Aphid infestation severity limits detection of sugarcane aphid stress on sorghum using UAS-derived vegetation index

Unmanned aircraft systems (UASs) and sensors are able to collect multi-temporal remote sensing data over the cropping season for detecting plant stress. In order to apply this technology for management of the sugarcane aphid on sorghum, pest severity was estimated by linking a normalized difference vegetation index (NDVI), a common measure of plant health, to ground measurements of insect abundance and yield. We introduce several UAS advancements in collecting imagery and processing designed to show the fine-scale of plant response to low aphid population levels needed for management purposes. We monitored structured low aphid populations on sorghum using UAS equipped with RGB and multispectral cameras. Analysis of post infestation aphid densities was able to detect significant effects of aphid manipulation on cumulative aphid days. However, the relationships between NDVI, sugarcane aphid abundance, and yield were variable for the aphid densities seen in this study. These results suggest the limitations on the effectiveness of NDVI to monitor aphid populations in sorghum may be more related to the form and severity of aphid feeding and plant response, as opposed to technological problems with UAS derived data.