Evaluation of macadamia felted coccid (Hemiptera: Eriococcidae) damage and cultivar susceptibility using imagery from a small unmanned aerial vehicle (sUAV), combined with ground truthing

Authors: PULAKKATU-THODI, I - Orise Fellow; DZURISIN, J - Colorado State University; Follett, Peter

Submitted to: Pest Management Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/11/2022
Publication Date: 7/27/2022
Citation: Pulakkatu-thodi, I., Dzurisin, J., Follett, P.A. 2022. Evaluation of macadamia felted coccid (Hemiptera: Eriococcidae) damage and cultivar susceptibility using imagery from a small unmanned aerial vehicle (sUAV), combined with ground truthing. Pest Management Science. 78(11):4533-4543. https://doi.org/10.1002/ps.7073.
DOI: https://doi.org/10.1002/ps.7073

Interpretive Summary: A small Unmanned Aerial Vehicle (sUAV) or “drone" was used to collect aerial high-resolution images of macadamia felted coccid (MFC) damage to macadamia nut trees in Hawaii orchards. Heavy infestations of MFC can kill branches, resulting in characteristic dead, copper-colored leaves, which are readily visible from the ground and from the air. RGB (red-green-blue) color images from UAV provided estimates of percent foliage damage which were compared with ground-based estimated (ground truthing). Spatial analysis showed clustering of foliage damage that reflected cultivar susceptibility to MFC infestation, with cultivars 344 and 856 being susceptible, and cultivars 800 and 333 being resistant. At foliage damage levels <10%, sUAV and ground truthing data were significantly positively correlated, suggesting sUAV may be useful in detecting early stages of MFC infestation and damage. Methods for post-collection processing of sUAV images to improve detection of low-level damage to the canopy are discussed.

Technical Abstract: Small Unmanned Aerial Vehicles (sUAV) or “drones,” combined with spatial data analysis, can provide growers with accurate and high-resolution detection of plant stress due to pest infestations. Macadamia felted coccid, Acanthococcus ioronsidei (Williams) (Hemiptera: Eriococcidae), is a significant pest of macadamia nut, Macadamia integrifolia Maiden & Betche (Protaceae), in Hawaii, and heavy infestations can kill branches, resulting in characteristic dead, copper-colored leaves. We investigated the feasibility of using RGB (red-green-blue) color images from sUAV to detect dieback caused by MFC infestation and compared sUAV estimates with ground-based damage estimates (ground truthing). Spatial analysis showed clustering of foliar damage that reflected cultivar susceptibility to MFC infestation, with cultivars 344 and 856 being susceptible, and cultivars 800 and 333 being resistant. sUAV and ground-based estimates of foliar damage were similar for the cultivar 344, but ground-based assessments were higher than sUAV for cultivar 856, possibly due to the differences in canopy architecture and significant early dieback in the lower canopy. At foliar damage levels <10%, sUAV and ground truthing data were significantly positively correlated, suggesting sUAV may be useful in detecting early stages of MFC infestation. Methods for post-collection processing of sUAV images to improve detection of low-level damage to the canopy are discussed.