GENOMICS AND PHENOMICS TO IDENTIFY YIELD AND DROUGHT TOLERANCE ALLELES FOR IMPROVEMENT OF CAMELINA AS A BIOFUEL CROP

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: TERMINATED
• Funding Source: AFRI COMPETITIVE GRANT
• Reporting Frequency: Annual
• Accession No.: 1010241
• Grant No.: 2016-67009-25639
• Project No.: CALW-2016-06845
• Proposal No.: 2016-06845
• Program Code: A6151
• Project Start Date: Sep 1, 2016
• Project End Date: Aug 31, 2022
• Grant Year: 2016

Project Director
Abdel-Haleem, H.

Recipient Organization
AGRICULTURAL RESEARCH SERVICE
800 BUCHANAN ST, RM 2020
BERKELEY,CA 94710-1105

Performing Department
Agricultural Research Service

Non Technical Summary
Plant oils represent an outstanding potential source of energy-dense hydrocarbons that can be used for fuels and industrial raw materials, but a major challenge is to produce these oils in non-food oilseed crops that have high yields and can grow under marginal and varied climatic conditions. In recent years, Camelina sativa has received considerable attention as a potential non-food biofuels crop, but significant challenges remain to develop stable, high-yielding, geographically adapted germplasm suitable for biofuels production. We will utilize advanced high-throughput phenotyping and genomics-based approaches to discover useful gene/alleles controlling seed yield and oil content and quality in camelina under water-limited conditions, and will identify high-yielding cultivars suitable for production in different geographical regions. The project includes three primary objectives: 1) Develop and apply automated, non-destructive high-throughput phenotyping (HTP) protocols to evaluate the phenotypic diversity and stress tolerance of a camelina panel consisting of 250 accessions, grown under well-watered and water-limited conditions. 2) Discover alleles/genes controlling morphological, physiological, seed, and oil yield properties using genome-wide association studies (GWAS). 3) Identify, test, and validate useful germplasm, including transgenic lines producing drop-in ready jet fuel components, under diverse environments and marginal production areas. Taken together, this project will significantly advance the utilization of non-food oilseed crops for biofuel production and provide guidance and insight for future studies of phenomics-based crop improvement.

Animal Health Component

0%

Research Effort Categories

Basic

80%

Applied

20%

Developmental

0%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
201	1899	1081	100%

Knowledge Area
201 - Plant Genome, Genetics, and Genetic Mechanisms;

Subject Of Investigation
1899 - Oilseed and oil crops, general/other;

Field Of Science
1081 - Breeding;

Keywords

camelina sativa

camelina population

cultivars

non-food oilseed crops

bioenergy

biofuel

high-throughput phenotyping

drought tolerance

genome-wide association studies

Goals / Objectives
The major goals of the project are to:1) Develop and apply automated, non-destructive high-throughput phenotyping (HTP) protocols to evaluate the phenotypic diversity and stress tolerance of a camelina panel consisting of 250 accessions, grown under well-watered and water-limited conditions.2) Discover alleles/genes controlling morphological, physiological, seed, and oil yield properties using genome-wide association studies (GWAS).3) Identify, test, and validate useful germplasm, including transgenic lines producing drop-in ready jet fuel components, under diverse environments and marginal production areas.

Project Methods
For Objective 1, a panel of 250 camelina lines, representing world-wide population diversity, will be cultivated under well-watered and water-limited conditions in both the field (Arizona and Nebraska) and greenhouse (Danforth Center) and analyzed using complimentary high-throughput phenotyping (HTP) approaches. The HTP platform in Arizona includes several high clearance tractors mounted with various sensors, GPS, and data capture equipment, while the Nebraska system features similar equipment using a cart-based vehicle. Three standardized phenotypes will be measured including canopy reflectance, canopy temperature, and crop height. Similar phenotypic data will be captured using a LemnaTec-based greenhouse facility available at the Danforth Center. Traditional phenotypic data, including seed yield components, will also be determined for field-grown plants. For Objective 2, the camelina panel will first be characterized using genotyping-by-sequencing, then GWAS will be conducted to discover SNP markers associated with trait data. By comparing GWAS results obtained from field-grown and greenhouse-grown plants, we will further address the question of how well greenhouse-based experiments translate to discovery of genes related to performance in the field. Lastly, for Objective 3, we will conduct advanced yield trails by growing 10 to 20 select plant lines in four different locations including Arizona, Nebraska, Florida, and Minnesota. We will also include several transgenic camelina lines engineered to produce medium chain fatty acids and/or terpenes, both of which are important jet fuel components. The project will be regularly evaluated by the group through quarterly teleconferences and webinars, and more frequent meetings between subgroups focused on each of the major project areas including field-based phenotyping, greenhouse-based phenotyping, and yield trials. Major milestones include development of standardized methods and protocols for collection of sensor-based data and conversion to "trait data", completion of genotyping-by-sequencing of the camelina population and identification of SNPs, and implementation of genomie-wide association studies to identify trait/SNP correlations. Additional milestones include standardization of experimental design and data collection methods for field studies conducted at multiple locations.

Progress 09/01/16 to 08/31/22

Outputs
Target Audience:Our research program aimed to develop and use advanced genomic and phenomic tools to improve camelina as a biofuel crop, with the end-goal of identifying genes and molecular markers associated with stability under stress conditions, and identifying specific camelina lines that are adapted to different regions of the country. Our work directly impacted several target groups, the first of which included farmers and producers who are interested in growing camelina as either a winter rotation crop, or a spring-grown crop, particularly in arid and semi-arid climates. The second group included scientists and breeders who focus their interest on understanding the basic mechanisms of drought tolerance in crop plants, with the goal to improve crop performance using molecular breeding approaches. The third group included scientists who are developing sensor-based technologies for monitoring crop growth and performance in the field, with applications in both precision agriculture and phenomics-based crop improvement. This latter group is multidisciplinary in nature and included mechanical, electrical, and electronics engineers, computational scientists, plant breeders, and plant physiologists, who often work in large collaborative teams. This area of research that generally referred to as "high- throughput phenotyping," not only shows great potential for increasing crop yields through improved precision agriculture, but also for accelerating genomics-based crop improvement. Thus, the final target group includes the public and all other users and consumers of agricultural products. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?? In Maricopa, 8 summer and part-time students, 3 technicians, one post-doc, and a visiting post-doc from Nebraska' team had the opportunity to learn about camelina crop, its agronomic practices and traits, its genetics and genomics techniques and handling and processing of HTP data. Students had also the opportunity to work on improvements and optimization of HTP data pipelines, including the development database using open-source SQL server software and usage of Python and GIS for geo-referencing. Additional protocols were developed for assessing the quality of sensor data, and work was continued to develop of graphical user interfaces, with a goal to distribute these open-source tools to the broader scientific community. Nebraska team trained two technicians and four undergraduates and one postdoc on HTP technologies and the associated data analysis techniques. Both undergraduates and postdoc had extensive opportunities to understand the agronomic practices involved in camelina crop production. They built a new phenocart. The postdoc had furthered his knowledge in how to use the phenotyping cart to analyze the phenome of the crop and was engaged in helping to solve some of the problems associated with phenotyping a crop with small leaves and only moderate canopy cover. In Minnesota, a visiting PhD student from France researching camelina agronomics spent two weeks at Morris, MN and was trained on research being done in the US to develop best management practices for camelina production and identify camelina genotypes best adapted to production under rainfed conditions. Also, an undergraduate student worked during the summer on research related to the project to develop camelina as biofuel crop that can help diversify US cropping systems. At the Danforth Center an undergraduate student had the opportunity to learn about accelerated data collection and its ability to improve agronomic traits in addition to learning about phenotyping. This student was trained to collect, interpret, and present plant yield and NMR data to a wider scientific audience. Also, Danforth team trained a Washington University graduate student with major in Evolution, Ecology, and Population Biology program. He had learned the basics of plant phenotyping and also analytical techniques for determination of fatty acids and glucosinolates using both destructive (HPLC and GC-MS) and non-destructive (NIR) methods. How have the results been disseminated to communities of interest?The overall results of the project were disseminated at nine scientific conferences and meetings. The phenotyping group in Maricopa, Arizona actively participated in an NSF-funded workshop, provided twice a year in Maricopa, to help educate other scientists, students, and industry partners on the practical uses and applications of FB-HTP. At summer of 2018, Nebraska group organized a Phenomic workshop mainly aimed at the University of Nebraska scientists interested in this area of research (https://biotech.unl.edu/unl-plant-phenomics-symposium). The symposium included internal and three external speakers from Wyoming, Illinois and Canada. The results of the phenocart data and yield data from 2018 as well as other efforts on technology development were presented by the group at the symposium which was attended by 70 - 100 students, postdocs and faculty from the University of Nebraska. During summer of 2018, Minnesota represented the project results, goals and objectives at a field day event in Morris, Minnesota. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? Goal 1: To develop and validate a high-throughput phenotyping (HTP) protocol suitable for detecting the variation in the camelina diversity panel, ten high diverse camelina varieties were planted during the fall2016. HTP data were collected on a weekly basis and analyzed using custom data pipelines. Traditional morphological data including plant height, flowering time, and chlorophyll content were also collected. At physiological maturities, plots were harvested, and seed yields were determined. Fifteen FB-HTP-related traits were collected and analyzed. Reflectance data were used to construct vegetation indices including Normalized difference vegetation index (NDVI) and Normalized difference red edge index (NDRE). Analyses of HTP data showed a temporal increase in trait values during the growing season. At the end of season, HTP values reached plateau suggesting collecting HTP data during vegetative stages. Analysis of variances (ANOVA) indicated variations among the tested varieties for HTP related traits, thus the HTP protocol proved a useful approach to characterize the genetic and phenotypic variation in the camelina. The 250 accessions of camelina diversity panel, plus ten commercial varieties, were planted in Maricopa, AZ under well-watered and water-limited conditions, for two years. The HTP protocols were used to collected data throughout the growing season. The HTP-related traits showed phenotypic variations between stressed and non-stressed conditions, as well as among genotypes. The genotypes responded differently to stress, where the differences in canopy temperature under two environments for the same genotype varied from 1.1°C to 3.2°C. In general, plant chlorophyll is degraded under stress, resulting in lower vegetation index estimates. There were also significant variations among genotypes for NDVI trait. Analysis of variances over environments and years indicated significant effects for genotypes, environments, but no GxE was observed in the studied HTP-related traits. The same was observed for other physiological traits such as seed yield, plant height, flowering time, fatty acids profiles and oil and protein contents were significant effects for genotypes, environments were indicated but no GxE interactions. A non-destructive greenhouse HTP experiment was carried out on a LemnaTec Scanalyzer 3-D available at the Danforth Center using the camelina diversity panel to estimate and measure HTP-related traits using RGB images that allow visualization and quantification of plant morphology traits such as leaf area, stem diameter and plant height. Data showed that GH-HTP protocols could be used to assays the phenotypic diversity within camelina panel. An experiment was conducted at Maricopa, Arizona using two growth chambers to implement a cost-effective automated irrigation and high throughput phenotyping. Results showed that the GC-HTP system was able to capture the phenotypic variation in tested camelina genotypes. The system was capable of controlling water management and collecting phenotypic data that can be scaled up for greenhouse and field irrigation and phenotyping applications. The leaf cuticular waxes, indicator of drought tolerance, were extracted and determined in the camelina diversity panel. Analyses of variances indicated that camelina diversity panel exhibited a wide range in total leaf wax contents, wax classes and constituents. High heritability values of the primary alcohols alkanes suggest the feasibility for selection of these traits during early generations in breeding programs. Goal 2: The knowledge of population structure and genetic diversity of camelina is important to conduct analyses to discover candidate genes controlling important traits using genome-wide Association analyses (GWAS) The population structure and genetic diversity in a camelina panel were characterized based on high throughput genotyping-by-sequencing (GBS) genotyping data. Population structure data were generated using the STRUCTURE algorithm that identified two subpopulations, which agreed with geographical classification of spring camelina panel and origin of those accessions. The GWAS analyses were performed on phenotypic data collected from 2-year field trial conducted in Maricopa, AZ using a FarmCPU algorithm implemented in GAPIT 3 software. GWAS analyses identified 57 SNPs significantly associated with fatty acidsd, oil content, protein content, plant height. Of the 57 identified SNPs, 6 SNPs shared associations between different traits, indicating putative genes related to fatty acid biosynthesis and transfer. The GWAS revealed a total of 42 significant SNP markers that were significantly associated with 15 leaf wax traits including major wax components such as total primary alcohols, total alkanes, and total wax esters as well as their constituents. These SNP markers are located on genes directly or indirectly related to wax biosynthesis. These loci could potentially serve as candidates for the genetic control involved in intracellular wax transport that might directly or indirectly facilitate leaf wax accumulation incamelina. Salinized soil is one of the abiotic stresses that affect camelina seed germination and thus yield. GWAS analyses identified 19 SNPs that were significantly associated with camelina seed germination rate and seedling dry weight for accessions were grown under salt stress. The significant SNPs were directly or indirectly related to phosphatase metabolism, signaling transduction, and cell membrane activities that are known to be affected by salt stress. Floral transition at the right time is one of the plant defenses against stress conditions, as well to maximize seed yield production in camelina as other corps. The high heritability of 0.893 for flowering time suggests that breeding could be effective in developing early mature camelina varieties. Using GWAS analyses, a total of 20 significant trait-associated SNP were identified to be colocalized within/or near a variety of transcription factors or protein families containing specific functional domains. These transcription factors interact with key regulatory genes in biological pathways to cooperatively regulate floral transition. Whole-genome prediction was also conducted to estimate the predictive ability to develop early flowering in camelina that will speed the exploration and genetic improvements of flower development and timing in camelina. Goal 3: Fifteen camelina genotypes with high productivity traits were selected from diversity panel and evaluated in multi-state experiment across Arizona, Florida, and Minnesota. Seed yield, seed oil and fatty acid profiles were estimated. Analysis of variances over locations showed that genotypes and locations significantly affected the variations in studied traits. Heritability estimates for the studies traits ranged from 0.94 (palmitic fatty acid) to 0.43 (seed oil content) indicating the response of those traits to different environments and regions. Phenotypic variation among genotypes coupled with their stability over different geographical and environmental growing conditions indicated the possibility to select genetically diverse genotype for high yield and stability targeting diverse environments. Further in-depth drought tolerance testing of selected camelina lines was conducted at phenotyping experiment on Nebraska' LemnaTech greenhouse. Twelve Camelina genotypes differ in responses to drought stress were chosen from the camelina diversity panel. This study was designed to provide information into the possible mechanisms of drought tolerance in these lines. There were two water treatments (drought vs. well-watered). Result indicated the strong effects of drought treatments on the fresh weights at each harvest time point, as well as differences in the twelve genotypes in yield due to drought stress conditions.

Publications

• Type: Conference Papers and Presentations Status: Published Year Published: 2020 Citation: Lyu B., Kim, J.Y., Abdel-Haleem, H., Szczepanek, A. E. (2020) High-throughput Seed Detection using Convolutional Neural Network: A Baseline Study. ASABE Annual International Virtual Meeting 2001454. (doi:10.13031/aim.202001454)
• Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Abdel-Haleem, H., Szczepanek, A.E., Luo, Z. 2021. Genetic diversity and population structure of USDA collection brassica juncea population. 32nd Association for the Advancement of Industrial Crops meeting, Bologna, 5-8 September 2021
• Type: Journal Articles Status: Published Year Published: 2023 Citation: Kim, J.Y., Abdel-Haleem, H.A., Luo, Z., Szczepanek, A.E. (2023) Open-Source electronics for plant phenotyping and irrigation in controlled environment. Smart Agricultural Technology. 3. Article 100093.
• Type: Journal Articles Status: Published Year Published: 2021 Citation: Luo, L.L., N. Fahlgren, T. Kutchan, D.P. Schachtman, Y. Ge, R.W. Gesch, S. George, J.M. Dyer, H.A. Abdel-Haleem.2021. Discovering candidate genes related to flowering time in the spring panel of Camelina sativa. Industrial Crops and Products. 173. Article 114104.
• Type: Journal Articles Status: Published Year Published: 2022 Citation: Abdel-Haleem, H.A., Luo, Z., Szczepanek, A.E. (2022) Genetic diversity and population structure of the USDA collection of Brassica juncea. Industrial Crops and Products. 187(2022). Article 115379.
• Type: Journal Articles Status: Published Year Published: 2022 Citation: Polydore, S., Fahlgren, N. 2022. Phenotypic analysis of a European Camelina sativa diversity panel. ESS Open Archive. February 05, 2022. DOI: 10.1002/essoar.10508336.3

Progress 09/01/20 to 08/31/21

Outputs
Target Audience:Our research program aims to develop and use advanced genomic and phenomic tools to improve camelina as a biofuel crop, with the end-goal of identifying genes and molecular markers associated with increased drought tolerance, and identifying specific camelina lines that are adapted to different regions of the country. Our work directly impacts several target groups, the first of which includes farmers and producers who are interested in growing camelina as either a winter rotation crop, or a spring-grown crop, particularly in arid climates. The second group includes scientists and breeders who are focused on understanding the basic mechanisms of drought tolerance in crop plants, with the goal of improving crop performance using molecular breeding approaches. The third group includes scientists who are developing sensor-based technologies for monitoring crop growth and performance in the field, with applications in both precision agriculture and phenomics-based crop improvement. This latter group is multidisciplinary in nature and includes mechanical, electrical, and electronics engineers, computational scientists, plant breeders, and plant physiologists, who often work in large collaborative teams. A major challenge is the collection, storage, processing, and analysis of the "big data" associated with these studies, and converting sensor-based data in biologically useful information. This area of research, generally referred to as "high-throughput phenotyping," not only shows great potential for increasing crop yields through improved precision agriculture, but also for accelerating genomics-based crop improvement. Thus, the final target group includes the general public and all other users and consumers of agricultural products. Changes/Problems:COVID-19 situation and in-place work restrictions affected the research activities and planned progress. Outdoor experiments were delayed with limited access to the facility and summer intern shortage that affected finalize the project objectives as outlined. What opportunities for training and professional development has the project provided?Arizona team continued to traina part-time student, two technicians and postdoc offield and lab experimental layout design and taking notes, and data recording and analyses. A trained technician is devoting his time to collect HTP data and analyses. Nebraska team trained two technicians and four undergraduates and one postdoc in high throughput phenotyping technologies and the associated data analysis techniques. In addition, the team learned how to collect physiological data related to drought tolerance studies in camelina. How have the results been disseminated to communities of interest?Results of these studies are disseminated at various scientific conferences and meetings. What do you plan to do during the next reporting period to accomplish the goals?The primary goal over the next year is to organize, statistically analyze, and publish results for the multi-location studies to evaluate agronomic and phenotypic responses of the spring camelina panel and 15 selected camelina accessions under field conditions using traditional and high through-out approaches, and the in-depth greenhouse-based to study potential drought mechanisms and responses under stress conditions. We will also continue to analyze both field and greenhouse-derived data to discover candidate genes controlling traits related to drought tolerance and productivity in camellia under stress conditions using a genome-wide association approach.

Impacts
What was accomplished under these goals? Goal 1: Develop and apply automated, non-destructive high-throughput phenotyping (HTP) protocols to evaluate the phenotypic diversity and stress tolerance of a camelina panel consisting of 250 accessions, grown under well-watered and water-limited conditions. A phenotyping experiment on Nebraska' LemnaTech greenhouse was conducted, where plants could be imaged regularly, and the seed could be harvested for yield analysis. The experiment was designed to study camelina genotypes that have been tested under field conditions across multi-years and drought treatments in Maricopa, AZ. Twelve Camelina genotypes differ in responses to drought conditions (six drought tolerant genotypes, and six drought sensitive genotypes) were chosen from the panel of 230 genotypes. This study was designed to provide information into the possible mechanisms of drought tolerance in these lines. There were two water treatments (drought vs. well-watered). The drought treatment was designed to maintain pot water contentat 35~40% field capacity; the well-watered treatment was watered daily to 80~85% field capacity. There were ten replicate pots for each genotype and water treatment. The total number of plants was 240. A subset of the plants was destructively sampled to collect ground truth data to correlate with image data. Destructive sampling occurred onceevery 2 weeks for 8weeks. At each time, 48plants were sampled(12 genotypes x 2 water treatments x2plants). Each destructive harvest was used to estimate total leaf area,leaf chlorophyll concentration, leaf VisNIR spectral reflectance using ASD spectrometer, total fresh weight and dry weight and leaf samples for sucrose analysis. Result indicated the strong effects of drought treatments on the fresh weights at each harvest time point, as well as differences in the twelve genotypes in yield due to drought stress conditions. Data are being further analyzed and will be published. An experiment was conducted at Maricopa, Arizona using two growth chambers set on different temperatures (35°C and 25°C) and differing amounts of water (40% and 90% water holding capacity) to implement a cost-effective automated irrigation and high throughput phenotyping (HTP) to help identify key alleles and associated molecular markers conditioning abiotic stress tolerance and agronomic traits. The HTP system was developed using a microcontroller and a single-board computer connected to soil moisture sensors and water pump control for automated irrigation. The HTP system consists of an Arduino Uno board, Raspberry Pi, a multispectral camera, mini LiDAR sensors, infrared thermometers and temperature\humidity sensor. The HTP system monitored plants every four seconds for canopy temperature, plant height, and soil moisture, captured an image every 10 minutes, and controlled the water pumps every 6 hours based on soil water levels measured by the soil moisture sensors. Data were wirelessly monitored by a smartphone and transferred to a computer for further statistical analyses. The preliminary results showed phenotypic variation due to tested camelina genotypes and heat and drought stresses. The system was capable of controlling water management and collecting phenotypic data that can be scaled up for greenhouse and field irrigation and phenotyping applications. Goal 2: Discover alleles/genes controlling morphological, physiological, seed, and oil yield properties using genome-wide association studies (GWAS). Floral transition at the right time is one of the plant defenses against stress conditions, as well to maximize seed yield production in camelina as other corps. GWAS analyses were performed to identify candidate genes controlling the variations in camelina flowering time. Flowering times data were recorded for spring camelina panel planted at Maricopa, AZ in 2-year multi-location field trials (goal #1). GWAS analyses were conducted using a newly developed FarmCPU algorithm implemented in GAPIT 3 software. Our data indicated wide variation in flowering time in spring camelina germplasm. The current study showed significant genotypic, environmental and year to year effects on flowering time. The high heritability of 0.893 for flowering time suggests that breeding will be effective in developing early flowering camelina varieties. A total of 20 significant trait-associated single nucleotide polymorphisms (SNPs) were identified to be colocalized within/or near a variety of transcription factors (e.g. SPA1, BIM1) or protein families containing specific functional domains (e.g. CCCH zinc finger protein family and B3-DNA binding domain containing protein family). These transcription factors interact with key regulatory genes in the four major pathways (i.e. photoperiod, autonomous, vernalization and gibberellic acid pathways) to cooperatively regulate floral transition. Whole-genome prediction was also conducted to estimate the predictive ability to develop early flowering in camelina that will speed the exploration and genetic improvements of flower development and timing in a camelina. Goal 3: Identify, test, and validate useful germplasm, including transgenic lines producing drop-in ready jet fuel components, under diverse environments and marginal production areas. To stay the camelina response to different U.S. geographical regions, fifteen diverse camelina genotypes selected from spring panel (goal #1) and evaluated in an advanced yield trial in multi-state experiment across Arizona, Florida, and Minnesota. Seed yield, seed oil and protein compositions and fatty acid profiles were estimated. Analysis of variances over locations and years showed that genotypes and locations significantly affected the variation in studied traits. Heritability estimates for the studies traits ranged from 0.94 (palmitic fatty acid) to 0.43 (seed oil content) indicating the response of those traits to different environments and regions. Phenotypic variation among genotypes coupled with their stability over different geographical and environmental growing conditions indicates the possibility to select genetically diverse genotype to start next breeding program for high yield and stability targeting diverse environments.

Publications

• Type: Journal Articles Status: Published Year Published: 2020 Citation: Luo, Z., Szczepanek, A.E., Abdel-Haleem, H.A. 2020. Genome-wide association study (GWAS) analysis of camelina seedling germination under salt stress condition. Agronomy. 10(9). https://doi.org/10.3390/agronomy10091444

Progress 09/01/19 to 08/31/20

Outputs
Target Audience:Our research program aims to develop and use advanced genomic and phenomic tools to improve camelina as a biofuel crop, with the end-goal of identifying genes and molecular markers associated with increased drought tolerance and identifying specific camelina lines that are adapted to different regions of the country. Our work directly impacts several target groups, the first of which includes farmers and producers who are interested in growing camelina as either a winter rotation crop, or a spring-grown crop, particularly in arid climates. The second group includes scientists and breeders who are focused on understanding the basic mechanisms of drought tolerance in crop plants, with the goal of improving crop performance using molecular breeding approaches. The third group includes scientists who are developing sensor-based technologies for monitoring crop growth and performance in the field, with applications in both precision agriculture and phenomics-based crop improvement. This latter group is multidisciplinary in nature and includes mechanical, electrical, and electronics engineers, computational scientists, plant breeders, and plant physiologists, who often work in large collaborative teams. A major challenge is the collection, storage, processing, and analysis of the "big data" associated with these studies and converting sensor-based data in biologically useful information. This area of research, generally referred to as "high-throughput phenotyping," not only shows great potential for increasing crop yields through improved precision agriculture, but also for accelerating genomics-based crop improvement. Thus, the final target group includes the general public and all other users and consumers of agricultural products. Changes/Problems:In Nebraska, the post-doc working on the project left in October of 2019 and we had to fill in with laboratory staff from Drs. Ge and Schachtman's labs. Hence, this slowed the expenditure of funds. What opportunities for training and professional development has the project provided?Arizona: Trained 3 part-time students, 2 technicians and a post-doc in field and lab experimental layout design and data collection and analysis. Trained a technician to collect HTP data and analyze the results. Danforth Center: The doctoral student Jordan Brock of Washington University in St. Louis (WUSTL) works with us on the Lemnatec phenotyping experiment. He also participated in the NIR analysis of seeds. Jordan is a doctoral student in the Evolution, Ecology and Population Biology program at WUSTL whose thesis in on the evolution of Camelina spp. Within this current project, Jordan has received training in the molecular genetics and biochemistry of Camelina, as well as in phenotyping techniques. How have the results been disseminated to communities of interest?Results of these studies were disseminated at various scientific conferences and meetings. What do you plan to do during the next reporting period to accomplish the goals?The primary goal over the next year is to organize, statistically analyze, and publish results for the multi-location study to evaluate agronomic and phenotypic responses of the 15 selected camelina accessions. We will also continue to analyze both field and greenhouse-derived data to evaluate camelina responses to stress, and identify genes underpinning traits of interest.

Impacts
What was accomplished under these goals? Goal 1: Develop and apply automated, non-destructive high-throughput phenotyping (HTP) protocols to evaluate the phenotypic diversity and stress tolerance of a camelina panel consisting of 250 accessions, grown under well-watered and water-limited conditions. Arizona: The camelina diversity panel including 250 lines plus 10 commercial varieties was planted for a second year in Maricopa, Arizona, with three replications each cultivated under well-watered or water-limited conditions. High-throughput phenotyping and traditional phenotypic data were collected throughout the growing seasons including seed yield, seed size, plant height, flowering time, fatty acids profiles and oil and protein contents, canopy temperature and vegetation indexes. Preliminary analyses over environments and years indicated significant effects for genotypes, environments, but no GxE was observed. Finalized analyses will be reported in the next report. Nebraska: Further in-depth drought tolerance testing of selected Camelina lines was completed. Lines were chosen based on two years of field data from experiments conducted in Arizona. The experiment was conducted in the Nebraska automated greenhouse, which allowed for precise measurement of soil water status of pots using a weighing and automated watering system to study lines that have been tested in the field across two years that differ in yield under drought. The system also allowed for frequent whole plant imaging using RGB, Infrared, hyperspectral and NIR cameras to gather both morphological and physiological data. Twelve camelina lines were tested with six being drought tolerant and six being drought sensitive. The study aimed to provide information regarding the possible mechanisms of drought tolerance in these lines. 12 camelina genotypes (6 drought tolerant lines, and 6 drought sensitive lines) were chosen from a panel of 230 lines that were field tested across two years by our collaborators in Maricopa, Arizona. Two water treatments (drought vs. well-watered). Droughted plants were maintained at 30~35% field capacity and well-watered plants at 80~85% field capacity. 10 replicates for each genotype and water treatment combinations for imaging and 8 replicates for destructive harvest at different time points were studied. Plants were randomized into 18 blocks with the well-watered and water-stressed plants randomized in each block. Ten empty pots were included to estimate evaporative water use and to calculate the water use efficiency of the lines. Imaging was done three times per week with all imaging modules. Imaging started when plants were three weeks old and ended imaging at post-flowering stage. A subset of the plants was destructively sampled to collect ground truth data to correlate with image data. Destructive sampling occurred every 2 weeks for 8 weeks. Each time, 48 plants were sampled (12 genotypes x 2 water treatments x 2 plants). The destructive harvests consisted of measuring: total leaf area; leaf chlorophyll with a lab analysis; leaf Vis-NIR spectral reflectance; total fresh weight; total dry weight; leaf sucrose; dry weight and leaf mineral analysis. Currently we are analyzing the plant image data and will work to write this up in the next 3 - 6 months. Danforth Center: The Kutchan lab at the Danforth Center performed NIR analysis on collected seed samples of Camelina that resulted from the spring 2019 Lemnatec phenotyping experiment and from Maricopa field trials. The results from these measurements provide information on seed composition that will be combined with data drawn from digital imaging RGB measurements to perform GWAS. Goal 2: Discover alleles/genes controlling morphological, physiological, seed, and oil yield properties using genome-wide association studies (GWAS). Preliminary GWAS analyses were performed by combining GBS data and phenotypic evaluations (goal 1) for 35 traits of interests in a 2-year field trial conducted in Maricopa, AZ. Phenotypic data were collected for 15 different fatty acids, 10 fatty acid ratios, and 4 agronomic traits. GWAS analyses were conducted using a newly developed FarmCPU algorithm implemented in GAPIT 3 software. GWAS analyses identified 57 SNPs significantly associated with palmitic acid, stearic acid, oleic acid, linoleic acid, linolenic acid, arachidic acid, eicosenoic acid, eicosadienoic acid, oil content, protein content, plant height and flowering date. Of the 57 identified SNPs, 6 SNPs shared associations between different traits, indicating putative genes related to fatty acid biosynthesis and transfer. Soil conditions may negatively affect camelina seed germination, which reduces optimal canopy stand and reduces final seed yield. Salinized soil is one of the abiotic stresses that affect camelina seed germination. A study was conducted to identify candidate genes controlling camelina seedling germination when irrigated with a 100 mM NaCl solution. GWAS analyses identified 19 SNPs that were significantly associated with camelina seed germination rate and seedling dry weight under salt stress. The significant SNPs were directly or indirectly related to phosphatase metabolism, signaling transduction, and cell membrane activities that are known to be affected by salt stress. With candidate gene discovery, the putative trait-associated SNPs can be used to unveil molecular mechanisms underlying secondary metabolite synthesis, plant growth and development, and genetically dissect important agronomic traits and to be used as molecular makers in breeding programs Goal 3: Identify, test, and validate useful germplasm, including transgenic lines producing drop-in ready jet fuel components, under diverse environments and marginal production areas. Fifteen camelina genotypes with diverse traits were selected and planted in an advanced yield trial in six repeated experiments across Florida, Minnesota and Arizona. The trial was designed to identify geographically adapted cultivars for different regions. Analysis of variances over locations and years showed that genotypes and environments significantly affected the studied traits. Varied phenotypic plasticity of studied genotypes indicated that some genotypes are stable over different geographical and environmental growing conditions. Minnesota: During the 2019 growing season, the 15 camelina accessions selected for improved drought tolerance were grown in Minnesota under rainfed conditions. The 2019 growing season in Minnesota received above normal precipitation. No drought or disease symptoms were observed. The flowering date among genotypes varied by as much as 4 days. Some differences in seed yield were observed among genotypes and yield ranged from 870 to 1237 kg/ha. Several of the accessions selected for improved drought tolerance yielded as much or more seed than the common control cultivars that were grown for comparison. Seedling emergence and hence final stand density varied substantially between replicated plots but did not greatly differ across most genotypes. However, one genotype, Cs058 stood out as consistently demonstrating reduced emergence and stand density as compared with the other accessions. For all the accessions grown in Minnesota, harvested seeds were dried and cleaned, and subsamples sent to Maricopa, AZ for analysis of oil content and fatty acid profile. Based on the evaluation of soil moisture from the surface to 1 m deep over the growing season, there were no differences observed in overall plant water use for four genotypes that were selected for measuring. However, there was an excess of precipitation, and hence soil moisture, particularly in the early part of the growing season.

Publications

• Type: Journal Articles Status: Published Year Published: 2019 Citation: Augustin, J.M., Brock, J.R., Augustin, M.M., Wellinghoff, R.L., Shipp, M., Higashi, Y., Kumssa, T.T., Cahoon, E.B., Kutchan, T.M. Field performance of terpene-producing Camelina sativa. Ind. Crop. Prod. 136:50-58 (2019).
• Type: Conference Papers and Presentations Status: Other Year Published: 2019 Citation: Luo, Z., T. Kutchan, M. Augustin, N. Fahlgren, D.P. Schachtman, Y. Ge, J. Dyer, and H. Abdel-Haleem (2019). Genome-wide association study (GWAS) helps unveil the genomic regions controlling yield components and fatty acids in Camelina sativa. Cruz, V.M.V. and M. Berti. (eds.) 2019. Advancing the Adoption of Industrial Crops through Innovation and Technology. 31st Annual Meeting of the Association for the Advancement of Industrial Crops (AAIC). Program and Abstracts. September 8-11, 2019, Tucson, Arizona, USA.
• Type: Conference Papers and Presentations Status: Other Year Published: 2020 Citation: Abdel-Haleem, H., L. Zinan, A.E. Szczepanek, M. Conley, H. Zhang, Y. Ge, D. Schachtman, M.M. Augustin, J. R. Brock, N. Fahlgren, T. Kutchan, S. George, R. Gesch and J. M. Dyer (2020). Genomics and phenomics to identify yield and drought tolerance alleles for improvement of camelina as a biofuel crop. Genomic Sciences Program Annual Principal Investigator Meeting, February 23-26, the Washington Hilton, Washington, DC, USA.

Progress 09/01/18 to 08/31/19

Outputs
Target Audience:Our research program aims to develop and use advanced genomic and phenomic tools to improve camelina as a biofuel crop, with the end-goal of identifying genes and molecular markers associated with increased drought tolerance, and identifying specific camelina lines that are adapted to different regions of the country. Our work directly impacts several target groups, the first of which includes farmers and producers who are interested in growing camelina as either a winter rotation crop, or a spring-grown crop, particularly in arid climates. The second group includes scientists and breeders who are focused on understanding the basic mechanisms of drought tolerance in crop plants, with the goal of improving crop performance using molecular breeding approaches. The third group includes scientists who are developing sensor-based technologies for monitoring crop growth and performance in the field, with applications in both precision agriculture and phenomics-based crop improvement. This latter group is multidisciplinary in nature and includes mechanical, electrical, and electronics engineers, computational scientists, plant breeders, and plant physiologists, who often work in large collaborative teams. A major challenge is the collection, storage, processing, and analysis of the "big data" associated with these studies, and converting sensor-based data in biologically useful information. This area of research, generally referred to as "high-throughput phenotyping," not only shows great potential for increasing crop yields through improved precision agriculture, but also for accelerating genomics-based crop improvement. Thus, the final target group includes the general public and all other users and consumers of agricultural products. Changes/Problems:In Minnesota, field prep and planting were slightly delayed (about two weeks) in Morris due to an extended winter and late arrival of spring combined with above normal precipitation. What opportunities for training and professional development has the project provided?In Maricopa, 4 summer students, 3 technicinas, one post-doc, and a visiting post-doc from Nebraska were all trained in methods of high-throughput phenotyping analysis of crops cultivated under normal and stress conditions. In Minnesota, a visiting PhD student from France researching camelina agronomics will spend two weeks with Dr. Gesch in Morris, MN to be mentored and learn about research being done in the US to develop best management practices for camelina production and identify camelina genotypes best adapted to production under rainfed conditions. Also, an undergraduate student will be working during the summer on the project and learning about research to develop camelina as biofuel crop that can help diversify US cropping systems. In Nebraska, the project has provided the opportunity for a postdoc and technician to gain extensive experience with plant phenotyping methods and data analysis. During the summer, 2 - 3 undergraduate students are involved in learning the agronomic side of field work and have learned some of the basics of using the phenocart. How have the results been disseminated to communities of interest?Results, project goals and pbjectives were presented at a field day event in Morris, Minnesota. What do you plan to do during the next reporting period to accomplish the goals?In Maricopa we will continue to analyze data to identify genes and markers associated with yield and drought-related traits in camelina. We will also complete an advanced yield trial to help identify regionally adapted cultivars that can be grown in different geographical locations. At the Danforth Center, seeds from the selected camelina plants grown more extensive drought conditions will be harvested, weighed, and analyzed by NIR to obtain approximate values for critical metrics including % oil, individual fatty acid content, glucosinolate, moisture, and nitrogen levels during the next reporting period. Seeds harvested from the 2019 Maricopa, AZ field trial will also be analyzed by NIR. The data will be compiled and compared to the previous LemnaTec drought study to look for lines that are consistently performing well in both the greenhouse and in the field, and the best genotypes will be highlighted and labeled with their best traits. Terpene-producing transgenic camelina grown with the 5 watering conditions will also be analyzed by GC-MS to determine terpene content when the seeds are dry and harvested. In Minnesota, seed and plant samples taken during the summer will be processed and analyzed in Morris, MN, including assessment of seed quality in the laboratory. A database will be built and developed to determine G x E interactions across the diverse environments used in the study. Agronomically best suited genotypes for a given environment will be identified. After the 2019 field season, the focus for the Nebraska group will be on finalizing the field data, completing data analysis and writing manuscripts from the work.

Impacts
What was accomplished under these goals? Goal 1: Develop and apply automated, non-destructive high-throughput phenotyping (HTP) protocols to evaluate the phenotypic diversity and stress tolerance of a camelina panel consisting of 250 accessions, grown under well-watered and water-limited conditions. The 250 camelina accessions were grown under well-watered and water-limited conditions in Maricopa, Arizona and year-one data were analyzed. Analyses of variances showed significant genetic and environmental effects but no GXE interactions among the 250 accessions for seed weight and seed yield. For example, the average 1,000 seed weights of camelina genotypes grown under stressed conditions (i.e., water-limited) were lighter by 14% compared to the same lines grown under well-water conditions. Within each treatment, genetic effects were significant for these traits. Some genotypes were stable, as their seed weight did not change much between the two environments even though a few genotypes had higher seed weights when planted under stress conditions. These genotypes could be good candidates for breeding for larger seeds under stress. HTP-related traits also showed phenotypic variations between stressed and non-stressed conditions, as well as among genotypes. For example, during HTP run #6, the average canopy temperate for plants grown under stress was higher than those grown under non-stress by 2.2°C, indicating the effect of environment on plant transpiration. The genotypes responded differently to stress, where the differences in canopy temperature under two environments for the same genotype varied from 1.1°C to 3.2°C. In general, plant chlorophyll is degraded under stress, resulting in lower vegetation index estimates. Our results indicated that camelina genotypes grown under stress tend to have lower NDVI values than those grown in well-watered conditions. There were also significant variations among genotypes. During the 2018-2019 season (year 2 of the experiment), the 250 lines and 10 commercial varieties were planted under well-watered and water-limited conditions, with three replications each. The 1,500 plots were maintained and harvested at suitable physiological maturities. HTP and traditional phenotyping data were collected throughout the growing season. Plots were harvested and will be used to estimate seed yield and weight. The 250 camelina lines were also planted and analyzed in Sidney, Nebraska. Four phenocart data collections were made but one week prior to harvest, the seed yield data were lost due to a hailstorm. Therefore, in the winter of 2019 the Nebraska group took their phenocart with four sensors (RGB, Multispectral, Lidar and Infrared Radiometer) to Maricopa, Arizona to phenotype the 250 camelina accessions being grown under well-watered and water-limited conditions. Phenocart data were collect eleven times during the season, along with soil moisture at three times during the season. The data are currently being analyzed and will be compared to results obtained using the Maricopa phenotyping system. The second major activity for the Nebraska group will be to repeat he 2018 field experiment in Sidney, NE using the 250 camelina accessions. This experiment will be completed by mid-August 2019 and will include phenocart measurements and yield data. The camelina panel was also previously grown in a greenhouse at the Danforth Center under well-watered and water-limited conditions and analyzed using an HTP LemnaTec Scanalyzer. From this study, 54 genotypes were selected for more comprehensive drought analysis. Phenotypic traits included: 1. Low erucic acid (22:1-anti-nutritive compound); 2. High seed yield in both well-watered and drought conditions; 3. High and low glucosinolate (GLS) producing genotypes (some GLS are anti-nutritive while others provide health benefits); and 4. High total oil (major industrial product) and genotypes with high α-linolenic acid (18:3- health promoting omega-3 fatty acid). The selected genotypes were grown for 4 weeks at water moisture levels of 100% (high), 80% (normal), 60% (slight drought), 40% (drought), and 20% (extreme drought). Each genotype was grown in triplicate and subject to VIS and NIR imaging with 2 side view/1 top view images taken each day. Mature plants were moved to a greenhouse where the drought study was continued until plant senescence. Seeds will be harvested and analyzed for multiple parameters including glucosinolate content, % oil, moisture, nitrogen, and fatty acid distribution using NIR calibrated specifically for camelina at the USDA ARS lab in Peoria, IL. Image data collected from the initial phenotyping experiment were analyzed using PlantCV, a plant phenotyping software package developed by co-PI Fahlgren. Plant traits for each plant at each timepoint from 8-37 days after sowing were measured, including total projected shoot area, plant height, and other shape and color characteristics. The Nebraska and Danforth groups are working together to further process and analyze the image data to compare greenhouse-based data to results obtained from HTP analysis of crops grown in the field. Goal 2: Discover alleles/genes controlling morphological, physiological, seed, and oil yield properties using genome-wide association studies (GWAS). Genome-wide association studies were conducted in Maricopa, AZ on year-one phenotypic data. Together, seven and nine significant SNP markers on different chromosomes were associated with plant height, seed weight and seed yield under water-limited and well-watered conditions, respectively, meaning that some genes might be actively triggered under either of the two conditions. No significant markers overlapped between environments. BLASTx annotation identified gene functions for 8 out of 16 genes where significant SNP markers resided. These functions were mainly related to general maintenance of cellular function and plant growth and development. These SNPs might be useful for marker-assisted selections. Data from more environments, years and locations will be combined for a more comprehensive GWAS analysis. Goal 3: Identify, test, and validate useful germplasm, including transgenic lines producing drop-in ready jet fuel components, under diverse environments and marginal production areas. Fifteen camelina genotypes with diverse traits were selected for an advanced yield trial in Florida, Minnesota, Maricop and Arizona to identify geographically adapted cultivars. In Florida, the 15 genotypes were planted and treated with irrigated and non-irrigated conditions. Data collection included stand count, days to flowering, lodging and scouted for diseases such as alternaria leaf spot and sclerotinia stem rot. All plots were harvested and seeds were dried, cleaned and weighed. Yield was calculated on the basis of 10% moisture. In Minnesota, the 15 camelina accessions were planted and evaluated under rainfed (non-irrigated) conditions. Periodically throughout the growing season, phenological data will be collected including final emergence (stand establishment), initial and 50% flowering time, plant height, incidence of disease if any, and SPAD measurements as an indicator of drought and/or other environmental stress. Additionally, one check line, a cultivar that has been grown extensively in Morris (CO46), and three experimental lines identified as being high yielding with improved drought resistance, will be used to monitor seasonal water use. Three replicate plots of each accession will be measured weekly for soil moisture from 0-100 cm depth. Camelina accessions will be harvested and used to determine yield. A subsample of seed will be thoroughly cleaned and used to measure seed oil content, total N and C content and fatty acid profiles. At harvest time, the dry weight of total biomass before threshing and dry seed weight after threshing will be used to determine harvest index.

Publications

• Type: Journal Articles Status: Published Year Published: 2018 Citation: Berry, J.C., Fahlgren, N., Pokorny, A.A., Bart, R.S., Veley, K.M. 2018. An automated, high-throughput method for standardizing image color profiles to improve image-based plant phenotyping. PeerJ. 6:e5727. DOI: 10.7717/peerj.5727.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Luo, Z., Brock, J., Dyer, J.M., Kutchan, T., Augustin, M., Scachtman, D., Ge, Y., Fahlgren, N., Abdel-Haleem, H.A. 2019. Genetic diversity and population structure of a Camelina sativa spring panel. Frontiers in Plant Science. 10:184.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Luo, Z., Tomasi, P., Fahlgren, N., Abdel-Haleem, H.A. 2019. Genome-wide association study (GWAS) of leaf cuticular wax components in Camelina sativa identifies genetic loci related to intracellular wax transport. Biomed Central (BMC) Plant Biology. 19:187.
• Type: Conference Papers and Presentations Status: Other Year Published: 2019 Citation: Abdel-Haleem, H. , Luo, Z., Tomasi, P,. Lohrey, G., Park, S., Wang, H., Fahlgren, N., Jenks, M., Dyer, J.M. 2019. Camelina leaf wax constituents and distributions, phenotypic variations and allele identification. 30th Annual Meeting of the Association for the Advancement of Industrial Crops (AAIC). London, ON.
• Type: Other Status: Other Year Published: 2019 Citation: Dyer, J.M. 2019. Genomics and phenomics to identify yield and drought tolerance alleles for improvement of camelina as a biofuel crop. Genomic Sciences Program Annual Principal Investigator Meeting, February 24-27, Tysons Corner, VA.
• Type: Other Status: Other Year Published: 2018 Citation: Fahlgren, N. 2018. Image analysis and data management with PlantCV. Wheat initiative training course on high-throughput wheat phenotyping. Bologna, Italy.
• Type: Conference Papers and Presentations Status: Other Year Published: 2018 Citation: Fahlgren, N. 2018. Machine learning methods in PlantCV for leaf tracking and more. 5th Annual Meeting of the Arkansas Bioinformatics Consortium. Little Rock, AR.
• Type: Conference Papers and Presentations Status: Other Year Published: 2018 Citation: Fahlgren, N. 2018. Raspberry pi-powered imaging and open source software for plant phenotyping. International Plant Phenotyping Symposium. Adelaide, Australia.
• Type: Other Status: Other Year Published: 2019 Citation: Fahlgren, N. 2019. A modular, community-driven framework for developing high-throughput plant phenotyping tools. Big Data Seminar Series, Southern Illinois University. Carbondale, IL.
• Type: Other Status: Other Year Published: 2019 Citation: Fahlgren, N. 2019. A modular, community-driven framework for developing high-throughput plant phenotyping tools. Plant Sciences Institute, Iowa State University, Ames, IA.
• Type: Conference Papers and Presentations Status: Other Year Published: 2019 Citation: Luo, Z., J. Brock, J. Dyer, T. Kutchan, M. Augustin, D.P. Schachtman, Y. Ge, N. Fehlgren, and H. Abdel-Haleem. 2019. Genetic diversity and population structure of Camelina sativa spring panel. 30th Annual Meeting of the Association for the Advancement of Industrial Crops (AAIC). September 23-26, 2018, London, ON.

Progress 09/01/17 to 08/31/18

Outputs
Target Audience:Our research program aims to develop and use advanced genomic and phenomic tools to improve camelina as a biofuel crop, with the end-goal of identifying genes and molecular markers associated with increased drought tolerance, and identifying specific camelina lines that are adapted to different regions of the country. Our work directly impacts several target groups, the first of which includes farmers and producers who are interested in growing camelina as either a winter rotation crop, or a spring-grown crop, particularly in arid climates. The second group includes scientists and breeders who are focused on understanding the basic mechanisms of drought tolerance in crop plants, with the goal of improving crop performance using molecular breeding approaches. The third group includes scientists who are developing sensor-based technologies for monitoring crop growth and performance in the field, with applications in both precision agriculture and phenomics-based crop improvement. This latter group is multidisciplinary in nature and includes mechanical, electrical, and electronics engineers, computational scientists, plant breeders, and plant physiologists, who often work in large collaborative teams. A major challenge is the collection, storage, processing, and analysis of the "big data" associated with these studies, and converting sensor-based data in biologically useful information. This area of research, generally referred to as "high-throughput phenotyping," not only shows great potential for increasing crop yields through improved precision agriculture, but also for accelerating genomics-based crop improvement. Thus, the final target group includes the general public and all other users and consumers of agricultural products. Changes/Problems:Given the late start at the beginning of our grant, we have requested, and received approval for, a one-year no-cost extension to help complete all of the proposed experiments. The new end date for this grant is 8/31/2020. In Nebraska, the collection of camelina lines was planted in Sidney NE, i.e., 234 of the 250 lines for which we could obtain seed. At the time of writing this report (July 31, 2018), four phenocart evaluations were conducted, stand counts were taken, but over the weekend, the seed on the entire field shattered due to hail resulting in a loss of yield information from this trial. We will repeat this experiment in the summer of 2019 summer growing out all the camelina lines and will harvest the crop earlier to avoid shattering. We will also use the crop in Maricopa over the winter to further calibrate and test our phenocart sensors. This will be done in collaboration with the USDA group in Maricopa. What opportunities for training and professional development has the project provided?In Maricopa, undergraduates and graduate students had the opportunity to learn about camelina, measurement of agronomic traits and the recording and processing of yield and HTP data. Students also had the opportunity to work on improvements and optimization of data pipelines, geographic information systems (GIS), and the development of an HTP database using open source SQL server software. Additional protocols were developed for assessing the quality of sensor data, in particular, by using Jupyter Notebooks. The end goal is to distribute these open source tools to the broader scientific community. In Nebraska, both undergraduates and postdocs had extensive opportunities to understand the agronomic practices involved in small seeded crop production. A postdoc, undergraduate and graduate student built a new phenocart and gained extensive experience using it in the camelina field. This has involved helping to solve some of the problems associated with electrical circuits and learning all the electronics behind the system developed by Dr. Ge. A LIDAR system was installed this year to measure canopy height, which required additional modification of the data collection software. At the Danforth Center an undergraduate REU student had the opportunity to learn about accelerated data collection and its ability to improve agronomic traits in addition to learning about phenotyping. This student had to collect, interpret, and present plant yield and NMR data to a wider scientific audience. Also at Danforth, a Washington University graduate student in the Evolution, Ecology, and Population Biology program has learned the basics of plant phenotyping and also analytical techniques for determination of fatty acids and glucosinolates using both destructive (HPLC and GC-MS) and non-destructive (NIR) methods. How have the results been disseminated to communities of interest?The phenotyping group in Maricopa, Arizona actively participates in an NSF-funded workshop, provided twice a year in Maricopa, to help educate other scientists, students, and industry partners on the practical uses and applications of field-based, high-throughput phenotyping. While this is not a direct goal of our NIFA-funded project, our core HTP group is actively involved in promoting the technology with the broader scientific community. On April 2, 2018 Dr. Schachtman organized a Phenomic workshop mainly aimed at the University of Nebraska scientists interested in this area of research (https://biotech.unl.edu/unl-plant-phenomics-symposium). The symposium included internal and three external speakers from Wyoming, Illinois and Canada. The results of the phenocart data and yield data from 2018 as well as other efforts on technology development were presented by Drs. Ge and Schachtman at the symposium which was attended by 70 - 100 students, postdocs and faculty from the University of Nebraska. What do you plan to do during the next reporting period to accomplish the goals?Over the next year, data from the first camelina field trials in Maricopa and Nebraska will be collected and analyzed. Seeds from camelina plants grown in both Nebraska and Maricopa will be sent to the Danforth Center for analysis by NIR to determine traits including general content of chlorophyll, fatty acids, glucosinolates, moisture, nitrogen, and oil. HTP-related data from the field experiments will be processed to obtain trait data representing plant height, canopy temperature, and canopy spectral reflectance. GWAS association analyses and allele discovery for all phenotypic data will be conducted. In Maricopa, the camelina field trial will be repeated for a second year of testing under well-watered and water-limited conditions. Both traditional and HTP-related data will be collected, as before. In Nebraska, 234 accessions of camelina will be planted under well watered and water stressed conditions and phenotypes will be measured to assess the relative drought tolerance of each accession in the summer of 2019. In the winter of 2019, the group from Nebraska will bring their phenocart and sensor package down to Maricopa to compare data collected with the Nebraska sensors to those used in Maricopa, to further investigate the cross platform datasets and ease of use of the two different methods for crop phenotyping. The LIDAR will also be calibrated for crop height. In St. Louis, selected genotypes will undergo another phenotyping experiment using the LemnaTec Scanalyzer to assess drought tolerance and its effect on agronomic traits. These seeds will ultimately be analyzed by NIR. Also in year 3, an advanced yield trial with approximately 10 selected camelina lines will be conducted at multiple locations including Maricopa, Arizona, Scotts Bluff, Nebraska, Quincy, Florida, and Morris, Minnesota, with the end goal of identifying regionally adapted cultivars. Several sites will also evaluate the performance of transgenic camelina lines engineered to produce either terpenes or medium-chain fatty acids, both of which are high value feedstocks for jet fuel production.

Impacts
What was accomplished under these goals? Goal 1: During the fall/winter 2017-2018 season, 241 accessions of the camelina spring diversity panel, plus ten commercial varieties, were planted in Maricopa, AZ under well watered and water limited conditions, with three reps each. A high-clearance tractor equipped with proximal sensors was used to collect field-based high-throughput phenotyping (FB-HTP) data on eight different dates throughout the growing season. Data captured included sonic measurements used to estimate crop height, canopy multi-spectral reflectances used to determine various vegetative indicies, and canopy temperature. The eight FB-HTP runs and data were stored and are awaiting final processing. At physiological maturities, plots were harvested and seed yields determined. Analyses of variance indicated significant effects for water stress treatments as well genotype plant heights, where plants tended to be shorter under water deficient conditions. For example, Calena's heights (check variety) were 93 cm and 88 cm under non-stressed and stressed conditions, respectively, and Ca001 (panel genotype) was 107 and 88 cm, respectively. These results revealed the high genetic variation for plant height among genotypes and the variable effects of drought stress on camelina. Some genotypes showed stability among the stress treatments (e.g., Celina showed 105 cm height under both well-watered and water-limited conditions). The heritability estimate for plant height was 0.97, indicating the possibility of a few genes controlling the trait. Flowering time of the panel ranged from 85 to 121 days after planting. Even though there were high significant variations among genotypes within each stress condition, there were no significant differences between stressed vs. non-stressed conditions nor GxE interaction. Preliminary observations showed variations in seed yield among genotypes and stress conditions, and final analyses will be reported in the coming year. In the prior reporting year, ten commercial varieties of camelina were planted at The High Plains Agriculture station near Sidney, NE. The lines were planted in six replicates for a well-watered treatment and six replicates for a water stressed treatment. Stand counts were also collected, which allowed for normalization of plot yield to number of plants in the plots. A mobile phenotyping system was designed and constructed to measure three plots at a time. The sensor modules included RGB cameras, ultrasonic height sensors, thermal infrared radiometers, and portable spectrometers. Two phenotyping surveys were conducted (6/16/2017 and 7/3/2017). Both HTP and yield data were collected. The analysis of yield data indicated that the water stressed plots yielded between 1 - 26% less than the well watered plots and validated that the environment in Sidney, NE was adequate to induce a drought decrease in yield. The ten varieties also showed a range of tolerance to water stress, providing confidence that variation in tolerance to drought would be found in the collections being screened. Phenotyping efforts also showed interesting results, including a correlation between green pixel fraction (GPF) and yield, which was highly significant and could in future be used as a predictor of yield. In late April of 2018, three reps and two treatments of 234 accessions of camelina were planted in Sidney, NE. The experiment was phenotyped four times using the phenocart, which in this year (2018) included a 3D scanning Lidar for estimating height. The collected data and traits will be used in future GWAS mapping studies, but due to seed shatter one week prior to harvest, the seed yield data were lost due to a hailstorm. A non-destructive HTP experiment was carried out on a LemnaTec Scanalyzer 3-D available at the Danforth Center using 285 genotypes of camelina that included 4 plants of each genotype. In this experiment, drought tolerance was tested with plants grown in duplicate under well-watered (80% full-capacity) and water-limited (40% full-capacity) conditions for 4.5 weeks (31 days). RGB images that allow visualization and quantification of plant color and structural morphology, such as leaf area, stem diameter and plant height, and NIR images that enable visualization of water content in plants in the near infrared spectrum of 900-1700 nm were collected for all plants daily for the entire 4.5 week period.The camelina plants were then transferred to a greenhouse to complete their life cycle where drought conditions were continued on the appropriate plants. 843 plants matured to the point of seed collection, including 271 genotypes. Seeds were harvested, cleaned, and weighed to determine yield. Seeds were also imaged to record variation in seed dimension for each genotype under well-watered and drought conditions. Seeds were then analyzed by bench top NMR to determine % oil content. Genotypes having at a minimum of 400 mg seeds were then analyzed by NIR to determine the percent content of chlorophyll, fatty acids, glucosinolates, moisture, nitrogen, and oil. Ground truthing was performed on camelina genotypes having high, medium, and low percentages of linolenic acid (18:3) and glucosinolates to validate the NIR data. These ground truthing analyses confirmed previous reports that the NIR data correlates well under the pretense of high, medium, and low amounts (qualitative) of glucosinolates and linolenic acid, but is not quantitative. Goal 2: The population structure and extent of genetic diversity in a spring camelina panel were characterized based on camelina GBS data provided by the Danforth Center. Data for heterozygosity and polymorphism information content (PIC) were collected. Population structure data were generated using the STRUCTURE algorithm. Two subpopulations were identified based on principal coordinates analysis, which was in agreement with geographical classification of spring camelina panel and STRUCTURE results. Analysis of molecular variance suggested that the majority of total variation (96%) was found within subpopulations while only a small portion (4%) was accounted for among subpopulation variation. Basic GWAS pipelines using GAPIT and TASSEL were established. GWAS analyses were subsequently performed using GBS data and the completed phenotypic evaluations for oil content and fatty acid composition, derived from the camelina panel grown at the Danforth Center, and leaf waxes, obtained from the camelina panel grown in Maricopa. GWAS analyses for fatty acid composition identified 26 SNP markers that were significantly associated with different fatty acid compositions under stress conditions. Under stress conditions, a significant SNP marker was identified and located on camelina chromosome 8 within the coding region of NAC domain-containing protein 78, which is known to play an important role in regulating plant response to drought stress. This marker accounted for approximately 13% and 16% of variation for oleic and linolenic acids, respectively. The diversity panel exhibited a wide range in total leaf wax contents, wax classes and constituents. GWAS revealed a total of 42 significant SNP markers that are putatively associated with wax-related traits, in which one significant marker was identified in each of three wax-related traits. The putative functions of these 42 SNP markers can be clustered into three major categories: peroxisome movement, defense in pathogen attack and repression of photomorphogenesis. Goal 3: Four transgenic terpene producing camelina lines were grown under well watered and water limited conditions in the greenhouse facility at the Danforth Center, then analyzed for terpene content. The results indicated that terpene production was not greatly affected by drought conditions. The lines included two transgenic camelina lines with genotype suneson and two with genotype licalla, each genotype having a (+)-δ-cadenine producing line and a S-(-)-limonene producing line.

Publications

• Type: Journal Articles Status: Published Year Published: 2018 Citation: Tomasi, P., Dyer, J.M., Jenks, M.A., Abdel-Haleem, H. 2018. Phenotypic variations in leaf cuticular wax classes and constituents in a spring Camelina sativa diversity panel. Industrial Crops and Products. 112:247-251.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Thompson, A.L., Thorp, K.R., Conley, M.A., Andrade-Sanchez, P., Heun, J.T., Dyer, J.M., White, J.W. 2018. Deploying a proximal sensing cart to identify drought-adaptive traits in upland cotton for high-throughput phenotyping. Frontiers in Plant Science. 9:507.

Progress 09/01/16 to 08/31/17

Outputs
Target Audience:Our research program aims to develop and use advanced genomic and phenomic tools to improve camelina as a biofuel crop, with the end-goal of identifying genes and molecular markers associated with increased drought tolerance, and identifying specific camelina lines that are adapted to different regions of the country. Our work directly impacts several target groups, the first of which includes farmers and producers who are interested in growing camelina as either a winter rotation crop, or a spring-grown crop, particularly in arid climates. The second group includes scientists and breeders who are focused on understanding the basic mechanisms of drought tolerance in crop plants, with the goal of improving crop performance using molecular breeding approaches. The third group includes scientists who are developing sensor-based technologies for monitoring crop growth and performance in the field, with applications in both precision agriculture and phenomics-based crop improvement. This latter group is multidisciplinary in nature and includes mechanical, electrical, and electronics engineers, computational scientists, plant breeders, and plant physiologists, who often work in large collaborative teams. A major challenge is the collection, storage, processing, and analysis of the "big data" associated with these studies, and converting sensor-based data in biologically useful information. This area of research, generally referred to as "high-throughput phenotyping," not only shows great potential for increasing crop yields through improved precision agriculture, but also for accelerating genomics-based crop improvement. Thus, the final target group includes the general public and all other users and consumers of agricultural products. Changes/Problems:Our NIFA grant had a start date of Sept. 1, 2016, but due to several delays and administrative issues, we did not obtain funding until early November, 2016. Partners did not receive their sub-awards until approximately March, 2017. This caused some delays in hiring, particularly for the Maricopa location, which was also subject to a government-wide hiring freeze. The inability to hire a post-doc in Maricopa has negatively impacted our spending rate, but the freeze has recently been lifted, and we are currently filling the post-doc position. Notably, we utilized as many in-house resources as possible to support the research this past fall, and made significant progress despite the vacancy. Because of the large amount of work involved and the logistics of managing such large numbers of plots, the Nebraska group plans to phenotype half (125 accessions) of the camelina collection next year and the other half in year three. In the end, the same amount of work will be done, but the study will not be replicated across two seasons. What opportunities for training and professional development has the project provided?In Maricopa, undergraduates and graduate students had the opportunity to learn about camelina, measurement of agronomic traits and the recording and processing of yield and HTP data. Students have also had the opportunity to work on improvements and optimization of data pipelines, including the development of an HTP database using open source SQL server software and usage of Python and GIS for geo-referencing. Additional protocols were developed for assessing the quality of sensor data, and work is continuing on the development of graphical user interfaces. The end goal is to distribute these open source tools to the broader scientific community. In Nebraska, both undergraduates and postdocs had extensive opportunities to understand the agronomic practices involved in small seeded crop production. Postdocs have furthered their knowledge in how to use the phenotyping cart to analyze the phenome of the crop and will be engaged in helping to solve some of the problems associated with phenotyping a crop with small leaves and only moderate canopy cover. How have the results been disseminated to communities of interest?The phenotyping group in Maricopa, Arizona actively participates in an NSF-funded workshop, provided twice a year in Maricopa, to help educate other scientists, students, and industry partners on the practical uses and applications of field-based, high-throughput phenotyping. While this is not a direct goal of our NIFA-funded project, our core HTP group is actively involved in promoting the technology with the broader scientific community. This workshop also provides a potential forum for highlighting results of our camelina study in the future. What do you plan to do during the next reporting period to accomplish the goals?In Maricopa, the entire camelina panel (250 lines) will be planted under well-watered and water-limited conditions in the fall of 2017, and both HTP-related and traditional traits will be measured. These activities will be conducted in support of goal #2. Initial statistical and GWAS analyses will be conducted and reported. In Nebraska, 125 accessions of Camelina will be planted under well watered and water stressed conditions and phenotypes will be measured to assess the relative drought tolerance of each accession.

Impacts
What was accomplished under these goals? Our research program aims to develop and utilize high-throughput phenotyping technologies, coupled with genomics-based approaches, to discover useful genes/alleles controlling seed yield and oil content and quality in camelina under water-limited conditions. Briefly, a panel of 250 camelina lines will be cultivated under well-watered and water-limited conditions in fields located in Arizona and Nebraska and analyzed using complimentary high-throughput phenotyping (HTP) approaches. The panel will also be grown and analyzed using a LemnaTec-based greenhouse facility available at the Danforth Center. Three standardized phenotypes will be measured including canopy reflectance, canopy temperature, and crop height. Traditional phenotypic data, including seed yield components, will also be determined. The camelina panel will be genotyped using genotyping-by-sequencing, then genome-wide association studies will be used to discover SNP markers associated with trait data. By comparing results obtained from field-grown and greenhouse-grown plants, we will also address the question of how well greenhouse-based studies translate to discovery of genes related to performance in the field. Lastly, we will conduct advanced yield trails by growing 10 to 20 select plant lines in four different locations including Arizona, Nebraska, Florida, and Minnesota. Collectively, our research project will accelerate the adoption of camelina as a non-food biofuel crop and provide robust high-throughput phenotpying tools that can be used to improve a broad array of food and non-food crops. Goal 1: Develop and apply automated, non-destructive high-throughput phenotyping (HTP) protocols to evaluate the phenotypic diversity and stress tolerance of a camelina panel consisting of 250 accessions, grown under well-watered and water-limited conditions. MARICOPA, AZ A high-clearance tractor was equipped with proximal sensors and imagers to collect field-based HTP (FB-HTP) data including crop height, canopy multi-spectral reflectances, and canopy temperature. To develop and validate an HTP protocol suitable for detecting the variation in the camelina diversity panel, ten camelina varieties (with maximal trait variation) were planted in a randomized complete block design, with 3 replicates, during the fall/winter 2016-2017. FB-HTP data were collected on a weekly basis (sixteen runs total) and data were stored, processed and analyzed using custom data pipelines. Traditional morphological data including plant height, flowering time, and chlorophyll content were also collected. At physiological maturities, plots were harvested and seed yields determined. The entire camelina panel (250 accessions) was also planted for seed increase, and flowering times were recorded to help in blocking accessions with similar flowering time in future experiments (Goals 2 and 3). Fifteen HTP-related traits were collected and analyzed. Reflectance data were used to construct vegetation indices including: Normalized difference vegetation index-red (NDVIR), Normalized difference vegetation index-amber (NDVIA), Physiological reflectance index (PRI), Normalized difference vegetation index-red-red edge (NDRRE), Normalized difference vegetation index-amber-red edge (NDARE), Normalized difference red edge index (NDRE), Canopy chlorophyll content index (CCCI), DATT index, Meris terrestrial chlorophyll index (MTCI) and Chlorophyll index vegetation index (CIRE). Analyses of HTP data showed a temporal increase in trait values during the growing season. For instance, NDVIA mean over the 10 varieties ranged from 0.243 (run 1) to 0.339 (run 9) to 0.477 (run 15), then dropped to 0.3275 at run 16. Plant height also showed a temporal increase with increasing camelina age. A drop in various values after run 15 could be explained by plant senescence, desiccation and partial plant lodging. Within a certain run, there were variations among the tested varieties, e.g., ANOVA analyses showed significant variation in NDVIA values of 0.5631 for the cultivar Midas compared to 0.3668 for cultivar Galena (run13). For plant height, Celine was significantly taller (87.45 cm) than other camelina varieties, while Robinson was shorter (64.72 cm) (LSD0.05=14.27 cm). HTP data were also compared to traditional phenotypic data, which revealed significant correlations between sensor-measured plant height and hand measurements (r=0.53604; Prob < 0.0048), and correlations between chlorophyll content and CIRE (r=0.458), DATTA (r=0.368), DATT (r=0.348), NDARE (r=0.426), NDRRE (r=0.390), NDRE (r=0.413), NDVIA (r=0.405) and NDVIR (r=0.422). With the significant correlations observed, the FB-HTP protocol will be useful for characterizing the genetic and phenotypic variation in the camelina panel in years two and three of the project. UNIVERSITY OF NEBRASKA The ten varieties of camelina were also planted at The High Plains Agriculture station near Sidney, Nebraska, on April 24, 2017. The lines were planted in 6 replicates for a well-watered treatment and 6 replicates for a water stressed treatment. Stand counts were also collected, which will allow for normalization of plot yield to number of plants in the plots. A mobile phenotyping system was designed and constructed to measure three camelina plots at a time. The sensor modules included RGB cameras, ultrasonic height sensors, thermal infrared radiometers, and portable spectrometers. Two phenotyping surveys were conducted (6/16/2017 and 7/3/2017). It took the team roughly 45 min to survey 100 field plots. The phenotyping system functioned robustly and reliably in the field, and no mechanical or electronic problems were encountered. One problem identified is that the ultrasonic height sensors did not measure the camelina height properly. To tackle this problem, the team is working to incorporate a 2D laser scanner into the system. Due to its scanning nature, the laser scanner can measure height at many points in the canopy. We are also working with the group in Maricopa to determine whether changes in the data capture or data processing methods might be useful for improving sonic-based estimations of canopy height. The cultivation of camelina in the dry climate of western NE was established and phenotyping equipment tested. In August we will have yield data that we will adjust by stand counts for the further analysis of difference in productivity due to different water conditions. DANFORTH CENTER A non-destructive HTP experiment on a LemnaTec Scanalyzer 3-D has been carried out on the 250 genotypes of camelina. In this experiment, drought tolerance was tested with plants grown in duplicate under well-watered (80% full-capacity) and water-limited (40% full-capacity) conditions for 4.5 weeks. The camelina plants were then transferred to a greenhouse to complete their life cycle. The plants are currently senescing and seeds are maturing. Seeds have been harvested from 30% of the plants to date. Seed yield, oil content and fatty acid analysis will begin upon completion of harvest. Red-Green-Blue (RGB) images that allow visualization and quantification of plant color and structural morphology, such as leaf area, stem diameter and plant height, and NIR images that enable visualization of water content in plants in the near infrared spectrum of 900-1700 nm were collected for all plants daily for 4.5 weeks.Collected seed will be imaged to record variation in seed dimension for each genotype under well-watered and drought conditions. Data analysis is currently underway. Goal 2: Discover alleles/genes controlling morphological, physiological, seed, and oil yield properties using genome-wide association studies (GWAS). No work conducted yet. Goal 3: Identify, test, and validate useful germplasm, including transgenic lines producing drop-in ready jet fuel components, under diverse environments and marginal production areas. No work conducted yet.

Publications