Phenotype, phenology, and disease pressure assessments in wild blueberry fields through the use of remote sensing technologies

Wild blueberry fields consist of the management of multiple Vaccinium species (V. angustifolium and V. myrtilloides) with different growth and development rates and susceptibility to diseases including Monilinia and Botrytis blossom blight diseases. To gain better insight into the population structure, floral and vegetative bud and canopy growth and development, research has been underway since 2018 on the use of unmanned aerial vehicles equipped with high resolution red-green-blue digital camera, multispectral, and most recently light detection and ranging (laser 3-D imaging) sensors. Statistical analyses procedures examining vegetative indices and growth parameters have included stepwise multilinear (SMLR), random forest (RF), and support vector machine (SVM) techniques. Results have illustrated that it is possible to identify and geospatially determine the presence of wild blueberry phenotypes of interest in wild blueberry fields from berry yield and disease susceptibility perspectives. Although the 3 statistical analysis procedures provided similar R² and concordance values, the SVM technique was the most effective in assessing phenological stages. The UAV system was also able to detect the incidence and severity of Monilinia and Botrytis blossom blight diseases and resulting changes to vegetation indices. These results have allowed for the generation of prescription maps depicting phenotype and growth stage specific regions of interest in wild blueberry fields which can result in site-specific, localized disease mitigation practices to be used.