The structures of roots play an essential role in plant growth, development, and stress responses. Minirhizotron imaging is one of the widely used approaches to capture and analyze root systems. After segmenting minirhizotron images, every individual root is separated from each other and the background. Root traits, like root lengths and diameter distributions, can provide information about the health of the plants. Current methods to analyze minirhizotron images usually rely on manually annotated labels and commercial software tools, which are time and labor-consuming. Unfortunately, these current methods usually generate a statistical analysis of the input image rather than the features of each root. In this work, we propose a pipeline to automatically use deep neural networks to segment roots from the background and then extract root features like lengths and diameter distributions from the individual segmented root. In detail, we first use a pre-trained U-Net to segment the roots in the minirhizotron images. Then, we separate each individual root with the help of connected component analysis. Finally, we extract the features like diameter distribution or root lengths of every individual root with morphological operations, like skeletonization. For evaluation, we conduct experiments on synthetic roots, which are made of strings and threads, and compare results against a benchmark root dataset (PRMI) of real switchgrass roots and compare the estimated results with the existing commercial software.

An organism's phenome results from expression of its genome (nature) under certain environment and management effects (nurture) and interactions between these factors, as well as measurement error. For over 30 years, DNA sequencing and genomics tools advanced to where it's now feasible to saturate genomes of segregating individuals, such that polymorphisms at nearly any position can be determined from other known positions. This is due to structure, linkage disequilibrium (LD), or linkage and is a powerful tool for genomic prediction and investigating biological phenomena. In contrast, most phenomics to date focuses on automating previously known "traits" as measurable and interpretable phenotypes; akin to focusing on measuring a single DNA marker rather than measuring an entire saturated genome. Viewing phenomics as a platform for discovery, similar to genomics, opens new methods for capturing phenomena in nature and nurture. Saturating a phenome would mean that an individual's fitness, performance, responses to environment and/or specific phenotypes could be accurately predicted in untested environments. To date, our experience with phenomic prediction for cumulative, complex phenotypes such as grain yield suggests it's possible to predict organismal performance in untested environments, possibly better than genomic methods despite less advanced tools and data. Factors limiting to saturating a phenome are evaluating enough individuals and environments, but more importantly, tools and methods to extract or "sequence" more phenomic features. Successfully saturating phenomes will impact every aspect of science and society, in biological disciplines from germplasm curators, physiologists to breeders, to education, the courtroom and policy.

Sara Tirado Tolosa1

Bioluminescence is used as marker e.g., in genetic or plant pathological studies. We developed a method to monitor bioluminescence at the whole plant level combined with phenotypic analysis of the plant. Using a CCD camera mounted in a cabinet shielding all external light we can image weak luminescence emissions from samples and map these to RGB images. Image processing delivers temporal and spatial data on the distribution of the luminescence together with phenotypic features of the plants. With this technology, microbial colonization of plants can be monitored. Arabidopsis plants were inoculated with Pseudomonas and Xanthomonas plant-pathogenic bacteria labelled with gene cassette for autonomous luminescence and disease progression was monitored over time. Luminescence imaging revealed accumulation of the bacteria in different plant tissues while the RGB images served to monitor plant growth and occurrence of disease symptoms. Applying this method, resistant plants could be selected from a mutant population. Disease responses of susceptible plants were compared to the responses of resistant plants. In the case of Pseudomonas, bacterial abundance reached its maximum during two to four days after inoculation, at a time when water soaking of the leaves could be observed as well with the RGB camera. At later stages-five to seven days after inoculation, disease symptoms in terms of leaf yellowing and tissue collapse occurred while bacterial populations appeared to decrease. With this method it was possible to monitor pathogen development and disease progression non-invasively at whole-plant level over time.

Field Based High Throughput Phenotyping Enables the Discovery of Loci Linked to Senescence and Grain Filling Period ORCiD: [Alper Adak; 0000-0002-2737-8041] Keywords: Grain filling period, field-based high throughput phenotyping, days to senescence, temporal phenotype. Senescence occurs progressively over time and is variable among different genotypes. To examine the temporal and genetic variation of senescence, 280 maize hybrids and 520 maize recombinant inbred lines (RILs) grown in 2017 and 2018 were investigated. Hybrids were grown in late and optimal planting trials; RILs were grown in irrigated and non-irrigated trials, both based on range-row design with two replications. Two types of Unmanned aerial systems (UAS, also known as UAV or drones) were flown over the germplasm between 14 and 20 times respectively. Temporal senescence of each row-plot in hybrids and RILs was scored visually according to percentile scoring using four to five rectified drone images between ~90 and ~130 days after planting. A mechanistic growth model was fit to each genotype using the temporal senescence scores, resulting in 0.94 and 0.97 R 2 for hybrids and RILs. Days to senescence (DTSE) and grain filling period (GFP) were calculated for each row plot using the developed mechanistic growth model. To predict the genotypic value for each RIL and hybrid, a mixed model with three-way interaction model (Genotype*Flight*Environment) was then run. Correlation was calculated ~0.84 and ~0.88 between grain yield and GFP and DTSE in hybrids. A major quantitative trait locus was also discovered on chromosome 1 (295.5 to 296.8 kb; 15% explained) linked to GFP in RILs. GFP is known to be physiologically important, UAS provided an easily scalable measure which can greatly increase the evaluation of variation in the field.

The research introduces a novel algorithm called HyperStressPropagateNet that uses deep neural network based time series modeling to illustrate the qualitative and quantitative propagation of drought stress in a plant using hyperspectral imagery. The hyperspectral cameras typically capture a broad range of wavelengths at very narrow intervals of a few nanometers creating a hyperspectral cube. HyperStressPropagateNet uses spectral band difference-based segmentation method to create the binary mask of the plant which is then used to segment the plant in all bands of a hyperspectral cube to create the reflectance spectra at each plant pixel. The algorithm uses convolutional neural networks to classify the reflectance spectra generated at each pixel into either stressed or unstressed categories to determine the temporal propagation of stress. The limited water availability in the soil is confirmed by changes in the soil water content (SWC) measured using a hand-held device. The excellent correlation between the SWC and the corresponding temporal progression of percentage of stress pixels computed by HyperStressPropagateNet demonstrates the efficacy of the method. The algorithm is evaluated on a dataset of image sequences of cotton plants captured by the hyperspectral camera in the LemnaTec Scanalyzer 3D High Throughput Plant Phenotyping Platform in the University of Nebraska-Lincoln, USA. The excellent performance of the method is established based on evaluations using various metrics, e.g., confusion matrix, precision-recall curve, and F1-score. The method has the potential to be generalized to any plant species to study the effect of abiotic stresses on sustainable agriculture.

Imaging of plants using multi-camera arrays in high-density growth environments is a strategy for affordable high-throughput phenotyping. In multi-camera systems, simultaneous imaging of hundreds to thousands of plants eliminates the time delay in measurements between plants seen in plant-to-camera or camera-to-plant systems, which allows for the analysis of plant growth, development, and environmental responses at a high temporal resolution. On the other hand, high plant density, camera-to-camera variation, and other trade-offs increase the complexity of data analysis. Here we present two recent updates to the PlantCV image analysis package to improve usability when working with multi-plant datasets. First, we introduce a method to automate detection of plants organized in a grid layout, reducing the need to make separate workflows for each camera in a multi-camera system. Second, we reduced the number of input and output parameters for functions handling the shape and location of plants and introduce automatic iteration over multiple objects of interest (e.g. plants), reducing the level of programming needed to build workflows.

Automated monitoring and evaluation systems for plant phenotyping are one of the keys to advance and strengthen crop breeding programs. In this study, the improvements of the camera-based sensor system and a weather station from a previous study-assembled mainly from Raspberry Pi products-board with dual cameras (RGB and NoIR) providing high spatial and temporal resolution data-is outlined. Hardware for the internet connection and the power supply system of the sensor were upgraded. Previously, the sensor could automatically capture plant images following user-defined time points; thus, an image processing algorithm (edge computing) was developed and installed to extract digital phenotypic traits from the images after capturing process. With the development, the new sensor system could be integrated with the internet, and a cloud server was configured to store data online (digital traits and raw images). A real-time monitoring system was created to visualize the time series data of a trait development and plant images throughout the season. With such a system, plant breeders will be able to monitor multiple trials for timely crop management and decision-making process, which is also resources efficiency.

Cover crops, plants grown during fallow periods between cash crops, are a promising solution to mitigating soil degradation induced by conventional agricultural practices and improving soil health. Cover crops can provide several beneficial ecosystem functions, such as soil structure remediation, soil microbial diversification, and nutrient recycling, depending on the plant species. Interactions between plant roots and the surrounding soil are key to the plant's ability to perform their ecosystem functions. The lack of data on cover crop roots inhibits our understanding of cover crop phenotype-ecosystem function relationships. We combine aboveground and belowground phenotyping measurements with physicochemical soil measurements to evaluate the field performance of 19 different plant species in monocultures and polycultures as winter cover crops in Missouri. Canopy cover imaging reveals significant differences in winter hardiness and weed suppression among cover crop varieties. Root biomass and root length density measured at depths up to 1 meter indicate differences in rooting behavior between cultivars suggesting the ability to breed cover crop varieties with improved root system architecture. I will also highlight our collaborative efforts utilizing remote sensing technologies (aerial RGB and hyperspectral imaging) to model carbon and nitrogen cycling in cover crop systems at a field scale. Finally, we have begun to characterize 3D root system architecture traits at the seedling stage using a gel-imaging system. Better understanding of cover crop rooting behavior will allow us to breed varieties with enhanced performance of beneficial ecosystem functions for sustainable agricultural systems.

The application of herbicides in agriculture has significantly increased in recent decades. While many herbicides improve the efficiency and efficacy of weed control, their excessive use at the wrong growth stage can cause crop foliar damage, higher input cost and negative environmental footprints. There are limited techniques to accurately monitor herbicide effects. Visual ratings are highly subjective and require extensive training or experience. High-throughput digital imaging is a promising tool to measure plant herbicide interaction in field crops. In this study, proximal and aerial based advanced sensors have been utilized to evaluate different herbicide modes-of-action in two model species, tame oat [Avena sativa; model for wild oat (Avena fatua)] and oriental mustard [Brassica juncea; model for wild mustard (Sinapis arvensis)]. The experimental trials were performed at three agro-climatic locations in Canada (Lethbridge (AB), Saskatoon (SK), and Lacombe (AB)). The proximal and UAV multispectral imagery data were collected for baseline (before treatment) and 1, 3, 7, 10, 14 and 21 days after treatments (DAT), alongside visual ratings. The Normalized Difference Vegetation Index (NDVI), Photochemical Reflectance Index, Chlorophyll Vegetation Index, and Optimized Soil Adjusted Vegetation Index were used to assess variation of different DAT pigment content (photosynthetic rate) and chlorosis (damage %) in plot vegetation. The variation in obtained temporal indices (NDVI) suggest that the developed technology has potential to replace visual ratings (R2 ≈0.65-0.94) and can be used as a rapid screening tool for herbicide activity. Therefore, remote sensing tools could improve the precision and consistency of future herbicide assessments.

The true bottleneck of artificial intelligence (AI) is not access to the data, but rather labeling this data. We have tons of raw agriculture image data coming from various sources and manual labelling remains to be a crucial step to keep the data well organized which requires considerable amount of time, money, and labor. This process can be made more efficient if we can automatically label the raw data. We propose contrastive learning representations for agriculture images (AgCLR) model that uses self-supervised representation learning approach on unlabeled real-world agriculture field data, to learn the useful image feature representations from the images. Contrastive learning is a self-supervised approach that enables model to learn attributes by contrasting samples against each other without the use of labels. AgCLR leverages the state-of-the-art SimCLRv2 framework to learn representations by maximizing the agreement between differently augmented views of same sample. We have incorporated critical enablers like mixed precision, multi-GPU distributed parallel computing, and use of Google Cloud's Tensor Processing Units (TPU) for optimizing the training process. We achieved 80.2% accuracy while classifying the test data. We further applied AgCLR to unrelated task to determine the alleys and rows in corn field videos for corn phenotyping and we observed two cluster formations for alleys and rows when plotted embeddings in a 3-dimensional space. We also developed a content-based image retrieval tool (pixel affinity) to identify similar images in our database and results were visually very promising.

The majority of domesticated plant species are herbaceous annuals and woody perennials, yet many herbaceous perennial species hold potential for future agricultural systems. In addition to multiyear harvests, herbaceous perennials provide many ecosystem services, including erosion control as a result of their large and persistent root systems. However, the multiyear lifespan of perennial species has been a barrier to rapid domestication as breeding cycles require phenotyping over multiple growing seasons. Using phenomic selection, high-dimensional secondary traits measured on seedlings could be used to develop relationship matrices among individuals which are then used to predict field traits. Additionally, these models can serve as the selection criteria to identify individuals to advance to the next (pre)breeding generation, thus shortening the breeding cycle. This project substitutes costly genomics data with high-dimensional phenomics data asking: Can elite individuals of perennial species be predicted by phenomic relatedness models based on high-dimensional traits recorded on seedlings? To date, we have imaged 2280 seedlings from each of the following three perennial crop candidate species: intermediate wheatgrass (Thinopyrum intermedium), sainfoin (Onobrychis viciifolia), and silphium (Silphium integrifolium) on the Bellwether Foundation Phenotyping Facility housed at the Danforth Center. The images were processed using PlantCV to generate high-dimensional color and near-infrared profiles for each plant on each image day. Additionally, profiles were generated with handheld spectrometers. This work re-imagines innovations in plant traits, kinship matrices, genomic selection, phenotyping centers, and ultimately domestication, in order to expedite the development of an emerging generation of climate resilient, ecologically sustainable crops.

A dry beans (Phaseolus vulgaris L.) cultivar must fit the environment in which it will be grown. Therefore, days to maturity (DM) is the most important physiological component affecting yield and grain quality outcomes. Additionally, dry bean stand count (SC) at early growth stages estimation provides useful information for agronomic decision-making and can measure root rot loss due to damping-off. The visual inspection to determine the accurate maturity date and the final number of emerged plants is labor-intensive, time-demanding, and tedious. Therefore, there is an increasing demand for alternative approaches to estimating DM and SC in a high-throughput phenotyping mode (HTP). In this study, we developed a Deep Learning (DL) HTP pipeline to capture the sequential behavior of time series data for estimating DM and to identify target plants in the early growth stage for SC estimation using field dry bean data obtained from aerial RGB images at the plot-level. A state-of-the-art hybrid model combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) was used to extract DM features and capture the sequential behavior of time series data. Faster R-CNN object detection method was deployed to SC. The DL model to estimate DM was tested on five different environments across years, and SC was done comparing different ground sample resolutions in two trials. Results suggest the effectiveness of the CNN-LSTM and Faster R-CNN models employed compared to traditional methods. Furthermore, this study highlighted the technical parameters that can influence the DL model results in the breeding program decision-making.

Most current phenotype plant research focuses primarily on above-ground traits, like leaves and flowers. Roots often get comparatively less attention because they are challenging to examine and image. Minirhizotron (MR) systems are one of the imaging approaches to studying plant roots underground. In MR systems, a tube is inserted into the ground to allow a camera to be inserted to capture the images of root systems. Unlike minirhizotron imaging, X-ray computed tomography (CT) captures the three-dimensional (3D) information of soil cores extracted from the soil. For a better analysis of roots, the first step is always to segment the roots from the background in the images or image sequences. The results of root segmentation play an essential role in further analysis like root diameter and length estimation. Current fully-supervised segmentation methods mainly use pixel/point-level annotated labels, which require much manual effort and time. In this work, we propose a weakly supervised root segmentation approach with graph convolutional networks. Our model only requires image-level annotations to segment roots from the images or image sequences. In detail, our model first constructs graphs for the neighboring pixels/points and then learns the distinguishable features used as hints for segmentation by training a classifier based on the image-level annotations. Finally, post-processing procedures like principal component analysis (PCA) are applied to refine the final segmentation results. We conduct experiments on the challenging 2D PRMI minirhizotron benchmark and 3D switchgrass root X-ray CT datasets for evaluation.

High-oil tobacco varieties have been recently engineered to produce increased leaf oil content for future food and fuel needs. An engineered variety of Nicotiana tabacum produces ~30 percent of leaf dry weight in lipids in the form of triacylglycerol (TAG), a significant increase relative to the less than 1 percent storage oil normally found in wild-type leaves. This high-oil tobacco also accumulates oil bodies in stomatal guard cells. In order to understand the impact of oil on guard cell shape, aperture, and dynamics, we have co-opted computer vision tools in PlantCV to create an accurate, flexible, and high-throughput method for microscopy image analysis of stomata. To this end, leaf impressions are made with silicone putty; clear nail polish peels of the putty impressions are imaged using light microscopy. Binary thresholding followed by point-and-click regions of interest and morphology calculations provide stomatal counts, aperture, and other shape characteristics. Applying this method to high-oil tobacco demonstrated reduced stomatal aperture but the same number of stomata per unit leaf area, providing a mechanistic explanation of high-oil tobacco responses to high temperature and water deficit stresses.

Predictive biology is the ability to predict a biological outcome from known inputs (and vice versa). Complex and urgent problems, such as climate change and a growing global population, requires a better grasp of predictive biology approaches. However, predictive biology requires a deep understanding of the genome-to-phenome relationship within an organism. The field of genomics has accelerated rapidly in the last few decades with technological advances that have helped reduced the costs of genomics research and easy-to-use computational tools. Phenomics technologies has not advanced at the same rate. Many phenotyping systems are expensive to own, require special training to use, focus on narrow areas of research, and produce large amounts of data with no standards of storage and documentation. We propose a design philosophy for high-throughput phenotyping systems called the OPEN Series. This philosophy focuses on systems that use off-the-shelf commercial products and open-source software to make high quality phenotyping systems efficiently and for use by general users. In addition, the OPEN Series focuses on integrating cloud-based image processing through the NSF-funded cyberinfrastructure CyVerse, thus allowing users to share and process data remotely. We've worked to integrate this philosophy into our own phenotyping systems, OPEN leaf and OPEN root, to great success. We hope to export our work in creating accessible and affordable phenotyping system to labs across the globe to accelerate our understanding of the genome-to-phenome relationship for predictive biology.

A document by Mao Li. Click on the document to view its contents.

Hyperspectral imaging is a promising method to predict traits in a high-throughput manner with the potential to unlock quantitative genetic studies. Researchers have successfully modeled physiological traits such as vegetative Nitrogen content, but scope of methodology and lack of truly novel testing data hinder large scale trust in the process. Here, I explore the ability to model leaf Nitrogen content from hyperspectral reflectance data collected with a LeafSpec imaging device on 22 maize hybrids. Three broad strategies based on different input feature sets are undertaken. Strategy one mines data for the most informative hyperspectral channels and then constructs a normalized index similar to NDVI as input features. Strategy two considers all 364 channels of hyperspectral data and makes predictions using various machine learning techniques; partial least squares regression(PLSR), random forest regression, and a feed-forward neural net regression. Strategy three aims to take advantage of the spatial distribution of hyperspectral data on the leaf surface by training a convolutional neural net(CNN). A normalized visual index constructed from bands most correlated with nutrient content out-performed established NDVI. PLSR was the most accurate algorithm, followed by feed-forward neural net and then CNN, based on coefficient of determination score. PLSR is well established as a robust method for hyperspectral prediction which is further evidenced by this study. This is one of the first applications of CNN for hyperspectral data. Despite not being the most accurate algorithm there remains room for hyper-parameter optimization.