We examined the following Data sets:
EnvvecFlyingFoxSurvey | EnvvecFlyingFoxSurvey | SCC Open Data (sunshinecoast.qld.gov.au)
On the Sunshine Coast Council Open Data Platform, a search for “FlyingFox” yields 3 results:
https://data.sunshinecoast.qld.gov.au/maps/envvecflyingfoxmanagementactions Layer representing management actions performed by Sunshine Coast Council at Flying Fox Roosts.
https://data.sunshinecoast.qld.gov.au/maps/envvecflyingfoxsurvey Publicly available Flying Fox Roost survey content for the Sunshine Coast Council area.
https://data.sunshinecoast.qld.gov.au/maps/envvecflyingfoxroosts Publicly available Flying Fox Roost location content for the Sunshine Coast Council area.
The City of Moreton Bay also has Flying Fox Observation Data https://www.moretonbay.qld.gov.au/Services/Environment/Local-Wildlife/Flying-Foxes/Observations
The individual flying fox and their habits:
We found the dataset inadequate to extrapolate data on habits of individual flying fox. We propose to introduce a Flying Fox identification program to track individual movements.
Initially, we proposed tracking mothers of juveniles by compiling photographic data of identifying wing prints taken at known roost sites across the Sunshine Coast Council area of mothers with young. This is a known method and can be found within existing literature.
This would expand the existing database with sightings of specific individuals. This will answer unknowns such as:
Do the flying foxes travel in the same groups?
Do the fling foxes return to the same locations year on year?
Are there any predictable or repetitive movements?
Answers to these and other questions can be added to the body of knowledge to assist Councils and other experts to educate the public to become more tolerant of flying fox populations as urban areas continue to expand around their habitats.
We were able to find an academic paper for identifying flying foxes by vein patterns on their wings. We found that the method of identifying individual flying foxes correctly identified 96% of the bats (flying foxes) (Amelon, S., Hooper, S.E., & Womack, K.M. ,2017).
In developing a unique solution, we then propose to use this data as a control set as we investigate sound as a method to track individual bat movements. As evidenced further below, one of the chief complaints from residents nearby Roosts is the noise. Can we record that noise at known Roost sites and, over time, compare those sounds using AI and Machine Learning from our control set of identified Flying Foxes to develop a more efficient method to track individual Flying Fox movements?
Educating the public
Sunshine Coast Council receive many calls from ratepayers whose quiet enjoyment of their properties is diminished due to the presence of flying foxes nearby.
Most complaints are from around a 30 to 100m radius of flying fox roosts.
The complaints are predominantly (in order of volume) about:
The vast majority are about noise and odour.
Most callers are seeking to move flying fox populations away (from them) – not much consideration is given as to where the Flying Foxes should go to considering its humans encroaching of habitats and much less of the other way around.
They (residents) are also not aware that flying fox populations are mobile – that an individual flying fox is rarely at any given site longer than 11 days – they just see the colony or Roost is there for a whole season.
As part of this project education of the public is critical. We go about this through a five-step plan we developed to handle the education of the public as well as response to public feedback and complaints.
When a complaint call regarding flying foxes is received from council it is important to recognise that the response from the individual is valid. Once details of the complaint have been recorded, ask the individual is they are willing to participate in a visit to inspect the roost and participate in the collection of data to understand flying fox habits.
If the individual expresses interest in a member of the council coming to inspect the roost location, ensure this is done as soon as possible. A member should be sent to inspect the roost and if the individual is willing, educate them on the program the council is running and the incentives and benefits of participation.
Education is critical to the success of this project. Education points involve how flying foxes normally do not stay in one roost but move around every 11 days.
Another point that should be mentioned is that when we understand their movements and preferred roosting options then we can create appetising roost locations away from private property and public thorough fares to minimise public conflict with the flying foxes
Explain to the individual that there is incentive for participation in the collection of the data for each successful contribution.
This could take the form of a discount up to a set amount from rates bills or offer other types of small sponsor gift voucher type rewards for correctly documented information that results in an effective tracking event. For example, the photo provided is sufficient to identify the animal and has supporting details of location and time taken. When the information is added to the database, if it matches and another entry, a micro deduction bounty is awarded to the uploader.
When the individual has expressed interest in the bounty, explain how the individual can participate to receive the reward.
This will involve the individual taking photos of the roost and submitting the photo through a portal to check for wing patterns.
If the wing pattern generates a match from the database, then the participant will receive a reward for the contribution as discussed in the bounty section.
Flyers and public education
Letterbox drops in neighbourhoods to distribute educational flyers when public visits are conducted will help spread public education.
Appealing to residents for their help in expanding our knowledge so we can better address community concerns and allay fears. Find a way to coexist.
Collecting the data
The data will be collected by wildlife cameras in fixed positions that have been installed by council around known roosts of flying foxes.
The camera takes a photo of the bat, this can be done in flight or while roosting. This photo is then uploaded automatically to the flying fox database.
The photo is then analysed using an algorithm which maps a triangular-shaped section of the flying fox’s wing. This section is then compared to other exiting mappings of flying foxes already in the data base.
If the section of the wing is not a match, then the algorithm creates a new entry for this flying fox which contains information about the flying fox like the species, which colony it was found in and whether it is a mother with juveniles.
If the section of the wing is a match, then the algorithm moves on to another section of the wing to narrow down and identify which flying fox in the data base the photo matches. Once the algorithm identifies a single match, the algorithm updates the sightings database which contains all the sightings of each flying fox.
The algorithm will also be able to distinguish between other animals which also populate trees like birds or even cats. Anything that isn’t a flying fox is ignored.
Our database contains fields to identify the flying fox (animal id number, species, the identifying wing features picture) and fields that describe the observation instance. Date, time, location, and any other relevant information like the name and contact details of the uploader or the council camera the photo was taken from.
Data will be collected through static cameras the local council has placed at roost locations and public photographs submitted through the website form.
Vulnerability of the cameras (keeping the lenses clear).
Cameras being stolen.
Flying foxes may not roost with their wings open.
Risk of people uploading fake photos.
It is anticipated that a predictive model could be developed whereby potential future roosting points can be identified. Any predictive ability would be improved by introducing further datasets, such as land use and development and establishing any correlation.
The database is fully customisable. Currently, it has limited sample fields which can be overwritten with future data. As data becomes available fields can be expanded. For example, if it is ascertained that bats remain loyal to a colony then movements can be tracked using a colony. A colony could be identified from a sufficient sample of individual flying-foxes.
The database can be expanded to network with other councils with the expectation to develop a national database.
The future and use of this data:
These letterbox drops and monitoring processes can then be expanded to locations surrounding areas deemed “a high probability” of having a Flying Fox Roost establish itself in the future.
Information can also be circulated to land developers to consider how their green space can be better formed to support peaceful cohabitation between humans and flying foxes.
Amelon, S., Hooper, S.E., & Womack, K.M. (2017). Bat wing biometrics: using collagen-elastin bundles in bat wings as a unique individual identifier. Journal of Mammalogy, 98, 744 - 751.