Taking stock on farming decisions: Meet Tom Keogh

By  Emily Lehmann 4 June 2024 2 min read

Despite growing up in urban Sydney, post-doctoral research fellow Tom Keogh has always had a strong connection to farming. Tom’s dad worked in policy research for agriculture and his relatives owned a sheep farm near Holbrook, in southern New South Wales.

A gap year spent out in the field seeded his journey into agricultural research. Tom likes crunching numbers so, when on the family farm, he found himself trying to solve problems using data.

"I learnt that pasture is not cheap when you consider the price of land. In fact, any sort of livestock feed costs a lot," Tom said.  

"It got me thinking about how you could better utilise the pasture supply by providing supplements to try and maximise lamb growth.

"There are so many ideas out there around how you might go about optimising lamb growth and so I wanted to analyse the data to find out how they stack up."

Tom went on to do his PhD, where he used lamb growth models to help answer this question. He looked at a range of scenarios and confirmed there’s no silver bullet to optimise animal growth. For animals to grow faster, they need to eat more. It’s as simple as that. But not every farming decision is so straightforward.

Data to inform destocking decisions during drought 

Tom is now building on this work by using models to inform new decision-making tools to help farmers during a drought.  For example, when graziers face feed shortages during dry conditions or a drought, they need to match livestock numbers with available feed. This can result in the difficult decision to destock a herd.

"Destocking involves making decisions around which animals to keep, sell or cull," Tom said. 

"Often these can be emotional decisions based on past experience or intuition, rather than data. Yet these decisions can have big consequences on future productivity when you look at the outputs from various modelled scenarios."

Tom is working on stock ranking tools to help farmers destock in a way that allows them to thrive when the drought breaks. 

“If we can make a simple, science-backed tool to guide the destocking process, then that could be really useful,” Tom said.

"For instance, a farmer may choose to keep animals that have a genetic advantage, which will help improve their farm’s longer-term profitability. 

"Our team is looking at what those attributes are and a way to rank them. Specifically, I’m exploring the usefulness of different models, including their strengths and weaknesses, to see how we can apply them to this problem."

Building our modelling expertise 

When Tom can’t find what he’s looking for in existing models, he uses data to change the equations to try and better reflect reality. 

"What I’ve found is that most existing livestock systems model focus on the average. For example, they may model climate variations over a 30-year period and look at the most profitable outcome," Tom said. 

"But we often get really different results when we focus on poor or good years.”

Tom is applying this thinking to analyse scenarios that could benefit the industry in both below and above average rainfall years. 

“Farming systems are complex, and they differ considerably across environments,” Tom said. 

“I’m trying to simplify destocking decisions using models so that we can assist producers all around Australia.”   

Tom’s research is part of a partnership with the University of New England and CQUniversity with funding from the Australian Government’s Future Drought Fund.

Our work on a destocking tool is contributing to our Drought Resilience Mission’s goal to reduce the impacts of drought.