'building Better Dogs' Seminar 11 Feb 2010

[Lablover]

Overall I enjoyed the day. Dogs can read minds?? I realise they are great body and environment readers, but read human minds?

My main focus is on working labradors. Lucky governments and/or colony breeders have been interested in health and working and behaviour including inheritability for a long period of time. Example this conference: http://www.pennvetwdc.org/Working_Dog_Conference.html

Edited February 12, 2010 by Lablover

[Rebanne]

There's nothing wrong with my chosen breeds or any breed for that matter, it all comes down to the quality of the owner.

:D

[mita]

Help, someone ..... does anyone understand what I am (trying) to say? Have I got it wrong? Am I really not seeing the trees for the forest, or is it the other way around?

Actually, Erny, you're on to something that's true for testing of humans....say, intelligence.

A test tests what the manual says it tests. It doesn't necessarily reflect what the testee does in real-life situations. And you've been talking about real-life situations for dogs.

Like, there's evidence that people who showed better everyday decision-making skills were those who'd had signficantly fewer negative events in their lives....regardless of differences in intelligence test scores. So life events count for how people learn to think their way thro' everyday life. You've actually been saying it'd be the same for dogs.

Yep....& some researchers are considering how to test for that with humans. But they're right in saying it would require a multimillion-dollar research program.

[Steve]

Hang on Slow down! It's been about temperament - fearfulness, agression, and over excitability. Not various breed characteristics.

Ill try and work off my notes but I think some of what you guys have to say isn't really relevant to what happened.

Pauline Bennett was well spoken and what she had to say was interesting though I don't think that with the main part of what she presented was anything that new to most breeders.

In the main it was about being sure to breed with a focus on temperament as well as anything else we may be selecting for.

Basically she said that if a dog's behaviour is predictable with well-described traits then it's easier to find suitable homes for them.

I didn't agree with a couple of her basic beliefs - 1 was That purebred breeders were constrained by breed standards because we focus on how the dog looks which are easily evaluated by the judges. It felt to me that she was saying that we muck it up because of the fact that we only care about the way the dog looks. I think she failed to explain and I have no way of knowing if she understands that showing a dog is only one part of the testing process for a breeding dog. That good breeders do select for temperament as well as health and conformation. A good ethical breeder wouldn't have a dog in their breeding program, which had temperament issues regardless of whether it had a perfect "look" or not.

She made a couple of comments I'm sorry we didn't have time to chat about with her. One was that ANKC breeders are not able to breed for the pet market when in fact as long as that's not all they do according to our codes of conduct they are able to do just that. There's no rule which says if you breed you have to show your dogs and what is required is that you consider future generations past the litter you are producing now so you have your eye on whets best for the breed as well as what's best for the pet buyers which are taking your dogs into their homes this litter.

The other was a question she posed in the after question time which was - Words to the effect that If people are breeding show dogs perhaps they should not breed pet dogs.

However, the thrust of it for us is that temperament issues do have a genetic component and we should be aware of that and consider it in our breeding programs - on that I agree with her.

Mike Goddard was also interesting and he discussed potential selection methods - specifically Estimated breeding values for what ever ever our goal for each litter was. He explained various modes of inheritance and explained that most genetic issues breeders have to deal with are not single genes [PRA for example] but rather many genes and environmental factors [HD AND EPILEPSY FOR EXAMPLE].

He also posed a question about whether people selecting for show dogs were perhaps not compatible with breeding pets.

He explained that no matter what we were aiming for in each litter that it was important to select the best dog within each breed and that because so many are polygenic that the best way to do that is to develop a data rich resource which breeders could use to pinpoint what dogs in what breeds would be most suitable to use with their bitches to aid them in reaching their goal. However, we have to make the info freely available and keep accurate records in order to do that and all of us would need to participate if it were to be a success. The basic message was - you get what you select for and most times you loose what you don't pay attention.

Sort of like - identify your goal, work out what you have to do to get there and then consider what you are prepared to compromise on to get that.

Paul McGreevy spoke of how some conformation traits may cause some breeds to suffer and that we should consider how what we do if we breed to extremes may impact on quality of life.

He spoke of how much we need accurate research which will help us to be more aware of prevalence in this country and discussed the LIDA program. This to me sound fantastic and a much needed resource to help us to sort out fact from assumption on what sort of things our dogs are suffering with and I believe it will track things like trends which may show up into the future to help us to be on top of what we need to test for and eliminate from our breeding programs.

I asked Paul if there was some method in this data base to identify who was breeding each animal that was being treated and diagnosed. This is simply because we are only interested in the dogs we breed and given that we only breed approx 10 % of puppies born in Australia each year we don't want to be pinged for things puppy farmers and cross bred breeders are breeding.

In fact if we cant identify what diseases dogs we breed turn up with in difference to any other then the data becomes useless to us and anyone else who wants to judge the fitness of purebred dogs.

He told me that he would allocate the task of sorting out issues such as these with a PHD student - which sounds good but he didn't really convince me that this was definitely a priority to be covered when I spoke with him privately. In my opinion any purebred dog breeder or dog registry such as Dogs NSW [who have pledged 30,000] to the project should ensure that this is something that will

occur BEFORE they hand over any cash because I think its something that will either make it great or a waste of a potentially great resource.

There were a couple of other things which I would like the opportunity to discuss with Paul.

He seems to think that purebred breeders are not able to have the stud books open when in fact in Australia from day one there has always been the ability for any breeder to apply for this to happen and there are many examples of when exactly this happened and in fact currently the mini bull terriers have an ability to use bull terriers in their breeding programs. Their studbook is open. That's because someone put the proposal together and was able to present how it would be done and why it needed to be done and any of us have the right and the ability to do the same and apply to the ANKC to open the stud books.

He spoke of the opening of the stud books in the UK for the Dalmatians as if it were a new thing and never done before. The way bob -tailed boxers were accepted in the UK was because someone opened the stud book. There are several examples Id like to talk over with him.

I think [almost sure] that he was saying we need to open our stud book and outcross and then back breed to allow our gene pools to be more diverse and Id love to discuss with him the technicalities of this and how he would advise this could be done, how often to what objective and how we can prevent us being in the same place we are but with mutant genes in the mix we didn't know to expect or test for. I thought he was really interesting and easy to talk to but I was disappointed I didn't have enough time or opportunity to narrow it all down a bit.

Next came Kate Scoeffel. I think what she is trying to do in bringing in commercial breeders who will agree to a code of conduct is great given that commercial dog breeding is a reality and its better in my opinion to try to differentiate someway between those who operate one way over another.

Some of her views are controversial, and I would like the opportunity to discuss some issues with her she raised which related to canine husbandry such as back to back litters, age for which dogs should be able to reproduce, how many litters they should have and what type of accommodations breeding dogs should be kept in and a couple of others. I thought it was unfortunate that in order to explain why she breeds first cross puppies for profit it was necessary to make negative comment on purebred breeders and their dogs - but it was a minor irritation.

While I think the concept is good there were some things which I felt needed perhaps a bit more thinking through.

The fact is I think the codes she presented on the whole are much lighter than most state's codes of practice and it will be interesting to see how many large scale commercial puppy farmers are prepared to step up to the plate and take back dogs etc.

Kate also said that she thought the pet shop should take the place of the breeder when the breeder sold their pups to the pet shop and that "perhaps pet shops could work in with shelters or something." In other words she wants her breeders to be accountable but they get let off the hook for that if they sell them via a pet shop or an agent etc. I'm not saying that's good or bad - simply repeating what was said.

She had some novel ideas when it came to working with purebred breeders which would probably stretch the friendship a bit.

She wants purebred breeders to work with cross bred breeders to enable us to get feedback on how our dogs perform when they are used for cross bred litters and she said she would like it if when she approached a purebred "beagle breeder" for a puppy for breeding first cross puppies that rather than the breeder throw her hands in the air and run away that they offered her a good quality dog which she could put in her breeding program.

I'm not sure how long it will take me to warm to that idea because I have to admit that this beagle breeders hands are in the air and I'm saying "not bloody likely" -I'm sure she would expect nothing less of me and would understand but I do think there are probably some things to be gained by treating each other with respect and sharing principals and science resources in canine husbandry. Id like to think that even though we have differences in our goals we could still benefit more by seeing what we might have in common rather than bring what we disagree on to the table.

Ill come back later and talk about what I got from the student presentations.

Edited February 12, 2010 by Steve

[Erny]

You CAN make a reliable claim about the number of times a chicken pecks once you have determined a method for counting those pecks.

Whether or not that is analogous to what Tammie is proposing is something I can't comment on.

But the fact that this is done under the name of "research" implies (to me, at least) that we can (or should be able to) use that information for ....... something.

But unless outside factors that influence how many times chickens peck are taken into account, isn't the fact that they peck 100 times or 10 times fairly useless, save that it scientifically proves that chickens peck?

Edited February 12, 2010 by Erny

[Are You Serious Jo]

You can't reliably make any claims about the current behaviour of an organism unless you know prior experience that shaped it

But that goes well beyond what KismetKat was suggesting. You CAN make a reliable claim about the number of times a chicken pecks once you have determined a method for counting those pecks.

Whether or not that is analogous to what Tammie is proposing is something I can't comment on.

That information is meaningless in isolation. You might end up with a mean rate of pecking per hour for that chicken. All you then know is the range and average times that chicken pecks, and if you have enough of a sample size, how many times chickens may peck. But it doesn't say anything about learning, motivation, global influences, interspecifc factors etc.

I've asked Tammy previously if she was trying add another trait to the personality indicies and measuring instrument we currently have for dogs but got no answer. I am a bit in the dark about aims as well as her previous thread wasn't clear. Is she attempting a new temperament test, a new ethogram, I can't work it out.

[mita]

in my opinion this seems like a worthwhile pursuit if handled objectively and used responsibly in future (which is not really Tammie's responsibility).

It's the creator's responsibility to provide a manual for proper use. Which would need to include a rationale on which the testing protocol is based.

And the rationale....not the test items....is the heart of the matter.

Edited February 12, 2010 by mita

[Lablover]

Erny,

Does this help, a published paper mentioned yesterday.

http://www.biol.ntnu.edu.tw/yuyinghs/yuyin...llJAVMA2003.pdf

[Sheridan]

You CAN make a reliable claim about the number of times a chicken pecks once you have determined a method for counting those pecks.

Whether or not that is analogous to what Tammie is proposing is something I can't comment on.

But the fact that this is done under the name of "research" implies (to me, at least) that we can (or should be able to) use that information for ....... something.

But unless outside factors that influence how many times chickens peck are taken into account, isn't the fact that they peck 100 times or 10 times fairly useless, save that it scientifically proves that chickens peck?

I'm still trying to work out how chicken pecks are analogous to dogs being amicable.

Given that my breed is supposed to have a sense of humour:

Good-tempered, spirited and game. Full of confidence and humour; a delightful, affectionate, intelligent companion.

I would have to ask how exactly the amicability test would measure a sense of humour. If I tell my dog a joke and he fails to chuckle, did he simply not find my joke funny or is it a failure of his laugh button?

[Aidan]

chicken pecks are something a chook does, it is part of being a chook, all chooks peck. A test would be easy to work out.

I think the example was chosen because it was so simple. There was a time when we didn't have units of measure for things like distance, or weight, or volume. Someone had to decide on a unit of measure, then have it accepted as a standard. Not quite so simple as we take for granted now.

Even testing how many times a dog barks in a vet clinic in 24 hrs as Aidan suggests is totally flawed unless those dogs have lived in a vacuum. I have 4 dogs and all react different while at the vets but all are very amiable, IMO.

My suggestion was not a possible test for "amicability", it was an example of something that we CAN quantify. We can count the number of times a dog barks in a vet clinic in the first 24 hours. If that information is useful (and I know of several experiments where it has been useful), then we have an objective measure.

It doesn't tell us how each dog got there and it does require an assumption that in-patient presentations to a vet clinic will follow some sort of statistical norm across a large enough population of dogs (e.g an average number of labs, an average number of terriers, an average number of dogs with broken legs, an average number of dogs who are blind etc etc) but we could test a hypothesis that "water spray collars and citronella collars are equally as effective in reducing barking in the veterinary clinic environment over the first 24 hours of admission" or "Classical music decreases barking in a vet clinic over the first 24 hours compared to no music at all" - and because we can objectively measure the behaviour, we can repeat experiments to see if they continue to provide similar results.

By comparison, we can't form any useful opinions about 'leadership style' used with "dogs who present to a veterinary behaviourist for dominance aggression" because we don't have a way to measure "leadership style" objectively (certainly not through owner reporting) and "dominance aggression" does not have a definition so we don't actually know if dominance is the cause of aggression, or even if the behaviour was even aggressive behaviour (believe it or not, "aggressive behaviour" doesn't even have an agreed-upon definition!)

[centitout]

SORRY,but i am not selling any of my girls for the express purpose of crossbreds-i get approached often enough now :D

[Erny]

Erny,

Does this help, a published paper mentioned yesterday.

http://www.biol.ntnu.edu.tw/yuyinghs/yuyin...llJAVMA2003.pdf

Thanks LL. It's a bit to go through, but on glance, it tells me that this research/exercise has been done before. I need to read further though.

[Aidan]

You can't reliably make any claims about the current behaviour of an organism unless you know prior experience that shaped it

But that goes well beyond what KismetKat was suggesting. You CAN make a reliable claim about the number of times a chicken pecks once you have determined a method for counting those pecks.

Whether or not that is analogous to what Tammie is proposing is something I can't comment on.

That information is meaningless in isolation. You might end up with a mean rate of pecking per hour for that chicken. All you then know is the range and average times that chicken pecks, and if you have enough of a sample size, how many times chickens may peck. But it doesn't say anything about learning, motivation, global influences, interspecifc factors etc.

KismetKat's analogy was to come up with a way to measure how many times a chicken pecks. A "peck counting method".

If, in future, someone wants to come up with a hypothesis for learning, motivation, global influence - on PECKING - then they will need a "peck counting method" to gather data for their experiment. How that data is used or how an hypothesis is formed is a separate issue.

[megan_]

You CAN make a reliable claim about the number of times a chicken pecks once you have determined a method for counting those pecks.

Whether or not that is analogous to what Tammie is proposing is something I can't comment on.

But the fact that this is done under the name of "research" implies (to me, at least) that we can (or should be able to) use that information for ....... something.

But unless outside factors that influence how many times chickens peck are taken into account, isn't the fact that they peck 100 times or 10 times fairly useless, save that it scientifically proves that chickens peck?

I've only been skimming this thread, so I could be way off, but my understanding of it is this: they are trying to determine whether you can measure behaviour. At this stage, what causes it is irrelevant. Once you know you can measure behaviour, you can then add the other factors.

A study that was trying to measure behaviour and causes (many of which are complicated and poorly defined) would be too complex and the information would most likely be useless. One of the hardest things to measure is causation - did factor X cause the chicken to peck more, or is it just a co-incidence? In order to have any hope of ascertaining that, factors need to be isolated and measured consistently.

If we want to answer the question of what factors influence behaviour and how do they impact behaviour we need to:

i) Determine if we can measure behaviour (Tammy's study), then

ii) Test what factors influence that behaviour

Erny - the information can be used for something....in the future. It should provide baseline of measuring behaviour. Then differnt scientists can test their causation hyphotheses against the baseline behaviour.

[Aidan]

You CAN make a reliable claim about the number of times a chicken pecks once you have determined a method for counting those pecks.

Whether or not that is analogous to what Tammie is proposing is something I can't comment on.

But the fact that this is done under the name of "research" implies (to me, at least) that we can (or should be able to) use that information for ....... something.

But unless outside factors that influence how many times chickens peck are taken into account, isn't the fact that they peck 100 times or 10 times fairly useless, save that it scientifically proves that chickens peck?

I do not know Tammie nor was I at the conference, all I can go off is what I have read here. If the aim is to find a "peck counting method" then what anyone does with information is up to them, it is their responsibility to stick to the scientific process. But until you have a ruler to tell you how long something is, or a beaker to tell you how much liquid it produces, you can't measure it.

[mlc]

I can't find the link :D

Hey there,

just reiterating my post from yesterday - we'll be posting the link to all presentations once Monash release the files and we've edited it into individual speaker's (plus question/answer/discussion sessions) - sorry for the delay, but this is likely to take around a week.

Everyone who registered would have received the full day's speaker abstracts (including recommended further reading for most talks) - if you didn't go but would like to see the abstracts, you can email Jeremy Skuse at [email protected] and ask him for a copy.

Cheers,

[Aidan]

So Tammie's thesis is about designing a standard test and recording the different behaviours of different dogs, but not really drawing any conclusions from it?

That is my understanding based on some of the things said in this thread and earlier. It would be good to get some clarification from Tammie herself, though!

[mita]

...If only dogs pecked. A peck is a peck is a peck. No degrees of 'peck'.

Interesting thing is that all the posters are actually filling in (from their own impressions) the rationale & aim which should be written to accompany this test, behavioral observation, or whatever the creator's named it.

Currently, the slants are all over the place. From a test to help pick out dogs with desired (or non-desired) behavioral patterns... to a thesaurus of definitions about dog behaviour interpreted by the researcher into observable actions (& therefore believed to be of use to later researchers).

No wonder people can't figure out what it's all about. They would need to read a rationale & aim.

Edited February 12, 2010 by mita

[Are You Serious Jo]

You can't reliably make any claims about the current behaviour of an organism unless you know prior experience that shaped it

But that goes well beyond what KismetKat was suggesting. You CAN make a reliable claim about the number of times a chicken pecks once you have determined a method for counting those pecks.

Whether or not that is analogous to what Tammie is proposing is something I can't comment on.

That information is meaningless in isolation. You might end up with a mean rate of pecking per hour for that chicken. All you then know is the range and average times that chicken pecks, and if you have enough of a sample size, how many times chickens may peck. But it doesn't say anything about learning, motivation, global influences, interspecifc factors etc.

KismetKat's analogy was to come up with a way to measure how many times a chicken pecks. A "peck counting method".

If, in future, someone wants to come up with a hypothesis for learning, motivation, global influence - on PECKING - then they will need a "peck counting method" to gather data for their experiment. How that data is used or how an hypothesis is formed is a separate issue.

Since I can't figure out what behaviours she'll be measuring and how then I can only guess. You seem to have a grasp on it, will she be measuring each behaviour on a scale and whether it occurs in a specific context? Pecks are an either or score, but other behaviour is on a gradient. Will her test of amicability include something like the strange situation test? What is she going to be able to do with this instrument? Is she purely gathering dogs and testing out different recording media for observer reliability?

Working out a method for measuring something is different for examining a "trait'. From what I read she is interested in the "trait" amicability, but finding a way to measure and record dog behaviours is different from investigating a specific question.

I can't work out the actual hypothesis being tested.

[Rebanne]

Even testing how many times a dog barks in a vet clinic in 24 hrs as Aidan suggests is totally flawed unless those dogs have lived in a vacuum. I have 4 dogs and all react different while at the vets but all are very amiable, IMO.

My suggestion was not a possible test for "amicability", it was an example of something that we CAN quantify. We can count the number of times a dog barks in a vet clinic in the first 24 hours. If that information is useful (and I know of several experiments where it has been useful), then we have an objective measure.

You still don't know why they are barking, is it fear, stress, seperation anxiety, over excited, etc, all of which should be handled differently or you worsen the problem.

It doesn't tell us how each dog got there and it does require an assumption that in-patient presentations to a vet clinic will follow some sort of statistical norm across a large enough population of dogs (e.g an average number of labs, an average number of terriers, an average number of dogs with broken legs, an average number of dogs who are blind etc etc) but we could test a hypothesis that "water spray collars and citronella collars are equally as effective in reducing barking in the veterinary clinic environment over the first 24 hours of admission" or "Classical music decreases barking in a vet clinic over the first 24 hours compared to no music at all" - and because we can objectively measure the behaviour, we can repeat experiments to see if they continue to provide similar results.

How do you objectively measure how effective a "treatment" is when each and every day there would be new dogs in the vet clinic barking? You might strike a day when they are all very mellow or a day when they all set each other off. A day when all the dogs admitted have very stable temperaments or days when they are all stress heads, or a day when they are a mixture. I'll ignore the use of a antibarking collar because I find that thought unbearable but you would have to play music for a long, long time to even come close to seeing if it would work.