan province, Iran

Abstract

On March 2019 a severe flood episode inundated large agricultural areas of Golestan province, in northern Iran. In this study, we assessed the impacts of the flooding episode on cotton bollworm, Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae). To do so, a network of pheromone-bait traps distributed throughout Golestan was installed and used to weekly record moth captures from May to September. The interpolated spatial distribution of cotton bollworm infestation by the Inverse Distance Weighted (IDW) interpolation method was overlaid by satellite-imagery flood maps. Flood extent maps were provided by seven mosaicked images of Sentinel-1. In addition to SAR products, a 15 m-NDWI of Landsat-8 was used. The flood extent based on NDWI (pan-sharped 15 m) was 52752 ha (47423 ha based on slope mask). Using ground control points of flooded and non-flooded areas within the supervised SVM classification procedure, the kappa coefficient (κ) was as 0.89 and 0.91 for VH polarization and NDWI, respectively. Our results showed a reduction in moth populations in the areas that were flooded in spring for the earlier generations of H. armigera. Highlights The impact of a flood episode on moth populations of H. armigera was investigated Change detection analysis from satellite images at two times allowed flood mapping Moth populations were drastically reduced in flooded areas Flood mainly affected the populations in the first and second generations Our results suggest that intentional flooding might be used for controlling H. armigera