iPlant Hosts Platform for Field-Based Imaging of Root Traits

Alexander Bucksch collects beans in the field to analyze their root traits, working to identify crops best suited to adapt to climate change. "Bean is the most important food legume in the world, and a major food source for 200 billion people in the sub-Saharan and 200 billion people in the tropics," he said.

By Shelley Littin, iPlant Collaborative

Researchers led by a team at the Georgia Institute of Technology (Georgia Tech) have developed an online platform that enables plant scientists to obtain quantitative phenotype information on the root systems of plants imaged in the field.

The platform, called Digital Imaging of Root Traits (DIRT), is now hosted by the iPlant Collaborative’s computational infrastructure, as described in a recent publication in the journal Plant Methods. Researchers anywhere in the world with an Internet connection can access the program by logging into an iPlant account.

The idea is to expedite and simplify the process of collecting measurements of plant roots in the field, said Alexander Bucksch, a research scientist in Joshua Weitz's group at Georgia Tech, and corresponding author on the paper. “Visual phenotypes of a plant can be computed reproducibly with imaging, including features impractical to measure by hand,” Bucksch explained.

The research team spent three years collecting plant root samples at field stations in the United States and South Africa and manually measuring dozens of traits including root density, angles, surface area, number of roots, and many more. They then configured calculations based upon these standardized measurements, to produce a program capable of giving highly accurate measurements for root traits based upon a photograph of roots.

Maize is another important crop that researchers are testing for survivability. "If you look at the soil you can visually guess that this crop is growing in a desert with low fertility soil," Bucksch said.

To use the platform, a scientist need only lay out the roots next to a marker for scale and take a photograph, and the program will provide many dozens of phenotype measurements. “DIRT provides a pipeline to move from a field-based image to quantitative data as part of studies by academics and breeders alike,” said Joshua Weitz, associate professor of biology and director of the Quantitative Biosciences Graduate Program at Georgia Tech.

In response to broad interest in using DIRT in the field, Bucksch and his colleagues approached iPlant to host the platform and make it freely available to researchers anywhere. Through iPlant, a user’s data are secure, so that only the account owner and their collaborators may see the results. This makes DIRT a secure, open-access, time saving tool for botanists everywhere.

“Field-based measurements are vital for quantifying traits,” said Nirav Merchant, co-principal investigator of the iPlant Collaborative at the University of Arizona’s BIO5 Institute, and director of Biocomputing at Arizona Research Laboratories. “Automating these methods is essential to support the high throughput nature of analysis. For scientists in the field, DIRT elegantly facilitates the analysis and management of image-based phenotype data by connecting them with scalable cyberinfrastructure and their global community of collaborators.”

Foreground: bean plants. Background: maize fields. Backdrop: Chiracahua mountains of Southern Arizona. When analyzing the specific phenotype traits of this many individual plants, Bucksch said, researchers need high throughput computing capability.

“Manually,” Bucksch said, “you cannot measure many dozens of traits in just five minutes, as you can with DIRT.”

The program currently works for nearly all plant root systems. “There are certain traits that only work on monocots or dicots, and we are currently exploring more about this,” Bucksch said, explaining that slight differences in the algorithms account for variation in the plant species.

Already the platform has a substantial user base, with several scientists regularly using it for their root measurements. “Undergraduate students are using DIRT, a Google group is providing user-to-user support, and at least one citizen scientist is currently using it,” said Bucksch.

"DIRT seems especially useful in my work because a plant is a lot like an iceberg: Most of it is totally hidden beneath the surface," said Tim Zebo, a recently retired electronics engineer turned hydroponics and aeroponics systems researcher. Dr. Zebo is analyzing the roots of plants grown in liquid nutrients because, in this era of rapid climate change and major droughts, those plants require less than 10 percent of the water needed for soil-grown plants.  He plans to use DIRT to better understand root system architectures to increase production and reduce time to harvest.

For plant scientists and breeders, Bucksch said, root traits are key. “The root is important to nutrient uptake,” he said, continuing “and understanding how environmental and growth factors influence root traits is vital to developing crops capable of surviving climate change.

Increased food availability and resilience is necessary to accommodate accommodate rapidly increasing global populations. Plant scientists must work together to understand how plant traits – including root structure and function – affect crop survivability and adaptability. Bucksch and co-authors have developed DIRT with this objective in mind and to enhance the science of root systems.

Their work was supported, in part, by grants from Georgia Tech, the National Science Foundation, the Howard G. Buffett Foundation and the Burroughs Wellcome Fund. Bucksch also thanks iPlant co-principal investigator Nirav Merchant and iPlant senior projects coordinator Martha Narro for their assistance integrating DIRT on iPlant infrastructure. Abhiram Das, a graduate student studying under Bucksch and Joshua Weitz at the Georgia Institute of Technology, led the effort to develop DIRT prior to receiving his doctorate in Bioinformatics earlier this year.