5 January 2024

Support sought for high-tech count of Kosciuszko's wild horses

| Edwina Mason
Join the conversation
wild horses standing in the bush in Kosciuszko National Park

State-of-the art technology that allows Australian farmers to view pests and crop stress from above, will be deployed to provide an independent count of the wild horses in Kosciuszko National Park. Image: Snowy Brumby Adventures with Michelle and Ian.

Brumby supporters are in the throes of calling for crowdfunding support to finance a new, highly sophisticated independent count of wild horse populations in Kosciuszko National Park (KNP).

As the NSW Government stands poised to commence aerial shooting of the horses, independent biostatistician Clare Galea has thrown her weight behind leading edge technology as an alternative measure of the numbers of horses in the park.

Using high-definition airborne cameras, she says high resolution imagery of open terrain in KNP combined with machine learning computer analytics and the latest artificial intelligence (AI) software should mean every horse in the surveyed area is accounted for.

When results are peer reviewed, the raw data will be made available to relevant parties to be cross-checked for integrity.

Developed by South Australian remote sensing and data analytics company, Airborne Logic, the technology has already been widely adopted for high quality analysis of crops, vegetation, assets and surface topography to provide actionable advice to farmers, government agencies and private companies.

Ms Galea says these advanced techniques would achieve far higher accuracy and open accountability in KNP.

The survey area will be limited to open terrain in the ‘Northern Block’ where most of the horses reside.

“The open terrain will provide clearer imagery and a very high level of accuracy when counting with software,” Ms Galea said.

READ ALSO New data triggers aerial shooting of wild horses in Kosciuszko National Park

Wild horse population surveys are prepared by Dr Stuart Cairns of the University of New England’s School of Environmental and Rural Science on behalf of NPWS using a method known as distance sampling.

These were undertaken in 2014, 2019, 2020, 2022 and October 2023 – the last of which estimated KNP’s wild horse population as between 12,934 and 22,536 horses, with a best estimate of 17,432 remaining.

This number has been widely disputed by locals and brumby advocates who have suggested the population now sits in the low thousands.

Foal in car

Efforts to rescue orphaned brumby foals continued over the Christmas period. Image: Snowy Brumby Adventures with Michelle and Ian.

Ms Galea says her primary goal is not to prove the current methodology inaccurate but to get the government to adopt more accurate and accountable methods of counting animals.

“This is not only for wild horses – as technology improves the same methods can be used on smaller species such as kangaroos, koalas, wild pigs and wild deer.”

She says the benefits of using advanced image recognition, machine learning and computer science for identification of animals is well documented.

“AI techniques which use imagery to accurately identify animals are rapidly advancing, helping scientists and data analytics companies to improve the overall success of animal identification,” she said.

“We will accurately count the overall number and the density of horses to provide imagery that can be independently assessed.”

READ ALSO Summernats 36 rolls into Canberra with high hopes (and a warning from police)

She’s confident the density will be significantly lower than NPWS states, which she says will demonstrate the need for the government to cease relying on the current counting methods.

“There has been much commentary and dispute on actual wild horse numbers in the park, and this proposal aims to make real ground on the issue,” she said.

“To ensure the NPWS do not cull to a number less than this, accurate counting methods must be adopted,” Ms Galea said.

Around two thirds of the $75,000 cost of commissioning the new count has been raised in the 10 days since the GoFundMe crowdfunding campaign was launched.

Crackenback local Rocky Harvey has teamed up with Ms Galea to organise the online fundraiser and he says donations continue to flow in.

“Obviously we can’t start a thing until the funds – which are going straight into a trust – are secured,” he explained, “but we’re hoping to reach our target in the next week and then lock down time frames.

“We would certainly like to think the imagery flights are run in mid January followed on by data analysis,” he said.

Original Article published by Edwina Mason on About Regional.

Join the conversation

All Comments
  • All Comments
  • Website Comments
Philip in Narooma10:39 am 08 Jan 24

As a Veterinarian who has been involved with statistical methodologies for over 40 years I would disagree with some of ‘An Ecologist’’s views.

Ms Galea’s qualifications in statistical analysis, I believe, are quite appropriate to make comment on this issue. Her discussion paper DOES raise some significant issues with the count and, in particular, with the methodology used.

Methodology in the Cairns (2022) paper (NOT peer-reviewed) is the issue that can be critiqued. Having read this paper in my opinion there are areas that would be in dispute amongst peer-reviewers. If the reference list in the Cairns (2022) paper is looked at then it becomes clear that NONE of the papers relating to feral horse populations in KNP are peer-reviewed .. they are all ‘reports’.

Finally what needs to be followed up is what is called ‘ground-truthing’. In other words cross checking the theoretical information with physical field surveys.

As to those who believe that ONE feral horse is one too many .. consider what will replace them? When the horses are removed has any scientists considered what will replace them .. I very much doubt it. These folks are the ‘lock it up and lock ‘em out’ fundamentalists.

I have been fishing in the Long plain, Geehi and Thredbo area of KNP since the mid 1960’s. In my opinion the effect of the rabbit on the Long plain area was severe . When Calicivirus wiped out 90% of the population in the mid ‘90s the effect was spectacular.

The effect of horses in the Long plain area is much less from personal observation. The horse population appears to have been fairly static in the Long plain area since the late ‘90’s, however the rise of pig populations and the deer population in the southern areas HAS been dramatic. The damage done by deer and pigs can be far more severe than horses because of behaviour aspects .. but that is a different story.

An Ecologist2:56 pm 09 Jan 24

Hi Philip of Narooma, so, both you and Galea lack ecological qualifications. In particular you both lack experience with the unusual and very learned statistical method of distance sampling. And it seems you have not read Galea’s report on the counting methods (18 pp). If you had, I suspect you may be less inclined to support her.

In fact the current count method IS published (https://www.jstor.org/stable/3784462). You’re correct that the 2022 report was not reviewed before release. But the earlier reports by Cairns are almost identical in nature, and have been extensively reviewed. After favourable reviews of previous reports I guess the NPWS and its minister did not want to delay the latest (2022) results while waiting for another review.

Even if doubts about the count were given credibility, and the count results were taken as merely an index of abundance, applying the same method consistently over 21 years shows that (1) the population increased steadily at 15% per year (in the face of control attempts); (2) by 2022 there were many times as many horses as when this method of counting commenced in 2001; and (3) that horses are now occupying a greater area.

Therefore many vulnerable animal and plant populations are encountering horses for the first time, or encountering high densities of them for the first time. Two scientific committees advising both the Commonwealth and NSW Governments have identified more than a score of such horse-vulnerable species, many of which occur nowhere but in Kosci NP. The clock needs to be turned back so those populations can survive even in the short term.

Ecologists have been warning about horses since at least 1953 and advising for decades on methods more humane than what was being used. If earlier advice from ecologists had been respected, we would not be in the current predicament.

Philip in Narooma4:52 pm 09 Jan 24

“An ecologist’ .. it appears you are unaware of what a peer-review means. It is very different to a ‘report’, even a report that has ‘favourable reviews. Perhaps I could take you back to 2003 and the catastrophe of the Grey Nurse shark count.

This count, a ‘mark recapture’, was done using ‘correct’ statistical methods (according to the authors) and the report presented to NSW DPI (fisheries). Subsequently the GNS was declared ‘critically endangered’ by the NSW Threatened Species Scientific Committee. This diverted millions of dollars fisheries funds into efforts to ‘Save the GNS’. Along with several others a case was brought by the NCC and the counts veracity was shown to be wrong. Instead of a number between 300 and 500 the actual count is more than 2700.

What went wrong? Nothing was wrong as far as the science of ‘mark recapture’. What was wrong was the methodology of the count. In fact many felt there was a corruption of the process that has probably soured many off so called ‘ecological’ methods.

You are correct that I do not have ‘ecological qualifications’ (whatever they are?). I have found the only folks who hammer that point in an argument are new graduates (or undergraduates) with minimal experience in the field. Notwithstanding my degree in Veterinary Science and my post grad. Qualifications, my near 40 years experience in statistical methodology in both Australia and Africa means I adopt a cautious approach when trying to count any wild animal.

My principal concerns about the feral horse issue is the amount of damage they are alleged to do compared to pigs, wild dogs and rabbits. The horses have been present in KNP since the mid 1800’s. Since the mid 1950’s virtually no control method has been performed and the number is alleged to be 18,000. So what was the population in 1960 at your figure of 15% increase (per year)?

If your focus is mainly on the damage done by horses in KNP, why haven’t you mentioned the significant body of scientific evidence showing horses specifically causing significant damage?

Damage that is fundamentally different to that caused by other feral animals and thus distinguishable by researchers in the area? Or is that research also in question?

I’m also wondering what the potential inaccuracy in the horse counts matters for, in overall effect?

Are you worried about a misallocation of resources to control feral animals? Or something else?

The other feral animal types you mention are controlled using similar methods, with thousands of wild deer, pigs etc being shot using helicopters in recent years. Rabbit populations also face significant control measures.

Seemingly the only bad thing that would occur is that they remove too many feral horses which would then be reflected in subsequent population counts, even if with wider confidence bounds than desired. It would also become significantly more difficult to hit expected cull quotas if the counts are wrong to such a high level as claimed.

If the horses are not the culprits for the environmental damage (despite the significant body of evidence showing the opposite), surely the environmental surveys would then show continued degradation of the areas? Leading to a need to adopt other methods and controls.

But I’m struggling to see a real downside of managing the feral horse populations, regardless of the claimed error in the counting methodology. Obviously improving those methods over time would be beneficial, but it doesn’t seem to be a reason to stop culling the feral animals in the short term. Also noting that using inaccuracy of counting is a common method used by activists to achieve their preferred outcome and stop population control measures.

An Ecologist8:31 am 10 Jan 24

Again Philip of Narooma, since it seems to matter to you, I too have had over 40 years experience, not that there is anything necessarily wrong with the thoughts of recent graduates. Nor is there any basis for an experienced veterinarian like you to try to cast shade on ecology. As with veterinary science, ecology is a well recognised science. It is defined as the scientific study of the relationships that determine abundance of living organisms. And there are numerous journals publishing ecology such as the Australian ‘Austral Ecology’ and ‘Wildlife Research’.

Returning to the topic of horse counts, there are multiple papers in peer reviewed journals, including ones comparing different methods of estimating horse abundance and numerous scientists have been involved in one way or another, over 21 years since the current counting method was first deployed, so the horse counts are not comparable to a single mistaken count of sharks.

Also regarding impacts there are multiple papers from authors in different research institutions showing horse impacts greater than the other species you mention. But even if the impact of horses was less than that of something else, it would be no basis to refrain from acting to reduce the horse impacts. And as Chewy14 argues, all the other species are being subjected to control.

I think this conversation has got to a point it is not worth persisting with.

Philip in Narooma5:39 pm 10 Jan 24

A summary of my concerns regarding the feral horse population
1. In KNP 4,500 deer, 670 pigs, 15 goats and 17 foxes were killed by aerial shooting in three years to end of 2022.
2. I have no qualms about reducing the number of horses if the following conditions are satisfied and correct.
a. The horses have been demonstrated to be the sole (even major) cause of the damage to the ecosystem , especially in the north of the park where they are most numerous. From personal observation their numbers are negligible
b. Consideration has been given to what will replace the horses. My concern relates to feral pigs, which are quite numerous in the north of the park. I don’t believe any thought has been given to this
3. Animal behaviour, of which I am very familiar, does not predicate that horses will convert stream and river edges to mud wallows, even cattle don’t. Pigs on the other hand are notorious for doing this. There may well be horse prints in a muddy area, in my opinion they did NOT do the original damage. Deer can be very damaging to river banks by the way the move over rough country – in single file. On our farm near Goulburn we were overrun with fallow deer at certain times of the year and caused erosion of Covan creek.
4. Weeds – AE claims the horses hoof means they carry weed seeds. I very much doubt that – horses can certainly pass weed seeds in their manure as they have a relatively simple stomach compared to a cow. But the weeds are already there, over the past 150 years!!
5. Carrying capacity. In terms of numbers of horses the concept that they will increase 15% year on year in a fixed environment (a very large paddock if you like) is completely foreign to animal husbandry and management. It is a furphy. Theoretically that is possible
a. Any large animal Veterinarian drives around calculating the carrying capacity of a paddock in terms of DSE. In other words 1 DSE is a 50kg merino wether and 1 DSE to the acre is the feed required in one acre to maintain his weight for a year.
b. A horse is the equivalent of 10 DSE, a pregnant mare is about 15.
c. The northern part of KNP varies between 0.5 to 1 DSE per acre. In other words one horse would require between 10 to 20 acres acres to maintain his weight. The southern part (Byadbo and Tin mine huts) would be 0.2 to 0.5.

In my opinion the horse population has probably varied by 10 to 30% up and down over the past 40 years

What a fantastic initiative! Finally we have an opportunity to establish an ACCURATE method of counting animal numbers. Once we have this proper census, then we can prepare appropriate action plans, whether that be horses, kangaroos or any other animal, native or introduced.
This hysteria over a single species, horses currently, is based on appallingly bad and manipulated statistics. I’m incredibly happy to be supporting this new counting measure. And looking forward to some rational and transparent decision making once results are available. Well done guys!

An Ecologist9:43 am 07 Jan 24

The horses occupy 3,700 sq km of the park according to NPWS. Much of it is rugged, as Banjo Paterson wrote in ‘The Man from Snowy River’. The task is nothing like counting horses in a paddock.

Many wildlife controversies involve disputes with scientists’ estimates of abundance. Unqualified protestors sometimes assert that they can do a better job themselves (Galea is a PhD student in education). The ecologists think that their PhD qualifications, training, years of experience, and hard-won papers in peer-reviewed ecology journals give them credibility. But sadly, some politicians seem to want to side with the passionate people who have as much right to count the horses as I do to wire a building for electricity or diagnose diabetes. In this way the horse controversy is like a score of controversies where science is pitted against alternative opinions, such as the biodiversity crisis, climate change, GMOs, vaccination, chemtrails, divining rods, crystal healing, flat earth and homeopathy.

As with counting, so too with humaneness. Highly qualified experts in veterinary science and ballistics have found that helicopter shooting is more humane than many other accepted methods that are applied to horses and other wildlife. And they set up binding Codes for how it must be done, which are publicly available. Yet many people with limited knowledge assert the contrary. For example the opinion that horse carcasses with multiple bullet holes indicates wounding is breathtakingly ignorant, because it is an absolute requirement of the relevant Codes to apply multiple shots to each animal, no matter what, and to do so more than once.

“The ecologists think that their PhD qualifications, training, years of experience, and hard-won papers in peer-reviewed ecology journals give them credibility.”

Where can we go then if we can’t rely on all of this experience and knowledge An Ecologist?

The method of the count is irrelevant.

If a count by any means gives a number greater than zero for any feral pest in the KNP then that pest needs to be eradicated in that area. By whatever means necessary.

John Kennedy1:39 pm 06 Jan 24

Just another inexperienced article about market research methodology; trying to muddy the waters yet again. Horses are not native and should be removed full stop. I would like future generations to experience Australian wildlife, not some hybrid mix of worldly ferals that are now found in most other continents.

Vicki Sprowles6:46 pm 06 Jan 24

You make me ashamed to be an aussie. All life is sacred. These horses were brought here by men and left alone in the inhospitable bush by men. They are a large part of our history and should be protected by law. If their numbers grow too large then vets should geld a number of the stallions. Problem solved. Aerial shooting is the most inhumane way to solve this or any other overpopulation problem. Not Acceptable to the Australian people.

You are embarrassing yourself. Horses are as welcome in a National Park as wild pigs, rabbits or cane toads. Most Australians support their total eradication.

Your argument appears to be that if men do something, even where it is damaging, that action should be protected by law.

Quite a few people might disagree with that weird attempt at logic.

“All life is sacred”
Nice to know your opinions are devout, though rationality has exited your cathedral..

Just shoot them all, the goal should be no feral horses.

Any feral horse numbers greater than zero is too many.

If feral horse advocates are so attached to them then why aren’t they taking them in themselves?

Splitting hairs over how many of these dear pink unicorns there may be – whether their souls depart this earth rapidly on a cloud lined with roses or die slowly in wounded pain, left to decompose back into the environment or are removed – is irrelevant. Feral horses should be eradicated from the Snowies just like deer, pigs, dogs, foxes, etc.

The only people supporting feral horses do so for personal financial gain. Talk to them, they have no interest in animal welfare, the environment or the community. I say the crowd sourcing initiative is great – anyone silly enough to donate deserves to be relieved of their money. The technology for this one is easy and free – even a count of zero feral horses means it will take decades for the flora and fauna of the Snowies to recover.

The NSW Government should be congratulated for getting on with the job despite the bleating of a few weekend pony patters.

Catherine Ehlefeldt6:28 pm 06 Jan 24

It sickens & disgusts me that many more of these horses are, once again, going to be slaughtered by aerial shooting & left to die an agonising death, shot to pieces by gun toting rednecks, with no conscience, shame on you

Nicole Melrose3:28 pm 05 Jan 24

You are all leaving your footprint on the earth in some way shape of form,
Who gives you the control over who has life and breaths
No matter what animal,
You have so many other things in life you could help the world
Starting with your very own wasteful existence
Fix your own backyard before trying to fix others
Then our world would be a much better place
… learn to mind your own business first, don’t buy new clothes, don’t use a car, grow your own food and don’t your dare crap in the toilet or use toilet paper…dont be a hypocritical human,
Control freaks
We are all introduced except First Nations, did they get a voice NO
Where are you all helping them???
Bloody joke

Inaction has consequences, as surely as action. Choices are unavoidable, whatever your fantasies.

Some people are trying to fix the horse pollution in the Kosciuszko back yard. Kindly take your nonsense elsewhere.

An Ecologist10:52 pm 06 Jan 24

No Nicole, humans evolved in Africa. They did not evolve in Australia. Aboriginal people also invaded Australia and contributed to the extinction of numerous species. Where should we defecate if not in the toilet? Where do you do it?

Colleen Krestensen1:15 pm 05 Jan 24

Please donate to this if you can. An independent aerial survey is vital to inform targeted management of brumbies and literally avert ‘overkill’. Over 1600 horses were killed in 2023 by NPWS. The Government’s aerial culling strategy is to shoot 7.5 bullets into each horse and kill the foals first. This is not good animal welfare and history has proven the cruelty of shooting horses from helicopters. Ms Galea is being quite transparent about the methodology/peer review.

Have we got them all yet? If not, there is under-kill, although half-bullets do seem inefficient.

I expect a few will survive, just like feral deer, pigs, foxes, cats, rabbits and cane toads. If you miss patting a horse, try any of the others listed.

It would be very helpful if Ms Galea would publish the methodology she plans to use, so that it can be discussed and critiqued before the survey occurs – in particular the method she plans to use to estimate the horses that are present in the survey area but not visible to the airborne cameras.

Daily Digest

Want the best Canberra news delivered daily? Every day we package the most popular Riotact stories and send them straight to your inbox. Sign-up now for trusted local news that will never be behind a paywall.

By submitting your email address you are agreeing to Region Group's terms and conditions and privacy policy.