speaker abstracts - plant · 2/11/2017 · that otherwise cannot be quantified by physical...
TRANSCRIPT
Speaker Abstracts
General Session: Phenomic Insights into Quantitative Traits February 11, 2017 | 8:00 AM – 12:15 PM
Increasing Yield Prediction Accuracy for Wheat Breeding through Combined Genomic Selection and Field-based High-throughput Phenotyping Jesse Poland, Kansas State University
We are combining genomic information and plant physiology measurements from high-throughput
phenotyping of canopy temperature and spectral reflectance to predict grain yield in small plot trials in
wheat breeding programs. Using combined multivariate models we can show a marked increase in the
prediction accuracy by combining genomic and phenomic information from correlated secondary traits.
When implemented across the breeding program, this becomes a powerful selection tool for breeders.
Phenotypes Phirst: What Genes Control Plant Responses to Environments? Tom Mitchell-Olds, Duke University
It is now feasible to dissect the molecular basis of genotype x environment interactions and to identify
the genes and pathways that control plant responses to environments. We report analyses of large,
multi-environment experiments from rice, maize, and soybean, and discuss genetic control of yield
tradeoffs across environments.
Role of High-throughput Phenotyping Tools to Increase Breeding Efficiency in Maize
Natalia de Leon, University of Wisconsin-Madison
Yield is the single most relevant measurement of a crop plant but, by its integration, this metrics does
not provide important details about the mechanism by which the specific genetic constitution of that
plant is able to achieve that end of the season response. Recent advances in plant high-throughput
phenotyping are expected to enhance current capabilities to dissect the genetic architecture of
economically relevant traits and integrate information from relevant related traits into prediction
models to improve breeding efficiency. Using maize as a model species, this presentation will focus on
describing examples of field based phenotyping tools able to assess plant morphological characteristics
that otherwise cannot be quantified by physical measurements and discuss their potential relevance for
productivity.
Uncovering the Maize Pan-genome to Identify Loci Associated with Relevant Agronomic Phenotypic
Traits
C Robin Buell, Michigan State University
Extensive genome variation exists between individual maize accessions. Using a 959 accession panel of
maize inbreds, we are cataloguing presence/absence variation to permit associations with phenotypes
related to biofuel feedstock traits.
Engineering Biomass for the Billion-ton Bioeconomy
Maureen McCann, Purdue University
By exploiting natural genetic variation, genetic engineering and synthetic biology approaches, the
biofuels research community is poised to enable use of a national resource of over 1B tons of
lignocellulosic biomass for fuels, chemicals and materials. In our Center for Direct Catalytic Conversion
of Biomass to Biofuels, we are re-designing cell wall architecture and biomass structure to optimize
recovery of liquid hydrocarbon fuels and high-value co-products.
Metabolome-wide Association Studies in Arabidopsis, Wheat and Hop Using Central, Specialized and
Lipid Metabolomes
David Riewe, IPK-Gatersleben
Janine Wiebach, IPK-Gatersleben; Alexander Feiner, Hopsteiner; Manuela Nagel, IPK-Gatersleben; Paul
Matthews, Hopsteiner; Andreas Boerner, IPK-Gatersleben; Thomas Altmann, IPK-Gatersleben
The metabolome is typically regarded as the penultimate phenotypic expression leading to variation in
such complex traits like germination, growth, or pathogen resistance. Studying the relationship between
metabolite levels and trait expression allows the identification of the underlying molecular processes.
Using large untargeted metabolomics datasets generated by gas and liquid chromatography hyphenated
to (high resolution) mass spectrometry, we identified known and unknown metabolites correlated to
agronomic important traits like seed viability, plant growth and downy mildew resistance in different
plant species. By combining genomic, transcriptomic, enzymatic, metabolomic and phenomic data from
different Arabidopsis thaliana populations, we could link a promotor polymorphism in the FUMARASE2
gene (FUM2) with control over malate-to fumarate conversion and biomass production at every step
from genome to phenome (Riewe et al., 2016). 90 wheat (Triticum aestivum) seed stocks stored for five
to fifteen years at different temperatures at the German Federal in-situ genebank were subjected to
seed germination assays and untargeted profiling of central and (oxidized) lipid metabolites. Using an
accurate mass based annotation strategy and high resolution LC-MSMS spectra at high coverage, we
identified hundreds of unmodified an (multiply) oxidized lipids. Correlation analysis revealed that lipid
oxidation and hydrolysis was found to be the most closely related phenomenon to coincide with losses
in germination frequency. In an F1 hop (Humulus lupulus) population with 192 individuals with variation
in downy mildew resistance, inoculation with Pseudoperonospora humuli led to the production of a
large number of specialized metabolites within 48 hours after inoculation. Phenotypic expression of
resistance of the F1 population was scored five days after inoculation. Using linear regression, a small
number of metabolites with potential protective function against downy mildew were identified and
mapped to the phenylpropanoid pathway.
High Throughput Phenotyping of Photosynthesis to Identify Relevant Quantitative Trait Loci and
Underlying Genes in Arabidopsis thaliana
Mark Aarts, Lab of Genetics, Wageningen University
Roxanne van Rooijen, Heinrich Heine University; Aina Prinzenberg, Wageningen University; Jeremy
Harbinson, Wageningen University
Natural genetic variation in plant photosynthesis efficiency is hardly studied even though breeding for
photosynthesis would be interesting to maintain increases in crop yields. One of the reasons this is not
studied is the notorious difficulty in adequately phenotyping photosynthesis parameters for genetic
research. We designed a versatile plant phenotyper, the Phenovator, capable of multispectral imaging of
1440 plants per experiment. Imaging includes chlorophyll fluorescence to determine the light use
efficiency of photosystem II electron transport (ΦPSII or Fq’/Fm’), and near infrared reflection to
measure projected leaf area. This system has been used to phenotype 350 genetically diverse
Arabidopsis thaliana accessions for genome wide association analysis. Plants were phenotyped at non-
stressful conditions and in response to cold treatment (5 °C) or a switch from 100 to 550 µmol m-2 s-1
irradiance. The observed genotypic variation was used to identify quantitative trait loci (QTL). Our latest
progress in identification and confirmation of QTL for ΦPSII, plant growth and epinastic leaf
movement will be presented, including detailed analysis of the allelic variation at the nucleotide level of
some of the genes underlying these QTL. Our work has shown that there is sufficient genetic variation
for PSII efficiency in Arabidopsis amenable for gene identification, which suggests the same will be the
case for crop species. Such would offer interesting opportunities for future crop photosynthesis
improvements.
Time Dependent Genetic Analysis Links Field and Controlled Environment Phenotypes in the Model C4
Grass Setaria
Max Feldman, Donald Danforth Plant Science Center
Rachel Paul, University of Illinois; Darshi Banan, University of Illinois; Jennifer Barrett, Donald Danforth
Plant Science Center; Jose Sebastian, Carnegie Institute for Science; Muh-Ching Yee, Carnegie Institute
for Science; Hui Jiang, Donald Danforth Plant Science Center; Alexander Lipka, University of Illinois; Tom
Brutnell, Donald Danforth Plant Science Center; Jose Dinneny, Carnegie Institute for Science; Andrew
Leakey, University of Illinois; Ivan Baxter, USDA – ARS
Vertical growth of plants is a dynamic process, which is influenced by both genetic and environmental
factors. Variation in plant height presents an intriguing ecological trade-off between access to solar
radiation, energetic demand for construction of lignified support tissue and the fecundity risk associated
with lodging. As such, plant height has a profound effect on overall plant architecture and biomass
composition. The primary goal of our studies have been to identify the major genetic loci that contribute
to the variation of plant height within two independent genetically structured populations of Setaria sp.
and how such phenotypes change temporally in response to differences in water availability and
planting density. To achieve this objective we have performed twelve growth trials of an interspecific S.
italica x S. viridis recombinant inbred line population in field and controlled environment settings in
addition to two experimental growth trials of a S. viridis natural diversity panel in the Bellweather
Phenotyping Facility at Donald Danforth Plant Science Center. Results from these studies identify
numerous genetic loci associated with variation in plant height shared across all experimental
treatments and locations but that the contribution of individual loci differs temporally. The
methodology presented in this study is applicable to any number of single or composite traits and is
used to deconvolute the contribution of plant height to other complex plant traits of agronomic
importance (biomass accumulation/composition and water use efficiency).
Emerging Technologies in Imaging
February 11, 2017 | 2:00 PM – 5:30 PM
Integrated Whole-Plant Fluxomics Aided by Imaging Technologies
Richard Ferrieri, University of Missouri
This talk will highlight some of the recent developments using short-lived radioactive isotopes coupled
with dynamic whole-plant Positron Emission Tomography (PET), Root Radiography, and Radiometabolite
Flux Analysis, as well as using stable isotopes coupled with MS-based imaging technologies to explore
the dynamic relationships that exist belowground between living plants and the surrounding
rhizosphere.
Advances in the Deployment and Integration of Imaging Sensors for HTP Plant Phenotyping
Xavier Sirault, CSIRO
From autonomous ground platforms combining multi-modality sensors to deploying sensors in nano-
satellites, the development of robotics and advanced non-invasive imaging technologies has opened up
new avenues for high-throughput screening of plants across scales.
This presentation aims to review the potential contribution of proximal and remote imaging to
phenotyping research but also to highlight that as with any new phenotyping technology, turning raw
data into biological information for real-world decision support outcomes remains key to realising the
adoption of these tools by the agricultural sector and the research community.
There’s a World Going on Underground: Imaging Technologies to Understand Root Growth Dynamics
and Rhizosphere Interactions
Chris Topp, Donald Danforth Plant Science Center
Root systems are the foundation of plant health and productivity, but are often ignored in plant
research. I’ll be presenting state-of-the-art technologies and approaches for understanding the “hidden-
half”.
Computed Tomography at the Fraunhofer EZRT Applied for Belowground Phenotyping
Stefan Gerth, Fraunhofer Institute for Integrated Circuits IIS – Development Center for X-ray Technology
EZRT
The Fraunhofer Development Center X-ray technology (EZRT) drives the technological development and
the application of Computed Tomography in a wide range of research areas. Within the presentation we
will demonstrate how the non-destructive nature of 3D data can be used for belowground phenotyping
for tubers, storage roots and maize root systems.
Synchrotron X-ray Fluorescence Imaging as a High-Precision Phenotyping Tool for Studies of the
Relationship between Mineral Nutrient Homeostasis, Fertility and Crop Yield
Olena Vatamaniuk, Cornell University
Ju-Chen Chia, Cornell University; Huajin Sheng, Cornell University; Jiapei Yan, Cornell University; Rong
Huang, Cornell High Energy Synchrotron Source; Louisa Smieska, Cornell High Energy Synchrotron
Source; Arthur Woll, Cornell High Energy Synchrotron Source
Transition metals such as iron (Fe), zinc (Zn) and copper (Cu) are required for the growth and
development of plants. These elements are also essential components of the human diet. In fact, Fe-
deficiency anemia affects over 30% of the world’s population, especially in poor populations across the
globe. While Cu is not as limiting in human diets as Fe, Cu interacts with Fe and is required for plant
fertility and seed/grain set. While the global demand for high-yield crops with a high density of Fe, Zn,
and Cu in edible tissues is increasing, limited Fe, Zn and Cu bioavailability in many agricultural soils
worldwide significantly impedes crop production, their mineral quality and thus, the future of food
security. Sustainable approaches for improving crop yield and mineral nutrient density relies on the
knowledge of Fe, Cu and Zn transport systems and their regulators that ensure adequate mineral uptake
into roots and root-to-shoot partitioning. In this regard, high-precision synchrotron x-ray fluorescent
microscopy (XRM)- based mapping of the spatial distribution of mineral elements at the tissue, cellular
and subcellular resolution in different plant genotypes emerges as a powerful tool connecting the
genotype-to-phenotype gap as pertains to improving mineral nutrient uptake, delivery to reproductive
organs and loading into grains. We will present our recent data showing the power of the integration of
XRM imaging with functional genomics tools for uncovering the function of transporters and
transcription factors in shoot-to-root signaling of iron Fe status and the role of Cu in male fertility in a
model plant Arabidopsis thaliana. We will also discuss the advantages of including this phenotypic
strategy as a gene mining and QTL-identification tool in molecular breeding efforts for crops with
improved growth and yield/spike characteristics in marginal soils and soils now in cultivation.
Graft Junction Formation: A Dynamic Investigation Using PET and Confocal Imaging
Margaret Frank, Donald Danforth Plant Science Center
Sergey Komarov, Mallinckrodt Institute of Radiology at Washington University School of Medicine; Qiang
Wang, Nevis Laboratories, Columbia University; Ke Li, Department of Electrical and Systems Engineering,
Washington University in Saint Louis; Halley Fowler, University of Missouri - Saint Louis; Viktoriya
Coneva, Donald Danforth Plant Science Center; Daniel Chitwood, Donald Danforth Plant Science Center;
Yuan-Chuan Tai, Mallinckrodt Institute of Radiology, Washington University School of Medicine
Grafting, a technique that involves the physical joining of roots and shoots from independent genotypes,
allows for desirable traits from disparate species to be harnessed within a single organism. This simple
yet effective strategy is used around the globe to combat a wide sweep of biotic and abiotic threats that
would otherwise have crippling effects on crop production. Despite the tremendous advantages that
grafting provides for crop protection and enhanced crop yield, successful graft combinations are still
identified on an inefficient, case-by-case basis. This is largely due to our present lack of understanding
regarding the fundamental mechanisms underlying the success of grafting as a technique. Here, we
focus on formation of the graft junction – a unique anatomical region that unites newly joined root and
shoot systems into a single vascular conduit, and thus plays a pivotal role in determining the functional
success of root-shoot combinations during grafting. We have developed an integrated imaging approach
using positron emission tomography (PET), and laser scanning confocal microscopy, which allows for in
vivo tracking of vascular conductance to be combined with high resolution anatomical reconstructions
that follow the cellular processes guiding graft junction formation. The synthesis of these two datasets
into a unified model for graft junction formation is being used to build a fundamental understanding of
the physiological and anatomical coordination that underlies this complex biological process. As a
preliminary result, we have established that 11CO2 is a dynamic and reliable marker for shoot-to-root
phloem conductance. Using this radionuclide, we demonstrate that phloem function is restored within
5-6 days post-grafting, and that this functional restoration coincides with the first anatomical
appearance of elongated phloem sieve cells within the junction region. The extension of this functional
framework to problematic grafting combinations may help elucidate the rules that govern successful
graft combinations.
Emerging Technologies in Quantitative Field Data
February 11, 2017 | 2:00 PM – 5:30 PM
Automated, High-Throughput, Time-Lapse Photography for Field-based Maize Phenomics
Patrick Schnable, Iowa State University
Networked systems of hundreds of computer-controlled cameras that enable high-throughput, high-
resolution, field-based time-lapse photography for studying maize growth/development and responses
to environmental stresses have been deployed at four locations over two years.
From Plots to Plants to Organs
James Schnable, University of Nebraska-Lincoln
Research at the University of Nebraska-Lincoln seeks to address three specific challenges in plant
phenotyping. Firstly, disambiguating data from the plot level to individual plants, and individual organs
of individual plants. Here computer-vision algorithms designed with an awareness of how corn plants
grow and develop can track individual leaves from one day to the next. Secondly, most new phenotyping
approaches require calibration using large validation datasets. As a result, in the short term, adopting
high throughput phenotyping methods can actually result in drastic increases in demand for manual
phenotyping measurements. We are developing a robotic platform which can identify individual leaves
and clamp a sensor package to the leave, enabling the automated collection of “ground truth” data. The
final challenge concerns the selection of appropriate numerical representations of phenotypes extracted
from image data. Early efforts in this area have focused on replicating transitional phenotypes which
have been manually measured under field conditions in the past. However, image-based phenotyping
also creates the possibility to identify and study the genetic basis of variation novel phenotypes which
would not be practical to measure manually, but which may be linked to agronomic performance under
field conditions. In order to stimulate algorithm development and statistical analysis in this area, we are
releasing high throughput RGB and hyperspectral image datasets collected from many of the same
genotypes used by the Genomes 2 Fields project (available at http://plantvision.unl.edu/dataset).
Quantifying Basic and Complex Plant Traits at the Tissue-, Plant-, and Canopy-Level Scales of
Organization: Theory, Results, and Applications
April Agee Carroll, Purdue University
Plant phenomics seeks to predict crop performance in context of plant genetic potential and environmental conditions. Grain yield at harvest across multiple locations and years will continue to be a gold standard measurement, but there is clear benefit and high demand for surrogate assays that can predict yield as quickly and accurately as possible, with less resources invested. This continuum forces compromise between accuracy of yield estimation, control over environmental conditions, and speed to actionable information. Assays that isolate different genetics under highly consistent environmental conditions are of interest, but by design controlled environment experiments do not replicate all conditions in the field, and to be predictive must be modeled against validated measurements. With improved sensing systems and computational methods, our group aims to increase the resolution of field measurements from the canopy scale to the individual plant level, and to inform and be informed by models created through controlled environments studies. This multi-scale translational approach reduces the need for compromise by generating reliable methods for early prediction of yield in the field and under controlled conditions. Results will be presented quantifying plant-to-plant variability with Unmanned Aerial Vehicles (UAVs), the PhenoRover, a mobile ground-based phenotyping platform, RGB, hyperspectral, and LiDAR sensors, ground reference data, novel georeferencing techniques, and various image and data analysis methods such as feature extraction and image segmentation. Such information provides insights into research plot and field quality, farm equipment performance, environmental stress tolerance, physiological plasticity, and fine-scale spatial variability. Challenges from phenotyping at the individual plant level using traditional and emerging techniques will also be discussed.
Development of a Ground Vehicle and a Mobile Observation Tower to Integrate Plant Phenotyping
from Plant Organ to the Canopy Level
Felix Fritschi, University of Missouri
A semi-autonomous ground vehicle and a mobile observation tower were developed and deployed in
the field in 2016. The mobile tower was equipped with a 360-degree robotic vision system to monitor a
large area of a field, primarily to obtain canopy-level information in at regular intervals over the course
of the growing season. The ground vehicle was deployed in semiautonomous fashion and includes a
robotic arm that can handle a broad range of sensors for more detailed, plant and organ level
characterization. This presentation will focus on describing the two platforms and will include the
presentation of preliminary results.
A High Throughput Leaf Gas Exchange Method Identified QTLs Responsible for Photosynthesis and
Stomatal Conductance Sensitivity to Ozone in Maize
Timothy Wertin, University of Illinois
Nicole Choquette, University of Illinois; Lauren McIntyre, University of Florida; Patrick Brown, University
of Illinois; Elizabeth Ainsworth, University of Illinois; Andrew Leakey, University of Illinois
An important challenge to current and future maize production is tropospheric ozone, a damaging air
pollutant which could further increase by 10-30 ppb by the end of this century. Currently, ozone is
estimated to cause losses of $14-26 billion to global agricultural production, including reducing U.S.
maize yields by ~10%. Ozone enters leaves through the stomata where it rapidly reacts to form other
reactive oxygen species that can overwhelm the detoxification potential of the leaf to impact physiology
and development. Signs of ozone stress include reduced photosynthesis and stomatal conductance. In
this study, we used high throughput leaf gas exchange to study the effects of elevated ozone
concentration (100 ppb) on photosynthesis and stomatal conductance in populations of hybrid and
inbred maize. A half diallel population (45 hybrids) made from 10 diverse maize inbreds previously
shown to have variable responses to elevated ozone was used to test trait heritability. A population of
100 near isogenic lines introgressing B73 and Mo17 was used to map loci associated with trait variation
and ozone sensitivity. Both populations were grown at ambient (~40 ppb) and elevated ozone (100 ppb)
in the field at a Free Air Concentration Enrichment (FACE) facility at the University of Illinois, Urbana-
Champaign (n=4). Leaves were cut early in the morning, transported to the lab and kept in a controlled
environment before light-saturated photosynthesis and stomatal conductance were measured. This
protocol was used to measure 2560 leaves over a 10-day period. Photosynthesis and stomatal
conductance were both highly heritable (H2 approximately 0.65). Distinct quantitative trait loci were
detected for response of both photosynthesis and stomatal conductance to O3 (α=0.05).Our
study describes a new method to rapidly measure leaf gas exchange on a large number of samples,
while the findings suggest that leaf gas exchange is under strong genetic control.
High-Throughput Phenotyping of Chenopodium Quinoa from Seed to Shoot
Malia Gehan, Donald Danforth Plant Science Center
Steven Callen, Donald Danforth Plant Science Center; Elizabeth Castillo, Donald Danforth Plant Science
Center; Jose Tovar, Donald Danforth Plant Science Center; Michael Miller, University of Nebraska,
Lincoln; Monica Tessman, Milwaukee School of Engineering; Willam Kezele, Donald Danforth Plant
Science Center
Chenopodium quinoa (quinoa) is a highly nutritious grain and vegetable crop reported to be highly
tolerant of salinity, with certain varieties able to grow with salt concentrations comparable to sea water.
The crop is also highly drought tolerant, cold tolerant, and can grow on rocky/sandy soil media making it
ideal for growth in arid and cold regions where crops aren’t typically produced. Despite quinoa’s many
advantages as an alternative crop, there are several challenges to its viability as a global crop. The
lifecycle of quinoa is long, ranging from 120 to 210 days, making the crop uneconomical to grow in some
contexts. Therefore, it is key to identify and develop fast maturing strains that can still withstand and
thrive on marginal land. However, a major bottleneck in plant science is the ability to efficiently and
non-destructively quantify plant traits (phenotypes) through time. This research develops high-
throughput phenotyping technologies and open-source platform-independent analysis tools (PlantCV;
http://plantcv.danforthcenter.org/) to quantify quinoa seed, and shoot growth through time. In a pilot
study, we quantify seed characteristics for 400 quinoa accessions, and examine the fresh weight, height,
color, growth rate, and yield for a subset of accessions in replicate over time.
Emerging Technologies in Robotics and Remote Sensing
February 11, 2017 | 2:00 PM – 5:30 PM
Robot-Assisted Phenotyping in the Wild
Joshua Peschel, Iowa State University
This talk will focus on the field deployment of a small, low-cost ground robot, TERRA-MEPP (Mobile
Energy-Crop Phenotyping Platform), to support the high-throughput characterization of agricultural row
crops. Current measurement approaches, such as unmanned aerial vehicles and automated LiDAR, are
restrictive and costly, leading to potentially suboptimal efficincies for quantifying the sub-daily growth of
crops over large areas. Experimental results will be presented to elucidate lessons learned and provide
recommendations for designing and deploying robotic technologies for agricultural field-based domains.
Mapping Dynamic Seedling Phenotypes to the Maize and Arabidopsis Genomes by Automated Image
Analysis
Edgar Spalding, University of Wisconsin-Madison
Automated image capture and analysis enables phenotypes to be treated as developmental time
courses rather than static traits. When scaled up to population-level studies, quantitative trait loci maps
with a developmental time axis are produced.
Semantic Mapping of Farms with Heterogenous Robotic Systems
Volkan Isler, University of Minnesota
In this talk, Volkan will present an overview of our work on building aerial, ground and surface robots
which can collect multi-modal data over multiple spatial scales. I will also present an overview of our
efforts to extract semantic information from this data.
Innovations toward Robotic Plant Phenotyping
Lie Tang, Iowa State University
In this presentation, Lie will emphasize their critical needs in enabling technologies and tools for high-
quality and high-throughput plant phenotyping, and showcase how innovations in automation and
robotics can open up new avenues toward that goal. He will convey to the audience that, to realize
robotic phenotyping, they are facing numerous challenges in the development of platforms, sensors,
and data processing algorithms. He will share his experiences with the development of the high-
throughput and high-accuracy Plant Stand Analyzer for large scale plant breeding programs, and his
ongoing efforts with two NSF MRI instrument development projects for indoor (Enviratron) and infield
(PhenoNet) robotic phenotyping.
Dimensionality and Throughput are Now Compatible in Automated Plant Phenotyping Systems
Nathalie Wuyts, VIB Department of Plant Systems Biology
Stien Mertens, VIB Department of Plant Systems Biology; Lennart Verbraeken, VIB Department of Plant
Systems Biology; Stijn Dhondt, VIB Department of Plant Systems Biology; Rafael Abbeloos, VIB
Department of Plant Systems Biology; Dirk Inze, VIB Department of Plant Systems Biology
The first generation of plant phenotyping systems typically focuses on plant growth-related traits in the
evaluation of genotype x environment interactions. High dimensionality, or the integration of a larger
diversity of plant traits in the analysis of the phenotype, is achieved by considering the shoot, organ and
cellular level for the retrieval of the structural basis of differences in plant growth and development.
This type of high-resolution or in-depth phenotyping does not allow for high-throughput, except for the
shoot and root system level where automated RGB-imaging and image analysis have revolutionized
plant phenotyping, often at the expense of dimensionality. Nowadays, both translational research and
the awareness about the major influence of the plant’s environment and growth conditions on its
functioning, and consequently its growth, have stimulated the development of second-generation
phenotyping systems for crops under greenhouse and field conditions, and the measurement of plant
physiological traits on a larger scale and in high-throughput by means of imaging. We are running and
developing the PHENOVISION was specifically designed to foster high-dimensionality in plant trait
measurement in an automated, high-throughput system for crops throughout development and under
greenhouse conditions. An imaging setup suited to the three-dimensional reconstruction of plants to
measure both shoot- and leaf-level traits with higher precision delivers an extended range of plant
growth-and architecture-related traits. To determine plant physiology-related traits in high-throughput,
thermal infrared and hyperspectral imaging systems were included. These are to date the most relevant
cameras for close-range and remote sensing under both controlled and field phenotyping conditions.
Recent progress in image and data analysis in the PHENOVISION phenotyping system and the prospects
for technology development will be discussed.
Virtual Reality: A New Tool for the Modern Crop Breeding Toolbox
Marco van Schriek, Keygene
Modern breeding process has an ever-expanding set of tools at its disposal, all of which capture valuable
information from the crop. While the technological advances to-date incorporate systems to address the
complexity of plant breeding as a science, most of the valuable, artisanal nuances that influence critical
decision-making in plant-breeding are largely ignored. A system that can efficiently capture and convert
crop plant data into quantitative metrics while enabling the plant breeder, to a certain degree, the
added-value of spatio-temporal qualitative analyses presents a valuable tool for the modern plant
breeder. This presentation describes how KeyGene integrated its expertise in data sciences, digital
phenotyping and plant breeding with expertise in Virtual Reality (VR) to address this challenge. This
proprietary VR based breeding tool allows users, for the first time, to become fully immersed in a
customized crop growth environment where VR breeders can access, visualize and analyze large
amounts of genotypic and phenotypic data, to make better data-driven decisions. Such an integrated
VR-based interface expands professional insight and enables better decision-making in crop breeding.
An Automated Phenotyping Platform for Trait Discovery and Characterization in Corn
Gerie van der Heijden, DuPont Pioneer
Transgenic or genome editing approaches for agronomic traits of corn, for example to improve drought
tolerance or nitrogen use efficiency, require an accurate and precise phenotyping pipeline. The small
genetic differences may lead to relatively small phenotypic differences, and hence we need to reduce
the random variation and increase the precision of the measurements.
The agronomic traits discovery pipeline of DuPont Pioneer consists of precision phenotyping possibilities
in the greenhouse as well as extensive field testing trials in managed stress environments. The managed
stress environment in the field offers accurate measurements of the end goal: yield under different
realistic stress conditions. Precision phenotyping in the greenhouse offers the possibility to test leads
year-round, to apply well-defined stresses at an individual plant basis and to study in detail plant growth
and development of each plant with advanced equipment, like fluorescence imaging and hyperspectral
imaging.
In this presentation we will show the automated greenhouse phenotyping platform for screening
drought tolerant phenotypes in corn.
General Session: Environmental/Stress Biology
February 12, 2017 | 8:30 AM – 12:15 PM
Adaptive Plasticity at Multiple Scales of Gene Regulation in Three Crops
Julia Bailey-Serres, University of California, Riverside
Water extremes can promote developmental reprogramming that involves a subset of root cells. We
have resolved responses using tools that access multiple layers of gene regulation and individual cell
types in rice, tomato and medicago.
Modifications of Source-Sink Relationships and Abiotic Stress Tolerance in Crop Plants
Eduardo Blumwald, University of California, Davis
Changes in plant hormone homeostasis promoted alterations in source/sink relationships that resulted
in enhanced plant tolerance to water-deficit stress. A number of genes associated with source
strengthening have been identified and their expression has been modified in a number of crop species
resulting in improved yields under stress.
The Root Phenome and 21st Century Agriculture
Jonathan Lynch, Penn State University
Crops with improved water and nutrient capture are urgently needed to sustain a growing population in
a changing climate. The root phenome is a challenge and an opportunity to achieve this goal.
Getting to the Root of the Matter: The Growing Importance of Root Biology to Global Food Security
Leon Kochian, University of Saskatchewan
There is a growing realization in the plant research community of the importance of root system
development and function to crop adaptation to soil-based abiotic stress. In this talk, Leon will discuss
our research integrating root phenotyping with molecular genetics and genomics to identify genes
underlying agronomically important crop traits, and how this information is being used to facilitate crop
improvement.
Metabolic Phenotyping of Germination and Respiration in Lettuce Seeds
Oliver Fiehn, University of California, Davis
Lettuce seeds exhibit thermoinhibition, an inability to germinate above a threshold upper temperature
limit. Following respiration measurements of different types of seeds, primary metabolites and lipids
were measured along germination time points at different temperatures.
Measuring Gene Expression Dynamics with High Temporal and Spatial Resolution using Hacked
Illumina Sequencers and Open Source Software
Nick Kaplinsky, Swarthmore College
High temperature stress result in protein denaturation and increased membrane fluidity, changes that
are deleterious to normal cellular function. The heat shock response (HSR) is a conserved transcriptional
response that results in the expression of heat shock proteins, protein chaperones which function to
ameliorate the effects of high temperatures. While a fair amount is understood about cellular responses
to heat shock, less is known about the cellular thermostats which regulate these responses. Surprisingly,
none of the genes shown to function as thermostats in Arabidopsis have been identified using forward
genetics, probably due to a combination of high levels of functional overlap within thermostat gene
families and a reliance on end-point phenotyping approaches.
To overcome this problem we have built a high content screening system using parts scavenged from
obsolete Illumina sequencers called the Rootscope. The Rootscope can quantitate gene expression with
high spatial and temporal resolution in 180 plants at a time. We are using this system to identify new
genes which regulate the HSR. The first gene identified in our genetic screen encodes a spliceosome
subunit that, when mutated, results in plants that respond to heat stress as if they have miscalibrated
thermostat mechanisms – they respond as if they are several degrees hotter than the actual heat shock
they are exposed to. This gene regulates the kinetics of the HSR and modulates the temperature
dependent splicing of HSFA2, a transcription factor known to regulate the HSR, as well as several other
spliceosomal proteins. These results suggest that the gene we have identified functions to regulate heat
stress dependent splicing and may provide mechanistic insights into how plants sense and respond to
high temperature stress.
Role of Mineral Nanoparticles in Improving Attachment and Colonization by Plant Growth Promoting
Rhizobacteria
Salme Timmusk, Uppsala BioCenter
The soil surrounding plant roots is one of the main sources of bacteria expressing plant-beneficial
activities (PGPR). Recently we have shown that the PGPR have great potential in protecting plants
against abiotic and biotic stress situations and restoring marginal lands (1-3). Multiple factors influence
isolate stability. The effectiveness of the inoculum is related to the formulation technology protecting
the cells from the surrounding environmental conditions and elimination of secondary effects. Titania
nanoparticle (TN) formulation form stable, large and thick bacterial clumps as a biofilm which influence
plant growth via facilitating root hair length and density and improving mulch biofilm formation (1-3).
This effect is especially pronounced in complex environments there several PGPR strains are used in
combination with TNs (1). The biofilm substantially improves root soil contact and enhances significantly
plant nutrient or biologically active compound acquisition from soil or bacterial origin. In addition,
improved organic matter content and porosity restores marginal lands (1)
The key to further advancing our treatment with microbial products is (i) surface grafting of sands and
intercalation of clays and employing hybrid biopolymer-mineral nanocomposites (ii) distinguishing
between genes and pathways that drive passenger events both in PGPR and crop plant.
1) S. Timmusk, G. A. Seisenbaeva, A. Muraya, J. Muthony, L. Behers, Nano titania aid multiple plant
growth promoting bacterial interactions under complex environments. Frontiers in Plant
Sci, (2016).
2) S. Timmusk, et al, Perspectives and challenges for microbial application for crop
improvement. Frontiers in Plant Sci, (2016).
3) S. Timmusk et al., PloS ONE DOI 10.1371/journal.pone.0096086, (2014).
Overlap between the Ad/Abaxial Developmental and ABA Signaling Networks
Tie Liu, Stanford University
Adam Longhurst, Carnegie Institution for Science; Franklin Talavera-Rauh, Carnegie Institution for
Science; Samuel Hokin, Carnegie Institution for Science; Kathryn Barton, Carnegie Institution for Science
The actions of the plant hormone abscisic acid (ABA) in the repression of germination and the inhibition
of water loss through stomata are well studied. Less well understood are the functions of ABA on
vegetative development including the mechanisms through which it influences growth of roots and
shoots. In our study of gene networks that control development of the adaxial (towards meristem/inner)
and abaxial (away from meristem/outer) domains of the leaf, we identified a set of genes oppositely
regulated by REVOLUTA and KANADI. Observing that this set of genes is enriched for genes encoding
ABA signaling components, we reasoned that genes of unknown function within this set would also be
involved in ABA signaling. Indeed, we find that a mutation in one of these, ABA INSENSITIVE GROWTH 1
(ABIG1), results in ABA resistant vegetative growth of shoot. Conversely, conditional overexpression of
ABIG1 mimics the application of ABA to wildtype plants: it causes leaf yellowing, reduces the rate of leaf
production and causes reduced root growth. Surprisingly, the ABA resistant abig1 mutants are also
drought resistant. We hypothesize that this reflects the ablation of the branch of the pathway for ABA
inhibition of growth and promotion of senescence while leaving intact the pathway for stomate closure.
To characterize ABIG1 induced senescence and meristem arrest, we carried out phenotyping and
quantifying analysis on the growth and development of meristem and young leaf primordia under at
various times post induction and drought treatments for both Arabidopsis and the orphaned crop
quinoa.
Analysis of Water Limitation Effects on the Phenome and Ionome of Arabidopsis at the Plant Imaging
Consortium
Argelia Lorence, Arkansas State University
Lucia Acosta, Arkansas State University; Suxing Liu, Arkansas State University; Erin Langley, Arkansas
State University; Zachary Campbell, Arkansas State University; Norma Castro-Guerrero, University of
Missouri; David Mendoza-Cozatl, University of Missouri; Argelia Lorence, Arkansas State University
Food security is currently one of the major challenges that we are facing as a species.
Understanding plant responses and adaptations to limited water availability is key to maintain or
improve crop yield, and this is even more critical considering the different projections of climate change.
This is an NSF-funded collaborative effort between the Mendoza and Lorence groups as part of the
activities of the Plant Imaging Consortium (http://plantimaging.cast.uark.edu/). In this work, we
combined two high-throughput -omic platforms (phenomics and ionomics) to begin dissecting time-
dependent effects of water limitation in Arabidopsis leaves and, ultimately, seed yield. As proof of
concept, we acquired high-resolution images with visible, fluorescence and near-infrared cameras and
used commercial and open source algorithms to extract the information contained in those images. At a
defined point, samples were also taken for elemental profiling. Our results show that growth, biomass
and photosynthetic efficiency were most affected under severe water limitation regimes, and these
differences were exacerbated at later developmental stages. The elemental composition and seed yield,
however, changed across the different water regimes tested and these changes included under- and
over-accumulation of elements compared to well-watered plants. Our results demonstrate that this
combination of phenotyping techniques can be successfully used to identify specific bottlenecks during
plant development that could compromise biomass, yield and the nutritional quality of plants.
Climate Metadata Analysis and Challenges
February 12, 2017 | 2:00 PM – 5:30 PM
Microbiome and Data Science: Harnessing the Unfulfilled Dream
Melissa Dsouza, University of Chicago
The advent of high-throughput sequencing technology has generated an unprecedented volume of
microbiome data. I will discuss and outline issues caused by the lack of an open-source, centralized
resource that processes and integrates all microbiome data and propose a vision to advance
microbiome research as a data-driven science.
Bringing Data Layers to Life in Precision Agriculture
Keith Koutsky, The Climate Corporation
In this presentation, Keith will discuss how at The Climate Corporation, they bring together many data
layers, millions of data points and a variety of disciplines to quantify what’s occurring on the farm and
how farmers can get to the best productivity going forward.
Translating Metadata in Microbiome Discoveries
Fan Yang, Iowa State University
Physicochemical measurements and microbiome analyses can inform ecosystem health. However,
either of them alone is not sufficient to predict the changes in ecological functions. To better
understand the microbial interactions and their ecological impacts, we integrated physicochemical and
microbial sequencing data to decode the complex microbial network in their associated environments.
Genetic Metadata Analysis and Challenges
February 12, 2017 | 2:00 PM – 5:30 PM
Finding Common Language to Describe Phenotypes in Plant Databases
Lisa Harper, US Department of Agriculture
Biocurators from 27 agriculturally related databases, both plant and animal, have come together and
formed the AgBioData Group. We are working together on many projects, including establishing and
implementing common metadata standards, data and code sharing strategies, and implementing a
common ontological language to describe plant phenotypes in a human and machine readable way. This
talk will outline our strategies, difficulties and progress.
Improved Methods for High-Confidence GO Annotation
Carolyn Lawrence-Dill, Iowa State University
Making a genome sequence accessible and useful involves three basic steps: genome assembly,
structural annotation, and functional annotation. Current datasets of GO functional annotations for
genes are high-coverage and low-confidence in most plant species. We describe our efforts to deploy
methods that produce high-confidence GO datasets for plants and outline the results of applying these
methods to maize genes.
Plant Comparative Phenomics
George Gkoutos, University of Birmingham
The presentation will reflect on experience from the biomedical domain to propose and describe
approaches for enabling plant comparative phenomics.
Open-Source Tools for High-throughput Plant Phenotyping
Noah Fahlgren, Donald Danforth Plant Science Center
The recent development of high-throughput plant phenotyping platforms has created challenges in data
analysis and management. In this talk I will discuss the software tools we have developed to address
these challenges and their application to different biological questions.
Transcriptional Rewiring: Synthetic and Evolutionary Processes Governing Plant Environmental Stress
Responses
Oliver Windram, Imperial College London
Plant defence responses are modulated by substantial transcriptional reprogramming, up to 40% of the
genome can be differentially expressed following pathogen challenge. High temporal resolution
differential expression analysis and subsequent network inference has revealed a large transcriptional
networks with a complex hierarchies. Overall this indicates that transcription factors play an important
role in the defence response. The complexity, diversity and redundancy in these defence response
networks further suggests interesting evolutionary principles may govern and give rise to these highly
structured and plastic networks. Previously we have shown that a synthetic technique known as genetic
rewiring, where the open reading frames of transcription factors are fused to different promoters
altering the natural expression of the regulators, can be used to synthetically evolve novel phenotypic
diversity. As a transcription factor can potentially regulate thousands of target genes, alteration of its
natural expression has the potential to generate radically new phenotypes through cascading regulatory
events. We have shown, in yeast, that heterologous protein expression can be significantly enhanced
using this rewiring approach. Furthermore, network analysis reveals that rewired open reading frames
and promoters possess characteristic topological network features that can serve as predictive
measures for future rewiring endeavours. In conjunction with a presentation of our work on rewiring
the yeast transcriptome I will discuss how we are using a combined systems and synthetic biology
approach to first construct large scale transcriptional networks responding to pathogen challenge in
plants. Secondly, I will show how through a process of topological network analysis we are now selecting
promoters and open reading frames which we will use to rewire the plant defence transcriptome. This
novel transciptome diversity will be used to help reveal key regulatory points and potential weaknesses
in these networks as well as highlighting potential solutions to help improve tolerance to plant
pathogens.
A Multispecies, Multi-trait GWAS Approach to Uncover Fundamental Genetic and Molecular
Mechanisms Underlying Responses to Phosphate Deficiency
Marco Giovannetti, Gregor Mendel Institute
Christian Goeschl, Gregor Mendel Institute; Santosh Satbhai, Gregor Mendel Institute; Stanislav Kopriva,
University of Cologne; Wolfgang Busch, Gregor Mendel Institute
Phosphate is an essential plant macronutrient and its low soil availability leads farmers to massive
application of phosphate-based fertilizer. Given the negative environmental impact of fertilizers, and
phosphate being a non-renewable resource, a better understanding of plant genetic and molecular
bases of phosphate absorption and utilization is needed. Much progress in that area has been made
over the past years, but most approaches focused on one species. However, it is unclear which of the
discovered mechanisms are conserved and which ones are specific to the particular species that is
studied. Here, we approach this problem by using a combination of high throughput root phenotyping
and IC-based phosphate quantification of natural populations of two different plant species, the
brassicaceae Arabidopsis thaliana and the legume Lotus japonicus. This allowed us to identify multiple
candidate genes regulating phosphate metabolism and root growth responses to varying phosphate
levels. We could determine multiple orthologs whose variants are significantly associated with these
phosphate related traits in both plant species. We are currently testing the involvement of high
confidence candidate genes in phosphate homeostasis and root growth responses upon phosphate
levels in both species.
Imaging Metadata Analysis and Challenges
February 12, 2017 | 2:00 PM – 5:30 PM
Workflows for Trait Extraction from Digital Data: Integrating Machine Learning, Crowdsourcing and
High throughput Computing
Baskar Ganapathysubramanian, Iowa State University
Baskar will present results from a few phenotyping projects where a combination of domain knowledge,
machine learning, image processing and high performance computing are used to create automated
workflows for trait extraction. This is a collaborative work by an interdisciplinary group that includes
engineers, data scientists, and plant scientists.
Early-Stage Maize Phenome
Nathan Miller, University of Wisconsin-Madison
Three image-based, high-throughput seed and seedling pipelines will be presented. Machine vision
coupled with readily available imaging technologies make for flexible, robust platforms useful for
quantifying maize seed imbibition, root gravitropism, and coleoptile emergence.
Data on High; Successes and Lessons from the Texas A&M AgriLife Unmanned Aerial Systems (UAS)
Project
Seth Murray, Texas A&M University
Multidisciplinary teams of 40 faculty, using five different UAS platforms, collected field phenotyping and
agronomic data across different crops and over time. Plant height of maize and sorghum, along with
southern rust disease in maize were successfully estimated by UAS but many challenges remain to make
successful measurements routine, especially UAS reliability, data management and data analysis.
Next Generation Data Management Applied to Image Data for Plant Phenotyping
Ramona Walls, University of Arizona
Next generation data management is a set of scalable practices that aims to support big data
throughout their life cycle to ensure discoverability, reusability, and persistence. Using imaging for plant
phenotyping as a case study, this presentation will describe the challenges involved in managing big
data, highlight some contemporary tools and techniques that can be used to address those challenges,
and point out gaps that still need to be filled.
Detecting Variation in Photosynthetic Rates with Hyperspectral Imaging
Katherine Meacham, University of Illinois
Christopher Montes, University of Illinois at Urbana Champaign; Liujun Li, University of Illinois at Urbana
Champaign; Stephen Long, University of Illinois at Urbana Champaign; Carl Bernacchi, University of
Illinois at Urbana Champaign
Improved photosynthetic rates have been shown to increase crop biomass, making improved
photosynthesis a focus for driving future grain yield increases. The RIPE project (Realizing Increased
Photosynthetic Efficiency) aims to increase crop biomass to improve food security in key food crops such
as rice and cassava by making improvements to the photosynthetic pathway, currently using nicotiana
(tobacco) as a model species. Gas exchange measurements are the most widely used method of
measuring photosynthesis in field trials but this process is laborious and slow, and requires further
modeling to estimate meaningful parameters and to upscale to the plot or canopy level. To determine
field scale efficiency of genetic modifications made, high throughput phenotyping technologies to detect
variation in photosynthetic rates are required. With hyperspectral imaging, we infer the carboxylation
rate of Rubisco (Vcmax) and electron transport rate (J) and detect photosynthetic variation between
cultivars with a partial least squares regression technique. Ground-truth measurements from
photosynthetic gas exchange, a full spectrum (400-2500nm) hyperspectral leaf clip, and extractions of
leaf pigments support the model.
Optimization of Cooking Time Prediction in Dry Beans (Phaseolus vulgaris L) using Vis/SNIR
Spectroscopy and Hyperspectral Imaging Techniques
Fernando Mendoza, Michigan State University
Karen Cichy, USDA – ARS; Jason Wiesinger, Cornell University; Renfu Lu, USDA – ARS; James Kelly,
Michigan State University
Automatized nondestructive sensing techniques are needed by dry bean breeders and processors to
predict cooking time rapidly, accurately and economically. The successful application of these
breakthrough technologies for cooking time estimations would substantially help in the objective
characterization and screening of bean germplasm. The main goal of this study was to implement and
optimize an experimental procedure and data analysis for evaluating cooking time in dry beans with
large genetic diversity using two optical sensing techniques: visible and short near-infrared spectroscopy
(Vis/SNIR) over the wavelength range of 400–2,480 nm, and an hyperspectral imaging system (HYPERS)
over the range of 450–1,000 nm. A diversity panel of Andean beans (Phaseolus vulgaris) grown in
Michigan (USA) in different years (2013, 2014 and 2015) that exhibited wide variability in cooking time
(soaked beans ranging from 17.1–120.3 min) was tested in this study. Spectral preprocessing methods,
image preprocessing and analysis, feature selection and multivariate statistical methods were critically
evaluated and optimized for improving cooking time predictions on soaked and unsoaked dry seeds.
Overall results showed that applying the two-band ratios preprocessing method and a single feature
selection method, Vis/SNIR and HYPERS models have great potential for predicting cooking time of dry
beans over a wide range of measurements. Based on partial least squares regression models, correlation
coefficients for prediction ( and standard error of prediction for soaked beans were significantly better
for HYPERS (ranging from = 0.856–0.889, = 8.1–7.7 min, respectively) than those for Vis/SNIR (ranging
from = 0.681–0.761, = 11.5–9.0 min, respectively). Also, as expected cooking times were longer for
unsoaked beans (from 80.1–147.0 min), and their predictions using HYPERS models were less robust and
accurate (ranging from = 0.667–0.708, = 11.1–10.6 min, respectively) than for soaked beans. It should
be noted, however, that the robustness of calibration models is impacted by genotypic diversity,
planting year and phenotypic used for model building, and hence, chemometric models should be
maintained and updated with new data.
Design of a Biotech Trait Discovery Phenotyping Program
Cory Christensen, Dow AgroSciences
With rapid advancements in “omics” technologies, it has become axiomatic that phenotyping is the new
frontier in need of technical innovation. Recent years have witnessed increased competition among
developers of phenotyping systems and new solutions are coming to market. In this environment,
researchers and leaders tasked with developing a phenotyping program have many options. This
presentation will focus on important design considerations, key experimental questions and critical
lessons learned in the context of an industrial-scale agronomic trait discovery program.
General Session: Plasticity in Plant Traits
February 13, 2017 | 8:30 AM – 12:15 PM
Transgene Mediated Epimutagenesis Leads to Heritable Phenotypic Variation
Bob Schmitz, University of Georgia
There exists extensive natural epigenetic and cryptic variation in plant genomes some of which is linked
to heritable phenotypic variation. Accessing this source of variation to improve plant traits is
cumbersome as it often relies on the identification and incorporation of rare events into breeding
programs. We describe a novel methodology termed “epimutagenesis” to rapidly generate phenotypic
variation by randomly demethylating the genome of Arabidopsis thaliana via transgenic expression of a
DNA demethylase. Not only does this system lead to widespread loss of DNA methylation, but it also
leads to the production of heritable phenotypic variation. Application of epimutagenesis to other plant
species will lead to expression of alleles typically silenced by DNA methylation, uncovering previously
hidden genetic variation.
Methylome Evolution in Plants
Frank Johannes, Techinical University of Munich
DNA methylation is a heritable epigenetic modification, with important roles in regulating the
expression of transposable elements, repeats and genes. Here I review new insights into the
evolutionary forces that shape DNA methylation patterns in plant genomes over short and long time
scales.
Searching for the Mechanisms that Drive Novel Gene Expression Responses to Abiotic Stress in Maize
Nathan Springer, University of Minnesota
Changes in gene expression are frequently used to tolerate abiotic stress in plants. We are studying the
processes that lead to natural variation in gene expression responses to abiotic stress with a focus on
the potential role for transposons in this process.
Blinded by the Light; An Alternative Type of Seed Dormancy
Zsuzsanna Merai, Gregor Mendel Institute
As well-described in textbooks, Arabidopsis seeds require light for germination. However, these
principles are not universal, some plant species germinate independently, others are actually inhibited
by light. We found another Brassicaceae species showing natural variation in light-regulated seed
germination, including light induced secondary seed dormancy. These traits might represent ecologically
relevant adaptations to specific environments and provide a great opportunity to investigate their
molecular regulation.
Genetics of Adaptive Plasticity: From Model Systems to Crops
John McKay, Colorado State University
The long-term goal of my research is to provide detailed functional knowledge of mechanisms and
mutations regulating genotype by environment interactions underlying adaptation. I have largely
focused on drought responses, including water use physiology and the sensitivity of growth, fitness and
yield to variation in available soil moisture. My research combines field trials, high-throughput and
detailed physiological screens in segregating families with statistical and evolutionary genomics to
identify causal polymorphisms underlying differences in drought adaptation.
I will report on results of using natural variation in drought adaptation among diverse lines of
Arabidopsis. I will also discuss ongoing collaborations with IRRI to apply field based high throughput
phenotyping towards discovering mutations underlying GxE to drought in rice.
Novel Plastid Behavior Associated with the MSH1 Effect
Jesus Beltran, University of Nebraska-Lincoln
Evan LaBrant, University of Nebraska-Lincoln; Mon-Ray Shao, University of Nebraska-Lincoln; Sally
Mackenzie, University of Nebraska-Lincoln
In plants, plastids serve as sensors of developmental and environmental stimuli, allowing cells to
modulate gene expression and likely to reprogram the epigenome in a heritable fashion. MSH1 is an
organellar DNA binding and thylakoid protein with profound roles in development, stress responses and
redox balance. Current evidence supports MSH1 as important for influencing transgenerational
epigenetic inheritance. MSH1, together with its interacting partner PPD3, localize to small, distinctive
plastids that display low autofluorescence and reside within the epidermis and vascular parenchyma.
We have termed these organelles ‘sensory plastids’. Our observations imply proteome heterogeneity in
different plastid types across discrete cell populations. To gain insight into the role of sensory plastids,
we combined Fluorescence-Activated Cell Sorting (FACS) and proteomics to characterize their
proteomes in Arabidopsis floral stems. We used Translating Ribosome Affinity Purification (TRAP)
analysis followed by RNA sequencing technology (TRAP-SEQ) to assess both MSH1-containing and
MSH1-depleted, cell-specific translating RNAs. Early results suggest that protein counts for responders
to abiotic stimulus are more prominent in rank in sensory plastids than in chloroplasts, which suggests
plastid specialized function. TRAP-SEQ experiments revealed that cells depleted of MSH1 actively
translate proteins for oxidation-reduction processes, cell redox homeostasis and auxin response, which
suggest a role of MSH1 in triggering retrograde signaling in specific cell types. Finding correlations
between cell-specific translatomes and the plastid proteome profiles within these cells would help to
understand the underpinnings of organellar functional specialization.
Assessment of Plant Performance Traits in Controlled Environments and Translation to the Field
Astrid Junker, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK)
Marc Heuermann, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK); Rongli Shi, Leibniz
Institute of Plant Genetics and Crop Plant Research (IPK); Henning Tschiersch, Leibniz Institute of Plant
Genetics and Crop Plant Research (IPK); Matthias Lange, Leibniz Institute of Plant Genetics and Crop
Plant Research (IPK); Daniel Arend, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK);
Jean-Michel Pape, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK); Rhonda Meyer,
Leibniz Institute of Plant Genetics and Crop Plant Research (IPK); Kathleen Weigelt-Fischer, Leibniz
Institute of Plant Genetics and Crop Plant Research (IPK); Michael Grau, Leibniz Institute of Plant
Genetics and Crop Plant Research (IPK); Andreas Boerner, Leibniz Institute of Plant Genetics and Crop
Plant Research (IPK); Uwe Scholz, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK);
Thomas Altmann, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK)
To meet the challenges in global food security requires the development of strategies towards the
optimization of yield formation and resource efficiency of crop plants under future climate scenarios. To
enable this, a deeper understanding of processes underlying plant acclimation to changing
environments is necessary. Using high throughput automated plant phenotyping systems at IPK, the
dynamics of plant growth and photosynthetic efficiency have been studied in accession panels of
Arabidopsis and maize under controlled conditions. This led to the identification of Arabidopsis and
maize candidate accessions with different acclimation strategies to high light and variations in biomass
yield and photosynthetic efficiency, respectively. To evaluate and enhance the prospects of lab-to field
translation, candidate maize accessions have furthermore been phenotyped for root- and shoot traits
under different cultivation conditions in the glasshouse and in the field. Modification of the standard
cultivation procedures improved the lab-to-field translation of phenotypic trait expression which will be
further optimized in the Plant Cultivation Hall currently being erected at IPK. This building will enable to
run specifically designed and reproducible cultivation scenarios that mimic field conditions. In this way,
trait expression similar to that in the field will be triggered and monitored using automated phenotyping
installations. Investigations involving precise variation of environmental parameters will yield a deeper
understanding of acclimation processes and underlying molecular mechanisms and genetic
determinants and will support the prediction of idiotypes with improved performance under future
climate scenarios. Furthermore recent achievements in phenotype data management, standardized
metadata representation (e.g. MIAPPE), and data publication will be presented.
Ectopic Expression of psNTP9, An Apyrase Gene from Pisum Sativum, Confers Drought Tolerance in
Transgenic Soybean
Roopadarshini Veerappa, University of Texas at Austin
Soybean (Glycine max (L.) Merr.) is one of the major resources of protein for human and animal
nutrition as well as a key source of vegetable oil. It is also considered a potential crop for the production
of biodiesel. Drought is a critical environmental factor that imposes water stress on crops, which is a
major constraint on plant growth and productivity, contributing to yield losses in crops, including
soybeans. Apyrases (Nucleoside Triphosphate-Diphosphohydrolases) are calcium- and magnesium-
activated enzymes that remove the terminal phosphate from nucleoside triphosphates (NTPs) and
nucleoside diphosphates (NDPs). By controlling the [NTP] and/or [NDP] in the Golgi, the ECM or other
subcellular sites, apyrases are known to play an essential role in plant growth, development and stress
responses. In the present study the pea apyrase gene (psNTP9), driven by CaMV 35S promoter, was
transferred into the soybean genome by Agrobacterium-mediated plant transformation. Two
independently transformed soybean lines expressing psNTP9 were produced. Segregation analyses
indicated single-locus insertion for both lines. Preliminary tests performed in the greenhouse evaluated
the phenotype of transgenic soybean plants under drought conditions. Transgenic plants outperformed
the wild-type plants when subjected to water deficit conditions, exhibiting notable drought tolerance
with a higher survival rate compared to the wild type. They also had a greater shoot and root biomass
and showed significantly higher seed production compared to non-transgenic plants under water deficit.
This is the first report showing that the ectopic expression of psNTP9 in transgenic soybeans confers
drought tolerance, highlighting the potential of this gene for molecular breeding.
Funded by grants from the National Science Foundation (IOS-1027514) and a private company to Drs.
Greg Clark and Stanley J. Roux at the University of Texas at Austin.
On the Statistical Mechanics of Cytosine DNA Methylation and the Dection of the Methylation
Regulatory Signal
Robersy Sanchez, University of Nebraska-Lincoln
DNA methylation patterning represents one feature of the epigenome that is highly responsive to
environmental stress and associates with transgenerational adaptation in plants. An information
thermodynamics theory on cytosine DNA methylation (CDM) was recently published. Results indicated
that most methylation changes occurring within cells are likely induced by thermal fluctuations to
ensure thermal stability of the DNA molecule, seemingly explainable by statistical mechanics laws.
Therefore, one does not observe a genome-wide relationship between methylation and gene
expression. Ignoring the statistical biophysics subjacent to CDM is a source of bias for the most common
statistical tests applied in methylome analysis. A key limitation to current genome-wide methylation
analysis is the inability to discriminate signal, induced by experimental treatment, from background
“noise” that emanates from this natural, dynamic methylation activity. Here, we provide novel insights
for application of statistical mechanics of CDM to detect methylation regulatory signal. Integration of
information thermodynamics of CDM and signal detection theory permits robust discrimination of
biological signal from physical noise-induced thermal fluctuations. The analytical steps summary follows:
i) Information thermodynamic theory computes the probabilities of signal plus noise, with use of
experimental controls to estimate the receiver’s threshold (cutoff Hellinger divergence value, HD) at
which the rate of false positive is minimal or at least acceptable for the experimental conditions. ii)
Potential differentially methylated positions (DMPs) are proposed based on the probabilities of their
corresponding HDs. iii) A logistic regression analysis is performed with the prior binary classification of
DMPs (from treatment versus control), and iv) a receiver operating curve (ROC) is built to estimate the
cutoff point for HD at which an observed methylation level represents a true DIMP. Our approach is
illustrated with examples from analysis of genome-wide CDM reprogramming induced by msh1 mutant
effect in Arabidopsis thaliana.
Climate and Crop Models to Predict Plant Productivity Based on
Environment
February 13, 2017 | 2:00 PM – 5:30 PM
Crops Models as Linking Technology Between Phenotyping and Genotype
Scott Chapman, CSIRO
This presentation will cover the role of dynamic models based on biological understanding as an avenue
to link molecular and whole plant levels of biological organization and how this offers advantages
beyond pure statistical methods in linking genotype to phenotype for plant breeding, especially when
confronted with genotype-by-environment interaction.
Improving Phenotypic Prediction through Crop Model-whole Genome Prediction Integration
Charlie Messina, DuPont Pioneer
Crop growth models integrate decades of knowledge in crop biology and provide frameworks to
integrate complex data to enable prediction. Crop model-whole genome prediction integration is
demonstrated in maize breeding for the US corn belt.
Global Trends in Crop Productivity
William Kolby Smith, University of Arizona
Feeding the world without destroying the environment is the biggest challenge of our time, and the
extent to which crop production can be increased will determine if we can feed a growing population
without land use change. Global-scale, empirical approaches can explore trends in crop productivity
and historical relationships between weather and yield. Here, we analyze high-resolution maps of yield
and harvested area for major crops, based on fusion of satellite and census data, to determine trends in
productivity, and impacts of climate on yields at the global scale. Examining yield gaps, defined as the
difference between current yields and realistically attainable yields, we find remarkable consistency at
the global scale, with substantial regional variation in closure of yield gaps. The absolute increase in
yield gaps suggests that global crop production can continue to increase for the foreseeable future.
Regional variations call for targeted intervention to improve management and seed quality.
Assessing Nitrogen Use Efficiency and Water Stress Interactions in Wheat Using High Resolution
Growth Analysis
Trevor Garnett, The University of Adelaide
Bettina Berger, The University of Adelaide; Nicholas Sitlington-Hansen, The University of Adelaide; Plett
Darren, The Australian Centre for Plant Functional Genomics; Mamoru Okamoto, The Australian Centre
for Plant Functional Genomics
Improved nitrogen use efficiency (NUE) in crop plants has the potential to significantly reduce fertilizer
application costs and increase crop yield. It is therefore important to develop efficient screening
methods for improved NUE. NUE is a complex, multi-component trait and screening for NUE in field
trials poses the challenge of low heritability and high environmental interaction. One of the major
restrictions to N uptake early in the season in southern Australia is lack of water, which restricts N
mobility in soil and hence N uptake. Phenotypic analysis in controlled environments offer the advantage
of focusing on individual components of NUE while controlling aspects of nitrogen (N)-availability, water
application and climate conditions. In results to be presented, the nitrogen response of wheat varieties
was investigated using a high throughput phenotyping platform. Treatments incorporated water
stresses early in the season and at flowering with the aim of better understanding the interaction
between these two important factors restricting grain crop yields. Plants were also grown on to maturity
to assess grain yield response. In terms of biomass there were large differences in N response between
varieties and major interactions with water stress. The shoot growth response of the varieties varied in
total biomass and also in the timing of growth. There were major differences in N response and
efficiency between varieties. There were good correlations between non-destructive measurements of
biomass and grain yields but this differed between water stress treatments. There were also good
correlations in performance between repeated experiments. In terms of developing a reliable,
controlled environment method for screening NUE in cereals the results are promising.
Phenotyping at Multiple Biological Scales Reveals Complex Interplay Between Traits and Environment
Elena Rice, Monsanto
Sarah Tucker, Monsanto; Frank Dohleman, Monsanto; Sean Yang, Monsanto; Shilpa Swarup, Monsanto;
Lex Flagel, Monsanto; Dmitry Grapov, Monsanto; Kimberly Wegener, Monsanto; Kevin Kosola,
Monsanto; Mike Hall, Monsanto; Ed Allen, Monsanto
There are increasing opportunities to improve crop yield using diverse approaches from traditional plant
breeding to recent technological developments such as genome editing, microbial enhancement, as well
as traditional biotechnology. Utilizing these technologies requires understanding of traits that can
contribute to an integrative phenotype like yield under diverse environmental conditions. The key
challenge here is the ability to measure large number of physiological and molecular traits using
sufficiently diverse and commercially-relevant populations, and the ability to integrate genotype,
phenotype, and environmental variables. Here, we systematically collected, 31 phenotypic traits
including yield, 83 annotated metabolites, over 21,000 transcripts on a set of 57 diverse, commercially
relevant maize hybrids across 3 environments, in the central corn belt of the USA, to understand two
central questions:
1) How do these traits vary over germplasm and environments?
2) How do traits relate to yield and to each other?
As expected, analyses showed a significant variability in measured phenotypic and molecular traits
across both pedigrees and environments and presented a complex picture of how groups of traits
interact and how they combine to produce yield. We also evaluated the opportunity to utilize
automated phenotyping system that combines throughput and resolution to characterize trait variability
for 52 commercially-relevant maize hybrids. Controlled environment maize phenotyping using high-
resolution and high-throughput can differentiate some phenotypic traits that contribute to increased
grain yield and provide and opportunity to answer specific questions under highly consistent
environmental conditions.
Integrating Multiscale Data into Models of Plant Growth and
Development
February 13, 2017 | 2:00 PM – 5:30 PM
Multi-scale Leaf Growth Phenotyping: Modelling the Relationships between Traits Over Time and in
Response to Environmental Stresses Gives Insight into Leaf Growth Control
Christine Granier, Laboratoire d’Ecophysiologie des Plantes sous Stress Environnementaux (LEPSE)
Phenotyping platforms for measuring plant traits at different scales are developed. Challenges and
opportunities for directions in multi-trait analyses will be presented with leaf development as a proof of
concept.
What Do We Learn from Phenotyping Tools that Bridge the Gaps between the Lab and the World?
David Kramer, Michigan State University
Some of the most pressing open questions about photosynthesis concern how it is adapted to function
in living organisms to balance the needs for efficient energy capture and the avoidance of toxic side
reactions. The importance of this balancing is especially critical to cope with the dynamic environmental
conditions experienced in the field. To address these questions, we need to observe the
biophysical/biochemical machinery of photosynthesis in action under the conditions of the field. David
will describe a series of enabling technologies designed to bridge the gaps between the lab and the field
that can be and how these tools can be used to network communities of researchers to address this
type of question. The technologies include: PhotosynQ.org - Community-driven plant phenotyping
platform for data-driven innovations in science and agriculture; MultispeQ – A PhotosynQ-enabled tool
for large-scale field phenotyping; CoralspeQ - A field-deployable phenotyping platform for corals, algae
and cyanobacteria; Dynamic Environmental Phenotype Imager (DEPI)- A platform for high throughput
dynamic plant phenotyping. David will describe new results using these tools to explore the
relationships between photochemistry and photoprotection in a wide range of wild type and mutant
plants, both in the lab and the field. The major outcomes of these experiments are the identification of a
biophysical “Achilles Heel” of photosynthesis; the identification of a large number of “ancillary
regulatory components” of photosynthesis to tune its responses to the environment; and practical
applications of new phenotyping tools to improve the breeding and management of crops, especially in
the developing world.
Full Stack Biology: Moving Freely Between Biological Layers with Databases, Statistics and Network
Modeling
Cody Markelz, University of California, Davis
We will discuss Cody’s research, which focuses on bridging the gap between computational and
experimental plant biology. This talk will focus on a framework for storing, collecting, modeling and
developing predictive models by integrating phenomics data.
Genome to Phenome in Plant Model Organisms and Foundation Species
Justin Borevitz, Australian National University
In this presentation, Justin will discuss working on genome variation and trait identification in
Arabidopsis and Brachypodium and translation to wheat and eucalyptus. New growth facilities and
hyperspectral image mapping will be shown.
The ’Genotype to Phenotypes’ Map
Ümit Seren, Gregor Mendel Institute of Molecular Plant Biology
Dominik Grimm, Machine Learning and Computational Biology Lab, Department of Biosystems Science
and Engineering; Joffrey Fitz, Max Planck Institute for Developmental Biology; Detlef Weigel, Max Planck
Institute for Developmental Biology; Karsten Borgwardt, Machine Learning and Computational Biology
Lab, Department of Biosystems Science and Engineering; Magnus Nordborg, Gregor Mendel Institute of
Molecular Plant Biology; Arthur Korte, Center for Computational and Theoretical Biology
In nature, organisms and specifically plants exhibit a diverse range of phenotypic variability.
Understanding how genetic variation translates into phenotypic variation has been a prime focus of
population geneticists, and we have made substantial progress in the Arabidopsis community.
Since the first Genome Wide Association Studies (GWAS; Atwell et. al 2010) many research groups have
generated and mapped high quality phenotypic data ranging from disease resistance phenotypes to
high-dimension molecular phenotypes — e.g., RNA expression, metabolites. Recently, a detailed map of
genetic variation was released by The 1001 Genomes Consortium for the Arabidopsis community. With
this high quality genotype and phenotype data on hand, we can construct a comprehensive “genotype-
to-phenotypes” map that allow us to perform meta-analyses of pleiotropy.
To that end, we must develop databases and tools to store and examine phenotypes, respectively, and
make them available to the community. In this talk, I will present a variety of these tools ranging from a
web-portal that allows researchers to easily carry out GWAS by simply providing phenotype data to a
central database, to an intuitive interface for storing and navigating through published phenotypes.
The Planteome Project: Annotating Plant Phenotypes using Ontologies for Data Integration
Laurel Cooper, The Planteome Project
Austin Meier, Oregon State University; Justin Elser, Oregon State University; Marie-Angélique Laporte,
Bioversity International; Justin Preece, Oregon State University; Chris Mungall, Lawrence Berkeley Lab;
Elizabeth Arnaud, Bioversity International; Pankaj Jaiswal, Oregon State University
The Planteome project (www.planteome.org) features a suite of reference ontologies for plants,
associated with a growing corpus of genomics data in a centralized online plant informatics portal. The
species-neutral references, the Plant Ontology, Plant Trait Ontology, and Plant Environment Ontology,
(developed by the project), along with those developed by collaborating groups, such as the Gene
Ontology and the Phenotype and Trait Ontology, and others, are mapped to species-specific controlled
vocabularies for crop plant traits and phenotypes. Traditional plant breeding methods for crop
improvement may be combined with next-generation analysis methods and automated scoring of traits
and phenotypes to develop improved varieties. Linking these analyses to the growing corpus of
genomics data generated by high-throughput sequencing, transcriptomics, proteomics, phenomics and
genome annotation projects requires common, interoperable, reference vocabularies (ontologies) for
the description of the data. Data annotations to ontology terms link phenotypes and germplasm to
genomics resources, and provide semantic integration of widely diverse datasets with the goal of plant
improvement. Analysis and annotation tools are being developed to facilitate studies of plant traits,
phenotypes, diseases, gene function and expression and genetic diversity data across a wide range of
plant species. The project database and the online resources provide researchers tools to search and
browse and access remotely via APIs for semantic integration in annotation tools and data repositories,
providing resources for plant biology, breeding, genomics and genetics. The project is supported by the
National Science Foundation award IOS #1340112
Statistical Analysis of Multivariate Data inc Machine/Deep Learning
February 13, 2017 | 2:00 PM – 5:30 PM
The Competing Environmental Influences on Plant Mating and Plant Floral Traits in a Common
Agricultural Weed
Gina Baucom, University of Michigan
The plant mating system is a labile trait that is strongly influenced by the environment. In this talk, I
discuss the relative role of both human-mediated and natural selection on plant mating patterns as well
as associated floral traits in the agricultural weed, Ipomoea purpurea.
Learning from Imprecise Labels for Remote Sensing and Phenotyping
Alina Zare, University of Florida
During this talk, Alina will introduce novel supervised and semi-supervised machine learning methods
that leverage imprecise, uncertain and incomplete labels. Application of these methods to remote
sensing and phenotyping will be presented and discussed.
Melding STEM and Plant Sciences towards Advanced Plant Phenotyping and Discovery
Jennifer Clarke, University of Nebraska-Lincoln
We will discuss the development of a cross-disciplinary research group to address computational and
analytical problems in plant phenomics and genomics. This development comes with associated
challenges, from cross-disciplinary communication to mentorship to balancing the interests and
expectations of various group members. We will highlight several successful projects associated with
plant phenotyping activities at the University of Nebraska-Lincoln. This presentation will conclude with
an introduction to the Midwest Big Data Hub focus area in Digital Agriculture and opportunities for
cross-institution and cross-sector collaborations.
Persistent Homology: A Powerful Tool for Mapping Variation in Complex Phenotypes
Washington Mio, University of Florida
I will discuss the concept of persistent homology and describe ways in which the method can be used to
construct compact, informative summaries of complex morphological and functional data. I also will
present an application of persistent homology to QTL analysis of tomato leaf shape.
A Meteorological Data Format that Makes Research Easier and will Work for Imaging Sensors
David LeBauer, University of Illinois
A common weather data format enabled a pipeline that matches many types of meterological data to
many crop and ecosystem simulation models. Extending this approach to imaging sensor data will
benefit the phenomics community.
Midwest Big Data Hub: Accelerating the Big Data Innovation Ecosystem
Melissa Cragin, Midwest Big Data Hub
The Midwest Big Data Hub (BDH) is one of four Big Data (BD) Innovation Hubs that were launched in
2015, with initial support from the National Science Foundation. The BD Innovation Hubs are intended
to accelerate and strengthen the data ecosystem, to develop effective, data-enabled cross-sector
networks to address scientific and societal problems of regional and national interest. The MBDH is a
growing organization of public and private partners investing in data, tools, infrastructure, and training
to improve access and use of shared and public data for scientific discovery and improved data-to-
decision systems. We will introduce the Big Data Hubs initiative, and report on some of the significant
activities underway in in the Midwest in Digital Agriculture, Food-Energy-Water, and Data Sharing, as
well as our collaborative demonstration projects. Finally, we will discuss opportunities to engage with
the Midwest Big Data Hub, and our growing networks of Public/Private partnerships.
General Session: Metabolomics and Large-scale Biochemical
Phenotyping
February 14, 2017 | 8:30 AM – 12:00 PM
Ionomics: Using Molecular Phenotypes to Dissect the Complex Interactions between Plant Genetics
and the Environment
Ivan Baxter, Donald Danforth Plant Science Center
Plants take up elements from their soil environment using processes controlled by the genome. The
complex nature of these interactions suggests that the elemental profile will be highly plastic with the
genetic factors influencing accumulation depending on which environment the plant is grown in. My lab
uses high throughput elemental profiling (‘ionomics’) as a tool for understanding plant adaptation. High-
throughput workflows allow a single inductively coupled plasma mass spectrometer (ICP-MS) to
precisely analyze hundreds of samples for more than 20 elements per day. We have used ionomic
profiling of 250,000+ maize kernels, soybeans and cotton seeds to detect the genetic and environmental
determinants of the ionome. I will focus on our attempts to understand the genetic by environmental
interactions that underlie elemental accumulation.
Unidentified Metabolomics Peaks and Metabolism’s Dark Underside
Andrew Hanson, University of Florida
Enzymes make mistakes and metabolites undergo spontaneous reactions; these facts are undeniable
but underappreciated. The damaged metabolites produced by enzyme errors and chemical side-
reactions are metabolism’s dark underside and can explain many unidentified metabolomics peaks.
Viewing Plants from the Inside-Out: Variation in the Sorghum Metabolome and Association to
Morpho-physiological Traits and Drought Tolerance
Courtney Jahn, Colorado State University
Changes in plant metabolism are instrumental to plant developmental processes, and thus underpin
many of the ways in which plants respond to the environment. Therefore, comprehensive study of plant
metabolism is valuable in evaluating phenotypic effects of abiotic stresses on plants and may uncover
potential breeding targets to improve sorghum for multiple uses in both benign and stressful
environments.
How Plants Became the Greatest Chemists - Capturing the Power of Diversity
Mark Lange, Washington State University
Plants synthesize and accumulate a breathtaking array of small molecules, with functions in all aspects
of plant life but also a myriad of uses in modern industries. This presentation will describe experimental
and computational approaches to gain insights into the unique metabolism of specialized tissues and
cell types in plants.
The PepSAVI-MS Pipeline for Natural Product Bioactive Peptide Discovery
Leslie Hicks, University of North Carolina at Chapel Hill
The PepSAVI-MS pipeline employs bioactivity assays, statistics, and mass spectrometry to identify
bioactive peptide targets from complex samples. While our initial focus demonstrates applicability for
antimicrobial and anticancer activities, the pipeline is broad-spectrum, high-throughput, and has the
potential to expedite the search for novel bioactive peptides.
Metabolomics and Large-scale Biochemical Phenotyping for Gene Discovery in Plant Specialized
Metabolism
Lloyd Sumner, University of Missouri
This presentation will describe our metabolomics approach and technologies utilized to generate high-
resolution biochemical phenotyping data. It will also illustrate how we are integrating this with other
omics technologies for the discovery of genes involved in plant specialized metabolite biosynthesis.
Poster Abstracts
Phenomic Insights into Quantitative Traits
100 Aiming for Perennial Wheat: Can Studying Root Systems Identify Key Traits for Perenniality?
Ian Clark, Washington State University
Kevin Murphy, Washington State University; Karen Sanguinet, Washington State University
What makes a successful perennial plant? Researchers have been attempting to convert common
annual grains, such as wheat, to a perennial life cycle for many decades. The development of perennial
wheat germplasm often involves the wide hybridization of annual wheat with its nearest perennial
relatives. It has previously been shown that perennial grasses have longer roots and larger root systems
than their annual relatives, but little is known about the spatiotemporal differences in root growth
across related annual and perennial genera. This study compares root system architecture throughout
development in annual fall seeded wheat (Triticum aestivum), tall wheatgrass (Thinopyrum ponticum)
and a hybrid of the two species. The hybrid was selected for wheat-like traits that expressed the post-
sexual cycle regrowth trait critical to the perennial growth habit. Roots are periodically monitored in situ
using a minirhizotron (to a depth of ~150 cm) in the Palouse region of Eastern Washington. Once scans
are taken, roots are mapped using the RootSnap! (CID Bio-Science, Inc.) software. Scans were taken
starting at germination (Autumn, 2015) at an interval of every 2-3 weeks. The annual wheat roots grew
deeper, initiated growth earlier in the season, and showed a steeper rooting angle. The roots exceeded
the measurable depth on April 15th compared to May 4th for the hybrid and May 25th for the
wheatgrass. All plots exceeded our measurable depth by May 25th, when the plants varied from Feekes
5 (T. ponticum) to 10 (T. aestivum). When this spatiotemporal aspect of root growth is added to the
general life history strategy of a perennial, we hope to better understand how our hybrids differ from
both parents and how this may affect survival throughout multiple years.
101 A Multispecies, Multi-trait GWAS Approach to Uncover Fundamental Genetic and Molecular
Mechanisms Underlying Responses to Phosphate Deficiency
Marco Giovannetti, Gregor Mendel Institute
Christian Goeschl, Gregor Mendel Institute; Santosh Satbhai, Gregor Mendel Institute; Stanislav Kopriva,
University of Cologne; Wolfgang Busch, Gregor Mendel Institute
Phosphate is an essential plant macronutrient and its low soil availability leads farmers to massive
application of phosphate-based fertilizer. Given the negative environmental impact of fertilizers, and
phosphate being a non-renewable resource, a better understanding of plant genetic and molecular
bases of phosphate absorption and utilization is needed. Much progress in that area has been made
over the past years, but most approaches focused on one species. However, it is unclear which of the
discovered mechanisms are conserved and which ones are specific to the particular species that is
studied. Here, we approach this problem by using a combination of high throughput root phenotyping
and IC-based phosphate quantification of natural populations of two different plant species, the
brassicaceae Arabidopsis thaliana and the legume Lotus japonicus. This allowed us to identify multiple
candidate genes regulating phosphate metabolism and root growth responses to varying phosphate
levels. We could determine multiple orthologs whose variants are significantly associated with these
phosphate related traits in both plant species. We are currently testing the involvement of high
confidence candidate genes in phosphate homeostasis and root growth responses upon phosphate
levels in both species.
102 From Phenome to Genome: Lessons from the Planteome
Dennis Stevenson, New York Botanical Garden
Dario Cavaliere, New York Botanical Garden; Brandon Sinn, New York Botanical Garden
A main goal of the Planteome, besides providing vocabularies for land plants, is to develop an ontologies
that can connect the phenome with the genes responsible for phenetic expression. Traditionally, our
vocabulary was developed as descriptors of the world around us, which is to describe what we see.
Thus, the same term is often allied strictly to appearance regardless of development and evolutionary
history, in a large part because the terminology was developed before the concept of evolution. For
example, the term ligule is applied to many tongue shaped plant structures, but many of these are
composed of entirely different organs and tissues. Lenticel is another example-- those on stems look
just like those on roots, but each are developmentally and structurally distinct. Today, morphological
terms need to be used for two very different purposes, one to provide universal descriptors for plant
identification and also to be used in the sense of homology to identify the genes responsible for the
development of these descriptors. The issue at hand is how to reconcile these without discarding
existing knowledge. The use of phylogenetic information from across the tree of life will be discussed as
applied to salient examples of conflicts between utility and homology when constructing ontologies.
These have been reconciled in Planteome ontologies, but many other such examples remain to be
addressed for the evo/devo world.
103 Identification and Characterization of Root System Architecture Mutants in Medicago truncatula
Chenglin Chai, The Samuel Roberts Noble Foundation
Parna Ghosh, The Samuel Roberts Noble Foundation; Sarah Oliver, Oklahoma State University; Yuhui
Chen, The Samuel Roberts Noble Foundation; Rujin Chen, The Samuel Roberts Noble Foundation
Root system architecture (RSA) refers to the spatial configuration of a root system in the soil. RSA plays
an essential role in foraging soil resources to maintain sustainable plant productivity. To maintain and
improve sustainable plant productivity particularly under suboptimal water and nutrient levels in the
soil, it is essential to understand how root system architecture is regulated at the molecular genetic level
and how this level of regulation is responsive to the environment (plasticity). To achieve this goal, we
employed three phenotyping methods and screened approximately 200 fast neutron bombardment
(FNB)-induced mutants in Medicago truncatula cv. Jemalong A17. Out of these mutants, we are focusing
on nine mutants with altered RSA traits including long roots (lr1~2), short roots (sr1~3), and fewer
lateral roots (flr1) and long lateral roots (llr1~3). Quantitative analysis revealed that short root
phenotype was due to either a low primary growth rate (in case of sr1 and sr3) or early termination of
primary root growth (in case of sr2); while a high growth rate of the primary root led to the long root
phenotype of the lr2 mutant. The total root length of all three sr mutants was significantly less than that
of the wildtype (A17). The sr1 also exhibited less total root length and higher number of lateral root than
the wildtype. By comparative genomic hybridization analysis, we identified three deletion mutations in
the sr1 mutant, which were confirmed by PCR. An F2 population derived from the cross between sr1 and
A17 will be used to establish genetic linkage between the responsible deletion and the short root
phenotype. The possible functions of deleted genes in the sr1 mutant are currently being investigated.
Detailed phenotyping and cloning of the responsible gene are also being conducted in these RSA
mutants.
104 High Throughput Phenotyping of Photosynthesis to Identify Relevant Quantitative Trait loci and
Underlying Genes in Arabidopsis thaliana
Mark Aarts, Wageningen University
Roxanne van Rooijen, Heinrich Heine University; Aina Prinzenberg, Wageningen University; Jeremy
Harbinson, Wageningen University
Natural genetic variation in plant photosynthesis efficiency is hardly studied even though breeding for
photosynthesis would be interesting to maintain increases in crop yields. One of the reasons this is not
studied is the notorious difficulty in adequately phenotyping photosynthesis parameters for genetic
research. We designed a versatile plant phenotyper, the Phenovator, capable of multispectral imaging of
1440 plants per experiment. Imaging includes chlorophyll fluorescence to determine the light use
efficiency of photosystem II electron transport (ΦPSII or Fq’/Fm’), and near infrared reflection to
measure projected leaf area. This system has been used to phenotype 350 genetically diverse
Arabidopsis thaliana accessions for genome wide association analysis. Plants were phenotyped at non-
stressful conditions and in response to cold treatment (5 °C) or a switch from 100 to 550 µmol m-2 s-1
irradiance. The observed genotypic variation was used to identify quantitative trait loci (QTL). Our latest
progress in identification and confirmation of QTL for ΦPSII, plant growth and epinastic leaf
movement will be presented, including detailed analysis of the allelic variation at the nucleotide level of
some of the genes underlying these QTL. Our work has shown that there is sufficient genetic variation
for PSII efficiency in Arabidopsis amenable for gene identification, which suggests the same will be the
case for crop species. Such would offer interesting opportunities for future crop photosynthesis
improvements.
105 Time Dependent Genetic Analysis Links Field and Controlled Environment Phenotypes in the Model
C4 Grass Setaria
Max Feldman, Donald Danforth Plant Science Center
Rachel Paul, University of Illinois; Darshi Banan, University of Illinois; Jennifer Barrett, Donald Danforth
Plant Science Center; Jose Sebastian, Carnegie Institute for Science; Muh-Ching Yee, Carnegie Institute
for Science; Hui Jiang, Donald Danforth Plant Science Center; Alexander Lipka, University of Illinois; Tom
Brutnell, Donald Danforth Plant Science Center; Jose Dinneny, Carnegie Institute for Science; Andrew
Leakey, University of Illinois; Ivan Baxter, USDA – ARS
Vertical growth of plants is a dynamic process, which is influenced by both genetic and environmental
factors. Variation in plant height presents an intriguing ecological trade-off between access to solar
radiation, energetic demand for construction of lignified support tissue and the fecundity risk associated
with lodging. As such, plant height has a profound effect on overall plant architecture and biomass
composition. The primary goal of our studies have been to identify the major genetic loci that contribute
to the variation of plant height within two independent genetically structured populations of Setaria sp.
and how such phenotypes change temporally in response to differences in water availability and
planting density. To achieve this objective we have performed twelve growth trials of an interspecific S.
italica x S. viridis recombinant inbred line population in field and controlled environment settings in
addition to two experimental growth trials of a S. viridis natural diversity panel in the Bellweather
Phenotyping Facility at Donald Danforth Plant Science Center. Results from these studies identify
numerous genetic loci associated with variation in plant height shared across all experimental
treatments and locations but that the contribution of individual loci differs temporally. The
methodology presented in this study is applicable to any number of single or composite traits and is
used to deconvolute the contribution of plant height to other complex plant traits of agronomic
importance (biomass accumulation/composition and water use efficiency).
106 Using Association Genetics to Identify Common and Species-specific Mechanisms of Nitrate
Dependent Root Growth Responses
Anna Malolepszy, Gregor Mendel Institute of Molecular Plant Biology
Wolfgang Busch, Gregor Mendel Institute of Molecular Plant Biology
Nitrogen availability is one of the most limiting factors for most organisms. While nitrogen gas is the
most abundant constituent of the Earth’s atmosphere, it isn’t available as nitrogen source for most
organisms, including plants. Legumes evolved the mechanisms to cope with a low level of nitrate by
establishing the beneficial interaction with rhizobia bacteria, which ensure the fixation of atmospheric
nitrogen to a plant-usable form. This unique adaptation strategy enables them to grow in the nitrate-
limiting condition. The formation of nodulation pathway clearly impacted the nitrate homeostasis and
signaling networks in legumes. We would like to understand how nitrate signaling and homeostasis were
modified to accommodate the evolutionary distinct way of obtaining nitrate via nodules. To get insights
into this adaptation we investigate the root plasticity of natural accessions of the model legumes Lotus
japonicus and Medicago truncatula under low and high nitrate conditions with and without rhizobia
inoculation. The root traits are quantified via an automated pipeline for high-throughput phenotyping
(BRAT) and are then used to conduct genome wide association studies (GWAS). Using a comparative
approach and analyzing GWAS data from Lotus japonicus, Medicago truncatula as well as Arabidopsis
thaliana, we can determine whether genetic variation underlying root growth responses to nitrate
abundance is located in the same or different sets of genes. Overall, using this multi-species comparison
we aim to shed light on the molecular basis of the nitrate economy in different species.
107 Phenotyping Solanaceous Crops for Abiotic Stress Tolerance Traits
Sridhar Gutam, Indian Council of Agricultural Research
The solanaceous crops like tomato, brinjal and potato are the most popular and delicious vegetables
grown and consumed in India and south of Asia as well. However these crops are being subjected to
various abiotic stresses due to the changing climatic conditions. And the most problematic abiotic stress,
drought stress is creating difficulties in cultivation of these crops in larger areas and producing in large
quantities to meet the demand. Therefore an attempt was made to study the root traits of these crops.
And in an experiment with tomato genotypes to study the root characteristics, there found a significant
differences in the genotypes with respect to the root phenotypic characters which can be suitably used
for the development of suitable genotypes of desired root characters or can be simply used as
rootstocks in grafting for the production of drought tolerant genotypes.
108 Improving the Stomatal Resistance to Water-loss and Water-use Efficiency in Sorghum bicolar for
Bioenergy Production
John Ferguson, University of Illinois at Urbana-Champaign
Victoria Scaven, University of Illinois at Urbana-Champaign; Patrick Brown, University of Illinois at
Urbana-Champaign; Tom Clemente, University of Nebraska-Lincoln; Edgar Spalding, University of
Wisconsin; Nathan Miller, University of Wisconsin; Andrew Leakey, University of Illinois at Urbana-
Champaign
Improving the water-use efficiency (WUE; ratio of biomass production to transpiration) of bioenergy
Sorghum by 40% is predicted to reduce its seasonal water requirement by 250 mm and improve
production by 27%. Such an improvement in WUE would make >9 million acres of marginal Midwestern
land economically viable for Sorghum bioenergy production. Mechanistic modelling has suggested that
reducing the number of stomata on leaves, thus reducing resistance to water loss, is a potentially
efficacious strategy through which WUE could be improved in Sorghum. Additionally, empirical evidence
from C4 crop species suggests that photosynthesis and biomass accumulation are not penalised when
concurrently reducing stomatal conductance, due to rising atmospheric [CO2].
As a precursor to the development of novel Sorghum varieties with improved WUE, we are assessing the
natural variation that exists for multiple physiological characteristics that underpin WUE, including
stomatal density, leaf-level gas exchange, and δ13C (integrated measure of WUE), in >800
Sorghum accessions. Initial findings reveal that substantial variation exists for stomatal number and
WUE proxies. Importantly, this variation demonstrates significant heritability, thereby providing
incentive that future association mapping should be successful in terms of identifying genetic loci linked
to the observed variation. In addition, we are genetically manipulating stomatal number via the
downregulation and overexpression of key stomatal development genes, such as Sb-EPF1 and Sb-
ERECTA. Imaging of the T1 generation of multiples transgenic events of the RNA interference of Sb-EPF1
demonstrates that disrupting stomatal development appears to reduce stomatal number. Future work
aims to further characterise the WUE and response to drought of the genetic manipulation stomatal
development and to assess leaf-level gas exchange and stomatal number from a larger number of
accessions during the 2017 growing season.
109 An Automated Phenotyping Platform for Trait Discovery and Characterization in Corn
Gerie van der Heijden, DuPont Pioneer
Transgenic or genome editing approaches for agronomic traits of corn, for example to improve drought
tolerance or nitrogen use efficiency, require an accurate and precise phenotyping pipeline. The small
genetic differences may lead to relatively small phenotypic differences, and hence we need to reduce
the random variation and increase the precision of the measurements.The agronomic traits discovery
pipeline of DuPont Pioneer consists of precision phenotyping possibilities in the greenhouse as well as
extensive field testing trials in managed stress environments. The managed stress environment in the
field offers accurate measurements of the end goal: yield under different realistic stress conditions.
Precision phenotyping in the greenhouse offers the possibility to test leads year-round, to apply well-
defined stresses at an individual plant basis and to study in detail plant growth and development of each
plant with advanced equipment, like fluorescence imaging and hyperspectral imaging. In this
presentation we will show the automated greenhouse phenotyping platform for screening drought
tolerant phenotypes in corn.
110 Quantitative Trait loci Analysis of Fiber Auality Traits using an Tntrogressed Random-mated
Recombinant Inbred Population
Khezir Bhatti, Kahramanmaras Sutcu Imam University
Adem Bardak, Kahramanmaras Sutcu Imam University; Halil Tekrek, Kahramanmaras Sutcu Imam
University; Ismail Akyol, Kahramanmaras Sutcu Imam University
Cotton is one of the most important natural fiber crops in the world. Cotton is of immense importance
for global economy by producing natural fiber for textile industry and edible oil for human consumption.
In recent decades, improvements in cotton fiber quality and yield have been stagnant and unable to
meet the demands of the modern textile industry. However, yield is often negatively correlated with
fiber quality in upland cotton. Therefore, yield must at least be maintained when improving fibre quality
for a cultivar to remain competitive. Molecular linkage maps provide essential tools for plant genetic
research, facilitating quantitative trait locus (QTL) mapping, marker-assisted selection and map based
cloning. Random-mating provides a potential methodology to break this correlation. The suite of fiber
quality traits that affect the yarn quality includes the length, strength, maturity, fineness, elongation,
uniformity and color. Identification of stable fiber quantitative trait loci (QTL) in Upland cotton is
essential in order to improve cotton cultivars with superior quality using marker-assisted selection
(MAS) strategy By crossing two Gossypium hirsutum parents, cv. Nazilli 84S and IS-4, and through
subsequent selfings, we obtained recombinant inbred line (RIL) population of 165 individuals. In the
present study, we used SSR and SRAP markers to construct a linkage map using 165 RIL individuals
derived from a cross between G. hirsutum x G. barbadense. The purpose of the present research to
detect quantitative trait loci (QTL) for fiber quality and provide information applicable to cotton
breeding.
111 Gene-based Crop Models: A Modular Approach to Model Early Vegetative and Reproductive
Development of the Common Bean (Phaseolus vulgaris L.)
Christopher Hwang, University of Florida
Melanie Correll, University of Florida; Salvador Gezan, University of Florida; Li Zhang, Texas University;
Mehul Bhakta, Monsanto; Kenith Boote, University of Florida; Jose Clavijo Michelangeli, Thomson
Reuters; James Jones, University of Florida
To assist plant breeders in designing new cultivars for increased yield and meet global food demands,
the next generation of mechanistic gene-based crop models offers the potential of predicting crop
vegetative and reproductive development based on genetic, environment, and management
information. Here, we illustrate an approach for developing a dynamic modular gene-based model to
simulate time to first flower, main stem node number, and leaf area by node position of common bean
(Phaseolus vulgaris L.). Crop characteristics in the model are functions of relevant genes (quantitative
trait loci (QTL)), the environment (E), and QTL x E interactions. The model was based on data from 187
recombinant inbred (RI) genotypes and the two parents grown at five sites (Citra, FL; Palmira, Colombia;
Popayan, Colombia; Isabela Puerto Rico; and Prosper, North Dakota). The model consists of four
dynamic QTL effect models for node addition rate (NAR, No. d-1), daily rate of progress from emergence
toward flowering (RF), daily maximum main stem node number (MSNODmax), and potential leaf size for
a node on the main stem (MSLAmax), that were integrated to simulate main stem leaf area for each
node vs. time, and date of first flower using daily time steps. Funding from the Borlaug Fellowship in
2016 allowed for an additional trial with seven RI lines to be conducted at the international center for
tropical agriculture (CIAT) for model evaluation. The use of mixed-effects models to analyze multi-
environment data from a wide range of genotypes holds considerable promise for assisting
development of dynamic QTL effect models capable of simulating vegetative and reproductive
development.
112 Phenomics of Reproductive Stage Drought Tolerance in Rice
Sudhir Kumar, IARI
Dhandapani Raju, IARI; Viswanathan V. – IARI
Drought stress adversely affect yield in about 45% of the rice grown area in India. In these area rice crop
experiences intermittent cycles of drought stress and recovery and/or terminal drought stress. Extensive
efforts are being made to improve drought tolerance and WUE of rice. Large-scale phenotyping under
field conditions often led to the identification and use of the drought avoidance genotypes which
mature early or genotypes with dehydration avoidance by water mining. An experiment with 60 diverse
rice genotypes was conducted, in automated Plant Phenomics Facility of ICAR-IARI, New Delhi, to
identify genotypic variation for reproductive stage water-deficit stress tolerance. Pot grown plants were
subjected to drought stress by withholding irrigation at booting stage. The evapotranspiration was
measured for both well watered and drought set plants by using automated weighing and watering
station. The drought set of plants were rewatered when the available soil moisture content decreased
to about 40%, and again the plants were subjected to the second cycles of stress. Plants were imaged by
using visual, IR, NIR and chlorophyll fluorescence imaging platforms at regular intervals. Temporal
responses in several image-based features measured by using visual, IR, NIR and chlorophyll
fluorescence imaging platforms showed considerable variations among rice genotypes. The potential
association of these image-based features with the component mechanisms of drought tolerance
viz.,cellular dehydration avoidance and cellular dehydration tolerance will be discussed.
113 Phenotyping Thousands of T-DNA Insertion Arabidopsis thaliana Mutants with the unPAK Project:
The Power of 200,000 Phenotypes
Courtney Murren, College of Charleston
Matthew Rutter, College of Charleston; Hillary Callahan, Barnard College/Columbia University; Jim
Leebens-Mack, University of Georgia; Michael Wolyniak, Hampden Sydney College; April Bisner, College
of Charleston; Allan Strand, College of Charleston
Determining how genes are associated with complex plant phenotypes is a major challenge in modern
biology. The unPAK project — undergraduates Phenotyping Arabidopsis Knockouts — has phenotyped
thousands of non-lethal insertion mutation lines from a single genomic background in the model
organism Arabidopsis thaliana. unPAK phenotypes are complex macroscopic vigor and fitness-related
traits of ecological, evolutionary, and agricultural importance. We contextualize the mutant line
variation with natural accessions (phytometers) for which whole genome information is available. With
our database, we evaluate phenotypic consequences of insertion mutations, to identify individual
mutant lines with distinct phenotypes that are candidates for further study and to describe broad
patterns in the effects of mutation on phenotype. We screen mutant lines across environments to
uncover context dependent mutational effects. We have detected mutational effects that change fruit
production (via increases or decreases), influence trait relationships, and are outside of the variation of
natural accessions (novel phenotypes). Undergraduate researchers across more than 15 institutions are
central to unPAK activities, both as research apprentices and as students participating in courses
through CURE (course based undergraduate research experiences). unPAK is providing students an
opportunity to obtain research experience while simultaneously answering central questions in plant
phenomics.
114 Nitrogen Utilization Efficiency of Arabidopsis Natural Variants
Shanhua Lin, Institute of Molecular Biology, Academia Sinica
Yi-Fang Tsay – Institute of Molecular Biology, Academia Sinica
Hundred million tons of nitrogen fertilizer was consumed each year. High demand of nitrogen fertilizer
not only makes the cost of farming and energy consumption increase, but also leads to eutrophication in
river and ocean and makes marine dead zone expanded. Improving nitrogen utilization efficiency (NUE)
for crops is an essential and urgent issue to increase crop production and at the same time reduce N
fertilizer demand. We would like to identify new strategy to improve NUE by genome-wise association
(GWAS) approach. This approach may help us to isolate NUE associated genes and to understand how
NUE is regulated. 95 Arabidopsis natural variations were grown with NH4NO3 at five different
concentrations, ranging from 0.2 mM (nitrogen limited) to 10 mM (nitrogen sufficient) for 21 days, and
then the biomass (dry weight) of rosette leaf were measured. Some ecotypes showed linear growth
increase when N was increased from 0.2 up to 10 mM. But, some ecotypes showed growth inhibition by
high N. The GWAS analyses of several traits, including the biomass at different nitrogen concentrations
and the ratio of biomass at different concentration of nitrogen were performed using the GWA-Portal
(https://gwas.gmi.oeaw.ac.at) platform to identify the potential genes associated with NUE at
vegetative stage. Twenty-seven SNPs with -log(p-value) >5 were observed. Further analysis of mutants
of these potential candidate genes nearby those SNPs will help us to find out novel regulatory
mechanism of NUE.
115 The National Crop Phenomics Center in Korea
Kyung-Hwan Kim, National Institute of Agricultural Sciences
Inchan Choi, National Institute of Agricultural Sciences; Hyeonso Ji, National Institute of Agricultural
Sciences; YongSeok Jung, Chungnam National University; TaekRyoun Kwon, National Institute of
Agricultural Sciences
It is well known that crop phenomics is an emerging area in the era of ‘omics’. Crop phenomics needs a
convergence of biotechnology, information technology, automated machine technology and so on, in
order to measure the morphological and physiological traits of crop plants. The National Institutes of
Agricultural Sciences (NIAS), Korea, has invested on agricultural biotechnology to create a new variation
of crop plants. So far there are many genes to be tested for their phenotypes. So NIAS sets up a high-
throughput crop phenotyping (HTCP) facility with size of 1,000 pots in this year. The HTCP facility
consists of an environmental control greenhouse, image acquisition devices, convey belt and data
management programs. At first, this center tried to analyze height, leaf area, and leaf color of rice plants
using RGB images. This center shall continue to analyze various important traits of rice plants such as
abiotic stress tolerances and yield components in the future.
116 Rapid Ground-based Phenotyping Methods for Canopy Cover and Canopy Reflectance Estimates in
a Durum Wheat Diversity Panel
Maria Newcomb, University of Arizona, School of Plant Sciences
Jeffrey White, USDA-ARS Arid Land Agricultural Research Center; Richard Ward, University of Arizona;
Michael Pumphrey, Washington State University; Roberto Tuberosa, University of Bologna; John Heun,
University of Arizona; Andrew French, USDA-ARS Arid Land Agricultural Research Center; William
Luckett, USDA-ARS Arid Land Agricultural Research Center; Pedro Andrade-Sanchez, University of
Arizona
There is increasing demand for simple, rapid methods to evaluate crop plants growing under field
conditions to separate and identify the interacting effects of genetics, environment and management on
crop growth and development. We assessed canopy cover using digital images and canopy reflectance
using spectral data collected from a tractor-mounted phenotyping system for a diversity panel of 260
accessions of durum wheat (Triticum turgidum Desf.), which was grown in 2016 at the Maricopa
Agricultural Center in Arizona. The diversity panel lines originated from different countries across the
wheat-growing areas, and was previously genotyped using a high-density single-nucleotide
polymorphism (SNP) assay. We manually photographed (Lumix DMC digital camera) each plot at two
positions within the plot, using a custom-built metal A-frame stand designed to ensure a nadir view from
a constant height and providing a field of view matched to the 0.76 m wide rows. Using a tractor-
mounted sensor array system, we measured canopy reflectance with a multi-spectral crop sensor
deployed on the front boom. Canopy cover was estimated from each digital camera image using a batch
script implemented in ImageJ (1.49V), with analysis of 600 images requiring approximately one hour.
Spectral reflectance measurements were recorded with a data logger and the data corresponding to the
red (670nm) and NIR (780nm) wavelengths were used to calculate the normalized difference vegetation
index (NDVI). Plot-level estimates of canopy cover and NDVI were assessed at different times during
crop growth. Comparisons of the two phenotypes showed good agreement but also high-lighted
potential confounding effects of plant population. Both methods show value as simple, low-cost options
for field-based phenotyping to facilitate genotype-phenotype association studies and better understand
the impacts of abiotic and biotic stresses on crops.
117 Functional Phenomics: Relating Phenes to Function using High-throughput Phenotyping and Data
Analytics
Larry York, Samuel Roberts Noble Foundation
The emergence of functional phenomics signifies the rebirth of physiology as a 21st century science
through the use of advanced sensing technologies and big data analytics. Functional phenomics
highlights the importance of phenotyping beyond only identifying QTL and gene candidates. Linking
form to function is imperative for human welfare, for while great advances have been made in genetics,
we have relatively less knowledge of what plants do. Which plant properties are important for
ecosystem services like forage production, grain yield, and soil carbon sequestration? Using vast
datasets of predictors (plant properties) and responses (yield) we can infer causal relationships using
multivariate and big data analytics. These data are generated by high-throughput phenotyping of
genetic mapping populations, so function and genetics are simultaneously and synergistically studied.
This presentation will outline a general approach for the theory and practice of functional phenomics.
First, the concept of phene will be explored as the fundamental unit of phenotype as a replacement for
more ambiguous terms such as ‘trait.’ The lack of semantics and grammar for phenomics hinders both
thought and experimental work. Next, the functional phenomics pipeline is proposed as a general
method for conceptualizing, measuring, and validating utility of plant phenes. The functional phenomics
pipeline begins with ideotype development. Second, a phenotyping platform is developed to maximize
the throughput of phene measurements. A mapping population is screened measuring target phenes
and indicators of plant performance such as yield and nutrient composition. Traditional forward genetics
allows genetic mapping, while functional phenomics links phenes to plant performance. Based on these
data, genotypes with contrasting phenotypes can be selected for smaller yet more intensive
experiments to understand phene-environment interactions in depth. Simulation modeling is also used
to understand the phenotypes and all stages of the pipeline feed back to ideotype and phenotyping
platform development.
118 Initial Assessment of Utilizing High-throughput Phenotyping in a Cotton Breeding Program
Alison Thompson, USDA-ARS
Kelly Thorp, USDA-ARS; Matthew Conley, USDA-ARS; Andy French, USDA-ARS
High-Throughput Phenotyping (HTP) is quickly becoming a highly sought after technique for rapid trait
measurements of plants in field and greenhouse settings. With changing technologies, the capabilities
of HTP increase each year, opening many opportunities for incorporation into different research
programs. One area of research that could greatly improve with the application of HTP is plant breeding,
but how best to incorporate this technology is still unclear. In 2009, a HTP tractor was developed in
Maricopa, AZ, through a joint effort between the US Arid Land Agricultural Research Center and The
University of Arizona, Maricopa Agricultural Center. Since 2012, the tractor has been used to evaluate
upland cotton for heat and water stress tolerance as part of a national breeders testing network. The
experimental design incorporated 35 cotton breeder lines as well as checks, which were planted in 2-
row, 12-meter long plots, spaced 1-meter apart. Well-watered and water-limited treatments were
applied when 50% of lines exhibited first flower, and continued until harvest. The objectives of this
research were to evaluate tractor based HTP in a breeding program and validate tractor measurements.
Plant height, canopy temperature, and spectral reflectance values were captured using proximal
sensors on the tractor once per week at solar noon. The proximal sensor data was analyzed using a
custom developed processing pipeline. Three cotton lines were identified as potentially drought tolerant
using plant canopy temperature and will undergo further evaluation. Challenges to incorporate tractor
based HTP in a breeding program include tractor logistics and engineering, field design, data processing
and management, data validation, and determining the best statistical approaches to analyze large
quantities of data.
119 Testing Character Displacement in the Belowground Root Traits of Two Closely Related Morning
Glory Species
Sara Colom, University of Michigan
Regina Baucom, University of Michigan
Character displacement, the processes whereby competing species evolve in their resource associated
traits as a response to competition, may be a crucial process underlying coexistence between closely
related plants, and further, may explain why such species are phenotypically divergent in regions where
they co-occur. Despite its potential to explain patterns of diversity and improve our understanding of
the adaptive value of diverse resource associated traits, research on character displacement in plant
systems is limited primarily to above-ground traits. As a consequence, consideration of character
displacement in belowground root traits is currently lacking. My dissertation research is aimed at
characterizing the belowground root structures in two species of morning glories, Ipomoea hederacea
and I. purpurea, and examines how variation in root traits between these two species may contribute to
their competitiveness. As a first step I conducted a preliminary common garden experiment where I
planted replicate seeds of multiple maternal lines of I.hederacea and I.purpurea individually in custom
built rhizotrons and phenotypically characterized their root structures during early growth. I measured
various root traits, examined how the species differ in their respective trait space and identified whether
maternal line variation existed in these traits. I found evidence for maternal variation in some individual
root traits, suggesting that under natural conditions selection from competition may influence root trait
evolution and subsequently lead to character displacement between the species.
120 Genetic Control of Maize Biochemical Defense Revealed by An Integrated Metabolomics,
Transcriptomics and Quantitative Genetics Approach
Shaoqun Zhou, Cornell University
Georg Jander, Boyce Thompson Institute
Appropriate selection and accurate measurement of quantitative phenotypes are two major challenges
in identifying the actual genetic basis of quantitative trait loci (QTL) related to plant disease resistance.
Classical QTL mapping with direct measurement of disease severity is often affected by highly variable
disease outcome and subjective severity scoring. In this study, we found that Fusarium graminearum-
resistant maize (Zea mays) inbred line Mo17 constitutively accumulates many metabolites that are
inducible by the same pathogen strain in the susceptible inbred line B73. We hypothesized that the
contrasting F. graminearum susceptibility levels of these two inbred lines could be attributed to
differences in their constitutive biochemical defense, and further, to the genes regulating the
accumulation of these defense-related metabolites. Hence, we conducted non-targeted metabolite
profiling by high resolution HPLC-MS and RNAseq transcript profiling of 80 B73 x Mo17 recombinant
inbred lines, thereby identifying significant metabolite QTL (mQTL) for over 700 mass features and
expression QTL (eQTL) for 1,200 genes that show constitutive differences between B73 and Mo17
seedling roots. Through this analysis, we identified an mQTL hotspot affecting over 70 mass features,
including ones known to be defense-related. To pinpoint the causative gene(s) underlying this mQTL
hotspot, we performed a correlative network analysis across the metabolomic and transcriptomic data,
and thereby identifying a gene that likely encodes a vesicular transport protein as the most probable
candidate. Transposon insertions in this gene result in significant depletion of DIMBOA-glucoside, a
major defense-related metabolite in maize seedling roots, which could cause compromised defense
against F. graminearum. In summary, we dissected the complex and highly variable disease severity
phenotype into an array of biochemical components that can be quantified more accurately.
Furthermore, we identified a gene affecting this complex disease resistance phenotype by combining
high-throughput transcriptomics, metabolomics, and quantitative genetics approaches.
121 Using field-based High-throughput Phenotyping for Genetic Discovery in Rice
Brook Moyers, Colorado State University
Paul Tanger, Colorado State University; Stephen Klassen, International Rice Research Institute; Julius
Mojica, Duke University; John Lovell, University of Texas, Austin; Brook Moyers, Colorado State
University; Marietta Baraoidan, International Rice Research Institute; Maria Elizabeth Naredo,
International Rice Research Institute; Kenneth McNally, International Rice Research Institute; Daniel
Bush, Colorado State University; Hei Leung, International Rice Research Institute; Jan Leach, Colorado
State University; John McKay, Colorado State University
To ensure food security in the face of population growth, decreasing water and land for agriculture, and
increasing climate variability, crop yields must increase faster than the current rates. Increased yields
will require implementing novel approaches in genetic discovery and breeding. We demonstrate the
potential of field-based high throughput phenotyping (HTP) on a large recombinant population of rice
(Oryza sativa) to identify genetic variation underlying important traits. We find that detecting
quantitative trait loci (QTL) with HTP phenotyping is as accurate and effective as traditional labor-
intensive measures of flowering time, height, biomass, grain yield, and harvest index. Genetic mapping
in this population, derived from a cross of an modern cultivar (IR64) with a landrace (Aswina), reveals
four alleles with negative effect on grain yield that are fixed in IR64, demonstrating the potential for HTP
of large populations as a strategy for the second green revolution.
122 Comparison of Genome-wide Association Study for First Flowring Time with Different Size of
Soybean Panel
Jung-Kyung Moon, National Acedemy Agriculture Science
Sookwon Park, National Institite of Crop Science; Sungtaeg Kang, Dankook university; Soon-Chun Jeong,
Korea Research of Bioscience & Biotechnology; Namshin Kim, Korea Research Institute of Bioscience &
Biotechnology; Yong Seok Jung, Chungnam National University
Recent achievements in soybean [Glycine max (Merr.) L.] genomics by the development of high
throughput genotyping technology have produced large amount of genomic polymorphism data and
therefore, genome-wide association studies (GWAS) become feasible tool to identify associations
between genomic variation and quantitative traits. Here, we provide the information about current
progress in soybean GWAS study using different size of core collection and original populations in
Korean soybean. The Korean soybean core collection was constructed using 391 and 814 soybean
varieties and recently developed 180K Axiom® Soybean Genotyping Array was employed to obtain the
genome-wide variation information. The trait of first flowering time was used to evaluate the GWAS
procedure using GAPIT, and its application to identify the novel genes. The result showed that precise
location of previously reported genes (e.g., GIGANTEA homologos, GIa, for flowering time) were
detected in small size of core collection. However, two other flowering genes, previously reported for
E2, E3, were well detected in bigger size of core collection set and original populations, Our results
suggested that GWAS in soybean with the size of 400 accessions could provide opportunity to discover
the novel genes/QTLs for various traits for sustainable crop breeding.
123 High Throughput Phenotyping and Genetic Dissection Reveal Plant Vigour and Water Saving
Traits are Co-mapped with Root “QTL-hotspot region” in Chickpea (Cicer arietinum L.)
Sivasakthi Kaliamoorthy, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT)
Mahendar Thudi, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT); Murugesan
Tharanya, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT); Sandeep Kale,
International Crops Research Institute for the Semi-Arid Tropics (ICRISAT); Jana Kholová, International
Crops Research Institute for the Semi-Arid Tropics (ICRISAT); Mahamat Halime, International Crops
Research Institute for the Semi-Arid Tropics (ICRISAT); Deepa Jaganathan, International Crops Research
Institute for the Semi-Arid Tropics (ICRISAT); Rekha Baddam, International Crops Research Institute for
the Semi-Arid Tropics (ICRISAT); Thiyagarajan Thirunalasundari, Bharathidasan University; Pooran Gaur,
International Crops Research Institute for the Semi-Arid Tropics (ICRISAT); Rajeev Varshney,
International Crops Research Institute for the Semi-Arid Tropics (ICRISAT); Vincent Vadez, International
Crops Research Institute for the Semi-Arid Tropics (ICRISAT)
Chickpea is the 2nd most important legume and terminal drought stress leads to substantial yield losses.
Water conservation at vegetative growth (due to canopy conductivity, size, and development) allows
plants to increase soil water extraction during grain filling and is hypothesised to help chickpea
adaptation to water limitation. With an aim of identifying genomic regions for these canopy traits, 232
RILs, derived from a cross between ICC 4958 and ICC 1882, were phenotyped at vegetative stage under
well water conditions using a high throughput phenotyping platform (LeasyScan). These data competed
other phenotype data for root traits and field-based assessment of agronomic traits collected in earlier
mapping studies. Different density genetic maps (241-SSR-Low density, 1007-SSR+SNPs-High density and
1557-SNPs-Ultra high density) were used for QTLs identification. Several major QTLs (M-QTLs) for plant
vigour related traits (3D-leaf area, shoot biomass, plant height and growth related traits) were identified
on CaLG04. These co-mapped with a previously identified major drought tolerance QTL-hotspot region
on CaLG04 (~300Kb). Furthermore, canopy conductance traits QTLs were mapped on CaLG03.
Comparative analysis of plant vigour and canopy conductance traits with different density genetic maps
resulted in decreased QTL size and increased LOD score and % of phenotypic variation considerably.
Majority of the QTL interactions identified by GMM at single locus were consistent with those detected
by CIM analysis of QTL cartographer. Most of plant vigour traits had alleles from ICC4958 whereas in
case of canopy conductance traits, favourable alleles were contributed by ICC1882. Plant vigour traits on
CaLG04 and canopy conductance related traits on CaLG03 provide opportunity to manipulate these loci
to tailor recombinants having lower transpiration rate and high plant vigour. This ideotype might be
useful in enhancing the water stress adaptation in chickpea. This work also demonstrated the value of
combining phenotyping at different levels of plant organization.
124 Determination of Cold Tolerance QTLs via High Throughput Photosynthetic Phenotyping
Donghee Hoh, Michigan State University
Isaac Osei-Bonsu, Michigan State University; Jeffrey Cruz, Michigan State University; Linda Savage,
Michigan State University; David Hall, Michigan State University; David Kramer, Michigan State
University
Global warming affects the crop yields, so there is increasing the need for improvement of the
efficiency, robustness, and sustainability of food crops. Cowpea (Vigna unguiculata (L.) Walp.) is a warm
season grain legume crop commonly cultivated in developing countries (Wellington, 2009). It is an
important source of protein and symbiotic N fixation. In addition, many varieties have resistance to
drought, low soil fertility (Ehlers ad Hall, 1997). However, heat stress during the flowering stage, insects
attacks in hot summer cause severe negative impacts on cowpea production. Early planting of cowpea
(when seasonal temperatures tend to be low) can be considered an opportunity to avoid the insect’s
attacks and heat stress. However, a major concern is how the chilling stress impacts performance/yield.
The goal of this research is to figure out chilling stress related genes and mechanisms. Dynamic
Environmental Photosynthesis Imager (DEPI) and photosynQ are powerful phenotype tools for
photosynthesis developed in Kramer Lab were used to collect photosynthetic data for QTL mapping.
Two approaches for this research are the mapping of quantitative trait loci (QTLs) to identify genes
related to chilling stress tolerance and identify biophysical/biochemical mechanisms of cold-tolerance in
photosynthesis. Low day/night temperature decreases ΦII and increases photoinhibition, and this
effect was observed significant genetic variance, so this condition was used for mapping QTLs related to
cold tolerance.
125 Studies on microRNAs in Selected Species of Zingiberales, Liliales, and Arecales: Monocot Orders
that Develop Fleshy Fruits.
Divya Nair, Texas Tech University
Chris Rock, Texas Tech University
Small RNAs such as microRNAs (miRNAs) are fundamental to the integrity and cohesion of the genome,
and a powerful influence on genome evolution. Economically important gingers, banana, and palms
develop fleshy fruits. Our study hypothesizes that fleshy fruit development is plesiomorphic (an ancient
basal trait) in monocots and mediated in part by MICRORNA828 (MIR828) which is a negative regulator
of anthocyanin and polyphenolic nutraceuticals biosynthesis. The implication is that species in those
monocot taxa with fleshy fruits will be proved in due course to have functional MIR828 genes, whereas
it is predicted that the most closely-related species that have recently evolved wind-dispersed seeds
such as Xerophyllum, vis-à-vis Trillium of the basal order Liliales, will not have a functional MIR828 locus.
The fact that gymnosperms and the single extant species (Amborella) of the sister lineage to all
flowering plants have a functional MIR828 gene is consistent with the hypothesis. We are integrating
genomic data across scales and species to explore the role of MIR828 and its target MYB transcription
factors' function in fruit development and nutraceutical metabolism. Results may impact our
understanding of homoplasy/convergent evolution: seemingly independent apomorphic appearances
based on cladistics of developmental similarity to aid endozoic seed dispersal by some monocots.
Emerging Technologies
200 Identifying Molecular Networks Underlying Root-root Interactions in High Plant Density Adapted
Maize
Nathanael Ellis, Donald Danforth Plant Science Center
Kari Miller, Donald Danforth Plant Science Center; Keith Duncan, Donald Danforth Plant Science Center;
Jody Edwards, USDA-ARS; Yuan-Chuan Tai, Donald Danforth Plant Science Center; Christopher Topp,
Donald Danforth Plant Science Center
Since the 1930’s, annual gains in U.S. maize production have largely been driven by the adaptation of
maize plants to increasing plant density. Two historically important maize populations, Iowa Stiff Stalk
Synthetic and Corn Borer, have been recurrently selected for increased hybrid grain yield, resulting in
density adapted plants. Shoot architectures, specifically increased leaf angles, are known outcomes of
selection during maize density adaptation, but little empirical data exists about the contribution from
roots. Preliminary work suggests that not only have root architectures changed in these populations, but
their sensing and growth responses to neighboring root systems have changed as well, pointing to a new
paradigm for understanding the key process of crop density adaptation. Selected cycles from
progressive density adapted germplasm will be assayed in several ways to quantify root system
architecture of plants grown individually or paired, to identify changes in root growth as a function of
density adaption. In this project, the effects of density adaptation on 3D root architecture and growth
will be assayed by Optical Projection Tomography, X-ray Computed Tomography, and Positron Emission
Tomography to study dynamic carbon allocation. These efforts will identify the most relevant time
points from which to sample leaf and root tissue for transcriptional profiling, as well as profiling of root
exudates via gas chromatography-mass spectrometry. This project will contribute to modeling the
relationship of carbon allocation and root growth in respect to root-root communication and to identify
the underlying genetic mechanisms controlling these processes. This research aims to identify the
molecular mechanisms of adaptation to high plant density in maize roots. Thus the research will
contribute to our basic understanding of how roots, the "hidden-half" of plants, communicate with one
another and exploit this knowledge for crop improvement. This material is based upon work supported
by the National Science Foundation under Award IOS number 1612033.
201 CHESS-U: A New Resource for Synchrotron-based X-ray Microscopy for High-precision Plant
Phenotyping
Arthur Woll, Cornell High Energy Synchrotron Source
Olena Vatamaniuk, Cornell University; Ju-Chen Chia, Cornell University; Haujin Sheng, Cornell University;
Jaipei Yan, Cornell University; Joel Brock, Cornell University
In the last decade, synchrotron-based x-ray microscopy has become increasingly recognized as a critical
tool for investigating metal distributions and speciation in plants and other biological systems. In
particular, such information has been utilized to determine or clarify genetic influences on metal
transport, signaling and storage. 2D and 3D implementations of x-ray fluorescence (XRF) mapping at the
micron scale can quickly and efficiently reveal quantitative variations in the total concentration and
distributions in a wide variety of samples with minimal sample preparation. Furthermore, x-ray
absorption fine structure (XAFS) permits obtaining metal speciation in vivo. Finally, high-energy phase
contrast imaging can image plant structure at the micron scale in sub-second time-scales. Yet, the utility
of these methods to the plant research community as a whole is limited both by available beamtime at
existing facilities, and, arguably, by a lack of specialized instrumentation to support particular needs of
plant research.
The Cornell High Energy Synchrotron Source (CHESS) has recently identified plant genotype / phenotype
interactions as one of seven science drivers motivating a two-year upgrade project, “CHESS-U,” to be
completed in late 2018. Recent applications of x-ray microscopy at CHESS will be described, both to
illustrate its utility and to emphasize future capabilities and opportunities enabled by CHESS-U.
202 Applying Hyperspectral Imaging to Studying Temperature Stress Responses in Maize
Tara Enders, University of Minnesota
Nathan Springer, University of Minnesota
Average yields of maize could drop substantially within the next century due to global climate change.
Understanding how maize varieties respond to temperature extremes will be instrumental in developing
varieties that can withstand future temperature stresses while still producing high yield. Hyperspectral
imaging of maize in response to abiotic stress is an unexplored concept, and is likely to provide useful
data to researchers interested in both abiotic stress and hyperspectral imaging. Hyperspectral imaging
tools are being developed to study phenotypes of temperature-stressed maize seedlings. Current efforts
include minimizing environmental and plant morphological effects on obtained data, as well as
developing methods for data analysis in MATLAB and R. We will document the variation of
hyperspectral traits in maize seedlings in response to both high and low temperature stresses in multiple
maize genotypes.Documenting spectra across genotypes and growth conditions will uncover the
dynamics of maize spectra in response to changing temperatures and allow for the discovery of genomic
loci that could provide improved tolerance.
203 Detection and Identification of Bitter Compounds in Alfalfa Protein Isolates using a Combination
of Sensory and Analytical Approaches
Pavel Aronov, Impossible Foods
David Ruiz, Impossible Foods; Harshada Natekar, Impossible Foods; Michelle Mai, Impossible Foods;
Ranjani Vardan, Impossible Foods; Pat Brown, Impossible Foods
Alfalfa is highly valued as a feed crop for cattle, but its consumption by humans is limited to alfalfa
sprouts because phytochemicals present in mature plants make it unpalatable (bitter and astringent) or
unsafe for human consumption (canavanine and other antimetabolites). Alfalfa protein isolate could be
suitable for human consumption if undesirable co-purifying metabolites were efficiently removed. To
achieve such a goal we need to identify and characterize the compounds responsible for off-flavors in
alfalfa protein extract. We prepared protein-enriched extracts from alfalfa leaves and fractionated them
by solid phase extraction (Oasis HLB, Waters) or semi preparative reversed phase Ultimate 3000 HPLC
with Hypersyl GOLD C18 column (Thermo Scientific). The fractions were evaporated, tested for trace
solvents, heavy metals and known toxic metabolites to ensure safety for human consumption and
subjected to a human sensory panel to identify bitter fractions. Bitter fractions were further analyzed by
Q Exactive Plus LC/MS to identify their small-molecule constituents. We also explored an alternative
metabolomics approach in which we subjected diverse alfalfa protein isolates to untargeted LC/MS
analysis and searched for correlations between variation among the isolates in the abundance of
individual small molecules and variation in the intensity of bitter taste. As a result, several candidate
bitter small molecules were identified in alfalfa and confirmed by sensory panel assays.
204 A Motorized High-Throughput & High-Accuracy Plant Stand Analyzer for Large Scale Maize
Breeding Programs
Lie Tang, FieldRobo LLC
This poster presentation reports a newly developed motorized high-throughput & high-accuracy plant
stand analyzer (PSA) for large scale maize breeding programs. At the core of the PSA technology is its
high-speed laser scanning and data processing unit which generates multiple plant stand phenotypic
data metrics including stand count, stand spacing, stalk size, tillers, and skips and doubles all in real-
time. Currently PSA can be towed by an ATV to measure 4 rows simultaneously or be hitched to a
tractor to form an 8-row system. PSA has a capacity of moving at 4 mph and sense maize plants from V4
to V7 when using ATV or tractor as its carrier. The productivity of a 4-row PSA will be equivalent to a
human crew of 23 tireless workers. If high ground clearance vehicles such as sprayers are used, PSA can
measure maize plants of up to R1 growth stage. PSA is equipped with a compressed air system to keep
the sensors clean from dust. Extensive field experiments were conducted to evaluate the performance
of PSA. On average PSA has achieved 98% stand counting accuracy with 1% repeatability error. A post-
pressing software program will fuse the RTK-GPS and wheel odometry data to accurately geo-tag each
plant stand. Subsequently the generated plant stand map can be imported into GIS software and
overlaid with planting polygons to aggregate stand data into rows and plots for statistical analysis.
Overall, PSA has for the first time made field-based plant stand analysis not only highly practical and
economical but also highly accurate for commercial maize breeding programs.
205 Indiana Corn and Soybean Innovation Center: Purdue’s Automated Field Phenomics Facility
Jason Adams, Purdue University
April Carroll, Purdue University
Purdue’s Indiana Corn and Soybean Innovation Center (ICSC) is a 25,500 square-foot research facility at
the Agronomy Center for Research and Education, a 1,400 acre research farm in West Lafayette,
Indiana. ICSC was designed to facilitate research on improved plant traits and varieties and to develop
profitable farming practices. These goals are achieved by facilitating collaborations between
agronomists, basic plant biologists, engineers, aviation scientists, data scientists, and disciplines beyond.
In addition to high-tech workspaces for diverse collaborators, the facility features an automated
threshing and shelling line for plant and seed processing as well as ample seed and plant processing labs.
Researchers also use phenotyping equipment such as root and leaf scanners, root washing stations, seed
counters and color sorters, as well as 3-D printers and scanners, ovens and grinders. The Center also
includes a 5,000 ft2 high bay phenomics tool development lab. This large flexible workspace is designed
for research and development of remote sensing platforms such as UAVs and Purdue’s PhenoRover, a
ground-based mobile platform, and for experimentation with novel measurement tools, cameras and
sensors. With plans to collect up to ten terabytes of data per week, a high speed fiber-optic cable
transfers data back to Purdue’s campus data storage and supercomputing clusters. ICSC is a core
component of the Institute for Plant Sciences, part of Purdue Moves, with investments over $20 million
for world-changing research. The Center is also supported with a combined $4 million investment from
the Indiana Soybean Alliance and Indiana Corn Marketing Council, and donors including AgReliant
Genetics, DuPont Pioneer, Ag Alumni Seed and ALMACO. The unique data collection capabilities in
addition to the multi-disciplinary collaborations at the Indiana Corn and Soybean Innovation Center
greatly enhance the ability of Purdue University researchers to address the grand challenge of food
security in the 21stcentury.
206 Introducing Moving Circular Spatial Adjustment for Large Scale High-Throughput Phenotypic Data
Extracted from Unmanned Aerial System (UAS) Imagery
Atena Haghighattalab, Kansas State University
Suchismita Mondal, CIMMYT; Ravi Prakash Singh, CIMMYT; Jesse Poland, Kansas State University
Unmanned aerial systems (UASs) have the potential to provide breeders with phenological data such as
plant height and vegetation indices (VIs) to help with more accurate selections on larger populations.
The phenotypes are determined by the genetics, but also influenced by the spatial variation across the
field. To improve the accuracy of selections, breeders need to account for non-heritable environmental
variation in the form of spatial field effects. There has been extensive investigation into field design and
statistical models to account for field effects, however this previous work was generally limited in the
overall treatment of the breeding field experiments as geospatial entities.
The objective of this study is to evaluate the use of coordinate based geostatistical spatial analysis for
controlling the heterogeneity in yield data as well as high throughput phenotypic data derived from UAS
imagery in a large wheat breeding nursery. We analyzed the temporal phenotypic data extracted from
UAS imagery for two environments during the growing season and detected the spatially non-random
patterns in the data. To correct the trends observed in the field, we used the coordinates of wheat plots
extracted from geo-referenced aerial imagery in a latitude-longitude grid to obtain an adjusted
phenotypic value of the plot in the center of the moving circular window. After spatial adjustment, we
found an increase in broad-sense heritability for VIs obtained from UAS platforms and grain yield in both
environments. The analysis implemented in this study enables breeders with collection and post-
processing of HTP measurements on larger populations and can provide increased accuracy for selection
in plant breeding. Overall the use of UAS imaging to improve selection in plant breeding has great
promise when combined with analysis and statistical methods that fully leverage the spatial nature of
the breeding trial data.
207 High-throughput Infrared Imaging: the New Crop Breeding Frontier
Smitha Kurup, Maharashtra Hybrid Seeds Company (Pvt) Ltd
Abhijeet Shillak, Maharashtra Hybrid Seeds Company Limited; Viajy Dalvi, Maharashtra Hybrid Seeds
Company Limited; Govindaraj K, Maharashtra Hybrid Seeds Company(Pvt) Limited; Bharat Char,
Maharashtra Hybrid Seeds Company(Pvt) Limited; Usha Zehr, Maharashtra Hybrid Seeds Company (Pvt)
Limited
The efficiency and adaptability of crops to various biotic and abiotic stresses must be improved
significantly to achieve global food demand by 2050. Towards this goal, phenomics has recently
emerged as a promising tool for accurate phenotyping of large set of genotypes. Phenomics includes
non-invasive techniques using sensors along with advanced computational methodologies for predictive
breeding. In the present study, thermal cameras were used for screening drought and salinity effects in
wheat and rice at field and greenhouse conditions. The experimental set up for detecting salinity effects
included rice (Oryza sativa) cultivars classified as two different categories ie., tolerant and susceptible
groups, subjected to varying levels of salt concentrations. Infrared image processing cope with
appropriate statistical analysis explicitly demonstrated that for salinity studies there is significant
difference between tolerant and susceptible cultivars at both 15 mM and 30 mM. This non-destructive
screening method will be helpful to identify salt tolerant as well as susceptible cultivars before the onset
of symptoms. Similarly, drought studies were performed on wheat (Triticum aestivum) accessions along
with checks at critical stages of drought. Field-based thermal profiling of wheat indicated a negative
correlation of canopy temperature with yield under stress condition. Infrared imaging enabled to screen
promising superior performing lines from large set of wheat lines. The relevant data of this study have
been integrated into an in-house developed phenomics database. With advancements in phenome data
capturing and analysing capabilities, phenomics-assisted breeding can be a reality in the near future.
Complementary use of a variety of sensors for screening a particular trait will increase the efficiency of
crop genetic improvement to meet future needs.
208 Stable Isotopic Labeling of Intact Plants for Molecular Turnover Measurement by Mass
Spectrometry: New Labeling Apparatus and Data Processing Approaches
Dana Freund, University of Minnesota
Calvin Peters, University of MInnesota; Aaron Rendahl, University of Minnesota; Erin Evans, University of
Minnesota; Jerry Cohen, University of Minnesota; Adrian Hegeman, University of Minnesota
There has been a recent rise in the use of high-throughput bioanalytical, ‘omics-scale’ techniques such
as metabolomics (metabolites) and proteomics (proteins). Stable (non-radioactive) isotopes as labels
can be applied for proteome or metabolome-scale labeling for the improvement of metabolite
annotation, metabolic pathway elucidation, quantification, and the measurement of molecular
dynamics. Plants can be grown under conditions where then metabolically incorporate either 15N from
mineral nutrients or 13C from labeled carbon dioxide gas. By switching between labeled and unlabeled
metabolic inputs and monitoring using mass spectrometry-based proteomics and metabolomics
approaches over time, it is possible to derive biomolecular turnover information for proteins and
metabolites. Here we discuss high-throughput computational tools for deriving turnover information
from high-resolution mass spectrometry data as well as some recent innovations in plant transient 13C-
labeling chambers. Automated modular 13C-carbon dioxide labeling chambers were designed to allow
plant materials to be harvested at various times following the initiation of labeling without perturbing
plants harvested at other time points. Proteins or metabolites are extracted from plant material and
analyzed by liquid chromatography tandem mass spectrometry (LC-MS/MS) using a high-resolution
accurate mass instrument. Here we present the results of our first plant growth experiments using the
modular chamber design. MS analysis of plant materials is used to demonstrate the extent of isotope
enrichment as a function of time during labeling experiments. In addition to labeling apparatus, we also
provide an update on the current status of our software tools (all open source and written in R) for
calculating turnover of proteins and metabolites in an automated high-throughput manner. The tools in
some cases use maximum likelihood estimation to derive fractional isotope enrichment and then fit
those enrichment values over time to obtain first order kinetic constants.
209 Detecting Northern Leaf Blight from Field Images using Machine Learning
Tyr Wiesner-Hanks, Cornell University
Chad DeChant, Columbia University; Siyuan Chen, Columbia University; Ethan Stewart, Cornell
University; Jason Yosinski, Geometric Intelligence; Michael Gore, Cornell University; Rebecca Nelson,
Cornell University; Hod Lipson, Columbia University
Northern leaf blight (NLB) is a fungal foliar disease of corn that has grown more severe in recent years,
causing roughly $2 billion in losses in the US in 2015. Geneticists, breeders, and growers all need to
accurately detect and quantify NLB infection in its early stages, but scouting for symptoms is very time-
consuming. Automatic detection of NLB lesions in images taken from the air or ground would be
preferable, but these images are often very noisy. We used human annotations to train a convolutional
neural network (CNN), a class of machine learning model, to detect NLB lesions in field images taken
from the ground with 97% accuracy. Testing of canopy images is ongoing, and early results are
promising.
210 High-throughput Phenotyping Platform for Compact Plants at WSU
Kiwamu Tanaka, Washington State University
Karen Sanguinet, Washington State University; Hans-Henning Kunz, Washington State University
Considering recent adverse environmental changes, increasing food demand, and predicted population
growth, we are in urgent need of designing more abiotic and biotic stress resistant crop species. A
complete understanding of the genomic mechanisms responsible for plant stress responses is required
to ensure sustainable food resources for society. High-throughput phenotyping will accelerate plant
research by overcoming the current technical limitations in phenomics. Washington State University
(WSU) envisions continued investment in phenomics. As we accumulate infrastructure for phenomics
research, we will start operating a high-throughput phenotyping service for compact plants other model
organisms as a core facility to enhance the capabilities of multiple researchers. The automated
phenomics platform, named Octopus (LemnaTec), will be built into a light/temperature/humidity/CO2-
controlled growth chamber. The platform carries a suite of cameras for visible RGB, photon counting,
pulsed chlorophyll fluorescence, and near-infrared imaging systems, as well as a Laser 3D Scanner. The
platform can accomplish phenomics for ~6000 Arabidopsis or ~500 crop seedlings in a single
autonomous experiment. High-content screening enabled by the broad detection power will result in an
in-depth understanding of phenotype to genotype relationships. Understanding genome-phenome
relationships accurately and in detail will help scientists unveil molecular mechanisms involved in plant-
environment interactions. For example, genomic-based prediction of plant performance under multiple
stress conditions and a complex genetic architecture of heritable phenotypic traits can be addressed
with this system. The goal is to establish this phenomics platform as a paid-use resource for the WSU
community and beyond.
211 A Novel Method for the Continuous and Direct Measurement of Water-stress in Tissues and Whole
Plants
Timothy Aston, University of Wyoming
Brent Ewers, University of Wyoming; Earl Wood, University of Wyoming; Carmela Guadagno, University
of Wyoming
Technological developments in our ability to phenotype plants has not kept pace with our ability to
sequence and edit genomes, resulting in whole-plant screening methods becoming the bottleneck to
applying plant biotechnology. The plant vascular system contains water with charged solutes, which
forms a network highly conductive to electricity. Previous work has tested the impact of electrical
signals through plants by assuming that the electrical resistance across plant tissue remains constant. In
this study we test a new approach that measures the electrical resistance across different parts of a
plant as a means of continually measuring the integrity of the plant transport system through tissues
and whole plants.
We used this method to monitor the impact of mild to lethal drought on plant physiology. In the well-
watered plants we observed a decrease in electrical resistance of the leaf blade, petiole and stem
through the morning the onset of which was well-correlated with light, likely as a result of increasing
tissue sugar concentrations as a consequence of photosynthesis and water loss. The electrical resistance
of the tissues of mildly droughted plants increased through the day, potentially as a result of hydraulic
failure. Previously, we identified chlorophyll fluorescence and membrane leakage as physiological
changes that co-occur in Brassica rapa when tissues and whole plants die from drought. Our
measurements of electrical resistance agreed well with the timing of these physiological changes,
suggesting that measuring plant electrical resistance may be an accurate method to measure when
tissues and whole plants are dead from drought. Because we used open source, inexpensive data
acquisition tools and the method is non-destructive and continuous it could be easily scaled up to
continually screen thousands of genotypes in high-throughput plant phenotyping studies.
212 Using Phenotypic Databases and Biotechnology to Unleash the Drug Discovery Potential of
Ethnobotany
John de la Parra, Northeastern University
A large database, covering thousands of years of human-plant interactions and containing culturally and
geographically diverse ethnobotanical information on the selection and use of medicinal plants for
neglected protozoan diseases has been created and mined for provable drug discovery leads. Candidate
species were prioritized and subsequent phytochemical characterization and bioassays were completed.
This study’s large-scale accounting for indigenous plant knowledge indicates that medicinal plants are
regularly chosen because of specific environmental conditions. Additionally, plants subjected to attack
by insect herbivores are often observed to have higher levels of more diverse specialized metabolites,
many with unique bioactivities. By querying the database, an initial test candidate species was chosen.
Pentalinon luteum, a traditional Seminole medicinal plant chosen with careful attention to
environmental conditions, regularly undergoes heavy seasonal herbivory by the larvae of the Oleander
moth, Syntomeida epilais. These larval predators consume and store unique cardiac glycosides that are
produced by their plant host. While these compounds are highly toxic to their predators, as the
ethnobotanical database in this study shows, plants with high cardiac glycoside content have had a long
history of medicinal value to humans. In addition, a very closely related traditional medicinal species
(Pentalinon andreuxii) mined from the database, was shown to be effective against the neglected
protozoan disease Leishmaniasis. To control for this variability of indigenous selection and herbivory-
induced stress responses, in vitro tissue cultures have been created to investigate the effects of the
addition of stress-signaling hormones as correlated to indigenous plant selection. By scanning a large
dataset of available ethnobotanical information, a robust study of the relatively unexplored aspects of
how humans choose medicinal plants is presented.
213 Planteye F500: Combine 3D and Multispectral Information in One Sensor
Gregoire Hummel, Phenospex
PlantEye is a high-resolution 3D laser scanner that computes a robust and validated set of morphological
plant parameters fully automatically. A core feature of PlantEye is that it can be operated in full sunlight
without any restrictions - crucial for plant phenotyping under field conditions or if you follow a “sensor-
to-plant-concept”. Phenospex has now developed a new sensor generation, that combines the actual
features of PlantEye on the fly with a 7-channel multispectral camera in the range between 400 –
900nm. This unique hardware-based sensor fusion concept allows us to deliver spectral information for
each data point of the plant in X, Y, Z-direction and we compute parameters like NDVI, color index and
many other vegetation indices. This new sensor generation opens a wide range of new possibilities in
plant phenotyping and increases it´s efficiency.
214 Measuring Plant Canopy Traits from UAV Imagery
Ethan Stewart, Cornell University
Nicholas Kaczmar, Cornell University; James Clohessy, Cornell University; Michael Gore, Cornell
University
The advent of affordable unmanned aerial vehicle (UAV) technology and recent relaxing of regulation make UAVs attractive tools for gathering plant data in the field. Common measures such as plant height are important for research and breeding but are time and labor intensive to collect. Vegetation indices such as the normalized difference vegetation index (NDVI) are typically gathered by plane or satellite platforms and do not offer the resolution needed to asses individual trial plots. UAVs offer the potential to gather large field data sets at high spatial and temporal resolution. Integration of GPS allows data to be accurately geolocated. UAVs fitted with RGB and multispectral cameras were flown at low elevation over corn fields consisting of small research plots. Images were stitched together and geo-referenced using ground control points of known positions to give geo-referenced orthomosaic images of the field. Plot boundaries were established from the orthoimages. 3D point clouds were produced from the RGB stiched images and used to calculate mean plant height for each plot. Plot level NDVI was calculated from the multispectral images.
Gathering field phenotypic data at high spatial and temporal resolution is beneficial to both plant research and breeding. Higher resolution datasets will allow the underlying genetic components to be investigated at finer detail than is currently feasible. The expansion of the platform to include measurements of other agronomically important traits is currently ongoing.
215 A high-throughput Plant Phenotyping System
Madan Bhattacharyya, Iowa State University
Dipak Sahoo, Iowa State University; Chimney Hegde, Iowa State University
High-throughput objective phenotyping with the aid of digital photography facilitates identification of
natural allelic variants as well as quantitative phenotyping of mutants for functional analyses of plant
genes. Our long-term goal is to identify novel genetic mechanisms for developing climate resilient crop
plants to secure food in the 21st century. The objective of this study is to develop a system for
phenotyping over 1,000 Arabidopsis ecotypes that have been sequenced. To accomplish this objective,
Arabidopsis Percival growth chambers (AR22LC9) were equipped with digital cameras (CropScore,
Tübingen, Germany) that are connected to a central server through Ethernets. We have developed and
validated a method to batch process images of hundreds of Arabidopsis seedlings to numerical data in a
spreadsheet-ready CSV file. We have developed a method for sowing Arabidopsis. We have phenotyped
157 Arabidopsis ecotypes for their responses to heat. By conducting GWAS of the phenotypic data, we
have identified nine genes. Among these genes, one encodes a putative RING/U-box-type E3 ubiquitin-
protein ligase, presumably involved in specific degradation of certain proteins through 26S proteasome
pathway under heat stress. This system will also be applicable in phenotyping seedlings of other crop
species including wheat, rice, maize, and alfalfa.
Environmental Stress Response in Plants
300 Tuber Size Distribution on One Stem of a Potato Plant (Morphological Approach to Study the
Ability of Individual Potato Tuber to Compete for Photosynthates)
Khaoula Mokrani, Higher Institut of Ariculture Chott Mariam
Wassim Saadaoui, Higher Agronomic Institute of Chott Mariam; Neji Tarchoun, Higher Agronomic
Institute of Chott Mariam
The main objective of this experiment was to study variation in tubers size on one potato stem and to
compare capacities of daughter tubers to import photoassimilates. The experiment was laid done in
factorial desing (3 x 3) during the main season of 2015 in Tunisia. Morphological parameters (Length of
stolons, median diameter of stolons, caliber of daughter tubers) were measured in eleven potato
cultivars (Solanum tuberosum L.), considering the five first stolons for each stem of each plant.
Parameters was measured four times during the development cycle of potato plant (49, 57, 71 and 82
day after plantation) and showed a high level of variation. Results showed Paramera and Dounia had the
highest length of stolons with 6.39 cm and 6.05 cm respectively. For caliber of daughter tubers, Evora
produced the biggest tubers with 31.78 mm, followed by Liseta (29.63 mm). The effect of measurment
dates on various characters, was very pronouced, specially for caliber of daughter tubers. High caliber of
daughter tubers was regestred during the fourth date of measure, with 37.66 mm, followed by the third
date (30.75 mm), then the second date (22.35 mm) and finally, the first date with value of 15.01 mm per
tuber. Results showed also that position of stolon on stem has a huge impact on caliber of daughter
tubers, first stolon produced the important tuber size with 30.09 mm, followed by the second stolon
(28.04 mm), then the third (24.24 mm), then the fourth (20.94 mm), while small tubers were produced
by the fifth stolon (15.49 mm).
301 Analysis of Water Limitation Effects on the Phenome and Ionome of Arabidopsis at the Plant
Imaging Consortium
Argelia Lorence, Arkansas State University
Lucia Acosta, Arkansas State University; Suxing Liu, Arkansas State University; Erin Langley, Arkansas
State University; Zachary Campbell, Arkansas State University; Norma Castro-Guerrero, University of
Missouri; David Mendoza-Cozatl, University of Missouri
Food security is currently one of the major challenges that we are facing as a species. Understanding
plant responses and adaptations to limited water availability is key to maintain or improve crop yield,
and this is even more critical considering the different projections of climate change. This is an NSF-
funded collaborative effort between the Mendoza and Lorence groups as part of the activities of the
Plant Imaging Consortium (http://plantimaging.cast.uark.edu/). In this work, we combined two high-
throughput -omic platforms (phenomics and ionomics) to begin dissecting time-dependent effects of
water limitation in Arabidopsis leaves and, ultimately, seed yield. As proof of concept, we acquired high-
resolution images with visible, fluorescence and near-infrared cameras and used commercial and open
source algorithms to extract the information contained in those images. At a defined point, samples
were also taken for elemental profiling. Our results show that growth, biomass and photosynthetic
efficiency were most affected under severe water limitation regimes, and these differences were
exacerbated at later developmental stages. The elemental composition and seed yield, however,
changed across the different water regimes tested and these changes included under- and over-
accumulation of elements compared to well-watered plants. Our results demonstrate that this
combination of phenotyping techniques can be successfully used to identify specific bottlenecks during
plant development that could compromise biomass, yield and the nutritional quality of plants.
302 Molecular Characterization of Ca2+/cation Antiporters (CaCA) Revealed Their Role in Stress
Response in Bread Wheat (Triticum aestivum L.)
Santosh Upadhyay, Panjab University
Mehak Taneja, Panjab University
Modulation of Ca2+ ion concentration is an important event during various stress responses. The
Ca2+/cation antiporters (CaCA) play the vital function in Ca2+ ion homeostasis, however, their
characterization is not performed in bread wheat (Triticum aestivum), which is an important crop plant.
Here, we identified thirty-four TaCaCA superfamily proteins in T. aestivum genome from four different
families (CAX, CCX, NCL, and MHX) based on their structural homology with known proteins from model
plants. Each A, B, and D-subgenome of allohexaploid T. aestivum genome contributed in the total
composition of TaCaCA superfamily in the form of homeologous genes, which consisted of comparable
gene and protein structure. About ten transmembranes and two Na_Ca_ex domains and α-repeat
regions were detected in the majority of TaCaCA proteins except for TaNCL, which consisted of single
α-repeat. The majority of genes showed differential expression during tissue developmental
stages, which indicated their role in development. Further, the modulated expression of a few genes
especially from CCX and NCL family during the fungal infestation, and heat, drought and salt stress
suggested their role during the stress response. The present study enlightened various characteristic
features of CaCA superfamily proteins in T. aestivum, however, the function of the individual gene needs
to be investigated in future studies.
303 Heterologous Expression of Serine hydroxymethyltransferase-3 from Rice Confer Tolerance to
Salinity Stress in Arabidopsis
Vandna Rai, National Research Centre on Plant Biotechnology
Pragya Mishra, National Research Centre on Plant Biotechnology; Ajay Jain, National Research Centre on
Plant Biotechnology; Teruhiro Takabe, Meijo University; Yoshito Tanaka, Meijo University; Nisha Singh,
National Research Centre on Plant Biotechnology; Neha Jain, National Research Centre on Plant
Biotechnology; Nagendra K Singh, National Research Centre on Plant Biotechnology
Among abiotic stresses, salinity stress adversely affects growth and development of rice. Contrasting
salt-tolerant and salt-sensitive rice varieties provided repository of genes that may be amenable for
manipulation through biotechnological interventions. In our earlier studies, omic (trasncriptomics and
proteomics) approach was employed in identifying salinity stress-induced serine hydroxyl
methyltransferase-3 (OsSHMT3). Overexpression of OsSHMT3 in E. coli conferred tolerance to
differentsalinity regime (0 mM to 700 mM NaCl). Augmented levels of osmolytes (glycine betane and
choline), and several amino acids (glycine and serine) in transformed E. coli could have possibly
contributed towards the observed phenotype. Transgenic Arabidopsis overexpressing OsSHMT3 (OEs)
also exhibited distinct tolerance towards salinity stress, and many of the vegetative and reproductive
morphological traits were comparable with the wild-type. A comprehensive omic approach (ionomic,
transcriptomic, metabolomic, and proteomic profiling) was employed for identifying the molecular
mechanism underlying OsSHMT3-mediated tolerance to salinity stress in Arabidopsis. The study
provided empirical evidence towards the potential role of SHMT3 in alleviating salinity stress in a
heterologous system. Efforts are now underway to explore the feasibility of alleviating salinity stress in
salt-sensitive rice varieties by overexpressing this gene.
304 Metabolome Analysis of Some Effective PGPR Strains Isolated from Rhizosphere of Arnebia
benthamii in North Western Himalaya
Javid Parray, University of Kashmir
Azra Kamii, University of Kashmir; Sumira Jan, University of Kashmir; Nowsheen Shameem, University of
Kashmir; Suhaib Bandh, University of Kashmir; Bashir Lone, University of Kashmir
Arnebia benthamii L. (Boraginaceae) commonly called as ‘Kahzaban’, critically endangered perennial
medicinal herb growing in the sub alpine and alpine zones of North West Himalaya at an altitude of 3500
–4000 m with limited density. Current study evaluates the diversity of culturable root-associated
bacteria from this plant which resulted in a total of 80 morphologically different isolates. The
composition of nutritive growth medium had a significant effect on the diversity of the isolated bacterial
isolates. The isolates were characterized for various metabolic, plant growth promoting (PGP) and other
useful activities, based on which they were clustered into groups. Majority of the isolates belonged to
Bacillus, Pseudomonas and ptoteabacteria and were the most dominant species of the total isolates.
Some isolates exhibited very high antioxidant activity and were able to protect the DNA damage as well
and few were able to inhibit the pathogenic fungal strains significantly.
305 Signal Transduction in Plant Stress – Perception and Response to Pathogens and Environmental
Stress at Molecular and System Levels
Sorina Popescu, Mississippi State University
Gizem Dimlioglu, Mississippi State University; Norbert Bokros, Mississippi State University; Priyardashini
Bhorali, Mississippi State University
Understanding response and adaptation to stress is a fundamental problem in plant biology. Our
research focuses on the molecular processes triggered by pathogens or environmental stress factors,
and how they shape the phenotypical response of the whole organism. In particular, we aim to
understand the organization of signal processing pathways and of the principles that govern their
activation and function in experimental models and crops. The prospect of generating synthetic
signaling pathways for broad stress tolerance in plants makes our work intriguing and exciting. Two
current projects in our lab exemplify my lab’s most recent work. One study is centered on the
calmodulin-binding protein INTEGRIN-LINKED KINASE1 (ILK1). ILK1 coordinates signaling during the
plant’s basic immune response and ionic stress triggered by high salt or osmolite concentration. We
demonstrate the importance of ILK1 at the confluence of biotic and abiotic stress response and highlight
the importance of the nutritional status of the plant and ion-mediated processes at the plasma
membrane. In a second study, an in vivo plant-pathogen interactomics screen identified tomato kinases
with critical functions in cellular responses associated with the immune response to bacterial pathogens.
In planta functional assays and bioinformatics analyses revealed general rules for the organization and
regulation of signaling networks in immunity and cell death, alongside strategies by which pathogen
virulence factors may impact the immune network.
306 Integrated ‘OMICs’ Approach to Tag Blister Blight Leaf Disease Resistance in Tea (Camellia
sinensis L.)
Kooragodage Mewan, Tea Research Institute of Sri Lanka
KH Tissa Karunarathne, University of Colombo; Jagath Weerasena, University of Colombo, Cumaratunga
Munidasa Mawatha, Colombo 03; E Nishantha Edirisinghe, Tea Research Institute of Sri Lanka; Pradeepa
Liyanage, Tea Research Institute of Sri Lanka; ISB Abeysinghe, Tea Research Institute of Sri Lanka
With the global concern on ‘high quality, clean and safe food’, development of resistant crop varieties to
biotic/ abiotic stresses is pivotal where accurate selection plays a fundamental role thus becoming
marker-assisted-selection as an indispensable tool in enhancing efficiency of conventional breeding
programs, especially for perennial crops like tea (Camellia sinensis L.). For tea, as the economically most
important plantation crop in Sri Lanka, genetic improvement for Blister Blight Leaf Disease (BB)
resistance, which causes 20-30% annual crop loss, is one of the foremost objectives in tea breeding
where breeders have to evaluate the ‘trait’ by laborious, environment dependent and time consuming
multi-location field assessments. To overcome above constraints, studies were extended to identify a
potential marker to BB resistance by integrating ‘Genomic’, ‘Metabolomic’ and ‘Phenomic’ tools.
307 RERJ1 – A Wound Responsive JA Dependent Early Inducible bHLH Transcription Factor – Is Involved
in the Rice JA-signaling System Together with OsMYC2 and OsJAZ
Ioana Valea, The University of Tokyo
Koji Miyamoto, Teikyo University; Okada Kazunori, The University of Tokyo
Pathogen and herbivore-induced damage in plants is the main cause for yearly yield losses and
subsequent economic ones. (Oerke, 2006) The jasmonic acid (JA) pathway is one of the significant
pathways shaping the fitness straining plant defense, as well as the balance to be established to growth
and of course plant growth itself among other influences on plant life stages. (Staswick et al., 1992)
RERJ1 (rice early responsive to jasmonates 1) has been previously shown to be a basic helix-loop-helix
(bHLH) transcription factor involved in the early JA signaling in rice. In response to wounding it reaches
its peak within 30 minutes after treatment. (Kiribuchi et al., 2004)
The main focus of our study is to characterize the function of RERJ1 within the JA signaling system in rice
with regard to OsMYC2, acting as an activator, and OsJAZ, the main known inhibitor group (Uji et al.,
2016). Our approach is on transcriptional level, tracking the transcriptional activity of RERJ1 in mutants
such as rerj1-TOS17 knock-down and OsMYC2 RNAi, and on protein level identifying direct interactions
between RERJ1 with OsMYC2 and OsJAZ.
A differential temporal pattern of expression until 4h after JA treatment and wounding in the rerj1-
TOS17 mutant shows that RERJ1 influences a wide range of the 15 known OsJAZ, several of which are
known to be involved in defense mechanisms. OsMYC2 transcript as well seems to be slightly affected
by RERJ1. In support the protein-protein interaction within a yeast two-hybrid assay shows RERJ1 being
able to physically interact with a wide range of OsJAZ, leading to the hypothesis that RERJ1 is directly
involved and shapes the known feedback loop between OsMYC2 and OsJAZ (Kazan and Manners, 2013).
Further transcriptional analysis in RERJ1/OsMYC2 overexpressing rice plants is currently on going.
308 Linking Duplication of the CALCINEURIN B-LIKE10 (CBL10) Calcium Sensor to Plant Salt Tolerance
Shea Monihan, University of Arizona
Choong-Hwan Ryu, University of Arizona; Courtney Magness, University of Arizona; Karen Schumaker,
University of Arizona
The Salt-Overly-Sensitive (SOS) pathway in Arabidopsis thaliana (Arabidopsis) functions to prevent the
toxic accumulation of sodium in the cytosol when plants are growing in saline conditions. In this
pathway, the SOS3 (roots) and CALCINEURIN B LIKE10 (CBL10, leaves) calcium sensors interact with the
SOS2 protein kinase to activate the SOS1 sodium/proton exchanger in the plasma membrane. CBL10
has been duplicated in Eutrema salsugineum (Eutrema), a salt-tolerant relative of Arabidopsis. Because
Eutrema maintains growth in salt-affected soils that kill most crop plants, the duplication of CBL10 in
Eutrema provides a unique opportunity to directly test the outcome of gene duplication and to link this
duplication to plant salt tolerance.
In Eutrema, down-regulation of the duplicated CBL10 genes (EsCBL10a and EsCBL10b) singly and in
combination reduces growth in saline conditions suggesting that both genes function in plant responses
to salt. Expression analyses and cross-species complementation assays in Arabidopsis and
Saccharomyces cerevisiae demonstrate that EsCBL10b is similar to AtCBL10 but with an enhanced
function; EsCBL10b is expressed in leaves, complements the Atcbl10 mutant, and has an enhanced
ability to activate the SOS pathway. EsCBL10a has a unique role in response to salt; it is expressed in
both leaves and roots, complements the Atcbl10 mutant, but only weakly activates the SOS pathway.
Based on its expanded expression into roots, EsCBL10a was expressed in the Atsos3 mutant in which
root growth is significantly reduced in saline conditions. In contrast to AtCBL10 or EsCBL10b, EsCBL10a
complemented the Atsos3 salt-sensitive phenotype, suggesting that EsCBL10a has a unique function.
These results suggest that both EsCBL10a and EsCBL10b function in plant responses to salt, but their
functions have diverged. EsCBL10a has a unique role in salt tolerance compared to AtCBL10 and
EsCBL10b, while EsCBL10b has an enhanced function in the SOS pathway relative to AtCBL10 and
EsCBL10a.
309 Ectopic Expression of psNTP9, An Apyrase Gene from Pisum Sativum, Confers Drought Tolorance in
Transgenic Soybean
Roopadarshini Veerappa, University of Texas at Austin
Soybean (Glycine max (L.) Merr.) is one of the major resources of protein for human and animal
nutrition as well as a key source of vegetable oil. It is also considered a potential crop for the production
of biodiesel. Drought is a critical environmental factor that imposes water stress on crops, which is a
major constraint on plant growth and productivity, contributing to yield losses in crops, including
soybeans. Apyrases (Nucleoside Triphosphate-Diphosphohydrolases) are calcium- and magnesium-
activated enzymes that remove the terminal phosphate from nucleoside triphosphates (NTPs) and
nucleoside diphosphates (NDPs). By controlling the [NTP] and/or [NDP] in the Golgi, the ECM or other
subcellular sites, apyrases are known to play an essential role in plant growth, development and stress
responses. In the present study the pea apyrase gene (psNTP9), driven by CaMV 35S promoter, was
transferred into the soybean genome by Agrobacterium-mediated plant transformation. Two
independently transformed soybean lines expressing psNTP9 were produced. Segregation analyses
indicated single-locus insertion for both lines. Preliminary tests performed in the greenhouse evaluated
the phenotype of transgenic soybean plants under drought conditions. Transgenic plants outperformed
the wild-type plants when subjected to water deficit conditions, exhibiting notable drought tolerance
with a higher survival rate compared to the wild type. They also had a greater shoot and root biomass
and showed significantly higher seed production compared to non-transgenic plants under water deficit.
This is the first report showing that the ectopic expression of psNTP9 in transgenic soybeans confers
drought tolerance, highlighting the potential of this gene for molecular breeding.
Funded by grants from the National Science Foundation (IOS-1027514) and a private company to Drs.
Greg Clark and Stanley J. Roux at the University of Texas at Austin.
310 Overlap Between the Ad/Abaxial Developmental and ABA Signaling Networks
Tie Liu, Stanford University
Adam Longhurst, Carnegie Institution for Science; Franklin Talavera-Rauh, Carnegie Institution for
Science; Samuel Hokin, Carnegie Institution for Science; Kathryn Barton, Carnegie Institution for Science
The actions of the plant hormone abscisic acid (ABA) in the repression of germination and the inhibition
of water loss through stomata are well studied. Less well understood are the functions of ABA on
vegetative development including the mechanisms through which it influences growth of roots and
shoots. In our study of gene networks that control development of the adaxial (towards meristem/inner)
and abaxial (away from meristem/outer) domains of the leaf, we identified a set of genes oppositely
regulated by REVOLUTA and KANADI. Observing that this set of genes is enriched for genes encoding
ABA signaling components, we reasoned that genes of unknown function within this set would also be
involved in ABA signaling. Indeed, we find that a mutation in one of these, ABA INSENSITIVE GROWTH 1
(ABIG1), results in ABA resistant vegetative growth of shoot. Conversely, conditional overexpression of
ABIG1 mimics the application of ABA to wildtype plants: it causes leaf yellowing, reduces the rate of leaf
production and causes reduced root growth. Surprisingly, the ABA resistant abig1 mutants are also
drought resistant. We hypothesize that this reflects the ablation of the branch of the pathway for ABA
inhibition of growth and promotion of senescence while leaving intact the pathway for stomate closure.
To characterize ABIG1 induced senescence and meristem arrest, we carried out phenotyping and
quantifying analysis on the growth and development of meristem and young leaf primordia under at
various times post induction and drought treatments for both Arabidopsis and the orphaned crop
quinoa.
311 Role of Mineral Nanoparticles in Improving Attachment and Colonization by Plant Growth
Promoting Rhizobacteria
Salme Timmusk, Uppsala BioCenter
Gulaim Seisenbaeva, Uppsala BioCenter
The soil surrounding plant roots is one of the main sources of bacteria expressing plant-beneficial
activities (PGPR). Recently we have shown that the PGPR have great potential in protecting plants
against abiotic and biotic stress situations and restoring marginal lands (1-3). Multiple factors influence
isolate stability. The effectiveness of the inoculum is related to the formulation technology protecting
the cells from the surrounding environmental conditions and elimination of secondary effects. Titania
nanoparticle (TN) formulation form stable, large and thick bacterial clumps as a biofilm which influence
plant growth via facilitating root hair length and density and improving mulch biofilm formation (1-3).
This effect is especially pronounced in complex environments there several PGPR strains are used in
combination with TNs (1). The biofilm substantially improves root soil contact and enhances significantly
plant nutrient or biologically active compound acquisition from soil or bacterial origin. In addition,
improved organic matter content and porosity restores marginal lands (1)
The key to further advancing our treatment with microbial products is (i) surface grafting of sands and
intercalation of clays and employing hybrid biopolymer-mineral nanocomposites (ii) distinguishing
between genes and pathways that drive passenger events both in PGPR and crop plant.
1) S. Timmusk, G. A. Seisenbaeva, A. Muraya, J. Muthony, L. Behers, Nano titania aid multiple plant
growth promoting bacterial interactions under complex environments. Frontiers in Plant
Sci, (2016).
2) S. Timmusk, et al, Perspectives and challenges for microbial application for crop
improvement. Frontiers in Plant Sci, (2016).
3) S. Timmusk et al., PloS ONE DOI 10.1371/journal.pone.0096086, (2014).
312 Novel Plastid Behavior Associated with the MSH1 Effect
Jesus Beltran, University of Nebraska-Lincoln
Evan LaBrant, University of Nebraska-Lincoln; Mon-Ray Shao, University of Nebraska-Lincoln; Sally
Mackenzie, University of Nebraska-Lincoln
In plants, plastids serve as sensors of developmental and environmental stimuli, allowing cells to
modulate gene expression and likely to reprogram the epigenome in a heritable fashion. MSH1 is an
organellar DNA binding and thylakoid protein with profound roles in development, stress responses and
redox balance. Current evidence supports MSH1 as important for influencing transgenerational
epigenetic inheritance. MSH1, together with its interacting partner PPD3, localize to small, distinctive
plastids that display low autofluorescence and reside within the epidermis and vascular parenchyma.
We have termed these organelles ‘sensory plastids’. Our observations imply proteome heterogeneity in
different plastid types across discrete cell populations. To gain insight into the role of sensory plastids,
we combined Fluorescence-Activated Cell Sorting (FACS) and proteomics to characterize their
proteomes in Arabidopsis floral stems. We used Translating Ribosome Affinity Purification (TRAP)
analysis followed by RNA sequencing technology (TRAP-SEQ) to assess both MSH1-containing and
MSH1-depleted, cell-specific translating RNAs. Early results suggest that protein counts for responders
to abiotic stimulus are more prominent in rank in sensory plastids than in chloroplasts, which suggests
plastid specialized function. TRAP-SEQ experiments revealed that cells depleted of MSH1 actively
translate proteins for oxidation-reduction processes, cell redox homeostasis and auxin response, which
suggest a role of MSH1 in triggering retrograde signaling in specific cell types. Finding correlations
between cell-specific translatomes and the plastid proteome profiles within these cells would help to
understand the underpinnings of organellar functional specialization.
313 Efficacies of Seventeen Organically made Northern Fertilisers on Sustainable Crops Production in
Acidic Soil of Food Security Under Climate Change Bangladesh Context
Durlave Roy, Northern Agro Services Ltd
Different field study was conducted to evaluate the efficacies of Northern organically made fertilizers on
yield and quality of crops in a study. BF (Rice) M-54, Yield increased by 17-20%, Urea saved by 50%, BF
(Wheat) M-51, Yield increased by 17-21%, BF (Potato) M-48, Yield increased by 25%, BF (Mustard) M-50,
Yield increased by 28.5 %, BF (Maize) M-53,Yield increased by 28%, BF (Onion) M-47, Yield increased by
5kg per decimal average, BF (Banana) M-49, Saving in Urea fertiliser per plant 50-100g, BF (Cauliflower)
M-66, Chemical fertiliser saves by 50%, BF (Sugarcane) M-63,Yield increased by 36%, BF (Garlic) M-201,
Yield increased by 5kg per decimal average, BF (Tomato) M-199, Chemical fertiliser saves by 50%, BF
(Cabbage) M-198, Chemical fertilisers saves by 50%,BF (Chili) M-200, Chemical fertiliser saves by 50%, BF
(Brinjal) M-403, Chemical fertiliser saves by 50%, BF (Watermelon) M-402, Saving in Urea per decimal by
15-200g, Shakti Fertiliser (Mango, Tea, Vegetables, Cotton), M-52, Yield increased by 45.8% of Mango,
15% of Tea, Chemical fertiliser saves by 50% of Rice and Vegetables, Higher yield growth of 2550kg per
hectare of Cotton, Organic Fertilizer (Potato, Sugarcane, Betel Leaf, Mango, Tea, Vegetables) M-65,Yield
increased by 20% of Mango, 9% of Tea, chemical fertilizer saves by 50% of Rice and Vegetables, Higher
yield growth of 68.41 tons per hectare of Sugarcane, 8850 betel Leafs per decimal, 23.68 tons per
hectare of Potato.
314 Heat and Drought Stress Effects on Seed Germination and Seedling Development of Sorghum
(Sorghum bicolor, L. Moench) in the Semi-Arid Sahel of Mali
Siaka Dembele, Institut d’Economie Rurale (IER)
Abiotic stresses such as extremes temperatures and drought, have detrimental effects on germination,
growth and yield of sorghum (Sorghum bicolor L. Moench) The objective of this pot trial was to evaluate
nine sorghum varieties on their potential of germination and seedling development when subjected to
different levels of early season drought and heat stresses. Treatments included 5 levels of heat stress
imposed by incorporating different layers (0, 0.5, 1, 2 and 3 cm) of charcoal powder on top of pots and
different levels of water stress (No water stress, one, two, three and four week water stress), or
combinations thereof. The experiment was set up as a complete randomized design with 3 replications.
Increasing drought and heat stress, in general, seed germination and seedling development decreased
as evidenced by various germination traits such as germination percentage at 8 days after sowing (DAS),
mean germination time (MGT), germination rate index (GRI), seedling vigour index (SVI), root vigour
index (RVI), and seedling dry weight (SDW). But despite of these abiotic stresses imposed. Nevertheless,
irrespective of the heat levels imposed, some varieties such as Banidoka, CSM63E, and Saba-tienda
showed good germination and seedling development as evidenced by a high GRI and SVI as compare to
the control. Differential varieties responses to heat and drought stress can prepare a suitable metabolic
reaction in seeds and can improve seed germination performance and seedling under drought and heat
prone area.
315 Reduced Germination of the siz1 Mutants is Caused by Low SLY1 Activity and Hyperdormancy
Hak Soo Seo, Seoul National University
Woo Seob Kim, Seoul National University; Beom Seok Choi, Seoul National University
Germination is a critical step for vegetative growth and seed production. Here, we examined the effect
of the E3 SUMO ligase AtSIZ1 on seed dormancy and germination. Result showed that the germination
rates of the siz1 mutants were less than 50%. But, their germination rates increased to wild-type levels
after cold stratification or long periods of ripening, and also exogenous gibberellin (GA) application. In
addition, suppression of AtSIZ1 caused rapid post-translational degradation of SLEEPY1 (SLY1) during
germination, and inducible AtSIZ1 overexpression led to increased SLY1 levels. The germination ratios
of siz1 mutant seeds in immature developing siliques were much lower than those of the wild type.
Moreover, SLY1 and DELAY OF GERMINATION 1 (DOG1) transcript levels were reduced in
the siz1 mutants, while the transcript levels of DELLA and ABSCISIC ACID INSENSITIVE 3 (ABI3) were
higher than those of the wild type. Taken together, these results indicate that the reduced germination
of the siz1 mutants results from impaired GA signaling due to low SLY1 levels and activity, as well as
hyperdormancy due to high levels of expression of dormancy-related genes including DOG1.
This work was supported by a grant from the Next-Generation BioGreen 21 Program (Plant Molecular
Breeding Center no. PJ01108701), Rural Development Administration, Republic of Korea.
316 Flowering Repression Activity of FLOWERING LOCUS C is Negatively Regulated by HIGH PLOIDY2
Hak Soo Seo, Seoul National University
Woosub Kim – Seoul National University
Sumoylation is an important post-translational modification for growth and development I plant and vertibrate. We thus examined the effect of sumoylation on Flowering Locus C (FLC)-mediated flowering. Here, we identified Arabidopsis HIGH PLOIDY2 (HPY2) as an E3 SUMO ligase for FLC. In vitro and vivo pull-down assays showed that FLC physically interacts with HPY2. In vitro assays showed that the stimulation of FLC sumoylation by HPY2 was dependent on SUMO-activating enzyme E1 and -conjugating enzyme E2, indicating that HPY2 was an E3 SUMO ligase for FLC. In transgenic plants, inducible HPY2 overexpression increased the concentration of FLC, indicating that HPY2 stabilized FLC through direct sumoylation. Flowering time in hpy2-2 mutants was shorter than in wild-type plants under long- and short-day conditions, with a greater effect under short-day conditions, and FLC was downregulated in hpy2-2 mutants. Taken together, these data indicate that HPY2 regulates FLC function and stability at both the transcriptional and post-translational levels through its E3 SUMO ligase activity. This work was supported by a grant from the Next-Generation BioGreen 21 Program (Plant Molecular Breeding Center no. PJ01108701), Rural Development Administration, Republic of Korea.
317 Phenotyping of Plants for Drought and Salt Tolerance using Infra-red Thermography
Taek-ryoun Kwon, The National Institute of Agricultural Sciences
Kyung-hwan Kim, The National Institute of Agricultural Sciences; Zamin Shaheed Siddiqui, Karachi
University
Drought and salinity are the major environmental constrains in global agricultural production. Plant
breeding for the drought and salt tolerance needs a proper assessment procedure to overcome stress
constrain. Fundamental understanding on the physiological nature of the plant tolerance provides
valuable information for the genetically modified crop’s development. Drought or salt stress induces
several common physiological responses in plants such as water relation and photosynthetic capacity. It
is because both stresses lead cellular dehydration in the plants, particularly, during the early phase of
stress imposition. Drought and salinity decrease CO2 availability for photosynthesis via stomatal
limitation as well as elevate leaf temperature due to partially closed stomata. In this scenario, stomatal
regulation and plant water status are important aspects in abiotic stress environment. These
physiological responses have a function to stabilize the temperature inside plant/leaf. Therefore
phenotyping through an infra-red thermography (heat sensitive sensor), could be a useful tool in the
selection of a tolerant genotypes. Infra-red thermography is a part of the electromagnetic spectrum
which emits a certain amount of radiation as a function of their temperatures. In general, the plants
which have less water, would have higher temperature and display more infra-red radiations. In abiotic
stresses such as drought and salinity, plant water status is affected and varied from the sensitive to
tolerant level. Infra-red images of plants are often linked with some of the physiological attributes to the
tolerance. This review covers the limits, advantages, linkages, comparison and other prospective of
using thermal images in modern phenotyping techniques.
318 Rice Bacterial Leaf Blight is Controlled by Rapid Systemic Movement of Niclosamide and its Effect
is Maintained for a Long Time
Beom Seok Choi, Seoul National University
Hak Soo Seo, Seoul National University
Bacterial leaf blight is one of the major diseases in rice and affects yields. Thus, various methods have been applied to protect rice from this disease. Here, we show systemic translocation of the human drug niclosamide (5-chloro-N-(2-chloro-4-nitrophenyl)-2-hydroxybenzamide) in rice and its long-term effect on prevention of rice leaf blight. The development of Xanthomonas oryzae pv. oryzae-induced rice leaf blight was effectively inhibited in untreated systemic leaves as in niclosamide-treated leaves, although its effect gradually decreased in a time-dependent manner. Time-course examination after niclosamide treatment showed that the niclosamide level was highest after 3 h in non-treated distal leaves, suggesting fast systemic movement of niclosamide from the treated local site to untreated distal regions. Our data indicate that niclosamide controls rice leaf blight by its rapid systemic movement and that its effect is maintained for a long time. This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (Grant Number: 2015057573).
319 Using RNAseq to Identify Key Player Genes Driving Soybean Drought Response in High and Low
Nitrogen Fixing Genotypes
Marianne Emery, University of Missouri
Arun Dhanapal, University of Missouri; Andy King, University of Arkansas; Larry Purcell, University of
Arkansas; Jeffrey Ray, USDA-ARS; James Smith, USDA-ARS; Felix Fritschi, University of Missouri; Christine
Elsik, University of Missouri
Patterns of gene expression provide insight into plant responses to stress conditions such as drought. In
legumes, the process of nitrogen (N) fixation has been shown to be sensitive to water deficit. A better
understanding of the gene(s) that drive major pathway(s) involved in soybean (Glycine max) response to
drought, and their effect on N fixation, will assist in the development of soybean cultivars better able to
adapt in drought conditions. Here we report a comparison of two soybean cultivars, identified
previously as high and low N fixing, exposed to varying levels of water availability. RNA extracted from
leaf, nodule, and root tissues were sequenced using Illumina Hi-Seq (100 bp, single-end) and quantified
at the gene level across environments and tissues. Gene expression values were used to build a mutual
information network resulting in nine gene clusters. The genes contained in each cluster were
characterized, and gene hubs, here defined as key player genes, were identified. We believe these key
player genes in each modeled cluster may be main drivers in soybean drought response and could be
putative targets for improvement in future breeding efforts.
320 Utilizing Cysteine Protease Inhibitor Gene to Enhance Stress Resistance in Soybean
Sophie LaRochelle, SUNY Cobleskill
Clara Richardson, SUNY Cobleskill; Angela Zhu, SUNY Cobleskill
Programmed cell death is a mechanism within plant cells in which the cell undergoes a series of changes
and dies due to a variety of factors. These factors include the cell cytoplasm decreasing, the nuclear
material condensing and the membrane beginning to disintegrate. Proteases within the plant play an
important role in the regulation of this process of programmed cell death. Within plant systems,
proteases are regulated at a variety of different levels including; transcription, translation, post
translation and specific protease inhibitor proteins. Soybean cells have inactivated proteases present in
order to oxidize stress through an indirect mechanism. Cysteine protease inhibitors may function in
modifying programmed cell death that is activated when a plant is under environmental stress or
pathogen attack. The ability to engineer soybean with the Cysteine Protease Inhibitor (CPI) gene can
provide resistance against stresses such as chilling, salt, drought and pathogen attacks.
321 PLANT U-BOX PROTEIN10 Regulates MYC2 Stability in Arabidopsis
Choonkyun Jung, Seoul National University
Pingzhi Zhao, Rockefeller University; Jun Sung Seo, Rockefeller University; Nobutaka Mitsuda, National
Institute of Advanced Industrial Science and Technology; Shulin Deng, Rockefeller University; Nam-Hai
Chua, Rockefeller University
MYC2 is an important regulator for jasmonic acid (JA) signaling, but little is known about its
posttranslational regulation. Here, we show that the MYC2 C-terminal region interacted with the PLANT
U-BOX PROTEIN10 (PUB10) armadillo repeats in vitro. MYC2 was efficiently polyubiquitinated by PUB10
with UBC8 as an E2 enzyme and the conserved C249 in PUB10 was required for activity. The inactive
PUB10(C249A) mutant protein retained its ability to heterodimerize with PUB10, thus blocking PUB10 E3
activity as a dominant-negative mutant. Both MYC2 and PUB10 were nucleus localized and
coimmunoprecipitation experiments confirmed their interaction in vivo. Although unstable in the wild
type, MYC2 stability was enhanced in pub10, suggesting destabilization by PUB10. Moreover, MYC2 half-
life was shortened or prolonged by induced expression of PUB10 or the dominant-negative
PUB10(C249A) mutant, respectively. Root growth of pub10 seedlings phenocopied 35S:MYC2 seedlings
and was hypersensitive to methyl jasmonate, whereas 35S:PUB10 and jin1-9 (myc2) seedlings were
hyposensitive. In addition, the root phenotype conferred by MYC2 overexpression in double transgenic
plants was reversed or enhanced by induced expression of PUB10 or PUB10(C249A), respectively.
Similar results were obtained with three other JA-regulated genes, TAT, JR2, and PDF1.2. Collectively,
our results show that MYC2 is targeted by PUB10 for degradation during JA responses.
322 Improving the Viability of Sorghum for Bioenergy Production
Sarah Pfaff, Rutgers University
Michael Pashos, Washington University; Christie Peebles, Rice University; Courtney Jahn, University of
Wisconsin
Sorghum is a drought tolerant crop that can be used as a feedstock for bioenergy production. Here we
report a three pronged approach to improve the viability of sorghum for conversion to bioenergy
through: (1) increased biomass production; (2) improved drought tolerance; (3) development of a
variety that is resistant to herbicides. Specific outcomes of this work include primers designed and
verified for Targeted Induced Local Lesions in Genomes analysis (TILLING) of a non-GMO mutant
sorghum population to identify mutations that enhance desired traits, such as increased biomass or
drought tolerance. Secondly, a plasmid was designed and assembled for use in sorghum plastid
transform to confer herbicide resistance. Lastly, a Recombinant Inbred Line (RIL) population between a
drought tolerant and drought susceptible sorghum line was planted and evaluated in the field to
determine Quantitative Trait Loci (QTL) for above and below ground traits that confer drought
tolerance. Preliminary results show success towards each of the three project goals.
323 Chlorophyll Fluorescence and Leaf Gas Exchange based Phenotyping of Heat Stress Tolerance in
Tepary, Common Bean and Cowpea at the Seedling and Juvenile Stages
Isaac Osei-Bonsu, Michigan State University
Donghee Hoh, Michigan State University; Wayne Loescher, Michigan State University; Dan TerAvest,
Michigan State University; David Kramer, Michigan State University
The recently prevalent consequences of climate change and the predicted rise in temperature
necessitates development of heat stress tolerant crops to be able to feed the growing population.
Photosynthesis is one of the processes negatively affected by high temperature (HT), the threshold of
which varies with crop species. Existing and new tools and platforms based on Chlorophyll fluorescence
(CF) and leaf gas exchange (LGE) techniques enable phenotyping of photosynthesis signatures for crops
of interest. In the current study we aimed to determine genotypic differences in CF and LGE-based
photosynthetic parameters under HT stress and explore the possibility of using tools based on these
techniques for phenotyping HS tolerance in genotypes of three legumes, viz tepary, common bean and
cowpea with differing HS tolerance levels. Seedlings grown at 30/25°C day/night temperatures were
subjected to a 5°C rise in both day and night temperatures every two to five days up to a maximum
temperature of 45/40°C at different growth stages starting from 3 d after germination. The common
bean line, Zorro, consistently emerged as the most heat susceptible of the three species irrespective of
the growth stage (seedling or juvenile/vegetative stage), whereas tepary beans and cowpea alternated
between high and medium HS tolerance depending on the line, growth stage and leaf type (unifoliate vs
trifoliate). Susceptibility of Zorro became more apparent after 2 d at 45/40°C with drastic decline in
ɸII (quantum yield of PSII), net assimilation (A), stomatal conductance (gs) and chlorophyll content,
whereas non photochemical quenching (NPQt) and yield of NPQ (ɸNPQ) increased. Results from
chlorophyll fluorescence parameters assessed with the multispeQ, connected to the photosynQ
platform (www.photosynq.org) corroborated that of LGE parameters measured with the Li-Cor,
indicating the utility of these two devices as tools for phenotyping HS tolerance in legumes.
324 Phenotyping Biochemical Bypasses to Photorespiration to Discover Improvements in
Photosynthesis and Crop Production
Amanda Cavanagh, University of Illinois
Paul South, USDA-ARS Photosynthesis Research Unit; Donald Ort, USDA-ARS Photosynthesis Research
Unit
As climate changes globally and human population increases, traditional methods of crop improvement
have become less effective in adapting and improving agricultural production. Improving photosynthetic
efficiency has been a long standing goal toward increasing crop yield, but has played a minor role in crop
improvements to date. In C3 crops such as wheat and rice approximately 25% of the fixed carbon
dioxide is lost to photorespiration, the energy expensive metabolic pathway that recycles toxic
compounds produced by Rubisco oxygenation reactions. Reducing photorespiratory yield losses by 5%
(i.e., to 31% for soybean and 15% for wheat) would be worth ~$500 million annually in the United
States. Synthetic biology has provided new opportunities to altering photorespiratory metabolism to
improve photosynthetic efficiency, but the phenotyping bottleneck hinders the speed at which benefits
can be realized. Using the Golden Gate synthetic biology approach, we have assembled and
transformed a series of multigene constructs that contain metabolic pathways to bypass
photorespiration. We developed a fluorescence based screen to phenotype single construct
transformed lines for rescue of photorespiratory deficient plants in the model plants Arabidopsis and
Tobacco. Our results indicate that large multigene constructs containing a metabolic bypass to
photorespiration can rescue changes in fluorescence caused by low CO2 stress in deficient plants. We
confirmed these results on plant phenotype through gas-exchange measurements of photosynthesis
under both greenhouse and field grown conditions. Determining robust photorespiratory bypass
constructs provide insight into next generation crops and our use of standard parts and fluorescent
screening to complement traditional photosynthetic phenotyping methods provide a tool kit for plant
synthetic biology to engineer improvements in photosynthetic efficiency.
325 Root Cortical Senescence Influences Metabolic Costs and Radial Water and Nutrient Transport in
Barley
Hannah Schneider, Penn State University
Tobias Wojciechowski, Forschungszentrum Juelich; Johannes Postma, Forschungszentrum Juelich;
Dagmar van Dusschoten, Forschungszentrum Juelich; Jonathan Lynch, Penn State University
Root cortical senescence (RCS) is a type of programmed cell death in cortical cells of many
Poaceaespecies. The functional implications of RCS formation are poorly understood, but studies
suggest that RCS formation confers both benefits and costs. The objectives of this research were to test
the hypotheses that: (1) genetic variation exists in RCS; (2) RCS reduces the metabolic cost of root tissue;
(3) RCS decreases radial water and nutrient transport. Using a Pitman chamber, radial water and
nutrient transport were measured from excised roots of barley using isotopes. We observed large
genetic variation for RCS between landraces and modern germplasm. RCS decreases the metabolic cost
of root tissue. Nitrogen and phosphorus deficient conditions increased the rate of RCS development.
RCS reduced root nitrogen content by 66%, phosphorus content by 63%, and respiration by 87%
compared to root segments with no RCS. RCS decreases radial water and nutrient transport. Barley roots
with complete RCS had 90% less radial water, 92% less radial nitrate, and 84% less radial phosphorus
transport compared to root segments with no RCS. RCS increased aliphatic suberin content by 30% in
the endodermis. RCS may be a useful adaptation to drought by reducing the metabolic costs of soil
exploration. As RCS progresses, less metabolic resources need to be invested in root maintenance of the
cortex, which could permit allocation of saved resources to the growth of shoots, other roots, and
reproduction. Reduced hydraulic conductivity induced by RCS may also be advantageous under drought
conditions by preventing desiccation of the root tip and surrounding soil. Under drought conditions,
plants must conserve the soil water throughout the growth season as a drought tolerance mechanism.
We propose that RCS merits investigation as a useful trait in improving the performance of barley and
other RCS forming species under edaphic stress.
326 Increasing Sugar Transport to Improve Soybean Response to Elevated [CO2]
Pauline Lemonnier, University of Illinois
Jennifer Quebedeaux, University of Illinois
Elevated atmospheric [CO2]causes a direct increase in instantaneousphotosynthesis and sugar
production in C3 plants, leading to a yield increase which is promising to meet future food demand.
However, previous studies have shown that soybean yield does not increase as much as predicted under
elevated [CO2]. This is possibly the result of feedback inhibition of photosynthetic capacity by
accumulated photoassimilates, suggesting insufficient sugar export capacity from the photosynthetic
source to sink tissues. Increasing sugar export capacity through over-expressing proton/sucrose
symporters involved in phloem loading has the potential to increase photosynthesis and yield under
elevated [CO2]. In this study, soybean was transformed to overexpress the sucrose
transporter AtSUC1H65K. Transgenic soybeans were grown at ambient (400 ppm) and elevated (600 ppm)
[CO2] at the Soybean Free Air Concentration Enrichment facility during 2016. Instantaneous light-
saturated photosynthesis and stomatal conductance were measured as well as the response of
photosynthesis to intercellular [CO2]. Leaf and pod tissues were collected to confirm the expression
of AtSUC1H65K, determine sucrose transport activity, and measure carbohydrate content. The effects of
altered sucrose transport and elevated [CO2] on seed yield were measured at the end of the growing
season.
327 Dissecting the Mechanism of Grain Yield Under Drought in Rice- Phenotypic and Genomic Insight
Supratim Basu, New Mexico Consortium
Miranti Rahmaningsih; Anuj Kumar; Venkategowda Ramegowda; Julie Thomas; Andy Pereira
Water deficit therefore strongly affects rice production. Among the major stages of rice growth and
development, the reproductive phase is the most affected by water deprivation. Water stress imposed
during this stage can reduce grain yield by up to 77% and longer periods of drought can reduce 90% of
grain yield. Plants are protected against drought by three different mechanisms: drought avoidance,
drought tolerance, and drought escape. To evaluate the genetic variation among rice accessions for
drought tolerance, fifteen rice genotypes from the USDA mini-core collection were randomly selected
for the study with N22 and Vandana and Nipponbare as the positive and negative control, respectively.
Water stress was applied 3-4 days before anthesis stage and the drought response was measured by
analysis of the reduction in yield parameters. Drought resistance levels were categorized by comparing
genotypes under drought and wellwatered control for measurements of panicle length, number of grain
per panicle, number of filled grain per panicle, 100-grain weight, and total grain weight per panicle.
Three genotypes, AMANE, WIR 3039, HKG98 identified from screening and resistant reference genotype
N22 were used further for gene expression analysis. Gene expression analysis revealed that
inflorescence tissue gives positive and higher correlation with phenotypic measurements than flag leaf
during reproductive stage drought. Expression levels of invertase genes and transcription factors exhibit
positive effects to drought resistance particularly in relation to number of grain per panicle and panicle
length. Therefore, our results will provide valuable information for dissecting the genetic basis of
drought resistance in rice as well as provide valuable resource for developing high yielding drought
resistant rice cultivars by breeding approaches.
328 Spatio-temporal Metabolome Profile in Tobacco and Soybean under Water Deficit Conditions
Roel Rabara, New Mexico Consortium
Prateek Tripathi, The Scripps Research Institute; Paul Rushton, 22nd Century Group Inc
Drought is a major environmental stress affecting crop production in a global scale. Understanding how
plants respond to water deficit is important in order to develop crops tolerant to drought. In this study,
we compare metabolome profile of tobacco and soybeans at different stages of water stress in both
leaves and roots. Comparisons between the two datasets reveal common responses between the two
species, responses specific to one of the species, responses that occur in both root and leaf tissues, and
responses that are specific to one tissue. Stomatal closure is the immediate response of the plant and
this did not coincide with accumulation of abscisic acid. A total of 116 and 140 metabolites were
observed in tobacco leaves and roots, respectively, while 241 and 207 were observed in soybean leaves
and roots, respectively. Accumulation of metabolites is significantly correlated with the extent of
dehydration in both species. Among the metabolites that show increases that are restricted to just one
plant, 4-hydroxy-2-oxoglutaric acid (KHG) in tobacco roots and coumestrol in soybean roots show the
highest tissue-specific accumulation. The comparisons of these two large nontargeted metabolomics
datasets provide novel information and suggest that KHG will be a useful marker for drought stress for
some members of Solanaceae and coumestrol for some legume species
329 The Development of a Predictive Model for Freezing Resistance in Zea mayusing In-situ
Measurements of Biochemical Modifications of Epicuticular Waxfollowing Chilling Treatment in
Controlled Environment and Field Conditions
Kaila Hamilton, University of Saskatchewan
Karen Tanino, University of Saskatchewan
The inability of Zea mays to resist yield-reducing frost damage has limited the opportunity for its
production in underutilized regions such as the Canadian prairies. Underutilized regions will be
especially important as climate change begins to dictate new landscapes. The in-situ surface lipid
composition of mature leaves in grain corn was studied under controlled environment and field
conditions using Attenuated Total Reflectance (ATR) Fourier Transform Infrared (FTIR)
Spectromicroscopy to evaluate chemical changes induced by chilling treatment in the epicuticular layer.
Plants were exposed to chilling treatment conditions over a 10 day (18℃/6℃) period and lethal freezing
temperature was determined using thermal imaging. We are developing a multifactor regression model
to use as a predictor to estimate lethal freezing temperature. The lipid composition, specifically aliphatic
compounds moderate leaf surface hydrophobicity. Hydrophobicity has been linked to increased freezing
resistance. Experimental FTIR results showed changes, following chilling treatment, in the asymmetrical
stretching region of the CH2 group associated with aliphatics which comprise the plant cuticle (cutin,
waxes and cutan). The ongoing development of this model in Zea mays appears to be a useful system
with practical applications for evaluating the correlation between abiotic stress induced epicuticular wax
modification on freezing resistance in a whole plant system.
Bioinformatics Tools, Data Standards, and Data Processing
400 Bioinformatics Tools, Data Standards, and Data Processing
Laurel Cooper, The Planteome Project
Austin Meier, Oregon State University; Justin Elser, Oregon State University; Marie-Angélique Laporte,
Bioversity International; Justin Preece, Oregon State University; Chris Mungall, Lawrence Berkeley Lab;
Elizabeth Arnaud, Bioversity International; Pankaj Jaiswal, Oregon State University
The Planteome project (www.planteome.org) features a suite of reference ontologies for plants,
associated with a growing corpus of genomics data in a centralized online plant informatics portal. The
species-neutral references, the Plant Ontology, Plant Trait Ontology, and Plant Environment Ontology,
(developed by the project), along with those developed by collaborating groups, such as the Gene
Ontology and the Phenotype and Trait Ontology, and others, are mapped to species-specific controlled
vocabularies for crop plant traits and phenotypes. Traditional plant breeding methods for crop
improvement may be combined with next-generation analysis methods and automated scoring of traits
and phenotypes to develop improved varieties. Linking these analyses to the growing corpus of
genomics data generated by high-throughput sequencing, transcriptomics, proteomics, phenomics and
genome annotation projects requires common, interoperable, reference vocabularies (ontologies) for
the description of the data. Data annotations to ontology terms link phenotypes and germplasm to
genomics resources, and provide semantic integration of widely diverse datasets with the goal of plant
improvement. Analysis and annotation tools are being developed to facilitate studies of plant traits,
phenotypes, diseases, gene function and expression and genetic diversity data across a wide range of
plant species. The project database and the online resources provide researchers tools to search and
browse and access remotely via APIs for semantic integration in annotation tools and data repositories,
providing resources for plant biology, breeding, genomics and genetics. The project is supported by the
National Science Foundation award IOS #1340112
401 On the Statistical Mechanics of Cytosine DNA Methylation and the Detection of the Methylation
Regulatory Signal
Robersy Sanchez, University of Nebraska-Lincoln
Sally Mackenzie, University of Nebraska-Lincoln
DNA methylation patterning represents one feature of the epigenome that is highly responsive to
environmental stress and associates with transgenerational adaptation in plants. An information
thermodynamics theory on cytosine DNA methylation (CDM) was recently published. Results indicated
that most methylation changes occurring within cells are likely induced by thermal fluctuations to
ensure thermal stability of the DNA molecule, seemingly explainable by statistical mechanics laws.
Therefore, one does not observe a genome-wide relationship between methylation and gene
expression. Ignoring the statistical biophysics subjacent to CDM is a source of bias for the most common
statistical tests applied in methylome analysis. A key limitation to current genome-wide methylation
analysis is the inability to discriminate signal, induced by experimental treatment, from background
“noise” that emanates from this natural, dynamic methylation activity. Here, we provide novel insights
for application of statistical mechanics of CDM to detect methylation regulatory signal. Integration of
information thermodynamics of CDM and signal detection theory permits robust discrimination of
biological signal from physical noise-induced thermal fluctuations. The analytical steps summary follows:
i) Information thermodynamic theory computes the probabilities of signal plus noise, with use of
experimental controls to estimate the receiver’s threshold (cutoff Hellinger divergence value, HD) at
which the rate of false positive is minimal or at least acceptable for the experimental conditions. ii)
Potential differentially methylated positions (DMPs) are proposed based on the probabilities of their
corresponding HDs. iii) A logistic regression analysis is performed with the prior binary classification of
DMPs (from treatment versus control), and iv) a receiver operating curve (ROC) is built to estimate the
cutoff point for HD at which an observed methylation level represents a true DIMP. Our approach is
illustrated with examples from analysis of genome-wide CDM reprogramming induced by msh1 mutant
effect in Arabidopsis thaliana.
402 Annotating Germplasm to Planteome Reference Ontologies
Austin Meier, Oregon State University
Justin Elser, Oregon State Univeristy; Laurel Cooper, Oregon State University; Pankaj Jaiswal, Oregon
State University; Marie-Angélique Laporte, Bioversity International
An expected use case of plant phenotype ontologies will be the identification of germplasm containing
particular traits of interest. If phenotype data from experiments is annotated using ontologies, it makes
sense to include annotations to that germplasm source. A lack of standardized data formatting reduces
the utility of these data. Standardizing germplasm data, including links to germplasm databases, or
distribution locations improves collaboration, and benefits both researchers and the scientific
community as a whole. All plant traits contained in the Planteome reference ontologies are searchable,
and interconnected through relationships in the ontology. All data annotated to these reference
ontologies will be displayed, shareable, and computable through the Planteome website
(www.planteome.org) and APIs. This manuscript will discuss the advantages of standardizing germplasm
trait annotation, and the semi-automated process developed to achieve such standardization.
403 Diversity Seek - A Community Driven Effort to Harness the Genetic Potential of Genebanks
Ruth Bastow, Global Plant Council
DivSeek is a community driven effort that aims to unlock the potential of crop diversity stored in
genebanks around the globe for food and nutritional security, and societal and economic benefits.
Assessing and identifying new sources of genetic variation is a critical part of any long-term strategy to
enhance the productivity, sustainability and resilience of crop varieties and agricultural systems.
Approximately seven million crop accessions are currently being conserved in genebanks collections
worldwide. This resource represents one of the greatest, largely untapped, opportunities for
accelerating yield gains and overcoming emerging crop productivity bottlenecks.
Many projects across the globe are characterizing crop diversity using genomic and phenomic
technologies, associated with cyberinfrastructure and high performance computing. As a result a great
many data sets are being generated. To ensure this data is findable, accessible, interoperable and
reusable (FAIR) the DivSeek initiative is working with partners across the globe to help bridge the
information requirements of gene bank curators, plant breeders and upstream biological researchers to
facilitate the use of plant genetic variation to accelerate the rate of crop improvement and furnish food
and agricultural products to the growing human population.
http://www.divseek.org/
404 Designing Intra-plot Variability Features from High-Throughput Sorghum Phenotypes
Karthikeyan Ramamurthy, IBM Research
Addie Thompson, Purdue University; Zhou Zhang, Purdue University; Fangning He, Purdue University;
Melba Crawford, Purdue University; Clifford Weil, Purdue University; Ayman Habib, Purdue University;
Mitchell Tuinstra, Purdue University
Traditional crop breeding approaches have relied on manually collecting ground measurements from
hundreds of phenotypes. Data collection protocols in automated, high throughput phenotyping have
resulted in more variety in the acquired data (RGB, hyperspectral, thermal images, and LiDAR point
clouds), more data points per plot, more genotypes and more time points.
Currently, machine learning methods are used to train trait prediction models using features extracted
from the image data as input and the desired traits as output. Good feature engineering using a
combination of domain knowledge as well as data considerations is crucial for accurate and robust high-
throughput phenotyping. The small sample size of manually collected output traits limits the use of fully
automated feature engineering methods.
Generally, summary features extracted from each plot are used for training such machine learning
models. For example, in our previous work we used primitive features such as height histograms of 3D
point clouds reconstructed from RGB images, and raw values, band ratios and morphological features
extracted from hyperspectral data, all summarized at the plot level. This was used to predict height,
stalk diameter, volume, and biomass of sorghum.
In this work, we develop principled feature engineering methods to represent the intra-plot variability of
such primitive features across different spatial dimensions. We model the spatial variability of different
heights in the 3D point cloud using eigen-spectral techniques. Since the hyperspectral data is registered
to the point cloud data at a pixel level, we use similar methods to capture its spatial variability across the
plot as well as along the height dimension. We assess the improvement obtained in phenotypic trait
prediction accuracy in sorghum by incorporating these intra-plot variability features. We also attempt to
intuitively comprehend the information provided by these features about plot-level variability.
405 A Sorghum Panicle Annotation and Counting Tool
Karthikeyan Ramamurthy, IBM Research
Peder Olsen, IBM Research; Upendra Chitnis, IBM Research; Addie Thompson, Purdue University; Naoki
Abe, IBM Research; Zhou Zhang, Purdue University; Melba Crawford, Purdue University; Fangning He,
Purdue University; Ayman Habib, Purdue University; Mitchell Tuintstra, Purdue University
Detecting and counting sorghum panicles from aerial images is of significant interest in automated high-
throughput phenotyping. The number of panicles in a plot correlates well with the number of fertile
tillers as well as the contribution of grain to biomass, both key phenotypes. However, panicle detection
is difficult due to the many variations in panicle appearance, camera viewing angle, occlusion, and
overlap. The appearance varies in size (20cm2 to 1000cm2 in area), shape (spindle, broom, cylinder or a
lax cone), color (chalky white, green, yellow, rusty brown or black), pose, and grain-size. Simple
classification methods based on color, size, or texture features do not work in panicle recognition. To
develop a robust machine learning model for recognizing and counting panicles, a well-annotated,
diverse data set is needed.
A data set collected by Purdue University for the ARPA-E TERRA program contains 400 RGB images over
a field with hundreds of sorghum varieties. Each image contains 8000x5000 pixels with 1cmx1cm
resolution. Rectangular or ellipsoidal bounding boxes are not appropriate for panicle annotation. Hence
we used super-pixels to annotate the extents of panicles. Super-pixels are natural and perceptually
meaningful segments of an image. We generate super-pixels automatically, and let a user click on a
super-pixel to select a part of a panicle. Typically, one or two clicks suffice to annotate a panicle, and we
derive panicle counts based on these annotations. The alignment between super-pixels and panicles
depend on how well the panicle stands out from the background. The preliminary annotation
performance is promising. Errors can be further reduced by allowing variation in super-pixel sizes. The
super-pixels in the first few images is then used in a template matching algorithm to automate the
annotation of future images. The user will then simply need to correct errors.
406 Predicting Cleavage Activity of CRISPR-Cas9 Across Diverse Maize Germplasm using Reference
Genome Analysis
Ian Braun, Iowa State University
Carolyn Lawrence-Dill, Iowa State University; Dipali Sashital, Iowa State University; Kan Wang, Iowa
State University; Jeffrey Wolt, Iowa State University
The CRISPR-Cas9 system enables sequence-specific editing at target sites within a genome. However,
editing may also occur at other points in the genome (off-target sites) where sequences are similar. In
particular, off-target cleavage presents a problem for specific editing in large, polyploid, or redundant
eukaryotic genomes such as that of maize. Exacerbating this problem, target sites for CRISPR-Cas9 often
are selected within the context of a reference genome, and then applied to different germplasm
altogether. In such cases, it would be valuable to estimate the number of expected off-target sites
unique to the new line. We compared the maize reference genome B73 with the genomes of four
additional sequenced lines (B104, W22, Mo17, and PH207) to estimate this relationship. The predicted
ratios between unique off-target sites and the number of off-target sites in the reference genome are as
large as 0.15, depending on the specific germplasm analyzed. These findings likely are influenced by
various confounding factors, including the relative quality of genomes analyzed. The extent to which
predicted off-target rates vary based on genuine genomic differences versus such artifacts is discussed.
Implications for experimental design are described.
407 An Efficient, Scalable Data Store for Automated High Throughput Phenotyping
Upendra Chitnis, International Business Machines
Karthikeyan Ramamurthy, International Business Machines; Upendra Chitnis, International Business
Machines
Modern crop breeding experiments generate several petabytes of data in different formats. UAVs and wheel-based systems collect petabytes of image and environment data that encode important phenomic characteristics of crops across spatio-temporal dimensions. Manual phenotyping and other surveys generate substantial amount of structured data. A mechanism to efficiently store and query these disparate data sets is necessary for developing downstream exploratory and predictive analytic approaches that make use of it. Such a data store, among other things, must have: 1. Scalable approaches for ingesting and processing petabytes of data in a reasonable amount of time. 2. Scalable file system for storing thousands to millions of files that vary widely in their sizes from a few MBs to 10s of GBs. 3. Data indexing features for efficient queries across spatio-temporal and other dimensions . 4. Data join capabilities across spatio-temporal and other dimensions. 5. Support for distributed computing architectures used by downstream exploratory and predictive analytics algorithms. No single off-the-shelf product supports all these features. Data stores such as MongoDB are scalable but provide very limited join capabilities for raster datasets. Big data stores such as HBase are very good at horizontal scaling, but do not provide spatial query capabilities. Some relational data stores, such as Postgres, provide good support for spatial queries but do not have join capabilities for raster images. In this presentation we will describe a scalable data platform solution for efficiently storing and querying high throughput phenotyping data. It is built on big data stack and supports distributed computing architecture necessary for downstream analytic applications. The solution includes linearly scalable custom algorithms to join different data layers such as soil, weather, elevation with custom phenotypic and image data sets. These data layers could be at different spatio-temporal resolutions and use different geographic coordinate systems (datums).
408 Quantitative Analysis of Cotton Canopy Growth and Development in Field Conditions using
Consumer-grade RGB-D Imaging
Yu Jiang, University of Georgia
Changying Li, University of Georgia; Tariq Shehzad, University of Georgia; Andrew Paterson, University of
Georgia
Field-based high throughput phenotyping (FB-HTP) systems with imaging capabilities can rapidly acquire
various imaging data of plants in field conditions, which enables the possibility of quantifying plant
canopies. The goal of this study was to develop approaches to quantitatively analyze cotton plant
canopy by extracting morphological traits from images collected in the field. An imaging-based FB-HTP
system (GPhenoVision) was used to scan a field that was planted with cotton plants in Watkinsville, GA,
in 2016. The field included four cotton cultivars of 24 plots each. Color and depth images with GPS
information were acquired in the field on four days covering growth stages from plant germination to
boll maturity. Based on GPS information, the collected color and depth images were segregated and
reconstructed into colored point clouds of individual plots. In the colored point clouds, cotton plant
canopies were segmented from the background using an excess-green (ExG) based color filter, and the
segmented canopy points were used to extract six morphological traits including canopy length, width,
height histogram, projection area, coverage ratio, and volume on each day. Changing rates of the six
traits were calculated as growth rates between two consecutive scanning dates, and thus a total of 42
phenotypic traits (6 traits per day by 4 days plus 6 growth rates per period by 3 periods) were extracted
to quantify canopy growth and development of cotton plants. Statistical analyses were conducted
between the extracted traits and cotton cultivars, and the results showed that certain traits were
statistically different among cultivars, indicating the efficacy of the proposed approach for quantification
of canopy growth and development. Additionally, the trait differences among cultivars could be
correlated with fiber yield for selecting genotypes with high yields.
409 Single Cell DNA-sequencing to Monitor Homoeologous Recombination Frequency in Wheat
Venkatesh Bollina, National Research Council Canada
Pankaj Bhowmik, National Research Council Canada; Erin Higgins, Agriculture and Agri-Food Canada;
Isobel Parkin, Agriculture and Agri-Food Canada; Andrew Sharpe, National Research Council Canada;
Sateesh Kagale, National Research Council Canada
To exploit the potential of wild relatives of bread wheat carrying beneficial traits to address biotic and
abiotic factors rely on successful crossover formation of homoeologous chromosomes during meiosis.
Meiotic recombination plays a vital role in acquiring new traits and genetic variation into commercial
crops. However, polyploidy crops like wheat and Brassica have a strict homologous pairing mechanism
controlled by a pairing homoeologous gene 1 (Ph1) and Ph1-like genes. Ph1 locus impedes the process
of modulation of homoeologous chromosomes between alien donor and a cultivated wheat variety.
There are several attempts to develop a Ph1 deletion mutant line through chemical and physical
mutagenesis and gene silencing approaches. Conventionally, laborious and time-consuming cytological
methods have been employed to study the homoeologous chromosomal pairing behaviour. Lack of a
fast and vigorous method to study the homoeologous recombination frequency hinders the progress of
exploiting genetic variation and elite traits in wild relatives. The main objective of this study is to
establish an easy and efficient method for monitoring the impact of modulation of recombination in
plants. We have devised a strategy to assess homoeologous recombination frequency in an F1 plant
which leverages the combined advantages of single cell whole genome sequencing technology. Single
cell haploid microspore from an F1 plant of an interspecific cross is the ideal material to quickly assess
homoeologous recombination frequencies as it is relatively easy to isolate thousands of microspores
carrying segregating genotypes. Capturing the F1 plant microspores in a high throughput manner using
fluorescence-activated cell sorting (FACS) or Fluidigm C1 single cell model followed by DNA isolation and
amplification workflow. And subsequent genotyping of multiple segregating microspores facilitates
assessment of the frequency of homoeologous recombination. The successful implementation of this
novel F1 microspore sequencing approach will provide a cost-effective breeding tool for genotyping and
assessment of recombination frequency.
410 The TERRA REF Hyperspectral Imagery Workflow
Jiarun Mao, University of California, Irvine
David LeBauer – University of Illinois at Urbana-Champaign, National Center for Supercomputing
Applications; Charles Zender – University of California, Irvine
Hyperspectral imaging provides insight into plant functioning that would be difficult or impossible to
measure using traditional field measurements. While there is much to be learned from these sensors,
widespread use requires expensive hardware as well as high technical expertise. The TERRA Reference
Phenotyping Platform (TERRA REF) is creating a high resolution public reference data set including
hyperspectral imaging sensors.
The TERRA REF team has deployed two Headwall hyperspectral sensors on a Lemnatec Field scanner in
Maricopa, AZ. These sensors cover the visible and infrared (300-2500 nm) portion of the light spectrum
at mm spatial and nm spectral resolution over a one acre field. Throughout the growing season, these
sensors scan diverse sorghum lines. Key features of the pipeline include self-describing output metadata
based on Climate and Forecasting (CF) standards, calibration from hyperspectral exposure image to
reflectance data, lossless compression to the raw imagery and self quality assurance. Using parallel
computing and optimal memory use in High Performance Computing environments (HPC), the pipeline
is capable of processing fifty 64 GB scans, each producing 256 GB of calibrated reflectance data. The
hyperspectral data pipeline requires the netCDF Operators (NCO) and Python netCDF4 libraries and uses
netCDF as the format of final products. The source code of the pipeline is available at
github.com/terraref/extractors-hyperspectral.
411 A Graphical Approach to Data Modeling and Storage
Mallory Lai, University of Wyoming
Brent Ewers, University of Wyoming
One of the great difficulties facing scientists today is the integration and curation of large sets of
disparate, yet intertwined, data. We propose a graph-centered approach to ‘omics data integration that
will provide scientists with effective models for analysis and the tools to manage large and complex data
sets. Here, we illustrate how we can use probabilistic graphical models, known as Dynamic Bayesian
Networks (DBN), to capture regulatory relationships using temporal transcriptomic and phenomic data
from Brassica rapa. These networks capture the data probabilistically and allow us to make predictions
about the plants using conditional and joint probabilities. Its graphical framework allows for subsequent
integration and storage into the open-source graph database Neo4j. As a graph database, Neo4j is
capable of handling diverse, heterogeneous data as directed graphs which can be accessed by
collaborators and other scientists. Once stored in Neo4j, the network can be easily queried and its
results returned in a visually intuitive form—a graph!
412 A Meteorological Data Format that Makes Research Easier and More Reliable will also Rork for
Imaging Sensors.
David LeBauer, University of Illinois
Elizabeth Cowdery, Boston University; Ankur Desai, University of Wisconsin; Michael Dietze, Boston
University; Rob Kooper, National Center for Supercomputing Applications; Shawn Serbin, Brookhaven
National Laboratory; Alexey Shiklomanov, Boston University; James Simkins, University of Wisconsin
Standard data formats make it easier to share and reuse data. There are dozens of formats for climate
data and dozens of algorithms with bespoke requirements. After we specified the Climate and
Forecasting standard data format for climate data, we could reduce the problem of matching n weather
data types to m weather-driven crop and ecosystem models from n × m to n + m. This means that we
can focus on fewer conversion, downscaling, gap-filling, and other algorithms while making them more
reliable and accessible. These tools have greatly reduced the technical overhead required to use
computer simulation models in plant and ecosystem research, and are available in the Predictive
Ecosystem Analyzer (PEcAn) R package for atmospheric data.
This talk will describe the PEcAn meteorological workflow and propose an analogous framework for
sensor data pipelines. The TERRA REF and PEcAn teams are extending the CF data format to cover
sensor data sets, starting with hyperspectral imagery. We suggest that this will make life easier for the
community. A common interchange format makes it easier to exchange data without requiring that all
data meets a single standard or specification.
413 Metadata Standards for a Large-scale Plant Phenotyping System
David LeBauer, University of Illinois
Craig Willis, NCSA/University of Illinois
A central goal of the TERRA-REF integrated plant phenotying system is to produce reference data that
are interoperable, extensible, and understandable by users. The TERRA-REF system combines
information from a variety of domains including agricultural field experiments, manually collected field
data, sensor-missions, derived trait data, and genome sequences.
This poster describes our approach combining multiple domain-specific vocabularies and ontologies
using the JSON-LD (JSON-based Serialization for Linked Data) format. This allows us to maximize
reusability and interoperability across domains, while enabling different levels of implementation
complexity. The TERRA-REF metadata standards are composed from the ICASA, Crop/Agronomy
Ontology, Open Geospatial Consortium (OGC), Climate & Forecast, and Gene Ontology families of
standards. JSON-LD combines the flexibility of the ubiquitous JSON data format for programmatic access
with the power of standard vocabularies and ontologies to fulfill our goals of interoperability,
extensibility, and understandability.
414 Plant Comparative Phenomics
George Gkoutos, University of Birmingham
Within the biomedical community, one of the most successful strategies for achieving standardization
and integration of biomedical knowledge, data and associated experiments was proposed more than a
decade ago with the advent the advent of the Gene Ontology. Since then, the biomedical community
has invested a considerable amount of effort, research and resources in the development of ontologies
that are now becoming and increasingly successful as information management and integration tools.
This presentation will focus on how phenotype data ontology-based standardisation can enable a variety
of applications. It will employ examples stemmed from the biomedical domain and attempt to provide
directions of how these could be applied in the plant domain.
415 Maize - GO Annotation Methods Evaluation and Review (Maize-GAMER)
Kokulapalan Wimalanathan, Iowa State University
Carson Andorf, USDA - ARS; Iddo Friedberg, Iowa State University; Carolyn Dill, Iowa State University
Making a genome sequence accessible and useful involves three basic steps: genome assembly,
structural annotation, and functional annotation. The quality of data generated at each step influences
the accuracy of inferences that can be made, with high-quality analyses producing better datasets that
result in stronger hypotheses for downstream experiments. Here we report a new, high-confidence
functional annotation of genes for the maize B73 reference genome. To develop this annotation set, we
used sequence similarity- and protein domain-based methods as well as mixed methods developed for
the Critical Assessment of Function Annotation (CAFA) competition. Individual annotation sets as well as
combined outputs of multiple methods were compared both to each other, and to the existing datasets
from Gramene (release 49) and Phytozome (release 12). Our new functional annotation increases the
number of genes that are assigned at least one functional annotation (GO term) as well as the quality of
functional assignments on average (based on F-measure). Annotations derived from the GO Annotation
Methods and Evaluation (GAME) pipeline will be made accessible via MaizeGDB
(http://www.maizegdb.org).
Plasticity in Plant Traits
500 Assessment of Plant Performance Traits in Controlled Environments and Translation to the Field
Astrid Junker, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK)
Marc Heuermann, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK); Rongli Shi, Leibniz
Institute of Plant Genetics and Crop Plant Research (IPK); Henning Tschiersch, Leibniz Institute of Plant
Genetics and Crop Plant Research (IPK); Matthias Lange, Leibniz Institute of Plant Genetics and Crop
Plant Research (IPK); Daniel Arend, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK);
Jean-Michel Pape, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK); Rhonda Meyer,
Leibniz Institute of Plant Genetics and Crop Plant Research (IPK); Kathleen Weigelt-Fischer, Leibniz
Institute of Plant Genetics and Crop Plant Research (IPK); Michael Grau, Leibniz Institute of Plant
Genetics and Crop Plant Research (IPK); Andreas Boerner, Leibniz Institute of Plant Genetics and Crop
Plant Research (IPK); Uwe Scholz, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK);
Thomas Altmann, Leibniz Institute of Plant Genetics and Crop Plant Research (IPK)
To meet the challenges in global food security requires the development of strategies towards the
optimization of yield formation and resource efficiency of crop plants under future climate scenarios. To
enable this, a deeper understanding of processes underlying plant acclimation to changing
environments is necessary. Using high throughput automated plant phenotyping systems at IPK, the
dynamics of plant growth and photosynthetic efficiency have been studied in accession panels of
Arabidopsis and maize under controlled conditions. This led to the identification of Arabidopsis and
maize candidate accessions with different acclimation strategies to high light and variations in biomass
yield and photosynthetic efficiency, respectively. To evaluate and enhance the prospects of lab-to field
translation, candidate maize accessions have furthermore been phenotyped for root- and shoot traits
under different cultivation conditions in the glasshouse and in the field. Modification of the standard
cultivation procedures improved the lab-to-field translation of phenotypic trait expression which will be
further optimized in the Plant Cultivation Hall currently being erected at IPK. This building will enable to
run specifically designed and reproducible cultivation scenarios that mimic field conditions. In this way,
trait expression similar to that in the field will be triggered and monitored using automated phenotyping
installations. Investigations involving precise variation of environmental parameters will yield a deeper
understanding of acclimation processes and underlying molecular mechanisms and genetic
determinants and will support the prediction of idiotypes with improved performance under future
climate scenarios. Furthermore recent achievements in phenotype data management, standardized
metadata representation (e.g. MIAPPE), and data publication will be presented.
501 The Impact of Short Petiole Length in Soybean Yield
Ursula Ruiz-Vera, University of Illinois at Urbana-Champaign
Benjamin Campbell, University of Minnesota; Venkatraman Srinivasan, University of Illinois at Urbana-
Champaign; Robert Stupar, University of Minnesota; Aaron Lorenz, University of Minnesota; Donald Ort,
USDA-ARS and University of Illinois at Urbana-Champaign
At the current rate of progress in yield improvement, agricultural will fall to feed a ~34% larger
population by 2050. The maximum yield (Yp) obtained for a crop is determine by the total incident solar
radiation available during the growing season multiplied by the light interception efficiency (Ei), the
energy conversion efficiency (Ec) and the partitioning efficiency (Ep). Ec is primarily determined by net
photosynthesis, does not operate near its theoretical maximum and is the remaining factor with the
potential to increase Yp by maximizing the proportion of solar energy intercepted by the crop and used
to build biomass. Model simulations indicate that one way to improve Ec is by altering the canopy
architecture. An important component of the canopy architecture are petioles but it is not known how
their length may affect light penetration into the canopy and thereby affect yield. In this experiment, 4
soybean NIL (near-isogenic lines) pairs of short and normal petiole length were grown in a randomized
block design experiment (4 blocks and row spacing of 0.254m) to determine the effect of petiole length
on photosynthetic physiology and yield. We hypothesized a higher yield in the short petiole compared to
the standard petiole soybean reasoning that short petioles at the top of canopy would permit more light
penetration to lower canopy layers where light could be more efficiently used in photosynthesis. There
were no differences in leaf photosynthesis, above-ground biomass, seed weight or harvest-index
between the short and non-short petiole plants within any given NIL pair. However, there were
differences among the four NIL groups for seed weight and harvest-index. The determinations of Ei, Ec
and Ep are still in progress, but the preliminary results suggest that a petiole size up to half of normal did
not increase yield compared to standard petiole soybean.
Modeling
600 Testing a Whole-plant Biophysical Model with High Throughput Electrical Conductivity
Measurements of Plant Hydraulics
Brent Ewers, University of Wyoming
Timothy Aston, University of Wyoming; Carmela Guadagno, University of Wyoming; Mallory Lai,
University of Wyoming; Cynthia Weinig, University of Wyoming; David Mackay, University at Bufffalo;
Jonathan Pleban, University at Buffalo
Biophysical models of plant response to stress help bridge the gap between the genome and phenome
and improve predictive understanding. The processes behind organ and whole plant carbon, water and
nutrient use are generally known, but a priori parameters of the governing equations are poorly
predicted for novel genotypes or varying environmental conditions. Many biophysical parameters are
directly related to expression of functional genes. The means of quantifying parameters vary from easy
yet destructive measurements of biomass to difficult continuous gas exchange and transport between
organs. We test the hypothesis that electrical conductivity measurements of tissues and whole plants
provide temporally rigorous proxies of input parameters, e.g. whole plant hydraulic conductance, and
output time series, e.g. stomatal conductance and predictions of plant growth. Measurement and data
curation of these traits on 100s to 1000s plants has severely limited phenotyping throughput. We make
an initial test of high throughput electrical conductivity measurements of droughted and well-watered
plants with TREES, a mechanistic model of rhizosphere-plant-atmosphere integrated water, carbon and
nutrient processes. We combined open source data acquisition and processing with the graph database
Neo4j to set up a workflow that enables inexpensive yet rigorous collection of electrical conductivity and
model tests. We also measured at discrete time points leaf photosynthesis, biomass partitioning and
environmental conditions, which are also needed to test the model on the herbaceous plant, Brassica
rapa. We found that 1) electrical conductivity provides a continuous estimate of plant hydraulic stress
that has been missing from herbaceous plant studies, 2) the graphical database improved the efficiency
and accuracy of data to model testing, and 3) the workflow accommodates other –omics data such as
RNAseq, metabolomics and UAV-acquired spectral estimates. All of these improve predictive
understanding of the genome to phenome in response to varying genotypes and stress.
Education & Outreach
700 Building High-throughput Plant Phenotyping Devices while Training the Next-generation of Users.
David Mendoza, University of Missouri
High-throughput plant phenotyping is an emerging and fast-moving field that requires active
collaboration across disciplines, including plants sciences, engineering and computer sciences. Currently,
there are commercial instruments that can be acquired to pursue large-scale high-throughput
phenotyping (HTP) experiments. However, there are also small labs interested in pursuing HTP screens
using diversity panels available for some plant species or mutant collections that require very specific
conditions. Moreover, some labs may have needs - in terms of design - that may not be commercially
available. So who is going to build these devices and how can we ensure that our students are well
trained to take advantage of HTP technologies?
At the University of Missouri, we have started an interdisciplinary program where teams of students
from engineering and plant sciences work together to design, build and test automatic platforms for
high-throughput phenotyping of plants. Furthermore, during the entire academic year (2015-2016) an
MU journalist followed the team and conducted interviews at regular intervals to document how
members from distant fields exchange ideas to solve the different problems that appeared at different
stages of the project.
Our first cohort of interdisciplinary students graduated in 2016 after successfully assembling S.P.I.P.
v1.0 [Small Plant Imaging Platform v1.0].
Learn more about the project @ NSF Broader Impacts: https://goo.gl/cl580t
701 NSF Research Traineeship – P3, Predictive Plant Phenomics
Carolyn Lawrence-Dill, Iowa State University
Julie Dickerson, Iowa State University; Thedore Heinde, Iowa State University l; Patrick Schnable, Iowa
State University
New methods to increase crop productivity are required to meet anticipated demands for food, feed,
fiber, and fuel. Using modern sensors and data analysis techniques, it is now feasible to develop
methods to predict plant growth and productivity based on information about their genome and
environment. However, doing so requires expertise in plant sciences as well as computational sciences
and engineering. Through P3, we bring together students with diverse backgrounds, including plant
sciences, statistics, and engineering, and provide them with data-enabled science and engineering
training. The collaborative spirit required for students to thrive in this unique intellectual environment
will be strengthened through the establishment of a community of practice to support collective
learning. This traineeship anticipates preparing forty-eight (48) doctoral students, including twenty-eight
(28) NRT funded doctoral students, with the understanding and tools to design and construct crops with
desired traits that can thrive in a changing environment.