Photosynthesis is the trait now being targeted globally to deliver future yield gains in bread wheat. Crucial to these efforts is the ability to measure and compare (phenotype) photosynthesis efficiency among the thousands of lines needed in breeding programs.

The work is building on an earlier Australian development, an analytical light tool called hyperspectral reflectance (HSR). This became the basis of a fast, high-throughput, paddock-deployable tool which can measure several key aspects in photosynthesis biochemistry in seconds.

Now, researchers working within the International Wheat Yield Partnership (IWYP) have used HSR to screen mapping populations that allow the detection of DNA markers associated with photosynthesis traits.

A key research focus was a population derived by crossing Seri and Babax cultivars, previously used to study drought-adaptive traits.

In this new project, HSR has proved to be a robust phenotyping tool for plants grown both in high-throughput glasshouses and in the field in Australia and at the International Maize and Wheat Improvement Centre (CIMMYT) in Mexico.

In parallel to this effort, UK-based researchers accepted the challenge to map differences in photosynthesis traits at the genetic level.

A new tool

This IWYP project, "Using Next Generation Genetic Approaches to Exploit Phenotypic Variation in Photosynthetic Efficiency to Increase Wheat Yield (IWYP64)", is headed by Professor Anthony Hall of the Earlham Institute in the UK and also involves researchers in Australia and Mexico.

The hyper-spectral reflectence device that scans light reflected from wheat leaves. Artificial intelligence then makes it possible to extract important information about leaf physiology, including photosynthetic capacity and the plant's nitrogen content. Photo: Professor Robert Furbank.

Among them are professors John Evans and Robert Furbank of the Australian National University, both of whom are world-leading specialists in the study of crop photosynthesis.

Professor Evans says photosynthesis is such a new target for breeding that foundational work was necessary to develop the ability to screen and select for photosynthesis-related traits.

"The key thing here is working up fast and reliable phenotyping abilities to measure traits in the field and, at the same time, to map the surveyed traits genetically," he says.

"This is such a new area that we need to actually demonstrate that there are genetic associations for the traits we are measuring with HSR."

While work continues on the genetic characterisation, the projects did find that HSR technology is a viable, robust and useful tool to phenotype wheat at the scale necessary to breed for greater radiation use efficiency, even when used by different operators in different continents.

"This project saw the maturation of an important new phenotyping technology," Professor Evans says.

"We now have a new tool to assess and compare wheat germplasm for diversity in photosynthesis traits."

A new trait

Phenotyping with HSR involves clipping a light-emitting sensor to a leaf. The light reflected from the leaf surface is analysed by a spectrometer, which quantifies the amount of reflected ultraviolet, visible, near infrared and shortwave infrared light.

This provides information at 2000 wavelengths rather than just two, as occurs with simpler reflectance instruments, such as the Greenseeker unit used for herbicide or fertiliser application.

Using this data, advanced computer analysis - increasingly referred to these days as artificial intelligence or AI - is used to model characteristics of the key enzyme in photosynthesis, Rubisco. These traits include:

• the amount of Rubisco in the leaf;
• nitrogen and dry mass per unit of leaf area; and
• the amount of Rubisco relative to the nitrogen content of the leaf.

The AI was 'trained' to model these traits by first providing HSR data along with information about Rubisco obtained with physiological tests.

"It's a trivial solution to achieve 20 per cent more Rubisco in the leaf if the increase is due to 20 per cent more nitrogen in the leaf," Professor Evans explains. "That's equivalent to simply applying more fertiliser to the crop.

"But we want 20 per cent more Rubisco enzyme activity - in the amount of carbon dioxide fixed - for the same amount of nitrogen in the leaf. That is what leads to gains in efficiency in the use of that nutrient resource."

As such, the focus is on improving resource efficiency: getting more carbon fixed per unit of nitrogen present in the leaf.

Ultimately, the researchers don't need an overly large increase to make a significant impact on crop productivity. This is due to quirks in photosynthesis biochemistry that allow gains to accumulate ... much like compound interest.

"The benefits can be in different forms," Professor Evans says.

"We could improve yield generally, or lift yield in proportion to the amount of nitrogen applied, or water availability."

The goal once the genetic analysis is completed, is to have DNA markers that can discern differences in this resource efficiency trait.

Future directions

HSR was originally developed by Dr Viridiana Silva Perez during a PhD project co-supervised by Professor Evans, Professor Furbank and CSIRO's Dr Tony Condon.

In the next phase, the researchers want to further progress this technology, especially by running the sensors above a canopy using a drone. This would allow for even faster data capture.

Professor Evans says the field of remote sensing is evolving rapidly and advances include the option to add other imaging instruments to boost HSR's analytical capacity. Included is the use of thermal cameras to measure canopy temperature and quantify transpiration (since open stomata allow for faster photosynthesis rates and more carbon dioxide uptake).

Professor Furbank says they are also leveraging off other IWYP projects such as the Australian-led Energy Use Efficiency project (IWYP 60) and GRDC-funded work at CSIRO led by Dr Condon (CSP 168) that use HSR technology.

This has provided data from other populations - such as a cross between two elite Australian cultivars, Kukri (PBR) and Excalibur (PBR) - and additional measurements at different stages of crop development across diverse environments.

In a promising finding, phenotypic data from plants grown at the Plant Accelerator in Adelaide during 2018 indicates that the Kukri-Excalibur population may provide better contrast between high and low-performing Rubisco traits than the Seri-Babax population.

"Basically all this foundational work is providing the tools that are needed to run fast and efficient breeding programs in which photosynthesis is the basis for yield gains," Professor Evans says.

"The ultimate goal is to put a bigger photosynthesis engine in the leaf while ensuring the plant has a suitable throttle to extract optimal productivity gains."

More information: John Evans, john.evans@anu.edu.au