A simple system for phenotyping of plant transpiration and stomatal conductance response to drought

Plant breeding for increased crop water use efficiency or drought stress resistance requires methods to quickly assess the transpiration rate (E) and stomatal conductance (gs) of a large number of individual plants. Several methods to measure E and gs exist, each of which has its own advantages and shortcomings. To add to this toolbox, we developed a method that uses whole-plant thermal imaging in a controlled environment, where aerial humidity is changed rapidly to induce changes in E that are reflected in changes in leaf temperature. This approach is based on a simplified energy balance equation, without the need for a reference material or complicated calculations. To test this concept, we built a double-sided, perforated, open-top plexiglass chamber that was supplied with air at a high flow rate (35 L min-1) and whose relative humidity (RH) could be switched rapidly. Measurements included air and leaf temperature as well as RH. Using several well-watered and drought stressed genotypes of Arabidopsis thaliana that were exposed to multiple cycles in RH (30-50 % and back), we showed that leaf temperature as measured in our system correlated well with E and gs measured in a commercial gas exchange system. Our results demonstrate that, at least within a given species, the differences in leaf temperature under several RH can be used as a proxy for E and gs. Given that this method is fairly quick, noninvasive and remote, we envision that it could be upscaled for work within rapid plant phenotyping systems.

Automated summary

To improve crop water use efficiency and drought stress resistance, a method for quickly assessing transpiration rate (E) and stomatal conductance (gs) of individual plants was developed using whole-plant thermal imaging. The method involves rapidly changing aerial humidity in a controlled environment to induce changes in leaf temperature. The results showed that leaf temperature correlated well with E and gs, making it a potential proxy for assessing these parameters. This noninvasive and remote method could be upscaled for use in rapid plant phenotyping systems.