Abstract
Effective surveillance has long been recognised as a key factor in managing and mitigating wildlife disease. But what exactly does effective surveillance look like and how does it contribute to the protection of wildlife health and the ecosystem services that wildlife provide?
We illustrate a model for a wildlife health surveillance system with two components: a disease detection system and a ecosystem health prediction system. Such a system is proposed as more effective at detecting wildlife disease events, enabling rapid and effective response, and more effective at identifying 'at risk' ecosystems, enabling proactive remedial action and/or targeted surveillance.
A mixed-user input and sorted output platform for disease detection will harness a much broader stakeholder group than previous wildlife disease surveillance models used in Australia, including non-expert participants ('citizens'). To create a sustainable and effective system, a sociological approach is leading design of the system (rather than a 'biomedical' or 'biosecurity' approach). Additionally, machine learning will be integrated into the system to allow real time curation and notification of events and biosecurity-sensitive sorting of information for release to participants. The effectiveness of this system will be tested against Australia's existing system and pathological investigation of 'significant' and random events.
The disease detection system will be integrated, using machine learning and data mining, into a broader 'ecosystem health' system, which draws on diverse data including climate, catchment, ecological, human movement and activity and other predictors of disease to create a real time map of ecosystems at risk of disease emergence. Disease detection through our mixed-user platform will enable feedback to our ecosystem health system for its continual improvement through machine learning. Likewise, real time notification of users in 'at risk' ecosystems will enable more targeted surveillance and ground truth alignment with in silico predictions.
Wildlife health is more than just the absence of disease. By taking a truly multidisciplinary approach and by putting the broader community at the heart of our system we plan to mainstream the concept of wildlife health and to enable and empower communities to care for their environment in a strategic and effective manner.
We illustrate a model for a wildlife health surveillance system with two components: a disease detection system and a ecosystem health prediction system. Such a system is proposed as more effective at detecting wildlife disease events, enabling rapid and effective response, and more effective at identifying 'at risk' ecosystems, enabling proactive remedial action and/or targeted surveillance.
A mixed-user input and sorted output platform for disease detection will harness a much broader stakeholder group than previous wildlife disease surveillance models used in Australia, including non-expert participants ('citizens'). To create a sustainable and effective system, a sociological approach is leading design of the system (rather than a 'biomedical' or 'biosecurity' approach). Additionally, machine learning will be integrated into the system to allow real time curation and notification of events and biosecurity-sensitive sorting of information for release to participants. The effectiveness of this system will be tested against Australia's existing system and pathological investigation of 'significant' and random events.
The disease detection system will be integrated, using machine learning and data mining, into a broader 'ecosystem health' system, which draws on diverse data including climate, catchment, ecological, human movement and activity and other predictors of disease to create a real time map of ecosystems at risk of disease emergence. Disease detection through our mixed-user platform will enable feedback to our ecosystem health system for its continual improvement through machine learning. Likewise, real time notification of users in 'at risk' ecosystems will enable more targeted surveillance and ground truth alignment with in silico predictions.
Wildlife health is more than just the absence of disease. By taking a truly multidisciplinary approach and by putting the broader community at the heart of our system we plan to mainstream the concept of wildlife health and to enable and empower communities to care for their environment in a strategic and effective manner.
Original language | English |
---|---|
Publication status | Published - Sept 2019 |
Event | 2019 Wildlife Disease Association Australasia (WDA-A) Conference - Gumleaves Bush Holidays, Little Swanport, Australia Duration: 29 Sept 2019 → 04 Oct 2019 https://www.wildlifedisease.org/wda/CONFERENCES/AustralasianConference.aspx https://researchoutput.csu.edu.au/admin/files/37391151/2019_WDAA_Program_Abstract.pdf (abstracts) |
Conference
Conference | 2019 Wildlife Disease Association Australasia (WDA-A) Conference |
---|---|
Country/Territory | Australia |
City | Little Swanport |
Period | 29/09/19 → 04/10/19 |
Other | Full papers only made available to conference registrants. Abstracts (pub avail) in link below. |
Internet address |