Joelle Claussen1, Antje Wolff2, Markus Rehak1, Johannes Wolff2, Bas van Eerdt3, Stefan Gerth1

Using magnetic resonance imaging (MRI), our established root phenotyping platform (van Dusschoten et al., 2016) can visualize and analyze plant roots in natural soil nondestructively (Pflugfelder et al., 2017). Using plant pots with 9 cm diameter and 30cm height, a root system can be scanned within 1h while roots down to diameters of 300µm can be detected and analyzed using our in-house root extraction software NMRooting (van Dusschoten et al., 2016). Thanks to automation with a pick-and-place robot the platform routinely achieves a throughput of 24 plants per day. All these values, however, are based on compromises between imaging speed and quality. In our system, the root detection limit is determined by the signal to noise ratio (SNR) of our images. The SNR can be increased by using smaller plant pots or by increasing the imaging time. In this contribution we investigate the potential gain in the root detection limit when sacrificing plant throughput in favor of image quality. We acquired low noise root images using repeated signal averaging during the measurement process. Using this approach, the root detection limit could be lowered, visualizing roots not detected by the standard imaging protocol.

Proper concentrations of several nutrients, such as iron (Fe), zinc (Zn) and potassium (K) in the soil are needed for plant growth. Thus, farmers sometimes test soils to determine fertiliser application rates. However, measuring soil and plant nutrient concentrations is relatively time-consuming and expensive. Therefore, usually, only a few samples are collected and analysed. This limits the scale of research but also poses some limitations on farming practices, as farmers cannot sample fields at high density in order to customize application rates. Cheap, fast, and high spatial resolution methods to measure soil nutrient concentrations would alleviate some of these limitations. This work aimed to determine the potential of hyperspectral imaging (HI) to predict the concentration of some soil nutrients. Soil samples were scanned by visible and near-infrared imaging systems with a total wavelength range of 450-1700 nm. Fe, Zn, and K were analyzed. Partial least-square regression models (PLSR) were used to correlate the relative reflectance of total Fe, Zn and K in the soil samples. The PLSR models could highly predict Fe concentration (R2 0.81, RMSE% 16.6) and performed moderately well for Zn (R2 0.30, RMSE% 0.9) and K (R2 0.47, RMSE% 5.58) concentrations. The overall results indicated that the hyperspectral technique coupled with PLSR could be an accurate and reliable method for determining soil nutrient concentrations.

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.

In natural ecosystems and low-input agriculture systems often the main source of nitrogen is biological nitrogen fixation by symbiotic coexistence with root colonizing microorganisms such as in root nodules in legumes. In return for this nutrient supply, plants allocate significant amount of photosynthetically fixed carbon (C) belowground, fueling activity and growth of the nodules. However, there is still a lack in understanding how plants modulate carbon allocation to a nodulated root system as a dynamic response to abiotic stimuli. Traditional approaches based on destructive sampling make investigations of localized carbon allocation dynamics difficult. Non-destructive 3D-imaging methods including Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) offers new perspective in analysing belowground processes on individual plants. MRI allows for repetitive measurements and quantification of root system architecture traits nodule structures while growing. PET was employed to follow the spatial distribution of leaf-supplied 11 C tracer to nodules and roots. Using Pisum sativum as model for legumes and applying nitrate as an additional N source we investigated short term C allocation dynamics in the root system. We found that the fraction of 11 C tracer arriving in the most active nodules decreased by almost 40% and remained stable between 16h and 42h after the N application. Our results highlight that the combination of MRI-PET enables deeper insights into short term C dynamics of roots and interactions with colonizing microbes. We expect that this modality has high potential for revealing mechanisms that relate to dynamic fitness traits supporting breedingprograms for future crops.

Roots are the "hidden half" of the plants, which play an essential role in plant development by drawing water and nutrients from the soil and providing anchorage to the plants. Global warming alters soil properties that limit root development and produces drought-like conditions that reduce water and nutrient uptake by the plant roots, ultimately affecting plant growth. It is, therefore, necessary to study Root System Architecture (RSA) traits and breed plants with better root architecture capable of resisting these conditions. RSA comprises various root traits including total root length, root angle, number of lateral roots, and many other traits. Due to the challenges in phenotyping these traits, the research on RSA is lacking behind. In this project, we are studying RSA traits in 500 late-maturity soybean accessions based on germinating the seeds on germination papers, transferring them to blue blotting papers, and imaging the roots 21 days after transfer. Image analysis for the root traits will be carried out using RootNav2.0 software. These phenotypic data and SNPs selected from Illumina Infinium SoySNP50K iselect SNP Beadchip will be used to perform Genome-wide Association Studies (GWAS) using Genome Association and Prediction Integrated Tool (GAPIT) package in R to identify genomic loci associated with each trait. The results from this study will help in discovery of new QTLS or genes that play a key role in root system architecture and help in breeding novel cultivars that are resistance to changing climate.

This study evaluates a hyperspectral imaging (HSI) technique to identify herbicide-resistant kochia (Bassia scoparia) biotypes to support weed management in cropping systems. The experiment was conducted under controlled-environment where glyphosate was applied to six different kochia populations. For each population (72 cell tray of plants), half of the plants were sprayed with Glyphosate 900 g ae ha-1 , while the other half remained an untreated control. Hyperspectral images were acquired over five time points spanning from glyphosate treatment to 15 days after treatment (DAT) using a proximal HSI system (Specim-IQ) with 204 spectral bands from 397nm to 1003nm. The average reflectances were extracted from plants that were characterized as glyphosate-resistant or-susceptible. We first analyzed the temporal variations of the spectra with and without the application of herbicide. The spectral profile exploits the advantages of temporal features in biotype discrimination. Random forest algorithms were used to classify the glyphosate-resistant and-susceptible populations, by using reflectance at optimal wavelengths (near-infrared) and various vegetation indices with high correlations with visual ratings. Based on the classification accuracy, the most important wavebands and vegetation indices were determined to classify the weed biotypes. Preliminary results show that:  1) For the untreated plants, the reflectance at red-edge to near-infrared reached the highest level on 8 DAT, revealing the highest chlorophyll content in the leaves. Then, the reflectance declined until 15 DAT.  2) In contrast, strong effects of glyphosate were captured on 8 DAT for the three herbicide-susceptible populations. For the three glyphosate-resistant populations, reflectance at red-edge to near-infrared did not increase from 1 to 8 DAT, which was opposite of the controlled plants.

ORCiD: [https://orcid.org/0000-0001-6665-6094] Plant phenotyping has been an essential aspect of crop science analytics that is saddled with tasks such as providing critical information about plants' genetics, traits, productivity, and other intricate details to gain insights about their survival under certain conditions and in a specific environment for various analyses. Various methods have been quantifying this information using various models culled from several kinds of datasets. In this study, we extract various phenotypic information about the soybean using UAS-based images captured over the growing fields within the selected experiment field. The DJI M300 unmanned aerial systems were equipped with the Zenmuse P1 and L1 sensors; both used to capture RGB and LiDAR images. In addition, the DJI P4 multispectral UAS was also used to collect multispectral information over these fields at various date intervals. The data captured is being processed using custom-developed algorithms and automated workflows to obtain biomass, vegetation indices, canopy cover, canopy height, and canopy volume. These indices would show variations in the traits of the crop under study as related to the soybean. This phenotypic information would be compared against the field measurements for validation.

Volume is an important phenotype and quality attribute of sweetpotato storage roots. Conventionally the volume of most agricultural products is measured by water displacement. This method, which requires submerging the products in a container of water and measuring the displacement of water in the container, is time-consuming and tedious. It would be beneficial for sweetpotato breeding programs and quality inspection if a rapid method is developed for measuring the volume of sweetpotatoes. This study is therefore to evaluate the feasibility of LiDAR (light detection and ranging) technology as a novel high-throughput approach to phenotyping and measurement of the volume of sweetpotatoes. LiDAR data will be acquired from sweetpotato storage roots using a consumer-grade sensor, Intel® RealSense™ L515, which is an RGB-D (red-green-blue-depth) camera. Ground-truth volume values will be obtained using the reference water displacement method. RGB images will be used to segment sweetpotatoes from background, and extract meaningful features (e.g., the major axis length and the center of mass), complement the point cloud data from depth images for volume estimation. The shape of the sweetpotatoes will be constructed by a series of three-dimensional coordinate points, the alpha shape method is to be used to envelop the boundary points of sweetpotatoes to obtain a non-convex body, and thereby the volume of the sweet potato will be calculated. The efficacy of the proposed method will be evaluated in terms of volume estimation accuracy.

Smart glasses are a rapidly emerging mobile data platform, which can be operated in a hands-free manner through voice commands, a heads-up display and a range of sensors and other digital features. As such, smart glasses enable crop scientists, horticulturalists and agronomists to capture, send and receive digital information, while leaving their hands free to carry out accompanying hands-on tasks or plant manipulations. Phenotypic data increasingly drives agricultural and horticultural development and breeding pipeline discovery. Real-world use cases from innovative agriculture and horticulture technology companies, such as Bayer Crop Science, demonstrate how smart glasses are: 1. serving as a digital phenotyping platform that complements established phenotyping platforms; 2. significantly increases efficiency in phenotypic data collection; 3. facilitate remote collaborations on experiments and other agronomic activities. Smart glass technology integrates easily into existing apps extending capabilities and workflows.

Tillers are shoots that arise from the base of a plant. When plants tiller, they place more carbon resources into vegetative growth as opposed to their grains. Understanding the environmental and genetic factors behind tillering is hampered by lack of high-throughput phenotyping for determining plant tiller count and angle. Currently, plant tiller counts are determined through manual inspection, which is laborious and low-throughput. In this study, we introduce a PlantCV (https://plantcv.danforthcenter.org/)-based algorithm for detecting tillers. This method uses OpenCV's line detection algorithm to detect lines that correspond to the tillers of the plant. From this, tiller count and angle of growth can be inferred. We use this method on Sorghum bicolor accessions from the TERRA-REF project that were grown for two weeks, cut back, and then allowed to regrow for two weeks. Of 200 randomly chosen images, this algorithm was able to accurately count within 1 tiller of the true number of tillers for 165 images. Furthermore, we find that these Sorghum bicolor accessions appear to place less resources into their tillers in the regrowth phase.

Minirhizotron imagery can be used to assess plant root health, and the amount of data for analysis motivates automation of root detection through use of neural networks. Building upon previous work, we show that we can use transfer learning from our PRMI dataset to assess root health across twelve classes of a new dataset to answer questions regarding how root health is affected by access of a tree by large herbivores, site infestation by Pheidole megacephala, and location of the tree. This dataset was collected from three paired sites at the Ol Pejeta Conservancy in Laikipia, Kenya, and consists of 20,000 images collected between September 2021 and May 2022. Each paired site represents four locations based on all four possible combinations of site infestation by Pheidole megacephala for at least 20 years and existence of herbivore-exclusion fence to keep large herbivores out. 1,332 images across all twelve classes of site and treatment combination were labeled with respective ground truths for model training. Our work uses the UNet architecture using pretrained weights on the network encoder and decoder which were obtained in 2019 in work which achieved over 99% accuracy on a dataset of peanut and switchgrass imagery. In our work, we found that training the model with our new dataset resulted in consistent performance across all classes of our new dataset, with over 99% accuracy for each class.

Reducing atmospheric CO 2 is essential to mitigating the effects of climate change. Marine macroalgae has garnered interest as a potential carbon sink to sequester atmospheric CO 2. Efforts to improve macroalgae for use as a carbon sink are just beginning, and understanding the phenotypic diversity of macroalgae is a critical part of identifying favorable characteristics for a marine carbon sink. In this study, we surveyed the natural variation of the marine macroalgae, Saccharina latissima (sugar kelp), from eight sites along the western coast of Iceland: three in Breiđafjörđur, four in Faxaflói, and one in the Atlantic Ocean south of Reykjanes. All collected sugar kelp samples were in the non-reproductive stage, and environmental data was collected at each site. All collected sugar kelp samples were phenotyped manually and through image analysis. Chlorophyll content, carbon and nitrogen content, and relative carbohydrate contents were also quantified. Strong correlations were observed between many morphological phenotypes and with environmental variables such as air temperature. While phenotypic differences between sampling locations were observed in some individual traits like blade length and chlorophyll content, principal component analysis did not reveal significant grouping of individuals by sampling location based on phenotype. These results suggest that phenotypic variation in sugar kelp populations across the Western Icelandic range is not strongly influenced by changes in environmental variation. These results are consistent with natural variation studies conducted in sugar kelp samples from New England, United States and indicate that sugar kelp can be a favorable choice for carbon sequestration across multiple environments.

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.

Urban agriculture has been broadly acknowledged for its potential to reduce carbon emissions, increase food security, and improve economic growth in some of the most vulnerable communities in the United States. Collard (B.oleracea var. viridis) is a diploid leafy green, grown on urban farms and community gardens across the country, including the St. Louis Metro region. Beyond their nutritional importance, collards provide urban and commercial agronomic systems with a plethora of important ecosystem services. They scavenge nitrogen and available resources, suppress weeds, and act as a biofumigant to control soil-borne pests and pathogens. Recently, The Heirloom Collard Project characterized the above-ground growth habits of 18 landrace collard varieties across 250 organic gardens and farms. Little work has been published to investigate collard root system architecture, which influences both quality traits and ecosystem services that contribute to sustainable crop production. The objectives of this research are to 1) quantify root spatial and temporal diversity across 18 landrace collard varieties, and 2) evaluate the relationship between root phenotype and urban farmer crowd-sourced data for key traits such as germination rate, disease resistance, vigor, yield, flavor, and winter hardiness. This work will lead to the development of a participatory framework for urban farmers and chefs to select varieties with improved root architecture based on regional needs.

X-ray tomography (XRT) is a powerful and versatile tool for generating detailed non-destructive three-dimensional (3D) image data of large and complicated structures. In particular, excavated, cleaned and dried maize root crowns can be rapidly scanned, and the resulting 3D volumes processed in a computational feature extraction pipeline to provide a wide range of root trait measurements. These measurements provide rich data that give insights into how roots occupy 3D space in ways not possible with any 2D imaging/measurement systems. Hundreds of root crowns can be scanned in a moderate-throughput system, and multivariate statistical analyses can provide valuable insight into the role that genes and quantitative trait loci play in selected root traits. Research presented will describe XRT scan parameter optimization and its impact on root trait data generated by the feature extraction pipeline.

Holistic assessment of fruit quality is an essential component of producing Strawberry varieties that will succeed in the marketplace and improve consumer satisfaction. However, several key quantitative traits are notoriously slow and expensive to assess using standard procedures, namely acidity and aroma, which require titration and gas chromatography and mass spectroscopy compared to others: brix, anthocyanins, and vitamin C, which are measured by refractometer and parallelized plate reader assays. Scaling up evaluations for acidity and aroma has been difficult as the techniques require 5 and 40 mins/sample, respectively, and sample preparation is equally intense, requiring multiple trained hands working for 10-hour sessions to create the sample series for 100 entries. We evaluated the ability (R 2 , RMSE) of a handheld near infrared (NIR) spectrometer, measuring 125 wavelengths between 800 and 1600 nm, and an electronic nose, measuring the reaction of 32 electrochemical sensors that respond to various compounds in gas samples, on 4,000 diverse strawberry accessions to determine if the 5 and 40 min/sample assays can be replaced with a 1 (0.33%) sec/sample (NIR) and 2 (5%) min/sample (E-nose) assay that require no additional sample prep. We also assess the NIR's ability to predict brix, anthocyanins, and vitamin C. With these two sensors, we will be able to increase the scale of early generation evaluation from hundreds to thousands of samples in early generations, produce full datasets prior to deadlines in the breeding program, and make more reliable genetic gains for quality traits affecting marketability and consumer acceptance.

Michelle Cho1, Zhengbin Liu2, Dhineshkumar Thirupatthi3, Tim Parker3, Shayla Gunn4,