Basel, Switzerland
May 2013
Dan McDonald is president and co-founder of the Phenotype Screening Corporation, located in Knoxville, Tennessee. After nearly 30 years developing scientific instrumentation at the Oak Ridge National Laboratory, he founded the Phenotype Screening Corporation together with Ronald Michaels in 2004. Their idea was to connect technology with humanity and to help challenge the frontier of feeding the world. They realized that the underground part of plants was probably highly related to yield, but also that very little was understood about that relationship due to the lack of technology that could measure and analyze the root system, especially in a nondestructive way.
We asked Dan Mc Donald what technology his company uses to overcome the difficulties of analyzing roots
The technology we use is soft-tissue X-ray imaging. This technology allows analyzing entire root systems in vivo when grown in an X-ray invisible medium. For this purpose, the plants are grown in Expanded Polystyrene, EPS, beads which also allow highly controlled water and nutrient management. The sizes of the containers in which we grow the plants are up to 1m deep. We use the same X-ray spectrum to analyze sampled roots from the field. The benefit of the soft X-ray image is that you have all the roots – even those that are hidden by others – in one image. For the analyses of those images, we developed special software that allows analyzing the entire root volume from the X-ray image and puts the characteristics of the root system into numbers. The smallest size that this technology can measured is 60 µm. This system allows us to study also root-pathogen interaction and to quantify, for example, certain nematodes within the root system, as well as their effect on root architecture.
What was the biggest challenge in developing this technology?
In the first projects, we were challenged a lot by the biologists with whom we worked because they felt that it was all very nice to be able to capture the characteristics of a root visually but wanted to have quantification, preferably, one single number to summarize the entire root system. It took us two years of software development to meet their basic quantification needs. In the end, we decided to provide three fundamental numbers: The total root length, the total projected root area and the total number of root transect crossings. While these are powerful summary values they can become even more useful when used as signatures. So we now provide a total root length signature, a total projected area signature and a total root crossing signature. These values are used to summarize the characteristics of a root system’s architecture:
- The total root length signature. The total root length value sums the length of all root segments, ignoring their thickness, and puts them together as one single length. We independently determine the total root length of seven different root size ranges. These seven numbers then become our total root length signature.
- The projected root area signature. That is the sum of the area of each root segment determined from the X-ray image and broken up into the root size ranges. These ranges can be customized for a specific crop. The projected root area combines the length and width of the root segments making up the root system.
- The total root crossings signature. For this value, the root system is analyzed in depths spaced in increments of 25 mm and for each depth the roots are counted. With this value you can get an idea of the root profile (i.e. does the plant have more roots in the middle of its root system than in deeper areas). We also determine a value for each root size range and provide a signature.
What is the challenge when analyzing root samples from the field?
For the analysis of roots from the field, we receive washed root systems shipped to us from field trials across the USA. We developed a special protocol for the shipment in order to keep the root systems intact during the shipment, but of course the sampling depends a lot on the soil type and on the plant. You might not be able to sample the entire root system because it extends too wide or too deep. One method that can be used in that case is to define a volume of soil that is sampled and to normalize your results for that volume. Another problem is maintaining the integrity of smaller root segments. It may be easy in sandy soils, but for other soils fine root segments often break off with the soil during washing. Many fine roots also desiccate quickly in air and it is imperative to wash your plant immediately after extraction (don’t let them sit exposed in the sun) and then to seal your root system sample as quickly as possible. One difference that we found between in situ analysis and field samples is that there are many fine horizontal roots developed by a plant in situ that collapse against the big roots when you are washing the plants, becoming difficult to distinguish. Through all of the projects we have been involved in over the years, we realized that root research is always a research of compromises which makes it a very exciting, challenging and a new frontier in agronomics.
What have you learned about roots
We have worked with many different crops over the years (corn, soy, wheat, cotton, rice, tomato, peppers, okra, squash, potato, prairie grasses and even trees) and analyzed about 250,000 images of root scans by now. What we have learned from it is that not only the plant species that have their certain characteristics with respect to their root system, but also the varieties, which might also perform differently in different environments. Precision farming is a new trend in agriculture and I could envisage that in the future a farmer would select a specific variety with specific root architecture for a given field location. However, although there is a growing interest in understanding the root architecture and its effects on yield, we are still far from understanding the root system. One breakthrough study with respect to corn root architecture was the paper by Graeme Hammer published in Crop Science (2009) where he correlated the yield increase in corn over the past 40 years with a change in root architecture. By selecting for a higher yield, somehow a deeper and narrower root system was selected as well allowing more plants to be grown per acre. However, this was a retrospective study. The breeders did not actively look for this trait at that time.
Phaenotype Screening Corporation:
http://www.phenotypescreening.com/
Article by Professor Graeme Hammer
https://www.crops.org/publications/cs/abstracts/49/1/299
Dan McDonald was interviewed by Dr. Melanie Goll, Network Community Manager for Syngenta.
Melanie moderates the Global Root Health Network, a group hosted by Syngenta on LinkedIn® professional networking services. This group brings together leading technical experts from private, public and nonprofit institutions worldwide to exchange knowledge about Root Health and the link to plant performance.
For more information, please contact melanie.goll@syngenta.com or explore the group on LinkedIn at http://www.linkedin.com/groups?gid=3773669
Species
