Replacing fossil fuels with cellulosic biofuels is a valuable component of reducing the drivers of climate change. This leads to a requirement to develop more productive bioenergy crops, such as Arundo donax with the aim of increasing above-ground biomass (AGB). However, direct measurement of AGB is time consuming, destructive, and labor-intensive. Phenotyping of plant height and biomass production is a bottleneck in genomics- and phenomics-assisted breeding. Here, an unmanned aerial vehicle (UAV) for remote sensing equipped with light detection and ranging (LiDAR) was tested for remote plant height and biomass determination in A. donax. Experiments were conducted on three A. donax ecotypes grown in well-watered and moderate drought stress conditions. A novel UAV-LiDAR data collection and processing workflow produced a dense three-dimensional (3D) point cloud for crop height estimation through a normalized digital surface model (DSM) that acts as a crop height model (CHM). Manual measurements of crop height and biomass were taken in parallel and compared to LiDAR CHM estimates. Stepwise multiple regression was used to estimate biomass. Analysis of variance (ANOVA) tests and pairwise comparisons were used to determine differences between ecotypes and drought stress treatments. We found a significant relationship between the sensor readings and manually measured crop height and biomass, with determination coefficients of 0.73 and 0.71 for height and biomass, respectively. Differences in crop heights were detected more precisely from LiDAR estimates than from manual measurement. Crop biomass differences were also more evident in LiDAR estimates, suggesting differences in ecotypes’ productivity and tolerance to drought. Based on these results, application of the presented UAV-LiDAR workflow will provide new opportunities in assessing bioenergy crop morpho-physiological traits and in delivering improved genotypes for biorefining.
Summary
High-throughput image-based technologies are now widely used in the rapidly developing field of digital phenomics and are generating ever-increasing amounts and diversity of data. Artificial intelligence (AI) is becoming a game changer in turning the vast seas of data into valuable predictions and insights. However, this requires specialized programming skills and an in-depth understanding of machine learning, deep learning, and ensemble learning algorithms. Here, we attempt to methodically review the usage of different tools, technologies, and services available to the phenomics data community and show how they can be applied to selected problems in explainable AI-based image analysis. This tutorial provides practical and useful resources for novices and experts to harness the potential of the phenomic data in explainable AI-led breeding programs.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.