Going after chickens

[Somerzby]

Anyone had a problem where their dog starts attacking chickens?

 

What's the best thing to do when this happens?

[persephone]

Bit scant on info  for us to think of anything ...

Your dog ? Your chooks ?
What breed/how old  is the dog ?
How do things normally work in regard to spaces for dogs/chooks? Seperate yards ? Do chooks free range , does dog free range , or ?

has the dog killed a chook/some chooks , or is it chasing them , or is it just not leaving them alone on the other side of the fence ? 

There are no "best things" which will suit every dog and chook problem- each dog is individual, as is every single situation involving a dog chasing/killing other animals/birds  

Lots of us have had this happen ..hopefully we can help . 


 

[Roova]

That's a tough one to work on because chasing chickens is pretty self rewarding to a dog and killing them which is what I presume you mean by attacking can be hard to come back from.  These are just my thoughts so Im definitely not saying they're correct.

 

Firstly a positive based behaviourist is always going to be the easiest avenue because they can see exactly what's going on and work with you to make a change.   If that's not an option you may need to work on management first, training second.

 

Management would be removing any option for the dog to practice the behaviour you don't want to see which might mean having him on lead or inside when the chickens are out. Any time he's allowed to stare or chase them you'll go backward in your training.

 

Training would begin with you deciding what behaviour you want to see instead and then heavily rewarding for this.  You'll have to work out where his threshold is for focusing on you or the chooks and don't push to hard or expect too much. You'll need the whole family on board with training\rewards too.  Once he realises its more rewarding to perform the behaviour you want, you can start changing reward delivery to be  intermittent but still regular to make it worth his while to show self control. 

 

If he's creating his own 'fun' activities you might need to look at what mental enrichment and training you're offering in general.  Can this be improved so he isn't creating his own entertainment?  I do wish you the best of luck with this.  I've been in your shoes and its a tough slog!

 

 

Edited January 17, 2018 by Roova

[talking dog]

19 hours ago, Somerzby said:

Anyone had a problem where their dog starts attacking chickens?

 

What's the best thing to do when this happens?

the problem with a "positive only" approach here is that your dog already found the reward that has the highest value for him - you will have problems to find something that has a significant higher value that allows you to use positive reinforcement as a working tool. Hence there are only 2 options:

 

a) you keep them separated, or

b) you use positive punishment.

 

I don't believe that a muzzle here is a good option as he can do severe damage also with paws if he tries to catch the chickens. 

 

Think about a puppy exploring the backyard for the first time and investigating a bee - usually they learn pretty fast without developing all those issues the "positive only" fanatics are afraid off.  In sheepdog herding training sometimes a whack with a pool noddle is used to teach a dog that is too ambitious to keep the required distance from the sheep. A whack with a pool noddle doesn't do any harm to the dog, and the sheep will appreciate it.

 

It is also important - for a successful training - that you provide alternative outlets for your dog's drive, e.g. flirt pole games where he can chase a lure. The more his appetite for chasing things is satisfied, the easier he will be to train.

[Roova]

 

I agree with some of what you said but did want to point out positive doesn't have to mean permissive.

I know everyone has their own methods but surely trainers realise if they aren't getting the result they want the dog either doesn't understand or isn't motivated enough?

 

I may be misreading your comments but are you suggesting a whack with a pool noodle might teach this dog to keep his distance?  If yes what happens when the owner isn't around or someone doesn't have a pool noodle in their hand?  How much will that damage the trust the dog currently has with his owner?

 

Im half way through Susan Garrett's recallers program and it's simply amazing how quickly you can teach any dog self control and a 100% reliable recall by  playing a variety of short games which the dogs go mad for. She is a positive only based trainer so all the games are a means to an end to get the result you want and the dog doesn't even know they're learning. Along the way dogs are taught anything from not chasing squirrels or rabbits, to not running off, to not jumping on guests or barking at the fence line and so on. You learn how to assess what behaviour you have, think of the behaviour you want and assess what training is needed to get there. 

 

No physical punishment needed.  I'm pretty sure she's taken half a dozen dogs, including a Jack Russell to the world's best level in agility!  Her dogs are walked around her property off lead without taking off to chase deer etc, so if she can get there without whacking her dogs it's not a necessary part of training.

 

I do like your take on alternative outlets for drive though! 

 

 

 

 

 

 

 

 

 

 

[talking dog]

Pending on the breed and for what purpose the dog was bred for, there is a huge difference between agility and normal obedience training on the one side, and herding training on the other side. You can't use  aversives for agility, agility is all about motivating and cheering the dog, it is all about using the principles of positive reinforcement - the moment you have to force a dog to do agility it is over, and I can't see any purpose in forcing a dog to do it. If I find my dog not in the mood for agility training (rare, but it can happen), then so be it, training cancelled - I don't force her to do it. For agility I need a dog on fire, not a dog I have to pressure to do it.

It is a total different animal (literally) wrt sheep or cattle herding (which is nothing else than hunting without the final kill) for a real herding dog. Agility is fun, but when you watch herding dogs doing their job driven by their instinct, you suddenly recognize that it is a dead serious business for the dog. It isn't hard to recognize the wolf in these dogs, all what we did via selective breeding is that we eliminated the final kill. And if I talk about real herding dogs I mean dogs that don't need any treats, games, toys, pats etc. as a reward - the reward is the work with the sheep or cattle. If the dog has a very strong drive, and if it was bred as a nipper (often preferred for cattle) it is pretty challenging to control and train the dog on sheep. Sometimes nippers with high confidence are preferred for sheep not trained previously with dogs, however, in trials you get immediately disqualified if the dog nips a sheep. 

Now people like Susan Garett do a fantastic job, but I believe - also based on my experience - there are horses for courses. Hunting and chasing prey animals is for every dog pretty self rewarding, once experienced, the jinn is usually out of the bottle. It is easy to correct this if you find a reward with a higher value, and you actually might find one, if the prey / herding drive in the dog isn't so strong. However, most dogs are dogs, and it is very likely that you won't find a reward with a higher value.

Consequently - regarding the OP's issue -  positive reinforcement doesn't work to train the dog not to attack the chickens as attacking the chickens follows already the principle of positive reinforcement, and there is likely no other reward that has more value for the dog compared with chasing these fluffy screaming fun balls.  Now we are back to the bee - the dog (assuming he is a smart one) won't chase the bees - why not?

The kind of aversive stimuli that would work for the OP depends on drive of the dog, localities, how deep the chasing is already reinforced etc. etc., and it will only work if used clinically. For every attempt attack there has to be a correction - with or without the owner close to the dog. For some dogs a towel ("bonker") thrown at the dog works, for others water hose, water filled balloons, some respond just to unpleasant sounds or loud noise. The most difficult part is to train the dog that the rules have to be obeyed also if the owner is away. Hide behind a bush, door, car and be ready if the dog makes an attempt. In addition to the positive punishment part of the training, it requires also work with the dog close to the chicken: let the dog approach the chicken (use a leash if this is the only way to control him), reward if it is in a controlled manner, let him sit and watch, drop, stand - exercise him close to the chicken and reward, reward. This is positive reinforcement for the good behaviour - but it won't work without the positive punishment part as it will otherwise always compete with the reward from chasing the chicken.

This training needs 100% commitment to be successful, there is no room for error. If the OP is not willing to spend the time and effort, then it is option a: keep them separated.

[talking dog]

10 hours ago, Roova said:

 

I may be misreading your comments but are you suggesting a whack with a pool noodle might teach this dog to keep his distance?  If yes what happens when the owner isn't around or someone doesn't have a pool noodle in their hand?  How much will that damage the trust the dog currently has with his owner?

 

The "whack" is more a friendly reminder that all the fun still has to comply with the T&Cs of the handler. You can do distance work with those herding dogs as much as you want and play with the trigger zones, but there is the time where you have to let the dog in "striking" distance to the sheep. Outside this zone the dog is the most well behaved dog, but inside this zone the nipping instinct can kick in overwriting all the other behaviour you trained before. The noodle helps the dog to find the balance between complying with the T&Cs and following his instincts. At the start the dog doesn't know that the nipping is wrong (or only wanted if the handler ask for it). However, you can't tell the dog when he is zoned into his instincts that this is wrong, he won't hear you. If you just walk away from the sheep with the dog (negative punishment) he can't differentiate which of his behaviours was wrong: too close? the nip? wrong sheep? too fast approach? The handler has to anticipate the strike for the perfect correction - perfect timing is the key. Later, when the communication line into the dogs own world is established, he will follow your cues even in such a high aroused state. And no, you don't lose any trust of the dog - working the sheep is so rewarding for these herding dogs (even without nipping) that you always will have a very happy and balanced dog after.

[The Spotted Devil]

Where Susan Garrett excels is her understanding of how to transfer value. I have working Gundogs - I hunt them on game and compete in retrieving/agility. Game makes their head explode and yes, I went through a stage where they would spit out a cheese reward because they were so set on either running or hunting. But I learned to transfer the value AWAY from the hunting scenario. If we have a blip I can fix it incredibly quickly because I am super consistent. I fence my chickens, however, because at home its very difficult to be consistent. And my OH is hopeless  

 

I don’t do herding so I will not comment on that. However. People told me I couldn’t train a working Gundog with food only and no punishment. I use a NRM or interrupt the behaviour only if continuing the behaviour will be reinforcing in itself. Am I a hard @r$e? Do I expect my dogs to work for their reinforcement? Yes, absolutely  

 

I am FAR from the perfect trainer and get lots of stuff wrong. However I don’t blame the science of learning theory. I blame the application. 

[talking dog]

@TSD: first let me tell you I really wish they had come up with different terms when they described operant conditioning - reinforcement is o.k., but using "positive", "negative" and "punishment" is just a recipe for confusion, heated debates and frustration. Wrt NRM: technically (please take it with a pinch of salt) if employed it represents a stimuli, and if you use it, and the outcome is a decreased unwanted behaviour, well, it is technically positive punishment. Some dogs doesn't like working with NRMs (understandable, you just signalized them that they missed the chance for a juicy treat), some dogs take it neutral or even appreciate the guidance it provides. So whether it is really positive punishment or something else depends actually on the dog.

My dog is too fast, respectively I'm too old or too fat for agility, so the only way to guide her around is by cues. I don't use a clinical NRM, but she knows from my changing voice wether she aims for the right or wrong obstacle, or does the right or wrong turn. Iiii, ahhh, no, uhhh. oooohhh etc. gives all the same result - she knows it is the wrong obstacle, listens, and picks (mostly) the right one. If she is unsure, she appreciates my guidance, and even looks for me to ask for it - if I'm too slow with my cue and have to correct her while she is already close to an obstacle she is actually pissed off. With sheep herding it is totally different. You could argue that the noodle is nothing else than a NRM, my dog doesn't bother, it just a reminder that nipping is the wrong choice. She won't take any treats, on a scale from 1-10 the treats she normally (obedience, agility) works for has a value of less than 1 while working with sheep would be 10 or even higher.

It all depends on the individual dog, his drives and instincts, the bond between the dog and his handler and the task. Knowing as much as possible about multiple trainings tools and behaviour science helps of course, but I will never settle on a certain approach assuming that this is the silver bullet for all trainings scenarios.

[The Spotted Devil]

I’m an animal behaviour scientist and I hate getting caught up in the quadrants. It’s a slippery slope to arguments that go round and round in circles.

 

My point is that I selectively reinforce the behaviours I like, split complex behaviours into tiny parts, set my dogs up for success, record keep (notes and video) my training in order to find out what I am doing right/wrong (am I really reinforcing what I think I am?) etc. Something like a NRM is a very rare thing for me to use and is used purposefully in a specific training focus. They are generally over used. I might not even say anything if my dog knocks a bar. I might just take ages to re-set the bar, tie my shoe lace etc. That is sufficient to take the dog out of that high arousal into a thinking space.

 

As for your example about you delivering a late cue and having to call the dog off and the dog becoming “pissed off” I will teach it differently. I have a cue that indicates to the dog that they must pull off the obstacle (again thank you SG). And they get reinforced heavily for demonstrating impulse control. I train it, I build confidence, I test it. Just a different way of approaching the same problem. 

 

I agree about treats vs live pheasant or rabbits or quail or whatever. Note I am talking about things I do not how I think they should be done. I have the SAME problems. However, as I said earlier I have learnt how to transfer the value and to use PERMISSIONS. I teach them to run PAST the duck to retrieve the dummy. Then they get to retrieve the duck. I’m always asking how well they understand the behaviour? What if I now add this distraction? I want to get them so high high high with excitement around game or an agility tunnel and then sit when cued quietly. That’s why I can whistle sit my older bitch on a rabbit at 100m. A little lateral thinking is all that’s required.

Edited January 19, 2018 by The Spotted Devil

[talking dog]

wrt "quadrants" respectively discussions about it: IMO it doesn't matter how we call it, or how different people class it; what really matters IMO is a) whether it works, and b) whether it is beneficial for the dog and its environment. My goal is not the employment of a specific method for the sake of whatever, my goal is the best outcome for the individual dog and the associated environment.

 

wrt "splitting complex behaviours into tiny parts": that's the challenging and interesting part in herding - I don't think it is possible. While working sheep I can try to take the steam out by letting the dog taking a holding position (drop, stand, hold, etc., ) putting everything "on hold", and despite that it might be a perfect balance, nothing moves: it won't get the dog into a thinking space or to a lower arousal level. She still listens to me, yes, she follows my cues, yes, but compared to agility it is like talking to someone who's mind is far, far away. If the dog has a high drive, you won't get the dog out of this hunting / herding mode if the sheep are still in "striking" distance. In agility I just ask her to drop and she switches from working mode in resting mode in the flick of a second. We can raise or lower the bars, change the course, she might watch, but doesn't bother, till I fire her up again.

 

Wrt knocking bars (happens rarely): it is not a failure for me; I know she always tries to clear it, so if it happens it doesn't matter to me and I never try to correct it. If it happens it is mostly my mistake anyway as I asked her to do it from an impossible angle or confused her with unclear hand signals.

 

[The Spotted Devil]

I get that herding is different. But working Gundogs on live game raises similar issues. You can absolutely split stuff up. Arousal is the main thing I work on.

 

I love how Bob Bailey sums it up....a behavioural scientist, one of the greatest trainers in the world who made his living training everything from chickens to crows to cats to military dogs to dolphins and everything in between. In essence he says that there are 100s of ways to train a dog. And the vast majority will get the result in the end. But how efficient is the method? How long will it take? How many errors? How precise is the behaviour? For him time = money. So he didn’t use positive reinforcement based training because he’s a nice guy. He used it - in the most precise and perfect way that he could - to get fluent behaviours as quickly as possible. But I’ve never seen anyone so adept at splitting behaviours into their smallest parts. Extraordinary. 

 

Knocking bars is not a failure I agree. But it happens for a reason - high arousal, not thinking, early take off, late cue, not focussed forward, worried, not clear on required behaviour. Review your videos. And a bar dropping protocol gives the dog every opportunity to figure it out for itself. 

Edited January 19, 2018 by The Spotted Devil

[ness]

My herding dog has been trained in herding using only positive reinforcement. We don't have stock access to practice consistently but she has managed a started title. Its mostly been about setting up what we are after and attaching a command when she is readily offering the behavior. There is absolutely no way I could use an external reinforcement - food or toy as she won't take either on stock. We use tugs, toys and food in other contexts so when we train agility and obedience but herding is purely access to stock and being allowed to work. It took a little longer to get a reliable stop and impulse control on her but with a bit of maturity it started to happen anyway. The key was setting her up for success and growing behaviors incrementally.

[talking dog]

wrt Bob Baily,  he also said in an interview about clicker training:

• "I'm not a clicker trainer. I have used a clicker, and quite successfully......In the modern use of the name Clicker Trainer, punishment, especially positive punishment, is disallowed. I allow myself to use punishment if I believe it is necessary to accomplish the task and the task merits the use of punishment. I rarely have need of punishment."

That's pretty much I would describe my approach.

 

@ ness: I presume your dogs are not bred as nippers? it makes training much more enjoyable if you don't have to be on the watch to prevent nipping.

[The Spotted Devil]

I recently did a workshop with Bob for 3 days. Working spot. His use of punishment is not what it sounds like. He VERY rarely uses it. He named ONE client. And couldn’t think of another example. He NEVER uses it to train. Only to absolutely proof. And not until the behaviour is 100% fluent. And he punishes extremely hard. Not a tap on the nose but so aversive as to completely extinguish the behaviour. He maintains that although this situation was life or death for the human handlers he still didn’t think it was necessary to make the training effective. 

[talking dog]

He (B.B.) must getting old and forgetful , here another quote from him: 

• Mind you, out of the many thousands of animals we have trained during many hundreds of training programs, we punished about a dozen times. The fact that I would do this at all disqualifies me from the ranks of what people call "clicker trainers." I don't mind. I have a clear conscience. I believe I did the best for animal and human kind.

Wrt nipping and punishment: the challenge here is that usually it is a wanted trait, for cattle anyway, and for sheep sometimes. When you look for herding dogs, they are actually advertised as nippers and not-nippers. And heelers are called so for a reason. So you actually don't want to extinguish it totally, you just want to control it.

 

 

[ness]

One of mine does actually go in fairly close and air snap but its harmless so we don't worry about it. She hasn't trialled. The one I trial is naturally a distance worker and no has never shown an inclination to put teeth on the sheep.

[Roova]

I thought I'd share this book for the OP (if they come back) as it has some positive reviews and might give them some tips.

 

"Chase! Managing your dogs predatory Instincts"

 

https://www.amazon.co.uk/gp/aw/d/1929242689/ref=cm_cr_srp_mb_bdcrb_top?ie=UTF8

 

This is one of the reviews:

A brilliant book! As a dog trainer and behaviourist (qualified) I gained alot from this book. It explains the merits of using positive reinforcement to change how dogs think and feel about their chosen prey (be it cars, rabbits, cyclists or cats). It also provides very good arguments (much better than I could have written myself) against using punishment, and aversives (things the dog may find frightening, or dislike intensely such as loud noises, or phsyical punishment such as an electric shock). I also feel I could recommend this book to clients without them getting to overwhelmed with science or terminology, nor is it too fluffy or full of anecdotes (just enough are used as examples to prove points). Overall the book aims to change your dog's behaviour, by training an alternative behaviour. From chasing its prey when he sees it, to sitting and looking at you. at which point you become way more fun than the prey item by providing food, attention and games. It also has some great exercies for teaching perfect recall and using longlines (great for allowing you dog freedom, whilst still being able to control his chasing). It would also be a good guide for preventing predatory chase in the first place for new owners of shepherds, border collies, lurchers and terriers.

Edited January 19, 2018 by Roova

[KobiD]

Interesting topic. And I agree with both sides to some extent.

 

I like to relate things back to the kids (toddlers specifically, as they don't have that higher thought process established yet), and in my experiences different things work for different ones in different situations. Every challenge is unique, and as such what works in one case may not for another.

 

I might be old fashioned, but I believe in action and consequence;  essentially what talking dog is saying about dogs and bees. With our youngest daughter, she has an awareness that the wooden spoon can be used for things other than cooking. It doesn't mean that I beat her with it.. but there is a cue which leads to a choice. If she offers behaviours I like I will reward her with treats, activities, games, stickers, etc. If the situation has arose that she is acting on impulse and making poor decisions I won't use positive rewards to encourage good choice, I may withhold and let her know she's missed out on occasion, or I more likely I will use the positive punishment cue (do I need the wooden spoon). Often she'll respond and change her actions, which then can be rewarded.. if not the follow through has to be consistent. On the other hand, there are times where she's clearly well over threshold, emotionally out of control, and all she needs is to be taking out from what she is doing and allowed to reset herself. As a parent you need to evaluate and work with what is in front of you. As a dog trainer it is the same.

 

In comparison, her older brother never really cared for aversive measures. Everything he was just like 'meh' and would be back to doing what you'd asked him to stop doing 5 minutes later. With him it is much more about controlling the environment/options to encourage the right choice. You could offer him 2 scenarios and let him pick one and he'd be as happy as a pig in it. Little did he realise that he'd played right into your game. His sister see's straight through this and is hell bent on what she wants regardless of what options we have. Different kids, different strategies. Again, as a dog trainer it's the same.

 

[talking dog]

4 hours ago, Roova said:

I thought I'd share this book for the OP (if they come back) as it has some positive reviews and might give them some tips.

 

"Chase! Managing your dogs predatory Instincts"

 

https://www.amazon.co.uk/gp/aw/d/1929242689/ref=cm_cr_srp_mb_bdcrb_top?ie=UTF8

 

This is one of the reviews:

A brilliant book! As a dog trainer and behaviourist (qualified) I gained alot from this book. It explains the merits of using positive reinforcement to change how dogs think and feel about their chosen prey (be it cars, rabbits, cyclists or cats). It also provides very good arguments (much better than I could have written myself) against using punishment, and aversives (things the dog may find frightening, or dislike intensely such as loud noises, or phsyical punishment such as an electric shock). I also feel I could recommend this book to clients without them getting to overwhelmed with science or terminology, nor is it too fluffy or full of anecdotes (just enough are used as examples to prove points). Overall the book aims to change your dog's behaviour, by training an alternative behaviour. From chasing its prey when he sees it, to sitting and looking at you. at which point you become way more fun than the prey item by providing food, attention and games. It also has some great exercies for teaching perfect recall and using longlines (great for allowing you dog freedom, whilst still being able to control his chasing). It would also be a good guide for preventing predatory chase in the first place for new owners of shepherds, border collies, lurchers and terriers.

the attached "review" highlights exactly why there is so much heated debates out there about using punishment for training: the reviewer condemns the use of punishment and aversives, and in the next sentence a longline becomes the perfect tool for recall training ! So how does he/she makes use of the longline for changing behaviour (dog not running off) at the end without applying any stimuli via the longline (and the stimuli would be force = aversives)? But hey, the reviewer is a qualified dog trainier and behaviourist and might use a fine twine as a longline which he/she is able to use without applying any tension.

"From chasing its prey when he sees it, to sitting and looking at you. at which point you become way more fun than the prey item by providing food, attention and games." ha, ha, that would be a pretty useless dog for sheep and cattle herding  ...the biggest fun I can give my dog is to allow her to work sheep, and no, she doesn't do it to please me, however she obeys my cues because she knows if she doesn't the fun would be over.

 

Edited January 20, 2018 by talking dog
reason for editing: spelling