Search

Stomatal conductance (gs) quantifies the rate of exchange of carbon dioxide for photosynthesis and water vapor for transpiration between plant leaves and the atmosphere. gs is usually measured by handheld devices like porometers, and readings are manually taken in the field, which is time-consuming and labor-intensive. In this study, we investigated the use of high-throughput phenotyping (HTP) data combined with weather data to estimate gs through machine-learning (ML) modeling. The experiment was conducted in a research field equipped with an HTP platform in 2020 and 2021 involving maize, sorghum, soybean, sunflower, and winter wheat. Weather variables including dew point temperature, wind speed, air temperature, solar radiation, and relative humidity were collected by an onsite weather station. Plot-level canopy temperature, soil temperature, and seven vegetation indices were acquired using a thermal infrared camera, a multispectral camera, and a visible near-infrared spectrometer integrated on the HTP platform. Three supervised ML methods (Partial Least Squares Regression (PLSR), Random Forest Regression (RFR), and Support Vector Regression (SVR)) were employed to train the estimation models for gs, and model performance was evaluated by Coefficient of Determination (R2) and Root Mean Squared Error (RMSE). The result showed that RFR and SVR outperformed PLSR in gs modeling. The RFR model achieved R2 of 0.63 and RMSE of 0.16 mol m−2·s−1 with the combination of phenotyping data and weather data. It outperformed the model using only the weather data (R2=0.35 and RMSE=0.21 mol m−2·s−1), or the model using only the phenotyping data (R2=0.46 and RMSE=0.19 mol m−2·s−1). This result suggested that high-throughput plant phenotyping data effectively complement weather data in estimating gs rapidly and non-destructively through ML. With the wide adoption of HTP technologies in aerial and ground-based platforms, this research provides a practical framework to estimate gs at large scale for crop breeding and irrigation management.

Two-dimensional (2-D), nonseparable, piecewise cubic convolution (PCC) for image interpolation is described and discussed.

Platforms and instrumentation for Field High-Throughput Plant Phenotyping (FHTPP) are well developed to measure important traits for crop breeding and agronomic studies. However, the research has focused on morphological and spectral traits; and approaches to estimate major physiological processes such as evapotranspiration (ET) for small experimental plots are lacking. In this study, we put forward a new analytical framework to estimate plot-scale ET by integrating frequent phenotyping data (multispectral and thermal infrared images, canopy reflectance, and LiDAR point clouds) from a FHTPP system (known as NU-Spidercam), the weather data, a simplified two-source energy balance model, and reference ET and crop coefficient calculation. The new plot-scale ET method was tested on five field experiments involving maize and soybean crops over two growing seasons, with the different treatment levels of irrigation water. Estimated plot-scale ET was accumulated across the growing reason for each plot, and its association with grain yield was investigated with regression analysis. The result showed that plot-scale accumulated ET captured the seasonal trend of plot water use and clearly differentiated the irrigation treatments. Strong linear correlations were observed between plot-scale ET and grain yield, with R2 values ranging from 0.35 to 0.93 (average R2 = 0.71). Plot-scale ET appeared to be a more steady and stronger predictor of grain yield across the seasons than several other morphological and spectral traits including crop height, green pixel fraction, canopy temperature depression, and red-edge normalized difference vegetation index. High spatial and temporal resolution of the field phenotyping data, along with the new analytical framework reported, successfully estimated ET at small plot scale, which is difficult to achieve with other systems or methods. Our work of estimating ET at the plot-scale can be adopt to other ground-based platforms and drones, thus empowers physiologists, breeders, and agronomists for high-throughput phenotyping of water-use related traits and drought response evaluation. • High-throughput phenotyping of plant physiological traits such as ET at the plot-scale is difficult. • We developed a new method to estimate plot-scale ET using high resolution phenotyping data. • The method also used weather data, two-source energy balance, reference ET, and crop coefficient. • Estimated plot-scale ET correlated strongly with corn and soybean grain yield with average R2=0.71. • This study filled an important gap to enhance research in crop water use and drought response. [ABSTRACT FROM AUTHOR]