Alan P. Gay

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.

Computer-vision based measurements of phenotypic variation have implications for crop improvement and food security because they are intrinsically objective. It should be possible therefore to use such approaches to select robust genotypes. However, plants are morphologically complex and identification of meaningful traits from automatically acquired image data is not straightforward. Bespoke algorithms can be designed to capture and/or quantitate specific features but this approach is inflexible and is not generally applicable to a wide range of traits. In this paper, we have used industry-standard computer vision techniques to extract a wide range of features from images of genetically diverse Arabidopsis rosettes growing under non-stimulated conditions, and then used statistical analysis to identify those features that provide good discrimination between ecotypes. This analysis indicates that almost all the observed shape variation can be described by 5 principal components. We describe an easily implemented pipeline including image segmentation, feature extraction and statistical analysis. This pipeline provides a cost-effective and inherently scalable method to parameterise and analyse variation in rosette shape. The acquisition of images does not require any specialised equipment and the computer routines for image processing and data analysis have been implemented using open source software. Source code for data analysis is written using the R package. The equations to calculate image descriptors have been also provided.