Really all depends on the species you wish to collar - for example, in polar bears, it is possible to collar females, but not males because their head is smaller than their neck so the collar can be relatively easily slipped off.  When I worked on badgers, our rule of thumb was that you had to be able to easily slip two fingers between the collar and the neck at the throat to ensure that it was tight enough to stay on, but would not constrict swallowing or breathing!

Thanks John,

The two fingers rule is applicable for many species. We also use it with almost all species.

While there are always exceptions like the polar bears, many species can be grouped by there neck characteristics - canines, gazelles etc.

I'm still hoping to find a review or research that empirically compared such rules of thumb.

here is an example with zebras:

http://onlinelibrary.wiley.com/doi/10.2193/2007-061/abstract

Hi Amir,

for wild boar: do not use collars:

if it is not tight enough (just one small finger fits between collar and neck) the collar will be lost usually within some few months.

If it is too tight the animal might (or better will) be injured. Me and some other colleagues will not use collars anymore due to animal welfare!

I will go on using ear-tag-devices on wild boar

for predators: seems to be nor problem with foxes, wild cat, lynx, wolf, martens (and badgers as John wrote). semi-aquatic like mink, otter or polecate should get implants!

If the animal has a normal neck, then it´s no problem! for red deer: two fingers should fit between collar and neck!

However, I do not know any overview. There are some old ones existing on the weight (percentage of bodyweight). But I do not find them again in my database.

Olli

I agree with Oliver Keuling.  There is no completely safe method unless the animal can be continuously and reliably viewed after fitting and there is a reliable remote release mechanism.  Have you considered non-invasive approaches that can be less costly?  A collar that is comfortable on fitting can quickly change fit with the condition of the animal and become too tight causing peripheral oedema and/or wounding, or too loose and fall off.  Animals can be more at-risk of predation, and less able to compete with con-specifics.  Even mild discomfort can change the natural behaviour of the animal and render the data collected unreliable.  Technology is progressing fast - it's time to move on.

Thanks guys, for your inputs and thoughts.

These are very helpful rules of thumbs. Still, it's very frustrating to know that after decades and thousands of trials (and saddening errors) all around the world, the knowledge that is easily accessible is  5% and two fingers as the common guidelines.

BTW, 5% of body mass to me seems inadequate, especially when I imagine men going around with ~ 3.5 kg collars around their necks.

I agree that the welfare of the animal is crucial, but since we use collaring to support conservation, you have to weigh the potential damage against the potential benefit. It's not an easy decision.

I think that in many (most?) cases, non-invasive methods can't give you the information you need.

I did find a few papers on the effects of collars on behavior and health,  and a few  that tested and suggest specific designs for different animals and conditions. Together with your advices - this short journey was very useful for me. I will be happy to share the papers, so contact me if you want them.

Cheers,

Amir

5% is an old rough rule. especially this has changed.

for smaller animals (especially mammals, it might still fit.

However, in most species it is much less! The heavier, the less percentage (as a rule ;-)

You may find some literature in

Millspaugh JJ & Marzluff JM 2001 Radio Tracking and Animal Populations, Academic Press

Kenward RE 2001 A Manual for Wildlife Radio Tagging, Academic Press

White GC & Garrott RA 1990 Analysis of Wildlife Radio-Tracking Data, Academic Press

Silvy NJ 2012 The Wildlife Techniques Manual - Research, Volume 1, 7th ed. John Hopkins University Press

looks like there is a lot of literature existing, but nobody did make a review out of it...

Seems to be your job to find all the literature and to make a review!

Olli

I found a lecture of one of my retired colleagues, there the % is between 0.2 and 5%, mainly below 3% (2%).

Unfortunately, I do not find the original source, but most of the data come from the 1960th and 1970th!

I fully agree with all Oliver's comments: the 5% rule is obsolete (and I just cannot understand how on earth could even be developed...). 

As a rule-of-thumb, definitely I suggest that a tag weight of 1% (or less) is the safest option. Over the last few decades, we have radiotagged martens, badgers, red foxes, porcupines, roe and red deer, chamois and common leopard, always sticking to the "no more than 1% rule". We never detected any trouble with our radiocollared animals.

As to the width of the collar, there are a few issues to consider:

is the animal a subadult and bound to grow? Then, you had better set the collar at a size which will not interfere with growth.

are you catching your animals when their body size is at a minimum, e.g. at the end of winter? Then, you should consider that their neck will grow slightly thicker later on.

do NOT be afraid to set the collar TIGHT: the 2-3 finger rule is alright (and the fingers had better be kept horizontal - not vertical: I have got evidence that this misunderstanding  may occur!).  The trouble that a loose collar can cause to an unfortunate animal cannot be guessed... from damaging the mandible of grazing herbivores to hamper a running/jumping animal.  In one case, one of our red foxes died a horrible death because the student who radiotagged her fit the collar too wide, despite my recommendations: we found this fox hanging from a bush, with the stunt of a low branch tucked between collar and neck. In another comparable case, a red deer hind stepped inside her collar with a front foot, presumably while grazing, and  moved around on three legs for several weeks (most fortunately, she managed to release her leg from the collar, eventually).

On the other hand, a too tight collar - even if certainly disturbing - is less likely to cause death or seriously hamper.  The neck of rutting male deer does increase its size, but this increase does not appear to bother radiocollared individuals.

One free-ranging dog, who escaped or was released by an irresponsible owner as a subadult, had still her old collar when she was rescued from the wild years later: her neck girth had adapted to the size of the collar and she had even reproduced at least once.

Have a good luck with your radiotracking and... KEEP COOL when you have the animal in your hands, to fit the radiocollar!

Cheers,

SANDRO

I agree with those above, 5% is an old rule and I doubt many people are using collars over 3% BM these days, 3% would also be quite a large collar. Every animal is different which is probably why there are no general rules but the 2 finger rule works well for most carnivores. Ungulates are more difficult because they generally have a tapered neck and because they graze and browse which results in a lot of head movement up and down and the same movement of the collar - this could be disturbing to the animal and requires a good snug fit. Animals moving uop and down cliffs or steep terrain have similar issues. with smallish animals it is generally possible to hold the animal while handling and see how the collar is moving so it can be adjusted accordingly.

I don't agree that collars always result in behavioural changes, discomfort etc. to the animal. There are definately cases of collaring where the disturbance to the animal is minimal and there are no ethical considerations, some collars are extremely small and don't move,  much the same as a dogs collar. however where the collar is definately noticed by the animal then having a very good reason to collar is important, collars are often the only way to get conservation information and are still the method of choice for understanding critical habitats, ecology and resource needs.       

Many good points have been mentioned already, so I will keep my comment short and just point out two additional points arising from our experiences with collaring banded mongooses:

1) Some animals loose weight and fatten up quite a bit across seasons. This needs to be taken into account when fitting a collar. A collar fitted normally (finger rule) in the lean season may well be too tight a couple of months later.

2) For collars with an antenna it is also worth thinking about positioning and length of the antenna. In our case, it proved to be worthwhile shortening the antenna, thereby reducing the risk that it gets caught and breaks off. This may not be a direct welfare issue, as the antenna typically breaks off quite easily, but it certainly affects effectivity of the collar (also, one can argue that putting an ineffective collar on an animal is ethically questionable).

As Corsin, I will keep  my comment short as most has been said.

From our experience on African buffalo and cattle (always good to test your collar on similar size/behavior domestic animal or pet):

1. Considering growth is very important (cf Sandro's comment).

2. For this mammal size --> 4 fingers thumb rule. Position of the head/neck during anesthesia can mislead a lot your perception of collar tightness --> try always to have the head in the same position.

3. Communication about your study --> some farmers once claimed they saw a buffalo in their village dying from strangulation from the collar. Was quite a story with national parks and we had to prove that it was not the case. Sometimes, it can be difficult to defend your case so put everything on the safer side for your protocol (better to have a few collar failures after falling than being expelled from the study site by stakeholders after contributing to the extinction of your study species!). Tourists can complain about collars need to be taken into account also.

4. We observed small injuries on cattle on the upper part of the neck but nothing important a part from damaging the skin (can be important for game/hunted species).

A review could be good to introduce new comers into the game.

Best,

AleX

Hi all, please remember that the 5% bodyweight is still used for very small mammals. If you want to track small mammals of 25gr you have to find collars between 0.5 and 1.5 gram, running into the trade-off between weight, battery life and detection range. So, the smaller the animal the closer to 5% bodyweight people go to extend range and battery size. Also it is important to remember, that virtually all small mammal females caught and collared get pregnant and gain a considerable amount of body weight, so multiple catch events to adjust the collar during a season should be considered. For small mammals, the collars have to be really tight, so animals should be monitored for some time before release to have the chance to readjust collars if they start to choke.

Cheers, Thilo

For new world monkeys of big and medium size (genus Cebus, Sapajus, Lagothrix, Alouatta) we use collars of synthetic materials.  The key is to put them not so loose because the risk to hook up a branch is high.  The distance between the neck and the collar should be the diameter of the little finger.  If you need that the collar will detach, you should fix it with absorbable surgical material.  For smaller species (genus Saimiri, Saguinus) we use backpacks instead of collars.

Well, I left the computer for a day to do field work, and found that half the chapter on terrestrial mammals has been written. Now we have to go to marine mammals, birds, reptiles, fish, insects...

Thank you all, and thank you ResearchGate - this is really a wonderful platform.

I think that a review is not the right way to gather experience like you shared. Most of it is not in papers anyhow, and I'm guessing most of it is not even well documented - with measures and outcomes  (including my own).

One option that I thought of is asking Movebank to add a section where people can describe what they did - design, weights (animal and tag), success/failure, etc. Preferably with pictures. Movebank can promote this by sending a request to all their users and post it on their homepage.

In the meanwhile, I will put all the papers I found so far in a Dropbox Folder (address at the end). Please upload what you find helpful.

Cheers,

Amir

https://www.dropbox.com/sh/z1r1zexlpaa91gz/AAANQlU4RmCDM3QGY61cBYK-a?dl=0

https://www.dropbox.com/sh/z1r1zexlpaa91gz/AAANQlU4RmCDM3QGY61cBYK-a?dl=0

Hi Amir,

Just added the Brooks, Bonyongo, Harris - 2008 - Effects of Global Positioning System Collar Weight on Zebra Behavior and Location Error to your dropbox.

I also worked on zebra and my main conclusion is that it depends on the morphology of the animal and its behavior.

For example with Zebra we had to put the collar on quite tight otherwise it would slide up and down the animal's neck and knock against it's jaw. Furthermore zebras are very rough when dust-bathing and can turn the collar upside down and the collar would then have very low success rates until it was turned back upside up.

Finally the material used may be important especially in hot environments it appears the collars become quite uncomfortable when the animal starts sweating.

For elephant, cattle and Buffalo we kept the collars quite loose (see Alex's comment above we work on the same project)

Did you contact the collar manufacturers? I am sure they would be very keen to develop these kinds of guidelines.

cheers

Hugo

Hi Amir,

it is a quite good idea to add experiences to a Databank like Movebank.

However, there are also other Databanks (like EURODEER) where this can be added.

Olli

Thanks Hugo.

I have a few more papers, didn't get the time to upload them.

I also thought about harnessing leading brands (or one) to such a project.

Olli - EURODEER (and EUROBOAR) are also very good ideas. If you have others, please share.

And if anyone is interested in helping me with this - please contact me and we get it coordinated. I'm quite swamped as it is...

Amir

Hi Amir,

just a short remark on the weight of the equipment attached to the animal. I think that we still need to differentiate between the situation of radio-telemetry collars and GPS-collars, as increased power consumption of the latter is asking for much bigger batteries. Depending on the animal size, fix frequency and study duration, the 5% threshold might still have its right to exist, particularly in animals below 5kg.

As I have been working with primates a few years ago, we also had to consider, if the animals were primarily tree-dwelling, as the movement (jumping) behavior might be influenced to a higher degree than in ground-dwelling species. 

Pascal 

I disagree on the 5% rule.  Where is the evidence to suggest that smaller animals might be less impacted?  Individual outcomes are rarely reported.  Consider that the weight of 5% around the neck is about 3.5kg for an average 70kg adult human.  I think most of us would agree that this would have a considerable impact on our daily activities.  The impact on quadrupeds might be expected to be greater.  In addition, for free-ranging animals that need to predate and avoid predation, any negative impact on movement could cause very significant reduction in fitness.

Putting collars on is the easy bit. Taking them off again is much harder, hence black hole in the literature.

Lots of black holes in this literature.  How often does a study using collars report the number where the transmitter failed, where the collar itself failed, how many animals collared were still transmitting at the end of the study, and the injury/mortality rates.  Transmitters that fail, but the collar remains on, are particularly hazardous for the animal.  More studies should attempt to compare the behaviour or ecology of the subject animal with unmarked/uninstrumented controls, or compare different weights of instrumentation, or designs of collar.

Here is an article I wrote for the Wild Felid Monitor 6(2) 2013.  Hope it helps 

A TOPIC EVERY RESEARCHER USING RADIOCOLLARS SHOULD CONSIDER

Rich Beausoleil, Bear & Cougar Specialist, Washington Department of Fish and Wildlife, [email protected]

In my career as a wildlife professional, I’ve been lucky enough to conduct field research with bear and cougar. I’ve been using breakaway cotton spacers (leather is also available) on radiocollars since graduate school. It has always been my standard operating procedure when conducting research. I remember my major advisor, Dr Michael Pelton, telling all the new graduate students to make sure they used cotton spacers and to read the manuscript written by Hellgren et al. (1988) titled

Use of Breakaway Cotton Spacers on Radiocollars (Wildlife Society Bulletin 16:216-218). They noted that the cotton spacers serve a dual function: 1) they are the weak-link in the collar belt that allow some stretch should a bear gain a significant amount of weight and, 2) the spacer will eventually rot and tear, allowing the collar to drop off the animal in 2-3 years. Use of a breakaway spacer provides an ethical insurance policy that the animal will not have to wear the radiocollar should it not be recaptured or the electronics in the radiocollar fail. Historically, radiocollar belts were 2 pieces of leather sewn together with a cotton thread; if a spacer wasn’t used the collar could potentially remain on an animal 6-8 years. Today, many collar manufacturers are using synthetic belts. I like the new collar materials because they allow the belt to be molded for a more comfortable fit for the animal, however, these belts can last on an animal well over a decade.

Over the years I’ve presented much of my research at conferences, and by talking to other researchers I have made 2 informal observations. First, the use of cotton spacers seems more common in bear than felid research. Second, it doesn’t seem as if graduate students are as aware of spacers as they should be. I may be incorrect in this assessment, but in my mind, the topics of animal welfare and ethics cannot be over-discussed. I think it’s worth broaching the subject and getting folks communicating, especially WFA members.

Recently, while attending the 10

th Mountain Lion Workshop in Montana in May of 2011, I was concerned to learn that very few researchers are using cotton spacers (or electronic break-away functions, more on that in a minute). While it’s true that mountain lions and other felids have a much more predictable and limited weight-gain compared to bears and therefore the ability of the collar to "stretch" is not as crucial, there remains the issue of electronic failure and recapture for collar-removal. It troubles me to think of any animal having to tote a collar around for the rest of its life without generating information for researchers. This is the reason for this short communication. It’s not to poke, criticize, or offend anyone: it’s strictly about ethics and animal welfare.

The more common reasons I’ve heard for not using cotton spacers include: (1) electronic reliably within the radiocollars has improved; (2) the animal will be recaptured; (3) I’m using an electronic breakaway; and (4) I don’t want to risk the collar being dropped prematurely and losing data. Responses #1-3 can be easily addressed: electronics will always fail and sometimes animals leave your study area. As researchers we need to insure that once we release a captured and collared animal we can be certain that the animal will not be wearing that collar forever.

My response to #4 is also straightforward;

we don’t always have to win. If we use cotton spacers, and the cotton rots off quicker than expected, that’s ok, move on to a 3 or 4-ply rather than a 2-ply spacer. But take solace in the fact that you put the animal’s welfare above informational needs. As wildlife professionals, we should strive to carry out our work with the highest ethical standards. Believe me, I know how immensely important research is and that information from every animal matters, especially when your study animal is a solitary, low-density carnivore of significant ecological and conservation importance, but we should live by the simple rule that as researchers we must always consider the long-term ramifications of our actions and place the animal above ourselves. We should also follow this same principle when attaching collars to the animals. By all means, securely attach the collar and make it fit appropriately, but leave room for growth and comfort. If we leave a bit too much room, and it gets pulled off, that’s ok; we don’t always have to win.

In closing, the use of electronic breakaways brings up a similar issue that I want to mention just briefly, collar weight. Electronic breakaways add weight to the collar, in some cases significantly, and collar weight should also be a concern. With more and more collar options being developed by manufacturers (e.g., larger batteries and cameras), the issue of weight will only increasingly need attention by researchers to strive and keep collar weight as low as possible. Please follow the simple rule that we don’t always have to win and always make ethics and animal welfare the most important study consideration.

As far as fit....i have heard researchers use terms like "2 fingers" or "3 fingers" when it comes to fitting a collar and deciding on how tight it should be.  That way too subjective.  I suggest putting the collar on and attaching the hardware (finger tight)where you think it'll work.  Then, working gently, place 1 ear underneath the collar and then the 2nd ear.  Then, try to remove the collar over the mandible/zygomatic arch .  If you cannot, undo the hardware and make the collar fit a bit looser by attaching it at the next notch.  If you can, then go a bit tighter.  Then, repeat the process. I suggest getting the collar fit to the point of being able to get the collar off but it would require some work to do so.  I have been using that method for over a decade and it has been extrememly succcessful.  As researchers we should always err on the side of the animal having the advantage. Also, we should always assume the collar will fail and insure the animal's well being will not be compomised.   

Thanks Richard for these wise words.

Can the Hellgren et al. (1988) paper be uploaded to the Dropbox folder?

It is at: https://www.dropbox.com/sh/z1r1zexlpaa91gz/AAANQlU4RmCDM3QGY61cBYK-a?dl=0

I'm not sure that the sharing settings are OK, if you don't manage, please send it to me and I will post it there.

Thanks,

Amir

https://www.dropbox.com/sh/z1r1zexlpaa91gz/AAANQlU4RmCDM3QGY61cBYK-a?dl=0

Hi,

Nov. 2015 update:

I contacted Movebank managers and had very interesting discussions with them.

Apparently, Movebank has a forum for asking questions and sharing knowledge:

https://www.movebank.org/forums/animal-tracking-equipment-and-methods

Yet, its efficiency is limited by the fact that there's no automatic notification of members when questions and/or answers are being posted.

 So, it seems to me that ResearchGate Q&A is currently the best channel for getting as much data as possible before designing, capturing, tagging etc.

Cheers,

Amir

https://www.movebank.org/forums/animal-tracking-equipment-and-methods

Hello,

Is the 5% rule actually defined/calculated anywhere?

Hi,

I think that no one knows the origin of this rule. i first heard it from people working with livestock.

It's obvious that it's inappropriate for many groups, if not all.

Review the answers above and the dropbox folder with papers.

If you find something interesting, let us know.

Cheers,

Amir

I found something:

Markham A 2008 On a Wildlife Tracking and Telemetry System: A Wireless Network Approach Doctoral thesis, University of Cape Town, South Africa

https://www.cs.ox.ac.uk/files/2112/ThesisSubmitv2.pdf

In the introduction he discusses the 5% rule with bibliographical references and different methods for proper use of tagging collars.

Does anybody have this:

W. Cochran, “Wildlife telemetry,” in Wildlife management techniques manual, S. Schemnitz, Ed. Random House, 1980.