S18 - Session P1 - Remote sensing tools to assist in phenotyping canopy architecture in pear rootstock breeding program

Authors: Sindhuja Sankaran *, Mugilan Raman, Zara York, Eduardo Carlos, Soon Li Teh, Kate Evans

The U.S. Pacific Northwest, a major pear production area, lacks cold-tolerant dwarfing rootstocks.

This limits establishment of the two-dimensional orchard structures that are critical for efficient fruit production and incorporation of engineering solutions to address the chronic shortage of labor. Breeding new improved rootstocks requires complex, challenging and time-consuming phenotyping of tree and canopy vigor, and related architectural traits in early breeding stages. In this project, our overall goal was to utilize remote sensing approaches (RGB and LiDAR) to assist in the extraction of digital features (canopy volume, height, branching pattern) that can assist in evaluating the overall canopy vigor associated with the rootstock germplasm. . About 450 seedling trees were imaged with a quadrocopter flight speed of 2.5 m/s from 15 m above ground level flying altitude. Multiple flight plans were incorporated at 45° and 65° in either direction, in addition to at nadir angleGround-based LiDAR data was also acquired to extract more detailed variability in the canopy architecture. Preliminary evaluation of data indicated that the canopy height estimated from integrated orthomosaic imagery (45°, 65°, nadir) was highly correlated to the canopy height estimated from orthomosaic imagery from data acquired at 45°. The extracted digital traits will be compared to the ground-reference data such as total height, trunk cross-sectional area, and total canopy growth to determine the potential for future phenotypic evaluation.