Weed-detection dogs on trial in Australian paddocks

SUBSCRIBE to our fortnightly email newsletter to receive more stories like this. Fikra Gedreamlak, an UNMAS Mine Detection Dog Handler, brings in his dog to check an area just cleared using radar and a metal detector: Now, Dr Cindy Hauser is trialling the use of trained detection dogs in weed control.

While we are all familiar with the valuable work trained detection dogs do at airports and in war zones, Dr Cindy Hauser, a research fellow at the University of Melbourne, is taking canine olfactory powers a step further, training detection dogs to sniff out pests and diseases in commercial crops, and invasive weeds in bushland.

Dr Hauser, the recipient of a 2016 Science and Innovation Award for Young People in Agriculture, Fisheries and Forestry, is using the $11,000 bursary from CSIRO’s division of Health and Biosecurity LINK to further her innovative sniffer-dog project.

“I want to develop guidelines to help work out how we can make the best of them in agricultural environments,” she says.

Dr Cindy Hauser won a Science and Innovation Award for her work training detection dogs for use in weed control.
Dr Cindy Hauser won a Science and Innovation Award for her work training detection dogs for use in weed control.
University of Melbourne

Hounds for hawkweed

Dr Hauser’s current focus is teaching dogs to sniff out a plant originally native to Eastern Europe, the noxious hawkweed.

Research on hawkweed – a flowering plant of the Hieracium genus, akin to asters or daisies, of which there are hundreds of species but only four, so far, in Australia – has found that humans have a poor record at detecting it.

Indeed, an expenditure of an average 25 hours of human search time per hectare has resulted in only a 20 percent success rate detecting this weed.

With this poor a track record, there’s a clear need for a more accurate and cost-effective hawkweed detection tool – and Dr Hauser thinks she may have found it in our canine companions’ noses.

Her initial research achieved great success: she was able to train a dog named Missy to distinguish three different species of hawkweed from various other plants in backyard conditions. Missy’s success suggests that trained dogs have the potential to be better than human search teams at detecting non-flowering plants amid crops and bushland.

The CSIRO-backed Science and Innovation award has allowed Dr Hauser to train two more dogs, whose weed-detection training regimes have benefitted from what was learnt from Missy’s pioneering efforts.

“I will compare their efforts to what we already know human searchers can achieve,” Dr Hauser says.

Orange hawkweed roots: hawkweeds are in the daisy family and are similar to dandelions.
Orange hawkweed roots: hawkweeds, part of the daisy family, are similar to dandelions.
John Hoskin, NSW Department of Primary Industries (DPI)

 

Smarter use of sniffer dogs

All this sounds great – but it’s important not to overplay even the best-trained dogs’ capabilities, Dr Hauser cautions.

“I'm really excited in the potential of detection dogs but I think we need to be smart about how we use them,” she says.

Whether a dog’s role is to detect drugs, invasive weeds, insects or pest-animal droppings, the first phase of training is much the same; however, expecting dogs to be able to detect and pinpoint the same ‘trained’ scent outdoors, especially in the wild, can be far more variable and complex a process.

Training dogs to detect insect and animal pests, invasive plants and pathogens in outdoor environs – even the relatively ‘tame’ confines of a paddock – presents problems that don’t exist in indoor areas, such as airports.

“When you get into the outdoors, there are all sorts of other distractions that the dog is faced with … the smells of other plants and animals, changing terrain, and shifting winds that carry the weed scent,” Dr Hauser explains.

Drug detection dog in action, sniffing passenger baggage at an airport - in outdoor enviironments, there are many additional challenges for detection dogs, says Dr Hauser.
Drug detection dog in action, sniffing passenger baggage at an airport - in outdoor enviironments, there are many additional challenges for detection dogs, says Dr Hauser.
Richy, Flickr CC

When you’re detecting weeds in bushland, there are additional issues. “When we're trying to cover huge areas of agricultural land or national park, we also need protocols to make sure a dog covers that area thoroughly,” she says.

Large tracts of outdoor vegetation, such as native bushland, present an array of challenges for detection dogs. They may be filled with distracting and tantalising odours. A change in wind can move scents around rapidly at a moment’s notice. Thick vegetation can be difficult to get through and may even trap scents away from the open air.

Dr Hauser’s research identifies what can be detected and what is likely to be missed when we go out into the field using ‘observational tools’, whether these be human eyes, DNA testing of water samples, motion-triggered cameras or, in this case, dogs’ noses.

“We need to understand what they can do well and what they, too, might be missing," she says.

The ultimate solution might be a multi-pronged one, using humans, dogs and drones.

Until that happens, Dr Hauser will continue her work with dogs, sharing her mathematical ‘detection models’ with government agencies to help them avoid missing what matters – including the noxious hawkweed, the spread of which could cost Australian agriculture a whopping one billion dollars.

Her work training weed-detection dogs is an important part of the wider effort to achieve better results for an agricultural sector that wages an ongoing, often costly, battle with invasive and noxious weeds.

About CSIRO Health and Biosecurity

With global trade and greater connections, Australia is subject to increasing biosecurity threats. Once they break through our biosecurity net, diseases, pests, invasive animals and plants can inflict massive damage to crops, livestock, farm profits, our unique natural environment and occasionally, human health.

CSIRO’s division of Health and Biosecurity assembles multidisciplinary research teams under the banner of one-Health to achieve “optimal health for people, animals, plants, the economy and the environment”; tackle biosecurity challenges; and “provide sustainable management strategies”, exploring new technologies for detection, surveillance, diagnosis and response.

The division’s overarching aim is to develop a biosecurity system for Australia that’s “pre-emptive, responsive, resilient, and based on cutting-edge surveillance, informatics and new technologies for integrated response”.

SUBSCRIBE to our fortnightly email newsletter to receive more stories like this.

Comments

Error | AgInnovators

Error

The website encountered an unexpected error. Please try again later.