Development of Light Sensors to Measure Defoliation

Abstract

Sensors have become valuable tools in agriculture when decisions regarding inputs require precision and speed. For example, factors in estimating defoliation in row crops, such as intensive labor and, in particular, subjectivity, are greatly reduced with the use of sensors that can remove these limitations and biases. Estimates of defoliation are almost always overestimated due to human error and biased, unconscious efforts to locate injury. To address these issues, the accuracy and preciseness of a light-based sensor to detect defoliation was tested by measuring simulated levels of defoliation (0-100%) on paper “leaves” at seven light intensities. Results indicated that higher lux values were detected through thinner paper (filter paper) than through thicker paper (cardstock), demonstrating that leaf thickness could potentially affect accuracy of the light-sensor system. Despite some light penetrating the thinner paper with simulated defoliation levels, the two light sensors tested yielded accurate and precise predictions of defoliation (R2 > 0.95). This light-sensor approach could potentially be used in the field to report real-time measurements of defoliation in row crops, such as soybeans, or in other plant-based systems where losses of leaf area require monitoring in order to prevent economic injury.