Artificial Intelligence in Miscanthus plant breeding

Kate Waddams at Aberystwyth University is head of knowledge exchange for a project looking at speeding up Miscanthus plant breeding with AI to help farmers grow more resilient crops in the future. In this guest blog, she tells us more…

For generations plant breeders have conducted the long and arduous task of breeding selected plants with the desired traits together to create a more valuable plant. However, this has taken considerable time, waiting for each generation to grow and establish before breeding in the next generation.

The hope now is that Artificial Intelligence (AI) can speed this process up significantly and could be used to breed the most resilient, productive, and nutritious crops to suit an ever-changing climate. It would be incredible if from inputting genetic data into computer models, the desired phenotype could be predicted before sowing a single seed.

The Importance of Miscanthus

Miscanthus is an efficient bioenergy crop, mostly used for heat and power. It can grow up to 4 metres every year and on land that is considered unsuitable for food crops. With its low input, high output properties, its carbon negative production and positive effects on soil health, Miscanthus is now an appealing option for many UK farmers. However, it takes 2-3 years to reach maturity therefore if AI can predict phenotype then it can reduce the length of each breeding cycle.

Miscanthus AI Project

Researchers at Aberystwyth University are using Miscanthus as a model crop in which to advance AI-generated plant breeding. The overall goals of the project are to increase the speed at which physical improvements are made during plant breeding, improve the desired traits of the plant, and increase the robustness of using AI to achieve these.

As well as using miscanthus as a model to test AI approaches to plant breeding, Aberystwyth University also aims to select varieties of Miscanthus that best tackle climate change. Developing varieties that are selected for their high-yielding properties can go some way to achieving this.  

The majority of the UK’s renewable electricity uses imported plant biomass. Miscanthus can be grown locally and on land unsuitable for arable crops. It has the potential to sequester more carbon than annual crops and its utilisation as a biofuel can be carbon-negative providing a promising solution to reduce our dependence on fossil fuels.

Miscanthus AI is one of a suite of research projects investigating the role that Artificial Intelligence can play in the UK reaching net zero carbon emissions.

What does this mean for farmers?

If the AI technology in the plant breeding sector proves successful, then arable farmers could benefit massively. At the moment, the unpredictable climate makes arable farming a challenging business. However, if farmers are sowing the most resilient crops that produce the greatest yield, then they can overcome some of the uncertainties they face every year.

So, if AI can speed up plant breeding, then it has the potential to keep up with the changing climate. Theoretical plant breeding could also suggest the desired phenotypes that would result from various genetic combinations of crops.

These advances in technology could change the way we farm in the future.

For more information, please visit https://www.aber.ac.uk/en/ibers/crop-science-and-plant-breeding/industrial-crops/#projects.