Research teams tackle the Great Barrier Reef’s second thorniest problem

As the crow flies, Minnesota lies 13,836 kilometres and half a world away from Brisbane.

But Dr Mary Bonin, who braved below-freezing temperatures every winter growing up near the lakes and prairies of the United States’ twelfth largest state, always knew she’d end up working on Queensland’s Great Barrier Reef.

“I was always that nerdy biology kid who had whale posters on the walls and was fascinated by coral reefs,” Bonin says.

That fascination led her to her current role with the not-for-profit Great Barrier Reef Foundation (GBRF), tackling one of the reef’s most serious threats next to climate change – the crown-of-thorns starfish (COTS).

“While they’re native to the Great Barrier Reef and many reefs across the world, they’re coral-eating machines which are one of the biggest drivers of coral loss over the last 40 years,” she says.

Due to their high reproductive potential – a single female starfish could release more than 100 million eggs in a breeding season – there are periodic outbreaks of COTS which can quickly strip a reef of 90% of its living coral tissue.

But just like cockroaches, rats, or other pests, the presence of the predatory starfish might not be apparent at first glance.

“If you’re out on the reef, they might be in a hole, or you might see just an arm sticking out,” says Bonin. “When you see the first one, you know there’s a lot more where that came from.”

As director of the GBRF’s COTS Control Innovation Program (CCIP) Bonin’s mission is to coordinate the efforts of more than 90 multidisciplinary experts from Australia’s leading science institutions as they seek to control COTS.

Their work has been given a $9.8 million boost by the partnership between the Australian Government’s Reef Trust and the GBRF.

The shot in the arm has allowed the GBRF to partner with the Australian Institute of Marine Science (AIMS), CSIRO, James Cook University and the University of Queensland to develop next-generation tools that will enhance management capacity to predict, detect and respond to deadly COTS outbreaks at large scale across the Great Barrier Reef.

Robotics, genetics, environmental DNA, pheromones and natural predators are among 21 new projects scientists are working on as part of the CCIP.

Funding controversy

Building the reef’s resilience is a high priority as our planet warms.

But the decision by the Turnbull government in 2018 to grant the GBRF $443.3 million, unleashed huge criticism.

The grant was part of $535.8 million in funding over five years to accelerate the delivery of activities set out in the Australian and Queensland Government’s Reef 2050 Plan.

Critics argued the grant should have been made through a competitive tender process or with consideration to other government agencies operating in this area.

“They’re coral-eating machines which are one of the biggest drivers of coral loss over the last 40 years.”

Dr Mary Bonin

However, during an interview conducted last year, Anna Marsden, the Managing Director of the GBRF, pointed out that the sheer size of the grant made the development of an overarching reef conservation strategy possible for the first time.

Previously, funds had been allocated to disparate projects without considering overlaps or synergies.

Having an integrated portfolio was thus something “very powerful and new” and was a catalyst for all the different players working in reef protection to come together, she said.

This collaborative spirit has carried through to the CCIP.

“We went through a pretty intensive design process with this program, so it was not your standard open call for proposals,” says Bonin.

“We wanted to break down those institutional silos, to ensure the research community working on this program could maximise their collective impact.”

Often, researchers send grant applications in, to receive a ‘yes’ or ‘no’ response, absent of any further dialogue.

The CCIP proceeded differently.

Over a six to eight-month long process, scientists from partner organisations worked together to identify key COTS knowledge and capability gaps, whittling down 52 promising possibilities to a more manageable research portfolio of 21 projects.

Research findings are helping hands-on crews manage COTS in the water.

The Great Barrier Reef Marine Park Authority’s Crown-of-thorns Starfish Control Program was established in 2012, and now has seven vessels and more than 100 people out on the reef, manually culling starfish using a lethal injection of either bile salts or household vinegar.

“We work closely with the scientists to feed information and knowledge and capability improvement into the COTS Control Program on an ongoing basis,” Bonin says.

“That kind of ongoing space and opportunity for regular routine dialogue across the scientists and managers – because we’re working as one big team – is what makes the approach special.”

Researchers who normally work from a laboratory, or behind a desktop, are also going out on COTS control vessels to learn from those who are hunting the starfish every day.

“With that deeper understanding of the people their science is serving, they can go back and iterate on their models, because they have that deeper, real-world understanding,” Bonin says.

“Testing ideas and knowledge with the people who are doing the work is, as a scientist, really rewarding and breaks the mould of where you work in your little bubble, complete your project and then present a final product to the world.”

The COTS Surveillance System is one of four projects which falls under the detection subprogram.

In a collaboration between AIMS and CSIRO, that project will this month trial a towed platform developed by AIMS.

The platform will be equipped with a camera which will take a continuous series of underwater images at four frames per second that can be analysed by machine learning models in real time.

Researchers are keen to see how this novel method of identifying COTS compares to traditional reef survey techniques such as manta tows, where a snorkeller being towed behind a boat makes regular observations.

It’s anticipated that the use of artificial intelligence rather than manual labour will revolutionise COTS mapping says Scott Bainbridge, leader of the technology development engineering team at AIMS.

This is crucial when you weigh factors such as human fatigue and distractibility against the 348,700 square kilometres that the Great Barrier Reef spans.

“The main problem we have with all of our monitoring [to date] is that it can’t scale,” Bainbridge says. “Changing the way we collect the data and analyse it is the game changer.”

Machine learning models which can flag the presence of COTS also translate well to other contexts says Dr Brano Kusy, principal research scientist and group leader of the distributed sensing systems with CSIRO.

Kusy has used machine learning to classify cattle behaviour, track fish movements, and detect the presence of different type of seagrass.

While he and his team have brought the machine learning expertise to the COTS project, he points out that AIMS has a proven track record building the electronics that are used in the water.

“We get to apply all of the science that we’re building in house … and then we can sit back and watch our software work on someone else’s hardware,” he says.

Bainbridge and Kusy agree that collaboration is crucial.

“These are big problems that we’re trying to deal with and they need to be [tackled] now – it’s not something where we can say, ‘I’ll come back to this in 20 years’ time’,” says Bainbridge.

“We’ve actually got to think differently in terms of the way the organisations work together by saying, ‘Okay, this is what we’ve got to do to get the job done.’ So, yes, we could do it individually. But we’re so much more effective when we work together.”