High-throughput canopy light interception using aerial LiDAR in maize

Canopy light interception is an important physiological trait that is traditionally measured using a light bar. However, this measurement is constrained by time and logistical challenges (days without clouds, only around noon, working in high temperatures), making it difficult to implement in large experiments such as those conducted in maize breeding programs. Recent advances in remote sensing, particularly the use of aerial LiDAR, have the potential for high-throughput characterization of canopy structures. Our objective was to evaluate whether aerial LiDAR can reliably estimate light interception and to determine its feasibility for deployment in commercial maize breeding experiments. To test this, aerial LiDAR data and light bar measurements were collected at five experiments varying in genotype, plant density, and irrigation treatments. Grain yield varied across experiments and treatments, from 3.8 to 24.6 Mg ha -1 , while maximum leaf area index ranged from 1.3 to 12.2 m 2 m -2 . Voxel canopy cover derived from LiDAR was strongly correlated with light bar light interception measurements across all environments and treatments (r = 0.96). A universal linear model, barely departing from the 1:1 line, showed high precision, low bias, and consistent performance across environments, with RMSE ranging from 5.5 to 7.9%. Experiment specific calibrations improved predictions by only 5% compared to the 1:1 (LiDAR voxel canopy cover equals light interception) and by 10% compared to the universal linear model, suggesting that individual experiment calibration is not needed. LiDAR light interception showed lower plot-to-plot error and higher treatment heritability than light bar light interception measurements. Furthermore, canopy light interception evolution curves over time derived from LiDAR data provided more accurate fits than those derived from light bar data in 98% of cases. We conclude that aerial LiDAR is an effective, high-throughput tool that can be used for measuring light interception in maize, providing more consistent and scalable data than traditional light bar measurements. A model in which LiDAR voxel canopy cover equals light interception works well across diverse conditions, and a workflow is described to support adoption. In the discussion, we further describe positive and negative aspects of the adoption of LiDAR compared to the traditional light bar, highlighting the value it provides for maize breeding, crop physiology, and overall field research.