ARTIFICIAL intelligence technology could help reliably predict the flystrike susceptibility of Australian sheep, according to a review that sets a five-year $10 million investment timeline toward a flystrike-resistant Australian flock.
The review identified that the Australian industry is still largely breeding sheep assuming that surgical and chemical means of reducing the risk of flystrike will always be available, while many consumer welfare concerns relate to blowfly infestations.
The authors of the review — ‘Strategic and novel approaches to reducing flystrike in sheep’ – that was commissioned by Meat & Livestock Australia, believe the industry should aim to prevent flystrike from occurring.
“A recommendation of this review is to focus the industries attention to a zero-flystrike target.”
The NeXtgen Agri team led by University of Sydney Emeritus Professor Herman Raadsma reviewed the current state of knowledge around flystrike and the developing technologies that might assist in the fight against the condition estimated to cost the Australian industry more than $170 million annually.
The authors listed three pillars of investment covering breeding sheep that are not susceptible to flystrike, providing farmers with information and knowledge to reduce flystrike to zero and promoting reliable insect control methods.
With predictions of milder winters and increased summer storm activity extending the fly season, the reviewers also recommended investment in using industry data with climate modelling to predict the likely impact of breech and body strike into the future.
Consumers will drive the cessation of surgical flystrike solutions
The reviewers said many of the welfare concerns that are likely to become the focus of consumer and activist pressure are directly related to flystrike.
“Reducing flystrike by non-chemical, non-surgical means has the triple impact of improving welfare, reducing chemical use and ensuring sustainable and ethical production.
“The goals are likely to be rewarded if not demanded by consumers,” the authors found.
“Whilst reducing the impact of flystrike has been the focus in industry, we believe that the target should be to prevent flystrike from occurring at all.
“We believe that with the appropriate application of current knowledge, combined with the deployment of new technologies in the pipeline, this is a realistic and attainable goal.”
The review found that significant investment has been made in understanding the sheep genetic factors that are important in reducing the susceptibility of sheep to flystrike.
“However, to date very little of this information has made it through to mainstream breeding programs.
“The industry is largely breeding sheep assuming that surgical and chemical means of reducing the risk of flystrike will always be available.”
The reviewers said the Australian wool industry has become reliant on a combination of effective management practices, opportunistic use of insecticides, and the use of suitable genotypes for the management of flystrike.
“It is the view by the authors in this report that the emphasis or need for additional/replacement blowfly control options will certainly increase in the future.
“Use of insecticides will decrease as consumer demands favour lower chemical residue contamination in wool, and blowflies are likely to develop genetic resistance to current insecticides,” they said.
“The use of surgical modifications as we know them today, will eventually cease on the grounds of animal welfare, driven by consumers.
“As such there will be a directional change in flystrike control methods as we currently see being used by industry.”
The review also found a lot of the knowledge around flystrike is limited to Merino sheep with very little information on non-Merino breeds available.
“Furthermore, there are also very few tools available to industry to breed towards lower flystrike in non-Merino breeds.
“It is essential for Meat and Livestock Australia to obtain better knowledge in this area.”
Artificial intelligence can help, but a vaccine….
The authors found there was enormous potential to deploy the latest artificial intelligence (AI) techniques to many of the problems and opportunities around flystrike susceptibility, detection and prevention.
But they cautioned against significant investment in the short-term in further understanding of the biology involved in the host-parasite interaction or heavy investment in the high-risk and high-cost areas of vaccines, genomics and gene-editing.
“We cannot see any likely biological pathway that will enable vaccination against flystrike in the time frame necessary for this investment portfolio.
“Despite developments in vaccine technology generally, the world has not seen vaccines against multi-cellular ecto-parasites which have high and lifelong efficacy (Stutzer et al., 2018).”
Australian Wool Innovation AWI has started a $2.5 million four-year research project to help develop a commercial vaccine to protect sheep from flystrike.
Use sheep genetics to move to zero flystrike
The review said the reduction of efficacy of fly chemicals posed a significant risk to sheep farming going forward. Their short-term (1-3 years) solutions included enhancing the awareness and availability of chemical resistance tests for flies, a chemical application audit and training module, improved identification and management of affected sheep, and the use of industry best5 practice champions.
The Nextgen team said it favoured a pragmatic approach of fully utilising the information that has been gained from the last century of research “and super-charging its application by the latest relevant technologies.”
The reviewers said they are highly in favour of the utilisation of quantitative genetics to achieve a phase shift in flystrike susceptibility in the Australian sheep population.
“We feel that this goal can be achieved quite rapidly with the identification and exploitation of resistant genotypes already available in industry.
“The recommendations from this review focus on moving towards an industry outcome where flystrike is reduced to zero,” they said.
The reviewers medium-term (3-8 years) flystrike solution recommendations focus on the development of genotypes and the exploitation of existing genotypes that are less susceptible to flystrike as “one of the few solutions that has serious long-term positive impacts on the industry.”
This included development of Australian Sheep Breeding Values for flystrike indicator traits, including for flystrike susceptibility, and incorporation into indexes for Merino, maternal and terminal sheep breeds.
The reviewers also recommended development of automated phenotype assessments with an image-based system to determine superior genotypes allowing genetic gain in flystrike indicator traits, enable early detection of sheep at risk of flystrike and identification of animals that would not require mulesing.
They also recommended a project to collect images recording the lifetime flystrike incidence in sheep to select animals less likely to be struck. This would enable greater selection precision than is possible with the current scoring of indicator traits.
“To unleash the full predictive power of artificial intelligence is perhaps the greatest opportunity of our generation to make a difference to flystrike in sheep.
“The data collection required for this project could be overlayed with other activities in the portfolio.”
The reviewers also recommended that a central progeny test site dedicated to gathering information on the indicator traits for flystrike as well as allowing the expression of flystrike itself would provide a very impactful resource for industry.
“This site could incorporate testing for resistance to worms as well.
“The site should be open to all woolled breeds of sheep,” they said.
“This work could form the basis for artificial intelligence training sets and consolidation of genetic parameters of traits linked to flystrike susceptibility in industry flocks.”
The Australian sheep industry is ideally placed to evaluate the impact of such technologies on profitability and advanced farming systems. In particular the prediction of flystrike susceptibility and occurrence stand to benefit from a significant investment in AI technologies.
AI could be applied to advanced prediction of current phenotypes for breech strike susceptibility using digital information capture, and breech and body strike susceptibility, using digital image capture. It could also be used to monitor fly abundance through trap data, or model abundance and high-density refuge based on topographical /vegetation data on farm and to develop predictive regional fly wave models.
The reviewers also said artificial intelligence could also be the gateway to monitor behavioural profiles of sheep through sensor-based alerts to distinguish fly-struck and non-struck animals.
Click here to read the full review report.