Plant phenotyping plays a crucial part in the development of new crop genotypes. In this study, the applicability of relatively simple commercially available digital phenotyping devices was tested and improved in the context of wheat variety testing. Aerial thermography is used to evaluate the performance of genotypes by measuring canopy temperature (CT). Because lightweight thermal cameras for drones are prone to significant thermal drift effects due to a lack of a signal stabilizing cooling, we propose a new approach to analyze drone based thermal images. Through the inclusion of covariates such as trigger timing and the position of the drone relative to measured plots, temporal trends and viewing-geometry related effects could be mitigated, which improved the CT measurements. Correlations between measurements on 270 experimental wheat plots taken within 20 min were very strong (R = 0.99) and highly genotype specific with generalized heritabilities > 0.95 in many cases. In a second experiment, autonomous PhenoCams mounted on poles 12 m above the field were evaluated for their suitability to track main phenological stages and senescence as a replacement for time consuming manual field scorings. Senescence and maturity of wheat could be tracked reliably in the field for three subsequent seasons with strong correlations between field-scorings and image-based estimates (R > 0.8). For emergence and heading, achieved correlations were poor. Both experiments demonstrated how image-based phenotyping with a comparably simple setup can be used to derive high quality data relevant in the evaluation of the performance of wheat genotypes in the field.

Maize is pivotal in supporting global agriculture and addressing food security challenges. To better understand the genetic factors that underpin maize growth, quantitative phenotyping of traits is essential. Root systems are challenging to phenotype given their below-ground, soil-bound nature. In addition, manual annotations of root images are tedious and can lead to inaccuracies and inconsistencies between individuals, resulting in data discrepancies. To address these issues we have developed an automated phenotyping pipeline utilizing Root Painter , Rhizovision , and R for maize root image analysis and efficient extraction of phenotypic data. This pipeline was tested on images of field-grown maize crown root systems (stages V6-V8) from the Wisconsin Diversity panel. By minimizing user input and increasing automation, these tools improve the consistency and accuracy of data metrics. Root Painter, a segmentation application based on U-Net with a user-friendly interface, specializes in identifying roots and nodes. 123 images were annotated in RootPainter's interface for training. Resulting in precise differentiation between roots and non-root structures, enabling unsupervised crown root phenotyping. Finally, these segmented images were subsequently processed using Rhizovision's batch image processor, extracting numerous key root traits, including total root length, network area, and volume. The output from Rhizovision was then analyzed using an R script, incorporating statistical and visualization packages. Comparing the results obtained from our automated phenotyping pipeline with manually measured root systems demonstrated increased accuracy and consistency across researchers. This integrated pipeline saves user time and reduces costs by harnessing open-source maize phenotyping software and robust data analysis techniques.

Breeding for improved, reliable cultivars despite growing environment irregularity can be challenging. Unmanned aircraft systems (UAS) are a popular high-throughput phenotyping technology that has been shown to help understand the mechanisms associated with crop productivity and create a potential for improved breeding strategy by providing unique insight into environmental response and cultivar productivity. Spectral reflectance indices (SRI), including both vegetation and water indices like NDVI, NDRE, and NWI were used to evaluate 11,593 Washington State University winter wheat breeding plots between 2019 and 2022. SRIs were then used with genomic data in univariate and multivariate gBLUP model predictions for grain yield. Prediction accuracy was evaluated on a leave-one-year-out validation strategy. Including SRI data as fixed effects in univariate genomic prediction models can improve prediction accuracy over the control but is unreliable across years. When used in multivariate models, SRIs improve prediction performance across years but minimally when considering the computationally more efficient base model. In univariate models, when test year NDVI data was used to calculate breeding values, prediction performance was at least 16% better than the control, ranging in prediction accuracy from 0.54 in 2019 to 0.93 in 2020. This study highlights the limitations of SRI and its use in genomic selection, especially when dealing with large breeding populations across environmental extremes. A significant application for the technology can be found in early season UAS data collection to aid accurate predictions in late season, a helpful tool in tight turnaround times commonly experienced in winter crop breeding programs.

Cold tolerance is of paramount importance for the survival of winter wheat, as failure to withstand low winter temperatures necessitates costly replanting during the spring season. Initial observations indicate that higher cold tolerance is associated with increased winter dormancy, resulting in slower spring growth and potentially lower grain yields. Understanding how quickly the plants grow during this critical period can provide insights into their overall health and yield potential. This study investigates the interaction between cold tolerance, winter dormancy, rate of spring growth, and final grain yield to develop cultivars with high cold tolerance and early spring growth. The plant material consists of 480 diverse soft white winter wheat varieties collected from Pacific Northwest (PNW) breeding programs. The population was cultivated for three years (2016, 2017, and 2018) at two PNW locations (Pullman, WA and Pendelton, OR). RGB and NIR images were captured weekly to generate growth curves and calculate NDVI values for each genotype. Additional data on tiller count, heading date, plant height, grain volume weight, and grain yield were collected. Modeling NDVI regression curves enabled us to quantitatively define plant growth attributes including spring growth rate, time of peak greenness, dry-down rate, time of total dry-down, time from heading to total dry-down. These data, correlated with final grain yield, will enable a better understanding of winter wheat growth patterns. This information is crucial for breeding regionally specialized winter wheat varieties that can maximize spring growth and yield potential while minimizing the risk of cold damage.

This study aimed to investigate the physiological and morphological responses of maize seedlings to environmental stressors. High-throughput imaging analysis was used to characterize the stress response phenotypes of 47 distinct maize genotypes exposed to water limitation (40% field capacity), heat (38 °C), and the combination of the two stresses. RGB and NIR images were collected daily, and analyzed with open-source, open-development software PlantCV. Our investigation focused on quantitative measurements of daily area, daily water loss, evaluation of estimated water use efficiency (WUE), and near-infrared (NIR) reflectance as an estimation of water content in tissues. Quantitatively comparing two non-normal distributions, like the NIR histogram data, can be challenging but important, since metrics like median mode values often do not capture variation across a sample. One of the primary obstacles lies in defining an appropriate metric to accurately quantify the variation between two distributions. To analyze NIR reflectance, we evaluated the dissimilarity between pairwise NIR histograms by utilizing the earth mover’s distance (EMD) analysis. The EMD quantifies the difference between two NIR reflectance distributions, and therefore indirectly evaluates the difference in leaf water content for stress vs control conditions. Overall, most genotypes displayed growth reduction under drought and heat. Interestingly, specific lines exhibited heightened WUE under water limitation, suggesting response to water scarcity. EMD results showing dissimilarities in pairwise NIR reflectance of control vs drought, can be used to describe variation in dynamic changes in water content levels among these groups.

Malting is the process of controlled germination of cereal grain that is steeped, germinated, and dry kilned to develop flavor compounds, fermentable sugars, the hydrolytic enzymes necessary for the brewing process. To meet the strict standards of the brewing industry, variation in the germination of barley grains need to be minimal but also maintain rapid germination and optimum levels of \(\alpha\) and \(\beta\) amylase. Preharvest sprouting poses a signinicant threat to barley cultivars prior to harvest, resulting in premature endosperm modification, reduced enzyme content, and poor malthouse pertormance. Pregerminated grain is ditticult to detect as there are often no signs of damage, and current methods such as the Hagberg falling/Stirring number, or pearling tests, each of which ultimately leads to the destruction of the seed. Here we used time-series hyperspectral imaging of barley seeds undergoing active and or pregerminated seed and used a deep neural network to predict stirring number and alpha amylase values, which are proxy measures for sprout damage. Prediction models were made for seven genotypes tested individually and with all genotypes combined. Stirring number assays ranged from 0 to 190 and our prediction models had mean average errors of 10.5 to 23.9 depending on the model. To increase the resolution and accuracy of the predictions we transitioned from using bulk conventional molecular assays to single seed assays and prediction models. These results demonstrate that hyperspectral imaging and machine learning models can be used to predict germinated grain in bulk and single seed assays.

Plants store carbon as oil in their seeds, which is a resource that needs to be mobilized in order to germinate. Camelina sativa is a prospective biofuel crop which has the potential to be used as a cover crop in corn and soy growing regions. Camelina plants have been engineered to produce medium-chain fatty acids useful in the production of biofuels, lubricants, and other products. To test suitability of these modified Camelina lines for growth in cool climates such as those found in the midwestern US, we performed germination tests on seeds in growth chambers at temperatures similar to those encountered in early season in corn-growing regions (6°C, 15°C, and 22°C). Controls were the untransformed background Suneson genotype, as well as plants from the far northern and southern ranges of C. sativa. Germinating seedlings were imaged using Raspberry Pis, and analyzed with PlantCV for the number of plants and the cotyledon area. To determine effects of seed fatty acid content later in the growth cycle, seedlings were transplanted and grown at 22°C until seed set. These plants were imaged every 20 days to track height, width, area, and solidity using PlantCV to analyze the data.

Photogrammetry is the science of obtaining a 3D scan of an object. Through this process, reliable information about the physical object's complex structure can be obtained, studied and analysed. A low-cost Structure from Motion (SfM) technique can be used to create 3D models using multiple 2D images from different viewpoints. A point cloud is a widely used 3D data form, which can be produced by depth sensors, such as LIDARs and RGB-D cameras. However, the cost of such scanners can be prohibitive, putting photogrammetry out of reach for many researchers and practitioners in the agriculture industry. We are developing a low-cost close-range photogrammetry rig that could be a beneficial tool for agronomists, plant scientists, and breeders. Our imaging system utilizes the Raspberry Pi to capture images with multiple cameras, and a commercial rotatory table to get images from different viewpoints. We discuss the development of extracting quantitative trait indices in wheat in order to automatically characterize planophile versus erectophile canopy architectures. Moving forward, we plan to use our photogrammetry rig for a variety of applications such as growth monitoring and extracting plant traits such as number of leaves, stem height, leaf length, leaf width, leaf area, and canopy volume. We also plan on developing bespoke, plug-and-play systems that are tailored to the specific needs of a researcher and can be operated with minimal expertise.

Wheat (Triticum aestivum L.) is one of the key staple crops worldwide. Even though future demand for wheat is estimated to increase by 6% by 2050, wheat production might drop by 30% due to climate change. The purpose of this research is to identify canopy architecture (light capture) and anatomical (stomatal) traits that significantly increase radiation use efficiency (RUE; dry weight biomass produced per unit radiation intercepted) and improve yield under high temperature and drought stress conditions due to the growing concern over food security. This research was conducted with five contrasting wheat genotypes in a new field-based high tunnel system with 4 treatments and three replications (control, heat stress, drought stress and heat x drought stress with 3 replications = 12 tunnels) in a randomized completed block design with stress applied at the heading stage. Canopy architecture was graded according to the visual scoring scale concerning UPOV and RUE was calculated in three growth stages. New techniques of high throughput imaging of stomatal number and size using a handheld digital microscope will also be presented.

Walnuts are the second most produced and consumed tree nut, with over 2.6 million metric tons produced in the 2022-23 harvest cycle alone. The United States is the second largest producer, accounting for 25% of the total global supply. Nonetheless, producers face an ever-growing demand in a more uncertain climate landscape, which requires effective and efficient walnut selection and breeding of new cultivars with increased kernel content and easy-to-open shells. Past and current efforts select for these traits using hand-held calipers and eye-based evaluations. Yet there is plenty of morphology that meets the eye but goes unmeasured, such as the volume of inner air or the convexity of the kernel. Here, we study the shape of walnut fruits based on X-ray CT (Computed Tomography) 3D reconstructions. We compute 49 different morphological phenotypes for 1264 individuals comprising 149 accessions. These phenotypes are complemented by traits of breeding interest such as ease of kernel removal and kernel weight. Through allometric relationships —relative growth of one tissue to another—, we identify possible biophysical constraints at play during development. We explore multiple correlations between all morphological and commercial traits, and identify which morphological traits can explain the most variability of commercial traits. We show that using only volume and thickness-based traits, especially inner air content, we can successfully encode several of the commercial traits.

We propose a solution with edge image processing and long-range connectivity named AICropCAM that can be used in drones, ground platforms, or as distributed sensor networks for plant phenotyping. We have successfully run multiple image classification, segmentation, and object detection models on this platform. Classification models help classify images based on image quality, crop type, and phenological stage. Object detection models could detect and count the number of plants, weeds, and insects and expand to count the flowers, fruits, and leaves. Segmentation models can separate the canopy from the background and potentially segment traits that indicate the nutrient deficit or disease. Canopy segmentation results help estimate leaf area index and chlorophyll content. Because the models run sequentially, like a decision tree, there is flexibility to select the most accurate model considering the crop type and the crop’s phenological stage that helps scan fields with multiple crops. The generated information is geo-tagged and transmitted through low throughput long-range communication protocol (e.g., LoRa) to cloud data storage. AICropCAM reduces 2-megabyte image files to around 100-byte actionable data, resulting in massive savings in data storage and transmission costs. This edge image capturing and processing system is open to improvement with new neural network predictive models and faster edge computers. This system provides plant scientists and crop breeders a low-cost, flexible phenotyping tool to extract multiple crop traits related to abiotic and biotic stress responses.

As the world’s population grows and the demand for food rises, more attention has been paid to increase crop yields and enhance global food security. Modern remote sensing technologies enable us to capture spectral features (such as NDVI) of crop canopy, which are widely used to assess crop growth, health, and stress conditions. We noticed in the literature that crop NDVI shows short-term variation within a day. Therefore, in this study, we leveraged the Spidercam Field Phenotyping Facility at University of Nebraska-Lincoln to measure and quantify the diurnal variation of canopy NDVI for corn and soybean crops. The experiments and data collection were conducted in 2022 and 2023, with canopy reflectance measured by a spectrometer (400-1000 nm) at multiple days covering different growth stages. In each day, measurements were taken at multiple time points within a time window ±3 hours centered around solar noon. Our analysis showed a clear concave-shaped, diurnal trend in NDVI for both crops, with the lowest NDVI at solar noon. More analyses will be performed to quantify this diurnal pattern and dissect the sources of variation due to solar angle and change in canopy morphology. This research will further improve the accuracy and relevance of NDVI in plant phenotyping and many other scientific disciplines and applications.

Global warming poses a substantial threat to food security, necessitating the development of crops resilient to the adverse effects of stress including, drought, heat, and their combined impact. To understand maize responses to individual and combined stress conditions of drought and heat at early vegetative stages, we did a comprehensive evaluation of the phenotypic responses of 47 diverse maize inbred lines. The plants were stressed for 13 days beginning at 7 days after planting, with heat stress conditions of 38/28°C day/night cycles, and the drought condition was achieved by reducing the water pot volume from 88 to 40% over the course of the experiment. The Bellwether Phenotyping Facility, a high-throughput phenotyping platform, was used to capture daily RGB images of the plants throughout the experiment. We extracted morphological and color-related traits from the images, and particular interest was placed on traits that exhibited high broad-sense heritability throughout the experiment. This approach allowed for the collection of a robust dataset of phenotypic responses. Our results revealed distinct responses among the maize genotypes under different stress conditions, with the combined drought and heat treatment leading to the most severe impairments in plant height and leaf area. To gain further insights, we applied a time series analysis to the extracted traits and created groups with dynamically-similar response patterns. This research is foundational for understanding maize stress responses to heat and drought conditions and will inform future work towards identifying important genes and molecular mechanisms, particularly in the context of combined stressors.

The Embedded Automated Generator of Labeled Images (EAGL-I) system is a tool for generating labeled images, particularly for data-driven methods, such as deep learning models. The system has already generated hundreds of thousands of images of weeds and crops. We present modifications made to the original system that are based on the experiences gathered from generating such large-scale datasets. The improvements relate to lighting conditions, ease of use, refined image segmentation, and pathfinding for camera-movements. To address lighting conditions, we made three major changes to the hardware. First, the blue keying fabric was replaced by solid black panels, mitigating reflections and achieving reliable color accuracy; second, sunlight entering the room through a window is diffused and partially blocked by a screen, achieving consistent and uniform lighting of the imaging environment; third, dimmable LED lights are installed allowing us to image with lower ISO and to reduce noise in the resulting images. A YOLO machine learning model was trained to replace the previous methods of estimating bounding boxes around the plants. This new way of creating bounding boxes adapts to different plant architectures, such as grasses or different kind of dicots. Finally, we implemented a version of the A* pathfinding algorithm to define save zones through which the camera will not be moved. Overall, these modifications improved system performance and image quality significantly, while making EAGL-I easier to use. We have extended potential applications of EAGL-I, particularly for plant phenotyping research and in fine-tuning machine learning models for image analysis.

Weed competition with crops is the number one cause of yield loss worldwide. Plant breeders combat this issue by breeding herbicide tolerant varieties. The Aggressor (Group1) and Beyond (Group 2) herbicides are commonly used to control weeds in the Pacific Northwest. Many weed species are developing tolerance to group 1 and 2 herbicides. Metribuzin is a group 5 herbicide that is labeled for use in wheat, but application results in severe plant injury. Release of winter wheat variety tolerant to Metribuzin would provide producers an alternate mode of action to control weeds. Historically breeders visually rate injury in breeding plots, these ratings can be variable and subject to individual bias. This study aims to improve the accuracy and efficiency of selecting herbicide tolerant lines in a breeding program by utilizing a drone mounted multispectral sensor. Multispectral data was collected on paired rows and advanced generation lines grown in yield trails. The vegetative indices calculated from the wavelengths measured by the multispectral sensor in this study include NDVI, NDRE, TCARI, NWI, and MTVI. Visual injury scores, plant height, and yield were also measured. Correlations between reflectance indices and yield were stronger than correlations between visual injury ratings and yield. There were also moderate to strong correlations observed between visual injury rating and vegetative indices as well as plant height difference and vegetative indices. The results of this study suggest that multispectral analysis on the plot level is an accurate indicator of herbicide injury in winter wheat.

Maximizing crop yield while conserving resources is a pressing challenge in modern agriculture. Maize (Zea mays L. ), a staple crop worldwide, relies heavily on photosynthesis, making radiation interception a pivotal factor in crop growth and yield. This study presents a novel approach to improve maize crop productivity by harnessing the principles of phyllotaxy and optimizing planting patterns to efficiently intercept solar radiation. Through the enhanced 3D maize model generation algorithm, the simulation incorporates critical factors such as curved surface leaf area, leaf arrangement, plant spacing, and solar angles, allowing us to quantify the radiation intercepted by the maize canopy. We investigate the most efficient phyllotaxy for individual plants and evaluate a range of planting patterns and their impact on radiation interception at various growth stages. Simulation results reveal that optimizing planting patterns based on phyllotactic principles can substantially enhance radiation capture compared to traditional planting methods. In conclusion, our research signifies a significant step towards harnessing the power of plant arrangement and optimizing planting strategies, including the use of an enhanced 3D maize model, to maximize radiation interception in maize crops. The optimal phyllotaxis and planting patterns establish a distinct phenological target for breeders. These findings hold promise for the development of more resilient and productive agricultural systems in an era of growing global food demand and resource constraints.

Climate change poses a significant threat to agricultural systems, with drought becoming increasingly prevalent in the Canadian prairies. This study addresses the urgent need to enhance crop resilience, focusing on Brassica carinata, a promising industrial feedstock crop used for the production of biofuels. Our research aims to comprehensively evaluate drought adaptive capacity in B. carinata through a combination of physiological and digital phenotyping methods. Under controlled conditions, we utilized a high-throughput phenotyping platform, the Plantarray system, to screen B. carinata germplasm. This system facilitated precise measurements of physiological traits, soil conditions, and atmospheric parameters, enabling the assessment of drought response. Concurrently, we conducted a field phenotyping experiment with 47 B. carinata Nested Association Mapping (NAM) founder lines and two B. napus checks, under irrigated and non-irrigated conditions. Aerial imagery obtained through Unmanned Aerial Vehicles (UAVs), complemented by phenological observations and manually recorded phenotypic data, was systematically gathered. Digital phenotypes extracted from aerial images are analyzed to identify a digital phenotype(s) for drought tolerance. Our study also explores the correlation between indoor physiological data and field performance of B. carinata lines, in an effort to identify parameters that can serve as reliable predictors of seed yield under drought stress. Overall, we believe this research provides valuable insights for enhancing crop resilience to drought.

Chenopodium quinoa is an important crop known for its salt tolerance. Salinity is an osmotic stress and ions accumulation in the root zone causes a reduction in soil water availability, affecting the uptake of essential nutrients, changing seed composition, and reducing biomass. Hence, the need for high-yield crops in poor soils. This research examines the effect of salt stress on quinoa photosynthetic efficiency and salt bladder development. Sensitive and tolerant quinoa lines were examined under salt stress conditions when a concentration of 155mM NaCl was applied. Soil conductivity was monitored for salt stress during the experiment. At approximately two months old, CropReporter images were taken and analyzed using PlantCV to estimate photosystem II efficiency, non-photochemical quenching (NPQ), chlorophyll content, and anthocyanin content. The analysis showed that the salt treatment did not negatively affect the plant photosynthetic efficiency (no changes in Fv/Fm, NPQ, Fq’/Fm’) but leaf area and chlorophyll content was statistically negatively affected by the treatment when comparing genotypes. Live tissue was also analyzed using reflection and fluorescence confocal microscopy, where epidermal salt bladders images were acquired, visualized and analyzed in 3-D and the salt tolerant line showed bigger bladder volumes compared with control conditions. A more high-throughput approach using PlantCV, an open-source image analysis software package targeted for plant phenotyping. This software helped count epidermal salt bladders using stereoscope images. A comprehensive understanding of the quinoa salt tolerant mechanisms by employing multidisciplinary approaches is necessary for their effective incorporation into salt-sensitive crops for better crop yields under stressful environments.