All-seeing ‘tractor’ takes crop science to the future

5 Aug 2016
Remote-seeing tractor developed by QAAFI that can drive over trial plots during different growth stages and compile a composite ‘picture’ from the visible to the infrared light spectrums

There is not much left to be known about a new sorghum plant ‘on trial’ after being straddled by the Gekko phenotyping tractor – an all-seeing mobile platform bristling with lenses and sensors probing for the visible and the invisible secrets of crop performance.

The tractor, which was developed by QAAFI, is a remote-sensing vehicle that can drive over trial plots during different growth stages and compile a composite ‘picture’ from the visible to the infrared light spectrums, from thermal and chemical imaging, and laser reflectance (or Lidar as it is known) for high-definition 3D profiles of the plant canopy.
This data can then be correlated with a crop’s production performance (for example, grain yield) by identifying links with these measured properties – such as plant structure, biochemical and physiological characteristics, photosynthesis efficiency – and the response overall to environmental conditions such as limited water or high temperatures.
The objective is to better understand why a particular variety (genotype) performs a certain way, then address identifiable causes for underperformance that can be rectified or isolated from other data so a genotype that might otherwise have potential is not simply discarded on the rudimentary evidence of grain yield alone.
For example, explains QAAFI’s Director for the Centre of Plant Science Professor Graeme Hammer, for a genotype that yields less than others would under the usual selection process be thrown out.
“But the data collected by the phenotyping tractor may tell us this particular genotype is growing more leaf area, so it runs out of water faster, gets hotter quicker, and that’s why it yields less.”
“So what if we take that leaf area influence away from the results? We might find we have a genotype of significant potential in other trait areas: possessing genetic potential that under current screening practises could be lost.”
Professor Hammer is a node leader with the Australian Research Council (ARC) Centre of Excellence for Translational Photosynthesis, and a specialist in crop modelling. He co-leads the phenotyping tractor project with the head of QAAFI’s sorghum breeding program and co-program leader in the ARC Centre of Excellence for Translational Photosynthesis, geneticist Professor David Jordan. QAAFI’s sorghum breeding program is a collaborative program with the Queensland Department of Agriculture and Fisheries and is supported by the Grains Research Development Corporation (GRDC).
Professor Hammer’s job is to run the data collected by the phenotyping tractor through analytical models to identify ‘confounding’ factors and stop them from obscuring the identification and potential use of other genetic gains that are bound to exist in the tens of thousands of genotypes that are screened.
Professor Jordan’s work is to try to find the genetic regions controlling variation in traits measured by the tractor. “What is exciting is that this equipment allows us to see how the crop develops over time because the tractor’s sensors allow us to measure non-destructively and regularly so we can see not just the end point, such as leaf area at maturity, but also how a plant gets to a certain end point. For this research we can progressively log the interactions that are determining, for example, a leaf’s final dimensions. It is important to know if leaf area, or shape, is affecting photosynthesis, which is a key determinant of yield.
“In effect we are now able to deconstruct a phenotype into its parts and from this learn how, through breeding or management, we can change the way a plant develops to optimise its use of land, light, water and nutrients.”
However, before then Professor Jordan says the challenge is to make sense of the vast amount of data coming in, which is why it needs to be a multidisciplinary effort: “My team is on the genetics, Graeme’s team is on the crop physiology and modelling, and we’re also working with the photosynthesis group at the Australian National University in Canberra.
“An example of how the physiological data collected by the phenotyping tractor and genetics research will come together could be, say, measuring leaf temperature at a particular crop growth stage. If we can find the gene associations with that growth stage we can delve into the species natural gene diversity to find even ‘better’ performing versions of those genes. In this way we can incrementally improve or optimise the different plant processes that contribute to a particular goal such as higher yield,” Professor Jordan explains.
“We are looking for the genetic changes we can make to improve photosynthesis, which drives productivity.”
Professor Hammer says the tractor can cover thousands of plots relatively quickly and take genotype selection to a whole new level of efficiency: “My team’s job is to use our crop-modelling expertise to make this massive amount of data useful for David’s team,” he says.
“The data has to be able to connect to selection strategies in the breeding program, from which David will package superior germplasm for use by commercial breeders.”
Sorghum is Queensland’s most valuable cereal crop, worth an estimated $432 per year. The phenotyping tractor, built onto the frame of a modified self-propelled sprayer, represents a $500,000 investment by UQ. Professor Jordan says that the tractor will also be applicable to other crops, including barley, wheat and pulses.
Sorghum is the fifth most important cereal worldwide and a critical food security crop in parts of Africa and Asia. The sorghum research in Queensland will directly contribute to lifting the crop’s climate resilience.
This project was funded by UQ and is being delivered by scientists jointly supported by the DAF, Australian Research Council, GRDC and UQ.