Monitoring Roseau cane die-back using drone imageries

Abstract

The Mississippi River delta, denoted as the world’s fifth largest delta and accounted as a stressed-out wetland has been witnessing loss of its dominant plant species, Phragmites Australis or Roseau cane. An invasive scale insect Nipponaclerda biwakoensis, commonly known as the Roseau cane scale is suspected of causing an early aging and die-back in the infested Roseau Canes. In this study, we have attempted a remote sensing approach to monitor the Roseau cane health to a spatial extent by identifying the scale infestation from the greenness of the plant. Field surveys were conducted in the Roseau cane die-back area of Pass a Loutre in the lower Mississippi River delta. High-resolution multi-spectral images were collected in different seasons from the RedEdge UAV drone camera at this location where the drone images having 8-centimeter resolution successfully identified the plant stress. We derived the Normalized Difference Vegetation Index (NDVI) from the orthophoto corresponding to the sampling site. Further, we collected sample Roseau cane stems from the stressed-out locations and counted the average scale count per stem. We performed a linear regression analysis between the average scale count per stem at each sampling site with NDVI and obtained an inverse relationship between the scales and vegetation health. Seasonal analysis showed a higher scale count per stem of Roseau cane in the fall season as compared to spring.