Improving wheat yields through better heat tolerance

Overview
Wheat generates $8.8 billion annually (FY24, ABS), but faces potential losses of 14% due to heat. This threat is increasing. A four-year multi-party GRDC funded project will accelerate the development of heat-tolerant lines.

Why
Development of new heat tolerant lines has been slow and resource intensive. Emerging technologies like compact sensors on drones, machine learning, high-throughput physiological and genetic screening will deepen understanding of plant heat tolerance. The multi-party partnership will accelerate introgression into new breeding lines.

Outcomes
Phenotyping for the wheat heat tolerance project has resulted in:

• A large dataset of photosynthetic leaf traits with associated hyperspectral measurements and yield data.
• The identification of several novel genetic features contributing to wheat heat tolerance.
• A comprehensive machine learning model for delivering novel predictions of dark respiration from hyperspectral data.

Impact
Australian wheat breeders and pre-breeders can access critical tools (germplasm, selection tools and value propositions) to develop wheat varieties with greater tolerance of high temperature conditions.

Potential benefit of up to AUD$1.2 billion (based on targeting the 14% potential yield loss of AUD$8.8 billion farm gate value in FY24).

Wheat is Australia’s primary cereal crop and grain export, underpinning our sovereign food security and generating $8.8 billion in farm gate value during 2023-24 (ABS). However, a warming climate is generating hotter conditions and earlier, longer heat wave events during critical growth periods for our winter wheat crop – causing potential yield losses of up to 14%.

APPN is supporting a four-year Grains Research & Development Corporation (GRDC) project (ANU2304-001RTX) to identify the genetic basis of the physiological mechanisms that underpin heat tolerance in wheat.  Findings from the project will help accelerate the development of new heat-tolerant commercial wheat varieties for Australian growers. 

Australia is a world leader in the production of high quality, food-grade wheat that is widely prized for flour milling, bread making and noodle production. 

Wheat is therefore an important food staple within Australia and our most significant cereal export. Between 65 and 75% of Australia’s total wheat production is exported each year – accounting for 10 to 20% of global wheat trade and generating AUD$8.8 billion in farm gate value during the 2023-24 financial year. (Source: ABS, 2025)

However, almost all Australian wheat is grown as a rainfed winter crop – making production vulnerable to a warming climate. 

Climate change is expected to bring higher extreme temperatures to Australia’s wheat-growing regions, resulting in potential yield losses of up to 14% - a significant risk to national food security and agricultural productivity.

Past research has identified breeding lines that can maintain relatively high yields under hot conditions. However, translating these into commercial wheat varieties has been slow and resource intensive. Traditional screening methods rely on markers that poorly predict crop performance under future climate conditions, and the physiological mechanisms behind heat tolerance remain poorly understood. 

Emerging technologies like compact sensors on drones and machine learning, combined with high-throughput physiological and genetic screening, are delivering the data needed by Australia’s wheat breeders, while also deepening scientific understanding of the key mechanisms behind plant heat tolerance.

As part of the wheat heat project, detailed physiological measurements were taken over three seasons at four sites in NSW and WA across a panel of 200 diverse wheat lines. These lines were part of the germplasm used for the earlier GRDC project ‘Increasing wheat yield and yield stability through improved heat tolerance during grain filling’, where yield was assessed at 40 sites across Australia. 

The results from both projects will provide detailed physiological insights to support decision-making in future heat-resilient breeding programs and form the basis for ongoing breeding work to develop new high yielding, heat-tolerant wheat varieties for Australian growers. 

So far, phenotyping for the wheat heat tolerance project has resulted in:

• A large dataset of photosynthetic leaf traits with associated hyperspectral measurements and yield data.
• The identification of several novel genetic features contributing to wheat heat tolerance in an Australian context.
• A comprehensive machine learning model for delivering novel predictions of dark respiration from hyperspectral data.

By 2027, the wheat heat project – including APPN’s contributions – will deliver novel insights into heat-tolerant wheat germplasm and screening methodologies for use by breeders and pre-breeders.

Being able to link heat tolerant germplasm to detailed yield and physiology data will help facilitate the uptake and adoption of new wheat varieties. 

APPN’s activities will also support the path to market for future heat tolerant wheat varieties by enabling breeders and pre-breeders to develop germplasm with improved genetic diversity for heat tolerance, use new phenomarker selection tools to screen for heat tolerance more efficiently, and conduct accurate cost-benefit analysis in screening and target genomic selection.

This national project resulted in the discovery of several gene candidates with high potential for increased heat tolerance in new wheat varieties.

As a result of these discoveries, Australian wheat breeders and pre-breeders are expected to have access to critical tools (germplasm, selection tools and value propositions) to develop wheat varieties with greater tolerance of high temperature conditions.

Ultimately, this project targets a potential farm gate value of up to AUD$1.2 billion (based on AUD$8.8 billion 2023-24 farm gate value multiplied by 14% potential yield losses).

The direct impact of APPN’s participation in the project included:

• Access to specialised research infrastructure. The project required access to high-value phenotyping equipment including spectroradiometers, porometers and high-precision controlled environments. Without APPN, it would have been very difficult to source this specialist equipment to enable data collection at scale.
• Time and money savings. By supporting the deployment of robotic Q2 fluorophore systems for photosynthesis and respiration measurements, APPN was able to help screen more than a hundred plant samples per day with just three people. Traditional methods using gas exchange systems would have been much more work-intensive and achieved fewer than half of these measurements in the same timeframe (using the same number of people).
• Field phenotyping innovations. APPN played a key role in advancing field-based spectral data collection through development of new methodologies. These methods are now being applied in practice, with the resulting data used to train machine learning models that predict physiological traits—accelerating the pace and precision of future high-throughput screening efforts.
• Unprecedented data generation. APPN enabled the capture of an unprecedented amount of hyperspectral, photosynthesis, respiration and stomatal conductance data on more than 200 diverse wheat lines over three field seasons. This wealth of physiology data is unmatched by previous collaborative research efforts, such as the International Wheat Yield Partnership and will underpin future research on wheat stress tolerance.
• Student opportunities. The project supported four PhD students whose theses will cover different aspects of the work. Additionally, casual staff were involved in data collection, giving them hands-on experience of applied science.  
• Capacity building. The APPN team trained students in the correct use of the intricate spectroradiometers to collect leaf reflectance in the field.  This propagates lifelong knowledge for future crop researchers and their research projects. 

Further flow-on impacts include: 

• Enhanced methodologies. The project facilitated the development of several innovative high-throughput phenotyping methodologies, including the development of FieldDino – a novel approach to screening stomatal morphology in the field and the first ever use of the high-throughput Q2 system to quantify photosynthetic oxygen evolution.
• Enhanced capabilities. The project enabled APPN to enhance our capabilities in high-throughput field phenotyping, campaign logistics, project management and research campaign support. These capabilities will directly benefit future research projects that APPN is involved with.
   

• Enhanced opportunities. Our work in this project made APPN an attractive partner for another GRDC investment requiring similar field campaigns: ‘Determining source to sink relationships in canola and identifying exploitable genetic diversity’. This brought in $1.543 million in funding and additional in-kind support to APPN-ANU.

The wheat heat project involved the following project partners:

• Australian National University (ANU)
• University of Sydney (USyd)
• University of Western Australia (UWA)
• University of New England (UNE)
• InterGrain 

International partners included the University of Essex (UK), the University of Lancaster (UK) and the International Maize and Wheat Improvement Center (CIMMYT). Statistical and modelling support was provided by Analytics for the Australian Grains Industry (AAGI) and the Australian Institute for Machine Learning (AIML)

The project is led by Professor Owen Atkin, Director of the ANU Agrifood Innovation Institute and Group Leader at the Research School of Biology, ANU. 

Phenotypic research infrastructure and expertise was provided by the APPN Nodes at ANU, USyd and UWA. APPN support also included access to field phenotyping equipment such as LICOR-6800s, spectroradiometers and LI-600 porometers; protocol and methodology development for high-throughput hyperspectral measurements in the field using the new SVC and Naturaspec spectroradiometers; and use of APPN growth capsules for controlled environment trials.