New varieties driven by world-class biometrics

GroundCover Live and online, stay up to date with daily grains industry news online, click here to read more

Underpinning new grain variety development in Australia are software and analysis methods that would not have been possible without the GRDC’s investment in biometricians since 1988.

Photo of Professor Brian Cullis, Dr Arthur Gilmour and Dr David Butler (From left) Professor Brian Cullis, Dr Arthur Gilmour and Dr David Butler were part of the development team for the ASReml software that underpins the evaluation of new grain varieties in Australia and overseas.

PHOTOS: SAGI

Biometrics is evolving at an unprecedented rate as researchers seek to answer more complicated questions, and with the genomics revolution they are able to generate and capture datasets of a size unimaginable 20 years ago.

Biometrics is both an art and a science requiring creativity and imagination to solve problems as well as a sound grasp of mathematics and statistics.

Australia's biometricians played a major role in the development of the ASReml software used by breeding companies around the world to increase the efficiency of variety selection.

Long-term GRDC investment has helped to maintain biometric capacity and stability. Since 2008 this investment has come under the banner of Statistics in the Australian Grains Industry (SAGI).

Over this time, computing capacity has increased to the point where massive datasets, known as ‘big data’, can be collated, while behind the scenes biometricians have developed complex solutions to understand this data and contribute to better agricultural strategies.

Top software

Photo of Alison Smith

Dr Alison Smith, from the University of Wollongong.

One such leap forward resulted from a chance encounter between the University of Wollongong’s (UOW) Professor Brian Cullis and Professor Robin Thompson from Rothamsted Research, UK, at a conference in 1986.

Working with Dr Arthur Gilmour (VSN International, and formerly NSW Department of Primary Industries) and later Dr David Butler (Queensland Department of Agriculture and Fisheries and UOW) and Dr Beverley Gogel (University of Adelaide), they developed the world’s leading software for analysing genetic datasets (ASReml).

Today ASReml is ubiquitous in plant breeding in Australia and around the world, and is used by the National Variety Trials (NVT) and plant breeding companies such as Australian Grain Technologies (AGT), Bayer and Monsanto. For AGT it is its go-to program for data analysis because the company is confident the software can handle thousands of entries across multiple sites and years and all of the inherent variability that comes with running field trials.

In the grains industry, rates of genetic gain have improved substantially as a result of working closely with biometricians using ASReml to improve selection accuracy in analysing field experiments and multi-environment trials.

Future solutions

Photo of ASReml software team

The team behind the ASReml software: (from left) Professor Robin Thompson, Professor Brian Cullis and Dr Arthur Gilmour.

With a view to developing the skill sets of the future, the GRDC has been a strong supporter of postgraduate students in biometrics. Two who have made significant contributions to the grains industry are Dr Gogel and Dr Alison Smith (UOW).

Dr Smith recognised that simple variety trial analysis failed to consider that the environmental variation at each field site is not the same as at another, and that the results for one field site would not necessarily correlate with another.

By using factor analytic models to account for genotype by environment (GxE) interactions, Dr Smith has been able to develop statistical models that make it easier to interpret GxE for multi-environment trials (METs). This allows growers to base variety selection decisions on more accurate predictions of performance relevant to their local conditions, and is the basis for the PV-Plus model used in the NVT Long Term Yield app. 

Dr Gogel has a passion for communicating statistics and believes everyone is capable of understanding the statistical principles that underpin good science. She takes her show on the road, talking to grower groups and agronomists about using statistics in on-farm trials. Likewise, the international uptake of ASReml by more than 2000 users can be tracked back to her comprehensive user documentation.

More information:

Professor Brian Cullis, University of Wollongong,
02 4221 5641,
bcullis@uow.edu.au

Next:

Delivering better varieties with predictive analysis

Previous:

Finding solutions to our biggest challenges

GRDC Project Code UW00005

Region North, Overseas, South